Oxford Textbook of Clinical Nephrology [4 ed.] 9780199592548, 0199592548

Table of contents :
Cover......Page 1
Oxford Textbook of Clinical Nephrology......Page 4
Copyright......Page 5
Summary of Contents......Page 8
SECTION 1: Assessment of the patient with renal disease......Page 26
1 Epidemiology of kidney disease......Page 28
2 Clinical assessment of the patient with renal disease: overview......Page 45
3 Presentations of renal disease......Page 47
4 Kidney disease-focused history taking......Page 52
5 Kidney disease-focused features on examination......Page 54
6 Urinalysis......Page 60
7 Assessment of renal function......Page 69
8 Tubular function......Page 87
9 Renal radiology: overview......Page 91
10 Ionizing radiation and radiation
protection......Page 93
11 Plain radiography, excretion radiography, and contrast radiography......Page 99
12 Intervention......Page 107
13 Ultrasound......Page 114
14 Computed tomography......Page 126
15 Magnetic resonance imaging......Page 134
16 Radioisotopes in diagnostic imaging in nephrology......Page 142
17 Immunological investigation of the patient with renal disease......Page 152
18 The renal biopsy......Page 167
19 Clinical trials: why and how in nephrology......Page 186
SECTION 2: The Patient with fluid, electrolyte, and renal tubular disorders......Page 194
20 An overview of tubular function......Page 196
21 Sodium transport and balance: a key role for the distal nephron......Page 214
22 Water homeostasis......Page 222
23 Potassium homeostasis......Page 229
24 Renal acid–base homeostasis......Page 243
25 Phosphate homeostasis......Page 250
26 Calcium homeostasis......Page 256
27 Magnesium homeostasis......Page 268
28 Approach to the patient with hyponatraemia......Page 274
29 Approach to the patient with hypernatraemia......Page 286
30 Approach to the patient with oedema......Page 297
31 Approach to the patient with salt-wasting tubulopathies......Page 307
32 Approach to the patient with polyuria......Page 316
33 Clinical use of diuretics......Page 324
34 Approach to the patient with hypo-/hyperkalaemia......Page 348
35 Approach to the patient with metabolic acidosis or alkalosis......Page 364
36 Approach to the patient with renal tubular acidosis......Page 388
37 Approach to the patient with hypercalcaemia......Page 397
38 Approach to the patient with
hypocalcaemia......Page 403
39 Approach to the patient with hypo-/hyperphosphataemia......Page 409
40 Approach to the patient with hypomagnesaemia......Page 422
41 Approach to the patient with renal Fanconi syndrome, glycosuria, or aminoaciduria......Page 437
SECTION 3: The patient with glomerular disease......Page 448
42 The glomerulus and the concept of glomerulonephritis......Page 450
43 The renal glomerulus: the structural basis of ultrafiltration......Page 461
44 Function of the normal glomerulus......Page 476
45 Mechanisms of glomerular injury: overview......Page 484
46 The patient with haematuria......Page 488
47 Loin pain haematuria syndrome......Page 494
48 Nutcracker syndrome and phenomenon......Page 498
49 Exercise-related pseudonephritis......Page 501
50 Proteinuria......Page 503
51 Postural proteinuria (benign orthostatic proteinuria)......Page 510
52 Nephrotic syndrome......Page 512
53 Pathophysiology of oedema in nephrotic syndrome......Page 521
54 Idiopathic nephrotic syndrome: overview......Page 524
55 Minimal change disease: clinical features and diagnosis......Page 526
56 Minimal change disease: treatment and outcome......Page 531
57 Primary focal segmental glomerulosclerosis: clinical features and diagnosis......Page 540
58 Primary focal segmental glomerulosclerosis: treatment and outcome......Page 550
59 Pathogenesis of proteinuria in minimal change disease and focal segmental glomerulosclerosis......Page 558
60 Membranous glomerulonephritis: overview......Page 562
61 Membranous glomerulonephritis: clinical features and diagnosis......Page 564
62 Membranous glomerulonephritis: treatment and outcome......Page 569
63 Secondary membranous glomerulonephritis......Page 582
64 Membranous glomerulonephritis: pathogenesis......Page 585
65 Immunoglobulin A nephropathy: overview......Page 590
66 Immunoglobulin A nephropathy: clinical features......Page 591
67 Immunoglobulin A nephropathy: diagnosis......Page 597
68 Immunoglobulin A nephropathy: treatment and outcome......Page 602
69 Immunoglobulin A nephropathy: pathogenesis......Page 611
70 Crescentic (rapidly progressive) glomerulonephritis......Page 617
71 Antiglomerular basement membrane disease: overview......Page 623
72 Antiglomerular basement membrane disease: clinical features and diagnosis......Page 624
73 Antiglomerular basement membrane disease: treatment and outcome......Page 631
74 Antiglomerular basement membrane disease: pathogenesis......Page 634
75 Alport post-transplant antiglomerular basement membrane disease......Page 644
76 Post-infectious glomerulonephritis: overview......Page 646
77 Post-streptococcal glomerulonephritis......Page 648
78 Immunoglobulin A-dominant post-infectious glomerulonephritis......Page 658
79 Glomerulonephritis associated with endocarditis, deep-seated infections, and shunt nephritis......Page 661
80 Membranoproliferative glomerulonephritis and C3 glomerulopathy......Page 666
81 Fibrillary and immunotactoid glomerulopathy......Page 674
82 Drug-induced and toxic glomerulopathies......Page 681
SECTION 4: The patient with interstitial disease......Page 692
83 Acute tubulointerstitial nephritis: overview......Page 694
84 Drug-induced acute tubulointerstitial nephritis......Page 703
85 Other toxic acute tubulointerstitial
nephritis......Page 712
86 Chronic tubulointerstitial nephritis: overview......Page 715
87 Drug-induced chronic tubulointerstitial nephritis......Page 720
88 Heavy metal-induced tubulointerstitial nephritis......Page 727
89 Aristolochic acid nephropathy caused by ingestion of herbal medicinal products......Page 734
90 Balkan endemic nephropathy......Page 739
91 Radiation nephropathy......Page 744
92 Urate nephropathy......Page 749
93 Immune-mediated tubulointerstitial nephritis......Page 751
SECTION 5: The patient with reduced renal function......Page 764
94 Chronic kidney disease: definition, classification, and approach to management......Page 768
95 Chronic kidney disease in the developed world......Page 780
96 Chronic kidney disease in developing countries......Page 787
97 Chronic kidney disease long-term outcomes: progression, death, cardiovascular disease, infections, and hospitalizations......Page 792
98 Cardiovascular disease and chronic kidney disease: overview......Page 802
99 Recommendations for management of high renal risk chronic kidney disease......Page 805
100 Hypertension as a cause of chronic kidney disease: what is the evidence?......Page 812
101 Diet and the progression of chronic kidney disease......Page 818
102 Lipid disorders of patients with chronic kidney disease......Page 825
103 Smoking in chronic kidney disease......Page 832
104 Analytical aspects of measurements and laboratory values in kidney disease......Page 838
105 Effect of lifestyle modifications on patients with chronic kidney disease......Page 849
106 Malnutrition, obesity, and undernutrition in chronic kidney disease......Page 855
107 Left ventricular hypertrophy in chronic kidney disease......Page 862
108 Sudden cardiac death in chronic kidney disease......Page 878
109 Epidemiology of calcium, phosphate, and parathyroid hormone disturbances in chronic kidney disease......Page 894
110 The role of inflammation in chronic kidney disease......Page 902
111 Vascular stiffness in chronic kidney disease: pathophysiology and implications......Page 909
112 Oxidative stress and its implications in chronic kidney disease......Page 920
113 Abnormal endothelial vasomotor and secretory function......Page 928
114 Endothelins and their antagonists in chronic kidney disease......Page 935
115 Chronic kidney disease-mineral and bone disorder: overview......Page 941
116 Imaging for detection of vascular disease in chronic kidney disease patients......Page 948
117 Pathophysiology of chronic kidney disease-mineral and bone disorder......Page 959
118 Management of chronic kidney disease-mineral and bone disorder......Page 964
119 Fibroblast growth factor 23, Klotho, and phosphorus metabolism in chronic kidney disease......Page 972
120 Vascular calcification......Page 982
121 Fractures in patients with chronic kidney disease......Page 995
122 Spectrum of bone pathologies in chronic kidney disease......Page 1001
123 Clinical aspects and overview of renal anaemia......Page 1008
124 Erythropoiesis-stimulating agents in chronic kidney disease......Page 1016
125 Iron metabolism in chronic kidney
disease......Page 1023
126 Iron management in renal anaemia......Page 1035
127 Pleiotropic effects of vitamin D......Page 1041
128 Immunity......Page 1063
129 The epidemiology of hepatitis viruses in chronic kidney disease......Page 1074
130 Gastroenterology and renal medicine......Page 1077
131 Cutaneous manifestations of end-stage renal disease......Page 1089
132 The patient with reduced renal function: endocrinology......Page 1097
133 Sexual dysfunction......Page 1116
134 Health-related quality of life and the patient with chronic kidney disease......Page 1124
135 Coagulopathies in chronic kidney disease......Page 1127
136 Mechanisms of progression of chronic kidney disease: overview......Page 1130
137 Proteinuria as a direct cause of progression......Page 1132
138 Nephron numbers and hyperfiltration as drivers of progression......Page 1137
139 Podocyte loss as a common pathway to chronic kidney disease......Page 1143
140 Disordered scarring and failure of repair......Page 1151
141 Modality selection for renal replacement therapy......Page 1161
142 Patient education and involvement in pre-dialysis management......Page 1167
143 Preparation for renal replacement
therapy......Page 1173
144 Choices and considerations for
in-centre versus home-based renal replacement therapy......Page 1182
145 Conservative care in advanced chronic kidney disease......Page 1190
146 Palliative care in end-stage renal disease......Page 1205
147 Patient selection when resources are limited......Page 1212
148 Acidosis in chronic kidney disease......Page 1217
SECTION 6: The patient with another primary diagnosis......Page 1222
149 The patient with diabetes mellitus......Page 1224
150 Kidney involvement in plasma
cell dyscrasias......Page 1273
151 The patient with cryoglobulinaemia......Page 1276
152 The patient with amyloidosis......Page 1289
153 The patient with myeloma......Page 1301
154 Light-chain deposition disease......Page 1317
155 Other consequences from monoclonal immunoglobulins/fragments: membranoproliferative glomerulonephritis and acquired Fanconi syndrome......Page 1324
156 The patient with sarcoidosis......Page 1329
157 The patient with vasculitis: overview......Page 1332
158 The patient with vasculitis: pathogenesis......Page 1340
159 The patient with vasculitis: clinical aspects......Page 1357
160 The patient with vasculitis: treatment and outcome......Page 1369
161 The patient with systemic lupus erythematosus: overview and
pathogenesis......Page 1381
162 The patient with systemic lupus erythematosus: clinical features, investigations, and diagnosis......Page 1390
163 The patient with systemic lupus erythematosus: treatment and outcome......Page 1405
164 The patient with antiphospholipid syndrome with or without lupus......Page 1414
165 The patient with scleroderma: systemic sclerosis......Page 1420
166 The patient with rheumatoid arthritis, mixed connective tissue disease, Sjögren syndrome, or polymyositis......Page 1428
167 The patient with sickle cell anaemia......Page 1438
168 The obese patient (metabolic
syndrome)......Page 1449
169 The patient with hepatorenal syndrome......Page 1458
170 Kidney/ear syndromes......Page 1467
171 Kidney/eye syndromes......Page 1475
172 The patient with renal cell cancer......Page 1482
173 The patient with Wilms tumour......Page 1489
174 The patient with haemolytic uraemic syndrome/thrombotic thrombocytopenic purpura......Page 1494
SECTION 7: The patient with urinary tract infection......Page 1516
175 Urinary tract infection in the adult: overview......Page 1518
176 Infection of the lower urinary tract......Page 1520
177 Upper urinary tract infection......Page 1531
178 Complicated urinary tract infection......Page 1539
179 Urinary tract infection in a patient with a kidney transplant......Page 1542
180 Urinary tract infection in infancy and childhood......Page 1545
181 Schistosomiasis: the parasite and the host......Page 1557
182 Schistosomiasis: clinical impact......Page 1563
SECTION 8: The patient with infections causing renal disease......Page 1568
183 Malaria......Page 1570
184 Leishmaniasis and trypanosomiasis......Page 1575
185 Hepatitis B......Page 1577
186 Hepatitis C......Page 1585
187 HIV and renal disease......Page 1592
188 Hantaviral infections......Page 1601
189 Dengue and other viral haemorrhagic fevers......Page 1606
190 Yellow fever, severe acute respiratory
syndrome virus, and H1N1 influenza infections......Page 1609
191 Leptospirosis......Page 1612
192 Syphilis......Page 1615
193 Rickettsiosis......Page 1617
194 Schistosomiasis......Page 1621
195 Nematode infections......Page 1627
196 Mycobacterial infections: tuberculosis......Page 1629
197 Mycobacterial infections: leprosy and environmental mycobacteria......Page 1645
198 Renal involvement in other infections: diarrhoeal diseases, salmonella, melioidosis, and pregnancy......Page 1650
SECTION 9: The patient with urinary stone disease......Page 1654
199 Epidemiology of nephrolithiasis......Page 1656
200 Approach to the patient with
kidney stones......Page 1662
201 Calcium stones......Page 1670
202 Uric acid stones......Page 1686
203 Cystine stones......Page 1691
204 Cell biology of nephrocalcinosis/nephrolithiasis......Page 1696
205 Medical management of nephrocalcinosis and nephrolithiasis......Page 1722
206 Imaging and interventional treatment: urolithiasis from the surgeon’s point of view......Page 1738
SECTION 10: The Patient with Hypertension......Page 1748
207 The structure and function of renal blood vessels......Page 1750
208 Regulation of vasomotor tone in the afferent and efferent arterioles......Page 1754
209 Tubuloglomerular feedback, renal autoregulation, and renal protection......Page 1763
210 The kidney and control of
blood pressure......Page 1767
211 The effect of hypertension on renal vasculature and structure......Page 1775
212 Ischaemic nephropathy......Page 1785
213 Renal artery stenosis: clinical features and diagnosis......Page 1791
214 Renal artery stenosis: diagnosis......Page 1801
215 Renal artery stenosis: management and outcome......Page 1807
216 Malignant hypertension......Page 1817
217 Resistant hypertension......Page 1824
218 The hypertensive child......Page 1827
219 Treatment of hypertension in children......Page 1850
SECTION 11: The patient with acute kidney injury (and critical care nephrology)......Page 1854
220 Definitions, classification, epidemiology, and risk factors of acute kidney injury......Page 1856
221 Pathophysiology of acute kidney injury......Page 1869
222 Clinical approach to the patient with acute kidney injury: diagnosis and differential diagnosis......Page 1897
223 The role of novel biomarkers in acute kidney injury......Page 1912
224 Prevention of acute kidney injury: overview......Page 1921
225 Prevention of acute kidney injury: non-pharmacological strategies......Page 1927
226 Prevention of acute kidney injury: pharmacological strategies......Page 1940
227 Prevention of acute kidney injury:
drug- and nephrotoxin-induced acute kidney injury......Page 1950
228 Non-dialytic management of the patient with acute kidney injury......Page 1958
229 Fluid overload in acute kidney injury......Page 1966
230 Electrolyte and acid–base disorders in AKI......Page 1970
231 Coagulation disturbances in acute
kidney injury......Page 1979
232 Renal replacement therapy in the patient with acute kidney injury: overview......Page 1981
233 Intermittent acute renal replacement therapy......Page 1987
234 Continuous renal replacement therapy......Page 2000
235 Peritoneal dialysis in acute kidney injury......Page 2014
236 Scoring systems in acute kidney injury patients......Page 2023
237 Overall outcomes of acute kidney injury......Page 2034
238 Renal outcomes of acute kidney injury......Page 2042
239 Acute kidney injury in children......Page 2049
240 Acute kidney injury in the elderly......Page 2063
241 Acute kidney injury in the tropics......Page 2072
242 Acute kidney injury and hantavirus disease......Page 2084
243 Community-acquired pneumonia and acute kidney injury......Page 2092
244 Acute kidney injury in severe sepsis......Page 2093
245 Cardiovascular surgery and acute kidney injury......Page 2101
246 Contrast-induced acute kidney injury......Page 2109
247 Renal failure in cirrhosis: pathogenesis, diagnosis, and treatment......Page 2116
248 Acute kidney injury in heart failure......Page 2134
249 Acute kidney injury in pulmonary
diseases......Page 2140
250 Acute kidney injury in pregnancy......Page 2154
251 Acute kidney injury in the
cancer patient......Page 2158
252 Acute kidney injury in polytrauma and rhabdomyolysis......Page 2168
253 Acute kidney injury in patients with severe burn injury......Page 2176
SECTION 12: The patient on dialysis......Page 2184
254 Uraemic toxins: overview......Page 2186
255 Haemodialysis: overview......Page 2198
256 Haemodialysis: vascular access......Page 2201
257 Haemodialysis: principles......Page 0
258 Haemodialysis: prescription
and assessment of adequacy......Page 2213
259 Haemodialysis: acute complications......Page 2232
260 Haemofiltration and haemodiafiltration......Page 2243
261 Dialysis withdrawal and palliative care......Page 2252
262 Frequent haemodialysis......Page 2259
263 Peritoneal dialysis: overview......Page 2274
264 Peritoneal dialysis: principles
and peritoneal physiology......Page 2277
265 Peritoneal dialysis: adequacy
and prescription management......Page 2288
266 Peritoneal dialysis: non-infectious complications......Page 2295
267 Overview of dialysis patient management and future directions......Page 2304
268 Cardiovascular complications in end-stage renal disease patients: pathophysiological aspects......Page 2310
269 Bacterial and fungal infections in patients on haemodialysis......Page 2318
270 Bacterial and fungal infections in patients on peritoneal dialysis......Page 2324
271 Viral infections in patients on dialysis......Page 2328
272 Cognitive function, depression,
and psychosocial adaptation......Page 2330
273 Volume assessment
and management in dialysis......Page 2338
274 Nutritional screening and nutritional management in dialysis patients......Page 2348
SECTION 13: The transplant patient......Page 2357
275 The evolution of kidney transplantation......Page 2359
276 Pre-transplant assessment
of the recipient......Page 2372
277 Organ donation......Page 2380
278 Donor and recipient kidney transplantation surgery......Page 2392
279 Immunology, sensitization,
and histocompatibility......Page 2404
280 Immediate post-transplant care
and surgical complications......Page 2419
281 Immunosuppression: drugs
and protocols......Page 2427
282 Renal transplant imaging......Page 2440
283 Rejection......Page 2455
284 Infection: prophylaxis, diagnosis, and management......Page 2467
285 Cardiovascular disease: prophylaxis, diagnosis, and management......Page 2477
286 Chronic allograft dysfunction......Page 2485
287 Cancer after kidney transplantation......Page 2497
288 Metabolic bone disease after renal transplantation......Page 2505
289 Recurrent renal disease: prophylaxis, diagnosis, and management......Page 2516
290 Paediatric renal transplantation......Page 2523
SECTION 14: Renal disease at different stages of life (infancy, adolescence, pregnancy, old age)......Page 2533
291 Growth and development......Page 2535
292 The adolescent with renal disease: transition to adult services......Page 2541
293 Contraception in patients
with kidney disease......Page 2544
294 Pregnancy and renal physiology......Page 2551
295 Pregnancy in patients with chronic kidney disease and on dialysis......Page 2563
296 Pre-eclampsia and related disorders......Page 2570
297 Acute kidney injury in pregnancy......Page 2580
298 Specific renal conditions in pregnancy......Page 2584
299 Pregnancy after renal transplantation......Page 2591
300 The kidney in ageing: biology, anatomy, physiology, and clinical relevance......Page 2594
SECTION 15: The patient with genetic renal disease......Page 2603
301 Ethical aspects of genetic testing......Page 2605
302 Antenatal diagnosis and pre-implantation genetic testing......Page 2608
303 The molecular basis of ciliopathies and cyst formation......Page 2611
304 The adult with renal cysts......Page 2631
305 The child with renal cysts......Page 2635
306 Autosomal dominant polycystic kidney disease: overview......Page 2639
307 Autosomal dominant polycystic kidney disease: clinical features......Page 2641
308 Autosomal dominant polycystic kidney disease: diagnosis......Page 2648
309 Autosomal dominant polycystic kidney disease: management......Page 2652
310 Management of intracranial
aneurysms......Page 2661
311 Management of cystic liver disease......Page 2663
312 Autosomal dominant polycystic
kidney disease in children
and young adults......Page 2667
313 Autosomal recessive polycystic kidney disease......Page 2670
314 Bardet–Biedl syndrome
and other ciliopathies......Page 2679
315 Hepatocyte nuclear factor-1B......Page 2685
316 Nephronophthisis and medullary cystic kidney disease: overview......Page 2688
317 Nephronophthisis......Page 2691
318 Autosomal dominant interstitial
kidney disease including
medullary cystic disease......Page 2696
319 Oral-facial-digital type 1 syndrome......Page 2699
320 The molecular basis of glomerular basement membrane disorders......Page 2702
321 Alport syndrome: overview......Page 2709
322 Alport syndrome: clinical features......Page 2711
323 Alport syndrome: diagnosis......Page 2714
324 Alport syndrome: management......Page 2720
325 Thin glomerular basement membrane nephropathy and other collagenopathies......Page 2723
326 Nail patella syndrome......Page 2725
327 Molecular basis of nephrotic syndrome......Page 2728
328 Molecular basis of renal tumour syndromes......Page 2736
329 WT1 and its disorders......Page 2741
330 Tuberous sclerosis complex renal disease......Page 2746
331 Hypoxia-inducible factor
and renal disorders......Page 2753
332 Von Hippel–Lindau disease......Page 2757
333 Molecular basis of complement-mediated renal disease......Page 2761
334 Inherited metabolic diseases and the kidney......Page 2771
335 Fabry disease: overview
and pathophysiology......Page 2781
336 Fabry disease: clinical features......Page 2784
337 Fabry disease: diagnosis......Page 2790
338 Fabry disease: management and outcome......Page 2794
339 Cystinosis......Page 2803
340 Mitochondrial diseases and the kidney......Page 2812
341 APOL1 and renal disease......Page 2822
342 MYH9 and renal disease......Page 2825
SECTION 16: The patient with structural and congenital abnormalities......Page 2827
343 Human kidney development......Page 2829
344 Kidney stem cells......Page 2836
345 Anatomical types of congenital anomalies: overview of obstruction......Page 2842
346 Renal agenesis......Page 2844
347 Renal dysplasia......Page 2845
348 Renal hypoplasia......Page 2846
349 Normal variation in nephron numbers......Page 2847
350 Renal tubular dysgenesis......Page 2849
351 Congenital solitary functioning kidney......Page 2850
352 Duplex, ectopic, and horseshoe
kidneys......Page 2852
353 Pelviureteric junction obstruction and megaureter in children......Page 2854
354 Posterior urethral valves......Page 2856
355 Primary vesicoureteric reflux and reflux nephropathy......Page 2858
356 The patient with urinary tract obstruction......Page 2868
357 Retroperitoneal fibrosis......Page 2875
358 Branchio-oto-renal syndrome......Page 2879
359 Townes–Brocks syndrome......Page 2881
360 Renal coloboma syndrome......Page 2883
361 Ante- and postnatal imaging to diagnose human kidney malformations......Page 2885
SECTION 17: Drugs and renal disease......Page 2897
362 Drug-induced nephropathies......Page 2899
363 Drug dosing in chronic kidney disease......Page 2925
364 Drug dosing in acute kidney injury......Page 2933
SECTION 18: Nephrology in the future......Page 2939
365 A global curriculum for training the next generation of nephrologists......Page 2941
A......Page 2947
B......Page 2954
C......Page 2956
D......Page 2964
E......Page 2967
F......Page 2969
G......Page 2971
H......Page 2973
I......Page 2978
K......Page 2981
L......Page 2982
M......Page 2984
N......Page 2987
O......Page 2990
P......Page 2991
R......Page 2997
S......Page 3001
T......Page 3004
U......Page 3008
V......Page 3010
Z......Page 3012

Citation preview

Oxford Textbook of

Clinical Nephrology

Free personal online access for 12 months Individual purchasers of this book are also entitled to free personal access to the online edition for 12 months on Oxford Medicine Online (www.oxfordmedicine.com). Please refer to the access token card for instructions on token redemption and access. Online ancillary materials, where available, are noted at the end of the respective chapters in this book. Additionally, Oxford Medicine Online allows you to print, save, cite, email, and share content; download high-resolution figures as Microsoft PowerPoint slides; save often-used books, chapters, or searches; annotate; and quickly jump to other chapters or related material on a mobile-optimised platform. We encourage you to take advantage of these features. If you are interested in ongoing access after the 12-month gift period, please consider an individual subscription or consult with your librarian.

VOLUME 1

Oxford Textbook of

Clinical Nephrology FOURTH EDITION Edited by

Managing Editors Neil Turner Norbert Lameire David J. Goldsmith Christopher G. Winearls Jonathan Himmelfarb Giuseppe Remuzzi

1

Section Editors William G. Bennett Jeremy R. Chapman Adrian Covic Marc E. De Broe Vivekanand Jha Neil Sheerin Robert Unwin Adrian Woolf

1 Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2016 The moral rights of the authors‌have been asserted First Edition Published in 1992 Second Edition Published in 1998 Third Edition Published in 2005 Fourth Edition Published in 2016 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2015938370 ISBN 978–0–19–870858–2 (volume 1) 978–0–19–870859–9 (volume 2) 978–0–19–870860–5 (volume 3) 978–0–19–959254–8 (set) Printed and bound in Great Britain by Bell & Bain, Glasgow. Oxford University Press makes no representation, express or implied, that the drug dosages in this book are correct. Readers must therefore always check the product information and clinical procedures with the most up-to-date published product information and data sheets provided by the manufacturers and the most recent codes of conduct and safety regulations. The authors and the publishers do not accept responsibility or legal liability for any errors in the text or for the misuse or misapplication of material in this work. Except where otherwise stated, drug dosages and recommendations are for the non-pregnant adult who is not breast-feeding Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Preface

We have almost completely rewritten the Oxford Textbook of Clinical Nephrology for its fourth edition. That was a big decision for a very successful book, but there were several important drivers. Huge developments in nephrology and medicine more broadly were only part of the reasoning. First, we wanted a text that would adapt easily to presentation in multiple formats—on paper and also electronically on different devices. Shorter (but more) chapters seemed an important part of this, but meant reviewing the organization of the entire book. However, this also helped us to meet our second objective, which was to make it easier to get quick answers to specific questions. And third, we wanted a structure that would aid updating in the future, both on paper and electronically.

Our authorship and editorship are substantially changed and even more international than before, with representation from every continent except Antarctica. Our coverage of global topics is substantially enhanced. There is an entirely new section on genetic diseases, and substantially increased coverage for the patient with a renal transplant. We have completely revised and combined previous sections on tubular disorders and electrolytes, achieving greater clarity and reduced duplication. There is new material on renal disease in childhood and old age, together with a completely new set of chapters on renal disease in pregnancy. The book is no longer but we believe it is substantially enhanced. This has been a huge project for the editors, project managers, and production team—but ultimately, a very rewarding one. We hope you will agree that it has been worthwhile.

Summary of Contents

VOLUME 1 SECTION 1

Assessment of the patient with renal disease  1 Edited by Christopher G. Winearls

SECTION 2

The patient with fluid, electrolyte, and renal tubular disorders  169 Edited by Robert Unwin and Pascal Houillier

SECTION 3

The patient with glomerular disease  423 Edited by Neil Turner

SECTION 4

The patient with interstitial disease  667 Edited by Adrian Covic

SECTION 5

The patient with reduced renal function  739 Edited by David J. Goldsmith

VOLUME 2 SECTION 6

The patient with another primary diagnosis  1197 Edited by Giuseppe Remuzzi

SECTION 7

The patient with urinary tract infection  1491 Edited by Neil Sheerin

SECTION 8

The patient with infections causing renal disease  1543 Edited by Vivekanand Jha

SECTION 9

The patient with urinary stone disease  1629 Edited by Marc E. De Broe

SECTION 10

The patient with hypertension  1723 Edited by Neil Turner

SECTION 11

The patient with acute kidney injury (and critical care nephrology)  1829 Edited by Norbert Lameire

VOLUME 3 SECTION 12

The patient on dialysis  2159 Edited by Jonathan Himmelfarb

viii

summary of contents

SECTION 13

SECTION 16

The transplant patient  2343

The patient with structural and congenital abnormalities  2813

Edited by Jeremy R. Chapman

Edited by Adrian Woolf

SECTION 14

Renal disease at different stages of life (infancy, adolescence, pregnancy, old age)  2519

SECTION 17

Drugs and renal disease  2883 Edited by William G. Bennett

Edited by Norbert Lameire and Neil Turner

SECTION 15

The patient with genetic renal disease  2589 Edited by Neil Turner

SECTION 18

Nephrology in the future  2925 Edited by Neil Turner

Contributors

Dwomoa Adu Department of Medicine, University of Ghana Medical School, University of Ghana, Accra, Ghana

Vicente Arroyo Liver Unit, Hospital Clinic, University of Barcelona, Barcelona, Spain

Behdad Afzali Medical Research Council Centre for Transplantation, King’s College London, London; NIHR Biomedical Research Centre, Guy’s & St Thomas’ NHS Foundation Trust, London; King’s College London, London, UK; and Lymphocyte Cell Biology Section, Molecular Immunology and Inflammation Branch, National Institutes of Arthritis, and Musculoskeletal and Skin Diseases, National Institutes of Health, Bethesda, MD, USA

Carla Maria Avesani Department of Applied Nutrition, Nutrition Institute, Rio de Janeiro State University, Rio de Janeiro, Brazil

Rajiv Agarwal Division of Nephrology, Department of Medicine, Indiana University School of Medicine, Indiana University; and Veterans Administration Medical Center, Indianapolis, IN, USA Tomas Thor Agustsson Department of Endocrinology and Metabolic Medicine, Landspitali; and The National University Hospital of Iceland, Reykjavik, Iceland Bassam Alchi Renal Unit, Addenbrooke’s Hospital, Cambridge, UK Helen Alderson Institute of Inflammation and Repair, University of Manchester, Manchester, UK

Fred E. Avni Department of Medical Imaging, University Clinics of Brussels—Erasme Hospital, Brussels, Belgium Seema Baid-Agrawal Division of Nephrology and Medical Intensive Care, Department of Medicine, Charite Medical University, Berlin, Germany Colin Baigent Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, University of Oxford, Oxford, UK Matthew A. Bailey The British Heart Foundation Centre for Cardiovascular Science, The University of Edinburgh, Edinburgh, UK Richard Baker Renal Unit, St. James’s University Hospital, Leeds, UK

Ased Ali Freeman Hospital, Newcastle upon Tyne, UK

Joanne Bargman University of Toronto, Toronto; University Health Network, Toronto; and Toronto General Hospital, Toronto, ON, Canada

Richard D. M. Allen University of Sydney, Sydney; and Royal Prince Alfred Hospital, Sydney, Australia

Rashad S. Barsoum Cairo University, The Cairo Kidney Centre, Cairo, Egypt

Michael Allon Division of Nephrology, University of Alabama at Birmingham, Birmingham, AL, USA

Philip Beales Molecular Medicine Unit, UCL Institute of Child Health, London, UK

Halima Amer King’s College London BHF Centre, The Rayne Institute, St Thomas’ Hospital, London, UK

Monica Beaulieu Division of Nephrology, University of British Columbia, Vancouver, BC, Canada

Corinne Antignac Molecular Genetics Laboratories and Paediatrics, Inserm U983, Hôpital Necker-Enfants Malades, Paris, France

Aminu K. Bello Division of Nephrology & Immunology, University of Alberta, Edmonton, AB, Canada

Mugurel Apetrii Clinic of Nephrology, ‘C. I. Parhon’ University Hospital, ‘Gr. T. Popa’ University of Medicine and Pharmacy, Iasi, Romania

Rinaldo Bellomo Department of Intensive Care, Austin Hospital, Heidelberg, VIC, Australia

xxxii

contributors Katherine Bennett-Richards Department of Nephrology, Whipps Cross University Hospital, London, UK Jo H. M. Berden Department of Nephrology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands Carsten Bergmann Center for Human Genetics, Bioscientia, Ingelheim; University Hospital Freiburg, Freiburg; and Department of Human Genetics, RWTH Aachen University, Aachen, Germany Jürg Biber Institute of Physiology, University of Zurich, Zurich, Switzerland Daniel G. Bichet University of Montreal; and Hôpital du SacréCœur de Montréal, Montréal, QC, Canada Scott D. Bieber Division of Nephrology, Department of Medicine, University of Washington, Seattle; and Northwest Kidney Centers, Seattle, WA, USA Patrick Biggar Division of Nephrology, Klinikum Coburg GmbH, Coburg, Germany Heiko Billing Department of Pediatrics I, University Children’s Hospital, Heidelberg, Germany Coralie Bingham Exeter Kidney Unit, Royal Devon and Exeter Hospital (Wonford), Exeter, UK Lesley-Anne Bissell Division of Rheumatic and Musculoskeletal Disease, Leeds Institute of Molecular Medicine, University of Leeds, Leeds, UK

Cormac Breen Department of Nephrology and Transplantation, Guy’s and St Thomas NHS Foundation Trust, London, UK Jeremiah R. Brown The Dartmouth Institute for Health Policy and Clinical Practice and Departments of Medicine and Community and Family Medicine, Audrey and Theodor Geisel School of Medicine at Dartmouth, Hanover, NH; and Providence Park Heart Institute, Novi, MI, USA Nele Brusselaers Department of Molecular Medicine and Surgery, Karolinska Institutet, Stockholm, Sweden; and Department of Intensive Care Medicine and Burn Unit, Ghent University Hospital, Ghent University, Ghent, Belgium Emmanuel A. Burdmann Division of Nephrology, University of Sao Paulo Medical School, Sao Paulo, Brazil Aine Burns University College London (UCL) Centre for Nephrology Royal Free Hospital, Royal Free NHS Trust, London, UK David A. Bushinsky University of Rochester School of Medicine; and Nephrology Division, University of Rochester Medical Center, Rochester, NY, USA Patrice Cacoub UPMC Univ Paris 06, UMR 7211, Paris; INSERM, UMR_S 959, Paris; CNRS, UMR 7211, Paris; AP-HP, Groupe Hospitalier Pitié-Salpêtrière, Department of Internal Medicine, Paris, France Beverly D. Cameron Diabetes Education Center, St. John Providence Hospital, Southfield, MI

John J. Bissler Division of Nephrology and Hypertension, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA

Stewart Cameron Department of Nephrology, Dialysis and Transplantation, King’s College London (Guy’s Hospital), London, UK

Christopher R. Blagg John Radcliffe Hospital, Oxford, UK; and Northwest Kidney Centers, University of Washington, Seattle, WA, USA

Angel Candela-Toha Department of Anaesthesiolgy and Reanimation, Hospital Universitario Ramón y Cajal, Madrid; Consorcio de investigación del Fracaso Renal Agudo de la Comunidad de Madrid (CIFRA), Madrid; and Instituto Ramón y Cajal de Investigación Sanitaria, Madrid, Spain

Judith Blaine Division of Renal Diseases and Hypertension, Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Detlef Bockenhauer Great Ormond Street Hospital for Children NHS Foundation Trust, London; and Institute of Child Health, University College London, London, UK Davide Bolignano Hypertension Unit, Department of Nephrology and Urology and CNR-Institute of Clinical Physiology, Ospedali Riuniti, Reggio Calabria, Italy Gregory L. Braden Tufts University School of Medicine, Baystate Medical Center, Springfield, MA, USA Kate Bramham Division of Women’s Health, King’s College London; and Women’s Health Academic Centre KHP, St Thomas’ Hospital, London, UK

Giovambattista Capasso Department of Internal Medicine and Department of Cardio-thoracic and Respiratory Science, Second University of Naples, Naples, Italy Andrés Cárdenas GI Unit, Institut Clinic de Malalties Digestives i Metaboliques, Hospital Clinic, and University of Barcelona, Institut d’Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Ciber de Enfermedades Hepáticas y Digestivas (CIBEREHD), Barcelona; and InstitutoReina Sofıa de Investigación Nefrológica (IRSIN), Spain Juan Jesús Carrero Divisions of Baxter Novum and Renal Medicine, Department of Clinical Sciences, Intervention and Technology, Karolinska Institutet, Stockholm, Sweden

 

contributors

Paul Carroll Consultant Endocrinologist, Guy’s & St. Thomas’ NHS Foundation Trust, London, UK

Lewis M. Cohen Division of Psychiatry, Baystate Medical Center, Tufts University School of Medicine, Springfield, MA, USA

Marie Cassart Department of Medical Imaging, University Clinics of Brussels—Erasme Hospital, Brussels, Belgium

Marie Condon Imperial College NHS Healthcare Trust Lupus Centre, Hammersmith Hospital, London, UK

Daniel C. Cattran University of Toronto, Toronto; Toronto General Research Institute, University Health Network; and Division of Nephrology, University Health Network, Toronto, ON, Canada

Thomas Connor Imperial College, London, UK

Steven Chadban Renal Unit, Royal Prince Alfred Hospital, Sydney, Australia

Jeff S. Coombes School of Human Movement Studies, University of Queensland, Brisbane, Australia

Christopher T. Chan Division of Nephrology, University Health Network, Toronto; and University of Toronto, Toronto, ON, Canada

Bruce A. Cooper Department of Renal Medicine, Kolling Institute of Medical Research, Royal North Shore Hospital, University of Sydney, St Leonards, Australia

Jung-San Chang Department of Renal Care, Kaohsiung Medical University, Kaohsiung, Taiwan

Cathy Corbishley Department of Cellular Pathology, St George’s Hospital, London, UK

Jeremy R. Chapman Renal Unit, Westmead Hospital, Westmead; and Centre for Transplant and Renal Research, Westmead Millennium Institute, University of Sydney, Westmead, Australia

Josef Coresh Cardiovascular Epidemiology, Comstock and Welch Centers, Johns Hopkins University, Bloomberg School of Public Health, Baltimore, MD, USA

Christos Chatziantoniou Hôpital Tenon, Paris; and Université Pierre et Marie Curie, Paris, France Dominique Chauveau Service de Néphrologie et Immunologie Clinique, CHU de Rangueil, Toulouse, France

Terry Cook Imperial College London, Hammersmith Campus, The Commonwealth Building, Hammersmith Hospital, UK

Adrian Covic Department of Nephrology, University of Medicine and Pharmacy ‘Gr. T. Popa’, Nephrology Clinic, ‘Dr. C. I. Parhon’ Hospital, Iaşi, Romania Sue Cox Guy’s & St Thomas’ NHS Foundation Trust, London, UK

Diana Chiu Salford Royal NHS Foundation Trust, Salford, UK

Gilles Crambert Metabolism and Renal Physiology Laboratory, Sorbonne Universités, UPMC Univ Paris 06, INSERM, Université Paris Descartes, Sorbonne Paris Cité, UMR_S 1138, CNRS ERL8226, Centre de Recherche des Cordeliers, F-75006, Paris, France

Bhavna Chopra Transplant Nephrology, Mayo Clinic, Rochester, MN; and Nephrology, University of Pennsylvania, Pittsburgh, PA, USA

Antonia J. Cronin Guy’s Hospital Renal Unit, Guy’s and St Thomas’ Foundation Hospital, King’s Health Partners AHSC, London, UK

Constantina Chrysochou Salford Royal NHS Foundation Trust, Salford, UK

Dinna N. Cruz Division of Nephrology-Hypertension, Department of Medicine, University of California, San Diego, San Diego, CA, USA; and International Renal Research Institute (IRRIV), Vicenza, Italy

Hung-Chun Chen Department of Renal Care, Kaohsiung Medical University, Kaohsiung, Taiwan

William R. Clark Renal Therapeutic Area, Baxter Healthcare, Deerfield, IL, USA Philip Clayton Renal Unit, Royal Prince Alfred Hospital, Sydney, Australia Jan Clement National Reference Laboratory for Hantavirus Infections, Clinical Virology, University Hospital Gasthuisberg, University of Leuven, Leuven, Belgium Anna Clementi Department of Nephrology, Policlinico Universitario, Catania, Italy Eric P. Cohen Medical College of Wisconsin, Milwaukee, WI, USA

Zhao Cui Renal Division, Peking University First Hospital, Institute of Nephrology, Peking University, Key Laboratory of Renal Disease, Ministry of Health of China, Beijing, P. R. China Gary C. Curhan Channing Laboratory and Renal Division, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA; and Division of Nephrology and Transplantation, Maine Medical Center, Portland, ME, USA Darshana Dadhania Division of Nephrology, Department of Medicine, Department of Transplantation Medicine, Weill Medical College of Cornell University, New York, USA

xxxiii

xxxiv

contributors Michel Daudon Service des Explorations Fonctionnelles, Hôpital Tenon, APHP, Paris, France Andrew Davenport UCL Centre for Nephrology, Royal Free Campus, University College London Medical School, London, UK Marc E. De Broe Department of Nephrology, University of Antwerp, Antwerp, Belgium Benjamin Dekel The Pediatric Stem Cell Research Institute and Division of Pediatric Nephrology, Edmond and Lili Safra Children’s Hospital, Chaim Sheba Medical Center, Tel Hashomer, Sackler Faculty of Medicine, Tel Aviv University, Israel Joris Delanghe Department of Clinical Chemistry, Ghent University Hospital, Gent, Belgium Olivier Devuyst Institute of Physiology, University of Zurich, Zurich; and Division of Nephrology, University Hospital Zurich, Zurich, SwitzerlandUniversité catholique de Louvain Medical School, Brussels, Belgium

Duska Dragun Department of Nephrology and Medical Intensive Care, Charité University Hospital, Campus Virchow-Klinikum, Berlin, Germany Joost P. H. Drenth Department of Gastroenterology and Hepatology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands Jeremy S. Duffield Division of Nephrology and Center for Lung Biology, Departments of Medicine & Pathology, and Institute for Stem Cell and Regenerative Medicine, University of Washington, Seattle, WA, USA Sara Dunsmore Toronto General Hospital, Toronto, ON, Canada Jean-Claude Dussaule Service de Physiologie HUPE, Assistance Publique-Hôpitaux de Paris (AP-HP), Paris; Université Pierre et Marie Curie, Paris; and INSERM U702, Hôpital Tenon, Paris, France John Eastwood Department of Renal Medicine and Transplantation, St George’s Hospital, London, UK

Patrick C. D’Haese Laboratory of Pathophysiology, University of Antwerp, Antwerp, Belgium

Tim Eisen Department of Oncology, Cambridge University Hospitals NHS Foundation Trust, Cambridge, UK

Ramya Dhandapani Royal Liverpool University Hospital, Liverpool, UK

Grahame J. Elder Department of Renal Medicine, Westmead Hospital, Westmead; and Osteoporosis and Bone Biology Division, Garvan Institute of Medical Research, Sydney, Australia

Neeraj Dhaun British Heart Foundation Centre for Cardiovascular Science, Queen’s Medical Research Institute, University of Edinburgh; and Department of Renal Medicine, Royal Infirmary of Edinburgh, Edinburgh, UK Meletios A. Dimopoulos Department of Clinical Therapeutics, National and Kapodistrian University of Athens School of Medicine, Athens, Greece Salvatore Di Filippo Department of Nephrology, Dialysis and Renal Transplant, Alessandro Manzoni Hospital, Lecco, Italy Bradley P. Dixon Division of Nephrology and Hypertension, Cincinnati Children’s Hospital Medical Center, Cincinnati, OH, USA

Marlies Elger Centre for Biomedicine and Medical Technology Mannheim, Department of Anatomy and Developmental Biology, Medical Faculty Mannheim, University of Heidelberg, Heidelberg, Germany Timothy Ellam Department of Cardiovascular Science, University of Sheffield Medical School, Sheffield, UK David H. Ellison Division of Nephrology & Hypertension, Oregon Health & Science University & Renal Section VAMC, Portland, OR, USA

Luis D’Marco Hospital Universitario Ruíz y Páez, Universidad de Oriente, Venezuela

Paul Emery Leeds Musculoskeletal Biomedical Research Unit, Leeds Teaching Hospitals NHS Trust, Leeds; and Rheumatic and Musculoskeletal Disease, Leeds Institute of Molecular Medicine, University of Leeds, Leeds, UK

Philippa Dodd Imperial College NHS Healthcare Trust Lupus Centre, Hammersmith Hospital, London, UK

Karlhans Endlich Department of Anatomy and Cell Biology, University Medicine Greifswald, Greifswald, Germany

Anthony Dorling MRC Centre for Transplantation, King’s College London, London; and Guy’s and St Thomas’ NHS Foundation Trust, Guy’s Hospital, London, UK

Fabrizio Fabrizi Division of Nephrology, Maggiore Policlinico Hospital, IRCCS Foundation, Milano, Italy

Alain Doucet Metabolism and Renal Physiology Laboratory, Sorbonne Universités, UPMC Univ Paris 06, INSERM, Université Paris Descartes, Sorbonne Paris Cité, UMR_S 1138, CNRS ERL8226, Centre de Recherche des Cordeliers, F-75006, Paris, France

Robert G. Fassett Royal Brisbane and Women’s Hospital, Brisbane; and School of Human Movement Studies, University of Queensland, Brisbane, Australia Bengt Fellström University of Uppsala, Uppsala, Sweden

 

contributors

Javier Fernández Liver Unit, Hospital Clinic Barcelona, IDIBAPS, Ciberhed, Barcelona, Spain

Michael J. Germain Baystate Medical Center, Tufts University School of Medicine, Springfield, MA, USA

Charles J. Ferro Department of Renal Medicine, Queen Elizabeth Hospital and University of Birmingham, Birmingham, UK

Morie A. Gertz Division of Hematology, Mayo Clinic, Rochester, MN, USA

Fernando C. Fervenza Division of Nephrology and Hypertension, Mayo Clinic, Rochester, MN, USA

Tom J. G. Gevers Department of Gastroenterology and Hepatology, Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands

Alexander Fichtner Division of Paediatric Nephrology, Centre for Paediatrics and Adolescent Medicine, University of Heidelberg, Heidelberg, Germany Fredric O. Finkelstein Yale University, New Haven, CT, USA Susan H. Finkelstein Yale University, New Haven, CT, USA Martin Flamant Hopital Bichat, Paris; and Université Diderot, Paris, France Stewart Fleming University of Dundee, Ninewells Hospital, Dundee, UK Frances Flinter Genetics Department, Guy’s & St Thomas’ NHS Foundation Trust, London, UK Danilo Fliser Department of Internal Medicine IV, Saarland University Medical Hospital, Homburg/Saar, Germany Lukas Foggensteiner The New Queen Elizabeth Hospital Birmingham, Edgbaston, Birmingham, UK Agnes B. Fogo Pathology Department, Vanderbilt University Medical Center, Nashville , TN, USA Ted A. Foster Oregon Health and Science, University Division of Internal Medicine, Portland, OR, USA Brunella Franco Telethon Institute of Genetics and Medicine (TIGEM), Department of Pediatrics Medical Genetics Services, Federico II University of Naples, Naples, Italy Valeska Frank Center for Human Genetics, Bioscientia, Ingelheim, Germany Benjamin J. Freda Tufts University School of Medicine, Baystate Medical Center, Springfield, MA; and Continuous Renal Replacement Therapies, Western New England Renal and Transplantation Associates, Springfield, MA, USA Simon J. Freeman Plymouth Hospitals NHS Trust, Radiology Department, Derriford Hospital, Crownhill, Plymouth, UK Daniel P. Gale UCL Centre for Nephrology, Royal Free Hospital School of Medicine, London, UK Timur A. Galperin St. Louis University, Department of Dermatology, Anheuser Busch Institute, St. Louis, MO, USA Giorgio Gentile IRCCS—Istituto di Ricerche Farmacologiche Mario Negri, Bergamo; and Unit of Nephrology, Azienda Ospedaliera Papa Giovanni XXIII, Bergamo, Italy

Pere Ginès Liver Unit, Institut Clinic de Malalties Digestives i Metaboliques, Hospital Clinic, and University of Barcelona, Institut d’Investigacions Biomèdiques August Pi-Sunyer (IDIBAPS), Ciber de Enfermedades Hepáticas y Digestivas (CIBEREHD), Barcelona; and InstitutoReina Sofıa de Investigación Nefrológica (IRSIN), Spain Hector Giral Division of Renal Diseases and Hypertension, University of Colorado at Denver and Health Sciences Center, Aurora, CO, USA Anna Giuliani Department of Nephrology Dialysis & Transplantation, San Bortolo Hospital, and International Renal Research Institute (IRRIV), Vicenza, Italy Richard J. Glassock Geffen School of Medicine at UCLA, Los Angeles, CA, USA James F. Glockner Department of Radiology, Mayo Medical School, Mayo Clinic, Rochester MN, USA Griet Glorieux Nephrology Division, University Hospital Ghent, Ghent, Belgium Luigi Gnudi Unit for Metabolic Medicine, Cardiovascular Division, King’s College London, London, UK Rishi M. Goel Department of Gastroenterology, Guy’s & St Thomas’ NHS Foundation Trust, London, UK Eric Goffin Department of Nephrology, Cliniques Universitaires Saint-Luc, Université Catholique de Louvain, Brussels, Belgium M. Refik Gökmen Department of Nephrology & Transplantation, Guy’s & St Thomas’ NHS Foundation Trust, London, UK Stanley Goldfarb Renal Electrolyte and Hypertension Division, University of Pennsylvania, Pittsburgh, PA, USA David J. Goldsmith Consultant Nephrologist, Guy’s and St Thomas’ Hospitals, London, UK; Professor, Cardiovascular and Cell Sciences Institute, St George’s University of London, London, UK Michael S. Goligorsky Renal Research Institute, New York Medical College, Valhalla, NY, USA Thomas A. Golper Vanderbilt University Medical Center, Medical Center North, Nashville, TN, USA Olga Gonzalez-Albarrán Endocrinology Dapartment, Hospital Universitario Ramón y Cajal, Madrid, Spain

xxxv

xxxvi

contributors Tim Goodship Institute of Human Genetics, Newcastle University, International Centre for Life, Newcastle upon Tyne, UK Paul Goodyer Montreal Children’s Hospital Research Institute, Montreal, QC, Canada Morgan E. Grams The Johns Hopkins Hospital, Baltimore, MD, USA John Grange London Clinic Cancer Centre B2, London, UK Darren Green University of Manchester, Manchester, UK Simon Gruenewald Departments of Nuclear Medicine and Radiology, Westmead Hospital, Westmead, Australia Marie Claire Gubler INSERM U983, Hôpital Necker-Enfants Malades, Paris; and APHP, Centre de référence des Maladies Rénales Héréditaires (MARHEA), Paris, France Victor Gueutin Department of Nephrology, Pitié-Salpêtrière Hospital, Paris, France Mónica Guevara Liver Unit, Hospital Clinic Barcelona, IDIBAPS, Ciberhed, Barcelona, Spain Theresa A. Guise Endocrinology and Metabolism, Department of Internal Medicine, Indiana University School of Medicine, Indianapolis, IN, USA Orlando M. Gutiérrez Division of Nephrology, Department of Medicine, School of Medicine; and Department of Epidemiology, School of Public Health, University of Alabama at Birmingham, Birmingham, AL, USA

Choli Hartono Division of Nephrology, Department of Medicine, Department of Transplantation Medicine, Weill Medical College of Cornell University, New York, USA Dietrich Hasper Department of Nephrology and Medical Intensive Care, Charité University Hospital, Campus VirchowKlinikum, Berlin, Germany Nick Hastie MRC Human Genetics Unit, University of Edinburgh Western General Hospital, Edinburgh, UK Jean-Philippe Haymann Department of Physiology, Hopital Tenon (AP-HP/UPMC/UMRS 702), Paris, France Richard Haynes Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, University of Oxford; and Oxford Kidney Unit, Oxford University Hospitals, Oxford, UK Peter Heeringa Department of Pathology and Medical Biology, University Medical Center Groningen, Groningen, The Netherlands Björn Hegner Department of Nephrology and Medical Intensive Care, Charité University Hospital, Campus Virchow-Klinikum, Berlin, Germany Laurence Heidet AP-HP, Service de Néphrologie Pédiatrique, Hôpital Necker-Enfants Malades, Paris; and APHP, Centre de référence des Maladies Rénales Héréditaires (MARHEA), Paris, France Udo Helmchen Department of Pathology, University of Hamburg, Hamburg, Germany

Mark Haas Department of Pathology and Laboratory Medicine, Cedars-Sinai Medical Center, Los Angeles, CA, USA

Heather Henderson Division of Nephrology, St. John Hospital and Medical Center, Detroit, MI, USA

Reza Hajhosseiny BHF Centre of Cardiovascular Excellence, Guy’s & St. Thomas’ NHS Foundation Trust, King’s College Academic Health Partners, London, UK

Lorna K. Henderson Renal Unit, Westmead Hospital, Westmead, Australia

Andrew Hall Institute of Anatomy, University of Zurich, Zurich, Switzerland Michèle Hall Department of Pediatric Nephrology, University Clinics of Brussels—Erasme Hospital, Brussels, Belgium

William G. Herrington Clinical Trial Service Unit, Richard Doll Building, Oxford; Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, Richard Doll Building, Oxford; and Oxford Kidney Unit, Oxford University Hospitals NHS Trust, Churchill Hospital, Oxford, UK

Mitchell L. Halperin Li Ka Shing Knowledge Institute in St. Michael’s Hospital, Toronto, ON, Canada

Irene J. Higginson King’s College London, London, UK

Mark Harber UCL Centre for Nephrology, Royal Free Hospital, London, UK

Jonathan Himmelfarb Division of Nephrology, Department of Medicine, Kidney Research Institute, University of Washington, Seattle, WA, USA

Lorraine Harper School of Immunity and Infection, Centre for Translational Inflammation Research, Queen Elizabeth Hospital Birmingham, Edgbaston, Birmingham, AL, USA

Hallvard Holdaas Rikshospitalet, Uniiversity of Oslo, Oslo, Norway

David C. H. Harris Sydney Medical School-Westmead, University of Sydney, Westmead Hospital, Westmead, Australia

Ewout Hoorn Division of Nephrology & Transplantation, Department of Internal Medicine, Erasmus Medical Center, Rotterdam, The Netherlands

 

Eric A. J. Hoste Department of Intensive Care Medicine, Ghent University Hospital, Ghent, Belgium Pascal Houillier Paris Descartes University; and Hopital Europeen Georges Pompidou, Assistance Publique-Hopitaux de Paris, Paris, France

contributors

Sabina Jelen Division of Renal Diseases and Hypertension, Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO, USA Vivekanand Jha Department of Nephrology, Postgraduate Institute of Medical Education and Research, Chandigarh, India

Frédéric A. Houssiau Service de Rhumatologie, Département de Médecine Interne, Cliniques universitaires Saint-Luc, Université catholique de Louvain, Bruxelles, Belgium

Achim Jörres Department of Nephrology and Medical Intensive Care, Charité University Hospital, Campus Virchow-Klinikum, Berlin, Germany

Erin J. Howden School of Human Movement Studies, University of Queensland, Brisbane, Australia

Dennis Joseph Division of Endocrinology, Department of Medicine, Indiana University School of Medicine, Indianapolis, IN, USA

Peter Howells Department of Medical Physics, Old Medical School, Leeds General Infirmary, Leeds, UK Jeremy Hughes University of Edinburgh, Renal Medicine, Edinburgh Royal Infirmary, Edinburgh, UK Beverley J. Hunt MRC Centre for Transplantation, King’s College London, London; and Guy’s and St Thomas’ NHS Foundation Trust, Guy’s Hospital, London, UK Alastair J. Hutchison Department of Renal Medicine, Manchester Academic Health Science Centre, Manchester, UK Young-Hwan Hwang Division of Nephrology, University Health Network and University of Toronto, Toronto, ON, Canada; and Department of Internal Medicine, Eulji General Hospital, Seoul, South Korea Lesley A. Inker Division of Nephrology, Tufts Medical Center and Tufts University School of Medicine, Boston, MA, USA Nadia Wasi Iqbal Division of Nephrology, Saint Louis University, St. Louis, MO, USA Hassan Izzedine Department of Nephrology, Pitié-Salpêtrière Hospital, Paris, France Bertrand L. Jaber Department of Medicine, Tufts University School of Medicine, St. Elizabeth’s Medical Center, Boston, MA, USA Michel Jadoul Department of Nephrology, Cliniques Universitaires Saint-Luc, Université Catholique de Louvain, Brussels, Belgium Kitty J. Jager Department of Medical Informatics, Academic Medical Center, Amsterdam, The Netherlands

Islam Junaid Department of Urology, The Royal London Hospital, London, UK Paul Jungers Département de Néphrologie, Hôpital Necker, APHP, Paris, France Philip A. Kalra Genomic Epidemiology Research Group, The Univeristy of Manchester, Manchester, UK Kamel S. Kamel Keenan Research Center in the Li Ka Shing Knowledge Institute in St Michael’s Hospital and Division of Nephrology, St. Michael’s Hospital, University of Toronto, Toronto, ON, Canada Mehmet Kanbay Division of Nephrology, Department of Medicine, Koc University School of Medicine, Istanbul, Turkey Duk-Hee Kang Division of Nephrology, Department of Internal Medicine, Ewha Womans University School of Medicine, Ewha Medical Research Center, Seoul, Korea Ea Wha Kang Division of Nephrology, Department of Internal Medicine, NHIC Ilsan Hospital, Goyang, Korea Efstathios Kastritis Department of Clinical Therapeutics, National and Kapodistrian University of Athens School of Medicine, Athens, Greece Akira Kawashima Department of Radiology, Mayo Medical School, Mayo Clinic, Rochester, MN, USA John A. Kellum Department of Critical Care Medicine, Center for Critical Care Nephrology, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

Mirian C. Janssen Radboud University Nijmegen Medical Centre, Nijmegen, The Netherlands

Claudia Kemper Medical Research Council Centre for Transplantation, King’s College London and NIHR Biomedical Research Centre at Guy’s & St Thomas’ NHS Foundation Trust and King’s College London, UK

Alan G. Jardine BHF Cardiovascular Research Centre, University of Glasgow, Glasgow, UK

Steven Kennish Department of Radiology, Royal Hallamshire Hospital, Sheffield

David Jayne Vasculitis and Lupus Clinic, Addenbrooke’s Hospital, Cambridge, UK

Markus Ketteler Division of Nephrology, Klinikum Coburg GmbH, Coburg, Germany

xxxvii

xxxviii

contributors Kaivan Khavandi BHF Centre of Cardiovascular Excellence, Guy’s & St. Thomas’ NHS Foundation Trust, King’s College Academic Health Partners, London, UK

Wolfgang Kühn Renal Division, University Freiburg Medical Center, Freiburg; and BIOSS Center for Biological Signalling Studies, University of Freiburg, Freiburg, Germany

J. Christopher Kingswood Sussex Renal Unit, The Royal Sussex County Hospital, Brighton, UK

Navneet Kumar Division of Cardiology, St. John Hospital and Medical Center, Detroit, MI, USA

Kazuhiro Kitajima Department of Radiology, Mayo Medical School, Mayo Clinic, Rochester MN, USA

Manjula Kurella Tamura Division of Nephrology, Stanford University School of Medicine, Stanford, CA; and VA Palo Alto Health Care System Geriatrics Research Education & Clinical Center, Palo Alto, CA, USA

Robert Kleta University College London, Royal Free Hospital, London, UK Alan S. Kliger Hospital of Saint Raphael, New Haven; and Yale University School of Medicine, New Haven, CT, USA Bertrand Knebelmann Service de Néphrologie, Hôpital NeckerEnfants Malades, Paris, France Simon R. Knight Nuffield Department of Surgical Sciences, University of Oxford, Oxford; and Centre for Evidence in Transplantation, Clinical Effectiveness Unit, The Royal College of Surgeons of England, London, UK Harbir Singh Kohli Post Graduate Institute of Medical Education and Research, Chandigarh, India

Dirk R. J. Kuypers Department of Nephrology and Renal Transplantation, University Hospitals Leuven, Leuven, Belgium Laura Labriola Department of Nephrology, Cliniques Universitaires Saint-Luc, Université Catholique de Louvain, Brussels, Belgium Helen J. Lachmann UK National Amyloidosis Centre, University College London School of Medicine, London, UK Robin Lachmann Charles Dent Metabolic Unit, National Hospital for Neurology and Neurosurgery, London, UK Kar Neng Lai University of Hong Kong, Hong Kong

Martin Konrad University Children’s Hospital, Pediatric Nephrology, Münster; and University Hospital Münster, Department of General Pediatrics, Pediatric Nephrology, Münster, Germany

Edmund J. Lamb Clinical Biochemistry, East Kent Hospitals University NHS Foundation Trust, Canterbury, Kent, UK

Stephen M. Korbet Department of Medicine, Rush University Medical Center, Chicago, IL, USA

Anne-Sophie Lambert Department of Medical Imaging, University Clinics of Brussels—Erasme Hospital, Brussels, Belgium

Maarit Korkeila Divisions of Renal Medicine and Baxter Novum, Department of Clinical Science, Intervention and Technology, Karolinska Institutet, Stockholm, Sweden Camille Nelson Kotton Infectious Diseases Division, Massachusetts General Hospital/Harvard Medical School, Boston, MA, USA Jeannette Kathrin Kraft Clarendon Wing Radiology Department, Leeds General Infirmary, Leeds, UK Wilhelm Kriz Centre for Biomedicine and Medical Technology Mannheim (CBTM), Anatomy and Developmental Biology, Medical Faculty Mannheim, University of Heidelberg, Mannheim, Germany Wilhelm Kriz Department of Anatomy and Developmental Biology, Medical Faculty Mannheim, Centre for Biomedicine and Medical Technology Mannheim, University of Heidelberg, Heidelberg, Germany Jean-Marie Krzesinski Department of Medicine, Nephrology Unit, B35, University of Liège, Belgium

Heather Lambert Newcastle University, Newcastle upon Tyne, UK

Norbert Lameire University Hospital Ghent, Ghent, Belgium Martin J. Landray Clinical Trial Service Unit and Epidemiological Studies Unit, Nuffield Department of Population Health, University of Oxford, Oxford, UK Christophe Legendre Université Paris Descartes and Hôpital Necker and INSERM U845, Paris; and Service de Transplantation Rénale Adulte, Hôpital Necker, Paris, France Rachel Lennon Wellcome Trust Centre for Cell-Matrix Research Michael Smith Building, University of Manchester, Manchester, UK Edgar V. Lerma Section of Nephrology, Department of Medicine, University of Illinois at Chicago College of Medicine, Chicago, IL, USA Andrew J. LeRoy Department of Radiology, Mayo Medical School, Mayo Clinic, Rochester, MN, USA Kieron S. Leslie Department of Dermatology, University of California San Francisco, San Francisco, CA, USA

 

Andrew S. Levey Tufts Medical Center and Tufts University School of Medicine, Boston; and US Department of Agriculture Jean Mayer Human Nutrition Research Center at Tufts University, Boston; and Clinical Research Graduate Program (PhD and MS), Sackler School of Graduate Biomedical Sciences, Boston, MA, USA Moshe Levi Division of Renal Diseases and Hypertension, Department of Medicine, University of Colorado Anschutz Medical Campus, Aurora, CO, USA

contributors

Rodger Loutzenhiser Department of Physiology and Pharmacology, University of Calgary, Calgary, AB, Canada Valerie A. Luyckx Division of Nephrology, University of Alberta, Edmonton, AB, Canada Bernadette Lynch Centre for Rheumatology, Royal Free Hospital, London, UK Naim Maalouf Department of Internal Medicine, University of Texas, Southwestern Medical Center, TX, USA

Adeera Levin Division of Nephrology, University of British Columbia, Vancouver; and BC Provincial Renal Agency, Vancouver, BC, Canada

Iain C. Macdougall Department of Renal Medicine, King’s College Hospital, London, UK

Elena N. Levtchenko Department of Pediatric Nephrology, University Hospitals Leuven, Leuven, Belgium

Iain M. MacIntyre Department of Clinical Pharmacology, Royal Infirmary of Edinburgh, Edinburgh, UK

Edmund J. Lewis Department of Medicine, Rush University Medical Center, Chicago, IL, USA

Friederike Mackensen Interdisciplinary Uveitis Center, Department of Ophthalmology, University of Heidelberg, Heidelberg, Germany

Fernando Liaño Consorcio de investigación del Fracaso Renal Agudo de la Comunidad de Madrid (CIFRA), Madrid; and Instituto Ramón y Cajal de Investigación Sanitaria, Madrid; and Department of Nephrology, Hospital Universitario Ramón y Cajal, Madrid; and Department of Medicine, Universidad de Alcalá, Alcalá de Henares, Spain

Finlay MacKenzie United Kingdom National External Quality Assessment Scheme (UKNEQAS), University Hospitals Birmingham NHS Foundation Trust, East Wing, Institute of Research and Development, Birmingham, UK

Aurora Lietor Department of Critical Care Medicine, Hospital Universitario Ramón y Cajal, Madrid, Spain Liz Lightstone Imperial College NHS Healthcare Trust Lupus Centre, Hammersmith Hospital, London; and Section of Renal Medicine and Vascular Inflammation, Department of Medicine, Imperial College London, London, UK Bengt Lindholm Divisions of Baxter Novum and Renal Medicine, Department of Clinical Sciences, Intervention and Technology, Karolinska Institutet, Stockholm, Sweden Agnès Linglart Department of Pediatric Endocrinology, Reference Center for Rare Disorders of the Mineral Metabolism, Paris, France Francois Lionnet Sickle cell Center, Hopital Tenon (AP-HP), Paris, France Jeffrey Lipman Department of Intensive Care Medicine, Royal Brisbane and Womens Hospital, Herston, Australia Francesco Locatelli Department of Nephrology, Dialysis and Renal Transplant, Alessandro Manzoni Hospital, Lecco, Italy Gerard M. London Department of Pharmacology, Georges Pompidou European Hospital, National Institute of Health and Medical Research U970, Paris; and Department of Nephrology, Manhès Hospital, Fleury Mérogis, France Graham M. Lord Department of Experimental Immunobiology, King’s College London, London; and Guy’s & St Thomas’ NHS Foundation Trust, London, UK

Piet Maes National Reference Laboratory for Hantavirus Infections, Clinical Virology, University Hospital Gasthuisberg, University of Leuven, Leuven, Belgium Julien Maizel Department of Nephrology, Medical Intensive Care Center, Amiens University Medical Center, Amiens, France Francesca Mallamaci Nephrology and Renal Transplantation Unit, Department of Nephrology and Urology and CNR-IBIM, Ospedali Riuniti, Reggio Calabria, Italy Jolanta Malyszko 2nd Department of Nephrology, Medical University, Bialystok, Poland Veena Manjunath Section of Nephrology, Yale University School of Medicine, New Haven, CT, USA Celestina Manzoni Department of Nephrology, Dialysis and Renal Transplant, Alessandro Manzoni Hospital, Lecco, Italy Stephen D. Marks Department of Paediatric Nephrology, Great Ormond Street Hospital for Children NHS Foundation Trust, Great Ormond Street, London, UK David Marples Institute of Membrane and Systems Biology, Faculty of Biological Sciences, University of Leeds, Leeds, UK Mark R. Marshall Department of Renal Medicine, Counties Manukau District Health Board; and Faculty of Medicine and Health Science, University of Auckland, Auckland, New Zealand Kevin J. Martin Division of Nephrology, Saint Louis University, St. Louis, MO, USA

xxxix

xl

contributors Philip Mason Oxford Kidney Unit, The Churchill Hospital, Oxford, UK

Elaine Murphy Charles Dent Metabolic Unit, National Hospital for Neurology and Neurosurgery, London, UK

Anne Massez Department of Medical Imaging, University Clinics of Brussels—Erasme Hospital, Brussels, Belgium

Fliss E. M. Murtagh King’s College London, London, UK

Patrick H. Maxwell Cambridge Biomedical Campus, University of Cambridge, Cambridge, UK Peter A. McCullough Baylor University Medical Center, Baylor Heart and Vascular Institute, Baylor Jack and Jane Hamilton Heart and Vascular Hospital, Dallas, TX; and The Heart Hospital, Plano, TX, USA Emily P. McQuarrie Renal Unit, Western Infirmary, Glasgow, UK Nicholas Medjeral-Thomas Imperial College, London, UK Rajnish Mehrotra Division of Nephrology, Department of Medicine, Kidney Research Institute, University of Washington, Seattle, WA, USA Ravindra L. Mehta Division of Nephrology and Hypertension, Department of Medicine, University of California San Diego, San Diego, CA, USA Giampaolo Merlini Amyloid Research and Treatment Center, Foundation Scientific Institute Policlinico San Matteo, University of Pavia, Pavia, Italy Alain Meyrier Department of Nephrology, University ParisDescartes, Hôpital Georges Pompidou and Broussais, Paris, France

Thangamani Muthukumar Division of Nephrology, Department of Medicine, Department of Transplantation Medicine, Weill Medical College of Cornell University, New York, USA Walter P. Mutter Department of Medicine, Division of Nephrology, Newton-Wellesley Hospital, Newton, MA Maarten Naesens Department of Nephrology and Renal Transplantation, University Hospitals Leuven, Leuven, Belgium Saraladevi Naicker Division of Nephrology, Department of Internal Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa Brian J. Nankivell Renal Unit, Westmead Hospital, Westmead, Australia John Neary Royal Infirmary, Little France, Edinburgh, UK Nathalie Neirynck Nephrology Division, University Hospital Ghent, Ghent, Belgium Catherine Nelson-Piercy St Thomas’ Hospital, London, UK Khai Ping Ng Department of Renal Medicine, Queen Elizabeth Hospital and University of Birmingham, Birmingham, UK Patrick Niaudet Hopital Necker-Enfants Malades, Paris, France

Eve Miller-Hodges MRC Human Genetics Unit, University of Edinburgh Western General Hospital, Edinburgh, UK

Peter W. Nickerson Canadian Blood Services, Winnipeg, MB, Canada

Christopher Mitchell Children’s Services, John Radcliffe Hospital, Oxford, UK (now retired)

Søren Nielsen Department of Biomedicine, The Water and Salt Research Centre, Aarhus University, Aarhus, Denmark

Orson Moe University of Texas Southwestern Medical Center, Dallas; and Charles and Jane Pak Center for Mineral Metabolism and Clinical Research, Dallas, TX, USA

Allen R. Nissenson DaVita Inc., Denver, CO; and David Geffen School of Medicine at UCLA

Gilbert W. Moeckel Department of Pathology, Yale University School of Medicine, New Haven, CT, USA Thomas Mone OneLegacy, Los Angeles, CA, USA John Moran DaVita Inc., Denver, CO, USA Peter J. Morris Centre for Evidence in Transplantation, Royal College of Surgeons of England, London, UK Marion Muche Division of Gastroenterology, Infectious Diseases and Rheumatology, Department of Medicine, Charite Medical University, Hindenburgdamm, Berlin, Germany Heini Murer Institute of Physiology, University of Zurich, Zurich, Switzerland

Marina Noris IRCCS-Istituto di Ricerche Farmacologiche Mario Negri, Bergamo, Italy Ali J. Olyaei Division of Nephrology and Hypertension, Oregon Health and Science University, Portland, OR, USA Albert C. M. Ong Department of Infection and Immunity, The University of Sheffield Medical School, Sheffield, UK Michael Oppert Emergency Department, Klinikum Ernst von Bergmann, Potsdam, Germany Stephan R. Orth Dialysis Centre, Bad Aibling, Germany Nadina Ortiz-Brüchle Department of Human Genetics, RWTH Aachen University, Aachen, Germany

 

Graham Paget Division of Nephrology, Department of Internal Medicine, Faculty of Health Sciences, University of the Witwatersrand, Johannesburg, South Africa Chirag R. Parikh Program for Applied Translational Research, Section of Nephrology, Department of Medicine, Yale School of Medicine, New Haven, CT, USA Kamal V. Patel Guy’s and St Thomas’ Hospitals London, UK Rajan K. Patel Institute of Cardiovascular and Medical Sciences, University of Glasgow, Glasgow, UK York Pei Division of Nephrology, University Health Network and University of Toronto, Toronto, ON, Canada

contributors

Robert Provenzano Division of Nephrology, Hypertensive and Transplant Medicine, St John Hospital and Medical Center, Detroit, MI; and Wayne State University, School of Medicine, Detroit, MI, USA John Prowle Adult Critical Care Unit, Royal London Hospital, Barts and the London NHS Trust, London, UK James M. Pullman Department of Pathology, Albert Einstein College of Medicine; and Montefiore Medical Center, Bronx, NY, USA May M. Rabadi Department of Anesthesiology, College of Physicians and Surgeons, Columbia University, New York, USA

Mark A. Perazella Section of Nephrology, Department of Medicine, Yale University, New Haven, CT, USA

Milan RadoviĆ School of Medicine, University of Belgrade, Belgrade; and Department of Nephrology, Cllinical Center of Serbia, Belgrade, Serbia

Lisa M. Phipps Department of Renal Medicine, Orange Base Hospital, Orange, Australia

Paolo Raggi Mazankowski Alberta Heart Institute, University of Alberta, Edmonton, AB, Canada

Rob Pickard Newcastle University, Newcastle upon Tyne; and Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK

Shamima Rahman Mitochondrial Research Group, Genetics and Genomic Medicine Programme, UCL Institute of Child Health, London, UK

Matthew C. Pickering Imperial College, London, UK

Ravindra Rajakariar Barts Health NHS Trust, London, UK

Michael L. Picton Department of Renal Medicine, Manchester Royal Infirmary and Manchester Academic Health Science Centre, Manchester, UK

Premil Rajakrishna Renal Medicine, Royal Infirmary, Edinburgh, UK

Arkadiy Pinkhasov Baystate Medical Center, Tufts University School of Medicine, Springfield, MA, USA Richard K. S. Phoon Westmead Clinical School, Westmead, New South Wales, Australia Yves Pirson Department of Nephrology, Cliniques Universitaires St Luc, Brussels, Belgium Henry C. C. Pleass University of Sydney, Sydney; and Westmead Hospital, Westmead, Australia Oren Pleniceanu The Pediatric Stem Cell Research Institute and Division of Pediatric Nephrology, Edmond and Lili Safra Children’s Hospital, Chaim Sheba Medical Center, Tel Hashomer, Sackler Faculty of Medicine, Tel Aviv University, Israel Anneleen Pletinck Nephrology Division, University Hospital Ghent, Ghent, Belgium Rutger J. Ploeg Nuffield Department of Surgical Sciences, University of Oxford, Oxford, UK Carol A. Pollock Department of Renal Medicine, Kolling Institute of Medical Research, Royal North Shore Hospital, University of Sydney, St Leonards, Australia Giuseppe Pontoriero Department of Nephrology, Dialysis and Renal Transplant, Alessandro Manzoni Hospital, Lecco, Italy

S. Vincent Rajkumar Division of Hematology, Mayo Clinic, Rochester, MN, USA Raja Ramachandran Department of Nephrology, Postgraduate Institute of Medical Education and Research, Chandigarh, India Wolfgang Rascher University Hospital Erlangen, Erlangen, Germany Brian B. Ratliff Department of Medicine, New York Medical College, Valhalla, NY, USA Gauthier Raynal Université Pierre et Marie Curie, Hôpital Tenon, Service d’urologie, Paris, France Lesley Rees Department of Paediatric Nephrology, Great Ormond Street Hospital for Children, London, UK Heather N. Reich University Health Network and University of Toronto, Toronto, ON, Canada Giuseppe Remuzzi Mario Negri Institute for Pharmacological Research, Bergamo, Italy Anna Richards Queen’s Medical Research Institute, University of Edinburgh, Edinburgh, UK James Ritchie Salford Royal NHS Foundation Trust, Salford, UK Ian S. D. Roberts Department of Cellular Pathology, John Radcliffe Hospital, Oxford, UK

xli

xlii

contributors Bernardo Rodriguez-Iturbe Universidad del Zulia, Hospital Universitario de Maracaibo, Maracaibo; and Laboratorio de Inmunobiología, Instituto Venezolano Investigaciones Científicas (IVIC)-Zulia, Venezuala Claudio Ronco Department of Nephrology Dialysis & Transplantation and International Renal Research Institute (IRRIV), San Bortolo Hospital, Vicenza, Italy

Franz Schaefer Centre for Pediatrics and Adolescent Medicine, Heidelberg, Germany Eva Schepers Nephrology Division, University Hospital Ghent, Ghent, Belgium Miet Schetz Department of Intensive Care Medicine, University Hospital KU Leuven, Leuven, Belgium

Pierre M. Ronco INSERM UMR_S 702, Tenon Hospital, Paris, France

Laurent Schild Department of Pharmacology & Toxicology, Lausanne University, Lausanne, Switzerland

Mitchell H. Rosner Division of Nephrology, University of Virginia Health System, Charlottesville, VA, USA

Adalbert Schiller Department of Nephrology, University of Medicine and Pharmacy ‘Victor Babeş’, Nephrology Clinic, Emergency County Hospital, Timişoara, Romania

Hansjörg Rothe Division of Nephrology, Klinikum Coburg GmbH, Coburg, Germany Piero Ruggenenti IRCCS—Istituto di Ricerche Farmacologiche Mario Negri, Bergamo; and Unit of Nephrology, Azienda Ospedaliera Papa Giovanni XXIII, Bergamo, Italy

Karl P. Schlingmann University Children’s Hospital, Pediatric Nephrology, Münster, Germany Michiel F. Schreuder Radboud University Medical Center, Nijmegen, The Netherlands

Luis M. Ruilope Nephrology Department, Hospital Doce de Octubre, Madrid, Spain

Melvin M. Schwartz Department of Pathology, Rush University Medical Center, Chicago, IL, USA

Andrew D. Rule Mayo Clinic and Medical School, Rochester, MN, USA

Vedat Schwenger Department of Nephrology, University of Heidelberg, Heidelberg, Germany

David N. Rush Health Sciences Centre, Winnipeg, MB, Canada

Victor F. Seabra Department of Medicine, Tufts University School of Medicine, St. Elizabeth’s Medical Center, Boston, MA, USA

Khashayar Sakhaee Division of Mineral Metabolism, University of Texas, Dallas, TX, USA Vinay Sakhuja Post Graduate Institute of Medical Education and Research, Chandigarh, India Alan D. Salama UCL Centre for Nephrology, Royal Free Hospital, London, UK

Liviu Segall Department of Nephrology, University of Medicine and Pharmacy ‘Gr. T. Popa’, Iaşi, Romania Jang Won Seo Division of Nephrology, Department of Internal Medicine, Hangang Sacred Heart Hospital, College of Medicine, Hallym University, Korea

Moin A. Saleem University of Bristol, Children’s Renal Unit, Bristol Royal Hospital for Children, Bristol, UK

Sanjeev Sethi Division of Anatomic Pathology, Department of Laboratory Medicine and Pathology, Mayo Clinic, Rochester, MN, USA

Richard Sandford Academic Laboratory of Medical Genetics, Addenbrooke’s Treatment Centre, Addenbrooke’s Hospital, Cambridge, UK

Mehmet Şükrü Sever Department of Internal Medicine/ Nephrology, Istanbul School of Medicine, Çapa/ Istanbul, Turkey

Menaka Sarav University of Chicago Pritzker School of Medicine, Chicago, IL; and Division of Nephrology and Hypertension, NorthShore University HealthSystem, Evanston, IL, USA

Gaurav Shah Midwest Heart Specialists/Advocate Medical Group, Naperville, IL, USA

Minnie M. Sarwal California Pacific Medical Center, San Francisco, CA, USA Anjali Bhatt Saxena School of Medicine, Stanford University, Stanford, CA; and Santa Clara Valley Medical Center, San Jose, CA, USA John A. Sayer Institute of Genetic Medicine, Newcastle University, Newcastle, UK

Ron Shapiro Thomas E. Starzl Transplantation Institute, University of Pittsburgh, Pittsburgh, PA, USA Neil Sheerin Institute of Cellular Medicine, Newcastle University, Newcastle upon Tyne, UK Anushree Shirali Section of Nephrology, Department of Medicine, Yale University, New Haven, CT, USA Seema Shrivastava MRC Centre for Transplantation, King’s College London, London; and St George’s Hospital, London, UK

 

Roslyn J. Simms Institute of Genetic Medicine, Newcastle University, Newcastle, UK

contributors

Bridget Sinnott University of Texas, Dallas, TX, USA

Eugene Teoh Department of Oncology, Oxford Cancer Imaging Centre, University of Oxford, Oxford, UK; and Radiology Department, Oxford University Hospitals NHS Trust, Churchill Hospital, Oxford, UK

Marijn Speeckaert Department of Nephrology, Ghent University Hospital, Gent, Belgium

Bernadette A. Thomas University of Washington, Seattle, WA, USA

Thimoteus Speer Department of Internal Medicine IV, Saarland University Hospital, Homburg/Saar, Germany

Nicola Thomas Public Health, Primary Care and Food Policy Department, School of Community and Health Sciences, City University, London, UK

Stuart M. Sprague University of Chicago Pritzker School of Medicine, Chicago, IL; and Division of Nephrology and Hypertension, NorthShore University HealthSystem, Evanston, IL, USA

Raj Thuraisingham Royal London Hospital, London, UK Marcello Tonelli University of Calgary, Calgary, Canada

Joanna Stachowska-Pietka Institute of Biocybernetics and Biomedical Engineering, Polish Academy of Sciences, Warsaw, Poland

Jonathan N. Townend Department of Cardiology, Queen Elizabeth Hospital and University of Birmingham, Birmingham, UK

Coen A. Stegeman Department of Nephrology, University of Groningen, University Medical Center Groningen, Groningen, The Netherlands

Olivier Traxer Université Pierre et Marie Curie, Hôpital Tenon, Service d’urologie, Paris, France

Peter Stenvinkel Division of Renal Medicine, Department of Clinical Science, Intervention and Technology, Karolinska Institutet, Stockholm, Sweden

Francesco Trepiccione Department of Cardio-thoracic and Respiratory Science, Second University of Naples, Naples, Italy

Douglas Stewart Freeman Heart and Vascular Institute, Joplin, MO, USA

Neil Turner Queen’s Medical Research Institute, University of Edinburgh, Edinburgh, UK; and Department of Renal Medicine, Edinburgh Royal Infirmary, UK

Arohan R. Subramanya Renal-Electrolyte Division, University of Pittsburgh School of Medicine, Pittsburgh, PA, USA

Andrew A. Udy Department of Intensive Care Medicine, Royal Brisbane and Womens Hospital, Herston, Australia

Rita Suri Division of Nephrology, University of Montreal, London, ON, Canada

Mark L. Unruh Renal-Electrolyte Division, University of Pittsburgh Medical Center, Pittsburgh, PA, USA

Manikkam Suthanthiran Division of Nephrology, Department of Medicine, Department of Transplantation Medicine, Weill Medical College of Cornell University, New York, USA

Ashish Upadhyay Kidney and Blood Pressure Center, Tufts University School of Medicine, Boston, MA, USA

Teena Tandon Division of Nephrology, Department of Medicine, Indiana University School of Medicine, Indiana University, Indianapolis, IN, USA Sydney C. W. Tang Department of Medicine, University of Hong Kong, Queen Mary Hospital, Hong Kong Eric N. Taylor Channing Laboratory, Department of Medicine, Brigham and Women’s Hospital and Harvard Medical School, Boston, MA; and Division of Nephrology and Transplantation, Maine Medical Center, Portland, ME, USA Karthik K. Tennankore Division of Nephrology, University Health Network, Toronto; and University of Toronto, Toronto, ON, Canada Teresa Tenorio Consorcio de investigación del Fracaso Renal Agudo de la Comunidad de Madrid (CIFRA), Madrid; and Instituto Ramón y Cajal de Investigación Sanitaria, Madrid; and Department of Nephrology, Hospital Universitario Ramón y Cajal, Madrid, Spain

Tushar J. Vachharajani Wake Forest University School of Medicine, Section of Nephrology, Winston-Salem, NC, USA Wim Van Biesen Renal Division, University Hospital Ghent, Ghent, Belgium Johan van der Vlag Department of Nephrology, Radboud Institute for Molecular Life Sciences, Radboud University Medical Center, Nijmegen, The Netherlands Raymond Vanholder Renal Division, University Hospital Ghent, Ghent, Belgium Radovan Vasko Department of Nephrology and Rheumatology, Goettingen University Medical School, Goettingen , Germany Nosratola D. Vaziri Division of Nephrology & Hypertension, Departments of Medicine, Physiology & Biophysics, Schools of Medicine and Biological Science, University of California, Irvine, CA, USA

xliii

xliv

contributors Anja Verhulst Laboratory of Pathophysiology, University of Antwerp, Antwerp, Belgium Benjamin A. Vervaet Laboratory of Pathophysiology, University of Antwerp, Antwerp, Belgium

I. David Weiner University of Florida College of Medicine, Gainesville, FL; and Nephrology and Hypertension Section,North Florida/South Georgia Veterans Health System, Gainesville, FL, USA

Udo Vester University of Essen, Essen, Germany

Toby Wells Abertawe Bro Morgannwg University Health Board, Radiology Department, Singleton Hospital, Swansea, UK

Katie Vinen King’s College Hospital NHS Foundation Trust, London, UK

Ulrich Wenzel Division of Nephrology, Department of Medicine, University Hospital Hamburg-Eppendorf, Hamburg, Germany

Sobhan Vinjamuri Royal Liverpool University Hospital, Liverpool, UK

Michael J. Weston Department of Radiology, St James’s University Hospital, Leeds, UK

Philip Vladica Departments of Nuclear Medicine and Radiology, Westmead Hospital, Westmead, Australia

Jack F. M. Wetzels Department of Nephrology, Radboud University Medical Center, Nijmegen, The Netherlands

Alexandra Voinescu Division of Nephrology, Saint Louis University, St. Louis, MO, USA

Caroline Whitworth Renal Medicine, Royal Infirmary of Edinburgh, Edinburgh, UK

Luminita Voroneanu Clinic of Nephrology, ‘C. I. Parhon’ University Hospital, ‘Gr. T. Popa’ University of Medicine and Pharmacy, Iasi, Romania

Thorsten Wiech Department of Pathology, University of Hamburg, Hamburg, Germany

Carsten A. Wagner Institute of Physiology, University of Zurich, Zurich, Switzerland

Kate Wiles Guy’s and St Thomas NHS Foundation Trust, and King’s College, London, UK

Stephen Waldek Sale, Greater Manchester, UK

Andrew D. Williams School of Health Sciences, University of Tasmania, Launceston, Australia

Stephen B. Walsh UCL Centre for Nephrology, Royal Free Hospital, London, UK

Christopher G. Winearls Oxford Kidney Unit, Oxford University Hospitals, Oxford, UK

Gerd Walz Renal Division, University Freiburg Medical Center, Freiburg; and BIOSS Center for Biological Signalling Studies, University of Freiburg, Freiburg, Germany

Charles S. Wingo Division of Nephrology, Hypertension, and Transplantation, Department of Medicine, University of Florida College of Medicine, Gainesville; and Nephrology and Hypertension Section, North Florida/South Georgia Veterans Health System, Gainesville, FL, USA

Haiyan Wang† Formerly, Renal Division, Department of Medicine, Peking University First Hospital; and Peking University Institute of Nephrology; and Key Laboratory of Renal Disease, Ministry of Health of China; and Key Laboratory of Chronic Kidney Disease Prevention and Treatment (Peking University), Ministry of Education, Beijing, China Jacek Waniewski Institute of Biocybernetics and Biomedical Engineering, Polish Academy of Sciences, Warsaw, Poland David J. Webb University/British Heart Foundation Centre for Cardiovascular Science, Queen’s Medical Research Institute, University of Edinburgh, Edinburgh, UK Catherine Weber Division of Nephrology, Department of Medicine, McGill University, Montreal, Quebec, Canada Stefanie Weber Pediatric Nephrology, University-Children’s Hospital Essen, Essen, Germany Angela C. Webster Sydney School of Public Health, University of Sydney, Sydney; and Centre for Transplant and Renal Research, Westmead Hospital, Sydney; and Centre for Kidney Research, Kids Research Institute, The Children’s Hospital at Westmead, Sydney, Australia

Paul Winyard Nephro-Urology Unit, UCL Institute of Child Health, London, UK Sarah Withers Cardiovascular Research Group, School of Biomedicine, University of Manchester, Manchester, UK Germaine Wong Sydney School of Public Health, University of Sydney, Sydney; and Centre for Transplant and Renal Research, Westmead Hospital, Sydney; and Centre for Kidney Research, Kids Research Institute, The Children’s Hospital at Westmead, Sydney, Australia Muh Geot Wong Department of Renal Medicine, Kolling Institute of Medical Research, Royal North Shore Hospital, University of Sydney, St Leonards, Australia Terry Wong Guy’s & St Thomas’ Hospital, London, UK Alexander Woywodt Lancashire Teaching Hospitals NHS Foundation Trust, Royal Preston Hospital, Preston, UK

 

Hong Xu Divisions of Baxter Novum and Renal Medicine, Department of Clinical Sciences, Intervention and Technology, Karolinska Institutet, Stockholm, Sweden Muhammad M. Yaqoob Queen Mary University of London, London; and Barts and the London NHS Trust, London, UK Soo Young Yoon Division of Nephrology, Department of Internal Medicine, Kwandong University College of Medicine, Goyang, Korea Nadia Zalunardo Division of Nephrology, University of British Columbia, Vancouver, BC, Canada Klaus Zerres Department of Human Genetics, RWTH Aachen University, Aachen, Germany

†We

contributors

Luxia Zhang Renal Division, Department of Medicine, Peking University First Hospital; and Peking University Institute of Nephrology; and Key Laboratory of Renal Disease, Ministry of Health, Beijing, China Ming-hui Zhao Renal Division, Peking University First Hospital, Institute of Nephrology, Peking University, Key Laboratory of Renal Disease, Ministry of Health of China, Beijing, P. R. China Robert Zietse Division of Nephrology & Transplantation, Department of Internal Medicine, Erasmus Medical Center, Rotterdam, The Netherlands Carmine Zoccali CNR-IBIM Clinical Epidemiology and Pathophysiology of Renal Diseases and Hypertension, Ospedali Riuniti, Reggio Calabria, Italy

record with sadness the death of Professor Hai-Yan Wang on 11 December 2014.

xlv

SECTION 1

Assessment of the patient with renal disease

1 Epidemiology of kidney disease  3 Aminu K. Bello, Marcello Tonelli, and Kitty J. Jager

2 Clinical assessment of the patient with renal disease: overview  20 Christopher G. Winearls

3 Presentations of renal disease  22 Christopher G. Winearls

4 Kidney disease-focused history taking  27 Christopher G. Winearls

5 Kidney disease-focused features on examination  29 Christopher G. Winearls

6 Urinalysis 35 Walter P. Mutter

7 Assessment of renal function  44 Marijn Speeckaert and Joris Delange

8 Tubular function  62 Marijn Speeckaert and Joris Delange

9 Renal radiology: overview  66 Michael J. Weston

10 Ionizing radiation and radiation protection  68 Jeannette Kathrin Kraft and Peter Howells

11 Plain radiography, excretion radiography, and contrast radiography  74 Akira Kawashima and Andrew J. LeRoy

12 Intervention 82 Steven Kennish

13 Ultrasound 89 Toby Wells and Simon J. Freeman

14 Computed tomography  101 Eugene Teoh and Michael J. Weston

15 Magnetic resonance imaging  109 Kazuhiro Kitajima, Akira Kawashima, and James F. Glockner

16 Radioisotopes in diagnostic imaging in nephrology  117 Ramya Dhandapani and Sobhan Vinjamuri

17 Immunological investigation of the patient with renal disease  127 Jo H. M. Berden and Jack F. M. Wetzels

18 The renal biopsy  142 Ian S. D. Roberts, Philip Mason, and Agnes B. Fogo

19 Clinical trials: why and how in nephrology  161 Richard Haynes, Martin J. Landray, William G. Herrington, and Colin Baigent

CHAPTER 1

Epidemiology of kidney disease Aminu K. Bello, Marcello Tonelli, and Kitty J. Jager Basic epidemiology principles in nephrology Introduction Epidemiology is the study of the distribution, determinants, and frequency of disease in populations or settings (Rothman, 1981, 2002). Therefore, epidemiological studies assess the extent of disease, risk/causal factors, natural history, prognosis, prevention/ treatment strategies, and the potential for new policies to prevent disease or improve outcomes (Rothman, 2002). Epidemiological research helps to inform evidence-based medicine by identifying risk factors for disease and to determine optimal treatment approaches; it is the cornerstone of public health research and of preventive medicine. The identification of unbiased causal relationships between exposures (risk factors or interventions) such as hypertension or the use of antihypertensive medication and outcomes like morbidity and mortality is therefore an important aspect of epidemiology. This section will discuss some epidemiological concepts, methods, and their application to clinical research in nephrology.

Research questions Defining an appropriate research question requires familiarity with knowledge gaps in the subject area to judge if a question is feasible, interesting, novel, ethical, and relevant (FINER criteria) (Hulley et al., 2007). Whereas the FINER criteria highlight general aspects of research questions, the development of a specific research question may follow the PICO format, in which P stands for the population (or problem) of interest, I for the intervention (or any other exposure), C for the comparison group, and O for the outcome of interest. Some suggest a PICOT approach, adding a T for the follow-up time to assess outcome (Haynes et al., 2006). An example of a question framed according to PICOT is ‘In dialysis patients (P), what is the effect of statins (I) compared to placebo (C) on cardiovascular mortality (O) after 4 years of follow-up (T)?’ A research question has implications for the choice of the study design—which will in turn determine the analytical methods.

Study designs Fig. 1.1 shows an algorithm for the classification of study designs in clinical research (Grimes and Schulz, 2002). Study design is an important aspect of study quality. Studies can be classified into experimental and observational ones depending on whether or not exposures like therapy were assigned by the investigators. Random allocation of exposures is important to prevent selection bias by the clinician (also known as ‘confounding by indication’) (Stel et  al., 2007)  occurring, when clinicians provide a specific

therapy because of preconceived ideas about which therapy is best. Therefore, when it comes to studies on the effects of therapy or other interventions, randomized controlled trials (RCTs) are the gold standard. This was exemplified by RCTs showing a lack of effect or even harmful effects of using high target haemoglobin level in chronic kidney disease (CKD) patients as opposed to using a lower target (Drueke et al., 2006; Singh et al., 2006), whereas the majority of observational studies had suggested that higher haemoglobin levels were associated with favourable outcomes. On the other hand, observational studies may answer questions on aetiology, diagnosis, prognosis, and adverse effects. In addition, they may provide answers on the effects of therapy where RCTs are not possible, inappropriate, inadequate, or unnecessary (Black, 1996). The effect of transplantation as compared to dialysis cannot be determined through an RCT, as allocation of renal grafts depends on other factors like HLA-matching. Where there is no comparison (control) group (as in case reports or case series), observational studies are called descriptive and where there is a comparison group they are referred to as analytical. Finally, the temporal direction of analytical observational studies determines the type of study. Cohort studies like the Dialysis Outcomes and Practice Patterns Study (DOPPS) determine the exposure of subjects to risk factors at the start of inclusion and then look forward in time to observe the occurrence of outcomes. They may provide a wealth of data which enable the investigator to study not only multiple outcomes but—unlike RCTs—also multiple exposures. In contrast, case–control studies compare cases (those with the disease or other outcome of interest) with controls (those without the outcome of interest) and then look back in time for exposures that might have caused the outcome. In nephrology, case–control studies are uncommon. Nevertheless, this study type is very efficient for studying potential risk factors for rare outcomes that may take a long time to develop, for example, CKD. By going back in time and looking for particular exposures like analgesics one may find associations between outcomes and these exposures. In such a case, prospective cohort studies are less efficient as one will need a very high number of subjects and a very long time to acquire an equal number of cases. Finally, cross-sectional studies examine the presence of an exposure and that of the outcome at the same moment in time. In most cases this simultaneity makes it difficult to determine which is the cause and which is the consequence, in other words, this design may induce a chicken-and-egg problem. Table 1.1 describes the strengths and weaknesses of different study designs (Jager et al., 2007).

4

Section 1  

assessment of the patient with renal disease Table 1.1  Strengths and weaknesses of frequently used study designs

Did investigator assign exposures?

Study design No

Yes Experimental study

Observational study

Random allocation?

Comparison group?

Yes

No

Randomized controlled trial

Exposure

Nonrandomized controlled trial

Analytical study

◆

First form of publication for new diseases, rare adverse events, or manifestations of disease ◆ Fast and inexpensive ◆ Hypothesis generating

◆

Cross-sectional study

◆

Can assess prevalence and burden of disease ◆ Fast and inexpensive ◆ Hypothesis generating

◆

Case–control study

◆

Efficient study design Very suitable for studying rare outcomes and outcomes that take a long time to develop ◆ Can study multiple exposures ◆ Relatively inexpensive ◆ Hypothesis generating

◆

Cohort study

◆

Can study multiple exposures, uncommon exposures and multiple outcomes ◆ Hypothesis generating

◆

Randomized controlled trial

◆

◆

No Descriptive study

Exposure and outcome at the same time Outcome Case–control study

Crosssectional study

Fig. 1.1  Algorithm for the classification of study designs in clinical research. Reprinted from The Lancet, 359(9300), Grimes DA, Schulz KF, An overview of clinical research: the lay of the land, 57–61, 2002, with permission from Elsevier.

Knowledge derived from different studies published on a specific topic can be summarized in a systematic review (Noordzij et  al., 2009). In contrast to narrative reviews, systematic reviews use explicit and reproducible methods for searching the literature and a critical appraisal of individual studies. This systematic methodology minimizes bias. Sometimes systematic reviews include a meta-analysis. This is a mathematical synthesis of the results of those individual studies in which more weight is given to results of studies with more events and sometimes to studies of higher quality. Criteria for the reporting of studies using particular designs have been summarized in statements like CONSORT (Schulz et al., 2010) for RCTs, STROBE (von Elm et al., 2007) for observational studies, and PRISMA (Moher et  al., 2009)  and MOOSE (Stroup et  al., 2000)  for their respective meta-analyses. Such statements assist authors in writing such reports, help editors and peer reviewers in reviewing manuscripts for publication, and aid readers in critically appraising the quality of published studies.

Weaknesses

Case report/ case series

Direction?

Outcome

Exposure Cohort study

Yes

Strengths

◆

Important potential to make causal inferences

Very limited potential to make causal inferences, unless in dramatic cases ◆ Selection bias

Very limited potential to make causal inferences, because the time order of exposure and outcome cannot be determined ◆ Selection bias ◆ Survival bias Some potential to make causal inferences ◆ Can study only one outcome at the time ◆ Choice of controls needs careful attention ◆ Selection bias ◆ Recall bias Some potential to make causal inferences ◆ If done prospectively, more expensive ◆ If done prospectively, may take a long time to complete ◆ Selection bias Can study multiple outcomes, but only one exposure ◆ Very expensive ◆ Limited generalizability when making use of restrictive in- and exclusion criteria ◆ Selection bias

Reprinted by permission from Macmillan Publishers Ltd: Kidney Int. Jager KJ, Stel VS, Wanner C, Zoccali C, Dekker FW. The valuable contribution of observational studies to nephrology, 72(6):671–5, 2007.

Measures of disease occurrence Different measures may be used to describe how often a disease (or other health outcomes) occurs in a population (Jager et al., 2007). Incidence expresses the development of new cases and is mostly used against the background of prevention, to assess disease aetiology or to determine risk factors. Depending on the study question, incidence may be reported as risk or as incidence rate. The latter is preferred when studying a dynamic population or when the observation period is sufficiently long for competing risks (like death from causes other than the outcome under investigation) or loss to follow-up to play a significant role. Determining the incidence rate of

a treatment like renal replacement therapy (RRT) is straightforward. When determining the incidence of a chronic condition like CKD, problems may arise, as it is unfeasible to identify newly developed CKD in all individuals who were initially free from this disease. Prevalence on the other hand is the number of existing cases. It reflects the burden of the disease in a population and can be used for the planning of healthcare facilities. Again, the assessment of the prevalence of a treatment like RRT is relatively uncomplicated. However, the assessment of the prevalence of stages 3–5 of CKD according to the National Kidney Foundation Kidney Disease

chapter 1 

Outcomes Quality Initiative (NKF-KDOQI) 2002 and Kidney Disease:  Improving Global Outcomes (KDIGO) 2004 definition only needs an estimated glomerular filtration rate (eGFR), whereas that of stages 1–2 CKD also requires albuminuria testing. Finally, measures of disease occurrence are used to study causes of disease or effects of risk factors. One way to estimate the size of such an effect is the calculation of the ratio between the disease occurrence (e.g. cardiovascular disease) in those being exposed to the risk factor (e.g. diabetes mellitus) and those not exposed to the risk factor. The resulting ‘relative risk’ is an example of this measure of effect.

epidemiology of kidney disease

Table 1.2  Types of regression analysis Outcome variable (dependent variable)

Exposure variable (independent variable)

Measure of effect

Linear

Continuous

Continuous or categorical

Relative risk

Logistic

Categorical

Continuous or categorical

Odds ratio

Survival analysis

Time-to-event

Continuous or categorical

Hazard ratio

Bias and confounding Bias is a systematic error in the design or conduct of a study (Tripepi et al., 2008). It may affect study validity in several ways, for example, through methods used by the investigator in recruiting study subjects or through factors affecting study participation (selection bias) or through systematic distortions in the collection of exposures or outcomes (information bias). Selection bias may, for example, occur when investigators performing an RCT use very strict inclusion criteria so that only relatively healthy subjects will be included in the study. The study results in this ‘selected’ group may not necessarily be generalized to very sick patients. Information bias on the other hand may, for instance, be induced when inaccurate instruments systematically over- or underestimate exposures like blood pressure or when in case–control study diseased individuals (cases) much better remember potentially harmful exposures than non-diseased individuals (controls). The latter is called recall bias. Confounding is a ‘blurring’ of effects, obscuring the ‘real’ effect of an exposure on outcome (Jager et al., 2008). For example, the inverse association between total cholesterol level and mortality in dialysis patients was likely the result of confounding by systemic inflammation and malnutrition (Liu et  al., 2004). For this reason, in aetiological studies, investigators do their best to prevent confounding, for example, by randomization, or to control confounding. A  frequently used method to control for confounding is including the confounding variable into multivariate analysis. However, before a variable qualifies as a confounder it should fulfil three criteria: (1) be a risk factor for the outcome studied, (2) be associated with the exposure, and (3) not be an effect of the exposure (be in the causal pathway). The last criterion is important to avoid ‘over-adjustment’ through which an investigator may take away part of the real effect of an exposure and may introduce bias instead of preventing it. For example, in a study on the association between body mass index (BMI) and the risk of ESRD (Hsu et al., 2006) the authors were right in not adjusting for blood pressure, as blood pressure can be considered in the causal pathway between BMI and end-stage renal disease (ESRD). Therefore, before including them in a multivariate analysis, the criteria for confounding should be checked carefully for every variable in every association studied.

Regression analysis This statistical technique describes the dependence of the outcome (‘dependent’) variable from the value of one or more exposure variables (‘independent’ variables). Regression can be used for univariate analysis (without controlling for confounding) and multivariate analysis (with adjustment for confounding). Table 1.2

shows that the choice for the type of regression analysis depends on the type of outcome studied, but not on the type of exposure variables that may be included in the analysis. Linear regression analysis demands that the outcome variable is continuous, for example, the 3-year risk of cardiovascular events. Logistic regression, on the other hand, is needed to study categorical outcomes such as whether or not patients are compliant with therapy prescription. Finally, survival analysis is used to study ‘time-to-event’ data like time to death, to first renal transplant, or to myocardial infarction. In survival analysis, the survival times of subjects not experiencing the event of interest are being censored. The Kaplan–Meier method, being the most popular method for survival analysis, provides estimates of survival probabilities and may compare survival between groups. The method can, however, only study the effect of one exposure at the time and it cannot provide an effect size. Multivariate survival analysis and effect estimation therefore need other techniques like Cox proportional hazards regression. The measure of effect of Cox regression is the hazard ratio. Both the odds ratio and the hazard ratio can be interpreted as relative risks.

Screening Investigating apparently healthy individuals to detect unrecognized early stages of disease allows measures to be taken to prevent or slow down progression and reduce (premature) death in those affected (Grootendorst et al., 2009). In 1968, the World Health Organization (WHO) published ten criteria to facilitate the selection of conditions that are suitable for screening (Wilson and Jungner, 1968) which have been supplemented with additional criteria thereafter (Andermann et al., 2008). These criteria should be taken into account when considering screening for conditions like CKD. Evaluation of screening programmes may suffer from bias. For example, those who volunteer to be tested for CKD may have a family history of kidney disease resulting in a high detection rate. In addition to this form of selection bias, those who are diagnosed by screening more frequently include those with slowly progressive disease, as those with poor prognosis are less likely to be picked up by screening because of their higher mortality risk. This length bias will provide a ‘better’ prognosis to those identified in a screening programme, even if screening has no effect on prognosis. Similarly, earlier diagnosis through screening may lead to an apparent increase in survival time. This phenomenon is known as lead-time bias (Grootendorst et al., 2009).

5

6

Section 1  

assessment of the patient with renal disease

Epidemiology of acute kidney injury (See also Chapter 220.)

Definition, classification, and evaluation of acute kidney injury Acute kidney injury (AKI) is a clinical syndrome characterized by a sudden onset of reduced kidney function, and is manifested by increased serum creatinine (SCr) or a reduction in urine output (Hou and Cohen, 1985; Bellomo et  al., 2004; Lameire et  al., 2005; Cerda, 2008). Until recently, there was no standard definition for AKI, leading to highly variable estimates of the incidence, prevalence, and prognosis of this syndrome (Lameire et al., 2005). Prior to 2004, the key definition in use was based on the WHO International Classification of Disease (ICD) codes, which are used mainly for administrative purposes in the healthcare systems of developed countries. Criteria based on various SCr values and/or the receipt of dialysis were also commonly used (especially by clinicians in hospital settings (Mehta and Chertow, 2003; Lameire et al., 2006))—although such definitions vary widely across settings, populations and institutions (Mehta and Chertow, 2003; Mehta et al., 2007). The absence of a standardized definition of AKI led to the development of consensus criteria by the Acute Dialysis Quality Initiative (ADQI), which are known as the Risk, Injury, Failure, Loss, and End-stage renal disease (RIFLE) 2004 criteria (Van Biesen et al., 2006; Mehta et al., 2007). The RIFLE scheme is based on two important parameters: changes in SCr or eGFR from baseline and urine output at specific points in time (Table 1.3) (Mehta et al., 2007). This classification was subsequently modified and the term ‘acute renal failure’ was replaced with ‘acute kidney injury’ to cover the entire spectrum of acute renal dysfunction from mild changes in renal function to the use of dialysis (Table 1.3) (Van Biesen et al., 2006). This standardization of terminology facilitates comparison between settings. However, the central role of quantitative changes in urine output in the scheme may make it most applicable to critical care units or other closely monitored settings. Although the RIFLE criteria were validated in over half a million AKI patients globally (Cerda et al., 2008b; Cruz et al., 2009), the system was subsequently revised by another international group (the Acute Kidney Injury Network (AKIN)). The AKIN scheme uses slightly broader criteria than the RIFLE system, and reflects the clinical significance of relatively small rises in SCr (Table 1.3) (Mehta et al., 2007; Cruz et al., 2009). Of note, it appears that RIFLE may be slightly better than AKIN for risk prediction and prognostication (Ricci et al., 2008). To reconcile these minor differences, the International Society of Nephrology (ISN) and KDIGO (Anonymous, 2012), has brought together international experts from multiple specialties to harmonize the RIFLE and AKIN criteria and produce a truly global definition and staging system (Anonymous, 2012)—further enabling future comparisons of the incidence, outcomes, and efficacy of therapeutic interventions for AKI (Table 1.3).

Limitations and pitfalls for AKI definition It is important to recognize that no single definition of a clinical syndrome is ideal; the purpose of establishing a consensus definition is to establish the presence/absence of disease, provide an index of severity, and relate these to prognosis and outcome

Table 1.3  Definitions/classification for AKI Classifications Criteria

Urine output criteria

KDIGO Stage 1

Increase ≥ 26 μmol/L within 48 h or increase ≥ 1.5 to 1.9 × reference SCr

< 0.5 mL/kg/h for > 6 consecutive h

Stage 2

Increase ≥ 2 to 2.9 × reference SCr

< 0.5 mL/kg/h for > 12 h

Stage 3

Increase ≥ 3 × reference SCr or increase 354 μmol/L or commenced on RRT irrespective of stage

< 0.3 mL/kg/h for > 24 h or anuria for 12 h

Risk

Increased creatinine × 1.5 or GFR decrease > 25%

< 0 .5 mL/kg/h × 6 h

Injury

Increased creatinine × 2 or GFR decrease > 50%

< 0 .5 mL/kg/h × 12 h

Failure

Increase creatinine × 3 or GFR decrease > 75% or creatinine 4 mg/dL (acute rise of 0.5 mg/dL)

< 0.3 mL/kg/h × 24 h (oliguria) or anuria × 12 h

Loss

Persistent ARF = complete loss of renal function > 4 weeks

ESRD

End-stage renal disease

RIFLE

AKIN Stage 1

Increase ≥ 26 μmol/L within 48 h or increase ≥ 1.5–1.9 × reference SCr

< 0.5 mL/kg/h for ≥ 6 h

Stage 2

Increase ≥ 2–2.9 × reference SCr

< 0.5 mL/kg/h for ≥ 12 h

Stage 3

Increase ≥3 × reference SCr or increase≥ 354 μmol/L with an acute rise of at least 44 μmol/L or initiation of RRT

< 0.3 mL/kg/h for ≥ 24 h or anuria ≥ 12 h

ICD 9 codes ARF

ARF without dialysis Any of the following: 584.5: ARF, with lesion of tubular necrosis 584.6: ARF, with lesion of renal cortical necrosis 584.7: ARF, with lesion of renal medullary (papillary) necrosis 584.8: ARF, with other specified pathologic lesion in kidney

ARF-D

ARF with dialysis: ARF code as above plus any of the following codes: 584.9: ARF, unspecified V39.95: haemodialysis V45.1: renal dialysis status (patient requires intermittent renal dialysis; presence of arteriovenous shunt) V56.0: extracorporeal dialysis (dialysis (renal) not otherwise specified) V56.1: fitting and adjustment of extracorporeal dialysis catheter

ARF = acute renal failure; ARF-D = acute renal failure with dialysis; GFR = glomerular filtration rate; KDIGO = Kidney Disease: Improving Global Outcomes; RRT = renal replacement therapy; SCr = serum creatinine.

chapter 1 

(Rothman, 2002). In addition to lack of standardization, definitions used before RIFLE/AKIN did not facilitate detection of milder forms of AKI—which are highly prevalent and also associated with adverse clinical outcomes (Yong et al., 2011). Despite the advantages of the RIFLE/AKIN criteria, these definitions also have some limitations (Van Biesen et al., 2006; Cerda et al., 2008b; Yong et al., 2011). First, the use of 6-hour and 12-hour urine output criteria make RIFLE difficult to use for retrospective studies of AKI, since such data are not routinely collected (Van Biesen et al., 2006; Srisawat et al., 2010). Second, there is not necessarily any correlation between the SCr and urine output criteria for severity of AKI, as patients often present with high SCr but good urine output or vice versa. Third, SCr/eGFR criteria are based on change from a baseline value (which is not always available). Fourth, eGFR is valid only in steady-state conditions—which is certainly not the case in AKI (Srisawat et al., 2010). Fifth, recent studies have shown that smaller changes in SCr than those specified by the RIFLE criteria (such as an absolute increase of 0.3 mg/dL) are also associated with poor outcomes (Srisawat et al., 2010). These limitations have been partially addressed by AKIN (Van Biesen et al., 2006; Srisawat et al., 2010)—which has eliminated the eGFR criterion and explicitly recognized the adverse prognostic values associated with very small changes in SCr (Van Biesen et al., 2006).

Risk factors and causes for AKI A partial list of predisposing factors (risk factors) and causes for AKI is shown in Fig. 1.2; the incidence and likelihood of which vary according to the clinical setting (hospital vs community; developed vs developing countries) (Joannidis and Metnitz, 2005; Lameire et al., 2005, 2006). For instance, acute tubular necrosis (ATN) due to sepsis is the most common cause of AKI in the critical care setting, accounting for up to 35–50% of all cases of AKI (Lameire et al., 2005). In a large-scale study, the Madrid Acute Renal Failure Study in Spain, ATN was the cause of AKI in 75.9% of intensive care unit (ICU) patients compared to 37.6% in non-ICU patients (Liano and Pascual, 1996). ATN in hospitalized patients is more likely to be multifactorial, with hypotension and nephrotoxins as important causes in addition to sepsis and surgery (Cerda et al., 2008b). For community-acquired AKI, prerenal or acute-on-chronic renal failure is common and usually occurs as the result of dehydration or drug-induced causes as seen with the use of non-steroidal anti-inflammatory drugs (NSAIDs), angiotensin-converting enzyme inhibitors (ACEI), and angiotensin receptor blockers (ARBs) (Cerda, 2008; Cerda et  al., 2008b). Elderly people and patients with multiple comorbidities such as diabetes are at particularly high risk of developing community-acquired AKI (Cerda, 2008)  (Fig. 1.2). AKI is being conceptualized as a spectrum as obtained in CKD from normal to increased risk, decreased function to kidney failure, and/or death (Fig. 1.3).

Descriptive epidemiology of AKI (See also Chapter 220.) There is a paucity of high-quality population-based data on the incidence of AKI—whether community- or hospital-acquired (Cerda et  al., 2008b; Yong et  al., 2011). Before the advent of the RIFLE/AKIN classification systems, the reported incidence of AKI varied from 1% to 30% across studies (Lameire et al., 2005; Waikar et al., 2008). As reviewed by Waikar et al. (2008), the incidence of AKI also shows considerable variation based on the clinical setting

epidemiology of kidney disease

(Lameire et al., 2005, 2006). The prevalence of hospital-acquired AKI is reported to be about five to ten times greater than community-acquired AKI, with reported rates of AKI of 5–7% of hospitalized patients (Hsu et  al., 2007; Waikar et  al., 2008; Yong et al., 2011). In country-specific data, the reported prevalence of AKI in the United States ranges from 1% (community acquired) to 7.1% (hospital acquired) of all hospital admissions (Lameire et al., 2005, 2006). The population incidence of AKI from UK data ranges from 172 pmp to 486–630 pmp (Yong et al., 2011). Early studies from the 1990s reported that the annual incidence rates of community-acquired AKI varied from 22 to 620 pmp, with most studies using the receipt of dialysis or SCr ≥ 300 or 500μmol/L to define AKI (Lameire et al., 2005; Yong et al., 2011). The reported incidence of AKI varies considerably and depends on how AKI was defined, the setting where it occurred, population studied, and also whether patients with de novo CKD developing AKI were included. For instance, the reported incidence varies from 140 to 620 pmp across countries (Feest et  al., 1993; Liano and Pascual, 1996; Stevens et al., 2001; Waikar et al., 2006). In the United Kingdom, the reported incidence was 620 pmp/year among patients with SCr ≥ 300 µmol/L (Stevens et al., 2001) and 140 pmp/year in those with SCr > 500 µmol/L (Feest et al., 1993). In Spain, the incidence of AKI was 209 pmp from data of patients in tertiary care hospitals and SCr > 177 µmol/L (Liano and Pascual, 1996). In the United States, the reported incidence was 288 pmp, but ICD-9 codes rather than SCr criteria were used to define AKI (Waikar et al., 2006). The incidence of community-acquired AKI appears to be increasing:  the incidence of non-dialysis-dependent AKI and dialysis-dependent AKI increased from 322.7 to 522.4 and 19.5 to 29.5 per 100,000 person-years, respectively, between 1996 and 2003 (Hsu et al., 2007). The rising incidence of community-based AKI over time likely reflects the ageing general population, the increasing prevalence of chronic comorbidity (including CKD), and increasing utilization of nephrotoxic agents (such as intravenous radiocontrast, aminoglycosides, NSAIDs, ACEIs/ARBs, and chemotherapeutic agents) among outpatients (Cerda, 2008; Yong et al., 2011). In ICU settings, an estimated 5–20% of patients experience an episode of AKI during the course of their illness, and AKI accounts for nearly 10% of all ICU bed-days (Joannidis and Metnitz, 2005; Yong et al., 2011). However, AKI usually coexists with other acute illnesses, since the incidence of AKI as single-organ failure in ICU patients is as low as 11% compared to 69% in non-ICU settings. There have been several studies on epidemiology of AKI in various settings (Waikar et al., 2008).

AKI in special populations (See also Chapters 239–241.) There are major differences in the epidemiology of AKI between geographic regions (especially in developing vs developed nations), and also across sociodemographic categories defined by age (elderly, children), gender, and social deprivation (Lameire et al., 2005, 2006, 2013; Zappitelli et al., 2008). Numerous reviews outline differences and similarities in the causes and consequence of AKI between developed and developing regions, analyse the practical implications of the identified differences, and make recommendations for management (Lameire et al., 2005, 2006, 2013). In developing countries, the reported incidence of hospital-based AKI is much lower than in developed countries (Lameire et al., 2006), likely because AKI is under-recognized in developing countries (Lameire et  al.,

7

Social demographics Old age Female Non-white Low SES

Risk Factors

Co-morbidities CVD (MI, PVD, HV arrhythmias Anaemia Chronic lung diseases Active cancer Hypoalbuminaemia sepsis Liver disease Vascular Renovascular disease Malignant hypertension Scleroderma Atheroemboli Microangiopathics (HUS, TTP) Vasculitides

AKI

Community

Health center

Medications Contrast agents NSAIDs Calcineurin inhibitors ACEI/ARBs Antibioitics/Antifungals (aminoglycosides, amphotericin B) Chemotherapy (methotrexate, cisplatin)

Hospital

Infections Acute infections (eg. streptococcal Chronic infections

Intensive care Poisonings Snake/bee stings Herbal remedies Insecticides

Non-intensive care

Environmental Heat stroke Natural disasters (eg. tsunamis) Pregnancy Septic abortions Pre-eclampsia/eclampsia Genetics G-6-P deficiency Mechanical BPH Neurogenic bladder Intra/extra ureteric obstruction Retroperitoneal fibrosis

Fig. 1.2  Risk factors for AKI.

chapter 1 

epidemiology of kidney disease

Conceptual model of acute kidney injury (AKI) ANTECENDENTS INTERMEDIATE STAGE AKI OUTCOMES

NORMAL

INCREASED RISK

COMPLICATIONS

DAMAGE

DECREASED GFR

KIDNEY FAILURE

DEATH

Cerdá J et al. CJASN 2008;3:881–886

©2008 by American Society of Nephrology

Fig. 1.3  Conceptual model for AKI. Drawn from Cerda, J., Lameire, N., Eggers, P. et al. Epidemiology of acute kidney injury. Clin J Am Soc Nephrol, 2008; 3:881–886.

2006; Cerda et  al., 2008a) due to reduced access to care, delayed referral to specialist services, and lack of dialysis services (Cerda et al., 2008a). Patients with AKI in developing nations tend to be substantially younger, are more likely to have infection-related glomerulonephritis, and are more likely to be female compared with those in the developed world (Lameire et al., 2006; Cerda, 2008)— the latter may be partially due to a high incidence of obstetrics complications such as septic abortion (Cerda et al., 2008a). In addition, seasonal increases in the risk of AKI may be observed in developing nations—due to corresponding variation in the risk of malaria, heat stroke, animal envenomation, and diarrhoeal diseases (Lameire et al., 2006; Cerda et al., 2008a; Lameire et al., 2013). Thus, the prerenal and toxic factors predominate as AKI risk factors in developing countries (Lameire et al., 2006). Preventive opportunities are often missed because of failure to recognize the risk factors and late presentation for treatment (Lameire et al., 2006). For instance, a firm aetiology for AKI cannot be established in many instances because of a lack of appropriate laboratory and technical support (Lameire et al., 2006, 2013). Regardless of the setting, most cases of AKI in children are secondary to prerenal causes from acute dehydration, major surgeries, and sepsis (Lameire et al., 2005). In contrast, the elderly are more susceptible to AKI due to multiple comorbidities and lower renal functional reserve associated with ageing (Lameire et al., 2006; Yong et al., 2011).

Prognosis and outcomes (See Chapter 237.) AKI in hospitalized patients is associated with poor prognosis, and mortality ranges from 10% to 80% depending on the population studied (Waikar et  al., 2008). Patients who present with uncomplicated AKI have a mortality of < 10% (Waikar et al., 2007). In contrast, patients presenting with AKI and other organ failure have been reported to have short-term mortality of > 50% (Ricci et al., 2008; Waikar et al., 2008; Alkandari et al., 2011; Yong et al., 2011; Singbartl and Kellum, 2012), especially if acute dialysis is required, which may be associated with short-term mortality of up to 80% (Lameire et al., 2005; Yong et al., 2011). As in CKD, lower

eGFR and heavier proteinuria independently increase the risk of AKI. However, the excess risk of mortality associated with an episode of AKI is actually lower among people with lower baseline eGFR and heavier proteinuria, as compared to those with normal kidney function (James et al., 2010).

Epidemiology of chronic kidney disease Definition, classification, and evaluation of CKD (See Chapter 94.) As for AKI, several terminologies such as ‘chronic renal failure’, ‘chronic renal insufficiency’, ‘chronic kidney failure’, ‘progressive renal failure or insufficiency’, and ‘pre-dialysis’ were used in the past to denote CKD, with further subclassification mainly by the underlying aetiology (Anonymous, 2002, 2007; Levey et al., 2011; Levey and Coresh, 2012). In 2002, the NKF-KDOQI released guidelines for the diagnosis and classification of CKD (Anonymous, 2002). These criteria standardized the nomenclature of CKD and the laboratory evaluation of kidney disease, documented the associations between level of kidney function and multiple complications, and facilitated risk stratification for adverse outcomes of CKD (Anonymous, 2002). The KDOQI criteria defined CKD by structural or functional abnormalities of the kidney for at least 3 months, manifested by either kidney damage (often persistent albuminuria) with or without a decreased eGFR to a value < 60 mL/min/1.73m2, or a decreased eGFR with or without other evidence of kidney damage for at least 3 months (Anonymous, 2002) (Box 1.1). This staging was later modified and endorsed by the international KDIGO in 2004 according to severity and treatment (Levey et al., 2011). This was further updated in 2012 following a decade of further research and practice since the publication of the initial guidelines (Stevens et al., 2013) (Table 1.4). The new classification system now incorporates a three-dimension operational definition for CKD that includes cause, GFR category, and albuminuria category (CGA). The underlying cause of CKD was added to the definition in order to highlight the importance of this information for management and prognostication (Table 1.5). Offsetting these

9

10

Section 1  

assessment of the patient with renal disease

Box 1.1  Stages of CKD by NKF-KDOQI definition ◆ Stage 1: patients with normal glomerular filtration rate (GFR), but some evidence of kidney damage as manifested by albuminuria/proteinuria, haematuria, or histological changes ◆ Stage 2: mild CKD characterized by GFR of 89–60 mL/min/1.73 m2 with some evidence of kidney disease as manifested by albuminuria/proteinuria, haematuria, or histological changes ◆ Stage

3: moderate CKD with GFR 59–30

mL/min/1.73 m2

Table 1.5 Classificationa of CKD based on presence or absence of systemic disease and location within the kidney of pathologic anatomic findings Examples of systemic diseases affecting the kidney

Examples of primary kidney diseases (absence of systemic diseases affecting the kidney)

Glomerular diseases

Diabetes, systemic autoimmune diseases, systemic infections, drugs, neoplasia (including amyloidosis)

Diffuse, focal or crescentic proliferative GN; focal and segmental glomerulosclerosis, membranous nephropathy, minimal change disease

Tubulointerstitial diseases

Systemic infections, autoimmune, sarcoidosis, drugs, urate, environmental toxins (lead, aristolochic acid), neoplasia (myeloma)

Urinary-tract infections, stones, obstruction

Vascular diseases

Atherosclerosis, hypertension, ischemia, cholesterol emboli, systemic vasculitis, thrombotic microangiopathy, systemic sclerosis

ANCA-associated renal limited vasculitis, fibromuscular dysplasia

Cystic and congenital diseases

Polycystic kidney disease, Renal dysplasia, Alport syndrome, Fa bry medullary cystic disease, disease podocytopathies

◆ Stage 4: severe CKD with GFR 29–15 mL/min/1.73 m2 ◆ Stage 5: CKD with GFR < 15 mL/min/1.73 m2, or where patient survival depends on provision of RRT in the form of dialysis or transplant. potential advantages are challenges associated with ascertaining the cause of CKD, especially in primary care. The addition of albuminuria to the definition has substantial benefits for prognostication and will undoubtedly enhance the utility of the new scheme. The evolution of these classification schemes over the last decade reflects our more refined understanding of CKD epidemiology. This trend is likely to continue in the future with identification of better markers for CKD and refinement in molecular diagnostic techniques. Although less sensitive than definitions based on SCr or eGFR, ICD-9-CM codes have been used to classify CKD (Anonymous, Table 1.4  KDIGO Classification of CKD 2012 (Stevens et al 2013; Anonymous 2013) Category

GFR, mL/ AER ACR Equivalent, Descriptor min 1.73 m2 mg/d mg/g

GFR

–

–

ANCA, antineutrophil cytoplasmic antibody; CKD, chronic kidney disease, GN, glomerulonephritis Genetic diseases are not considered separately because some diseases in each category are now recognized as having genetic determinants.

G1

≥90

–

–

Normal or high

a Note that there are many different ways in which to classify CKD. This method of

G2

60–89

–

–

Mildly decreased*†

separating systemic diseases and primary kidney diseases is only one, proposed by the Work Group, to aid in the conceptual approach.

G3a

45–59

–

–

Mildly to moderately decreased

G3b

30–44

–

–

Moderately to severely decreased

G4

15–29

–

–

Severely decreased

Definition of CKD stage 5 and ESRD

G5

2200 mg/d [ACR >2220 mg/g])

◆ 585.5: chronic kidney disease, stage 5

Modified by NT 7th Aug 2015

◆ 585.6: end-stage renal disease

(Note that the other table on the page submitted with proof corrections was unchanged table 1.5 - NT)

◆ 585.9: chronic kidney disease, unspecified.

chapter 1 

(Levey et al., 2011). CKD stage 5 and ESRD are therefore not synonymous because not all patients with CKD stage 5 receive RRT, some patients with asymptomatic stage 4 CKD (GFR 15–29.9 mL/ min/1.73 m2) may receive RRT, and kidney transplant recipients often have an eGFR > 15 mL/min/1.73 m2. Most available data on the epidemiology of ESRD are derived from registries of patients receiving chronic RRT. Because of different access to RRT across regions and healthcare systems, the incidence and prevalence of ESRD may not be fully captured by renal registries, which actually focus on ‘receipt of RRT’ (discussed below).

Pitfalls and limitations of the definition/staging system for CKD The NKF-KDOQI classification scheme has been widely adopted around the world (Eckardt et al., 2009; Ikizler, 2009; Gansevoort and de Jong, 2010). This important initiative permitted a common nomenclature for clinicians and researchers, and facilitated an international effort to educate the public about the significance of CKD. Despite these benefits, the scheme also has limitations. For example, individuals with only minimal functional or radiological abnormalities might be classified with stage 1 CKD despite the uncertain clinical relevance of these findings; elderly individuals with only mild reductions in eGFR could be classified as having stage 3 CKD—raising concerns about inappropriate labelling of healthy individuals as diseased (Chen and Hsu, 2003). Moreover, the prognosis of earlier stage disease is not necessarily better than those with more advanced stages, and progression from lower to higher stages is not necessarily inevitable. In fact, little is known about the natural history of stages 1 and 2 CKD and increasing evidence is accruing that a substantial proportion of patients with stages 3 and 4 CKD have stable eGFR (Keith et al., 2004). It has been suggested as more appropriate to divide patients into those with certain urinary abnormalities such as isolated haematuria or microalbuminuria and those with impaired kidney function (Winearls and Glassock, 2009). The latter may warrant subclassifications based on the presence or absence of progression and associated risk factors such as hypertension and proteinuria (Glassock and Winearls, 2008a; Winearls and Glassock, 2009). Guidelines from the United Kingdom by the National Institute for Health and Care Excellence (NICE) have attempted to address some of these criticisms by dividing CKD stage 3 into 3A and 3B (eGFR of 45–59 and 30–55 respectively), and the addition of the suffix ‘p’ to CKD stages to denotes significant proteinuria (Anonymous, 2007). These issues have been the subject of vigorous debate in recent years (Levey et al., 2011), and revision to the NKF-KDOQI scheme by an international KDIGO working group has recently been published (Anonymous, 2013).

Risk factors and causes for CKD A large body of epidemiological and clinical evidence has linked certain risk factors to the initiation and progression of CKD. These can be classified into distinct categories based on the presence or absence of established causation; factors that have been proven to be causal (risk factors) and those that are associated with CKD in the absence of established causal relationship (risk markers) (Anonymous, 2002; Levey et al., 2011). Causes of CKD include diabetes mellitus, hypertension, ischaemia, infection, obstruction, toxins, and autoimmune and infiltrative diseases (Anonymous, 2002; Levey et  al., 2011). Although

epidemiology of kidney disease

it is important to identify the cause(s) of CKD so that specific therapy can be instituted, adverse outcomes (including cardiovascular events and progression to ESRD) often occur despite appropriate treatment and irrespective of the underlying cause (Levey et al., 2011). Risk factors for development of CKD have traditionally been classified as susceptibility factors and initiation factors. Susceptibility factors increase susceptibility to CKD, for example, older age, male gender, and familial/genetic predisposition. Initiation factors directly initiate kidney damage; such factors include diabetes, hypertension, chronic infections, drugs, and toxins. The progression factors are risk factors associated with worsening of already established kidney damage, such as high levels of proteinuria, hypertension, poor glycaemic control in diabetes, obesity, and smoking (Anonymous, 2002, 2007; Levey et al., 2011). However, the difficulty of detecting the early stages of CKD makes it difficult to determine whether the risk factors so far identified in the population relate more to susceptibility, initiation, or even progression. Therefore, this traditional classification may be somewhat artificial. These risk factors may also be subclassified as potentially modifiable/preventable and non-modifiable (Table 1.6). Finally, risk factors can also be classified as clinical (diabetes, hypertension, autoimmune diseases, systemic infections, drugs) or sociodemographic (age, race, poverty/low income, toxins).

Global CKD epidemiology Initial population-based estimates of the prevalence of CKD were obtained from the National Health and Nutrition Examination Survey (NHANES) in the United States. The prevalence of eGFR 60 mL/min (Glassock and Winearls, 2008a). Some of these limitations are being addressed with development of more accurate and precise estimating equations—for instance, the CKD Epidemiology Collaboration (CKD-EPI) equation performs better at higher true GFR, and leads to slightly lower population-based estimates of CKD prevalence. For example, using the CKD-EPI equation rather than the MDRD equation reduced the estimated prevalence of CKD in the United States to 11.5% (95% confidence interval (CI), 10.6–12.4%) from 13.1% (CI, 12.1–14.0%) using the MDRD equation (Matsushita et al., 2010a; Stevens et al., 2010).

CKD in special populations The epidemiology of CKD may differ in special populations such as ethnic minorities, aboriginal people, Roma, the homeless, and children (Esbjorner et al., 1997; Ardissino et al., 2003; Harambat et al., 2012)—perhaps due to differences in genetic, behavioural, and sociodemographic characteristics (Foster, 2008; Huong et al., 2009; Hoy et al., 2010). There have been several population-wide surveys on the prevalence of CKD in indigenous people particularly in Australia and Canada (Gao et al., 2007; Hoy et al., 2010). Aboriginal people tend to have a higher burden of CKD than found in the general population (Katz et al., 2006; Gao et al., 2007; Hoy et al., 2010). For instance, on a remote Aboriginal island community, 26% of adults had microalbuminuria and 24% had overt albuminuria when screened in 1992–1995 (Hoy et al., 1998). Data on CKD epidemiology among ethnic black people is predominantly US based (KEEP, 2002) and information about the incidence and prevalence of CKD in Africa is sparse. In both the United States and the few communities in Africa where studies were conducted, black people tend to have a higher burden of CKD than white people. For instance, in a community survey conducted in adults attending primary care clinics in Soweto, 35% of the people screened had proteinuria; 10% needed referral to a tertiary hospital, and of these, most had stages 3–5 CKD

epidemiology of kidney disease

(Katz et al., 2006). A recent conservative estimate on prevalence of CKD in a sub-Saharan African country was 36% (Sumaili et al., 2009), as compared to 13.1% in the United States (Coresh et al., 2007).

CKD screening: measures and utility Currently, screening for CKD is accepted practice only in people with hypertension or diabetes (Anonymous, 2002). The UK CKD guidelines also recommend at least annual screening of all adults at risk of obstructive kidney disease and those with prevalent CVD, while the US KDOQI guidelines extend screening to all those aged > 60  years with testing for albuminuria (using urine albumin/ protein), SCr, and estimation of GFR (Anonymous, 2002, 2007). Recommendations from the ISN are more liberal, and advocate proactive screening for markers of CKD in all subjects visiting general practitioners, similar to the screening for high blood pressure or cholesterol concentrations (Levey et  al., 2011). However, all these recommendations are based mostly on expert consensus rather than high quality evidence, and doubt remains as to whether population or targeted screening is justifiable and cost-effective (Boulware et al., 2003; Manns et al., 2010). In cardiovascular disease (CVD), it has been argued that population screening would be more effective than targeted approaches as most cases are not derived from the minority at highest risk (Levin and Stevens, 2011). However, CVD is more prevalent than CKD, and it is uncertain whether the same arguments apply. Before population-based screening for CKD could be recommended, further research is required to determine the true prevalence of early stages of CKD in different populations; the prevalence of associated risk factors; the attributable risk associated with such risk factors; their amenability to treatment; and the incremental benefit compared to case-finding by means other than organized screening.

Prognosis and outcome in CKD Overall, even a mild reduction in estimated eGFR is associated with adverse clinical outcomes, as is increased urinary protein excretion (reviewed in (Matsushita et al., 2010b; Gansevoort et al., 2011; van der Velde et al., 2011). Among subjects with normal kidney function, proteinuria is independently associated with an increased risk of poor outcomes, which is further amplified in the setting of reduced eGFR—and has been observed in multiple populations including those at high CVD risk (Gerstein et al., 2001; Mann et al., 2001; Rahman et al., 2006; Solomon et al., 2007; Yokoyama et al., 2008; Anand et al., 2009) and in the general population (Grimm et al., 1997; Clausen et al., 2001; Kasiske, 2001; Hillege et al., 2002; Go et al., 2004; Klausen et al., 2004; Tonelli et al., 2006; Hemmelgarn et al., 2010; Matsushita et al., 2010b).

Epidemiology of renal replacement therapy Definition of RRT (See Sections 12 and 13.) RRT comprises the different forms of haemodialysis, peritoneal dialysis, and kidney transplantation and may be used to overcome a short period of kidney failure. The majority of RRT patients, however, receive chronic RRT for the treatment of ESRD (Levey et al., 2011). As mentioned earlier, not all patients with ESRD receive RRT—for example, due to differences in access to RRT across healthcare systems, or because of cultural differences in referral

13

Section 1  

assessment of the patient with renal disease Rates

Rates 2000 Rate per million population

Rate per million population

2000 1500 1000 500 0

0–19 20–44 45–64 65–74 75+ All

1500 1000 500 0

80 84 88 92 96 00 04 08

94 96 98 00 02 04 06 08

Fig. 1.4  Incidence of RRT by age category in USRDS (left) and ERA-EDTA registry (right) 1998–2012 (ERA-EDTA Registry, 2014; US Renal Data System, 2014).

patterns and patient attitudes towards RRT. Therefore, estimates based on the incidence and prevalence of chronic RRT will underestimate the burden of ESRD.

Risk factors and causes of ESRD Risk factors for ESRD are similar to those for CKD—with the caveat that characteristics that reduce the risk of mortality may also increase the risk of ESRD (since only those who survive long enough can receive RRT). Worldwide, diabetes mellitus and hypertension are the leading causes of ESRD treated with RRT. Renal registries use different systems for coding primary renal diseases underlying ESRD. The US Renal Data System (USRDS) makes use of ICD-9-CM coding, other registries use ICD-10, but the majority make use of (modifications of) the European Renal Association–European Dialysis and Transplant Association (ERA-EDTA) coding system for primary renal diseases (van Dijk et al., 2001; Venkat-Raman et al., 2012).

Global epidemiology of RRT for ESRD The incidence of RRT as reported by registries is determined by the age and sex distribution of the general population, the prevalence of diseases underlying ESRD, factors related to progression of CKD, survival from competing risk like CVD, access to RRT (Caskey et al., 2011), the timing of starting RRT in relation to GFR, and the completeness of data within those registries. The prevalence of RRT results from the incidence of RRT and the survival of RRT patients. Registry data show substantial changes in the incidence rates of RRT over the last 15 years. Figs 1.4 and 1.5 provide examples of trends in the United States and in Europe by age and primary renal disease. Rates 160 120 80 40 0

80 84 88 92 96 00 04 08

The initial rapid growth in incidence rates was likely due to a combination of population ageing, increased prevalence of diseases underlying ESRD, and an increased acceptance of older and sicker patients. In addition, a substantial rise of the number of patients starting dialysis above 10 mL/min/1.73m2 has provided a significant contribution to this increase (Rosansky et al., 2009). Recent data indicate that in the United States and in Europe the incidence rate of RRT has decreased since 2009 (US Renal Data System, 2014; ). Current differences in the incidence of RRT across the world are striking (Table 1.8). Mexico, Taiwan, and the United States have the highest incidence rates, whereas Hungary, Portugal, and Greece rank first in Europe. The prevalence of RRT has continued to increase by 1–2% per year (Kramer et al., 2009; US Renal Data System, 2014). Although in some countries this may in part be due to increased patient survival on RRT, the increase in prevalence may merely reflect higher numbers of patients starting RRT than numbers of patient deaths. Countries with the highest burden of RRT patients include Taiwan, Japan, and the United States. In general, patients on RRT have an unfavourable prognosis with remaining life expectancies being reduced to 20–35% in dialysis patients and to 70–80% in transplant recipients (ERA-EDTA Registry, 2014; US Renal Data System, 2014). In dialysis patients both cardiovascular (Foley et al., 1998; de Jager et al., 2009) and non-cardiovascular (de Jager et  al., 2009; Sarnak and Jaber, 2000) mortality are highly increased compared with those in the general population. Determinants of patient survival on RRT include age and sex, the cause of ESRD, timing of referral to a nephrologist, genetic factors, general population mortality, access to Rates

Rate per million population

Rate per million population

14

160 120

Diabetes Hypertension Glomerulonephritis Cystic disease

80 40 0 94 96 98 00 02 04 06 08

Fig. 1.5  Incidence of RRT by primary renal disease in USRDS (left) and ERA-EDTA registry (right) 1998–2012 (ERA-EDTA Registry, 2014; US Renal Data System, 2014).

Table 1.8  Incidence and prevalence estimates of RRT in different countries in 2012 Country

Incidence rate of RRT (pmp)

Incidence of DM (%)

Incidence of DM (pmp)

Prevalence of RRT (pmp)

Prevalence of RRT Prevalence of HD as (pmp) on transplant % of dialysis )

Albania

67

10

7

325

73

Argentina

159

36

57

836

168

95

Australia

112

36

40

916

411

81

Austria

140

26

36

1023

516

91

Bahrain

258

36

93

328

51

88

Belgium

188

20

37

1225

514

92

Bosnia and

125

29

36

718

52

96

720

216

91

38

59

1183

500

83

1263

203

95

Brazil

172

Canada

156

Chile

170

Croatia

158

24

38

1033

384

93

Denmark

124

28

35

872

411

80

Estonia

81

22

17

554

328

86

Finland

81

34

27

808

482

82

France

154

22

33

1139

507

93

Georgia

200

24

47

546

49

97

Greece

210

26

54

1136

231

93

484

Hong Kong

165

48

79

1192

Hungary

234

39

91

633

27

Iceland

59

0

0

683

Indonesia

191

26

50

265

Iran

105

22

23

621

299

95

395

94

86 440

73 96

Israel

183

49

89

1125

Japan

285

45

128

2365

Latvia

89

20

18

538

290

74

Malaysia

225

61

137

1056

64

91

Mexico (Jalisco)

467

59

276

1409

526

50

Montenegro

24

27

7

310

135

94

Netherlands

120

16

20

923

539

85

New Zealand

116

49

57

901

344

69

Norway

103

17

17

887

639

84

695

331

92

97

Oman

110

48

53

Philippines

117

44

51

Poland

135

25

34

748

254

94

Portugal

220

31

69

1670

602

93

Qatar

83

280

2

77

Rep. of Korea

221

51

113

1353

272

87

Romania

151

13

20

766

58

89

Russia

48

17

8

214

44

92

95

(continued)

16

Section 1  

assessment of the patient with renal disease

Table 1.8 Continued Country

Incidence rate of RRT (pmp)

Incidence of DM (%)

Incidence of DM (pmp)

Prevalence of RRT (pmp)

Prevalence of RRT Prevalence of HD as (pmp) on transplant % of dialysis )

Saudi Arabia

126

39

49

730

245

91

Serbia

123

24

30

752

106

91

Singapore

285

66

188

1741

368

88

Slovenia

122

28

34

996

310

97

Spain

120

25

30

1092

555

89

Sweden

115

23

26

933

530

79

Taiwan

450

45

203

2902

Thailand

221

38

84

906

89

90 77

Ukraine

25

12

3

147

18

85

United States

359

44

159

1976

594

91

United Kingdom

107

24

25

867

435

86

Uruguay

150

33

40

1073

316

90

Sources: ERA-EDTA 2014 and USRDS 2014; where cells are empty, data are unavailable.

transplantation, and other aspects of quality of RRT care (Kramer et al., 2012). There exists considerable international variation in the survival of patients starting dialysis (Kramer et al., 2012). Differences in patient age, sex, primary renal disease, and the presence of co-morbidities explain only a small part of that international variation (Goodkin et al., 2003; van Manen et al., 2007). Even when general population mortality rates and differences in treatment characteristics are taken into account, a major part of the variation in dialysis mortality across countries remains unexplained (van Dijk et al., 2007). Recent data suggest that the mortality of dialysis patients is higher in countries with high expenditure on healthcare. Complementary explanations for the international variation in mortality on dialysis may therefore include a more liberal acceptance policy among richer nations, differences in access to transplantation as well as different patterns of healthcare spending (Kramer et al., 2012).

RRT in special populations (See Chapter 95.) The epidemiology of RRT differs in specific populations like ethnic minorities and children. Data from the USRDS (2014) show that in 2012 the age- and sex-adjusted incidence rates of RRT for African Americans, Native Americans, and Asians were 908, 412, and 379 pmp. This is 3.3, 1.5, and 1,4 times greater than the rate of 279 found in white people. Incidence rates from Europe, albeit lower, have shown an increased incidence rate of RRT in Asians and black people (Roderick et al., 1996)  and in non-Caucasians in general (van den Beukel et  al., 2010), whereas in Canada, Australia, and New-Zealand incidence rates are highly increased in indigenous people (Dyck, 2001; McDonald, 2010; McDonald et al., 2010). Increased incidence rates in ethnic minorities frequently come together with better patient survival rates. US and UK black patients (Roderick et al., 2009; US Renal Data System, 2014), UK South Asian (Roderick et al., 2009), and non-Caucasian Dutch RRT patients (van den Beukel et  al.,

2008) have better survival rates than their white and non-Caucasian counterparts. In contrast, in Aboriginal Australians and New Zealand Maoris patient survival is lower compared with that of non-indigenous people (McDonald and Russ, 2003). The background of these survival differences requires further elucidation. In 2008, the median incidence rate of RRT in children aged 0–19 years was 9 (range 4–18) per million age-related population (pmarp) with the highest incidence rates in adolescents (Harambat et al., 2012). In children the pattern of diseases underlying ESRD is quite different from that in adults. In developed countries congenital disorders, including congenital anomalies of the kidney and urinary tract (CAKUT) and hereditary nephropathies, are responsible for more than half of all cases starting RRT. Differences in the incidence and causes of RRT across races and ethnic minorities are already reflected in paediatric populations (Harambat et  al., 2012). The prevalence of RRT was around 65 pmarp in Australia, Canada, Malaysia, and Western-Europe, whereas the United States had a higher (85 pmarp) and Japan a lower (34 pmarp) prevalence. Although mortality of children on RRT is relatively low compared with that in adults (2-year survival 96% in Europe (ERA-EDTA Registry, 2014) and 95% in the United States (US Renal Data System, 2014), their risk of death is about 30 times higher than in their healthy peers (Groothoff et al., 2002; McDonald and Craig, 2004).

Future challenges A key future challenge is to improve the definition and characterization of various kidney disease entities (Gansevoort and de Jong, 2010). This is an area of intense interest: better methods for SCr measurement, calibration, and standardization have evolved over the years, and higher-performance estimating equations are being developed (Stevens et al., 2010). In addition, newer filtration markers that are independent of muscle mass (such as cystatin C) are gaining increasing recognition and are being tested as a future alternative to SCr. Other potential biomarkers such as neutrophil gelatinase-associated lipocalin, kidney injury molecule-1, monocyte

chapter 1 

chemotactic peptide, netrin-1, and interleukin-18 among several others are becoming available for early identification of AKI (Waikar et al., 2008; Soni et al., 2011). It seems likely that some of these new tools will become more widely available, and will help to address the limitations of existing epidemiological studies. A second key challenge is to develop epidemiologic surveillance systems for both AKI and the various aspects of CKD—especially non-dialysis-dependent disease (Levey et al., 2011), and in developing nations (Barsoum, 2006; Nugent et al., 2011). Third, better information is needed on who to screen for kidney disease (as well as what test should be used and when screening should be performed)—aiming to identify people in whom early intervention to prevent adverse outcomes is feasible and cost-effective. These initiatives will help to establish the prevention and treatment of CKD as an important public health issue, and potentially as part of the WHO strategy on chronic non-communicable diseases (Couser and Riella, 2011).

References Adler, A. I., Stevens, R. J., Manley, S. E., et al. (2003). Development and progression of nephropathy in type 2 diabetes: the United Kingdom Prospective Diabetes Study (UKPDS 64). Kidney Int, 63, 225–32. Alkandari, O., Eddington, K. A., Hyder, A., et al. (2011). Acute kidney injury is an independent risk factor for pediatric intensive care unit mortality, longer length of stay and prolonged mechanical ventilation in critically ill children: a two-centre retrospective cohort study. Crit Care, 15, R146. Anand, I. S., Bishu, K., Rector, T. S., et al. (2009). Proteinuria, chronic kidney disease, and the effect of an angiotensin receptor blocker in addition to an angiotensin-converting enzyme inhibitor in patients with moderate to severe heart failure. Circulation, 120, 1577–84. Andermann, A., Blancquaert, I., Beauchamp, S., et al. (2008). Revisiting Wilson and Jungner in the genomic age: a review of screening criteria over the past 40 years. Bull World Health Organ, 86, 317–19. Anonymous (2002). K/DOQI clinical practice guidelines for chronic kidney disease, evaluation classification and stratification. Kidney Disease Outcome Quality Initiative. Am J Kid Dis, 39 Suppl 2, S1–S246. Anonymous (2005). New ICD-9-CM Code for Beneficiaries with Chronic Kidney Disease. [Online] Anonymous (2007). Chronic Kidney Disease in Adults. UK Guidelines for Identification, Management and Referral. [Online] Anonymous (2012). Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO Clinical Practice Guideline for Acute Kidney Injury. Kidney Int Suppl, 2, 1–138. Anonymous (2013). Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group. KDIGO 2012 clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int Suppl, 3, 1–150. Ardissino, G., Dacco, V., Testa, S., et al. (2003). Epidemiology of chronic renal failure in children: data from the ItalKid project. Pediatrics, 111, e382–7. Barsoum, R. S. (2006). Chronic kidney disease in the developing world. N Engl J Med, 354, 997–9. Bellomo, R., Ronco, C., Kellum, J. A., et al. (2004). Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Crit Care, 8, R204–12. Black, N. (1996). Why we need observational studies to evaluate the effectiveness of health care. BMJ, 312, 1215–18. Boulware, L. E., Jaar, B. G., Tarver-Carr, M. E., et al. (2003). Screening for proteinuria in US adults: a cost-effectiveness analysis. JAMA, 290, 3101–14.

epidemiology of kidney disease

Brown, W. W., Peters, R. M., Ohmit, S. E., et al. (2003). Early detection of kidney disease in community settings: the Kidney Early Evaluation Program (KEEP). Am J Kidney Dis., 41(1):22–35. Caskey, F. J., Kramer, A., Elliott, R. F., et al. (2011). Global variation in renal replacement therapy for end-stage renal disease. Nephrol Dial Transplant, 26, 2604–10. Cerda, J. (2008). World Kidney Day and acute kidney injury. Kidney Int, 73, 1441. Cerda, J., Bagga, A., Kher, V., et al. (2008a). The contrasting characteristics of acute kidney injury in developed and developing countries. Nat Clin Pract Nephrol, 4, 138–53. Cerda, J., Lameire, N., Eggers, P., et al. (2008b). Epidemiology of acute kidney injury. Clin J Am Soc Nephrol, 3, 881–6. Chen, M. L. and Hsu, C. Y. (2003). Should the K/DOQI definition of chronic kidney disease be changed? Am J Kidney Dis, 42, 623–5. Clausen, P., Jensen, J. S., Jensen, G., et al. (2001). Elevated urinary albumin excretion is associated with impaired arterial dilatory capacity in clinically healthy subjects. Circulation, 103, 1869–74. Coresh, J., Astor, B. C., Greene, T., et al. (2003). Prevalence of chronic kidney disease and decreased kidney function in the adult US population: Third National Health and Nutrition Examination Survey. Am J Kidney Dis, 41, 1–12. Coresh, J., Selvin, E., Stevens, L. A., et al. (2007). Prevalence of chronic kidney disease in the United States. JAMA, 298, 2038–47. Couser, W. G., and Riella, M. C. (2011). World Kidney Day 2011: protect your kidneys, save your heart. Kidney Int, 79, 483–5. Cruz, D. N., Ricci, Z., and Ronco, C. (2009). Clinical review: RIFLE and AKIN—time for reappraisal. Crit Care, 13, 211. De Jager, D. J., Grootendorst, D. C., Jager, K. J., et al. (2009). Cardiovascular and noncardiovascular mortality among patients starting dialysis. JAMA, 302, 1782–9. Drueke, T. B., Locatelli, F., Clyne, N., et al. (2006). Normalization of hemoglobin level in patients with chronic kidney disease and anemia. N Engl J Med, 355, 2071–84. Dyck, R. F. (2001). Mechanisms of renal disease in indigenous populations: influences at work in Canadian indigenous peoples. Nephrology, 6, 3–7. Eckardt, K. U., Berns, J. S., Rocco, M. V., et al. (2009). Definition and classification of CKD: the debate should be about patient prognosis— a position statement from KDOQI and KDIGO. Am J Kidney Dis, 53, 915–20. ERA-EDTA Registry (2014). ERA-EDTA Registry Annual Report 2012. Amsterdam: Academic Medical Center. Esbjorner, E., Berg, U., and Hansson, S.( 1997). Epidemiology of chronic renal failure in children: a report from Sweden 1986-1994. Swedish Pediatric Nephrology Association. Pediatr Nephrol, 11, 438–42. Feest, T. G., Round, A., and Hamad, S. (1993). Incidence of severe acute renal failure in adults: results of a community based study. BMJ, 306, 481–3. Foley, R. N., Parfrey, P. S., and Sarnak, M. J. (1998). Clinical epidemiology of cardiovascular disease in chronic renal disease. Am J Kidney Dis, 32, S112–19. Foster, M. (2008). Treating patients with CKD who are violent or homeless: a need for further study. Nephrol Nurs J, 35, 96. Gansevoort, R. T., and De Jong, P. E. (2010). Challenges for the present CKD classification system. Curr Opin Nephrol Hypertens, 19, 308–14. Gansevoort, R. T., Matsushita, K., van der Velde, M., et al. (2011). Lower estimated GFR and higher albuminuria are associated with adverse kidney outcomes. A collaborative meta-analysis of general and high-risk population cohorts. Kidney Int, 80, 93–104. Gao, S., Manns, B. J., Culleton, B. F., et al. (2007). Prevalence of chronic kidney disease and survival among aboriginal people. J Am Soc Nephrol, 18, 2953–9. Gerstein, H. C., Mann, J. F., Yi, Q., et al. (2001). Albuminuria and risk of cardiovascular events, death, and heart failure in diabetic and nondiabetic individuals. JAMA, 286, 421–6.

17

18

Section 1  

assessment of the patient with renal disease

Glassock, R. J. and Winearls, C. (2008a). An epidemic of chronic kidney disease: fact or fiction? Nephrol Dial Transplant, 23, 1117–21. Glassock, R. J. and Winearls, C. (2008b). The global burden of chronic kidney disease: how valid are the estimates? Nephron Clin Pract, 110, c39–46; discussion c47. Go, A. S., Chertow, G. M., Fan, D., et al. (2004). Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med, 351, 1296–305. Goodkin, D. A., Bragg-Gresham, J. L., Koenig, K. G., et al. (2003). Association of comorbid conditions and mortality in hemodialysis patients in Europe, Japan, and the United States: the Dialysis Outcomes and Practice Patterns Study (DOPPS). J Am Soc Nephrol, 14, 3270–7. Grimes, D. A., and Schulz, K. F. (2002). An overview of clinical research: the lay of the land. Lancet, 359, 57–61. Grimm, R. H., Jr., Svendsen, K. H., Kasiske, B., et al. (1997). Proteinuria is a risk factor for mortality over 10 years of follow-up. MRFIT Research Group. Multiple Risk Factor Intervention Trial. Kidney Int Suppl, 63, S10–14. Grootendorst, D. C., Jager, K. J., Zoccali, C., et al. (2009). Screening: why, when, and how. Kidney Int, 76, 694–9. Groothoff, J. W., Gruppen, M. P., Offringa, M., et al. (2002). Mortality and causes of death of end-stage renal disease in children: a Dutch cohort study. Kidney Int, 61, 621–9. Harambat, J., Van Stralen, K. J., Kim, J. J., et al. (2012). Epidemiology of chronic kidney disease in children. Pediatr Nephrol, 27, 363–73. Haynes, R. B., Sackett, D., Guyatt, G., et al. (2006). Clinical Epidemiology: How to do Clinical Practice Research (3rd ed.). Philadelphia, PA: Lippincott Williams & Wilkins. Hemmelgarn, B. R., Manns, B. J., Lloyd, A., et al. (2010). Relation between kidney function, proteinuria, and adverse outcomes. JAMA, 303, 423–9. Hillege, H. L., Fidler, V., Diercks, G. F., et al. (2002). Urinary albumin excretion predicts cardiovascular and noncardiovascular mortality in general population. Circulation, 106, 1777–82. Hou, S. H. and Cohen, J. J. (1985). Diagnosis and management of acute renal failure. Acute Care, 11, 59–84. Hoy, W. E., Kincaid-Smith, P., Hughson, M. D., et al. (2010). CKD in Aboriginal Australians. Am J Kidney Dis, 56, 983–93. Hoy, W. E., Mathews, J. D., Mccredie, D. A., et al. (1998). The multidimensional nature of renal disease: rates and associations of albuminuria in an Australian Aboriginal community. Kidney Int, 54, 1296–304. Hsu, C. Y., McCulloch, C. E., Fan, D., et al. (2007). Community-based incidence of acute renal failure. Kidney Int, 72, 208–12. Hsu, C. Y., McCulloch, C. E., Iribarren, C., Darbinian, J., et al. (2006). Body mass index and risk for end-stage renal disease. Ann Intern Med, 144, 21–8. Hulley, S. B., Cummings, S., and Browner, W. (2007). Designing Clinical Research: An Epidemiologic Approach (3rd ed.). Philadelphia, PA: Lippincott Williams & Wilkins. Huong, N. T., Long, T. D., Bouissou, F., et al. (2009). Chronic kidney disease in children: the National Paediatric Hospital experience in Hanoi, Vietnam. Nephrology (Carlton), 14, 722–7. Ikizler, T. A. (2009). CKD classification: time to move beyond KDOQI. J Am Soc Nephrol, 20, 929–30. Jager, K. J., Stel, V. S., Wanner, C., et al. (2007). The valuable contribution of observational studies to nephrology. Kidney Int, 72, 671–5. Jager, K. J., Zoccali, C., Macleod, A., et al. (2008). Confounding: what it is and how to deal with it. Kidney Int, 73, 256–60. James, M. T., Hemmelgarn, B. R., Wiebe, N., et al. (2010). Glomerular filtration rate, proteinuria, and the incidence and consequences of acute kidney injury: a cohort study. Lancet, 376, 2096–103. Joannidis, M. and Metnitz, P. G. (2005). Epidemiology and natural history of acute renal failure in the ICU. Crit Care Clin, 21, 239–49. Kasiske, B. L. (2001). The kidney in cardiovascular disease. Ann Intern Med, 134, 707–9. Katz, I. J., Hoy, W. E., Kondalsamy-Chennakesavan, S., et al. (2006). Chronic kidney disease management—what can we learn from South African and Australian efforts? Blood Purif, 24, 115–22.

Keith, D. S., Nichols, G. A., Gullion, C. M., et al. (2004). Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med, 164, 659–63. Klausen, K., Borch-Johnsen, K., Feldt-Rasmussen, B., et al. (2004). Very low levels of microalbuminuria are associated with increased risk of coronary heart disease and death independently of renal function, hypertension, and diabetes. Circulation, 110, 32–5. Kramer, A., Stel, V. S., Caskey, F. J., et al. (2012). Exploring the association between macroeconomic indicators and dialysis mortality. Clin J Am Soc Nephrol, 7(10), 1655–63. Kramer, A., Stel, V., Zoccali, C., et al. (2009). An update on renal replacement therapy in Europe: ERA-EDTA Registry data from 1997 to 2006. Nephrol Dial Transplant, 24, 3557–66. Lameire, N., Van Biesen, W., and Vanholder, R. (2005). Acute renal failure. Lancet, 365, 417–30. Lameire, N., Van Biesen, W., and Vanholder, R. (2006). The changing epidemiology of acute renal failure. Nat Clin Pract Nephrol, 2, 364–77. Lameire, N. H., Bagga, A., Cruz, D., et al. (2013). Acute kidney injury: an increasing global concern. Lancet, 382,170–9. Levey, A. S. and Coresh, J. (2011). Chronic kidney disease. Lancet, 379, 165–80. Levey, A. S., De Jong, P. E., Coresh, J., et al. (2011). The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report. Kidney Int, 80, 17–28. Levin, A. and Stevens, P. E. (2011). Early detection of CKD: the benefits, limitations and effects on prognosis. Nat Rev Nephrol, 7, 446–57. Liano, F., and Pascual, J. (1996). Epidemiology of acute renal failure: a prospective, multicenter, community-based study. Madrid Acute Renal Failure Study Group. Kidney Int, 50, 811–18. Liu, Y., Coresh, J., Eustace, J. A., et al. (2004). Association between cholesterol level and mortality in dialysis patients: role of inflammation and malnutrition. JAMA, 291, 451–9. Mann, J. F., Gerstein, H. C., Pogue, J., et al. (2001). Renal insufficiency as a predictor of cardiovascular outcomes and the impact of ramipril: the HOPE randomized trial. Ann Intern Med, 134, 629–36. Manns, B., Hemmelgarn, B., Tonelli, M., et al. (2010). Population based screening for chronic kidney disease: cost effectiveness study. BMJ, 341, c5869. Matsushita, K., Selvin, E., Bash, L. D., et al. (2010a). Risk implications of the new CKD Epidemiology Collaboration (CKD-EPI) equation compared with the MDRD Study equation for estimated GFR: the Atherosclerosis Risk in Communities (ARIC) Study. Am J Kidney Dis, 55, 648–59. Matsushita, K., van der Velde, M., Astor, B. C., et al. (2010b). Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet, 375, 2073–81. McDonald, S. (2010). Incidence and treatment of ESRD among indigenous peoples of Australasia. Clin Nephrol, 74 Suppl 1, S28–31. McDonald, S., Excell, L., and Jose, M. (2010). End-stage kidney disease among Indigenous peoples of Australia and New Zealand. In ANZDATA Registry Report 2010, pp. 12-1–12-19. Adelaide: Australia and New Zealand Dialysis and Transplant Registry. McDonald, S. P. and Craig, J. C. (2004). Long-term survival of children with end-stage renal disease. N Engl J Med, 350, 2654–62. McDonald, S. P. and Russ, G. R. (2003). Current incidence, treatment patterns and outcome of end-stage renal disease among indigenous groups in Australia and New Zealand. Nephrology (Carlton), 8, 42–8. Mehta, R. L. and Chertow, G. M. (2003). Acute renal failure definitions and classification: time for change? J Am Soc Nephrol, 14, 2178–87. Mehta, R. L., Kellum, J. A., Shah, S. V., et al. (2007). Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care, 11, R31. Moher, D., Liberati, A., Tetzlaff, J., et al. (2009). Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med, 6, e1000097. Noordzij, M., Hooft, L., Dekker, F. W., et al. (2009). Systematic reviews and meta-analyses: when they are useful and when to be careful. Kidney Int, 76, 1130–6.

chapter 1 

Nugent, R. A., Fathima, S. F., Feigl, A. B., et al. (2011). The burden of chronic kidney disease on developing nations: a 21st century challenge in global health. Nephron Clin Pract, 118, c269–77. Rahman, M., Pressel, S., Davis, B., et al. (2006). Cardiovascular outcomes in high-risk hypertensive patients stratified by baseline glomerular filtration rate. Annals of Internal Medicine, 144, 172–180. Ricci, Z., Cruz, D., and Ronco, C. (2008). The RIFLE criteria and mortality in acute kidney injury: a systematic review. Kidney Int, 73, 538–46. Roderick, P., Byrne, C., Casula, A., et al. (2009). Survival of patients from South Asian and Black populations starting renal replacement therapy in England and Wales. Nephrol Dial Transplant, 24, 3774–82. Roderick, P. J., Raleigh, V. S., Hallam, L., et al. (1996). The need and demand for renal replacement therapy in ethnic minorities in England. J Epidemiol Community Health, 50, 334–9. Rosansky, S. J., Clark, W. F., Eggers, P., et al. (2009). Initiation of dialysis at higher GFRs: is the apparent rising tide of early dialysis harmful or helpful? Kidney Int, 76, 257–61. Rothman, K. (2002). Epidemiology: An Introduction. Oxford: Oxford University Press. Rothman, K. J. (1981). The rise and fall of epidemiology, 1950–2000 A.D. N Engl J Med, 304, 600–2. Sarnak, M. J., and Jaber, B. L. (2000). Mortality caused by sepsis in patients with end-stage renal disease compared with the general population. Kidney Int, 58, 1758–64. Schulz, K. F., Altman, D. G., and Moher, D. (2010). CONSORT 2010 statement: updated guidelines for reporting parallel group randomised trials. BMJ, 340, c332. Singbartl, K., and Kellum, J. A. (2012). AKI in the ICU: definition, epidemiology, risk stratification, and outcomes. Kidney Int, 81(9), 819–25. Singh, A. K., Szczech, L., Tang, K. L., et al. (2006). Correction of anemia with epoetin alfa in chronic kidney disease. N Engl J Med, 355, 2085–98. Solomon, S. D., Lin, J., Solomon, C. G., et al. (2007). Influence of albuminuria on cardiovascular risk in patients with stable coronary artery disease. Circulation, 116, 2687–93. Soni, S. S., Pophale, R., and Ronco, C. (2011). Review: new biomarkers for acute renal injury. Clin Chem Lab Med, 49(8), 1257–63. Srisawat, N., Hoste, E. E., and Kellum, J. A. (2010). Modern classification of acute kidney injury. Blood Purif, 29, 300–7. Stel, V. S., Jager, K. J., Zoccali, C., et al. (2007). The randomized clinical trial: an unbeatable standard in clinical research? Kidney Int, 72, 539–42. Stevens, L. A., Schmid, C. H., Greene, T., et al. (2010). Comparative performance of the CKD Epidemiology Collaboration (CKD-EPI) and the Modification of Diet in Renal Disease (MDRD) Study equations for estimating GFR levels above 60 mL/min/1.73 m2. Am J Kidney Dis, 56, 486–95. Stevens, P. E., Levin, A., Kidney Disease: Improving Global Outcomes Chronic Kidney Disease Guideline Development Work Group (2013). Evaluation and management of chronic kidney disease: synopsis of the kidney disease: improving global outcomes 2012 clinical practice guideline. Ann Intern Med, 158(11), 825–30. Stevens, P. E., Tamimi, N. A., Al-Hasani, M. K., et al. (2001). Non-specialist management of acute renal failure. QJM, 94, 533–40. Stroup, D. F., Berlin, J. A., Morton, S. C., et al. (2000). Meta-analysis of observational studies in epidemiology: a proposal for reporting. Meta-analysis Of Observational Studies in Epidemiology (MOOSE) group. JAMA, 283, 2008–12. Sumaili, E. K., Cohen, E. P., Zinga, C. V., et al. (2009). High prevalence of undiagnosed chronic kidney disease among at-risk population in Kinshasa, the Democratic Republic of Congo. BMC Nephrol, 10, 18. Tonelli, M., Wiebe, N., Culleton, B., et al. (2006). Chronic kidney disease and mortality risk: a systematic review. J Am Soc Nephrol, 17, 2034–47.

epidemiology of kidney disease

Tripepi, G., Jager, K. J., Dekker, F. W., et al. (2008). Bias in clinical research. Kidney Int, 73, 148–53. US Renal Data System (2014). USRDS 2014 Annual Data Report: An Overview of the Epidemiology of Kidney Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Van Biesen, W., Vanholder, R., and Lameire, N. (2006). Defining acute renal failure: RIFLE and beyond. Clin J Am Soc Nephrol, 1, 1314–19. Van den Beukel, T. O., Dekker, F. W., and Siegert, C. E. (2008). Increased survival of immigrant compared to native dialysis patients in an urban setting in the Netherlands. Nephrol Dial Transplant, 23, 3571–7. Van den Beukel, T. O., Jager, K. J., Siegert, C. E., et al. (2010). Racial minority groups on dialysis in Europe: a literature review. Clin Nephrol, 74 Suppl 1, S78–84. Van der Velde, M., Matsushita, K., Coresh, J., et al. (2011). Lower estimated glomerular filtration rate and higher albuminuria are associated with all-cause and cardiovascular mortality. A collaborative meta-analysis of high-risk population cohorts. Kidney Int, 79, 1341–52. Van Dijk, P. C., Jager, K. J., De Charro, F., et al. (2001). Renal replacement therapy in Europe: the results of a collaborative effort by the ERA-EDTA registry and six national or regional registries. Nephrol Dial Transplant, 16, 1120–9. Van Dijk, P. C., Zwinderman, A. H., Dekker, F. W., et al. (2007). Effect of general population mortality on the north-south mortality gradient in patients on replacement therapy in Europe. Kidney Int, 71, 53–9. Van Manen, J. G., Van Dijk, P. C., Stel, V. S., et al. (2007). Confounding effect of comorbidity in survival studies in patients on renal replacement therapy. Nephrol Dial Transplant, 22, 187–95. Venkat-Raman, G., Tomson, C. R., Gao, Y., et al. (2012). New primary renal diagnosis codes for the ERA-EDTA. Nephrol Dial Transplant, 27(12), 4414–19. Von Elm, E., Altman, D. G., Egger, M., et al. (2007). The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Lancet, 370, 1453–7. Waikar, S. S., Curhan, G. C., Wald, R., et al. (2006). Declining mortality in patients with acute renal failure, 1988 to 2002. J Am Soc Nephrol, 17, 1143–50. Waikar, S. S., Liu, K. D., and Chertow, G. M. (2007). The incidence and prognostic significance of acute kidney injury. Curr Opin Nephrol Hypertens, 16, 227–36. Waikar, S. S., Liu, K. D., and Chertow, G. M. (2008). Diagnosis, epidemiology and outcomes of acute kidney injury. Clin J Am Soc Nephrol, 3, 844–61. Wilson, J. M. and Jungner, G. (1968). Principles and Practice of Screening for Disease. Public Health Paper No. 34. Geneva: World Health Organization. Winearls, C. G. and Glassock, R. J. (2009). Dissecting and refining the staging of chronic kidney disease. Kidney Int, 75, 1009–14. Yokoyama, H., Oishi, M., Kawai, K., et al. (2008). Reduced GFR and microalbuminuria are independently associated with prevalent cardiovascular disease in Type 2 diabetes: JDDM study 16. Diabetic Medicine, 25, 1426–32. Yong, K., Dogra, G., Boudville, N., et al. (2011). Acute kidney injury: controversies revisited. Int J Nephrol, 2011, 762634. Zappitelli, M., Parikh, C. R., Akcan-Arikan, A., et al. (2008). Ascertainment and epidemiology of acute kidney injury varies with definition interpretation. Clin J Am Soc Nephrol, 3, 948–54. Zhang, Q. L., and Rothenbacher, D. (2008). Prevalence of chronic kidney disease in population-based studies: systematic review. BMC Public Health, 8, 117.

19

CHAPTER 2

Clinical assessment of the patient with renal disease: overview Christopher G. Winearls Introduction The next chapters describe the presentations of renal diseases, their assessment from the history and physical examination, and their investigation by the use of laboratory tests, imaging, and histopathology. Nephrologists are referred patients by their colleagues in primary care or in other specialties when they believe that there is a kidney problem needing an expert. That judgement is not always correct, for example, oedema or haematuria often have causes outside the kidney. The reasons for referral are in three categories: 1. To explain abnormalities attributable to kidney disease that have been found in asymptomatic individuals including those at risk of familial conditions. 2. Symptomatic renal disease: renal failure, either acute or chronic; abnormalities of urination including poly- and oliguria, visible haematuria; unexplained loin pain; the classic renal syndromes (nephritic and nephrotic). 3. Renal consequences of systemic conditions: metabolic, inflammatory, infectious; drugs, malignancy, pregnancy, organ failure (especially cardiac and hepatic). When responding to referrals coming by letter, telephone, or email there are two questions to be asked: 1. Should one be involved in this patent’s care and if so how? In other words, will one be able to add value? The problems seldom come neatly packaged. They have often emerged into prominence in the context of complex other diseases, both acute and chronic. Sometimes the referral is for diagnosis, sometimes it is management, and sometimes it is both. These dilemmas are particularly poignant for nephrologists who have at their disposal two powerful tools. They can examine the diseased organ directly by examining tissue obtained by biopsy and can, theoretically, replace its function indefinitely. The question should sometimes be not, ‘Can I?’ but, ‘Should I be involved?’ It takes some courage to admit that one has nothing to offer, for example, the patient with oliguria and terminal heart failure or malignant disease. It is hubristic to suggest interventions directed at a consequence of irremediable disease which may prolong suffering for no more than a few days and delay a merciful death and so one should pause before doing so. The supplementary question is; ‘If I should be involved, should I be the clinician in overall charge?’ When the renal disease dominates the clinical problem, for example,

in acute dialysis dependent renal failure, there are advantages in one team determining the priorities and supervising drug prescription. There are others in which a supportive role is sufficient. To take charge of every patient with a renal component to their illness would overwhelm the service. This is a particular problem in patients with renal failure as an additional complication of complicated surgery. Such patients need to be managed in the service responsible for treating the root cause. Without so doing, the renal problem will not resolve either. 2. How soon does one need to be involved? Contact may need to be immediate, for example, when there is an acute uraemic emergency or delayed as a routine outpatient assessment. This ‘triaging’ is not always straightforward. Some patients with life- and kidney-threatening conditions can appear deceptively well. Two examples come to mind—acute cast nephropathy in myeloma and rapidly progressive glomerulonephritis. One should not be deceived by the automatic labelling of a patient with an abnormal creatinine that has been translated into an estimated glomerular filtration rate (eGFR), as having chronic kidney disease usually abbreviated in correspondence to ‘CKD’. Having accepted the referral the nephrologist should clear his/ her mind of the potentially misleading glib statements within referral letters and clinical notes, assumptions of the diagnosis, and the prejudices of the referrer or indeed the patient. Who has not seen the oedematous ‘nephrotic’ patient without proteinuria who actually has heart failure or the patient with a reduced eGFR labelled as having ‘CKD’ explained by anything from bodybuilding to hypothyroidism? A  good rule is to assume that the referral lacks several pieces of highly relevant information and includes some that are actually wrong—‘no nephrotoxic drugs’, ‘normal-sized kidneys’, and ‘no relevant family history’ are common false assertions. It is then time to apply the ‘clinical method’ which comprises the following steps: ◆ Identification of the presenting complaint and its immediate history. The history should be methodically retaken including the general history of the patient including past medical, drug, social, travel, illness in the family, and a systematic enquiry of the function of other systems. ◆ Performing a general physical examination. ◆ Arranging laboratory and imaging investigations—screening and targeted.

chapter 2  ◆ Constructing

a differential diagnosis. ◆ Formulating a plan for management: specific and general. ◆ Communicating the conclusions to the patient, referring colleagues and others who will be needed to contribute to care. There is no substitute for ‘the clinical method’. It is a false economy to move straight to investigations not only because these may not provide the answer but because their interpretation will depend

clinical assessment of patients with renal disease on the clinical context. Radiologists and renal pathologists rightly expect nephrologists to describe the problem and how the findings will alter management. Accurate diagnosis is paramount for it informs the medical management exactly but knowing everything else will dictate how this will be delivered. The way the clinical method is applied depends on the clinical presentation—these are described in Chapter 3.

21

CHAPTER 3

Presentations of renal disease Christopher G. Winearls Introduction It is not patients who seek the advice and help of nephrologists but primary care physicians and medical and surgical colleagues, who have identified a problem that they believe calls for the expertise of a nephrologist. What are these circumstances?

Abnormalities in asymptomatic patients raising the suspicion of renal disease The referral usually follows the finding of an abnormality on clinical or laboratory examination that is not causing symptoms. These arise at routine medical examinations at school, for employment, military service, life and salary insurance, immigration, major surgery, potential live kidney donation, registration with a new family doctor, ‘well woman/man’ clinics and as part of health screening in at-risk individuals with systemic disease, at booking for pregnancy care, and assessment of the risks of methods of contraception. The potential for causing anxiety is significant, especially if the patient attends a renal unit for evaluation, walking past signs to the dialysis and transplant wards. This concern has to be balanced against the wish to make an early diagnosis of renal disease in case it is possible to halt it, delay its progress, and defer or prevent the onset of renal failure. We will consider first the common reasons patients are referred to a nephrologist for the first time.

Abnormal findings in patients at risk of renal disease and complications Many referrals will come from colleagues in other disciplines who are on the lookout for renal problems associated with the condition they are treating or monitoring. Indeed they will have been advised to measure the estimated glomerular filtration rate (eGFR) more or less frequently. ◆ Primary care physicians are encouraged to screen for renal dysfunction in patients with diabetes, hypertension, heart failure, and vascular disease (ischaemic heart disease, strokes, and peripheral vascular insufficiency) and those on long-term treatment with drugs known to cause kidney injury, for example, lithium, non-steroidal anti-inflammatory drug, and calcineurin inhibitors. ◆ Urologists will refer patients with neurogenic bladders, urinary diversion, kidney stones, retroperitoneal fibrosis, and even relieved obstruction. ◆ Gastroenterologists will refer patients with malabsorption who are prone to oxalate nephropathy, and those on 5-aminosalicylic acid drugs used for inflammatory bowel disease which can cause interstitial nephritis at any time after they are instituted.

◆ Cardiologists

and vascular surgeons will refer patients with atherosclerotic disease and abdominal aortic aneurysms may also suffer from renal vascular disease which comes to light when renin–angiotensin–aldosterone system (RAAS) blockade is instituted. ◆ Haematologists will refer patients with paraprotein disorders. ◆ Rheumatologists will refer patients with vasculitides in whom they are suspicious of relapse in the hope that a renal biopsy will confirm this and justify escalation of immunosuppression. ◆ Radiologists will often add as addendum to their reports of incidental findings of renal cysts, kidney size asymmetry, and parenchymal calcification, ‘suggest referral to a nephrologist’.

Invisible haematuria (microscopic haematuria) (See Chapter 46.) This is by definition asymptomatic and is found when urine is tested with urine dipsticks as part of clinical assessment. This problem is not referred to nephrologists in the first instance.

Asymptomatic proteinuria (See Chapter 50.) This may have been detected by dipstick testing either as a part of a routine assessment or under abnormal conditions, for example, an intercurrent illness. Dipsticks can be misleading because they are not calibrated for urinary concentration. An early morning spot urine protein or albumin creatinine concentration should be requested and the presence of orthostatic proteinuria excluded before the patient is formally referred. The threshold for full evaluation of significant proteinuria (defined as > 300 mg/day or 30 mg/ mmol creatinine) up to and including renal biopsy will depend on the clinical context and the ‘need to know’.

Abnormal constituents of urine (See Chapter 175.) These include asymptomatic bacteriuria, and leucocyturia (see Chapter 175). Asymptomatic bacteriuria without leucocyturia suggests sample contamination, with leucocytes, an asymptomatic infection or urinary tract colonization. Isolated leucocyturia is found after treatment for an infection, in subjects with renal stones, interstitial nephritis, renal tuberculosis, or papillary necrosis.

An ‘abnormal’ eGFR (See Chapter 94.) This has become one of the most common reasons for referral to a nephrologist. This has arisen from the application of the Modification of Diet in Renal Disease and Chronic Kidney Disease Epidemiology

chapter 3 

presentations of renal disease

Collaboration eGFR equations to routine plasma creatinine measurements which are then reported as an eGFR and an interpretation as to a possible stage of chronic kidney disease (CKD), for example, ‘an eGFR of 45–59 mL/min/1.73m2 indicates stage 3A kidney disease’. Such statements are neither useful nor legitimate not least because a diagnosis of CKD requires the abnormal GFR to have been present for at least 3 months but also because the explanation may be acute kidney injury (AKI). It is also misleading because an eGFR of > 60 mL/min/1.73m2 is not necessarily ‘normal’. Indeed, there is no definition of a normal eGFR. Although an eGFR of < 60 mL/min is likely to indicate impairment of kidney function across the age range, the significance is different in the young compared to the old. A decision as to whether individuals with a reduced eGFR need a formal nephrological assessment will depend on their underlying co-morbid conditions, age, and the rate of change. The new KDIGO CKD guideline has the explicit requirement that CKD should be described according to cause, GFR, and albumin excretion—the CGA system (see Chapter 94). When the GFR is unequivocally abnormal (< 45mL/ min/1.73m2) or lower than predicted for the age of the patient there will need to be a decision on the urgency of the assessment. These dilemmas are explored and recommendations made in the KDIGO clinical practice guideline (Kidney Disease:  Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group, 2012).

consequence of RAAS blockade with angiotensin converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs), or potassium-sparing diuretics such as spironolactone or amiloride. This is becoming a common problem in patients receiving such combinations as part of their treatment for heart failure. When renal function is relatively preserved, these abnormal K+ concentrations, usually approximately 6.5 mmol/L do not cause problems.

A raised blood pressure

Urinary symptoms

Patients are referred if there are indications that this may be secondary to renal disease (reduced eGFR, haematuria, and or proteinuria). They will require a full nephrological assessment. (See Chapter 210.)

Screening for an inherited disorder known to be present in the family Common examples include autosomal dominant polycystic kidney disease (ADPKD), Alport syndrome, reflux nephropathy, urate nephropathy, and nephronophthisis-medullary cystic disease complex. (See Sections 15, 16.)

Hyper- and hyponatraemia (See Chapters 28 and 29.) Hypernatraemia is seldom asymptomatic but hyponatraemia (sodium < 135 mmol/L) often is. This is usually a consequence of salt wasting with mainly water replenishment, psychogenic polydipsia, and the syndrome of inappropriate antidiuretic hormone secretion (SIADH) a full description of this problem is provided in Chapter 28.

Symptomatic urinary disease The patient has symptoms and signs that strongly suggest an underlying kidney or urinary tract disease. Patients in this group of referrals may be suffering urinary symptoms or from one of the recognized renal syndromes that trigger automatic referral to a nephrologist. These are listed in Table 3.1. Patients with problems of urination (micturition) are not referred to nephrologists in the first instance but these may be relevant in cases referred with haematuria, recurrent urinary tract infection. Such patients will be referred either in the hospital setting or as requests for advice from the primary care. There is often overlap with urology and choice and direction of the referral is often perverse and random. Table 3.1  Urinary syndromes and symptoms

Incidental electrolyte abnormalities

Syndrome

Definition

(See Section 2.)

Nephrotic

Oedema associated with hypoalbuminaemia and heavy proteinuria (usually > 3 g/day)

Nephritic

Hypertension and oedema associated with haematuria and proteinuria

Acute kidney injury

This is based on urine output and change in plasma creatinine concentration (Ftouh et al., 2013)

Chronic kidney disease

Abnormalities of kidney structure or function present for > 3 months with implications for health.

Chronic kidney failure

An eGFR < 15 mL/min causing uraemia requiring symptom control, dialysis, and or renal transplantation

Hypokalaemia (K+ < 3.5 mmol /L) (See Chapter 34.) This will often cause concern. When associated with hypertension, a number of disorders will have to be considered. These include renovascular hypertension and Cushing, Conn, and rarely Liddle syndromes. If the blood pressure is normal the simple explanations should be considered first, for example, thiazide and loop diuretic use, diarrhoea, and purgative use. The abuse of such drugs can be hard to prove. A rare explanation is spurious hypokalaemia caused by delay in potassium (K+) measurement in patients with a membrane pump abnormality. This is also seen in patients with acute myeloblastic leukaemia. All these may have been excluded before the nephrologist is asked to confirm or deny the presence of Gitelman or Bartter syndromes.

Hyperkalaemia (K+ > 5.5 mmol/L) (See Chapter 34.) This also causes concern and the risk-averse doctor will refer patients to hospital urgently for further blood tests. The explanations range from the trivial, a haemolysed or old blood sample especially if taken with a rare tight tourniquet, to the relevant, for example, a

Symptoms: Urinary symptoms Recurrent macroscopic haematuria Loin pain ± haematuria Hypertensive renal disease Dysuria, polyuria

23

24

Section 1  

assessment of the patient with renal disease

Dysuria This is an umbrella term covering different sensations of discomfort with micturition and experienced, therefore, several times a day. In men, burning in the penile urethra suggests urethritis but a deeper pain suggests problems in the prostate or bladder. If there are systemic symptoms and signs, the prostate should be considered and samples taken after prostatic massage and imaging organized. Acute prostatitis can progress to septicaemia and the treatment is different from cystitis. Cystitis is uncommon in men unless there is an underlying structural or predisposing cause such as a stone, outflow obstruction, or malignancy. In women, dysuria is often associated with urinary urgency and frequency suggesting a diagnosis of cystitis. This occurs more commonly in women during their sexually active years but also after menopause when the effects of oestrogen deficiency reduce the defences of the bladder. Dysuria also describes difficulty in actually passing urine. This is more common in men who will describe a collection of distressing symptoms pointing to bladder outflow obstruction. Because bladder emptying is incomplete, they notice frequency, including nocturia, urgency, difficulty in initiating micturition (waiting up to a minute for flow to start), a poor stream, and then dribbling after micturition is thought to be finished. The finding of a full bladder often palpable to the umbilicus, a large volume of post-micturition residual urine, and a thick-walled bladder on bladder scan give the diagnosis. The causes include urethral structure and prostatic occlusion of the urethra. This diagnosis is often made when these men, often elderly, are referred unexamined with ‘CKD’. Strangury is the symptom of very painful and difficult micturition often caused by a bladder stone at the internal urethral meatus or in the urethra itself.

Urinary frequency, polyuria, and nocturia It is essential to establish what the patient actually means when they describe ‘Having to go to the toilet all the time’. ‘How often and how much urine?’ are the questions that need answering. It is best to get a record of a day’s worth of micturition, charting time and volumes passed. A kitchen measuring jug will suffice. Frequent large volumes of urine point to a concentration defect and frequent small volumes to a micturition problem or bladder irritation or a contracted bladder volume, setting off detrusor contraction despite the presence of relatively small volumes. The causes of true polyuria > 3 L/day are many (see Table 3.2). The diuresis may be the result of excess water consumption

Table 3.2  Causes of polyuria Cause

Diagnosis

Primary polydipsia

History and/or water deprivation

Solute diuresis

Urinary glucose

Drugs

Diuretic use

Central diabetic insipidus

Water deprivation test

Nephrogenic diabetes insipidusa

History, kidney imaging, and a water deprivation test

a Common causes: post-obstructive, hypercalcaemia, long-term lithium use or excess dose

acutely, and many forms of interstitial kidney injury. (See Chapter 32.)

(polydipsia), obligate loss caused by solutes such as glucose, or failure of ‘antidiuresis’ at a central (posterior pituitary) or peripheral (the kidney) level. These are simply distinguished by direct questioning and occasionally a water deprivation test with and without administration of arginine vasopressin. Primary or congenital diabetic insipidus is rare and will manifest in early life. Nocturia is the disturbing need to wake and micturate during the sleep cycle. Provided the bladder has been emptied before retiring and fluids have been avoided in the preceding 1–2 hours, most healthy subjects can last until morning perhaps because sleep is associated with antidiuretic hormone (ADH) secretion. When the volumes of urine passed are low, the causes are those of frequency described earlier. Large volumes may be the continuation of the problem of polyuria but a reversed problem of micturition, that is, more at night than in the day is more problematic. It implies that there is more urine production at night than during the day and is attributed to preferential renal perfusion when the other calls on cardiac output are reduced. It is commonly seen in patients with CKD, congestive heart failure, and nephrotic syndrome.

Anuria and oliguria (See Chapter 219.) It is uncommon for patient to report these and curious that they do not notice that they have ‘stopped going’. Anuria and oliguria (urine output < 400 mL/24 hours or < 0.5 mL/kg per hour) are emergencies and in the otherwise symptom free patient point to relatively few causes. Anuria The differential diagnosis includes acute renal vascular occlusion, catastrophic renal injury, for example, antiglomerular basement membrane (anti-GBM) disease, and bilateral complete urinary obstruction or obstruction of a single kidney. Oliguria There are many causes, including those of anuria, but there is usually evidence of a systemic disorder with or without a further aggravant or precipitant, for example, dehydration in a patient with heart failure and an ACEI and diuretics. Both oliguria and anuria are observed in the hospital setting and will trigger an entry into an AKI algorithm examining prerenal, renal, and postrenal causes (Lameire et al., 2013).

Macroscopic haematuria This is an alarming symptom and because of association of ‘red’ with danger, is seldom ignored. In the United Kingdom, any adult patient with painless macroscopic haematuria would be referred to a urologist to be seen within 2 weeks to exclude malignancy. Referral to a nephrologist is usually after the common urological causes have been excluded. The obvious ‘medical’ renal causes include ADPKD and over-anticoagulation but three glomerular diseases can manifest as macroscopic haematuria: Immunoglobulin A (IgA) nephropathy, Alport syndrome, and anti-GBM disease. Patients with the nephritic syndrome describe brown cloudy urine which is less alarming than truly bloody urine.

Loin pain This problem is usually referred when a urological cause has already been sought but it is worth rehearsing the symptoms and causes associated with loin pain.

chapter 3 

Renal colic is a very severe pain during which no normal activity can be undertaken. It is sudden in onset, comes in waves, radiates anteriorly and into the genitalia, and is associated with nausea and vomiting. It can abate quite suddenly. The urine will often but not always test positive for blood. This description implies that a stone or clot or papilla is in the ureter which is trying to move it on by peristalsis. These episodes are usually managed in the community and are only referred if the problem is recurrent or there are ‘red flags’. These include fever (implying the possibility of infection behind an obstructing stone), known solitary kidney, pain resistant to standard analgesia, pregnancy, renal dysfunction, oliguria, or poor social support. It is more difficult to attribute pain confined to the loin to the presence of a stone. There are other causes such as bleeds into renal cysts, pyelonephritis, renal infarcts, pelvi-ureteric junction obstruction, and the loin pain haematuria syndrome. This is a curious condition in which patients present with very severe chronic loin pain with and without visible haematuria and few if any abnormalities are found by imaging or even renal biopsy; the description of the pain is vivid and by the time of referral many patients are taking very large doses of opiate analgesics. Occasionally a factitious cause of haematuria is proved. On examination the patients are exquisitely tender during attempts to palpate the kidney bi-manually. There will often be a request for a surgical solution ranging from auto transplantation to nephrectomy. It is considered a form of somatoform pain disorder (Winearls and Bass, 1994) (See Chapter 47). Acute glomerulonephritis is occasionally associated with loin discomfort but seldom with severe pain. Severe loin pain has also been attributed to the ‘nutcracker phenomenon’ when the left renal vein is compressed between the aorta and the superior mesenteric artery.

Nephrotic syndrome (See Chapters 48, 52.) This is a shorthand term for the combination of oedema, heavy proteinuria, and hypoalbuminaemia. Cut-off concentrations as diagnostic criteria are unhelpful as there is a poor correlation with the effects of the syndrome. Hypercholesterolaemia is an epiphenomenon and not helpful diagnostically. Usually the plasma albumin is < 30 g/L, the urine protein loss > 3 g/24 hours or > 350 mg/mmol creatinine. The diagnosis may be missed if the latter two components are not sought and the oedema misattributed to immobility, heart failure, and venous insufficiency. Adult patients need prompt assessment by a nephrologist and almost all will require a renal biopsy. The cost of guessing the pathology by the known hierarchy of causes in age groups is too high to be allowed.

presentations of renal disease

used today as most patients with this combination have recognized glomerular diseases such as IgA nephropathy, Henoch–Schönlein purpura (HSP), systemic lupus erythematosus (SLE) with a diffuse proliferative glomerulonephritis, or a systemic vasculitis.

Acute kidney injury (See Section 11.) AKI is defined as any of the following:  an increase in serum creatinine by > 0.3 mg/dL (> 26.5 μmol/L) within 48 hours; or an increase in serum creatinine to > 1.5 times baseline, which is known or presumed to have occurred within the prior 7 days; or a urine volume < 0.5 mL/kg/h for 6 hours (Kidney Disease:  Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group, 2012). In practice, AKI covers a spectrum from transient minor to catastrophic kidney injury arising in both hospital and the community and always represents a nephrological emergency. There are two tasks: one is to attempt to halt or reverse the injury and the other is to provide support to the patient to compensate for the effects of renal failure (Kidney Disease:  Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group, 2012; Lameire et al., 2013).

Symptomatic chronic kidney disease (See Section 5.) The overt symptoms of chronic disease are usually late in the illness and deceptively non-specific. Once CKD or kidney failure have been found by the reporting of the eGFR the nephrologist has to decide whether these symptoms are indeed a consequence of the kidney disease. The lower the eGFR or more advanced the stage the more likely this is true. Patients with CKD stage 5 are almost always symptomatic though they often only acknowledge the severity after they have been dialysed, but those with CKD stage 3A and 3B are not. At one extreme the symptom complex will include the full range of uraemic consequences affecting almost all systems: ◆ Dyspnoea

is explained by pulmonary oedema, anaemia, and acidosis ◆ Anorexia and weight loss ◆ Pruritus ◆ Cognitive decline ◆ Sexual dysfunction by the central effects of the elusive uraemic toxins ◆ Skeletal discomfort and proximal weakness by secondary hyperparathyroidism

Nephritic syndrome (See Chapter 46.) This term describes the combination of oedema, hypertension, glomerular haematuria, and proteinuria (not in the ‘nephrotic’ range) with or without a reduction in GFR. A  looser definition is glomerular haematuria with any of the features. Unlike the nephrotic syndrome the patient has evidence of a significantly expanded extracellular volume with a raised jugular venous pressure. This and the oedema are attributed to sodium and water retention caused by an acute inflammatory injury to the glomeruli. This is a classical complication of beta haemolytic streptococcal infection in children. It is rare in adults. The term is not much

The patient has a systemic disorder known to be complicated by renal involvement There are many conditions in which the kidney is the victim of collateral damage and this can be severe enough to mean that the nephrologist has to take responsibility for overall care.

Metabolic diseases and inherited disorders (See Chapters 149, 167, 334.) The most important is diabetes mellitus that is now the single most common cause of end-stage renal disease. CKD clinics

25

26

Section 1  

assessment of the patient with renal disease

are now dominated by patients with earlier stages of diabetic nephropathy hoping that good conservative management will prevent or delay progression. Unfortunately many have passed beyond the point of reversibility so the major contribution of the nephrology clinic is in helping to find a tolerable and effective combination of blood pressure-lowering drugs. Tuberous sclerosis, sickle cell disease, and other rarer disorders such as Anderson–Fabry disease or cystinosis also cause renal failure and their care has to be shared with experts in their other manifestations.

Malignancy (See Chapters 60, 63, 150, 172, 251.) Apart from the obstructing effects of solid tumours on the renal tract there are non-metastatic effects such as membranous and membranoproliferative glomerulonephritis and hypercalcaemia. Chemotherapy with agents such as cisplatin has adverse effects on the kidney which if extreme make renal replacement necessary. The tumour lysis syndrome is less common now that the risks have been recognized but still occurs in patients with high tumour burdens (especially leukaemias) responding to effective chemotherapy. This is a renal emergency requiring prolonged dialysis to control potassium, urate, and phosphate concentrations. The most common joint malignancy-kidney problem is in paraprotein disorders the effects of which range from acute cast nephropathy to the deposition disorders such as AL amyloid, lightand heavy-chain deposition disease.

Infection (See Sections 8 and 11.) Sepsis and shock are the most common cause of AKI in the hospital setting but there are community acquired infections that present with life-threatening renal injury too. The best examples are falciparum malaria, Escherichia coli O157 causing haemolytic uraemic syndrome (HUS), leptospiral infection causing Weil disease, hantavirus infection, and post-infectious proliferative glomerulonephritis (usually a consequence of beta haemolytic streptococcal infection). Subacute and chronic infections can lead to glomerular injury too:  hepatitis B and C, HIV, and bacterial endocarditis are the most common. Renal tuberculosis is quite rare in the developed world but not so in emerging economies.

Auto-immune inflammatory disorders (See Chapters 156, 162, 165, 166.) The management of the renal consequences of the vasculitides especially SLE, HSP, systemic sclerosis, the polyangiitides are a significant part of the acute and long-term workload of clinical nephrologists. Sarcoidosis and Sjögren syndrome can involve the kidney and evidence of interstitial inflammation obtained from

renal biopsy provides justification for immunosuppressive treatment which is otherwise usually held in reserve.

Effects of drugs (See Chapter 362.) Most drugs are two-edged swords and the prescribers dread finding changes in renal function especially if they are uncertain as to whether the drug is really the cause. This is a particular problem in oncology (cisplatin and intravenous pamidronate), rheumatology, and infectious disease (antiretroviral and antituberculosis drugs, high-dose aciclovir and sulphonamides, and amphotericin are prime examples). Although drug withdrawal is an option, a definite diagnosis of the nature of the kidney injury is preferable. An allergic interstitial nephritis will require active treatment not just stopping the agent.

Pregnancy (See Chapter 250.) Pregnancy is a systemic state in which the kidneys are vulnerable especially if there is any underlying renal disease. The catastrophic effects of the later specific complications such as eclampsia, haemolysis, elevated liver enzymes, low platelets (HELLP) syndrome, and HUS almost always require nephrology input.

Other system failure (See Chapters 247–9.) Patients with the end stages of heart and liver failure are usually referred with oliguria, disproportionate plasma urea concentrations, and hyperkalaemia, all explained by a combination of the system failure and the valiant pharmacological attempts to ameliorate their condition with various diuretics, ACEIs, and spironolactone. They are usually hypotensive and the decision on whether to offer renal support is finely balanced, especially if the underlying cause is irremediable.

References Ftouh, S., Thomas, M., on behalf of the Acute Kidney Injury Guideline Development Group (2013). Acute kidney injury: summary of NICE guidance. BMJ, 347, f4930. Hole, B., Whittlestone, T., and Tomson, C. (2014). BMJ, 349, g6768. Kidney Disease: Improving Global Outcomes (KDIGO) (2012). KDIGO clinical practice guideline for glomerulonephritis. Kidney Int Suppl, 2, 1–274. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group (2012). KDIGO Clinical Practice Guideline for Acute Kidney Injury. Kidney Int Suppl, 2, 1–138. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group (2013). KDIGO clinical practice guideline for the evaluation and management of chronic kidney disease. Kidney Int Suppl, 3, 1–150. Lameire, N. H., Bagga, A., and Cruz, D., et al. (2013). Acute kidney injury: an increasing global concern. Lancet, 382, 170–9. Winearls, C. G. and Bass, C. (1994). The loin pain haematuria syndrome. Nephrol Dial Transplant, 9, 1537–9.

CHAPTER 4

Kidney disease-focused history taking Christopher G. Winearls Introduction Each referral must be approached with an open mind. No assumptions should be made and every statement should be checked. One should not believe or disbelieve what has been asserted.

The presenting complaint This will usually be in the pick list described in Chapter 3. One should start with the simple introduction, ‘I have been asked to see you because [presenting complaint] but I would like to know what you believe to be the problem.’ What concerns the patient is not always what concerns the doctor. For example, a patient consults a primary care physician because of tiredness, and is found to be anaemic and have renal impairment. It is easy to accept this sequence as causes and effect but it is often not so. Ten per cent of the population have ‘chronic kidney disease’ (CKD) but a minority are anaemic so one has to dissect the problem not just agree with the proposition.

The history of the presenting complaint This involves a detailed exploration of the onset, duration, progression, alleviating and aggravating features, and associated symptoms. This is well illustrated by the visible haematuria of mesangial immunoglobulin (Ig)-A disease which is of acute onset, painless, may follow an infection, and is of short duration. It is quite different from that of a bladder tumour. The patient’s understanding of medical words should be explored. What do they mean by ‘UTIs’, cystitis, ‘kidney pain’, migraine, angina? The same applies to the phrases in the referral which will often include vague and often misleading terms such as chronic pyelonephritis, essential hypertension, and ‘PET’.

The past medical history One is greatly assisted by a good general practice record of attendances or the hospital notes which should be examined from the first page. The past history may come as a cryptic computer printout that has been assembled over years and passed from one family practitioner to another. Each should be asked about and the basis of the diagnosis confirmed. This needs to be exhaustive, with prompting because patients have

variable memories and notions of what constitutes a significant medical problem, for example, diabetes may be considered a longstanding background nuisance rather than a relevant condition ‘because I have had it for so long and it does not cause me any problem’. Childhood illnesses are particularly difficult as the adult patient will not be accompanied by a parent and may have ‘grown out’ of the problem. Reflux nephropathy is often suspected from the story of the individual being a sickly child, with frequent fevers, courses of antibiotics, or enuresis which eventually resolved. Every operation, every medication, every hospitalization should be recorded. In women the history of pregnancies can be very revealing, for example, proteinuria at booking or blood pressure concerns that do not fit pre-eclampsia. Men have fewer opportunistic medical assessments but findings at employment, insurance, or military service medicals are helpful.

The drug (medication) history One is often informed of the currently prescribed drugs, not those that may have triggered the problem or exacerbated it and have been discontinued. Examples of the missing culprits include lithium for bipolar disorder; 5-aminosalicylic acid for ulcerative colitis; proton pump inhibitors; gold or penicillamine for rheumatoid arthritis; intravenous bisphosphonates for breast cancer; cisplatin; non-steroidal anti-inflammatories; angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin receptor blockers (ARBs); and calcineurin inhibitors for non-transplant indications. There is seldom mention of the ‘over-the-counter’ drugs but the referrer may be unaware of these so the patient should be quizzed. Perhaps the most notorious example of this problem is Chinese herbal remedies which contain aristolochic acid.

The family history This is perhaps one of the most neglected elements of history taking, for doctors rely on the catch-all question, ‘Do any diseases run in your family?’ Instead one should ask about causes of death and disability in three generations. Adult polycystic kidney disease commonly declares itself because a parent is on renal replacement treatment but there are families in which the affected parent has died prematurely, has been estranged or has not yet been diagnosed.

28

Section 1  

assessment of the patient with renal disease

The social history Patients are quite surprised that one wants to know so much about their ethnicity, where they were born, where they have lived, where they have travelled, what work they have done, what their habits—personal, sexual, smoking, drug use illicit and recreational—are. These provide clues to disorders such as HIV-associated nephropathy, focal segmental glomerulosclerosis, Chinese herb nephropathy, renal tract tumours, and atherosclerotic renal arterial disease. Of course serious renal disease, especially when renal replacement treatment is required, has far reaching implications.

One has to explore employment, relationships and domestic arrangements.

The review of systems This can be tedious and unrewarding especially if the patient is trying to be helpful and agrees that they do have symptom such as chest pain or breathlessness ‘occasionally’. It is best to leave the question open allowing them to declare and describe a symptom spontaneously. ‘How is your breathing?’ is preferred to ‘Do you ever get breathless?’ This said, one does have to be quite direct on sensitive issues such as sexual function.

CHAPTER 5

Kidney disease-focused features on examination Christopher G. Winearls Introduction This depends on the context and clinical presentation and the emphasis will be different too. Tell-tale signs are often unnoticed in the general examination of the eyes (lecithin cholesterol acyltransferase deficiency (LCAT) deficiency, Fabry disease, corneal calcification), the skin (vasculitis, Anderson–Fabry disease), the optic fundus (haemorrhages and exudates, papilloedema), and the hands (nail patella syndrome, splinter haemorrhages of SLE and subacute bacterial endocarditis) (see Figs 5.1 and 5.2). In the cardiovascular system, one is most interested in the volume status (the pulse, jugular venous pressure, and presence of oedema and its extent); blood pressure measured accurately with the an appropriate-sized cuff (the patient rested and lying down for 10 minutes and then standing); the left ventricular (LV) impulse for signs of LV hypertrophy (LVH), murmurs, and a pericardial friction rub; and the peripheral pulses. In the respiratory system one is looking for pleural effusions, signs of chronic suppurative lung disease, and clues to bronchial malignancy. In the abdomen it is the prominence of the kidneys, the bladder, and the presence of renal bruits that one wants to record. Pelvic examination is not usually routine but the slightest hint of bladder outflow problems or obstruction requires an examination of the prostate or the uterine cervix. In the nervous system one is interested in cognitive function reflecting uraemia; peripheral neuropathy in uraemia and diabetes; and autonomic neuropathy in diabetes and AL amyloid.

Investigations These will be arranged after the completion of the examination but there are general principles that need stating (see Chapters 6–18). The rule is to start with the simple and avoid requesting a comprehensive and unselected set of blood tests, the results of which may be distract or be difficult to interpret. Before the consultation ends, one will want to know what the urine dipsticks show, for example, haematuria, proteinuria, leucocytes, and nitrites, and, if appropriate, urine microscopy (see Chapter 6). A full blood count and biochemical screen including calcium, phosphate, and liver enzymes, and a C-reactive protein are reasonable routine requests because they may provide a clue to common disorders that have rather non-specific symptoms or may be clinically silent.

Once the problem is described the investigations are directed very specifically at diagnosis and consequences. For example, if one is sure one is dealing with true chronic kidney disease (CKD) one wants imaging of the renal tract, a further estimate of GFR, urine protein quantification, an electrocardiogram and echocardiogram looking for LVH, and a parathyroid hormone test for evidence of mineral and bone disorder (MBD). The consultation is closed with a description of the working diagnosis, the way it will be resolved but avoiding a description of specific treatments, unless generic or justifiable, or indeed prognosis. This is more often the business of the follow-up. It is wise to counsel patients about the risks of searching the Internet which can cause unnecessary anxiety. There are good informative websites to which patients can be referred, for example, the website of the Edinburgh Renal Unit ().

Specific consultations The acute uraemic emergency (See Chapter 222.) In the hospital setting, the nephrologist will usually be consulted either because the diagnosis is obscure or there is a need for renal replacement treatment. This will need to be instituted urgently if there are any of the life-threatening consequences of renal failure: pulmonary oedema, metabolic acidosis, hyperkalaemia, and encephalopathy. The same rules apply to community-referred patient after which the diagnostic algorithm is followed seeking prerenal, renal, and postrenal causes. As emphasized earlier, the clinical method takes precedence over investigation especially as physical examination is not usually diagnostic. Laboratory tests tend to be confirmatory rather than immediately diagnostic except in the silent renal diseases such cast nephropathy and anti-glomerular basement membrane (GBM) disease. There should be a low threshold for performing a renal biopsy unless there is a very obvious cause.

The patient with CKD stages 3–5 (See Section 5.) These referrals are made for one of two reasons. First because an estimated GFR (eGFR) has been interpreted as a stage of CKD, for example, ‘an eGFR of 45–59 mL/min/1.73m2 indicates stage 3A kidney disease’. A decision as to whether individuals with a reduced eGFR need a formal nephrological assessment will depend on their

30

Section 1  

assessment of the patient with renal disease

(A)

(B)

(C)

(D)

(F)

(I)

(J)

(E) (G)

(K) (H)

Fig. 5.1  Physical signs on general examination of the hands and peripheries. (A) White bands in the nails—evidence of hypoalbuminaemia in a patient who had an episode of nephrotic syndrome. (B) Palmar crease hyperpigmentation in a patient referred with hyperkalaemia caused by adrenal failure (Addison’s disease). (C) Broad fingers in a potential living donor found to be hypertensive and acromegalic. (D) White shiny hands in a patient with scleroderma. (E) Limb purpura in a patient with the nephritic syndrome caused by type 2 essential cryoglobulinaemia. (F) A digital infarct in a patient with microscopic polyangiitis. (G) Widespread purpura in a young man with AKI and disseminated intravascular coagulation caused by meningococcal septicaemia. (H) Gouty tophus in a patient on long-term diuretics. (I) Finger pulp infarct in a patient with nephritic syndrome caused by endocarditis. (J) Digital ischaemia in a patient with AKI and disseminated intravascular coagulation following a dog bite. The infecting organism was Capnocytophaga canimorsus. (K) Angiokeratoma corporis diffusum I in the bathing trunk distribution in a patient with Anderson–Fabry disease.

underlying co-morbid conditions, age, and the rate of change. Is this actually acute kidney injury (AKI) in an early stage or CKD? The common trap is the older patient with non-specific symptoms and a lower than predicted GFR and a bland urine sediment attributed to ‘CKD’ who actually has AKI caused by interstitial nephritis or myeloma. These dilemmas are explored and recommendations made in the Kidney Disease: Improving Global Outcomes (KDIGO) clinical practice guideline (Lameire et al., 2013).

The second reason is that a patient has presented with symptoms, for example, of anaemia and routine testing has shown that renal function is more or less abnormal.

The first referral The first task is to confirm that the patient does indeed have CKD. The eGFR may misrepresent the true renal function so one needs to take account of ethnicity, intercurrent events, the consumption of

chapter 5 

(A)

(B)

(D)

(E)

(F)

kidney disease-focused features on examination (C)

(H)

(G)

(I)

(J)

Fig. 5.2  Signs in the eyes and the face in patients with kidney disease. (A) The puffy eyelid of patient with unexplained CKD G3a A1 who was found to have myxoedema. (B) Limbic calcification in a young woman on long-term haemodialysis with autonomous hyperparathyroidism. (C) Collapsed nasal bridge (saddle nose) in a patient with granulomatosis with polyangiitis. (D) Subconjunctival haemorrhages in a patient with Weil’s disease. (E) Skin ecchymoses and jaundice in a man with AKI and Weil’s disease. (F) Cloudy corneas in a woman with LCAT deficiency causing CKD. Courtesy of Prof. A. J. Rees. (G) Prominent subconjunctival blood vessels in a patient with Anderson–Fabry disease, who presented with proteinuria and acroparaesthesia. (H) Differing skin pigmentation in a young woman with membranous nephropathy who had used mercury containing skin-lightening cream. (I) Facial lipodystrophy in a woman with nephrotic syndrome caused by minimal-change glomerulonephritis. (J) The taut skin and telangiectasiae in a patient with scleroderma and AKI caused by hypertension as part of a scleroderma crisis.

drugs that alter creatinine concentration but not GFR (trimethoprim and cimetidine), or those that do alter GFR reversibly such as non-steroidal anti-inflammatory drugs. The second is to establish the chronicity of the CKD and its immediate relevance to the health of the patient. CKD has greater implications in younger subjects than in much older people with co-morbidity. The third is to find out the

rate of change which will inform the conversation about prognosis. One can establish a cause in approximately 80% of patients but in the older group one often has to resort to suggesting vascular and hypertensive disease. Intensive attempts to make a diagnosis are only justified if a remediable or modifiable cause is possible. Finally one documents the presence or absence of CKD complications.

31

32

Section 1  

assessment of the patient with renal disease

The follow-up patient Patients with CKD of at different stages constitute a majority of the attendees at general nephrology clinics. They seldom fail to attend and until the more advanced stages have few symptoms attributable to their kidney disease. They will nevertheless share a variety of problems fearing that they are manifestations of CKD or because they have been told so. When asked how they are feeling, they will generally say, ‘I am fine’. It is advisable to press them for symptoms of uraemia and if possible engage their accompanying relative. Some patients are in denial, others have accommodated their disabilities, and a few fear that any complaint will lead to the institution of dialysis. One needs to move on to individual elements of the CKD syndrome including its cause, the complications, and then the plans for the future. Much of this will rely on the laboratory tests, best undertaken before the actual visit. Measuring the blood pressure, the weight, and the demeanour are a good neutral start, setting the patient at their ease. That done, a conversation on their understanding of the nature of CKD, its potential for progression, and what one might do when ‘the kidneys are no longer doing enough to keep you healthy’ is needed. As patients dread reaching end stage, one reviews the serial eGFRs, answering the inevitable question, ‘What can I do to avoid the need for dialysis?’ This allows one to reinforce the need for good blood pressure control, adherence to a sensible diet, and forgoing tobacco. One then checks the complications including anaemia, MBD, and cardiovascular risk. Cardiovascular disease occurs prematurely in patients with CKD and many patients will die of it before reaching end-stage renal disease. Consideration should be given to measures to reduce cardiovascular risk such as lowering low-density lipoprotein (LDL) cholesterol and controlling blood pressure (which is already a focus of slowing progression of their kidney disease). Do not be misled by a ‘normal’ LDL cholesterol concentration: this predicts cardiovascular risk poorly in patients with CKD and lowering LDL cholesterol is effective regardless of the starting concentration. Depending on the stage (usually late stage 4)  one talks about options ranging from transplantation, through various forms of dialysis, to medical care without dialysis (sometimes called maximal conservative care—not a useful term because it suggests that there is minimal conservative care). In some services the patients approaching stage 5 are referred to ‘low clearance clinics’. These have the advantage of reliable adherence to process, easy access to dietetic, dialysis, vascular access and transplantation, and other advice and a well-mapped pathway. It does, however, separate the patient from his or her nephrologist who may have cared for the patient for many years. It is possible for each stage 4/5 patient to have a pro forma or their own logbook to record and ensure that all the basic elements of CKD care have been covered. Their drug regimens should be reviewed particularly if there may have been additions made by other disciplines. As the GFR falls it is worth asking whether the ACEI should be continued, especially if the K+ is drifting above and beyond the upper limit of normal.

The patient with unexplained haematuria (invisible and visible) The patient will be referred to a nephrologist in the first instance if the primary care physician has excluded a trivial or non-urinary tract cause (gynaecological), confirmed that the abnormality is persistent, the patient is under 50 years of age, and if there is collateral

evidence of kidney diseases such as proteinuria or renal dysfunction. Referral to a urologist first is advised in patients over the age of 34–40 years or if they are younger if there is no obvious renal cause. They will seek to exclude stones, infections including in the prostate, non-infective inflammation, and tumours. If a urologist has excluded an obvious structural cause, the issue for the nephrologist is to decide whether this finding is significant, that is, is there an underlying disorder of the kidney requiring a renal biopsy for diagnosis? This will be decided on the basis of the red cell appearances being suggestive of a glomerular origin, the presence of any collateral evidence of renal disease, a ‘need to know’, and the policies of the medical system (Hole, Whittlestone and Tomson, 2014). A nephrologist is not usually involved until primary care physicians and urologists have exhausted common causes which are infection, stones, inflammation, and a malignancy anywhere from the renal parenchyma to the urethra. The timing of the haematuria helps a little: observed early and then clearing suggests a urethral problem; at the end of micturition, a bladder cause; and throughout, an upper urinary tract cause. The appearance is important—fresh red haematuria with clots points to bleeding from the urinary tract, muddy brown (oxidized blood) suggests a renal parenchymal cause such as glomerulonephritis. When such patients do reach the nephrologist he/she will want to be reassured that the cause of ‘haematuria’ is indeed blood and not haemoglobin or myoglobin. The absence of red cells on urine microscopy will be the alert. There are other causes of discoloured urine but they do not look like haematuria. Patients on rifampicin produce orange urine; in porphyria the urine has to stand before changing colour. The common three renal causes of macroscopic haematuria are immunoglobulin (Ig)-A disease, especially during periods of infection; anti-GBM disease; and adult polycystic kidney disease. Finally there is the problem of factitious haematuria which may be part Munchhausen syndrome. The complaint of haematuria is intermittent and inconveniently (for the doctor) difficult to confirm. An immediate cystoscopy revealing clear urine within the bladder and from both ureteric orifices within hours of a bloody urine being provided is the best way to establish the diagnosis. The handling of this finding is challenging but it allows both urologist and nephrologist to desist from further intrusive investigation. This presentation is sometimes in combination with loin pain in what is blandly called the ‘loin pain haematuria syndrome’. This is a common entity in which the pain is out of proportion to the examination, imaging, and pathology findings which are minimal and non-specific. It is resistant to conventional analgesia and the doses and duration of opioid analgesics exceptional. The descriptions of the pain are graphic and appear exaggerated as the pain is claimed to be present at the time (Winearls and Bass, 1994).

The patient with proteinuria or the nephrotic syndrome (See Chapters 50 and 52.) This consultation will concentrate on the indication for a renal biopsy and dealing with the consequences of the nephrotic state. The key question in the asymptomatic patient with proteinuria is whether a biopsy will provide useful prognostic information and change management. The case for a biopsy is very strong in nephrotic patients as it is dangerous to offer trials of treatment with corticosteroids except

chapter 5 

in children. Serological tests can offer clues but seldom certainty. When tests for antiphospholipase A2 receptor antibodies become widely available the decision will be more difficult in those adults suspected of suffering from membranous nephropathy who have a positive test. The nephrotic patient will want relief from uncomfortable oedema and the nephrologist will have to make a judgement about the need or otherwise for anticoagulation KDIGO clinical practice guideline for glomerulonephritis.

The patient with diabetes mellitus (See Chapter 149.) There are specific extra tasks in patients with diabetes mellitus and renal disease. The first is to test the assumption that the renal disease is indeed a consequence of diabetes. It is assumed to be so especially if the patient has longstanding diabetes, retinopathy, and proteinuria. Even when these are present the explanation can be another disease. A sudden change in renal function, ‘nephrotic’ range proteinuria, haematuria, and new systemic symptoms should alert the nephrologist to another cause. Diabetes occurs in approximately 5% of the population so it is no surprise when they are found to have conditions such as IgA nephropathy or steroid-sensitive nephrotic syndrome. Control of diabetes in patients with reduced renal function can be difficult. Insulin requirements fall and sulphonylurea drugs have prolonged half-lives. The patients fearing hypoglycaemia may eat more to prevent this. Metformin can rarely cause lactic acidosis if the dose is not modified, or an additional illness supervenes. The advice on metformin use is different in the United States from Europe. Glipizide is the preferred sulphonylurea in renal failure because it is not excreted by the kidney. Tight control of diabetes does not improve outcomes in patients with established nephropathy and may actually have an adverse effect on survival so haemoglobin A1c targets are less stringent. Of all patients with CKD, diabetics have the worst cardiovascular risk so this aspect of management—smoking cessation, statin use, good blood pressure control, and a low threshold for investigation of ischaemic heart disease—should be emphasized. In older patients, target blood pressures should be a matter for clinical judgement and are > 120/80 mmHg. One should emphasize to the patient that blood pressure control offers the best chance of delaying the need for renal replacement and reducing the risk of cardiovascular events. This may require combinations of drugs but not angiotensin receptor blockers and angiotensin-converting enzyme inhibitors (ACEIs) as so-called dual renin–angiotensin–aldosterone system blockade is associated with an increased risk of adverse events. See the KDIGO guideline for further information (Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group (2013).

The review of the haemodialysis patient (See Chapter 255.) Depending on the frequency of review this can be a short or a long consultation. The patient and the nephrologist are likely to have different agendas and it is best to start with theirs. One wants to gain an understanding of how dialysis is affecting their lives and whether they are actually gaining the maximum benefit. A  ‘Good Dialysis Index’ proposed by John Agar () can be adapted. For example, one can start with the open-ended questions to allow the patient to tell one how treatment is affecting their everyday life: ◆ ‘How are you feeling today?’ ◆ ‘What symptoms do you have during and after your treatment?’ ◆ ‘Can you eat and drink what you want?’ ◆ ‘How does having a dialysis treatment affect what you want to do?’ ◆ ‘If you are not already retired, are you working?’ ◆ ‘Have you been in hospital recently? If so, for what?’ One can then move to the nephrologist’s agenda: ◆ ‘How is your vascular access?’ ◆ ‘How many hours of treatment are you actually having?’ ◆ ‘How much weight do you need to take off on your first dialysis of the week?’ ◆ ‘How often do you have a problem with low blood pressure during treatment?’ ◆ ‘How many blood pressure tablets are you taking?’ ◆ ‘How much “Epo” do you need? Is your haemoglobin in range?’ ◆ ‘Is your phosphate a problem? How many binders do you need?’ ◆ (If appropriate) ‘Are you on the transplant waiting list?’ ◆ (If appropriate) ‘Can we talk about home haemodialysis?’ ◆ Has your cardiovascular risk been addressed? Are you on a statin?’ The examination is often neglected in busy clinics but should include a look at the fistula, the skin, the pulses, the muscles and the ease of rising from a chair, the heart sounds to be sure that calcific aortic stenosis is not stealthily developing, and the lung bases for signs of chronic pulmonary congestion. Only then should one scan the laboratory results reflecting good treatment of anaemia, bone mineral balance, and nutrition.

The review of the peritoneal dialysis patient (See Chapter 263.) These consultations are generally conducted with the peritoneal dialysis (PD) nurse specialists who will review the practical issues. They may have undertaken this in the patient’s home before nephrology review. ◆ Is the catheter working and is the exit site clean? ◆ Is the bowel habit regular? ◆ Is the regimen correct and controlling the concentration of uraemic markers and fluid balance? This is assisted by the peritoneal equilibration test and an ‘adequacy test’. They will assess nutrition both under and over. ◆ Is the technique being practised properly and is a refresher course of instruction needed? The nephrologist will concentrate on the overall well-being of the patient, checking that they are still content with their choice of modality and reassuring them that PD is delivering good renal replacement but mindful of the possibility that a switch to haemodialysis may become necessary in the future as residual function declines or the function of the peritoneal membrane wanes. The blood pressure and cardiovascular risk factors will be scrutinized along with the medication list, specifically asking whether the

33

34

Section 1  

assessment of the patient with renal disease

cardiovascular risk been addressed and that prescription of a statin has been considered. The aim of this is to reduce the tablet load, particularly blood pressure agents, analgesics, and often unnecessary drugs such as proton pump inhibitors (PPIs) and H2 antagonists. The need for and dose of an erythropoiesis-stimulating agent, iron repletion, and markers of MBD will be examined. Finally they will want to be sure that renal transplantation has been considered and if so the patient is listed

The review of the stable transplant patient (See Section 13.) These routine visits require a systematic review of the function of the renal transplant and the health of the patient. Unless specific issues are volunteered one moves to the checklist. The threshold for investigating for opportunistic infections and malignancy should be low. ◆ Graft function is assessed by the plasma creatinine and compared to the best achieved and the last measurement. The eGFR has not replaced the plasma creatinine in routine care. Urine is not routinely examined for evidence of infection but dipsticks for proteinuria and invisible haematuria are advisable if only to be alert to recurrent disease. ◆ A review of the medication, taking the opportunity of stopping drugs such as H2 antagonists, PPIs that have hung over from the early postoperative phase, the dose of the main immunosuppressive drugs based on trough levels, and white cell count. Is the patient on appropriate agents to reduce cardiovascular risk (such as a statin) and/or bone protection? Can the blood pressure medications be reduced or simplified?

◆

The blood pressure should be measured properly, that is, after the patient has been sitting quietly for 10 minutes. ◆ There should be a brief look at the skin and the organization of the annual dermatology review confirmed.

The assessment of the potential living kidney donor (See Chapter 277.) This is a strange consultation for a nephrologist because the patient is usually healthy and one’s task is to confirm this but one has to be sure not to miss a hitherto unrecognized reason not to proceed. One starts by satisfying oneself that the potential donor is not under any duress and is well informed about the procedure and its consequences. There are three issues: ◆ Is it safe for the recipient? This is essentially asking whether there are any chronic blood-borne viral infections or potential abnormalities of the kidney that may preclude transplantation, for example, vascular abnormalities or renal tumours. ◆ Is the kidney healthy enough to be transplanted—is the GFR satisfactory, is there proteinuria or any evidence of CKD? This is a concern for both recipient and donor. ◆ Is the donor healthy enough to undergo non-essential surgery and are the risks of being uni-nephric acceptable? Occasionally one has to rely on a 24-hour ambulatory blood pressure monitoring, genetic testing in ADPKD, or even renal biopsy if there is invisible haematuria.

CHAPTER 6

Urinalysis Walter P. Mutter

Urinalysis aids in the diagnosis of renal disease especially in cases when a renal biopsy is not immediately available or is contraindicated. It is most informative when done by the treating physician with knowledge of the clinical context (Fogazzi et al., 1998). Inspection is done by eye. Routine chemical analysis is done by dipstick but urine microscopy is essential for it may reveal abnormalities even when chemical evaluation is normal (Szwed and Schaust, 1982).

Osmolality is a measure of dissolved particles in solution with a normal range of 50–1200 mOsm/kg. Specific gravity is generally directly proportional to urine osmolality but may be disproportionately higher in the presence of larger molecules such as glucose and iodinated contrast when measured by refractometry. Urine dipsticks are usually used to estimate specific gravity as they are rapid and convenient. They do not detect non-ionic substances such as radiocontrast, urea, or glucose (Siemens, 2010). Alkaline urine may cause falsely low readings. Proteinuria in the range of 100–750 mg/dL may cause elevated readings (Siemens, 2010).

Inspection

Protein

Colour

(See Chapter 7 for further detail.) Proteinuria may be quantified by dipstick, protein to creatinine ratio, albumin to creatinine ratio, or by a 24-hour urine collection. Reagent strips are most sensitive to albumin. If non-albumin protein is suspected, analysis of total urine protein using turbidometric or dye-binding techniques is recommended (Fogazzi et al., 2008a). Sensitivity is in the range of 15–30 mg/dL albumin. Microalbuminuria, defined as urine albumin excretion between 30 and 300 mg/day or 20 and 200 mg/L, may not be detected by dipstick testing. Very alkaline urine and gross haematuria may yield false-positive results (Pugia et al., 1999).

Introduction

Normal urine colour varies from almost clear to dark yellow as a result of urobilin pigment. The depth of the colour depends on the urine concentration. Table 6.1 lists typical causes for abnormal urine colour.

Clarity Cloudy or turbid urine is most commonly caused by urinary tract infections but may be due to contamination from vaginal secretions, faecal material, gross haematuria, crystals, or lipids as in chyluria (Fogazzi et  al., 2008a). Larger particles including cells, crystals, and casts will usually sediment with centrifugation.

Odour Ammonia ions give urine its typical odour. A foul or feculent odour may signify urinary tract infection. Sweet or fruity smelling urine may be present in diabetes, ketosis, or maple syrup urine disease. Phenylketonuria gives off a musty odour, isovaleric academia a sweaty foot odour, while hypermethioninaemia gives off a rancid or fishy odour (Fogazzi et al., 2008a).

Chemical analysis Chemical analysis of the urine is typically performed using a chemical reagent dipstick composed of absorbent squares that change colour after application of urine. Colour change may be interpreted by eye or by a colorimeter. A standard dipstick will measure up to 10 urine parameters including protein, blood, leucocytes, nitrite, glucose, ketones, pH, specific gravity, bilirubin, and urobilinogen.

Urine concentration Urine concentration may be measured as specific gravity or urine osmolality.Specific gravity is the weight of a solution divided by the weight of water. The normal range for urine is 1.005–1.030.

Blood The urine dipstick detects haemoglobin by measuring its peroxidase activity. Haemoglobin catalyses the oxidation of a chromogen to produce a coloured product. Myoglobin also catalyses this reaction, resulting in a positive test. A positive dipstick in the absence of red blood cells (RBCs) in the urine sediment should prompt consideration of haemolysis or rhabdomyolysis. The urine dipstick is quite sensitive to the presence of blood in the urine detecting haemoglobin levels > 0.015 mg/dL and in excess of three RBCs/μL (Siemens, 2010). False positives are possible if microbial peroxidase is present. False negatives may be seen in patients taking ascorbic acid (vitamin C) (Brigden et al., 1992).

Glucose Small amounts of glucose, usually 10 leucocytes/μL (Siemens, 2010). Sensitivity may be decreased by concentrated urine that prevents cell lysis as well as cephalexin, glucose, oxalic acid, and tetracycline (Fogazzi et  al., 2008a; Siemens, 2010).

Urine microscopy Identification of casts, cells, and crystals relies on the experience of the observer. Not surprisingly there is significant variation in the interpretation of urinary sediment findings (Wald et  al., 2009). For example, there is no consensus among nephrologists as to how many granular or renal tubular cell casts are required to support a clinical diagnosis of acute tubular necrosis (ATN) (Chawla et al., 2008). Inter-observer variability among nephrologists of common urinary findings has been reported. Ten nephrologists in a single centre agreed on the presence of dysmorphic RBCs only 31% of the time (Wald et al., 2009). However, nephrologists may perform better than hospital-based technologists in the identification of some characteristics including renal tubular epithelial (RTE) cells, granular casts, RTE casts, and dysmorphic RBCs (Tsai et al., 2005). Urine chemistry including pH, specific gravity, blood, and protein should be obtained in advance of the microscopic examination. Interval re-examination of the urine sediment is advised in patients with changing clinical status or where the initial findings are discordant with the clinical presentation.

Microscope selection Phase contrast microscopy may have advantages for cellular identification and is recommended in some guidelines (European Confederation of Laboratory Medicine, 2000; Kouri et al., 2000). However, many nephrologists use bright field microscopy and this may be the only available option (Fogazzi and Grignani, 1998). Both methods have been used for research and either is acceptable. A polarizing filter is helpful for identifying lipids and certain crystals.

Urine collection Unfortunately, there is a lack of standardization in the way urine is collected and processed in clinical practice (Fogazzi and Grignani, 1998). The following approach, adapted from published guidelines, is recommended (Fogazzi et al., 1998, 2008a). Every patient presenting to the nephrology clinic should undergo at least one complete urinalysis (Fogazzi et  al., 2008a). Written instructions should be provided to the patient in advance of a clinic visit. Ideally urine should not be collected within 72 hours after strenuous exercise or during menstruation in women. Men should retract the foreskin when present and women should spread the labia. The use of cotton wipes for urethral opening in both genders is reasonable (Fogazzi et al., 2008a). Urinary catheter specimens should be avoided. Urine should be collected midstream in a sterile plastic container. Urine from the first morning void is generally more concentrated, increasing the sensitivity for detecting abnormalities. However, prolonged stasis in the bladder may lead to lysis or degradation of cells or casts (Fogazzi et al., 2008a). Analysis within 2–3 hours is recommended as urine pH and temperature change

chapter 6 

urinalysis

rapidly after voiding. Formed elements such as cells and casts may degrade, especially in alkaline or dilute urine. Refrigeration at 4ºC is reasonable for short-term storage when urine cannot be examined immediately, particularly if culture is to be performed. However, this may lead to precipitation of calcium or uric acid crystals.

Preparation of the sample for microscopy Once collected, 10 mL of urine is spun in a plastic tube in a table top centrifuge at 400 G (2000 RPM for 10 minutes) then 9.5 mL of urine is aspirated or decanted and the remaining 0.5 mL agitated gently using a pipette. One drop, about 50 µL, is applied to a glass slide and covered with a glass cover slip. Examination begins at low power (100 ×) looking for cells, casts, and crystals ideally including 20 microscopic fields. Prolonged examination of 50 low power fields is appropriate for urine with few formed elements or when looking for the presence or absence of RBC casts or dysmorphic RBCs (Fogazzi et al., 2008a). Increased magnification up to 400 × is used for more detailed examination and cell identification. Urine elements may be quantified using counting chambers containing a fixed volume of urine. In the alternative, the nephrologist reports the number of objects per HPF at 400 ×. When done carefully, this technique correlates reasonably well with counting chambers (Fogazzi et al., 1998).

Crystals Crystals are commonly observed in urine, occurring in approximately 8% of unselected samples (Fogazzi, 1996). They are likely to be clinically significant when associated with renal failure such as calcium oxalate crystals in ethylene glycol poisoning, uric acid crystals in tumour lysis syndrome, and various drug crystals such as aciclovir in drug-associated renal failure (Fogazzi et al., 1998). Studies of crystalluria for establishing the diagnosis in nephrolithiasis or monitoring treatment have yielded variable results because of the overlap with the normal condition (Fogazzi, 1996). Both high and low urine pH and a low temperature promote crystal formation (Fogazzi, 1996; Fogazzi et al., 2008a). Acid urine is associated with calcium oxalate, uric acid, and amorphous urate crystals. Alkaline urine is associated with calcium phosphate, triple phosphate, and amorphous phosphates (Fogazzi et al., 1998). Examination of urine for crystals should be performed on urine as close to 37ºC as possible. If amorphous crystals are present, acidification or alkalinization of the urine may be helpful to dissolve them, clearing the field of view to observe other urine sediment findings (Fogazzi and Cameron, 1996). Calcium oxalate and uric acid crystals are soluble in alkaline solutions, whereas triple phosphate and calcium phosphate crystals are soluble in acidic solutions (Fogazzi, 1996).

Fig. 6.1  Monohydrated (biconvex disc) and bihydrated (bipyramidal) calcium oxalate crystals.

may appear as joined semicircles with central indentation known as the ‘dumbbell’ form (Martinez-Giron, 2008). They may be seen in a broad pH range of < 5.4–6.7. Factors favouring formation include low pH, concentrated urine, and high oxalate ingestion or absorption.

Uric acid (Fig. 6.2) Uric acid may form several types of crystals in patients with uric acid nephrolithiasis, hyperuricosuria, and tumour lysis syndrome.

Calcium oxalate (Fig. 6.1) Calcium oxalate crystals may be present in normal urine but are frequently seen in patients with nephrolithiasis, on a high-oxalate diet, metabolic defects, vitamin C in high doses, and ethylene glycol poisoning. Calcium oxalate dihydrate usually forms square or rectangular shapes with characteristic diagonal highlights forming a bi-pyramidal shape known as the ‘envelope’ form. Ovoid shapes are also seen (Fogazzi, 1996). Calcium oxalate monohydrate crystals

Fig. 6.2  Uric acid crystals (lozenge variety). Inset: the same under polarized light.

37

38

Section 1  

assessment of the patient with renal disease

Flat diamond or rhomboidal ‘lemon’-shaped sheets are most common. Polarization reveals beautiful multicolour birefringence. Amorphous urates appear as granular debris in urine of low pH, < 5.4–5.8 (Fogazzi, 1996).

initiating appropriate therapy if renal failure is present (Perazella, 2003; Martinez-Giron, 2008).

Phosphates

Red blood cells (Figs 6.3a and 6.3b)

Phosphate crystals include calcium phosphate, triple phosphate, and amorphous phosphates. Triple phosphate crystals are rectangular prism-shaped crystals known as ‘coffin lids’. Calcium phosphate crystals may also appear as stars or needles. Amorphous phosphates appear as fine granular clumps.

Miscellaneous crystals Miscellaneous crystals include cholesterol crystals which are rectangular plates and cystine crystals which are hexagonal. Tyrosine and leucine crystals may be observed in liver disease.

Crystals in renal failure Tubular damage and obstruction by crystals may cause renal failure. Examples are acute uric acid nephropathy, calcium oxalate in ethylene glycol poisoning, and various drugs including aciclovir, triamterene, indinavir, methotrexate, and sulphonamides (Fogozzi, 1996; Perazella, 1999; Ting et al., 2000). Drug crystals should be suspected in subjects receiving a high-dose, intravenous infusion, and those with volume depletion or renal failure. Pleomorphic and atypical shapes may suggest drug-induced crystals. Commonly observed drug crystals include aciclovir (needles), protease inhibitors such as indinavir (plates and stars), and amoxicillin (needles, sheaves of wheat) (Fogazzi et al., 2008a). The presence of crystals in the urine is not diagnostic of crystal-induced renal failure but should prompt the clinician to consider altering therapy, removing the offending drug, or

Cells RBCs appear as translucent biconcave disks of between 4 and 7 μm in diameter. RBCs may be isomorphic (similar to circulating RBCs) or dysmorphic (abnormal in shape (poikilocytosis) or size (anisocytosis)) (Thal et al., 1986). Isomorphic RBCs are associated with non-glomerular haematuria from a genitourinary or external source. Urinary catheterization is a common cause in the hospitalized patient (Hockberger et al., 1987). RBCs may appear dysmorphic as a result of changes in pH, osmolality, and protein concentration (Wandel and Kohler, 1998). For example, shape transformation from discocyte (normal) to echinocyte (crenated) to stomatocyte (indented ball) by altering urine chemistry is rapid and reversible (Wandel and Kohler, 1998). These changes may not indicate renal disease. Forms of dysmorphia associated with renal disease include anulocytes, ghost cells, schizocytes, codocytes, and knizocytes (Kohler et al., 1991). The cause of dysmorphic RBCs is unknown but might arise from passage through the glomerular barrier or tubules (Schuetz et al., 1985; Kubota et al., 1988). Dysmorphic RBCs are associated with glomerular bleeding but may also be seen in non-glomerular and tubulointerstitial disease (Birch and Fairley, 1979; Fairley and Birch, 1982; Schramek et  al., 1989; Kohler et  al., 1991). A  review of five studies from 1982 to 1984 showed that dysmorphic RBCs were seen in 95% of patients with a glomerular lesion and only 8% of patients with a non-glomerular lesion (Thal et al., 1986). However, other data in patients with haematuria and proteinuria showed lower sensitivity

(a) (b) Fig. 6.3  (a) Isomorphic erythrocytes (dark cells have lost their haemoglobin content). (b) Different types of dysmorphic erythrocytes.

chapter 6 

and specificity (Favaro et al., 1997). Clinical application is hampered by the lack of a standard definition of dysmorphism (Thal et al., 1986). Acanthocytes (also known as G1 cells) are a unique form of dysmorphic RBC characterized by a round shape with one or two smaller, round protrusions, vesicles, or blebs attached (Kohler et al., 1991; Kitamoto et al., 1993; Lettgen et al., 1995). Crenated RBCs (echinocytes) which do not signify pathology have numerous, symmetric spiculations or spikes and may be seen in concentrated urine. They may be confused with acanthocytes but the appendages are smaller and more numerous. Although there is no standard cut-off for the number of dysmorphic RBCs or acanthocytes required to diagnose glomerular haematuria, the presence of acanthocyturia, defined as a urine containing ≥ 5% acanthocytes, may be useful. In one study acanthocyturia showed 98% specificity and 52% sensitivity for a glomerular lesion on biopsy. The sensitivity increased to 84% if four urine samples were examined (Kohler et  al., 1991; Kohler and Wandel, 1993). In patients with isolated microscopic haematuria, the presence of dysmorphic RBCs and acanthocytes is helpful. In a review of 16 patients (10 children and 6 adults) the sensitivity and specificity for the presence of ≥ 40% dysmorphic RBCs was 59% and 91% respectively for glomerular disease. Acanthocyturia ≥ 5% was 69% and 85% sensitive and specific (Fogazzi et al., 2008b). Acanthocyturia is particularly useful in identifying non-diabetic glomerular disease in diabetics where haematuria is common. In 68 patients with biopsy-proven diabetic nephropathy, haematuria was present in 62% but acanthocyturia was present in only 4%. In contrast, acanthocyturia was present in 40% of those with a superadded glomerulonephritis (Heine et al., 2004). A reasonable definition of glomerular haematuria is ≥ 40% dysmorphic RBCs or ≥ 5% acanthocytes (Fogazzi et al., 2008a).

Leucocytes (Fig. 6.4) Leucocyturia is a non-specific finding requiring clinical interpretation. Leucocytes may be seen in urinary tract infection, glomerulonephritis, tubulointerstitial nephritis, and renal transplant rejection (Fogazzi et al., 1998). Sterile pyuria, or leucocytes in the absence of bacterial urinary tract infection, is characteristic of renal tuberculosis but may also be seen in nephrolithiasis, urinary tract malignancy, analgesic nephropathy, and treated infections. Most nephrologists do not attempt to differentiate leucocytes in unstained urine. Neutrophils are the most common leucocyte observed in urine. They have a diameter of 7–13  μm with a granular cytoplasm and lobulated nucleus (Fogazzi et al., 2008a). Eosinophils may be identified by Wright’s or Hansel’s stain (Nolan et al., 1986; Nolan and Kelleher, 1988). Eosinophils have been reported in association with acute interstitial nephritis (AIN), cholesterol emboli, and glomerulonephritis (Nolan and Kelleher, 1988). Some authors have questioned the utility of looking for urine eosinophils in AIN given the low specificity (Nolan and Kelleher, 1988; Fogozzi et  al., 1998, 2008a). Macrophages are large cells 15 to > 100  μm in diameter and may be seen in various glomerular diseases or as oval fat bodies containing phagocytized lipids (Fogazzi et al., 2008a). Lymphocytes are seen with cellular rejection in renal transplant patients but require special staining (Fogazzi et al., 2008a).

urinalysis

Fig. 6.4  Neutrophils with their lobulated nucleus and granular cytoplasm (phasecontrast microscopy, original magnification 400).

Squamous epithelial cells Squamous epithelial cells are large, flat, 45–65 μm in diameter, and with a single small nucleus. They may come from the bladder trigone, urethra, or genital skin. They are frequently observed, especially in women, and do not indicate pathology. When observed with bacteria and white cells, genital contamination of the urine should be suspected.

Renal tubular cells RTE cells are 11–15 μm in diameter and may be round or rectangular with a large nucleus (Fogazzi et al., 2008a). They are usually seen in severe renal disease such as ATN, glomerulonephritis, AIN, or cellular rejection of renal transplants (Mandal et al., 1985; Fogazzi et al., 1998). The presence of renal tubular cells in the absence of other pathological urinary findings is rare (Fogazzi et al., 1998). In nephrotic patients, RTE cells may contain fat droplets and appear as ‘oval fat bodies’ similar to macrophages.

Uroepithelial or transitional cells Uroepithelial cells come from the urothelial lining of the urinary tract from the bladder, ureter, or renal pelvis. They may be round, oval, or rectangular. Cells from the superficial layers are 20–40 μm in diameter with a large, clear cytoplasm while those from the deeper layers tend to be smaller, 13–20 μm in diameter, oval, with a small and granular cytoplasm (Fogazzi et al., 2008a). Transitional cells from the deeper layer are strongly associated with urologic disease including stones, cancer, and hydronephrosis (Fogazzi et al., 1995, 1998, 2008a).

Microorganisms and miscellaneous cells Bacteria are common and, of course, require culture for identification and antibiotic sensitivities. Yeasts are clear circular cells, smaller than

39

40

Section 1  

assessment of the patient with renal disease

RBCs, sometimes identified by their budding or dividing. The hyphal forms of yeast form long, branching tubular structures similar to tree branches. Trichomonas vaginalis may be observed as a flagellated, motile, single-cell protozoan (Mutter and Brown, 2011). Schistosoma haematobium eggs may be present and associated with non-glomerular haematuria (Fogozzi et  al., 2008a). Non-genitourinary cells need to be considered, including intestinal epithelia in patients with ileal bladders (Gai et al., 2007). Spermatozoa are also commonly seen in subjects with retrograde ejaculation. Unusual cells that are large or multinucleated might indicate a malignancy.

Oval fat bodies Lipid may be present in the urine and is common with heavy proteinuria as in nephrotic syndrome and diabetic nephropathy (Fairley and Birch, 1993). Lipid droplets are small, about 1–5  μm in diameter. When incorporated into tubular cells or macrophages they are referred to as ‘oval fat bodies’ (Fairley and Birch, 1993).

Casts Casts are linear tubular structures that form in the renal tubules, usually the loop of Henle, distal tubules, or collecting ducts. The presence of casts in the urine is defined as cylinduria. Tamm–Horsfall mucoprotein provides the scaffolding and cells present in kidney or leaked into the urine during filtration become incorporated to form a cast (Rutecki et al., 1971; Hoyer and Seiler, 1979). The presence of certain casts is strongly correlated with the presence of renal disease identified by histology. The number of casts may predict the severity of injury (Gyory et al., 1984). There are seven major types of casts: hyaline, epithelial, granular, fatty, broad, waxy, and red or white blood cells. Unusual casts such as bilirubin casts in liver failure and bacterial casts in pyelonephritis have also been observed (Lindner et al., 1980). In renal conditions that wax and wane over time such as lupus nephritis, vasculitis, and cryoglobin-associated renal disease, the presence of casts is used to monitor disease activity (Fogazzi and Leong, 1996; Fogazzi et al., 1998). For example, RBC and WBC casts may precede clinical relapse in lupus nephritis (Hebert et al., 1995).

Fig. 6.5  A hyaline cast with a ‘fluffy’ appearance due to the fibrillary substructure of Tamm–Horsfall glycoprotein.

Hyaline casts (Fig. 6.5) Hyaline casts are concretions of Tamm–Horsfall protein and may be seen in normal urine. Other proteins such as immunoglobulins, fibrin, and complement may also be present (Fairley et al., 1983). Numerous hyaline casts are seen in volume contraction, congestive heart failure, and also seen after exercise or administration of diuretics (Imhof et al., 1972).

Epithelial cell casts Epithelial cell casts may be seen in both ATN and glomerulonephritis.

Granular casts (Fig. 6.6) Granular casts consist of degenerated tubular cells and strongly suggest the presence of renal pathology (Rutecki et  al., 1971). Coarse granular casts contain partially degenerated cells while fine granular casts result from further cellular degeneration. Muddy brown casts are pigmented granular casts observed in ATN. Small

Fig. 6.6  A finely granular cast.

numbers of granular casts may be seen in any form of chronic kidney disease.

Fatty casts Fatty casts may be granular or hyaline in appearance with small circular lipid droplets. Polarization reveals characteristic Maltese crosses. Fatty casts may also contain epithelial cells.

chapter 6 

Broad casts Broad casts are wider than other varieties and are believed to form in the larger collecting tubules. They may appear granular, waxy, or epithelial.

Waxy casts Waxy casts are broad, well demarcated, with flat sides and ends resembling melted wax. Sometimes they may appear in a corkscrew form. They are thought to be formed from disintegrated tubular cells and are associated with advanced kidney disease.

Red blood cell casts (Fig. 6.7) RBC casts suggest intrarenal inflammation. They may be present in pyelonephritis, allergic interstitial nephritis, and particularly severe glomerulonephritis. They are relatively infrequent, appearing in 22–38% of patients with glomerulonephritis but when present are specific for inflammatory glomerular pathology (Fogazzi et al., 1998). They are particularly useful to distinguish a glomerular source of haematuria in those with isolated haematuria. In a review of 16 patients with isolated microscopic haematuria, RBC casts were 100% sensitive for glomerular disease (Fogazzi et al., 2008b).

Leucocyte/white blood cell casts WBC casts may be seen in pyelonephritis, interstitial nephritis, and glomerulonephritis. Other urinary findings help determine the clinical significance. The finding of WBC casts may be particularly helpful when AIN is suspected and there is no evidence of urinary tract infection.

Clinical interpretation Urine sediment findings need to be interpreted within the clinical context. A bland urinary sediment does not exclude the presence of

urinalysis

significant renal disease. It is a common finding in vascular disease and may be consistent with a recovering renal injury (Goorno et al., 1967; Eiser et al., 1979; Fogazzi et al., 1998).

Tubulointerstitial disease Patients with tubulointerstitial disease may have a bland urine sediment. AIN may be characterized by pyuria and WBC casts in the absence of urinary tract infection. However, in a case series of 21 patients with biopsy-proven AIN, only 14% showed WBC casts. Interestingly 29% had RBC casts. WBCs were present in 57% and almost 50% had haematuria (Fogazzi et al., 2012). Eosinophiluria is generally not sensitive or specific enough to be clinically useful (Nolan and Kelleher, 1988; Fogazzi et al., 2008a).

The sediment in nephrotic syndrome The nephrotic syndrome is associated with oval fat bodies, fatty casts, and cholesterol crystals (Moriggi et al., 1995). Patients with nephrotic syndrome caused by diseases such as membranous nephropathy and focal and segmental glomerulosclerosis may have haematuria but RBC casts are rarely seen (Fogazzi et al., 2008a). Changes in the sediment such as the appearance of RBCs might indicate a clinical change such as renal vein thrombosis. In diabetes and minimal change disease, there is usually little to find on microscopy in spite of heavy proteinuria.

The sediment in nephritic syndrome Patients with diseases such as lupus nephritis, immunoglobulin A  nephropathy, and post-infectious glomerulonephritis usually have nephritic sediment characterized by dysmorphic RBCs including acanthocytes and RBC casts. RBC casts are suggestive of glomerular inflammation but may be seen in non-inflammatory conditions such as malignant hypertension (Moriggi et  al., 1995), embolic renal infarction (Blakely et  al., 1994), and pre-eclampsia (Leduc et  al., 1991). Changes in the urine sediment including the number of RBCs and the number and character of casts are sometimes used as a marker of disease activity. WBC casts, RTEs, and leucocytes are common, especially in necrotizing vasculitis. The number of RBC casts has been shown to correlate with disease activity in granulomatosis with polyangiitis (formerly called Wegener’s granulomatosis) (Fujita et al., 1998).

Urinary findings in acute tubular necrosis

Fig. 6.7  An erythrocyte cast. Inset: a haemoglobin cast.

ATN is generally associated with granular casts often referred to as ‘muddy brown casts’ thought to result from degraded renal tubular cells. RTE cells and casts are often present (Schentag et al., 1979; Mandal et al., 1985; Marcussen et al., 1995). However, the urine sediment may be normal in ATN. A  scoring system based on the number of casts in the urine has been proposed as a marker of ATN severity but is not in standard use (Chawla et  al., 2008). Other urinary findings can suggest the aetiology of ATN such as uric acid crystals (urate nephropathy) and oxalate crystals (ethylene glycol poisoning) and drug crystals such as those formed by aciclovir (Fogazzi et  al., 1998). Myoglobin casts are seen in rhabdomyolysis, while bilirubin casts may be seen in renal failure associated with hyperbilirubinaemia (van Slambrouck et al., 2013).

41

42

Section 1  

assessment of the patient with renal disease

Urinary tract infection Common findings include foul odour and cloudy urine. Bacteria, leucocytes, isomorphic erythrocytes, superficial transitional cells, and leucocyte casts may be present (Fogazzi et al., 1998).

References Birch, D. F. and Fairley, K. F. (1979). Haematuria: glomerular or non-glomerular? Lancet, 2, 845–6. Blakely, P., Cosby, R. L., and McDonald, B. R. (1994). Nephritic urinary sediment in embolic renal infarction. Clin Nephrol, 42, 401–3. Brigden, M. L., Edgell, D., McPherson, M., et al. (1992). High incidence of significant urinary ascorbic acid concentrations in a west coast population—implications for routine urinalysis. Clin Chem, 38, 426–31. Chawla, L. S., Dommu, A., Berger, A., et al. (2008). Urinary sediment cast scoring index for acute kidney injury: a pilot study. Nephron Clin Pract, 110, c145–50. Eiser, A. R., Katz, S. M., and Swartz, C. (1979). Clinically occult diffuse proliferative lupus nephritis. An age-related phenomenon. Arch Intern Med, 139, 1022–5. European Confederation of Laboratory Medicine (2000). European urinalysis guidelines. Scand J Clin Lab Invest Suppl, 231, 1–86. Fairley, J. K., Owen, J. E., and Birch, D. F. (1983). Protein composition of urinary casts from healthy subjects and patients with glomerulonephritis. Br Med J (Clin Res Ed), 287, 1838–40. Fairley, K. F. and Birch, D. F. (1982). Hematuria: a simple method for identifying glomerular bleeding. Kidney Int, 21, 105–8. Fairley, K. F. and Birch, D. F. (1993). Microscopic urinalysis in glomerulonephritis. Kidney Int Suppl, 42, S9–12. Favaro, S., Bonfante, L., D’Angelo, A., et al. (1997). Is the red cell morphology really useful to detect the source of hematuria? Am J Nephrol, 17, 172–5. Fogazzi, G. B. (1996). Crystalluria: a neglected aspect of urinary sediment analysis. Nephrol Dial Transplant, 11, 379–87. Fogazzi, G. B. and Cameron, J.S. (1996). Urinary microscopy from the seventeenth century to the present day. Kidney Int, 50, 1058–68. Fogazzi, G. B., Carboni, N., Pruneri, G., et al. (1995). The cells of the deep layers of the urothelium in the urine sediment: an overlooked marker of severe diseases of the excretory urinary system. Nephrol Dial Transplant, 10, 1918–19. Fogazzi, G. B., Edefonti, A., Garigali, G., et al. (2008b). Urine erythrocyte morphology in patients with microscopic haematuria caused by a glomerulopathy. Pediatr Nephrol, 23, 1093–100. Fogazzi, G. B., Ferrari, B., Garigali, G., et al. (2012). Urinary sediment findings in acute interstitial nephritis. Am J Kidney Dis, 60, 330–2. Fogazzi, G. B. and Grignani, S. (1998). Urine microscopic analysis—an art abandoned by nephrologists? Nephrol Dial Transplant, 13, 2485–7. Fogazzi, G. B. and Grignani, S., and Colucci, P. (1998). Urinary microscopy as seen by nephrologists. Clin Chem Lab Med, 36, 919–24. Fogazzi, G. B. and Leong, S. O. (1996). The erythrocyte cast. Nephrol Dial Transplant, 11, 1649–52. Fogazzi, G. B. and Passerini, P. (1994). Urinary sediment—its use in recognizing complications of the nephrotic syndrome. Nephrol Dial Transplant, 9, 70–1. Fogazzi, G. B., Verdesca, S., and Garigali, G. (2008a). Urinalysis: core curriculum 2008. Am J Kidney Dis, 51, 1052–67. Fujita, T., Ohi, H., Endo, M., et al. (1998). Level of red blood cells in the urinary sediment reflects the degree of renal activity in Wegener’s granulomatosis. Clin Nephrol, 50, 284–8. Gai, M., Motta, D., Giunti, S., et al. (2007). Urinalysis: do not forget this type of cell in renal transplantation. J Nephrol, 20, 94–8. Goorno, W., Ashworth, C. T., and Carter, N.W. (1967). Acute glomerulonephritis with absence of abnormal urinary findings. Diagnosis by light and electron microscopy. Ann Intern Med, 66, 345–53.

Gyory, A. Z., Hadfield, C., and Lauer, C. S. (1984). Value of urine microscopy in predicting histological changes in the kidney: double blind comparison. Br Med J (Clin Res Ed), 288, 819–22. Hebert, L. A., Dillon, J. J., Middendorf, D. F., et al. (1995). Relationship between appearance of urinary red blood cell/white blood cell casts and the onset of renal relapse in systemic lupus erythematosus. Am J Kidney Dis, 26, 432–8. Heine, G. H., Sester, U., Girndt, M., et al. (2004). Acanthocytes in the urine: useful tool to differentiate diabetic nephropathy from glomerulonephritis? Diabetes Care, 27, 190–4. Hockberger, R. S., Schwartz, B., and Connor, J. (1987). Hematuria induced by urethral catheterization. Ann Emerg Med, 16, 550–2. Hoyer, J. R. and Seiler, M. W. (1979). Pathophysiology of Tamm-Horsfall protein. Kidney Int, 16, 279–89. Imhof, P. R., Hushak, J., Schumann, G., et al. (1972). Excretion of urinary casts after the administration of diuretics. Br Med J, 2, 199–202. Kitamoto, Y., Tomita, M., Akamine, M., et al. (1993). Differentiation of hematuria using a uniquely shaped red cell. Nephron, 64, 32–6. Kohler, H. and Wandel, E. (1993). Acanthocyturia detects glomerular bleeding. Nephrol Dial Transplant, 8, 879. Kohler, H., Wandel, E., and Brunck, B. (1991). Acanthocyturia—a characteristic marker for glomerular bleeding. Kidney Int, 40, 115–20. Kouri, T. T., Gant, V. A., Fogazzi, G. B., et al. (2000). Towards European urinalysis guidelines. Introduction of a project under European Confederation of Laboratory Medicine. Clin Chim Acta, 297, 305–11. Kubota, H., Yamabe, H., Ozawa, K., et al. (1988). Mechanism of urinary erythrocyte deformity in patients with glomerular disease. Nephron, 48, 338–9. Leduc, L., Lederer, E., Lee, W., et al. (1991). Urinary sediment changes in severe preeclampsia. Obstet Gynecol, 77, 186–9. Lettgen, B. and Wohlmuth, A. (1995). Validity of G1-cells in the differentiation between glomerular and non-glomerular haematuria in children. Pediatr Nephrol, 9, 435–7. Lindner, L. E., Jones, R. N., and Haber, M. H. (1980). A specific urinary cast in acute pyelonephritis. Am J Clin Pathol, 73, 809–11. Mandal, A. K., Sklar, A. H., and Hudson, J. B. (1985). Transmission electron microscopy of urinary sediment in human acute renal failure. Kidney Int, 28, 58–63. Marcussen, N., Schumann, J., Campbell, P., et al. (1995). Cytodiagnostic urinalysis is very useful in the differential diagnosis of acute renal failure and can predict the severity. Ren Fail, 17, 721–9. Martinez-Giron, R. (2008). Crystal-like structure in urine sediment. Diagn Cytopathol, 36, 252. Moriggi, M., Vendramin, G., Borghi, M., et al. (1995). Nephritic urinary sediment: not only in proliferative glomerulonephritis but also in malignant hypertension. Nephron, 70, 131. Mutter, W. P. and Brown, R. S. (2011). Point-of-care photomicroscopy of urine. N Engl J Med, 364, 1880–1. Nolan, C. R., 3rd, Anger, M. S., and Kelleher, S. P. (1986). Eosinophiluria—a new method of detection and definition of the clinical spectrum. N Engl J Med, 315, 1516–19. Nolan, C. R., 3rd and Kelleher, S. P. (1988). Eosinophiluria. Clin Lab Med, 8, 555–65. Perazella, M. A. (1999). Crystal-induced acute renal failure. Am J Med, 106, 459–65. Perazella, M. A. (2003). Drug-induced renal failure: update on new medications and unique mechanisms of nephrotoxicity. Am J Med Sci, 325, 349–62. Pugia, M. J., Lott, J. A., Profitt, J. A., et al. (1999). High-sensitivity dye binding assay for albumin in urine. J Clin Lab Anal, 13, 180–7. Rutecki, G. J., Goldsmith, C., and Schreiner, G. E. (1971). Characterization of proteins in urinary casts. Fluorescent-antibody identification of Tamm-Horsfall mucoprotein in matrix and serum proteins in granules. N Engl J Med, 284, 1049–52.

chapter 6 

Schentag, J. J., Gengo, F. M., Plaut, M. E., et al. (1979). Urinary casts as an indicator of renal tubular damage in patients receiving aminoglycosides. Antimicrob Agents Chemother, 16, 468–74. Schramek, P., Schuster, F. X., Georgopoulos, M., et al. (1989). Value of urinary erythrocyte morphology in assessment of symptomless microhaematuria. Lancet, 2, 1316–19. Schuetz, E., Schaefer, R. M., Heidbreder, E., et al. (1985). Effect of diuresis on urinary erythrocyte morphology in glomerulonephritis. Klin Wochenschr, 63, 575–7. Siemens (2010). Multistix and Labstix Reagent Strips Product Guide. Tarytown, NY: Siemens Healthcare Diagnostics. Szwed, J. J. and Schaust, C. (1982). The importance of microscopic examination of the urinary sediment. Am J Med Technol, 48, 141–3. Thal, S. M., DeBellis, C. C., Iverson, S. A., et al. (1986). Comparison of dysmorphic erythrocytes with other urinary sediment parameters of renal bleeding. Am J Clin Pathol, 86, 784–7.

urinalysis

Ting, S. M., Ching, I., Nair, H., et al. (2009). Early and late presentations of ethylene glycol poisoning. Am J Kidney Dis, 53, 1091–7. Tsai, J. J., Yeun, J. Y., Kumar, V. A., et al. (2005). Comparison and interpretation of urinalysis performed by a nephrologist versus a hospital-based clinical laboratory. Am J Kidney Dis, 46, 820–9. Van Slambrouck, C. M., Salem, F., Meehan, S. M., et al. (2013). Bile cast nephropathy is a common pathologic finding for kidney injury associated with severe liver dysfunction. Kidney Int, 84, 192–7. Wald, R., Bell, C. M., Nisenbaum, R., et al. (2009). Interobserver reliability of urine sediment interpretation. Clin J Am Soc Nephrol, 4, 567–71. Wandel, E. and Kohler, H. (1998). Acanthocytes in urinary sediment—a pathognomonic marker? Nephrol Dial Transplant, 13, 206–7.

43

CHAPTER 7

Assessment of renal function Marijn Speeckaert and Joris Delanghe Estimation of glomerular filtration rate Exogenous compounds Assessment of kidney function in the clinical laboratory is usually based on the measurement of the clearance of various substances by the kidneys. The renal clearance of a substance is defined as ‘the volume of plasma from which the substance is completely cleared by the kidneys per unit time’. The clearance of a substance S is given by: Cs = (Us × V)/Ps. In most circumstances, an estimation of the glomerular filtration rate (eGFR) based on formulas incorporating serum creatinine is sufficient for clinical decision-making. However, for patients in whom GFR estimates based on serum creatinine are likely to be inaccurate or may have adverse consequences, measured GFR (mGFR) may be helpful. In the following sections we describe clinical situations in general medicine and nephrology where measurement of GFR should be considered (Stevens and Levey, 2009a).

Exogenous markers GFR can be measured as the clearance of exogenous or endogenous filtration markers. Non-radioactive compounds used as exogenous markers to measure GFR include iohexol, inulin, iodoacetate, and diethylenetriaminepentaacetic acid (DTPA). Clearances based on these exogenous markers are very accurate but are expensive and rather impractical and therefore are mainly restricted to research use. (See Table 7.1.)

Inulin The reference method to determine GFR is the urinary clearance of the fructose polymer inulin during a continuous intravenous infusion. Alternatively, the plasma clearance of inulin can be determined, which does not require urine collection (Ferguson et al., 1950). 51Cr-EDTA

The chromium-51 ethylenediaminetetraacetic acid (51Cr-EDTA) marker is not commercially available in the United States, but there is an extensive European experience with this marker. The urinary clearance of 51Cr-EDTA consistently underestimates inulin clearance by 5–15% in most, although not all studies, suggesting a degree of tubular reabsorption (BrochnerMortensen, 1978).

DTPA DTPA, an analogue of EDTA is usually labelled with technetium-99m (99mTc). Advantages include a short half-life (6 hours), minimizing

radiation exposure, and a high counting efficiency of 99mTc. DTPA is thought to be freely filtered at the glomerulus, with minimal tubular reabsorption, but it may undergo extrarenal elimination. Its major limitation is the potential of dissociation of 99mTc from DTPA and binding to plasma proteins, leading to underestimation of the GFR. The extent of dissociation is not predictable, leading to imprecision and bias. In addition, chelating kits and Tc generators are not standardized, making comparisons of mGFR with DTPA 99mTc among different institutions difficult (Stevens and Levey, 2009b). The magnetic resonance imaging contrast agents, gadolinium-DTPA or gadolinium-tetraazacyclododecane tetraacetic acid (gadolinium-DOTA), have recently been introduced as novel exogenous filtration markers because of their wide availability and their low rate of allergic reactions (Schuhmann-Giampieri and Krestin, 1991). In addition, a highly sensitive, novel, immunoassay technique is easily performed in most clinical laboratories and needs very low doses (1/40th of dose used for contrast) (BioPAL, Inc., 2007). The problem of systemic nephrogenic fibrosis has diminished enthusiasm for using this agent in subjects with significantly reduced GFRs, but because of the very low dose needed this complication is very unlikely to occur (Swaminathan and Shah, 2007).

Iohexol and iothalamate Iohexol and iothalamate are radiographic contrast agents and can be used as exogenous GFR markers comparable to inulin and 51Cr-EDTA (Schwartz and Furth, 2007). Concern about radiation led to the use of the non-radioactive radiographic contrast agent, iohexol (Omnipaque®) (Schwartz et al., 2006). Following iohexol administration, its content in collected plasma and urine samples is measured by high-performance liquid chromatography (HPLC). A  stable molecule such as iohexol would enable quality assessment schemes to be developed for GFR analysis. Iohexol accurately measures GFR with a four-point plasma disappearance curve (10, 30, 120, and 300 minutes) or, in most cases, with a two-point disappearance time (120 and 300 minutes) (Brandt et al., 2006). Iothalamate is commonly administered as a radioactive iodine label, but can also be administered in its non-radioactive form and is measured with HPLC (Notghi et  al., 1986). In the radioactive form, it is most commonly administered as a bolus subcutaneous injection. Iodine-125 (125I)-iothalamate has been widely adopted for measurement of GFR. To block thyroidal uptake, cold iodine is administered at the time of 125I-iothalmate administration, thus precluding its use in people with known allergies to iodine. Most studies comparing urinary clearance of iothalamate to inulin

chapter 7 

Table 7.1  Strengths and limitations of some exogenous GFR markers (Stevens and Levey, 2009a) Marker

Advantages

Disadvantages

Inulin

Gold standard

Expensive

Cr-EDTA

Widely available in Europe

Probable tubular reabsorption

99Tc-DTPA

Widely available in the United States

Iohexol

Not radioactive

Possible tubular reabsorption or protein binding Expensive Not used in patients with allergies to iodine

Iothalamate

Inexpensive

Probable tubular secretion

131I-hippuran

30% protein binding

showed a small positive bias, probably because of tubular secretion of iothalamate.

Endogenous markers Creatinine and creatinine standardization Creatinine is by far the most commonly used biochemical marker (Fig. 7.1) of renal function. Creatinine originates from the non-enzymatic hydrolysis of creatine and phosphocreatine, two substances found almost exclusively in the muscle (Wyss and Kaddurah-Daouk, 2000). This reaction occurs at a constant rate. Creatinine is a small molecule (molecular mass of 113 Da) that does not bind plasma proteins and is mainly eliminated by the kidney in

assessment of renal function

individuals with normal renal function. Creatinine is freely filtered by the glomerulus and is not reabsorbed throughout the tubules. Therefore its clearance can be measured as an indicator of GFR. Its production rate is decreased in patients with hepatic diseases (Cocchetto et  al., 1983). Serum or plasma creatinine concentrations do not only depend on the GFR, but also on muscle mass and on dietary protein intake (Crim et al., 1975; Heymsfield et al., 1983; Lew and Bosch, 1991). Tubular secretion, intestinal exchange, and the method and analytical standards used are also important variables. In cases of severe renal dysfunction, the GFR based on creatinine clearance rate will be ‘overestimated’, because the active tubular secretion of creatinine will account for a larger fraction of the total creatinine cleared. Ketoacids, cimetidine, and trimethoprim reduce creatinine tubular secretion and therefore increase the accuracy of the GFR estimate, particularly in severe renal dysfunction. Creatinine is rather unstable and should be measured without delay. Serum or plasma creatinine concentrations only increase beyond the reference values when kidney function is reduced by > 50%. Below this threshold, there is the so-called creatinine blind range, in which a measured creatinine clearance may provide more diagnostic information. In order to convert from conventional to SI units, the following conversion factor can be used: 1 mg/dL corresponds to 88.5 µmol/L. Creatinine can be determined using various modifications of the Jaffe principle (alkaline picrate reaction) (Delanghe and Speeckaert, 2011), by enzymatic dry chemistry (Toffaletti et al., 1983)  or by enzymatic colorimetric methods. The most commonly used assay is based on the so-called Jaffe reaction (Jaffe, 1886). In his landmark paper, Jaffe discussed that the alkaline picrate reaction could also occur, be it to a much lesser extent with a number of organic compounds (e.g. acetone, glucose). (These compounds were later designated as pseudo-chromogens and are a source of unspecificity in the Jaffe reaction). In the earliest

Creatine + creatine phosphate pools (± 120 g)

c

c c

Anabolism

Serum creatinine

Catabolism Creatinine pool Glomerular filtration

Extrarenal creainine Tubular secretion Urinary creatinine load: ± 1700 mg/24 h × 1.73 m2

Fig. 7.1 

45

46

Section 1  

assessment of the patient with renal disease

methods, serum creatinine was assayed by reactions based on alkaline picrate after deproteinization or dialysis, which eliminated the pseudo-chromogen effect of proteins (Levey, 1990). Today, however, analysers use untreated serum or plasma, making creatinine assays using alkaline picrate reaction prone to the so-called protein error. On average, this effect produces a positive difference of 27  µmol/L creatinine in Jaffe assays (Wuyts et  al., 2003). Because urine contains relatively little or no protein, the protein error affects only creatinine determinations in serum or plasma resulting in underestimation of the creatinine clearance when creatinine methods affected by protein error are used. Since Jaffe only observed a complexation between picric acid and creatinine in alkaline environment and never described an analytical method, variation amongst Jaffe methodological ‘recipes’ is broad. Notwithstanding stricter regulations, between-laboratory variation of Jaffe-based methods has not decreased over the last decade, despite technical progress in laboratory automation (Delanghe et al., 2008). Despite the known limitations, methods based on the Jaffe reaction are still extensively used for measuring serum creatinine. Haemolysed sera induce falsely increased values and the Jaffe reaction is negatively interfered by bilirubin. The ‘true’ creatinine value can be determined enzymatically. So-called compensated creatinine methods allow similar results as obtained by enzymatic creatinine assays. These values are about 20% lower compared with the non-compensated creatinine values (Junge et al., 2004). Standardization of serum creatinine measurements is very important because of the central role of this biomarker for the calculation of creatinine clearance, and the use of creatinine values for estimation of GFR (Levey, 1990). Estimations of GFR can be obtained using equations that empirically combine all of the average effects from biological factors that affect serum creatinine concentrations in serum other than GFR (Stevens and Levey, 2004). Assays not calibrated in agreement with the kinetic alkaline picrate method, used in the eGFR formula, introduce a source of error into GFR estimates. Implementing traceability of serum creatinine assays to gas chromatography-isotope dilution mass spectrometry (GC-IDMS) or liquid chromatography-isotope dilution mass spectrometry (LC-IDMS) leads to changes in the clinical decision-making criteria currently used for serum creatinine concentrations and creatinine clearance since calibration to an IDMS reference produces a lowering of serum creatinine values by 10–30% for most methods. (See Tables 7.2 and 7.3.)

Creatinine in urine For measuring creatinine in urine, a timed collection is necessary for measuring creatinine clearance. In most cases, the preferred method is the 24-hour collected urine, compensating for the variation in creatinine excretion during the day. Reference ranges are 0.87–2.41 g/24 hours (7.7–21.3 mmol/24 hours) for men and 0.67–1.59 g/24 hours (5.9–14.0  mmol/24 hours) for women. In contrast to the serum values, reference values for urinary creatinine are relatively independent from the method used.

Creatinine: recent analytical issues The National Kidney Disease Education Program (NKDEP), the College of American Pathologists (CAP), and the National Institute for Standard and Technology (NIST) have collaborated

to prepare a serum-creatinine reference material (NIST 967) with demonstrated commutability with native clinical specimens in routine methods. These materials are value-assigned with the GC-IDMS and LC-IDMS reference measurement procedures (Dodder et al., 2007). Nearly all clinical laboratory methods are now expected to have calibration traceable to an IDMS reference measurement procedure. The MDRD Study equation used for estimating GFR in adults has been updated to use coefficients appropriate for such methods (Levey et al., 2007). Routine reporting of eGFR for adults based on an IDMS-traceable creatinine has been proposed as clinical standard for patient care (Kallner and Khatami 2008; Narva, 2009). Variability in creatinine results may, however, contribute to substantial uncertainty in estimating GFR (Delanghe et al., 2008). Implementing traceability of serum creatinine assays to GC- or LC-IDMS has led to changes in the clinical decision-making criteria currently used for serum creatinine concentrations and creatinine clearance (Delanghe et al., 2011). Completeness of a 24-hour urine collection can be estimated by determining the excreted amount of creatinine. In order to obtain reliable test results, precise instructions dealing with the precise scenario of the timed urine collection should be provided to the patient and the nursing staff. When creatinine clearance is measured in patients who have been administered 3  × 800 mg cimetidine (a blocker of tubular secretion of creatinine) the day preceding the blood sampling, the effect of tubular secretion can be corrected. The cimetidine protocol, with creatinine clearance derived from a 2-hour urine collection, allows estimating of glomerular filtration rate in a clinical setting (Hellerstein et al., 1998). However, it cannot be used on a wide scale.

Creatinine clearance A classical creatinine clearance determination requires a serum or plasma specimen and a timed urine collection. The clearance is considered to be the amount of liquid filtered out of the blood. It is based on a ‘normal’ adult body surface of 1.73 m2. A small quantity of creatinine that appears in the urine (7–10%) is due to tubular secretion. By consequence, creatinine clearance usually exceeds inulin GFR by a factor of 1.1–1.2 at clearances above 80–90 mL/ min. The adult reference range (values standardized to 1.73 m 2 body surface) is 66–143 mL/min for the enzymatic method and 71–151 mL/min for Jaffe-based compensated tests. From the age of 40 onwards, creatinine clearance drops by about 8.5 mL/min per decade. From 2009 onward, most commercially available creatinine assays have been standardized to the reference method IDMS. This standardization change led to a worldwide reduction of serum creatinine values (Ceriotti et al., 2008). In routine clinical practice, 24-hour urine collections are an important source of error which impairs reliable calculation of the creatinine clearance. Consequently, practical formulas have been developed, which allow an estimation of the creatinine clearance without timed urine collections. The classical creatinine clearance is calculated as follows: creatinine clearance = urine creatinine ( mg / dL ) × urine volume ( mL / 24 h ) plasma creatinine ( mg / dL ) × 1440



chapter 7 

assessment of renal function

Table 7.2  Paediatric reference intervals for creatinine concentrations in serum, calculated with a non-parametric (NP) method and using fractional polynomials (FP) (Schlebusch et al., 2002) NP percentile Age group

N

2.5th

97.5th

Cord sera

51

46

86

Preterm neonates 0–21 days

58

28

87

Term neonates 0–14 days

69

27

2 months to < 1 year

41

1 to < 3 years 3 to < 5 years

FP percentile Midpoint age

2.5th

97.5th

—

—

At birth

29

90

81

7 days

22

73

14

34

7 months

11

34

45

15

31

2 years

15

30

41

23

37

4 years

21

34

5 to < 7 years

43

25

42

6 years

26

40

7 to < 9 years

46

30

48

8 years

31

46

9 to < 11 years

47

28

57

10 years

35

53

11 to < 13 years

42

37

63

12 years

38

59

13 to < 15 years

38

40

72

14 years

41

65

Cord sera

51

0.52

0.97

—

—

Preterm neonates 0–21 days

58

0.32

0.98

At birth

0.33

1.01

Term neonates 0–14 days

69

0.31

0.92

7 days

0.26

0.81

2 months to < 1 year

41

0.16

0.39

7 months

0.15

0.37

1 to < 3 years

45

0.17

0.35

2 years

0.18

0.33

3 to < 5 years

41

0.26

0.42

4 years

0.24

0.39

5 to < 7 years

43

0.29

0.48

6 years

0.29

0.46

7 to < 9 years

46

0.34

0.55

8 years

0.34

0.53

9 to < 11 years

47

0.32

0.64

10 years

0.39

0.60

11 to < 13 years

42

0.42

0.71

12 years

0.43

0.67

13 to < 15 years

38

0.46

0.81

14 years

0.47

0.73

µmol/L

mg/dL

Table 7.3  Reference intervals for creatinine concentrations in serum in adults (ages 18–74 years) (Junge et al., 2004) Percentile (90% CI) N Men Women

120 120

2.5th

97.5th

µmol/L

64 (63–66)

104 (99–107)

mg/dL

0.72 (0.71–0.75)

1.18 (1.12–1.21)

µmol/L

49 (46–55)

90 (83–103)

mg/dL

0.55 (0.52–0.62)

1.02 (0.94–1.17)

Conversion factor: 1 mg/dL = 88.5 µmol/L.

creatinine clearance =

(140 − age) × body weight ( kg ) 72 × serum creatinine ( mg / dL )

In women: creatinine clearance =

0.85 × (140 − age ) × body weight ( kg ) 72 × serum creatinine ( mg / dL )

In children, the Schwartz formula is frequently used: creatinine clearance =

0.55 × length ( cm )

serum creatinine ( mg / dL )

In term newborns and infants during the first year of life: In adults, the Cockcroft and Gault formula is used for the estimation of the creatinine clearance, which only requires the serum creatinine value, patient’s age, and body weight:

creatinine clearance =

0.45 × length ( cm )

serum creatinine ( mg / dL )

47

48

Section 1  

assessment of the patient with renal disease

The mentioned coefficients of the Schwartz formula must not be used in combination with enzymatic or so-called compensated creatinine assays, because this will lead to a serious overestimation of the creatinine clearance. For enzymatic IDMS calibrated creatinine assays a coefficient value of 0.42 instead of 0.55 is used (Schwartz et al., 2009). In contrast, enzymatic and compensated Jaffe-methods are well adapted for calculations with the Cockcroft and Gault formula. The coefficients of the Cockcroft and Gault formula do not match non-compensated creatinine determinations.

Calculation of GFR Only when relative GFR loss exceeds 50%, will serum creatinine values exceed the upper reference range (so-called creatinine-blind range). The serum creatinine concentration also depends on age, gender, and muscle mass. The best parameter for the determination of renal function is the GFR. International scientific societies now recommend the application of a new formula for the estimation of GFR. This formula has been validated in the framework of the Modification of Diet in Renal Diseases Study Group Study (MDRD Study) (Levey et al., 1993). The original MDRD Study equation was based on six variables: age, sex, ethnicity, and serum levels of creatinine, urea, and albumin (Levey et al., 1999). Subsequently, a four-variable equation consisting of age, sex, ethnicity, and serum creatinine levels was proposed to simplify clinical use (Levey et al., 2000). This equation is now widely accepted, and many clinical laboratories are using it to report GFR estimates. Extensive evaluation of the MDRD Study equation shows good performance in populations with lower levels of GFR but variable performance in those with higher levels. Variability among clinical laboratories in calibration of serum creatinine assays introduces error in GFR estimates and may account in part for the poorer performance in this range. The MDRD Study equation has now been re-expressed for use with standardized serum creatinine assays, allowing GFR estimates to be reported by using standardized serum creatinine and overcoming this limitation to the current use of GFR estimating equations (Levey et al., 2009):

(

)

GFR mL / min / 1.73 m 2 = 175 (serum creatinine)

−1.154

×

(age) × (0.742 in women) × (1.21 in African Americans) −0.2 033



The calculation (using coefficients adapted to IDMS standardization) may only be used in combination with corresponding IDMS-traceable methods. The MDRD Study equation is superior to the Cockcroft and Gault formulation in predicting kidney function in most people (Poggio et al., 2005). In 2009, a new formula, named the Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) was also introduced (Levy et al., 2009), which provides a more accurate GFR estimation for patients with GFR values between 60 and 90 mL/ min. In this equation, the values of the constants of a, b, and c vary on the basis of race, sex, and serum creatinine. GFR = a × (serum creatinine / b ) × ( 0.993) c

age

The variable a takes on the following values on the basis of race and sex: for African Americans: women = 166; men = 163. For white/

other, the coefficients were respectively 144 for women and 141 for men. The variable b takes on the following values on the basis of sex: women = 0.7 and men = 0.9. The variable c takes on the following values on the basis of sex and creatinine measurement: for women: if serum creatinine ≤ 0.7 mg/dL = −0.329; if serum creatinine > 0.7 mg/dL = −1.209; for men: if serum creatinine ≤ 0.9 mg/ dL = −0.411; if serum creatinine > 0.9 mg/dL = −1.209. The investigators compared the abilities to identify patients with various degrees of kidney disease of the CKD-EPI formula with the MDRD formula. Although both formulas performed similarly well, the cases in which subjects were misclassified in terms of category of kidney disease, the CKD-EPI formula was more often correct (P < 0.001). In the separate subset of patients designated to validate the new formula, the CKD-EPI formula demonstrated less bias and improved precision and greater accuracy than the MDRD formula (P < 0.001), particularly for those patients with an estimated GFR > 60 mL/min. The median difference between measured and estimated GFR was 2.5 mL/min for the CKD-EPI formula and 5.5 mL/ min for the MDRD formula (P < 0.001). Because of the reduced error of overestimation of kidney function, the CKD-EPI equation yielded a lower estimated prevalence of kidney disease in the National Health and Nutrition Examination Survey (NHANES) dataset (You et al., 2011). Where the MDRD equation estimated the prevalence of kidney disease at 13.1%, the CKD-EPI formula estimated it slightly lower at a prevalence of 11.5%. This limited extent to which the equation overestimates GFR reflects the major advantage of this newer formula. The determination of the creatinine clearance is commonly used in routine diagnostics to estimate the GFR, when serum creatinine value is still within reference limits. The Cockcroft and Gault formula must not be used in patients with acute kidney insufficiency and/or an unstable renal function, very obese patients or in presence of pronounced oedema (overestimation of clearance). Multiple conditions are characterized by increased GFR and kidney weight such as diabetes, pregnancy, unilateral nephrectomy, and high protein intake. In these conditions, an increase of GFR due to increase of effective renal plasma flow is the primary event that over a period of days is followed by kidney growth (Bak et al., 2000). In a recent study, data from cross-sectional studies that compared two or more creatinine-based GFR estimating equations to a reference GFR measurement were analysed. Studies from North America, Europe, and Australia showed a better performance of the CKD-EPI equation at higher GFRs (approximately > 60 mL/ min per 1.73 m2). However at lower GFRs, the MDRD Study equation was superior. As the CKD-EPI and MDRD Study equations were developed in North American and European populations that mainly consisted of black people and white people, the equations had to be modified in Asian and African populations to improve their performance by adding or removing a ‘race/ethnicity’ coefficient. Those coefficients are not useful beyond the local population, possibly because of differences in GFR measurement methods or differences in populations in addition to race or ethnicity. In conclusion, neither the CKD-EPI nor the MDRD Study equation is optimal across all populations and GFR ranges. New equations should focus on other filtration markers instead of or in addition to creatinine (Earley et al., 2012).

GFR and drug dose calculation Estimation of the creatinine clearance is a key element in the calculation of the correct dose of many drugs which are

chapter 7 

characterized by a narrow therapeutic index and a renal elimination pathway (Nagao et al., 2005). The implementation of eGFR reporting and standardization of creatinine measurements has created some uncertainty and confusion concerning the assessment of kidney function for drug dosing adjustment (Narva, 2009). Pharmaceutical companies have used for many years the Cockcroft and Gault equation (Cockcroft and Gault, 1976) to estimate creatinine clearance as the basis for drug dose adjustments. There is no modified Cockcroft and Gault equation available for use with the IDMS-traceable creatinine results. Consequently, creatinine clearance estimated from the Cockcroft and Gault equation will be increased upon restandardization. The original paper by Cockcroft and Gault does not contain detailed information regarding the creatinine assay used. The Cockcroft and Gault formula estimates creatinine clearance, which is not synonymous for GFR, since creatinine clearance is partly influenced by a tubular secretion of the compound. On the other hand, MDRD and CKDP-EPI formulas provide an estimate for GFR. Because the MDRD Study equation is relatively new, eGFR data is not part of drug safety information or package inserts. There is a concern about dosing toxic drugs with narrow therapeutic indices (e.g. carboplatin). Some clinicians request back-calculation to a non-standardized creatinine which can then be used in the Cockcroft and Gault equation. Back-calculation appears to be unnecessary (Stevens et  al., 2009). Concordance for drug dosing recommendations based on measured GFR for cleared drugs cleared by the kidneys was best for MDRD (88%) compared to the Cockcroft and Gault equation adjusted for ideal body weight (CGIBW) (82%) and the unadjusted Cockcroft and Gault (85%). Of the three estimating methods, the Cockcroft and Gault equation was most likely to generate higher recommended drug dosages, and CGIBW was most likely to generate lower recommended drug dosages. The MDRD produced recommendations that were lower than Cockcroft and Gault in 9% of the study population and higher in 10% when the CGIBW was used. The use of the MDRD equation for drug dosing adjustment in adults is encouraged (Narva 2009). Implementation of the CKD-EPI formula (Levey et al., 2009) in place of MDRD does not have much clinical impact on this specific application since drug dose adjustment is rarely needed for patients with GFR > 60 mL/min. The National Kidney Disease Education Program (NKDEP) (National Kidney Disease Educational Program, 2011) has published an educational advisory on estimating kidney function for drug dosing purposes. The advisory encourages the use of MDRD or Cockcroft and Gault estimating equations and, when there is concern that estimated kidney function is not adequate for patient safety or when there is a distinct difference in recommended dose between the two methods, suggests consideration of measured creatinine clearance or assessment of GFR by the clearance of exogenous compounds. Special attention should be paid to the elderly. MDRD has only been validated for patients aged between 18 and 70 years. MDRD overestimates renal function as age increases. While the optimized Cockcroft and Gault equation underestimates renal function, this was of a smaller magnitude, consistent across age, and thus better suited for dose calculation in the elderly (Roberts et al., 2009). Therefore, it is important not to extend the use of MDRD to the age groups for which the formula has not been validated. The new CKD-EPI formula has been validated over a wider range of ages and seems to be more accurate, also in the elderly patients (Aucella et al., 2010; Michels et al., 2010).

assessment of renal function

GFR estimation in paediatrics Creatinine restandardization should be a major concern in paediatrics due to the lower reference ranges for serum creatinine in infants and children (Delanghe, 2008). For calculating GFR, this systematic positive bias has been greatly compensated by the overestimation attributable to tubular secretion of creatinine, which is relatively more important in children. Although some Jaffe manufacturers try to correct for protein error through the use of a fixed compensation factor for the protein content in adults (Wuyts et  al., 2003), this procedure overcorrects with paediatric samples because of children’s low serum protein concentrations. Children, particularly younger children, have lower albumin and immunoglobulin (IgG) plasma concentrations than adults. In consequence, use of restandardized Jaffe-type assays results in overcompensation when used in children or infants. This overcorrection produces inaccurate GFR estimates, especially in neonates and children. Compensating calibration in Jaffe assays to IDMS standards results in an underestimation of serum or plasma creatinine due to the lower reference values for total protein in younger children. The enzymatic methods manage to measure serum creatinine more correctly (Delanghe et al., 2008). The IFCC Scientific Division recommends that more specific creatinine methods be adopted (Panteghini and IFCC Scientific Division, 2008). In enzymatic creatinine methods, analytical non-specificity is largely eliminated. The lower enzymatic creatinine result (when the result has not been adjusted to Jaffe-like results) leads to a marked increase of creatinine clearance and GFR estimations because of the increased effect of tubular secretion on test results. The use of enzymatic assays emphasizes the relative proportion of tubular secretion of creatinine which makes serum or plasma creatinine less suitable as a GFR marker in paediatric medicine, despite the analytical improvement (Delanghe, 2008). For estimating GFR in children and infants, the Schwartz (Schwartz et al., 1976, 1984; Schwartz and Gauthier, 1985) and the Counahan–Barratt equations (Counahan et  al., 1976)  are used. Both provide GFR estimates based on a constant value multiplied by the child’s height divided by the serum creatinine concentration. The values (SI units) for the constant used in both equations differ considerably: k = 38 (Counahan et al., 1976) and k = 48.7 (Schwartz et al., 1984; Delanghe et al., 2008). Since these formulas were validated 35 years ago, reassessment of classical formulas for estimating creatinine clearance and GFR using modern creatinine assays is necessary. As a result of restandardization, the original Schwartz-formula overestimates GFR. In the revised equation (Schwartz et al., 2009), with height measured in cm, using the calculation of 0.413 × (height/serum creatinine) (conventional units) or 46.67  × (height/serum creatinine; µmol/L), provides a good approximation of the estimated GFR formula. A newly developed CKiD equation uses serum creatinine, urea, and cystatin C (Cys C), plus height and sex, to estimate GFR, yielding the equation:

(

)

GFR mL / min per 1.73 m2 = 39.1 height ( m ) / Scr ( mg / dL ) [30 / BUN mg / dL)  0.169 1.099  male

0.516

× 1.8 / Cys C ( mg / L )  height ( m ) / 1.4]0.188 . 

( )

0.294



The new equation gives a better agreement with measured GFR than the updated Schwartz equation. There is also the Lund–Malmö

49

50

Section 1  

assessment of the patient with renal disease

equation which was validated for both adults and children (Nyman et  al., 2008). These formulas are only valid in combination with enzymatic methods with calibration traceable to an IDMS reference procedure. However, as the majority of the manufacturers of in vitro diagnostics will prefer general solutions based on a recalibration setup at the upper reference limit of an adult population, there is a risk of overcompensation because of the differences in serum matrix. So far, the updated Schwartz formula is the only paediatric GFR-estimating equation with coefficients suitable for use with IDMS-traceable creatinine methods. The Counahan equation was found to have nearly the same performance as the updated Schwartz equation (Miller, 2009). Because of the difficulties in adapting creatinine assays to the new calibrators in a uniform way, Cys C offers a promising alternative for calculating GFR in children. Compared to serum creatinine, low-molecular-weight proteins have a better diagnostic sensitivity for detecting an impaired GFR in children (Keevil et al., 1998). Although some caveats must be taken into account for a correct interpretation of the results, (Manetti et al., 2005; Hari et al., 2007), the protein-based GFR calculations only require serum values. The progress in the standardization of these protein assays will enable the wide-scale use of these methods.

Epidemiological studies As a result of creatinine restandardization, measured and calculated creatinine clearance values will increase, and the corresponding reference interval will be different. In an attempt to reduce late referral and to improve the care of patients with chronic kidney disease (CKD), different organizations have issued guidelines on the timely referral of patients to a nephrologist (National Kidney Foundation, 2005). Most suggest referral of patients with a GFR < 60 mL/min/1.73 m2, and mandate referral if the GFR is < 30 mL/min/1.73 m2. KDIGO published updated guidelines in 2013 (Kidney Disease:  Improving Global Outcomes (KDIGO) CKD Work Group, 2013). Important differences in classifications were obtained when different correction formulae for creatinine (the basis for GFR calculation) were applied.

Urea Urea was the very first marker of renal dysfunction, and renal failure is still referred to as uraemia. Urea is a small molecule (molecular mass: 60 Da) produced from the metabolism of amino acids. Its daily production varies largely with protein intake and is also dependent on protein degradation in the intestines as observed during gastrointestinal bleeding (Cottini et al., 1973; Maroni et al., 1985; Chalasani et  al., 1997). In the United States, often BUN (blood urea expressed as nitrogen) is used as a parameter, which corresponds to the urea-N-value. Urea value (mg/dL) is obtained by multiplying the BUN value by 2.14. More than 90% of urea is excreted through the kidneys. Urea is filtered freely by the glomeruli; 40–60% of the filtrated urea moves passively out of the renal tubules a process depending on urine flow rate. When water reabsorption is increased in renal tubules, or when intravascular volume is depleted, increased urea reabsorption leads to high serum urea concentration. Protein intake strongly correlates with serum urea concentration. In catabolic states, the loss of endogenous protein nitrogen results in urea production. Rapid muscle breakdown can also lead to a temporary increase in the serum urea concentration.

Because of these effects of protein metabolism upon urea concentration, serum creatinine is a more specific marker for renal function. The BUN/creatinine ratio may be helpful in the discrimination between prerenal and postrenal azotaemia. The reference for the BUN/creatinine ratio is between 10:1 and 20:1. Lower values usually denote acute tubular necrosis, low protein intake, starvation, or severe liver disease. High ratios with normal creatinine levels may be observed when there is tissue breakdown, prerenal azotaemia, high protein intake, or after gastrointestinal haemorrhage. High ratios associated with elevated creatinine concentration are found in either postrenal obstruction or prerenal azotaemia. Because urea is the major nitrogen compound produced by protein catabolism it allows an estimation of the amount of catabolized protein (1g N ~ 6.25 g protein). The reference range for urea in serum or plasma is between 20 and 50 mg/dL (3.3–8.3 mmol/L). For converting conventional into SI units, the conversion factor:  1 mg/ dL = 0.1667 mmol/L can be used. Urea clearance is a poor indicator of GFR, as its production rate is dependent on several non-renal factors, including diet. In addition, the variable level of back-diffusion will influence both plasma and urine urea concentrations. Urine urea excretion allows a good estimate of the protein content in the diet: 1 g of protein corresponds with ± 300 mg (50 mmol) of urea synthesis. The reference range for a daily urine urea excretion is between 1500 and 2600 mg/dL (250–433  mmol/L). Ammonia contamination (e.g. in urinary tract infection caused by Proteus spp.) leads to falsely increased values.

Cystatin C Serum concentrations of low-molecular weight marker proteins are primarily determined by GFR. An ideal marker has to have a constant production rate and should not vary in its concentration in situations with an acute-phase reaction. Cys C seems to share these ‘ideal’ properties. Cys C is a cysteine-proteinase inhibitor with a low molecular mass (13,300 Da), and is produced at a constant rate by all nucleated cells. Its small size and high pI (9.2) enable it to be freely filtered at the glomerulus. In normal conditions, serum Cys C is almost completely filtered by the renal glomerulus and largely catabolized by proximal tubular cells. Since serum Cys C concentration is closely correlated with the GFR, its determination has been introduced as a marker of renal function (Delanghe 2008). The Cys C concentration in serum is therefore determined by the GFR and in contrast to creatinine, Cys C concentrations are not affected by muscle mass and these values can also be used in children (Table 7.4). Unlike creatinine, serum Cys C reflects renal function independent of age, gender, and body composition. In the routine clinical laboratory, Cys C can be assayed using immunonephelometric or immunoturbidimetric tests. As the interindividual variation of serum Cys C-concentration is lower than that of creatinine, this protein marker has a higher diagnostic sensitivity and specificity in particular in patients with a moderate reduction of the GFR (between 40 and 80 mL/min). Cys C also offers a promising alternative for calculating GFR in children, as only a determination in serum or plasma is required. Especially in the so-called blind range of creatinine or in patients with low muscular mass, Cys C proves to be a superior marker compared with serum creatinine. Formulas have been developed

chapter 7 

assessment of renal function

Beta-trace protein

Table 7.4  Cystatin C reference values Age group

(mg/L)

< 1 month

1.37–1.89

1–12 months

0.73–1.17

Adult men

0.50–0.96

Adult women

0.57–0.96

allowing reliable estimation of GFR based on Cys C (Bökenkamp et  al., 1998). GFR-estimating equations based on Cys C are currently limited to the laboratory method that was used to derive the equation. Thyroid dysfunction affects serum Cys C concentration, possibly influencing the production rate of the protein (Manetti et al., 2005). Cys C is upregulated in certain tumours. Also pharmacological factors (e.g. glucocorticoid treatment) may influence the concentrations of serum Cys C (Delanghe 2008). In 2010, the IFCC Working Group for the Standardisation of Cystatin C (WG-SCC), in collaboration with the Institute for Reference Materials and Measurements (IRMM), announced the availability of the new certified reference material ERMDA471/IFCC. The new certified reference material will allow a global standardization of this analyte (Grubb et al., 2010). Because of its low individual variability, Cys C has fewer inherent limitations as a screening test for detecting a declining GFR than serum creatinine. However, serum creatinine is probably still the better assay for following sequential changes in an individual with confirmed renal disease (Keevil et al., 1998). Serum creatinine concentrations are lower in malnourished children and lead to overestimation of GFR, while Cys C levels are unaffected. Agreement between the Schwartz formula and gold standard GFR shows considerable bias, with a mean difference of 10.8% and a trend towards overestimation of the GFR by the Schwartz formula with lower GFRs. Cys C-based GFR estimates show significantly less bias and serves as a better estimate for GFR in children. Recently, combined GFR estimates based on both creatinine and Cys C have been presented (Stevens and Levey, 2009b).

Beta-trace protein (BTP) is a low-molecular-weight glycoprotein with a molecular weight of 23–29 kDa, depending on the degree of glycosylation. BTP (prostaglandin D synthase) belongs to the lipocalin protein family. It transports small lipophilic substances. BTP can be measured using immunonephelometry. Recently BTP has been introduced for the measurement of kidney function in the creatinine-blind range. BTP, like Cys C, is independent of age and gender. The upper reference limits of serum BTP and Cys C are 1.01 and 1.20 mg/L, respectively. Above 2 years, the reference range of BTP is constant at 0.43–1.04 mg/L (Filler et  al., 2002). These data confirm the potential of BTP as an endogenous GFR marker. International standardization of BTP is still lacking.

Beta-2 microglobulin Beta-2 microglobulin (B2M; 11.3 kDa) has been advocated as a predictor of GFR (Tables 7.5 and 7.6), but its serum concentration can increase as an acute-phase reactant in disorders such as lupus nephritis, that clearly require adequate assessment of GFR. B2M has the disadvantage of being increased in patients with several malignancies, particularly lymphoproliferative disorders including multiple myeloma. Like Cys C and BTP, B2M has the advantages of age and muscle mass independence (Filler et al., 2002).

Markers of acute kidney injury Recently, a number of biomarkers have been proposed for the early diagnosis, predicting the prognosis and differential diagnosis of acute kidney injury (AKI) (see Chapter 222) (Fig. 7.2).

Neutrophil gelatinase-associated lipocalin Neutrophil gelatinase-associated lipocalin (NGAL) or lipocalin 2 (LCN2) is a 25-kDa protein covalently bound to matrix metalloproteinase-9 from neutrophils (Devarajan 2010). It serves as a natural immunity protein with antibacterial properties. NGAL is also expressed by other cells among which are kidney epithelial and tubular cells (Devarajan, 2010; Ichino et al., 2009). NGAL expression is induced in injured cells, and both serum and urine NGAL is a promising biomarker of AKI (Devarajan 2010).

Table 7.5  Various methods for assessing GFR using endogenous markers (Delanghe, 2008) Marker (molecular weight, kDa)

Standardization

Remark

Creatinine (0.131)

NIST SRM 967

Partially determined by tubular secretion, dependent on muscle mass

Conventional creatinine clearance

NIST SRM 967

Biased due to urine collection problems. Overestimation of GFR due to tubular secretion

Paediatric estimation formulas (e.g. Schwartz, Counahan)

No longer compatible with new NIST SRM 967 creatinine standard. Partly depends on body constitution. Large variation for constant factor reported

Cystatin C (13)

International standardization is in progress

Ciclosporin and corticosteroids may interfere; some tumours upregulate cystatin C. Results may be affected by thyroid dysfunction

Beta trace protein (± 26)

Not available

Limited data

Beta 2 microglobulin (11.3)

Not available

Increased in some inflammatory conditions and some malignancies

51

52

Section 1  

assessment of the patient with renal disease

Table 7.6  Formulas for assessing GFR using endogenous markers Formula

Calculation

Remark

Cockcroft–Gault

eCCr = (140 – age) × mass (kg) × (0.85 if female)/72 × serum creatinine (mg/dL)

Estimates creatinine clearance

CCr-corrected = CCr × 1.73/BSA

A corrected formula for body surface area (BSA)

eGFR = 186 × serum creatinine−1.154 × age−0.203 ×

May only be used in subjects aged 18–70. Quantitative expression not valid when GFR vales > 60 mL/min

MDRD

(1.212 if black) × (0.742 if female) CKD-EPI

eGFR = 141 × min (SCr/k,1)a × max (SCr/k,1)−1.209 × 0.993age × (1.018 if female) × 1.159 if black)

Estimates GFR

Schwartz

eGFR = 0.413 × (height/serum creatinine) (conventional units)

Used in children

Counahan–Barratt

eGFR = 0 − 43 height (cm)/serum creatinine (mg/100 mL).

Used in children

Lund–Malmö formula

eGFR = eX − 0.00587 × age + 0.00977 × LBM

Cystatin C based, valid over a broad age range

Urinary NGAL can be assayed using commercial immunoassays (Bennett et al., 2008; Grenier et al., 2010). The cut-off value for urinary NGAL is 100 micrograms/L (Hirsch et al., 2007; Bennett et al., 2008). In urinary tract infections, increased urinary NGAL concentrations have been observed (Yilmaz et al., 2009). As urinary neutrophils can be regarded as a potential source of urinary NGAL, leucocyturia contributes to urinary NGAL concentrations and affects thus the interpretation of urinary NGAL results (Decavele et al., 2011). In urine specimens with a high white blood cell (WBC) count, (e.g. intensive care patients) (Bagshaw et al., 2010; Cruz et al., 2010), the use of a fixed cut-off value for interpreting urinary NGAL data may therefore lead to false-positive results. On average, one leucocyte corresponds with 12 pg of urine NGAL. Although the contribution of neutrophil NGAL to urinary NGAL is generally small, the effect may become important if there is a high leucocyte count and urinary NGAL levels may exceed 100 micrograms/L. A mathematical correction compensating neutrophil NGAL contribution may increase the diagnostic efficiency of NGAL in presence of increased NGAL levels and high urinary WBC counts: Corrected NGAL ( micrograms / L ) = NGAL − 0.12 urinary WBC coun t, 109 cells / L .

(

)

In tubular dysfunction, tubular impairment may also contribute to elevated urine NGAL concentrations (Decavele et al., 2011). In patients with CKD, an inverse correlation with estimated GFR has been described for both serum NGAL and urinary NGAL, suggesting that under these particular conditions this protein may also represent a surrogate index of residual renal function (Bolignano et al., 2009).

Kidney injury molecule-1 In animal models, kidney injury molecule-1 (KIM-1) is one of the most highly induced proteins in the kidney after AKI, and a proteolytically processed ectodomain of KIM-1 is easily detected in the urine soon after AKI. Enzyme-linked immunosorbent assay (ELISA)-based assays for KIM-1 have been developed, but are not yet commercially available. In a small human cross-sectional study,

KIM-1 was markedly induced in proximal tubules from kidney biopsies in patients with established AKI, and urinary KIM-1 distinguished ischaemic AKI from prerenal azotaemia and chronic renal disease. In a case–control study of children undergoing cardiac surgery, urinary KIM-1 levels were increased in subjects who developed AKI, with an AUC of 0.83 at the 12-hour time point (Han et al., 2008). In a larger prospective cohort study of hospitalized patients with established AKI, urinary KIM-1 was associated with adverse clinical outcomes, including dialysis requirement and death (Liangos et al., 2007). An advantage of KIM-1 as a urinary biomarker is that, in contrast to NGAL, its expression seems to be limited to the injured or diseased kidney since no systemic source of KIM-1 has been described. KIM-1 is induced in the kidney and upregulated in the urine by a large number of nephrotoxins, including ciclosporin, cisplatin, gentamicin, and heavy metals.

Liver fatty acid binding protein Liver fatty acid binding protein (L-FABP) is a 14-kDa protein normally expressed in the proximal convoluted and straight tubules of the kidney, and upregulated in animal models of AKI. An ELISA for this analyte is commercially available. It was used to demonstrate that urinary L-FABP levels were significantly increased prior to the increase in serum creatinine in those patients who developed AKI post contrast dye. In a recent prospective study of children undergoing cardiac surgery, urine L-FABP increased at 4 hours post-bypass, with an AUC of 0.810 for a cut-off value of 486 ng/mg creatinine (Portilla et al., 2008). However, increased urinary L-FABP levels are also found in patients with non-diabetic CKD, early diabetic nephropathy, idiopathic focal glomerulosclerosis and polycystic kidney disease. Additionally, L-FABP is also abundantly expressed in the liver, and urinary L-FABP may be influenced by serum L-FABP levels.

Proteinuria/differentiation of proteinuria Proteinuria Excessive urinary protein excretion is a cardinal sign of renal disease and a pathogenetically important factor in the progression of

chapter 7 

assessment of renal function

Tubular proteins Alpha-1 microglobulin (A1M) Beta-2 microglobulin (B2M)

New biomarkers

Retinol binding protein (RBP)

Neutrophil gelatinase-associated lipocalin (NGAL)

Fatty acid binding protein (FABP)

Kidney injury molecule-1 (KIM-1) Cystatin C

Biomarkers for diagnosis of acute kidney injury

Delayed biomarkers for kidney injury

Renal lysosomal enzymes

Creatinine

N-Acetyl-β-D-glucosaminidase (NAG)

Cystatin C

Fig. 7.2  Biomarkers for early diagnosis, predicting prognosis and differential diagnosis of acute kidney injury (AKI).

renal and cardiovascular disease. Further analysis of the type of proteinuria provides information about the possible underlying renal disease (e.g. glomerulonephritis, interstitial nephropathy). In the step-by-step diagnostic work-up of renal disease, knowledge about the type of proteinuria (low molecular/high molecular mass) is useful. Proteinuria may be due to prerenal (overflow proteinuria), renal (glomerular, tubular), or postrenal (bleeding or infection of the kidneys, ureter, bladder, or urethra) causes. A protein excretion exceeding approximately 200 mg/24 hours (based on a daily urinary volume of 1.5 L) is considered clinically relevant. The reference range for urinary protein excretion is lower: < 150 mg in 24 hours. Under physiological conditions, urinary protein excretion does not exceed 150 mg/day for adults and 140 mg/m2 of body surface for children. The daily physiological proteinuria contains mucoprotein (e.g. Tamm–Horsfall glycoprotein, 70 mg), blood group-related substances (35 mg), albumin (16 mg), immunoglobulins (6 mg), mucopolysaccharides (16 mg), and very small amounts of other proteins such as hormones and enzymes (King and Boyce, 1963). There are four categories of pathological proteinuria:  ‘glomerular’, ‘tubular’, ‘overload’, and ‘benign’ (Abuelo, 1983). Apart from this, proteinuria can also be factitious as a presentation of Munchausen’s syndrome. Glomerular proteinuria is due to an increased glomerular permeability to proteins and occurs in primary and secondary glomerulopathies. It can be further classified into selective and non-selective proteinuria. Selective proteinuria is characterized by the presence of predominantly anionic proteins with a molecular weight of 50–80 kDa (e.g. albumin). In the case of non-selective proteinuria, larger proteins with a molecular weight > 50–80 kDa are detected, in addition to albumin. This is indicative for a more severe damage of the glomeruli. Tubular proteinuria is due to a decreased tubular reabsorption of proteins

normally present in the glomerular filtrate. It is seen in tubular and interstitial disorders, including those which develop in the course of chronic glomerular diseases. Overload proteinuria is secondary to an increased production, or release of low-molecular weight proteins, which are usually reabsorbed by the proximal tubular cells, such as immunoglobulin light chains (which are increased in monoclonal gammopathies), lysozyme (which is increased in some leukaemias), or myoglobin (which is increased in rhabdomyolysis). Benign proteinuria includes functional proteinuria, as seen in fever or after exercise, idiopathic transient proteinuria, and orthostatic proteinuria. Factitious proteinuria may originate from addition of exogenous proteins to urine specimens. This is sometimes observed in Münchausen syndrome (Boelaert et al., 1984), psychiatric disorders known as factitious disorders wherein those affected feign disease, illness, or psychological trauma in order to draw attention or sympathy to themselves. Once pathological proteinuria has been detected, it must be quantified and submitted to qualitative analysis to discover to which of the above four categories it belongs.

Detection Qualitative evaluation Dipsticks See also ‘Dry chemistry’ and ‘Albuminuria’. Although dipstick urinalysis is characterized by a low cost, a wide availability, and a short procedure time, to date there is no consensus among the guidelines regarding the use of dipsticks as a front-line screening test for proteinuria in asymptomatic individuals (National Kidney Foundation 2002; Konta et al., 2007; Graziani et al., 2009; Lamb et al., 2009). Recent studies show a high negative predictive value of dipstick testing and a minimal risk of falsely

53

54

Section 1  

assessment of the patient with renal disease

negative cases of macroalbuminuria, whereas positive results require further confirmatory testing (White et al., 2011).

Precipitation methods Precipitation methods evaluate the turbidity occurring after proteins are precipitated by sulphosalicylic acid, trichloracetic acid, or by heat and acetic acid-sodium acetate buffer. Turbidimetric methods detect all urinary proteins, but alpha-1-acid glycoprotein and Tamm–Horsfall mucoprotein are not precipitated by sulphosalicylic acid (Shiba et al., 1985). They are very sensitive and may detect protein concentrations as low as 2.5 mg/L. With these methods, the protein concentration is expressed on a semiquantitative scale, from 0 to +++ or ++++. Many drugs can cause positive interference, for example, large amounts of penicillin or cephalosporin analogues, miconazole, tolbutamide, or sulphonamide metabolites. False-positive results can also be caused by the presence in the urine of radiocontrast media.

Quantitative evaluation Urine proteins can be quantified by several methods (Table 7.7), but as urine can contain a large range of proteins and also many interfering substances, all available methods are imperfect. Benzethonium chloride (Iwata and Nishikaze, 1979), trichloroacetic acid precipitation (Henry et al., 1956), dye binding methods with Brome-phenol blue (Hemmingsen, 1972), Coomassie brilliant blue (Bradford, 1976), Ponceau red (Pesce and Strande, 1973) and pyrogallol-red, nephelometry, and turbidimetry have been suggested. All these methods can be automated, in contrast to the biuret examination (Weicheselbaum, 1946). Determination of total protein is a compromise because no procedure detects all the proteins in urine. The upper limits of reference intervals in healthy individuals are method-dependent. Several intervals are cited:  pyrogallol red method:  180 mg/day (Orsonneau et  al., 1989), for pregnant women: 260 mg/day (Higby et al., 1994); biuret method: 150 mg/ day; turbidimetric method:  75 mg/day (Henry et  al., 1956), turbidimetric method:  protein/creatinine 8 g/mol (second morning urine), or 70 mg/g in conventional units. Turbidimetric methods Although these methods are among the most frequently used (e.g. benzethonium chloride in alkaline medium), they suffer from major drawbacks. Trichloroacetic acid is unaffected by protein composition and can detect low protein concentrations, if the temperature of urine ranges between 20 and 25°C. However, numerous drugs can interfere with this method (Lorentz and Weiss 1986). The

method based on benzethonium chloride gives a stable turbidity and is less dependent on temperature than the previous methods. However, it produces 11–31% less turbidity for gammaglobulins than for albumin, and gives unevenly dispersed aggregates in the presence of large protein concentrations, with consequent falsely low protein estimates (McElderry et al., 1982). Dye-binding techniques These are based on the interaction between proteins and a dye, which causes a shift in the absorption maxima (measured photometrically) of the dye. Coomassie brilliant blue G250 (Bradford, 1976) is very sensitive, but with diluted samples significant deviations from linearity are possible. Moreover, underestimation of tubular proteinuria and interference from various metabolites, drugs, and preservative compounds has been observed. The method is somewhat improved by the addition of sodium dodecylsulphate. Ponceau S (Pesce and Strande, 1973) is equally sensitive to albumin and globulins, but falsely low values can result from the loss of precipitate during the decantation step while falsely high values may derive from the contamination of the precipitate by unbound dye. Aminoglycoside antibiotics interfere positively. Pyrogallol red-molybdate method can easily be implemented on automated analysers. This method has been improved by the addition of sodium dodecylsulphate (Orsonneau, 1989). Biuret methods These methods are based on the interaction between copper ions and the carbamide group of proteins (Lorentz and Weiss, 1986). They have the same sensitivity for all proteins. The protein in urine must be concentrated before the biuret reaction occurs. Interference from drugs, radiographic contrast media, and coloured metabolites is minimal (Bradford, 1976; Lorentz and Weiss, 1986). A modification of the biuret method, which is now the reference method recommended by the American Association for Clinical Chemistry, utilizes gel filtration to exclude small interfering compounds.

The protein:creatinine ratio The 24-hour urine collection is considered as the ‘gold standard’ for the quantitative evaluation of proteinuria. However, it is time-consuming, subject to error, and is frequently associated with significant collection errors, which are mainly due to improper timing and missed samples. The calculation of the protein:creatinine ratio on spot urine samples, which corrects for variations in urinary concentration due to hydration, is an alternative to the 24-hour

Table 7.7  Some quantitative methods for proteinuria Method Turbidimetric Trichloroacetic Benzethonium chloride Dye binding Coomassie brilliant blue Ponceau Biuret

Analytical sensitivity (mg/L)

Linearity (mg/L)

Distinctive features

20 10 2.5 20

20–2400 10–1600 5–1500 100–1600

Same sensitivity for albumin and globulins, many drugs can interfere Albumin overestimation, underestimation of increased protein concentrations High sensitivity, underestimation of tubular proteins Same sensitivity for albumin and globulins, positive interference with aminoglycoside antibiotics

chapter 7 

collection (Ginsberg et al., 1983). The measurements obtained with this method correlated well with those obtained in the classical way. All patients with proteinuria of > 3.5 g/24 hours had a ratio of greater than 3.5 in single voided samples, and all patients with a proteinuria of < 0.2 g/24 hours had a ratio of < 0.2. The applicability and reliability of this method has been confirmed by several investigators (Rodby et al., 1995; Ruggenenti et al., 1998), and is now recommended by the American Guidelines (National Kidney Foundation, 2002). The protein:creatinine ratio is preferably calculated on the first morning urine. This avoids the possible variations caused by the circadian rhythm of protein excretion, which is maximal during the day and minimal during the night, compared to the urinary excretion of creatinine, which is fairly constant over the day (Koopman et al., 1989).

Further exploration and selectivity of proteinuria In the case of a positive qualitative protein test, the cause of proteinuria should be investigated. Urine protein electrophoresis allows a differentiation of proteinuria. Electrophoretic techniques (Fig. 7.3) are essential in detecting the clonality of serum paraproteins (M components) and monoclonal immunoglobulin light chains in urine (Bence Jones proteins) (Merlini et al., 2001). Today immunofixation of urine or serum is the gold standard technique in defining the isotype of the myeloma clone. However, it can occasionally give misleading ‘pseudo-oligoclonal’ patterns in urine specimens that contain polyclonal immunoglobulins (Harrison et al., 1992). Some authors recommend immunofixation electrophoresis of all specimens because of the sensitivity of the method. As an alternative to screening and in monitoring of proteinuria, turbidimetric quantitation of urinary kappa or lambda chain excretion is also suggested, combined with immunofixation in case of a positive finding with unknown clonality. Another way to distinguish the different types of proteinuria is based on the measurement of specific proteins (e.g. albumin, transferrin, alpha-1-microglobulin (A1M), alpha-2-macroglobulin (A2M)). Electrophoresis provides useful information in problem cases (e.g. Von Münchausen syndrome or Münchausen by proxy syndrome—falsification of laboratory results by adding substances to urine).

1

2

3

4

5

6

Fig. 7.3  Various types of proteinuria: serum protein pattern for comparison (lane 1), selective (lane 2) and non-selective (lane 3) glomerular proteinuria, tubular proteinuria (lane 4), Bence Jones proteinuria (lanes 5 and 6).

assessment of renal function

Agarose-electrophoresis after protein concentration and the use of a very sensitive staining (e.g. gold stains) is one of the most widely used methods. Size fractionation of urinary proteins on sodium dodecyl sulphate-acrylamide gel electrophoresis (SDS-PAGE) has been used to classify prerenal, renal, and post-renal proteinuria before quantitative measurements of specific proteins became widely available. SDS-PAGE is a laborious procedure and is therefore rarely used in the routine laboratory. There is a tendency towards using specific urinary proteins as marker proteins. In overflow proteinuria, excessive amounts of small plasma proteins exceeding the reabsorption capacity of the proximal tubulus result in a protein excretion in the final urine (overflow proteinuria). Typical examples are haemoglobin monomers, for example, intravascular haemolysis and immunoglobulin light chains (Bence Jones proteins) in a monoclonal gammopathy. Postrenal proteinuria occurs when proteins enter the urine because of a haemorrhage or an inflammation within the lower urinary tract or genital tract. As a marker protein in such situation, A2M is mainly determined, as it cannot be filtered by the glomerulus because of its 725-kDa molecular mass. The selectivity of glomerular proteinuria is classically evaluated by the ratio of the clearance of IgG (molecular weight 150 kDa) to the clearance of transferrin (molecular weight 80 kDa). When the ratio is < 0.1 the proteinuria is defined as selective (Cameron and Blandford, 1966). In children with nephrotic syndrome, this strongly suggests the presence of minimal change disease responsive to corticosteroids. Other approaches are based on the ratio of the clearance of IgM to that of albumin (Bakoush et al., 2001).

Albuminuria The excretion of small amounts of albumin in urine was historically called ‘microalbuminuria’. According to the American Diabetes Association (2014), low-grade albuminuria corresponds to an albumin excretion rate of approximately 30–300 mg/24 hours. However, albuminuria is characterized by a large intraindividual variability. Low-grade albuminuria is an excellent marker for assessing vascular damage and early renal disease. Generalized atherosclerosis leads to medullary hypoperfusion and hypoxia, which induces renal tubular dysfunction. An impaired reabsorption of albumin and other tubular proteins is observed. Increased albumin excretion is regarded as the earliest marker for detecting changes in the glomerular basal membrane. At this stage, the glomerulopathy can still be modified by taking appropriate measures (strict metabolic control in diabetics, blood pressure reduction in hypertensives by association of angiotensin-converting enzyme inhibitors or angiotensin II-receptor blockers), preventing progression towards renal insufficiency. False-low results may occur in extreme proteinuria because of an antigen excess. Screening is not indicated in the presence of urinary tract infections, following strenuous physical exercise, uncontrolled diabetes mellitus or hypertension, cardiac insufficiency, fever, and menses. These factors may induce a temporary increase in albumin excretion. In contrast to cases of overt proteinuria or macroalbuminuria, 80% of the urinary proteins in normo- and microalbuminuria are of non-albumin nature. Further research is required to clarify the importance of specific non-albumin proteins in urine (Methven et al., 2011). The 24-hour urine collection has been considered the ‘gold standard’ method for the measurement of albuminuria due to

55

56

Section 1  

assessment of the patient with renal disease

the circadian rhythm of urinary protein excretion (Hansen et al., 2002). However, the urine albumin:creatinine ratio (ACR) on random urine samples can also be used (Assadi, 2002), as it has been demonstrated that total urinary protein and albumin excretion are equally powerful predictors of all-cause mortality and renal outcomes in patients with CKD attending a general nephrology clinic (Methven et al., 2011). As the urinary albumin concentration depends on the hydration status, the constant urinary excretion of creatinine ensures that the urinary ACR adjusts for hydration changes (Heerspink et al., 2010). Moreover, this ratio incorporates the predictive value of poor health as reflected by low muscle mass and decreased creatinine excretion (Nangaku, 2006). The most common methods to measure albuminuria are nephelometric (Stamp 1988) and turbidimetric assays, but those measurements are expensive, have a delay in the provision of results and require a specialist biochemistry laboratory (Shukla et  al., 1988; White et al., 2011). Urinary albumin measurements are currently not standardized due to a lack of a reference method and reference (primary and secondary (matrix)) material. Multiple molecular forms of albumin in urine are identified (Fig. 7.4). Modification of albumin by proteolysis during passage through the urinary tract and chemical modification during specimen storage leads to the formation of albumin fragments. Multiple methods have been developed to quantify albuminuria and significant different results are reported dependent on the available assay. The current point of view of the National Kidney Disease Education Program—IFCC Working Group on Standardization of Albumin considers the immunoassay with polyclonal sera as the primary method of quantifying urine albumin. At present there are no hard arguments for measuring immunochemically non-reactive albumin in urine.

Immunoassays using polyclonal antisera for the detection of urinary albumin remain the gold standard. The development of a reference measurement procedure remains one of the challenges for the future (Speeckaert et al., 2011). Test strips are also available, but the use of those strips as a screening tool is controversial due to the limited evaluation of the diagnostic accuracy of dipsticks for the general population. In a large Australian population-based cohort study, a dipstick (Bayer Multistix 10 SG) reading of ≥ 1+ identified adults with ACR ≥ 30 mg/g with high specificity and was characterized by a strong negative predictive value and a tendency toward a higher false-positive rate. There was a minimal risk of missed cases of macroalbuminuria (the identification of ACR ≥ 300 mg/g had a false-negative rate of < 0.1%). However, a large variation in positive predictive values was observed across low- and high-risk subgroups according to albuminuria prevalence. The diagnostic accuracy was not significantly influenced by the presence of diabetes, hypertension, or age. An inferior test sensitivity was reported in women, which may be explained by the use of creatinine to correct for concentration in the measurement of the reference standard (White et al., 2011). There may be an important difference between reagent strips produced by different manufacturers in detecting the same threshold of albuminuria or in calibration to indicate a positive test result (trace, 1+, or > 1+) (Lamb et al., 2009). As a significant heterogeneity in diagnostic test performance across different populations has been observed, evaluation of the test performance in a large population group remains essential (Konta et al., 2007; White et al., 2011). It is recommended that a positive strip test is followed by a quantitative determination (American Diabetes Association, 1994). New studies should focus on the impact of treatment interventions in individuals

Renal tubule Albumin

Albumin fragments

Cubilin Megalin Amnionless Proximal tubular cell

Early endosome Lysosome/ late endosome hybrid

Late endosome Reabsorption of albumin

Degradation of albumin Blood vessel

Fig. 7.4  Overview of the albumin fragmentation process. Derived from Speeckaert et al., (2011).

Lysosome

chapter 7 

with low-grade albuminuria and update the cost-effectiveness of dipsticks as a screening tool (White et al., 2011).

Tubular proteins Tubular proteinuria occurs when glomerular function is normal, but the proximal tubules have a diminished capacity to reabsorb and to catabolize proteins, causing an increased urinary excretion of the low molecular mass proteins that normally pass through the glomerulus, such as retinol-binding protein, A1M, and B2M. Since these products are present in the plasma in low concentrations, urine protein excretion in tubular proteinuria is usually < 1.5 g/day. Nephrotoxins that damage the proximal tubule include heavy metals (e.g. cadmium and lead) and some drugs (e.g. aminoglycosides, ciclosporin, cytostatic agents, and analgesics). Tubular proteinuria is also associated with acute and chronic pyelonephritis, renal vascular diseases, kidney transplant rejection, Fanconi syndrome, and Balkan nephropathy.

Alpha-1-microglobulin A1M is a 27-kDa glycoprotein produced by the liver and has been given other names, for example, complex-forming glycoprotein, heterogeneous in charge, protein HC (Penders and Delanghe, 2004). In plasma, approximately 50% of A1M forms a complex with IgA. Free, monomeric A1M passes through the glomerulus out into the primary urine, from which it is reabsorbed by the proximal tubules where it is catabolized. The reference range of the urinary concentration is 1000 mg/g creatinine), does the A1M concentration tend to be elevated with increasing albuminuria (Hoffmann et al., 1995). A1M is a useful marker for assessing renal damage in urological diseases (e.g. obstructive uropathy). Assaying urinary A1M provides a diagnostic tool for the diagnosis and monitoring of urinary tract disorders. A1M excretion was related to the grade of vesicoureteral reflux. The urinary A1M/creatinine ratio was highly sensitive and specific and correlated with DMSA scintigraphy. In contrast to cystitis, elevated urinary A1M excretion is seen in acute pyelonephritis (Everaert et al., 1998). Determination of urinary A1M has also been used in intoxications with heavy metals (cadmium (Jung et  al., 1993; Fels et  al., 1994), lead (Lim et al., 2001), and mercury-induced renal dysfunction (Kobal et al., 2000) at an early stage). Urinary A1M as a tubular marker and albumin as a glomerular marker were successfully used as an early marker for Balkan nephropathy screening (Cvorisćec, 2000). Urinary A1M can allow an early diagnosis of tubular proteinuria. In idiopathic membranous nephropathy, excretion of

assessment of renal function

IgG and A1M was associated with the extent of tubulointerstitial damage. A1M is also useful for the early detection of nephropathy in diabetic subjects. The proximal tubules are damaged early during the period of subclinical diabetic nephropathy (Marczewski et al., 1996).

Beta-2-microglobulin B2M (molecular weight:  11800)  has been used as a marker protein for tubulopathy (Guder et al., 1998, Delanghe, 2007). B2M is filtered by the glomerulus and almost completely reabsorbed and catabolized in the proximal tubules. B2M is a marker of tubular reabsorption. However, the protein is unstable at urinary pH values < 6. In tubulointerstitial diseases, the activity of urinary proteases may be increased, which may accelerate the breakdown of B2M. The reference range is < 0.3 mg/L or < 0.2 mg/g creatinine. B2M is a less reliable urine marker than other microproteins since various diseases (multiple myeloma, Waldenström’s disease, B-cell lymphomas, primary systemic amyloidosis, HIV-infection) raise its serum level, which in turn increases renal elimination. B2M can be measured using nephelometry or turbidimetry.

Retinol binding protein Because of its low molecular mass (~ 21 kD), retinol binding protein (RBP) is almost completely filtered by the glomerulus. In the proximal tubules > 99% is reabsorbed. Its reference range in urine is < 0.5 mg/L. In case of an impaired tubular protein resorption, increased RBP values are found in urine. The protein can be measured using immunonephelometry.

Alpha-2-macroglobulin For the evaluation of a significant haematuria (RBC > 25 cells/μL), another specific protein (A2M) can be used as a marker for postrenal bleeding. A2M is a high molecular weight (725 kDa) endoprotease inhibitor, present in plasma. The protein cannot pass the glomerular basement membrane, even in glomerular nephropathy. Since it is not related to or dependent of the urinary RBC-concentration, A2M can provide additional information in haematuria cases. The reference range of A2M/creatinine in urine is < 10 mg/g (or 1.13 mg/mmol). High values are observed in postrenal bleeding (Guder et al., 1998).

N-Acetyl-beta-D-glucosaminidase N-Acetyl-beta-D-glucosaminidase (ß-NAG) is a lysosomal enzyme showing a high activity in the proximal tubulus. Increased urinary ß-NAG activities are observed following toxic damage to the proximal tubular and interstitial cells (e.g. tubular dysfunction, nephritis, nephrotoxic substances (aminoglycosides, cytostatic drugs, contrast media)). ß-NAG is stable at room temperature for approximately 3 hours, at 4°C up to 50 days. When urine pH values exceed 8, enzyme activity drops. In the presence of high glucose concentrations, activity diminishes because of non-enzymatic glycation. The reference range is lower than < 6.3 U/L or < 5 U/g creatinine. One possible approach is based on the simultaneous measurement in the same specimen of different proteins both of glomerular (albumin and IgG) and tubular origin (A1M). The simultaneous measurement of IgG and A1M can be used to evaluate the course and response to treatment in glomerular diseases such as membranous nephropathy (Bazzi et al., 2001).

57

58

Section 1  

assessment of the patient with renal disease

Enzymuria Normal individuals excrete small amounts of enzymes located in the cells of the renal tubules in their urine (Pesce and First, 1979). Increased excretion of these enzymes is a sensitive index of renal damage (Marhun, 1979). These measurements have the advantage of being easily made and have a great sensitivity, but unfortunately are not specific for the disease processes and are not therefore very useful for diagnostic purposes. A number of brush-border enzymes have been studied in normal and pathological urines. Most data relate to NAG, a hydrolytic enzyme present in lysosomes (Sherman et  al., 1983). The excretion of tubular enzymes or NAG increases in a great variety conditions e.g. use of aminoglycoside antibiotics (Mondorf et al., 1984), cisplatin (Rojanasthien et al., 2001), contrast media (Erley et al., 1999), heavy metals and organic solvents (Tassi et al., 2000), glomerulonephritis (Valles et al., 2000), acute interstitial nephritis (Wolf et al., 1990), and extracorporeal shock-wave lithotripsy (Weichert-Jacobsen et al., 1998). The assay of urinary NAG or other tubular enzymes may detect minimal degrees of renal dysfunction, which is both an advantage and a disadvantage for clinical use. The problem in using these assays diagnostically is that, for example, modest renal insults, intercurrent urinary tract infections, increased urinary flow, and therapeutic doses of aminoglycosides will all increase urinary enzyme excretion. Thus, the clinical value of urinary NAG estimations in patients is limited.

References Abuelo, J. G. (1983). Proteinuria: diagnostic principles and procedures. Ann Intern Med, 98, 186–91. American Diabetes Association (2014). Standards of medical care in ­diabetes—2014. Diabetes Care, Suppl, 37, S14–80. Assadi, F. K. (2002). Quantitation of microalbuminuria using random urine samples. Pediatr Nephrol, 17, 107–10. Aucella, F., Guida, C. C., Lauriola, V., et al. (2010). How to assess renal function in the geriatric population. J Nephrol Suppl, 23, S46–54. Bagshaw, S. M., Bennett, M., Haase, M., et al. (2010), Plasma and urine neutrophil gelatinase-associated lipocalin in septic versus non-septic acute kidney injury in critical illness. Intensive Care Med, 36, 381–4. Bak, M., Thomsen, K., Christiansen, T., et al. (2000). Renal enlargement precedes renal hyperfiltration in early experimental diabetes in rats. J Am Soc Nephrol, 11, 1287–92. Bakoush, O., Torffvit, O., Rippe, B., et al. (2001). High proteinuria selectivity index based upon IgM is a strong predictor of poor renal survival in glomerular diseases. Nephrol Dial Transplant, 16, 1357–63. Bazzi, C., Petrini, C., Rizza, V., et al. (2001). Urinary excretion of IgG and alpha(1)-microglobulin predicts clinical course better than extent of proteinuria in membranous nephropathy. Am J Kidney Dis, 38, 240–8. Bennett, M., Dent, C. L., Ma, Q., et al. (2008). Urine NGAL predicts severity of acute kidney injury after cardiac surgery: a prospective study. Clin J Am Soc Nephrol, 3, 665–73. BioPAL, Inc. (2007). Laboratory Manual: Functional Immunoassay Technology (FITTM), Glomerular Filtration Rate (GFR), FIT-GFRTM Kit (Gd-DTPA). Worcester, MA: BioPal, Inc. Boelaert, J. R., Delanghe, J. R., Schurgers, M. L., et al. (1984). Red urine due to factitious myoglobinuria. Nephron, 38, 67–8. Bökenkamp, A., Domanetzki, M., Zinck, R., et al. (1998). Cystatin C—a new marker of glomerular filtration rate in children independent of age and height. Pediatrics, 101, 875–81. Bolignano, D., Lacquaniti, A., Coppolino, G., et al. (2009). Neutrophil gelatinase-associated lipocalin (NGAL) and progression of chronic kidney disease. Clin J Am Soc Nephrol, 4, 337–44.

Bradford, M. M. (1976). A rapid and sensitive method for the quantitation of microgram quantities of protein utilizing the principle of protein-dye binding. Ann Biochem, 72, 248–54. Brandt, J. R., Wong, C. S., Hanrahan, J. D., et al. (2006). Estimating absolute glomerular filtration rate in children. Pediatr Nephrol, 21, 1865–72. Brochner-Mortensen, J. (1978). Routine methods and their reliability for assessment of glomerular filtration rate in adults, with special reference to total [51Cr]EDTA plasma clearance. Dan Med Bull, 25, 181–202. Cameron, J. S. and Blandford, G. (1966). The simple assessment of selectivity in heavy proteinuria. Lancet, ii, 242–7. Ceriotti, F., Boyd, J. C., Klein, G., et al. (2008). Reference intervals for serum creatinine concentrations: assessment of available data for global application. Clin Chem, 54, 559–66. Chalasani, N., Clark, W. S., and Wilcox, C. M. (1997). Blood urea nitrogen to creatinine concentration in gastrointestinal bleeding: a reappraisal. Am J Gastroenterol, 92, 1796–99. Cocchetto, D. M., Tschanz, C., and Bjornsson, T. D. (1983). Decreased rate of creatinine production in patients with hepatic disease: implications for estimation of creatinine clearance. Ther Drug Monit, 5, 161–8. Cockcroft, D. W. and Gault, M. H. (1976). Prediction of creatinine clearance from serum creatinine. Nephron, 16, 31–41. Cottini, E. P., Gallina, D. L., and Dominguez, J. M. (1973). Urea excretion in adult humans with varying degrees of kidney malfunction fed milk, egg or an amino acid mixture: assessment of nitrogen balance. J Nutr, 103, 11–19. Counahan, R., Chantler, C., Ghazali, S., et al. (1976). Estimation of glomerular filtration rate from plasma creatinine concentration in children. Arch Dis Child, 51, 875–8. Crim, M. C., Calloway, D. H., and Margen, S. (1975). Creatine metabolism in men: urinary creatine and creatinine excretions with creatine feeding. J Nutr, 105, 428–38. Cruz, D. N., de Cal, M., Garzotto, F., et al. (2010). Plasma neutrophil gelatinase-associated lipocalin is an early biomarker for acute kidney injury in an adult ICU population. Intensive Care Med, 36, 444–51. Cvorisćec, D. (2000). Early diagnosis of endemic nephropathy. Clin Chim Acta, 297, 85–91. Decavele, A., Dhondt, L., De Buyzere, M.L., et al. (2011). Increased urinary neutrophil gelatinase associated lipocalin in urinary tract infections and leukocyturia. Clin Chem Lab Med, 49, 999–1003. Delanghe, J. (2007). New screening diagnostic techniques in urinalysis. Acta Clinica Belgica, 62, 155–61. Delanghe, J. (2008). How to establish glomerular filtration rate in children. Scand J Clin Lab Invest Suppl, 68, 46–51. Delanghe, J. R., Cobbaert, C., Galteau, M. M., et al. (2008). Trueness verification of actual creatinine assays in the European market demonstrates a disappointing variability that needs substantial improvement. An international study in the framework of the EC4 creatinine standardization working group. Clin Chem Lab Med, 46, 1319–25. Delanghe, J.R., Cobbaert, C., Harmoinen, A., et al. (2011). Focusing on the clinical impact of standardization of creatinine measurements: a report by the EFCC Working Group on Creatinine Standardization. Clin Chem Lab Med, 49, 977–82. Delanghe, J. R. and Speeckaert, M. M. (2011). Creatinine determination according to Jaffe-what does it stand for? NDT Plus, 4, 83–86. Devarajan, P. (2010). Review: neutrophil gelatinase-associated lipocalin. A troponin-like biomarker for human kidney injury. Nephrology 15, 419–28. Dodder, N. G., Tai, S.S., Sniegoski, L.T., et al. (2007). Certification of creatinine in a human serum reference material by GC-MS and LC-MS. Clin Chem, 53, 1694–9. Earley, E., Miskulin, D., Lamb, E. J., et al. (2012). Estimating equations for glomerular filtration rate in the era of creatinine standardization. Ann Intern Med, 156, 785–95. Everaert, K., Delanghe, J., Vande Wiele, C., et al. (1998). Urinary alpha 1-microglobulin detects uropathy. A prospective study in 483 urological patients. Clin Chem Lab Med, 36, 309–15.

chapter 7 

Fels, L. M., Bundschuh, I., Gwinner, W., et al. (1994). Early urinary markers of target nephron segments as studied in cadmium toxicity. Kidney Int Suppl, 46, S81–88. Ferguson, M. H., Olbrich, O., Robson, J.S., et al. (1950). The use of inulin clearance as a measure of glomerular filtration. Exp Physiol, 35, 251–79. Filler, G., Priem, F., Lepage, N., et al. (2002). ß-trace protein, cystatin C, ß2-microglobulin, and creatinine compared for detecting impaired glomerular filtration rates in children. Clin Chem, 48, 729–36. Ginsberg, J. M., Chang, B.S., Matarese, R.A., et al. (1983). Use of single voided urine samples to estimate quantitative proteinuria. N Engl J Med, 309, 1543–6. Graziani, M. S., Gambaro, G., Mantovani, L., et al. (2009). Diagnostic accuracy of a reagent strip for assessing urinary albumin excretion in the general population. Nephrol Dial Transplant, 24, 1490–4. Grenier, F. C., Ali, S., Syed, H., et al. (2010). Evaluation of the architect urine NGAL assay: assay performance, specimen handling requirements and biological variation. Clin Biochem, 43, 615–20. Grubb, A., Blirup-Jensen, S., Lindström, V., et al. (2010). First certified reference material for cystatin C in human serum ERM-DA471/IFCC. Clin Chem Lab Med, 48, 1619–21. Guder, W. G., Ivandic, M. and Hofmann W. (1998). Physiopathology of proteinuria and laboratory diagnostic strategy based on single protein analysis. Clin Chem Lab Med, 36, 935–9. Han, W. K., Waikar, S.S., Johnson, A., et al. (2008). Urinary biomarkers in the early diagnosis of acute kidney injury. Kidney Int, 73, 863–9. Hansen, H. P., Hovind, P., Jensen, B.R., et al. (2002). Diurnal variations of glomerular filtration rate and albuminuria in diabetic nephropathy. Kidney Int, 61, 163–8. Hari, P., Bagga, A., Mahajan, P., et al. (2007). Effect of malnutrition on serum creatinine and cystatin C levels. Pediatr Nephrol, 22, 1757–61. Harrison, H. H., Godsey, T., Bedford, K., et al. (1992). Comparison of microheterogeneity patterns of purified monoclonal light chains (Bence Jones proteins) and polyclonal free light chains that produce the pseudo-oligoclonal (‘ladder light chain’) pattern in immunofixation studies of urine. Clin Chem, 38, 963. Heerspink, H. J. L., Gansevoort, R.T., Brenner, B.M., et al. (2010). Comparison of different measures of urinary protein excretion for prediction of renal events. J Am Soc Nephrol, 21, 1355–60. Hellerstein, S., Berenbom, M., Alon, U. S., et al. (1998). Creatinine clearance following cimetidine for estimation of glomerular filtration rate. Pediatr Nephrol, 12, 49–54. Hemmingsen, L. (1972). A simple and rapid method for quantitative determination of protein in urine. Clin Chim Acta, 36, 185–8. Henry, R. J., Segalove, M. and Sobel, C. (1956). Turbidimetric determination of proteins with sulfosalicylic and trichloracetic acids. Proc Soc Exp Biol Med, 92, 748–51. Heymsfield, S. B., Arteaga, C., McManus, C., et al. (1983). Measurement of muscle mass in humans: validity of the 24-hour urinary creatinine method. Am J Clin Nutr, 37, 478–94. Higby, K., Suiter, C. R., Phelps, J. Y., et al. (1994). Normal values of urinary albumin and total protein excretion during pregnancy. Am J Obstet Gynecol, 171, 984–9. Hirsch, R., Dent, C., Pfriem, H., et al. (2007). NGAL is an early predictive biomarker of contrast-induced nephropathy in children. Pediatr Nephrol, 22, 89–95. Ichino, M., Kuroyanagi, Y., Kusaka, M., et al. (2009). Increased urinary neutrophil gelatinase associated lipocalin levels in a rat model of upper urinary tract infection. J Urol, 181, 2326–31. Iwata, J. and Nishikaze, O. (1979). New micro-turbidimetric method for determination of protein in cerebrospinal fluid and urine. Clin Chem, 25, 1317–9. Jaffe, M. (1886). Ueber den Niederschlag welchen Pikrinsa¨ure in normalen Harn erzeugt und über eine neue reaction des Kreatinins. Zeitschrift für Physiologische Chemie 10, 391–400. Jung, K., Pergande, M., Graubaum, H. J., et al. (1993). Urinary proteins enzymes as early indicators of renal dysfunction in chronic exposure to cadmium. Clin Chem, 39, 757–65.

assessment of renal function

Junge, W., Wilke, B., Halabi, A., et al. (2004). Determination of reference intervals for serum creatinine, creatinine excretion and creatinine clearance with an enzymatic and a modified Jaffe method. Clin Chim Acta, 344, 137–48. Kallner, A. and Khatami, Z. (2008). How does the MDRD Study equation compare with serum creatinine in routine healthcare? Anatomy of MDRD-eGFR. Scandinavian Journal of Clinical and Laboratory Investigation, 68, 39–45. Keevil, B. G., Kilpatrick, E. S., Nichols, S. P., et al. (1998). Biological variation of cystatin C: implications for the assessment of glomerular filtration rate. Clin Chem, 44, 1535–9. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group (2013). KDIGO clinical practice guideline for the evaluation and management of chronic kidney disease Kidney Int Suppl, 3, 1–150. King, J. S. and Boyce, W. H. (1963). High Molecular Weight Substances in Human Urine. Springfield, IL: Thomas. King, L. S., Choi, M., Fernandez, P. C., et al. (1998). Defective urinary-concentrating ability due to a complete deficiency of aquaporin-1. N Engl J Med, 345, 175–9. Kobal, A. B., Flisar, Z., Miklavcic, V., et al. (2000). Renal function in miners intermittently exposed to elemental mercury vapour. Arch Indust Hygiene Toxicol, 51, 369–80. Konta, T., Hao, Z., Takasaki, S., et al. (2007). Clinical utility of trace proteinuria for microalbuminuria screening in the general population. Clin Exp Nephrol, 11, 51–5. Koopman, M. G., Krediet, R. T., Koomen, G. C., et al. (1989). Circadian rhythm of proteinuria: consequences of the use of urinary protein: creatinine ratios. Nephrol Dial Transplant, 4, 9–14. Lamb, E. J., MacKenzie, F., and Stevens, P. E. (2009). How should proteinuria be detected and measured? Ann Clin Biochem, 46, 205–17. Levey, A. S. (1990). Measurement of renal function in chronic renal disease. Kidney Int, 38, 167–84. Levey, A. S., Bosch, J. P., Lewis, J. B, et al. (1999). A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med, 130, 461–70. Levey, A. S., Coresh, J., Greene, T., et al. (2007). Chronic Kidney Disease Epidemiology Collaboration. Expressing the Modification of Diet in Renal Disease Study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clin Chem, 53, 766–72. Levey, A. S., Greene, T., Kusek, I., et al. (2000). A simplified equation to predict glomerular filtration rate from serum creatinine [Abstract]. J Am Soc Nephrol, 11, A0828. Levey, A. S., Greene, T., Schluchter, M. D., et al. (1993). Glomerular filtration rate measurements in clinical trials. Modification of Diet in Renal Disease Study Group and the Diabetes Control and Complications Trial Research Group. J Am Soc Nephrol, 4, 1159–71. Levey, A. S., Stevens, L. A., Schmid, C. H., et al. (2009). A new equation to estimate glomerular filtration rate. Ann Intern Med, 150, 604–12. Lew, S. W. and Bosch, J. P. (1991). Effect of diet on creatinine clearance and excretion in young and elderly healthy subjects and in patients with renal disease. J Am Soc Nephrol, 2, 856–65. Liangos, O., Perianayagam, M. C., Vaidya, V. S., et al. (2007). Urinary N-acetyl-beta-(D)-glucosaminidase activity and kidney injury molecule-1 level are associated with adverse outcomes in acute renal failure, J Am Soc Nephrol, 18, 904–12. Lim, Y. C., Chia, K. S., Ong, H. Y., et al. (2001). Renal dysfunction in workers exposed to inorganic lead. Ann Acad Med Singapore, 30, 112–7. Lorentz, K. and Weiss, T. (1986). Protein bestimmung intramuscular Urin-Eine kritische Uebersicht. J Clin Chem Clin Biochem, 24, 309–23. Manetti, L., Pardini, E., Genovesi, M.., et al. (2005). Thyroid function differently affects serum cystatin C and creatinine concentrations. J Endocrinol Invest, 28, 346–9. Marczewski, K. T., Dziemidok, P., Grzywna, R., et al. (1996). Microalbuminuria or micro-alpha-1-microglobulinuria: which one is better predictor of the development of the secondary nephropathy? Clin Nephrol, 45, 136–7.

59

60

Section 1  

assessment of the patient with renal disease

Marhun, D. (1979). Evaluation of urinary enzyme patterns in patients with kidney diseases and primary benign hypertension. Curr Probl Clin Biochem, 9, 135–9. Maroni, B. J., Steinman, T. I. and Mitch, W. E. (1985). A method for estimating nitrogen intake of patients with chronic renal failure. Kidney Int, 27, 58–65. McElderry, L. A., Tarbit, I. F. and Cassells-Smith, A. J. (1982). Six methods for urinary protein compared. Clin Chem, 28, 2294–6. Merlini, G., Marciano, S., Gasparro, C., et al. (2001). The Pavia approach to clinical protein analysis. Clin Chem Lab Med, 39, 1025–8. Methven, S., MacGregor, M. S., Traynor, J. P., et al. (2011). Comparison of urinary albumin and urinary total protein as predictors of patient outcomes in CKD. Am J Kidney Dis, 57, 21–8. Michels, W. M., Grootendorst, D. C., Verduijn, M., et al. (2010). Performance of the Cockcroft-Gault, MDRD, and new CKD-EPI formulas in relation to GFR, age, and body size. Clin J Am Soc Nephrol, 5, 1003–9. Miller, W. G. (2009). Estimating equations for glomerular filtration rate in children: laboratory considerations. Clin Chem, 55, 402–3. Mondorf, A. W., Scherberich, J. E., Falkenberg, F. W., et al. (1984). Brush border enzymes and drug nephrotoxicity. In K. Solez and A. Wellthon (eds.) Acute Renal Failure, pp. 281–98. New York: Dekker. Nagao, S., Fujiwara, K., Imafuku, N., et al. (2005). Difference of carboplatin clearance estimated by the Cockroft-Gault, Jelliffe, Modified-Jelliffe, Wright or Chatelut formula. Gynecol Oncol, 99, 327–33. Nangaku, M. (2006). Chronic hypoxia and tubulointerstitial injury: a final common pathway to end-stage renal failure. J Am Soc Nephrol, 17, 17–25. Narva, A. S. (2009). Assessment of kidney function for drug dosing. Clin Chem, 55, 1609–11. National Kidney Disease Education Program, http://nkdep.nih.gov/ resources/ckd-drug-dosing-508.pdf National Kidney Foundation (2002). K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis Suppl, 39, S1–S266. National Kidney Foundation (2005). K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Part 1. Executive summary. Am J Kid Dis Suppl, 39, S17–S30. Notghi, A., Merrick, M. V., Ferrington, C., et al. (1986). A comparison of simplified and standard methods for the measurement of glomerular filtration rate and renal tubular function. British Journal of Radiology 59, 35–9. Nyman, U., Björk, J., Lindström, V., et al. (2008). The Lund-Malmö creatinine-based glomerular filtration rate prediction equation for adults also performs well in children. Scandinavian Journal of Clinical Laboratory Investigation, 68, 568–76. Orsonneau, J. L., Douet, P., Massoubre, C., et al. (1989). An improved pyrogallol red-molybdate method for determining total urinary protein. Clin Chem, 35, 2233–6. Panteghini, M. and IFCC Scientific Division. (2008). Enzymatic assays for creatinine: time for action. Clin Chem Lab Med, 46, 567–72. Payn, M. M., Webb, M. C., Lawrence, D., et al. (2002). Alpha1-microglobulin is stable in human urine ex vivo. Clin Chem, 48, 1136–8. Penders, J. and Delanghe, J. R. (2004). Alpha 1-microglobulin: clinical laboratory aspects and applications.Clin Chim Acta, 346, 107–18. Pesce, A. J. and First, R. M. (1979). Proteinuria. An Integrated Review. New York: Dekker. Pesce, M. A. and Strande, C. S. (1973). A new micromethod for determination of protein in cerebrospinal fluid and urine. Clin Chem, 19, 1265–7. Poggio, E. D., Wang, X., Greene, T., et al. (2005). Performance of the Modification of Diet in Renal Disease and Cockcroft-Gault equations in the estimation of GFR in health and in chronic kidney disease. J Am Soc Nephrol, 16, 459–66.

Portilla, D., Dent, C., Sugaya, T., et al. (2008). Liver fatty acid-binding protein as a biomarker of acute kidney injury after cardiac surgery. Kidney Int, 73, 465–72. Roberts, G. W., Ibsen, P. M. and Schiøler, C. T. (2009). Modified diet in renal disease method overestimates renal function in selected elderly patients. Age Ageing, 38, 698–703. Rodby, R. A., Rohde, R. D., Sharon, Z., et al. (1995). The urine protein to creatinine ratio as a predictor of 24-hour urine protein excretion in type 1 diabetic patients with nephropathy. The Collaborative Study Group. Am J Kidney Dis, 26, 904–9. Rojanasthien, N., Kumsorn, B., Atikachai, B., et al. (2001). Protective effects of fosfomycin on cisplatin-induced nephrotoxicity in patients with lung cancer. Int J Clin Pharmacol Ther, 39, 121–5. Ruggenenti, P., Gaspari, F., Perna, A., et al. (1998). Cross sectional longitudinal study of spot morning urine protein: creatinine ratio, 24 urine protein excretion rate, glomerular filtration rate, and end stage renal failure in chronic renal disease in patients without diabetes. Br Med J, 316, 504–9. Schlebusch, H., Liappis, N., Kalina, E., et al. (2002). High sensitive CRP and creatinine: reference intervals from infancy to childhood. J Lab Med, 26, 341–6. Schuhmann-Giampieri, G. and Krestin, G. (1991). Pharmacokinetics of Gd-DTPA in patients with chronic renal failure. Invest Radiol, 26, 975–79. Schwartz, G. F., Feld, L. G. and Langford, D. J. (1984). A simple estimate of glomerular filtration rate in full-term infants during the first year of life. J Pediatr, 104, 849–54. Schwartz, G. J. and Furth, S. L. (2007). Glomerular filtration rate measurement and estimation in chronic kidney disease. Pediatr Nephrol, 22, 1839–48. Schwartz, G. J., Furth, S., Cole, S. R., et al. (2006). Glomerular filtration rate via plasma iohexol disappearance: pilot study for chronic kidney disease in children. Kidney Int, 69, 2070–7. Schwartz, G. F. and Gauthier B. (1985). A simple estimate of glomerular filtration in adolescent boys. J Pediatr, 106, 522–26. Schwartz, G. J., Haycock, G. B., Edelmann, C. M. Jr, et al. (1976). A simple estimate of glomerular filtration rate in children derived from body length and plasma creatinine. Pediatrics, 58, 259–63. Schwartz, G. J., Muñoz, A., Schneider, M. F., et al. (2009). New equations to estimate GFR in children with CKD. J Am Soc Nephrol, 20, 629–37. Sherman, R. L., Drayer, D. E., Leyland-Jones, B. R., et al. (1983). N-acetyl-β-glucosaminidase and β2-microglobulin: their urinary excretion in patients with parenchymatous renal diseases. Arch Intern Med, 143, 1183–5. Shiba, K. S., Kanamori, K., Harada, T., et al. (1985). A cause of discrepancy between values for urinary protein as assayed by the Coomassie Brilliant Blue G-250 method and the sulphosalycilic acid method. Clin Chem, 31, 1215–8. Shukla, P. S., Endeman, H. J. and Kortlandt, W. (1988). Effects of five different antisera on the immunoturbidimetric determination of urinary albumin. Clin Chem, 34, 430. Speeckaert, M. M., Speeckaert, R., Van De Voorde, L., et al. (2011). Immunochemically unreactive albumin in urine: fiction or reality? Crit Rev Clin Lab Sci, 48, 87–96. Stamp, R. J. (1988). Measurement of albumin in urine by end-point immunonephelometry. Ann Clin Biochem, 245, 442–5. Stevens, L. A. and Levey A. S. (2004). Clinical implications for estimating equations for GFR. Ann Intern Med, 141, 959–61. Stevens, L. A. and Levey, A. S. (2009a). Measured GFR as a confirmatory test for estimated GFR. J Am Soc Nephrol, 20, 2305–13. Stevens, L. A. and Levey, A. S. (2009b). Current status and future perspectives for CKD testing. Am J Kid Dis, 53 (S3), S17–S26. Stevens, L. A., Nolin, T. D., Richardson, M. M., et al. (2009). Comparison of drug dosing recommendations based on measured GFR and kidney function estimating equations. Am J Kid Dis, 54, 33–42. Swaminathan, S. and Shah, S. V. (2007). New insights into nephrogenic systemic fibrosis. J Am Soc Nephrol, 18, 2636–43.

chapter 7 

Tassi, C., Abbritti, G., Mancuso, F., et al. (2000). Activity and isoenzyme profile of N-acetyl-beta-d-glucosaminidase in urine from workers exposed to cadmium. Clin Chim Acta, 299, 55–64. Toffaletti, J., Blosser, N., Hall, T., et al. (1983). An automated dry-slide enzymatic method evaluated for measurement of creatinine in serum. Clin Chem, 29, 684–7. Vaidya, V. S., Ferguson, M. A. and Bonventre, J. V. (2008). Biomarkers of acute kidney injury. Annu Rev Pharmacol Toxicol, 48, 463–93. Valles, P., Peralta, M., Carrizo, L., et al. (2000). Follow-up of steroid-resistant nephrotic syndrome: tubular proteinuria and enzymuria. Pediatr Nephrol, 15, 252–8. Weichert-Jacobsen, K., Stöckle, M., Loch, T., et al. (1998). Urinary leakage of tubular enzymes after shock wave lithotripsy. Eur Urol, 33, 104–10. Weicheselbaum, T. E. (1946). An accurate and rapid method for the determination of proteins in small amounts of blood serum and plasma. Am J Clin Pathol, 10, 40–9.

assessment of renal function

White, S. L., Yu, R., Craig, J. C., et al. (2011). Diagnostic accuracy of urine dipsticks for detection of albuminuria in the general community. Am J Kidney Dis, 58, 19–28. Wolf, G., Scherberich, J. E., Nowack, A., et al. (1990). Urinary excretion of dipeptidyl aminopeptidase IV in patients with renal diseases. Clin Nephrol, 33, 136–42. Wuyts, B., Bernard, D., Van den Noortgate, N., et al. (2003). Reevaluation of formulas for predicting creatinine clearance in adults and children, using compensated creatinine methods. Clin Chem, 49, 1011–14. Wyss, M. and Kaddurah-Daouk, R. (2000). Creatine and creatinine metabolism. Physiol Rev, 80, 1107–213. Yilmaz, A., Sevketoglu, E., Gedikbasi, A., et al. (2009). Early prediction of urinary tract infection with urinary neutrophil gelatinase associated lipocalin. Pediatr Nephrol, 24, 2387–92. You, L., Zhu, X., Shrubsole, M. J., et al. (2011). Renal function, bisphenol A, and alkylphenols: results from the National Health and Nutrition Examination Survey (NHANES 2003–2006). Environ Health Perspect, 119, 527–33.

61

CHAPTER 8

Tubular function Marijn Speeckaert and Joris Delanghe Introduction Assessment of tubular function is more complicated than the measurement of the glomerular filtration rate (GFR), because tubular functions are more complex and differ along the nephron. Tubular functions cannot be studied as a whole, but must be divided according to the different segments of tubule. Most investigations of the tubular functions also necessitate an accurate measurement of the GFR.

Investigating proximal tubular functions The first postglomerular segment of the nephron is the proximal tubule. This segment reabsorbs about 60% of the filtered water and the solutes in a heterogeneous manner. Some functions are specific for the proximal tubule and their alterations indicate proximal tubule dysfunction. (See also Chapters 20 and 41).

Proximal tubule dysfunction Several rare syndromes are characterized by isolated renal phosphate wasting. A generalized impairment of the proximal tubule is known as the Fanconi syndrome, in which hypophosphataemia, renal glucosuria, hypouricaemia, aminoaciduria, and proximal renal tubular acidosis due to bicarbonate loss in the urine are observed (Clarke et al., 1995). The classic example is X-linked hypophosphataemic rickets, in which the defect in proximal tubular phosphate transport is due to a mutation in the PHEX gene.

Concentrating and diluting urine The kidney plays a central role in water homeostasis by its ability to concentrate or dilute urine according to water intake and extrarenal losses. Water is absorbed in the proximal tubule, the thin descending limb, and the collecting tubule. In the proximal tubule, water movement from lumen to blood is iso-osmotic and driven by active solute transports:  solutes accumulate at the external side of the basolateral cell membranes, generating a transepithelial osmotic pressure difference that drives water flow. Consequently, when an osmotically active solute is present in the tubular fluid and cannot be reabsorbed, water reabsorption in the proximal tubule is decreased leading to an osmotic diuresis. In the proximal tubule transepithelial water transport uses two routes: transcellular and paracellular. Water crosses proximal cell membranes through the type 1 aquaporin, which is constitutively expressed at the apical and basolateral sides of the cells. In the thick ascending limb and the distal convoluted tubule, no net water reabsorption occurs; only solutes are retrieved from the

urine. The loop of Henle creates the countercurrent multiplication process and the corticomedullary osmotic gradient (Morello and Bichet, 2001). In the distal convoluted tubule, the lumen osmolality decreases to 60 mOsm/kg H2O. The collecting duct has extremely low water permeability in the absence of the antidiuretic hormone vasopressin. Under this condition, no water transport is observed along the collecting duct and the final urine is maximally diluted. In the presence of vasopressin, transepithelial movements of water occur across the principal cells. Three types of aquaporins are involved in water transport in the principal cells. Aquaporin 3 and 4 expression are restricted to the basolateral plasma membrane and are not regulated by vasopressin. They provide the pathways that permit water to enter the interstitium. Aquaporin 2 is expressed at the apical plasma membrane exclusively in the presence of vasopressin, thus increasing water permeability. The transepithelial difference in osmolality drives water from the lumen to the interstitium. When the level of vasopressin in plasma is high enough, water reabsorption occurs along the entire collecting duct and the final urine is maximally concentrated (Nielsen et al., 2002).

Testing the concentrating ability The concentrating ability of the kidney depends not only on the action of vasopressin but also on the presence of an adequate osmotic corticomedullary gradient. This gradient is altered at the early stage of renal insufficiency, and is profoundly reduced in particular in pathology with medullary damage such as adult polycystic kidney disease, nephronophthisis, sickle cell disease, reflux nephropathy, tubular necrosis, or transplanted kidneys. The assessment of the concentrating capacity of the kidney is based on the measurement of the urine osmolality. Plasma osmolality must be measured simultaneously for a better interpretation of the results. Urine and plasma osmolalities are measured by the depression of the freezing point. The study of the maximal capacity of the kidney to concentrate urine is important whenever a form of diabetes insipidus is suspected. This defect can be of a genetic origin or can be acquired. It can be located in the kidney (nephrogenic diabetes insipidus) or due to the lack of vasopressin secretion. The maximal concentrating capacity of the kidney is determined by a fluid restriction test (Davis and Zenser, 1993). The patient has to stop all fluid and food intake from midnight the day prior to the investigation. The next morning, plasma and urine osmolalities are measured. In normal subjects, plasma osmolality ranges between 280 and 295 mOsm/kg H2O and urine osmolality is > 600 mOsm/ kg H2O. The maximal urine concentration, that usually exceeds 1000 mOsm/kg H2O, is achieved after water deprivation from 18 to 24 hours (Miles et al., 1954; Rowe et al., 1976). The maximum

chapter 8 

concentrating ability is lower in infants and elderly people (Rowe et  al., 1976; Monnens et  al., 1981; Aperia et  al., 1983; Davis and Davis, 1987). During the investigation, the urinary flow falls below 0.4 mL/min, blood pressure and urine osmolality are measured every hour and plasma osmolality every 2 hours. The test is usually stopped when urine osmolality exceeds 800 mOsm/kg H2O, which excludes major defects in the urine concentrating process, or when plasma osmolality is above the upper normal values (296 mOsm/kg H2O). This will distinguish between nephrogenic diabetes insipidus and central deficiency of vasopressin secretion. In central diabetes insipidus, DDAVP (desmopressin) increases urine osmolality above 750 mOsm/kg H2O within the first 2 hours following the injection (Radó et al., 1976; Curtis and Donovan, 1979; Moses et  al., 1985; Williams et  al,. 1986). In nephrogenic diabetes insipidus, DDAVP injection does not increase urine osmolality further. A blood sample is usually drawn before DDAVP injection to allow plasma vasopressin concentration measurement in case of central deficiency. This test can be used in infants and children (Monnens et al., 1981). In compulsive water drinkers, urine osmolality rises above 600 mOsm/kg H2O. Surreptitious water ingestions induce a fall in urine osmolality between two successive collections accompanied by an increase in urinary flow rate.

Testing the diluting capacity The diluting capacity of the kidney can also be tested. This investigation is useful in patients who present with unexplained hyponatraemia. The lowest urinary osmolality is obtained when vasopressin secretion is completely suppressed. Diuretics, by decreasing sodium reabsorption especially in the distal convoluted tubule (thiazide), impair the diluting capacity. The diluting capacity requires normal solute intake. At the beginning of the investigation, plasma and urinary osmolality are measured, then the patient drinks a water load of 20 mL/kg body weight in 20 minutes (DeRubertis et al., 1971; Davis and Zenser, 1993). In normal subjects, the plasma osmolality remains within the normal range after the water load and urine osmolality rapidly decreases below 100 mOsm/kg H2O within 2 hours and 80% of the ingested water is excreted within 4 hours. An alteration of the diluting capacity can be due to an incomplete inhibition of vasopressin secretion, or to adrenal insufficiency, hypothyroidism, potassium depletion, or liver disease.

tubular function

For urine, the specific gravity is a function of the number and weight of the dissolved solute particles However, in some conditions there is a remarkable divergence between osmolality and specific gravity. If larger-molecular-weight substances are present, there will be divergence between specific gravity and osmolality. It used to be measured with a hydrometer, commonly called a ‘urinometer’, a weighted float marked with a scale for specific gravities from 1.000 to 1.060. The urinometer is simple and quick to use, but has some drawbacks:  it requires a volume of urine not always available in spot samples. There may be difficulty in reading the meniscus and there may be a tendency for the device to cling to the side of the container. Moreover, specific gravity varies with temperature and urinometers may be inaccurately calibrated. Urine specific gravity should be corrected for protein and glucose, if either is present (Schrier, 2007). Proteins increase the specific gravity by 0.001 for each 0.4 g/dL and glucose by 0.001 for each 0.27 g/dL (Hofmann et al., 1995).

Refractometry Refractometry is the most reliable way to measure specific gravity. This method is based on measurement of the refractive index, which is related to the weight of solutes per unit volume of urine. Several types of refractometers are available, which are very simple to use and require only one drop of urine. In some automated test strip readers, a refractometer is integrated in the instrument.

Osmolality

Relative density

Osmolality is the measure of solute concentration, defined as the number of osmoles (Osm) of solute per litre (L)  of solution (osmol/L). Osmolality is a measure of the osmoles of solute per kilogram of solvent (osmol/kg or Osm/kg). Measurements of the osmotic concentration of urine are considered more valid than specific gravity measurements. Osmolality is a function of the number of particles in solution and is usually measured through the effect of colligative properties on the freezing point of the solution. Numerous manual or fully automated osmometers are available commercially. Measurement of osmolality offers definite advantages over other methods. These include (a) no temperature correction is necessary, (b) only small volumes of urine are required, and (c) there is no interference from proteins or other macromolecules. However, high glucose concentrations contribute significantly to the measured osmolality (10 g/L of glucose = 55.5 mOsm) (Hofmann et al., 1995). (See Table 8.1.)

Effects of dilution may be measured by four methods: specific gravity, refractometry, osmolality, and dry chemistry (dipsticks).

Dry chemistry

Assessment of renal concentration ability The capacity of the kidneys to conserve water can be assessed by measuring the solute concentration of urine. After fluid deprivation of 18 hours or more, urine osmolality exceeding 850 mOsm/kg is considered normal. In cases of a complete antidiuretic hormone (ADH) deficiency or in nephrogenic diabetes insipidus (where the normal response to ADH is lacking), urine osmolality rarely exceeds 300 mOsm/kg.

Specific gravity Specific gravity is a measure of the content of solids in a solution. The measurement of urine specific gravity is related to osmolality.

Alternatively, specific gravity can be measured as a urine dipstick test. After intravenous administration of iodine-containing contrast media, extraordinarily high values (> 1.040) may be temporarily obtained.

Urine conductivity The conductivity can be used as a substitute for osmolality (reference range: 8–32 mS/cm). The conductivity measurement is related to the concentration of electrolytes in the urine. In contrast to osmolality and specific gravity, uncharged particles (e.g. glucose, urea) do not contribute to conductivity. There is a good agreement between conductivity, urine density, and osmolality (r = 0.8 or higher) (Hofmann et al., 1995). Conductivity correlates better

63

64

Section 1  

assessment of the patient with renal disease Table 8.1  Serum and urine osmolality Serum osmolality Ref. values: 280–296 mOsm/kg

Urine osmolality Ref. values: 500–800 mOsm/kg

Clinical significance

Normal or increased Decreased Normal Normal or increased

Increased Decreased Decreased Decreased (without increased fluid intake) Increased

Volume deficit Volume excess Increased fluid intake or diuretics use Kidneys unable to concentrate urine or lack of ADH (diabetes insipidus) SIADH (syndrome of inappropriate ADH secretion)

Decreased

with osmolality than the creatinine concentration (r = 0.84 and 0.68, respectively) and therefore might serve as a better indicator of hydration status ((Hofmann et al., 1995; Hannemann-Pohl and Kampf, 1999).

Osmolar clearance and free-water clearance The quantitative ability of the kidney to excrete or retain water can be estimated by the calculation of osmolar and free-water clearance. Conceptually, the rate of water excreted in urine can be divided into two components. One contains all the urinary solutes and the amount of water required to obtain a solution iso-osmotic to plasma. This defines the osmolar clearance. The other component is made of solute-free water and is called free water clearance: CH2O = V – Uosm × V/Posm. This theoretical flow is positive when urine is hypotonic to plasma (water has been ‘added in excess’ to osmoles in urine) and negative when final urine is hypertonic to plasma (water has been ‘substracted in excess’ to osmoles to concentrate urine). The contribution of urea to water excretion and the use of urea concentration to calculate osmolality are detailed in Soroka et al. (1997).

transporter (Moe et al., 2000). This transporter is also expressed in the apical membrane of the collecting duct principal cells together with the amiloride sensitive sodium channel (ENaC) (Kim et al., 1998). In this segment, sodium reabsorption is regulated by the mineralocorticoid hormone aldosterone and coupled to K+ secretion. A global measurement of sodium absorption by the tubules is obtained by measuring the fractional excretion of sodium (FENa) which is expressed as the ratio of the sodium excretion rate to the sodium filtered load: where is the urinary concentration of sodium, the urine flow rate. In various clinical conditions, physicians are interested in knowing what proportion of sodium is being reabsorbed in the proximal and distal parts of the nephron. The measurement of sodium absorption in the proximal tubule is based on the assumption that lithium is absorbed to the same extent as sodium in this segment but is not handled in the distal parts of the nephron (Thomsen, 1990). Measurement of the urine sodium concentration provides information on the integrity of tubular reabsorptive function. The following equation can be used to calculate the fractional excretion of sodium: FE Na = 100 × U Na × PCrn / PNa × U Crn .

Fractional excretion of sodium



Sodium reabsorption occurs along the entire nephron by various mechanisms. In the proximal tubule, 60% of the filtered sodium is recovered from the initial urine by active transcellular movements and passive paracellular pathways. In the first mechanism, sodium enters the apical pole of proximal cells coupled to organic (glucose, amino acids, etc.) or inorganic (phosphate) solutes via specific sodium-dependent carriers or is exchanged for H+ ions. In all tubule segments, sodium is actively pumped out of the cells via the basolateral Na+/K+-ATPase. The passive paracellular sodium flux is due to osmotic water flow from lumen to blood that entrains sodium through the lateral intercellular space (Moe et al., 2000). The thick ascending limb reabsorbs about 30% of the sodium filtered at the glomerulus. Sodium is transported across the apical membranes via the Na+/K+/Cl2−-cotransporter and the Na+/H+exchanger. The K+-recycling at the apical membrane generates a lumen positive transepithelial voltage driving additional paracellular sodium reabsorption. Furosemide and bumetanide inhibit the Na+/K +/Cl 2−cotransporter and sodium absorption in the thick ascending limb. The distal convoluted tubule reabsorbs about 10% of the filtered sodium load via the apical thiazide-sensitive sodium chloride

A FENa < 1% suggests a prerenal cause, because volume depletion or a decrease in the effective circulating volume induce an increase in sodium reabsorption. A value > 2–3% is indicative of acute tubular necrosis or injury. In renal tract obstruction, values may be either > or < 1%. The value is lower in early disease, but with renal damage from the obstruction, the value becomes higher (Steiner, 1984).

References Aperia, A., Broberger, O., Herin, P., et al. (1983). Postnatal control of water and electrolyte homeostasis in pre-term and full-term infants. Acta Paediatr Scand Suppl, 305, 61–5. Clarke, B. L., Wynne, A. G., Wilson, D. M., et al. (1995). Osteomalacia associated with adult Fanconi’s syndrome: clinical and diagnostic features. Clin Endocrinol, 43, 479–90. Curtis, J. R. and Donovan, B. A. (1979). Assessment of renal concentrating ability. Br Med J, 1, 304–5. Davis, B. B. and Zenser, T. V. (1993). Evaluation of renal concentrating and diluting ability. Clin Lab Med, 13, 131–4. Davis, P. J. and Davis, F. B. (1987). Water excretion in the elderly. Endocrinol Metab Clin North Am, 16, 867–75. DeRubertis, F. R., Michelis, M.F., Beck, N., et al. (1971). ‘Essential’ hypernatremia due to ineffective osmotic and intact volume regulation of vasopressin secretion. J Clin Invest, 50, 97–111.

chapter 8 

Hannemann-Pohl, K. and Kampf, S. G. (1999). Automation of urine sediment examination: a comparison of the Sysmex UF-100 automated flow cytometer with routine manual diagnosis (microscopy,test strips, and bacterial culture). Clin Chem Lab Med, 37, 753–4. Hofmann, W., Edel, H. and Guder, W. G. (1995). A mathematical equation to differentiate overload proteinuria from tubulo-interstitial involvement in glomerular diseases. Clin Nephrol, 44, 28–31. Kim, G. H., Masilamani, S., Turner, R., et al. (1998). The thiazide-sensitive Na-Cl cotransporter is an aldosterone-induced protein. Proc Natl Acad Sci U S A, 95, 14552–7. Miles, B. E., Paton, A., and De Wardener, H. E. (1954). Maximum urine concentration. Br Med J, 16, 901–5. Moe, O. W., Berry, C. A., and Rector, F. C. J. (2000). Renal transport of glucose, aminoacids, sodium, chloride and water. In B. M. Brenner (ed.) The Kidney, pp. 375–415. Philadelphia, PA: W. B. Saunders Company. Monnens, L., Smulders, Y., van Lier, H., et al. (1981). DDAVP test for assessment of renal concentrating capacity in infants and children. Nephron, 29, 151–4. Morello, J. P. and Bichet, D. G. (2001). Nephrogenic diabetes insipidus. Annu Rev Physiol, 63, 607–30. Moses, A. M., Scheinman, S. J., and Schroeder, E. T. (1985). Antidiuretic and PGE2 responses to AVP and dDAVP in subjects

tubular function

with central and nephrogenic diabetes insipidus. Am J Physiol, 248, F354–F359. Nielsen, S., Frøkiaer, J., Marples, D., et al. (2002). Aquaporins in the kidney: from molecules to medicine. Physiol Rev, 82, 205–44. Radó, J. P., Marosi, J., Szende, L., et al. (1976). The antidiuretic action of 1-deamino-8-D-arginine vasopressin (DDAVP) in man. Int J Clin Pharmacol Biopharm, 13, 199–209. Rowe, J. W. Shock, N. W. and DeFronzo, R. A. (1976). The influence of age on the renal response to water deprivation in man. Nephron, 17, 270–8. Schrier, G. (2007). Diseases of the Kidney and Urinary Tract. Philadelphia, PA: Lippincott Williams & Wilkins. Soroka, S. D., Chayaraks, S., Cheema-Dhadli, S., et al. (1997). Minimum urine flow rate during water deprivation: importance of the nonurea versus total osmolality in the inner medulla. J Am Soc Nephrol, 8, 880–6. Steiner, R. (1984). Interpreting the fractional excretion of sodium. Am J Med, 77, 699–702. Thomsen, K. (1990). Lithium clearance as a measure of sodium and water delivery from the proximal tubules. Kidney Int Suppl, 28, S10–S16. Williams, T. D., Dunger, D. B., Lyon, C. C., et al. (1986). Antidiuretic effect and pharmacokinetics of oral 1-desamino-8-D-arginine vasopressin. 1. Studies in adults and children. J Clin Endocrinol Metab, 63, 129–32.

65

CHAPTER 9

Renal radiology: overview Michael J. Weston Brief introduction to radiology chapters Imaging of form and function is an integral part of modern medical practice essential for diagnosis, treatment, and monitoring. Moreover, image-guided biopsies, drainage procedures, and ablations have circumvented the need for many invasive surgical procedures. There are a multitude of different techniques available so their diversity makes an understanding of radiological practice essential for the clinicians who rely upon them. Chapters  10–16 will highlight the uses of plain radiography, fluoroscopy, ultrasound, computed tomography (CT), magnetic resonance imaging, radionuclide studies, and image-guided intervention. All imaging studies work best if a specific question is asked. This helps to choose both the best modality and protocol to answer the question. The clinical information given will often assist the interpretation of the findings. The more vague the indication for a scan, the less likely that useful information will be provided. Both the requesting clinician and the radiologist need to be clear how the scan result will alter management. Performing imaging procedures that will not alter the outcome is wasteful and unkind to the patient. It is important to understand the adverse effects of imaging tests. Some, such as the effects of ionizing radiation, are well described but often poorly understood. A  surprisingly high proportion of doctors who are not radiologists have been found in surveys to be ignorant of which radiological techniques involve ionizing radiation and are unable to quantify the relative risks involved. Adverse effects of contrast medium administration are particularly pertinent to the nephrologist. Other adverse effects are less appreciated, for example, the burden imposed by incidental findings discovered on tests done for other reasons. It is not unknown for patients to embark on a series of tests requested to clarify incidental imaging findings, causing them to suffer both psychological and physical harm, without, in the end, any benefit to their health. The clinician should remember the need to treat the patient and not the scan. An understanding of the difference between form and function is essential. For instance, dilation of the renal collecting system does not equate to obstruction. Ultrasound of the kidneys in someone who has a urinary conduit will often show hydronephrosis because an ultrasound cannot distinguish obstruction from chronic reflux. Renal function will clearly have a role in deciding if further imaging is needed. Isotope renography or a conduitogram are both potential next imaging steps. Similarly, pregnant women often show

physiological renal upper tract dilation so finding hydronephrosis will not distinguish pregnancy-related dilation from stone disease in someone with flank pain. Magnetic resonance urography can solve this problem whilst avoiding ionizing radiation to the fetus. Many radiological tests do combine some measure of function along with a depiction of anatomy. CT with intravenous contrast medium enhancement will give a broad indication of the kidney’s ability to excrete. This is not as reliable as isotope renography in assessing relative function, though both tests are contraindicated or unhelpful in renal failure. Positron emission tomography (PET) in its commonest manifestation looks at glucose metabolism and when superimposed on or combined with a CT scan done at the same time can clearly highlight areas of high metabolism—most commonly looking for metastases or lymph nodes involved in post-transplant lymphoproliferative disorder. An important task for the referring clinician is preparation and patient consent for the radiological test. Iodinated contrast agents are nephrotoxic; it is not enough to assure the radiologist of appropriate prior patient hydration, the patient should also give informed consent. This is more than just understanding the risk of worsening renal failure but also considering alternate ways of investigating the problem. Clearly, patient consent is even more of an issue when an image-guided biopsy is proposed. Proper consent requires input from both the clinician requesting the biopsy and the radiologist who will perform the procedure. Incorrect patient preparation may also prevent a scan from being done and waste an appointment slot. For instance, having breakfast just before attending for a PET/CT scan will mean the scan will not be done (because of the effect on glucose metabolism). Finally, the imaging history needs to be kept in mind. The answer to a problem may not lie in asking for a new imaging test but in a reappraisal of the previous imaging. The following chapters will show how to choose the right test, to recognize their limitations and risks, and to understand what the findings mean. The patient needs to be properly involved in the decision-making process so that they are prepared both mentally and physically for the procedure and its purpose. It is important for clinicians to establish a good relationship with the radiology department and the radiologists to ensure that the best service can be delivered. In summary: 1. Only request investigations if the result can be interpreted and used to alter management.

chapter 9 

renal radiology: overview

2. Choose the right test by consulting the radiology colleague in advance with a description of the problem.

6. Be aware of potential adverse effects.

3. Always review previous imaging.

8. Ensure that the patient is properly prepared for the investigation, for example, bowel preparation, full bladder, fasted for PET scans, and hydrated for contrast administration.

4. Involve the patient in the discussion and take proper consent. 5. Do not confuse anatomical and functional findings.

7. Treat the patient and not the scan.

67

CHAPTER 10

Ionizing radiation and radiation protection Jeannette Kathrin Kraft and Peter Howells Overview Ionizing radiation is a natural phenomenon and we are constantly exposed to small amounts of radiation from natural and artificial sources. The largest amount of man-made radiation we receive as a population is from medical imaging. Radiation exposure in medical imaging can be expressed as effective dose measured in millisieverts (mSv). As effective dose is difficult to measure in practice, indicators of dose such as dose area product (DAP) and dose length product (DLP) are commonly used. It is usual to compare the amount of radiation patients receive through imaging with natural background radiation (2.4 mSv per year). Ionizing radiation has potential adverse effects on the human body. It may cause acute effects due to cell death called tissue reactions or deterministic effects. Damage to the cell’s DNA may lead to random or stochastic effects such as cancer induction. Available data on the risk of inducing cancer are extrapolated from the high-dose exposures of atomic bomb survivors or radiation accidents. It is difficult to estimate cancer risk accurately at the low radiation doses commonly used in medical imaging. Therefore models such as the linear non-threshold model are used to extrapolate risk at low-level radiation exposure. As medical imaging has the potential for harm, the risk to the patient has to be managed and controlled by all heath care professionals. The Ionising Radiation (Medical Exposure) Regulations 2000 (IR(ME)R 2000) impose obligations on hospital authorities to minimize radiation doses to patients in the United Kingdom (Department of Health, 2000). This can be achieved through careful imaging and technique selection, justification, and optimization using the ALARA principle in keeping the dose ‘as low as reasonably achievable’ or practicable. Repeat studies should be avoided and diagnostic reference levels (DRLs) regularly audited. Both healthcare professionals and patients need to be educated about the risks and benefits of imaging. Employees working with ionizing radiation are regularly monitored. Benefit from medically indicated imaging almost always outweighs the risk especially in emergency situations such as assessment of trauma. Radiation should not be feared.

Introduction In 1895, the German physicist Wilhelm Conrad Röntgen discovered X-rays when he noted that a fluorescent screen began to glow while he was experimenting with a vacuum tube. Soon after, in 1896, Becquerel noted that photographic plates stored with uranium were

fogged, thereby discovering natural occurring radiation. By the early 1900s, X-rays were used widely in medical imaging. However, soon after their discovery adverse health effects became apparent and many early radiologists died of radiation-induced sickness and cancer (Berry, 1986). In 1915, the British Röntgen Ray Society adopted radiation protection recommendations, probably representing the first joint effort at radiation protection. In the 1920s, film badges were introduced for routine monitoring of personnel and in 1925 the first exposure limits were set. In the 1950s, the study of biological effects continued with the long-term follow-up of the subjects exposed to radiation from the two atomic bombs in Japan in 1945 and is still ongoing today (Richardson et al., 2009; Ozasa et al., 2012). Radiation accidents such as Chernobyl in 1986 and Fukushima in 2011 increased public awareness. After the advent of computed tomography (CT) in the 1970s, the use of ionizing radiation in medical imaging has steadily increased. Approximately 62  million CT examinations are performed each year in the United States and approximately 4 billion radiological examinations are performed worldwide (Brenner and Hall, 2007; United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), 2011). The first direct studies of cancer risk in children who have undergone CT scanning were published in 2012 and 2013 (Pearce et al., 2012; Mathews et al., 2013). Although the risk for any individual may be small, the increasing radiation exposure to the population may be a health risk in the future (Brenner and Hall, 2007). However, there is still a lack of awareness of radiation exposure and associated risks among medical practitioners (Shiralkar et al., 2003).

Definition of ionizing radiation Radiation is a form of energy transmitted through space. There are many types including heat, light, radio waves, microwaves, X-rays, and gamma rays. The frequency of the wave describes its position in the electromagnetic spectrum. Low-energy radio waves with low frequencies are at one end of the electromagnetic spectrum; high-frequency, high-energy waves such as X-rays or gamma rays are at the other end. High-frequency waves that carry a large amount of energy can penetrate material and transfer energy to an atom that may cause displacement of an electron from its orbit around the nucleus. This process is called ionization. Ionizing radiation can be electromagnetic radiation such as X-rays or gamma rays or particulate radiation such as alpha particles, beta particles,

chapter 10 

or neutrons. It can be produced in a generator such as an X-ray machine or come from radioactive material as used in nuclear medicine or radiotherapy.

ionizing radiation and radiation protection

Measuring radiation

constituents of the soil. The average annual effective dose from natural sources in the world is estimated to be 2.4 mSv (UNSCEAR, 2010). When describing radiation exposure from medical examinations we often compare the dose received from that examination to natural background radiation.

Absorbed dose

Man-made radiation

When ionizing radiation passes through tissue it deposits energy causing ionization and excitation of the matter. The amount of energy deposited divided by the mass of tissue exposed is called the ‘absorbed dose’. The unit of absorbed dose is joule per kilogram (J/ kg) but for convenience it has been given a special name: gray (Gy). This is an easy quantity to measure but is not a useful indicator of long-term risk.

Radiation exposure may be increased in several ways by human activity and technological processes, for example, air travel (increased levels of cosmic radiation), nuclear testing, and nuclear accidents such as Chernobyl in 1986, nuclear power production, occupational exposure, consumer products such as electronic equipment, fire alarms, and luminous numbers on watches and clocks. However, the largest amount of man-made radiation exposure is through medical imaging procedures.

Equivalent dose Different kinds of radiation cause different biological effects for the same amount of energy deposited. For instance, neutrons or alpha particles cause more damage per unit absorbed dose than X-rays or gamma rays. To allow for this, a quality factor for radiation was introduced leading to the expression ‘equivalent dose’. For X-rays and gamma rays the absorbed dose equals the equivalent dose. The unit of equivalent dose is also J/kg but to avoid confusion it has been given a different name: sievert (Sv).

Effective dose In addition, different tissues have variable sensitivity to radiation explained by cell turnover rate and tissue-specific characteristics. To account for this a tissue weighting factor was added to modify equivalent dose to ‘effective dose’. The unit for the effective dose stays the same: Sv. Doses in medical imaging and radiation protection are usually organ specific and refer to effective doses. As they are often small, most are expressed as millisievert (mSv).

Indicators of dose In practice, it is not possible to measure effective dose directly. Within a procedure other indicators of dose are generally used, such as the DAP in radiography and fluoroscopy or DLP in CT. DAP measures the dose in air multiplied by the X-ray beam area at a point close to the X-ray unit. This quantity is a constant for any distance from the X-ray tube. DLP measures the dose per rotation of the CT scanner multiplied by the length of the patient included in the scan. Effective dose can be inferred from each of these values but only with knowledge of other parameters.

Sources of ionizing radiation Radiation is a natural phenomenon which we are exposed to continuously from natural and artificial man-made sources. Low levels of radiation cannot be seen or felt, so people are not usually aware of it.

Natural background radiation Natural background radiation is ubiquitous and includes high-energy cosmic rays, radioactive nuclides from the earth’s crust such as uranium and thorium, and radon gas. The annual effective dose from natural background radiation varies substantially throughout the world due to variation in altitude and the

Medical exposures Medical exposures account for 98% of the contribution from all artificial sources and are now the second largest contributor to the population dose worldwide, representing approximately 20% of the total from all sources. It is estimated that from 1997 to 2007, approximately 3.6 billion medical radiation procedures were performed annually worldwide, an increase of 40% on the last decade (UNSCEAR, 2010). This increase is demonstrated particularly for high-dose examinations, such as CT scanning, the use of which is increasing rapidly. Medical radiation procedures were 65 times more frequent in developed countries compared to underdeveloped countries reflecting an imbalance of radiation exposure and healthcare provision. In several countries medical radiation exposure is now larger than exposure from natural sources (UNSCEAR, 2010).

Potential health effects of ionizing radiation Biological effects As previously described, high-energy ionizing radiation causes displacement of an electron from its orbit around the nucleus. This electron can cause direct damage when it hits a strand of DNA or indirect damage when the electron reacts with surrounding water molecules producing free radicals or ions. A free radical is a molecule with unpaired electrons that in turn avidly undergoes chemical reactions with surrounding molecules such as DNA. Damage to a cell’s DNA may have several consequences. Single-strand breaks may simply be repaired without detrimental effects to the cell. However, a double-strand break can be repaired incorrectly which may have hereditary effects if it occurs in gonadal cells, or produce malignancy through activation of oncogenes (UNSCEAR, 2011). The amount of damage to an individual or organ depends on the type and amount of radiation and therefore the energy transferred to the tissues, the duration of exposure, the distance from the radiation source, and whether the exposure was continuous or intermittent. The probability of a tissue or organ suffering effects from ionizing radiation is called radiosensitivity. Tissues with high rates of mitosis and undifferentiated cells are the most sensitive. Therefore, bone marrow cells, germ cells, lens cells, and bowel mucosal cells are highly sensitive whereas bone and neural cells are relatively less radiosensitive.

69

70

Section 1  

assessment of the patient with renal disease

Tissue reactions (deterministic effects) Ionizing radiation may cause acute effects which are a result of cell death. They require a threshold dose of manifestation which is not the same for each individual or tissue. These are also known as deterministic or non-stochastic effects and are usually seen at high radiation doses such radiotherapy or radiation accidents. Because a threshold can be identified, radiation protection measures and occupational dose limits can eliminate such effects for workers. Deterministic effects include skin erythema, ulceration and burns, bone marrow depression, cataracts, acute radiation sickness, sterility, and fetal death (Mettler and Upton, 2008b). Healing may occur but permanent damage to tissues may result in fibrosis or necrosis. Although relatively rare, such effects have been observed from medical imaging such as CT scanning and interventional procedures (Shope, 1996; De Olazo Banaag and Carter, 2005; Bogdanich, 2010).

Random (stochastic) effects Damage to a cell’s DNA increases the probability of cancer induction. If germ cells are affected, hereditary effects may be observed. These effects are random and long term and are also known as ‘stochastic’ or ‘non-deterministic effects’. There are two important studies that investigated the direct risk of cancer induction from medical imaging. These have shown a small increased risk of leukaemia and brain cancer after cranial CT in children (Pearce et al., 2012; Mathews et al., 2013). Lifespan studies on atomic bomb survivors indicate increased risk for development of leukaemia and solid cancers including stomach, lung, liver, colon, breast, gallbladder, oesophagus, bladder, and ovary after exposure to ionizing radiation (Richardson et al., 2009; Ozasa et al., 2012).

Estimating risk from medical examinations Cancer risk At the low radiation doses (effective doses < 100–200 mSv) commonly used in medical imaging, statistical limitations make it difficult to estimate cancer risk accurately. Very large numbers of patients would need to be followed up over a lifetime to quantify risk at low doses of radiation exposure (Brenner et al., 2003). Most of the available data on cancer induction was therefore extrapolated from high radiation doses received over a short time interval following the survivors of the detonation of the atomic bombs in 1945 and radiation accidents. The assumption that radiation risk continues in a linear fashion when extrapolated from high to low doses is termed the ‘linear non-threshold model’ (Brenner et al., 2003; National Academy of Sciences, 2006). At present this model is the most widely accepted theory to describe radiation effects at low doses. It assumes that no level of radiation exposure is safe and that even the smallest dose of radiation has a potential to cause harm. Data from atomic bomb survivors and radiation accidents have been used to develop risk models and theories. Such models predict that for a radiation exposure of 0.1 Sv above background radiation, 1 person in 100 would be expected to develop cancer. Similarly, for a radiation exposure of 10 mSv, which is approximately the effective dose that would be received from a CT scan of the abdomen, 1 in 1000 people could be expected to develop cancer (National Academy of Sciences, 2006). At present it is not possible to determine whether a cancer was induced by radiation or any other cause.

In a population, radiation effects especially from low-dose radiation can be obscured by the large number of non-radiation cancers occurring. One in three people will develop cancer from natural causes and one in six people are predicted to die of such cancer (National Academy of Sciences 2006). There is also a considerable latent period between radiation exposure and the induction of cancer. The mean latent period for leukaemia is 7–10  years, 10–15  years for bone tumours, and approximately 20 years for most solid tumours (Mettler and Upton, 2008a). This time lag of cancer induction explains the higher risk for younger patients and the relatively low risk for elderly patients.

Hereditary defects Excess hereditary effects due to radiation exposure have not been demonstrated in humans and were not detected in atomic bomb survivors of Hiroshima and Nagasaki (National Academy of Sciences, 2006; UNSCEAR, 2011). Minor effects are likely to be so small that current studies are unable to differentiate them from naturally occurring mutational effects. However, radiation-induced mutations can theoretically occur in reproductive cells (egg and sperm) resulting in hereditable diseases. Such effects have been previously demonstrated in animal studies, mainly mice (UNSCEAR, 2001).

Risks to the fetus In utero radiation of the fetus causes great anxiety. Children and the fetus are more sensitive to ionizing radiation than adults because of the larger number of dividing cells and the longer lifespan available to develop cancer. Radiation exposure very early on in a pre-implantation stage could lead to an all-or-nothing phenomenon of spontaneous abortion or normal pregnancy. High doses delivered during organogenesis may cause central nervous system abnormalities inducing microcephaly or mental retardation. Malformations are thought to have a threshold of 100–200 mGy absorbed dose which is rarely reached in medical diagnostic imaging. During the third trimester, the organs develop and the risk of congenital malformations is low but there is still a small risk of inducing cancer or leukaemia. However, diagnostic radiological procedures should not normally exceed 50–100 mGy and with such an exposure there is a 99% chance that the child will not develop cancer or leukaemia (Brent, 1984; International Commission on Radiological Protection, 2000).

Absolute risk to a patient An estimation of the subject’s risk from a medical procedure can be made by applying the average effective dose of the procedure and taking the age at the time of exposure, the organ irradiated, and the sex into account (Wall et al., 2011). As risk from radiation exposure is cumulative, radiation doses from several procedures are added to calculate risk. Because of the benefits of imaging in diagnosing illness and injury, a medically indicated imaging study will almost always outweigh the risk associated with it. This is especially apparent when evaluating risks associated with common daily activities such as driving a car. Those activities carry a much higher risk of dying than the cancer risk of 1:1000 associated with a single CT scan of abdomen delivering a dose of 10 mSv (Health Protection Agency, 2001; National Academy of Sciences, 2006; National Safety Council, 2010). Therefore, when assessing radiation exposure at

chapter 10 

Table 10.1  Typical effective doses to adults from selected radiological procedures (National Radiological Protection Board, 1994; Hart et al., 2012) Radiograph

3rd quartile DAP per radiograph (Gy × cm2)

Effective dose estimated from NRPB coefficients (mSv)

Chest PA

0.3

0.03

Abdomen AP

4.4

1.1

IVU

14

2.5

AP = anterior–posterior; DAP = dose area product; NRPB = National Radiological Protection Board; PA = posterior–anterior.

very low radiation doses such as, for instance, for single exposure of a chest radiograph (0.01 mSv), the risk of radiation-induced cancer persists but in practice becomes negligible (Health Protection Agency, 2001).

Quantifying radiation in medical imaging When describing radiation doses received by patients the comparison is with natural background radiation (2.4 mSv per year). Effective doses for most radiological examinations have been published (Mettler et al., 2008). Plain film radiographic examinations vary by a large factor from 0.001 mSv to 10 mSv. CT examinations have relatively high average effective doses between 2 and 20 mSv and interventional procedures can give effective doses between 5 and 70 mSv (Mettler et  al., 2008). For comparison, a round-trip flight from London to New  York will result in a radiation exposure from increased cosmic radiation of about 0.1 mSv (Brenner et al., 2003). From time to time, the Health Protection Agency publishes the results of national patient dose surveys performed in the United Kingdom, for radiological examinations and for CT (Shrimpton et  al., 2005; Hart et  al., 2012). The doses are typically shown as entrance dose, DAP, or DLP for CT, but effective doses can be estimated from tables published by the National Radiation Protection Board (1994). Doses of selected examinations are shown in Tables 10.1 and 10.2.

Managing risk to patients Even though uncertainties remain about the absolute risk of low-dose radiation, the risk is not entirely negligible. Concerns Table 10.2  Typical effective doses to adults from selected CT procedures (Shrimpton et al., 2005; Hart et al., 2012) CT scan

Reference DLP per scan (Gy × cm)

Effective dose estimated from NRPB coefficients (mSv)

Chest (high resolution)

170

2.4

Abdomen

470

7.1

Chest, abdomen, and pelvis

940

14.1

DLP = dose length product; NRPB = National Radiological Protection Board.

ionizing radiation and radiation protection

about radiation and induction of cancer are likely to be in the public focus for the foreseeable future. It is therefore the responsibility of all healthcare professionals to manage and control the risk to protect patients. This point is reinforced by government authority through legislation.

Ionising Radiation (Medical Exposure) Regulations 2000 The IR(ME)R 2000 impose obligations on hospital authorities and clinicians to minimize radiation doses to patients (Department of Health, 2000). The three principles governing patient protection in medical imaging are justification, optimization, and DLRs.

Imaging selection Not all imaging modalities use harmful radiation. The same information might be acquired with an imaging modality that does not use ionizing radiation such as sonography or magnetic resonance imaging (MRI). Sometimes it can be beneficial just to observe the patient for a time period for symptoms to unmask. This is part of a decision process to justify the examination.

Justification For each examination performed there should be careful risk/benefit analysis. It is the responsibility of the clinician as the ‘Referrer’ (term under IRMER) to supply sufficient information for the radiology professional as the ‘Practitioner’ to decide whether the examination is justified. The iRefer: Making the Best Use of Clinical Radiology publication by The Royal College of Radiologists can help clinicians to select appropriate imaging (The Royal College of Radiologists, 2012). For most examinations, the radiographer or ‘Operator’ working within departmental protocols will decide whether to proceed and perform the examination. This is not unique to the risk assessment associated with ionizing radiation. Similar consideration are made, for instance, when considering the risk of bleeding or infection from a diagnostic biopsy, the risk of general anaesthesia, or simply whether it is safe for the patient to be transferred to a different department.

Technique selection and optimization If imaging using ionizing radiation is necessary, the radiology professional will aim to use the lowest possible radiation dose giving an accurate diagnosis. This attempt at optimization refers to the ALARA principle of reducing the radiation dose to ‘as much as reasonably achievable or practicable’. This includes only scanning the area indicated (e.g. renal area rather than the whole abdomen), working under local protocols, adjusting scanning parameters for children, avoiding multiple phase scanning (e.g. non-contrast, arterial, and portal venous phase CT scans), and controlling scan time in fluoroscopy. Optimization also includes quality assurance programmes, regular audits, and planned replacement of elderly or outdated equipment.

Repeat studies Repeat studies can be avoided. Patients with chronic or long-term problems such as young patients, for example, with nephrolithiasis or kidney transplants should be carefully managed as radiation dose is thought to be cumulative over time. These patients should not be repeatedly subjected to CT scans each time they present.

71

72

Section 1  

assessment of the patient with renal disease

Perhaps the clinical question can be answered with an ultrasound scan, reserving CT scanning for problem-solving. Discussion with the radiology professional should be encouraged.

Diagnostic reference levels Under IR(ME)R 2000, DRLs should be established for each standard examination and type of equipment at each imaging department. These levels indicate the radiation dose a typical patient is likely to receive from a standard radiological examination. However, DRLs are not strictly dose limits as the radiation dose received by a patient depends on factors such as their size, the technique selected, and, for fluoroscopic examinations, the screening time. DRLs may be seen as action thresholds and are regularly audited. If they are regularly exceeded, further optimization of technique or particular justification is needed. Radiation doses from interventional procedures can be high, particularly to the skin where the same area is persistently irradiated during an extended procedure. For such procedures, similar action thresholds, in terms of DAP or total screening time, can be used to inform the radiology professional of the possibility of tissue reactions and possible injury that would be likely to result if the same body area is continuously irradiated.

Communication and education Communication between the clinician and the radiology professional is essential and will help to tailor a radiological investigation to the patient’s needs, reducing the amount of radiation. Educating clinicians and patients about the risk of radiation exposure from medical procedures is the aim of campaigns such as ‘Image Gently’ for children and ‘Image Wisely’ for adults (Alliance for Radiation Safety in Pediatric Image Imaging, n.d.; American College of Radiology, n.d.).

Minimizing risk to staff Principles Any measures that reduce patient dose also reduce potential occupational dose to staff. During fluoroscopy, dose can be reduced by reducing the beam size through collimation, using pulsed fluoroscopy and last image hold instead of continuous screening, and by keeping exposure times as short as possible. Although the operator may stay away from the primary beam directed at the patient, a significant dose from scattered radiation is delivered to the operator when performing fluoroscopic examinations in theatre. The inverse square law demonstrates that doubling the distance to the source reduces the dose by a factor of four. Therefore stepping back during exposure or using a remote control will significantly reduce staff doses. Structural shielding and protective clothing such as aprons, thyroid shields, and goggles reduce the dose received by staff. Local rules and procedures are in place to protect staff and should be regularly reinforced through continuous education and training.

Dose monitoring Usually, employees working with ionizing radiation are regularly monitored. The International Commission on Radiological Protection has recommended dose limits for staff working with ionizing radiation and these are reiterated in radiation protection

legislation in the United Kingdom (UK Government, 1999). For employees over 18 years of age the annual limit on effective dose is 20 mSv, for trainees and employees under 18 years 6 mSv, and 1 mSv per year for the general public (UK Government ,1999; Heron et al., 2010). Additional annual limits on dose are set at 500 mSv equivalent dose for extremities, such as the fingers and hands of those involved in interventions, and for each 1  cm2 of skin. An annual limit on dose to the lens of the eye is set at 150 mSv, but is under review at the time of writing, with the expectation that it will be reduced to 20 mSv. In practice, occupational doses are constrained below these annual limits with a requirement to ‘classify’ radiation workers whose doses are expected, following risk assessment of their work, to exceed 3/10 of any annual limit.

Conclusion Medical imaging using ionizing radiation is invaluable in evaluating the sick patient. Radiation should not be feared, especially in emergency situations when there is a clear benefit from scanning. Radiation cannot be entirely avoided but it can be managed by following the ALARA principle. Medical exposures should be justified, duplication of scans should be avoided, and the lowest radiation dose consistent with the diagnostic objective should always be used.

References Alliance for Radiation Safety in Pediatric Image Imaging (n.d.). Image Gently. [Online] American College of Radiology (n.d.). Image Wisely. [Online] . Berry, R. J. (1986). The radiologist as guinea pig: radiation hazards to man as demonstrated in early radiologists and their patients. J R Soc Med, 79, 506–9. Bogdanich, W. (2010). After stroke scans, patients face serious health risks. N Y Times, 31 July. Brenner, D. J., Doll, R., Goodhead, D. T., et al. (2003). Cancer risks attributable to low doses of ionising radiation: assessing what we really know. Proc Natl Acad Sci USA, 100(24), 13761–6. Brenner, D. J. and Hall, E. J. (2007). Computed tomography—an increasing source of radiation exposure. N Engl J Med, 357, 2277–84. Brent, R. L. (1984). The effects of embryonic and fetal exposure to x-ray, microwave and ultrasound. Clin Obstet Gynecol, 26(2), 484–510. De Olazo Banaag, L. and Carter, M. L. (2005). Radionecrosis induced by cardiac imaging procedures: a case study of a 66-year-old diabetic male with several comorbidities. J Invasive Cardiol, 20(8), E2336. Department of Health (2000). The Ionising Radiation (Medical Exposure) Regulations 2000 (together with notes on good practice). London: Department of Health. Hart, D., Hillier, M. C., and Shrimpton, P. C. (2012). Doses to Patients from Radiographic and Fluoroscopic X-Ray Imaging Procedures in the UK-2010 Review. HPA-CRCE-034. Chilton: Health Protection Agency.

Health Protection Agency (2001). National Radiation Protection Board. X-rays How safe are they? [Patient information leaflet] Heron, J. L., Padovani, R., Smith, I., et al. (2010). Radiation protection of medical staff. Eur J Radiol, 76, 20–3. International Commission on Radiological Protection (2000). Pregnancy and medical radiation. ICRP Publication 84. Ann ICRP, 30(1). Mathews, J. D., Forsythe, A. V., Brady, Z., et al. (2013). Cancer risk in 680,000 people exposed to computed tomography scans in childhood

chapter 10 

or adolescence: data linkage study of 11 million Australians. BMJ, 346, f2360. Mettler, F. A., Huda, W., Yoshizumi, T. T., et al. (2008). Effective doses in radiology and diagnostic nuclear medicine: a catalogue. Radiology, 248(1), 254–3. Mettler, F. A. and Upton, A. C. (2008a). Cancer induction and dose-response models. In Medical Effects of Ionising Radiation (3rd ed.), pp. 71–116. Philadelphia, PA: Saunders Elsevier. Mettler, F. A. and Upton, A. C. (2008b). Deterministic effects of radiation. In Medical effects of Ionising Radiation (3rd ed.), pp. 285–388. Philadelphia, PA: Saunders Elsevier. National Academy of Sciences. (2006). Health Risk from Exposure to Low Levels of Ionising Radiation: BEIR VII Phase 2. [Free Executive Summary] [Online] National Radiological Protection Board (1994). Estimation of Effective Dose in Diagnostic Radiology from Entrance Surface Dose and Dose-Area Product Measurements. NRPB-R262. London: National Radiological Protection Board. National Safety Council (2010). Injury Fact. Lifetime Odds of Death for Selected Causes, United States 2006. [Online] Ozasa, K., Shimizu, Y., Suyama, A., et al. (2012). Studies of the mortality of atomic bomb survivors, report14, 1950–2003: an overview of cancer and non-cancer diseases. Radiat Res, 177, 229–3. Pearce, M. S., Salotti, J. A., Little, M. P., et al. (2012). Radiation exposure from CT scans in childhood and subsequent risk of leukaemia and brain tumours: a retrospective cohort study. Lancet, 380(9840), 499–505. Richardson, D., Sugiyama, H., Nishi, N., et al. (2009). Ionizing radiation and leukemia mortality among Japanese atomic bomb survivors, 1950–2000. Radiat Res, 172, 368–82.

ionizing radiation and radiation protection

Shiralkar, S., Rennie, A., Snow, M., et al. (2003). Doctors’ knowledge of radiation exposure: questionnaire study. BMJ, 327(7411), 371–2. Shope, T. B. (1996). Radiation-induced skin injury from fluoroscopy. RadioGraphics, 16, 1195–9. Shrimpton, P. C., Hillier, M. C., Lewis, M. A., et al. (2005). Doses from Computed Tomography (CT) Examinations in the UK—2003 Review. NRPB-W67. Chilton: National Radiological Protection Board. The Royal College of Radiologists (2012). iRefer: Making the Best Use of Clinical Radiology. London: The Royal College of Radiologists.

UK Government (1999). The Ionising Radiations Regulations 1999. SI 1999 No. 3232. London: Her Majesty’s Stationery Office.

United Nations Scientific Committee on the Effects of Atomic Radiation 2001 (2001). Report: Hereditary Effects of Radiation, Annex. New York: UNSCEAR. United Nations Scientific Committee on the Effects of Atomic Radiation 2008 (2010). Report: Sources of Ionising Radiation, Annex A: Medical Radiation Exposures. New York: UNSCEAR. United Nations Scientific Committee on the Effects of Atomic Radiation 2010 (2011). Report: Summery of Low Dose Radiation Effects on Health. New York: UNSCEAR. Wall, B. F., Haylock, R., Jansen, J. T. M., et al. (2011). Radiation Risks from Medical X-Ray Examinations as a Function of Age and Sex of the Patient. HPA-CRCE-028. Chilton: Health Protection Agency.

73

CHAPTER 11

Plain radiography, excretion radiography, and contrast radiography Akira Kawashima and Andrew J. LeRoy Introduction Radiology is an essential component in the evaluation of the urinary tract. For decades, excretion radiography following intravenous (IV) administration of iodinated contrast media (also referred to as intravenous urography (IVU), intravenous pyelography (IVP), and excretory urography (EXU)), has been the primary imaging modality of the urinary tracts. In recent years, other imaging modalities such as ultrasonography (US), computed tomography (CT), and magnetic resonance imaging (MRI) have superseded excretion radiography. However, these cross-sectioned imaging examinations also have their limitations. Plain radiography, contrast radiography, and EXU still remain important in the diagnosis of some urinary tract diseases (Table 11.1). Plain radiography is both a primary abdominal examination technique and an essential initial component of subsequent excretion and contrast radiographic studies. Contrast radiography is obtained following direct injection of contrast material into the urinary tracts in an antegrade or retrograde fashion. The urographic imaging sequence with IV contrast material or direct contrast injection is designed to optimize depiction of specific portions of the urinary tract during maximal contrast opacification, and a tailored urographic study may provide diagnostic detail beyond the current capabilities of other imaging modalities (Hattery et al., 1988; Banner, 2001; Dyer et al., 2001) This chapter will outline basic uroradiologic imaging with each subsection providing information on technique, indications, complications, and normal anatomical findings.

image prior to IV or direct contrast injection for conventional urography. This scout radiograph helps establish appropriate radiographic exposure technique factors, evaluate the patient’s positioning, and confirms that residual oral contrast material from previous studies has been eliminated from the gut. The indications for the examinations and patient risk factors should be reviewed carefully. A  menstrual history should be obtained in women of childbearing age. Radiation exposure should always be limited appropriately. The presence of a fetal skeleton on the scout image justifies postponing the planned contrast study. A  scout image is imperative prior to contrast studies because stones and calcifications can be obscured by contrast material in the urinary tract on subsequent images.

Anatomical landmarks Several abdominal and pelvic organs can be identified on plain radiographic images. Organs such as the liver, spleen, kidneys, and bladder can be outlined on the image based on the contrasting densities of the organs themselves compared to surrounding retroperitoneal or mesenteric fat. The outline of the kidneys is helpful for assessing renal size, focal cortical scarring or mass, and for delineating intrarenal calcifications. The psoas muscle margins usually can be seen on the plain radiograph (see Fig. 11.1), but may be obscured by retroperitoneal fluid or mass. Plain films can demonstrate abnormalities of the bowel gas pattern (e.g. ileus, bowel obstruction, and faecal impaction) and also allow for evaluation of skeletal anomalies, such as sacral agenesis and spinal dysraphism.

Plain radiography

Excretion radiography

Plain radiography of the abdomen and pelvis is most often referred to as a KUB (Kidneys, Ureters, and Bladder) film or an abdominal ‘flat plate’ (Fig. 11.1), and can provide important information about radiopaque urinary calculi (Figs 11.2 and 11.3), calcifications, renal masses (Fig. 11.2), and renal size and contour (Dyer et  al., 1998; Barbaric and Pollack, 2000). A  combination of projection radiographs and non-contrast coronal tomography is valuable in outlining the kidneys and visualizing calculi. An anteroposterior (AP) projection radiograph of the abdomen in supine position may also be referred to as a scout (preliminary)

Excretion radiography is frequently used as a non-invasive screening procedure for the entire urinary tract because it provides both anatomic and functional information (Friedenberg and Harris, 2000).

Patient preparation A thorough pre-examination ‘bowel prep’ with a mild laxative (e.g. 10 ounces (28 g) of magnesium citrate solution) the night before the procedure is advisable to rid the colon of stool, which may interfere with clear visualization of the urinary tract. Withholding of fluid

Table 11.1  Glossary of terminology in uroradiography Terms (abbreviation)

Definitions

Plain abdominal radiography, KUB, plain film, conventional radiograph, flat plate, scout

Film-screen or digital projection radiograph without use of iodinated contrast medium

Excretion radiography, excretory urography (EXU), intravenous Imaging of the kidneys and urinary tracts before and after the IV administration of iodinated urography (IVU), intravenous pyelography (IVP) contrast medium Contrast radiography Retrograde pyelography

Imaging of the upper urinary tracts before and after the direct retrograde injection of contrast medium

Antegrade pyelography

Imaging of the upper urinary tracts before and after the direct antegrade injection of contrast media with percutaneous needle puncture of a calyx

Cystography Static cystography

Imaging of the bladder before, during and after the direct contrast administration either in a retrograde fashion through a transurethral Foley catheter or in an antegrade manner through a suprapubic tube

Voiding cystourethrography (VCUG)

Imaging of the bladder and urethra during micturition after the direct administration of contrast medium into the bladder. The procedure is monitored under fluoroscopy and recorded with spot films or video recording.

Retrograde urethrography (RUG) Loopography ileal conduit study

Opacification of the ileal loop before, during and after the direct injection of contrast medium with reflux of into the upper urinary tracts. The procedure is monitored under fluoroscopy and recorded with spot and overhead films.

Pouchgraphy

(A)

(B)

(C)

Fig. 11.1  Plain radiography and excretion radiography. (A) Plain radiograph of the abdomen and pelvis reveals normal contour of the kidneys and the medial margin of the psoas muscles (arrows). L = liver; LK = left kidney; RK = right kidney; UB = urinary bladder; UT = uterus. (B) Magnified view of the left kidney with abdominal compression 8 minutes after IV contrast material administration. Calyx (white arrows), infundibulum (I), renal pelvis (RP), proximal ureter (U). The calyx which project posteriorly is seen en face (white arrowhead). Normal fold of the ureter at the ureteropelvic junction (black arrow). (C) Abdominal radiograph after release of ureteral compression reveals normal ureters and urinary bladder (UB). The left ureter at the ureteropelvic junction (arrow) is better distended and straight. UT = uterus.

76

Section 1  

assessment of the patient with renal disease demonstrate the calyces, pelves, and proximal ureters (pyelographic phase) (Fig. 11.1). The ureters are generally well visualized on the 10-minute radiograph after release of external compression. Films of the bladder (often including a post-void film) conclude the examination.

Diuretic excretion radiography

Fig. 11.2  Patient with renal stones and renal cell carcinoma. Plain radiograph demonstrates bilateral renal calculi (short arrows) with the largest over the left renal hilum (arrowhead). A round mass is projected over the lower pole of the left kidney (long arrow).

and solid food overnight improves renal concentration and excretion of the contrast medium. An empty stomach also decreases the possibility of aspiration of solid food by a patient vomiting after IV contrast administration.

Filming procedure A plain abdominal radiograph prior to the administration of IV contrast media is essential. After contrast medium is injected, lower abdominal compression is applied anteriorly. Approximately 2–3 minutes after the contrast material has been administered, nephrotomograms (usually three) are obtained to visualize the renal parenchyma (nephrographic phase) (Hattery et  al., 1988). Radiographs obtained 8–10 minutes after the contrast injection

(A)

(B)

This modification of excretion radiography is reserved for those patients suspected of having volume-dependent hydronephrosis in whom the initial (dehydrated) excretion radiography revealed no obstruction. With a brisk diuresis after furosemide injection (Lasix®, 20 mg IV), a borderline ureteropelvic junction or ureteral narrowing may become inadequate to allow free flow of the increased urine volume and may reveal in hydronephrosis and even reproduce or worsen the pain associated with obstruction.

Contrast media IV iodinated contrast material is excreted by glomerular filtration with little or no tubular secretion, with subsequent concentration in the tubules and collecting ducts and then progressive opacification of the urinary tract. Urine concentrations are determined by the dose administered, the glomerular filtration rate, and the renal tubular function. As the urine is concentrated in the renal collecting ducts, the relative concentration of the contrast media is enhanced 50–100-fold.

Contrast dosage Contrast medium can be administered either as a bolus injection or as a drip infusion. The bolus injection allows for a rapid and dense nephrogram when images are obtained 2–3 minutes after the injection. In the average-size adult, 60–100 mL of contrast medium is usually administered.

(C)

Fig. 11.3  Patient with renal stones and transitional cell carcinoma. (A) Magnified view of plain radiograph demonstrates two adjacent stones projected over the lower pole of the left kidney. (B) Excretory radiograph shows the left renal stones in the lower intrarenal collecting system. Note irregular narrowing of the left renal pelvis (arrow) with associated caliectasis. Kidneys are congenitally malrotated bilaterally. (C) Magnified view of antegrade pyelogram following a percutaneous needle puncture of a lower pole calyx shows a large filling defect in the mid to upper left renal pelvis with urothelial irregularity (arrows), characteristics of urothelial tumour. The two stones (arrowheads) lie in the lower left collecting system.

chapter 11 

Adverse effects The majority of adverse effects of iodinated contrast media are mild or moderate and only observation, reassurance, and support are required. However, the more severe adverse effects are often preceded by mild or moderate symptoms or prodrome. Reactions to contrast agents usually occur within 20 minutes following IV injection. The most threatening contrast reactions are those unanticipated events that result in sudden compromise of critical body functions, particularly systemic anaphylactic reactions or profound cardiovascular collapse. The widely used newer non-ionic, low-osmolar contrast media (LOCM) offer a slightly more comfortable, safer examination to patients who are sensitive to the older high-osmolar contrast media (HOCM) and produce fewer adverse effects. The risk of rare life-threatening reactions is not, however, eliminated by the use of these LOCM (Palmer, 1988; Katayama et  al., 1990). Appropriate training and vigilance by healthcare workers are therefore necessary in clinical areas where contrast media are administered. Anaphylactoid (idiosyncratic) reactions are generally classified as (1) mild, (2) moderate, and (3) severe (Cohan et al., 2010). Minor side effects, do not usually require therapy. Nausea and vomiting are the most common adverse reactions. Other minor reactions include rash, itching, hives (less than four), swelling, headache, dizziness, shaking, nasal congestion, pallor, flushing, chills, sweats, and anxiety. These symptoms are usually self-limiting, but observation is required to confirm resolution and lack of progression. Only supportive treatment is necessary. Patient reassurance is usually helpful. Moderate reactions include tachycardia, bradycardia, hypertension, mild hypotension, vasovagal reactions, dyspnoea, bronchospasm, wheezing, and pronounced cutaneous reaction such as extensive hives (four or more) or diffuse erythema. The observed clinical signs and symptoms of moderate reactions should be considered as indications for immediate treatment (Table 11.2). These situations require close, careful observation of possible progression into a life-threatening event. Patients with moderate symptoms such as systemic urticaria or facial oedema are treated with antihistamines and subcutaneous epinephrine. Severe types of reactions include laryngeal oedema, and/or, severe hypotension, require small doses of 1:10,000 dilution of epinephrine intramuscularly. The most severe reactions include convulsions, arrhythmias, unresponsiveness, and cardiopulmonary arrest. Prophylactic corticosteroids are strongly recommended for ‘at-risk’ patients who require contrast media (Lasser, 1988; Greenberger and Patterson, 1991). They should also receive LOCM. However, no regimen has eliminated repeat reactions completely. Pre-testing is not predictive, and may itself be dangerous, so is not recommended (Yamaguchi et al., 1991). Alternative means of imaging to obtain the required information should be considered before embarking on this potentially dangerous path.

Contrast-induced acute kidney injury (See Section 11.) IV contrast media may aggravate pre-existing renal dysfunction in a small percentage of patients. The mechanism is not known. There is no standard definition for reporting contrast media-induced nephrotoxicity. The definition of ‘significant’ changes varies considerably among studies. Acute contrast-induced nephrotoxicity has been defined as an increase in the baseline creatinine values of 20– 50% or an absolute increase from 0.5 mg/dL to 1.0 mg/dL (Katzberg,

plain, excretion, and contrast radiography

1997). Most cases of contrast media-induced nephrotoxicity are unrecognized because the serum creatinine concentration is not systematically checked after procedures. This nephrotoxicity is usually self-limiting. Serum creatinine usually peaks within 3–5 days, and usually returns to baseline within 10–14 days (Katzberg, 1997). Nephrotoxicity is extremely rare when kidney function is normal. Patients with chronic kidney disease caused by diabetic nephropathy appear to be the most at risk of developing contrast-induced nephropathy. They are more susceptible when dehydrated and when exposed to relatively large volumes of contrast agents. Therefore in high-risk patients, hydration is advised. Because treatment options are limited once oliguric renal failure has developed, most clinical effort has been aimed at prevention of contrast-induced acute kidney injury. Alternatively, CT urography with oral or IV hydration and with reduced contrast dose could be performed. In addition to active hydration, several studies have suggested various pharmacologic agent administrative regimens may be of benefit in preventing contrast-induced acute kidney injury, including N-acetylcysteine (Tepel et al., 2000) and bicarbonate. In patients with underlying chronic renal failure who undergo regular dialysis, contrast media is readily cleared by dialysis because contrast media molecules are not protein bound. Unless there is significant underlying cardiac dysfunction, or very large volumes of contrast media have been administered, there is no need for urgent dialysis. Patients with multiple myeloma may occasionally develop acute kidney injury following IV contrast administration. Dehydration should always be avoided in patients with myeloma. In diabetic patients with normal renal function taking the oral agent metformin (Glucophage®), discontinuing metformin is advised for at least 48 hours after administration of IV contrast material to minimize the risk of developing metformin-associated lactic acidosis. Inadvertent subcutaneous extravasation of IV administered contrast medium is not uncommon. Such extravasation may result in local pain, erythema, and swelling. These symptoms usually resolve with local therapy including both warm and cold compression and elevation of the affected extremity (Cohan et  al., 2010). An immediate surgical consultation is indicated when patients develop increased swelling or pain after 2–4 hours, altered tissue perfusion, changes in sensation in the affected extremity, or skin ulceration.

The normal excretion urogram The renal parenchyma is best assessed during the nephrographic phase of urography. The normal kidney may range from 9 to 13 cm in cephalocaudal dimension depending on sex and age. The left kidney is frequently larger that the right kidney by approximately 0.5  cm. The kidneys have sharp margins and a smooth contour. Congenital fetal lobulation, a common normal variant, can be differentiated from scars, prior renal infarction or inflammation by their smooth contour and regular spacing and relationship to normal calyces and is often bilateral. Renal parenchyma measures 3–3.5 cm in thickness in the polar regions and 2–2.5 cm in the interpolar regions. The number of the calyces varies considerably. Each calyx (minor calyx) is deeply cupped and surrounds one papilla. The peripheral portion of the calyx is called the fornix. A group of calyces, termed compound calyces, drains two to four papillae and is frequently seen in the polar regions. Two or more infundibula (major calyces), each leading to single or multiple calyces, arise

77

78

Section 1  

assessment of the patient with renal disease

Table 11.2  Reactions to iodinated contrast material and treatment Type of reactions

Treatment

Hives

Contrast injection discontinued if not completed. Minimal (less than four): observation If diffuse or symptomatic, diphenhydramine PO/IV 25–50 mg If severe, epinephrine IV (1:10,000) 0.1 mL/kg slow push over 2–5 min

Angio-oedema

Closely monitor the airway, O2 by mask Isotonic IV fluids Epinephrine SQ: (1:1000) 0.1–0.3 mL(0.1–0.3 mg) If severe or associated with hypotension or airway compromise, epinephrine IV (1:10,000) 0.1 mL/kg slow push over 2–5 min, up to 3 mL/dose. Repeat in 5–30 min as needed

Diffuse erythema

Isotonic IV fluid, 0.9% normal saline or lactated Ringer’s 1–2 L rapidly If hypotensive: epinephrine IV (1:10,000), 0.1 mg (1 mL) slowly. Repeat up to 1 mg Hydrocortisone (Solu-Cortef®) IV 200–300 mg push over 1–2 min

Bronchospasm

O2 10 L/min by mask Isotonic IV fluids Beta-agonist inhalers, metaproterenol (Alupent®) and albuterol (Proventil®), 2 puffs and inhale. Repeat as necessary Epinephrine SQ (1:1000), up to 0.3 mL. Repeat up to 1 mg total. If hypotensive, add epinephrine (1:10,000), slowly IV 0.1 mg (1 mL) slowly Repeat up to 1 mg total

Laryngeal oedema

O2 6–10 L/min by mask Isotonic IV fluids Epinephrine IV (1:10,000), 0.1 mg (1 mL) slowly. Repeat up to 1 mg total

Pulmonary oedema

Elevate torso O2 10 L/min by mask Furosemide (Lasix®) 10–40 mg IV over 2 min Morphine 1–3 mg IV Hydrocortisone (Solu-Cortef®) IV 200–300 mg push over 1–2 min Transfer to intensive care unit or emergency department

Hypotension with tachycardia (anaphylactic shock)

Legs elevated 60º or more, or Trendelenburg position O2 by mask Isotonic Ringer’s lactate or normal saline IV fluids Epinephrine IV (1:10,000) 0.1 mg (1 mL) slowly. Repeat up to 1 mg total Transfer to intensive care unit or emergency department

Hypotension with bradycardia (vagal reaction)

Legs elevated 60º or more, or Trendelenburg position O2 by mask Isotonic lactated Ringer’s or normal saline IV fluids, 1–2 L rapidly Atropine 0.5–1.0mg IV push slowly. Repeat up to 3 mg

Hypertensive crisis

O2 10 L/min by mask Secure IV access Furosemide (Lasix®) 40 mg slowly over 2 min. Nitroglycerine 0.4 mg tablet sublingual. Repeat every 5–10 min Transfer to intensive care unit or emergency department

Seizures/convulsions

O2 6–10 L/min by mask Lorazepan (Ativan®) 2–4 mg IV Transfer to emergency room for further evaluation and workup

Modified from Cohan et al. (2010).

separately from the renal pelvis. Conventionally, all branches from the renal pelvis, whether single or multiple, are termed infundibula. Normal infundibula are straight without bowing or displacement. The renal pelvis sometimes appears to be outside of the confines of

the kidneys, where it often has a distended appearance (the extrarenal pelvis). The upper ureter usually begins as a smooth extension from the renal pelvis and descends lateral to the transverse processes of

chapter 11 

plain, excretion, and contrast radiography

the upper lumbar vertebrae. The middle third of the ureter is usually superimposed on the transverse processes of the lower lumbar vertebrae. The ureter crosses anterior to the iliac vessels at a slightly higher position on the right than the left. The distal ureter courses posterolaterally and then anteromedially to enter the bladder. Peristaltic activity may change the size and shape of the calyces, pelvis, and ureter from image to image. When the bladder is progressively distended, it is smoothly marginated and appears roughly spherical. On a post-void image, the mucosal pattern of the bladder is frequently identified.

CT urography Cross-sectional imaging studies including US, CT, and MRI are now used more often to assess the renal parenchyma because it has been shown that the sensitivity of detection of small renal masses with excretory urography is greatly decreased. With introduction of helical CT, CT urography has been increasingly used as a more definitive study for the evaluation of the urinary tract. The renal parenchyma is evaluated with CT scans before and after IV contrast administration, and then the collecting system is visualized by reformatted images generated from thin-section multidetector helical CT images obtained during excretory phase of contrast enhancement (Chow and Sommer, 2001; Caoili et al., 2002; Kawashima et  al., 2004a). CT urography has replaced standard excretion radiography in the majority of patient evaluations with urologic indications.

Contrast radiography Retrograde pyelography Retrograde pyelography is the opacification of the ureter and pelvicaliceal system by the direct retrograde injection of contrast media at the time of cystoscopy (Fig. 11.4) (Imray et al., 2000). The examination may be done in a cystoscopic suite, or the ureter may be cannulated and the patient may be subsequently brought to the radiology department for the examination. The examination is best performed with fluoroscopy and appropriate spot and overhead images. When a urothelial lesion is suspected, subsequent endoscopy with brushing or biopsy of the lesion for a histological diagnosis is performed under fluoroscopic control. Retrograde pyelography is performed to investigate lesions of the ureter and renal collecting system that cannot be defined adequately by less invasive imaging or, to visualize the collecting systems and ureters when iodinated contrast media is contraindicated. The procedure is performed with sterile technique and is contraindicated in a patient with a urinary tract infection. Retrograde pyelography cannot be performed when patients cannot or should not undergo cystoscopy (e.g. patients recovering from recent bladder or urethral surgery). The procedure may be impossible or incomplete when it is difficult to cannulate the ureter (e.g. patients with very large prostates). Delayed images can be obtained after retrograde pyelography to evaluate drainage of the collecting system. If significant obstruction is identified during retrograde pyelography then ureteral stent placement should be considered to avoid the risk of bacterial spread into the upper tract above the obstruction. Other complications of retrograde pyelography include ureteral perforation and contrast reaction. The most common ureteral injury during retrograde pyelography is perforation, occurring

Fig. 11.4  Retrograde pyelogram reveals normal contrast filling of the right ureter and intrarenal collecting system.

during advancement of the catheter or guide wire. These injuries are usually managed with either observation or stent placement depending upon the extent of the injury. Up to 10–15% of contrast media can be absorbed during retrograde pyelography. Caution is therefore advised in patients with a known contrast allergy. Fluoroscopic monitoring of retrograde pyelography is helpful to avoid excess contrast volume injection, reducing the amount of extravasation from the distended upper collecting system.

Antegrade pyelography Antegrade pyelography is performed to visualize the upper tracts and to delineate the site or nature of upper urinary tract obstruction when excretion radiography is unsatisfactory, retrograde pyelography cannot be performed (e.g. ureteral diversion), or alternative imaging techniques (US, CT, and MRI) are not definitive. This technique is indicated to determine whether a dilated collecting system is obstructed or not when there is renal dysfunction after kidney transplantation. Pyelography is an essential component of upper urinary tract urodynamic testing (Whitaker test). Antegrade pyelography is contraindicated in patients with uncorrectable bleeding diatheses, diffuse skin infection over the puncture site, or anatomic anomalies which preclude safe renal puncture. Non-dilated collecting systems are not a contraindication, but it is much more difficult to puncture a non-dilated collecting system percutaneously. A peripheral calyx or the renal pelvis is punctured percutaneously with a 21-guage, thin-walled needle from a posterior or posterolateral approach (Fig. 11.3). Renal localization is provided by means of contrast excreted after an IV injection or, in the event of a non-visualizing kidney, with US guidance. The procedure is carried out under fluoroscopic control and spot images are obtained.

79

80

Section 1  

assessment of the patient with renal disease

The side effect of this procedure is inadvertent puncture of adjacent intra-abdominal structures. Although puncture of the renal vein, kidney parenchyma, liver, spleen, or colon is possible, few complications ensue because of the small size of the needle.

Cystography Static cystography Static cystography provides information on bladder volume, contour, position, and integrity. Static cystography is performed to assess suspected bladder rupture, to demonstrate bladder diverticula, delineate vesicoenteric fistulae, and to assess postoperative healing following bladder or distal ureteral surgery. The normal cystogram The distended bladder is a smooth-walled organ with either a round or an oval shape. The oval-shaped bladder is often aligned vertically in the female and horizontally in the male. In the newborn the bladder lies above the symphysis pubis and descends as the child grows. In the older child and young adult the bladder lies at or below the level of the symphysis pubis.

Voiding cystourethrography Voiding cystourethrography (VCUG) provides anatomic information about the lower urinary tract during the physiologic act of micturition. VCUG is performed to diagnose vesicoureteral reflux, to assess bladder emptying, and to evaluate the urethra for posterior urethral valves in the infant male, urethral stricture disease and urethral diverticula in female patients. VCUG is useful in assessing certain types of voiding dysfunction (e.g. detrusor-external sphincter dyssynergia and neuropathic bladder) and demonstrating reflux into an ectopic ureter which inserts into the urethra. The bladder is filled with contrast material using a transurethral catheter as for a static cystogram. Once filled, the older child or adult patient is asked to void in the upright position. The procedure is monitored with videofluoroscopy and recorded with either spot images or video recording. In male patients, the voiding images should be obtained with the pelvis in a 45º oblique position similar to retrograde urethrography (RUG), so that the entire length of urethra is better demonstrated (Fig. 11.5.)

Fig. 11.5  Voiding cystourethrogram demonstrates normal bladder and female urethra (arrow). No vesicoureteral reflux.

or small bowel segment (pouch) is the most common method of establishing permanent urinary diversion (Spring and Deshon, 2000; Banner, 2001). The isolated but otherwise intact bowel loop serves as a simple conduit for urinary flow, transporting urine outward toward the stoma in a continuous, rhythmic, isoperistaltic manner. The detubularized pouch, on the other hand, lacks the contractivity to propel urine to the outside, thus becoming a reservoir (e.g. neobladder, continent diversion) rather than a conduit.

Retrograde urethrography RUG provides detailed visualization of the anterior urethra in the male. Unlike a voiding cystourethrogram, RUG often incompletely visualizes the posterior urethra because of the resistance to retrograde flow provided by the external urethral sphincter. Complete evaluation of the entire urethra often requires both procedures, which may be performed at separate intervals. RUG is rarely indicated in female patients. RUG is most frequently indicated to assess suspected or known urethral stricture disease, suspected urethral trauma, and to demonstrate urethral diverticula, fistulae, and neoplasms (Kawashima et al., 2004b). Retrograde urethrograms should be performed in all patients with pelvic trauma prior to cystography in order to minimize further urethral injury with planned bladder drainage catheter insertion.

Loopography and pouchography After cystectomy, anastomoses of the ureters to an isolated intact segment (loop) of ileum, transverse colon, or a detubularized large

Fig. 11.6  Loopogram shows contrast filling of the ileal conduit (IC) with reflux into the distal ureters bilaterally. A large filling defect in the right lower intrarenal collecting system with urothelial irregularity (arrow) is characteristics of urothelial tumour.

chapter 11 

Radiographic examination of loops or pouches is referred to as loopography or pouchgraphy, respectively. A loopogram is performed to visualize the bowel conduit and subsequently the upper urinary tracts by reflux of injected contrast media (Fig. 11.6). This procedure is utilized in postoperative patients with progressive renal failure to assess conduit integrity and to exclude ureteral obstruction. In such patients, the absence of reflux may indicate ureteral anastomotic obstruction. Alternatively, neobladder and continent diversion ureteral anastomoses are often created with antirefluxing techniques. Therefore, the absence of reflux into the ureters on pouchograms may not indicate urinary obstruction. Loopography is essential to delineate the anatomy of a urinary diversion and probable site of ureteral implantation prior to any planned endourologic or interventional procedures on the diverted ureters.

References Banner, M. P. (2001). Diagnostic uroradiology. In P. Hanno, S. B. Malkowicz, and A. J. Wein (eds.) Clinical Manual of Urology, pp. 87–133. New York: McGraw-Hill. Barbaric, Z. L. and Pollack, H. M. (2000). Abdominal plain radiography. In H. M. Pollack, B. L. McClennon (ed.) Clinical Urography (Vol. 1), pp. 67–146. Philadelphia, PA: W. B. Sanders. Caoili, E. M., Cohan, R. H., Korobkin, M., et al. (2002). Urinary tract abnormalities: initial experience with multi-detector row CT urography. Radiology, 222, 353–60. Chow, L. C., Sommer, F. G. (2001). Multidetector CT urography with abdominal compression and three-dimensional reconstruction. AJR Am J Roentgenol, 177, 849–55. Cohan, R. H., Jafri, S., Choyke, L.P., et al. (2010). Manual on Contrast Media, Version 7. Reston, VA: American College of Radiology. Dyer, R. B., Chen, M. Y., and Zagoria, R. J. (1998). Abnormal calcifications in the urinary tract. Radiographics, 18, 1405–24. Dyer, R. B., Chen, M. Y., and Zagoria, R. J. (2001). Intravenous urography: technique and interpretation. Radiographics, 21, 799–821.

plain, excretion, and contrast radiography

Friedenberg, R. M. and Harris, R. D. (2000). Excretory urography. In H. M. Pollack and B. C. McClennon (eds.) Clinical Urography (Vol. 1), pp. 147–281. Philadelphia, PA: W.B. Sanders. Greenberger, P. A. and Patterson, R. (1991). The prevention of immediate generalized reactions to radiocontrast media in high-risk patients. J Allergy Clin Immunol, 87, 867–72. Hattery, R. R., Williamson, B., Jr., Hartman, G. W., et al. (1988). Intravenous urographic technique. Radiology, 167, 593–9. Imray, T. J., Lieberman, R. P., and Pollack, H. M. (2000). Retrograde pyelography. In H. M. Pollack, B. L. McClennon (eds.) Clinical Urography (Vol. 1), pp. 282–302. Philadelphia, PA: W. B. Sanders. Katayama, H., Yamaguchi, K., Kozuka, T., et al. (1990). Adverse reactions to ionic and nonionic contrast media. A report from the Japanese Committee on the Safety of Contrast Media. Radiology, 175, 621–8. Katzberg, R. W. (1997). Urography into the 21st century: new contrast media, renal handling, imaging characteristics, and nephrotoxicity. Radiology, 204, 297–312. Kawashima, A., Vrtiska, T. J., LeRoy, A. J., et al. (2004a). CT urography. Radiographics, 24(Suppl. 1), S35–54 Kawashima, A., Sandler, C. M., Wasserman, N. F., et al. (2004b). Imaging of urethral disease: a pictorial review. Radiographics, 24(Suppl. 1), S195–216. Lasser, E. C. (1988). Pretreatment with corticosteroids to prevent reactions to i.v. contrast material: overview and implications. AJR Am J Roentgenol, 150, 257–9. Palmer, F. J. (1988). The RACR survey of intravenous contrast media reactions. Final report. Australas Radiol, 32, 426–8. Spring, D. B. and Deshon, G. E., Jr. (2000). Radiology of vesical and supravesical urinary diversions and orthotopic bladder replacements. In H. M. Pollack and B. L. McClennon (eds.) Clinical Urography (Vol. 1), pp. 357–77. Philadelphia, PA: W. B. Sanders. Tepel, M., van der Giet, M., Schwarzfeld, C., et al. (2000). Prevention of radiographic-contrast-agent-induced reductions in renal function by acetylcysteine. N Engl J Med, 343, 180–4. Yamaguchi, K., Katayama, H., Takashima, T., et al. (1991). Prediction of severe adverse reactions to ionic and nonionic contrast media in Japan: evaluation of pretesting. A report from the Japanese Committee on the Safety of Contrast Media. Radiology, 178, 363–7.

81

CHAPTER 12

Intervention Steven Kennish Introduction Imaging technology and interventional techniques are constantly being improved and refined. The percutaneous insertion of a nephrostomy drain was the first of these and remains the most commonly performed and useful intervention. Natural extensions of this technique include both percutaneous stone removal and antegrade ureteric stenting. Nephrologists must be aware of what these fundamental procedures entail, their indications, the techniques involved in performing them, and their limitations and complications.

Percutaneous nephrostomy Introduction Radiologically guided percutaneous drainage of the kidney was first described in 1955, and has become the key procedure in the management of renal tract obstruction (Goodwin et al., 1955).

Indications Indications vary depending on local expertise and preferences. The main reasons for inserting a percutaneous nephrostomy (PCN) are given in Box 12.1. The nephrologist is most likely to become involved in organizing a PCN for a patient initially referred with renal failure in whom subsequent investigations have demonstrated an obstructive ‘postrenal’ cause. This may be congenital, developmental, stone related, malignant, or iatrogenic. The acutely obstructed kidney becomes ischaemic, and a PCN is a quick and effective method of decompressing the collecting system, draining infected urine and relieving pressure on the nephrons (Rickards et al., 1993; Heyman et al., 1997). A PCN is often performed before an exact diagnosis has been made. Bilateral renal obstruction, obstruction of a single functioning kidney, systemic features of sepsis, or refractory hyperkalaemia necessitate urgent intervention. Nephrologists will, however, do as much as they can to improve the patient’s condition to make the procedure safe and easier. This will include dialysis, blood transfusion, and correction of acidosis. There is little evidence that a PCN is superior to a retrograde stent as the primary treatment, even for an infected obstructed collecting system, but it does avoid a general anaesthetic and ureteric manipulation (Türk et al., 2012). A urology opinion should be sought as to the most appropriate management. In practice, a PCN is usually preferred, especially in the sick patient. Malignant ureteric obstruction may be secondary to infiltrative primary tumour (often of the bladder or prostate), malignant nodal enlargement (including lymphoma), metastatic disease, or

retroperitoneal fibrosis. PCN drainage of the kidneys preserves nephrons and can prevent the development of urosepsis, Iatrogenic ureteric occlusion or injury is an infrequent complication of major laparoscopic surgery (Palaniappa et al., 2012). Percutaneous urinary diversion allows healing of a leak at a site of injury, or a defective ureteroileal anastomosis. Fistulating disease within the pelvis, incontinence, and irritative bladder symptoms post radiotherapy are also recognized indications for urinary diversion necessitating bilateral PCN insertion. The transplant kidney may require a PCN in the early postoperative period if the ureteric stent has become occluded, or if there is urine extravasation.

Techniques PCN insertion techniques are very much tailored by individual radiologists. The basic principles of the procedure are outlined in Box 12.2 The risks quoted for haemorrhage, infection, and procedural failure should be based on individual operator audit data (see Table 12.1). Pre-procedural analgesia is recommended, and the patient’s coagulation profile should be normal. Antibiotics are administered if there is a clinical suspicion of infection or there are risk factors such as stone disease, recent instrumentation or an indwelling foreign body such as a ureteric stent. PCN insertion is usually performed with ultrasound and fluoroscopic guidance with the patient in either the prone or supine/ oblique position. A  posterior axillary line approach allows a puncture along the avascular plane of Brödel, a natural watershed territory between the dorsal and ventral branches of the renal arteries, so as to minimize vascular injury (Dyer et  al., 2002). The transplant kidney should be approached from a skin puncture lateral to the iliac fossa incision, to try to avoid crossing the peritoneum. Ultrasound allows for the visualization and avoidance of nearby intra-abdominal structures. The dark anechoic fluid-filled hydronephrotic pelvicalyceal system is very clearly demonstrated against the echo-bright renal sinus fat (Fig. 12.1). A subcostal approach avoids crossing the pleura, which is painful and runs the risk of complications. Most radiologists use ultrasound alone to guide the initial puncture. Typically a thin 22-G trocar needle or a larger 19-G needle with a 15-G outer sheath can be used for calyceal puncture. In expert hands the hydronephrotic pelvicalyceal system is usually accessed safely by a single puncture while the subject holds their breath. The calyces of the non-hydronephrotic kidney are much more of a challenge.

chapter 12 

Box 12.1  PCN indications ◆ Relief of ureteric obstruction

intervention

Table 12.1  Complication rates associated with PCN insertion (Lewis and Patel, 2004; Wah et al., 2004)

◆ Drainage of infected urine

Complication

Rate

◆ Pain relief in ureteric colic

Haemorrhage

< 4%

Urosepsis

< 4%

◆ Urinary diversion to allow the healing of leaks or fistulae ◆ Preparation for percutaneous renal tract surgery.

Limitations Spinal deformities and ectopic kidneys make the procedure difficult, but it is uncommon for overlying pleura or bowel to preclude a percutaneous approach completely. The only absolute contraindication to PCN is an uncorrected bleeding diathesis, and whilst a haematology opinion is useful, retrograde stents may be more appropriate. Uraemic and anaemic patients have poor platelet function so these problems should be dealt with before proceeding. Extra-long puncture needles have been developed to deal with the morbidly obese patient. Ultrasound transducer-mounted needle guides are helpful in securing a trajectory predicted by an on-screen pre-calibrated tramline target (Fig. 12.1). A single-step, 8-Fr needle trocar-mounted drain insertion may be appropriate in the grossly hydronephrotic infected kidney, reducing manipulation and therefore the risk of bacteraemia. A 0.038-inch J-tipped guidewire can be adequately visualized with ultrasound alone, if fluoroscopy is not immediately available, or undesirable, such as in the pregnant patient (Fig. 12.2). Often a retrograde ureteric stent is preferable. Children require general anaesthesia.

Complications PCN is a safe procedure. Nevertheless, as with any intervention, complications can occur. PCN insertion performed at night and by radiologists who do not regularly perform the procedure have higher complication rates (Lewis and Patel, 2004; Wah et al., 2004). Frank haematuria is normal after nephrostomy insertion. It usually clears within 24–48 hours, but often, reassuringly, before the patient leaves the radiology suite.

Renal pelvic injury

< 1%

PCN dislodgement

< 15%

Persistent or worsening haematuria is an indication for a nephrostogram. If the nephrostomy tube has migrated and drain sideholes lie within the vascular renal parenchyma, then a replacement tube to tamponade the track is needed. Pseudoaneurysms and arteriovenous fistulae are uncommon, and the need for embolization for bleeding post PCN is exceptionally rare (Farrell and Hicks, 1997). Despite best efforts to avoid manipulation within, and overdistension of an infected collecting system, systemic urosepsis can arise due to pyelovenous backflow. Inadvertent enteric injury is rare and minimized by the use of ultrasound (Zagoria and Dyer, 1999). A permanent external urine drainage bag is never popular with a patient, unless there is no other option, and the PCN is understood to be preserving or improving quality of life. Internalized drainage with a ureteric stent is always a more attractive option psychologically, but may not provide adequate drainage, or may cause side effects. There is little data on the consequences of a PCN drain/s on quality of life, but the chances of preserving renal function and prolonging life in the terminally ill patient is something that always requires careful consideration. Ultimately the decision should be taken by the patient.

Percutaneous nephrolithotomy Introduction Fernström and Johansen first described what is now known as percutaneous nephrolithotomy (PCNL) in 1976 (Fernström and Johansson, 1976).

Box 12.2  Principles of PCN insertion technique 1. Aseptic skin preparation and draping 2. Infiltration of local anaesthetic down to the renal capsule 3. Ultrasound-guided puncture of a target calyx 4. Aspiration of urine 5. Careful contrast administration under fluoroscopy guidance 6. Guidewire placement through the puncture sheath in to the pelvicalyceal system 7. Dilatation of a percutaneous track 8. PCN drain insertion and deployment in the pelvicalyceal system 9. Fixation of the PCN drain with dressings and connection to a drainage bag.

Fig. 12.1  The dark anechoic hydronephrotic pelvicalyceal system contrasts with bright hyperechoic renal sinus fat. The pre-calibrated needle guide tramlines can be used to assist renal puncture.

83

84

Section 1  

assessment of the patient with renal disease Table 12.2  Approach to pelvicalyceal stone burden

Fig. 12.2  The 0.038-inch wire can be easily visualized with ultrasound alone to confirm position.

PCNL has replaced open stone surgery, offering similar stone clearance rates with reduced morbidity. PCNL remains a major undertaking, not least because extracorporeal shock wave lithotripsy (ESWL) and modern endourological techniques are pushing the PCNL case mix exclusively towards the most complex cases.

Indications The main indications for PCNL are stones > 2.5 cm in diameter, obese patients with a poor response to ESWL, stones associated with distal renal tract obstruction, lower-pole stones, and stones associated with indwelling foreign bodies, such as encrusted stents. Failed primary or repeated ESWL and/or ureteroscopy treatment are also common reasons for trying this technique.

Techniques Typically both an expert interventional radiologist and urologist are needed. Pre-procedural imaging should be carefully evaluated in a combined endourology meeting. A planning computer tomography (CT) scan should be used to evaluate all but the simplest of stones. An unenhanced phase demonstrates the stone burden and a delayed urographic phase details the pelvicalyceal anatomy, essential for planning access. Three-dimensional reconstructions are particularly helpful. Aberrant pelvicalyceal anatomy can be identified, and the pleural reflections and position of the colon can be taken into account when considering the approach. Because stones are colonized with bacteria (especially Proteus spp.), administration of antibiotics pre-procedure is recommended. Either a prone or supine approach is possible. A supine approach with the patient in the lithotomy position allows for combined percutaneous and retrograde intrarenal stone surgery. Supine techniques have been shown by some to be more time efficient (Hoznek et  al., 2012; Kumar et  al., 2012). The urologist usually places a retrograde ureteric catheter up to the renal pelvis to allow for pelvicalyceal opacification as preparation for the percutaneous approach. The expert radiologist interprets and cross references the preoperative CT and on-table fluoroscopic and ultrasound imaging to plan the safest and best access to the collecting system. The usual approach is to target a calyx, which is in line with the long axis of the stone. The aim is to create as few tracks as possible to clear the maximum stone volume (see Table 12.2). Although

Stone location

Track

Isolated lower pole

Direct lower pole puncture

Multiple lower pole

Upper pole puncture

Isolated or multiple upper pole

Lower pole puncture

Staghorn

Lower pole ± further tracks

flexible instruments and laser lithotripsy can be invaluable, an optimally placed track will allow for much more effective rigid ultrasonic or pneumatic lithotripsy. Occasionally large staghorn stones require multiple tracks. Tracks of between 24 and 30 Fr diameter can be created by telescopic metal dilators or serial fascial dilators, but there is a risk of kinking the heavy-duty access wire. Although expensive, the balloon dilatation system is generally preferred despite the PCNL audit in the United Kingdom demonstrating a non-significant trend towards a higher blood transfusion rate with this method (Tomaszewski et al., 2010; Armitage et al., 2012). (See Fig. 12.3.) Occasionally intraoperative bleeding will halt progress. A large-bore nephrostomy tube can be placed, and a second-look procedure can be arranged after the bleeding has settled. A post PCNL nephrostomy is not always necessary (Table 12.3).

Limitations and complications Retrograde catheter placement by the urologist is not always possible in patients with complex lower tract anatomy or urinary diversions. If the stone burden is large enough to warrant PCNL, a direct ultrasound-guided puncture on to a calyx or calyceal stone is usually still possible. There is little agreement in the literature about how to assess the presence of residual post-PCNL fragments, and what constitutes a ‘stone-free’ state (Armitage et al., 2012). Comparison studies are

Fig. 12.3  A 30-Fr access sheath, safety wire, and retrograde catheter are required for percutaneous management of this staghorn stone.

chapter 12 

Table 12.3  Advantages of post-PCNL nephrostomy placement versus ‘tubeless PCNL’ Post-PCNL nephrostomy

Tubeless

Allows drainage

Reduced post-op pain

Allows post-op nephrostogram

Reduced post-op hospital stay

Tamponades track

No increased risk of bleeding

limited by uncertainty over stratification of case mix and complexity, but the Guy’s Stone Score has been shown to predict post PCNL stone free rate accurately (Thomas et al., 2011). Complications are usually related to haemorrhage, urosepsis, or urine leak and have been classified by severity on the basis of the nature of further treatment required, using a modified Clavien grading system (Tefekli et al., 2008) (Table 12.4). If there is clinical concern for significant postoperative bleeding, an urgent CT angiogram should be arranged after appropriate resuscitation. Post-PCNL imaging follow-up should be tailored to the individual patient, with rapid staghorn formers and those with underlying metabolic disorders (e.g. cystinuria) requiring a more intense schedule. Repeat unenhanced CT is a cumulative radiation hazard, and the combination of plain film and ultrasound is usually sufficient if recurrent stones are to be treated with ESWL in the first instance. Less traumatic, specially designed miniaturized dilators, sheaths, and instruments are preferred in the paediatric population but are associated with longer operative times in the adult (Wah et al., 2013). Fluid balance and prevention of hypothermia are particular concerns in the anaesthetized child. The PCNL population is a high-risk group for further stone formation and intervention. Longer-term renal impairment and hypertension cannot confidently be attributed to post-PCNL parenchymal scarring, although there is no control group for comparison.

intervention

Table 12.5  Causes of ureteric obstruction Non-malignant

Malignant

Stone disease

Bladder

Tuberculosis

Prostate

Post radiotherapy

Cervical

Retroperitoneal fibrosis

Breast

Pelviureteric junction obstruction

Retroperitoneal

Hydronephrosis of pregnancy

Nodes and lymphomas (move to right)

Indications Ureteric stents drain the pelvicalyceal system in cases of ureteric obstruction (Table 12.5) and allow fistulae and leaks to resolve by providing a frame around which ureteral epithelialization is facilitated. Ureteric stents used to protect the transplant kidney ureteric anastomosis are placed intraoperatively. An antegrade (as opposed to a retrograde) approach to stenting offers a greater success rate, especially for distal ureteric strictures. Even if a contrast injection demonstrates complete ureteric occlusion, it is often still possible to negotiate the stricture in an antegrade fashion with a torque-controlled hydrophilic wire (Fig. 12.4). Antegrade ureteric stent placement requires percutaneous renal access, which can be secured in advance with a PCN, or gained at the same sitting.

Techniques The basic outline of an antegrade ureteric stent insertion is described in Box 12.3. An upper pole or interpolar calyceal track provides the best angle of approach towards the pelviureteric junction if there is the luxury of planning the original PCN. A mature access track of 8 Fr aids the

Antegrade ureteric stent insertion Introduction The placement of a ureteric stent in an antegrade fashion is a natural extension of the PCN technique. Most patients find internal drainage more convenient and cosmetically acceptable (Banner, 1998). A ureteric stent is less likely to be inadvertently displaced than a PCN drain, and usually has a lower morbidity (Holmes et al., 1993).

Table 12.4  A UK PCNL audit based on 1028 procedures documents complication rates (Armitage et al., 2012) Complication

Rate

Haemorrhage requiring transfusion

2.5%

Haemorrhage requiring embolization

0.4%

Post procedural fever

16%

Post procedural sepsis

2.4%

Visceral injury

0.4%

Fig. 12.4  Even very tight distal ureteric strictures can be easily crossed with a guidewire and catheter.

85

86

Section 1  

assessment of the patient with renal disease

Box 12.3  Antegrade ureteric stent insertion ◆ Patient consent ◆ Preparation including antibiotics and sedation with analgesia ◆ Aseptic skin preparation and draping ◆ Local anaesthesia ◆ Creation of a percutaneous track or exchange of a pre-existing nephrostomy over a guidewire for an angiographic catheter ◆ Nephro-ureterogram to demonstrate the anatomy ◆ Guidewire and catheter manipulation through the obstruction ◆ Exchange of hydrophilic wire for stiff wire and removal of catheter ◆ Stent insertion and deployment ◆ Placement of a covering non-locking loop drain.

procedure significantly, but single-step percutaneous puncture and stent deployment can be undertaken. Gentle continuous rotation of the hydrophilic wire is often all that is needed to allow passage through very tight strictures. An acute sense of urinary urgency is often felt by the patient, which may be eased by distending the bladder with a saline-contrast mix to distance the guidewire tip from the trigone. The hydrophilic wire is exchanged for a 0.038-inch working wire via a catheter. Occasionally a stricture cannot be crossed by the softer stent material, which may buckle. The stent then needs to be carefully withdrawn to allow balloon dilatation if appropriate. The tip of the stiff wire is left within the pelvicalyceal system when the proximal stent pigtail is deployed, so as to facilitate the placement of a covering PCN tube. This can be used if there is early stent failure, which can occur if a percutaneous track has been created at the same procedure, due to fresh occlusive blood clots. Covering PCN tubes can be capped off to allow renal function monitoring. The use of non-locking tubes is advised as locking sutures can become entangled with the proximal stent pigtail. After several days, non-locking PCN drains can be removed on the ward, or under fluoroscopic guidance once a nephroureterogram confirms satisfactory drainage.

Limitations and complications Advances in the development of polyurethane and silicone copolymers have led to the creation of soft stents, which are relatively resistant to occlusion, fracture, and migration, and are well tolerated. Nevertheless these factors remain problematic. Stent encrustation is a particular problem in a subset of patients whose risk factors include a long indwelling time, urinary sepsis, history of stone disease, and metabolic abnormalities (Holmes et al., 2010). Ureteric stents should not be placed in patients with an active urinary tract infection. Co-polymer stent patency rates have been demonstrated to fall from 95% at 3 months post insertion, to 54% at 6 months (Lu et al., 1994). Full-length metal stents provide increased patency rates for up to 12 months in patients with malignant ureteric strictures. Nevertheless, both co-polymer and metallic stents remain vulnerable to failure from extrinsic malignant compression of the ureter. Nephrostograms demonstrate that most flow occurs around, rather

than through a ureteric stent. If problematic, vesicorenal reflux can be minimized by bladder catheterization. Irritative bladder symptoms due to stimulation of the trigone by the distal stent pigtails can be minimized by careful stent positioning. Short, thermo-expandable titanium-nickel stents, which do not contact the trigone, are an option in the patient with a short malignant stricture. In the United Kingdom, registering a stent insertion with the British Association of Urological Surgeons (BAUS) national stent register has become best practice. The clinician is alerted by email when stent exchange is required; avoiding the forgotten severely encrusted stent, which often requires both an endourological and percutaneous approach to retrieval (British Association of Urological Surgeons, n.d.). Co-polymer ureteric stents should be changed by the urologists in a retrograde manner at cystoscopy initially after 6 months, and if uneventful, annually thereafter. Individual patient circumstances and risk factors for encrustation should however be considered and stent changes timed to pre-empt complications such as encrustation or occlusion.

Renal biopsy (See also Chapter 14.)

Introduction Real-time imaging is essential for a safe needle biopsy of the kidney. Ultrasound or CT can be utilized. Biopsies are performed both by radiologists and nephrologists. The former are called upon in challenging cases, for example, obese subjects.

Indications Non-focal renal parenchymal biopsy is utilized to diagnose a diffuse nephropathy or to exclude renal transplant rejection. Targeted renal biopsy is being increasingly utilized for assessing the indeterminate, small, solid renal lesion, although its use remains somewhat controversial. With modern advancements in histological techniques, however, the targeted biopsy of small renal lesions is gaining popularity as a way of avoiding the unnecessary morbidity of either surgery or ablation for benign lesions (Beland et al., 2007).

Technique Written informed consent and a normal patient clotting profile are again mandatory prerequisites. A posterior approach to the native kidney avoids crossing the peritoneum. Ultrasound is preferred as it offers continuous real-time imaging and the flexibility to alter the planned trajectory at any time. An angled approach with the needle allows ribs, pleura, and bowel to be avoided. Aseptic technique and draping are employed. A  sterile ultrasound probe cover is utilized. Infiltration of local anaesthesia can be guided by ultrasound to anaesthetize a track down to the renal cortex. A small skin incision to allow the insertion of a spring-loaded core biopsy needle is then made. Suspended patient respiration stabilizes the target kidney. The goal of non-focal renal biopsy is to maximize the amount of cortex obtained whilst avoiding damage to the renal hilum. The fewest cores possible should be obtained to minimize bleeding complications. Immediate assessment with microscopy is useful to ensure enough material is obtained for light, immune-, and electron microscopy.

chapter 12 

intervention

Patients should have bed rest and regular assessment of blood pressure and heart rate for at least 8 hours post procedure. If there are any concerns for significant bleeding an unenhanced and postcontrast arterial phase CT should be arranged in tandem with resuscitative measures. Renal angiography and embolization is the gold standard treatment for active bleeding (Fig. 12.5). The needle track for iliac fossa transplant kidneys should pass lateral to the incision scar to avoid crossing the peritoneum.

Limitations and complications Ninety-five per cent of patients undergoing non-focal renal biopsy should have an adequate diagnostic sample taken at one visit. Targeted renal biopsy has the greatest sensitivity (97%) and negative predictive value (89%) for larger lesions of between 4 and 6 cm in size (Rybicki et al., 2003; Uppot et al., 2010). Renal biopsy is a safe procedure but minor bleeding is very common. Small perinephric haematomas are seen in approximately 50% of patients and up to a third have frank haematuria post procedure. Severe bleeding is uncommon, however, with transfusion rates of approximately 1% and embolization required in far fewer. Iatrogenic arteriovenous fistulae tend to be self-limiting if small. Tumour seeding of the renal biopsy needle track is very rare with only a handful of case reports in the literature.

Drainage of collections Introduction Percutaneous imaging-guided drainage offers a safe and rapid way of dealing with troublesome fluid collections. It is well tolerated by the patient, and may avoid the need for further surgery.

Indications A renal or perirenal abscess may present acutely as a source of systemic sepsis. Specific post-transplant collections include urinomas, haematomas, and lymphocoele. If a fluid collection is believed to be infected, painful or otherwise detrimental to the patients’ progress it is appropriate to request drainage as long as it is deemed to be safely accessible by the radiologist. Decisions about the management of a postoperative collection must be taken by the surgical team.

Fig. 12.5  A urinoma consequent to a ureteroileal anastomotic defect is drained percutaneously with CT guidance.

Damage to adjacent organs is also a possibility. Pre-procedural antibiotics and a normal clotting profile are required. Antiseptic skin preparation, draping, and local anaesthesia are the next steps. Local anaesthetic infiltration can be imaged with ultrasound or CT to ensure that sensitive structures such as the peritoneum are anaesthetized adequately, and that the needle trajectory is appropriate. Long 20-G needles can be used to infiltrate local anaesthetic and act as pathfinders down to the collection to guide subsequent trocar mounted drain placement. A tandem technique can be used which allows for single step track dilatation. An 8-Fr drain is inserted alongside the long needle (Uppot, 2011). Final positioning is confirmed with CT before advancing the drain over the inner trocar and aspirating. A modified Seldinger technique is usual practice. A  15-G sheath mounted on a 19-G needle is inserted in to the collection.

Techniques Radiologists often favour either ultrasound or CT based on expertise and familiarity. There are, however, specific factors which would favour using one modality over the other. It is obviously vital to avoid iatrogenic injury to important overlying or adjacent structures such as large blood vessels and bowel. If a safe percutaneous route is not available in an axial plane then ultrasound is favoured over CT as the modality to guide drainage. The ultrasound probe can be angled steeply and used to negotiate overlying structures. Ultrasound also allows safe continuous real-time imaging where the progress of the advancing percutaneous needle tip can be followed. CT on the other hand involves ionizing radiation and even with CT fluoroscopy the radiologist has to advance the needle, pause, and re-image. (See Fig. 12.6.) CT more accurately depicts a gas- and liquid-filled collection from gas- and liquid-filled adjacent bowel. It is often difficult to make this distinction with ultrasound as highly echogenic gas bubbles prevent the ultrasound beam from progressing to (and therefore imaging) deeper tissues. Discomfort, bleeding, and subsequent drain complications such as blockage and migration should be included in patient consent.

Fig. 12.6  Embolization coils are placed in to the feeding artery for a post-nephrectomy bleeding pseudoaneurysm.

87

88

Section 1  

assessment of the patient with renal disease

Withdrawing the inner needle allows fluid aspiration prior to inserting a 0.038 J-tipped wire. Positioning can be confirmed with ultrasound or CT prior to serial fascial dilatation and subsequently drain placement over the wire. The initial aspirate can be sent for analysis. Special drain dressings can be supplemented with sutures for security. Tension on the drain should be avoided. Migration and eventually drain displacement leaves a partially drained collection, which is much more difficult to target for re-drainage. Irrigation of an infected collection is not routinely advocated, as it has a tendency to promote bacteraemia. Once several doses of appropriate antibiotics have been administered the drain can be flushed if required.

Limitations and complications Haemorrhage, adjacent organ injury, sepsis, and infecting a previously sterile fluid collection are recognized complications. Drains can be painful and may kink, block, and be pulled out. It is important when draining a urinoma that urinary diversion away from the leak is also considered.

References Ahallal, Y., Khallouk, A., and El Fassi, M. J. (2010). Risk factor analysis and management of ureteral double-J stent complications. Rev Urol, 12(2–3), 147–51. Armitage, J. N., Irving, S. O., and Burgess, N. A. (2012). Percutaneous nephrolithotomy in the United Kingdom: results of a prospective data registry. Eur Urol, 61, 1188–93. Banner, M. P. (1998). Antegrade and retrograde ureteral stent placement. In M. P. Banner (ed.) Radiologic Interventions: Uroradiology, pp, 96–127. Baltimore, MD: Williams and Wilkins. Beland, M. D., Mayo-Smith, W. W., Dupuy, D. E., et al. (2007). Diagnostic yield of 58 consecutive imaging-guided biopsies of solid renal masses: should we biopsy all that are indeterminate? Am J Roentgenol AJR, 188, 792–97. British Association of Urological Surgeons (BAUS). Stent Registry. [Online]

Dyer, R. B., Regan, J. D., Kavanagh, P. V., et al. (2002). Percutaneous nephrostomy with extensions of the technique: step by step. Radiographics, 22, 503–25. Farrell, T. A. and Hicks, M. E. (1997). A review of radiologically guided percutaneous nephrostomies in 303 patients. J Vasc Interv Radiol, 8(5), 769–74. Fernström, I. and Johansson, B. (1976). Percutaneous pyelolithotomy. a new extraction technique. Scand J Urol Nephrol, 10(3), 257–9. Goodwin, W. E., Casey, W. C., and Woolf, W. (1955). Percutaneous trocar (needle) nephrostomy in hydronephrosis. JAMA, 157, 891–4.

Heyman, S. N., Fuchs, S., Jaffe, R., et al. (1997). Renal microcirculation and tissue damage during acute ureteral obstruction in the rat: effect of saline infusion, indomethacin and radiocontrast. Kidney Int, 51, 653–63. Holmes, S. A., Christmas, T. J., and Rickards, D. (1993). Ureteric stents. In D. Rickards, S. Jones, K. R. Thomson, et al. (eds.) Practical Interventional Radiology, pp. 53–65. London: Edward Arnold. Hoznek, A., Rode, J., Ouzaid, I., et al. (2012). Modified supine percutaneous nephrolithotomy for large kidney and ureteral stones: technique and results. Eur Urol, 61(1), 164–70. Kumar, P., Bach, C., Kachrillas, S., et al. (2012). Supine percutaneous nephrolithotomy (PCNL): ‘in vogue’ but in which position? BJU Int, 110(11), 1118–21. Lewis, S. and Patel, U. (2004). Major complications after percutaneous nephrostomy—lessons from a department audit. Clin Rad, 59, 171–9. Lu, D. S., Papanicolaou, N., and Girard, M. (1994). Percutaneous internal ureteral stent placement: review of technical issues and solutions in 50 consecutive cases. Clin Rad, 49(4), 256–61. Palaniappa, N. C., Telem, D. A., Ranasinghe, N. E., et al. (2012). Incidence of iatrogenic ureteral injury after laparoscopic colectomy. Arch Surg, 147(3), 267–71. Rickards, D., Jones, S., and Kellett, M. (1993). Percutaneous nephrostomy. In D. Rickards, S. Jones, K. R. Thomson, et al. (eds.) Practical Interventional Radiology, pp. 23–33. London: Edward Arnold. Rybicki, F. J., Shu, K. M., Cibas, E. S., et al. (2003). Percutaneous biopsy of renal masses: sensitivity and negative predictive value stratified by clinical setting and size of masses. Am J Roentgenol AJR, 180, 1281–7. Tefekli, A., Ali Karadag, M., Tepeler, K., et al. (2008). Classification of percutaneous nephrolithotomy complications using the modified clavien grading system: looking for a standard. Eur Urol, 53(1), 184–90. Thomas, K., Smith, N. C., and Hegarty, N. (2011). The Guy’s stone score-grading the complexity of percutaneous nephrolithotomy procedures. Urology, 78(2), 277–81. Tomaszewski, J. J., Smaldone, M. C., Schuster, T., et al. (2010). Factors affecting blood loss during percutaneous nephrolithotomy using balloon dilatation in a large contemporary series. J Endourol, 24(2), 207–11. Türk, C., Knoll, T., Petrik, A. et al. (2012). Guidelines on Urolithiasis. [Online] European Association of Urology. Uppot, R. N. (2011). Nonvascular interventional radiology procedures. In E. Quaia (ed.) Radiological Imaging of the Kidney, pp. 271–87. Heidelberg: Springer. Uppot, R. N., Harisinghani, M. G., and Gervais, D. A. (2010). Imaging-guided percutaneous renal biopsy: rationale and approach. Am J Roentgenol AJR, 194, 1443–9. Wah, T. M., Kidger, L., Kennish, S., et al. (2013). MINI PCNL in a Pediatric Population. Cardiovasc Intervent Radiol, 36(1), 249–54. Wah, T. M., Weston, M. J., and Irving, H. C. (2004). Percutaneous nephrostomy insertion: outcome data from a prospective multioperator study at a UK training centre. Clin Rad, 59, 255–61. Zagoria, R. J. and Dyer, R. B. (1999). Do’s and don’t’s of percutaneous nephrostomy. Acad Radiol, 6(6), 370–7.

CHAPTER 13

Ultrasound Toby Wells and Simon J. Freeman Ultrasound physics Greyscale ultrasound Ultrasound creates an image using short-duration pulses of high-frequency sound waves generated by electrically stimulating a piezoelectric crystal in a handheld ultrasound transducer. The pressure wave generated is transmitted to the patient using a coupling gel between the transducer and skin surface. Returning echoes hit the same crystal and the vibrations induced are reconverted to an electrical signal which is then processed to create the image. Different tissues reflect the waves to varying degrees when exposed to the ultrasound beam. This behaviour is largely determined by tissue density and is expressed as the ‘acoustic impedance’. At the boundaries between soft tissues of different acoustic impedance the sound wave is partially reflected. If the interface is close to perpendicular with the ultrasound beam (more than 60°) the reflected echo will return to the transducer and be detected and its strength determines the brightness of the interface on the display. At soft tissue/air or soft tissue/bone interfaces the difference in acoustic impedance is so great that almost all the sound is reflected and none left to image deeper structures. The inability of ultrasound to penetrate gas-filled organs (such as bowel or lung) and bone can cause difficulty in obtaining a useful diagnostic ultrasound study. Other echoes are derived from tiny tissue structures that are similar in size to the ultrasound wavelength because instead of reflecting the ultrasound beam they cause it to be scattered in all directions. This phenomenon is of particular importance in generating a Doppler signal from moving red blood cells. Ultrasound is assumed to have a constant velocity in soft tissue (about 1540 m/s), so the time interval between transmission of the ultrasound pulse and reception of the returning echo can be used to calculate the distance of the reflecting interface from the transducer. The crystals within the transducer are fired sequentially and the returning reflected and backscattered echoes analysed for their strength and delay to construct a real-time two-dimensional image, referred to as the greyscale or B (brightness)-mode. Higher-frequency sound produces images of higher resolution, but the sound is attenuated more quickly and so tissue penetration is reduced. This should be taken into account when selecting the appropriate transducer for an ultrasound study. A high-frequency transducer (7–11 MHz) gives better resolution and is ideal for imaging superficial structures such as the testes, or for endocavity use such as in transrectal ultrasound. A lower-frequency transducer (3.5–5 MHz) is usually required to visualize abdominal structures such as the native kidneys. Selecting the appropriate

transducer is therefore a trade-off between image quality and depth penetration. Resolution is optimized by ensuring that the focus of the beam is adjusted to the level of the region of interest. High-frequency transducers usually have a flat surface (linear array) producing a parallel ultrasound beam of the width of the transducer. Lower-frequency abdominal and endocavity transducers usually have a curved surface producing a diverging ultrasound beam and a wide ultrasound sector. Harmonic imaging and spatial compounding are techniques that are now routinely employed by most manufacturers to produce images with fewer artefacts and greater resolution or clarity.

Doppler ultrasound The Doppler effect is the change in frequency that occurs when a sound source or detector are moving relative to one another. This is a familiar phenomenon to us in our everyday lives; for example, the change in tone of a siren as an emergency vehicle passes a static observer. The same principle can be employed in diagnostic ultrasound to demonstrate and measure blood flow. The frequency shift of an ultrasound wave that has been scattered by cells in moving blood is proportional to the velocity of the scattering object. Blood flowing towards the transducer will increase the frequency; conversely, flow away from the transducer will decrease the frequency. The frequency shift that results is small in comparison with the transmitted frequency; usually in the audible range. The Doppler frequency shift also depends on the transmitted ultrasound frequency and the cosine of the angle between the ultrasound beam and the direction of blood flow. The greatest Doppler frequency shift is generated when blood is flowing directly towards or away from the transducer, so there is no Doppler signal if the vessel is parallel to the transducer. The ultrasound practitioner must therefore optimize probe positioning and, where possible, utilize beam steering to ensure a satisfactory angle. If measurement of flow velocity is required, the angle of blood flow relative to the ultrasound beam must be manually selected using angle correction which permits the flow velocity to be calculated from the Doppler frequency shift. In this situation the angle should be 60° or less. Doppler information may be displayed as either a plot of velocity against time on a pulse wave (spectral) graph, or direction of flow and velocities and may be translated into different colours on a colour flow map. Power Doppler is another mode that assigns colour to blood flow. It uses the power of the Doppler signal rather than the frequency shift to display blood flow and is more sensitive to low flow than colour Doppler but gives no information about velocity or direction of flow.

90

Section 1  

assessment of the patient with renal disease

(A)

(B)

Fig. 13.1  (A) Colour flow Doppler allows identification of an interlobar artery in a renal transplant. The spectral trace obtained is normal—note the forward flow in diastole with a trace likened to a ‘ski jump’, and resistive index (RI) of 0.65. (B) The spectral Doppler trace from this renal transplant demonstrates reversed diastolic flow, in this case due to renal vein thrombosis.

The normal spectral arterial renal Doppler trace has high velocity flow in systole but also persistent forward flow throughout diastole, with a trace likened to a ‘ski slope’ (Fig. 13.1). Various parameters have been described to quantify this pattern. One of these, the resistance index (RI), is described in the renal transplant ultrasound section of this chapter.

Ultrasound contrast imaging Ultrasound contrast agents are increasingly being used in many different areas of ultrasound practice. The most recently available agents are based on stabilized microbubbles of perfluorocarbon gas in lipid shells. These have an excellent safety profile and no renal toxicity (Correas et al., 2001). Ultrasound takes advantage of the resonant properties of the microbubbles in the sound beam, enhancing visualization of blood flow. Contrast-enhanced ultrasound (CEUS)

studies are performed at very low power settings using harmonic techniques. This requires specific software which is now available on most mid- and high-range systems. There are many potential applications of ultrasound contrast agents, including characterization of complex renal cysts, renal vascular disorders, infection, transplant assessment, differentiation between complex renal cysts and solid lesions, and between renal pseudomasses and tumours (McArthur and Baxter, 2012). There are further applications for CEUS in scrotal, bladder, and prostatic ultrasound (Piscaglia et al., 2012) (Fig. 13.2).

Safety Ultrasound is a safe imaging technique which does not involve ionizing radiation. Despite the enormous numbers of ultrasound studies performed there are no confirmed deleterious medical side

Fig. 13.2  Microbubble contrast-enhanced image of the right kidney. Split screen view: the image to the reader’s right is a conventional greyscale image; the image to the left is a contrast-specific image showing (normal) intense enhancement of the kidney (arrowheads) and overlying liver in the nephrogram phase. The small area of unenhanced tissue (arrow) is a renal cyst.

chapter 13 

effects of this technique but it is good practice to minimize exposure by adjusting the power output to the lowest possible levels to obtain a diagnostic quality image, and minimizing scan time. This is of particular importance in obstetric imaging when scanning the fetus. Energy is transmitted to the patient during ultrasound examinations and this may cause mechanical or thermal effects. The main mechanical concern is cavitation (the collapse of gas-filled bubbles liberating energy) with a risk of local tissue damage. An estimate of the probability of cavitation occurring is displayed as the mechanical index (MI) on the ultrasound machine display. As gas bubbles are not normally found in the body, cavitation is generally only of concern when ultrasound contrast microbubble agents are used at high acoustic power settings and at gas/tissue interfaces. The probability of tissue heating occurring is displayed as the thermal index (TI). For a more comprehensive review of clinically based ultrasound principles and physics see Correas et al. (2001).

ultrasound

Fig. 13.3  A normal kidney showing echogenic renal sinus fat centrally, hypoechoic medullary pyramids, and intermediate echogenicity renal cortex peripherally of similar echogenicity to the adjacent liver. There is a small column of Bertin (normal variant) seen in the upper third of the kidney (arrow).

Renal ultrasound Introduction Ultrasound is the primary imaging modality in most kidney disorders. It is the first-choice investigation of renal functional impairment, in particular to exclude obstruction.

Technique and normal appearances To achieve accurate diagnosis requires attention to good ultrasound technique. The kidneys are orientated such that their upper poles are situated posteromedially to the lower poles. This should be borne in mind when orientating the transducer to obtain a true long-axis image of the kidneys. Close anatomical relation to the bowel means the lower poles are frequently obscured by bowel gas, and the upper poles may be obscured by ribs. Scanning usually begins with the patient in the supine position using a 3–5-MHz curvilinear transducer. Views of the right kidney can be facilitated by using the liver as an acoustic window. Scanning intercostally but angling down towards the feet may allow visualization of the lower poles when they would otherwise be obscured by bowel gas and compression can be used to displace gas. Rolling the patient into the lateral decubitus position and deep inspiration may also be of assistance. Fasting is unnecessary preparation for renal ultrasound. It should be remembered that although the kidneys, adrenals, and perinephric fat are enclosed in Gerota’s fascia, the two layers may not be fused inferiorly and medially to the kidneys. This can allow considerable mobility of the kidneys inferiorly and across the midline when rolling some patients onto their side. The renal cortex, medulla, and sinus can be differentiated on ultrasound. The central renal sinus fat is usually echogenic and the medullary pyramids are echo poor. The cortex covers the bases of the pyramids, extends down between them in the columns of Bertin, and is of intermediate echogenicity (Fig. 13.3). The relative reflectivity of the cortex is often compared to the adjacent liver or spleen, with the normal renal cortex being less echogenic (darker) than the normal liver and spleen. Much of ultrasound assessment of the kidney is subjective, but a commonly used quantitative measurement is renal length. Normal adult renal length is 10–12 cm, typically slightly larger on the left side and in male patients. Renal length correlates with the patient’s

height and a decrease in size is seen in old age (Emamian et al., 1993). Other measurements such as renal parenchymal thickness can be performed (where a value of > 10  mm is normal) (Dubbins, 2006). Direct visualization of renal arteries and veins can be achieved at the renal hila or by scanning the aorta and retroperitoneum from a variety of positions. Assessment of the spectral Doppler wave form in the main renal artery is often possible; a peak systolic flow velocity of > 180 cm/s is considered by many to be an indicator of haemodynamically significant renal artery stenosis (Strandness, 1994), while other authors have found the ratio of renal artery to aortic peak velocity more reliable (Li et  al., 2006). Ultrasound assessment of native renal artery stenosis is however frequently technically difficult or impossible and computed tomography (CT) or magnetic resonance imaging (MRI) angiography are more reliable techniques (Halpern et al., 1998). Renal perfusion can usually be readily demonstrated by colour or power Doppler so that even vessels as small as the interlobular arteries can be resolved in slim patients and in transplant kidneys.

Renal parenchymal disease Ultrasound is usually the first imaging investigation requested in patients presenting with acute or chronic kidney injury. The main role of ultrasound in this setting is to exclude renal obstruction, measure renal size, and evaluate parenchymal thickness and echogenicity. Ultrasound is insensitive in differentiating between the different causes of renal parenchymal disease and predicting prognosis. Sonographic changes are inconsistent and non-specific, whatever the underlying pathology. In patients with glomerular disease, ultrasound is often normal in the early stages, but sonographic abnormalities are often seen in patients with tubular-interstitial nephropathies (Quaia and Bertolotto, 2002) (Fig. 13.4). In acute kidney injury, renal size is occasionally increased due to inflammation and oedema but eventually the kidneys will reduce in size and renal parenchymal thickness will be lost. Focal cortical scarring may indicate reflux nephropathy or segmental infarction

91

92

Section 1  

assessment of the patient with renal disease

Fig. 13.4  This patient presented with acute kidney injury requiring dialysis. Ultrasound demonstrated increased cortical echogenicity (relative to the liver parenchyma) and loss of corticomedullary differentiation. Renal biopsy was performed showing acute interstitial nephritis.

and a small kidney with a smooth outline suggest the possibility of ischaemic renal disease. There is often increased echogenicity of the renal cortex. Initially the medullary pyramids may be unaffected and become more conspicuous, but the pyramids are also frequently involved leading to loss of corticomedullary differentiation. The ‘end-stage kidney’ appears as a small, shrunken, and echogenic organ that may even be difficult to identify within adjacent perinephric fat. Spectral Doppler ultrasound is unhelpful in determining the aetiology of parenchymal renal disease as all processes tend to show an increase in intrarenal vascular resistance (RI) (Mostbeck et al., 1991). Patients with end-stage renal failure, usually on long-term dialysis, may develop acquired cystic kidney disease (ACKD) (Levine, 1996). Ultrasound will demonstrate small kidneys containing multiple cysts. ACKD is associated with an increased risk of renal cancer with 2–7% of patients ultimately developing a renal carcinoma (Choyke, 2000).

occasionally also seen on ultrasound as an enlarged, elongated unilateral kidney with two distinct renal sinus complexes, often with a different orientation (Goodman et al., 1986). Unilateral renal agenesis has an approximate incidence of 1 in 1000, and is usually associated with compensatory hypertrophy of the contralateral kidney. Duplex kidneys are one of the more common congenital abnormalities found in 2.7–4.2% of patients on urographic series (Mascatello et al., 1997). Although ultrasound cannot usually make a definitive diagnosis, the condition can be inferred when the kidney is large in comparison with the contralateral side with a band of renal cortex dividing the sinus in the interpolar region (Fig. 13.5). The upper moiety may be hydronephrotic due to obstruction by a ureterocele (Nussbaum et al., 1986), which should be sought on bladder scanning. The lower moiety may have a dilated collecting system or cortical scarring secondary to reflux. Hypertrophied columns of Bertin represent a protrusion of normal parenchymal tissue into the renal sinus and are a regular finding on ultrasound, seen in almost half of kidneys in one study (Lafortune et al., 1986). The column is isoechoic with adjacent renal cortex and there should be no contour abnormality of the overlying renal tissue. Splenic (dromedary) humps are a focal bulge of normal renal parenchyma arising from the interpolar region of the kidney, usually on the left side. Many patients also show regular indentations of the renal contour representing persistent fetal lobulation. These conditions need to be recognized as a normal variant rather than a renal mass or evidence of scarring (Bhatt et al., 2007). Ultrasound contrast agents can be very helpful in confirming that pseudomasses are normal renal tissue (Mazziotti et al., 2010).

Ureteric obstruction and renal stone disease (See Chapter 200.) Ultrasound is requested to exclude obstruction, either as a cause of renal functional impairment or in patients with flank pain. Ultrasound is very accurate in detecting hydronephrosis (Ellenbogen et al., 1978), but it should be remembered that ultrasound demonstrates anatomical rather than functional changes (Webb, 2000). Pelvicalyceal dilatation may occur in conditions other than obstruction, such as vesicoureteric reflux, and there may be no dilatation for several hours following the onset of ureteric

Congenital abnormalities and pseudomasses The numerous congenital abnormalities and pseudomasses of the kidneys can be demonstrated with ultrasound. Renal ectopia is most commonly due to incomplete ascent. Pelvic kidneys are found in between 1 in 2200 and 1 in 3000 patients; are more common on the left side and are prone to ureteropelvic junction obstruction and formation of calculi (Cinman et al., 2007). The pelvic kidney can usually be identified by ultrasound but will be malrotated and often hydronephrotic and may be misinterpreted as a pathological mass by the unwary ultrasonographer who fails to note the absence of a kidney in the ipsilateral renal fossa. The most common fusion abnormality is the horseshoe kidney, which occurs in 1 in 400 individuals. The isthmus of tissue joining the lower poles is not always visible on ultrasound as it may be very thin or obscured by bowel gas, and ultrasound diagnosis may be difficult (Strauss et  al., 2000). This condition should be suspected if the kidneys have an abnormal orientation with medial displacement and poor visualization of the lower poles; anterior position of the renal pelvis and a curved renal shape. Crossed fused renal ectopia is

Fig. 13.5  A duplex kidney—it is large and has renal cortex extending into the renal sinus.

chapter 13 

(A)

ultrasound

(B)

Fig. 13.6  (A) Hydronephrosis—the anechoic collecting system is dilated. (B) The cause of the hydronephrosis is identified; a calculus at the vesicoureteric junction (arrowed). Note the dilated ureter proximal to this.

obstruction (Laing et al., 1985). Hydronephrosis is recognized when the echogenic renal sinus is replaced by communicating fluid-filled anechoic spaces, representing the dilated pelvicalyceal system (Fig. 13.6). It is important to demonstrate this communication as parapelvic cysts can be mistaken for dilated calyces (Fig. 13.7). There is also a delay in the restoration of normal appearances after relief of an obstruction, often for several days and a degree of pelvicalyceal dilatation may persist indefinitely, particularly after relief of longstanding or severe obstruction. A number of grading systems have been proposed for the severity of hydronephrosis but simply stating mild, moderate, or severe is adequate and avoids confusion. The accuracy of ultrasound in diagnosing renal obstruction can be improved by the use of Doppler techniques. Elevated intrarenal vascular resistance in an obstructed kidney can be demonstrated by spectral Doppler examination (Platt et al., 1989, 1993). Colour Doppler ultrasound can also be used to demonstrate absence or abnormality of jets of urine entering the bladder in patients with ureteric obstruction (Burge et al., 1991). These techniques can be of particular value in the evaluation of pregnant patients in whom differentiation between physiological

(A)

hydronephrosis and hydronephrosis due to stone obstruction can be particularly difficult and where the use of imaging investigation using ionizing radiation (CT of the kidneys, ureters, and bladder (KUB) and intravenous urography (IVU)) are undesirable (Hertzberg et al., 1993; Haddad et al., 1995). MRI can be very helpful in difficult cases and is safe after the first trimester (Spencer et al., 2004). Renal stones appear as bright echogenic foci. Within the kidney it can be difficult to distinguish a small stone from the echogenic renal sinus. Sensitivity is improved by using a higher-frequency transducer and ensuring the focus is located at the level of the suspected calculus. Ultrasound has a reported sensitivity of 96% and specificity of 89% for detection of stones in the pelvicalyceal system (with tomography as the gold standard) (Middleton et al., 1988), but the figure is as low as 37% for ureteric calculi (Aslaksen et al., 1990), largely due to overlying bowel gas. In female patients, transvaginal ultrasound can be used to demonstrate distal ureteric stones (Laing et al., 1994). CT has now become the investigation of choice for renal stone disease and should be the first-line investigation for most patients with acute onset loin pain performing

(B)

Fig. 13.7  (A) Apparent hydronephrosis—multiple anechoic structures in the renal pelvis easily mistaken for a dilated collecting system on ultrasound, but subsequently shown to represent parapelvic cysts. (B) On CT urogram, renally excreted contrast is seen in the true collecting system, compressed by the adjacent cysts.

93

94

Section 1  

assessment of the patient with renal disease

much better than ultrasound (and IVU) in this situation (Yilmaz et al., 1998).

Solid and cystic renal masses Ultrasound is excellent at differentiating solid from cystic lesions of the kidney. Simple cysts are very common and increase in incidence with age (Ravine et al., 1993). Sonographically, cysts are anechoic, normally spherical, and have posterior acoustic enhancement—the area behind the cyst appears brighter than at the equivalent depth elsewhere on the image. They may be parapelvic, intraparenchymal, or exophytic, in which case observing movement with the kidney on respiration is helpful in making the diagnosis. Benign cysts must be distinguished from cystic malignancy. Septations, loculations, or solid components increase the likelihood of malignancy. The Bosniak classification categorizes cystic lesions by complexity based on their appearances at CT and is used to guide management (Bosniak, 2012). Central to this classification is the degree of enhancement following contrast medium administration. This is easily measured on CT, but only possible on ultrasound using microbubble contrast media (Tamai et  al., 1995). Caution should therefore be applied when extrapolating the Bosniak system to ultrasound practice and most complex cystic renal masses will require further evaluation with CT. A number of other non-malignant renal lesions may appear as complex cystic masses on ultrasound including multilocular cystic nephroma, hydatid disease, abscess, haematoma, and xanthogranulomatous pyelonephritis. The hereditary renal cystic conditions are often encountered during abdominal ultrasound examination. The most common is autosomal dominant polycystic kidney disease (ADPKD) which results in enlarged kidneys with multiple cysts of varying size, randomly distributed throughout the kidney. Ultrasound is sometimes requested to screen a first-degree relative of a patient known to have ADPKD. Diagnostic criteria exist relating the number of cysts to the patient’s age (Ravine et al., 1994; Pei et al., 2009). Screening ultrasound may give a false negative result if performed below 30 years of age (Nicolau et al., 1999). (See Chapters 306–309.) The ultrasound appearances of renal cell carcinoma are variable. Large tumours tend to be heterogenous but predominantly

(A)

hypoechoic. Smaller tumours may be hyperechoic and impossible to distinguish from angiomyolipomas (Vallancien et al., 1990). The presence of vascular invasion into the renal vein and inferior vena cava is readily assessed on ultrasound, although CT is required for complete staging (Fig. 13.8). Other solid renal masses include angiomyolipomas, which are usually brightly echogenic on ultrasound due to the presence of fat. Oncocytomas are usually impossible to distinguish from renal cell carcinoma on ultrasound, although a central scar is said to be characteristic and colour Doppler ultrasonography may show central radiating vessels (Fan et al., 2008). Upper tract transitional cell carcinomas are hard to see on ultrasound; the sensitivity of CT urography or IVU is much greater (O’Connor et al., 2010). Lymphomas have variable appearances which include diffuse infiltration and focal masses and they are typically hypoechoic.

Infection (See Chapters 175–179.) Ultrasound imaging is often requested in cases of suspected pyelonephritis. The findings may include reduced parenchymal echogenicity and increase in renal size. Perfusion defects may be detected using colour or power Doppler; however, the scan is frequently normal. In focal pyelonephritis (lobar nephronia) part of the kidney may be enlarged or of altered echogenicity, either increased, mixed, or decreased (Farmer et al., 2002). The main role of ultrasound is to exclude pyonephrosis or abscess formation. Emphysematous pyelonephritis has a characteristic appearance on ultrasound. The renal parenchyma contains gas which produces bright echoes with heterogeneous (‘dirty’) posterior acoustic shadowing. When the perinephric soft tissues are involved the entire kidney may be difficult to identify. Chronic pyelonephritis secondary to vesicoureteric reflux may cause renal scarring, seen on ultrasound as focal cortical thinning overlying a dilated or clubbed calyx. Dimercaptosuccinic acid (DMSA) scintigraphy is used more frequently in this setting due to its greater sensitivity (Roebuck et al., 1999). In xanthogranulomatous pyelonephritis chronic low grade infection is usually secondary to a central obstructing stone. There are no pathognomonic ultrasound features except that stones may be identified, the affected kidney is

(B)

Fig. 13.8  (A) A large renal cell carcinoma at the upper pole of the right kidney (arrows). (B) Views of the inferior vena cava with Doppler show extension of the tumour into the cava as a filling defect.

chapter 13 

frequently enlarged and contains cystic spaces and fluid collections in the perinephric soft tissues (Tiu et al., 2001). (See Chapter 355.)

Trauma Improved access to CT scanning with modern multidetector row machines has meant that ultrasound is now rarely the first-line investigation for renal trauma. However, the portability of ultrasound means it is still occasionally useful in the emergency department, and it does have a role in follow-up. The key findings are loss of continuity of the renal cortex, perinephric collections of blood or urine, and disruption of renal perfusion. The use of ultrasound contrast agents may improve accuracy but as they are not excreted in the urine cannot be used to detect collecting system injuries.

Renal transplants (See Chapter 282.) About a half of all patients with end-stage renal disease (ESRD) rely on a functioning renal transplant (462 pmp in the United Kingdom). Ultrasound is vital in detection of the complications of renal transplantation, and is used both for regular surveillance and as the first line imaging investigation for graft dysfunction. Renal transplants are typically placed in either iliac fossa, but the axis of the kidney is variable and the scan plane must be tailored to accommodate this. The superficial position of the transplant usually allows it to be assessed easily and a higher-frequency transducer can often be used. The scan technique involves a thorough greyscale assessment of the transplant and evaluation of the surrounding soft tissues for fluid collections. Global renal perfusion is then assessed with colour Doppler followed by a spectral Doppler assessment of several interlobar arteries. Finally the main vascular pedicle and iliac vessels should be examined. In the early postoperative period, peritransplant collections may represent urinomas, haematomas, lymphocoeles, or abscess formation. Ultrasound-guided needle aspiration may be necessary to differentiate between them. Renal artery thrombosis (global or segmental) is easily recognized by an absence of perfusion of all or part of the transplant. Renal vein thrombosis may be difficult to visualize directly but can be inferred when the normal low resistance intrarenal spectral Doppler pattern is replaced by a high-resistance pattern with reversal of diastolic flow (Fig. 13.1). Arteriovenous fistulas and pseudoaneurysms are usually related to previous biopsies and are readily identified with Doppler techniques. A mild degree of hydronephrosis is a common finding and does not necessarily indicate ureteric obstruction but progressive hydronephrosis in the face of deteriorating graft function may require antegrade pyelography or nephrostomy drainage to exclude obstruction. The common causes of early graft dysfunction are acute tubular necrosis (ATN) and rejection. Despite early optimism (Rifkin et al., 1987), ultrasound has been unsuccessful in accurately differentiating between acute rejection and ATN (Kelcz et al., 1990; Chow et al., 2001). In both conditions the intrarenal vascular resistance increases (Saarinen, 1991)  (usually measured by calculating the RI from a spectral Doppler trace of an interlobar artery; the upper limit of normal is < 0.7). Ultrasound-guided biopsy is usually required. Ultrasound can also identify some of the causes of late transplant failure. Peak systolic flow velocities of > 250 cm/s in the main transplant artery indicate a haemodynamically significant stenosis (Patel et al., 2003). Increasing hydronephrosis suggests a ureteric stricture.

ultrasound

RI measurements are not useful in the diagnosis of chronic allograft nephropathy (chronic rejection) although CEUS shows some promise in this area (Schwenger and Zeier, 2006).

Bladder ultrasound The bladder is usually evaluated through a suprapubic approach with a standard abdominal curvilinear transducer. To achieve useful information requires the bladder to be well distended. Quantification of bladder volume is easily achieved and is of particular importance in patients with lower urinary tract symptoms. This requires measuring distances in both transverse and the longitudinal planes, and is often performed pre and post micturition. A  commonly used formula for converting these linear measurements to an approximate volume is: V = ( H × W × L ) × 0.625 where V is volume, H is height, W is width, and L is length of the bladder. Most ultrasound systems will automatically perform these calculations but the result tends to underestimate bladder volume slightly; accuracy is, however, acceptable for routine clinical use (Nwosu et al., 1998). Three-dimensional ultrasound may give more accurate measurements (Byun et al., 2003) but is not widely available. The normal bladder wall measures 3–5 mm in thickness which may be useful in predicting the severity of bladder outflow obstruction, but measurements depend on the degree of bladder distension and which part of the bladder wall is measured (Oelke et al., 2007). Many bladder abnormalities can be identified and assessed by ultrasound including diverticula, urachal abnormalities, stones, and blood clots. In patients with cystitis, bladder ultrasound is usually normal (but bladder emptying can be assessed). In chronic cystitis there may be bladder wall thickening. The presence of echogenic foci with acoustic shadowing (representing gas) in the bladder wall should alert the ultrasound practitioner to the possibility of emphysematous cystitis. Gas in the bladder lumen, in the absence of instrumentation, is suggestive of a vesicoenteric fistula. Ureterocoeles can be elegantly demonstrated on bladder ultrasound. They are seen as cystic structures projecting into the bladder lumen from the ureteric orifices which can frequently be seen to fill and empty on real-time scanning (Fig. 13.9A). Detection of an ectopic ureterocoele arising from the upper moiety of a duplex kidney inserting below the bladder neck can be facilitated by scanning through the perineum (Vijayaraghavan, 2002). Transitional cell bladder cancer can often be identified on ultrasound as a soft tissue projection into the bladder lumen. Predictably polypoid tumours are much more reliably identified than sessile ones. Tumours < 5 mm in size and those located on the anterior wall or bladder neck are frequently more difficult to identify (Itzchak et  al., 1981)  and ultrasound cannot replace cystoscopy (Ozden et al., 2007). Ultrasound has a limited ability to differentiate transitional cell carcinoma from other bladder masses including other types of bladder tumour, invasive prostate carcinoma, adherent clots, and endometriosis–cystoscopic inspection and biopsy is always required. The simple manoeuvre of rolling the patient onto their side will differentiate dependent debris and clot from posterior tumour. The presence of flow within a bladder mass on Doppler interrogation usually indicates a bladder tumour (Fig. 13.9B).

95

96

Section 1  

assessment of the patient with renal disease

(A)

(B)

Fig. 13.9  (A) A ureterocoele, seen as a cystic structures projecting into the bladder lumen from the ureteric orifice. (B) A polypoid transitional cell carcinoma of the bladder with internal blood flow demonstrated by colour Doppler.

Paediatric renal ultrasound ( See Sections 15 and 16.) Ultrasound is the primary imaging modality in paediatric radiology for most conditions, largely because it does not involve ionizing radiation. However, to obtain good images requires the cooperation of the child and parents, warm coupling gel, a relaxed atmosphere, and patience and skill on the part of the ultrasound practitioner. The use of higher-frequency transducers is often possible and transducers with a smaller ‘footprint’ are often used. Urinary tract abnormalities diagnosed in infancy and childhood are sometimes associated with other extrarenal abnormalities or syndromes, for example the ‘VACTERL’ collection of non-randomly associated birth defects. Neonate kidneys are of greater relative size than adults, and charts are available to relate renal size to age (Rosenbaum et al., 1984). The renal cortex is brighter than in adult kidneys, medullary pyramids are also relatively large, corticomedullary differentiation greater, and there is little renal sinus fat (Hricak et al., 1983). By 1 year of age these morphological differences will have resolved and the kidneys resemble those of adults. Antenatal detection of renal tract abnormalities by scans performed in pregnancy is relatively common. Postnatal imaging for antenatal detected hydronephrosis is best performed at 48–72 hours after birth, as relative dehydration in the early postnatal period may produce a false negative result. The diameter of the renal pelvis is often used as a reproducible measure of the degree of hydronephrosis and can be used to guide further management (Sidhh et al., 2006). In most patients renal pelvic dilatation is due to an unobstructed extrarenal pelvis but postnatal ultrasound is required to identify the subgroup of patients with diseases that may require surgical intervention. These conditions include ureteropelvic junction obstruction (UPJO), congenital megaureter, vesicoureteric reflux, posterior urethral valves, duplex kidney, multicystic dysplastic kidney, and prune belly syndrome. Additional imaging investigations such as the micturating cystourethrogram and nuclear medicine diuretic isotope renogram are sometimes required (Belarmino and Kogan, 2006). Autosomal recessive and autosomal dominant polycystic kidney disease can both present in childhood (See Chapter 313). The recessive form also involves the liver leading to hepatic fibrosis

in some patients. In young children with this condition the renal cysts are often tiny and impossible to resolve with ultrasound, the kidneys appearing enlarged and globally echogenic but with time cyst formation may become apparent (Blickman et  al., 1995). ADPKD typically presents in adult patients, but enlarged kidneys containing multiple cysts may occasionally be seen in older children. Renal cysts may also be seen in patients with tuberous sclerosis but characteristically these patients have multiple renal angiomyolipomas, appearing as echogenic nodules within the renal parenchyma. Multicystic dysplastic kidney is another cause of a unilateral cystic kidney; it is usually detected on prenatal ultrasound and may be associated with contralateral UPJO in some patients. In children with urinary tract infection, an ultrasound study is usually performed to exclude a structural abnormality. In infants this is frequently accompanied by a micturating cystogram to exclude vesicoureteric reflux and a DMSA scan to detect renal scarring. The National Institute of Health and Care Excellence (NICE) have produced comprehensive guidelines on the management of urinary tract infection in childhood (NICE, 2007). Wilms tumour (nephroblastoma) is the commonest malignancy in childhood usually occurring in children aged between 2 and 5 years. The commonest presentation is with an abdominal mass. On ultrasound, Wilms tumour is seen as a solid mass replacing all or part of the kidney, often with hypoechoic areas of necrosis (De Campo, 1986)  (Fig. 13.10). Five to 10% are bilateral and there are a number of associated urogenital abnormalities (Breslow et al., 1993). Staging requires CT or MRI. Wilms tumour cannot be reliably differentiated from the more benign entity mesoblastic nephroma on ultrasound. Bladder abnormalities such as neurogenic bladder, stones, cystitis, urachal abnormalities. or rhabdomyosarcoma may be seen on ultrasound. Scanning with a full bladder is usually required. Because of its larger relative size, the normal neonatal adrenal is usually seen on routine ultrasound, unlike adults. There is progressive cortical involution in the first few weeks of life and tables are available for normal adrenal size in the newborn (Scott et al., 1990). Adrenal masses such as neuroblastoma and adrenal haemorrhage may be identified on ultrasound and Doppler may help to differentiate between these two conditions (Deeg et al., 1998).

chapter 13 

Fig. 13.10  Wilms tumour—a large heterogeneous mass arising from and replacing the kidney in a 2-year-old child. With thanks to Dr J. Foster, Derriford Hospital, Plymouth, UK.

ultrasound

insertion of a needle or biopsy trochar under direct visualization. Experience is required to keep the needle in the ultrasound sector and identify the position of the needle tip. Needle holders are available to assist this—they attach to the transducer and ensure the path of the needle remains within the field covered by the beam. The depth of the needle track can be adjusted. Bowel is less easily visualized (and so avoided) on ultrasound than on CT. Many recommend biopsying transversely across the lower pole of the kidneys as this avoids hitting the large vessels at the renal hilum, and yields more nephrons for the histopathologist to assess. Ultrasound is now routinely used in the placement of central venous lines/dialysis catheters, as it allows accurate infiltration of local anaesthetic and confident cannulation of vessels. A chest radiograph will still be required, however, to ensure that the tip is in the desired location—usually at the level of the lower superior vena cava, above the right atrium. Ultrasound has a further role in aiding the safe placement of peritoneal dialysis catheters.

Ultrasound for procedures

Assessment of dialysis fistulae

(See Chapters 12 and 18.) In addition to its extensive diagnostic role, ultrasound is ideally suited to guide interventions (Frede et al., 2001). It allows the

(See Chapter 256.) Ultrasound has a major role in pre-assessment, monitoring, and assessing poorly functioning fistulae.

(A)

(B)

(C)

Fig. 13.11  (A) Doppler ultrasound shows an increased peak systolic flow rate at a stenosis in the draining vein of an arteriovenous dialysis fistula, confirmed by subsequent angiography (B), and treated by angioplasty (C).

97

98

Section 1  

assessment of the patient with renal disease

A dialysis arteriovenous fistula (AVF) is a subcutaneous anastomosis between artery and vein, usually in the forearm but occasionally in the leg. A graft is usually made of polytetrafluoroethylene (PTFE) and is a synthetic communication formed in a similar position. AVFs generally have better long-term patency, fewer complications, less morbidity, and improved flow in comparison to grafts (Hirth et al., 1996; Rosas et al., 2003). There is some evidence that the systematic use of preoperative ultrasound to assess the arterial and venous anatomy of a proposed site of fistula formation increases the success rate (Brimble et al., 2002; Ferring et al., 2010). This is often referred to as ‘vein mapping’ and is used to confirm adequate vessel size and patency, absence of significant arterial disease, and exclude vascular anomalies. There is also evidence that prospective monitoring of fistulae and grafts for haemodynamically significant stenoses reduces morbidity, thrombosis, and intervention, improving long-term patency (Smits et al., 2001). Other authorities reserve assessment for those with clinical problems (Teodorescu et al., 2012). Monitoring can take several different forms, including clinical examination, flow measurement during dialysis using dilutional techniques, or Doppler ultrasound to assess the volume of flow. The flow rate in a healthy fistula or graft is up to 2000 mL/min, and there is a risk of thrombosis if the flow rate drops below a critical level. Further investigation and, where necessary, intervention is usually indicated in a graft when the flow volume is below 600 mL/min or drops more than 20–25% on serial ultrasound measurement and in a fistula when flow volume is below 300 mL/min. The ultrasound examination should also assess the entire vascular access circuit to identify abnormalities such as stenoses, thrombosis, intimal flaps and aneurysm/pseudoaneurysm formation in the supplying arteries and draining veins. The accuracy of Doppler ultrasound to assess overall flow rate through a fistula or graft as a screening test for dysfunction is debated. Some studies suggest comparable accuracy to fistulography, but an experienced operator is required and there is considerable interobserver variability (Akoh and Hakim, 1999; Schwarz et al., 2003). It is, perhaps, more universally accepted in assessing a fistula in which dysfunction is already suspected (Teodorescu et al., 2012). In these cases stenosis is visualized as a local increase in flow velocity. A peak systolic velocity of > 400 cm/s or greater than threefold focal increase in peak systolic velocity are parameters used to indicate a significant stenosis (Older, 1998). Ultrasound can often directly visualize thrombosis within the fistula or supplying vessels and this can be confirmed by incomplete or absent luminal filling on colour Doppler examination. The length of thrombus can also be measured and is useful in planning intervention. Formal angiography remains the gold standard for assessment of fistula or graft dysfunction, and allows intervention (Fig. 13.11). Good fistula function requires good arterial inflow and a low pressure venous outflow. Venous stenosis is less common with AVFs than grafts but can be more severe. It usually occurs as a consequence of neointimal hyperplasia at the venous anastomosis or in the draining vein, but can occur within the lumen of synthetic grafts.

Further reading Allan, P. L., Baxter, G. M., and Weston, M. L. (eds.) (2011). Clinical Ultrasound (3rd ed.). Edinburgh: Churchill Livingstone. Pozniak, M. A. and Allan, P. L. (eds.) (2013). Clinical Doppler Ultrasound (3rd ed.). Edinburgh: Churchill Livingstone, Elsevier.

References Akoh, J. A. and Hakim, N. S. (1999). Preserving function and long-term patency of dialysis access. Ann R Coll Surg Engl, 81, 339–42. Aslaksen, A. and Göthlin, J. H. (1990). Ultrasonic diagnosis of ureteral calculi in patients with acute flank pain. Eur J Radiol, 11, 87–90. Belarmino, J. M. and Kogan, B. A. (2006). Management of neonatal hydronephrosis. Early Hum Dev, 82, 9–14. Bhatt, S., MacLennan, G., and Dogra, V. (2007). Renal pseudotumors. AJR Am J Roentgenol, 188, 1380–7 Blickman, J. G., Bramson, R. T., and Herrin, J. T. (1995). Autosomal recessive polycystic kidney disease: long-term sonographic findings in patients surviving the neonatal period. AJR Am J Roentgenol, 164, 1247–50. Bosniak, M. A. (2012). The Bosniak renal cyst classification: 25 years later. Radiology, 262(3), 781–5. Breslow, N., Olshan, A., Beckwith, J. B., et al. (1993). Epidemiology of Wilms tumor. Med Pediatr Oncol, 21(3), 172–81. Brimble, K. S., Rabbat, C. G., Treleaven, D. J., et al. 2002). Utility of ultrasonographic venous assessment prior to forearm arteriovenous fistula creation. Clin Nephrol, 58, 122–7. Burge, H. J., Middleton, W. D., McClennan, B. L., et al. (1991). Ureteral jets in healthy subjects and in patients with unilateral calculi: comparison with color Doppler US. Radiology, 180, 437–42. Byun, S. -S., Kim, H. H., Lee, E., et al. (2003). Accuracy of bladder volume determinations by ultrasonography: are they accurate over entire bladder volume range? Urology, 62(4), 656–60. Chow, L., Sommer, F. G., Huang, J., et al. (2001). Power Doppler imaging and resistance index measurements in the evaluation of acute renal transplant rejection. J Clin Ultrasound, 29(9), 483–90. Choyke, P. L. (2000). Acquired cystic kidney disease. Eur Radiol, 10(11), 1716–21. Cinman, N. M., Okeke, Z., and Smith, A. D. (2007). Pelvic kidney: associated diseases and treatment. J Endourol, 21(8), 836–41. Correas, J. M., Bridal, L., Lesavre, A., et al. (2001). Ultrasound contrast agents: properties, principles of action, tolerance, and artifacts. Eur Radiol, 11(8), 1316–28. De Campo, J. F. (1986). Ultrasound of Wilms’ tumor. Pediatr Radiol, 16(1), 21–4. Deeg, K. H., Bettendorf, U., and Hofmann, V. (1998). Differential diagnosis of neonatal adrenal haemorrhage and congenital neuroblastoma by colour coded Doppler sonography and power Doppler sonography. Eur J Pediatr, 157(4), 294–7. Dubbins, P. (2006). The kidney. In D.L. I. Cochlin, P.A. Dubbins, B. B. Goldberg, et al. (eds.) Urogenital Ultrasound: A Teaching Atlas (2nd ed.), pp. 1–104. Oxford: Taylor and Francis. Ellenbogen, P. H., Scheible, F. W., Talner, L. B., et al. (1978). Sensitivity of grey scale ultrasound in detecting urinary tract obstruction. AJR Am J Roentgenol, 130, 731–3. Emamian, S. A., Nielsen, M. B., Pedersen, J. F., et al. (1993). Kidney dimensions at sonography: correlation with age, sex, and habitus in 665 adult volunteers AJR Am J Roentgenol, 160(1), 83–6. Fan, L., Lianfang, D., Jinfang, X., et al. (2008). Diagnostic efficacy of contrast-enhanced ultrasonography in solid renal parenchymal lesions with maximum diameters of 5 cm. J Ultrasound Med, 27(6), 875–85. Farmer, K. D., Gellett, L. R., and Dubbins, P. A. (2002). The sonographic appearances of acute focal pyelonephritis 8 years experience. Clin Radiol, 57(6), 483–7. Ferring, M., Claridge, M., Smith, S. A., et al. (2010). Routine preoperative vascular ultrasound improves patency and use of arteriovenous fistulas for hemodialysis: a randomized trial. Clin J Am Soc Nephrol, 5(12), 2236–44. Frede, T., Hatzinger, M., and Rassweiler, J. (2001). Ultrasound in endourology. J Endourol, 15(1), 3–16. Goodman, J. D., Norton, K. I., Carr, L., et al. (1986). Crossed fused renal ectopia: sonographic diagnosis. Urol Radiol, 8(1), 13–16.

chapter 13 

Haddad, M. C., Abomelha, M. S., and Riley, P. J. (1995). Diagnosis of acute ureteral calculus obstruction in pregnant women using colour and pulsed Doppler sonography. Clin Radiol, 50, 864–6. Halpern, E. J., Rutter, C. M., Gardiner, G. A., Jr, et al. (1998). Comparison of Doppler US and CT angiography for evaluation of renal artery stenosis. Acad Radiol, 5(8), 524–32. Hertzberg, B. S., Carroll, B. A., Bowie, J. D., et al. (1993). Doppler US assessment of maternal kidneys: analysis of intrarenal resistivity indexes in normal pregnancy and physiologic pelvicaliectasis. Radiology, 186, 689–92. Hirth, R. A., Turenne, M. N., Woods, J. D., et al. (1996). Predictors of type of vascular access in hemodialysis patients. JAMA, 276, 1303–8. Hricak, H., Slovis, T. L., Callen, C. W., et al. (1983). Neonatal kidneys: sonographic anatomic correlation. Radiology, 147, 699–702. Itzchak, Y., Singer, D., and Fischelovitch, Y. (1981). Ultrasonographic assessment of bladder tumors. 1. Tumor detection. J Urol, 126(1), 31–3. Kelcz, F., Pozniak, M. A., Pirsch, J. D., et al. (1990). Pyramidal appearance and resistive index: insensitive and nonspecific sonographic indicators of renal transplant rejection. AJR Am J Roentgenol, 155, 531–5. Lafortune, M., Constantin, A., Breton, G., et al. (1986). Sonography of the hypertrophied column of Bertin. AJR Am J Roentgenol, 146(1), 53–6. Laing, F. C., Benson, C. B., DiSalvo, D. N., et al. (1994). Distal ureteric calculi: detection with vaginal US. Radiology, 192, 545–8. Laing, F. C., Jeffrey, R. B., Jr., and Wing, V.W. (1985). Ultrasound versus excretory urography in evaluating acute flank pain. Radiology, 154, 613–16. Levine, E. (1996). Acquired cystic kidney disease. Radiol Clin North Am, 34, 947–64. Li, J.-C., Wang, L., Jiang, Y.-X., et al. (2006). Evaluation of renal artery stenosis with velocity parameters of Doppler sonography. J Ultrasound Med, 25, 735–42. Mascatello, V. J., Smith, E. H., Carrera, G. F., et al. (1997). Ultrasonic evaluation of the obstructed duplex kidney. AJR Am J Roentgenol, 129, 113–20. Mazziotti, S., Zimbaro, F., Pandolfo, A., et al. (2010). Usefulness of contrast-enhanced ultrasound in the diagnosis of renal pseudotumors. Abdominal Imaging, 35(2), 241–5. McArthur, C. and Baxter, G. M. (2012). Current and potential renal applications of contrast-enhanced ultrasound. Clin Radiol, 67(9), 909–22. Middleton, W. D., Dodds, W. J., Lawson, T. L., et al. (1988). Renal calculi: sensitivity for detection with US. Radiology, 167, 239–44. Mostbeck, G. H., Kain, R., Mallek, R., et al. (1991). Duplex Doppler sonography in renal parenchymal disease. Histopathologic correlation. J Ultrasound Med, 10(4), 189–94. National Institute for Health and Clinical Excellence (2007). Guidelines on Urinary Tract Infection: Diagnosis and Long Term Management of Urinary Tract Infection in Children. Nicolau, C., Torra, R., Badenas, C., et al. (1999). Autosomal dominant polycystic kidney disease types 1 and 2; assessment of US sensitivity for diagnosis. Radiology, 213, 273–6. Nussbaum, A. R., Dorst, J. P., Jeffs, R. D., et al. (1986). Ectopic ureter and ureterocele: their varied sonographic manifestations. Radiology, 159, 227–35. Nwosu, C. R., Khan, K. S., Chien, P. F. W., et al. (1998). Is real-time ultrasonic bladder volume estimation reliable and valid? A systematic overview. Scand J Urol Nephrol, 32(5), 325–30. O’Connor, O. J., Fitzgerald, E., and Maher, M. M. (2010). Imaging of hematuria. AJR Am J Roentgenol, 195(4), 263–7. Oelke, M., Höfner, K., Jonas, U., et al. (2007). Diagnostic accuracy of noninvasive tests to evaluate bladder outlet obstruction in men: detrusor wall thickness, uroflowmetry, postvoid residual urine, and prostate volume, Eur Urol, 52(3), 827–34. Older, R. A. (1998). Haemodialysis access stenosis: early detection with color Doppler US. Radiology, 207, 161–4. Ozden, E., Turgut, A. T., Turkolmez, K., et al. (2007). Effect of bladder cancer location on detection rates by ultrasonography and computed tomography. Urology, 69(5), 889–92.

ultrasound

Patel, U., Khaw, K. K., and Hughes, N. C. (2003). Doppler ultrasound for detection of renal transplant artery stenosis-threshold peak systolic velocity needs to be higher in a low-risk or surveillance population. Clin Radiol, 58(10), 772–7. Pei, Y., Obaji, J., Dupuis, A., et al. (2009). Unified criteria for ultrasonographic diagnosis of ADPKD. J Am Soc Nephrol, 20(1), 205–12. Piscaglia, F., Nolsøe, C., Dietrich, C. F., et al. (2012). The EFSUMB guidelines and recommendations on the clinical practice of contrast enhanced ultrasound (CEUS), Update 2011 on non-hepatic applications. Eur J Ultrasound, 33, 33–59. Platt, J. F., Rubin, J. M., and Ellis, J. H. (1989). Distinction between obstructive and non-obstructive pyelocaliectasis with duplex Doppler sonography. AJR Am J Roentgenol, 153, 997–1000. Platt, J. F., Rubin, J. M., and Ellis, J. H. (1993). Acute renal obstruction: evaluation with intrarenal doppler and conventional US. Radiology, 186, 685–8. Quaia, E. and Bertolotto, M. (2002). Renal parenchymal diseases: is characterisation feasible with ultrasound? Eur Radiol, 12(8), 2006–20. Ravine, D., Gibson, R. N., Donlan, J., et al. (1993). An ultrasound renal cyst prevalence survey: specificity data for inherited renal cystic diseases. Am J Kidney Dis, 22(6), 803–7. Ravine, D., Sheffield, L. J., Danks, D. M., et al. 91994). Evaluation of ultrasonograhic diagnostic criteria for autosomal dominant polycystic kidney disease 1. Lancet, 343, 824–7. Rifkin, M. D., Needleman, L., Pasto, M. E., et al. (1987). Evaluation of renal transplant rejection by duplex Doppler examination: value of the resistive index. AJR Am J Roentgenol, 148(4), 759–62. Roebuck, D. J., Howard, R. G., and Metreweli, C. (1999). How sensitive is ultrasound in the detection of renal scars? Br J Radiol, 72(856), 345–8. Rosas, S. E., Joffe, M., Burns, J. .E, et al. (2003). Determinants of successful synthetic hemodialysis vascular access graft placement. J Vasc Surg, 37, 1036–42. Rosenbaum, D. M., Korngold, E., and Teele, R. L. (1984). Sonographic assessment of renal length in normal children. Am J Roentgenol, 142, 467–70. Saarinen, O. (1991). Diagnostic value of resistive index of renal transplants in the early postoperative period. Acta Radiol, 32(2), 166–9. Schwarz, C., Mitterbauer, C., Boczula, M., et al. (2003). Flow monitoring: performance characteristics of ultrasound dilution versus color Doppler ultrasound compared with fistulography. Am J Kidney Dis, 42, 539–45. Schwenger, V. and Zeier, M. (2006). Contrast-enhanced sonographu as early diagnostic tool of chronic allograft nephropathy. Nephrol Dial Transplant, 21, 2694–6. Scott, E. M., Thomas, A., McGarrigle, H. H., et al. (1990). Serial adrenal ultrasonography in normal neonates. J Ultrasound Med, 9(5), 279–83. Sidhh, G., Beyene, J., and Rosenblum, N. D. (2006). Outcome of isolated antenatal hydronephrosis: a systematic review and meta-analysis. Pediatr Nephrol, 21(2), 218–24. Smits, J. H., van der Linden, J., Hagen, E. C., et al. (2001). Graft surveillance: venous pressure, access flow, or the combination? Kidney Int, 59, 1551–8. Spencer, J. A., Chahal, R., Kelly, A., et al. (2004). Evaluation of painful hydronephrosis in pregnancy: magnetic resonance urographic patterns in physiological dilatation versus calculous obstruction. J Urol, 171, 256–60. Strandness, D. E. Jr. (1994). Duplex imaging for the detection of renal artery stenosis. Am J Kidney Dis, 24(4), 674–8. Strauss, S., Dushnitsky, T., Peer, A., et al. (2000). Sonographic features of horseshoe kidney: review of 34 patients. J Ultrasound Med, 19, 27–31. Tamai, H., Takiguchi, Y., Oka, M., et al. (2005). Contrast enhanced ultrasonography in the diagnosis of solid renal tumors. J Ultrasound Med, 24(12), 1635–40.

99

100

Section 1  

assessment of the patient with renal disease

Teodorescu, V., Gustavson, S., and Schanzer, H. (2012). Duplex ultrasound evaluation of hemodialysis access: a detailed protocol. Int J Nephrol, 2012, Article 508956. Tiu, C. -M., Chou, Y. -H., Chiou, H. -J., et al. (2001). Sonographic features of xanthogranulomatous pyelonephritis. J Clin Ultrasound, 29(5), 279–85. Vallancien, G., Torres, L. O., and Gurfinkel, E. (1990). Incidental detection of renal tumours by abdominal ultrasonography. Eur Urol, 18(2), 94–6.

Vijayaraghavan, S. B. (2002). Perineal sonography in diagnosis of an ectopic ureteric opening into the urethra. J Ultrasound Med, 21(9), 1041–6. Webb, J. A. (2000). Ultrasonography and Doppler studies in the diagnosis of renal obstruction. BJU Int, 86(1), 25–32. Yilmaz, S., Sindel, T., Arslan, G., et al. (1998). Renal colic: comparison of spiral CT, US and IVU in the detection of ureteric calculi. Eur Radiol, 8, 212–17.

CHAPTER 14

Computed tomography Eugene Teoh and Michael J. Weston Introduction Computed tomography (CT) is the most widely used cross-sectional imaging modality. Introduced in the 1970s, CT served as a means of generating images in the transverse plane (axial images) of the internal organs of the patient. It has since evolved to be used to obtain volumes of image data of the relevant body section, which may be manipulated with image processing software to view the internal organs on sub-millimetre sections in multiple planes, and many other applications including surface rendering to allow three-dimensional views. The rudimentary construct of the CT scanner involves an X-ray tube and a row of detectors built into a rotatable gantry. The patient lies on a couch that can move longitudinally through the gantry and images are acquired from relevant sections of the patient’s body when they are in position within the gantry (see Fig. 14.1). The gantry rotates around the patient and X-rays are exposed to acquire data from different angles, which are used to construct the image using image reconstruction algorithms. In modern-day helical CT scanners, this process is rapid, with the gantry in constant rotation and constant exposure and image acquisition being performed as the body part of interest moves through the gantry. The speed of acquisition is increased by the use of multiple rows of detectors, allowing the whole body to be imaged in a matter of seconds. The constructed CT image is laid out as a matrix of pixels, each designated its own CT number or Hounsfield unit (HU). The CT number corresponds to the average density of tissues (and thus level of grey) within the pixel, with a positive correlation between density and CT number. For example, air has a CT number of −1000, water of zero, and bone 500–1500. Since the advent of multidetector CT (MDCT), its use in the evaluation of the urinary tract has expanded widely and it has now replaced intravenous urography (IVU) and the plain radiograph. A variety of CT techniques exist for evaluation of the urinary tract and the choice of technique depends largely on the indication, taking into account other patient factors (e.g. age and contraindications to iodinated contrast).

Contrast agents and practical techniques Contrast agents Iodine is the principal component of contrast medium used in CT, as it provides the radiographic density. Contrast agents may be classified according to their osmolarity (high or low) and composition

(ionic or non-ionic). Both characteristics contribute to the agent’s osmolality (and hence the propensity to cause fluid shifts), the highest being high-osmolar, ionic agents and lowest being low-osmolar, non-ionic agents. In modern-day CT, low-osmolar, non-ionic contrast agents are used, because they are five to ten times safer than their ionic counterparts (The Royal College of Radiologists, 2010). Prior to the examination, checks for contraindications to contrast administration and risk factors for contrast-related adverse effects (drug reactions, nephrotoxicity, and lactic acidosis) must be undertaken (Table 14.1). Essential points to cover are a history of previous contrast reaction, asthma, renal function, diabetes mellitus, allergies, and metformin therapy (The Royal College of Radiologists, 2010). The renal function of the patient must be checked prior to the examination. This should be in the form of the estimated glomerular filtration rate (eGFR), taken within a 3-month period prior to the examination in clinically stable patients, or a 7-day period in patients who have an acute illness or have a history of renal disease (The Royal College of Radiologists, 2010). Special precautions should be applied in patients with renal impairment. This is set arbitrarily as an eGFR of < 60 mL/min/1.73m2 although thresholds and practice will vary depending on local protocols (The Royal College of Radiologists, 2010). The Royal College of Radiologists has published guidance on recommended action for each risk factor (see ‘Safety issues and contraindications’). As with the administration of any drug, patient consent must be obtained before administration. (The Royal College of Radiologists, 2010)

Unenhanced CT KUB CT scan of the kidneys, ureters, and bladder (CT KUB) is the most accurate and now the primary investigation of suspected ureteric colic (The Royal College of Radiologists, 2012). Accuracy rates for this technique are high with reported sensitivities of 96–100% and specificities of 95.5–100% (Tamm et al., 2003). The patient is normally scanned in the prone position with images obtained from the top of the kidneys to the bladder base. The use of oral or intravenous contrast medium is not required. A full urinary bladder is the only prerequisite as this allows more sensitive detection of bladder calculi. Low-dose techniques are now routine, having been shown to maintain high rates of accuracy in detecting stones despite reduced quality of the images of soft tissue structures. With MDCT, the images can be acquired in a single breath-hold but reconstruction of thin sections is essential to avoid missing small calculi, particularly those of low density (Saw et al., 2000).

102

Section 1  

assessment of the patient with renal disease

X-ray tube

Detector array

Fig. 14.1  Basic schematic of CT scanning.

Renal CT Renal CT is the first-line investigation for the characterization of a renal mass (The Royal College of Radiologists, 2012), previously detected by a different imaging investigation, for example, ultrasound or CT KUB. The examination requires the administration of intravenous contrast medium and routinely involves two phases: an unenhanced phase and a nephrographic phase. The unenhanced phase is acquired before contrast administration. It provides baseline characterization of the density of the lesion to establish enhancement patterns (if any) and detection of high-density

components which may be otherwise obscured with the administration of contrast medium. The nephrographic phase is acquired 100 seconds after administration of intravenous contrast. It produces homogeneous enhancement of the renal parenchyma, allowing for reliable detection of renal masses (Israel and Bosniak, 2005).

Renal transplant CT Although ultrasound examination is the first-line imaging test in patients with a renal transplant, CT is often the next investigation, This is most commonly needed in the postoperative setting (Fig. 14.2).

Table 14.1  Recommended action in patients with increased risk of adverse effects Risk factor

Action

Previous reaction

◆

Establish nature of reaction and causative agent Reassess the need for contrast, risk vs benefit of its use, and consider the use of other investigations ◆ If absolutely necessary, use a different low osmolality agent ◆

Asthma

◆

If symptomatic/not well controlled, defer examination if not urgent, subject to improved management of asthma ◆ If well controlled, reassess the need for contrast and consider the use of alternative investigations

Multiple allergies/ single severe allergy

◆

Renal disease/ diabetes mellitusa

◆

Metformin

◆

◆

If deemed necessary: ◆ Proceed with a low osmolality agent ◆ Maintain close medical supervision for 30 minutes with IV access maintained ◆ Ensure all other safety measures are in place

Establish the nature and sensitivity of allergy Reassess the need for contrast, risk vs benefit of its use, and consider the use of alternative investigations.

Considering the severity of renal impairment, reassess the need for contrast, risk vs benefit of its use, and potential use of other investigations ◆ Use the smallest possible dose of low osmolality agent ◆ Ensure the patient is optimally hydrated before and after contrast injection ◆

If renal function is normal, no further action is required If renal function is abnormal, cessation of metformin for 48 hours is recommended in consultation with the referring doctor

a Coexistent diabetes mellitus in patients with renal impairment carries significant risk.

chapter 14 

(A)

computed tomography

(C)

(B)

Fig. 14.2  Renal transplant CT. (A) Coronal image of a transplanted kidney in the left iliac fossa acquired in the arterial phase, producing intense cortical enhancement compared to the medulla. The venous phase on the other hand produces a more homogeneous pattern of parenchymal enhancement. (B) The arterial anatomy is clearly delineated in the arterial phase. (C) Reconstructed oblique sagittal image demonstrating the entire length of the transplant artery (and its point of anastomosis onto the external iliac artery).

In most situations, the renal transplant CT involves three phases—an unenhanced phase, followed by an arterial and then a venous phase post contrast administration (Sebastia et al., 2001). In addition to demonstrating the presence and anatomical relationships of common postoperative complications such as collections, each phase has a specific role in the assessment of the transplant patient. The unenhanced phase is helpful in showing areas of haemorrhage or calcification, which are of high density on CT and might otherwise be masked after contrast medium administration. The arterial and venous phases are used to delineate the vascular anatomy and in particular to assess patency. The pattern of graft enhancement on these phases can also demonstrate areas of infarction.

CT urography CT urography (CTU) is the best test for detecting renal calculi, renal masses, and upper urinary tract urothelial cell carcinoma

(A)

(UUT-UCC) (The Royal College of Radiologists, 2012)  and is therefore the ‘one-stop’ imaging investigation of haematuria in the appropriate patient group (discussed in detail in ‘Indications’). The technique involves imaging the urinary tract in three phases:  unenhanced, nephrographic, and excretory, as for the sequence of radiographs performed in IVU (see Fig. 14.3). The unenhanced phase is performed as a CT KUB. The nephrographic phase acquires images of the kidneys (as previously described) after administration of intravenous contrast, allowing detection of renal masses. The excretory phase of the study assesses the collecting systems, ureters and, to a lesser extent, the urinary bladder when optimal luminal opacification of these structures has taken place following contrast excretion by the kidneys. It is best acquired 10–12 minutes after intravenous contrast administration.

(B)

Fig. 14.3  The three phases of CTU. (A) Unenhanced. (B) Nephrographic. (C) Excretory.

(C)

103

104

Section 1  

assessment of the patient with renal disease

Renal colic CT KUB is the most accurate, and therefore the primary investigation for suspected ureteric colic. In addition to the high accuracy rates for detection of calculi, it has superseded IVU by eliminating the need for intravenous contrast medium and by being quick to perform. The presence of obstructing calculi and the level of obstruction can be clearly defined. The stone size can be accurately measured, information which influences patient management. Furthermore, extra-genitourinary causes of the patient’s symptoms may be seen as unenhanced CT, for example, appendicitis and diverticulitis (Tamm et al., 2003). Most calculi appear as multifaceted, high-density foci, and the entirety of the kidneys, both ureters, and the bladder should be scrutinized for these. In the absence of calculi, ancillary signs of urolithiasis should be sought, such as hydronephrosis and/or hydroureter caused by distal obstruction, and perinephric and/or periureteric fat stranding indicating inflammation (see Fig. 14.4). These features may indicate recent passage of a stone or the presence of radiolucent calculi (particularly when there is a clear transition point of ureteric obstruction) (Smith et al., 1996).

Complications following renal transplantation Fig. 14.4  Reformatted image of the right kidney in a patient with significant stone disease previously treated with lithotripsy. There is hydronephrosis and hydroureter secondary to the formation of ‘steinstrasse’ (street of stones) within the distal ureter (arrowheads). There are also residual renal calculi predominantly within the lower pole calyx (arrow).

(See Chapter 280.) CT has a role in the evaluation of both immediate and delayed complications of renal transplantation. In the immediate postoperative period, abnormal Doppler waveforms on a transplant ultrasound may indicate the need for further assessment with CT. Arterial and venous thromboses, arterial stenosis, and infarcted grafts can be demonstrated on the triple-phase transplant CT (see Fig. 14.5). The

(A)

(B)

(C)

(D)

Fig. 14.5  Renal transplant complications. (A) Stenosis of the transplant artery (arrow) demonstrated on CT in a postoperative patient with low urine output and abnormal findings on Doppler examination. (B) Diffuse infarction of the lower pole of a transplant kidney characterized by decreased density of the renal parenchyma (arrowheads), compared to the perfused upper pole. (C) Thrombus was also demonstrated as a low-density filling defect (bold arrowheads) within the transplant vein of the same patient, with further areas of infarction (arrows). The distal aspect of the vein remains patent (arrowheads). (D) In a clinical caveat similar to that described in (A), the finding of a kink in the main transplant artery (arrow) explains poor graft perfusion seen on Doppler examination.

chapter 14 

cause and anatomical relationships of other structural abnormalities seen on ultrasound, such as transplant hydronephrosis and collections, can also be characterized in more detail with CT. During long-term follow-up, complications from immunosuppression include opportunistic infection, occult sources of sepsis, and post-transplant lymphoproliferative disorder (PTLD). The use of CT is commonplace in the detection of such complications and in the latter, for staging and follow-up.

Haematuria (See Chapter 46.) The major anxiety in the patient presenting with haematuria is that it is a sign of urological malignancy. The choice of imaging investigation is dictated by the pre-test probability, influenced by two main factors: age and the nature of haematuria. CTU is indicated in older patients presenting with frank haematuria, where the possibility of malignancy is high. There is no clear and agreed age cut-off although evidence from large series support the age of 40 years as a suitable threshold for this indication (Khadra et al., 2000; Edwards et al., 2006). Urolithiasis is the most common benign cause of haematuria in younger patients presenting with frank haematuria, CT KUB is definitely indicated to evaluate for stone disease. The choice of imaging investigation is less clear-cut in subjects with microscopic haematuria. Most cases of upper tract cancers demonstrate persistent haematuria and in the older patient with microscopic haematuria, interval testing to establish persistence should inform the decision on the use of CTU (Edwards et al., 2006). In younger patients with microscopic haematuria, many factors should be considered before requesting CT. Persistence of haematuria, and the presence of pain justify performing a CT. Because of the higher prevalence of renal parenchymal disease in this younger group, suggestive features on clinical and urine dipstick examination, and abnormal renal function should lead first to nephrological assessment rather than CT. A CT KUB is justifiable when there is a corroborating history of renal colic or if ultrasound examination does not show signs of parenchymal disease.

Characterization of renal masses Renal CT is the first-line investigation for the characterization of a renal mass (The Royal College of Radiologists, 2012), usually first detected by a different imaging investigation. Based primarily on its enhancement characteristics and morphological features, renal CT should inform the clinician as to whether the mass is one that needs surgical removal or if it has characteristics justifying follow-up observation (Israel and Bosniak, 2005). When evaluating a lesion, it is also helpful to decide whether it is solid or cystic. Renal cell cancers (RCCs) are the commonest renal parenchymal tumours and are mostly solid (Fig. 14.6A) although 2–5% are predominantly cystic. RCCs vary widely in size, uniformity, and extent on CT, which explains why it cannot be definitively diagnosed using this investigation. It can also not be distinguished from an oncocytoma on CT. Other solid renal masses include angiomyolipomata, lymphomas, and metastases. Of these, an angiomyolipoma can be definitively diagnosed by the detection of fat within the mass (Fig. 14.6B) (fat has a CT number of –60 to –150 HU). Lymphomata and metastases typically feature as multiple bilateral solid masses.

computed tomography

When evaluating a cystic renal mass, the Bosniak renal cyst classification is applied (Israel and Bosniak, 2005). It allows the lesion to be categorized into one of five groups based on morphology and enhancement patterns. These suggest the probable nature of the lesion (Table 14.2). Each category has accompanying clinical management recommendations which should, however, not be adhered to strictly. They should instead serve as a guide to management of individual cases, taking into account factors such as history, co-morbidity, and patient choice, particularly in patients with category III lesions (Fig. 14.7) (Israel and Bosniak, 2005).

Characterization of adrenal masses Adrenal masses are a relatively common incidental finding on CT studies, with adenomas being the commonest and featuring in up to 1.5% of examinations (Dèahnert, 2011). In general, an adrenal mass can be diagnosed as an adenoma on unenhanced CT if its average density is ≤ 10 HU, owing to abundance of lipid content. If the mass is > 10 HU, it should be regarded as indeterminate and

(A)

(B)

Fig. 14.6  (A) An example of an enhancing predominantly solid renal mass (arrows), subsequently diagnosed to be RCC. (B) A fat-containing mass with an average density of −69 HU (arrows), in keeping with an angiomyolipoma.

105

106

Section 1  

assessment of the patient with renal disease (A)

(B)

(C)

(D)

Fig. 14.7  (A) A Bosniak category I simple cyst within the upper pole of the right kidney (arrow). (B) A Bosniak category II cyst with a hairline thin calcified septa (arrow). (C) A Bosniak category III cyst within the lower pole of the right kidney with an enhancing thickened and nodular inferior wall (arrow). (D) A Bosniak category IV cystic mass with an enhancing soft tissue component independent of its enhancing thickened wall (arrow).

can be further characterized by assessing its enhancement washout (Korobkin, 2000). Enhancement washout refers to the drop in density between a contrast-enhanced phase and a delayed phase (normally acquired 15 minutes after contrast medium administration). An enhancement washout of > 50% and a delayed phase density < 35 HU would classify the mass as a lipid-poor adenoma. Masses with enhancement washouts < 50% and delayed phase densities > 35 HU remain indeterminate and require further investigation. Depending on various factors such as pre-existing malignancy and size, these take the form of biopsy, surgery, follow-up studies, or radionuclide imaging. Importantly, no specific CT features can differentiate between a functioning and non-functioning adenoma (Korobkin, 2000).

Detection and characterization of urothelial lesions CTU has superseded IVU for imaging the upper urinary tract because of its increased sensitivity and specificity for urothelial abnormality (Anderson et al., 2007). As a cross-sectional technique, it depicts extraluminal pathology and where relevant, pathology within the urachus. UUT-UCC can be seen as enhancing soft tissue density masses and tend to cause filling defects within the opacified urinary tract (Fig. 14.8A, B). Some UUT-UCCs can manifest as circumferential wall thickening which may not even distort the lumen (Anderson et al., 2007).

Bladder tumours can manifest as filling defects within a contrast-opacified bladder (Fig. 14.8C), focal masses, or asymmetric wall thickening, whereas uniform wall thickening tends to represent benign disease (Cohan et al., 2009). It is important to note that in the specific context of suspected bladder cancer, CTU is only an adjunct to cystoscopy (Blick et al., 2012).

Detection and characterization of urinary tract injury The excretory phase of CTU (± unenhanced phase depending on prior imaging) is a non-invasive means of evaluating selected cases with this indication, particularly in cases of suspected ureteric injury where a conventional cystogram might not yield the required information. An example is illustrated in Fig. 14.9. Where there is suspicion of urinary tract injury in a patient who has undergone a trauma series CT, this can be performed with relative ease if the suspicion is raised early as the patient can be scanned after the appropriate excretory phase delay without having to receive a further dose of intravenous contrast (stressing the importance of the secondary survey in trauma). With CT, a better spatial representation of the pattern and extent of injury can be depicted compared to cystography.

Cancer staging As is the case for PTLD and urological cancer, CT is widely used in cancer imaging to establish the presence and extent of distant disease. Staging CT of the chest, abdomen, and pelvis is performed routinely in patients with RCC, 28% of whom have disseminated

chapter 14 

(A)

(B)

computed tomography

(C)

Fig. 14.8  (A) Axial image demonstrating multiple filling defects within the calyceal system (arrows), subsequently diagnosed to be transitional cell carcinoma. (B) Coronal image on the same patient, demonstrating further disease in the proximal ureter (arrow). (C) Multiple filling defects within the bladder (arrowheads) confirmed to be multifocal bladder cancer.

disease at presentation. The lungs, lymph nodes, liver, and bone are common sites of involvement (Dèahnert, 2011). In bladder cancer, magnetic resonance imaging is the most accurate imaging modality for local staging. Staging CT is indicated for cases of confirmed muscle-invasive bladder cancer, including CTU for complete examination of the upper urinary tracts (Stenzl et al., 2009).

The role of CT in other nephrological disease Beyond the spectrum of transplantation, malignant, and surgical pathology, CT has a more limited role in the evaluation of ‘medical’ renal disease. It may demonstrate features of some disease entities such as nephrocalcinosis in renal tubular acidosis, cortical and papillary necrosis, and AIDS-related nephropathy, amongst others.

(A)

(B)

Safety issues and contraindications (Table 14.1) Contrast medium administration (See Chapter 11.) Several practical measures must be put in place before the administration of contrast (The Royal College of Radiologists, 2010). An appropriately trained doctor should be immediately available to manage severe reactions. Within the confines of the radiology department, an appropriately trained individual should be immediately available to identify and manage severe contrast reactions. Resuscitation facilities should be readily available. A patient should not be left unsupervised within 5 minutes of contrast administration, and should remain on the premises for 15 minutes after (30 minutes in patients with an increased risk of contrast reaction) (The Royal College of Radiologists, 2010). Low-osmolar, iodinated contrast agents are generally associated with a low rate of adverse effects (0.153% in large series data) (Hunt et  al., 2009), the majority of which are mild and can be treated within the radiology department. Severe reactions (anaphylactic spectrum) are rare, constituting 3.3% of all adverse effects (Hunt et  al., 2009). Nonetheless, where identified (see ‘Techniques’), appropriate precautions need to be undertaken to reduce the risk of the adverse effects from contrast administration. These are summarized in Table 14.2. The responsibility for deciding on contrast medium use eventually lies with the supervising radiologist (The Royal College of Radiologists, 2010).

Radiation protection

Fig. 14.9  An example of CTU delineating ureteric injury in a patient presenting with vaginal discharge of urine 3 weeks post pelvic surgery. (A) Contrast opacification of a blind-ending left distal ureter (arrowheads) with a fine track of contrast leading into the vaginal cavity (arrow) which is opacified. (B) A separate communication between the vagina and rectum (arrow) is also demonstrated.

(See Chapter 10.) The main disadvantage of CT is the ionizing radiation dose to the patient. The total effective dose of an optimized CTU protocol (low-dose unenhanced scan, followed by a combined nephrographic and excretory phase) acquired by a modern 64-slice CT scanner is 20.1 mSv (Vrtiska et  al., 2009). This is equivalent to just over 9 years of background radiation in the United Kingdom or an added risk of fatal cancer of 1 in 1000. Hence, despite its superior accuracy, CT examinations of the urinary tract should not be undertaken lightly and as with all diagnostic ionizing radiation examinations, should be justified by the clinical

107

108

Section 1  

assessment of the patient with renal disease

Table 14.2  The Bosniak renal cyst classification system Category

Criteria

Management

I (benign)

Non-enhancing, low attenuation (< 20 HU) simple cyst with a hairline-thin wall. No septations, solid components or calcification

No further intervention required

II (benign)

Cystic lesion with one No further of more of the following intervention required characteristics: a few hairline-thin septations which may perceptibly (but not measurably) enhance; thin calcification or short segment of slightly thick calcification along its wall or septa; small (< 3 cm), non-enhancing, high attenuation cyst

IIF (follow-up)

Cystic lesion with one or more of the following characteristics: multiple hairline-thin/minimally thickened septations which may perceptibly (but not measurably) enhance; thick or nodular calcification along its wall or septa; large (> 3 cm), non-enhancing large (> 3 cm), non-enhancing high attenuation cyst

Imaging follow-up to establish stability

III (indeterminate)

Cystic mass with enhancing thickened irregular or smooth walls or septa

Surgery, biopsy

IV (malignant)

Category III features and distinct enhancing soft-tissue components independent of the wall or septa

Surgery

Source data from Israel and Bosniak (2005).

benefit balanced against the radiation risk (The Royal College of Radiologists, 2012). As previously discussed, appropriate selection of patients with a high pre-test probability of disease is key to achieving this balance. This is particularly true in the context of haematuria. This requires clear referral guidelines to be established and adhered to by radiologists and the referrer to have a sound understanding of indications for each examination. Communication between both

parties should always be maintained, to decide on the best imaging pathway for complex cases.

References Anderson, E. M., Murphy, R., Rennie, A. T., et al. (2007). Multidetector computed tomography urography (MDCTU) for diagnosing urothelial malignancy. Clin Radiol, 62(4), 324–32. Blick, C. G., Nazir, S. A., Mallett, S., et al. (2012). Evaluation of diagnostic strategies for bladder cancer using computed tomography (CT) urography, flexible cystoscopy and voided urine cytology: results for 778 patients from a hospital haematuria clinic. BJU Int, 110(1), 84–94. Cohan, R. H., Caoili, E. M., Cowan, N. C., et al. (2009). MDCT urography: exploring a new paradigm for imaging of bladder cancer. AJR Am J Roentgenol, 192(6), 1501–8. Dèahnert, W. (2011). Radiology Review Manual (7th ed.). Philadelphia, PA: Lippincott Williams & Wilkins. Edwards, T. J., Dickinson, A. J., Natale, S., et al. (2006). A prospective analysis of the diagnostic yield resulting from the attendance of 4020 patients at a protocol-driven haematuria clinic. BJU Int, 97(2), 301–5. Hunt, C. H., Hartman, R. P. and Hesley, G. K. (2009). Frequency and severity of adverse effects of iodinated and gadolinium contrast materials: retrospective review of 456,930 doses. AJR Am J Roentgenol, 193(4), 1124–7. Israel, G. M. and Bosniak, M. A. (2005). How I do it: evaluating renal masses. Radiology, 236(2), 441–50. Khadra, M. H., Pickard, R. S., Charlton, M., et al. (2000). A prospective analysis of 1,930 patients with hematuria to evaluate current diagnostic practice. J Urol, 163(2), 524–7. Korobkin, M. (2000). CT characterization of adrenal masses: the time has come. Radiology, 217(3), 629–32. Saw, K. C., McAteer, J. A., Monga, A. G., et al. (2000). Helical CT of urinary calculi: effect of stone composition, stone size, and scan collimation. AJR Am J Roentgenol, 175(2), 329–32. Sebastia, C., Quiroga, S., Boye, R., et al. (2001). Helical CT in renal transplantation: normal findings and early and late complications. Radiographics, 21(5), 1103–17. Smith, R. C., Verga, M., Dalrymple, N., et al. (1996). Acute ureteral obstruction: value of secondary signs of helical unenhanced CT. AJR Am J Roentgenol, 167(5), 1109–13. Stenzl, A., Cowan, N. C., De Santis, M., et al. (2009). The updated EAU guidelines on muscle-invasive and metastatic bladder cancer. Eur Urol, 55(4), 815–25. Tamm, E. P., Silverman, P. M., and Shuman, W. P. (2003). Evaluation of the patient with flank pain and possible ureteral calculus. Radiology, 228(2), 319–29. The Royal College of Radiologists (2010). Standards for Intravascular Contrast Agent Administration to Adult Patients (2nd ed.). London: The Royal College of Radiologists. The Royal College of Radiologists (2012). iRefer: Making the Best Use of Clinical Radiology. London: The Royal College of Radiologists. Vrtiska, T. J., Hartman, R. P., Kofler, J. M., et al. (2009). Spatial resolution and radiation dose of a 64-MDCT scanner compared with published CT urography protocols. AJR Am J Roentgenol, 192(4), 941–8.

CHAPTER 15

Magnetic resonance imaging Kazuhiro Kitajima, Akira Kawashima, and James F. Glockner Introduction Since its introduction in the mid 1980s, magnetic resonance imaging (MRI) has become a powerful imaging tool for the evaluation of patients with renal disease. It is a useful alternative to computed tomography (CT) in patients in whom the use of iodinated contrast media is contraindicated or in patients at risk for radiation exposure. MRI has its particular advantages over other imaging modalities. Many individual properties of the magnetic resonance (MR) signal, such as proton density, relaxation rates (T1 and T2), flow, chemical shift, diffusion and perfusion, contribute to image contrast. The individual weighting of these contributions can be altered by changing the pulse sequence. It demonstrates excellent contrast detail in normal tissues (e.g. kidneys, bladder, and prostate) and pathological lesions are frequently very conspicuous against the background of normal tissue. A variety of MR acquisitions can be performed quickly, within a comfortable breath-hold, and these reduce or eliminate motion artefact as well as acquiring dynamic images following contrast administration. It is an excellent tool for imaging arteries (MR angiography (MRA)) and veins (MR venography (MRV)). A  variety of techniques are available, some of which do not require gadolinium contrast agents. Specific information which can be obtained with MRI include blood flow in arteries or veins, the presence of fat or lipid within a lesion, and apparent diffusion coefficients of normal and pathological tissue.

Principles MRI depends on the magnetic properties of nuclei, predominantly hydrogen, in the form of water molecules. Water protons behave like very small spinning magnets, and in the presence of a large magnetic field will align along the direction of the field, thereby minimizing their energy state and also generating a small net magnetization within the patient. When radiofrequency (RF) pulses of the proper frequency are applied, the water protons absorb energy and alter the orientation of their magnetic vectors with respect to the magnetic field. When the RF pulse is turned off, the protons release energy and relax back to their ground state. The energy released is also in the form of RF waves, which are detected by the antennas or coils within the MRI system and converted into an image. This is akin to spectroscopy. The spatial information required for generation of an image is provided by imposing very

small gradients on the large static magnetic field, thereby changing the resonant frequency of water protons according to their spatial location. There are few but important contraindications to MRI. Patients with pacemakers, ferromagnetic intracranial aneurysm clips, cochlear implants, and metallic ocular foreign bodies cannot be imaged safely. Open or large-bore magnets, sedation, or general anaesthesia may be necessary for examination of paediatric patients or those with claustrophobia. Large patients may be excluded because of limitations in the diameter of the gantry opening or the weight limits of the examining table. MRI can be considered during pregnancy only if the benefit of the procedure outweighs the benefits of alternative non-ionizing diagnostic imaging studies.

Pulse sequences Superior soft tissue contrast resolution is one of the key advantages of MRI over CT. Unlike CT, MRI contrast arises from a complex relationship of many different factors including proton density, magnetic relaxation (T1 and T2 relaxation time constants), magnetic susceptibility, and flow. It is possible to ‘weight’ the relative contribution of these factors and thereby create images with different tissue contrast (i.e. T1 weighted, T2 weighted, proton density weighted, etc.). A pulse sequence is a computer program specifying the details of image acquisition by the MRI system, and by changing various parameters within the sequence it is possible to generate images whose contrast is weighted to different properties of tissue. Two relaxation times, for example, describe the return to the equilibrium state following the initial RF pulse: the T1 relaxation time, or spin lattice relaxation time, is a measure of a proton’s ability to exchange energy with its surrounding chemical matrix, and determines how quickly a tissue can become magnetized. The T2 relaxation time, termed the spin-spin or transverse relaxation time conveys how quickly a given tissue loses its magnetization. By altering two parameters within the pulse sequence, TR (repetition time) and TE (echo time), it is possible to change the image contrast to reflect predominant T1, T2, or proton density weighting.

T1-weighted imaging An image in which the contrast is primarily determined by differences of T1 relaxation values within the tissues is produced by choosing a relatively short TR of 250–600 msec and a relatively short TE of 10–20 msec for a spin echo or fast spin echo pulse sequence. In a T1-weighted image, tissue with short T1 relaxation times appears

110

Section 1  

assessment of the patient with renal disease

bright. T1-weighted images are a good way to assess morphology and anatomy of organs because of relatively stronger signal intensity (Fig. 15.1A). Breath-hold T1-weighted images can be obtained with three- (3D) or two-dimensional (2D) T1-weighted, spoiled gradient echo pulse sequences, which can improve image quality by reducing motion.

times (e.g. kidney) appears brighter (Fig.  15.1B). Pathological conditions such as tumours, inflammation, or oedema will have long relaxation times and will appear bright on T2-weighted images. Static fluid such as urine appears markedly bright on T2-weighted images due to its very long T2 relaxation times (Table 15.1).

T2-weighted imaging

Gadolinium-based magnetic resonance contrast agents and their side effects

A long TR (2000–4000 msec), long TE (60–120 msec) spin echo or fast spin echo pulse sequence produces a T2-weighted image. In a T2-weighted image, tissue with long T2 relaxation

Intravascular MR contrast agents are used to provide additional image contrast during MRI (Fig.  15.1C–E). Gadolinium is a

(A)

(B)

(C)

(D)

(E)

Fig. 15.1  Normal kidneys. (A) Axial T1-weighted spoiled gradient echo (SPGR) image without fat suppression obtained at the level of the mid kidneys reveals kidneys of intermediate signal intensity are surrounded by retroperitoneal fat, which appears bright. The renal sinus appears bright because of fat. A = abdominal aorta; C = inferior vena cava; D = duodenum; PH = pancreatic head. (B) The kidneys appear hyperintense on T2-weighted fast spin echo image with fat suppression. The retroperitoneal fat signal intensities are suppressed. Urine in the renal collecting system (arrows = intrarenal collecting systems; P= renal pelves) appears markedly bright similar to cerebrospinal fluid and fluid in the bowel loops. A = abdominal aorta; C = inferior vena cava; D = duodenum; PH = pancreatic head. (C) Breath-hold dynamic enhanced T1-weighted SPGR image with fat suppression 40 seconds after starting IV administration of gadolinium contrast demonstrates increased enhancement of the cortices (straight arrow) with the corticomedullary interface representing the cortical nephrographic phase. a = proximal main renal arteries; v = proximal left main renal vein; A = abdominal aorta; C = inferior vena cava. (D) Dynamic-enhanced T1-weighted SPGR image with fat suppression 70 seconds after gadolinium injection demonstrates homogeneous renal enhancement representing the nephrographic phase. a = proximal main renal arteries; v = proximal left main renal vein; A = abdominal aorta; C = inferior vena cava. (E) Delayed-enhanced T1-weighted image with fat suppression demonstrates excreted contrast material in the renal collecting systems bilaterally, representing the pyelographic phase. P= renal pelves.

chapter 15 

magnetic resonance imaging

Table 15.1  Relative T1 and T2 values for tissues and body fluid in the genitourinary system T1-weighted image T2\T1 T2-weighted image

Absent (markedly low SI)

Long T1 (low SI)

Intermediate T1 (intermediate SI)

Short T1 (high SI)

Long T2 (markedly high SI)

Urine, simple cysts, oedema, inflammation, neoplasm

Proteinaceous fluid (complicated cysts, abscess)

Subacute haemorrhage (extracellular methaemoglobin)

Long T2 (high SI)

High free water tissue (kidneys, testes, peripheral gland of prostate, seminal vesicle, penis)

Subacute haemorrhage (extracellular methaemoglobin)

Intermediate T2

High bound water tissue (adrenal, central gland of prostate, muscle)

Fat (adipose tissue, fatty bone marrow)

Short T2 (low SI) Absent (markedly low SI)

Gadolinium contrast agent

Air, gas, calculi, compact bone, haemosiderin, ion deposition, highly concentrated gadolinium contrast

SI = signal intensity.

lanthanide metal with a k-shell configuration of unpaired electrons, which results in its paramagnetic character. Gadolinium in the free state is extremely toxic; however, when bound to larger stable molecules it is safe for human use. Gadolinium chelates, following intravenous injection, are predominantly eliminated by the kidneys via glomerular filtration. Gadolinium shortens T1-relaxation times of adjacent tissue. Therefore, T1-weighted pulse sequences are used after intravenous (IV) administration of gadolinium-based MRI contrast agents (GBCAs) (0.1–0.2 mmol/kg). Dynamic enhanced imaging, using breath-hold, 3D or 2D T1-weighted, spoiled gradient echo pulse sequences, permits evaluation of the enhancement properties of parenchymal tissue or masses and can also be used to assess renal arteries and veins (Fig. 15.2).

(A)

Adverse reactions Adverse reactions are encountered with a much lower frequency than is observed after administration of iodinated contrast media. The frequency of all adverse events after IV injection of GBCAs ranges from 0.07% to 2.4% (Cohan et al., 2010). The vast majority of these reactions are mild, including coldness at the injection site, nausea with or without vomiting, headache, warmth or pain at the injection site, paraesthesia, dizziness, and itching. Although most adverse events resulting from administration of iodinated contrast media occur within 30 minutes of administration, reactions to GBCA injection can develop after more than 1 hour after GBCA. Reactions resembling an ‘allergic’ response are very unusual and vary in frequency from 0.004% to 0.7%. Rash, hives, and urticaria are the most

(B)

Fig. 15.2  Renal vein thrombus in a 49-year-old man. Enhanced T1-weighted SPGR images with fat suppression in axial (A) and coronal (B) planes reveal filling defect in the right renal vein. Benign simple cyst is present in the right kidney (A). Specimen obtained from percutaneous needle biopsy of the kidney demonstrated membranous glomerulonephritis.

111

112

Section 1  

assessment of the patient with renal disease

frequent of this group, and very rarely there may be bronchospasm. Severe, life-threatening anaphylactoid or non-allergic anaphylactic reactions are exceedingly rare (0.001–0.01%). Fatal reactions to GBCAs are very rare. Treatment of moderate to severe adverse reactions to GBCAs is similar to that of moderate or severe reactions to iodinated contrast media. Personnel must be trained and the equipment readily available for management and/or resuscitation of patients receiving intravenous GBCAs for MRI. The frequency of acute adverse reactions to GBCA is about eight times higher in patients with a previous reaction to GBCA. Second reactions to GBCAs can be more severe than the first. Persons with asthma and various other allergies, including to other medications or foods are also at greater risk, with reports of adverse reaction rates as high as 3.7%. Although there is no cross-reactivity, patients who have had previous allergic-like reactions to iodinated contrast media are also in this category. In the absence of any widely accepted policy for dealing with patients with prior contrast reactions (especially to GBCAs) and the need for subsequent exposure to MR agents, it does seem prudent to take precautions in a patient who previously had a reaction to GBCAs. It should be determined if is GBCA necessary, if a different brand could be used, and if 12–24 hours of premedication with corticosteroids and antihistamines should be initiated. This protocol is particularly applicable in patients who had prior moderate to severe reactions.

Nephrogenic systemic fibrosis GBCAs previously were widely believed to be well tolerated and non-nephrotoxic, even in patients with impaired renal function. However, exposure to GBCAs, often used in MRI, in patients with renal failure and those maintained on dialysis has been linked with the development of nephrogenic systemic fibrosis (NSF). NSF is a rare disorder with a scleroderma-like presentation, which may progress to potentially fatal multiorgan-system fibrosing disease (Fig 15.3). NSF was first noted in 1997 and reported to the medical community in 2000 (Cowper et al., 2000). Between 1997 and 2007,

Fig. 15.3  Patterned plaques in nephrogenic systemic fibrosis. Plaques can be seen to be red to purple coloured, thin, and fixed. Reproduced from Girardi et al. J Am Acad Dermatol (2011) 65; 6:1095–1106 with permission.

> 500 cases of NSF in patients with severe renal insufficiency (glomerular filtration rate (GFR) < 30 mL/min/1.73 m2) were reported (Abu-Alfa, 2011), and no documented cases of NSF have occurred in patients with a GFR of > 30 mL/min/1.73 m2 without acute kidney injury. Additional major risk factors include use of high doses or repeat doses of GBCAs a proinflammatory state, acute kidney injury of any severity associated with hepatorenal syndrome, or during the perioperative period after liver transplantation. Because treatment options are currently limited, there has recently been an emphasis on prevention and identification of patients who are at increased risk of developing NSF prior to any GBCA injection. The US Food and Drug Administration (FDA) now recommends that specific types of GBCAs including gadodiamide, gadoversetamide, and gadopentetate dimeglumine should not be used for patients with acute kidney injury or chronic, severe renal insufficiency with a GFR < 30 mL/min/1.73 m2. It is advisable to avoid GBCAs in dialysis-dependent patients unless the possible benefits clearly outweigh the potential risk, and to limit the type and amount in patients with estimated GFR rates < 30 mL/min/1.73m2.

Magnetic resonance angiography MRA is accepted as an accurate and non-invasive technique for evaluation of patients with suspected renal artery stenosis and other renovascular diseases. Contrast-enhanced MRA utilizing a 3D spoiled gradient echo sequence is a time efficient and safe test when compared with conventional catheter-directed arteriography (Fain et al., 2001). Improved hardware and acquisition techniques have led to shorter acquisition times and improved spatial resolution, and state-of-the-art MRA now provides detailed anatomical information of the renal arteries (Fig. 15.4) and has proven highly sensitive and specific for detection of renal artery disease, including

Fig. 15.4  Gadolinium-enhanced three-dimensional MRA reveals normal single main renal arteries bilaterally.

chapter 15 

Fig. 15.5  Bilateral renal artery aneurysms in a patient with Ehlers–Danlos syndrome. Gadolinium-enhanced 3D MRA demonstrates aneurysmal dilation of the bilateral renal arteries at the renal artery bifurcation.

fibromuscular dysplasia, and aneurysms (Figs. 15.5 and 15.6). The visibility of the intrarenal vessels remains limited when compared to conventional angiography. Although false-negative studies are rare on MRA, slight overestimation of the degree of stenosis can

(A)

magnetic resonance imaging

occur and may lead to false-positive diagnoses. The combination of an anatomic MRA acquisition with a velocity-dependent technique (phase-contrast imaging) can help to minimize the tendency to overestimate stenoses (Schoenberg et al., 2002). As with conventional angiography, MRA is an anatomic test, and provides little information regarding the functional significance of a stenosis. It is highly accurate in determining the number of renal arteries, the size of the kidneys, and the presence of any anatomic variants. MRI can be combined with a functional test similar in concept to captopril nuclear renography (Chapter 16). This test, termed MR renography together with MRA may be helpful to determine functional significance in the work-up of renovascular hypertension (Bokacheva et al., 2009). Non-contrast techniques were the first MRA methods developed and include time-of-flight and phase-contrast MRA. These had several limitations, however, including motion artefact, long acquisition times, and insensitivity to in-plane flow, and were largely surpassed by 3D contrast-enhanced methods; however, the advent of NSF has stimulated a renewed interest in non-contrast MR angiography. A recently developed technique employing 3D steady-state free precession (SSFP) with arterial spin labelling has advantages including high vascular signal to noise ratio, fairly rapid acquisition times, and inherent flow compensation (Glockner et  al., 2010; Miyazaki and Akahane, 2012). This has shown sensitivities comparable to contrast-enhanced MRA for detection of renal artery stenosis in limited studies. Non-contrast MRA also tends to overestimate the degree of stenosis by signal loss due to post-stenotic turbulent rapid flow.

Magnetic resonance urography MR urography is an alternative to CT urography (Chapter 14), but is generally not used as the first-line examination for work-up of patients with haematuria. MRI is insensitive to urolithiasis

(B)

Fig. 15.6  Bilateral renal artery stenoses. (A) Gadolinium-enhanced 3D MRA demonstrates a high grade stenosis of the single right main renal artery (short arrow) and focal stenosis of the accessory left renal artery (long arrow). The left main renal artery is negative for stenosis. (B) Digital subtraction catheter-directed aortogram corresponds to the MR angiographic findings of bilateral renal artery stenoses indicated with short and long arrows.

113

114

Section 1  

assessment of the patient with renal disease

Fig. 15.7  Coronal view of maximal intensity projection of MR urogram utilizing maximal intensity projection of respiratory gated, heavily T2-weighted fast recovery fast spin echo sequence with use of contrast media. Urine (U) in the right renal collecting system and ureter appears markedly bright similar to fluid filled structures including cerebrospinal fluid (CSF), fluid-filled gastric fundus (S) and bowel loops, and biliary system. Prior left nephroureterectomy and cystectomy with ileal conduit urinary diversion for urothelial carcinoma.

because stones do not have signal characteristics that allow them to be readily detected. Several different approaches are available. One common technique obtains fast heavily T2-weighted hydrographic images using (a) 2D half-Fourier transformation fast spin echo sequences acquired in a single breath-hold and (b) 3D fast recovery fast spin echo sequence with respiratory gating (Fig. 15.7). These static–fluid MR urography techniques provide excellent images of the urinary tract, particularly when dilated and/ or obstructed. Another popular method is gadolinium-enhanced excretory MR urography. This technique is similar in concept to CT urography, and relies on visualizing the excretion of a gadolinium contrast agent in the urinary tract with a 3D, spoiled gradient echo T1-weighted acquisition. Visualization of the collecting system and ureters is further improved with a simultaneous dose of furosemide (5–10 mg) and is particularly useful in assessing non-obstructed urinary tracts (Takahashi et al., 2010). One advantage of MR urography in comparison to CT urography (Fig. 15.8) is that multiple images can be acquired over time until optimal visualization of the urinary tract has been achieved, without any concerns regarding patient radiation.

Functional magnetic resonance imaging Diffusion-weighted imaging (DWI) is a technique employing strong bipolar gradients to create a sensitivity of the MR signal to the thermally induced Brownian (or random walk) motion of water molecules and allowing in vivo measurement of molecular diffusion. The apparent diffusion coefficient (ADC) is a quantitative

Fig. 15.8  Coronal view of maximal intensity projection of MR urogram using breathold, T1-weighted 3D spoiled gradient echo sequence obtained during excretory phase of contrast enhancement. Prior right nephroureterectomy and cystectomy with ileal neobladder urinary diversion for urothelial carcinoma. Neoblad = ileal neobladder.

parameter calculated from DWI which is used as a measure of diffusion. Because the ADC is also dependent on capillary perfusion and water diffusion in the extravascular space, alteration of the ADC provides information regarding microstructural changes. ADC values of the kidneys are higher than ADC values of other abdominal organs, most likely due to high water content and abundant blood supply of the kidneys, with possible contributions from flow in the tubular system. In most studies, the ADC value of the renal cortex is higher than that of the renal medulla, presumably because of the higher perfusion component in the renal cortex. Measurements of renal ADC may be useful for characterization of diffuse disease as well as focal lesions. Restricted diffusion, characterized by high signal intensity on DWI and decreased ADC value, tends to indicate higher cell density (e.g. malignant neoplasm) and increased viscosity (e.g. purulent material in abscess and pyonephrosis). Other functional imaging techniques including diffusion tensor imaging (DTI), blood oxygen level-dependent (BOLD), and arterial spin labelling (ASL) are currently under investigation (Notohamiprodjo et  al., 2010). MR DTI might provide further information about the renal microstructure. It features diffusion measurements along at least six different directions from which the full diffusion tensor and thus the main diffusion direction can be calculated. BOLD MRI is used for non-invasive assessment of the intrarenal oxygen content and uses the paramagnetic effect of deoxyhaemoglobin for indirect depiction of renal oxygenation. ASL techniques can be used to study tissue perfusion without administration of contrast agents. Most ASL techniques, like flow-sensitive alternating inversion-recovery, apply arterial pre-saturation pulses to invert or saturate the magnetization of arterial blood flowing into a recorded section.

chapter 15 

Interpretations and indications Kidney On T1-weighted images, the kidney is outlined by the surrounding retroperitoneal fat (Fig. 15.1A) and the renal cortex is isointense or slightly higher in signal intensity than the medulla, with the degree of corticomedullary differentiation depending on the patient’s age and hydration status. On T2-weighted images, the kidney appears bright (Fig. 15.1B) and the cortex is isointense or slightly hypointense compare to the medulla. The most common uses of MR in the evaluation of the kidneys include MRA of the renal arteries, MRI of indeterminate renal masses shown on other imaging studies (Pedrosa et al., 2008), and staging of known renal malignancies (Reznek, 2004). MR evaluation of indeterminate renal masses should begin with axial T1-weighted images (with and without fat saturation) and axial fast spin echo T2-weighted images. These images allow for proper identification of the mass as well as providing vital tissue characterization with regard to potential fat content or haemorrhagic components. The most vital portion of an MR examination of an indeterminate renal mass is a dynamic gadolinium-enhanced T1-weighted spoiled gradient recalled echo (SPGR) pulse sequence. This is generally performed in an axial or coronal plane in a single breath-hold utilizing a 3D- or 2D-SPGR sequence (Fig. 15.1C–E). The images are acquired before gadolinium administration, at 40 seconds (the cortical nephrographic phase), at 70 seconds (the early homogeneous nephrographic phase), and at 120 seconds (the late homogeneous nephrographic phase) after gadolinium administration. Alternatively, a small test dose of contrast may be administered to determine the timing delay following contrast injection to achieve optimal visualization of the cortical phase. Excreted contrast medium starts to appear in the collecting system (the pyelographic phase) 3 minutes following IV contrast injection. When a renal mass is seen on MRI, the most important characteristic indicating the presence of a malignancy is enhancement of the mass by more than 15% above baseline after IV contrast enhancement (Ho et al., 2002). Subtraction of the pre-contrast from the post-contrast images may be helpful to detect contrast enhancement of a renal mass when the lesion contains haemorrhagic or proteinaceous contents and appears bright on pre-contrast T1-weighted images. MRI is the most accurate method for the detecting renal vein and inferior vena cava (IVC) tumour thrombi in patients with renal cell carcinoma compared to ultrasonography and enhanced CT (Reznek, 2004). MRI is also accurate in determining whether the filling defects in the renal vein and IVC represent bland thrombus or tumour thrombus (Aslam Sohaib et al., 2002). Gadolinium-enhanced MRA and MRV in addition to cross-sectional and 3D display can also provide pre-surgical ‘road mapping’ prior to surgery. This can be especially helpful in nephron sparing surgery. MRA depicts renal arterial and venous lesions with exquisite detail and is also becoming a primary imaging study in patients with suspected renal aneurysms, arteriovenous communications, and renal vein occlusion. These acquisitions can be combined with functional techniques that can be used to make rough assessment of overall and split renal function as well as renal blood flow.

Ureter The intrarenal collecting systems are generally not well visualized on routine imaging unless dilated. The renal pelvis appears dark on T1-weighted images and markedly bright on T2-weighted images,

magnetic resonance imaging

characteristics of a fluid filled structure. The renal pelvis and ureter can be identifiable on T1-weighted images by the surrounding retroperitoneal fat. The normal ureteral wall is not well visualized. MR urography is an alternative to CT urography and excretory urography to assess the upper urinary tract. T2-weighted MR urography utilizing breath-hold heavily T2-weighted images is helpful in depicting a dilated upper urinary tract and in identifying the level of urinary obstruction. Gadolinium-enhanced MR urography is useful to improve detection of upper tract urothelial carcinoma (Takahashi et al., 2010).

Bladder The bladder wall is well demarcated by perivesical fat but is often indistinguishable from the low-intensity urine on T1-weighted images. On T2-weighted images, the low-intensity bladder wall is well outlined by high-intensity urine. With its multiplanar capability and various pulse sequences that allow for fast dynamic enhanced imaging, MRI has the potential to become the imaging modality of choice in staging bladder cancer. With MRI, all of the observations previously possible with CT are not only possible with MRI, but in addition, MRI can identify macroscopic perivesical fat invasion and may be helpful in differentiating between superficial and deep muscle invasion of the bladder wall. MR can detect enlarged abdominal and pelvic lymph nodes. MRI is also useful for the evaluation of patient with stress urinary incontinence. Pelvic floor laxity and abnormalities of the supporting fascia can be demonstrated in incontinent women by obtaining fast sagittal and coronal T2-weighted images at rest and at maximal pelvic floor strain, allowing for the detection and delineation of cystocoeles, rectocoeles, enterocoeles, and uterine prolapse.

Prostate and seminal vesicles The prostate is well outlined by fat and is homogeneous in signal intensity on T1-weighted MR images. The glandular prostate is subdivided into the peripheral gland and the central gland. T2-weighted images provide the best depiction of the prostate zonal anatomy and capsule. On T2-weighted images, the peripheral gland appears as a high-signal-intensity area in the posterior and posterolateral aspect of the gland surrounding the intermediate to low-intensity central gland. In young men, the peripheral gland generally constitutes 70% of the glandular tissues and is very bright on T2-weighted images. In older men, the central gland is composed mainly of the enlarged transitional zone from benign prostatic hyperplasia, and becomes heterogeneous on T2-weighted images. Prostatic cancer typically appears as areas of decreased signal in the peripheral gland on T2-weighted images. Unfortunately, prostatitis, post-biopsy haemorrhage, and atrophy can have a similar appearance. Cancer in the transition zone is shown as a homogeneous low signal mass with indistinct margins on T2-weighted images. Additional techniques including DWI, dynamic contrastenhanced (DCE) imaging, and proton (1H) MR spectroscopy can help further increase the level of confidence in detection and characterization of prostate cancer by providing functional and molecular information (Hoeks et al., 2011). DWI is an emerging clinical tool which does not require contrast administration: recent reports have demonstrated that ADC maps can improve detection and estimation of tumour aggressiveness. DEC imaging is the most common method for direct depiction of tumour vascularity and perfusion characteristics and can improve detection of cancer.

115

116

Section 1  

assessment of the patient with renal disease

(A)

(B)

Fig. 15.9  Left suprarenal mass (arrow) appears significantly dark on coronal out-of-phase T1-weighted spoiled gradient echo image (A) compared to that on in-phase image (B), characteristic of benign adenocortical adenoma with abundant intracellular lipids.

1H MR spectroscopy of the prostate is able to depict spectral profiles of cellular metabolites and characterize cancerous lesions by demonstrating reduced citrate and elevated choline peaks. MR spectroscopic imaging combined with MRI can improve tumour localization and volume measurement compared with conventional MRI alone. However, MR spectroscopy needs specialized software and a high level of expertise in scanning and interpretation. The seminal vesicles appear as symmetrically paired ovoid cystic structures. MRI is valuable for the evaluation of patients with ejaculatory dysfunction (e.g. haematospermia, painful ejaculation) and useful in detecting stones, masses, and obstruction of the seminal vesicles and ejaculatory ducts.

Urethra High-resolution MRI with phased-array pelvic and endorectal coils has dramatically enhanced the ability to visualize abnormalities of the female urethra and periurethral tissue. MRI is accurate in diagnosing diverticula of the female urethra. The male urethra is rarely visualized on routine images unless a transurethral Foley catheter is inserted. MRI is useful in staging carcinoma of the male and female urethra.

Adrenal gland (Fig. 15.9) Normal adrenal glands appear as an inverted V or Y configuration. Most normal adrenal glands are well outlined by the retroperitoneal fat on axial T1-weighted images; this may be augmented by thin sectioning or coronal or sagittal views. The gland appears hypointense to the liver and isointense to skeletal muscle. T2-weighted images are unnecessary if the adrenal glands appear normal on high quality T1-weighted images; differentiation of the adrenal cortex from the medulla is generally not possible on routine MR images. Out-of-phase and in-phase chemical shift gradient echo pulse sequences can differentiate benign adrenocortical adenomas from metastases. Benign adrenal adenomas naturally contain a moderate amount of intracellular lipid (lipid rich adenoma), whereas metastases do not. Out-of-phase MR images result in a cancellation of signal in lipid-rich benign adenomas because the signal contributions from intracellular lipid and water oppose each other and cancel the resultant signal within the pixel. The characteristic MR findings of phaeochromocytomas include extremely high signal intensity on T2-weighted images and increased contrast

enhancement. Unfortunately, not all phaeochromocytomas have these imaging characteristics.

References Abu-Alfa, A. K. (2011). Nephrogenic systemic fibrosis and gadolinium-based contrast agents. Adv Chronic Kidney, 18, 188–92. Aslam Sohaib, S. A., Teh, J., Nargund, V. H., et al. (2002). Assessment of tumor invasion of the vena caval wall in renal cell carcinoma cases by magnetic resonance imaging. J Urol, 167, 1271–5. Bokacheva, L., Rusinek, H., Zhang, J. L., et al. (2009). Estimates of glomerular filtration rate from MR renography and tracer kinetic models. J Magn Reson Imaging, 29, 371–82. Cohan, R. H., Jafri, S., Choyke, L. P., et al. (2010). Manual on Contrast Media, Version 7. Reston, VA: American College of Radiology. Cowper, S. E., Robin, H. S., Steinberg, S. M., et al. (2000). Scleromyxoedema-like cutaneous diseases in renal-dialysis patients. Lancet, 356, 1000–1. Fain, S. B., King, B. F., Breen, J. F., et al. (2001). High-spatial-resolution contrast-enhanced MR angiography of the renal arteries: a prospective comparison with digital subtraction angiography. Radiology, 218, 481–90. Glockner, J.F., Takahashi, N., Kawashima, A., et al. (2010). Non-contrast renal artery MRA using an inflow inversion recovery steady state free precession technique (Inhance): comparison with 3D contrast-enhanced MRA. J Magn Reson Imaging, 31, 1411–18. Ho, V. B., Allen, S. F., Hood, M. N., et al. (2002). Renal masses: quantitative assessment of enhancement with dynamic MR imaging. Radiology, 224, 695–700. Hoeks, C. M., Barentsz, J. O., and Hambrock, T. (2011). Prostate cancer: multiparametric MR imaging for detection, localization, and staging. Radiology, 261, 46–66. Miyazaki. M. and Akahane, M. (2012). Non-contrast enhanced MR angiography: established technique. J Magn Reson Imaging, 35, 1–19. Notohamiprodjo, M., Reiser, M. F., and Sourbron, S. P. (2010). Diffusion and perfusion of the kidney. Eur J Radiol, 76, 337–47. Pedrosa, I., Sun, M. R., Spencer, M., et al. (2008). MR imaging of renal masses: correlation with findings at surgery and pathologic analysis. Radiographics, 28, 985–1003. Reznek, R. H. (2004). CT/MRI in staging renal cell carcinoma. Cancer Imaging, 14, 25–32. Schoenberg, S. O., Knopp, M. V., Londy, F., et al. (2002). Morphologic and functional magnetic resonance imaging of renal artery stenosis: a multireader tricenter study. J Am Soc Nephrol, 13, 158–69. Takahashi, N., Glockner, J, F., Kawashima, A., et al. (2010). Gadolinium enhanced magnetic resonance urography for upper urinary tract malignancy. Urology, 183, 1330–6.

CHAPTER 16

Radioisotopes in diagnostic imaging in nephrology Ramya Dhandapani and Sobhan Vinjamuri Introduction to nuclear medicine Nuclear medicine uses functional techniques to assess the different physiological processes of the kidneys. The early understanding of kidney function relied heavily on nuclear medicine imaging. Advances in other modalities such as ultrasound, computed tomography (CT), and magnetic resonance imaging (MRI) are mow preferred for identification of structural abnormalities. There are still some fundamental clinical questions that can be quickly and conclusively answered by isotope imaging techniques such as renal plasma flow, glomerular filtration rate (GFR), renal transit time, and assessment of outflow obstruction. Although the exposure to ionizing radiation should be considered, risk:benefit analysis in individual patients usually favours the clinical value of a test justifying a small additional radiation burden.

Introduction to radioactivity Atoms of all elements are composed of known arrangements of protons, neutrons, and electrons which characterize them as individual nuclides. Nuclides containing the same number of protons, the same atomic number, and the same chemical properties are known as isotopes. The most prevalent isotopes are usually stable and comprise naturally occurring elements. Radioisotopes are unstable members of the group of radioisotopes of an element. The nuclei of radioisotopes undergo rearrangement and change to a stable form, emitting radiation in the process. A radionuclide is a specific radioactive atom, designated by indicating the element and its atomic mass such as iodine-131 or technetium (Tc)-99m. The radiation emitted by a radionuclide enables the detection of extremely small masses, below the limits of chemical detection. It is thus possible to use radionuclides as true tracers for substances without introducing excessive amounts. The radiation emitted by radionuclides is characteristic and unique in terms of the rate of decay, type of radiation, and the energy of that radiation. Radionuclides used in nuclear medicine may emit alpha particles, beta particles, or gamma radiation. Alpha particles are large particles that are emitted from the nucleus with high energy levels. It is a charged particle with a charge of +2 as it has lost two of its electrons and is essentially a helium ion. Beta particles are similar in terms of charge and other physical properties to electrons. While ordinary electrons are found in the electronic shells orbiting the nucleus, beta-minus particles are emitted from the nuclei. They have high kinetic energy. Gamma rays are analogous to energy

packets emitted from the nucleus as part of its emissions to enable the achievement of a more stable physical state. Alpha and beta emitters are commonly used for radionuclide therapy. Gamma rays are more readily transmitted through tissue and therefore the internal administration of gamma emitting radionuclides allows external measurements and imaging of patients using a gamma camera. When the radionuclides are combined with a chemical or a pharmaceutical compound with particular physiological properties, the resultant compound is a radiopharmaceutical. These compounds are subject to strict pharmaceutical controls as is every other medicine suitable for human use. Clinical nuclear medicine usually involves the detection and quantification of ionizing radiation emitted from radioactive substances. A gamma camera produces an image corresponding to the distribution of radioactive substances in the body. Computer display and enhancement of the images with numerical assessment is frequently employed. Tc-99m is the most commonly used radionuclide in nuclear medicine imaging. It has a half-life of 6. 02 hours and this allows patients to attend for outpatient procedures and travel to home after the procedure without any strict radiation precautions. It has a gamma ray energy of 140 keV which is optimally primed for use with the modern gamma camera. The image quality is therefore of a very high quality. It can be readily complexed with a range of compounds to assess different physiological functions in the body and does not have any pharmacological action by itself. It is not toxic and does not elicit an immune response when injected into humans.

Radiation exposure and effective dose Radiation is a property of matter that is all around us and is much more common than people realize. Radiation is analogues to the passing of energy through matter. Most of the radiation does not change the environment it passes through and is called non-ionizing radiation. However, when radiation interacts with the matter as it passes through, then it is called ‘ionizing radiation’. When the energy associated with the radiation is deposited in a particular tissue it is termed as ‘absorbed dose’. This property is useful when a therapeutic benefit is required, however, in most diagnostic settings, this is not helpful and the risk to the individual should be weighed against the benefit of a result that can have a useful impact on patient management.

118

Section 1  

assessment of the patient with renal disease

Effective dose is an estimate of the total radiation burden to the patient from the exposure to the radioactive compound. This is a useful term that also allows comparison of relative radiation exposure across all modalities involving exposure to different forms of ionizing radiation including X-rays, CT scans, nuclear medicine procedures, and positron emission tomography (PET) scans. Absorbed organ doses in patients and effective doses are not measurable quantities but are based on estimates. Most of what is known about the carcinogenic effects of radiation is derived from studies of atomic bomb survivors. There are additional studies involving medical uses of radiation which have provided some epidemiological data. Radiation exposure is commonly known to be associated with some dose-dependent toxicities such as hair loss and acute radiation sickness. The development of cancer was believed to be a non-dose-dependent association of radiation. However, recent studies have strongly suggested a dose-dependent increase in cancer mortality. Some cancers are caused in children when the exposure occurred during pregnancy (teratogenic), or the risk has passed on to the next generations when the exposure occurred while the patient was not pregnant (genetic).

Principles of imaging the renal tract using radionuclides The three broad functions of the kidney that can be assessed using radionuclides include glomerular filtration, tubular secretion, and cortical function/activity. Agents that are physiologically inert and cleared exclusively by glomerular filtration are used to assess the GFR. These include chromium (Cr)-51 ethylenediaminetetraacetic acid (EDTA) and Tc-99m diethylenetriaminepentaacetic acid (DTPA). Tc-99m MAG3 is a tubular agent that has a higher extraction efficiency from the blood and is easy to prepare. Tc-99m dimercaptosuccinic acid (DMSA) is predominantly a cortical agent that provides an index of cortical function or activity (see Table 16.1).

Routine clinical indications for isotope renography (See Table 16.2) ◆ Assessment of differential kidney function ◆ Assessment of possible pelviureteric obstruction ◆ Assessment of renal damage post pyelonephritis ◆ Renal transplant imaging ◆ Assessment of vesicoureteric reflux ◆ Assessment of renal artery stenosis.

Patient preparation The most important prerequisite for the renogram is an adequate state of hydration of the patient which has to be maintained during the test. It is important to avoid an oliguric state since the result can (because of sluggish urine flow) mimic an obstructive pattern. The bladder is emptied immediately before the examination and the patient is positioned in a sitting or prone position with the gamma camera against their back. It is important to position the patient comfortably as they need to maintain the same posture for 20–40 minutes. After intravenous injection of the appropriate dose of the radiopharmaceutical, images are acquired for 20–40 minutes. The patient is then asked to empty their bladder at the end of the procedure to reduce the radiation dose to the pelvic organs.

Interpretation Studies are analysed by producing clear, summed computer images and defining regions of interest over each kidney and the urinary bladder. Curves are then obtained from the detected count rate against time for each region and the obtained curves are expressed as a percentage of the injected dose. The normal renogram shows three classic phases. Immediately following

Table 16.1  Some common isotopes and their routine clinical indications. Please see text for further details Radioisotope

Common indication(s)

Comment

Cr-51 EDTA

Glomerular filtration

Gold standard for estimation of renal function for research, oncology, and transplant donor evaluation

Tc-99m MAG3

Differential renal function Effective renal plasma flow Captopril renography Transplant renography

Useful to assess function and upper tract obstruction

Tc-99m DMSA

Renal scarring and reflux

Provides information on relative renal function, particularly when kidneys are rotated or at different depths

Tc-99m DTPA

Glomerular filtration

Used in non-obstructed kidneys

F-18 fluoride PET-CT scans

Staging of cancers

Estimate of osteoblastic bone lesions

F-18 FDG PET-CT scans

Screening tool

In systemic illness such as PUO, immunocompromised, vasculitis, metastases, and to assess tumour recurrence post nephrectomy It can also be used to stage post-transplant lymphoproliferative disease

Ga-67 citrate, In-111-labelled leucocytes

Suspected infection

Intense uptake in infection of the kidneys

Table 16.2  Some common clinical questions and suggested procedures. Please see text for further details Clinical question(s)

Procedure to request

Comment

Need accurate estimate of relative renal function

DMSA renal scan

In situations where accuracy is paramount or where renographic estimate is likely to be difficult (e.g. in infants). No information is provided on the status of outflow tract

Suspected upper tract obstruction

Diuresis MAG3 renogram

Estimate of relative function will be routinely provided

Suspected renal scarring

DMSA renal scan

Estimate of relative renal function will be routinely provided

Need accurate estimate of absolute GFR in mL/min

GFR measurement (Cr-51 EDTA)

Could also use Tc-99m DTPA

Need estimate of both relative and absolute GFR

Basic DTPA renogram with GFR

In principle, both can be measured by a single injection of Tc-99m DTPA. However, in children (or adults with compromised renal function) it is preferable to inject Tc-99m MAG3 and Cr-51 EDTA simultaneously. Absolute GFR is measured by blood sampling

Assessment of possible renovascular hypertension

Captopril renogram

Both DTPA and MAG3 can be used

Higher confidence in possibility of acute tubular necrosis

MAG3 renogram

Good perfusion, progressive uptake, and no excretion

Suspected kidney infection

In-111 WBCs or Ga-67 citrate

In-111 white blood cells are not normally excreted by the renal tract. Hence any renal uptake is usually abnormal

Suspected systemic condition such as vasculitis or carcinoma

F-18 FDG PET-CT scanning

FDG PET-CT scans have high sensitivity but lower specificity. Hence further focused testing of abnormal areas is required for confirmation

(A) RENOGRAM

(B) 10.8

Left Kidney Rel Uptake 50.9 Right Kidney Rel Uptake 49.1

% U p t a k e

0 RT KIDNEY 3 MIN % UPTAKE 16 MIN % UPTAKE MAX % UPTAKE TIME TO PEAK (Mins) 3 MIN RELATIVE % UPTAKE

10.4

LT KIDNEY

2

4

6

8 10 Time (mins)

10.7

2.9

2.4

10.5

10.8

3.5

3.2

49.1

50.9

Fig. 16.1  (A) Renogram showing normal tracer activity in the bladder. (B) Graph showing normal tracer activity in the bladder.

12

14

16

120

Section 1  

assessment of the patient with renal disease (A) RENOGRAM (B) 5.0

Left Kidney Rel Uptake 15.5 Right Kidney Rel Uptake 84.5

% U p t a k e

0 RT KIDNEY

LT KIDNEY

3 MIN % UPTAKE

5.0

0.9

16 MIN % UPTAKE

2.3

1.1

MAX % UPTAKE

5.0

1.1

3.2

15.8

84.5

15.5

TIME TO PEAK (Mins) 3 MIN RELATIVE % UPTAKE

2

4

6

8 10 Time (mins)

12

14

16

Fig. 16.2  (A) Renogram showing abnormal tracer activity in the bladder. (B) Graph showing abnormal tracer activity in the bladder.

the intravenous injection of the radiopharmaceutical, there is a rapid rise which reflects the vascular supply to the kidney. The second phase is a more gradual slope which corresponds to the renal handling of the tracer by the kidneys and is dependent on various factors such as supply rate, extraction efficiency, intra luminal transit, and excretion. In a normal kidney, the curve reaches a peak at 2–5 minutes and activity starts to leave the renal area which is the beginning of the third phase. At this point the tracer activity starts to appear in the bladder and this is now predominantly the excretory phase. See Fig. 16.1 for normal appearances and Fig. 16.2 for examples of abnormal uptake and excretion.

Limitations of routine isotope renography The value of assessing function and physiology of the renal tract by isotope renography is tried and tested and the benefits usually outweigh the small radiation-associated risks. Because the dynamic phase of renal function needs to be assessed, it is vital that the radioactive tracer or any related medication such as diuretics should not be extravasated into the interstitial tissue at the site of intravenous injection. Insufficient hydration, either self-induced or a result of medication or co-morbidities, can result in delayed uptake and excretion by one or both kidneys and may be misinterpreted as poor function. Significant patient motion during the study may introduce errors in the activity–time curve analysis.

DTPA versus MAG3 renography Tc-99m-labelled DTPA is a physiologically inert compound that is predominantly excreted via the glomeruli (approximately 90% excreted by 4 hours). Up to 10% of the injected activity is bound by plasma proteins and is not available for excretion and therefore the net value may represent a slight underestimation. It does not get secreted or filtered by the tubules and also does not localize to the parenchyma/cortex. Tc-99m labelled MAG3 is a predominantly tubular secretory agent (95% tubular secretion vs 5% glomerular clearance). Although we cannot measure GFR with this agent, another aspect of renal function, that is, effective renal plasma flow can be measured and this represents a surrogate marker of global renal function. Because of the relative ease of preparation of Tc-99m MAG3, better visualization of the renal parenchyma, unpredictable nature of protein binding of Tc-99m DTPA, better image quality, and relative reproducibility of tubular function at poorer renal function, Tc-99m MAG3 is, for most routine indications, the preferred test over Tc-99m DTPA renography.

Choosing MAG3 or DMSA Tc-99m DMSA scanning provides information mainly on cortical or parenchymal activity. This has been widely considered a better test to estimate differential kidney function mainly because it is a parenchymal agent and routine imaging includes anterior as well as posterior. It is particularly useful when both kidneys are located at different depths or rotated with respect to one another. It is also the

Chapter 16 

radioisotopes in diagnostic imaging in nephrology

Fig. 16.3  Tc-99m DMSA scanning example.

preferred agent when one kidney is functioning poorly and there is a requirement to assess the differential function to a higher degree of accuracy. Tc-99m DMSA is clearly the agent of choice in suspected pyelonephritis and to assess the extent of cortical damage or scarring. See Fig. 16.3. Tc-99m MAG3 is the preferred agent of choice for any assessment of excretion or possible obstruction. The early parenchymal phase of the Tc-99m MAG3 renogram can be used to assess the parenchymal integrity and the differential function. In routine clinical practice, where there is an expectation that both kidneys are located at the same depth and morphologically similar, Tc-99m MAG3 does provide a good estimate of the renal function. Due to the high radiation burden associated with Tc-99m DMSA renography and the fact that Tc-99m MAG3 provides additional information about renal function, Tc-99m MAG3 can be used to provide an index of differential function in most clinical situations.

Differential renal function Biochemical tests provide a good estimate of overall renal function, but it is useful to know individual kidney function in selected situations such as assessment of potential living kidney donors. Anatomical tests such as an ultrasound and CT rely on size symmetry between the two kidneys and in most cases this should be adequate. However, it is now well accepted that even if the kidneys have the same size, they may contribute differently to the overall renal function. It is important to quantify this asymmetry and isotope tests provide quick, easy and reproducible information.

All three of the commonly used isotopes (Tc-99m DTPA, Tc-99m MAG3, and Tc-99m DMSA) can be used to provide an estimate of the differential function. However, in routine clinical practice, Tc-99m MAG3 is preferred and in patients with known asymmetry of depth or rotated kidneys, Tc-99m DMSA is preferred.

Chromium-51 EDTA estimation GFR is the accepted standard measure of renal function. It is routinely measured using tracers that are cleared exclusively by glomerular filtration, the most common being Cr-51 EDTA and Tc-99m DTPA. GFR is best assessed by Cr-51 EDTA. Clearance values are typically 85% of those using inulin which although gold standard is difficult and time-consuming and is inappropriate for routine clinical use. Since Cr-51 EDTA is expected to be excreted solely by the glomerular filtration route, it has been shown that the rate of disappearance of this compound from the blood is proportional to the renal clearance. Hence there is no need to obtain urinary samples. The commonly used protocols involve at least two timed venous blood samples at 2–3 hours post intravenous injection of the radioactive tracer. The more samples that are obtained, the higher the mathematical accuracy. It is commonly used for monitoring the effect of nephrotoxic drugs, calculating dose in chemotherapy, detection of renal failure with inconclusive serum creatinine values, and in the assessment of potential live donors for transplantation. The values should be interpreted with caution in patients with ascites, oedema, or other expanded body fluid space and in patients receiving intravenous fluids where the renal clearance value may be overestimated. Final values may be corrected for body surface area

121

122

Section 1  

assessment of the patient with renal disease (A)

(B)

2.1

0.2

Left Kidney Rel Uptake 47.4 Right Kidney Rel Uptake 52.6

%

%

U p t a k e

U p t a k e

0

2

4

6

8 10 Time (mins)

12

14

16

0

Left Kidney Rel Uptake 41.3 Right Kidney Rel Uptake 58.7

2

4

6

8 10 Time (mins)

12

14

16

Fig. 16.4  (A) Graph showing the captopril phase. (B) Graph showing the captopril phase 2.

rather than for weight, as in the paediatric population, the body surface area corrected GFR increases with age from birth up to 2 years and remains constant into adulthood thereafter.

Captopril renography Captopril renography has been of proven value in the identification of renovascular hypertension in patients with high blood pressure. This test also has a value in the evaluation of asymmetrically sized kidneys and to confirm or exclude functional impairment secondary to renal artery stenosis. The test has two phases: a ‘baseline phase’ to assess the baseline differential function and excretory patterns; a second ‘Captopril phase’ study is done on another day after oral administration of captopril and repeating the renogram. A reduction in renal uptake (unilateral reduction by at least 10%) or prolonged retention of the radiotracer in the renal parenchyma after administration of captopril in comparison to the baseline study is considered a positive indicator of renal artery stenosis causing renovascular hypertension. The captopril phase can also be performed with other angiotensin-converting enzyme inhibitors such as lisinopril or enalapril and the longer-acting medication does not need to be stopped for the ‘captopril phase’ (see Fig. 16.4).

Renal failure Tc-99m MAG3 is the isotope of choice in renal failure as better renal extraction improves visualization of the kidneys. The pattern of the renogram curve can be helpful to identify the cause of acute kidney injury. Prerenal causes will affect the initial perfusion part of the curve. Renal parenchymal diseases affect the tubular function part of the curve and postrenal causes affect the excretory part of the curve. In acute arterial occlusion, there is severely impaired blood flow, nil or poor tracer uptake, with poor or absent excretion. Acute renal vein thrombosis shows an enlarged kidney with poor perfusion and prolonged cortical tracer retention. In causes of prerenal failure such as prolonged hypoperfusion or hypotension, there is normal tracer uptake with delayed excretion and drainage. Parenchymal diseases such as acute glomerulonephritis show symmetrically enlarged kidneys with poor MAG3

clearance. Acute tubular necrosis and urinary tract obstruction show good tracer uptake in the flow phase, as the arterial flow is well established, but no tracer excretion. Correlation with other imaging modalities such as ultrasound/CT may then exclude the presence of a dilated collecting system and aid diagnosis. Renograms in patients with chronic renal failure usually demonstrate small kidneys with little or no tracer uptake, prolonged excretion, and poor drainage.

Renography in transplanted kidneys The main indication for an isotope renogram in post-transplant evaluation is the differentiation of acute tubular necrosis from rejection. Rejected kidneys have poor perfusion as well as poor tubular excretion, resulting in poor visualization of the transplant kidney on the renogram. In acute tubular necrosis, there is normal perfusion/visualization of the transplant kidney but the excretory phase shows no, or minimal, excretion. The urinary bladder is frequently not visualized despite the kidney being visualized. Unfortunately ciclosporin toxicity can also appear similar on a functional aspect and this common differential diagnosis needs exclusion when a transplant renogram is suggestive of acute tubular necrosis. Serial isotope renograms post transplant are helpful in identifying subacute and chronic rejection. See Fig. 16.5.

SPECT-CT and PET-CT scans SPECT (single-photon emission computed tomography) uses radioisotopes that emit gamma rays which are then measured directly using a gamma camera. PET involves injection of a known amount of positron-emitting radiopharmaceuticals and imaging of the whole body or specific organs using a PET scanner. Due to the need for anatomical localization of sites of abnormal tracer activity on PET scan, there was recognition of the need to develop fusion imaging whereby anatomical context can also be provided. Initial efforts were focused on registering PET and CT data obtained separately and using advanced computer software to merge or fuse the two scans. However, this approach was superseded by combining the PET and CT technologies into one

Chapter 16 

(A)

radioisotopes in diagnostic imaging in nephrology (B) 10.5

RENOGRAM

Transplant Bladder

% U p t a k e

8 Time (mins)

BLADDER 3 MIN % UPTAKE

TRANSPLANT

5.9

2.0

16 MIN % UPTAKE

10.5

1.9

MAX % UPTAKE

10.5

2.1

75.1

24.9

3 MIN RELATIVE % UPTAKE

16

Fig. 16.5  (A) Post-transplant renogram showing subacute and chronic rejection. (B) Graph showing post-transplant subacute and chronic rejection.

scanner. The patient therefore undergoes one examination at one sitting and there is minimal patient motion and other technical problems. The CT scan is typically done without intravenous contrast medium enhancement. More recently combined PET-MRI imaging has been introduced with the added advantage of reducing radiation exposure. PET-CT scanners now form the main bulk of functional scanners, with manufacturers slowly phasing out PET only scanners and promoting advances in PET-CT and PET-MRI scanning. One limitation of current PET/CT technology is that data are acquired sequentially rather than simultaneously. This then limits the ability to study functional processes that change between the two scans. Another factor for consideration is the radiation dose to patients. Artefacts caused by patient motion or organ motion between the two sets of images coupled with different breathing protocols for PET and CT can result in problems for data interpretation. Mis-registration between the two sets of images is a recognized disadvantage which is factored in while reporting PET-CT scans. Data acquisition with new hybrid PET-MRI scanners provides the option of real-time simultaneous studies. MRI offers better soft tissue contrast and the challenge of developing viable PET-MRI systems focused on addressing apparent technical incompatibilities such as the fact that PET detectors are based on scintillation crystal blocks which are highly sensitive to even small magnetic fields; and space constraints of fitting in two technologies in one ergonomic system suitable for patient use and convenience. Therefore the

type of sequential imaging performed with PET-CT is not a viable option for PET-MRI and the way forward is to combine both systems into one gantry system.

Renal infection and inflammation Tc-99m DMSA scintigraphy is a highly sensitive and specific tool to detect and confirm renal involvement associated with lower urinary tract infections, especially in the paediatric population. Renal sequelae such as scarring secondary to vesicoureteric reflux can also be assessed in the post -acute or chronic phases. Tc-99m DMSA scintigraphy is recommended for the evaluation of the extent of scarred renal tissue in both children and adults with chronic pyelonephritis. SPECT scanning has been shown to increase the diagnostic confidence for detecting cortical defects associated with pyelonephritis. Renal abscess and focal pyelonephritis may appear as a focal defect with increased uptake in the vascular phase. CT and ultrasound are usually used for correlation. Pyonephrosis on the renal scintigraphy will show no tracer uptake as there is very little viable renal tissue. Gallium (Ga)-67 citrate and indium (In)-111-labelled leucocytes are used in imaging for suspected renal infections. Ga-67 shows non-specific uptake in the kidneys, but has been shown to be particularly useful in acute focal bacterial nephritis which has non-specific ultrasound and CT findings. This commonly occurs in diabetic patients and shows unilateral or multifocal areas of

123

124

Section 1  

assessment of the patient with renal disease

Fig. 16.6  Whole-body FDG PET-CT scan.

increased uptake in the kidneys. Unlike Ga-67, In-111-labelled leucocyte scans do not show activity within the urinary tract and are more specific for acute renal infections. These scans can differentiate perinephric abscess from pyelonephritis. Whole-body fluorodeoxyglucose (FDG) PET-CT scans can be used as a screening tool to detect active infection or inflammation involving either the kidney alone or as part of a systemic process in conditions such as vasculitis, glomerulonephritis, polycystic kidney disease, pyrexia of unknown origin (PUO), and in immunocompromised patients. In PUO, PET-CT provides useful diagnostic information in about 40% of patients. Active FDG uptake is seen in the blood vessels affected by vasculitis for which PET has positive predictive value of 93% and negative predictive value of 80%. See Fig. 16.6.

Post-transplant lymphoproliferative disorder FDG PET-CT scan at diagnosis of post-transplant lymphoproliferative disorder will show uptake in all affected lesions and may result in the detection of lesions not seen at conventional CT. This may upstage the disease. Repeating the scan after treatment will also help with prognosis as FDG uptake will have gone from all successfully treated lesions. Persistence of uptake indicates a partial response to treatment. Absence of any uptake in all lesions is a strong predictor of remission. FDG PET-CT has the advantage over conventional contrast medium-enhanced CT in avoiding contrast-related nephrotoxicity in patients with impaired renal function.

Renal tumours Most renal tumours are now detected incidentally when patients undergo cross-sectional imaging for other indications. Nuclear

Fig. 16.7  Example of very high levels of glucose metabolism in FDG PET-CT scan.

medicine and PET tracers are not specific for routine use in the detection and diagnosis of renal cell cancers. FDG hyperactivity is not specific for malignancy. Extensively necrotic primary tumours and lymph nodes show low FDG uptake and may be missed on PET-CT scans. However, when the abnormalities are marked, such as very high levels of glucose metabolism, then there is a higher likelihood of malignancy. See Figs 16.7 and 16.8. FDG PET-CT scans are extremely useful to identify extrarenal sites of renal cancer, tumour recurrence post nephrectomy, and lymph node involvement. Renal metastases are predominantly lytic and may not be detected as hot spots on the Tc-99m methylene diphosphonate isotope bone scan. Scanning with 18-Fluoride is particularly effective in the detection of distant bone metastases thereby influencing therapeutic decisions for staging, assessing recurrence, or new metastasis after initial therapy. See Fig. 16.9.

Renal trauma CT is the gold standard in trauma imaging. Although renal scintigraphy is not routinely performed in the acute setting, it is useful in patients with mild renal injury to monitor and follow-up renal function, presence of urine leak, or acute renovascular compromise. Haematomas are seen as photopenic areas on the renogram. Severe contusions appear as a small functioning kidney while mild contusions may be demonstrated as an enlarged kidney. Persistent focal photopenic defects on follow-up scans are usually areas of infarction. MAG3 renograms have been shown to identify urine leak post renal transplant surgery and post intervention procedures such as percutaneous stenting of renal arteries. However, Tc-99m DTPA

Chapter 16 

radioisotopes in diagnostic imaging in nephrology

DMSA planar versus SPECT Tc-99m DMSA is the isotope of choice to demonstrate the cortical integrity of the kidneys. SPECT is used in correlation with Tc-99m DMSA and other imaging modalities in specific conditions which need to differentiate pathological from non-pathological renal tissue. When Tc-99m DMSA renography shows a cortical abnormality such as a pseudotumour, that is, column of Bertin, dromedary hump, or fetal lobulation, SPECT is useful to demonstrate normal renal tissue uptake at that site, helping to confirm the diagnosis. Photopenic defects on the Tc-99m DMSA may be due to reflux or scarring, cysts, abscess, or neoplasms. These are shown as a true cortical defect on SPECT which then can be evaluated further with other imaging techniques as CT or ultrasound.

Conclusion Nuclear medicine techniques provide a range of diagnostic options in assessing the various physiological and pathological processes of the kidney.

Further reading

Fig. 16.8  Very high levels of glucose metabolism in FDG PET-CT scan in a patient with malignancy.

identifies urine extravasation from the kidneys better than Tc-99m MAG3 which may miss a small subtle leak. There have been case reports in which SPECT scan was useful to identify the exact site of urine leak post renal transplant due to the added advantage of anatomical correlation.

Fig. 16.9  Scanning with fluorine-18 shows bone metastasis.

Achong, D. M. and Tenozrio, L. E. (2003). Abnormal MAG3 renal scintigraphy resulting from dehydration. Clin Nucl Med, 28, 683–4. Aktas, A., Aras, A., Colak, T., et al. (2006). Comparison of Tc-99m DTPA and Tc-99m MAG3 perfusion time-activity curves in patients with renal allograft dysfunction. Transplant Proc, 38, 449–53. Alavi, A., Mavi, A., Basu, S., et al. (2007). Is PET-CT the only option? Eur J Nucl Med Mol Imaging, 34, 819–21. Bar-Shalom, R., Yefremov, N., Guralnik, L., et al. (2006). SPECT/CT using 67Ga and 111In-labelled leukocyte scintigraphy for diagnosis of infection. J Nucl Med, 47, 587–94. Bianchi, E., Pascual, M., Nicod, M., et al. (2008). Clinical usefulness of FDG-PET/CT scan imaging in the management of posttransplant lymphoproliferative disease. Transplantation, 85(5), 707–12. Bleeker-Rovers, C. P., Van Der Meer, J. W., and Oyen, W. J. (2009). Fever of unknown origin. Semin Nucl Med, 39, 81–7. Broekuezen-de Gast, H. S., Tiel-Van Buul, M. M. C., and Van Beek, E. J. R. (2001). Severe hypertension in children with renovascular disease. Clin Nucl Med, 26, 6–609. Caglar, M., Moretti, J. L., Buchet, P., et al. (1998). Enalapril plus frusemide MAG3 scinitgraphy in hypertensive patients with atherosclerosis and moderate renal insufficiency. Nucl Med Commun, 19, 1135–40. Dondi, M., Fanti, S., and De Fabritis, A., et al. (1992). Prognostic value of captopril renal scintigraphy in renovascular hypertension. J Nucl Med, 33, 2040–4. Dondi, M., Monetti, N., Fanti, S., et al. (1991). Use of technetium-99m-MAG3 for renal scintigraphy after angiotensin converting enzyme inhibition. J Nucl Med, 32, 424–8. Esteves, F. P. Taylor, A., Maantunga, A., et al. (2006). 99Tcm MAG3 Renography: normal values for MAG3 clearance and curve parameters, excretory parameters, and residual urine volume. AJR, 187, W610–7. Fanti, S., Dondi, M., Corbelli, C., et al. (1993). Evaluation of hypertensive patients with a solitary kidney using captopril renal scintigraphy with Tc-99m-MAG3. Nucl Med Commun, 14, 969–75. Freeman, L. and Blaufox, M. D. (1999). Renal nuclear medicine including consensus reports. Semin Nucl Med, 29, 146–88. Goldberg, M. A., Mayo-Smith, W. W., Papanicolau, M., et al. (1997). FDG PET characterisation of renal masses: preliminary experience. Clin Radiol, 52, 510–15. Kahn, D., Ben-Haim, S., Bushnell, D. L., et al. (1994). Captopril enhanced Tc-99m MAG3 renal scintigraphy in subjects with suspected renovascular hypertension. Nucl Med Commun, 15, 515–28.

125

126

Section 1  

assessment of the patient with renal disease

Kapoor, V., McCook, B. M., and Torok, F. S. (2004). An introduction to PET-CT imaging. Radiographics, 24, 523–43. Karesh, S. M. (1999). Principles of radiation. In R. H. Wagner, S. M. Karesh, and J. R. Halama (eds.) Questions and Answers in Nuclear Medicine, pp. 1–48. St. Louis, MO: Mosby. Kiratli, P. O., Caner, B., Altun, B., et al. (2001). Superiority of Tc-99m MAG3 to Tc-99m DTPA in treating a patient with mild renal artery stenosis. Ann Nucl Med, 15, 45–8. Love, C. and Palestro, C. J. (2004). Radionucleide imaging of infection. J Nucl Med Tech, 32, 47–57. Love, C., Tomas, M. B., Tronco, G. G., et al. (2005). FDG PET of infection and inflammation. Radiographics, 25, 1357–68. Majima, T., Hattori, Y., Funahashi, Y., et al. (2012). 99Tcm MAG3 renography to monitor renal transplant function among kidneys from donors after cardiac death. Transplant Proc, 44, 49–53. Mankoff, D. A., Muzi, M., and Zaidi, H. (2006). Quantitative analysis in nuclear oncologic imaging. In H. Zaidi (ed.) Quantitative Analysis of Nuclear Medicine Images, pp. 494–536. New York: Springer. Mettler, F. A. and Guiberteau, M. J. (eds.) (2006). Essentials of Nuclear Medicine Imaging (5th ed.). Philadelphia, PA: Saunders and Elsevier. Murray, I. P. C. and Ell, P. J. (eds.) (1994). Nuclear Medicine in Clinical Diagnosis and Treatment. London: Churchill Livingstone. Nally, J. V. and Barton, D. P. (2001). Contemporary approach to diagnosis and evaluation of renovascular hypertension. Urol Clin North Am, 28, 781–91. Palmer, M. R., Donohue, K. J., Francis, J. M., et al. (2011). Evaluation of relative renal function for patients who had undergone simultaneous liver-kidney transplants using Tc-99m-MAG3 scintigraphy with attenuation correction from anatomical images and SPECT/CT. Nucl Med Commun, 32, 738–44. Piepsz, A., Colarina, P., Gordon, I., et al. (2000). Guidelines for Glomerular Filtration Determination in Children. [Online]

Prigent, A., Cosgriff, P., Gates, G. F., et al. (1999). Consensus report on quality control of quantitative measurements of renal function obtained from the renogram: International Consensus Committee from the Scientific Committee of Radionuclides in Nephrourology. Semin Nucl Med, 29, 146–59. Rajadhyaksha, C. D., Parker, J. A., Barbaras, L., et al. (2005). Normal and Benign Pathologic Findings in 18-FGD-PET and PET-CT. An Interactive Web Based Image Atlas. Boston, MA: Joint programme in Nuclear Medicine, Harvard Medical School. Ramdave, S., Thomas, G. W., Berlangieri, S. U., et al. (2001). Clinical role of F-18 fluorodeoxy glucose positron emission tomography for the detection and management of renal cell carcinoma. J Urol, 166, 825–30. Rossleigh, M. A. (2001). Renal cortical scintigraphy and diuresis renography in infants and children. J Nucl Med, 42, 91–5. Safaei, A., Figlin, R., Hoh, C. K., et al. (2002). The usefulness of F-18 deoxyglucose whole body positron emission tomography (PET) for re-staging of renal cell cancer. Clin Nephrol, 57, 56–62. Sandler, M. P., Coleman, R. E., Patton, J. A. (2002). Diagnostic Nuclear Medicine (4th ed.). Philadelphia, PA: Lippincott, Williams and Wilkins. Schreij, G., van Es, P. N., van Kroonenburgh, M. J. P. G., et al. (1996). Baseline and postcaptopril renal blood flow measurements suspected of renal artery stenosis. J Nucl Med, 37, 1652–5. Seto, E., Segall, G. M., and Terris, M. K. (2000). Positron emission tomography detection of osseous metastasis of renal cell carcinoma not identified on bone scan. Urology, 55, 286. Sfakianakis, G. N., Al-Skeikh, W., Heal, A., et al. (1982). Comparison of scintigraphy with In-111 leucocytes and Ga-67 in the diagnosis of occult sepsis. J Nucl Med, 23, 618. Shao, Y., Cherry, S. R., Farahani, K., et al. (1997). Simultaneous PET and MR imaging. Phys Med Biol, 42, 1965–70. Wagner, R. H., Karesh, S. M., and Halama, J. R. (eds.) (1999). Questions and Answers in Nuclear Medicine. St Louis, MO: Mosby.

CHAPTER 17

Immunological investigation of the patient with renal disease Jo H. M. Berden and Jack F. M. Wetzels Introduction The kidney is an important site of immune injury mediated by both the innate and adaptive immune systems. Our understanding of the pathogenesis of immunological renal diseases has been improved by study of the time course of changes in kidney morphology in animal models. A good example is the renal injury that occurs after injecting bovine serum albumin (BSA)-anti-BSA complexes into rabbits, a model of serum sickness, and a prototypical example of an immune complex disease. It is now well established that many primary glomerular diseases are immune mediated. Furthermore, the kidney is the target in various infectious diseases, systemic auto-immune disorders and the vasculitides (Couser, 2012; Nasr et al., 2013). Renal injury may be evoked by humoral (antibody) or cellular immune responses. Three mechanisms can be defined. The kidney may be the target of antibodies that are directed against kidney-specific antigens, exciting a type II immunological reaction. Anti-glomerular basement membrane (GBM) disease, caused by antibodies directed against collagen IV, an intrinsic component of the GBM and membranous nephropathy, caused by antibodies directed at podocyte antigens are the best known examples. More often, the kidney is involved in a type III immunological reaction caused by the deposition of immune complexes. These immune complexes may either be formed in the circulation or in situ, by binding of an antibody to an antigen already planted in the glomerular capillary wall. The size, charge, and composition of the immune complexes determine their final localization. The site at which these deposits are formed will determine the pathological changes. Mesangial deposits induce mesangial matrix expansion and haematuria; subendothelial deposits a proliferative glomerulonephritis with capillary loop necrosis and crescent formation; while subepithelial deposits mostly evoke the nephrotic syndrome. The role of cellular immunity in inducing glomerular pathology is less well understood. In glomerular diseases such as minimal change disease, focal segmental glomerulosclerosis, and antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis injury is probably also mediated by cellular responses. Although in vasculitis various auto-antibodies may be present in the circulation there is no evidence of antibody or complement deposition in the kidneys. Laboratory assays which evaluate the immune response, in particular the identification of (auto)-antibodies are valuable tools in establishing a diagnosis and/or monitoring of the activity of the

disease. An alphabetical overview of various immunological markers and associated renal diseases is given in Table 17.1.

Laboratory techniques These assays depend on a number of laboratory techniques. The practising clinician should be aware that test results are dependent on the technique used and be aware of pitfalls.

Electrophoresis Electrophoresis is a simple technique, where a mixture of proteins is applied to an agar gel and separated by applying a charge over the gel. Protein electrophoresis of serum shows a typical pattern with separation of albumin, α1, α2, ß, and γ globulins. More specific assays are based on the principle that antigen and antibody form precipitates. Examples are immunoelectrophoresis or immunofixation, where proteins are separated by charge and then precipitated by antibodies that diffuse into the gel. If specific antibody-antigen reactions occur these are visible as precipitation lines in the gel. In immunonephelometry the antigen and antibody form insoluble complexes in solution that are detected by light scatter. The specificity of these techniques is dependent on the specificity of the antibodies, which can be either monoclonal or polyclonal. When there is antigen excess the assays may provide negative results unless samples are appropriately diluted (Daval et al., 2007).

Indirect immunofluorescence With this technique, antibodies directed at tissue antigens can be visualized. Tissue slides (varying from kidney to specific cell lines) are exposed to various anti-sera. Antibodies directed to antigens that are expressed in the tissue/cells will bind. These antibodies are visualized by adding a fluorescent second antibody directed at human immunoglobulins. The fluorescent pattern and the tissue type provide clues to the nature of the involved antigens, for example, ANCA will cause on granulocytes either cytoplasmic or perinuclear staining and antinuclear antibodies (ANAs) will react with the nuclei of cell lines. Indirect immunofluorescence (IF) on kidney sections together with serological tests are essential tools to diagnose the cause of glomerulonephritis. The indirect IF technique is subject to several pitfalls: the specificity of the antigens is unclear, and most laboratories use only anti-immunoglobulin (Ig)-G antibodies as secondary antibodies, which will not detect IgA or IgM class antibodies.

Table 17.1  Overview of immunological markers in renal diseases Condition ♯

Associated systemic or renal disease

ANA

+

SLE; different auto-immune diseases

ANCA

+

Granulomatosis with polyangiitis; microscopic polyangiitis; renal limited vasculitis; eosinophilic granulomatosis with polyangiitis

Beta-2-glycoprotein 1

+

APS

Cardiolipin

+

APS

Centromere

+

Cutaneous sclerosis

C1q

+

SLE nephritis

DNAse B

+

PSGN

dsDNA

+

SLE

ENA

+

MCTD; Sjögren’s syndrome; scleroderma; SLE

α–fodrin

+

Sjögren’s syndrome

GBM

+

anti-GBM disease; Goodpasture syndrome

Histone

+

drug-induced SLE; SLE

MPO

+

See ANCA

Nucleosome

+

SLE nephritis

Phospholipid

+

See anti-cardiolipin/lupus anticoagulant

PLA2R

+

Membranous nephropathy

Proteinase 3

+

See ANCA

RNP

+

MCTD

SS-A (Ro)

+

SLE; Sjögren’s syndrome

SS-B (La)

+

Sjögren’s syndrome

Streptolysin O (ASO)

+

PSGN

Topo-isomerase I

+

systemic scleroderma

C1q

D

SLE (80%; ♀ > ♂)

CH50

↓/D

See C3/C4

C2

D

SLE (25%, ♀ > ♂), IgA nephropathy, HSP, MPGN type 1

C3

↓/M

SLE (25%, MPGN type 1, DDD, PSGN, aHUS

D

SLE (seldom)

↓

SLE, cryoglobulinaemia

D

SLE (75%)

Factor H

↓/D/M

aHUS, DDD

Factor B

↓/D/M

aHUS, DDD

Factor I

↓/D/M

aHUS, DDD

C3 nephritic factor

+

DDD, partial lipodystrophy

C5-C9

D

SLE (seldom)

+

Cryoglobulinaemia, SLE, Sjögren syndrome, lymphoma, HCV

Immunological marker Screening assays

Antibody specificity

Complement

C4

Miscellaneous Cryoglobulins

(Continued)

chapter 17 

immunological investigation of renal disease

Table 17.1 Continued Condition ♯

Associated systemic or renal disease

↑

IgA nephropathy

D

SLE

Lupus anticoagulant

+

APS, SLE

Paraproteins

+

Amyloidosis, MIDD, cast-nephropathy

vWF protease activity

↓

TTP

Immunological marker IgA

a Meaning of condition: + = present; D = deficiency; M = monitoring; ↓ = decreased; ↑ = increased.

ANA = antinuclear antibodies; ANCA = anti-neutrophil cytoplasmic antibodies; ASO = antistreptolysin O; D = deficiency; DDD = dense deposit disease; ENA = extractable nuclear antigens; GBM = glomerular basement membrane; HCV = hepatitis C virus; HSP = Henoch–Schönlein purpura; HUS = haemolytic uraemic syndrome; M = mutation; MCTD = mixed connective tissue disease; MIDD = monoclonal immunoglobulin deposition disease; MPGN = membranoproliferative glomerulonephritis; MPO = myeloperoxidase; PLA2R = M type phospholipase A2 receptor; PR3 = proteinase 3; PSGN = poststreptococcal glomerulonephritis; TTP = thrombotic thrombocytopenic purpura;

Enzyme-linked immunosorbent assay In the direct enzyme-linked immunosorbent assay (ELISA), antigens are coated on a plate. Antibodies will bind, and can be visualized with a ligand, mostly a secondary antibody labelled with an enzyme. The enzyme converts a colourless substrate into a chromogen that can be measured. In the ELISA, the sensitivity and specificity are determined by the nature and conformational shape of the coated antigens, the sort of ligand, and the epitope specificity of the antibody. The specificity and sensitivity of an ELISA can be increased by an indirect technique, the so-called antigen capture ELISA. In this technique, plates are coated with specific antibodies which bind the antigen under investigation. This technique allows the antigen to remain in its native conformation. Subsequently, antibodies reactive with the captured antigen can then be detected.

Immunoblotting This technique is used to further characterize the antigen specificity of an antibody. The protein mixture is separated depending on size or charge by gel chromatography and the molecules are then transferred to a nitrocellulose membrane. These blots are incubated with serum samples. After binding, antibodies are visualized by enzyme-labelled secondary antibodies. The blotted proteins are identified in a marker lane by labelled antibodies of known specificity, which allows by comparison determination of the antibody specificities in the sample under investigation.

Immunoglobulins Immunoglobulin A (See Chapters 65–69.) Immunoglobulin A  (IgA) comprises 15–20% of all serum immunoglobulins. In serum, most IgA is present in its monomeric form (molecular weight 160 kD). Two classes of IgA are recognized: IgA1 (80%) and IgA2 (20%). Plasma concentrations of IgA may be increased in up to 50% of patients with IgA nephropathy (D’Amico, 1988). Thus, normal values do not exclude the presence of IgA nephropathy. Moreover, elevated levels of IgA can be found in patients with hepatitis, liver cirrhosis, and Sjögren’s disease. A recent study compared serum IgA in Caucasian patients with IgA nephropathy and healthy controls (Moldoveanu et al., 2007). Although mean serum IgA level was increased in the patients, the sensitivity was only 33%. Although specificity was high at 90%,

the comparator group consisted of healthy controls and lacked patients with relevant comorbidity. Recent studies have pointed to the role of aberrantly glycosylated, galactose-deficient IgA1 (GD-IgA1) and the formation of glycan-specific anti-GD-IgA1 antibodies in the pathogenesis of IgA nephropathy. The importance of altered glycosylation is supported by the finding that this altered form of IgA is preferentially present in the mesangial IgA deposits of patients with IgA nephropathy (Allen et  al., 2001). GD-IgA1 is less effectively cleared by the sialoglycoprotein receptor, demonstrates increased binding and activation of mesangial cells, and activates complement through the mannose binding lectin pathway. Moldeveanu and colleagues used a lectin from the garden snail Helix aspersa that specifically binds GD-IgA1 to develop an ELISA assay for the measurement of serum GD-IgA1 (Moldoveanu et al., 2007). They reported increased levels (> 90th percentile of values in 150 healthy controls) of GD-IgA1 in 117 of 153 patients with IgA nephropathy and calculated a sensitivity of 76% and a specificity of 90% for predicting a diagnosis. In this study, serum GD-IgA1 did not correlate with clinical activity. In contrast, Zhao et al. evaluated 275 patients with IgA nephropathy and reported a higher risk of progressive renal failure in patients with higher levels of GD-IgA (hazard ratio 4.76 for the fourth quartile vs first quartile) (Zhao et  al., 2012). In IgA nephropathy the development of disease is dependent on the presence of anti-GD-IgA1 antibodies (mostly of the IgG class). Indeed, in a study of 60 patients with IgA nephropathy, serum levels of IgG anti-GD-IgA1 antibodies were increased and predicted the diagnosis with sensitivity of 88% and specificity of 95% (Suzuki et al., 2009). Moreover, the antibody titre correlated with proteinuria, which strongly supports their pathogenetic role. The association between auto-antibody levels and progression was confirmed recently (Berthoux et al., 2012). IgA deficiency is the most common form of Ig deficiency, occurring in 1 in 800 people. IgA deficiency is associated with an increased incidence of systemic auto-immune diseases including SLE.

Paraproteins Paraproteins are monoclonal immunoglobulins, synthesized and secreted by abnormal clones of B lymphocytes. Paraproteins can consist of the intact immunoglobulin or contain only one part of the immunoglobulin molecule, either the heavy chain (G, A, M, D, or E) or the light chain (κ or λ). The light chains are also known as Bence Jones proteins.

129

130

Section 1  

assessment of the patient with renal disease

Paraproteins can be recognized as a single band on serum or urine protein electrophoresis. However, a normal pattern of protein electrophoresis does not exclude the presence of a paraprotein. Protein electrophoresis has a low sensitivity for free light chains (FLC), the amount of paraprotein may be too small and the paraprotein may be masked in the ß-globulin band (Heher et al., 2010). Therefore, if a paraprotein is suspected additional immunofixation electrophoresis of serum and urine is recommended for detecting of a paraprotein and confirming its identity. Various renal diseases can be associated with a monoclonal gammopathy (Heher et  al., 2010). Examples include cast nephropathy, amyloidosis, monoclonal immunoglobulin deposition disease (MIDD), fibrillary and immunotactoid nephropathy, and membranoproliferative glomerulonephritis (MPGN) (see Chapters  60–64). Light microscopic evaluation of the kidney biopsy may already have raised the suspicion of a paraprotein related kidney disease. IF may be pathognomonic by revealing the mere presence of κ or λ light chains. However, IF studies may be misleading. In amyloidosis the paraproteins are modified, and specific epitopes may be lost, resulting in negative light chain staining. Alternatively, the paraprotein may interact with normal, polyclonal IgG (see ‘Cryoglobulins’). In patients with cast nephropathy, free light chains are always found in the urine. In fact, in the majority of patients the light chains are present in relatively large amounts (> 1 g) and account for > 70% of total urinary protein. Of note, urine dipsticks do not detect light chains, therefore in patients in whom myeloma is suspected, the urine must be screened by biochemical techniques. Immunofixation electrophoresis (IFE) lacks 100% accuracy for the diagnosis of paraprotein-related disorders. Using serum and urine IFE the sensitivity for detecting a monoclonal B cell disorder was 92% (Vermeersch et al., 2008b). Sensitivity is even lower in patients with amyloidosis or MIDD. In patients with AL amyloidosis sensitivity of serum IFE is 74–80% (Katzmann et al., 2009; Palladini et  al., 2009)  and of urine IFE 67–83% (Palladini et  al., 2009, Dispenzieri et al., 2009). In patients with MIDD, even when serum and urine immunofixation are combined a paraprotein may be undetected in a significant proportion of patients (4–22%). In a recent study in patients with glomerular patterns of injury (membranous nephropathy or MPGN) and characterized by the presence of monoclonal IgG on IF, a paraprotein was discovered in serum and/or urine by IFE in only 30% of patients (Guiard et al., 2011). The development of a nephelometric assay for FLC’s has allowed the detection and quantitation of small amounts of free κ and λ chains (Katzmann et al., 2002). However, absolute values of FLCs may be misleading. In healthy persons serum κ and λ FLCs increase with age. The κ/λ ratio was independent of age and a diagnostic reference interval of 0.26–1.65 was established (Katzmann et al., 2002). Similarly, absolute levels of FLCs are increased in patients with polyclonal hypergammaglobulinaemia, however the κ/λ ratio is always normal. Use of this assay improved diagnostic accuracy for paraprotein related disorders when compared to IFE:  in 67 patients with multiple myeloma, amyloidosis or MIDD, serum IFE electrophoresis was positive in only 72% of patients, whereas an abnormal κ/λ FLC ratio was found in 96%. Of note, differences were less impressive when urine IFE was used as a comparator, 89% of patients being positive (Katzmann et al., 2002). Similar results were reported by Bochtler et al. (2008), who studied 133 subsequent patients with AL amyloidosis. An abnormal FLC ratio was observed in 116 (87%) of patients and the FLC assay was

significantly better than serum IFE (abnormal in 69%), but not superior to urine IFE (abnormal in 86%) or the combination (abnormal in 92%). Although the FLC ratio may not improve diagnostic accuracy, serum free light measurements may aid in predicting prognosis in AL amyloidosis: median overall survival was 16 months in 644 patients with an abnormal κ/λ ratio, and 64 months in the remaining 86 patients with normal κ/λ ratio (Kumar et al., 2010). Quantitation of FLCs at baseline also improved risk prediction in patients with MGUS, and smouldering multiple myeloma (Rajkumar et al., 2005; Dispenzieri et al., 2009), and changes in FLCs during treatment predicted prognosis in AL amyloidosis (Lachmann et al., 2003). Since free light chains are cleared by glomerular filtration, serum concentrations of free light chains are dependent on glomerular filtration rate (GFR). Indeed, Hutchison et al reported that serum free light chains levels are inversely related to estimated GFR (eGFR) (Hutchison et  al., 2008). In patients with severe chronic kidney disease (CKD) (eGFR < 15 mL/min) serum levels are increased almost tenfold. There are slight differences in sieving characteristics between κ and λ light chains, sieving coefficient being lower for λ chains, that more often exist as dimers in the circulation. As a result, the κ/λ FLC ratio increases with decreasing eGFR. For patients with CKD new reference intervals have been defined: mean ratio 1.12 (95% confidence interval 0.37–3.1).

Cryoglobulins Cryoglobulins are Igs or combinations thereof that precipitate in the cold and re-dissolve upon rewarming to 37°C (Dammacco et  al., 2001). Cryoglobulins are potent activators of the classical complement pathway, which may explain the vasculitis that occurs in medium and small vessels. Renal involvement is seen in up to 25% of patients, and typically manifests as a MPGN (Alpers and Smith, 2008). Cryoglobulins are classified according to their component Igs. Type I cryoglobulins are composed of a single monoclonal Ig, type II cryoglobulins (also called mixed cryoglobulins) comprise a complex of a monoclonal Ig (typically IgM with antiglobulin, i.e. rheumatoid factor activity) and polyclonal IgG, and type III contain polyclonal IgG and/or IgM antiglobulin and polyclonal IgG (see Chapter 151). Renal involvement and the syndrome of ‘essential mixed cryoglobulinaemia’ is seen predominantly in subjects with type II cryoglobulins (Alpers and Smith, 2008). Type II cryoglobulins can be associated with a variety of underlying diseases such as auto-immune diseases, lymphoproliferative disorders, and hepatitis C virus (HCV) infection. In Mediterranean countries, up to 90% of patients with type II cryoglobulinaemia are HCV positive. However, reports differ from countries with a lower prevalence of HCV infections, with a prevalence of HCV ranging from 0% to 22% (Tervaert et al., 2007), (Weiner et al., 1998). In patients with non-HCV cryoglobulinaemia often no underlying cause is found and these 25–40% of patients are referred to as having true essential mixed cryoglobulinaemia. For the detection of cryoglobulins in the serum meticulous precautions at the time of blood collection must be taken, since improper sample handling may prevent detection of cryoglobulins (Allen et al., 2001; Vermeersch et al., 2008a; Motyckova and Murali, 2011). Blood must be drawn into a warm tube, and kept at 37°C during clotting. The serum is then placed in a refrigerator for at least 3 days (type I cryoglobulins usually precipitate within 24 hours, type II and type III cryoglobulins may need up to 1 week).

chapter 17 

The cryoprecipitate can be seen as a white precipitate in the bottom of the tube. After washing to remove passively trapped proteins, the precipitate is dissolved at 37°C and analysed by immunofixation techniques to establish its composition. There is no standard way of reporting and quantitation of cryoglobulins. Cryoglobulins are reported qualitatively (positive or negative) or quantitatively, using cryocrit, total protein content, or the immunoglobulin concentration of the cryoprecipitate. The cryocrit is the percentage of the centrifuged serum that is cryoglobulin. Since there is a lack of correlation between cryoglobulin concentration and symptoms, sensitive methods for the detection of cryoglobulins are advised. Assays now can detect very low concentrations of immunoglobulins in cryoprecipitate of healthy and disease controls. The 97.5th percentile of IgA, IgG, and IgM in diseased controls was proposed as the upper reference value and amounted 2 mg/L, 11 mg/L, and 34 mg/L respectively (Vermeersch et al., 2008a). The serum of patients with cryoglobulinaemia usually contains rheumatoid factor activity, resulting from the presence of an IgM with anti-IgG specificity. Serum complement C4 levels are usually very low, whereas C3 levels may be normal or low. If in a patient with cold insensitivity, purpura, livedo reticularis and/or recurrent skin ulcera no serum cryoglobulins are detected, a diagnosis of cryofibrinogenaemia must be considered (Blain et al., 2000). Cryofibrinogen consists of fibrinogen-fibrin complexes, and precipitates only from plasma (Vermeersch et al., 2008a).

immunological investigation of renal disease

on C3i. Measurements of complement breakdown products such as C3d may be more specific indicators of complement activation. However, local complement activation has a limited effect on systemic complement levels. Conversely, decreased levels of complement do not always reflect immunological injury. Temporary reductions of complement C3 and/or C4 can be found in various

Antibody on microbe

C1r

Classical pathway

MASP

C1s

Lectin pathway

C4 C4a C4b

Alternative pathway

C2 C4b2b P

C3 C2a

H

Microbial cell walls

I

Activation

C3bBb

C3bH B Ba

C3a C5a

D C5

Attaches to C3b receptors on phagocytes

Opsonization

Deactivation

C3b

Complement The complement system is a group of tightly controlled plasma proteins, which after activation evoke potent biological effects. The activation pathways of the complement system, its most important components and regulatory proteins, and the biological effects are shown in Fig. 17.1. However, the crucial regulation of complement activation occurs at the membrane level, either by the constitutive membrane bound membrane co-factor protein (MCP; CD46) or by factor H, which binds to membrane glycoproteins. The complement system can be activated in three different ways, the classical pathway, the alternative pathway, and the MBL (mannose binding lectin) pathway. All three pathways lead to the generation of an enzymatically active C3-convertase, which cleaves the central component C3 into C3a and C3b. C3b converts the C3-convertase to a C5 convertase, which cleaves C5 to C5a and C5b, the initiator of the formation of the C5b–C9 complex, the so-called membrane attack complex. The alternative pathway of complement is constantly active, partly mediated by the activating factor B. To prevent overactivation of the complement system, regulatory (inhibitory) proteins are present such as factors H, I, and MCP. Levels of individual components of the complement system can be measured by immunochemical methods. C3, C4, and factor B are the most frequently measured, and allow differentiation of alternative (low C3, normal C4) and classical (low C3 and low C4) pathway activation. Complement activation is not always reflected by decreased levels of C3 or C4. Sometimes the increased production of complement components, as a result of an acute phase response, counterbalances the increased breakdown. Alternatively, activation of C3 may go unrecognized since antibodies are used that recognize an epitope that is present on intact C3 as well as

MBL

C1q

C5b C9 C9 C9 C9 C9 C9 C9

C6

PMN chemotaxis Mast cell degranulation

C7 C8

Vascular permeability

Membrane attack complex

Phagocytosis

Lysis

Inflammation

Fig. 17.1  Overview of the complement system. C4 is activated either by the classical pathway or by the lectin pathway. The classical pathway is activated by binding of C1q to antigen–antibody complexes. This activates C1r and C1s which cleave C4. The lectin pathway is activated if the mannan binding lectin (MBL) binds to mannose groups on bacteria. This activates the serine proteases MASP 1 and 2 (MBL associated serine protease). MBL and MASP are homologous to C1q, C1r, and C1s. Next, C2 can bind to C4b and then becomes susceptible for cleavage by C1s. After cleavage the classical C3 convertase (C4b2b) is formed, which cleaves C3 into C3b and C3a. The formation of C3b can also result from alternative pathway activation. Spontaneous C3 activation occurs spontaneously at a low level. If this activated C3b binds to factor B, the complex becomes susceptible for cleavage by factor D. This results in the formation of the alternative pathway C3 convertase C3bBb. However, factor H has a higher affinity for C3b than factor B. If factor H binds to C3b the complex is cleaved by factor I resulting in inactivated C3b. Certain microbial surfaces favour the association between C3b and factor B and therefore activate the alternative pathway. Cell surfaces promote the binding of factor H to C3b thereby inhibiting alternative pathway activation. Generation of C3b is a central feature of complement activation, because of the biological functions associated with C3b formation namely opsonization/phagocytosis and pro-inflammatory responses via C3a and C5a. In addition it forms the initial event in the formation of the so-called membrane attack complex C5b–C9 which drills holes in bacteria and cells. Adapted from Playfair JHL, Lydyard PM. Medical Immunology for Students. Churchill and Livingstone 1995; page 8.

131

132

Section 1  

assessment of the patient with renal disease

conditions such as atheroembolic disease, sepsis, pancreatitis, severe burns, following the use of contrast agents or inulin, cuprophan haemodialysis, cardiopulmonary bypass, haemolytic crisis in malaria, and porphyria (Hebert et al., 1991). The CH50 activity assay is a functional assay that measures the overall activity of the complement system after activation of the classical pathway. Absent or very low CH50 activity indicates a congenital deficiency of one of the complement components of the classical pathway such as C1q, C2, or C4. We now know that dysregulation of the complement system is often not detected by the standard immunochemical quantitative measurement of the involved proteins. Most abnormalities are caused by genetic mutations that alter the function of the protein, without affecting its production or expression. Function of complement regulatory proteins may also be affected by the presence of specific antibodies, the best example being antibodies against factor H. The complement system is frequently involved in renal diseases. Although for many decades the role of the classical pathway has been emphasized, abnormalities in alternative pathway activation have recently attracted much attention as cause of renal diseases such as the haemolytic uraemic syndrome (HUS) and dense deposit disease (DDD) (Taylor et al., 2010; Pickering and Cook, 2011). In Table 17.2 we provide an overview of complement abnormalities in various renal diseases. Congenital deficiencies of complement components deserve special consideration. There are known associations between complement deficiencies and the development of various renal diseases (Table 17.1). Deficiencies of C1q, C2, and C4 are frequently associated with systemic lupus erythematosus (SLE). C3 nephritic factor is an auto-antibody that is associated with DDD. This auto-antibody prevents the deactivation of C3b by factor H, either by binding to C3b or to factor H. Some patients with DDD and C3 nephritic factor have a typical gaunt appearance with partial lipodystrophy, the absence of fat tissue in the face and upper part of the body. The pathogenesis has been recently clarified: the adipocytes in the upper part of the body produce a protein called adipsin which is identical to factor D. In the presence of C3 nephritic factor complement will be activated and mediate lysis of the adipocytes at these sites.

Antinuclear antibodies ANAs are directed against nuclear constituents, either nucleic acids or nucleoproteins. ANA can be detected in an indirect IF test by applying serum to fixed cell lines. Most commonly used are Hep-2 cells. A positive ANA on indirect IF can be found in up to 10% of aged healthy persons. Various defined specificities of autoantibodies can cause a positive ANA (Table 17.1). The nuclear staining pattern can already provide an indication for the specificity of the ANA (see Fig. 17.2). A homogenous or peripheral staining pattern is mainly caused by antibodies against the nucleosome or its components, double-stranded DNA (dsDNA) or histones. A speckled pattern is caused by antibodies against nucleoproteins like Sm, RNP, SS-A (Ro), or SS-B (La). A centromere or nucleolar staining is suggestive for various scleroderma-associated autoantibodies. The antibody specificities can be more precisely determined by additional testing. Specificity for dsDNA can be determined qualitatively by an indirect IF assay on Crithidia luciliae, a flagellate which contains a kinetoplast with pure dsDNA. For quantitative testing the Farr assay is used, a modified RIA that particularly detects high avidity antibodies. Anti-DNA ELISA systems are available but their specificity is lower because (non-pathogenic) low avidity antibodies are also detected. In recent years it has become apparent that antinucleosome antibodies have a higher sensitivity (especially in lupus nephritis) than anti-dsDNA and an equal specificity (van der Vlag and Berden, 2011). In lupus patients negative for anti-dsDNA 54% were positive for antinucleosome antibodies. In one study it was found that all anti-dsDNA negative patients with lupus nephritis were antinucleosome positive. There are now ELISAs available that test both anti-DNA and antinucleosome antibodies (Biesen et al., 2011). The specificity for other defined nucleoproteins, as suggested by a speckled ANA pattern, can be documented by a number of techniques including an immunodiffusion technique using purified extracted nuclear antigens (ENAs), an immunoblotting technique using either an extract of nuclear and/or cytoplasmic antigens or recombinant autoantigens, or an ELISA using these autoantigens. Well-known specificities include anti-histone, anti-Sm, anti-RNP, anti-SS-A, anti-SS-B, and anti-topo-isomerase I.

Table 17.2  Complement levels in various renal diseases C4

C3

Factor B

PSGN

N(–↓)

↓↓

N?

Only in acute phase

Lupus nephritis

↓↓

↓↓

N

Proportionally decreased

MPGN type I MPGN type II (DDD)

N-↓ N

↓ ↓↓

N ↓

Due to presence C3NeF

GN associated with chronic infections

↓

↓

N

Cryoglobulinaemic GN

↓–↓↓

N or ↓

N

ANCA associated GN

N

N

N

Anti-GBM disease

N

N

N

Haemolytic uraemic syndrome

N

N or ↓

N

↓ = decrease; ↑ = increase; C3NeF = C3 nephritic factor; DDD = dense deposit disease; GN=glomerulonephritis, MPGN = membranoproliferatieve glomerulonephritis; N = normal; PSGN = poststreptococcal glomerulonephritis.

chapter 17 

(A)

(B)

(C)

(D)

immunological investigation of renal disease

Fig. 17.2  The four common patterns of antinuclear staining. Homogenous (A), peripheral (B), speckled (C) and nucleolar (D). Reproduced with permission from Dieppe PA, Bacon PA, Bamji AN, Watt I. Slide Atlas of Rheumatology, vol. 9 SLE. Gower, London, 1984; page 9.6.

Anti-dsDNA (anti-nucleosome) and anti-Sm antibodies are almost pathognomonic for SLE. Both specificities are included in the American College of Rheumatology criteria for SLE. Various other ANAs are also frequently found in patients with SLE and correlate with specific disease associations (see Chapter 163).

cryoglobulinaemia, and MPGN (60–80%). Therefore, their specificity for a particular disease is low. Moreover, anti-C1q antibodies are also present in healthy controls (5%), the prevalence increasing with age (> 70 years, 18% positive).

Anti-C1q antibodies

Antineutrophil cytoplasmic antibodies

The popularity of the use of assays for measurement of circulating immune complexes in the past using C1q as a substrate, revealed the presence of autoantibodies to C1q in many disease conditions (Kallenberg, 2008). These antibodies can be measured by using immobilized C1q either in ELISA or a RIA. To prevent binding of immune complexes or dsDNA to C1q these tests are performed at an increased ionic strength, mostly 0.3 M NaCl. These anti-C1q antibodies are frequently found in patients with proliferative lupus nephritis (30–80%). In many lupus patients anti-C1q antibodies are related to disease activity with higher prevalences and higher titres in active renal disease compared to inactive LN or active non-renal lupus (Kallenberg, 2008). In two studies, anti-C1q levels rose significantly prior to relapses of renal disease (Coremans et al., 1995; Moroni et al., 2001). In the latter study, a sensitivity of 87% and a specificity of 92% was found. However, this was also found in patients with hypocomplementemic urticarial vasculitis (100%) mixed connective tissue disease, polyarteritis nodosa (± 30%),

The discovery of ANCAs has been a major breakthrough when the differential diagnosis includes vasculitis and rapidly progressive glomerulonephritis (RPGN). ANCAs are directed against proteins that are predominantly present in the azurophilic granules of neutrophils. The screening assay for ANCA is indirect IF on ethanol fixed neutrophils. Under these conditions three different patterns can be seen:  cytoplasmic ANCA (c-ANCA), perinuclear ANCA (p-ANCA), or atypical ANCA (x-ANCA) (Fig. 17.3) The difference in staining pattern between c-ANCA and p-ANCA is due to a fixation artefact. Ethanol fixation disrupts the lysosomal membrane and as a result the lysosomal proteins will leak into the cytoplasm. Cationic proteins (targeted by p-ANCA) will be attracted to the negatively charged nuclear membrane giving rise to the perinuclear staining pattern, whereas anionic or neutral proteins will be more evenly distributed in the cytoplasm (c-ANCA). On formalin-fixed neutrophils there are no differences, and only a c-ANCA staining pattern is seen.

133

134

Section 1  

assessment of the patient with renal disease

(A)

(B)

(C)

Formalin fixation Strong cationic proteins (e.g. MPO) C-ANCA

Weakly cationic or neutral proteins (e.g. PR3)

Ethanol fixation

P-ANCA

C-ANCA Antibodies to strong cations

Antibodies to neutral proteins or weak cations (e.g. PR3)

Fig. 17.3  Antinuclear cytoplasmic antibodies (ANCA). (A and B) Indirect immunofluorescence techniques using ethanol-fixed neutrophils demonstrate perinuclear (p) (A) and cytoplasmic (c) (B) staining patterns. (C) The use of crosslinking fixatives (e.g. formalin) during the preparation of the neutrophil substrate leads to alterations in granule membranes, allowing positively charged constituents to migrate to the negatively charged nuclear membrane. Because this does not occur with formalin fixation, use of both types of fixation allows the distinction of a true p-ANCA from an antinuclear antibody (ANA). With formalin-fixed neutrophils, antibodies that would cause p-ANCA patterns (on ethanol-fixed cells) will display diffuse granular cytoplasmic staining, whereas nuclear staining would indicate the presence of an ANA. MPO = myeloperoxidase; PR3 = proteinase 3.

P-ANCA must be differentiated from ANAs that also give a perinuclear staining pattern. The effects of fixation can provide a clue. Whereas p-ANCA staining will change to a c-ANCA pattern on formalin fixed neutrophils, the perinuclear staining of ANAs will remain unchanged. Furthermore, ANA staining can be observed when using other cells or tissues, whereas p-ANCA antibodies are neutrophil specific. Most c-ANCA are directed against a 29kD protein called proteinase-3 (PR3). P-ANCA are most frequently directed against myeloperoxidase (MPO), although there are other antigenic targets such as elastase, lactoferrin, lysosyme, cathepsin G, and bactericidal/permeability-increasing protein (BPI). The specificities of x-ANCA are not fully elucidated yet although binding has been observed both to the neutrophil protein tubulin-B5 and the bacterial cell division protein (FtsZ). This cross-reaction is perhaps implicated in the pathogenesis of inflammatory bowel disease (Terjung and Spengler, 2009). This is in line with the observation in a French study in which 71.8% of patients with ulcerative

colitis and 11% of patients with Crohn’s disease were x-ANCA positive, while healthy blood donors showed a prevalence of 0% (Desplat-Jego et al., 2007). The antigenic specificity of ANCA should be tested by the widely available ELISA assays for PR3 and MPO. These commercial assays only detect IgG class antibodies. Although there is concordance between c-ANCA staining and anti-PR3 positivity in ELISA and between p-ANCA staining and anti-MPO positivity, the results are not always congruent and depend on the assay used (Wang et  al., 1997). Of c-ANCA positive sera, 44–84% were positive in a PR3 ELISA, whereas p-ANCA positivity was confirmed by anti-MPO ELISA in 25–75% of cases (Wang et  al., 1997). In the latter study, only 35% of p-ANCA positive sera proved MPO-ANCA positive in capture ELISA, immunoblot and inhibition assays, indicating the frequent existence of other specificities. A  European multicenter analysis revealed that the IF test is more sensitive but the ELISA is more specific (Hagen et al., 1998). However, one should realize that sensitivity and specificity vary between laboratories. Up to 5% of patients with vasculitis are negative in IIF but positive by a MPO or PR3 specific ELISA (Savige et  al., 2000). Granulomatosis with polyangiitis (formerly Wegener’s granulomatosis) is predominantly associated with PR3-ANCA, while microscopic polyangiitis is primarily associated with MPO-ANCA. Of note 20% of patients have the alternative ANCA and 10% are negative. Recently, a new autoantibody specificity was described in ANCA-associated vasculitis. The antibodies were directed against lysosome-associated membrane protein-2 (LAMP-2) (Kain et al., 2008). Intriguingly, LAMP-2 shows partly homology with a bacterial protein FimH present on Gram-negative bacteria. Immunization with FimH in rats led to induction of anti-LAMP-2 and a pauci-immune GN. However, whether anti-LAMP-2 is prevalent in ANCA-positive patients remains to be proven since the results are very conflicting (Kain et al., 2012; Roth et al., 2012). Therefore, at present screening for anti-LAMP-2 is not recommended in patients with suspected ANCA vasculitis.

Anti-GBM antibodies Anti-GBM disease is the prototypic example of antibody mediated renal disease. In anti-GBM disease the glomerulus is severely damaged with a characteristic picture of severe extracapillary proliferation with capillary wall necrosis. Patients present with the clinical picture of a RPGN. In some patients, especially in smokers, the lungs may be involved too. Classical Goodpasture’s syndrome is a pulmonary renal syndrome of RPGN with haemoptysis. A renal biopsy in patients with anti-GBM disease discloses a typical picture on IF, with linear IF of Ig along the glomerular capillary wall. In most cases the causative antibody is of the IgG class, however, patients with IgA and IgM class anti-GBM antibodies have also been described. The antibodies in anti-GBM glomerulonephritis are directed against a conformational epitope on the so called non-collagenous (NC1) domain of the collagen IV α3 chain (Hellmark et al., 1997). This epitope becomes accessible if the dimeric form is converted to the monomeric form. Based on this knowledge, sensitive and specific ELISA assays have been developed using recombinant protein for the quantitative measurement of anti-GBM antibodies. These antibodies can be detected in high titres in almost every patient

chapter 17 

with anti-GBM disease. The sensitivity and specificity of the assay is very high, although low titres can be observed in other unrelated conditions. The positive predictive value (PPV) and negative predictive value (NPV) are therefore very high and are ≥ 95% (Sinico et al., 2006). Of note, the ELISA regularly only detects IgG antibodies, thus the rare patient with IgA or IgM anti-GBM antibodies may remain formally undiagnosed as suffering from anti-GBM disease. In approximately 30% of patients with anti-GBM antibodies, ANCA antibodies are also found (two-thirds with anti-MPO specificity). Although the co-occurrence of ANCA and anti-GBM antibodies is mostly associated with a more favourable course of the disease and in these patients treatment is more efficacious, this is not a uniform finding (Rutgers et al., 2005). The epitope of the anti-GBM antibodies in this subset of patients is not different (Hellmark et  al., 1997). ANCA-activated neutrophils unmasking the cryptic epitope of the NC1-domain is an intriguing, but not proven hypothesis. Recently, it was shown that ANCAs precede the formation of anti-GBM antibodies, lending further support for the pathogenic significance of ANCA in these patients (Olson et al., 2011) Anti-GBM antibodies can also be produced in patients with Alport’s disease after renal transplantation (Brainwood et al., 1998; Kalluri et al., 2000). Alport’s disease is caused by mutations in genes encoding the α3, α4, or α5 chains of collagen IV. These patients do not express these collagen chains in their native forms in their kidneys. After renal transplantation, some patients may develop an aggressive form of anti-GBM disease, causing renal failure. Recent studies have indicated that anti-GBM antibodies can be detected in almost all patients, linear deposits of IgG along the glomerular capillary wall are seen in 20–50% of patients, whereas only a minority develop manifest anti-GBM disease. Several studies have looked for the epitopes to which these antibodies are directed. Although the results are equivocal, many patients have antibodies that are directed against the NC1 domain of collagen α5, especially in patients with X-linked Alport syndrome (Borza, 2007). These antibodies may not be detected by commercial anti-GBM ELISA assays, that use the specific NC1α3 antigen as substrate.

Antipodocytic antibodies It has long been suggested that idiopathic membranous nephropathy (iMN) can be caused by the binding of a circulating antibody to an antigen that is present on podocytes, but the responsible antigen in human membranous nephropathy was elusive. In 2009, the M-type phospholipase A2 receptor (PLA2R) was identified as an important antigenic target (Beck et al., 2009). The authors showed that antibodies against PLA2R, primarily of the IgG4 subclass, were present in approximately 70% of patients with iMN. This study finally proved that iMN is an auto-immune disease. A Western blot technique had been used for the detection of antibodies against PLA2R, but an indirect IF technique is now commercially available. For this assay slides containing biochips of HEK 293 cells transfected with cDNA coding for PLA2R, or non-transfected cells as control are used (Hoxha et al., 2011). A commercial ELISA assay is now available (Dahnrich et al., 2013). Recent studies suggest that antibodies against PLA2R are specific for membranous nephropathy, and particularly the idiopathic form. However, the accuracy of this assay to diagnose iMN and to exclude secondary causes of MN will require prospective studies.

immunological investigation of renal disease

A quantitative assay may be needed for prediction of prognosis and to guide treatment (Hofstra et al., 2012). The role of antibodies against other podocytic antigens such as superoxide dismutase, aldosereductase, and alpha-enolase needs further evaluation.

Antiphospholipid antibodies Antiphospholipid (aPL) antibodies are associated with the antiphospholipid syndrome (APS), which is characterized by (recurrent) arterial and/or venous thromboses and recurrent miscarriages (Levine et  al., 2002). The kidney is often involved, with vascular lesions and variable thrombotic microangiopathy (TMA) (Nochy et al., 1999). APS can be primary (without systemic auto-immunity) or secondary in patients with auto-immune diseases such as SLE (Daugas et al., 2002). The first recognized aPL antibodies were the lupus anticoagulant (LAC) and the anticardiolipin (aCL) antibodies. More recently antibodies to β2-glycoprotein-1 have been identified. The LAC is defined as a serum factor that prolongs phospholipid-dependent coagulation tests in platelet-poor plasma. Examples are the activated partial thromboplastin time (APTT), the kaolin clotting time, or diluted Russell’s viper venom clotting time (DRVVT). Tests are considered positive if the coagulation defect cannot be corrected by addition of normal plasma, but becomes normal after addition of excess anionic phospholipids or freeze-thawed thrombocytes. There are a number of different methods to detect LAC activity (Jacobsen et al., 2001). By only using the standard APTT test, with a cut of value of 67 seconds, 47% of lupus anticoagulant positive samples will be missed (Jacobsen et al., 2001). DRVVT is considered the most sensitive test so the presence of LAC must be ascertained by using at least two independent tests but acknowledging that the accuracy and reproducibility of the test procedures is low. ACL antibodies are directed against cardiolipin, and can be detected in ELISA. Although some aCL antibodies have LAC activity, the results do not always overlap. Thus when searching for aPL antibodies both tests are necessary. Recently, it has been demonstrated that aPL antibodies may be directed at various plasma proteins that have affinity for anionic phospholipids. In fact, aCL antibodies recognize a conformational epitope formed between CL and β2-glycoprotein-1 (β2GP-1). Therefore, β2GPI is an essential co-factor in these assays. Because of this, antibodies to β2-glycoprotein-1 can also be tested directly in ELISA. This is especially necessary in patients in whom there is a high suspicion for APS but negative aCL or LAC tests. For a diagnosis of APS at least one clinical manifestation should have occurred (thrombosis or pregnancy morbidity) together with positive tests 12 weeks apart for one of the aPL assays (aCL, LAC or β2GP-1). The rate of thrombosis increases significantly when more aPL tests are positive: single positivity, 27.6%; double, 38.8; or triple, 66.7% (Lee et al., 2003). Additional specificities include proteins such as prothrombin, protein C, protein S, and annexin V. Of note, these proteins are only recognized in the assays if bound to anionic surfaces. In ELISA antigens must therefore be coated specifically on γ-irradiated polystyrene plates. Although the association between the presence of aPL antibodies and thrombosis is well recognized, the pathogenicity of the antibodies is unclear because aCL antibodies may be detectable in

135

136

Section 1  

assessment of the patient with renal disease

2–4% of healthy controls. The incidence is higher (15%) in patients with rheumatoid arthritis or SLE (Merkel et al., 1996). Other specificities are also found with high frequency in patients with SLE. A 21–56% prevalence of antibodies to ß2GP-1, prothrombin, protein C, protein S, and annexin V has been reported in patients with SLE (Nojima et al., 2001). Patients with positive lupus anticoagulant had significantly more frequent antibodies against prothrombin and ß2GP-1, with no difference in the other antibody specificities. The association between LAC, aCL, and aβ2GP-1 antibodies and the APS is well established (Levine et al., 2002). In general, LAC activity is more specific whereas aCL and aβ2GP-1 antibodies show higher sensitivity. However, it was found that the presence of one of the latter antibodies has a poor association with thrombosis (Galli et al., 2003). Understanding of the APS and the role of the aPL antibodies in various clinical manifestations should improve with the development of more specific and better standardized assays. An association was observed between antibodies against ß2GP-1 and prothrombin and arterial thrombosis; between antibodies against protein S and venous thrombosis; and between anti-annexin V and fetal loss (Nojima et al., 2001). These results need confirmation in prospective studies. Patients with an established APS need life-long full oral anticoagulation. For the monitoring of the intensity of oral anticoagulation the international normalized ratio (INR) is used aiming at an INR between 2.0 and 3.0 (Crowther et al., 2003). It should be realized that monitoring of the INR may be difficult in patients with LAC since some available tests may be influenced. In the presence of LAC, values of INR may be falsely elevated, and the level of anticoagulation may be overestimated (Moll and Ortel, 1997).

Antimicrobial responses (See Chapter 164.) The kidney can be damaged in the course of many infectious diseases (see Chapters 77–9). Poststreptococcal glomerulonephritis (PSGN), hepatitis B, and hepatitis C are three important examples.

Antibodies against streptococcal antigens The epidemiology of PSGN has changed in the past three decades (Rodriguez-Iturbe and Musser, 2008). In central Europe, the incidence is low, the elderly being more affected than children. The disease most often occurs in epidemics, and a high incidence is reported in rural communities with low socioeconomic status. PSGN was once regarded the best example of an immune-complex mediated glomerular disease, caused by entrapment in the glomerular capillary wall of circulating immune complexes that consisted of streptococcal antigens and antistreptococcal antibodies. Alternative hypotheses included in situ formation of immune complexes by binding of antibodies against bacterial antigens planted in the GBM, or by binding of antibodies to intrinsic glomerular antigens that mimic bacterial epitopes. The focus has shifted recently to two newly defined, potential nephritogenic antigens, that is, the nephritis-associated plasmin receptor (NAPlr) and the streptococcal pyrogenic exotoxin B (SPEB) (Rodriguez-Iturbe and Musser, 2008). Both NAPlr and SPEB have been detected in glomerular deposits and antibodies against these antigens were present in a high percentage of patients with PSGN, although results differed markedly between populations, suggesting that more than one antigen may be responsible for development of PSGN. The diagnosis of

PSGN requires proof of a prior streptococcal infection. Most studies evaluate the production of antibodies directed against streptococcal antigens. Although antibodies against NAPlr and SPEB may be most relevant, thus far no large study has evaluated their accuracy in diagnosing PSGN. Antibodies against streptolysin O (ASO) are not very accurate, since many patients, in particular those with a streptococcal skin infection, do not produce ASO antibodies. The detection of antibodies directed to other streptococcal antigens, such as deoxyribonuclease B (DNAse B), streptokinase, hyaluronidase, zymogen, GAPDH, and NAD+ glycolase, may be more sensitive and specific for the diagnosis of post-streptococcal infection. In children with acute PSGN the sensitivity of ASO, anti-DNAse and antizymogen titres has been studied (Parra et al., 1998). When compared to controls ROC curves showed superiority of anti-zymogen > anti-DNAse B > ASO. Sensitivity and specificity was 42% and 84% for ASO, 73% and 89% for anti-DNAse B, and 87% and 85% for anti-zymogen. Of note, increased titres of these antibodies were also observed in patients with upper respiratory tract infections or impetigo who did not experience a glomerulonephritis. Thus, the above-mentioned specificity of the assays does not hold if patients with infections, without evidence of renal involvement, are included.

Hepatitis C Hepatitis C is held responsible for various renal diseases such as MPGN type I, membranous nephropathy, and cryoglobulinaemic glomerulonephritis (Lauer and Walker, 2001; Charles and Dustin, 2009). The detection of hepatitis C is based on the detection of antibodies against hepatitis C antigens. An enzyme immunoassay (EIA) is used for screening. This test is very sensitive when used in low-risk population (99%). However, false-positive results are common, thus positive results should always be confirmed by a recombinant immunoblot assay (Chandler, 2000). False-negative results occur particularly in patients with end-stage kidney disease and patients using immunosuppression. Reported sensitivity of an EIA was only 53% and 72% in two studies in dialysis units (Schroter et al., 1997; Kalantar-Zadeh et al., 2005). In patients with high suspicion for hepatitis C, HCV RNA can be detected by rtPCR, a very sensitive method that allows detection of virus as early as 2 weeks after infection. Since viral RNA is unstable, blood samples must be processed within 3 hours. In patients with cryoglobulinaemia, HCV viral RNA may be predominantly contained in the cryoprecipitate. Special care must be taken to draw the blood at 37oC and that the rtPCR is performed on the cryoprecipitate (Charles and Dustin, 2009).

Hepatitis B Various serological markers can be used to evaluate a hepatitis B infection, for example, HBsAg, HBe antigen, and anti-HBs, anti-HBc, and anti-HBe antibodies. The test results provide an indication on the time sequence and the infectivity of the blood (Edey et al., 2010). The presence of HBsAg and HBe suggests infectivity and anti HBs indicates immunity. However, HBsAg and anti-HBs antibody negative patients may be infectious, particularly if organs are used for transplantation. There is a risk of transmission particularly if an organ is transplanted into a non-immune recipient, the risk being higher with liver than with kidney transplantation (Delmonico and Snydman, 1998). It has been suggested that immunosuppressed patients (including patients on haemodialysis) may

chapter 17 

not develop anti-HBs antibodies after exposure to the viral antigen. If suspicion is high, HBV DNA can be detected by sensitive PCR techniques. There is continuous uncertainty about the presence and prevalence of occult HBV infection in HBsAg negative dialysis patients, with estimates ranging from 0% to 26.6% (Fabrizi et al., 2005; Di Stefano et al., 2009).

Immunological studies in patients with specific renal syndromes: guidelines Few studies have specifically assessed the value of testing specific immunological markers in establishing a diagnosis or monitoring disease activity in patients with renal diseases. We will briefly discuss the data that are most relevant for the important clinical renal syndromes and several specific disease entities.

Nephrotic syndrome Two studies have argued against the routine testing of patients with proteinuria or a nephrotic syndrome (Howard et  al., 1990; Chew et  al., 1999). Measurement of ANA, rheumatoid factor, C3, C4, CH50, HbsAg, Venereal Disease Research Laboratory (VDRL) tests, cryoglobulins, and serum and urine electrophoresis did not improve diagnostic accuracy. Therefore, in almost all cases a renal biopsy is needed to establish a diagnosis (Richards et al., 1994). Based on the renal biopsy results (and sometimes on clinical grounds), additional serological tests can be performed that can strengthen the diagnosis and sometimes are helpful during follow-up. If validated, measurement of anti-PLA2R may become the first accurate diagnostic test for a specific cause of the nephrotic syndrome.

Rapidly progressive glomerulonephritis/acute nephritic syndrome (See Chapters 70–4.) In patients with an RPGN glomerular damage occurs often very rapidly and is irreversible if treatment is postponed. Therefore, a rapid diagnosis and early treatment is urgently needed. Serological markers can be very helpful in establishing a diagnosis in these patients. It is important to stress that an examination of the urinary sediment for the presence of dysmorphic erythrocytes and/ or erythrocyte casts remains the most important diagnostic tool to differentiate between RPGN and other causes of acute kidney injury. The presence of high titres of anti-GBM antibodies in patients with a RPGN is virtually pathognomonic for anti-GBM disease. By

Table 17.3  Sensitivity and specificity of ANCA using combined IIF and ELISA in systemic vasculitis Sensitivity

Specificity

Granulomatosis with polyangiitis (formerly Wegener’s disease; N = 97)

73%

98%

Microscopic polyangiitis (N = 44)

67%

98%

Idiopathic RPGN (N = 12)

82%

98%

Adapted from Hagen et al. (1998). Numbers in parentheses represent number of patients in the study cohort. Specificity data are based on 184 disease controls.

immunological investigation of renal disease

using the widely available, specific, and sensitive ELISA, a diagnosis can be made within a day. Although most centres still prefer to perform a renal biopsy, such an approach can be challenged as it may delay institution of treatment with plasma exchange. In patients with anti-GBM antibodies, a search for the presence of ANCA is also warranted. Clinicians should also be aware of the possibility that rare cases of anti-GBM disease are caused by non-IgG antibodies, that will not be detected in the standard assays. Regular assessment of the anti-GBM antibody titres during follow-up may guide treatment. In general, antibodies will disappear within 12–16 weeks after start of treatment, and at that time point immunosuppressive treatment can be rapidly tapered and stopped. Renal transplantation in a patient with anti-GBM disease should be postponed until no anti-GBM antibodies have been detected for at least 3 months. The finding of ANCA has a high predictive value in patients with RPGN. In fact, some authors have argued that in patients with extrarenal symptoms that fit with a multisystemic vasculitis, and who present with acute kidney injury, the detection of ANCA is sufficient for the diagnosis and a renal biopsy may be omitted. Data on the sensitivity and the specificity of ANCA for the diagnosis of systemic or renal limited vasculitis are given in Table 17.3 (Hagen et al., 1998). The diagnostic accuracy is best if ANCA are analysed by both IIF and ELISA techniques. In a meta-analysis, combined testing for c-ANCA/αPR3 and p-ANCA αMPO showed a sensitivity of 85.5% and a specificity of 98.6% for the diagnosis of small vessel vasculitis (Choi et al., 2001). One should realize that the high positive and negative predictive values in this and other studies only hold for patients with a (rapidly) deteriorating renal function, in whom the prevalence of systemic vasculitis is high. Predictability is much lower in patients with moderate renal failure (Falk et al., 2000) (Table 17.4). In such patients a renal biopsy is always needed to establish a diagnosis. Recently, it was demonstrated that during active phases of ANCA-associated vasculitis plasma levels of C3a, C5a, C5b-9, and Bb were elevated while properdin levels were diminished. Activation of the classical pathway was not observed since C4d levels were not different between active and non-active disease. Factor Bb levels were significantly associated with the amount of cellular crescents and clinical disease activity (Gou et al., 2013). These data indicate activation of the alternative pathway (Kallenberg and Heeringa, 2013) It is still unknown whether evaluation of ANCA activity is of value in the follow-up of patients. In 50–60% of patients titres fluctuate parallel to disease activity. In 83% of patients a favourable response to immunosuppressive treatment is accompanied by a negative ANCA test after a median of 12 weeks (Jayne et al., 1995). The predictive value of a rise in ANCA titre (if measured monthly) for a clinical relapse has a sensitivity of 57% and a specificity of 45%. The specificity increases to 72% if transient increases after conversion from cyclophosphamide to azathioprine and rises at the moment of clinical relapse are excluded. These results were confirmed in a recent prospective study. The predictive value for a relapse of a rise in c-ANCA titre in IF was 57% and 71% if assessed in a PR3-ELISA (Boomsma et  al., 2000). However, one should realize that disease quiescence is possible with persistently positive ANCA titres and that relapses can occur while the ANCA test remains negative. Serial measurements of ANCA is therefore of limited use (Tomasson et al., 2012).

137

138

Section 1  

assessment of the patient with renal disease

Table 17.4  Estimated positive (PPV) and negative (NPV) predictive values of ANCA for pauci-immune crescentic glomerulonephritis Clinical picture

PPV

NPV

RPGN

98%

80%

Nephritic syndrome and Serum creatinine > 270 μmol/L (A)

92%

80%

Serum creatinine 135–270 μmol/L (B)

77%

98%

Serum creatinine < 135 μmol/L (C)

47%

99%

Adapted from Falk et al. (2000). Data are for adult patients with observed prevalences of pauci-immune crescentic glomerulonephritis 47% for RPGN, 21% for A, 7% for B, and 2% for C.

In patients with a RPGN that are anti-GBM and ANCA negative, additional serological tests may be done based on clinical symptoms or biopsy results (Fig. 17.4). In these patients abnormalities in serum complement levels, the presence of auto-antibodies, or the detection of cryoglobulins or paraproteins are not enough sensitive or specific to establish a diagnosis, however, a positive finding may be helpful in confirming a diagnosis or be used to guide treatment. Abnormalities in complement levels in renal diseases are given in Table 17.2. For most diagnoses there is no evidence that serial measurements of complement can be used to monitor disease activity.

Systemic lupus erythematosus Laboratory tests are pivotal in establishing the diagnosis and may be helpful in monitoring the renal disease activity in patients with SLE. The ANA test is a valuable screening test with high sensitivity but low specificity. If positive, additional tests must be performed to establish antibody specificities. The presence of anti-dsDNA and anti-Sm antibodies establishes a diagnosis of SLE with 85% sensitivity and 93% specificity. Anti-nucleosome antibodies (van der Vlag and Berden, 2011) and antibodies to C1q (Kallenberg, 2008) may be better predictors of renal involvement in SLE; however, prospective data are limited. Complement levels are not helpful in the diagnosis of SLE, but persistently low CH50

values raise the possibility of a deficiency of one of the early complement factors. Antinuclear antibody titres are valuable in monitoring SLE nephritis disease activity during follow-up. If anti-dsDNA is measured monthly by the Farr assay, onset or flares of lupus nephritis will be accompanied by increases of anti-dsDNA titres in > 90% of patients (ter Borg et al., 1990). Measuring of anti-C1q antibodies might be more accurate in predicting renal relapses (Moroni et al., 2001). The likelihood of renal involvement in SLE patients who are negative for anti-C1q or anti-dsDNA antibodies is very low (specificity between 92% and 100%) (Trendelenburg et al., 1999; Moroni et al., 2001). Unfortunately, the anti-C1q assay is not available for routine use. Whether measurement of complement is valuable for monitoring disease activity is still unclear. Monitoring of C4 is not useful, although a low C4 value (< 110 mg/L) at the moment of renal remission, was identified as a risk marker for a flare (Illei et  al., 2002). Measurement of C3 appears more useful, with sensitivities ranging from 20% to 95% and specificities between 74% and 94% (Esdaile et al., 1996). However, in many patients with renal disease activity normal C3 levels are found, whereas decreased C3 levels may be present in patients without renal disease activity. In Table 17.5 the experience in a 6-year prospective study in 228 patients is given (Moroni et al., 2009). It underscores the low PPV of serology for confirming renal lupus flares.

Thrombotic microangiopathy TMA is a descriptive diagnosis and associated with various clinical syndromes (see Chapter 174). Dysregulation of the complement system is an important cause of TMA, a condition also defined as atypical HUS. In patients who present with the triad of Coombs-negative haemolytic anaemia, thrombocytopenia, and renal insufficiency an extensive search for an underlying abnormality is needed. This includes the measurement of the levels of complement proteins, a search for a mutation in the relevant proteins, and the detection of antibodies. Table 17.6 summarizes the recent guidelines (Taylor et al., 2010). Alternative diagnoses must be considered. The presence of aPL antibodies is characteristic for the APS.

Serologic test/IF pattern renal biopsy ANCA+ C3/C4 norm

Anti-MPO+

No extra-renal sympt PI-CGN

ICx markers + C3 low C4 low or norm

Anti-GBM+ C3/C4 norm

Anti-Pr3+

Syst vasc

Asthma eosinophilia

Resp. granulomea

MPA

EGPA

GPA

Anti-GBM

ANA+ anti-dsDNA+ anti-C1q+

[IgA] elevated purpura

Cryoglob + HCV+

Antimicrob resp+

LN

HSP

CGN

PIGN

Fig. 17.4  Diagnostic approach of the patient with a rapidly progressive glomerulonephritis. Anti-GBM = anti-GBM nephritis; CGN = cryoglobulinaemic glomerulonephritis; CSS = Churg–Strauss syndrome; HCV = hepatitis C virus; HSP = Henoch–Schönlein purpura; IC = immune-complex; [IgA] = serum IgA concentration; LN = lupus nephritis; MPA = microscopic polyangiitis; PI-CGN = pauci-immune crescentic glomerulonephritis; PIGN = post-infectious glomerulonephritis; WG = Wegener’s granulomatosis.

chapter 17 

Table 17.5  Positive (PPV) and negative predictive values (NPV) of serological test for lupus nephritis flare Test

Sensitivity

Specificity

PPV

NPV

OR

Anti-C1q

81

71

38

94

11.8 (4.9–18.1)

Anti-dsDNA

70

67

31

91

ND

(reduced) C3

79

51

28

93

2.99 (1.5–5.8)

(reduced) C4

74

64

31

92

3.3 (1.7–6.5)

OR = odds ratio (with 95% confidence interval).

Table 17.6  Investigations in patients with atypical haemolytic uraemic syndrome Exclude other causes Infections: Shiga toxin-producing bacteria; Streptococcus pneumonia;HIV Malignancy SLE, antiphospholipid syndrome Drugs (mitomycin, vincristine, calcineurin inhibitors) Exclude thrombotic thrombocytopenic purpura ADAMTS13 activity < 5% (genetic or immune) Disorders of complement regulation Serum: C3, C4, C3d, factor H, factor I Expression of CD46 on PMBC by FACS analysis Mutation screening: factor H, factor I, factor B, C3, CD46 Auto-antibody screening: factor H, factor B CD46 = membrane cofactor protein; PMBC = peripheral blood mononuclear cells; SLE = systemic lupus erythematosus. Adapted from Taylor et al. (2010).

Thrombotic thrombocytopenic purpura (TTP) is caused by abnormalities in the activity of serum von Willebrand factor (vWF) protease. This protease is a zinc metalloprotease called ADAMTS13, encoded by a gene on chromosome 9q34. This protease rapidly degrades large vWF multimers that are formed in the endothelial cells. A decreased activity of the protease results in the persistence of large vWF multimers in the circulation. These large multimers cause endothelial damage and promote thrombosis. Familial forms of TTP have been associated with a deficiency of this vWF protease. In several families mutations in the ADAMTS13 gene have been found (Levy et  al., 2001). The acquired form of TTP has been associated with a functional loss of vWF protease activity due to circulating antibodies. TTP typically is associated with an ADAMTS13 activity < 5% (Furlan et al., 1998). Tests for these abnormalities are only available in specialist laboratories.

References Allen, A. C., Bailey, E. M., Brenchley, P. E., et al. (2001). Mesangial IgA1 in IgA nephropathy exhibits aberrant O-glycosylation: observations in three patients. Kidney Int, 60, 969–73. Alpers, C. E. and Smith, K. D. (2008). Cryoglobulinaemia and renal disease. Curr Opin Nephrol Hypertens, 17, 243–9.

immunological investigation of renal disease

Beck, L. H., Jr., Bonegio, R. G., Lambeau, G., et al. (2009). M-type phospholipase A2 receptor as target antigen in idiopathic membranous nephropathy. N Engl J Med, 361, 11–21. Berthoux, F., Suzuki, H., Thibaudin, L., et al. (2012). Autoantibodies targeting galactose-deficient IgA1 associate with progression of IgA nephropathy. J Am Soc Nephrol, 23, 1579–87. Biesen, R., Dahnrich, C., Rosemann, A., et al. (2011). Anti-dsDNA-NcX ELISA: dsDNA-loaded nucleosomes improve diagnosis and monitoring of disease activity in systemic lupus erythematosus. Arthritis Res Ther, 13, R26. Blain, H., Cacoub, P., Musset, L., et al. C. (2000). Cryofibrinogenaemia: a study of 49 patients. Clin Exp Immunol, 120, 253–60. Bochtler, T., Hegenbart, U., Heiss, C., et al. (2008). Evaluation of the serum-free light chain test in untreated patients with AL amyloidosis. Haematologica, 93, 459–62. Boomsma, M. M., Stegeman, C. A., van der Leij, M. J., et al. (2000). Prediction of relapses in Wegener’s granulomatosis by measurement of antineutrophil cytoplasmic antibody levels: a prospective study. Arthritis Rheum, 43, 2025–33. Borza, D. B. (2007). Autoepitopes and alloepitopes of type IV collagen: role in the molecular pathogenesis of anti-GBM antibody glomerulonephritis. Nephron Exp Nephrol, 106, e37–43. Brainwood, D., Kashtan, C., Gubler, M. C., et al. (1998). Targets of alloantibodies in Alport anti-glomerular basement membrane disease after renal transplantation. Kidney Int, 53, 762–6. Chandler, L. (2000). Diagnostic tests for hepatitis C virus. Clin Microbiol Newsletter, 22, 145–9. Charles, E. D. and Dustin, L. B. (2009). Hepatitis C virus-induced cryoglobulinaemia. Kidney Int, 76, 818–24. Chew, S. T., Fitzwilliam, J., Indridason, O. S., et al. (1999). Role of urine and serum protein electrophoresis in evaluation of nephrotic-range proteinuria. Am J Kidney Dis, 34, 135–9. Choi, H. K., Liu, S., Merkel, P. A., et al. (2001). Diagnostic performance of antineutrophil cytoplasmic antibody tests for idiopathic vasculitides: metaanalysis with a focus on antimyeloperoxidase antibodies. J Rheumatol, 28, 1584–90. Coremans, I. E., Spronk, P. E., Bootsma, H., et al. (1995). Changes in antibodies to C1q predict renal relapses in systemic lupus erythematosus. Am J Kidney Dis, 26, 595–601. Couser, W. G. (2012). Basic and translational concepts of immune-mediated glomerular diseases. J Am Soc Nephrol, 23, 381–99. Crowther, M. A., Ginsberg, J. S., Julian, J., et al. (2003). A comparison of two intensities of warfarin for the prevention of recurrent thrombosis in patients with the antiphospholipid antibody syndrome. N Engl J Med, 349, 1133–8. D’Amico, G. (1988). Clinical features and natural history in adults with IgA nephropathy. Am J Kidney Dis, 12, 353–7. Dahnrich, C., Komorowski, L., Probst, C., et al. (2013). Development of a standardized ELISA for the determination of autoantibodies against human M-type phospholipase A2 receptor in primary membranous nephropathy. Clin Chim Acta, 421, 213–8. Dammacco, F., Sansonno, D., Piccoli, C., et al. (2001). The cryoglobulins: an overview. Eur J Clin Invest, 31, 628–38. Daugas, E., Nochy, D., Huong, D. L., et al. (2002). Antiphospholipid syndrome nephropathy in systemic lupus erythematosus. J Am Soc Nephrol, 13, 42–52. Daval, S., Tridon, A., Mazeron, N., et al. (2007). Risk of antigen excess in serum free light chain measurements. Clin Chem, 53, 1985–6. Delmonico, F. L. and Snydman, D. R. (1998). Organ donor screening for infectious diseases: review of practice and implications for transplantation. Transplantation, 65, 603–10. Desplat-Jego, S., Johanet, C., Escande, A., et al. (2007). Update on Anti-Saccharomyces cerevisiae antibodies, anti-nuclear associated anti-neutrophil antibodies and antibodies to exocrine pancreas detected by indirect immunofluorescence as biomarkers in chronic inflammatory bowel diseases: results of a multicenter study. World J Gastroenterol, 13, 2312–8.

139

140

Section 1  

assessment of the patient with renal disease

Di Stefano, M., Volpe, A., Stallone, G., et al. (2009). Occult HBV infection in hemodialysis setting is marked by presence of isolated antibodies to HBcAg and HCV. J Nephrol, 22, 381–6. Dispenzieri, A., Kyle, R., Merlini, G., et al. (2009). International Myeloma Working Group guidelines for serum-free light chain analysis in multiple myeloma and related disorders. Leukemia, 23, 215–24. Edey, M., Barraclough, K., and Johnson, D. W. (2010). Review article: Hepatitis B and dialysis. Nephrology (Carlton), 15, 137–45. Esdaile, J. M., Joseph, L., Abrahamowicz, M., et al. (1996). Routine immunologic tests in systemic lupus erythematosus: is there a need for more studies? J Rheumatol, 23, 1891–6. Fabrizi, F., Messa, P. G., Lunghi, G., et al. (2005). Occult hepatitis B virus infection in dialysis patients: a multicentre survey. Aliment Pharmacol Ther, 21, 1341–7. Falk, R., Nachman, P. H., Hogan, S. L., et al. (2000). ANCA vasculitis and glomerulonephritis: a Chapel Hill perspective. Semin Nephrol, 20, 233–43. Furlan, M., Robles, R., Galbusera, M., et al. (1998). von Willebrand factor-cleaving protease in thrombotic thrombocytopenic purpura and the hemolytic-uremic syndrome. N Engl J Med, 339, 1578–84. Galli, M., Luciani, D., Bertolini, G., et al. (2003). Lupus anticoagulants are stronger risk factors for thrombosis than anticardiolipin antibodies in the antiphospholipid syndrome: a systematic review of the literature. Blood, 101, 1827–32. Gou, S. J., Yuan, J., Chen, M., et al. (2013). Circulating complement activation in patients with anti-neutrophil cytoplasmic antibody-associated vasculitis. Kidney Int, 83, 129–37. Guiard, E., Karras, A., Plaisier, E., et al. (2011). Patterns of noncryoglobulinaemic glomerulonephritis with monoclonal Ig deposits: correlation with IgG subclass and response to rituximab. Clin J Am Soc Nephrol, 6, 1609–16. Hagen, E. C., Daha, M. R., Hermans, J., et al. (1998). Diagnostic value of standardized assays for anti-neutrophil cytoplasmic antibodies in idiopathic systemic vasculitis. EC/BCR Project for ANCA Assay Standardization. Kidney Int, 53, 743–53. Hebert, L. A., Cosio, F. G., and Neff, J. C. (1991). Diagnostic significance of hypocomplementemia. Kidney Int, 39, 811–21. Heher, E. C., Goes, N. B., Spitzer, T. R., et al. (2010). Kidney disease associated with plasma cell dyscrasias. Blood, 116, 1397–404. Hellmark, T., Segelmark, M., and Wieslander, J. (1997). Anti-GBM antibodies in Goodpasture syndrome; anatomy of an epitope. Nephrol Dial Transplant, 12, 646–8. Hofstra, J. M., Debiec, H., Short, C. D., et al. (2012). Antiphospholipase A2 receptor antibody titer and subclass in idiopathic membranous nephropathy. J Am Soc Nephrol, 23, 1735–43. Howard, A. D., Moore, J., Jr., Gouge, S. F., et al. (1990). Routine serologic tests in the differential diagnosis of the adult nephrotic syndrome. Am J Kidney Dis, 15, 24–30. Hoxha, E., Harendza, S., Zahner, G., et al. (2011). An immunofluorescence test for phospholipase-A(2)-receptor antibodies and its clinical usefulness in patients with membranous glomerulonephritis. Nephrol Dial Transplant, 26, 2526–32. Hutchison, C. A., Harding, S., Hewins, P., et al. (2008). Quantitative assessment of serum and urinary polyclonal free light chains in patients with chronic kidney disease. Clin J Am Soc Nephrol, 3, 1684–90. Illei, G. G., Takada, K., Parkin, D., et al. (2002). Renal flares are common in patients with severe proliferative lupus nephritis treated with pulse immunosuppressive therapy: long-term followup of a cohort of 145 patients participating in randomized controlled studies. Arthritis Rheum, 46, 995–1002. Jacobsen, E. M., Barna-Cler, L., Taylor, J. M., et al. (2001). The evaluation of clotting times in the laboratory detection of lupus anticoagulants. Thromb Res, 104, 275–82. Jayne, D. R., Gaskin, G., Pusey, C. D., et al. (1995). ANCA and predicting relapse in systemic vasculitis. Q J Med, 88, 127–33. Kain, R., Exner, M., Brandes, R., et al. (2008). Molecular mimicry in pauci-immune focal necrotizing glomerulonephritis. Nat Med, 14, 1088–96.

Kain, R., Tadema, H., McKinney, E. F., et al. (2012). High prevalence of autoantibodies to hLAMP-2 in anti-neutrophil cytoplasmic antibody-associated vasculitis. J Am Soc Nephrol, 23, 556–66. Kalantar-Zadeh, K., Miller, L. G., and Daar, E. S. (2005). Diagnostic discordance for hepatitis C virus infection in hemodialysis patients. Am J Kidney Dis, 46, 290–300. Kallenberg, C. G. (2008). Anti-C1q autoantibodies. Autoimmun Rev, 7, 612–15. Kallenberg, C. G. and Heeringa, P. (2013). Complement is crucial in the pathogenesis of ANCA-associated vasculitis. Kidney Int, 83, 16–18. Kalluri, R., Torre, A., Shield, C. F., 3rd, et al. (2000). Identification of alpha3, alpha4, and alpha5 chains of type IV collagen as alloantigens for Alport posttransplant anti-glomerular basement membrane antibodies. Transplantation, 69, 679–83. Katzmann, J. A., Clark, R. J., Abraham, R. S., et al. (2002). Serum reference intervals and diagnostic ranges for free kappa and free lambda immunoglobulin light chains: relative sensitivity for detection of monoclonal light chains. Clin Chem, 48, 1437–44. Katzmann, J. A., Kyle, R. A., Benson, J., et al. (2009). Screening panels for detection of monoclonal gammopathies. Clin Chem, 55, 1517–22. Kumar, S., Dispenzieri, A., Katzmann, J. A., et al. (2010). Serum immunoglobulin free light-chain measurement in primary amyloidosis: prognostic value and correlations with clinical features. Blood, 116, 5126–9. Lachmann, H. J., Gallimore, R., Gillmore, J. D., et al. (2003). Outcome in systemic AL amyloidosis in relation to changes in concentration of circulating free immunoglobulin light chains following chemotherapy. Br J Haematol, 122, 78–84. Lauer, G. M. and Walker, B. D. (2001). Hepatitis C virus infection. N Engl J Med, 345, 41–52. Lee, E. Y., Lee, C. K., Lee, T. H., et al. (2003). Does the anti-beta2-glycoprotein I antibody provide additional information in patients with thrombosis? Thromb Res, 111, 29–32. Levine, J. S., Branch, D. W., and Rauch, J. (2002). The antiphospholipid syndrome. N Engl J Med, 346, 752–63. Levy, G. G., Nichols, W. C., Lian, E. C., et al. (2001). Mutations in a member of the ADAMTS gene family cause thrombotic thrombocytopenic purpura. Nature, 413, 488–94. Merkel, P. A., Chang, Y., Pierangeli, S. S., et al. (1996). The prevalence and clinical associations of anticardiolipin antibodies in a large inception cohort of patients with connective tissue diseases. Am J Med, 101, 576–83. Moldoveanu, Z., Wyatt, R. J., Lee, J. Y., et al. (2007). Patients with IgA nephropathy have increased serum galactose-deficient IgA1 levels. Kidney Int, 71, 1148–54. Moll, S. and Ortel, T. L. (1997). Monitoring warfarin therapy in patients with lupus anticoagulants. Ann Intern Med, 127, 177–85. Moroni, G., Radice, A., Giammarresi, G., et al. (2009). Are laboratory tests useful for monitoring the activity of lupus nephritis? A 6-year prospective study in a cohort of 228 patients with lupus nephritis. Ann Rheum Dis, 68, 234–7. Moroni, G., Trendelenburg, M., Del Papa, N., et al. (2001). Anti-C1q antibodies may help in diagnosing a renal flare in lupus nephritis. Am J Kidney Dis, 37, 490–8. Motyckova, G. and Murali, M. (2011). Laboratory testing for cryoglobulins. Am J Hematol, 86, 500–2. Nasr, S. H., Radhakrishnan, J., and D’agati, V. D. (2013). Bacterial infection-related glomerulonephritis in adults. Kidney Int, 83, 792–803. Nochy, D., Daugas, E., Droz, D., et al. (1999). The intrarenal vascular lesions associated with primary antiphospholipid syndrome. J Am Soc Nephrol, 10, 507–18. Nojima, J., Kuratsune, H., Suehisa, E., et al. (2001). Association between the prevalence of antibodies to beta(2)-glycoprotein I, prothrombin, protein C, protein S, and annexin V in patients with systemic lupus erythematosus and thrombotic and thrombocytopenic complications. Clin Chem, 47, 1008–15.

chapter 17 

Olson, S. W., Arbogast, C. B., Baker, T. P., et al. (2011). Asymptomatic autoantibodies associate with future anti-glomerular basement membrane disease. J Am Soc Nephrol, 22, 1946–52. Palladini, G., Russo, P., Bosoni, T., et al. (2009). Identification of amyloidogenic light chains requires the combination of serum-free light chain assay with immunofixation of serum and urine. Clin Chem, 55, 499–504. Parra, G., Rodriguez-Iturbe, B., Batsford, S., et al. (1998). Antibody to streptococcal zymogen in the serum of patients with acute glomerulonephritis: a multicentric study. Kidney Int, 54, 509–17. Pickering, M. and Cook, H. T. (2011). Complement and glomerular disease: new insights. Curr Opin Nephrol Hypertens, 20, 271–7. Rajkumar, S. V., Kyle, R. A., Therneau, T. M., et al. (2005). Serum free light chain ratio is an independent risk factor for progression in monoclonal gammopathy of undetermined significance. Blood, 106, 812–17. Richards, N. T., Darby, S., Howie, A. J., et al. (1994). Knowledge of renal histology alters patient management in over 40% of cases. Nephrol Dial Transplant, 9, 1255–9. Rodriguez-Iturbe, B. and Musser, J. M. (2008). The current state of poststreptococcal glomerulonephritis. J Am Soc Nephrol, 19, 1855–64. Roth, A. J., Brown, M. C., Smith, R. N., et al. (2012). Anti-LAMP-2 antibodies are not prevalent in patients with antineutrophil cytoplasmic autoantibody glomerulonephritis. J Am Soc Nephrol, 23, 545–55. Rutgers, A., Slot, M., Van Paassen, P., et al. (2005). Coexistence of anti-glomerular basement membrane antibodies and myeloperoxidase-ANCAs in crescentic glomerulonephritis. Am J Kidney Dis, 46, 253–62. Savige, J., Davies, D., Falk, R. J., et al. (2000). Antineutrophil cytoplasmic antibodies and associated diseases: a review of the clinical and laboratory features. Kidney Int, 57, 846–62. Schroter, M., Feucht, H. H., Schafer, P., et al. (1997). High percentage of seronegative HCV infections in hemodialysis patients: the need for PCR. Intervirology, 40, 277–8. Sinico, R. A., Radice, A., Corace, C., et al. (2006). Anti-glomerular basement membrane antibodies in the diagnosis of Goodpasture syndrome: a comparison of different assays. Nephrol Dial Transplant, 21, 397–401. Suzuki, H., Fan, R., Zhang, Z., Brown, R., et al. (2009). Aberrantly glycosylated IgA1 in IgA nephropathy patients is recognized by IgG antibodies with restricted heterogeneity. J Clin Invest, 119, 1668–77.

immunological investigation of renal disease

Taylor, C. M., Machin, S., Wigmore, S. J., et al. (2010). Clinical practice guidelines for the management of atypical haemolytic uraemic syndrome in the United Kingdom. Br J Haematol, 148, 37–47. Ter Borg, E. J., Horst, G., Hummel, E. J., et al. (1990). Measurement of increases in anti-double-stranded DNA antibody levels as a predictor of disease exacerbation in systemic lupus erythematosus. A long-term, prospective study. Arthritis Rheum, 33, 634–43. Terjung, B. and Spengler, U. (2009). Atypical p-ANCA in PSC and AIH: a hint toward a ‘leaky gut’? Clin Rev Allergy Immunol, 36, 40–51. Tervaert, J. W., Van Paassen, P., and Damoiseaux, J. (2007). Type II cryoglobulinaemia is not associated with hepatitis C infection: the Dutch experience. Ann N Y Acad Sci, 1107, 251–8. Tomasson, G., Grayson, P. C., Mahr, A. D., et al. (2012). Value of ANCA measurements during remission to predict a relapse of ANCA-associated vasculitis--a meta-analysis. Rheumatology (Oxford), 51, 100–9. Trendelenburg, M., Marfurt, J., Gerber, I., et al. (1999). Lack of occurrence of severe lupus nephritis among anti-C1q autoantibody-negative patients. Arthritis Rheum, 42, 187–8. Van Der Vlag, J. and Berden, J. H. (2011). Lupus nephritis: role of antinucleosome autoantibodies. Semin Nephrol, 31, 376–89. Vermeersch, P., Gijbels, K., Knockaert, D., et al. (2008a). Establishment of reference values for immunoglobulins in the cryoprecipitate. Clin Immunol, 129, 360–4. Vermeersch, P., Van Hoovels, L., Delforge, M., et al. (2008b). Diagnostic performance of serum free light chain measurement in patients suspected of a monoclonal B-cell disorder. Br J Haematol, 143, 496–502. Wang, G., Csernok, E., De Groot, K., et al. (1997). Comparison of eight commercial kits for quantitation of antineutrophil cytoplasmic antibodies (ANCA). J Immunol Methods, 208, 203–11. Weiner, S. M., Berg, T., Berthold, H., et al. (1998). A clinical and virological study of hepatitis C virus-related cryoglobulinaemia in Germany. J Hepatol, 29, 375–84. Zhao, N., Hou, P., Lv, J., Moldoveanu, Z., et al. (2012). The level of galactose-deficient IgA1 in the sera of patients with IgA nephropathy is associated with disease progression. Kidney Int, 82(7), 790–6.

141

CHAPTER 18

The renal biopsy Ian S. D. Roberts, Philip Mason, and Agnes B. Fogo Introduction The first renal biopsies from living patients were obtained surgically as early as 1896 usually during the course of operations to decapsulate the kidneys in patients with Bright disease, to relieve the ‘intrarenal oedema’, a procedure popular during the first two decades of the twentieth century (Cameron and Hicks, 1997). The first needle renal biopsies were obtained accidentally at attempted liver biopsy in the 1940s. This led Alwall to use the technique to target the kidney directly using the aspiration biopsy needle developed by Iversen in 1944 (but reported in 1952 (Alwall, 1952)). It was the landmark paper by Iversen and Brun in 1951 that popularized the use of the percutaneous renal biopsy (Iversen and Brun, 1951). The technique was refined including positioning the patient prone (originally patients were biopsied sitting) and using a Vim–Silverman (non-aspiration) needle with improved success and tissue adequacy (Muehrcke et  al., 1955). Originally imaging included plain films before and during the procedure, insufflation of perirenal air to aid localization, fluoroscopy, and the use of intravenous urography. Biopsies are now generally performed using real-time ultrasound and a disposable needle with or without a biopsy gun. Initially some pathologists were sceptical that interpretation of such small tissue fragments would be useful. However, after many of the pioneer renal pathologists attended a CIBA foundation meeting in 1955, it became widely accepted. This coincided with the introduction of new techniques for treating specimens for light microscopic, immunohistochemical, and later electron microscopic examination. The percutaneous renal biopsy is now a routine and essential diagnostic tool guiding diagnosis and prognosis.

Indications for renal biopsy Inevitably there is variation between clinicians in the threshold for recommending a renal biopsy. The guiding principle should be whether the biopsy findings are likely to change management or to make a diagnosis that will have some prognostic value. The risk:benefit ratio for each individual patient needs to be taken into account, especially when the risks are higher, for example, in an obese or uncooperative patient or if clotting is impaired. Box 18.1 lists the indications, which are discussed in more detail.

Nephrotic syndrome Most nephrologists agree that biopsy of adults with nephrotic syndrome is essential to establish the diagnosis and plan treatment. It has been argued that adults with nephrotic syndrome should all be treated with a course of corticosteroids (for presumptive

steroid responsive minimal change disease (MCD)) and only those not responding should be biopsied. Two decision analysis papers supporting this policy concluded that routine treatment with high-dose steroids would be associated with fewer complications and lower mortality (Hlatky, 1982; Levey et  al., 1987). However, these were based on old pre-ultrasound complication data and assumed that corticosteroids would be given for MCD, membranous nephropathy, and possibly also membranoproliferative glomerulonephritis. Furthermore, since some adults with MCD may take up to 12 weeks to respond to corticosteroids, the potential risks of high-dose steroids in up to 75% of patients without MCD are themselves significant. Repeated biopsy may be indicated, for example, in patients whose first biopsy suggested MCD but did not respond, or become corticosteroid dependent or resistant to exclude an alternative diagnosis, for example, focal segmental glomerulosclerosis (FSGS) (see Chapter 57). The situation is different in children in whom > 75% (> 90% aged 1–10 years) have steroid sensitive nephrotic syndrome. Most paediatricians recommend biopsy only in those who have not responded to corticosteroids or who have atypical features such as low complement concentrations or microscopic haematuria (although this latter finding is not be particularly discriminative (International Study of Kidney Disease in Children, 1978)).

Acute kidney injury Most acute kidney injury (AKI) is due to prerenal disease, ‘acute tubular injury’ (ATI), or obstruction (see Chapter  220) and the underlying cause is usually obvious and biopsy is unnecessary. Only approximately 4% of AKI is secondary to glomerulonephritis (with or without vasculitis) or acute interstitial nephritis. If the cause is unclear, a biopsy will clarify the diagnosis and inform treatment which might prevent irreversible renal injury. This is particularly important for glomerulonephritis associated with vasculitis, lupus, anti-glomerular basement membrane (GBM) disease and myeloma-associated kidney injury. It is important to remember that patients with these underlying conditions may be systemically ill and therefore also have ATI. Clinical features of systemic diseases, laboratory tests (e.g. antineutrophil cytoplasmic antibodies (ANCAs), antinuclear antibodies (ANAs), complements, anti-GBM antibodies, and myeloma screen) or heavy proteinuria may suggest one of these underlying conditions. A biopsy may also be indicated in patients who fail to recover after an episode of AKI with an identified precipitating cause in order to exclude another diagnosis or to determine the likelihood of recovery.

chapter 18 

Box 18.1  Indications for native renal biopsy ◆ Nephrotic syndrome (except children) ◆ Acute kidney failure without clear cause ◆ Chronic kidney impairment especially with proteinuria and/or haematuria ◆ Non-nephrotic proteinuria ◆ Microscopic haematuria ◆ Systemic diseases with abnormal function and/or heavy proteinuria.

Systemic disease

the renal biopsy

patients are keen to have a diagnosis and are prepared to accept the small risks of the procedure to obtain this.

Microscopic haematuria The arguments for and against biopsy in this setting are similar to those for ‘non-nephrotic’ proteinuria and both may coexist. Isolated haematuria (without collateral evidence of renal disease such as proteinuria), especially in those aged over 40–50  years, needs a urological diagnostic work-up, but renal biopsy is unlikely to change management. Biopsies in this setting most often reveal immunoglobulin (Ig)-A nephropathy IgA or other conditions that are unlikely to progress (Topham et  al., 1994). Individuals with microscopic haematuria being worked up as potential kidney transplant donors should be biopsied since it is important to exclude a glomerulonephritis such as IgA nephropathy.

Patients with known systemic diseases including systemic vasculitis, anti-GBM disease, and systemic lupus erythematosus (SLE) are often biopsied. It has been argued that patients who test positive for ANCA or anti-GBM antibodies do not need a biopsy, since management will not be altered. However, a biopsy may be helpful to assess the acuteness of the illness and the degree of chronic and irreversible damage. Although not very predictive of response to treatment in ANCA-positive disease, and accepting potential sampling bias, the renal biopsy may guide management, for example, in an elderly patient with evidence of long-standing disease (with fibrocellular or fibrous crescents) and much interstitial fibrosis in whom continued immunosuppression may carry significant and unwarranted risks. The decision to biopsy must always be made on an individual basis.

Transplant kidneys

Diabetic patients

Before renal biopsy, the patient should have a renal ultrasound to establish the presence of two normal-sized unobstructed kidneys. Biopsy of a small kidney is associated with higher risks of bleeding and is less likely to provide diagnostic information (small scarred kidneys often show sclerosed glomeruli and interstitial fibrosis with no features of the original disease). The size of the kidney needs to be interpreted in the context of the patient’s size and the cortical thickness should be assessed. Biopsy of a kidney with a cortical thickness of < 1 cm has a higher risk of complications and a lower chance of obtaining diagnostically useful tissue. Anatomic abnormalities, for example, ectopic kidneys, crossed fused ectopia, and horseshoe kidneys, increase the risk of bleeding. Any of these factors need to be taken into account when estimating the balance of risk over benefit for the individual patient. The patient should be tested for any bleeding tendency and have normal clotting tests, a platelet count ≥ 100,000 and not be severely anaemic (haemoglobin < 9 g/dL). The value of the bleeding time test has been controversial. Although there have been reports that it is predictive of complication (Stratta et al., 2007) (although this was only significant for haematoma > 1 cm and not other major or minor bleeding complications), most studies have not found this to be so (Peterson et al., 1998; Manno et al., 2004). Bleeding time in uraemia can be partially reversed with an infusion of DDAVP (0.3 micrograms/kg). The effect is maximal after 30–60 minutes, wearing off by 24 hours. Although not evidence based, many recommend this for patients with poor renal function (e.g. urea > 20 mmol/L or creatinine > 200 µmol/L). Patients with ischaemic heart disease should not be given DDAVP because of the small risk

Diabetic nephropathy commonly occurs in patients with long-standing diabetes. The UKPDS study found that in people with type 2 diabetes the prevalence of microalbuminuria, macroalbuminuria, or an elevated creatinine/renal failure was 25%, 5%, and 0.8% respectively at 10 years following diagnosis (Adler et al., 2003) (see Chapter 149). Patients with diabetic nephropathy usually have evidence of retinopathy and/or neuropathy and when these are present, a biopsy is not usually indicated. In the absence of these complications or with rapid-onset nephrotic syndrome, a short duration of (diagnosed) diabetes and possibly with significant haematuria or systemic disease, a biopsy is indicated to determine if another treatable condition is present. Even when another cause for renal disease is found there is often evidence of diabetic nephropathy as well (Pham et al., 2007).

Proteinuria without the nephrotic syndrome Opinion and threshold for biopsy of patients with ‘non-nephrotic proteinuria’ varies. It is difficult to argue benefit in patients with relatively low-level proteinuria (< 1.5 g/day or PCR of < 150 mg/mmol) and normal renal function. It is true that low-level proteinuria may precede the nephrotic syndrome for most causes of this clinical syndrome (except for MCD). Management of such patients with good blood pressure control and maximizing angiotensin-converting enzyme inhibitor/angiotensin receptor blocker therapy would be the same unless nephrotic syndrome and/or renal impairment were to develop. The threshold for biopsy is lowered as the proteinuria increases but the arguments are essentially the same. Sometimes

Some units routinely take biopsies during implantation either before and/or after reperfusion. This may be helpful in defining the presence or degree of existing disease in the donor kidney, particularly important in older donors. Patients with delayed graft function should be biopsied regularly, even weekly, until function returns to exclude the development of rejection as ATN recovers. This policy may vary depending on the induction therapy (early rejection is unlikely with anti-T-cell-agent induction). Later in the post-transplant course the kidney is biopsied to investigate dysfunction or heavy proteinuria (see Chapters 286 and 289).

The procedure Pre-biopsy evaluation

143

144

Section 1  

assessment of the patient with renal disease

Box 18.2  Safety conditions for renal biopsy ◆ Haemoglobin > 9 g/dL ◆ Platelets >100,000 × 109/L ◆ Prothrombin (PT) and activated partial thromboplastin time (APPT) < 1.2 × control ◆ Blood pressure preferably < 160/95 mmHg ◆ Sterile urine.a a

See text.

of coronary artery thrombosis. The use of the thromboelastogram (TEG) has been reported in a single study to be predictive of bleeding following transplant biopsies although the evidence was a 4% drop in haematocrit and only 2% of patients had gross haematuria or a haematoma on ultrasound (Davis and Chandler, 1995). Antiplatelet agents including aspirin, clopidogrel, non-steroidal anti-inflammatory drugs, omega-3 fatty acids, glycoprotein IIb/ IIIa inhibitors, and dipyridamole, all of which increase bleeding time, should be stopped a week before a routine biopsy. Biopsy of patients on aspirin and clopidogrel should be avoided but if absolutely essential then platelet transfusion will reduce the risk of bleeding. Anticoagulants should be reversed or stopped before biopsy and the decision to change to heparin needs to be considered on an individual patient basis. Unfractionated heparin should be stopped at least 6 hours and low-molecular-weight heparin 24 hours before biopsy. Correction of clotting abnormalities with fresh frozen plasma or clotting factors may be necessary for an urgent biopsy. Box 18.2 lists the recommended criteria and Box 18.3 the relative contraindications for renal biopsy. The concern about performing renal biopsy in patients with urinary tract infection is the risk of a post-biopsy perinephric haematoma becoming infected. Anxious patients may need sedation (e.g. with midazolam) to improve the safety of the procedure. A single kidney is often listed as a contraindication but this is relative (all transplant biopsies are of course single functioning kidneys) and for most patients the risk is lower than an open biopsy. Pregnancy is not a contraindication but it is important to be clear that the biopsy will substantially affect management before delivery.

Renal biopsy technique Consent should be taken before the procedure but counselling should be offered as early as possible. An advice leaflet allows the Box 18.3  Relative contraindications for renal biopsy ◆ Single kidney ◆ Small kidneys with thin cortex ◆ Abnormal anatomy including cysts ◆ Obesity ◆ Uncooperative patient ◆ Uncontrolled hypertension ◆ Infection, especially acute pyelonephritis/perinephric abscess.

patient to consider the options at leisure and make a note of questions. The possible complications and their incidence should be provided. These would include the following: ◆ Mild to moderate back/loin pain which usually settles with simple analgesia such as paracetamol ◆ Visible haematuria (up to 5%) which usually clears spontaneously within 24 hours ◆ Local bleeding always occurs but is usually minor and self-limiting; more severe bleeding requiring transfusion (~ 1%) ◆ Bleeding requiring angiographic intervention (≤ 0.5%). See Appendix 18.1 for a standard patient information leaflet which is used in the author’s unit. Biopsies are generally performed under direct ultrasound control using a disposable Tru-Cut® needle or biopsy gun. Biopsy under direct vision has a higher success and lower complication rate than using ultrasound merely for localization. Choice of needle gauge (G) is a matter of personal preference; it is customary to use a 16-G needle for native and 18-G for transplant biopsies. These generally provide adequate sample size with a trend to lower bleeding rates in these single kidneys which may require multiple biopsies. Either kidney may be biopsied, but the left kidney is usually more convenient as it is usually lower so access to the pole is easier. The patient lies prone on one or two firm pillows and the operator first determines the optimum entry site with ultrasound and in which phase of respiration the biopsy will best be performed, aiming for the lower pole through the cortex. The skin is sterilized, sterile gel applied, and the probe covered with a sterile sheath. Local anaesthetic is administered and then, using a spinal (21-G) needle, more local anaesthetic is delivered down to the capsule of the kidney. The biopsy needle is then advanced to the renal capsule under ultrasound visualization, the patient stops breathing as the device is fired and immediately removed, after which the patient can breathe normally. Generally two passes are made to obtain two cores of tissue. Patients unable to lie prone may be biopsied in the sitting position as for a pleural drain. Transplant biopsies are usually more straightforward because the kidney is superficial and does not move with respiration. However, transplant kidneys sometimes lie in unusual orientations and more care needs to be taken to avoid bowel (especially with intraperitoneal kidneys). It is also useful to use colour Doppler to identify large blood vessels, such as the iliac artery and vein, which may lie very close to the lower pole. Patients are kept on bed rest for 4 hours with pulse and blood pressure monitoring and the urine tested for the presence of blood. Renal biopsy is usually performed as a day-case procedure. Patients are discharged home approximately 6 hours after the procedure provided there are no signs of bleeding and the patient will not be alone for the first night. Patients with macroscopic haematuria are kept under observation until this settles (and usually overnight). Very obese patients and those in whom ultrasound localization is difficult may be biopsied using computed tomography (CT) guidance. Other techniques have been described including open surgical and laparoscopic biopsy and, in patients with uncorrectable clotting abnormalities, biopsy is performed via the internal jugular vein but bleeding complications can still occur and the samples are small and often difficult to interpret.

chapter 18 

Complications of renal biopsy Reported complication rates vary widely depending on the era and techniques used. Needle sizes vary from 14 to 18 G and may be automatic or manual with imaging varying from fluoroscopy to ultrasound used for localization or in real time, making comparison unreliable. The best estimates are from a systematic review and meta-analysis, which analysed 34 studies including nearly 9500 adult native kidney biopsies using an automatic needle (14–18 G) under real-time ultrasound (Corapi et al., 2012) and a similar sized review from the Norwegian Kidney Biopsy Register (Tondel et al., 2012) (Table 18.1). As might be expected, apart from needle size, significant haemorrhage is more likely in patients with bleeding abnormalities, poorer renal function (especially acute kidney injury) and higher blood pressures, emphasizing the importance of ensuring that the benefits of biopsy outweigh the risks and that risks are minimized as much as possible. Pain: most patients experience some discomfort in the loin once the local anaesthetic wears off and simple analgesia with paracetamol is usually sufficient. More severe pain might develop with a large haematoma or renal colic or bladder pain might result from significant macroscopic haematuria especially with clots. Haemorrhage: approximately 25% of the cardiac output flows to the kidneys and biopsy inevitably results in some bleeding. Usually the haematoma results in tamponade preventing a large bleed. Bleeding usually settles spontaneously. About 0.5% of patients require a blood transfusion with 16- or 18-G needles (2% with 14-G needles). Macroscopic haematuria with clots should be managed with an irrigation catheter. If the bleeding does not settle, the patient should be referred for angiography with embolization. This rarely fails but surgery and, at the last resort, nephrectomy are very occasionally necessary (Table 18.1). Arteriovenous fistula: arteriovenous fistulae have been reported in up to 18% (in very old literature) but are probably relatively common if looked for in the early post-biopsy period (e.g. with Doppler (Hubsch et al., 1990; Merkus et al., 1993)) but most of these settle

spontaneously within days or weeks. Occasionally a fistula increases in size causing macroscopic haematuria (often with clots) or renal impairment. These require radiological embolization. Other complications: other organs may be inadvertently biopsied, including bowel, liver, spleen, and gall bladder. These are uncommon with the use of ultrasound and an experienced operator. Occasionally a pneumothorax or haemopneumothorax may develop if the pleural cavity breached. Occasionally intracapsular bleeding, which is very painful, has been reported to cause tamponade and persistent renin–angiotensin-driven hypertension (the ‘Page kidney’) (McCune et al., 1991).

Laboratory handling of the renal tissue obtained by biopsy Once the decision to perform a biopsy has been made, it is essential that laboratory and diagnostic procedures are optimized to maximize the information obtained from the biopsy. For the diagnosis of native renal diseases, biopsy and laboratory protocols aim to produce an adequate core of paraffin-embedded renal tissue for light microscopy, snap-frozen renal cortex for immunostaining, and cortex containing at least one glomerulus fixed for electron microscopy (EM) if required. Ideally, the specimen should be examined under a dissecting microscope to identify renal cortex which can be divided while fresh. If this is not possible, for example, when the renal unit and laboratory are in different hospitals, the specimen may be transported in suitable fixatives for light microscopy (LM) and EM, and buffer or a transport medium if frozen tissue for immunostaining is required.

Best practice guidelines in renal pathology Guidelines for the handling and reporting of renal biopsies have been produced in the United Kingdom by the Association of Clinical Pathologists (Furness, 2000)  and The Royal College of Pathologists, and more recently by the Renal Pathology Society

Table 18.1  Incidence of complications of native renal biopsy in adults Complication

Incidence, % (N = 9474) (95% confidence interval) Meta-analysis, Corapi et al. (2012)

Incidence, % (N = 9288) Norwegian Registry (Tondel et al., 2012)

Macroscopic haematuria

3.5 (2.2–5.1)

1.9

Transfusion

0.9 (0.4–1.5)

0.9

Angiographic intervention

0.6 (0.4–0.8)

0.2a

Nephrectomy

0.01

NR

Deathb

0.02

0

Haematomac

85c

NR

Arteriovenous fistulad

NR

a Angiographic intervention and surgery recorded together; nephrectomy, if any, not separately reported. b Two deaths from 8971 procedures reported in a meta-analysis. c Haematoma incidence is unhelpful since small haematomas are almost inevitable and not reported in

meta-analysis or the Norwegian study but 85% in one prospective CT study (Rosenbaum et al., 1978). d Arteriovenous fistula not reported in meta-analysis or the Norwegian study (see text).

NR = not reported.

the renal biopsy

145

146

Section 1  

assessment of the patient with renal disease

(Chang et  al., 2012). The use of histological data in patient management is also included in several of the KDIGO guidelines (Kidney Disease:  Improving Global Outcomes (KDIGO) Glomerulonephritis Work Group, 2012). There is considerable variation between laboratories in methodology used, reflected by an audit of practice in the United Kingdom (Roberts et  al., 2001). Whilst this may in part be a result of differences in expertise and resources, it also reflects a lack of evidence base for many recommendations.

What is an adequate biopsy? An adequate biopsy is one that is sufficient for its purpose, whether that is to make a diagnosis or to provide information to allow disease classification or determination of disease activity, chronicity, and prognosis. How much tissue is required for the accurate diagnosis depends on the disease. For example, a single glomerulus may be sufficient for the diagnosis of glomerular diseases, such as IgA nephropathy or membranous glomerulonephritis, in which the lesions or immune deposits are diffuse, whilst the detection of more focal lesions, in conditions such as primary FSGS, require rather more tissue. Histological classifications of renal diseases frequently include a statement regarding how much tissue is required for application of the classification. This is usually expressed as number of glomeruli, and in many cases was decided arbitrarily. For example, the Banff schema of renal allograft pathology (Racusen et al., 1999; Solez et al., 2008) defines an adequate biopsy as containing at least ten glomeruli and two arterial cross sections. When this was formally tested, it was found that a Banff ‘adequate’ biopsy with two arteries would miss almost a half of vascular rejections, due to sampling artefact (McCarthy and Roberts, 2002). In contrast, the Oxford classification of IgA nephropathy provides an evidence-based recommendation for biopsy adequacy; a minimum of eight glomeruli should be present for glomerular scoring, a value that was determined by analysis of the 265 biopsies on which this classification was based (Roberts et al., 2009).

Stains for light microscopy Commonly used stains for renal biopsies are haematoxylin and eosin (H&E), periodic acid–Schiff (PAS), Jones’ methenamine silver, trichrome (such as Masson’s trichrome (MT), or elastic van Gieson (EVG)), and Congo red. H&E is used for overall evaluation of the biopsy and is particularly valuable in the assessment of cellular infiltrates. PAS detects polysaccharides and glycoproteins. It is used for the demonstration of basement membranes, the brush border of the proximal tubular epithelium, hyaline subendothelial deposits, and assessing mesangial cellularity. Jones’ methenamine silver stain provides detailed visualization of the glomerular basement membrane, revealing abnormalities such as membrane ‘spikes’ in membranous glomerulonephritis. Trichrome stains produce three colours, distinguishing matrix, cytoplasm, and nuclei. They are of value in assessing the amount of fibrosis; mature collagen stains blue with MT and red with EVG. EVG additionally highlights the elastica of arteries, staining it black. This is of value in the assessment of chronic vascular pathology. Congo red is used for the demonstration of amyloid deposits, which it stains pink, producing anomalous colours under polarized light (Howie et al., 2008).

The diagnostic process What is a diagnosis? A potential source of confusion in nephrology is that renal diagnoses are based on different types of information. As a result, multiple terms may be used to describe one condition. Diagnostic labels are based variously on clinical presentation, immunology, histology, and pathogenesis. For example, an ANCA-positive vasculitis (immunological) presents as a rapidly progressive glomerulonephritis (RPGN, clinical) with a pauci-immune necrotizing glomerulonephritis (histological) on renal biopsy. It is important to appreciate the overlap of these different diagnostic labels; in this example, ANCA-positive vasculitis is just one cause of RPGN and a necrotizing glomerulonephritis. The kidney has a limited number of ways to respond to injury and a single morphology may be seen in a number of very different conditions. For example, a nodular glomerulosclerosis may be seen in diabetic nephropathy, light chain deposition disease and in idiopathic nodular sclerosis. In chronic kidney disease, whatever the primary insult, the final common pathway is renal fibrosis with glomerulosclerosis and tubular atrophy. At this stage, histological changes are frequently non-specific and the challenge for the pathologist is to identify clues to the underlying disease process. Equally, one condition can produce diverse morphologies. This is particularly true for lupus nephritis and IgA nephropathy, which can result in almost any pattern of glomerular disease. Many histological ‘diagnoses’ are morphological patterns rather than diagnoses. This is best illustrated by glomerular diseases; morphological labels such as acute proliferative glomerulonephritis (GN), membranoproliferative GN, and FSGS each have a number of potential underlying causes. For glomerulonephritides, there is a link between the target of injury, the morphology, and clinical manifestation. Those conditions that selectively damage the permeability barrier (such as membranous nephropathy and MCD) produce the nephrotic syndrome, with little or no glomerular inflammation or proliferation. Necrotizing glomerular injury with rupture of capillary walls (as in vasculitic glomerulonephritis and anti-GBM disease) results in haematuria and a marked reduction in glomerular filtration rate, producing a RPGN. The exudation of fibrin and cytokines results in reactive extracapillary proliferation, producing a crescentic morphology. Those conditions (such as lupus nephritis and IgA nephropathy) associated with mesangial or subendothelial immune deposits produce glomerular inflammation with the nephritic syndrome and a mesangial or endocapillary proliferative morphology. This latter group show the most variation in morphology and clinical manifestations; when there is associated damage to the permeability barrier there may be heavy proteinuria and the nephrotic syndrome, and if there is necrosis they may present with a RPGN.

Morphological definitions and illustrations The first stage of histological diagnosis is to describe the morphological pattern of disease; to identify the type of lesion and its distribution. To be of clinical value, descriptions must be precise. Vague terminology, such as ‘proliferative glomerulonephritis’ and ‘focal sclerosis’, is to be avoided. Terms such as focal and diffuse, or mild, moderate, and severe, lack precision and reproducibility; lesions should be quantified wherever possible. For example, when a glomerular lesion is described, there should be an indication of the

chapter 18 

the renal biopsy

The terms focal and diffuse are used to describe the proportion of glomeruli involved, whereas segmental and global refer to the extent of involvement within individual glomeruli. Unfortunately, terminology is not standardized and different definitions are used for focal versus diffuse, and segmental versus global, according to whether sclerosis or proliferation is being described. When describing sclerosis, a focal lesion is one involving some but not all glomeruli and a segmental lesion part of but not all of a tuft. When describing proliferation, a focal lesion is one involving < 50% of glomeruli (rather than < 100%), and a diffuse lesion is one involving most glomeruli, > 50% (rather than all). Similarly, a segmental proliferative lesion is one involving < 50% of a glomerular tuft and a global lesion one involving > 50% of the glomerulus.

hypercellularity:  an increased number of cells within glomerular capillary lumina. These may be endothelial cells or intravascular leucocytes. ◆ Extracapillary proliferation: hypercellularity/proliferation within Bowman’s space, producing more than two cell layers between the capillary tufts and Bowman’s capsule. This appearance is commonly referred to as a cellular crescent which is usually a result of necrosis with exudation of fibrin and cytokines. A collapsing glomerulopathy may be associated with extracapillary proliferation, producing a similar appearance to a cellular crescent. ◆ Necrosis: disruption of the glomerular basement membrane (best appreciated on a silver-stained section) with fibrin exudation and karyorrhexis. The latter may not be evident and the minimum requirement for the definition of a necrotizing lesion is extracapillary fibrin exudation. ◆ Membranoproliferative pattern (also termed mesangiocapillary pattern):  mesangial hypercellularity with thickening of capillary walls. This produces a lobular appearance to the glomerulus. Capillary wall thickening is due to duplication of the glomerular basement membrane, as a reaction to interposed cells and subendothelial or intramembranous deposits.

Proliferation

Sclerosis

percentage of glomeruli in the biopsy showing this lesion. Equally, tubular atrophy should be quantified according to the percentage of cortex involved. In this section we provide a glossary of terms and illustrate common lesions.

Glomerular lesions The normal glomerulus is illustrated in Fig. 18.1.

Distribution of lesions

Proliferation is used to describe an increase in glomerular cells that may result from infiltrating leucocytes or proliferation of endogenous glomerular cells. For this reason, the term hypercellularity is sometime more appropriate than proliferation. Hypercellularity/proliferation is subclassified according to the part of the glomerulus involved. The site of proliferation gives an indication of the underlying cause and is of prognostic and therapeutic importance. The term ‘proliferative glomerulonephritis’ is not recommended. Whilst it is frequently used to describe endocapillary proliferation, it lacks precision and is a potential source of misunderstanding. The various proliferative lesions are illustrated in Fig. 18.2 and defined as follows: ◆ Mesangial hypercellularity: more than three mesangial cells in a peripheral mesangial area in a standard 2–3-micron thick paraffin section. The central stalk of the tuft should not be used for assessing cellularity. Thicker sections give an artefactual impression of hypercellularity.

(A)

◆ Endocapillary

Sclerosis is an increase in extracellular matrix within the glomerulus. In segmental and global glomerulosclerosis, the excess matrix is associated with obliteration of capillary lumina. Globally sclerosed glomeruli that are expanded and solidified by matrix are seen in advanced diabetic glomerulopathy or in a chronic glomerulonephritis. In mesangial sclerosis and nodular glomerulosclerosis, capillaries are patent. An adhesion is continuity with matrix material between glomerular basement membrane and Bowman’s capsule that is separate from an area of sclerosis, that is, capillaries associated with an adhesion are patent. The various types of segmental sclerosing lesions are illustrated in Fig. 18.3. Glomerular obsolescence is collapse of the glomerular tuft with fibrosis in Bowman’s space, and is typical of ischaemic injury. Hyalinosis is the accumulation of non-matrix proteins, that is, insudation of plasma proteins, between the endothelium and the glomerular basement membrane. The hyaline has an amorphous

(B)

Fig. 18.1  The normal glomerulus (A = H&E; B = methenamine silver stain), showing vascular pole and juxtaglomerular apparatus at the bottom. H&E illustrates the normal mesangial cellularity with ≤ 3 mesangial cell nuclei per segment. Silver stain demonstrates the normal glomerular basement membrane that appears smooth and delicate. Note the absence of circular capillary profiles; these are only seen when the capillary wall is abnormally thickened and rigid, as in membranous nephropathy.

147

148

Section 1  

assessment of the patient with renal disease (A)

(B)

(C)

(D)

(E)

(F)

Fig. 18.2  Proliferative glomerular lesions. (a) (PAS) IgA nephropathy showing mesangial hypercellularity in which there are ≥ 4 mesangial cells in a peripheral mesangial area. (b) (H&E) Post-infectious glomerulonephritis showing endocapillary hypercellularity, in which the capillary lumina are filled with infiltrating leucocytes. (c) (H&E silver) Anti-GBM disease showing extracapillary proliferation or cellular crescent, in which there is partial tuft collapse and proliferation of cells within Bowman’s space. (d) (H&E silver) ANCA-associated vasculitis, showing necrosis with capillary wall rupture and fibrin exudation. (e, f) (H&E and silver) Membranoproliferative pattern, showing a lobular appearance of the glomerular tuft with mesangial hypercellularity and thickened capillary walls, with GBM duplication evident on the silver stain.

eosinophilic appearance that is best appreciated in PAS or silver/ H&E-stained sections.

Tubulointerstitial lesions Normal tubules are illustrated in Fig. 18.4. In the normal renal cortex tubules are back to back without significant intervening space. The interstitium comprises peritubular capillaries and a delicate network of myofibroblasts that are not appreciated without immunohistochemistry (IH).

Atrophy and fibrosis Tubular atrophy: atrophic proximal tubules have thick irregular basement membranes and decreased diameter. Atrophic

distal tubules are frequently increased in diameter and contain Tamm–Horsfall protein casts. Sheets of tubules with this appearance are sometimes described as ‘thyroidization’ due to their resemblance to thyroid follicles. Interstitial fibrosis is an increase in extracellular matrix that separates the tubules. There is pronounced proliferation of myofibroblasts in early fibrosis that precedes matrix production. The distribution of interstitial fibrosis and tubular atrophy gives an indication of its aetiology and is illustrated in Fig. 18.4. Sharply delineated segmental atrophy/fibrosis is seen in reflux nephropathy or following healing of cortical infarcts. A multifocal distribution is typically seen in chronic glomerulonephritis or small vessel disease. Diffuse peritubular fibrosis follows inflammatory tubular

chapter 18 

(A)

(B)

(C)

(D)

(E)

(F)

the renal biopsy

Fig. 18.3  Sclerosing glomerular lesions. (A) (H&E silver) Glomerular tip lesion, showing adhesion and hyalinosis at the tubular pole. This lesion reflects severe proteinuria and may be seen in a number of nephrotic conditions, including primary FSGS, membranous nephropathy and diabetic nephropathy. Patients with tip lesions and otherwise normal glomeruli by light microscopy have a clinical course similar to minimal change disease. (B) (H&E silver) Collapsing lesion, showing tuft collapse with proliferation and swelling of visceral epithelial cells, an indicator of severe podocyte injury. (C) (PAS) Perihilar segmental sclerosis, typical of secondary FSGS, associated with hyperfiltration injury. (D) (H&E) Segmental sclerosis, not otherwise specified, typical of primary nephrotic FSGS. (E) (H&E) A broad-based bland segmental scar, typical of sclerosis following segmental necrosis in vasculitic glomerulonephritis. (F) (H&E) Nodular glomerulosclerosis in diabetic nephropathy, showing patent capillary loops around the mesangial nodules of matrix.

injury, such as a tubulointerstitial nephritis (TIN) in native kidneys or tubulointerstitial rejection in transplants.

Inflammation The significance of interstitial inflammation depends upon the nature of the cells, their distribution and the presence of tubulitis (infiltration of tubules by the inflammatory cells). Examples are illustrated in Fig. 18.5. Tubular atrophy and interstitial fibrosis are usually accompanied by an infiltrate of small lymphocytes, frequently with lymphoid aggregates, whatever the aetiology of the

chronic damage; such infiltrates are not a marker of TIN, but may still be contributing to the fibrotic process. Features of an active TIN are a mixed infiltrate, frequently including plasma cells and eosinophils, around non-atrophic tubules with an associated lymphocytic tubulitis. The type of inflammation may give an indication to the aetiology. Plasma cell-rich infiltrates are typically seen in TIN associated with systemic autoimmune diseases such as Sjögren syndrome and IgG4-associated disease. In the latter, there are numerous IgG4-positive plasma cells and an expansile fibroproliferative interstitial reaction that may result in renal mass lesions on

149

150

Section 1  

assessment of the patient with renal disease (A)

(B)

(C)

(D)

(E)

(F)

Fig. 18.4  Tubular atrophy and interstitial fibrosis. (a, b) Normal renal tubules on H&E/silver and PAS stains. The proximal tubules have abundant eosinophilic cytoplasm. PAS stain demonstrates the brush border on the luminal surface. The tubular basement membranes are closely apposed with very little interstitium evident. (c, d) (H&E and PAS) Focal tubular atrophy, typically seen in chronic glomerular disease. Atrophic proximal tubules are shrunken and have thick basement membranes, highlighted on PAS stain. (e) (H&E silver) Chronic calcineurin inhibitor toxicity showing a striped pattern of atrophy and fibrosis, reflecting the small vessel disease. (f) (H&E) Diffuse peritubular fibrosis in chronic tubulointerstitial nephritis.

imaging. Eosinophils are frequently prominent in drug-associated TIN, and there may also be granulomata. Multiple, often confluent discrete epithelioid granulomata are typical of renal sarcoidosis. Occasional intratubular neutrophils are common in TIN, and do not necessarily indicate the presence of an infective nephritis. Features of acute infective nephritis/pyelonephritis are interstitial oedema with a peritubular infiltrate of neutrophils, a neutrophilic tubulitis and plugs of neutrophils within tubular lumina (see below).

Tubular casts Examples of tubular casts and crystals are illustrated in Fig. 18.6. Red blood cell casts are a marker of glomerular haematuria, unless there has been renal trauma. However, rare such casts may also be

seen in acute TIN. The red cells are frequently dysmorphic and show lysis. Large numbers of red blood cell casts may result in acute kidney injury, even in the absence of severe necrotizing glomerular lesions. Protein casts: the most frequent protein casts are formed largely by Tamm–Horsfall protein (uromodulin), produced in the thick ascending limb of the loop of Henle. These are seen in atrophic distal tubules or in association with filtered proteins in proteinuric states. These casts have a uniform hyaline appearance and, as uromodulin is a glycoprotein, are PAS positive. Other frequently seen casts include (a) light chain casts in myeloma/light chain cast nephropathy, which are PAS negative and have a hard crystalline, fractured appearance with an associated inflammatory reaction;

chapter 18 

(A)

(B)

(C)

(D)

(E)

(F)

the renal biopsy

Fig. 18.5  Tubulointerstitial inflammation. (a) (H&E) A non-specific infiltrate of small lymphocytes associated with an area of scarring. (b) (H&E) Tubulointerstitial nephritis, showing a mixed interstitial infiltrate of lymphocytes, plasma cells, eosinophils, and neutrophils. (c) (H&E) Renal sarcoidosis, characterized by non-necrotizing epithelioid granulomata. (d) (H&E) A severe neutrophilic tubulitis with neutrophil casts, indicative of ascending bacterial infection. (e, f) IgG4-associated tubulointerstitial nephritis showing a plasma cell-rich interstitial infiltrate. IH (F) demonstrates numerous IgG4-positive plasma cells.

and (b) myoglobin or haemoglobin casts associated with rhabdomyolysis and intravascular haemolysis respectively. These typically have a granular or beaded appearance. Greenish pigmented bile casts with associated ATI may be present in patients with severe liver disease. Crystals: the most frequently seen intratubular crystals are calcium oxalate and calcium phosphate. Calcium oxalate crystals are often sheave-shaped and do not take up standard histochemical stains and are best visualized under polarized light. When extensive, they are indicative of hyperoxalaemia and hyperoxaluria, but are seen in small numbers following ATI. Tubular deposits of calcium phosphate are strongly haematoxyphilic and do not polarize. They are seen in hypercalcaemic states or with large loads of

ingested phosphate, and less frequently following severe ATI, when intraluminal cell debris acts as a nidus for calcification.

Acute tubular injury This may be ischaemic or toxic in aetiology (Fig. 18.7). In ischaemic ATI, there is tubular dilatation and simplification of the epithelium without tubular basement membrane thickening. In toxic injury, there may be cytopathic changes within tubular epithelium, such as cytoplasmic swelling and vacuolation. In severe injury, degenerated epithelial cells are shed into the lumen.

Viral inclusions Certain viral infections, such as polyoma viruses, cytomegalic virus (CMV), and adenovirus, produce characteristic inclusions within

151

152

Section 1  

assessment of the patient with renal disease (A)

(B)

(C)

(D)

(E)

(F)

Fig. 18.6  Tubular casts and crystals. (A) (PAS) PAS-positive Tamm–Horsfall protein (uromodulin) casts in distal tubules. (B) (PAS) Light chain cast nephropathy, characterized by PAS-negative fractured casts with associated inflammation and tubular epithelial injury. (C) (H&E) Granular eosinophilic haemoglobin casts in a patient with AKI associated with intravascular haemolysis. (D) (H&E) Red cell casts, indicative of glomerular haematuria, in a patient with vasculitic glomerulonephritis. (E) (H&E) Calcium phosphate deposits within tubules in a patient with hypercalcaemia and nephrocalcinosis. (F) (H&E polarized light) Calcium oxalate crystals in a patient with AKI associated with enteric hyperoxaluria.

tubular epithelial nuclei. These may be associated with evidence of epithelial injury and an inflammatory reaction (viral nephritis). Special immunostains can identify the specific virus.

Vascular lesions A normal intravenal supply is illustrated in Fig. 18.8. Arcuate arteries at the corticomedullary junction give off radial branches, the interlobular arteries that extend towards the renal capsule. From these branch afferent arterioles, which supply blood to the glomeruli. Arterial and arteriolar lesions may be acute or chronic; the latter may reflect vascular remodelling following earlier acute injury.

Thrombotic microangiopathy describes the morphology of acute microvascular injury, also termed malignant vascular injury (Fig. 18.8). Endothelial activation and injury result in a subendothelial exudate that frequently contains fibrin and red blood cells. There may be luminal thrombosis, particularly of arterioles and glomerular capillaries. The chronic phase is characterized by a proliferative response within the intima (‘onion skin proliferation’) and progressive intimal fibrosis. Arteriosclerosis describes remodelling of the arterial intima, largely comprising elastic fibres, also termed fibroelastosis (Fig. 18.8). This is a ‘normal’ age-related change, accelerated in association with long-standing hypertension.

chapter 18 

(A)

(B)

(C)

(D)

(E)

(F)

the renal biopsy

Fig. 18.7  Acute tubular injury (H&E). (A) Ischaemic acute tubular injury showing widespread tubular dilatation and epithelial simplification. (B) Cytotoxic acute tubular injury, with marked epithelial swelling and degenerative changes, in this case secondary to gentamicin toxicity. (C) Isometric tubular epithelial vacuolation in a renal transplant patient with Tacrolimus toxicity. (D–F) Viral infections in renal transplant patients. Polyoma virus infection (D) with basophilic nuclear inclusions. Adenovirus infection (E) characteristically produces marked tubular necrosis. CMV infection (F) with eosinophilic nuclear inclusions.

Arteritis is characterized by focal necrotizing vascular lesions. A  necrotizing glomerulonephritis, frequently with crescents, is the most frequent lesion in renal involvement by systemic small vessel vasculitis (granulomatosis with polyangiitis, microscopic polyangiitis, eosinophilic granulomatosis with polyangiitis (formerly known as Churg–Strauss vasculitis)), with arterial involvement seen in approximately 20% of cases. Less common is renal involvement in polyarteritis nodosa that involves the larger arcuate and segmental arteries, producing segmental necrosis and aneurysms.

Quantification of histological lesions The renal biopsy provides important information regarding disease activity and chronicity and may therefore be used to guide therapy. Indicators of active disease are necrosis, inflammation, and proliferation, lesions that are potentially responsive to immunosuppressive therapy. Chronic injury is characterized by glomerulosclerosis, interstitial fibrosis, tubular atrophy, and arteriosclerosis, lesions that have been regarded as largely irreversible. It is important for the biopsy report to include measures of the extent of acute and chronic lesions.

153

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

Fig. 18.8  Vascular lesions. (A) (H&E) A normal interlobular artery. (B–D) Arterial TMA. In the acute phase (B), endothelial injury results in marked intimal oedema. This progresses to a healing phase, initially with a proliferative intimal reaction (C) and finally obliterative intimal fibrosis (D, trichrome). (E) (elastic van Gieson) Arteriosclerosis (fibroelastosis), associated with essential hypertension, showing reduplication of the internal elastic lamina. (F) (H&E) A necrotizing arteritis in a patient with ANCA-positive vasculitis. (G) (H&E) Normal afferent and efferent arterioles at the glomerular hilum. (H) (H&E) Marked arteriolar hyalinosis in a patient with diabetic nephropathy.

chapter 18 

Glomerular lesions The total number of glomeruli, together with the number showing each of the glomerular lesions described above, are routinely included in the biopsy report. The percentage of glomeruli showing necrosis, proliferation, crescents, and segmental or global sclerosis has been shown to be of prognostic value in many diseases. For some conditions, this quantitative data forms the basis of a classification, the details of which are disease specific. Widely used schemas include the ISN/RPS classification of lupus nephritis (Weening et  al., 2004)  and the Oxford classification of IgA nephropathy (Cattran et al., 2009).

Tubulointerstitial lesions The extent of tubular atrophy/interstitial fibrosis should be assessed semiquantitatively and is most commonly expressed as a percentage of the cortical area present, to the nearest 10%. Various methods of measuring interstitial fibrosis and tubular atrophy are used (Farris and Colvin, 2012). Tubular atrophy is best visualized on PAS- or silver-stained sections which highlights the thickened basement membranes of atrophic tubules. Trichrome stains demonstrate established interstitial fibrosis but do not reveal the earlier stages characterized by proliferation of interstitial myofibroblasts and production of an immature matrix. Use of quantitative image analysis methods to assess fibrosis may help to improve reproducibility (Farris et al., 2011). The reproducibility of percentage interstitial fibrosis and tubular atrophy depends in part on the pattern of chronic tubulointerstitial damage. For conditions, such as chronic glomerulonephritis, that are associated with a multifocal pattern of atrophy and fibrosis, reproducibility is high, despite observers using different approaches to measurement. In contrast, reproducibility is low for conditions in which there is diffuse cortical involvement such as in chronic damage associated with TIN. Tubular atrophy is a process that develops over time, and the poor reproducibility of the diffuse pattern of chronic tubular injury reflects the difficulty in defining at what point a tubule should be regarded as atrophic. It is also recommended that the extent of interstitial inflammation is assessed. Inflammation is of different significance depending on whether it involves fibrotic or non-fibrotic cortex. In renal transplant biopsies, mononuclear inflammatory cell infiltration in non-fibrotic cortex is a criterion used in the diagnosis of T-cell-mediated rejection, although it has been shown to be poorly reproducible, particularly in the presence of chronic damage (Furness and Taub, 2001).

Arterial lesions It is recommended that lesions of arteries are scored based on the most severe lesions. As there are usually few arteries in a renal biopsy, assessment of arterial lesions is subject to the greatest sampling artefact. Intimal thickening is scored by comparing the thickness of the intima to that of the media in the same segment of vessel, thus allowing for variation in thickness resulting from oblique sectioning of the vessel. The intima is scored variously as normal, and thickened to more or less than the thickness of the media.

Immunohistology Immunohistology is routinely used to demonstrate deposits of immunoglobulins, light chains, and complement. In renal

the renal biopsy

transplantation it is valuable in the diagnosis of viral infections such as polyoma virus and CMV and for detection of complement C4d as a marker of antibody-mediated rejection. Other uses include the typing of leucocyte infiltrates, classification of amyloid deposits, identification of myoglobin casts, and characterization of abnormal matrix proteins. Immunofluorescence microscopy (IF) on frozen sections and IH on paraffin sections may both be used for the demonstration of Ig and complement deposits. IF has the advantage of greater sensitivity and it enables quantitative histology; unlike IH, the level of fluorescence signal correlates closely with the amount of antigen present. Demonstration of linear IgG deposits in anti-GBM disease, and light chain restriction in light chain deposition disease, are frequently only possible using IF on frozen sections. IH has the advantage that it is performed on the same paraffin block as is used for routine light microscopical stains and may be used when there is no frozen core of tissue available for IF. The greater detail of tissue structure in paraffin sections enables more precise localization of deposits using IH, and unlike IF, IH sections are permanent. IH does suffer from greater laboratory artefacts, such as false positivity resulting from plasma proteins in the tissue sections, and false negative stains resulting from masking of epitopes by formalin fixation. Various methods of specimen washing and antigen retrieval are described to minimize these artefacts. Glomerular deposits seen on immunohistology are described by their location (mesangial vs capillary wall; segmental vs global; focal vs diffuse) and appearance (linear, finely granular, coarse granular, pseudolinear). Tubular and vascular deposits are described when present. Clonal cast staining is described. Examples are illustrated in Fig. 18.9.

Electron microscopy EM is valuable, particularly in the diagnosis of glomerular diseases and recurrent disease post transplantation. EM blocks should be prepared for all native renal biopsies. In some centres, EM reports are issued for all cases, whilst in others the pathologist selects which to examine ultrastructurally on the basis of the LM and IF findings. EM is essential for diagnosis in approximately one-quarter of native renal biopsies (Pearson et al., 1994). The ultrastructure of the normal glomerulus is illustrated in Fig. 18.10. EM provides precise localization of glomerular deposits, and reveals internal structure to deposits. It is essential for the diagnosis of fibrillary and immunotactoid glomerulonephritis, conditions that are defined by their ultrastructural appearance. Mesangial, subendothelial, subepithelial and intramembranous deposits are illustrated in Fig. 18.10. In addition to demonstration of deposits, EM is used to quantify the extent of podocyte foot process effacement in proteinuric conditions. Abnormalities of the GBM usually require EM for diagnosis. Whilst localized thinning of the lamina densa may be seen at sites of injury, diffuse thinning is indicative of thin membrane nephropathy or early stage or a carrier state of Alport syndrome. In established renal involvement in Alport syndrome, there is multilayering of the thin lamina densa, producing a basket-weave appearance. Amorphous thickening of basement membranes is a feature of diabetic glomerulopathy. EM plays a central diagnostic role in the diagnosis of certain storage

155

156

Section 1  

assessment of the patient with renal disease (A)

(B)

(C)

(D)

(E)

(F)

Fig. 18.9  Immunofluorescence. (a) Linear GBM positivity for IgG in anti-GBM disease. (b) Granular capillary wall positivity for IgG in membranous nephropathy. (c) Mesangial positivity for IgA in IgA nephropathy. (d) Mesangial and capillary wall positivity for C3 in C3 glomerulonephritis. (e) Mesangial and tubular basement membrane positivity for kappa light chains in light chain deposition disease. (f) Positivity for lambda light chains in tubular casts in light chain cast nephropathy.

diseases, including Fabry disease and lecithin cholesterol acyltransferase deficiency.

Diagnostic algorithms In Table 18.2, the characteristic features of some selected glomerular diseases are summarized. In the final section, we present a series of diagnostic algorithms for glomerular diseases (Fig. 18.11), based

on selected appearances at LM. These illustrate how morphology is refined to a diagnosis using IF, EM, and clinical data. The algorithms are not intended to be comprehensive and do not include some of the rarer causes of these patterns of disease. In the first algorithm, in which glomeruli are normal by LM, the range of diagnoses considered is dependent upon the clinical features and urinary findings, emphasizing the importance of clinicopathological correlation in the diagnostic process.

(A)

(B)

(C)

(D)

(E)

(F)

(G)

(H)

Fig. 18.10  Electron microscopy. (A) The normal capillary wall, comprising lumen at the bottom, lined by a fenestrated endothelium. The outer aspect of the basement membrane is covered by podocyte foot processes. The podocyte cell body is at the top. (B) Minimal change disease, showing diffuse effacement of podocyte foot processes. (C) Paramesangial deposits in IgA nephropathy. (D) Subepithelial deposits with membrane spikes in membranous nephropathy. (E) Subendothelial deposits in C3 glomerulonephritis. (F) Intramembranous electron dense deposits in dense deposit disease. (G) Mesangial deposits of IgG with a fibrillary internal structure in fibrillary glomerulonephritis. (H) Lipid inclusions within podocytes and endothelial cells in a patient with Fabry disease.

Table 18.2  Summary of characteristic abnormalities of selected glomerular diseases Disease and typical clinical presentation

LM pattern

IF

EM, other findings

Mesangia

Subepithelial

Subendothelial

+

+ (wire loop)

Mesangial and capillary wall deposits Reticular aggregates in endothelial cells

−

Mesangial ± subendothelial deposits

Haematuria/nephritis Lupus ISN/RPS class IV (diffuse proliferative)

Endocapillary proliferation in > 50% of glomeruli, wire loops

+

IgA nephropathy

Mesangial proliferation

+ IgA dominant

−

Post-infectious GN

Endocapillary proliferation intracapillary neutrophils

−

+ Coarse chunky C3 dominant

Haemolytic-uraemic syndrome

Arteriolar/glomerular thrombosis Endothelial swelling

−

−

−

Endothelial injury Expansion of lamina rara interna

Minimal change disease

Normal

−

−

−

Effacement of podocyte foot processes

Focal segmental glomerulosclerosis

Segmental glomerulosclerosis

± IgM, C3

−

−

Effacement of podocyte foot processes, no deposits

Diabetic nephropathy

Increased mesangial matrix, ± nodules, thick GBM, hyalinized arterioles

−

−

−

Thick GBM without deposits

Membranous GN

Spikes on Jones’ silver stain

−

+ granular IgG, C3

−

Subepithelial deposits

Amyloid

Pink, acellular material in glomeruli, Congo Red +

−

−

−

Fibrils

All Igs, C3, C4

Irregular, hump-like deposits on top of GBM

Nephrotic syndrome

RPGN Anti-GBM disease

Global or segmental necrosis of linear staining of GBM glomeruli, crescents IgG, C3

Granulomatosis with polyangiitis

Focal segmental necrosis of glomeruli, crescents

−

−

−

No deposits

Microscopic polyangiitis Focal segmental necrosis of glomeruli, crescents

−

−

−

No deposits

EM = electron microscopy; IM = immunofluorescence microscopy; LM = light microscopy.

No deposits

(A)

Normal Glomerulus by LM

Isolated Haematuria IgA

Nephrotic proteinuria Negative IMF

IgAN

Thin GBM

Normal GBM

Thin GBM Alport early/carrier

Non-renal haematuria

(B)

Negative IMF

IgG + C3

MCD or unsampled FSGS

MN Stage I

Endocapillary proliferation

IgA

C3 dominant

IgAN

IgG + C3

Full House IgG,A,M,C3,C1q

Non-organised deposits

Fibrillary deposits

Infection/other ICGN

Fibrillary GN

Subepithelial humps, no DD

Subendothelial deposits

Postinfectious GN

C3 GN

Monoclonal Ig

Organised deposits Lupus nephritis

Proliferative GN with monoclonal deposits

Linear IgG

Dense deposit disease

Anti-GBM disease

Glomerular crescents

Negative IMF

Pauci-immune vasculitic GN

Cryoglobulinaemic GN

Intramembranous linear dense deposits

(D)

(C)

IgM + IgG

Double contours of GBM Positive IMF

Granular disease specific IMF

Various immune complex GN

Negative IMF

Subendothelial deposits

Subepithelial/ intramembranous deposits

Widening of LRI, no deposits

Various diseases, see endocapillary proliferation

Chronic MN

Chronic TMA

Fig. 18.11  Algorithms for analysis of renal biopsies. (A.) Normal glomeruli by light microscopy in a patient with renal disease (B.) Glomeruli show endocapillary proliferation (C.) Glomeruli show crescents (See also chapter 70) (D.) Glomerular basement membrane shows double contours (see also chapter 80). Key: Light microscopy (LM) Pink; Immunofluorescence Green; Urinalysis Yellow; Diagnoses shown in Blue. Abbreviations C3: Third component of Complement. DD: Dense deposits. GBM: glomerular basement membrane. ICGN: Glomerulonephritis characterised by immune complexes. GN: Glomerulonephritis. Ig: Immunoglobulin (hence IgA for Immunoglobulin A, IgG etc). IgAN: IgA nephropathy. IMF: immunofluorescence. LM: light microscopy. LRI: Lamina rara interna. MCD: Minimal change disease. MN: Membranous nephropathy (= membranous glomerulonephritis). TMA: Thrombotic microangiopathy.

160

Section 1  

assessment of the patient with renal disease

References Adler, A. I., Stevens, R. J., Manley, S. E., et al. (2003). Development and progression of nephropathy in type 2 diabetes: the United Kingdom Prospective Diabetes Study (UKPDS 64). Kidney Int, 63(1), 225–32. Alwall, N. (1952). Aspiration biopsy of the kidney, including a report of a case of amyloidosis diagnosed through aspiration biopsy of the kidney in 1944 and investigated at an autopsy in 1950. Acta Med Scand, 143, 430–5. Cameron, J. S. and Hicks, J. (1997). The introduction of renal biopsy into nephrology from 1901 to 1961: a paradigm of the forming of nephrology by technology. Am J Nephrol, 17, 347–58. Cattran, D. C., Coppo, R., Cook, H. T., et al. (2009). The Oxford classification of IgA nephropathy: rationale, clinicopathological correlations, and classification. Kidney Int, 76, 534–45. Chang, A., Gibson, I. W., Cohen, A. H., et al. (2012). A position paper on standardizing the nonneoplastic kidney biopsy report. Clin J Am Soc Nephrol, 7, 1365–8. Corapi, K. M., Chen, J. L., Balk, E. M., et al. (2012). Bleeding complications of native kidney biopsy: a systematic review and meta-analysis. Am J Kid Dis, 60(1), 62–73. Davis, C. L. and Chandler, W. L. (1995). Thromboelastography for the prediction of bleeding after transplant renal biopsy. J Am Soc Nephrol, 6(4), 1250–5. Farris, A. B., Adams, C. D., Brousaides, N., et al. (2011). Morphometric and visual evaluation of fibrosis in renal biopsies. J Am Soc Nephrol, 22, 176–86. Farris, A. B. and Colvin, R. B. (2012). Renal interstitial fibrosis: mechanisms and evaluation. Curr Opin Nephrol Hypertens, 21, 289–300. Furness, P. N., Taub, N., and Convergence of European Renal Transplant Pathology Assessment Procedures (CERTPAP) Project (2001). International variation in the interpretation of renal transplant biopsies: report of the CERTPAP Project. Kidney Int, 60, 1998–2012. Furness, P. N. (2000). ACP. Best practice no 160. Renal biopsy specimens. J Clin Pathol, 53, 433–8. Hlatky, M. A. (1982). Is renal biopsy necessary in adults with nephrotic syndrome. Lancet, 2(8310), 1264–8. Howie, A. J., Brewer, D. B., Howell, D., et al. (2008). Physical basis of colors seen in Congo red-stained amyloid in polarized light. Lab Invest, 88, 232–42. Hubsch, P. J., Mostbeck, G., Barton, P. P., et al. (1990). Evaluation of arteriovenous fistulas and pseudoaneurysms in renal allografts following percutaneous needle biopsy. Color-coded Doppler sonography versus duplex Doppler sonography. J Ultra Med, 9(2), 95–100. International Study of Kidney Disease in Children (1978). Nephrotic syndrome in children: prediction of histopathology from clinical and laboratory characteristics at time of diagnosis. A report of the International Study of Kidney Disease in Children. Kidney Int, 13(2), 159–65. Iversen, P. and Brun, C. (1951). Aspiration biopsy of the kidney. Am J Med, 11, 324–30. Kidney Disease: Improving Global Outcomes (KDIGO) Glomerulonephritis Work Group (2012). KDIGO Clinical Practice Guideline for Glomerulonephritis. Kidney Int Suppl, 2, 1–274.

Levey, A. S., Lau, J., Pauker, S. G., et al. (1987). Idiopathic nephrotic syndrome. Puncturing the biopsy myth. Ann Internal Med, 107(5), 697–713. Manno, C., Strippoli, G. F., Arnesano, L., et al. (2004). Predictors of bleeding complications in percutaneous ultrasound-guided renal biopsy. Kidney Int, 66(4), 1570–7. McCarthy, G. P. and Roberts, I. S. (2002). Diagnosis of acute renal allograft rejection: evaluation of the Banff 97 guidelines for slide preparation. Transplantation, 73, 1518–21. McCune, T. R., Stone, W. J., and Breyer, J. A. (1991). Page kidney: case report and review of the literature. Am J Kidney Dis, 18(5), 593–9. Merkus, J. W., Zeebregts, C. J., Hoitsma, A. J., et al. (1993). High incidence of arteriovenous fistula after biopsy of kidney allografts. Brit J Surg, 80(3), 310–12. Muehrcke, R. C., Kark, R. M., and Pirani, C. L. (1955). Technique of percutaneous renal biopsy in the prone position. J Urol, 74(3), 267–77. Pearson, J. M., McWilliam, L. J., Coyne, J. D., et al. (1994). Value of electron microscopy in diagnosis of renal disease. J Clin Pathol, 47, 126–8. Peterson, P., Hayes, T. E., Arkin, C. F., et al. (1998). The preoperative bleeding time test lacks clinical benefit: College of American Pathologists’ and American Society of Clinical Pathologists’ position article. Arch Surg, 133(2), 134–9. Pham, T. T., Sim, J. J., Kujubu, D. A., et al. (2007). Prevalence of nondiabetic renal disease in diabetic patients. Am J Nephrol, 27(3), 322–8. Racusen, L. C., Solez, K., Colvin, R.B., et al. (1999). The Banff ‘97 working classification of renal allograft pathology. Kidney Int, 55, 713. Roberts, I. S., Cook, H. T., Troyanov, S., et al. (2009). The Oxford classification of IgA nephropathy: pathology definitions, correlations, and reproducibility. Kidney Int, 76, 546–56. Roberts, I. S. D. and Davies, D. R. (2001). Handling of renal biopsies: different approaches reflect a lack of evidence for what constitutes ‘best practice’. J Clin Pathol, 54, 413–14. Rosenbaum, R., Hoffsten, P. E., Stanley, R. J., et al. (1978). Use of computerized tomography to diagnose complications of percutaneous renal biopsy. Kidney Int, 14(1), 87–92. Solez, K., Colvin, R. B., Racusen, L. C., et al. (2008). Banff 07 classification of renal allograft pathology: updates and future directions. Am J Transplant, 8, 753–60. Stratta, P., Canavese, C., Marengo, M., et al. (2007). Risk management of renal biopsy: 1387 cases over 30 years in a single centre. Eur J Clin Invest, 37(12), 954–63. The Royal College of Pathologists (n.d.). Datasets and Tissue Pathways. [Online] www.rcpath.org/publications-media/publications/datasets/ datasets-TP.htm Tondel, C., Vikse, B. E., Bostad, L., et al. (2012). Safety and complications of percutaneous kidney biopsies in 715 children and 8573 adults in Norway 1988-2010. Clin J Am Soc Nephrol, 7(10), 1591–7. Topham, P. S., Harper, S. J., Furness, P. N., et al. (1994). Glomerular disease as a cause of isolated microscopic haematuria. Q J Med, 87(6), 329–35. Weening, J. J., D’Agati, V. D., Schwartz, M. M., et al. (2004). The classification of glomerulonephritis in systemic lupus erythematosus revisited. Kidney Int, 65, 521–30.

CHAPTER 19

Clinical trials: why and how in nephrology Richard Haynes, Martin J. Landray, William G. Herrington, and Colin Baigent Introduction Randomized trials are an indispensable tool for nephrologists seeking to improve outcomes for their patients. Randomized trials are the best method for identifying (and quantifying) the benefits and risks of interventions in clinical practice, and the only method that can (when properly conducted) eliminate bias (Collins and MacMahon, 2001). The incremental improvements in survival seen in cardiology and cancer medicine are due in part to the widespread acceptance and conduct of large randomized trials. Nephrology lags far behind most specialities in medicine in its evidence base (Strippoli et al., 2004) to the detriment of patients under the care of nephrologists and also the larger number of patients with more moderate kidney disease who are cared for by doctors from other specialities. Furthermore, most trials in nephrology have been too small to provide reliable answers and therefore to change clinical practice. Many commonly used treatments currently recommended in nephrology guidelines have never been tested in an adequately sized randomized trial (e.g. phosphate binders, immunosuppressants, and dialysis techniques) and therefore their effects are not known. In some cases, they could be doing substantial harm. This lack of a robust evidence base will become more important in the coming decades as the global population ages and the epidemic of type 2 diabetes mellitus matures, with a consequent rise in the numbers of patients with kidney disease. In order to provide the best possible care for our current and future patients, there is an urgent need to design, fund, and conduct large randomized trials capable of addressing important clinical questions in nephrology.

Why do we need large randomized trials in nephrology? Moderate effects are the best that can be expected Some treatments have large, and hence obvious, effects on survival: for example, it was clear without the need for any randomized trials that prompt treatment of diabetic coma or ventricular fibrillation can save lives. However, in part because of these striking successes but also because of large effects of treatments on intermediate disease markers (such as antihypertensives on blood pressure), the hopes of large treatment effects on major clinical outcomes (such

as end-stage renal disease or mortality) in nephrology, have been unrealistically high. In general, where substantial uncertainty remains about the efficacy of a treatment on major clinical outcomes in nephrology, its effects on such outcomes are probably either negligibly small, or only moderate (i.e. a 15–20% proportional reduction in risk), rather than large (Box 19.1). This may be because a variety of pathophysiological mechanisms are responsible for progression of chronic kidney disease (CKD) or its complications (e.g. cardiovascular disease), only one of which may be appreciably influenced by any one particular therapy. The existence of (at best) moderate effects of most treatments is strongly suggested by meta-analyses of various therapies in nephrology (Jafar et al., 2001; Webster et al., 2004) and by the typically modest strength of epidemiological relationships between known causal risk factors that are modifiable by treatment (e.g. blood pressure) and outcomes that are common in CKD (such as heart failure). In nephrology, failure to acknowledge that treatment effects are likely to be, at best, moderate has led to randomized trials of promising treatments that have almost always been too small: many apparently negative trials may have missed potentially worthwhile benefits (e.g. EVOLVE, which tested cinacalcet in haemodialysis patients with secondary hyperparathyroidism (Chertow et al., 2012)) whilst on the other hand, exaggerated claims of efficacy have proved to be unrealistically large and have not changed clinical practice (e.g. trials of antioxidants in haemodialysis patients (Boaz et al., 2000; Tepel et al., 2003)).

What types of studies are required to assess promising treatments in nephrology? If moderate differences in outcome resulting from promising treatments are to be detected reliably, then any errors in studies of such treatments need to be much smaller than this. This requirement necessitates a study design that both excludes biases and minimizes random error. Randomization eliminates bias by ensuring that each type of patient can be expected (but for the play of chance) to have been allocated in similar proportions to different treatment strategies. This means that only the treatment effect and random differences should affect the final comparisons of outcome. The way to guarantee very small random differences is to design studies that include large numbers of relevant events that are potentially preventable by the treatment under consideration.

162

Section 1  

assessment of the patient with renal disease

Box 19.1  Reasons why the effects of treatment on clinical outcomes are usually moderate (but still worthwhile) ◆ Clinical outcomes have multiple causes: diseases such as atherosclerosis or progressive kidney disease have multiple causal risk factors. On the other hand, treatments are usually specific to one particular cause, so can only mitigate the risk attributable to that cause, but not the others. Therefore, claims of large treatment effects are usually not plausible. ◆ Large treatment effects on intermediate outcomes do not usually translate into large effects on clinical outcomes:  certain drugs do have large effects on measurements such as blood pressure, proteinuria, or acute rejection rates. Although these effects may be large, for the reasons given above, they are unlikely to translate into large effects on clinically important outcomes such as development of end-stage renal disease, transplant failure, or death. The easy availability of measuring such outcomes and their association with clinical outcomes often leads to the false assertion that treatments which reduce such intermediate outcomes will improve clinical outcomes. ◆ There are also some reasons specific to nephrology that may limit the magnitude of expected treatment effects: for example, the dose of drug used may be limited by either risk of nephrotoxicity or systemic toxicity due to reduced renal clearance, and patients with kidney disease are prescribed many medications which may limit compliance to study treatment. It is often claimed that by collecting enough information about various prognostic features, it is possible to make statistical adjustments to correct for any such differences between the types of patients who, in an observational (i,e. non-randomized study), receive the different treatments that are to be compared. Such methods, which are often carried out on routinely collected healthcare data (such as dialysis registries, for example), aim to achieve comparability between those entering the different treatment groups, but they cannot be guaranteed to do so, and they often fail seriously (Deeks et al., 2003). The difficulty is that some important prognostic factors may be unrecorded, while others may be difficult to assess exactly and hence difficult to adjust for reliably. Although there are examples of non-randomized studies in which the estimated effects of treatment appear quantitatively close to those observed in analogous randomized trials, there are many examples where they are not, being either quantitatively incorrect—so that drugs appear either misleadingly promising or of misleadingly low efficacy—or even qualitatively incorrect, when a harmful drug might appear effective (or vice versa) (Grady et al., 1992; Rossouw et al., 2002). This residual bias means such analyses should not be used to guide treatment decisions, however large the dataset. Random errors are minimized by ensuring that there are sufficient patients recruited to yield the number of ‘endpoints’ that is required for reliable statistical inference. However, it is not sufficiently widely appreciated just how large clinical trials need to be in order to detect moderate differences reliably. This can be illustrated by the recent Study of Heart and Renal Protection (SHARP): the 17% relative reduction in major atherosclerotic events (non-fatal myocardial infarction, coronary death, ischaemic stroke, or revascularization procedure) was statistically significant among around

9000 patients, but if this trial had only been half as large (say, 4500 patients) then it would have had about a one in three chance of missing this clinically worthwhile treatment benefit (at a type 1 error rate of 5%) (Baigent et al., 2011). When trials are not large enough to detect such treatment effects, their results can be very misleading. First, since it is mathematically impossible for small trials to be statistically significant unless they yield large treatment estimates, small trials are useless unless the drug under study is a miracle cure (in which case a randomized trial would not be necessary to identify its effect). In all other cases, such trials will not help to distinguish between effective (hence useful) treatments and those that are useless or even harmful. It is common for positive results from small single-centre trials not to be replicated when larger multicentre trials are completed. In nephrology, for example, two small trials of renal replacement therapy (RRT) for acute kidney injury appeared to show a reduction in mortality in those allocated to more intensive RRT (Ronco et al., 2000; Schiffl et al., 2002), but in subsequent multicentre trials, no such difference was found (Palevsky et al., 2008; Bellomo et al., 2009). Small trials can even be misleading about the direction of treatment effect. For example, a small trial of the inotrope vesnarinone suggested that it halved the risk of death in patients with heart failure (13 vesnarinone versus 33 placebo deaths, P = 0.002) (Feldman et al., 1993). However, when the same regimen was tested among a larger population of similar patients, mortality was in fact increased (292 vesnarinone versus 242 placebo deaths, P = 0.02) (Cohn et al., 1998). Very large trials may also have the benefit of being able to address efficacy and safety in a wide range of different types of patients, in whom treatment effects may differ.

How do we achieve large, successful randomized trials in nephrology? Assuming that a new treatment is moderately effective among patients with CKD, four main requirements need to be satisfied by a randomized trial if it is to be able to establish this fact reliably (Table 19.1): 1. Randomization of a large sample of patients (to produce a large number of endpoints) 2. Maintenance of compliance with the randomized treatment allocation (to maintain study power) 3. Ascertainment of relevant study outcomes 4. Appropriate statistical analysis.

Randomization of a large sample of patients In order to recruit large numbers of patients into trials, it is necessary to streamline trial procedures to ensure wide eligibility criteria and efficient recruitment.

Trial eligibility: using the uncertainty principle Randomization can be offered only if both doctor and patient feel substantially uncertain as to which of the trial options is best. The question then arises:  ‘Which categories of patients about whose treatment there is such uncertainty should be offered randomization?’ The answer is all of them, welcoming the heterogeneity that this will produce. (For example, either the treatment of choice will turn out to be the same for men and women, in which case the

chapter 19 

clinical trials: why and how in nephrology

Table 19.1  Main requirements for successful large clinical trials and methods to achieve them Requirement

Study design

Study conduct

Effective recruitment

Consider a pilot study Use uncertainty principle to determine eligibility Avoid unnecessary exclusion criteria Heterogeneity is a strength in a randomized controlled trial population

Pre-screen Simple consent process Detailed baseline phenotyping (including biological samples)

Good compliance

Aim for as large a difference in relevant and measurable intermediate outcome (such as biomarkers) as possible Use pre-randomization run-in period

Use procedures that maintain investigator and participant compliance Leave patient management as much as possible to managing physician

Reliable ascertainment of outcomes

Make trial long enough to accrue sufficient efficacy and safety information

Use simple case report forms

Employ on-site monitoring only where necessary

Minimize number of study visits and biological samples

Central statistical monitoring

Use routine collected data where possible to supplement study data

Review by unblinded Data Monitoring Committee

Use electronic data capture methods that check and validate data entry and point of entry

Select a clinically important outcome

Analyse by the principle of intention to treat

Ensure outcome is modifiable by study treatment

Consider appropriate methods to analyse subgroups (tests of heterogeneity and trend with allowance for multiplicity of testing)

Relevant outcomes and appropriate analysis

trial might as well include both, or it will be different, in which case it is particularly important to study both.) This approach of randomizing the full range of patients in whom there is substantial uncertainty as to which treatment option is best was used in the ASTRAL trial of renal arterial revascularization for renal artery stenosis (Fig. 19.1) (Wheatley et al., 2009): ◆ If the doctor or the patient were reasonably certain, for any reason, that they did wish to revascularize the affected kidney(s) in that particular patient, the patient was not eligible for entry into ASTRAL. ◆ Conversely, if the doctor or the patient were reasonably certain, for any reason, that they did not wish to revascularize the affected kidney(s) in that particular patient, the patient was likewise not eligible for entry into the trial. ◆ If, but only if, the doctor and patient were substantially uncertain what to recommend, the patient was eligible for randomization between the renal revascularization (by angioplasty with or without stenting) versus standard medical care alone. In ASTRAL, there were substantial differences between individual nephrologists in the types of patients about whom they were uncertain. For example, most (but not all) were convinced that patients with severe bilateral renal artery stenosis should be revascularized, but there were differing views about the precise threshold of stenosis above which a patient should be offered such treatment and whether it should be used among those with stable kidney function. The use of the uncertainty principle enabled the trial to yield at least some direct evidence in a broad range of patients. As a result of the wide and simple entry criteria adopted by ASTRAL, 806 patients were randomized (which is almost four times more than all previous trials of renal revascularization combined), no patient received a treatment which they or their doctor believed to be contraindicated, and the study was able to provide some evidence about the

effects of renal revascularization with which to plan the treatment of future patients. Other trials in nephrology have also taught us the importance of embracing uncertainty. During the course of the placebo-controlled TREAT trial of erythropoietin, the investigators were initially criticized for choosing a placebo comparator, but as other data emerged during the trial they were then criticized for the haemoglobin target selected in the active arm. Eventually, the TREAT trial provided reliable data to show that a policy of increasing haemoglobin from 9 g/dL to about 13 g/dL does not reduce cardiovascular disease in patients with diabetes and CKD and may in fact cause harm (Pfeffer et al., 2009).

Efficient recruitment The chief focus of a large-scale trial designed to detect effects on major clinical outcomes should be to address its main hypothesis. In order to maximize the probability of a successful trial, it may be prudent to limit expenditure on measurements or substudies which are not strictly necessary. Such expenditure takes resources away from the primary purpose of the trial, and may be a poor investment if the trial fails to answer its primary question due to insufficient size. When designing a new study, it is wise to consider every proposed visit, measurement and assay critically. It is often possible to reduce costs without loss of scientific information. In SHARP, for example, it was not considered necessary to collect blood and urine for analysis in the central laboratory at every 6-month study visit, and it was possible to estimate changes in blood lipid profile among a random 10% subsample of patients at 1 year and 4 years after randomization (SHARP Collaborative Group, 2010). Moreover, knowledge of the results of some blood assays whilst the trial is in progress may not be necessary, and can be done more cheaply in batches at a later date in stored frozen samples.

163

164

Section 1  

assessment of the patient with renal disease Renal artery stenosis detected

Should patient be offered immediate revascularization?

Doctor(s) or patient reasonably certain that revascularization is not appropriate: patient is ineligible

Doctor(s) or patient reasonably certain that revascularization is appropriate: patient is ineligible

Doctor(s) and patient substantially uncertain whether to revascularize: uncertainty implies eligibility

Group 1: allocated to medical therapy only

Group 2: allocated to medical therapy plus revascularization

6% get revascularization (median delay 20 months)

83% get revascularization (median delay 1 month)

Fig. 19.1  Demonstration of uncertainty principle in recruitment. The ASTRAL trial investigated renal artery revascularization in renal artery stenosis (Wheatley et al., 2009). Because individual clinicians differed in the types of patients about whom they were uncertain, no category of patient about which there was widespread uncertainty was excluded, so the trial result is widely applicable.

Pilot studies can be helpful in testing and refining processes and determining what data are essential to collect (and conversely what are unnecessary) (Box 19.2). They are also better suited to providing more detailed assessments of the treatments on intermediate variables such as blood pressure or cholesterol. It is feasible to study such measures in detail in a small number of closely followed patients and provide both accurate and reliable results, whereas such close observation is neither feasible nor necessary in a larger trial assessing clinical outcomes. One important aspect of the trial that can be piloted is the identification, invitation and recruitment of trial participants; time spent testing and modifying these in a pilot study can yield substantial rewards in terms of more rapid recruitment into the subsequent larger trial. One method of enhancing recruitment that can be very time- and cost-effective is pre-screening. For many renal trials this can be achieved by identifying potentially eligible patients from renal unit databases while other aspects of the study are still being set-up (e.g. before full ethics or regulatory approvals). Potentially eligible patients are asked if they would consider entering a trial so that there is a pool of patients who can be rapidly invited when the study can formally begin at a site. Pre-screening also acts as an assessment of a site’s performance and of whether eligibility criteria need to be modified.

Box 19.2  Information that can be obtained from pilot studies ◆ Recruitment: ease and rates of recruitment; test methods (e.g. pre-screening). ◆ Trial materials: assess acceptability of patient information leaflets and consent forms. ◆ Tolerability of study treatment:  will inform likely compliance in larger study which will have a major impact on statistical power. ◆ Efficacy on intermediate outcomes:  whilst assessment of effects on an intermediate outcome (e.g. cholesterol) may help in estimating likely effects on clinical outcomes for the main trial, in the absence of a suitable (i.e. causally relevant) intermediate outcomes effect size estimates in pilot studies will generally be of little value (since they are subject to large random errors). ◆ Trial procedures: identify trial procedures that patients or trial clinic staff find hard to adhere to, allowing their modification (or, perhaps even their removal if they are not absolutely necessary).

chapter 19 

Maintenance of compliance A high incidence of participants randomly allocated to active treatment stopping that treatment (‘drop-outs’) or those allocated to placebo starting open-label active treatment (‘drop-ins’) quickly erodes a trial’s ability to test the study hypothesis (that is, in statistical terms, it reduces study power): as much effort should be put into ensuring trial participants remain on their study treatment as goes into recruiting them in the first place. In general, every one treatment drop-out or drop-in post-randomization has the equivalent effect to recruiting two fewer patients in the first place. Such considerations were a key part of the design of SHARP, which included a pre-randomization 6-week placebo run-in period. Such a period allows participants, investigators, and other doctors time to reconsider a particular patient’s participation, enables tests to check eligibility to be conducted, and is a test of compliance so that those participants who were unlikely, or unable, to take study treatment long-term can be excluded prior to randomization (Lang et al., 1991). Similarly, the exclusion of patients with prior myocardial infarction or coronary revascularization was intended to reduce the chances of ‘drop-in’ statin therapy among such patients. Compliance can also be improved by minimizing the impact of the study on the patients’ daily routine and clinical care. Trials should also have procedures to identify participants who discontinue study treatment so any modifiable reasons can be addressed.

Reliable ascertainment of relevant study outcomes The reliable collection of study outcomes is ensured by careful study design. To prevent incomplete outcome ascertainment introducing bias, all participants should also continue to attend study visits even if they have stopped treatment or had a study outcome. Where feasible, a study may be ‘double blind’ so that both investigators and participants do not modify their reporting of events (consciously or subconsciously) by knowledge of treatment allocation. Alternatively, open-label studies can minimize these biases, providing that study endpoints are not subjective and that information is collected and assessed systematically (e.g. using registry data), blinded to the randomized treatment allocation. In order to facilitate complete follow-up, study visits should aim to collect the minimum information required from patients. Collecting vast amounts of data just because it might prove interesting both increases costs and makes patients less likely to attend visits, both of which work against a successful outcome. Using electronic case report forms is also an efficient method to check completeness (and internal validity) of data during the study clinic visit, rather than through time-consuming (and costly) checks of data later. To ensure high-quality follow-up, trial site staff should receive adequate training and support, and key aspects of study performance should be monitored (Landray et al., 2012). Traditionally, ‘on-site monitoring’, where a member of the trial team visits a study site, has been considered to be a standard requirement. However, the detection of missing data (or fraud) is more effectively—and much more efficiently—achieved by review of centrally collected data (Pogue et  al., 2013). In addition, statistical techniques can often be used to identify sites which are ‘outliers’ in measures of performance (e.g. study visit duration and frequency of adverse event reporting) and therefore require further scrutiny (perhaps including a targeted on-site visit) (Valdes-Marquez et al., 2011). It

clinical trials: why and how in nephrology

is important not to waste large amounts of resources trying to identify every item of data. Small amounts of missing or incorrect data are likely to be randomly distributed between arms and therefore, provided the study is large enough, unlikely to reduce the reliability of the results. Ultimately, the most important monitoring during a trial is that performed by an independent data monitoring committee, which regularly reviews unblinded interim analyses of the study’s emerging findings. These analyses are a more informative means of reviewing the safety of the study treatment than individual reports of adverse events.

Appropriate choice of study outcomes and statistical analyses Appropriate choice of study outcomes In any trial, it is important to select a primary outcome that is both clinically important and sensitive to the study treatment. Most ‘surrogate’ outcomes (e.g. albuminuria) are not true surrogates (Prentice, 1989) and trials based on these can be misleading: for example, although dual renin–angiotensin system blockade reduces albuminuria more than monotherapy (Jennings et  al., 2007), no trials have yet identified improvements in clinical outcomes (but larger trials have shown harm) with this strategy (Parving et al., 2012; Fried et  al., 2013). When combining endpoints, outcomes should be of broadly similar clinical significance and likely to be affected in the same direction. For example, trials of antithrombotic therapy need to consider ischaemic and haemorrhagic stroke subtypes separately for net benefit in different types of participant to be accurately estimated, whereas a trial of blood pressure-lowering treatment might reasonably combine them (since the epidemiological associations of blood pressure and each of the stroke subtypes are qualitatively similar). Total mortality is one of the least sensitive endpoints and is rarely an appropriate outcome because an effective treatment is likely to yield a moderate benefit on just a few causes of death, and as a result is likely to have only a small effect on total mortality. Any effect is also very dependent on the patterns of mortality in the population tested and the result may not be generalizable to a different population.

Appropriate statistical analysis Even after proper randomization of a large number of patients, bias can be introduced by inappropriate statistical analysis. For example, excluding patients who do not adhere to their study treatment will introduce bias if the prognosis of those excluded from one treatment group differs from that of those excluded from another (which is highly likely as the side-effects of active treatment are often more pronounced in those that are sicker). The Coronary Drug Project provides an example of this: patients who took at least 80% of their allocated clofibrate had lower mortality after 5 years than those who did not (15.0% versus 24.6% respectively, P = 0.0001). However, the difference was even more striking between patients who did or did not take at least 80% of their allocated placebo (15.1% versus 28.3%, P < 0.00001) (The Coronary Drug Project Research Group, 1980). It is therefore essential that the primary analysis of any trial should compare the outcome among all those originally allocated one treatment (regardless of how long they took the treatment for or even if they ever did) with the outcome among all those allocated to the comparator group (i.e. according to the ‘intention-to-treat’ principle).

165

166

Section 1  

assessment of the patient with renal disease

When faced with an overall trial result, doctors frequently want to know for which group of patients is this treatment particularly helpful (or harmful). Looking at the treatment effect in the subgroup of interest in isolation is not appropriate; the correct question is whether the treatment effect observed in that subgroup differs significantly from the overall treatment effect. In other words, the overall treatment effect becomes the new ‘null’ hypothesis and the relevant test is whether the subgroup of interest differs from this (using tests for heterogeneity or trend to test this hypothesis). In the absence of evidence of heterogeneity, the best resolution is to emphasize the overall results of the trial (Collins and MacMahon, 2001). It is often asked whether patients who respond better to a given treatment (e.g. have a larger than average blood pressure reduction with an antihypertensive) gain more benefit in terms of clinical outcomes. Not infrequently, these subgroups are defined after randomization and this can introduce bias. Groups of patients defined by different post-randomization responses are very unlikely to differ only randomly from each other and it is highly possible that some of these other differences might be related to the risk of the clinical outcome of interest. Inferences drawn from such non-randomized comparisons of ‘responders’ versus ‘non-responders’ are biased and could be misleading. For example, the RENAAL study of losartan versus placebo in diabetic nephropathy showed that allocation to losartan reduced the risk of death, dialysis or doubling of creatinine by 16% (P = 0.02) (Brenner et al., 2001). A subsequent analysis of the study suggested that patients who had a larger reduction in albuminuria derived the most renoprotective benefit (Eijkelkamp et al., 2007). However, the groups of patients were defined by post-randomization changes in albuminuria and therefore the analyses are potentially biased. Such analyses can be performed if response to treatment is defined prior to randomization. For example, in the Heart Protection Study, all potential participants entered a run-in phase with during which they received simvastatin 40 mg. Measurement of blood lipids at the beginning and end of run-in allowed for a pre-randomization assessment of low-density lipoprotein-lowering ‘responsiveness’. Therefore, post-randomization outcomes could be compared between active and control therapies within each response group without affecting the balance of randomization (Heart Protection Study Collaborative Group, 2002).

Summary If clinical outcomes for patients with kidney disease are to improve it is essential that nephrologists embrace the concept of large randomized trials, the concept of being uncertain, and join collaborative studies of the key clinical questions. Both randomization and large sample sizes are vital if the trial results are to be a reliable test of the intervention and if the results are to change clinical practice. Large trials require careful design and conduct to ensure that doctors can be confident that the answers they give are relevant to their practice. A doctor treating an individual patient may feel that the large trial results deny the individuality of their patient. However, one of the main reasons trials need to be so large is precisely because patients are so different from one another. Only by randomizing large numbers can the effect of treatment be distinguished from natural variation in clinical course. Trial populations may not be representative of all the patients doctors meet in routine clinical practice, but

an unrepresentative trial population can still provide a generalizable trial result. It is very unlikely that the patients recruited into a large randomized trial are so different from ‘normal’ patients that the proportional effect of treatment would differ in some meaningful way. From such data, the effect of treatment on ‘real-world’ observed absolute risk can be predicted. Large randomized trials therefore remain the only way to reliably answer questions about the efficacy and safety of many treatments in nephrological practice and are key to improving clinical outcomes.

References Baigent, C., Landray, M. J., Reith, C., et al. (2011). The effects of lowering LDL cholesterol with simvastatin plus ezetimibe in patients with chronic kidney disease (Study of Heart and Renal Protection): a randomised placebo-controlled trial. Lancet, 377, 2181–92. Bellomo, R., Cass, A., Cole, L., et al. (2009). Intensity of continuous renal-replacement therapy in critically ill patients. N Engl J Med, 361, 1627–38. Boaz, M., Smetana, S., Weinstein, T., et al. (2000). Secondary prevention with antioxidants of cardiovascular disease in endstage renal disease (SPACE): randomised placebo-controlled trial. Lancet, 356, 1213–8. Brenner, B. M., Cooper, M. E., de Zeeuw, D., et al. (2001). Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy. N Engl J Med, 345, 861–9. Chertow, G. M., Block, G. A., Correa-Rotter, R., et al. (2012). Effect of cinacalcet on cardiovascular disease in patients undergoing dialysis. N Engl J Med, 367, 2482–94. Cohn, J. N., Goldstein, S. O., Greenberg, B. H., et al. (1998). A dose-dependent increase in mortality with vesnarinone among patients with severe heart failure. Vesnarinone Trial Investigators. N Engl J Med, 339, 1810–6. Collins, R. and MacMahon, S. (2001). Reliable assessment of the effects of treatment on mortality and major morbidity, I: clinical trials. Lancet, 357, 373–80. Deeks, J. J., Dinnes, J., D’amico, R., et al. (2003). Evaluating non-randomised intervention studies. Health Technol Assess, 7, iii–x, 1–173. Eijkelkamp, W. B., Zhang, Z., Remuzzi, G., et al. (2007). Albuminuria is a target for renoprotective therapy independent from blood pressure in patients with type 2 diabetic nephropathy: post hoc analysis from the Reduction of Endpoints in NIDDM with the Angiotensin II Antagonist Losartan (RENAAL) trial. J Am Soc Nephrol, 18, 1540–6. Feldman, A. M., Bristow, M. R., Parmley, W. W., et al. (1993). Effects of vesnarinone on morbidity and mortality in patients with heart failure. Vesnarinone Study Group. N Engl J Med, 329, 149–55. Fried, L. F., Emanuele, N., Zhang, J. H., et al. (2013). Combined angiotensin inhibition for the treatment of diabetic nephropathy. N Engl J Med, 369(20), 1892–903. Grady, D., Rubin, S. M., Petitti, D. B., et al. (1992). Hormone therapy to prevent disease and prolong life in postmenopausal women. Ann Intern Med, 117, 1016–37. Heart Protection Study Collaborative Group (2002). MRC/BHF Heart Protection Study of cholesterol lowering with simvastatin in 20,536 high-risk individuals: a randomised placebo-controlled trial. Lancet, 360, 7–22. Jafar, T. H., Schmid, C. H., Landa, M., et al. (2001). Angiotensin-converting enzyme inhibitors and progression of nondiabetic renal disease. A meta-analysis of patient-level data. Ann Intern Med, 135, 73–87. Jennings, D. L., Kalus, J. S., Coleman, C. I., et al. (2007). Combination therapy with an ACE inhibitor and an angiotensin receptor blocker for diabetic nephropathy: a meta-analysis. Diabet Med, 24, 486–93. Landray, M. J., Grandinetti, C., Kramer, J. M., et al. (2012). Clinical trials: rethinking how we ensure quality. Drug Information Journal, 46, 657–60. Lang, J. M., Buring, J. E., Rosner, B., et al. (1991). Estimating the effect of the run-in on the power of the Physicians’ Health Study. Stat Med, 10, 1585–93.

chapter 19 

Palevsky, P. M., Zhang, J. H., O’Connor, T. Z., et al. (2008). Intensity of renal support in critically ill patients with acute kidney injury. N Engl J Med, 359, 7–20. Parving, H. H., Brenner, B. M., McMurray, J. J., et al. (2012). Cardiorenal end points in a trial of aliskiren for type 2 diabetes. N Engl J Med, 367, 2204–13. Pfeffer, M. A., Burdmann, E. A., Chen, C.-Y., et al. (2009). A trial of darbepoetin alfa in type 2 diabetes and chronic kidney disease. N Engl J Med, 361(21), 2019-32. Pogue, J. M., Devereaux, P. J., Thorlund, K., et al. (2013). Central statistical monitoring: detecting fraud in clinical trials. Clin Trials, 10, 225–35. Prentice, R. L. (1989). Surrogate endpoints in clinical trials: definition and operational criteria. Stat Med, 8, 431–40. Ronco, C., Bellomo, R., Homel, P., et al. (2000). Effects of different doses in continuous veno-venous haemofiltration on outcomes of acute renal failure: a prospective randomised trial. Lancet, 356, 26–30. Rossouw, J. E., Anderson, G. L., Prentice, R. L., et al. (2002). Risks and benefits of estrogen plus progestin in healthy postmenopausal women: principal results from the Women’s Health Initiative randomized controlled trial. JAMA, 288, 321–33. Schiffl, H., Lang, S. M., and Fischer, R. (2002). Daily hemodialysis and the outcome of acute renal failure. N Engl J Med, 346, 305–10.

clinical trials: why and how in nephrology

SHARP Collaborative Group (2010). Study of Heart and Renal Protection (SHARP): randomized trial to assess the effects of lowering low-density lipoprotein cholesterol among 9,438 patients with chronic kidney disease. Am Heart J, 160, 785–794 e10. Strippoli, G. F., Craig, J. C., and Schena, F. P. (2004). The number, quality, and coverage of randomized controlled trials in nephrology. J Am Soc Nephrol, 15, 411–9. Tepel, M., Van Der Giet, M., Statz, M., et al. (2003). The antioxidant acetylcysteine reduces cardiovascular events in patients with end-stage renal failure: a randomized, controlled trial. Circulation, 107, 992–5. The Coronary Drug Project Research Group (1980). Influence of adherence to treatment and response of cholesterol on mortality in the coronary drug project. N Engl J Med, 303, 1038–41. Valdes-Marquez, E., Hopewell, J. C., Landray, M., et al. (2011). A key risk indicator approach to central statistical monitoring in multicentre clinical trials: method development in the context of an ongoing large-scale randomized trial. Trials, 12, A135. Webster, A. C., Playford, E. G., Higgins, G., et al. (2004). Interleukin 2 receptor antagonists for renal transplant recipients: a meta-analysis of randomized trials. Transplantation, 77, 166–76. Wheatley, K., Ives, N., Gray, R., et al. (2009). Revascularization versus medical therapy for renal-artery stenosis. N Engl J Med, 361, 1953–62.

167

SECTION 2

The Patient with fluid, electrolyte, and renal tubular disorders

20 An overview of tubular function  171 Matthew A. Bailey

21 Sodium transport and balance: a key role for the distal nephron  189 Laurent Schild

22 Water homeostasis  197 David Marples and Søren Nielsen

23 Potassium homeostasis  204 Alain Doucet and Gilles Crambert

24 Renal acid–base homeostasis  218 Carsten A. Wagner and Olivier Devuyst

25 Phosphate homeostasis  225 Heini Murer, Jürg Biber, and Carsten A. Wagner

26 Calcium homeostasis  231 Francesco Trepiccione and Giovambattista Capasso

27 Magnesium homeostasis  243 Pascal Houillier

28 Approach to the patient with hyponatraemia  249 Ewout J. Hoorn and Robert Zietse

29 Approach to the patient with hypernatraemia  261 Robert Zietse

30 Approach to the patient with oedema  272 Jean-Marie Krzesinski and Eric P. Cohen

31 Approach to the patient with salt-wasting tubulopathies  282 Detlef Bockenhauer and Robert Kleta

32 Approach to the patient with polyuria  291 Daniel G. Bichet

33 Clinical use of diuretics  299 David H. Ellison and Arohan Subramanya

34 Approach to the patient with hypo-/hyperkalaemia  323 Charles S. Wingo and I. David Weiner

35 Approach to the patient with metabolic acidosis or alkalosis  339 Mitchell L. Halperin and Kamel S. Kamel

36 Approach to the patient with renal tubular acidosis  363 Stephen B. Walsh

37 Approach to the patient with hypercalcaemia  372 Dennis Joseph and Theresa A. Guise

170

38 Approach to the patient with hypocalcaemia  378 Agnès Linglart and Anne-Sophie Lambert

39 Approach to the patient with hypo-/ hyperphosphataemia  384 Judith Blaine, Hector Giral, Sabina Jelen, and Moshe Levi

40 Approach to the patient with hypomagnesaemia  397 Martin Konrad and Karl P. Schlingmann

41 Approach to the patient with renal Fanconi syndrome, glycosuria, or aminoaciduria  412 Detlef Bockenhauer and Robert Kleta

CHAPTER 20

An overview of tubular function Matthew A. Bailey Introduction

The proximal tubule

Of the approximately 180 L of water that is filtered by the renal glomeruli per day, only 1–2 L is excreted. Similarly, the filtered sodium (Na+) load of approximately 25,000 mmoL/day is mostly reabsorbed with only approximately 150 mmol/day being excreted. Chloride (Cl−), bicarbonate (HCO3−), calcium (Ca2+), and magnesium (Mg+) have a proportionally similar excretory profile and although secretion is important for the homeostasis of some electrolytes (particularly potassium and acid–base balance), the metabolic activity of the kidney is chiefly concerned with reabsorption. Most of the transport processes are coupled to the reabsorption of sodium, either directly through specialized transport proteins, or indirectly due to transepithelial electrochemical gradients. The ‘pump-leak’ model is evident in most nephron segments: the heteromeric Na+/K+-ATPase extrudes three sodium ions across the basolateral membrane in exchange for two potassium ions, generating a low intracellular sodium concentration and a steep electrochemical gradient for passive entry across the apical membrane via a selection of specialized transport proteins (Fig. 20.1). This process accounts for most of the energy used by the kidney, which is derived chiefly through oxidative metabolism. Renal tissue, which accounts for approximately 0.5% of total body mass, consumes approximately 7% of whole-body oxygen. Renal blood flow is about 20% of the cardiac output and the renal cortex is maintained at the high partial pressure of oxygen required for aerobic metabolism. The proximal tubule is the powerhouse for reabsorptive processes. It reclaims all of the filtered glucose, amino acids, and small proteins, 80% of HCO3− and phosphate (PO4), and 70% of Na+, Cl−, potassium (K+), and water. The loop of Henle contributes a further 25% to overall reabsorption of Na+, Cl−, and K+ with significant amounts of Ca2+ and Mg+ also being reabsorbed, particularly in the thick ascending limb of Henle (TALH). The distal tubule and collecting duct can fine-tune overall urinary excretion in accordance with systemic balance requirements: Na+, Cl− Ca2+, and Mg2+ are all reabsorbed here, whereas the net movement of K+ is secretory. Water reabsorption in the distal nephron is variable, being influenced by vasopressin and thus hydration, and to a lesser extent, volume status. In the first part of this chapter, the major transport pathways and regulatory features for each nephron segment is described. The focus here is on the transepithelial flux of Na+,K+, and water. In the second part, other important aspects of renal homeostasis, including urine concentration and acid–base balance, are summarized. Throughout the chapter, the key transport proteins are given and, where different, the name of the encoding gene is also provided.

The main function of the proximal tubule is the bulk reabsorption of the glomerular filtrate. The majority of this occurs in the proximal convoluted tubule. Two structural features of the proximal tubule cells enhance the surface area for transepithelial flux. First, microvilli project into the tubule lumen from the apical membrane (the brush border membrane); and second, the basolateral membrane is extensively invaginated. Transport in the proximal tubule is driven by the Na+/K+-ATPase, expressed in the basolateral membrane. Hydrolysis of ATP by this transporter underpins the high metabolic requirements of the proximal tubule, which has mitochondria densely packed to the basolateral membrane. The proximal tubule lies mainly in the cortex and relies predominantly on aerobic metabolism. This contrasts with those segments in the medulla that are in a relatively hypoxic environment and therefore have capacity for anaerobic metabolism. The proximal tubule is prone to hypoxic insult and mitochondrial dysfunction, which may reflect intrinsic differences in the mitochondrial population, compared to the distal tubule (Hall et al., 2009). In the proximal tubule, the NHE3 isoform of Na+/H+ exchange is the main transporter responsible for Na+ entry into the cells but a whole battery of specialized transporters are also expressed in this membrane, which couple sodium entry to movement of other species (Fig. 20.2A). Despite the prominence of NHE3, mutations in the encoding gene (SLC9A3) have not yet been associated with human renal disease. This reflects the redundancy of renal sodium transport:  NHE3 knockout mice have increased distal sodium reabsorption (Bailey et al., 2004) due to activation of the renin–angiotensin–aldosterone system (RAAS); the mild hypotension and alkalosis reflect an absorptive defect in the colon. Aquaporin-1 (AQP1) water channels are constitutively expressed in both membranes, and contribute to the high water hydraulic permeability of the proximal segment. Deletion of AQP1 in mice diminishes substantially transcellular flux and this channel is required for near isosmolar reabsorption observed in the proximal tubule (Vallon et al., 2000). However, water transport in the proximal tubule as a whole remains robust and the low-resistance paracellular shunt provides a route, perhaps the major route, of transport for water across the proximal tubule. Experiments in rodents suggest that the tubular fluid is relatively hypotonic, the transepithelial osmotic gradient of < 10 mosmol/kg being sufficient to drive large amounts of water reabsorption (Green and Giebisch, 1989). It is possible, however, that reabsorption can occur in the absence of an osmolar gradient due to ‘electro-osmosis’ or current-induced fluid movement (Fischbarg, 2010).

172

Section 2  

fluid, electrolyte, and renal tubular disorders

Lumen

Interstitial fluid

Na+ Na+ Na+

K+ K+

X Na+ Y

Apical membrane

Basolateral membrane

Fig. 20.1  Prototypical renal tubule cell. The Na+/K+-ATPase or ‘sodium pump’ in the basolateral membrane maintains a low intracellular Na+ concentration, generating an electrochemical gradient for passive entry of Na+ across the apical membrane. The transport protein facilitating movement of Na+ across the apical membrane varies along the nephron. Cotransport (symport and antiport) systems use the kinetic energy of Na+ entry to move another species against its electrochemical gradient. A simple system—an ion channel for Na+—is found in the distal nephron.

The apical membrane of the proximal tubule also expresses a range of enzymes that change the composition of luminal fluid. Carbonic anhydrase type IV, for example, is vital to bicarbonate reabsorption (see later) and enzymes such as neutral endopeptidase (Walter et  al., 1997)  and ecto-5′-nucleotidase can influence the tubular concentration of paracrine agents (Shirley et al., 2009)  and thereby influence tubular function in downstream segments.

Transport processes The initial short segment of the proximal tubule (S1 segment) reclaims most of the filtered glucose, amino acids, and phosphate by cotransporters directly coupled to sodium entry into the cell down its electrochemical gradient (Fig. 20.2A). Glucose crosses the apical membrane via the low-affinity, high-capacity, sodium-dependent transporter SGLT2 (SLC5A2), and exits via GLUT2. The reabsorption of glucose (and amino acids) is electrogenic, generating a small, lumen negative, transepithelial potential difference (PD) (approximately −2mV) in the S1 proximal tubule. Although this should favour the paracellular reabsorption of Cl−, the reflection coefficient of the S1 PT for Cl− is 0.9 (a reflection coefficient of 1.0 indicates complete impermeability of the S1 epithelium to the solute in question). Cl− reabsorption therefore lags behind that of Na+ and water, causing the tubular fluid Cl− concentration to rise slightly. Thus, Na+ reabsorption in the S1 PT is preferentially ‘coupled’ to that of HCO3−. This process is indirect: sodium entry into the cell via the NHE3 exchanger causes secretion of a proton into the tubule lumen. An apical H+-ATPase also contributes to proton secretion. In the tubule fluid, protons combine with the filtered HCO3− to produce carbonic acid and ultimately carbon dioxide (CO2) and water. This is discussed in more detail in the later section on HCO3−. The remainder of the proximal convoluted tubule is the S2 segment. In the cortex, the proximal straight tubule also consists of S2 cells, with S3 cells being limited to the medulla. The preferential reabsorption in the S1 segment of NaHCO3 relative to NaCl creates a modest tubular-fluid to plasma concentration gradient for

Cl−. The gap junctions between S2 cells have a high permeability for Cl−, allowing paracellular reabsorption down its concentration gradient. Since electrogenic Na+ reabsorption in the S2 segment is now limited by the low tubular fluid concentration of glucose and amino acids, paracellular Cl− flux causes a reversal of the transepithelial PD (now +2 mV). This lumen positivity favours paracellular reabsorption of cations (Fig. 20.2B) but the majority of sodium transport is thought to be transcellular via apical NHE3. The gradient also favours anion secretion and there is substantial plasma to tubule fluid concentration gradient for HCO3−. However, this gradient is not dissipated due to low paracellular permeability for HCO3− backflux. Micropuncture evidence suggests that the S1 and S2 proximal tubule reabsorbs approximately 70% of the filtered Cl−. Not all of this occurs by paracellular diffusion as some transporters, coupled ultimately to the basolateral Na+/K+-ATPase, help to reabsorb Cl− by the transcellular route (Fig. 20.3). The apical entry path is either secondary or tertiary active transport involving members of Slc26 anion exchanger family (Sindic et al., 2007). These transporters are multifunctional, capable of transporting, in addition to Cl−, anions such as sulphate, iodide, formate, oxalate and HCO3−. CFEX (SLC26A6) has an important role here. Expressed on the apical membrane, CFEX is capable in expression systems of chloride/formate, chloride/oxalate and chloride/bicarbonate exchange (Aronson, 2006). Studies in Slc26a6 null mice indicate that the primary mode in vivo is to mediate oxalate-dependent Cl− reabsorption in the proximal tubule. The chloride exits via a basolateral K+Cl− cotransport mechanism (probably KCC1) or through Cl− channels. K+ reabsorption pathways in the proximal convoluted tubule are not completely resolved. Much of the potassium that enters the cell via the Na+/K+-ATPase is recycled across the basolateral membrane via K+ channels (channels of KCNK and KCNJ gene families have been identified) or K+Cl− cotransport. The isoforms KCC3 (SLC12A6) and KCC4 (SLC12A7) have been identified (Jentsch, 2005). These basolateral exit routes for K+ are crucial for maintaining efficient transepithelial Na+ flux, stabilizing intracellular K+ concentration (and cell volume) in the face of fluctuating rates of transepithelial Na+ transport (Hamilton and Devor, 2012). K+ channels have also been identified in the apical membrane but the electrochemical gradient would favour secretion of K+ into the tubular fluid. A small secretory component has been identified but the physiological purpose of this is unknown. There is some evidence for active transport mechanisms, particularly the non-gastric isoform of H+/K+-ATPase (ATP12A), but the contribution that this pathway makes under normal circumstances is probably minor (Malnic et al., 2013). K+ reabsorption lags behind that of Na+ in the S1 segment, causing the tubular fluid concentration to rise slightly. Combined with the lumen positive PD in the S2 segment, paracellular reabsorption by simple diffusion is sufficient to account for the majority of K+ reabsorption by the proximal tubule. A certain proportion of K+ reabsorption is mediated by paracellular solvent drag. Since the S2 segment has a high K+ permeability (reflection coefficient of < 0.4), significant amounts of this solute can be entrained by vectoral fluid flux. The proximal straight tubule, or pars recta, has similar mechanisms of transport for the major electrolytes as the PCT and reabsorption here is also effectively isosmolar. Notably, the proximal straight tubule expresses in addition to AQP1, AQP7 (Sohara et al.,

chapter 20 

an overview of tubular function

(A)

S1 Interstitial fluid

Lumen Na+ glucose Na+ amino acids Na+ NPT2a phosphate Na+ NHE3 H+ SGLT2

Na+ K+ K+ K+ Cl–

–2mV

H2O AQP1 AQP1 (B) Lumen

S2

SGLT1

Interstitial fluid

Cl–

+

Na+ NHE3

Na+

K+

H+ K+

Na+ K+

K+ Cl–

H2O AQP1 AQP1

Fig. 20.2  The major Na+ transport pathways in the proximal tubule. (A) In the S1 segment, Na+ transport is coupled to that of glucose, amino acids, and bicarbonate, generating a small lumen negative transepithelial potential difference. (B) In the S2 segment, this potential becomes lumen positive..

2009), which contributes both to transepithelial water flux and the generation of a concentrated medulla due to its permeability to glycerol. Studies in isolated tubule segments have identified a lower Na+/K+-ATPase activity per unit length and the capacity for Na+ transport is approximately 50% that of the PCT. The straight segment reclaims (‘mops up’) the remaining 10% of the filtered glucose load with the high-affinity, low-capacity Na+-dependent cotransporter, SGLT1 (SLC5A1), providing entry across the apical membrane; GLUT1 facilitates basolateral exit. A major difference in the proximal straight tubule is that net K+ flux is secretory, which may reflect the diminution (or reversal) of the concentration gradient for paracellular K+ diffusion coupled to transcellular secretion via apical K+ channels. The PST is also an important site for the secretion of organic acids and bases, via multiple members of the SLC22A gene family (Fig. 20.4). For organic acids, the first step of the secretory process is basolateral anion exchange, in which

the exit of dicarboxylate (chiefly α-ketoglutarate) down a concentration gradient is coupled to entry of organic acids (Rizwan and Burckhardt, 2007). The main transporters are OAT1 and OAT3. Apical efflux utilizes a range of transporters, including OAT4 (humans only) and URAT-1, and is not rate limiting. For bases, the overall pathway is similar, with the basolateral entry step being rate limiting and mediated by the selective exchangers, OCT1 and OCT2 (Jonker and Schinkel, 2004). Clinically important organic acids (e.g. non-steroidal anti-inflammatory drugs, diuretics, penicillin) and bases (e.g. amiloride, cimetidine and atropine) are substrates for this secretory system, which is often, therefore, defined from a pharmacological viewpoint. Indeed, the prototypic substrate for organic acid secretion is p-aminohippurate, this being the central tenet for its clearance being used as an index of effective renal plasma flow. Creatinine is secreted via the organic base pathway, which can lead

173

174

Section 2  

fluid, electrolyte, and renal tubular disorders

S2

Lumen

NHE3 H+

Na+

Na+

H+

K+ K+ Cl–

HCOO–

HCOO– CFEX

Interstitial fluid

K+ Cl–

Cl–

Fig. 20.3  Tertiary active transport facilitating transepithelial Cl− reabsorption.

to an overestimation of creatinine clearance and thus glomerular filtration rate (GFR). The physiological role of this system is less well characterized. URAT1 contributes to the urate homeostasis. Urate is the end product of purine metabolism in humans (rodents express uricase allowing metabolism of urate to allantoin) and hyperuricaemia is a risk factor for chronic kidney and cardiovascular disease (Filiopoulos et al., 2012). URAT1 reabsorbs urate in exchange for lactate or nicotinate and is therefore an attractive therapeutic target in patients with hyperuricaemia/gout. Corticosteroids have affinity for the OAT proteins and dopamine and adrenalin are transported via the organic base route. These systems may therefore contribute to intrarenal recycling of physiological active hormones. Gene deletion studies in mice also indicate that OAT3 may contribute to blood pressure control (Vallon et al., 2008), suggesting a role in Na+ and fluid homeostasis.

Major control mechanisms Autoregulation of blood flow is an intrinsic property of the vasculature that stabilizes renal perfusion in the face of fluctuating

S3 Lumen OAT4

Cl–

URAT1

OA–

Interstitial fluid α-KG2–

OAT1

–

OAT3

OA

Na+ K+ H+ OB+

OCT1 OB+

OCT3

Fig. 20.4  The S3 proximal tubule is a major site for the secretion of organic acids (OA) and bases (OB), which include several classes of clinically important drugs. The physiological role of this system is less well defined.

blood pressure. Autoregulation is so efficient that renal blood flow can be largely independent of blood pressure in the physiological range (Cupples, 2007). Whole-kidney autoregulation is governed through the combined influence of at least two mechanisms: tubuloglomerular feedback (see ‘The macula densa’) and the intrinsic myogenic response of the vascular smooth muscle. These regulatory systems have different, but overlapping, operational frequencies. Of the two major components, only the intrinsic myogenic response is both necessary and sufficient for full whole-kidney autoregulation (Cupples, 2007). The myogenic response operates along the preglomerular vascular tree, responding to increased transmural pressure by channel-mediated calcium influx and reflex vasoconstriction of the vascular smooth muscle. The exact signalling mechanisms are not defined, but local release of ATP is implicated, causing vasoconstriction through activation of P2X1 channels (Shirley et al., 2013). Glomerulotubular balance (GTB) is a mechanism intrinsic to the haemodynamic and structural properties of the proximal tubule. It ensures that fluctuations in GFR, and therefore filtered solute load, are matched by near proportionate changes in proximal tubular reabsorption such that the fractional reabsorption of the proximal tubule is held almost constant. This serves to prevent loss of solute in the event of increased GFR; as GFR drops it maintains delivery of sodium to the distal nephron, permitting efficient regulation of potassium and proton secretion. Mechanistically, the balance of peritubular Starling forces is a major component of GTB. Thus, when filtration rate rises due to an increased filtration fraction, the oncotic pressure in the peritubular capillaries is elevated, stimulating reabsorption. When GFR rises without a change in filtration fraction (influenced by glomerular plasma flow), flow-dependent reabsorption may contribute to GTB. The underlying mechanism of flow-dependence couples mechanical forces exerted on the microvilli to altered intracellular calcium signalling (Weinbaum et  al., 2010)  and modulation of paracrine agents, such as ATP (Shirley et al., 2013), dopamine (Du et al., 2012b) or angiotensin II (Du et al., 2012a). Pressure natriuresis is the process through which increases in arterial blood pressure lead to an increase in renal sodium excretion. This process has infinite feedback gain, that is, it continues to

chapter 20 

function until blood pressure returns to set point. Pressure natriuresis is central to the integrated maintenance of sodium homeostasis and thus the long-term regulation of blood pressure (Wadei and Textor, 2012). When kidney function is normal, sustained elevation of arterial pressure of only a few mmHg can evoke large changes in sodium excretion. An impaired pressure natriuresis response is a hallmark of hypertensive states or conditions of sodium sensitivity of blood pressure (Ivy and Bailey, 2014) (Fig. 20.5). Pressure natriuresis is ascribed to a diminution of sodium reabsorption in the proximal tubule following an acute increase in blood pressure. The main mechanism is an increase in renal interstitial hydrostatic pressure, particularly in the medulla. Decapsulation of the kidney prevents the rise in interstitial pressure and blunts the natriuresis, indicating that the response is largely determined by physicochemical factors. Pressure natriuresis occurs in the absence of large changes in renal blood flow due to autoregulation. This at first seems counterintuitive, since the increase in systemic arterial pressure must be transduced to the kidney by changes in renal perfusion pressure. One possibility is that medullary vasculature autoregulates less well than the cortex and small increases in whole-kidney blood flow exert large changes in the medulla. Nevertheless, factors that directly affect tubular sodium transport, particularly those that determine nitric oxide bioavailability are important (Menzies et al., 2015). Renal sympathetic nerve activity (RNSA) can directly stimulate proximal tubule sodium transport by increasing NHE3 and Na+/ K+-ATPase activity. RSNA is increased following a reduction in circulating volume or a modest (< 5 mmol/L) elevation of plasma sodium, such as can occur following a meal. RNSA thus contributes to an integrated view of blood pressure control (Kopp, 2011). Renal sympathetic overactivity has long been linked to sodium retention in experimental hypertension and recent clinical data indicate that bilateral sympathetic efferent denervation effects sustained reductions in blood pressure in selected hypertensive patients (Esler et al., 2010). Plasma angiotensin II, within the physiological range stimulates sodium transport by activation of NHE3; pharmacological levels

Salt excretion

Sodium intake (× Normal)

8 6

B High salt intake

4 A

2

C

50

100

150

Normal salt intake 200

MABP (mmHg)

Fig. 20.5  Renal function curve showing the relationship between arterial blood pressure (MABP) and renal sodium excretion. (A) The equilibrium pressure that is maintained through adjustment in sodium balance. (B) On sustained increases in salt intake, the function curve shifts to the left to give a higher level of excretion at any given pressure. (C) If natriuretic capacity is impaired, the curve shifts to the right and is flattened, so that a higher equilibrium pressure is required to match sodium output to input.

an overview of tubular function

are inhibitory. These effects are mediated by AT1 receptors on the basolateral membrane. Glucocorticoids, acting via the glucocorticoid receptor, stimulate the reabsorption of sodium bicarbonate due to activation of NHE3 and the Na+HCO3− cotransporter on the basolateral membrane. From an integrated standpoint, however, glucocorticoids promote natriuresis, kaliuresis and phosphaturia. Some of this reflects inhibition of transporters but the majority reflects the haemodynamic effects of increased GFR (Hunter et al., 2014). Intraluminal factors produced by proximal tubule cells will regulate transport by autocrine/paracrine signalling. The concentration of angiotensin II in tubular fluid is much higher than that in plasma due to the intrarenal RAAS (Ramkumar and Kohan, 2013). Angiotensin II can either stimulate (via AT1R) or inhibit (via AT2R) proximal transport by regulating NHE3 activity. Similarly, the proximal tubule has a local dopaminergic system (Zhang et al., 2012); activation of D1 receptor contributes to the natriuretic effect of dopamine. Finally, intraluminal nucleotides inhibit NHE3 activity via P2Y1 receptors (Shirley et al., 2013).

The loop of Henle The loop of Henle comprises the proximal straight tubule (see above), the thin descending limb, the thin ascending limb, and the cortical and medullary segments of the TALH. The thin descending limb has a high expression of AQP1 in both apical and basolateral membranes and has, therefore, a large hydraulic permeability. The permeability to sodium varies among species and transepithelial transport processes are not defined. In contrast to the proximal tubule, the thin limb is simple epithelial of flat cells with short microvilli and few mitochondria. Consistent with this, only a low level of Na+/K+-ATPase activity has been detected and few other transport proteins detected:  there is evidence for Cl− and K+ channels and K+Cl− cotransport and a sodium-borate (or potentially Na+HCO3− transporter), encoded by SLC4A11, has been identified (Groger et al., 2010). UT-A2 (SLC14A2) transporters have been identified in the thin descending limb, giving this segment a high permeability to urea. The transition from thin descending limb to thin ascending limb is evident only in the deep nephron population and occurs just before the bend of the loop. The thin ascending limb is a simple epithelium, having low water permeability due to absence of any aquaporin. Although Na+/K+-ATPase activity is low, the epithelial permeability to sodium, potassium and chloride is high. A  bumetanide-sensitive pathway, presumably a Na+-K+-2Cl− cotransporter has been identified in several species (Nishino et  al., 2007)  and mRNA for NHE1, NHE3, H+-ATPase and H+/ K+-ATPase is expressed (Pannabecker et al., 2002). In addition, Cl− channels are identified in both apical and basolateral membrane (Liu et al., 2002). The thin descending limb has significant permeability to urea, allowing rapid equilibrium of the tubular fluid and medullary interstitium.

Thick ascending limb of Henle The TALH is a major site of Na+Cl− reabsorption, accounting for approximately 20% of the filtered load. Cells of the TALH contain a high number of mitochondria, reflecting the significant activity of Na+/K+-ATPase. Sodium crosses the apical membrane of TALH

175

176

Section 2  

fluid, electrolyte, and renal tubular disorders

Lumen

TALH

Interstitial fluid

ROMK K+

Na+ K+ 2Cl–

NKCC-2

K+

Na+

Na+

NHE2 NHE3 +

K+

H+ Na+ K+ Ca2+ Mg2+

Cl–

ClC-Kb

Cl– Barttin

ClC-Ka

K+ Cl–

KCC1

Fig. 20.6  Major transport pathways in the thick ascending limb of Henle.

cells by several routes (Fig. 20.6). The NHE3 and NHE2 exchangers but their contribution to transcellular Na+ flux is marginal (Shirley et al., 1998). More recently mRNA for a sodium-activated sodium channel has been found in the TALH. Expression is regulated by dietary sodium content and although it is hypothesized that this may be the elusive renal sodium sensor, its physiological function is unclear (Lara et al., 2012). The main route of entry is via NKCC2 (Fig. 20.6). This protein is inhibited by loop diuretics, which compete with Cl− at one of its binding sites. NKCC2 is encoded by a single gene, SLC12A1, mutations in which cause the salt-wasting Bartter syndrome type 1 (Seyberth and Schlingmann, 2011) (Fig. 20.7). Alternative splicing of SLC12A1 produces variants of NKCC2, which differ in their ion affinity and in their distribution along the TALH. Physiologically, the distribution is such that the affinity matches the concentration

of the transported ions in the tubular fluid, thereby permitting maximum rates of Na+Cl− reabsorption (Ares et al., 2011). Chloride exits the basolateral membrane via K+Cl− cotransport or through a heteromeric chloride channel assembled from CLC-Kb and Barttin (BSDN). Mutations in these subunits cause Bartter’s type 3 and 4, respectively (Fig. 20.7). Most of the potassium that enters the cell via NKCC2 diffuses back across the apical membrane through apical K+ channels ROMK/ KCNJ1, driven by the electrochemical gradient. This recycling of K+ has two consequences. First it ensures the continued availability of K+ to the cotransporter. Na+ and K+ must bind in a 1:1 stochiometry and the K+ concentration in the fluid delivered into the TALH is an order of magnitude lower than that of Na+. Second, it confers electrogenicity on a transporter that is intrinsically electroneutral. The charge separation caused by movement of K+ back into the tubule

Lumen

TALH

Interstitial fluid

Na+ Type 2

K+

K+ Type 1

K+ 2Cl– Na+

Na+ K+ Ca2+ Mg2+

K+ Cl–

Type 3

Cl– Type 4

Fig. 20.7  Mutations causing Bartter syndrome impair the transcellular reabsorption of Na+Cl−, leading to a diminution of the gradient for paracellular reabsorption of cations.

chapter 20 

fluid and Cl− out across the basolateral membrane generates a trans­ epithelial PD that is lumen positive (~ +10 mV). Theoretically, the transepithelial PD can drive paracellular reabsorption of cations or secretion of anions. The tight junctions of the TALH do have a high ionic conductance (and a very low hydraulic conductance) but are selective for cations. Approximately 50% of the Na+ reabsorbed in the TALH is via this voltage-driven paracellular shunt. This also provides a major route for the reabsorption of Ca2+ and Mg2+. The TALH reabsorbs approximately 25% of the filtered K+ load. Despite a significant K+ conductance in both apical and basolateral membrane, the majority of this reabsorption is paracellular and under conditions of high Na+ reabsorption, the net transcellular K+ flux can be secretory. Nevertheless the apical K+ channels are a vital for the efficient functioning of the TALH. Patch-clamp studies have identified two K+ channels in this membrane. The smaller conductance channel has the electrophysiological ‘fingerprint’ of the cloned ROMK channel, encoded by KCNJ1. Mutations in this gene cause Bartter syndrome type 2 and small molecule inhibitors of ROMK are in development as novel diuretics (Bhave et al., 2011). Although the molecular identity of the second channel is unknown, it is not expressed in the TALH of KCNJ1 knockout mice (Lu et al., 2004), suggesting that it is formed as a heteromeric complex in which ROMK is essential for physiological function.

The macula densa The macula densa, a small plaque of approximately 20 cells found at most distal segment of the cTALH, is the senor unit of the juxtaglomerular apparatus and influences the secretion of renin from granular cells of the afferent arteriole (Kurtz, 2011). Macula densa cells also exert a countervailing influence on GFR in the glomerulus of origin (and perhaps also in neighbouring nephrons) via vasoactive effects on the afferent arteriole and mesangial cells. This process, called tubule-glomerular feedback, serves to stabilize the moment-to-moment delivery of NaCl into the distal nephron and thereby optimize the fine-tuning of Na+ (and K+/H+) homeostasis by the kidney: aberrant tubuloglomerular feedback has been implicated in the progression of diseases such as diabetic nephropathy (Vallon and Thomson, 2012). Macula densa cells express the high-affinity isoform of NKCC2 in the apical membrane. The exit pathways for Na+ are unclear and intracellular Na+ concentration is not well buffered compared to other renal epithelial cells, rising and falling along with apical entry. As Cl− exits through channels, the membrane becomes depolarized, activating a maxi-anion channel through which ATP is released. The release of ATP, directly proportional to Na+ delivery to the apical membrane, is the critical effector of tubule-glomerular feedback, promoting vasoconstriction of the afferent arteriole either directly or following metabolism to adenosine (Bell et al., 2009). The TALH is also the site for the production and secretion of Tamm–Horsfall protein (aka uromodulin). This glycoprotein is the most abundant protein in normal urine, with up to 150 mg/ day being excreted. The physiological roles are poorly understood. Mutations in the encoding gene, UMOD, cause a cluster of rare autosomal dominant kidney diseases, characterized by progressive tubule-interstitial damage, hyperuricaemia, urinary concentrating defects, cyst formation, and progressive renal failure. Genome-wide association studies have linked common UMOD variants to chronic kidney disease, hypertension, and type 2 diabetes. Studies in umod knockout mice provide a more mechanistic

an overview of tubular function

link to physiological function. Tamm–Horsfall protein may guard against infection by inhibiting adherence of bacteria to tubular cells and act as a constitutive inhibitor of calcium crystallization in tubular fluid (Rampoldi et al., 2011).

Major control mechanisms As with the proximal tubule, the loop of Henle is well endowed with sympathetic nerve endings: the medullary thick limb is the most densely innervated of any tubule segment. Increased RNSA increases Na+ reabsorption in the loop of Henle as a whole and β-adrenergic stimulation in vitro stimulates Na+ flux in isolated TALH. In vivo bilateral denervation studies suggest that RNSA increases NKCC2 by a cyclic AMP-dependent mechanism (Torp et al., 2012). Aldosterone has a stimulatory effect on Na+ and K+ reabsorption in superficial loops of Henle perfused, although this has not been a consistent finding. NKCC2 does not seem to be an aldosterone-induced protein, but studies report stimulation of basolateral Na+/K+-ATPase (Grossman and Hebert, 1988), which would increase the driving force for Na+ transport reabsorption. The receptor through which these effects are mediated is undefined:  both mineralocorticoid and glucocorticoid receptors are expressed in the thick limb but 11β-hydroxysteroid dehydrogenase type 2 (11β-HSD2), the enzyme that governs ligand access, is absent (Ackermann et al., 2010). Circulating vasopressin stimulates Na+ reabsorption in the TALH through V2 receptor-mediated activation of adenylate cyclase and increased production of cAMP. This cascade culminates in a coordinated activation of key components in the transcellular reabsorptive pathway: apical K+ and basolateral Cl− channels are activated and there is increased trafficking of NKCC2 to the apical membrane Vasopressin also phosphorylates two threonine residues in the N-terminal of NKCC2, which may contribute to enhanced transporter activity (Ares et al., 2011). Other hormones that lead to increased intracellular cAMP such as parathyroid hormone (PTH), calcitonin, and glucagon increase NKCC2 activity and stimulate reabsorption in the TALH. Intraluminal factors inhibiting reabsorption in the loop of Henle include eicosanoids (e.g. prostaglandin E2), produced intrarenally from arachadonic acid metabolism by cyclooxygenase, endothelin-1 (ET-1) and extracellular ATP. The TALH is second only to the collecting duct in terms of ET-1 synthesis. ET-1, via ETB receptors, inhibits sodium transport via an effect on NKCC2 (Ramseyer et al., 2011). ATP is released by TALH cells in response to increased flow and inhibits NKCC2 activity via P2 receptors in both the apical and basolateral membranes. Nitric oxide production (through NOS3) may be the final inhibitory mediator for these paracrine factors (Garvin et al., 2011). Increased extracellular Ca2+ inhibits Na+ transport in the TALH, reducing the driving force for calcium reabsorption by the paracellular cation shunt pathway. The increased interstitial Ca2+ is detected by the G-protein coupled Ca2+-sensing receptor, localized on the basolateral membrane. Activation of the receptor increases the production of 20-HETE and thereby inhibits NKCC2, ROMK, and the basolateral Na+/K+-ATPase. The cognate ligand is Ca2+ but the receptor is activated by a number of other divalent and trivalent cations, including Mg2+ (Gamba and Friedman, 2009). Mutations in this receptor can cause a bartterlike syndrome, sometimes called Bartter type 5, but this nomenclature is not widely accepted.

177

178

Section 2  

fluid, electrolyte, and renal tubular disorders

Distal tubule and collecting duct The distal tubule lies between the macula densa and the point of confluence with another distal tubule at which a collecting duct is formed. The distal tubule incorporates several subsections: the distal convoluted tubule (DCT), the connecting tubule (CNT), and a short section of the cortical collecting tubule, often referred to as the initial collecting tubule. The collecting duct is divided into the cortical (CCD), outer medullary (OMCD), and inner medullary collecting duct (IMCD). In addition to a cell type characteristic of each subsegment, intercalated cells are found in much of the distal nephron. In general, adjacent cells have a much deeper contact than in the proximal tubule and the tight junctions are much less ‘leaky’:  paracellular flux is not so common and the epithelia is able to maintain a larger transepithelial PD.

Distal convoluted tubule The DCT makes up approximately 50% of the total distal tubule. The first part (DCT1) consists exclusively of DCT cells while DCT2 also includes a small proportion of intercalated cells. The DCT has the highest density of mitochondria along the nephron. These cluster along the basolateral membrane, which has a considerably higher surface area than that of the apical membrane. The epithelial of the DCT is very plastic and will amplify or regress depending on sodium delivery. The exact cue to this is unknown but transepithelial flux of Na+ is likely to be a major factor. In the apical membrane, DCT cells express a Na+Cl− cotransporter (TSC/NCC) sensitive to thiazide diuretics binding to the chloride-binding site. The apical membrane also contains the epithelial calcium channel (ECaC or TRPV5), a ROMK-like potassium channel, a K+Cl− cotransporter and NHE2 (Fig. 20.8). In keeping with a central role for the DCT in Na+ homeostasis, basolateral Na+/K+-ATPase is greater than in any other nephron segment. The basolateral membrane also has a chloride exit channel (CLC-K2) and a Ca2+-ATPase. An inwardly rectifying potassium channel (Kir4.1) mediates K+ recycling across the basolateral membrane. The importance of this process is underscored by the description

DCT

of the salt wasting EAST syndrome that results from mutations in encoded by KCNJ10, the encoding gene (Bockenhauer et al., 2009). The major route for Na+Cl− transport across the apical membrane is NCC, coupled to exit across the basolateral membrane via Na+/K+-ATPase and ClC-K2 respectively. Loss of function mutations in the encoding gene, SLC12A3, cause Gitelman syndrome (Seyberth and Schlingmann, 2011), which presents as a milder form of Bartter syndrome. A point of differential diagnosis, however, is that Gitelman syndrome presents with hypocalciuria and hypomagnesaemia, while Bartter syndrome patients have normal or hypercalciuria and typically normal Mg2+ levels. The basis for these differences is explained in the section on Ca2+ transport. Pseudohypoaldosteronism type 2 (Gordon disease or familial hypertension with hyperkalaemia (FHH)) presents as a gain of function of NCC: hypertension in these patients is very sensitive to low-dose thiazide diuretics. Human genetic screening identified mutations in two members of a novel family of regulatory kinases, the with-no-lysine (WNK) kinases isoforms 1 and 4. The expression of WNK4 is limited to the aldosterone-sensitive distal nephron. In the DCT is normally exerts an inhibitory effect on thiazide-sensitive Na+ transport, partly by reducing expression of NCC at the apical membrane. This inhibitory effect is no longer observed in PHAII-WNK4 mutants, giving rise to dysregulated NCC activity. The physiological effects of WNK1 are not well understood. It is expressed along the nephron but there is no evidence to suggest a direct interaction with NCC and it is likely that WNK1 affects the activity of other regulatory kinases (Gamba, 2009). The DCT is not a major site of potassium secretion: the absence of electrogenic sodium reabsorption in DCT1 results in an apical membrane potential close to the K+ equilibrium potential and there is thus little driving force for ROMK-mediated section. DCT2 contains hybrid cells that also express the epithelial sodium channel (ENaC) in the apical membrane (Nesterov et  al., 2012). ENaC-mediated Na+ entry depolarizes the apical membrane, sustaining some K+ secretion. Potassium secretion can also occur via K+Cl− cotransport (KCC1) but only if the luminal chloride concentration is unusually low. This can happen if the delivery of HCO3− is high (Amorim et al., 2003), a situation that also favours

Lumen

Interstitial fluid

3Na+

Na+ NCC

Kir4.1 K+

K+

KCC1

Kv1.1

2K+

Cl–

Cl–

Cl–

K+

TRPV-5 Na+ ENaC

Fig. 20.8  Major transport pathways in the distal convoluted tubule.

Ca2+ Ca2+

ClC-Kb

Na+ Ca2+

NCX1

chapter 20 

NHE2-mediated NaHCO3 reabsorption in this segment (Bailey et al., 2004).

Major control mechanisms In order to be physiologically active, NCC must be phosphorylated on threonine and serine residues in a regulatory domain in the N-terminus (Rafiqi et al., 2010) and correctly trafficked to the apical membrane of the DCT. Recent studies indicate that angiotensin II, via activation of AT1R, can increase NCC phosphorylation (van der Lubbe et al., 2011) and stimulate thiazide-sensitive Na+ transport (Ashek et al., 2012). In vivo studies have shown that corticosteroids can increase NCC-mediated sodium transport in the DCT (Velazquez et  al., 1996). The exact mechanisms of action are not clear and both MR and GR are expressed in the DCT, but not 11β-HSD2 (Ackermann et al., 2010). Other hormones, such as vasopressin and insulin can phosphorylate NCC and increase apical expression. In the chronic setting, the kidney escapes from the Na+-retaining effects of aldosterone (aldosterone escape) by downregulation of NCC (Turban et al., 2003).

The connecting tubule The CNT consists of two cell types: CNT cells, the majority, and intercalated cells, the remainder (~ 30% of the total). Like DCT cells, CNT cells exhibit basolateral amplification and have mitochondria along this membrane. The majority of Na+ transport in the CNT is amiloride-sensitive. These cells express NHE2, which can be inhibited by amiloride analogues, but the major route for sodium reabsorption, is via the ENaC in the apical membrane, coupled to basolateral Na+/K+-ATPase activity. Notably, ENaC activity in the CNT does not seem to be regulated by aldosterone (Nesterov et al., 2012).

CCD

an overview of tubular function

There is functional evidence for thiazide-sensitive Na+ transport, but NCC expression is virtually zero. This may reflect a novel, pathway for Na+ reabsorption in the intercalated cell (see Fig. 20.14), in which Cl− entry through pendrin is recycled on a newly identified apical transporter, NDCBE (Na+-driven, Cl−, bicarbonate exchanger (SLC4A8)), in exchange for Na+ (Leviel et al.). This system is best described in the collecting duct and may explain the long-standing observation that a significant proportion of Na+ transport in the cortical collecting duct is sensitive to thiazides. The remainder is amiloride-sensitive and in the CNT, ENaC-mediated Na+ entry depolarizes the apical membrane with respect to the basolateral membrane, generating a large-lumen negative PD of 30–40 mV. The apical membrane also expresses two K+ channels in the apical membrane: the low conductance (~ 35 pS) K+ channel, ROMK (KCNN1) and a larger (~ 100 pS), big potassium (BK) (KCNMA1) channel (Malnic et  al., 2013). The high K+ permeability of the apical membrane favours K+ secretion and exerts a repolarizing influence on the apical membrane potential, thereby sustaining ENaC-mediated Na+ transport. The CNT is also an important site for Ca2+ reabsorption, accounting for approximately 15% of the filtered load and exerting the final regulation of urinary excretion under the control of PTH (see below). CNT is the most proximal site of vasopressin-regulated water transport, expressing both AQP2 water channels and the V2 receptor. The route of Cl− transport is not clear. A basolateral exit pathway (ClC-K2) is available and Cl− might enter the cell via NDCBE in the β-intercalated cell (Leviel et al.). Alternatively, chloride might be reabsorbed paracellularly.

The collecting duct The initial collecting duct is the final segment of the distal tubule and has the same structural makeup as the CNT, comprising mainly

Principal cell

Lumen

Interstitial fluid

Na+ Na+

ENaC

K+ K+ K+

ROMK

H2O AQP-2

Fig. 20.9 Na+ reabsorption in the principal cell of the connecting tubule and collecting duct.

AQP-3 or 4

179

180

Section 2  

fluid, electrolyte, and renal tubular disorders Cortisol

11βHSD2

Aldosterone

MR

Transcription MR

Na+ ENaC

Cortisone Na+

Protein synthesis

Na+ K+

Fig. 20.10  Aldosterone specificity in the distal nephron is governed by the enzyme 11β-dehydroxysteroid dehydrogenase type 2 (11β-HSD2), which converts glucocorticoids into metabolites that are not able to activate the mineralocorticoid receptor (MR).

of principal cells with intercalated cells making up the remainder. The OMCD is much the same as the CCD, both in appearance and cell structure. There is a diminution in the proportion of intercalated cells during the transition from the outer to the inner stripe and are absent from the terminal IMCD. In this segment, structurally distinct IMCD cells replace principal cells. In the initial collecting tubule (ICT), CCD, and OMCD, the major entry pathway for Na+ is apical ENaC. Expression of this channel, and of the basolateral Na+/K+-ATPase, declines progressively along the collecting duct and the principal cells become less densely packed with mitochondria. ENaC is not found beyond the first third of the IMCD, being replaced by an amiloride-sensitive, non-selective cation channel. Mechanistically, Na+ reabsorption in the CCD proceeds via the same mechanism as in the CNT: entry through ENaC is electrophysiologically coupled to K+ secretion via ROMK/BK channels. The OMCD and IMCD do not express apical K+ channels: ENaC-mediated Na+ reabsorption is, therefore, self-limiting as depolarization of the apical membrane attenuates the electrochemical gradient for Na+ entry (Fig. 20.9). Under sodium replete conditions, ENaC reabsorbs approximately 1–2% of the filtered Na+ load (Ashek et al., 2012). Sodium transport in this segment is plastic and strongly influenced by the RAAS. In the short term (< 4 hours), aldosterone stimulates Na+/K+-ATPase activity and regulates ENaC trafficking, prolonging the half-life of the channel complex in the apical membrane. In the long term, MR activation increases the expression of the basolateral pump, ENaC and ROMK potassium channels. Constitutive activation of ENaC causes the hypertensive Liddle syndrome. ENaC is composed of three subunits (alpha, beta, and gamma), encoded by distinct genes (SCNN1a, SCNN1b, and SCNN1c). Genetically, Liddle syndrome arises from mutations in β or γ subunits that impair the removal of the channel complex from the apical membrane (Palmer et al., 2012). Conversely, the salt-wasting disorder of pseudohypoaldosteronism type 1 is caused by loss-of function mutations in the mineralocorticoid receptor (PHA1A) or by mutations in the ENaC genes (PHA1B), rendering the channel non-functional (Mullins et al., 2006). In mice, deletion of Scnn1a, Scnn1b, or Scnn1c causes perinatal lethality. More sophisticated genetic approaches have deleted specifically the αENaC subunit from cells of collecting duct lineage (Rubera et al., 2003) or from cells of the CNT (Christensen et al.,

2010). These studies strongly suggest that sodium balance, even under conditions of dietary sodium restriction, is critically dependent on ENaC in the late DCT and CNT.

Aldosterone action and 11β-hydroxysteroid dehydrogenase type 2 In vitro, MR can be activated with equal potency both by aldosterone and cortisol; in vivo, ligand access to MR is determined by co-localization with 11β-HSD2 (Fig. 20.10). By catalysing the rapid conversion of cortisol into cortisone, which does not activate MR, 11β-HSD2 confers upon MR the specificity to aldosterone that it inherently lacks (Funder et  al., 1988). MR and 11β-HSD2 have overlapping distributions in the collecting duct, helping to define the ‘aldosterone-sensitive distal nephron’. The DCT, however, does not express the enzyme (Ackermann et al., 2010) and the control by corticosteroids of sodium transport is not fully understood. In rats, pharmacological inhibition of 11β-HSD2 increased sodium reabsorption by the collecting duct (Bailey et al., 2001). In humans, excess intake of liquorice, which contains the 11β-HSD2 inhibitor glycyrrhetinic acid, causes a sodium-dependent hypertension. Inactivating mutations in the encoding gene (HSD11B2) cause the hypertensive syndrome of apparent mineralocorticoid excess (AME): a mouse model of this disorder presents with hypertension and severely hypokalaemia and impaired renal Na+ excretion due to activation of ENaC (Bailey et al., 2008). AME is an extreme phenotype and, like the other Mendelian blood pressure disorders, is very rare. Nevertheless, these disorders illustrate the fundamental role of renal Na+ transport in blood pressure. Moreover, mild mutations in these same genes may be prevalent in the essential hypertensive population (Wagner, 2008), particularly in those individuals with low-renin or salt-sensitive hypertension. For example, several association studies link polymorphisms in HSD11B2 to blood pressure and mice with 50% of the normal levels of enzyme salt-sensitive blood pressure due to an impaired ability regulate ENaC activity (Craigie et al., 2012).

Major control mechanisms The major regulator of Na+ reabsorption and K+ secretion in the distal nephron is aldosterone, which is secreted from the zona

chapter 20 

glomerulosa of the adrenal cortex in response to plasma angiotensin II (reflecting effective plasma volume) and K+ concentration. The glucocorticoid receptor is also expressed throughout the distal nephron and provides an alternative pathway for ENaC regulation, particularly when the hypothalamic–pituitary–adrenal axis is activated (Bailey et al., 2009). Vasopressin is emerging as a powerful regulator of Na+ reabsorption in the collecting duct, acting synergistically with aldosterone activating ENaC via V2 receptor activation (Stockand, 2010). Another neurohypophysial hormone, oxytocin, is natriuretic due to nitric oxide-dependent inhibition of tubular reabsorption and by stimulating the release of atrial natriuretic peptide (Gimpl and Fahrenholz, 2001). The natriuretic peptides are a complex family of peptides influencing cardiovascular and renal function. Atrial natriuretic peptide, released from atrial myocytes in response to stretch, increases sodium excretion through coordinated inhibition of the apical cation channel and basolateral Na+/K+-ATPase in the IMCD (Beltowski and Wojcicka, 2002). Natriuretic peptides also act as antagonists of the RAAS. A wide-range of paracrine factors influences collecting duct function. Extracellular nucleotides regulate Na+ and water transport through activation of P2X and P2Y receptors (Bailey and Shirley, 2009). Endothelins also influence ENaC activity (Kohan et al., 2011). These systems have complex levels of interaction and probably serve to modify the overall ‘tone’ set by the RAAS. Proteolytic cleavage of the γ subunit by serine proteases (e.g. prostatin, trypsin, chymotrypsin, and elastase) will increase ENaC activity (Rossier and Stutts, 2009). ENaC cleavage is regulated in vivo by aldosterone and may also underpin the sodium retention observed in proteinuric states such as nephrotic syndrome. Several of the key transporters in the distal nephron have marked circadian rhythm of expression (Stow and Gumz, 2012). ENaC, for

an overview of tubular function

example, is under control of the clock gene per1 and disruptions in circadian control have been linked to non-dipper patterns of blood pressure (nocturnal hypertension), a risk factor for cardiovascular and kidney disease.

Urine concentration and dilution (See Chapter 22.) The loop of Henle reabsorbs a considerable amount of sodium, potassium, and water. In addition to these transepithelial fluxes, a major function of the loop of Henle is the generation and maintenance of the interstitial osmotic gradient that increases from approximately 290 mOsm/kg in the renal cortex to approximately 1200 mOsm/kg the tip of the medulla. The driving force for this is the reabsorption in the TALH of solute without water, which generates a ‘horizontal’ osmotic gradient of approximately 200 mOsm/kg at any point between the tubule fluid and interstitium. This ‘single osmotic effect’ also exists at any given level between the ascending and descending limb because the latter has high osmotic permeability and is in equilibrium with the surrounding interstitium. The hairpin structure of the loop, in which flow in the ascending limb is in the opposite direction to that in the descending limb, multiplies the single effect (countercurrent multiplication), creating a much larger ‘vertical’ or corticomedullary gradient (Fig. 20.11). The highest degree of urine concentration is found in mammals with the longest renal papilla. This partly reflects the increased length of the loop of Henle, which in species adapted to arid climates can multiply the single effect to 11 Osmol/L/kg at the papillary tip. The countercurrent multiplier function of the loop of Henle dilutes the tubule fluid (fluid exiting the TALH has an osmolarity of ~ 100mOsm/kg) and concentrates the medullary interstitium. This creates the potential to produce hypotonic or hypertonic urine

Na+ H2O

Na+ Isotonic

290

H2O

290

Hypertonic

Na+

Na+ 100

290

Na+ 600

600

Na+ 400

H2O

900

Na+

Inner medulla 700

900

Hypertonic 1200

Outer medulla

600

H2O

900

Cortex

1200

Fig. 20.11  Countercurrent multiplication of the ‘single osmotic effect’ in the loop of Henle.

1200

181

182

Section 2  

fluid, electrolyte, and renal tubular disorders

in response to homeostatic requirements, by varying the permeability to water of the collecting duct system. The key hormone is vasopressin (antidiuretic hormone), released from the hypothalamus, via the posterior pituitary, in response to increased plasma osmolarity or decreased circulating volume. Vasopressin binds the V2 receptor, activating a signalling cascade that leads ultimately to insertion of AQP2 into the apical membrane of the principal cell. The basolateral membranes of these cells constitutively express AQP3 and AQP4: apical insertion of AQP2 is the rate-limiting step for transepithelial water reabsorption. When plasma vasopressin is high, the hypotonic fluid delivered into distal tubule becomes isotonic by the ICT and progressively hypertonic as it descends through the MCD. In conditions of hydration, when vasopressin is low, little water is extracted during passage through the distal nephron and the final urine can be further diluted by the continued reabsorption of sodium chloride.

The role of urea The thin limbs of the loop of Henle are permeable to urea (ascending > descending), the TALH and distal nephron are impermeable up to the terminal section of the IMCD. In this part of the nephron, vasopressin-dependent water reabsorption has led to a high urea concentration within the tubule fluid. Vasopressin also determines the urea permeability of the terminal IMCD: V2 receptor activation increases the expression of urea transporters (UT-A1 and UT-A3) in the apical and basolateral membrane respectively, facilitating the passive reabsorption of urea into the inner medullary interstitium (Fenton and Knepper, 2007). Some of this urea enters the vasa recta and some the S3 segment of the proximal tubule and the descending and ascending thin limbs of the loop of Henle. This is then returned to the IMCD to be reabsorbed. The net result of this urea ‘recycling’ process is to add urea to the inner medullary interstitium, thereby increasing interstitial osmolality. The high urea concentration within the MCD is rendered osmotically ineffective since it is balanced by a similarly high urea concentration in the medullary interstitium. This allows large quantities of urea to be excreted without obligate osmotic diuresis. Moreover, the concentration of urea in the medullary interstitium increases water abstraction from the thin descending limbs of deep nephrons, raising the intraluminal Na+ concentration within these structures. Thus urea recycling also creates a concentration gradient that in theory could account for passive Na+ reabsorption from the thin ascending limbs, which lack Na+/K+-ATPase. However, this concept is challenged by studies in mice with genetic deletion of UT-A1 and UT-A3, which have reduced urea concentration in the inner medullary interstitium but a normal interstitial Na+Cl− gradient (Fenton and Knepper, 2007). Thus, the mechanisms responsible for the inner medullary electrolyte gradient are still undefined. An interesting theory is that peristaltic contractions observed in the renal pelvis compress rhythmically the hyaluronic acid matrix in the inner medulla, generating a hydrostatic pressure gradient to create the single effect: paralysis of the papillary wall reduces the osmolarity of the inner medulla (Pruitt et al., 2006). It is worth emphasizing, however, that the ultimate driving force for countercurrent multiplication is active Na+ reabsorption in the TALH. This is underscored by the disruption of the osmotic gradient when loop diuretics are given.

Countercurrent exchange in the vasa recta If the capillaries supplying the renal medulla had the usual anatomical arrangement of a capillary network, then medullary blood would rapidly dissipate the medullary osmotic gradient as the hypertonic interstitium equilibrated with isotonic capillary blood. This does not happen to any appreciable extent, because the vasa recta also have a special anatomical arrangement, embracing the epithelial structures in a U-shape. The blood does indeed equilibrate with the neighbouring interstitium but solute entry and water loss in the descending vasa recta are offset by solute loss and water entry in the ascending vasa recta. Although countercurrent exchange is a passive process, contractile cells, called pericytes, control vasa recta flow. This epithelial vascular cross-talk is modulated by a variety of autocrine/paracrine agents (e.g. nitric oxide, eicosanoids, adenosine ATP), and helps to match blood flow to transport in the TALH. Countercurrent exchange applies also to oxygen, which diffuses from descending to ascending vasa recta. This ‘shunting,’ combined with ongoing energy-dependent salt transport in the TALH, renders medullary tissue relatively hypoxic. Nevertheless, medullary cells are adapted to this hostile environment. They have a higher capacity for glycolysis than do cells in the cortex and hypoxia and hyperosmolarity induce, via the transcription factor TonEBP (NFAT5), the expression of proteins that protect the cell and inhibit apoptosis (Burg et al., 2007).

Renal H+/HCO3− transport (acid–base balance)

In individuals on a typical acid-producing diet, the kidneys must reabsorb essentially all the filtered HCO3− (> 4000  mmol/day) and add sufficient extra HCO3− to the plasma to regenerate the buffer anions consumed in buffering the daily acid load (normally ~ 50 mmol/day).

Acid–base transport (See Chapter 24.)

Bicarbonate reabsorption The bulk of filtered HCO3− (~ 80%) is reabsorbed in the proximal tubule, largely in the S1 and S2 segments. About half of the remainder is reabsorbed in the loop of Henle, the rest in the distal tubule and collecting duct. HCO3− reabsorption is indirect: H+ and HCO3− ions are generated in tubular cells (facilitated by intracellular carbonic anhydrase (type II)); the H+ ions are secreted into the lumen, whereas the HCO3− ions enter the plasma. In the proximal tubule, H+ is secreted mainly via apical NHE3 but also via apical H+-ATPase. The secreted H+ ions combine with filtered HCO3− ions to form H2CO3, which is rapidly converted to CO2 and H2O in the presence of apical carbonic anhydrase (type IV); CO2 and H2O diffuse into the cell; both moieties are able to use AQP1 water channels (Endeward et al., 2006). The HCO3− ions generated within the cell enter the interstitial fluid (and thence plasma) via the basolateral Na+HCO3− cotransporter (NBC-1, SLC4A4), which carries three HCO3− ions to one Na+ ion. The net result of these processes is that a filtered HCO3− ion is removed while another one replaces it in plasma (Fig. 20.12). Bicarbonate reabsorption in the loop of Henle takes place mainly in the TALH. Secretion of H+ into the lumen is largely via NHE3, although H+−ATPase makes a modest contribution. There is no apical carbonic anhydrase in the loop of Henle. Intracellularly

chapter 20 

Lumen

NHE3

Interstitial fluid 3Na+

Na+

2K+

H+

H+ H2CO3 CA IV CO2 + H2O AQP-1

K+

HCO3–

H+

HCO3– + H+

an overview of tubular function

3HCO3–

H2CO3

Na+ NBC1A

CAII CO2 + H2O

Fig. 20.12 HCO3− reabsorption in the proximal tubule is dependent on carbonic anhydrase (CA), types 2 and 4.

generated HCO3− enters the interstitial fluid via a basolateral Na-3 HCO3 cotransporter, as in the proximal tubule, and possibly a Cl/ HCO3 anion exchanger (AE2; SLC4A2). In the early DCT, there is evidence for some H+ secretion through apical NHE2 and H+-ATPase (Bailey et  al., 2004), but the cells responsible for H+ handling beyond the early distal tubule DCT2 to IMCD are the intercalated cells. As their name implies, these are interspersed among the majority cell types in each segment. Intercalated cells comprise approximately 30% of the cells in the distal nephron and come in three varieties, all of which contain carbonic anhydrase (type II) and generate H+ and HCO3− ions: type A (the predominant form), type B (present only in the distal tubule and CCD), and non-A, non-B cells, which might be able to switch their function between the other two types according to the prevailing acid–base status. Type A  intercalated cells have an apical H+-ATPase, which secretes H+ into the lumen; HCO3− crosses the basolateral membrane using a Cl−/HCO3− exchanger (AE1; SLC4A1A). The Cl− ions entering the cell by this route are recycled largely through Cl− channels (CLCNKB) and partly through a K+Cl− cotransporter (KCC4; SLC12A7). Although the H+-ATPase is the major player, the apical membrane contains a second type of proton pump: H+/ K+-ATPase, whose activity is upregulated during potassium depletion (Fig. 20.13). Type B intercalated cells are essentially the reverse of type A intercalated cells: H+ ions are pumped across the basolateral membrane via H+-ATPase, and HCO3− ions enter the lumen via an anion exchanger, in this case pendrin (Fig. 20.14). Although intercalated cells have generally been regarded as being concerned solely with acid–base balance, the Cl− ions entering via pendrin can exit the basolateral membrane through Cl− channels, thus providing a mechanism for transepithelial Cl− reabsorption. Moreover, as described above, recent studies suggest the existence of a thiazide-sensitive, Na+-driven Cl−/HCO3− exchanger (Leviel et al., 2010).

Lumen

Interstitial fluid Na+

NH4+ 2Cl–

H+

NKCC-1 AE-1

Cl–

H+ K+

K+ K+ Cl– K+

NH3 RhCG

NH3 RhCG

Fig. 20.13  An acid-secreting (type A) intercalated cell.

This is achieved by the generation of H+ and HCO3− in tubular cells, by the same mechanism as for HCO3− reabsorption, and the addition of HCO3− to the peritubular plasma. The problem then becomes how to deal with the extra H+ ions generated simultaneously with the extra HCO3− ions. There are two principal means of doing so.

Titratable acid excretion Some of the extra H+ ions are secreted into the lumen (via Na+/ H+ exchange and H+-ATPase), where they react with buffer anions (principally filtered HPO42−); any buffer that escapes reabsorption effectively eliminates H+ ions in the urine. The quantity of H+ lost in this way, determined by back-titrating the urine with strong base to pH 7.4 (hence the term ‘titratable acid’), normally amounts to approximately one-third of overall acid excretion. (A few free H+ ions appear as such in the urine, but these are quantitatively insignificant since the minimum urine pH is ~ 4.5.)

Addition of extra bicarbonate to plasma

Ammonium excretion

As indicated above, the kidneys not only reabsorb virtually all filtered HCO3−, but also add further HCO3− to the renal circulation.

The other means of dealing with the extra H+ ions requires the production of NH4+ ions in the proximal tubule. Proximal tubular cells

183

184

Section 2  

fluid, electrolyte, and renal tubular disorders

CD

Lumen

Interstitial fluid H+

HCO3–

Pendrin

Cl– Cl–

NDCBE

Cl–

Na2HCO3–

ClC-KB

Fig. 20.14  A base-secreting (type B) intercalated cell, which also expresses the thiazide sensitive Na+-driven chloride-bicarbonate exchanger (NDCBE).

take up the amino acid glutamine and deaminate it, through reactions catalysed by two enzymes (glutaminase and glutamate dehydrogenase), to NH4+ and α-ketoglutarate (Fig. 20.15). The NH4+ ions are secreted into the tubular lumen largely by substituting for H+ on NHE3 (although some intracellular NH4+ dissociates into NH3, which diffuses across the apical membrane to be ‘trapped’ in the lumen by recombining with secreted H+). The α-ketoglutarate is largely metabolized to glucose, through a series of reactions that consume H+ ions. Most NH4+ ions secreted into the proximal tubule and delivered to the loop of Henle are reabsorbed in the TALH. This reabsorption is partly paracellular, driven by the lumen-positive transepithelial PD, and partly transcellular. Apical uptake of NH4+ is mainly via NKCC2, on which NH4+ can substitute for K+, and basolateral exit is largely via a Na/H exchanger (NHE4; SLC9A4), on which NH4+ can substitute for H+ (Weiner and Verlander, 2011). As a result of events in the TALH, NH4+, in equilibrium with NH3, accumulates in the medullary interstitium. The final, and critical, stage of NH4+ excretion is the transference of this interstitial

NH4+/NH3 to the lumen of the collecting duct. Until recently, this was thought to occur by simple diffusion of NH3 across the basolateral and apical membranes, with ‘diffusion trapping’ of NH4+ in the lumen as a result of H+ secretion. However, it is now clear that simple diffusion cannot account for the NH3 movement; instead, specific transport proteins are involved, notably the rhesus glycoprotein RhCG (Weiner and Verlander, 2011). Genetically engineered mice lacking RhCG have a much reduced capacity to excrete NH4+, owing to reduced permeability to NH3 of apical and basolateral membranes in the CD (Fig. 20.13). Moreover, increased expression of RhCG is seen in the OMCD (in both intercalated and principal cells) of rats subjected to metabolic acidosis, though as yet the mechanism of this apparent adaptation is not known (Wagner et  al., 2011). Although rhesus glycoproteins are the major players, other transporters in the CD may have a subsidiary role in mediating NH4+ excretion. The basolateral membrane of A-type intercalated cells contains a Na+-K+-2Cl−- cotransporter (NKCC1; SLC12A2), on which NH4+ can compete with K+; similarly, NH4+ can compete with K+ on a basolateral Na+/K+-ATPase,

Lumen NHE3

Interstitial fluid

Na+ NH4

+

3Na+ 2K+

+

NH4

K+ Glutamine

α-ketoglutarate2– glucose H+

HCO3– H2CO3

CAII CO2 + H2O

Fig. 20.15  Ammonium secretion by the proximal tubule.

3HCO3– Na+

NBC1A

chapter 20 

but the physiological importance of these transporters in relation to acid–base balance is probably minor (Weiner and Verlander, 2011).

Major control mechanisms Chronic acidosis (of non-renal origin) enhances H+ secretion and HCO3− addition to the plasma, while chronic alkalosis results in HCO3− excretion. A  variety of mechanisms are involved. First, changes in systemic pH are paralleled by changes in intracellular pH, affecting H+ secretion directly. Second, NHE3 is directly upregulated by chronic acidosis. Third, chronic acidosis increases the trafficking of H+-ATPase to the apical membrane and that of AE1 to the basolateral membrane, whilst the activity of β-intercalated cells is downregulated. In chronic alkalosis, the opposite occurs. Finally, chronic acidosis will increase ammoniagenesis by the proximal tubule. In addition to the factors described above, acid–base disorders of respiratory origin affect intracellular PCO2 and thereby influence the generation of protons and bicarbonate within tubular cells. Stimulation of the RAAS by acidosis increases H+ secretion by both the proximal and distal nephron. Direct actions of angiotensin II on NHE3 activity are described. Aldosterone can increases H+-ATPase activity, both directly and due to increased lumen negative PD secondary to ENaC activity. The metabolic alkalosis resulting from hypokalaemia is due to the combined action of several factors. First, hypokalaemia causes a compensatory loss of K+ across the basolateral membrane and a reciprocal movement of H+ into the cell. Second, hypokalaemia stimulates renal ammoniagenesis; third, hypokalaemia increases insertion of H+-ATPase into the apical membrane of the α-intercalated cell (Bailey et al., 1998) and activates H+/K+-ATPase.

Renal handling of some specific solutes Phosphate transport (See Chapter 25.) The kidney plays a central role in phosphate homeostasis. Inorganic phosphate is filtered as HPO42− and H2PO4−, normally in the ratio 4:1. Micropuncture studies in superficial nephrons indicate that approximately 80% of the filtered phosphate is usually reclaimed in the proximal tubule. The urinary excretion normally amounts to approximately 10% of the filtered load, suggesting that a small proportion is reabsorbed beyond the proximal tubule. There is some evidence for phosphate reabsorption in the distal tubule but this is controversial and an alternative possibility is that the proximal tubules of deep nephrons that are inaccessible to micropuncture have high reabsorption rates. Proximal tubular reabsorption of phosphate is transcellular (Fig. 20.2A). Entry across the apical membrane uses one of two Na+-phosphate cotransporters, NPT2a (SLC34A1) and NPT2c (SLC34A3):  NPT2a is the major player, expression being subject to physiological regulation (Prie et al., 2011). Transport across the basolateral membrane is not defined at the molecular level, although there is functional evidence for a Na+-phosphate cotransporter, a phosphate/anion exchanger, and possibly a phosphate channel.

Control of phosphate reabsorption Dietary phosphate is a major factor in the control of absorption. A high phosphate intake lowers, and a low phosphate intake raises, the number of NPT2a cotransporters in the apical membrane.

an overview of tubular function

PTH and glucocorticoids reduce the number of NPT2a cotransporters in the apical membrane and thereby increases phosphate excretion (Biber et al., 2009). Phosphotonins are a group of phosphaturic factors (MEPE, SFRP-4, FGF-23) that inhibit reabsorption in the proximal tubule (Shirley et al., 2010) by reducing the abundance of NPT2a cotransporters in the apical membrane. The intracellular signalling events are not defined. The active metabolite of vitamin D, calcitriol (1.25 dihydroxycholecalciferol) has profound effects on phosphate homeostasis. Evidence supports both stimulation and inhibition of tubular phosphate reabsorption. Inhibitory actions might be indirect, via phosphotonin; stimulatory effects might be direct via a response element in the NPT2a promotor (Biber et al., 2009). Disturbances of acid–base balance affect phosphate excretion:  alkalosis stimulates, whilst chronic acidosis inhibits, apical Na+/phosphate co transporters, causing corresponding changes in excretion rates.

Calcium transport (See Chapter 26.) Around 60% of the filtered Ca²+ is reabsorbed in the proximal tubules, mainly in the PCT but also in the pars recta. No significant Ca²+ transport occurs in the thin descending or thin ascending limbs of Henle, owing to their low permeability to Ca²+: the TALH reabsorbs approximately 25% of the filtered load. The remaining Ca²+ reabsorption takes place in the distal tubule (~ 10% of the filtered load); very little is reabsorbed in the collecting duct (Lambers et al., 2006a). Usually 1–2% of the filtered load of Ca²+ is excreted, the actual figure being closely regulated by the requirements for overall Ca²+ balance.

Proximal tubule In the S1 segment the intratubular Ca²+ concentration increases slightly (by 10–20%), creating a small concentration gradient across the S2 epithelium. Together with the small lumen-positive transepithelial PD, this gradient is sufficient to drive passive paracellular Ca²+ reabsorption; a small proportion may be reabsorbed by solvent drag. A small component of proximal Ca²+ reabsorption is active and transcellular, but little information is available on the molecular mechanisms.

Thick ascending limb of Henle At least half the Ca²+ reabsorption in the TALH is passive and paracellular, driven by the lumen-positive transepithelial PD. Loop diuretics or Bartter syndrome abolish this gradient and are calciuric. The remainder is transcellular, most likely due to passive entry through as-yet-unidentified apical Ca²+ channels, coupled with active exit across the basolateral membrane via Ca²+-ATPase.

Distal tubule Ca²+ is reabsorbed in both the DCT and CNT exclusively through a transcellular route (Fig. 20.8). These segments express the Na+/ Ca²+ exchanger (NCX1; SLC8A1) and a Ca²+-ATPase in the basolateral membrane but the epithelial Ca2+ channel (ECaC or transient receptor potential vanilloid 5 (TRPV5) channel) in the apical membrane is rate-limiting (de Groot et al., 2008). DCT cells exhibit the highest Ca²+-ATPase activity of any nephron segment, and in the DCT region it is the sole mode of basolateral Ca²+ efflux,

185

186

Section 2  

fluid, electrolyte, and renal tubular disorders

whereas in DCT and in CNT cells the basolateral membrane also contains Na+/Ca²+ exchangers. Transcellular Ca²+ reabsorption is facilitated by calbindin D28k, an intracellular Ca²+ binding protein expressed predominantly in the DCT and CNT. By binding Ca²+, calbindins help to maintain the extremely favourable electrochemical gradient for apical entry; they are also thought to help ‘shuttle’ Ca²+ from apical to basolateral membrane via a direct interaction with TRPV5 (Lambers et al., 2006b).

Regulation of calcium reabsorption Ca²+ reabsorption in the proximal tubule is essentially unregulated; physiological control of Ca²+ excretion is exerted at the TALH and the distal tubule. The principal hormone involved in the regulation of Ca²+ reabsorption is PTH. Although PTH may have a small inhibitory effect in the proximal tubule, it more than compensates for this by stimulating Ca²+ reabsorption in the TALH (at least in the cortical segment) and distal tubule. PTH appears to exert its effect on transcellular, rather than paracellular, reabsorption by stimulating apical Ca²+ uptake and basolateral Na+/Ca+ exchange activity. The effects of the hormone calcitonin on renal Ca²+ reabsorption are somewhat paradoxical. Despite its generally hypocalcaemic action, calcitonin stimulates Ca²+ reabsorption in the TALH and distal tubule by cAMP-dependent mechanisms. Calcitriol (1.25-dihydroxycholecalciferol) targets mainly intestine and bone, but may also affect Ca²+ handling in the kidney: there is evidence that it can stimulate Ca²+ reabsorption in the distal tubule, either directly or by potentiating the effect of PTH. A Ca²+/Mg²+ sensing receptor (see above) in the basolateral membrane of the TALH and distal tubule is activated by increased plasma concentrations and inhibits reabsorption of these cations. Renal Ca²+ excretion is influenced by acid–base status: acidosis increases, and alkalosis reduces, Ca²+ excretion rates. Although this can be attributed partly to changes in the filtered load of Ca²+ (acidosis reduces the proportion of plasma calcium bound to albumin and thereby increases ultrafilterable Ca²+), and partly to non-specific changes in proximal tubular reabsorption, these effects cannot account fully for the phenomenon. A specific inhibitory effect of acidosis on Ca²+ reabsorption in the distal tubule has been documented, and it is thought that the mechanism might involve pH sensitivity of the distal tubular apical Ca²+channel.

Magnesium transport (See Chapter 27.) Mg²+ is largely an intracellular ion, found mostly in bone; only approximately 1% is in extracellular fluid. Typically, approximately 10 mmol of Mg2+ is consumed per day, with approximately 6 mmol being lost in the faeces and the remaining approximately 4% in the urine. In contrast to the situation with Ca2+, bone does not appear to play a role in the control of Mg2+ levels, and there is little evidence for control of intestinal uptake, so Mg2+ balance depends entirely on the regulation of urinary excretion. Approximately 75% of serum Mg2+ is ultrafilterable, and overall tubular Mg2+ reabsorption usually amounts to approximately 97% of the filtered load. Intriguingly, the major site of reabsorption is the loop of Henle: only 15–20% of filtered Mg2+ is reabsorbed in the proximal tubule, whereas up to 70% is reabsorbed in the loop, mainly in the TALH; the remaining 10–15% is reabsorbed in the distal tubule.

There is no evidence for transcellular Mg2+ transport in either the proximal tubule or the TALH. In the proximal tubule S2 segment, both concentration and electrical gradients favour paracellular Mg2+ reabsorption, but the permeability is low. Mg²+ reabsorption is driven by the lumen-positive transepithelial voltage in the TALH (Fig. 20.6). Recent evidence indicates that the integrity of the cation-selective paracellular pathway is dependent on the interaction of at least two tight-junction proteins: claudin-16 and claudin-19 (Hou and Goodenough, 2010). Mutations in the gene for either protein result in the syndrome of familial hypomagnesaemia with hypercalciuria and nephrocalcinosis (Haisch et al., 2011). The final nephron segment in which Mg2+ is reabsorbed is the distal tubule, where its transport is transcellular (Ferre et al., 2011). Entry across the apical membrane may be mediated by the transient receptor potential channel melastatin (TRPM), members 6 and 7, but the concentration gradient for entry is much smaller than that for calcium. It is not yet known whether an intracellular binding protein exists for Mg2+; nor has the mechanism of exit across the basolateral membrane been identified, though it is usually assumed to be a Na+/Mg²+ exchanger (San-Cristobal et al., 2010). Clearly, anything inhibiting normal reabsorptive processes in the TALH (e.g. loop diuretics) is likely to be magnesiuric, and anything stimulating reabsorption in the TALH (e.g. peptide hormones) antimagnesiuric, but the factors controlling Mg2+ excretion are ill understood. It appears that epidermal growth factor (EGF), released locally, has a stimulatory effect on TRPM6 activity, though how (or if) this is regulated remains to be determined. Other factors that can influence Mg2+ reabsorption in the distal tubule include the apical K+ channel Kv1.1 and the basolateral K+ channel Kir4.1 (KCNJ10), the latter often found as a heteromeric complex with Kir5.1. Mutations in the gene encoding Kv1.1 are associated with inappropriately high Mg2+ excretion. This channel normally hyperpolarizes the apical membrane, so its dysfunction reduces the driving force for Mg2+ entry (San-Cristobal et al., 2010). Mutations that disable the basolateral Kir4.1/5.1 channel, by preventing K+ recycling, reduce the activity of the basolateral Na+/K+-ATPase and also result in hypomagnesaemia, though in this case the explanation is elusive.

References Ackermann, D., Gresko, N., Carrel, M., et al. (2010). In vivo nuclear translocation of mineralocorticoid and glucocorticoid receptors in rat kidney: differential effect of corticosteroids along the distal tubule. Am J Physiol Renal Physiol, 299, F1473–85. Amorim, J. B., Bailey, M. A., Musa-Aziz, R., et al. (2003). Role of luminal anion and pH in distal tubule potassium secretion. Am J Physiol Renal Physiol, 284, F381–8. Ares, G. R., Caceres, P. S., and Ortiz, P. A. (2011). Molecular regulation of NKCC2 in the thick ascending limb. Am J Physiol Renal Physiol, 301, F1143–59. Aronson, P. S. (2006). Essential roles of CFEX-mediated Cl(-)-oxalate exchange in proximal tubule NaCl transport and prevention of urolithiasis. Kidney Int, 70, 1207–13. Ashek, A., Menzies, R. I., Mullins, L. J., et al. (2012). Activation of thiazide-sensitive co-transport by angiotensin II in the cyp1a1-Ren2 hypertensive rat. PLoS One, 7, e36311. Bailey, M. A., Fletcher, R. M., Woodrow, D. F., et al. (1998). Upregulation of H+-ATPase in the distal nephron during potassium depletion: structural and functional evidence. Am J Physiol, 275, F878–84.

chapter 20 

Bailey, M. A., Giebisch, G., Abbiati, T., et al. (2004). NHE2-mediated bicarbonate reabsorption in the distal tubule of NHE3 null mice. J Physiol, 561, 765–75. Bailey, M. A., Mullins, J. J., and Kenyon, C. J. (2009). Mineralocorticoid and glucocorticoid receptors stimulate epithelial sodium channel activity in a mouse model of Cushing syndrome. Hypertension, 54, 890–6. Bailey, M. A., Paterson, J. M., Hadoke, P. W., et al. (2008). A switch in the mechanism of hypertension in the syndrome of apparent mineralocorticoid excess. J Am Soc Nephrol, 19, 47–58. Bailey, M. A. and Shirley, D. G. (2009). Effects of extracellular nucleotides on renal tubular solute transport. Purinergic Signal, 5, 473–80. Bailey, M. A., Unwin, R. J., and Shirley, D. G. (2001). In vivo inhibition of renal 11beta-hydroxysteroid dehydrogenase in the rat stimulates collecting duct sodium reabsorption. Clin Sci (Lond), 101, 195–8. Bell, P. D., Komlosi, P., and Zhang, Z. R. (2009). ATP as a mediator of macula Densa Cell Signalling. Purinergic Signal, 5, 461–71. Beltowski, J. and Wojcicka, G. (2002). Regulation of renal tubular sodium transport by cardiac natriuretic peptides: two decades of research. Med Sci Monit, 8, RA39–52. Bhave, G., Chauder, B. A., Liu, W., et al. (2011). Development of a selective small-molecule inhibitor of Kir1.1, the renal outer medullary potassium channel. Mol Pharmacol, 79, 42–50. Biber, J., Hernando, N., Forster, I., et al. (2009). Regulation of phosphate transport in proximal tubules. Pflugers Arch, 458, 39–52. Bockenhauer, D., Feather, S., Stanescu, H. C., et al. (2009). Epilepsy, ataxia, sensorineural deafness, tubulopathy, and KCNJ10 mutations. N Engl J Med, 360, 1960–70. Burg, M. B., Ferraris, J. D., and Dmitrieva, N. I. (2007). Cellular response to hyperosmotic stresses. Physiol Rev, 87, 1441–74. Christensen, B. M., Perrier, R., Wang, Q., et al. (2010). Sodium and potassium balance depends on alphaENaC expression in connecting tubule. J Am Soc Nephrol, 21, 1942–51. Craigie, E., Evans, L. C., Mullins, J. J., et al. (2012). Failure to downregulate the epithelial sodium channel causes salt sensitivity in hsd11b2 heterozygote mice. Hypertension, 60, 684–90. Cupples, W. A. (2007). Interactions contributing to kidney blood flow autoregulation. Curr Opin Nephrol Hypertens, 16, 39–45. De Groot, T., Bindels, R. J., and Hoenderop, J. G. (2008). TRPV5: an ingeniously controlled calcium channel. Kidney Int, 74, 1241–6. Du, Z., Wan, L., Yan, Q., et al. (2012a). Regulation of glomerulotubular balance II: impact of angiotensin II on flow-dependent transport. Am J Physiol Renal Physiol, 303(11), F1507–16. Du, Z., Yan, Q., Wan, L., et al. (2012b). Regulation of glomerulotubular balance. I. Impact of dopamine on flow-dependent transport. Am J Physiol Renal Physiol, 303, F386–95. Endeward, V., Musa-Aziz, R., Cooper, G. J., et al. (2006). Evidence that aquaporin 1 is a major pathway for CO2 transport across the human erythrocyte membrane. FASEB J, 20, 1974–81. Esler, M. D., Krum, H., Sobotka, P. A., et al. (2010). Renal sympathetic denervation in patients with treatment-resistant hypertension (The Symplicity HTN-2 Trial): a randomised controlled trial. Lancet, 376, 1903–9. Fenton, R. A. and Knepper, M. A. (2007). Mouse models and the urinary concentrating mechanism in the new millennium. Physiol Rev, 87, 1083–112. Ferre, S., Hoenderop, J. G., and Bindels, R. J. (2011). Insight into renal Mg2+ transporters. Curr Opin Nephrol Hypertens, 20, 169–76. Filiopoulos, V., Hadjiyannakos, D., and Vlassopoulos, D. (2012). New insights into uric acid effects on the progression and prognosis of chronic kidney disease. Ren Fail, 34, 510–20. Fischbarg, J. (2010). Fluid transport across leaky epithelia: central role of the tight junction and supporting role of aquaporins. Physiol Rev, 90, 1271–90. Funder, J. W., Pearce, P. T., Smith, R., et al. (1988). Mineralocorticoid action: target tissue specificity is enzyme, not receptor, mediated. Science, 242, 583–5.

an overview of tubular function

Gamba, G. (2009). The thiazide-sensitive Na+-Cl− cotransporter: molecular biology, functional properties, and regulation by WNKs. Am J Physiol Renal Physiol, 297, F838–48. Gamba, G. and Friedman, P. A. (2009). Thick ascending limb: the Na(+):K (+):2Cl (-) co-transporter, NKCC2, and the calcium-sensing receptor, CaSR. Pflugers Arch, 458, 61–76. Garvin, J. L., Herrera, M., and Ortiz, P. A. (2011). Regulation of renal NaCl transport by nitric oxide, endothelin, and ATP: clinical implications. Annu Rev Physiol, 73, 359–76. Gimpl, G. and Fahrenholz, F. (2001). The oxytocin receptor system: structure, function, and regulation. Physiol Rev, 81, 629–83. Green, R. and Giebisch, G. (1989). Osmotic forces driving water reabsorption in the proximal tubule of the rat kidney. Am J Physiol, 257, F669–75. Groger, N., Frohlich, H., Maier, H., et al. (2010). SLC4A11 prevents osmotic imbalance leading to corneal endothelial dystrophy, deafness, and polyuria. J Biol Chem, 285, 14467–74. Grossman, E. B. and Hebert, S. C. (1988). Modulation of Na-K-ATPase activity in the mouse medullary thick ascending limb of Henle. Effects of mineralocorticoids and sodium. J Clin Invest, 81, 885–92. Haisch, L., Almeida, J. R., Abreu da Silva, P. R., et al. (2011). The role of tight junctions in paracellular ion transport in the renal tubule: lessons learned from a rare inherited tubular disorder. Am J Kidney Dis, 57, 320–30. Hall, A. M., Unwin, R. J., Parker, N., et al. (2009). Multiphoton imaging reveals differences in mitochondrial function between nephron segments. J Am Soc Nephrol, 20, 1293–302. Hamilton, K. L. and Devor, D. C. (2012). Basolateral membrane K+ channels in renal epithelial cells. Am J Physiol Renal Physiol, 302, F1069–81. Hou, J. and Goodenough, D. A. (2010). Claudin-16 and claudin-19 function in the thick ascending limb. Curr Opin Nephrol Hypertens, 19, 483–8. Hunter, R. W., Ivy, J. R., and Bailey, M. A. (2014). Glucocorticoids and renal sodium transport: implications for hypertension and salt-sensitivity. J Physiol, 592, 1731–44. Ivy, J. R. and Bailey, M. A. (2014). Pressure natriuresis and the renal control of arterial blood pressure. J. Physiol, 592, 3955–67. Jentsch, T. J. (2005). Chloride transport in the kidney: lessons from human disease and knockout mice. J Am Soc Nephrol, 16, 1549–61. Jonker, J. W. and Schinkel, A. H. (2004). Pharmacological and physiological functions of the polyspecific organic cation transporters: OCT1, 2, and 3 (SLC22A1-3). J Pharmacol Exp Ther, 308, 2–9. Kohan, D. E., Rossi, N. F., Inscho, E. W., et al. (2011). Regulation of blood pressure and salt homeostasis by endothelin. Physiol Rev, 91, 1–77. Kopp, U. C. (2011). Endothelin in the control of renal sympathetic nerve activity. Contrib Nephrol, 172, 107–19. Kurtz, A. (2011). Renin release: sites, mechanisms, and control. Annu Rev Physiol, 73, 377–99. Lambers, T. T., Bindels, R. J., and Hoenderop, J. G. (2006a). Coordinated control of renal Ca2+ handling. Kidney Int, 69, 650–4. Lambers, T. T., Mahieu, F., Oancea, E., et al. (2006b). Calbindin-D28K dynamically controls TRPV5-mediated Ca2+ transport. EMBO J, 25, 2978–88. Lara, L. S., Satou, R., Bourgeois, C. R., et al. (2012). The sodium-activated sodium channel is expressed in the rat kidney thick ascending limb and collecting duct cells and is upregulated during high salt intake. Am J Physiol Renal Physiol, 303, F105–9. Leviel, F., Hubner, C. A., Houillier, P., et al. (2010). The Na+-dependent chloride-bicarbonate exchanger SLC4A8 mediates an electroneutral Na+ reabsorption process in the renal cortical collecting ducts of mice. J Clin Invest, 120, 1627–35. Liu, W., Morimoto, T., Kondo, Y., et al. (2002). Analysis of NaCl transport in thin ascending limb of Henle’s loop in CLC-K1 null mice. Am J Physiol Renal Physiol, 282, F451–7. Lu, M., Wang, T., Yan, Q., et al. (2004). ROMK is required for expression of the 70-pS K channel in the thick ascending limb. Am J Physiol Renal Physiol, 286, F490–5. Malnic, G., Giebisch, G., Muto, S., et al. (2013). Regulation of K+ excretion. In R. J. Alpern, M. J. Caplan, O. W. and Moe (eds.) Seldin & Gieibsch’s The Kidney, pp. 1659–715. St Louis, MO: Elsevier, Inc.

187

188

Section 2  

fluid, electrolyte, and renal tubular disorders

Menzies, R. I., Unwin, R. J., and Bailey M. A. (2015). Renal P2 receptors and hypertension. Acta Physiol, 213(1), 232–41. Nesterov, V., Dahlmann, A., Krueger, B., et al. (2012). Aldosterone dependent and independent regulation of the epithelial sodium channel (ENaC) in mouse distal nephron. Am J Physiol Renal Physiol, 303(9), F1289–99. Nishino, M., Morimoto, T., Nishio, T., et al. (2007). Gestational length affects a change in the transepithelial voltage and the rNKCC2 expression pattern in the ascending thin limb of Henle’s loop. Pediatr Res, 61, 171–5. Palmer, L. G., Patel, A., and Frindt, G. (2012). Regulation and dysregulation of epithelial Na+ channels. Clin Exp Nephrol, 16, 35–43. Pannabecker, T. L., Brokl, O. H., Kim, Y. K., et al. (2002). Regulation of intracellular pH in rat renal inner medullary thin limbs of Henle’s loop. Pflugers Arch, 443, 446–57. Prie, D., Torres, P. U., and Friedlander, G. (2011). Phosphate handling: new genes, new molecules. Horm Res Paediatr, 76 Suppl 1, 71–5. Pruitt, M. E., Knepper, M. A., Graves, B., et al. (2006). Effect of peristaltic contractions of the renal pelvic wall on solute concentrations of the renal inner medulla in the hamster. Am J Physiol Renal Physiol, 290, F892–6. Rafiqi, F. H., Zuber, A. M., Glover, M., et al. (2010). Role of the WNK-activated SPAK kinase in regulating blood pressure. EMBO Mol Med, 2, 63–75. Ramkumar, N. and Kohan, D. E. (2013). Proximal tubule angiotensinogen modulation of arterial pressure. Curr Opin Nephrol Hypertens, 22(1), 32–6. Rampoldi, L., Scolari, F., Amoroso, A., et al. (2011). The rediscovery of uromodulin (Tamm-Horsfall protein): from tubulointerstitial nephropathy to chronic kidney disease. Kidney Int, 80, 338–47. Ramseyer, V. D., Cabral, P. D., and Garvin, J. L. (2011). Role of endothelin in thick ascending limb sodium chloride transport. Contrib Nephrol, 172, 76–83. Rizwan, A. N. and Burckhardt, G. (2007). Organic anion transporters of the SLC22 family: biopharmaceutical, physiological, and pathological roles. Pharm Res, 24, 450–70. Rossier, B. C. and Stutts, M. J. (2009). Activation of the epithelial sodium channel (ENaC) by serine proteases. Annu Rev Physiol, 71, 361–79. Rubera, I., Loffing, J., Palmer, L. G., et al. (2003). Collecting duct-specific gene inactivation of alphaENaC in the mouse kidney does not impair sodium and potassium balance. J Clin Invest, 112, 554–65. San-Cristobal, P., Dimke, H., Hoenderop, J. G., et al. (2010). Novel molecular pathways in renal Mg2+ transport: a guided tour along the nephron. Curr Opin Nephrol Hypertens, 19, 456–62. Seyberth, H. W. and Schlingmann, K. P. (2011). Bartter- and Gitelman-like syndromes: salt-losing tubulopathies with loop or DCT defects. Pediatr Nephrol, 26, 1789–802. Shirley, D. G., Bailey, M. A., Wildman, S. S. P., et al. (2013). Extracellular nucelotides and renal function. In R. J. Alpern, M. J. Caplan, and O. W. Moe (eds.) Seldin & Gieibsch’s The Kidney, pp. 511–37. St Louis, MO: Elsevier, Inc. Shirley, D. G., Faria, N. J., Unwin, R. J., et al. (2010). Direct micropuncture evidence that matrix extracellular phosphoglycoprotein inhibits proximal tubular phosphate reabsorption. Nephrol Dial Transplant, 25, 3191–5. Shirley, D. G., Vekaria, R. M., and Sevigny, J. (2009). Ectonucleotidases in the kidney. Purinergic Signal, 5, 501–11.

Shirley, D. G., Walter, S. J., Unwin, R. J., et al. (1998). Contribution of Na+H+ exchange to sodium reabsorption in the loop of Henle: a microperfusion study in rats. J Physiol, 513 (Pt 1), 243–9. Sindic, A., Chang, M. H., Mount, D. B., et al. (2007). Renal physiology of SLC26 anion exchangers. Curr Opin Nephrol Hypertens, 16, 484–90. Sohara, E., Uchida, S., and Sasaki, S. (2009). Function of aquaporin-7 in the kidney and the male reproductive system. Handb Exp Pharmacol, 290, 219–31. Stockand, J. D. (2010). Vasopressin regulation of renal sodium excretion. Kidney Int, 78, 849–56. Stow, L. R. and Gumz, M. L. (2012). The circadian clock in the kidney. J Am Soc Nephrol, 22, 598–604. Torp, M., Brond, L., Nielsen, J. B., et al. (2012). Effects of renal denervation on the NKCC2 cotransporter in the thick ascending limb of the loop of Henle in rats with congestive heart failure. Acta Physiol (Oxf), 204, 451–9. Turban, S., Wang, X. Y., and Knepper, M. A. (2003). Regulation of NHE3, NKCC2, and NCC abundance in kidney during aldosterone escape phenomenon: role of NO. Am J Physiol Renal Physiol, 285, F843–51. Vallon, V., Eraly, S. A., Wikoff, W. R., et al. (2008). Organic anion transporter 3 contributes to the regulation of blood pressure. J Am Soc Nephrol, 19, 1732–40. Vallon, V. and Thomson, S. C. (2012). Renal function in diabetic disease models: the tubular system in the pathophysiology of the diabetic kidney. Annu Rev Physiol, 74, 351–75. Vallon, V., Verkman, A. S., and Schnermann, J. (2000). Luminal hypotonicity in proximal tubules of aquaporin-1-knockout mice. Am J Physiol Renal Physiol, 278, F1030–3. Van Der Lubbe, N., Lim, C. H., Fenton, R. A., et al. (2011). Angiotensin II induces phosphorylation of the thiazide-sensitive sodium chloride cotransporter independent of aldosterone. Kidney Int, 79, 66–76. Velazquez, H., Bartiss, A., Bernstein, P., et al. (1996). Adrenal steroids stimulate thiazide-sensitive NaCl transport by rat renal distal tubules. Am J Physiol, 270, F211–9. Wadei, H. M. and Textor, S. C. (2012). The role of the kidney in regulating arterial blood pressure. Nat Rev Nephrol, 8, 602–9. Wagner, C. A. (2008). How much is blood pressure in the general population determined by rare mutations in renal salt-transporting proteins? J Nephrol, 21, 632–4. Wagner, C. A., Devuyst, O., Belge, H., et al. (2011). The rhesus protein RhCG: a new perspective in ammonium transport and distal urinary acidification. Kidney Int, 79, 154–61. Walter, M., Unwin, R., Nortier, J., et al. (1997). Enhancing endogenous effects of natriuretic peptides: inhibitors of neutral endopeptidase (EC.3.4.24.11) and phosphodiesterase. Curr Opin Nephrol Hypertens, 6, 468–73. Weinbaum, S., Duan, Y., Satlin, L. M., et al. (2010). Mechanotransduction in the renal tubule. Am J Physiol Renal Physiol, 299, F1220–36. Weiner, I. D. and Verlander, J. W. (2011). Role of NH3 and NH4+ transporters in renal acid-base transport. Am J Physiol Renal Physiol, 300, F11–23. Zhang, M. Z., Yao, B., Yang, S., et al. (2012). Intrarenal dopamine inhibits progression of diabetic nephropathy. Diabetes, 61, 2575–84.

CHAPTER 21

Sodium transport and balance: a key role for the distal nephron Laurent Schild Introduction (Na+)

Sodium is the most important contributor to the osmolality of the extracellular fluid (ECF), and hence is a major determinant of the ECF volume. The kidneys constantly maintain the ECF volume by regulating urinary Na+ excretion. The urinary Na+ excretion has to precisely balance the daily Na+ intake to avoid changes in ECF volume. The intravascular volume of the ECF is critically dependent on changes in body Na+ content and represents an important determinant of systemic blood pressure. The arterial pressure of a normal adult is kept within a narrow range and rarely deviates by > 10–20%, even between diurnal and nocturnal periods. A number of pressure control systems are necessary to maintain such constancy in blood pressure; they include baroreceptors and neural reflex systems that respond within seconds or minutes to abrupt changes in blood pressure. This short-term regulation of blood pressure relies mainly on the heart, the blood vessels, and the adrenal medulla. The kidney maintains blood pressure within hours or days by controlling the ECF volume. A.  C. Guyton first described this renal control system for the long-term regulation of blood pressure, and postulated the existence of a unique mean arterial blood pressure called the equilibrium pressure, the pressure at which Na+ intake and output are in balance (Guyton, 1992). If arterial pressure rises above the equilibrium pressure, then the urinary Na+ excretion becomes greater than the net Na+ intake, and circulating volume decreases until pressure returns to equilibrium. This pressure control never stops functioning to balance Na+ intake and output to maintain blood pressure at equilibrium. Of course, this leaves unexplained the reasons and the mechanisms that cause blood pressure to rise in the majority of the hypertensive patients. Elevated blood pressure may result from a high salt intake that exceeds the ability of the kidneys to eliminate Na+. The prolonged ingestion of large quantities of salt increases blood pressure in the dog, rabbit, baboon, and chimpanzee. In humans, a relation between salt intake and blood pressure is well documented from epidemiological studies (Denton et  al., 1995; Elliott et  al., 1996; Meneton et al., 2005). The effect of Na+ on blood pressure is certainly complex and may include factors other than the Na+ handling by the kidney. However, this chapter will focus mainly on regulated Na+ absorption along the nephron, which influences Na+ homeostasis and the maintenance of blood pressure.

Sodium handling by the nephron (See Chapter 20.) More than 99% of the filtered load of Na+ is reabsorbed along the nephron and in the collecting tubule. The major fraction (90%) of the filtered Na+ is reabsorbed along the proximal tubule and in the thick ascending limb (TAL). The remaining 8–10% of the filtered Na+ is reabsorbed in the distal convoluted tubule (DCT), the connecting tubule (CNT), and the collecting duct; the Na+ absorption in the distal nephron and the collecting duct is tightly regulated by aldosterone and vasopressin. In the proximal tubule, the electroneutral Na+/hydrogen (H+) exchanger links Na+ reabsorption to that of bicarbonate. In addition, Na+ absorption is coupled with the uptake of solutes such as glucose, amino acids, phosphate, sulphate, and lactate by different cotransporter systems. The thin descending limb of Henle is impermeable to Na+ ions, but the TAL contributes to the reabsorption of 20–30% of the filtered Na+. Two major transport pathways contribute to the Na+ absorption in this segment, the electroneutral Na-K-2Cl cotransporter and the Na+/H+ exchanger (Fig. 21.1). A particularity of the TAL is the presence of a lumen positive electrical potential generated by the recycling of potassium (K+) across the apical membrane; because of the selective cationic permeability of the tight junctions in the TAL, this lumen positive potential contributes to approximately 50% of Na+ absorption along a favourable driving force for diffusion of Na+ and divalent cations through the paracellular pathway. In the DCT, Na+ absorption is mediated by the electroneutral Na-Cl cotransporter (NCC) that is specifically expressed in the apical membrane of this nephron segment. Further downstream, the epithelial Na+ channel (ENaC) is responsible for electrogenic Na+ absorption. In the late portion of the distal DCT (DCT2), the expression of NCC and ENaC overlap. Further downstream in the CNT, the cortical (CCD) and medullary portions of the collecting duct (MCD), ENaC is found without NCC, and its expression follows a decreasing axial gradient from the CNT down to the MCD (Loffing and Kaissling, 2003). ENaC allows the electrogenic entry of Na+ into the cell along a favourable electrochemical gradient (Fig. 21.1). The resulting depolarization of the apical membrane provides the driving force for K+ secretion through the K+ channels (ROMK) that co-localize with ENaC at the apical membrane of principal cells (Loffing and Kaissling, 2003; Loffing and Korbmacher, 2009). Thus, in the CNT and CCD, the electrogenic Na+ absorption mediated by

Section 2  

fluid, electrolyte, and renal tubular disorders The distal convoluted tubule Thiazides

Na+

+ K+ K

CI–

MR

P

Na+

CI–

The cortical collecting duct The thick ascending limb Furosemide

P

Amiloride

CI–

Spironolactone

ENaC

Na+ ROMK

K+

+ + K+

ROMK

+

Na+

Ca++ Mg++

MR

K+

P

Na+

K+

–

K+

CI –

Na+

Na+ 2 CI– K+

K+ K+

K+

190

CI–

Fig. 21.1  Cellular mechanisms contributing to Na+ absorption in the thick ascending limb (red) the distal convoluted tubule (grey), the connecting tubule, and the collecting duct (yellow). In the thick ascending limb, Na+ absorption is essentially mediated by the Na-K-2Cl co-transporter (NKCC); this cotransport system serves Na+ absorption and K+ recycling across the apical membrane. In the distal convoluted tubule (DCT), the electroneutral Na-Cl cotransporter (NCC) is responsible for Na+ absorption. In the principal cells of the connecting tubule and the collecting duct, Na+ absorption is electrogenic via the epithelial sodium channel (ENaC). In these segments, the negative electrical potential in the lumen provides a favourable driving force for K+ secretion. The transporters NKCC, NCC, and ENaC are the target of the diuretics furosemide, thiazides and amiloride respectively. Spironolactone is an antagonist of the mineralocorticoid receptor (MR).

ENaC is coupled with K+ secretion, and is critically dependent on whole-body K+ status (Frindt et al., 2011). Although it represents only a small fraction of the filtered load, the Na+ absorption in the distal nephron and the collecting tubule is under the tight control of the renin–angiotensin–aldosterone system (RAAS). Following a reduction of the effective circulating volume and/or a dietary salt restriction, the RAAS is activated, leading to a decrease in urinary Na+ excretion. Aldosterone increases Na+ absorption in the DCT, CNT, and CCD, where it increases the abundance of both transporters NCC and ENaC at the apical membrane of principal cells (Velazquez et al., 1996; Kim et al., 1998). Angiotensin II also enhances Na+ absorption in the DCT (Wang and Giebisch, 1996). In addition, arginine vasopressin (AVP) stimulates Na+ absorption in the DCT, CNT, and CCD (Ecelbarger et al., 2000; Pedersen et al., 2010). These nephron segments share in common the expression of the mineralocorticoid receptor (MR), the vasopressin receptor V2R, and the enzyme 11-β hydroxysteroid dehydrogenase type 2 (Bostanjoglo et al., 1998; Bachmann et al., 1999; Mutig et al., 2007). The understanding of the detailed molecular and cellular mechanisms involved in the regulation of Na+ excretion by the kidney has greatly progressed with the identification of the genetic basis of

Mendelian disorders featuring alterations in Na+ homeostasis and elevated blood pressure.

Human diseases The importance of the RAAS in controlling renal Na+ excretion, Na+ balance, and blood pressure is largely supported by use of drugs that lower blood pressure such as diuretics, angiotensin converting enzyme inhibitors, or angiotensin II receptor antagonists. Beside this pharmacological evidence, recent genetic studies have identified renal and adrenal genes responsible for monogenic forms of hypertension. The glucocorticoid-remediable aldosteronism (GRA) and the syndrome of apparent mineralocorticoid excess (AME) are two examples of Mendelian forms of elevated blood pressure (Lifton, 1996). GRA is a defect of the regulated synthesis of aldosterone associated with severe hypertension and elevated plasma aldosterone levels. The genetic defect is a gene duplication of the aldosterone synthase and 11β-hydroxylase, generating a novel gene that places the synthesis of aldosterone under the control of corticotrophin (ACTH), instead of its physiological secretagogue angiotensin II. The high plasma levels of aldosterone can be normalized by suppression of ACTH release with glucocorticoids.

chapter 21 

AME is an autosomal recessive disorder due unphysiological stimulation of the MR by cortisol. This disorder is characterized by a moderate to severe hypertension with very low plasma aldosterone levels, but sensitive to MR antagonists such as spironolactone. The genetic defects of the AME syndrome are loss of function mutations in 11β-hydroxysteroid-dehydrogenase, a key enzyme present in aldosterone-sensitive kidney cells that protects the MR from stimulation by cortisol (Stewart et al., 1987). The MR has a high affinity for cortisol, which is present in the plasma at a higher concentration than aldosterone. Other Mendelian disorders have established the critical role of the NCC or ENaC in the maintenance of salt balance. Pseudohypoaldosteronism type II (PHA-II also named familial hyperkalaemic hypertension or Gordon syndrome) is characterized by a salt-sensitive hypertension exquisitely sensitive to thiazide diuretics, associated with hyperkalaemia, and hyperchloraemic acidosis (Wilson et al., 2001). These clinical features strongly suggest that this disorder is due to an overactivity of the NCC in the DCT. Genome-wide linkage studies showed that PHA-II is genetically heterogeneous: they identified overlapping deletions in chromosome 1 located in the first intron of a gene encoding a serine-threonine protein kinase WNK1. Expression studies of WNK1 in lymphocytes of affected patients showed a fivefold increase in WNK1 expression compared with unaffected patients, suggesting a gain-of-function mutation of WNK1 (Wilson et al., 2001). The WNK1 gene encodes two WNK1 isoforms, a ubiquitous long isoform L-WNK1, and a kidney-specific KS-WNK1, expressed predominantly in the DCT, and missing most of the kinase domain. In another PHA-II locus on chromosome 17, mutations were found in a gene encoding a WNK isoform, WNK4; these mutations target a highly conserved segment of negatively charged residues distal to the kinase domain. Although strong genetic evidence supports the notion that the mutations in the WNK4 gene are the cause in PHA-II, there is no clear evidence that favours loss- or gain-of-function mutations of WNK4 kinase. WNK4 is localized in the aldosterone-sensitive distal nephron at the tight junctions and at the subapical membrane region of the DCT, in the cytoplasm of CNT and CCD (Wilson et al., 2001; Kahle et al., 2004). These genetic studies strongly suggest that the WNK1 and WNK4 kinases are important regulators of NCC in the distal nephron. Liddle syndrome (or pseudoaldosteronism) is an autosomal dominant form of salt-sensitive hypertension associated with low plasma aldosterone, low plasma renin activity, hypokalaemia, and metabolic alkalosis. In the early 1960s, G. W. Liddle reported a case of pseudoaldosteronism, and described the syndrome as ‘a disorder in which the renal tubules transport ions with such abnormal facility that the end result simulates that of a mineralocorticoid excess’ (Liddle et al., 1963). Blood pressure in these patients could be normalized with amiloride and dietary salt restriction, but spironolactone was not effective. The genetic defects are mutations in the last exon of the SCNN1B and SCNN1G genes encoding the β and γ ENaC subunits that delete a conserved proline-rich motif in the cytosolic C-terminus (Shimkets et al., 1964). In vitro experiments could establish that these mutations are gain-of-function mutations, as postulated by G.W. Liddle, leading to an increase in the abundance and in the activity of ENaC at the cell surface (Schild et  al., 1995; Firsov et  al., 1996). The important role of NCC and ENaC in Na+ homeostasis is further demonstrated by Mendelian disorders with mirror images of Liddle syndrome or PHA-II.

sodium transport and balance

Pseudohypoaldosteronism type-1 (PHA-1) is a rare disease of mineralocorticoid resistance associating hyponatraemia, hyperkalaemia and metabolic acidosis with high levels of plasma aldosterone. Two forms of PHA-1 have been identified:  an autosomal dominant form with usually mild symptoms restricted to the kidney, and associated with heterozygous mutations in the NR3C2 gene encoding for the MR (Chang et al., 1996; Geller et al., 1998). The generalized PHA-I form, also called autosomal recessive PHA-I, is a multisystem disorder characterized by salt wasting from the kidney, the colon, the sweat gland, and a reduced capacity to reabsorb Na+ in the airways leading to rhinorrhoea, pulmonary congestion, and recurrent pulmonary infections. Mutations have been identified in the SCNN1A, SCNN1B, and SCNN1G genes encoding for the α, β and γ ENaC subunits leading to loss of function of ENaC. Patients with Gitelman syndrome exhibit hypokalaemic alkalosis, hypocalciuria, hypomagnesaemia, and low blood pressure. Gitelman syndrome is associated with loss-of-function mutations in the SLC12A3 gene encoding the NCC transporter (Simon et al., 1996). These genetic studies confirm Guyton’s hypothesis that the kidney plays a central role in the maintenance of Na+ balance and blood pressure. In addition, the identification of the genetic basis of Mendelian forms of hypertension and salt-losing nephropathies greatly helped to identify the distal nephron and the collecting tubules as the critical sites for the fine regulation of Na+ absorption and for the maintenance of a Na+ balance.

Cellular and molecular aspects of Na+ transport in the ASDN Na+ absorption in the aldosterone-sensitive distal nephron (ASDN) and collecting duct results from the concerted activity of the NCC and the ENaC.

NCC function and regulation in vitro The NCC cotransporter belongs to the solute carrier family 12 and allows the electroneutral entry of Na+ with Cl− from the tubule lumen into the cell. NCC is the pharmacological target for the thiazide diuretics. The identification of mutations causing PHA-II in the WNK kinases led to the discovery of a novel regulatory pathway that controls NCC activity at the cell surface. In vitro co-expression experiments could demonstrate that WNK4 reduces the NCC activity; this NCC inhibition is due in part to a reduced NCC trafficking to the plasma membrane and a decrease in NCC abundance at the cell surface. WNK4 harbouring the PHA-II mutations (WNK4D561A) no longer inhibit NCC activity, leading to a hyperactivity of the transporter (Wilson et al., 2003; Yang et al., 2003). It could also be shown that WNK1 interacts with WNK4 and suppresses its inhibitory effect on NCC, resulting in an upregulation of NCC (Fig. 21.2) (Yang et al., 2005). Finally, a kidney isoform of WNK1 (KS-WNK1) lacking the kinase domain was found to be a negative regulator of WNK1, preventing inhibition of WNK4 (Delaloy et  al., 2003; Subramanya et  al., 2006). These in vitro experiments identify a kinase inhibitory pathway on NCC activity involving successively the KS-WINK1, the downstream WNK1, and the WNK4. Except for the KS-WNK1 isoform, the regulation of the NCC but also the Na-K-2Cl (NKCC) cotransporters by WNK kinases is usually dependent on their catalytic activity. However, it is still

191

192

Section 2  

fluid, electrolyte, and renal tubular disorders Urine

NCC

NCC

ENaC

ENaC Ub

P Nedd4-2

Nedd4-2

Nedd4-2 SPAK/OSR1

WNK4 SGK1

P

P MAPK/ERK

WNK4D561A

WNK1 Aldosterone

P

ENaC Ub

SGK1 AGT II

Aldosterone

Fig. 21.2  Regulated Na+ absorption mediated by the Na-Cl cotransporter (NCC; left) and epithelial sodium channel (ENaC; right). (Left) Phosphorylation of the NCC increases its transport activity at the cell surface. Kinases that contribute to the NCC phosphorylation include SPAK/OSR1 kinase under the control of angiotensin II, and the aldosterone-induced SGK-1 kinase. The NCC is negatively controlled by WNK4 kinase and ubiquitin-ligase Nedd4-2. (Right) ENaC activity at the apical membrane of principal cells is mainly controlled by aldosterone, which increases the number and the activity of the channel. The aldosterone-induced SGK-1 kinase plays a minor role is this regulation. The ubiquitin-ligase Nedd4-2 via a direct interaction with ENaC promotes the retrieval of active channels for the cell surface.

unknown if WNK4 directly phosphorylates NCC, and how WNK4 inhibits NCC. Serine or threonine phosphorylation in the N-terminus of NCC increases the transporter activity without changing its abundance at the cell surface (Pacheco-Alvarez et al., 2006; Glover et al., 2009). Beside the WNKs, a number of kinases in the DCT have been identified as potential candidates for NCC phosphorylation and regulation. Among them, the STE20/SPS1-related proline alanine-rich kinase (SPAK) and the oxidative stress responsive protein type 1 (OSR1) may represent the missing link between WNK4 and NCC (Fig. 21.2). Serine or threonine residues in the N- terminus of NCC or NKCC are phosphorylation sites for OSR1 and SPAK and when phosphorylated increase the activity of NCC (Piechotta et al., 2002; Moriguchi et al., 2005; Vitari et al., 2005). In vitro it could be shown that WNK1 and WNK4 phosphorylate SPAK and OSR1 kinases (Moriguchi et al., 2005; Vitari et al., 2005). Other kinases that are components of the aldosterone-signalling pathway regulate Na+ absorption in the ASDN. The serum and glucocorticoid-induced kinase1 (SGK1) is an early aldosterone-induced protein. The co-expression of SGK1 with WNK4 in heterologous expression systems decreases the inhibitory effect of WNK4 on NCC activity (Rozansky et al., 2009) (Fig. 21.2). Another potential signalling pathway for NCC regulation by aldosterone involves the ubiquitin ligase Nedd4-2 known to regulate ENaC at the cell surface. In vitro Nedd4-2 stimulates NCC ubiquitylation and reduces its activity at the cell surface. SGK1 prevents, in a kinase-dependent manner, the inhibition of NCC by Nedd4-2 (Arroyo et al., 2011).

NCC: regulation in vivo and contribution to sodium homeostasis The genetic disruption of NCC in mice leads to mild perturbations in fluid and electrolyte homeostasis; no apparent hypokalaemic alkalosis is observed, nor hypovolaemia, or change in arterial blood

pressure; however, hypocalciuria, and hypomagnesaemia, together with a sharp reduction in the number of DCT cells, could be demonstrated (Schultheis et al., 1998). To maintain Na+ balance in the absence of a functional NCC transporter, compensatory mechanisms for Na absorption need to be evoked that likely include the upregulation of NKCC in the thick ascending limb or ENaC in the CNT. The stimulation of the renin–angiotensin–aldosterone cascade by dietary Na+ restriction increases Na+ absorption in the DCT, NCC activity, and NCC phosphorylation. In mice lacking the aldosterone-induced kinase SGK1, a dietary Na+ restriction attenuates NCC expression and phosphorylation (Vallon et  al., 2009). The effects of dietary salt restriction on NCC in the DCT can potentially be mediated either by aldosterone or by angiotensin II. Angiotensin II independently of aldosterone, increases Na+ absorption in the DCT together with NCC phosphorylation, and also increases the intracellular abundance of SPAK kinase (van der Lubbe et al., 2011). Thus, this upregulation of NCC following dietary salt restriction also involves the intermediary kinases SPAK/ OSR1 (Chiga et al., 2008). The central role of SPAK in the regulation of Na+ balance and blood pressure is further supported by the study of a knock-in mouse model carrying a loss-of-function mutation in the kinase domain of SPAK; these mice exhibit marked hypotension on a normal salt diet, hypomagnesaemia, and hypocalciuria associated with reduced expression and phosphorylation of both NCC and NKCC cotransporters (Rafiqi et al., 2010). This phenotype contrasts with the mild phenotype observed in the NCC knockout mice. The physiological role of WNK4 and WNK1 in mediating the aldosterone or angiotensin II effects on NCC has not yet been clearly established in vivo. No changes in the expression of WNK1 or WNK4 proteins were observed on dietary NaCl restriction or after aldosterone or angiotensin II infusion (van der Lubbe et al., 2011).

chapter 21 

Nevertheless, WNK4 plays a central role in the regulation of NCC and in the pathogenesis of PHA-II. Mouse carrying two additional transgene copies of WNK4 wildtype, or the WNK4 mutant causing PHA-II (WNK4D561A/+), show opposite effects on blood pressure, and on K+ or Ca2+ homeostasis (Lalioti et al., 2006). The mice expressing the WNK4 mutant recapitulate the essential features of the human PHA-II, including an elevated blood pressure sensitive to thiazides, hyperkalaemia, and hypercalciuria. Mice overexpressing WNK4 wild type exhibit a discrete phenotype similar to that observed in the NCC KO mice, with a slightly lower blood pressure compared with wild type, hypercalciuria, and hypokalaemia on a low K+ diet. A knock-in transgenic mouse model, heterozygous for the mutation WNK4D561A/+ found in PHA-II, shows a constitutive activation of NCC as the primary cause of PHA-II (Yang et al., 2007). Indeed, during the first 3 months of life these mice developed an elevated blood pressure, hyperkalaemia with a reduced fractional excretion of K+, and metabolic acidosis; these changes could be corrected by the NCC inhibitor hydrochlorothiazide. These changes also correlated with increased renal expression of phosphorylated NCC at the apical membrane of DCT cells, and enhanced phosphorylation of OSR1/SPAK kinases. This model also provides further evidence for the role of OSR1/SPAK kinases in the regulation of NCC. It should be mentioned that the role of the kidney-specific KS-WNK1 kinase in the regulation of NCC remains uncertain, since the KS-WNK1−/− mouse model shows no clear phenotype similar to PHA-II (Hadchouel et al., 2010). These in vivo experiments establish the critical role of WNK4 and SPAK/OSR1 interacting kinases in the pathogenesis of PHA-II. This is further supported by the use of a triple knock-in mouse approach, demonstrating that in PHA-II associated with the WNK4D561A mutation, the enhanced NCC phosphorylation is fully dependent on the integrity of the kinase domain of SPAK and OSR1 (Chiga et al., 2011). The in vitro and in vivo experiments investigating the regulation of NCC by WNK kinases illustrate the complexity of the signalling cascades that modulate Na+ absorption in the DCT (Fig. 21.2). From in vitro experiments we learn that WNK4 negatively regulates NCC independently of SPAK/OSR1 kinases and that the WNK4D561A mutant associated with PHA-II no longer inhibit NCC. In vivo experiments show that mice carrying the WNK4D561A mutation recapitulate the essential features of PHA-II; this phenotype correlates with an increase in NCC phosphorylation and NCC expression that requires the integrity of SPAK and OSR1 kinases. To reconcile these observations, one has to postulate a dual role for WNK4: WNK4 under the control of the WNK1 kinases inhibits NCC; alternatively, WNK4 can escape WNK1 control, as does the WNK4D591A mutant, and stimulate the SPAK/OSR1 kinases to activate NCC, as postulated in PHA-II (Fig. 21.2).

ENaC function and regulation in vitro ENaC belongs to a family of cation channels called the ENaC/ degenerins ion channel family (Kellenberger and Schild, 2002). This family comprises proton-gated acid sensing ion channels expressed in the mammalian central and peripheral nervous system, or touch-sensitive ion channels (degenerins) expressed in Caenorhabditis elegans. ENaC is a heteromultimeric channel made of three homologous α, β, and γ subunits; the α-ENaC subunit is absolutely required for channel activity, and the co-expression

sodium transport and balance

of three αβγ subunits is necessary for the maximal expression of ENaC-mediated current at the cell surface (Canessa et al., 1994). ENaC is highly selective to Na+ ions, and is constitutively open when present at the apical membrane of principal cells. ENaC is the pharmacological target of the K+-sparring diuretics such as amiloride or triamterene that act as pore channel blockers. Each ENaC subunit is made of two α-helices that constitute the transmembrane part of the channel pore with the selectivity filter and the amiloride binding site; the transmembrane helices are separated by a large extracellular loop that makes up more than half of the mass of the protein. The amino- and carboxy-termini of the protein are intracellular (Canessa et al., 1994). ENaC is constitutively open and fluctuates between an open and a closed state. A number of factors influence ENaC gating. An acute increase in intracellular Na+ reduces the channel openings and channel current, likely to prevent a massive entry of Na+ ions into the cell when luminal Na+ concentration is increasing. Other intracellular factors associated with cellular stress such as a decrease in pH, increase in oxidative stress or a rise in Ca2+ ions decrease channel open probability (Palmer and Frindt, 1987; Chraibi and Horisberger, 2002; Kellenberger et  al., 2005; Anantharam et al., 2006). ENaC is activated by soluble proteases such trypsin, chymotrypsin, kallikrein, or elastase. When co-expressed with GPI-anchored proteases such as CAP-1, an orthologue of the human prostasin, or CAP-2, an orthologue of the human transmembrane protease serine 4 (TRPMSS4), ENaC activity was significantly increased (Vallet et al., 1997; Chraibi et al., 1998; Vuagniaux et al., 2000; Harris et al., 2007). In the kidney and in heterologous expression systems the αand the γ-ENaC subunit are found in two molecular forms, a highand a low-molecular-weight form, resulting from the cleavage of the subunits by proteases, possibly furin (Masilamani et al., 1999; Hughey et al., 2003; Ergonul et al., 2006; Harris et al., 2007; Frindt and Palmer, 2009). The molecular mechanisms by which proteases activate ENaC are still not completely understood. Another important way to regulate ENaC-mediated Na+ absorption in the ASDN is to control the number and the stability of ENaC channels at the cell surface. ENaC ubiquitylation is an important mechanism that determines the stability of the channel at the cell surface. Ubiquitylation is a general process that labels proteins with ubiquitin in the cell or at the cell surface and targets them for endocytosis and degradation. Nedd4-2 is a protein ubiquitin ligase that belongs to the HECT family (Rotin and Schild, 2008). It mediates their mono-ubiquitylation of β- and γ-ENaC subunits on binding specifically to conserved PY motifs present in their cytosolic C- terminus (Staub et al., 1997). The ubiquitylated ENaC at the cell surface undergoes clathrin-mediated endocytosis and degradation. In vitro Nedd4-2 efficiently suppresses ENaC activity. Mutations in the cytosolic PY motifs of ENaC subunits that are found in Liddle syndrome alter the interaction between Nedd4-2 and ENaC, leading to retention of active channels at the cell surface (Firsov et al., 1996; Schild et al., 1996). By contrast, deubiquitylation enzymes such as Usp2-45 increase ENaC activity (Ruffieux-Daidie et al., 2008). In vitro experiments support the idea that aldosterone stabilizes ENaC at the cell surface by inhibition of ENaC ubiquitylation and endocytosis. Among the aldosterone-induced proteins identified in vivo, it was found that the phosphatidylinositide 3′-kinase (PI3K)-dependent kinase SGK-1 (serum- and glucocorticoid-regulated kinase 1), increases the abundance of

193

194

Section 2  

fluid, electrolyte, and renal tubular disorders

the active ENaC at the cell surface (Verrey et  al., 2008). SGK1 phosphorylates Nedd4-2, promoting the interaction between Nedd4-2 and the 14-3-3 scaffolding proteins, preventing the Nedd4-2-dependent ubiquitylation of ENaC. As for SGK1, the 14-3-3 proteins are induced by aldosterone (Debonneville et al. 2001). In addition to the SGK1/Nedd4-2 pathway that stabilizes ENaC at the cell surface, a number of kinases have been reported to modulate ENaC activity in vitro. The Raf-1-MAPK/ERK kinases act to inhibit the cell surface expression of ENaC by stimulating the interaction between Nedd4-2 and ENaC (Nicod et al., 2002; SHI et al., 2002; Falin and Cotton, 2007). Finally, WNK4 partially inhibits ENaC activity in vitro, but not the WNK4 mutant causing PHA-II; this ENaC inhibition was independent of the active kinase domain (Ring et al., 2007a, 2007b). WNK1 increases ENaC activity via the activation of SGK that phosphorylates and inactivates Nedd4- 2 (Xu et al., 2005). These studies performed in vitro in heterologous expression systems or in cortical collecting duct cell lines identify Nedd4-2 as a critical convergence point for the regulation of ENaC at the cell surface.

ENaC regulation in vivo The physiologically most important regulator of ENaC is aldosterone. In the kidney, aldosterone increases the biosynthesis of α-ENaC, whereas the β- and the γ-ENaC subunits are constitutively expressed. The upregulation of the synthesis of α-ENaC by aldosterone leads to an increase in the expression of active ENaC channels at the cell surface. It is not yet clear whether the increase in the biosynthesis of the α-ENaC is sufficient for trafficking of the multimeric αβγ-ENaC channels at the cell surface or whether additional regulators of ENaC trafficking to the apical membrane are necessary. A variety of genetic mouse models have been generated to study the physiological and pathophysiological roles of ENaC in the maintenance of whole-body salt balance, and in the control of blood pressure. Several ENaC KO mouse models recapitulate PHA-I. The constitutive inactivation of either α or β or γ leads to a severe renal phenotype, including increased Na+ excretion, hyperkalaemia, and elevated plasma aldosterone levels (Hummler et al., 1997). Interestingly, the selective invalidation of ENaC in the different segments of the ASDN revealed the predominant role of ENaC in the CNT for the maintenance of the salt balance and K+ homeostasis. Specific ENaC knockout in only the collecting duct did not result in any alteration in the Na+ and K+ homeostasis (Rubera et al., 2003). By contrast ENaC invalidation at the distal end of the DCT and in the CNT replicated the PHA-I phenotype (Christensen et al., 2010). Mice deficient in the MR have normal prenatal development, but die soon after birth from dehydration and hyperkalaemia. This severe PHA-I phenotype of MR knockout mice further confirms the critical roles of NCC and ENaC in maintaining Na+ balance (Berger et al., 1998). The severity of this phenotype contrasts with the milder renal phenotype of mice lacking the aldosterone-induced protein SGK1 (Fejes-Toth et  al., 2008). The SGK deficient mice show higher natriuresis only on a low Na+ diet, which is not related to decreased ENaC activity (Fejes-Toth et al., 2008). Furthermore, under chronic aldosterone treatment, ENaC activity is identical in the SGK deficient and the wild-type littermates. This suggests that in vivo SGK is not essential for the long-term regulation of ENaC by aldosterone. Interestingly the SGK knockout mouse model also

reveals that other Na+ transporters such as NCC located upstream of ENaC are also likely targets of SGK-1 kinase. To gain more insight into the role of ENaC in regulating ECF volume and blood pressure, salt-sensitive mouse models of Liddle syndrome or pseudohypoaldosteronism were generated. Mice carrying Liddle’s mutation in the SCNN1B gene with deletion of the C-terminus of β-ENaC have been generated; when maintained on a high-salt diet these mice recapitulate Liddle syndrome with an elevated blood pressure, low plasma levels of aldosterone and renin, hypokalaemia, and metabolic alkalosis (Pradervand et al., 1999). The deletion of the proline-rich motif in the C-terminus of β-ENaC, and impaired interaction with Nedd4-2, does not compromise the ability of the channel to respond to aldosterone (Dahlmann et al., 2003). Electrophysiological measurements of ENaC activity in microdissected tubules from Liddle’s mice revealed that under normal dietary salt ENaC activity was not detectable, but under low dietary salt and with high plasma aldosterone, the affected mice show a dramatic increase in ENaC activity at the cell surface, which even more than in wild-type litter-mates. Thus, neither SGK1 nor Nedd4-2 appears to be critical limiting factors for the stimulation of ENaC-mediated Na+ reabsorption by aldosterone. Nedd4-2 is an important downregulator of ENaC, but also of NCC. The knock-out of the Nedd4-2 gene in mice results in an elevated blood pressure associated with low plasma aldosterone levels; effects that correlate with an increase in the expression of both ENaC and NCC (Shi et al., 2008). Thus, in vitro and in vivo experiments on ENaC regulation have not yet unambiguously identified the molecular and cellular mechanisms involved in the aldosterone-signalling pathway.

Conclusion In summary, the fine-tuning of Na+ reabsorption in the ASDN and collecting duct is critical for the maintenance of Na+ balance, ECF volume, and blood pressure. The identification of genes responsible for Mendelian forms of hypertension and the generation of transgenic mouse models, together with in vitro approaches, have provided us with an unprecedented understanding of the molecular and cellular mechanisms involved in hormonally regulated Na+ absorption in the kidney. Future research is needed to address the functional interactions between these newly identified regulatory pathways that control Na+ absorption and the transport of other ions such as K+ or calcium in these nephron segments. Another question of pathophysiological relevance raised by these recent studies on genetically modified mice models is how the kidney develops compensatory mechanisms to maintain Na+ homeostasis when one regulatory pathway for Na+ is deficient or defective.

Acknowledgement I would like to thank Olivier Staub for critically reviewing the manuscript.

References Anantharam, A., Tian, Y., and Palmer, L. G. (2006). Open probability of the epithelial sodium channel is regulated by intracellular sodium. J Physiol, 574, 333–47.

chapter 21 

Arroyo, J. P., Lagnaz, D., Ronzaud, C., et al. (2011). Nedd4-2 modulates renal Na+-Cl- cotransporter via the aldosterone-SGK1-Nedd4-2 pathway. J Am Soc Nephrol, 22, 1707–19. Bachmann, S., Bostanjoglo, M., Schmitt, R., et al. (1999). Sodium transport-related proteins in the mammalian distal nephron—distribution, ontogeny and functional aspects. Anat Embryol (Berl), 200, 447–68. Berger, S., Bleich, M., Schmid, W., et al. (1998). Mineralocorticoid receptor knockout mice: pathophysiology of Na+ metabolism. Proc Natl Acad Sci U S A, 95, 9424–29. Bostanjoglo, M., Reeves, W. B., Reilly, R. F., et al. (1998). 11Beta-hydroxysteroid dehydrogenase, mineralocorticoid receptor, and thiazide-sensitive Na-Cl cotransporter expression by distal tubules. J Am Soc Nephrol, 9, 1347–58. Canessa, C. M., Schild, L., Buell, G., et al. (1994). Amiloride-sensitive epithelial Na+ channel is made of three homologous subunits. Nature, 367, 463–7. Chang, S. S., Grunder, S., Hanukoglu, A., et al. (1996). Mutations in subunits of the epithelial sodium channel cause salt wasting with hyperkalaemic acidosis, pseudohypoaldosteronism type 1. Nat Genet, 12, 248–53. Chiga, M., Rafiqi, F. H., Alessi, D. R., et al. (2011). Phenotypes of pseudohypoaldosteronism type II caused by the WNK4 D561A missense mutation are dependent on the WNK-OSR1/SPAK kinase cascade. J Cell Sci, 124, 1391–5. Chiga, M., Rai, T., Yang, S. S., et al. (2008). Dietary salt regulates the phosphorylation of OSR1/SPAK kinases and the sodium chloride cotransporter through aldosterone. Kiney Int, 74, 1403–9. Chraibi, A. and Horisberger, J. D. 92002). Na self inhibition of human epithelial Na channel: temperature dependence and effect of extracellular proteases. J Gen Physiol, 120, 133–45. Chraibi, A., Vallet, V., Firsov, D., et al. (1998). Protease modulation of the activity of the epithelial sodium channel expressed in Xenopus oocytes. J Gen Physiol, 111, 127–38. Christensen, B. M., Perrier, R., Wang, Q., et al. (2010). Sodium and potassium balance depends on alphaENaC expression in connecting tubule. J Am Soc Nephrol, 21, 1942–51. Dahlmann, A., Pradervand, S., Hummler, E., et al. (2003). Mineralocorticoid regulation of epithelial Na+ channels is maintained in a mouse model of Liddle’s syndrome. Am J Physiol Renal Physiol, 285, F310–18. Debonneville, C., Flores, S.Y., Kamynina, E., et al. (2001). Phosphorylation of Nedd4-2 by Sgk1 regulates epithelial Na+ channel cell surface expression. EMBO J, 20, 7052–9. Delaloy, C., Lu, J., Houot, A. M., et al. (2003). Multiple promoters in the WNK1 gene: one controls expression of a kidney-specific kinase—defective isoform. Mol Cell Biol, 23, 9208–21. Denton, D., Weisinger, R., Mundy, N. I., et al. (1995). The effect of increased salt intake on blood pressure of chimpanzees. Nature Med, 1, 1009–16. Ecelbarger, C. A., Kim, G. H., Terris, J., et al. (2000). Vasopressin-mediated regulation of epithelial sodium channel abundance in rat kidney. Am J Physiol Renal Physiol, 279, F46–53. Elliott, P., Stamler, J., Nichols, R., et al. (1996). Intersalt revisited: further analyses of 24 hour sodium excretion and blood pressure within and across populations. Intersalt Cooperative Research Group. BMJ, 312, 1249–53. Ergonul, Z., Frindt, G., and Palmer, L. G. (2006). Regulation of maturation and processing of ENaC subunits in the rat kidney. Am J Physiol Renal Physiol, 291, F683–93. Falin, R. A. and Cotton, C. U. (2007). Acute downregulation of ENaC by EGF involves the PY motif and putative ERK phosphorylation site. J Gen Physiol, 130, 313–28. Fejes-Toth, G., Frindt, G., Naray-Fejes-Toth, A., et al. (2008). Epithelial Na+ channel activation and processing in mice lacking SGK1. Am J Physiol Renal Physiol, 294, F1298–305. Firsov, D., Schild, L., Gautschi, I., et al. (1996). Cell surface expression of the epithelial Na channel and a mutant causing Liddle syndrome: a quantitative approach. Proc Natl Acad Sci U S A, 93, 15370–5.

sodium transport and balance

Frindt, G., Houde, V., and Palmer, L. G. (2011). Conservation of Na+ vs. K+ by the rat cortical collecting duct. Am J Physiol Renal Physiol, 301, F14–20. Frindt, G. and Palmer, L. G. (2009). Surface expression of sodium channels and transporters in rat kidney: effects of dietary sodium. Am J Physiol Renal Physiol, 297, F1249–55. Geller, D. S., Rodriguez-Soriano, J., Vallo Boado, A., et al. (1998). Mutations in the mineralocorticoid receptor gene cause autosomal dominant pseudohypoaldosteronism type I. Nat Genet, 19, 279–81. Glover, M., Zuber, A. M., and O’Shaughnessy, K. M. (2009). Renal and brain isoforms of WNK3 have opposite effects on NCCT expression. J Am Soc Nephrol, 20, 1314–22. Guyton, A. C. (1992). Kidneys and fluids in pressure regulation. Small volume but large pressure changes. Hypertension, 19, I2–8. Hadchouel, J., Soukaseum, C., Busst, C., et al. (2010). Decreased ENaC expression compensates the increased NCC activity following inactivation of the kidney-specific isoform of WNK1 and prevents hypertension. Proc Natl Acad Sci U S A, 107, 18109–14. Harris, M., Firsov, D., Vuagniaux, G., et al. (2007). A novel neutrophil elastase inhibitor prevents elastase activation and surface cleavage of the epithelial sodium channel expressed in Xenopus laevis oocytes. J Biol Chem, 282, 58–64. Hughey, R. P., Mueller, G. M., Bruns, J. B., et al. (2003). Maturation of the epithelial Na+ channel involves proteolytic processing of the alpha- and gamma-subunits. J Biol Chem, 278, 37073–82. Hummler, E., Barker, P., Talbot, C., et al. (1997). A mouse model for the renal salt- wasting syndrome pseudohypoaldosteronism. Proc Natl Acad Sci U S A, 94, 11710–15. Kahle, K. T., Gimenez, I., Hassan, H., et al. (2004). WNK4 regulates apical and basolateral Cl- flux in extrarenal epithelia. Proc Natl Acad Sci U S A, 101, 2064–9. Kellenberger, S., Gautschi, I., Pfister, Y., et al. (2005). Intracellular thiol-mediated modulation of epithelial sodium channel activity. J Biol Chem, 280, 7739–47. Kellenberger, S. and Schild, L. (2002). Epithelial sodium channel/degenerin family of ion channels: a variety of functions for a shared structure. Physiol Rev, 82, 735–67. Kim, G. H., Masilamani, S., Turner, R., et al. (1998). The thiazide-sensitive Na-Cl cotransporter is an aldosterone-induced protein. Proc Natl Acad Sci U S A, 95, 14552–7. Lalioti, M. D., Zhang, J., Volkman, H. M., et al. (2006). Wnk4 controls blood pressure and potassium homeostasis via regulation of mass and activity of the distal convoluted tubule. Nat Genet, 38, 1124–32. Liddle, G. W., Bledsoe, T., and Coppage, W. S. (1963). A familial renal disorder simulating primary aldosteronism but with negligible aldosterone secretion. Trans Assoc Am Physicians, 76, 199–213. Lifton, R. P. (1996). Molecular genetics of human blood pressure variation. Science, 272, 676–80. Loffing, J. and Kaissling, B. (2003). Sodium and calcium transport pathways along the mammalian distal nephron: from rabbit to human. Am J Physiol Renal Physiol, 284, F628–43. Loffing, J. and Korbmacher, C. (2009). Regulated sodium transport in the renal connecting tubule (CNT) via the epithelial sodium channel (ENaC). Pflugers Arch, 458, 111–35. Masilamani, S., Kim, G. H., Mitchell, C., et al. (1999). Aldosterone-mediated regulation of ENaC alpha, beta, and gamma subunit proteins in rat kidney. J Clin Invest, 104, R19–23. Meneton, P., Jeunemaitre, X., de Wardener, H. E., et al. (2005). Links between dietary salt intake, renal salt handling, blood pressure, and cardiovascular diseases. Physiol Rev, 85, 679–715. Moriguchi, T., Urushiyama, S., Hisamoto, N., et al. (2005). WNK1 regulates phosphorylation of cation-chloride-coupled cotransporters via the STE20-related kinases, SPAK and OSR1. J Biol Chem, 280, 42685–93. Mutig, K., Paliege, A., Kahl, T., et al. (2007). Vasopressin V2 receptor expression along rat, mouse, and human renal epithelia with focus on TAL. Am J Physiol Renal Physiol, 293, F1166–77.

195

196

Section 2  

fluid, electrolyte, and renal tubular disorders

Nicod, M., Michlig, S., Flahaut, M., et al. (2002). A novel vasopressin-induced transcript promotes MAP kinase activation and ENaC downregulation. EMBO J, 21, 5109–17. Pacheco-Alvarez, D., Cristobal, P. S., Meade, P., et al. (2006). The Na+:Cl− cotransporter is activated and phosphorylated at the amino-terminal domain upon intracellular chloride depletion. J Biol Chem, 281, 28755–63. Palmer, L. G. and Frindt, G. (1987). Effects of cell Ca and pH on Na channels from rat cortical collecting tubule. Am J Physiol, 253, F333–9. Pedersen, N. B., Hofmeister, M. V., Rosenbaek, L. L., et al. (2010). Vasopressin induces phosphorylation of the thiazide-sensitive sodium chloride cotransporter in the distal convoluted tubule. Kiney Int, 78, 160–9. Piechotta, K., Lu, J., and Delpire, E. (2002). Cation chloride cotransporters interact with the stress-related kinases Ste20-related proline-alanine-rich kinase (SPAK) and oxidative stress response 1 (OSR1). J Biol Chem, 277, 50812–19. Pradervand, S., Wang, Q., Burnier, M., et al. (1999). A mouse model for Liddle’s syndrome. J Am Soc Nephrol, 10, 2527–33. Rafiqi, F. H., Zuber, A. M., Glover, M., et al. (2010). Role of the WNK-activated SPAK kinase in regulating blood pressure. EMBO Mol Med, 2, 63–75. Ring, A. M., Cheng, S. X., Leng, Q., et al. (2007a). WNK4 regulates activity of the epithelial Na+ channel in vitro and in vivo. Proc Natl Acad Sci U S A, 104, 4020–4. Ring, A. M., Leng, Q., Rinehart, J., et al. (2007b). An SGK1 site in WNK4 regulates Na+ channel and K+ channel activity and has implications for aldosterone signaling and K+ homeostasis. Proc Natl Acad Sci U S A, 104, 4025–9. Rotin, D. and Schild, L. (2008). ENaC and its regulatory proteins as drug targets for blood pressure control. Curr Drug Targets, 9, 709–16. Rozansky, D. J., Cornwall, T., Subramanya, A. R., et al. (2009). Aldosterone mediates activation of the thiazide-sensitive Na-Cl cotransporter through an SGK1 and WNK4 signaling pathway. J Clin Invest, 119, 2601–12. Rubera, I., Loffing, J., Palmer, L. G., et al. (2003). Collecting duct-specific gene inactivation of alphaENaC in the mouse kidney does not impair sodium and potassium balance. J Clin Invest, 112, 554–65. Ruffieux-Daidie, D., Poirot, O., Boulkroun, S., et al. (2008). Deubiquitylation regulates activation and proteolytic cleavage of ENaC. J Am Soc Nephrol, 19, 2170–80. Schild, L., Canessa, C. M., Shimkets, R. A., et al. (1995). A mutation in the epithelial sodium channel causing Liddle disease increases channel activity in the Xenopus laevis oocyte expression system. Proc Natl Acad Sci U S A, 92, 5699–703. Schild, L., Lu, Y., Gautschi, I., et al. (1996). Identification of a PY motif in the epithelial Na channel subunits as a target sequence for mutations causing channel activation found in Liddle syndrome. EMBO J, 15, 2381–7. Schultheis, P. J., Lorenz, J. N., Meneton, P., et al. (1998). Phenotype resembling Gitelman’s syndrome in mice lacking the apical Na+-Clcotransporter of the distal convoluted tubule. J Biol Chem, 273, 29150–5. Shi, H., Asher, C., Chigaev, A., et al. (2002). Interactions of beta and gamma ENaC with Nedd4 can be facilitated by an ERK-mediated phosphorylation. J Biol Chem, 277, 13539–47. Shi, P. P., Cao, X. R., Sweezer, E. M., et al. (2008). Salt-sensitive hypertension and cardiac hypertrophy in mice deficient in the ubiquitin ligase Nedd4-2. Am J Physiol Renal Physiol, 295, F462–70.

Shimkets, R. A., Warnock, D. G., Bositis, C. M., et al. (1994). Liddle’s syndrome: heritable human hypertension caused by mutations in the beta subunit of the epithelial sodium channel. Cell, 79, 407–14. Simon, D. B., Nelson-Williams, C., Bia, M. J., et al. (1996). Gitelman’s variant of Bartter’s syndrome, inherited hypokalaemic alkalosis, is caused by mutations in the thiazide-sensitive Na-Cl cotransporter. Nat Genet, 12, 24–30. Staub, O., Gautschi, I., Ishikawa, T., et al. (1997). Regulation of stability and function of the epithelial Na+ channel (ENaC) by ubiquitination. EMBO J, 16, 6325–36. Stewart, P. M., Wallace, A. M., Valentino, R., et al. (1987). Mineralocorticoid activity of liquorice: 11-beta-hydroxysteroid dehydrogenase deficiency comes of age. Lancet, 2, 821–4. Subramanya, A. R., Yang, C. L., Zhu, X., et al. (2006). Dominant-negative regulation of WNK1 by its kidney-specific kinase-defective isoform. Am J Physiol Renal Physiol, 290, F619–24. Vallet, V., Chraibi, A., Gaeggeler, H. P., et al. (1997). An epithelial serine protease activates the amiloride-sensitive sodium channel. Nature, 389, 607–10. Vallon, V., Schroth, J., Lang, F., et al. (2009). Expression and phosphorylation of the Na+-Cl- cotransporter NCC in vivo is regulated by dietary salt, potassium, and SGK1. Am J Physiol Renal Physiol, 297, F704–12. Van der Lubbe, N., Lim, C. H., Fenton, R. A., et al. (2011). Angiotensin II induces phosphorylation of the thiazide-sensitive sodium chloride cotransporter independent of aldosterone. Kiney Int, 79, 66–76. Velazquez, H., Bartiss, A., Bernstein, P., et al. 91996). Adrenal steroids stimulate thiazide-sensitive NaCl transport by rat renal distal tubules. Am J Physiol, 270, F211–19. Verrey, F., Fakitsas, P., Adam, G., et al. (2008). Early transcriptional control of ENaC (de)ubiquitylation by aldosterone. Kidney Int, 73, 691–6. Vitari, A. C., Deak, M., Morrice, N. A., et al. (2005). The WNK1 and WNK4 protein kinases that are mutated in Gordon’s hypertension syndrome phosphorylate and activate SPAK and OSR1 protein kinases. Biochem J, 391, 17–24. Vuagniaux, G., Vallet, V., Jaeger, N. F., et al. Activation of the amiloride-sensitive epithelial sodium channel by the serine protease mCAP1 expressed in a mouse cortical collecting duct cell line. J Am Soc Nephrol, 11, 828–34, 2000. Wang, T. and Giebisch, G. (1996). Effects of angiotensin II on electrolyte transport in the early and late distal tubule in rat kidney. Am J Physiol, 271, F143–9. Wilson, F. H., Disse-Nicodeme, S., Choate, K. A., et al. (2001). Human hypertension caused by mutations in WNK kinases. Science, 293, 1107–12. Wilson, F. H., Kahle, K. T., Sabath, E., et al. (2003). Molecular pathogenesis of inherited hypertension with hyperkalemia: the Na-Cl cotransporter is inhibited by wild-type but not mutant WNK4. Proc Natl Acad Sci U S A, 100, 680–4. Xu, B. E., Stippec, S., Chu, P. Y., et al. (2005). WNK1 activates SGK1 to regulate the epithelial sodium channel. Proc Natl Acad Sci U S A, 102, 10315–20. Yang, C. L., Angell, J., Mitchell, R., et al. (2003). WNK kinases regulate thiazide-sensitive Na-Cl cotransport. J Clin Invest, 111, 1039–45. Yang, C. L., Zhu, X., Wang, Z., et al. (2005). Mechanisms of WNK1 and WNK4 interaction in the regulation of thiazide-sensitive NaCl cotransport. J Clin Invest, 115, 1379–87. Yang, S. S., Morimoto, T., Rai, T., et al. (2007). Molecular pathogenesis of pseudohypoaldosteronism type II: generation and analysis of a Wnk4(D561A/+) knockin mouse model. Cell Metab, 5, 331–44.

CHAPTER 22

Water homeostasis David Marples and Søren Nielsen Introduction The kidney filters some 180 L/day of fluid from blood plasma: it follows that the vast majority of this filtrate needs to be reabsorbed. Most of this reabsorption occurs passively and constitutively, as water is drawn across the nephron wall by the osmotic gradient produced by active solute transport. In the proximal tubule and descending thin limb of Henle’s loop, the combination of leaky tight junctions and the presence of specific water channels (aquaporin 1; AQP1) means that this reabsorption occurs rapidly, and in the presence of a very small osmotic gradient. In contrast, the low water permeability of the ascending limb and distal convoluted tubule allows salt transport to establish a substantial osmotic gradient, with the tubular fluid being substantially diluted as solute is removed. This is important when water needs to be lost without excessive salt depletion. The accumulation of salt in the medullary interstitium, and the dilute tubular fluid, together provide a driving force for water reabsorption from the collecting ducts, allowing the production of very concentrated urine. Whether the urine is dilute or concentrated is determined primarily by plasma levels of the antidiuretic hormone vasopressin, which alters the permeability of the collecting duct by causing the insertion of AQP2 water channels into the apical membrane of the tubular cells. Arginine vasopressin (AVP) is a peptide hormone released from the posterior pituitary. The main trigger for its release is a rise in plasma osmolality. However, large changes in circulating volume and/or blood pressure can also cause AVP release, as can nausea and a number of other stimuli, including some drugs. Of course, antidiuresis cannot reverse a rise in plasma osmolality: it can only stop it getting worse. Ultimately, there needs to be the intake of additional water, and the major stimuli for thirst are similar to those for AVP release. However, the threshold for thirst is higher than that for AVP release. Disorders of water balance can arise due to failures of normal organ function, regulatory mechanisms, or because of alterations in salt or water intake. The primary defect may be in the renal system, or may be a secondary consequence of problems elsewhere:  for example, low effective circulating volume due to heart disease.

Body fluid composition About half to two-thirds of body weight is made up of water, depending on body composition: muscle bulk increases the fraction, while fat decreases it. Of this water, two-thirds is intracellular, as a potassium-rich solution, while the remaining third is extracellular, and is predominantly a saline solution, with a range of other solutes depending on the compartment. Plasma makes

up about a quarter of the extracellular fluid: typically about 5% of body weight. Typically water moves freely between compartments. Small solutes move freely between plasma and interstitial fluid, but movement across cell membranes is tightly controlled, as is movement into and out of some specialized compartments such as the cerebrospinal fluid. Under normal circumstances plasma proteins cannot move out of the vasculature, and therefore exert an osmotic pressure, called the oncotic pressure, tending to draw water out of the interstitium back into the plasma, and opposing the effects of the hydrostatic pressure. The balance of these Starling forces result in turnover of fluid in normal capillary beds. These forces are exaggerated in the renal peritubular capillaries, where the hydrostatic pressure is lower than in most capillaries because of the resistance of the efferent arterioles, while the oncotic pressure is higher than usual, because of the essentially protein-free glomerular filtrate removed from the plasma. This facilitates the movement of the fluid reabsorbed from the filtrate back into the bloodstream.

Fluid balance Under normal circumstances, water intake will be about 2.5 L/day, from a combination of drinking, water present in food, and water created by metabolism of food. Balancing this, there will be insensible losses from the lungs and skin, loss in the faeces, and loss in the urine, with the latter typically amounting to around 1.5 L/day, which equates to about 1 mL/min. A more detailed breakdown is shown in Table 22.1, but it is important to recognize that all these figures are very dependent on circumstance:  a hot, dry environment will greatly increase losses through ventilation and perspiration, for example. Conversely, social circumstances, or compulsive polydipsia, can result in greatly increased fluid intake. The kidneys, under hormonal control, can reduce urinary loss to as little as about 400 mL/day, or increase it to 20 L/day, to maintain a balance.

Renal handling of water The regulated excretion of water is handled by the kidney, but it has to coordinate this role with its excretion of waste substances and the regulation of salt balance. The first step of the process is glomerular filtration (see Chapter 43). About 20% of the plasma entering the kidney is filtered into the nephrons, with the exception that nearly all plasma proteins are retained within the blood vessels. All solutes with a molecular weight lower than about 7 kDa are freely filtered, while larger substances are progressively less freely filtered, until, at a molecular weight around 70 kDa (depending on the charge and shape of the molecule) there is virtually no filtration

198

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 22.1  Normal water handling in adults in a temperate climate Route

Volume (mL/day)

Modifying factors

Intake Oral fluids

1200

Social, cultural, environmental

Water in food

1000

Diet (fasting)

300

Diet (fasting)

Metabolic water total

2500

Output Urine

1500

Requirements for solute excretion, fluid intake (ADH)

Insensible (skin and lungs)

800

Exercise, environment

Sweat

100

Temperature, exercise

Faeces

100

Diarrhoea

Total

2500

(Maddox and Brenner, 2000). This results in the entry of about 180 L/day of filtrate into the tubules, and a rise in the oncotic pressure of the blood leaving the glomeruli. This blood then flows through a portal system to the peritubular capillaries, where the consequent low hydrostatic pressure and high oncotic pressure facilitate fluid uptake from the peritubular space. To reduce 180 L of filtrate to 1.5 L of urine, it is necessary for the tubules to reabsorb > 99% of the filtrate: even at maximal diuresis around 90% is being absorbed. This water is being drawn passively out of the tubule, down the osmotic gradient created by the active solute transport described in Chapter 20. For such reabsorption to be efficient, it is important that the water permeability of the epithelium is high, and this is achieved by the expression of specific water channels, called aquaporins (AQPs), in both the apical and basolateral membranes of the cells in water-permeant segments of the tubule.

AQP1 in the proximal tubule and descending limb In the proximal tubule, where there is rapid reabsorption of both solutes and water, there is abundant expression of AQP1 in both the apical brush border and the basolateral membranes, with very little seen intracellularly (Nielsen et al., 1993), suggesting that following synthesis it is rapidly inserted into the membrane, with little or no regulation by shuttling from stores. This is consistent with the largely constitutive activity of the proximal tubule. Evidence from knockout studies in mice suggests that AQP1 is responsible for about 80% of the water permeability of the proximal tubule (Schnermann et  al., 1998), and allows reabsorption of about two-thirds of the solute and water filtered to occur in this segment. By allowing water to move out of the tubules in response to a very modest osmotic gradient, it maximizes the efficiency of solute transport too, by limiting the back-flux that would otherwise occur through the leaky tight junctions between cells in this nephron segment. In knockout mice the osmotic gradient increased 3-4 fold, despite which water reabsorption fell by about half. These mice also had a substantially reduced GFR, presumably to limit the excess urinary water loss.

Furthermore, these mice had a substantially impaired urinary concentrating capacity: they could not produce urine concentrated to > 650 mOsm/kg, despite water deprivation that raised their plasma tonicity significantly (Ma et al., 1998). This is thought to reflect impairment of the countercurrent multiplication mechanism:  AQP1 is also expressed in the descending thin limbs of long–loop nephrons (but probably not most short-loop nephrons), and this is thought to be important in allowing the concentration of solutes in the descending limbs. Water permeability in the descending limbs of AQP knockout mice was reduced by about 90% (Chou et al., 1999). (See Chapter 20 for details on the generation of the corticomedullary osmotic gradient.) Despite these apparently important functional defects seen in AQP1 knockout mice, humans (and indeed mice) lacking AQP1 appear remarkably unaffected under normal circumstances (King et al., 2001). AQP1 is expressed on erythrocytes, and acts as a blood grouping antigen (Colton). Thus, Colton-null people lack AQP1 expression. This is exceptionally rare, but such people have no overt disease, although lab investigation does reveal a decrease in urinary concentrating capacity. Unlike in mice, their proximal tubular water reabsorption appears normal, and the defect is probably due to failure to generate a normal medullary osmotic gradient. This study also found the AQP1-null humans had a normal GFR, suggesting that, although AQP1 is expressed in human (but not rat) glomeruli (Maunsbach et al., 1997), this is of limited importance in the filtration process. While we have described the reabsorption of solute and water in the proximal tubule as constitutive, it is certainly the case that there is hormonal modulation of this process (Reeves and Andreoli, 2007). As part of this modulation, there are changes in AQP1 expression in response to physiological changes (angiotensin levels, interstitial tonicity), and also in pathological conditions such as during ureteric obstruction (Li et al., 2003). These may contribute to the altered renal water handling seen in these conditions.

AQP7 in the late proximal tubule A second aquaporin, AQP7, is expressed in the proximal straight tubule (Ishibashi et al., 2000; Nejsum et al., 2000). AQP7 is permeable to a number of other solutes, including glycerol and urea, as well as to water. In mice, knocking out this gene leads to glyceroluria, but not to a concentrating defect, although mice lacking both AQP1 and AQP7 have a more severe concentrating defect than those lacking AQP1 alone. These results suggest that AQP7 plays only a modest role in water reabsorption in the late proximal tubule, but is important in the reabsorption of glycerol (Hara-Chikuma et al., 2005). This may be significant for intrarenal glucose production.

Ascending limb of Henle’s loop and distal convoluted tubule These segments of the nephron have low water permeability, and little water reabsorption from the tubular fluid occurs. Consistent with this, there are no aquaporins expressed in these nephron regions. Indeed, the very active salt uptake in these regions tends to result in the production of a dilute tubular fluid (Moe et  al., 2000). This is important for water balance in two ways: first, the salt extracted contributes substantially to the accumulation of salt in the renal medulla, thus providing an osmotic gradient for (regulated) water reabsorption from the collecting ducts, as discussed

chapter 22 

below, and second, because it makes it possible to excrete a water load with relatively little salt loss. Drugs that block salt uptake, such as loop diuretics and thiazides, cause an increase in urine output, but actually blunt the ability to excrete free water (because salt is lost at the same time).

The connecting tubule and collecting duct—the site of regulated water permeability Although only a small fraction (about 15%) of the filtered water reaches the later parts of the renal tubule, this still represents some 20–25 L of water per day. As noted in the previous paragraph, this fluid is markedly hypotonic, with a typical osmolality of around 100 mOsm/kg. H2O. The water permeability of the connecting tubule and collecting duct are regulated by the antidiuretic hormone AVP, which causes shuttling of AQP2 water channels from an intracellular store to the apical plasma membrane of collecting duct granular cells. In the absence of these water channels this apical plasma membrane is highly impermeable to water (Nielsen et al., 1995), and large volumes of dilute urine can be excreted, allowing the clearance of a water load. When vasopressin levels rise, the insertion of AQP2 water channels into the apical plasma membrane allows the entry of water from the tubular fluid (Nielsen et al., 1995). Because this fluid has been diluted by salt extraction in the ascending limb of Henle’s loop and distal convoluted tubule, much of the water can be reabsorbed in the cortical collecting ducts (Ward et al., 1999), particularly if there is concomitant salt reabsorption occurring. In the cortex there is an abundant vasculature that can pick up this water, and the osmotic gradient is maintained by this high blood flow, which effectively clamps the interstitial tonicity to that of plasma. Thus only a modest amount of water flows down into the medullary collecting ducts, where further water can be extracted by the interstitial tonicity built up by the loop of Henle, allowing the production of small volumes of highly concentrated urine, without flooding the inner medulla with large amounts of water, which would flush out the salt and other solutes accumulated by the countercurrent multiplication process, etc. Depending on the levels of ADH, and the amount of solute that needs to be excreted, urine output can vary from about 0.5 to 20 L (Barlow and de Wardener, 1959). In contrast to the regulated permeability of the apical plasma membrane, the basolateral membrane of the cells is consistently highly permeable to water, reflecting the constitutive expression of AQP3 and AQP4. Both provide exit pathways for water entering the cells through AQP2, but AQP3 is expressed predominantly in the cortical and outer medullary collecting ducts (Ecelbarger et  al., 1995), while AQP4 is more abundant in the inner medulla (Terris et  al., 1995). Mice with these channels knocked out give important confirmation of the relative contributions of the cortical and inner medullary collecting ducts to the control of urine volume: defects in AQP3 lead to a very marked polyuria, since cortical water reabsorption is severely impaired (Ma et  al., 2000), while a lack of AQP4 only causes marginal polyuria, with a modest impairment of maximal concentrating capacity (Ma et al., 1997). However, interpretation of the results is complicated by the observation that knocking out AQP3 also caused a substantial decrease in AQP2 expression (Ma et al., 2000), perhaps because of a drop in intracellular concentration and/or cell swelling.

water homeostasis

Cellular mechanisms of AQP2 regulation As noted above, the major regulator of collecting duct water permeability is the posterior pituitary hormone vasopressin (arginine vasopressin in man, although the porcine equivalent, lysine vasopressin, is sometimes used clinically). AVP is carried in the bloodstream to the basolateral surface of the collecting duct principal cells, where it interacts with V2 receptors, which are GPCRs linked to Gs (Lolait et al., 1992; Seibold et al., 1992). They consequently cause the production of the second messenger cAMP, and the activation of protein kinase A (PKA). Amongst the proteins phosphorylated by PKA is AQP2 itself, which undergoes phosphorylation of serine 256 in its C-terminus. This acts as a trigger for shuttling of AQP2 stored in vesicles to the cell surface (Nielsen et al., 1993b; Marples et al., 1995), although the details of how this is initiated remain to be established. Mutant forms of AQP2 where this phosphorylation site is abolished (e.g. S256A) are unable to traffic to the surface (Fushimi et al., 1997), while mutations which are effectively constitutively phosphorylated at serine 256 (S256D) traffic spontaneously to the cell surface in cell models of the collecting duct (Kamsteeg et al., 2000), suggesting that phosphorylation of AQP2 alone is sufficient to provoke the delivery process, but it remains possible that other targets of PKA also play an important part. There is evidence for the involvement of both microtubule-based transport of AQP2-bearing vesicles towards the apical plasma membrane (Phillips and Taylor, 1989; Marples et al., 1998), and the reorganization of the actin cytoskeleton (Kachadorian et al., 1979; Ding et al., 1991), which may act as a barrier to vesicular movement, and may also act as a substrate for myosin-based transport of the vesicles (Chou et al., 2004). Targeting recognition proteins similar to those involved in the docking of synaptic vesicles also appear to be important in the delivery of the vesicles specifically to the apical plasma membrane (Nielsen et al., 1995b). The retrieval of AQP2 from the cell surface also appears to be regulated:  activation of protein kinase C increases the endocytosis, although this is not dependent on the phosphorylation of AQP2 itself; nor does it require the dephosphorylation of S256 (van Balkom et al., 2002). Conversely, cAMP decreases the rate of retrieval, slowing the background endocytic rate, and thus enhancing the effects of the increased insertion described above. The mechanisms behind this regulation of retrieval remain to be determined, but it is interesting that it may be prevented by phosphorylation of AQP2 at serine 269 (Moeller et al., 2010). The fate of the endocytosed AQP2 also remains unclear: while some cell models demonstrate that the AQP2 can be repeatedly recycled back into the exocytic pathway (Katsura et al., 1996; Gustafson et al., 2000), there is also evidence that it can be transported to multivesicular bodies, thought to be part of a degradative pathway (Nielsen et al., 1993b; van Balkom et  al., 2009), and also that it is subsequently shed into the urine. This shedding of AQP2 does not appear to reflect loss of apical plasma membrane, or of whole cell fragments, but appears to be a specific process relating to vasopressin activity and AQP2 turnover (Wen et al., 1999). This loss of AQP2 may be an important component of the longer-term regulation that allows the body to adapt to chronic alterations in water balance. While vasopressin is clearly the major trigger for AQP2 shuttling, it has been shown, at least in research models, that other stimuli can induce shuttling. For example, a rise in cGMP can lead to AQP2 phosphorylation via PKG, and cause its exocytosis (Bouley

199

200

Section 2  

fluid, electrolyte, and renal tubular disorders

et al., 2005). Similarly, inhibition of phosphatases has been shown to induce shuttling (Valenti et al., 2000). Drugs (such as sildenafil) which can activate these pathways may be useful in the management of some forms of nephrogenic diabetes insipidus (Sanches et al., 2012) (see Chapter 32). The water permeability of the collecting ducts at any given time reflects the balance between AQP2 insertion and retrieval, together with the total amount of AQP2 available in the cell. Thus it is best thought of as a dynamic equilibrium, with increasing levels of AVP shifting the equilibrium towards greater plasma membrane AQP2, while falling levels (or other signals that increase endocytosis) shift the equilibrium towards intracellular AQP2, and thus decrease the water permeability. The connecting tubule represents a transitional zone between distal convoluted tubule and the collecting duct proper. It expresses both V2 receptors and AQP2 (Kishore et al., 1996), and is able to shuttle AQP2 to the apical surface of the cells in response to AVP, like the collecting duct. However, unlike the cortical collecting ducts nearby, the connecting tubule can also insert a substantial amount of AQP2 into its basolateral membrane (Jeon et al., 2003), although this is not affected by AVP levels; the functional significance of this remains to be determined. It is clear that water reabsorption via AQP2 in the connecting tubule is functionally important (at least in mice), since mice lacking AQP2 altogether died of dehydration within a few days after birth (Yang et al., 2001), while those which expressed AQP2 in the connecting tubules but not the collecting ducts survived, albeit with severe polyuria (Rojek et al., 2006).

Long-term adaptation of the collecting duct The shuttling mechanisms described above respond rapidly (within a few minutes) to changes in circulating AVP levels, thus providing a responsive system that can maintain good water balance despite substantial acute changes. It has become clear that there are also longer-term, adaptive changes, which allow the body to modulate the response to vasopressin. For example, chronic water restriction leads to an increase in AQP2 synthesis (Nielsen et al., 1993b), thus increasing the total store of AQP2 present in the principal cells. If a certain level of AVP results in the delivery of a certain fraction of total AQP2 to the plasma membrane, it is clear that increasing total AQP2 will result in higher plasma membrane permeability at any given AVP level. Thus a high degree of urinary concentration can occur at lower AVP levels. This has two advantages: the body needs less AVP to maintain a given degree of antidiuresis, and it retains a flexible response: further dehydration has the capacity to elicit a further response. Conversely, chronic water loading leads to a downregulation of AQP2 (Ecelbarger et al., 1997), reducing the collecting duct permeability at any given AVP level. Indeed, this decreased expression provides an opportunity to ‘escape’ partially from the water retention caused by SIADH. These changes provide a mechanism for the changes in human concentrating capacity brought about by alterations in body fluid balance (de Wardener and Herxheimer, 1957; Jones and de Wardener, 1956). One signal driving the expression of AQP2 is vasopressin itself: Brattleboro rats, which lack significant levels of AVP, have substantially reduced AQP2 levels, which normalize following AVP infusion for 5  days (DiGiovanni et  al., 1994). Similarly, infusion of AVP in normal rats increases AQP2 expression (Terris et  al., 1996). Conversely, treatment with V2 receptor antagonists leads to a fall in AQP2 levels (Marples et al., 1998). Like the acute shuttling

response to AVP, this also appears to be mediated by cAMP, and the 5′ untranslated region of the AQP2 gene includes a cAMP-response element (Yasui et al., 1997). Thus dehydration will cause an increase in AQP2 expression at least partly by causing a rise in AVP levels, while water loading will suppress AVP release. However, it is clear that there are other signals that are at least as potent. The most direct evidence comes from ‘vasopressin escape’ experiments: Rats which are given a continuous infusion of AVP are nonetheless able to downregulate AQP2 expression when they are water-loaded for several days: this does not represent a loss of renal responsiveness to AVP, because their remaining AQP2 is still targeted to the apical plasma membrane (Ecelbarger et al., 1997). Further evidence comes from experiments in which rats deprived of water showed much greater increases in AQP2 expression than could be achieved by AVP infusion, suggesting that there are additional, possibly synergistic, signals (Terris et al., 1996). In vitro studies have suggested that the osmolality of the environment may be a significant factor, and a hypertonicity response element has been shown to be associated with the AQP2 gene (Kasono et al., 2005; Hasler et al., 2005), but in vivo studies have not really supported this hypothesis (Marples et al., 1996, 1998). Another possible signal could be tubular flow rates, but again in vivo models are not really consistent with this (Marples et al., 1998). It is likely that a number of paracrine and/or systemic hormones may play a role, but remain to be identified.

Disordered regulation of aquaporins A number of diseases are associated with impaired water balance, and these may cause or result from changes in aquaporin expression or function. We have seen that a genetic defect in AQP1 leads to a subclinical defect in concentrating capacity (King et al., 2001):  in contrast, mutations in AQP2 (or the V2 receptor) lead to severe congenital nephrogenic diabetes insipidus (NDI) (Deen et al., 1994). No AQP3-null humans have yet been identified, but mouse models suggest that they too would suffer from severe NDI (Ma et al., 2000). These mutations are rare, but acquired forms of NDI are much more common:  electrolyte disturbances such as hypokalaemia (Marples et  al., 1996)  and hypercalcaemia (Earm et al., 1998), drugs such as lithium (Marples et al., 1995), and pathological conditions such as urinary tract obstruction (Frøkiaer et al., 1996) have been shown to be associated with decreases in AQP2 expression, which explain at least part of the polyuria seen in these circumstances (see Chapter 32). In contrast, overexpression of AQP2 can lead to water retaining states: for example, during pregnancy (Ohara et al., 1998), SIADH, and cirrhosis (Fujita et  al., 1995), and congestive heart failure (Nielsen et al., 1997). In many cases, these seem to be due to excess AVP release due to non-osmotic stimuli (Schrier et al., 1998) (see Chapter 28).

Systemic control of water balance Control of antidiuresis As we have seen, in the absence of a central antidiuretic signal the kidney will produce and excrete a large volume of dilute urine. The main stimulus for antidiuresis is the hormone vasopressin, which is synthesized by the magnocellular neurons of the hypothalamus. Their cell bodies lie in the supraoptic and paraventricular nuclei of the hypothalamus, while their axons project to the posterior

chapter 22 

pituitary (neurohypophysis). Vasopressin is a cyclic nonapeptide, but is synthesized as a preprohormone, which is then cleaved to yield AVP itself, together with neurophysin II and copeptin, within the secretory vesicles (Brownstein, 1983). Genetic defects in either the AVP or neurophysin components lead to autosomal dominantly inherited forms of central diabetes insipidus (Babey et al., 2011), suggesting that neurophysin plays a significant role in the packaging and/or release of AVP.

Table 22.2  Factors affecting AVP release Factors increasing AVP release

Factors inhibiting AVP release

Physiological/pathophysiological

Physiological/pathophysiological

Hypertonicity/hyperosmolality

Hypotonicity/hypo-osmolality

Hypotension

Hypertension

Volume depletion (total or effective)

Volume expansion

Osmotic stimuli The main trigger for the release of AVP is plasma osmolality. The magnocellular neurons themselves are osmosensitive, but the main osmosensory cells appear to lie in the organum vasculosum laminae terminalis (OVLT), a region of the hypothalamus which lies functionally outside the blood–brain barrier (Bourque, 2008). Lesions in this region lead to both central diabetes insipidus and an impairment of thirst, suggesting that it is important for osmosensing in both systems, although different cells (with different osmotic thresholds) appear to be involved. Infusion of this region with hypertonic solutions leads to an increased neuronal firing rate, which in turn leads to increased AVP release. Conversely, hypotonic solutions are able to suppress the basal activity. The outcome is a system which produces essentially no AVP release below a plasma osmolality of about 280 mOsm/Kg H2O, but a rapid rise thereafter, leading to a maximally antidiuretic level of AVP when plasma osmolality reaches about 295: thus the entire dynamic range of the system is triggered by a change of about 5%. Within this range, the relationship between osmolality and plasma AVP levels is roughly linear (Robertson et al., 1982). There is some evidence that the action of central osmoreceptors can be modulated by input from peripheral sensors: gut osmosensors, as well as temperature sensors in the mouth and pharynx. These can result in the cessation of drinking, and a fall in AVP secretion, before there has been any detectable drop in plasma osmolality. Presumably this information is carried via vagal afferents, and signals passed to the hypothalamus via medullary nuclei (Bourque, 2008).

Non-osmotic stimuli A second stimulus for antidiuresis is a low blood pressure or effective circulating volume. In contrast to the effects of osmolality, this response shows an exponential pattern: decreases of up to 5% in

water homeostasis

Nausea Pain/stress Hypoxia and hypercapnia Hypothyroidism Hypoglycaemia Hormones: angiotensin II, bradykinin, histamine Circadian rhythm (AVP levels higher at night) Drugs

Drugs

µ-opioids and narcotics

κ-opioids

Alpha-adrenoceptor agonists (via BP?)

Beta-adrenoceptor agonists

Nicotine (in non-smokers)

Dopamine antagonists (via suppressed nausea?)

Barbiturates

Ethanol

General anaesthetics

Glucocorticoids

Antipsychotics and antidepressants

circulating volume or blood pressure have very little effect on AVP release, while once the levels have fallen by 10% the effect rapidly accelerates (Berl and Robertson, 2000). In addition to causing release of ADH in their own right, these stimuli appear to enhance the responsiveness to an osmotic challenge (in other words, they increase the rise in AVP seen for a given rise in osmolality). The signals for these stimuli come from atrial and vascular baroreceptors, signalling via medullary nuclei.

Table 22.3  Causes of the syndrome of inappropriate secretion of antidiuretic hormone (SIADH) (note that the drugs mentioned in Table 22.2 that stimulate AVP release will mimic this condition) Malignancy

Central nervous system disorders

Pulmonary disease

Others

Pulmonary/mediastinal

Acute trauma/haemorrhage/stroke

Infections: bacterial, viral or fungal

Chronic inflammation

Gastrointestinal, including pancreas

Mass lesions (tumour, haematoma, abscess, hydrocephalus)

Cystic fibrosis

HIV

Genitourinary, including uterus, prostate and bladder

Inflammatory and demyelinating disease

Acute respiratory failure

Prolonged severe exercise

Leukaemia and lymphoma

Acute psychosis

Chronic obstructive pulmonary disease

Idiopathic

Delirium tremens

Positive pressure ventilation

Pituitary stalk section

201

202

Section 2  

fluid, electrolyte, and renal tubular disorders

A range of other stimuli and drugs (see Table 22.2) have also been shown to affect AVP release. The most powerful of these is nausea, which can cause very large rises in AVP secretion. Similarly, drugs which cause nausea can be potent stimuli. Other stimuli such as pain and stress may act by causing nausea, or there may be independent pathways. A rare, but potentially clinically important, cause of high AVP secretion is SIADH, which may be due to ectopic production, or to disease causing excessive release from the hypothalamic cells (Table 22.3).

Destruction of AVP AVP has a half-life in the circulation of 10–35 minutes (Sharman and Low, 2008), so the system can respond fairly rapidly to changes in fluid balance. AVP is degraded mainly in the liver and kidneys, where peptidases break the cyclic structure and then progressively cleave the peptide. Some is also lost directly in the urine, and for the artificial analogue desmopressin this is the main route of loss: this explains its much longer duration of action.

Thirst Ultimately, renal water retention can only stop a water deficit from getting worse: in order to reverse the problem further water intake is required. As noted above, a rising plasma osmolality stimulates thirst, at least partly through osmoreceptors located in the OVLT (Bourque, 2008). However, the sensation of thirst does not develop until plasma osmolality has reached a level at which antidiuresis is nearly maximal: typically about 290 mOsm/Kg H2O (Robertson et  al., 1982). The consequence is that water throughput is minimized. As with AVP release, thirst can also be stimulated by haemodynamic factors, but again these only become significant once quite large changes have been experienced.

References Babey, M., Kopp, P., and Robertson, G. L. (2011). Familial forms of diabetes insipidus: clinical and molecular characteristics. Nat Rev Endocrinol, 7(12), 701–14. Barlow, E. D. and de Wardener, H. E. (1959). Compulsive water drinking. Q J Med, 8(110), 235–58. Berl, T. and Robertson, G. L. (2000). Pathophysiology of water metabolism. In B. M. Brenner (ed.) Brenner and Rector’s The Kidney (6th ed.), pp. 866–924. Philadelphia, PA: Saunders. Bouley, R., Pastor-Soler, N., Cohen, O., et al. (2005). Stimulation of AQP2 membrane insertion in renal epithelial cells in vitro and in vivo by the cGMP phosphodiesterase inhibitor sildenafil citrate (Viagra). Am J Physiol Renal Physiol, 288(6), F1103–12. Bourque, C. W. (2008). Central mechanisms of osmosensation and systemic osmoregulation. Nat Rev Neurosci, 9, 519–31. Brownstein, M. J. (1983). Biosynthesis of vasopressin and oxytocin. Annu Rev Physiol, 45, 129–35. Chou, C. L., Knepper, M. A., Hoek, A. N., et al. (1999). Reduced water permeability and altered ultrastructure in thin descending limb of Henle in aquaporin-1 null mice. J Clin Invest, 103(4), 491–6. Chou, C. -L., Christensen, B. M., Frische, S., et al. (2004). Non-muscle myosin II and myosin light chain kinase are downstream targets for vasopressin signaling in the renal collecting duct. J Biol Chem, 279(47), 49026–35. Deen, P. M., Verdijk, M. A., Knoers, N. V., et al. (1994). Requirement of human renal water channel aquaporin-2 for vasopressin-dependent concentration of urine. Science, 264(5155), 92–5. De Wardener, H. E. and Herxheimer, A. (1957). The effect of a high water intake on the kidney’s ability to concentrate the urine in man. J Physiol (Lond), 139(1), 42–52.

DiGiovanni, S. R., Nielsen, S., Christensen, E. I., et al. (1994). Regulation of collecting duct water channel expression by vasopressin in Brattleboro rat. Proc Natl Acad Sci U S A, 91(19), 8984–8. Ding, G. H., Franki, N., Condeelis, J., et al. (1991). Vasopressin depolymerizes F-actin in toad bladder epithelial cells. Am J Physiol, 260(1 Pt 1), C9–16. Earm, J. H., Christensen, B. M., Frokiaer, J., et al. (1998). Decreased aquaporin-2 expression and apical plasma membrane delivery in kidney collecting ducts of polyuric hypercalcemic rats. J Am Soc Nephrol, 9(12), 2181–93. Ecelbarger, C. A., Nielsen, S., Olson, B. R., et al. (1997). Role of renal aquaporins in escape from vasopressin-induced antidiuresis in rat. J Clin Invest, 99(8), 1852–63. Ecelbarger, C. A., Terris, J., Frindt, G., et al. (1995). Aquaporin-3 water channel localization and regulation in rat kidney. Am J Physiol, 269(5 Pt 2), F663–72. Frøkiaer, J., Marples, D., Knepper, M. A., et al. (1996). Bilateral ureteral obstruction downregulates expression of vasopressin-sensitive AQP-2 water channel in rat kidney. Am J Physiol, 270(4 Pt 2), F657–68. Fujita, N., Ishikawa, S. E., Sasaki, S., et al. (1995). Role of water channel AQP-CD in water retention in SIADH and cirrhotic rats. Am J Physiol, 269(6 Pt 2), F926–31. Fushimi, K., Sasaki, S., and Marumo, F. (1997). Phosphorylation of serine 256 is required for cAMP-dependent regulatory exocytosis of the aquaporin-2 water channel. J Biol Chem, 272(23), 14800–4. Gustafson, C. E., Katsura, T., McKee, M., et al. (2000). Recycling of AQP2 occurs through a temperature- and bafilomycin-sensitive trans-Golgi-associated compartment. Am J Physiol, 278(2), F317–26. Hara-Chikuma, M., Sohara, E., Rai, T., et al. (2005). Progressive adipocyte hypertrophy in aquaporin-7-deficient mice: adipocyte glycerol permeability as a novel regulator of fat accumulation. J Biol Chem, 280(16), 15493–6. Hasler, U., Vinciguerra, M., Vandewalle, A., et al. (2005). Dual effects of hypertonicity on aquaporin-2 expression in cultured renal collecting duct principal cells. J Am Soc Nephrol, 16(6), 1571–82. Ishibashi, K., Imai, M., and Sasaki, S. (2000). Cellular localization of aquaporin 7 in the rat kidney. Exp Nephrol, 8(4–5), 252–7. Jeon, U. S., Joo, K. W., Na, K. Y., et al. (2003). Oxytocin induces apical and basolateral redistribution of aquaporin-2 in rat kidney. Nephron Exp Nephrol, 93(1), e36–45. Jones, R. V. H. and de Wardener, H. E. (1956). Urine concentration after fluid deprivation or pitressin tannate in oil. BMJ, 1, 271–4. Kachadorian, W. A., Ellis, S. J., and Muller, J. (1979). Possible roles for microtubules and microfilaments in ADH action on toad urinary bladder. Am J Physiol, 236(1), F14–20. Kamsteeg, E. J., Heijnen, I., van Os, C. H., et al. (2000). The subcellular localization of an aquaporin-2 tetramer depends on the stoichiometry of phosphorylated and nonphosphorylated monomers. J Cell Biol, 151(4), 919–30. Kasono, K., Saito, T., Saito, T., et al. (2005). Hypertonicity regulates the aquaporin-2 promoter independently of arginine vasopressin. Nephrol Dial Transplant, 20(3), 509–15. Katsura, T., Ausiello, D. A., and Brown, D. (1996). Direct demonstration of aquaporin-2 water channel recycling in stably transfected LLC-PK1 epithelial cells Am J Physiol, 270, F548–53. King, L. S., Choi, M., Fernandez, P. C., et al. (2001). Defective urinary-concentrating ability due to a complete deficiency of aquaporin-1. N Engl J Med, 345(3), 175–9. Kishore, B. K., Mandon, B., Oza, N. B., et al. (1996). Rat renal arcade segment expresses vasopressin-regulated water channel and vasopressin V2 receptor. J Clin Invest, 97(12), 2763–71. Li, C., Wang, W., Knepper, M. A., et al. (2003). Downregulation of renal aquaporins in response to unilateral ureteral obstruction. Am J Physiol, 284(5), F1066–79. Lolait, S. J., O’Carroll, A. M., McBride, O. W., et al. (1992). Cloning and characterization of a vasopressin V2 receptor and possible link to nephrogenic diabetes insipidus. Nature, 357(6376), 336–9.

chapter 22 

Ohara, M., Martin, P. Y., Xu, D. L., et al. (1998). Upregulation of aquaporin 2 water channel expression in pregnant rats. J Clin Invest, 101(5), 1076–83. Ma, T., Song, Y., Yang, B., et al. (2000). Nephrogenic diabetes insipidus in mice lacking aquaporin-3 water channels. Proc Natl Acad Sci U S A, 97(8), 4386–91. Ma, T., Yang, B., Gillespie, A., et al. (1997). Generation and phenotype of a transgenic knockout mouse lacking the mercurial-insensitive water channel aquaporin-4. J Clin Invest, 100(5), 957–62. Ma, T., Yang, B., Gillespie, A., et al. (1998). Severely impaired urinary concentrating ability in transgenic mice lacking aquaporin-1 water channels. J Biol Chem, 273(8), 4296–9. Maddox, D. A. and Brenner, B. M. (2000). Glomerular ultrafiltration. In B. M. Brenner (ed.) Brenner and Rector’s The Kidney (6th ed.), pp. 319–374. Philadelphia, PA: Saunders. Marples, D., Frokiaer, J., Dorup, J., et al. (1996). Hypokalemia-induced downregulation of aquaporin-2 water channel expression in rat kidney medulla and cortex. J Clin Invest, 97(8), 1960–8. Marples, D., Knepper, M. A., Christensen, E. I., et al. (1995). Redistribution of aquaporin-2 water channels induced by vasopressin in rat kidney inner medullary collecting duct. Am J Physiol, 269(3 Pt 1), C655–64. Marples, D., Schroer, T. A., Ahrens, N., et al. (1998). Dynein and dynactin colocalize with AQP2 water channels in intracellular vesicles from kidney collecting duct. Am J Physiol, 274(2 Pt 2), F384–94. Maunsbach, A. B., Marples, D., Chin, E., et al. (1997). Aquaporin-1 water channel expression in human kidney. J Am Soc Nephrol, 8(1), 1–14. Moe, O. W., Berry, C. A., and Rector, F. C., Jr. (2000). Renal transport of glucose, amino acids, sodium, chloride, and water. In B. M. Brenner (ed.) Brenner and Rector’s The Kidney (6th ed.), pp. 413–52. Philadelphia, PA: Saunders. Moeller, H. B., Praetorius, J., Rutzler, M. R., et al. (2010). Phosphorylation of aquaporin-2 regulates its endocytosis and protein-protein interactions. Proc Natl Acad Sci U S A, 107(1), 424–9. Nejsum, L. N., Elkjaer, M., Hager, H., et al. (2000). Localization of aquaporin-7 in rat and mouse kidney using RT-PCR, immunoblotting, and mmunocytochemistry. Biochem Biophys Res Comm, 277(1), 164–70. Nielsen, S., Chou, C. L., Marples, D., et al. (1995). Vasopressin increases water permeability of kidney collecting duct by inducing translocation of aquaporin-CD water channels to plasma membrane. Proc Natl Acad Sci U S A, 92(4), 1013–17. Nielsen, S., DiGiovanni, S. R., Christensen, E. I., et al. (1993b). Cellular and subcellular immunolocalization of vasopressin-regulated water channel in rat kidney. Proc Natl Acad Sci U S A, 90(24), 11663–7. Nielsen, S., Marples, D., Birn, H., et al. (1995b). Expression of VAMP-2-like protein in kidney collecting duct intracellular vesicles. Colocalization with Aquaporin-2 water channels. J Clin Invest, 96(4), 1834–44. Nielsen, S., Smith, B. L., Christensen, E. I., et al. (1993). CHIP28 water channels are localized in constitutively water-permeable segments of the nephron. J Cell Biol, 120(2), 371–83. Nielsen, S., Terris, J., Andersen, D., et al. (1997). Congestive heart failure in rats is associated with increased expression and targeting of aquaporin-2 water channel in collecting duct. Proc Natl Acad Sci U S A, 94(10), 5450–5.

water homeostasis

Reeves, W. B. and Andreoli, T. E. (2007). Tubular sodium transport. In R. W. Schrier (ed.) Diseases of the Kidney and Urinary Tract (8th ed.), pp. 124–59. Philadelphia, PA: Lippincott Williams & Wilkins. Robertson, G. L., Aycinena, P., and Zerbe, R. L. (1982). Neurogenic disorders of osmoregulation. Am J Med, 72, 339–53. Rojek, A., Fuchtbauer, E. M., Kwon, T. H., et al. (2006). Severe urinary concentrating defect in renal collecting duct-selective AQP2 conditional-knockout mice. Proc Natl Acad Sci U S A, 103(15), 6037–42. Sanches, T. R., Volpini, R. A., Massola Shimizu, M. H., et al. (2012). Sildenafil reduces polyuria in rats with lithium-induced NDI. Am J Physiol Renal Physiol, 302(1), F216–25. Schnermann, J., Chou, C. L., Ma, T., et al. (1998). Defective proximal tubular fluid reabsorption in transgenic aquaporin-1 null mice. Proc Natl Acad Sci U S A, 95(16), 9660–4. Seibold, A., Brabet, P., Rosenthal, W., et al. (1992). Structure and chromosomal localization of the human antidiuretic hormone receptor gene. Am J Hum Genet, 51(5), 1078–83. Phillips, M. E. and Taylor, A. (1989). Effect of nocodazole on the water permeability response to vasopressin in rabbit collecting tubules perfused in vitro. J Physiol (Lond), 411, 529–44. Sharman, A. and Low, J. (2008). Vasopressin and its role in critical care. Cont Ed Anaesth Crit Care Pain, 8 (4), 134–7. Schrier, R. W., Fassett, R. G., Ohara, M., et al. (1998). Vasopressin release, water channels, and vasopressin antagonism in cardiac failure, cirrhosis, and pregnancy. Proc Assoc Am Phys, 110(5), 407–11. Terris, J., Ecelbarger, C. A., Marples, D., et al. (1995). Distribution of aquaporin-4 water channel expression within rat kidney. Am J Physiol, 269(6 Pt 2), F775–85. Terris, J., Ecelbarger, C. A., Nielsen, S., et al. (1996). Long-term regulation of four renal aquaporins in rats. Am J Physiol, 271(2 Pt 2), F414–22. Van Balkom, B. W., Boone, M., Hendriks G. et al. (2009). LIP5 interacts with aquaporin 2 and facilitates its lysosomal degradation. J Am Soc Nephrol, 20(5), 990–1001. Van Balkom, B. W. M., Savelkoul, P. J. M., Markovich, D., et al. (2002). The role of putative phosphorylation sites in the targeting and shuttling of the aquaporin-2 water channel. J Biol Chem, 277(44), 41473–9. Valenti, G., Procino, G., Carmosino, M., et al. (2000). The phosphatase inhibitor okadaic acid induces AQP2 translocation independently from AQP2 phosphorylation in renal collecting duct cells. J Cell Sci, 113 (Pt 11), 1985–92. Ward, D. T., Hammond, T. G., and Harris, H. W. (1999). Modulation of vasopressin-elicited water transport by trafficking of aquaporin2-containing vesicles. Ann Rev Physiol, 61, 683–97. Wen, H., Frokiaer, J., Kwon, T. H., et al. (1999). Urinary excretion of aquaporin-2 in rat is mediated by a vasopressin-dependent apical pathway. J Am Soc Nephrol, 10(7), 1416–29. Yang, B., Gillespie, A., Carlson, E. J., et al. (2001). Neonatal mortality in an aquaporin-2 knock-in mouse model of recessive nephrogenic diabetes insipidus. J Biol Chem, 276(4), 2775–9. Yasui, M., Zelenin, S. M., Celsi, G., et al. (1997). Adenylate cyclase-coupled vasopressin receptor activates AQP2 promoter via a dual effect on CRE and AP1 elements. Am J Physiol, 272(4 Pt 2), F443–50.

203

CHAPTER 23

Potassium homeostasis Alain Doucet and Gilles Crambert Introduction Potassium (K+), the most abundant cation in the organism, is involved in a large variety of physiological functions. The equilibrium between the concentration of K+ in the extracellular space (low) and the intracellular compartment (high) is crucial for maintaining the electrical properties of excitable and non-excitable cells, because it determines the membrane resting potential. The high intracellular concentration of K+ (120–140 mmol/L) also contributes to the intracellular osmolarity, a determinant of cell volume. It is therefore crucial to finely tune both extracellular and intracellular K+ concentrations. To achieve this goal, there is a coordinated regulation between processes/mechanisms that store/release K+ from internal stores (internal balance) and those that retain/excrete K+ (external balance) (Fig. 23.1).

Internal potassium balance K+ compartments in the organism The total K+ content of the body is maintained constant around 50–55 mmol of K+/kg of body weight (excepted during growth and pregnancy). The most important part (98%) of the body K+ is intracellular and the remaining 2% is mainly present in the extracellular fluid (around 70 mmol; extracellular space). Skeletal muscle is the main store of intracellular K+ (around 3000 mmol), whereas liver and red blood cells account for 200 mmol each. To maintain a constant equilibrium, intracellular storage compartments may either ‘pump’ from or release K+ into the extracellular space; whereas the release of K+ into the extracellular space is a passive process through K+ channels and carriers, its pumping into muscle cells is active and mediated by the Na+,K+-ATPase. Therefore, regulatory factors (thyroid hormone, insulin, exercise, etc.) that modify Na+,K+-ATPase density in skeletal muscles have a direct impact on extracellular K+ concentration by modifying the rate of K+ entry into muscle cells.

Postprandial and fasting K+ transfers Because of its low K+ level, the extracellular space is the compartment of the organism that is the most susceptible to K+ concentration variations. Indeed, the amount of K+ ingested daily (around 70–100  mmol) is similar to the plasma pool of K+. Therefore, one challenge that the organism has to face is to maintain plasma K+ constant (between 3.5 and 5  mmol/L) whatever the circumstances (after a meal or during fasting for instance). The inability to do so would modify the extracellular/intracellular K+ ratio and affect physiological functions. A complex regulatory system, involving both internal (K+ exchange from intracellular stores and

extracellular space; internal balance) and external mechanisms (K+ excretion in urine and stools; external balance), is therefore in charge of maintaining constant the plasma K+ level. It is crucial that both internal and external K+ balance mechanisms work in coordination to avoid harmful fluctuations of the extracellular K+ level. The absorption of a meal, which is primarily dictated by the need for an energy source (under the final form of glucose) is normally accompanied by a large intake of K+ because natural products, from meat to vegetables, crops, or fruits are K+-rich sources. Organisms have selected the same regulatory mechanism, which involves mainly insulin, to clear the plasma of both glucose and K+ and avoid the harmful consequences of increasing their plasma concentration. Indeed, this hormone acts on skeletal muscles by increasing the density of glucose transporters (glut-4) and Na,K-ATPase at the cell surface that move glucose and K+ into the cells respectively. During the fast period, this intracellular reserve of K+ is then slowly released into the plasma compartment from where it is eliminated by the kidney, allowing the body to maintain a constant plasma K+ value. This process is described by Youn and McDonough (2009) as an ‘altruistic specialization to donate’ intracellular K+ to extracellular compartments.

External balance Intestinal K+ excretion The colon contributes to K+ homeostasis by its ability to either reabsorb or secrete K+. Under a normal K+ diet, < 10% of the K+ intake is excreted in the faeces which, compared with kidney excretion, seems negligible. However, under special dietary conditions or pathological status like end-stage renal failure, the colonic contribution to K+ homeostasis becomes more crucial. Colonic reabsorption of K+ is mediated by the H+,K+-ATPase type 2 (HKA2, see below for details), which is expressed at the apical (luminal) side of the cells. Lack of HKA2 increases the faecal excretion of K+ twofold under a normal diet and this enhancement is even stronger (fivefold) during K+ restriction. The secretion of K+ is mediated at the basolateral side by the Na+,K+-ATPase and the Na+-K+-2Cl− cotransporter NKCC1; at the apical side, potassium exits into the lumen via Ca2+-dependent K+ channels known as big K+ (BK) channels. This secretion pathway is upregulated through a mechanism involving stimulation of BK activity by aldosterone and adrenaline when K+-rich food is provided to animals.

Urine K+ excretion Renal excretion of K+ results from glomerular filtration and transport of K+ along the renal tubule. The daily filtered load of K+ (~

chapter 23 

Internal balance

potassium homeostasis

External balance

Intracellular compartments (98% of total body K+)

Liver 200 mmol

RBC 200 mmol Extracellular compartment (2% of total body K+)

Others 400 mmol 70 mmol

Food intake 70–100 mmol K+

Fecal excretion 5–10 mmol K+

Muscles 3000 mmol Renal excretion 70 mmol K+ Total body K+ = 55 mmol/kg

Fig. 23.1  Schematic representation of K+ repartition in the different compartments of the organism. Ninety eight per cent of body K+ is present in intracellular compartments, mainly in muscles. The priority of the organism is to maintain the extracellular K+ concentration within a narrow range despite the variations of K+ intake. For this purpose, K+ may be pumped or released from internal stores (internal balance) or may be reabsorbed or secreted by kidneys and intestine (external balance).

750 mmol) far exceeds normal K+ intake (~ 70–100 mmol/day), demonstrating a massive capacity for tubular K+ reabsorption. Nevertheless, in some circumstances, for example, when glomerular filtration rate (GFR) is decreased or K+ intake is high, the rate of K+ excretion may exceed its rate of filtration, revealing the existence of a tubular secretion mechanism. However, whatever the condition and the net flux of K+ within the kidney, urinary K+ excretion results from both reabsorption and secretion processes that originate in specific segments of the nephron. As schematically shown in Fig. 23.2, the initial segments of the nephron reabsorb the bulk of filtered K+, whereas the late portions either reabsorb or secrete K+ so as to match homeostasis requirements. Superimposed on these processes are mechanisms that allow for the recycling of K+ within the kidney medulla. Before analysing the cellular mechanisms of K+ transport in the different segments of the nephron, we will review briefly the main properties of the proteins involved in the transport of K+.

encompass the eight-amino acid signature of this family. A special feature of the X,K+-ATPase subgroup is their requirement for an additional subunit (called the β-subunit) in addition to the P-type ATPase itself (the α-subunit) to form a mature and functional transporter. The common mechanism of ion transport by P-type ATPases involves the transient phosphorylation of an aspartyl residue during the functional cycle, and the transition between two main and opposite conformations: the E1 conformation which exhibits a high affinity for intracellular Na+ or H+ (i.e. the ion that has to be moved out of the cell against its gradient) and ATP, and the E2 conformation which exhibits a high affinity for K+ (which has to be moved into the cell against its gradient). The passage from one conformation to the other depends on the phosphorylation of the P-ATPase. Two types of X,K+-ATPase participate in the renal regulation of the K+ balance, namely the Na+,K+-ATPase and the H+,K+-ATPase.

Potassium transporters

The Na+,K+-ATPase is an ubiquitous plasma membrane enzyme that transports two K+ into and three Na+ out the cells by using the energy of the hydrolysis of one molecule of ATP. The Na+,K+-ATPase is composed of two obligatory subunits, the catalytic subunit (called α-subunit), a 10-transmembrane-spanning domain protein, and the β-subunit, a single transmembrane domain, type II glycoprotein with a large ectodomain. The Na+,K+-ATPase is inhibited by a family of compounds known as cardiac glycosides or digitalis (e.g. digoxin and ouabain). The presence of the β subunit is necessary for the structural and functional maturation of the pump and influences the kinetic properties of the α subunit. The existing 4 α (α1 to α4) and 3 β (β1 to β3) isoforms exhibit a different tissue distribution and can assemble to produce Na+,K+-ATPase isozymes with different transport and pharmacological properties. The renal Na+,K+-ATPase consists of an α1β1 isozyme that is present all along the nephron at the

X,K+-ATPases Primary active transporters utilize the energy released by the hydrolysis of ATP to move ions against their electrochemical gradient. Membrane proteins able to achieve this goal therefore display the properties of ion transporters (ion binding and movement through a hydrophilic environment) and of enzymes (binding and hydrolysis of ATP), both properties being interconnected. These active transporters are referred to as pumps or as ATPases. X,K+-ATPases belong to a large family of proteins, the P-type ATPases, that share common topogenic motifs and use a similar mechanistic pathway to transport ions across the membrane. All P-type ATPases possess a core structure comprising three pairs of transmembrane domains and the connecting loops which together

X,K+-ATPases

205

206

Section 2  

fluid, electrolyte, and renal tubular disorders

CNT

S1

Cortex

DCT CTAL

CCD

S2

OMCDo

S3

Outer stripe Outer medulla

MTAL OMCDi

Inner stripe

tDL IMCDi tAL IMCDt

Inner medulla

Fig. 23.2  Schematic representation of K+ movements along the nephron. This diagram shows the nomenclature of nephron segments used in this chapter and the net movement of potassium in each of them: blue and grey arrows represent secretion and reabsorption of K+ respectively. S1, proximal convoluted tubule; S2 and S3 cortical and medullary proximal straight tubule; tDL and tAL, thin descending and ascending limbs of Henle’s loop; MTAL and CTAL, medullary and cortical thick ascending limb; DCT, distal convoluted tubule; CNT, connecting tubule; CCD and OMCD, cortical and outer medullary collecting duct (o and i subscripts refer to portions of OMCD located in the outer and inner stripes of the outer medulla); IMCD, inner medullary collecting duct (i and l subscripts refer to the initial and late half of IMCD).

basolateral side of the cells. This isozyme exhibits a high affinity for extracellular K+ (around 1 mmol/L) and internal Na+ (around 9 mmol/L) and a rapid turnover rate (50 transported charges/s at 20°C). These properties indicate that the concentration of intracellular Na+ is the rate-limiting factor of the renal α1β1 Na+, K+-ATPase. The β3 subunit has also been observed in the kidney but its presence does not modify the main kinetic properties of the Na+,K+-ATPase. Na+,K+-ATPase activity is two to six times higher in the thick ascending limb of Henle’s loop and the distal convoluted tubule than in proximal and terminal segments (Katz et al., 1979). In addition to these obligatory subunits, members of a family of small, one-transmembrane domain proteins interact with the Na+,K+-ATPase and modulate its kinetic properties. These proteins, called FXYD in reference to a common amino-acid motif, exhibit specific expression profiles and affect differentially Na+ and K+ affinities. The two main FXYD family members expressed in the kidney are FXYD2 isoforms a and b (previously known as the γa and γb subunits of the Na+,K+-ATPase) and FXYD4 (known as the corticoid hormone-induced factor (CHIF)) (Sweadner et al., 2003). It seems established that FXYD2 isoforms are mainly expressed in the thick ascending limb (TAL) of Henle’s loop, whereas FXYD4 is expressed exclusively in the distal part of nephron (Shi et al., 2001). The functional relevance of these regulatory subunits is still under investigation. There is evidence that FXYD4 plays a role in the regulation of the renal K+ balance. Its expression is upregulated during K+ loading and the association of FXYD4 with Na+,K+-ATPase decreases its affinity for external K+. Combined together, these observations suggest that FXYD4 helps to excrete K+ by allowing the Na+,K+-ATPase to function more efficiently when extracellular

K+ is high. However, when placed under a high-K+ diet, FXYD4 null-mice exhibit only a mild phenotype featuring a higher urine volume, but no clear K+ excretion defect (Aizman et al., 2002).

X,K+-ATPases Like the Na+,K+-ATPase, the H+,K+-ATPase is an heterodimer composed of an α and a β subunit. There are two isoforms of the α subunit that were originally distinguished by their tissue expression and their pharmacological properties. The H+,K+-ATPase type 1 (known as the gastric H+,K+-ATPase; HKA1) is mainly present in the parietal cells of the stomach where it serves to acidify the gastric content. It is associated with a specific β subunit (the HKβ) and is sensitive to pharmacological compounds like omeprazole and Schering 28080, but insensitive to cardiac glycoside (Sachs et al., 1995). The H+,K+-ATPase type 2 (known as the non-gastric or the colonic H+,K+-ATPase; HKA2) is abundantly expressed in the colon and is sensitive to both Schering 28080 and ouabain. In heterologous expression systems, all β subunits (subunits β1, β2, β3 of the Na,K-ATPase and the HKβ) may be coupled with the α subunit of the H+, K+-ATPase type 2 with no obvious modification of its kinetic properties. However, the nature of the β subunit associated with H+, K+-ATPase type 2 in vivo is not well defined and could be tissue-specific. As opposed to HKA1, which specifically exchanges intracellular protons against extracellular K+ ions, HKA2 seems more flexible in terms of the ions that it can transport. Indeed, since its first characterization, it has been established that the H+ and K+ fluxes are not equal and that another cation in addition to H+ should be simultaneously transported to maintain electroneutrality. Further investigations have shown that HKA2 expression not only modifies

chapter 23 

the intracellular content of Na+, but also that the HKA2-mediated 86Rb flux is Na+-dependent with an affinity around 9 mmol/L. The in vivo consequences of the ability to transport Na+ have not been elucidated yet. Another particularity of the HKA2 is the presence of a truncated form (HKα2b) that lacks 108 amino acids at its N terminus. This truncated form arises from an alternative splicing that fuses the exon 1 to the exon 2. The short form of HKA2 is expressed with the long form both in colon and kidney, but here again its physiological relevance has not been elucidated. Both types of H+,K+-ATPases are expressed in the distal part of the nephron (mainly cortical collecting duct (CCD), outer medullary collecting duct (OMCD), and inner medullary collecting duct (IMCD)) and some reports suggest the presence of the H+,K+-ATPase type 2 also in the TAL and the connecting tubule (CNT) (for review, see Silver and Soleimani, 1999). As opposed to the Na+,K+-ATPase, both H+,K+-ATPases are located at the apical side of the cells. In the distal nephron, HKA1 seems restricted to the intercalated cells, whereas HKA2 has been identified in both principal and intercalated cells.

K+ channels K+ channels are the most widely distributed ion channels and display a large diversity of biophysical characteristics. However, they all share a common tetrameric structure (homo- or heteromers) with each subunit showing a common core structure that consists of two transmembrane domains separated by a P-loop. There are four main families of K+ channels: calcium-activated K+ channels, inwardly rectifying K+ channels (through which K+ passes more easily in the inward direction), tandem pore domain K+ channels that are constitutively open or display a basal activity, and voltage-gated K+ channels that open or close in response to changes in the membrane voltage. Members of all these families of K+ channels are present in the kidney where they ensure a large variety of functions. At the basolateral side, they contribute to: (a) hyperpolarizing the cell membrane and thereby generating a driving force for other transport systems, (b) the maintenance of the cell volume, and (c) the recycling of K+ ions accumulated by the Na+,K+-ATPase. On the opposite side, they participate in apical K+ recycling in TAL cells (along with NKCC2) and in K+ secretion in the distal part of the nephron. In view of the scope of this chapter, we will mainly focus on K + channels that have a direct impact on K+ homeostasis, namely those expressed in the distal part of the nephron: renal outer medulla K+ (ROMK) (Kir1.1), BK channel (or Maxi K channel), and Kir 4.1/ Kir 5.1.

ROMK The ROMK channel, the first K + channel to be functionally cloned, belongs to the family of KCNJ inwardly rectifying K+ channels (Kir). Its structure is characterized by two transmembrane domains with intracellular NH2- and COOH- termini. The extracellular loop that connects both transmembrane segments forms a P-loop that contains the K+ selectivity filter. A functional channel requires the assembly of four ROMK monomers. There are three ROMK isoforms (from ROMK1 to ROMK3, or Kir1.1a to Kir 1.1c) that originate from different splicings, differ in the sequence of their NH2- terminal domain, but exhibit similar biophysical properties. They also differ in their location along the

potassium homeostasis

nephron. ROMK1 has the most restricted localization, from DCT to OMCD. ROMK3 has the largest distribution, from the medullary TAL to the OMCD. ROMK2 displays a similar localization as ROMK3 but is not present in the TAL (Boim et al., 1995). The physiological relevance of the existence of these three isoforms is not well understood. The generation of ROMK-deficient mice finally confirmed the molecular signature of the small K+ conductance described originally in the TAL and the CCD. It also provides the first model of Bartter syndrome type 2 (Lorenz et al., 2002). The biophysical characteristics of ROMK channels expressed in heterologous systems correspond to those of the predominant apical K+ current measured in TAL, namely a conductance of 35-pS with a high open probability (about 0.9) and a high sensitivity to intracellular pH (acidification leading to inhibition of ROMK). However, some discrepancies have also been observed. For instance, the 35-pS conductance is barely rectified and is regulated by ATP and sensitive to glibenclamide. These native properties suggest that ROMK is associated with proteins that may belong to the ATP-binding cassette protein family (ABC protein). It has been shown that expression in heterologous system of ROMK along with the sulfonylurea receptor, a subunit of the Kir6.1 K+ channel, or with CFTR (two ABC proteins), leads to characteristics that correspond to those of the native channel. Interestingly, in CFTR-null mice the ROMK conductance is not regulated by intracellular ATP and is not sensitive to glibenclamide anymore, indicating that CFTR and ROMK interact functionally in vivo.

Big K+ channel The BK channel story started with the identification of an apical K+ conductance of 90-pS that was activated by an elevated intracellular Ca2+ concentration ([Ca2+]i) (Hunter et al., 1984). In addition to its large conductance, this channel displays a low open probability and a high sensitivity to iberiotoxin. Because of these characteristics (low Po and activation by high [Ca2+]i), it has been proposed that this channel is not functional under basal conditions but more likely activates after the rise of [Ca2+]i secondary to a depolarization, an hypo-osmotic stress or a membrane stretch. The molecular identification of this conductance was first done in 1993 by cloning mouse orthologues of the Drosophila gene slo. More recent studies have clarified the structure of BK channels. They consist of two subunits, the pore-forming α subunit (encoded by the slo1 gene) and a β-regulatory subunit which is devoid of channel function but interferes with α subunit properties. In the kidney, there is one isoform of the α-subunit (slo1) and two of the β-subunit (β2 and β4). The BK channel conductance has been observed in the CNT and the CCD, both in principal and intercalated cells. As mentioned above, the characteristics of this channel do not support a constitutive activation. In conventional patch-clamp experiments on split-open tubules, BK channel activity is not observed. Other experiments have shown that BK activity is activated by fluid flow (Taniguchi and Imai, 1998), which was further confirmed by the demonstration that flow-stimulated K+ secretion is absent in BK knockout mice (Pluznick et al., 2003). Renal epithelial cell are equipped with a cilium that acts as a ‘flow-sensor’ and it is now proposed that flow-induced bending of this cilium activates apical stretch-sensitive Ca2+-channels and increases [Ca2+]i which in turn activates BK.

207

208

Section 2  

fluid, electrolyte, and renal tubular disorders

Basolateral K+ channels in the collecting duct: example of Kir4.1/Kir5.1 The presence of K+ conductances in the basolateral membrane of CCD epithelial cells is required for (a) enabling K+ recycling (which allows the Na+,K+-ATPase to function efficiently), (b) setting the basolateral membrane potential, and (3)  determining the direction of the net basolateral K+ flux. In vivo studies demonstrate the presence of at least three different K+ conductances (20-pS, 40-pS, and 76-pS) in the basolateral membranes of native mouse CCD (Lachheb et al., 2008). Among them, the 40-pS conductance turns out to be the most abundant and to have biophysical properties similar to those of heterotetrameric Kir4.1/Kir5.1 K+ channels. This K+ channel, as its name implies, is formed of two proteins, Kir4.1 (encoded by the KCNJ10 gene) and Kir5.1 (encoded by the KCNJ16 gene). Kir4.1 was originally cloned from brain tissue on the basis of sequence similarity with ROMK channels. It shows a high expression level at the basolateral side of DCT, and CNT and CCD principal cells. The biophysical characteristics of Kir4.1 as determined in Xenopus oocytes revealed a channel with a high open probability and a small (around 25-pS) strongly rectified conductance. The channel displays a complex sensitivity to intracellular pH: acidification reduces its open probability, but increases its conductance. In an effort to identify K+ channels similar to ROMK, another member of this gene family, referred to as Kir5.1, was discovered and characterized. When expressed alone in Xenopus oocytes, Kir5.1 does not produce any ionic current. However, it specifically assembles with Kir4.1 and modifies its conductance properties. The most striking effects of Kir5.1 on the properties of Kir4.1 are the increase in the current amplitude, the stronger rectification and the appearance of a time-dependent component affecting the current amplitude. The channel formed by Kir4.1 and Kir5.1 is a heterotetramer and its biophysical properties depend on the ratio of both subunits and their position in the tetramer. Assembly of Kir5.1 with Kir4.1 also provides a higher sensitivity to intracellular pH. In the kidney, in contrast to Kir4.1, Kir5.1 has a broad localization along the nephron. Kir4.1 and Kir5.1 genes are of particular interest since recent studies provide evidence for their involvement in human pathologies. Polymorphisms in KCNJ10, the human genes encoding Kir4.1, are associated with the SeSAME/EAST syndrome, a rare disease leading to neurological and renal dysfunctions (Scholl et al., 2009) whereas Kir5.1-null mice exhibit a renal phenotype exactly opposite to that of SeSAME/EAST patients (Paulais et al., 2011). These recent findings outline the importance of a functional basolateral K+ channel in the distal part of the nephron.

K+ carriers: NKCC and KCC The electroneutral cation-Cl− cotransporters belong to the solute carrier family slc (slc12 subfamily). These carriers utilize the Na+ or K+ gradient created by the Na+,K+-ATPase to move chloride ions across the plasma membrane. This family is divided into three large subgroups, the Na+-Cl− cotansporters (NCC), the K+-Cl− cotransporters (KCC), and the Na+-K+-Cl− cotransporters (NKCC). For the purpose of this chapter we will focus on the K+-dependent members of this family, the KCC and NKCC transporters that are expressed in the kidney.

KCC K+-Cl− cotransporters were identified first as mediators of swelling-activated K+ efflux in red blood cells. Further studies led

to the cloning and molecular characterization of 4 isoforms of KCC (KCC1 to 4) encoded by the slc12a4 to slc12a7 genes. KCC transporters, as all slc12 family members, have 12 transmembrane-spanning domains and intracellular N- and C-terminal parts. More specifically, they exhibit a large and heavily glycosylated extracellular loop between transmembrane domains 5 and 6.  KCC2 is a neuronal-specific isoform but the three other isoforms are present in the kidney: KCC1 is expressed all along the nephron; KCC3 is mainly found in the proximal tubules where it participates in the regulation of cell volume and KCC4 is confined to the basolateral membranes of TAL, DCT, and CCD α-intercalated cells. In addition to deafness, targeted disruption of the gene encoding KCC4 leads to renal tubular acidosis (Boettger et al., 2003).

NKCC Two genes, slc12a1 (NKCC2) and slc12a2 (NKCC1), were found to encode for the Na+-K+-Cl− cotransporters. These isoforms exhibit different tissue and cell patterns of expression. NKCC1 is present in almost all tissues and organs and, in polarized epithelia; it is located at the basolateral membrane. Conversely, NKCC2 is a kidney-specific isoform that displays a restricted expression at the apical side of TAL cells. NKCC uses the energy of the Na+ gradient to move K+ and Cl− into the cells with a stoichiometry of 1Na+:1K+:2Cl−. NKCC resembles KCC with 12 transmembrane-spanning domains and intracellular N- and C-terminal extremities. However, NKCCs exhibit a large extracellular loop connecting transmembrane domains 7 and 8. NKCCs are also characterized by their sensitivity to loop diuretics such as bumetamide or furosemide. The molecular diversity of NKCCs is increased by the existence of alternate splicing and dual polyadenylation sites. Alternate splicing generates three variants (A, B, and F) that differ in the sequence of a 32 amino-acid cassette at the beginning of the second transmembrane domain. NKCC2A, B or F exhibit specific localization in the TAL with expression of the F form in the deep part of the medullary TAL, that of the A form extending all along the TAL to the macula densa, and that of the B form being mostly cortical. Interestingly, expression of the different isoforms in Xenopus oocytes revealed kinetic differences, in terms of Na+, K+, and Cl− affinities for instance. However, their physiological relevance remains to be determined. In the TAL, the presence of NKCC2 enables the net reabsorption of Na+ and Cl−. Human mutations in the NKCC2 gene that lead to a loss of function are related to Bartter syndrome type 1 (Simon et  al., 1996) Other human polymorphisms in this gene have been associated with protection against hypertension (Ji et al., 2008); however the functional effect of these polymorphisms on NKCC2 activity remains to be determined.

Cellular mechanisms of potassium transport along the nephron Proximal tubule The first two-thirds of the proximal tubule (approximately the S1 and S2 segments) reabsorbs 30–70% of filtered K+ (Malnic et al., 1964) whereas the last third (the S3 segment) secretes K+ (Jamison, 1987). K+ secretion along the S3 segment is thought to occur via passive diffusion through the paracellular pathway, owing to the presence of a high K+ concentration in the medulla interstitium (Jamison, 1987). At least three mechanisms account for K+

chapter 23 

reabsorption along the initial proximal tubule:  solvent drag and diffusion through the paracellular route and active transcellular reabsorption. Active reabsorption of K+ in the proximal tubule is supported by indirect in vivo pieces of evidence and by direct measurement of the electrochemical gradient in rabbit PCT perfused in vitro. K+ transport proteins known to be present in proximal tubule cells include K+ channels at the apical membrane, and Na+,K+-ATPase, K+ channels and a K-Cl cotransporter at the basolateral membrane (Fig. 23.3A). Thus, in the absence of an active mechanism accounting for apical uptake of K+, and given the transepithelial voltage prevailing along the proximal tubule (from slightly lumen-negative in the early S1 to slightly lumen-positive in the S2 and S3 segments (Fromter, 1984)), these cells appear equipped to secrete rather than to reabsorb K+. Two mechanisms may circumvent this apparent paradox. Firstly, a H+,K+-ATPase activity has been described in the proximal tubule which, if efficient at the apical membrane, might account for active entry of K+. Secondly, Weinstein has proposed an elegant model (Weinstein, 1988) in which he considers a fourth compartment constituted by the interspace limited by the lateral membranes of proximal tubule cells, the tight junction and the basement membrane (Fig. 23.3B). According to this model,

(A)

the Na+,K+-ATPase located on lateral membranes pumps K+ from the interspace into the cell, thereby providing a driving force for K+ diffusion across the tight junction (assuming that K+ diffusion across the basement membrane is slower). Intracellular K+ then diffuses across the basal membrane. Whatever the mechanism, active K+ reabsorption would account for only 20% of proximal tubule reabsorption. Evidence for solvent drag comes from the observations that (a)  K+ reabsorption by proximal tubules is dependent on fluid reabsorption, and (b) experimentally inducing water secretion by increasing the osmolality of the luminal fluid induces K+ secretion. Although proximal tubule reabsorbs approximately 60–70% of filtered fluid, solvent drag-mediated reabsorption is thought to account for only 20% of K+ reabsorption. This stems from the fact that the bulk of water is reabsorbed via the transcellular route through K+-impermeable aquaporins and not via the paracellular pathway. Many pieces of evidence support passive diffusive reabsorption of K+ in the late portion of the proximal convoluted tubule and S2 segment. In these portions of the proximal tubule, the favourable gradient for passive diffusion results from the slightly lumen-positive voltage and from the fact that K+ concentration is slightly higher in

(C) Apical

(E) Apical

Basal

–

Na+ K+

Basal

Apical

Na+ K+

K+

K+

Cl–

Cl–

K+

K+ + (B)

Apical

Basal

K+

(D)

–

Apical

+

K+

Na+ K+ H+ K+

Cl–

Cl– –

Basal

Apical

Na+

K+

Cl–

– (F)

Basal

Na+ Cl–

Basal

Na+

2Cl–

Cl–

+

potassium homeostasis

+

–

K+

+

Fig. 23.3  Cell models of K+ transport. Black circles represent primary active transporters (ATPases), grey circles are cotransport of counter-transport systems, and rectangles are channels. (A) Proximal convoluted tubule cell. Transepithelial voltage varies from slightly lumen-negative value in early convolutions to slightly negative in late convolutions. This model cannot account for the massive reabsorption of K+ originating along the S1 segment, as it suggest that only K+ secretion can occur. (B) Proximal convoluted tubule cell model developed by Weinstein to account for K+ reabsorption. This model integrates an additional intercellular compartment limited by lateral membranes, intercellular junctions, and basement membrane, and considers that Na+,K+-ATPase is mainly expressed at the lateral membrane. (C) Thick ascending limb of Henle’s loop. K+ reabsorption in this segment may occur through the transcellular route (considering that some K+ entering via NKCC2 leaves the cells through the basolateral membrane rather than being recycled across the apical membrane) and the para cellular route (owing to the lumen-positive transepithelial voltage generated by transcellular reabsorption of NaCl). (D) Initial distal convoluted tubule (DCT1). The lumen-negative transepithelial voltage favours K+ secretion rather than its recycling across the basolateral membrane. (E) Late distal convoluted tubule (DCT2). The presence of a diffusive Na+ entry pathway in parallel with electroneutral NCC depolarizes the apical membrane and increases the driving force for K+ secretion. (F) Connecting tubule and collecting duct principal cell. The high lumen-negative transepithelial voltage brought about by diffusive apical entry of Na+ generates a high driving force for K+ secretion. Under K+ restricted conditions, apical H+,K+-ATPase energizes K+ reabsorption.

209

210

Section 2  

fluid, electrolyte, and renal tubular disorders

the tubular fluid than in plasma (owing to the primary iso-osmotic reabsorption of sodium bicarbonate and sodium chloride prevailing in the early proximal convoluted tubule). Paracellular diffusive reabsorption would account for 60% of proximal tubule K+ reabsorption. Although it is quantitatively important, it is worth mentioning that the rate of K+ reabsorption along the proximal tubule is not a major determinant of the urinary excretion of K+. As a matter of fact, the rate of K+ reabsorption along the loop of Henle increases proportionally with K+ delivery to this segment so that the fraction of K+ leaving the loop of Henle is approximately 10% of filtered K+, whatever the proximal tubule reabsorption rate.

mutations of either NKCC2 or ROMK, as well as mutations affecting the apical Cl− conductance (Hebert, 2003). It is thus legitimate to assume that most K+ reabsorption along the TAL takes place through the paracellular pathway. Although the cellular mechanism of Na+, K+, and Cl− reabsorption is identical in the medullary and cortical portions of the TAL, these two segments do not play the same role regarding ion balance. Due to the striking differences between blood circulation within the kidney cortex and medulla, most ions reabsorbed along the medullary thick ascending limb are recycled within the medulla whereas those reabsorbed within the cortex can be drained out of the kidney.

Loop of Henle

Distal convoluted tubule, connecting tubule, and cortical collecting duct

The loop of Henle consists of the proximal straight tubule, the thin descending limb of Henle’s loop, the thin ascending limb (only in juxtamedullary nephrons) and the thick ascending limb (TAL) with its medullary and cortical parts. Thin segments of the loop of Henle show no apical K+ conductance, indicating that any transepithelial transport occurs via the paracellular pathway which is highly permeable to K+ (Imai et al., 1987). As the tubule fluid flows along the thin descending limbs, it enters regions with increasing concentrations of peritubular K+ whereas the opposite occurs along the thin ascending limb. This peritubular K+ gradient is responsible for passive secretion and reabsorption of K+ along the thin descending and ascending segments respectively. Although these respective fluxes are not quantifiable, it is assumed that they are similar and therefore that the load of K+ leaving the proximal tubule equals that delivered to the TAL. As already mentioned, the TAL reabsorbs variable amounts of K+, depending on the amount delivered (Greger, 1985). In some circumstances, for example, in response to loop diuretic treatment, it can also secrete K+. TAL cells display both active and passive K+ transport systems at both borders (Fig. 23.3C). At the apical membrane, they express an electroneutral Na+-K+-2Cl− cotransporter (NKCC2), which allows for active entry of K+ and Cl− coupled to passive entry of Na+, as well as K+ channels including ROMK. The basolateral membrane expresses Na+,K+-ATPase pumps, K+ and Cl− channels and possibly a K+-Cl- cotransporter These transport systems make possible the massive transcellular reabsorption of NaCl which is associated with depolarization of the basolateral membrane (due to the electrogenic diffusion of chloride) and hyperpolarization of the apical membrane (due to the electrogenic diffusion of K+); they thereby generate a lumen-positive transepithelial voltage (Greger, 1985). In turn, the transepithelial voltage drives passive reabsorption of Na+ and K+ via the paracellular pathway. The K+ ions that have accumulated within the cell above equilibrium by both the Na+,K+-ATPase and NKCC2 leave the cell via either apical or basolateral membranes. Thus, the net transcellular flux of K+ depends on the balance between apical versus basolateral efflux. It should be recalled that apical recycling of a major fraction of K+ entering the cell via NKCC2 is a requirement for sustained NaCl reabsorption (Greger, 1985): because K+ concentration in the tubular fluid is approximately 30-fold lower than that of sodium and chloride, the activity of NKCC2 would stop rapidly unless K+ is re-introduced in the tubular fluid. This is well illustrated by the fact that Bartter syndrome, a monogenic disease featuring inhibition of ion transport in the TAL, can be induced by loss-of-function

The nephron portion located downstream the cortical thick ascending limb and the macula densa, referred to as the distal tubule, is complex both by its architecture and its cell composition (see Fig. 23.2). Classically, it is subdivided into a distal convoluted tubule (DCT), a connecting tubule (CNT), and a cortical collecting duct (CCD), but (a)  there are marked differences between superficial and juxtamedullary nephrons, the latter showing a much longer CNT, and (b) this subdivision does not reflect properly the axial functional heterogeneity of the distal tubule. The DCT is traditionally considered to be made of a single type of cells, but the cells from the initial and distal parts of this segment (respectively called DCT1 and DCT2) display different transport systems. Conversely, the CNT and CCD are both made of principal cells (called CNT cells in case of the CNT) and intercalated cells, and each of these cell types displays similar transport systems in the CNT and CCD. In addition, the quantitative contributions of these different structures to ion transport are not known because in vivo flux measurements by micropuncture studies either apply to combined structures (late DCT and early CNT) or cannot be done (late CNT and CCD). For these reasons, we find it more convenient to address the mechanisms of K+ transport along the distal tubule at the level of the different cell types. Globally, the distal tubule is the main site of K+ transport adaptation. Under ‘normal’ conditions (of dietary K+ ingestion and kidney function) the distal tubule secretes small amounts of K+ to balance the load delivered at the DCT (~ 10% of the filtered load) with the dietary intake. It also shows a great ability to secrete large amounts of K+ in response to K+ overload. Conversely, it is able to reabsorb almost all the K+ that is delivered by the TAL in case of severe K+ restriction. Cells from the DCT1 display Na+,K+-ATPase and K+ and Cl− channels at their basolateral pole, and K+ channels (mainly ROMK) and K+-Cl− cotransporters in parallel with Na-Cl cotransporters (NCC) at their apical border (Fig. 23.3D). The K+ ions which have accumulated in the cell by the Na+,K+-ATPase should preferentially leave the cell across the basolateral membrane, owing to the membrane depolarization brought about by conductive Cl− exit. However, part of the Cl− that has accumulated in the cells by NCC may recycle across the apical membrane by the K+-Cl− cotransporter and thereby drive K+ secretion. DCT2 cells have the same transport systems as DCT1 cells except that they also display an apical Na+ channel (ENaC) which is responsible for part of Na+ reabsorption (Fig. 23.3E). This electrogenic

chapter 23 

Na+ entry depolarizes the apical membrane and thereby facilitates K+ secretion through apical K+ channels. It also tends to generate a slightly lumen-negative transepithelial voltage which can drive K+ secretion through the paracellular pathway. CNT cells and CCD principal cells differ from DCT cells by the facts that (a) ENaC is the main, if not unique, pathway for apical entry of Na+, (b) they express ROMK and BK channels at their apical membrane, and (c) there is no system for apical Cl− entry (Fig. 23.3F). Thus, Na+ reabsorption by principal cells strongly depolarizes the apical membrane which generates a driving force for K+ exit. It also generates a lumen-negative transepithelial voltage, but the paracellular back diffusion of Na+ and K+ is negligible because the intercellular junctions are mostly impermeable to ions (tight epithelium). It is important to stress that, according to this mechanism, K+ secretion by principal cells, the main pathway for K+ secretion along the distal tubule, is dependent on Na+ reabsorption and transepithelial voltage. Principal cells are also responsible for K+ reabsorption during K+ depletion because, under these conditions, they express the type 2 H+,K+-ATPase at their apical border. Intercalated cells are classically considered responsible for the regulation of acid-base balance. Type A  intercalated cells can secrete protons and ammonium whereas type B intercalated cells can secrete bicarbonate. Recently, it was also shown that type B intercalated cells mediate NaCl reabsorption (Leviel et al., 2010). Under basal conditions, intercalated cells are not equipped for transepithelial K+ transport and are considered to be silent. However, because proton secretion by type A intercalated cells is mediated by the electrogenic V-type H-ATPase which generates a lumen-positive transepithelial voltage, activation of proton secretion during metabolic acidosis reduces the driving force for secretion by principal cells. Intercalated cells express type 2 H+,K+-ATPase at their apical border during K+ depletion, and thereby may participate in K+ reabsorption although no basolateral K+ conductance has been described.

Medullary collecting duct Based on topographical, morphological and functional criteria, the medullary collecting duct is divided into an outer and an inner medullary segment (OMCD and IMCD respectively), each of which is further subdivided into two sections, the outer and inner stripes for the OMCD (OMCDo and OMCDi respectively), and the initial (first 50%) and late (last 50%) portions for the IMCD (IMCDi and IMCDl; see Fig 23.2). The OMCDo consists of the same cell types as the CCD and it has the same K+ transport systems and functional properties, except that the rates of Na+ reabsorption and K+ secretion are smaller. The OMCDi only displays passive paracellular K+ transport, either secretion or reabsorption depending on the transepithelial electrochemical gradient. K+ reabsorption along the OMCDi is the primary motor of K+ recycling. The IMCD secretes K+ except during dietary K+ depletion when it reabsorbs it (Backman and Hayslett, 1983). Besides the Na+,K+-ATPase, the basolateral membrane of IMCD cells display a Na+-K+-2Cl− cotransporter (NKCC1) that accumulates K+ within the cell, and K+ and bicarbonate conductances. The apical membrane contains an amiloride-sensitive cation channel, with a similar conductance to Na+, K+ and ammonium that likely mediates part of

potassium homeostasis

K+ secretion. The IMCD has a low transepithelial resistance which enables paracellular K+ secretion. The molecular mechanism of K+ reabsorption during dietary K+ depletion remains unknown.

Potassium recycling Medullary K+ recycling makes it possible to highly concentrate K+ in the inner medulla interstitium and to thereby excrete K+ via secretion in the IMCD (Jamison, 1987). For example, during water deprivation, water reabsorption in the OMCD increases the concentration of K+ in the luminal fluid and enables its passive reabsorption along the OMCDi and its secretion into the S3 segment of the proximal tubule and into the thin descending limb. The K+ that has accumulated in the fluid of the thin descending limb is then reabsorbed passively along the thin ascending limb and accrues in the inner medulla interstitium. This prevents K+ reabsorption along the IMCD and enables the excretion of the K+ that was secreted upstream in the CNT and CCD. In the absence of this recycling, the K+ ions that are concentrated in the urine during water deprivation would be reabsorbed in part along the IMCD, and K+ excretion would be reduced. Note that K+ reabsorption along the TAL may also contribute to K+ recycling.

Regulation of urinary potassium excretion The urinary excretion of K+ is modulated by many factors which control K+ secretion and/or K+ reabsorption mostly beyond the DCT. Among these factors, which include dietary K+ intake, plasma K+ concentration, acid–base status, hormones, tubular flow rate, and tubular sodium concentration, aldosterone is considered to play a central role because it participates in a feedback regulatory loop: an increase in the plasma concentration of K+ triggers the adrenal secretion of aldosterone which promotes K+ secretion by the distal nephron. Aldosterone is also involved in the complex regulatory pathways triggered by several of the above listed factors that control urinary K+ excretion.

Mineralocorticoids Mechanism of action of mineralocorticoids Aldosterone is an adrenal steroid that controls the balance of Na+, Cl−, K+, and H+ and therefore controls extracellular volume, blood pressure, acid–base balance, and plasma K+ concentration. Because they modulate the metabolism of minerals, aldosterone and its derivatives are referred to as mineralocorticoids. The main renal effect of aldosterone is to induce Na+ retention and K+ secretion. The mineralocorticoid receptor (MR) belongs to the family of nuclear receptors, a specific group of ligand-activated transcription factors. The MR displays similar affinities for aldosterone and glucocorticoids. Since the latter are present in the plasma at concentrations approximately 1000-fold higher than aldosterone, they should saturate the MR. The permanent activation of the MR by glucocorticoids is prevented by the presence of a type 2 11β-hydroxysteroid dehydrogenase (HSD2) which converts glucocorticoids into metabolites that have no affinity for the MR In the kidney, MR is expressed from the DCT to the OMCD, a nephron portion consequently defined as the aldosterone-sensitive distal nephron (ASDN). Specifically, MR is expressed in DCT cells and in the principal cells of the CNT and collecting duct. HSD2 is co-expressed

211

212

Section 2  

fluid, electrolyte, and renal tubular disorders

with MR in all these nephron segments except the DCT1 which therefore should be under permanent, positive MR control. As for its targets, aldosterone modulates the expression of several genes that contribute to Na+ reabsorption. Aldosterone has an ‘early’ effect enabling the over expression of regulators of Na+ transport such as the kinase SGK1 (serum and glucocorticoid induced kinase) (Chen et al., 1999), which increases the abundance of Na+,K+-ATPase, NCC and ENaC at the cell surface. A ‘late’ effect of aldosterone is to directly induce the expression of these transporters and therefore to increase their total number (for references, see Verrey, 1999). Aldosterone also induces the hypertrophy and hyperplasia of ASDN (reviewed in Stanton, 1989), but this is an indirect effect that seems to be dependent on the stimulation of Na+ reabsorption. All together, these early and late aldosterone effects stimulate a net transepithelial Na+ flux. Increasing NCC activity has no direct effect on K+ transport whereas stimulation of ENaC and Na+,K+-ATPase in the CNT/CCD increases the driving force for K+ secretion. Therefore, the main (perhaps the only) mechanism by which aldosterone induces K+ secretion is through the stimulation of the electrogenic Na+ transport system in CNT/CCD principal cells. Indeed, neither BK channel expression nor ROMK expression is directly activated by aldosterone.

The aldosterone paradox Hypovolaemia and hyperkalaemia are the main factors that induce aldosterone secretion by adrenals. Thus, the paradox is that to maintain homeostasis in these two conditions, aldosterone has to be able to either maximally increase Na+ reabsorption while minimally increasing K+ secretion during hypovolaemia, or maximally increase K+ secretion while minimally increasing Na+ reabsorption in response to hyperkalaemia. Conceptually, this can be achieved by two means: (1) by modulating the relative effects of aldosterone on electrogenic Na+ reabsorption (which is coupled to K+ secretion) on the one hand and on electroneutral NaCl reabsorption (which is uncoupled to K+ secretion) on the other hand, and/or (2)  by dissociating K+ secretion from electrogenic Na+ reabsorption in CNT/CCD principal cells. Whatever the mechanism, the solution to the paradox requires that some factor be differentially expressed during hypovolaemia and hyperkalaemia and modulate aldosterone effects. Angiotensin II is likely to play such a role because (a) its plasma level increases during hypovolemia but not hyperkalaemia, (b) its receptor is expressed along the whole ASDN, and (c)  it increases Na+ reabsorption in the DCT1 and inhibits ROMK throughout the ASDN (Kahle et al., 2008). A key molecule in the modulation of aldosterone action by angiotensin II was identified in patients with Gordon syndrome (or pseudohypoaldosteronism type II, PHAII), a disease characterized by hypertension and hyperkalaemia, owing to constitutively active Na+ reabsorption and inhibition of K+ secretion. The disease is caused by mutations in genes encoding for ‘with-no-lysine’ (WNK) kinases (Wilson et al., 2001). One member of this family, WNK4, is mainly expressed in the distal nephron (DCT to OMCD) and is of particular interest; further investigations have demonstrated its role as a ‘molecular switch’ enabling the kidney either to reabsorb Na+ or to secrete K+. This kinase displays different conformations in response to aldosterone and/or angiotensin II. Under basal conditions, WNK4 is not phosphorylated and inhibits NCC, ROMK and ENaC (Kahle et al., 2003), impeding both Na+ reabsorption and

K+ secretion. During hypovolaemia, angiotensin II promotes the WNK4-PHAII conformation also found in PHAII, which removes NCC and ENaC inhibition and promotes their activation by aldosterone, but which amplifies ROMK inhibition (Kahle et al., 2003). During hyperkalaemia, SGK1 phosphorylates WNK4 which, under this new conformation, maintains its inhibitory action on NCC whereas it activates ENaC and ROMK.

Potassium load Because the amount of K+ ingested daily is of the same order of magnitude as the total amount of K+ in extracellular fluids, the first challenge the organism has to face is to maintain its plasma K+ concentration constant after each meal. The kidneys are not only able to excrete the daily K+ intake under normal feeding conditions, but they can also efficiently excrete a large overload of K+ while maintaining the plasma K+ concentration within tight limits. In addition, their ability to excrete an acute K+ load is increased following chronic K+ loading, a feature known as K+ adaptation. Classically, the regulation of renal K+ excretion is considered to proceed via a feedback mechanism in which dietary K+ ingestion increases the plasma K+ level, which in turn stimulates pathways, in particular the secretion of aldosterone, that clear plasma K+. Given the small magnitude of changes in plasma K+ levels observed after a meal, an alternate, feedforward regulation mechanism has been proposed (Rabinowitz, 1996). According to this mechanism, the digestive tract senses the amount of K+ ingested during a meal and sends messages towards target organs to stimulate their capacity to clear plasma K+ and to thus anticipate any change in the plasma K+ concentration. Feedback and feedforward mechanisms of K+ balance regulation are not mutually exclusive.

Postprandial regulation of K+ excretion Following a ‘normal’ meal, the ingested load of K+ is rapidly stored in muscles and liver so that its extracellular concentration increases only slightly (< 0.5 mmol/L). Insulin, the secretion of which increases rapidly during a meal, is the main effector of increased uptake of K+ by liver and muscles, owing to its stimulatory action on Na+,K+-ATPase Thereafter, K+ is slowly released from these storage organs into the extracellular compartment from where it is subsequently excreted, via the kidneys. Because changes in the plasma level of K+ during this phase are too small to induce secretion of aldosterone and its kaliuretic effects, other regulation mechanisms were searched for. Pieces of evidence supporting feedforward regulation of K+ (Fig. 23.4) excretion were obtained in contexts resembling postprandial conditions. Initially, it was shown that a meal inducing a 0.5 mmol/L rise in plasma K+ concentration increased the excretion of K+, whereas intravenous administration of K+ promoting the same increase in plasma K+ level had no effect on K+ excretion (Rabinowitz et  al., 1985). This finding rules out the role of hyperkalaemia and aldosterone in the postprandial increase in K+ excretion and points to the involvement of a factor of gastrointestinal origin. In a latter study (Lee et al., 2007), it was found that an intragastric infusion of K+ did not alter the plasma level of K+ but increased its excretion whereas intravenous injection of the same amount of K+ raised both the plasma level of K+ and its excretion, confirming that the postprandial excretion of K+ is independent of hyperkalaemia but dependent instead on a factor present in the gut.

chapter 23 

potassium homeostasis

PC –

ENaC

Na+

ERK p38 c-src

–

ROMK

K+

Aldosterone 3NA+ NKA

K+

2K+ HKA2

H+

Low–K+ diet

ROS

+ PR +

K+ HKA2

Progesterone

H+ IC

Fig. 23.4  Regulation of K+ transport in the collecting duct during dietary K+ depletion. Potassium depletion decreases ROMK activity in principal cells (PC) (a) through a decrease in aldosterone that reduces the electrogenic driving force for K+ secretion and (b) by an increased production of reactive oxygen species (ROS) leading to ROMK endocytosis. In parallel, K+ reabsorption is activated in principal and intercalated cells (PC and IC) through progesterone-induced expression of H+,K+-ATPase type 2 (HKA2).

Besides this gut factor, the nature of which remains unknown, other factors may participate in the feedforward regulation of postprandial renal K+ secretion. Insulin increases urinary K+ secretion (Hoekstra et al., 2012), but its mechanism of action is uncertain: insulin rapidly stimulates Na+,K+-ATPase in the CCD (Feraille et  al., 1992), but a recent study in collecting duct cells shows that it also stimulates ROMK endocytosis (Cheng and Huang, 2011). Recently, it was shown that after a meal, the renal excretion of tissue kallikrein increased along with K+ excretion, whereas the plasma levels of K+ and aldosterone remained unchanged. Conversely, the postprandial plasma K+ level increased in kallikrein-deficient mice. Furthermore, kallikrein was shown to inhibit H+,K+-ATPase activity in the collecting duct (El Moghrabi et al., 2010). Altogether these data suggest that kallikrein may contribute to the feedforward regulation of postprandial K+ excretion by inhibiting K+ reabsorption in the distal nephron. However, the mechanism responsible for the postprandial simulation of kallikrein release by distal tubule remains unknown. Other factors may be involved in response to specific meals. In case of a protein-rich (and therefore K+-rich) meal, glucagon is released by the pancreas and acts on the liver where it induces the massive production of its second messenger cAMP, which is then released in the blood. A glucagon-mediated increase in blood cAMP concentration is thought to trigger K+ secretion along the distal nephron by promoting the renal hyperaemia and the increased GFR that are observed after a protein-rich meal, which per se increases K+ secretion, but also through direct tubular effects on K+ secretion (Ahloulay et al., 1996). In case of a salt-rich meal, the small intestine releases the natriuretic peptide uroguanylin. Uroguanylin has been reported to induce either kaliuretic (Fonteles et al., 1998) or antikaliuretic effects (Moss et al., 2010); the response seems to depend on two signalling mechanisms within the collecting duct that lead to either inhibition of ROMK or activation of BK.

Altogether, these factors contribute to the excretion of the dietary K+ load on a day-to-day basis while maintaining the extracellular fluid concentration of K+ almost constant. The amazing and unexplained feature of these regulatory mechanisms is that kidneys excrete the exact amount of K+ ingested, and no more, even though muscles continually release K+ in the extracellular fluid at each contraction.

Acute K+ loading In response to an acute K+ load (e.g. a single K+-enriched meal), the above regulatory mechanisms of K+ excretion are triggered. However, because a large influx of K+ in the extracellular fluid overcomes the storage capacity of muscles and liver, at least transiently, the extracellular concentration of K+ increases more than after a regular meal. An increase in extracellular K+ concentration stimulates renal K+ excretion both directly and indirectly, in particular by increasing aldosterone secretion (Fig. 23.5). Increasing the extracellular K+ concentration stimulates Na+,K+-ATPase, which is the motor for K+ secretion in CCD/CNT principal cells, by a substrate effect. However, this effect is likely limited because normal extracellular K+ concentration stimulates Na+,K+-ATPase almost maximally. A high plasma K+ concentration also reduces Na+ and water reabsorption along the proximal tubule and loop of Henle and subsequently increases the delivery of Na+ and the flow rate in the distal nephron, both of which stimulate K+ secretion (see ‘Flow rate and Na+ delivery’). Hyperkalaemia is the most potent stimulus of aldosterone secretion by adrenal glands and, accordingly plasma aldosterone rises rapidly in response to an acute K+ load. In the short term, aldosterone increases the abundance of ENaC and Na+,K+-ATPase at the apical and basolateral membrane respectively by recruiting pre-existing units (see ‘Mechanism of action of mineralocorticoids’).

213

214

Section 2  

fluid, electrolyte, and renal tubular disorders +

Urine flow

High–K+ diet

+ 3Na+ K+

BK

NKA FXYD4

+ Na+

2K+

+

ENaC

Aldosterone

+

+

MR K+

ROMK

PC

Fig. 23.5  Regulation of K+ transport in the collecting duct during K+ loading. Potassium loading induces increase of urinary K+ secretion through two parallel mechanisms. Adrenal production of aldosterone is highly sensitive to increased plasma K+ level. This hormone acts on principal cells (PC) by activating K+ entry at the basolateral cells through activation of Na+,K+-ATPase and over-expression of its regulator FXYD4. Aldosterone also favours K+ exit at the apical side by increasing the electrogenic driving force through activation of ENaC. In parallel of aldosterone action, the increase in urinary flow activates big K+-channels (BK) through a Ca2+-dependent pathway.

Chronic K+ loading and K+ adaptation Chronic feeding a K+-rich diet induces renal and extrarenal adaptations that make it possible to survive what would be a lethal K+ overload if it were acute (Thatcher and Radike, 1947). The extrarenal adaptation mostly stems from an increased ability of muscles to accumulate extracellular K+, owing to over-expression of Na+,K+-ATPase. The development of renal adaptations is mainly dependent on hyperaldosteronaemia but, when adapted, kidneys display an increased ability to secrete K+ at any plasma levels of aldosterone and K+. Renal adaptations lead to the increased ability of the CNT and CCD to secrete K+ and to the recruitment of OMCDo to secrete rather than reabsorb K+, which allows the kidney to excrete large acute overdoses of K+. Adaptation of the CNT and CCD results from increased synthesis and membrane expression of Na+,K+-ATPase, FXYD4 (which decreases the Na+,K+-ATPase affinity for K+) and ENaC, all of which increase the driving force for K+ secretion as well as the redistribution of ROMK and BK channels to the apical membrane (with no change in the total amount of proteins), which increases the K+ conductance. In addition, part of chronic K+ loading-induced K+ secretion in the CCD is independent of ENaC (Frindt and Palmer, 2009). These adaptations are associated with a marked enlargement of the surface area of the basolateral membrane of principal cells and enhanced density of mitochondria. This hypertrophic effect is in part due to hyperaldosteronaemia.

K+ deprivation Besides adjustments in their internal K+ balance, organisms confront dietary K+ restriction by reducing the intestinal and renal losses of K+. The renal adaptation includes two synergistic mechanisms, the inhibition of K+ secretion and the activation of K+ reabsorption in the distal nephron.

Inhibition of K+ secretion K+

Inhibition of secretion in the distal part of the nephron is the result of direct effects on ROMK and BK channels, and a decreased

driving force. K+ depletion decreases the circulating aldosterone level, ENaC activity, and hence the driving force for K+ secretion in the CNT/CCD. Inhibition of ROMK involves a complex signalling cascade leading to endocytosis of ROMK channels. The initial event in this cascade is the production of reactive oxygen species (ROS). On the one hand, ROS activates ERK and p38 MAP kinases, which leads to endocytosis of ROMK. On the other hand, ROS activate c-src tyrosine kinases which, in turn, phosphorylate ROMK channels and facilitate their internalization (Babilonia et al., 2005). Interestingly, it has been shown that the production of ROS and its action on ROMK are not dependent on a hypokalaemic status.

Activation of K+ reabsorption Micropuncture studies have revealed that the distal tubules of rats placed under a low-K+ diet reabsorb K+ (Malnic et al., 1964). Identification of K+ depletion-induced H+,K+-ATPase activity in the distal part of the nephron provided a molecular support to this reabsorption process. Later, this was confirmed by showing that K+ reabsorption was inhibited by the H+,K+-ATPase inhibitor SCH28080. Among the two isoforms of H+,K+-ATPase that are expressed in the distal nephron, only type 2 (corresponding to the so-called colonic or nongastric HKA; HKA2 see ‘H+,K+-ATPases’) is overexpressed under conditions of K+ restriction. The hormonal trigger of HKA2 stimulation was recently identified as the progesterone that is produced by adrenal glands (Elabida et al., 2011). The net production of this steroidogenesis intermediate increases because it is generated at a higher rate and consumed at a lower rate. The presence of the nuclear progesterone receptor (PR) in the distal part of the nephron (CCD, OMCD) and the sensitivity of the progesterone-induced HKA2 stimulation to the PR antagonist RU486 suggest the involvement of this receptor.

Hypertrophy It was reported very early that dietary K+ restriction increases kidney weight mainly as a result of enlargement of the OMCD. It is assumed that increasing the membrane surface area for ion

chapter 23 

exchange increases the capacity for K+ reabsorption. OMCD enlargement results from both hypertrophy (cell growth), which occurs during the first days of K+-depletion, and hyperplasia (cell proliferation), which starts after a week of dietary restriction. The signals that triggers these effects are not fully characterized yet but growth factors such as the insulin-growth factor 1 (IGF-1) and the growth differentiation factor 15 (GDF15) are known to be involved. The renin–angiotensin system is also involved in the process of OMCD hypertrophy, since treatment with enalapril (an angiotensin converting enzyme inhibitor) impedes K+ restriction-induced OMCD enlargement.

Flow rate and Na+ delivery In vivo micropuncture studies have shown that extracellular volume expansion, infusion of Na+ salts with poorly absorbable anions, or administration of proximal tubule, and/or loop or distal tubule diuretics increases K+ secretion along the late distal nephron (Duarte et al., 1971). Free-flow micropuncture confirmed that increasing fluid delivery to the distal tubule increases K+ secretion (Kunau et al., 1974). Conversely, low GFR, Na+ depletion, and extracellular volume contraction decrease K+ secretion (Davidson et al., 1958). The common characteristic of these manoeuvres is that they alter the fluid flow rate in the distal nephron, and/or the Na+ load delivered therein. In these experimental set-ups, it is difficult to determine which parameter among fluid flow rate, Na+ load, and Na+ concentration in the tubule fluid is responsible for the control of K+ secretion. Alternately, the possible effect of these parameters and their mechanism of action can be inferred from our current knowledge of the molecular and cellular mechanisms of K+ secretion along the CNT and CCD. Increasing the concentration of Na + in the luminal fluid increases the driving force for Na+ uptake via ENaC which in turn depolarizes the apical membrane and facilitates K+ secretion. Increased Na+ entry also enhances its intracellular concentration and the activity and abundance of Na+,K+-ATPase, which increases in turn intracellular K+ concentration. This also favours K+ secretion. An increased in tubular fluid flow rate is sensed by the apical cilium and translated into a rise in intracellular free calcium concentration. In turn, calcium activates BK channels. An increase in fluid flow rate also dilutes the secreted K+ and decreases its luminal concentration, which increases the chemical gradient favourable to K + secretion. The Na+ load delivered to the distal nephron is thought to modulate K+ secretion through changes in Na+ concentration and/or flow rate. All these parameters have synergistic effects on K+ secretion because they affect the driving force and the membrane conductance respectively.

Acid–base balance The acid–base status modulates K+ transport beyond the DCT. Changes in acid–base status alter K+ excretion both directly, by altering the activity of K+ transport systems in the CNT/DCT and the driving force for K+ secretion, and indirectly through changes in fluid/Na+ delivery and aldosterone status. Direct and indirect effects alter K+ transport in opposite directions. In the short term, direct effects prevail and metabolic acidosis decreases K+ excretion whereas metabolic alkalosis increases it. Conversely, indirect effects prevail during chronic metabolic acidosis and lead to increased K+ excretion and hypokalaemia.

potassium homeostasis

Acute metabolic acidosis decreases K+ excretion at any level of plasma K+ concentration, and alkalosis does the opposite. A low pH reduces the activity of Na+,K+-ATPase and of ROMK in the CNT/ CCD. It also inhibits the activity of ENaC and thereby hyperpolarizes the apical membrane which reduces the driving force for apical K+ exit. Metabolic acidosis also promotes proton secretion by type A  intercalated cells through the activation of V type H-ATPase. This electrogenic ATPase tends to reduce the lumen-negative transepithelial voltage generated by Na+ reabsorption by principal cells, and thereby reduces the driving force for K+ secretion. Finally, acidosis also increases the activity of H+,K+-ATPase and promotes K+ reabsorption. The increased K+ excretion that is observed in chronic metabolic acidosis is accounted for by increased fluid delivery to the distal nephron and a high plasma aldosterone level. The low plasma bicarbonate concentration prevailing during metabolic alkalosis reduces the filtered load of bicarbonate, which in turn reduces the reabsorption of Na+ and water in the proximal tubule. This proximal tubule defect increases fluid delivery to the distal nephron and induces volume depletion which triggers aldosterone secretion.

Other Vasopressin In vitro, vasopressin increases K+ secretion by CNT/CCD principal cells via activation of its V2 receptors and the production of cAMP (Tomita et al., 1986). This effect is accounted for by increased activity of apical ENaC (which depolarizes the apical membrane) and of Na+,K+-ATPase (which increases intracellular K+ concentration), the conjunction of which increases the electrochemical gradient favourable to K+ secretion. Vasopressin also increases K+ secretion by the distal nephron via activation of its V1 receptors located at the apical membrane (Amorim and Malnic, 2000). These receptors are coupled to Ca2+ which may therefore activate BK channels. In vivo, an increase in plasma vasopressin levels is mainly observed during the transition from diuresis to antidiuresis, a condition that does not alter K+ excretion. As a matter of fact, the stimulatory effect of vasopressin on the electrochemical gradient favourable to K+ secretion is likely blunted by the decreased fluid flow in the CNT/CCD that is associated with vasopressin-induced water reabsorption.

Adrenergic agents Epinephrine decreases K+ excretion by two means: (1) it stimulates K+ uptake by muscle and liver and thereby decreases plasma K+ concentration (Bia and DeFronzo, 1981), and (2) it exerts direct effect on the renal handling of K+ (DeFronzo et  al., 1983). This latter effect takes place beyond the DCT and is mediated mainly through activation of α-adrenergic receptors, which are located in type B intercalated cells. Activation of α-adrenergic receptors and the subsequent production of cAMP in type B intercalated cells stimulate bicarbonate secretion and generate a lumen-positive transepithelial voltage (Siga et al., 1996) that can in turn inhibit K+ secretion by principal cells. It also increases H+,K+-ATPase activity and may thus enhance K+ reabsorption. The antikaliuretic effect of epinephrine in the distal nephron may also be mediated in part through activation of α2-adrenergic receptors which antagonizes the effects of vasopressin and other cAMP-producing hormones in principal cells.

215

216

Section 2  

fluid, electrolyte, and renal tubular disorders

Diuretics (See Chapters 28 and 29.) Diuretics that primarily act on nephron segments localized from the proximal tubule to the distal convoluted tubule increase K+ excretion whereas those acting on the CNT/CCD do not affect K+ excretion. The kaliuretic effect of diuretics mostly results from inhibition of K+ reabsorption at their site of action and/or increased secretion of K+ along the CNT/CCD secondarily to increased delivery of Na+ and water (flow) to these segments. In the long term, an increase in plasma aldosterone levels brought about by volume depletion also triggers K+ secretion in the CNT/CCD. Diuretics acting on the proximal tubule include osmotic agents (e.g. mannitol or urea) that prevent water reabsorption and inhibitors of carbonic anhydrase (e.g. acetazolamide) that inhibit Na+ reabsorption coupled to proton secretion and consequently water reabsorption. Because K+ reabsorption in the proximal tubule is tightly coupled to water reabsorption (see Chapter 21), proximal tubule diuretics reduce K+ reabsorption along the proximal tubule (Velazquez and Giebisch, 1988). However, this effect is mostly compensated by increased K+ reabsorption along the loop of Henle, and the kaliuretic effect of this class of diuretics mainly stems from flow-induced K+ secretion along the CNT/CCD. Through inhibition of NKCC2 in the TAL, loop diuretics (e.g. furosemide or bumetanide) primarily inhibit Na+ reabsorption (natriuretic effect) which secondarily blunts the generation of the corticomedullary osmolarity gradient and thereby reduces water reabsorption along the medullary collecting duct (diuretic effect). Their kaliuretic effect results both from the inhibition of K+ reabsorption along the thick ascending limb of Henle’s loop (inhibition of apical K+ entry and abolition of the transepithelial voltage) and from increased Na+ delivery to the distal tubule (Greger, 1997). Thiazides, the most widely used diuretics in the treatment of hypertension, mainly inhibit NCC in the DCT. This inhibition has no primary effect on K+ transport in the DCT but it increases Na+ delivery to the CNT/CCD and thus K+ secretion therein (Velazquez and Giebisch, 1988). Thiazides also inhibit the electroneutral NaCl reabsorption in CNT/CCD intercalated cells (Leviel et al., 2010), but this has no effect on K+ reabsorption. Potassium-sparing diuretics (Horisberger and Giebisch, 1987) acting on the CNT/CCD include ENaC inhibitors (e.g. triamterene and amiloride) and antagonists of the aldosterone receptor (e.g. spironolactone). Inhibition of apical Na+ entry in principal cells hyperpolarizes the apical membrane which reduces the driving force for K+ secretion. Antagonists of aldosterone receptors antagonize the effects of aldosterone on ENaC and on K+ secretion.

References Ahloulay, M., Dechaux, M., Hassler, C., et al. (1996). Cyclic AMP is a hepatorenal link influencing natriuresis and contributing to glucagon-induced hyperfiltration in rats. J Clin Invest, 98, 2251–8. Aizman, R., Asher, C., Fuzesi, M., et al. (2002). Generation and phenotypic analysis of CHIF knockout mice. Am J Physiol Renal Physiol, 283, F569–77. Amorim, J. B. and Malnic, G. (2000). V(1) receptors in luminal action of vasopressin on distal K+ secretion. Am J Physiol Renal Physiol, 278, F809–16. Babilonia, E., Wei, Y., Sterling, H., et al. (2005). Superoxide anions are involved in mediating the effect of low K intake on c-Src expression and renal K secretion in the cortical collecting duct. J Biol Chem, 280, 10790–6.

Backman, K. A. and Hayslett, J. P. (1983). Role of the medullary collecting duct in potassium conservation. Pflugers Arch, 396, 297–300. Bia, M. J. and Defronzo, R. A. (1981). Extrarenal potassium homeostasis. Am J Physiol, 240, F257–68. Boettger, T., Rust, M. B., Maier, H., et al. (2003). Loss of K-Cl co-transporter KCC3 causes deafness, neurodegeneration and reduced seizure threshold. EMBO J, 22, 5422–34. Boim, M. A., Ho, K., Shuck, M. E., et al. (1995). ROMK inwardly rectifying ATP-sensitive K+ channel. II. Cloning and distribution of alternative forms. Am J Physiol, 268, F1132–40. Chen, S. Y., Bhargava, A., Mastroberardino, L., et al. (1999). Epithelial sodium channel regulated by aldosterone-induced protein sgk. Proc Natl Acad Sci U S A, 96, 2514–9. Cheng, C. J. and Huang, C. L. (2011). Activation of PI3-kinase stimulates endocytosis of ROMK via Akt1/SGK1-dependent phosphorylation of WNK1. J Am Soc Nephrol, 22, 460–71. Davidson, D. G., Levinsky, N. G., and Berliner, R. W. (1958). Maintenance of potassium excretion despite reduction of glomerular filtration during sodium diuresis. J Clin Invest, 37, 548–55. Defronzo, R. A., Stanton, B., Klein-Robbenhaar, G., et al. (1983). Inhibitory effect of epinephrine on renal potassium secretion: a micropuncture study. Am J Physiol, 245, F303–11. Duarte, C. G., Chomety, F., and Giebisch, G. (1971). Effect of amiloride, ouabain, and furosemide on distal tubular function in the rat. Am J Physiol, 221, 632–40. El Moghrabi, S., Houillier, P., Picard, N., et al. (2010). Tissue kallikrein permits early renal adaptation to potassium load. Proc Natl Acad Sci U S A, 107, 13526–31. Elabida, B., Edwards, A., Salhi, A., et al. (2011). Chronic potassium depletion increases adrenal progesterone production that is necessary for efficient renal retention of potassium. Kidney Int, 80, 256–62. Feraille, E., Marsy, S., Cheval, L., et al. (1992). Sites of antinatriuretic action of insulin along rat nephron. Am J Physiol, 263, F175–9. Fonteles, M. C., Greenberg, R. N., Monteiro, H. S., et al. (1998). Natriuretic and kaliuretic activities of guanylin and uroguanylin in the isolated perfused rat kidney. Am J Physiol, 275, F191–7. Frindt, G. and Palmer, L. G. (2009). K+ secretion in the rat kidney: Na+ channel-dependent and -independent mechanisms. Am J Physiol Renal Physiol, 297, F389–96. Fromter, E. (1984). Viewing the kidney through microelectrodes. Am J Physiol, 247, F695–705. Greger, R. (1985). Ion transport mechanisms in thick ascending limb of Henle’s loop of mammalian nephron. Physiol Rev, 65, 760–97. Greger, R. (1997). Why do loop diuretics cause hypokalaemia? Nephrol Dial Transplant, 12, 1799–801. Hebert, S. C. (2003). Bartter syndrome. Curr Opin Nephrol Hypertens, 12, 527–32. Hoekstra, M., Yeh, L., Lansink, A. O., et al. (2012). Determinants of renal potassium excretion in critically ill patients: The role of insulin therapy. Crit Care Med, 40(3), 762–5. Horisberger, J. D. and Giebisch, G. (1987). Potassium-sparing diuretics. Ren Physiol, 10, 198–220. Hunter, M., Lopes, A. G., Boulpaep, E. L., et al. (1984). Single channel recordings of calcium-activated potassium channels in the apical membrane of rabbit cortical collecting tubules. Proc Natl Acad Sci U S A, 81, 4237–9. Imai, M., Taniguchi, J., and Tabei, K. (1987). Function of thin loops of Henle. Kidney Int, 31, 565–79. Jamison, R. L. (1987). Potassium recycling. Kidney Int, 31, 695–703. Ji, W., Foo, J. N., O’Roak, B. J., et al. (2008). Rare independent mutations in renal salt handling genes contribute to blood pressure variation. Nat Genet, 40, 592–9. Kahle, K. T., Ring, A. M., and Lifton, R. P. (2008). Molecular physiology of the WNK kinases. Annu Rev Physiol, 70, 329–55. Kahle, K. T., Wilson, F. H., Leng, Q., et al. (2003). WNK4 regulates the balance between renal NaCl reabsorption and K+ secretion. Nat Genet, 35, 372–6.

chapter 23 

Katz, A. I., Doucet, A., and Morel, F. (1979). Na-K-ATPase activity along the rabbit, rat, and mouse nephron. Am J Physiol, 237, F114–20. Kunau, R. T., Jr., Webb, H. L., and Borman, S. C. (1974). Characteristics of the relationship between the flow rate of tubular fluid and potassium transport in the distal tubule of the rat. J Clin Invest, 54, 1488–95. Lachheb, S., Cluzeaud, F., Bens, M., et al. (2008). Kir4.1/Kir5.1 channel forms the major K+ channel in the basolateral membrane of mouse renal collecting duct principal cells. Am J Physiol Renal Physiol, 294, F1398–407. Lee, F. N., Oh, G., Mcdonough, A. A., et al. (2007). Evidence for gut factor in K+ homeostasis. Am J Physiol Renal Physiol, 293, F541–7. Leviel, F., Hubner, C. A., Houillier, P., et al. (2010). The Na+-dependent chloride-bicarbonate exchanger SLC4A8 mediates an electroneutral Na+ reabsorption process in the renal cortical collecting ducts of mice. J Clin Invest, 120, 1627–35. Lorenz, J. N., Baird, N. R., Judd, L. M., et al. (2002). Impaired renal NaCl absorption in mice lacking the ROMK potassium channel, a model for type II Bartter’s syndrome. J Biol Chem, 277, 37871–80. Malnic, G., Klose, R. M., and Giebisch, G. (1964). Micropuncture study of renal potassium excretion in the rat. Am J Physiol, 206, 674–86. Moss, N. G., Riguera, D. A., Fellner, R. C., et al. (2010). Natriuretic and antikaliuretic effects of uroguanylin and prouroguanylin in the rat. Am J Physiol Renal Physiol, 299, F1433–42. Paulais, M., Bloch-Faure, M., Picard, N., et al. (2011). Renal phenotype in mice lacking the Kir5.1 (Kcnj16) K+ channel subunit contrasts with that observed in SeSAME/EAST syndrome. Proc Natl Acad Sci U S A, 108, 10361–6. Pluznick, J. L., Wei, P., Carmines, P. K., et al. (2003). Renal fluid and electrolyte handling in BKCa-beta1-/- mice. Am J Physiol Renal Physiol, 284, F1274–9. Rabinowitz, L. (1996). Aldosterone and potassium homeostasis. Kidney Int, 49, 1738–42. Rabinowitz, L., Sarason, R. L., and Yamauchi, H. (1985). Effects of KCl infusion on potassium excretion in sheep. Am J Physiol, 249, F263–71. Sachs, G., Shin, J. M., Briving, C., et al. (1995). The pharmacology of the gastric acid pump: the H+,K+ ATPase. Annu Rev Pharmacol Toxicol, 35, 277–305. Scholl, U. I., Choi, M., Liu, T., et al. (2009). Seizures, sensorineural deafness, ataxia, mental retardation, and electrolyte imbalance (SeSAME

potassium homeostasis

syndrome) caused by mutations in KCNJ10. Proc Natl Acad Sci U S A, 106, 5842–7. Shi, H., Levy-Holzman, R., Cluzeaud, F., et al. (2001). Membrane topology and immunolocalization of CHIF in kidney and intestine. Am J Physiol Renal Physiol, 280, F505–12. Siga, E., Houillier, P., Mandon, B., et al. (1996). Calcitonin stimulates H+ secretion in rat kidney intercalated cells. Am J Physiol, 271, F1217–23. Silver, R. B. and Soleimani, M. (1999). H+-K+-ATPases: regulation and role in pathophysiological states. Am J Physiol, 276, F799–811. Simon, D. B., Karet, F. E., Hamdan, J. M., et al. (1996). Bartter’s syndrome, hypokalaemic alkalosis with hypercalciuria, is caused by mutations in the Na-K-2Cl cotransporter NKCC2. Nat Genet, 13, 183–8. Stanton, B. A. (1989). Renal potassium transport: morphological and functional adaptations. Am J Physiol, 257, R989–97. Sweadner, K. J., Arystarkhova, E., Donnet, C., et al. (2003). FXYD proteins as regulators of the Na,K-ATPase in the kidney. Ann N Y Acad Sci, 986, 382–7. Taniguchi, J. and Imai, M. (1998). Flow-dependent activation of maxi K+ channels in apical membrane of rabbit connecting tubule. J Membr Biol, 164, 35–45. Thatcher, J. S. and Radike, A. W. (1947). Tolerance to potassium intoxication in the albino rat. Am J Physiol, 151, 138–46. Tomita, K., Pisano, J. J., Burg, M. B., et al. (1986). Effects of vasopressin and bradykinin on anion transport by the rat cortical collecting duct. Evidence for an electroneutral sodium chloride transport pathway. J Clin Invest, 77, 136–41. Velazquez, H. and Giebisch, G. (1988). Effect of diuretics on specific transport systems: potassium. Semin Nephrol, 8, 295–304. Verrey, F. (1999). Early aldosterone action: toward filling the gap between transcription and transport. Am J Physiol, 277, F319–27. Weinstein, A. M. (1988). Modeling the proximal tubule: complications of the paracellular pathway. Am J Physiol, 254, F297–305. Wilson, F. H., Disse-Nicodeme, S., Choate, K. A., et al. (2001). Human hypertension caused by mutations in WNK kinases. Science, 293, 1107–12. Youn, J. H. and McDonough, A. A. (2009). Recent advances in understanding integrative control of potassium homeostasis. Annu Rev Physiol, 71, 381–401.

217

CHAPTER 24

Renal acid–base homeostasis Carsten A. Wagner and Olivier Devuyst The kidney in systemic acid–base balance The kidney has a central role in the maintenance and restoration of systemic acid–base balance. It does so in concert with other organs such as lung, bone, intestine, liver, or skeletal muscles. Daily metabolism produces approximately 1 mmol protons per kilogram body weight in a healthy adult on a standard diet and average physical activity, which needs to be buffered or eliminated. Elimination and buffering of these protons depend critically on the kidney. As depicted in Fig. 24.1, we absorb approximately 20 mmol acid from our diet and lose additional 10  mmol base equivalents from intestinal secretions which in total need to be replenished. The major acid load comes from daily metabolism liberating 15,000 mmol of volatile acid in the form of carbon dioxide (CO2) that can be eliminated by ventilation, and 40  mmol of non-volatile acids that require buffering (mostly by bicarbonate) or direct renal excretion. The kidney contributes to buffering of acids by reabsorbing virtually all filtered bicarbonate and the de novo synthesis of bicarbonate from ammoniagenesis, leading to the excretion of approximately 40  mmol of ammonium into urine. In addition, the kidneys excrete free protons that have to be buffered by urinary buffers, so-called titratable acids (mostly phosphate and ammonium), thereby buffering and eliminating a total of 70 mmol of acid/day (Curthoys, 2008; Hamm et al., 2008). Thus, the kidney contributes to acid–base balance with three major functions:  (1)  the reabsorption of filtered bicarbonate, (2) replenishing of bicarbonate buffers through ammoniagenesis, and (3) excretion of protons, involving ammonium and titratable acids as urinary buffers to increase the capacity to eliminate sufficient amounts of protons in a relatively small volume of urine. Loss of these functions leads to various forms of metabolic acidosis seen in rare syndromes of inherited forms of renal tubular acidosis, more common forms of renal acidosis due to poisoning, hormone deficiencies, or as unwanted drug side effects, and are very common in patients with chronic kidney disease.

Acid–base handling along the nephron The various nephron segments contribute in different ways to this task of the kidney. Proximal tubule segments are involved in bicarbonate reabsorption, ammoniagenesis, and determination of urinary excretion of titratable acids, whereas the thick ascending limb of the loop of Henle (TALH) reabsorbs mostly bicarbonate, and the collecting ducts excrete protons and ammonium, and together are the main sites of renal acid–base control and adaptation.

Bicarbonate reabsorption Normal plasma bicarbonate concentrations are in the range of approximately 25 mmol/L; assuming a normal glomerular filtration rate of 120 mL/min, approximately 4300–4500 mmoles of bicarbonate are filtered into urine (tubular fluid). The normal urine is practically devoid of bicarbonate, which means that efficient mechanisms responsible for the reabsorption of all filtered bicarbonate operate in the proximal tubule, TALH, and initial portions of the distal convoluted tubule (DCT).

Proximal tubular bicarbonate reabsorption The proximal tubule is the major site of bicarbonate reabsorption and accounts for about 80% of filtered bicarbonate. Both active transcellular and passive paracellular processes contribute to the efficient removal of bicarbonate from the ultrafiltrate. Removal of bicarbonate from urine and the secretion of protons result in a fall in urine pH from 7.4 (initial ultrafiltrate) to about pH 6.8 by the end of the proximal tubule (pars recta). The capacity of the proximal tubule to reabsorb bicarbonate is limited and is saturated if plasma bicarbonate levels exceed 26–28 mmol/L (Pitts et al., 1949). This threshold is explained by the maximal transport rates and abundance of proteins involved in bicarbonate reabsorption. Since bicarbonate reabsorption is achieved by Na+/H+ exchange and is thereby linked to Na+ and volume status, extracellular volume contraction or expansion shifts the threshold for bicarbonate reabsorption by the proximal tubule, causing more or less bicarbonate to be reabsorbed.

Luminal mechanisms of bicarbonate transport The reabsorption of bicarbonate in the proximal tubule is initiated by the secretion of protons into urine which is mediated by sodium/proton (Na+/H+)-exchangers (NHEs) and proton pumps located in the luminal brush border membrane (Fig. 24.2). About 80% of proton secretion is mediated by Na+/H+ exchange, whereas H+ secretion by proton pumps contributes for about 20%. Several NHEs are expressed in the proximal tubule, NHE3 (SLC9A3) is the major isoform in adults (Orlowski and Grinstein, 2004). NHE3 contributes also as a major mechanism to sodium reabsorption in the proximal tubule. Protons for NHEs and proton pumps are provided by the activity of intracellular carbonic anhydrases (mostly carbonic anhydrase II (CAII)). The secreted protons combine in the luminal fluid with filtered bicarbonate to form CO2 and water (H2O). This reaction is catalysed by membrane anchored extracellular carbonic anhydrases (mostly carbonic anhydrase IV (CAIV)). CO2 enters proximal tubule cells by diffusion. Whether passage through aquaporin 1 water channels contributes to facilitate CO2 entry has not been fully clarified.

chapter 24 

renal acid–base homeostasis

Diet 20 mmol H+/day Nonvolatile acid 40 mmol H+/day

Absorbed 20 mmol H+/day

ECF pH 7.4

Secreted 10 mmol OH–/day

Feces 10 mmol OH–/day

Reabsorbed 4300 mmol HCO3–/day + New HCO3– 70 mmol/day

Volatile acid 15,000 mmol CO2/day

Metabolism

Expired air 15,000 mmol CO2/day

Filtered 4300 mmol HCO3–/day

Ammonium 40 mmol H+/day

Titratable acids 30 mmol H+/day

Urine 70 mmol H+/day

Fig. 24.1  Acid–base fluxes in a healthy adult of 70-kg body weight. Adapted from Giebisch and Windhager (2009).

Activity of NHE3 and proton pumps at the luminal membrane is stimulated by acidosis, endothelin, and angiotensin II. Carbonic anhydrase inhibitors such as acetazolamide inhibit proximal tubular bicarbonate and sodium reabsorption by blocking CAII and CAIV activity.

Basolateral exit of bicarbonate Bicarbonate formed by rehydration of CO2 by intracellular carbonic anhydrases leaves proximal tubular cells across the basolateral

membrane into blood via the electrogenic Na+-bicarbonate (HCO3) cotransporter NBCe1 (SLC4A4) (Romero et al., 1999, 2004). The transporter is stimulated by acidosis and angiotensin II (Geibel et al., 1990). Mutations in SLC4A4 underlie a severe form of inherited (autosomal recessive) proximal renal tubular acidosis (RTA, type II) often associated with ocular keratopathy, dental malformations, basal ganglia calcifications, and mental retardation (Igarashi et al., 1999).

Interstitium

Urine

Na+/K+– ATPase K+

Na+ NHE3 NBCe1

Na+ HCO3–

H+ CAII H2O + CO2

V-ATPase

Na+ H+ HCO3–

H+ CAIV CO2 + O2

Fig. 24.2  Transcellular bicarbonate reabsorption in the proximal tubule. Absorption of filtered bicarbonate is initiated by the secretion of protons across the luminal brush border membrane of proximal tubular cells by the Na+/H+ exchanger NHE3 and by V-type H+-ATPases. Luminal carbonic anhydrase IV (CAIV) catalyses the formation of CO2 and water, CO2 diffuses into the cells, where it is rehydrated by intracellular carbonic anhydrase II (CAII). The resulting protons are recycled across the luminal membrane, whereas the bicarbonate ions are transported into interstitium/blood by the basolateral NBCe1 Na+-bicarbonate cotransporter.

219

220

Section 2  

fluid, electrolyte, and renal tubular disorders

Another form of RTA is caused by mutations in the gene encoding CAII. This autosomal recessive form of RTA (type III or mixed RTA) is characterized by loss of bicarbonate from the proximal tubule and a combined urinary acidification defect of the collecting duct (see below). Patients suffer also from osteopetrosis and mental retardation due to cerebral calcifications (Roth et al., 1992).

Paracellular movement of bicarbonate The low rate of chloride reabsorption in the initial parts of the proximal tubule creates a lumen-negative potential that drives reabsorption of anions, including bicarbonate, through the paracellular pathway in the later parts of the proximal tubule. This movement of ions is further enhanced by solvent drag (movement of water and ions through the paracellular space) driven by the osmotic gradient between luminal tubular fluid and the interstitium, created by the active reabsorption of solutes. Solvent drag and lumen-negative potential provide mechanisms for massive bicarbonate reabsorption that do not directly use up energy in the form of ATP.

Bicarbonate reabsorption by the TALH and DCT The filtered bicarbonate not reabsorbed by the proximal tubule is reabsorbed in the TALH and DCT. Thus, tubular fluid entering the collecting duct system contains only minute amounts of bicarbonate under conditions of acid–base balance. The mechanisms accounting for bicarbonate reabsorption in the TALH are similar to the proximal tubule, also involving NHE3 and proton pumps on the luminal membrane. The exit pathways for bicarbonate across the basolateral membrane are not well defined. The electroneutral Na+-HCO3 cotransporter NBCn1 is localized at the basolateral membrane, but is thought to mediate bicarbonate uptake from blood that is required for ammonium absorption by the TALH (Jakobsen et al., 2004). The late section of the DCT is characterized by the presence of the first intercalated cells, which express a luminal proton pump (also called V-type H+-ATPase) and a basolateral chloride-bicarbonate exchanger (AE1).

Ammoniagenesis Ammoniagenesis occurs only in the proximal tubule and serves to eliminate protons, as well as de novo generation of bicarbonate from the metabolism of glutamine. Glutamine is extracted from peritubular capillaries, mediated in part by the glutamine transporter SNAT3 (SLC38A3) expressed in the basolateral membrane of proximal tubule cells (Moret et al., 2007). Filtered glutamine, reabsorbed by luminally localized amino acid transporters, may also contribute to ammoniagenesis, although the supply of glutamine is clearly not sufficient during the renal response to acid loading or acidosis. In proximal tubule cells, glutamine is imported into mitochondria and metabolized by mitochondrial phosphate-dependent glutamine and glutamate dehydrogenase to yield α-ketoglutarate. These processes liberate two NH3 and one HCO3− ion per glutamine. Alpha-ketoglutarate can then be used for gluconeogenesis or further metabolized to yield an additional two HCO3− ions (Curthoys, 2008). The HCO3− synthesized during ammoniagenesis exits the cell via the basolateral NBCe1 bicarbonate transporter, whereas NH3 either diffuses into urine and is trapped as NH4+ after protonation or binds intracellular protons and may be excreted into urine by the NHE3 exchanger instead of a proton. Some NH3/NH4+ may also be transported into venous blood and must be detoxified

by the liver. The buffering of intracellular protons by NH3 contributes to the further elimination of protons and the replenishing of bicarbonate. Thus, complete metabolism of glutamine to glucose can thereby generate three HCO3− ions and two NH3, which eliminates another two protons. Ammoniagenesis and gluconeogenesis are stimulated during acidosis and by various hormones. Low intracellular pH leads to stabilization of mRNAs of various ammoniagenic and gluconeogenic enzymes (Curthoys, 2008; Ibrahim et al., 2008). Insulin and angiotensin II may stimulate some steps in the ammoniagenic pathway or the excretion of NH3/NH4+ into urine (Chobanian and Hammerman, 1987). Prostaglandin F2-alpha (PGEF2α) on the other hand may inhibit ammoniagenesis (Sahai et al., 1995). Most of the ammonium excreted into urine at the level of the proximal tubule is reabsorbed in the TALH by the NKCC2 cotransporter, the target of loop diuretics. NH4+ is then accumulated in the interstitium with a corticopapillary gradient and high concentrations of ammonium in the papilla (Wagner et al., 2011a). The high medullary concentrations of ammonium are required to maintain a gradient for ammonium secretion into urine along the collecting duct system. The ability of the medullary interstitium to accumulate high concentrations of ammonium may depend on the presence of specific negatively charged sulpholipids that bind ammonium with high affinity. Reabsorption of NH4+ by the TALH and accumulation in the medulla is increased during acidosis.

Acid–base excretion by the collecting duct system The collecting duct system comprises the connecting tubule (CNT), the cortical collecting duct (CCD), the outer medullary collecting duct (OMCD), and inner medullary collecting duct (IMCD). Acid–base transport occurs in all segments down to the initial part of the IMCD, whereas the last part of the IMCD does not participate in acid–base handling. The collecting duct system serves for fine-tuning of renal acid–base excretion and is critical for normal acid–base balance, as is evident from genetic disorders affecting transport proteins exclusively expressed in this segment (see below and Chapter 36). At least two types of intercalated cell mediate acid or alkali excretion, respectively. Type A intercalated cells secrete protons and ammonium, whereas type B intercalated cells excrete bicarbonate.

Proton and ammonium excretion Type A intercalated cells generate bicarbonate from the hydration of CO2 and the subsequent excretion of protons by V-type proton pumps (V-ATPases or H+-ATPases) into urine (Fig. 24.3, upper panel) (Wagner et al., 2004). The role of H+/K+-ATPases in urinary acidification, beyond their function in potassium conservation, is unclear. The process of bicarbonate generation is catalysed by the intracellular carbonic anhydrase II (CAII). The newly formed bicarbonate is secreted into the interstitium/blood across the basolateral membrane by the anion exchanger 1 (SLC4A1, AE1) (Wagner et al., 2009). However, there may be additional anion exchangers involved, such as SLC26A7. Type I or distal RTA (dRTA I) is caused by mutations in the AE1 (SLC4A4) gene (mostly autosomal dominant) or by mutations in the ATP6V1B1 (B1), or ATP6V0A4 (a4) V-ATPase subunits. Mutations in V-ATPase subunits are inherited in an autosomal

chapter 24 

Interstitium CO2 CI

–

K+ (NH4+) H+

CAII HCO3–

H+

+

Na NH4+ NKCC1

NH4+

RhCG

NH3

RhBG

NH3

Interstitium CO2

Urine

H+/K+– ATPase

CO2 + H2O AE1

Na+/K+– ATPase K+(NH4+) Na+ K+(NH4+) 2CI–

renal acid–base homeostasis

V-ATPase H+ NH3

RhCG

H+

NH3

H+ + TA

NH4+

Urine CO2 + H2O CAII

V-ATPase

CO2

+

H

HCO3–

Pds

CI–

HCO3–

Pds

CI–

CO2 CAII + H2O

H+

V-ATPase

H+ + HCO3–

Fig. 24.3  Acid–base transport by intercalated cells in the collecting system. (Upper panel) Type A intercalated cells generate bicarbonate catalysed by carbonic anhydrase II (CAII). The bicarbonate is exchanged for extracellular chloride by the basolateral anion exchanger AE1, whereas the proton is secreted into urine, mostly by V-type H+-ATPases (V-ATPases). Intercalated cells also secrete NH3 that is taken up from the interstitium by different transport routes, including Na+/K+-ATPases and NKCC1 where NH4+ substitutes for K+, as well as by the RhBG and RhCG NH3 transporters. At the luminal side, NH3 is excreted into urine and is trapped after protonation to NH4+. (Lower panel) Type B and non-A/non-B intercalated cells are characterized by the expression and function of the luminal Pendrin (Pds) chloride/bicarbonate exchanger mediating bicarbonate secretion into urine. Bicarbonate secretion is driven by bicarbonate synthesis facilitated by carbonic anhydrase II (CAII) and proton secretion by V-ATPases. V-ATPases are localized on the basolateral side (type B intercalated cells) and/or luminal side (non-A/non-B intercalated cells), thereby mediating either net bicarbonate secretion or net chloride reabsorption, respectively.

recessive manner and are often associated with progressive bilateral sensorineural deafness due to expression of these V-ATPase subunits in the inner ear (Fry and Karet, 2007). Proton secretion drives ammonium excretion along the collecting duct. Previously thought to be a passive process, ammonium excretion is now recognized to be mediated by specific transport proteins of the Rhesus (Rh) protein family (Weiner and Hamm, 2007; Wagner et al., 2009, 2011a). Two members are expressed in kidney:  RhBG and RhCG, with RhBG found only in basolateral membranes, whereas RhCG is predominantly expressed in the luminal membrane, but also functional in the basolateral membrane. Both proteins are found not only in type A  intercalated cells, but they have also been detected in type B intercalated cells and in principal cells. If RhCG is critical when maximal ammonium excretion is required, the role of RhBG is less clear, being potentially involved in the uptake of ammonium from the medullary interstitium. RhCG is remarkable, because it may function

as a gas channel permeable only to NH3, requiring intracellular de-protonation of NH4+ and then binding of protons in the urine by NH3. The activity and expression of RhCG is upregulated during acidosis in parallel with enhanced urinary ammonium excretion (Weiner and Hamm, 2007).

Titratable acids Titratable acids are alkali buffers binding and neutralizing protons in urine. The name refers to the process of analytical determination of titratable acids by titration of an acidified urine sample with alkali buffers. The major urinary titratable acids are inorganic phosphate (HPO32− and H2PO3−) and citrate (Hamm et  al., 1987). Other substances such as creatinine and uric acid can contribute to the buffering capacity of urine. The amount of available titratable acids depends mostly on their reabsorption in the proximal tubule, since both phosphate and citrate are freely filtered and actively reabsorbed to some extent by the proximal

221

222

Section 2  

fluid, electrolyte, and renal tubular disorders

tubule. The rate of reabsorption of phosphate by its major transporter NaPi-IIa, and citrate by NaDC1, is highly pH-sensitive (Aruga et al., 2000; Nowik et al., 2008). In the case of NaPi-IIa, protons directly block the transporter, increasing phosphate excretion during acidosis. NaDC1 expression is stimulated during acidosis, thereby reducing urinary citrate availability (with consequences for titratable acids and less ability to complex urinary calcium during acidosis). Increasing amounts of titratable acids, mostly phosphate, are mobilized by stimulated intestinal phosphate absorption and increased dissolution of bone matrix, releasing calcium, phosphate, and bicarbonate during prolonged periods of acidosis. Titratable acids are required to buffer protons along the collecting duct. Proton secretion by proton pumps acidifies urine, but this process is limited by the inability to pump protons against a proton gradient larger than 3.5 pH units (intracellular pH ~ 7.2 versus maximal urinary pH 4.5). Approximately 40 mmol of protons need to be excreted, whereas 1 L of unbuffered urine at pH 4.5 contains only 30 micromoles; without buffering, several hundred litres of urine would be required to excrete a sufficient amount of protons.

Bicarbonate excretion HCO3− excretion is mediated by type B and non-A/non-B intercalated cells that are present in the late DCT, CNT, and CCD. These cells are characterized by the luminal expression of the chloride-bicarbonate exchanger pendrin (SLC26A4) secreting bicarbonate into urine in exchange for urinary chloride (Fig. 24.3, lower panel) (Royaux et  al., 2001). Thus, pendrin may serve a second role in chloride absorption and blood pressure regulation (thereby providing a molecular link between plasma chloride and bicarbonate concentrations). Pendrin is critical for bicarbonate excretion, at least in mice, as indicated by in vivo balance and in vitro microperfusion studies with pendrin KO mice (Royaux et  al., 2001). However, in humans with Pendred syndrome (due to mutations in pendrin/SLC26A4), no obvious renal phenotype has been reported to date. The activity of pendrin depends on the formation of bicarbonate by intracellular carbonic anhydrase II and even more on the activity of basolateral H+-ATPases providing the driving force for bicarbonate secretion (Pech et al., 2006). In addition, pendrin activity may be stimulated by adrenergic agonists and aldosterone (Verlander et al., 2003; Azroyan et al., 2012; Pelzl et al., 2012; Mohebbi et al., 2013). Expression and luminal abundance are increased by alkali loading, metabolic alkalosis, aldosterone, and chloride depletion, whereas acidosis or hyperchloraemia may downregulate pendrin expression (Wagner et al., 2011b).

Regulation of renal acid–base handling The kidneys adapt to changes in systemic acid–base status (acidaemia and alkalaemia), metabolic rates of acid or alkali production, dietary intake, or physical activity by excreting more acid equivalents and increasing ammoniagenesis or stimulating bicarbonate secretion, respectively. The exact mechanisms by which the kidney (and extrarenal organs) senses the changes in acid–base status are not clear (Brown and Wagner, 2012). However, several hormones regulate, at least in part, the adaptive process, which also involves pronounced morphological changes.

Hormonal adaption and regulation Angiotensin II and aldosterone Circulating levels of angiotensin II and aldosterone increase during metabolic acidosis (Schambelan et al., 1987; Gyorke et al., 1991). Blockade of the angiotensin-converting enzyme (ACE) or angiotensin II type 1 receptors delays the renal adaption to acidosis in healthy humans, as well as in various animal models (Henger et al., 2000). Angiotensin II acts on renal acid excretion by stimulating NHE3-, NBCe1- and V-ATPase-dependent bicarbonate absorption, as well as ammonium excretion in the proximal tubule, and by stimulating V-ATPase-mediated urinary acidification in the collecting duct (Geibel et al., 1990; Wagner et al., 1998; Nagami, 2002; Rothenberger et al., 2007). Angiotensin II may also increase phosphate reabsorption, thereby decreasing its delivery to the collecting duct and availability as a buffer (Riquier-Brison et al., 2010). Aldosterone stimulates urinary acidification and type A  intercalated cell function. It may act directly on V-ATPases and AE1 activity (Stone et  al., 1983; Winter et  al., 2004, 2011). Moreover, aldosterone stimulates sodium absorption by neighbouring principal cells through the amiloride-sensitive epithelial sodium channel (ENaC), creating a more lumen-negative potential that facilitates proton secretion by intercalated cells (Kovacikova et al., 2006). The latter mechanism underlies the urinary acidification test that uses a combination of loop diuretics (e.g. furosemide) and an aldosterone analogue (e.g. fludrocortisone), which increases sodium delivery from the upstream loop of Henle and stimulates sodium absorption by the collecting duct (Walsh et al., 2007). Inhibition of aldosterone’s actions by mineralocorticoid receptor antagonists (e.g. spironolactone or eplerenone) reduces renal acid excretion. Acquired or inherited disorders of aldosterone synthesis or signalling, underlie type IV hyperkalaemic distal RTA (Sebastian et al., 1980; Karet, 2009).

Endothelin The production and release of endothelin is stimulated during acidosis (Wesson et al., 1998). Endothelin enhances renal acid excretion directly by acting on various nephron segments and possibly indirectly by increasing aldosterone release (Wesson and Dolson, 1997; Khanna et  al., 2004, 2005). Endothelin stimulates NHE3 activity in the proximal tubule and urinary acidification in distal nephron segments, which may involve V-ATPases (Eiam-Ong et al., 1992; Walter et al., 1995; Laghmani et al., 2001; Licht et al., 2004). In healthy humans, the function of endothelin in renal acid excretion is not entirely clear, and its effects may depend on salt balance (more active during salt depletion/restriction). Moreover, endothelin may reduce renal ammoniagenesis and acid excretion, at least during chronic acidosis. Since the above data were obtained under blockade of both major endothelin receptors (ET-A/ET-B receptors), which may have antagonistic effects (Pallini et al., 2012), experiments with selective agonists or antagonists are required to clarify the picture.

Other hormones and factors A variety of other factors influence the renal capacity to excrete acid or base equivalents. Among these factors are the acid–base status, electrolyte intake and balance (most notably for chloride and potassium), volume status, and additional hormonal factors such as insulin, prostaglandins, norepinephrine, glucocorticoids, and many more (Hamm et al., 2008).

chapter 24 

Mirroring the influence of pH on potassium balance, potassium depletion, and hypokalaemia are associated with metabolic alkalosis (Aronson and Giebisch, 2011). In the kidney, potassium depletion stimulates ammoniagenesis, proximal tubular bicarbonate reabsorption, as well as urinary acidification and acid excretion along the collecting duct. These effects involve glucocorticoids to some extent (Sicuro et al., 1998).

Morphological adaption and plasticity Stimulation of renal acid or base excretion is associated with morphological changes that occur mostly in the proximal tubule and in the collecting duct. Chronic acid loading and acidosis induce hypertrophy of the kidney (Bento et al., 2005). The exact mechanisms governing such hypertrophy during acidosis are unknown. Subtle changes occur in the connecting tubule and cortical collecting duct where the proportion of the various subtypes of intercalated cells changes in response to acid or alkali loading. In particular, the relative abundance of type A intercalated cells is increased during acidosis, whereas the density of type B intercalated cells increases during alkalosis (Schwartz et al., 1985; Al-Awqati, 1996). Potential mechanisms for such plasticity include inter-conversion of type A and B intercalated cells (Al-Awqati, 2011), proliferation of the respective differentiated cell type (Duong Van Huyen et al., 2008; Welsh-Bacic et al., 2011), as well recruitment of new cells from a pool of progenitor cells. These processes may involve extracellular matrix proteins such as hensin (DMT1) (Al-Awqati, 2011), secreted factors such as GDF15 (Duong Van Huyen et al., 2008), Notch signalling (Jeong et al., 2009; Quigley et al., 2011), and transcription factors like Foxi1 or CP2L1 (Blomqvist et  al., 2004; Yamaguchi et al., 2006). However, the exact regulation and role of these proteins in ontogenesis of the collecting duct, as well as plasticity, have not been clarified.

References Al-Awqati, Q. (1996). Plasticity in epithelial polarity of renal intercalated cells: targeting of the H+-ATPase and band 3. Am J Physiol, 270, C1571–80. Al-Awqati, Q. (2011). Terminal differentiation in epithelia: the role of integrins in hensin polymerization. Annu Rev Physiol, 73, 401–12. Aronson, P. S. and Giebisch, G. (2011). Effects of pH on potassium: new explanations for old observations. J Am Soc Nephrol, 22, 1981–9. Aruga, S., Wehrli, S., Kaissling, B., et al. (2000). Chronic metabolic acidosis increases NaDC-1 mRNA and protein abundance in rat kidney. Kidney Int, 58, 206–15. Azroyan, A., Morla, L., Crambert, G., et al. (2012). Regulation of pendrin by cAMP: possible involvement in beta-adrenergic-dependent NaCl retention. Am J Physiol Renal Physiol, 302, F1180–7. Bento, L. M., Carvalheira, J. B., Menegon, L. F., et al. (2005). Effects of NH4Cl intake on renal growth in rats: role of MAPK signalling pathway. Nephrol Dial Transplant, 20, 2654–60. Blomqvist, S. R., Vidarsson, H., Fitzgerald, S., et al. (2004). Distal renal tubular acidosis in mice that lack the forkhead transcription factor Foxi1. J Clin Invest, 113, 1560–70. Brown, D. and Wagner, C. A. (2012). Molecular mechanisms of acid-base sensing by the kidney. J Am Soc Nephrol, 23(5), 774–80. Chobanian, M. C. and Hammerman, M. R. (1987). Insulin stimulates ammoniagenesis in canine renal proximal tubular segments. Am J Physiol, 253, F1171–7. Curthoys, N. P. (2008). Renal ammonium ion production and excretion. In R. J. Alpern and S. C. Hebert (eds.) Seldin and Giebisch’s The Kidney: Physiology and Pathophysiology (4th ed.), pp. 1601–20. Burlington, MA: Elsevier.

renal acid–base homeostasis

Duong Van Huyen, J. P., Cheval, L., Bloch-Faure, M., et al. (2008). GDF15 triggers homeostatic proliferation of acid-secreting collecting duct cells. J Am Soc Nephrol, 19, 1965–74. Eiam-Ong, S., Hilden, S. A., King, A. J., et al. (1992). Endothelin-1 stimulates the Na+/H+ and Na+/HCO3- transporters in rabbit renal cortex. Kidney Int, 42, 18–24. Fry, A. C. and Karet, F. E. (2007). Inherited renal acidoses. Physiology (Bethesda), 22, 202–11. Geibel, J., Giebisch, G., and Boron, W. F. (1990). Angiotensin II stimulates both Na+-H+ exchange and Na+/HCO3- cotransport in the rabbit proximal tubule. Proc Natl Acad Sci U S A, 87, 7917–20. Giebisch, G. and Windhager, E. (2009). Transport of acids and bases. In W. F. Boron and E. L. Boulpaep (eds.) Medical Physiology, pp. 851–65. Philadelphia, PA: Saunders Elsevier. Gyorke, Z. S., Sulyok, E., and Guignard, J. P. (1991). Ammonium chloride metabolic acidosis and the activity of renin-angiotensin-aldosterone system in children. Eur J Pediatr, 150, 547–9. Hamm, L. L., Alpern, R. J., and Preisig, P. A. (2008). Cellular mechanisms of renal tubular acidification. In R. J. Alpern and S. C. Hebert (eds.) Seldin and Giebisch’s The Kidney: Physiology and Pathophysiology (4th ed.), pp. 156–213. Philadelphia, PA: Academic Press. Hamm, L. L. and Simon, E. E. (1987). Roles and mechanisms of urinary buffer excretion. Am J Physiol, 253, F595–605. Henger, A., Tutt, P., Riesen, W. F., et al. (2000). Acid-base and endocrine effects of aldosterone and angiotensin II inhibition in metabolic acidosis in human patients. J Lab Clin Med, 136, 379–89. Ibrahim, H., Lee, Y. J., and Curthoys, N. P. (2008). Renal response to metabolic acidosis: role of mRNA stabilization. Kidney Int, 73, 11–18. Igarashi, T., Inatomi, J., Sekine, T., et al. (1999). Mutations in SLC4A4 cause permanent isolated proximal renal tubular acidosis with ocular abnormalities. Nat Genet, 23, 264–6. Jakobsen, J. K., Odgaard, E., Wang, W., et al. (2004). Functional up-regulation of basolateral Na+-dependent HCO3- transporter NBCn1 in medullary thick ascending limb of K+-depleted rats. Pflugers Arch, 448, 571–8. Jeong, H. W., Jeon, U. S., Koo, B. K., et al. (2009). Inactivation of Notch signaling in the renal collecting duct causes nephrogenic diabetes insipidus in mice. J Clin Invest, 119, 3290–300. Karet, F. E. (2009). Mechanisms in hyperkalemic renal tubular acidosis. J Am Soc Nephrol, 20, 251–4. Khanna, A., Simoni, J., Hacker, C., et al. (2004). Increased endothelin activity mediates augmented distal nephron acidification induced by dietary protein. J Am Soc Nephrol, 15, 2266–75. Khanna, A., Simoni, J., and Wesson, D. E. (2005). Endothelin-induced increased aldosterone activity mediates augmented distal nephron acidification as a result of dietary protein. J Am Soc Nephrol, 16(7), 1929–35. Kovacikova, J., Winter, C., Loffing-Cueni, D., et al. (2006). The connecting tubule is the main site of the furosemide-induced urinary acidification by the vacuolar H+-ATPase. Kidney Int, 70, 1706–16. Laghmani, K., Preisig, P. A., Moe, O. W., et al. (2001). Endothelin-1/ endothelin-B receptor-mediated increases in NHE3 activity in chronic metabolic acidosis. J Clin Invest, 107, 1563–9. Licht, C., Laghmani, K., Yanagisawa, M., et al. (2004). An autocrine role for endothelin-1 in the regulation of proximal tubule NHE3. Kidney Int, 65, 1320–6. Mohebbi, N., Perna, A., Van Der Wijst, J., et al. (2013). Regulation of two renal chloride transporters, AE1 and pendrin, by electrolytes and aldosterone. PLoS One, 8, e55286. Moret, C., Dave, M. H., Schulz, N., et al. (2007). Regulation of renal amino acid transporters during metabolic acidosis. Am J Physiol Renal Physiol, 292, F555–66. Nagami, G. T. (2002). Enhanced ammonia secretion by proximal tubules from mice receiving NH(4)Cl: role of angiotensin II. Am J Physiol Renal Physiol, 282, F472–7. Nowik, M., Picard, N., Stange, G., et al. (2008). Renal phosphaturia during metabolic acidosis revisited: molecular mechanisms for decreased renal phosphate reabsorption. Pflugers Arch, 457, 539–49.

223

224

Section 2  

fluid, electrolyte, and renal tubular disorders

Orlowski, J. and Grinstein, S. (2004). Diversity of the mammalian sodium/proton exchanger SLC9 gene family. Pflugers Arch, 447, 549–65. Pallini, A., Hulter, H. N., Muser, J., et al. (2012). Role of endothelin-1 in renal regulation of acid-base equilibrium in acidotic humans. Am J Physiol Renal Physiol, 303, F991–9. Pech, V., Kim, Y. H., Weinstein, A. M., et al. (2006). Angiotensin II increases chloride absorption in the cortical collecting duct in mice through a pendrin-dependent mechanism. Am J Physiol Renal Physiol, 292(3), F914–20. Pelzl, L., Pakladok, T., Pathare, G., et al. (2012). DOCA sensitive pendrin expression in kidney, heart, lung and thyroid tissues. Cell Physiol Biochem, 30, 1491–501. Pitts, R. F., Ayer, J. L., and Schiess, W. A. (1949). The renal regulation of acid-base balance in man; the reabsorption and excretion of bicarbonate. J Clin Invest, 28, 35–44. Quigley, I. K., Stubbs, J. L., and Kintner, C. (2011). Specification of ion transport cells in the Xenopus larval skin. Development, 138, 705–14. Riquier-Brison, A. D., Leong, P. K., Pihakaski-Maunsbach, K., et al. (2010). Angiotensin II stimulates trafficking of NHE3, NaPi2, and associated proteins into the proximal tubule microvilli. Am J Physiol Renal Physiol, 298, F177–86. Romero, M. F. and Boron, W. F. (1999). Electrogenic Na+/HCO3- cotransporters: cloning and physiology. Annu Rev Physiol, 61, 699–723. Romero, M. F., Fulton, C. M., and Boron, W. F. (2004). The SLC4 family of HCO 3—transporters. Pflugers Arch, 447, 495–509. Roth, D. E., Venta, P. J., Tashian, R. E., et al. (1992). Molecular basis of human carbonic anhydrase II deficiency. Proc Natl Acad Sci U S A, 89, 1804–8. Rothenberger, F., Velic, A., Stehberger, P. A., et al. (2007). Angiotensin II stimulates vacuolar H+-ATPase activity in renal acid-secretory intercalated cells from the outer medullary collecting duct. J Am Soc Nephrol, 18, 2085–93. Royaux, I. E., Wall, S. M., Karniski, L. P., et al. (2001). Pendrin, encoded by the Pendred syndrome gene, resides in the apical region of renal intercalated cells and mediates bicarbonate secretion. Proc Natl Acad Sci U S A, 98, 4221–6. Sahai, A., Nissim, I., Sandler, R. S., et al. (1995). Prostaglandin F2 alpha- and 12-O-tetradecanoylphorbol-13-acetate-induced alterations in the pathways of renal ammoniagenesis. J Am Soc Nephrol, 5, 1792–8. Schambelan, M., Sebastian, A., Katuna, B. A., et al. (1987). Adrenocortical hormone secretory response to chronic NH4Cl-induced metabolic acidosis. Am J Physiol, 252, E454–60. Schwartz, G. J., Barasch, J., and Al-Awqati, Q. (1985). Plasticity of functional epithelial polarity. Nature, 318, 368–71. Sebastian, A., Sutton, J. M., Hulter, H. N., et al. (1980). Effect of mineralocorticoid replacement therapy on renal acid-base homeostasis in adrenalectomized patients. Kidney Int, 18, 762–73.

Sicuro, A., Mahlbacher, K., Hulter, H. N., et al. (1998). Effect of growth hormone on renal and systemic acid-base homeostasis in humans. Am J Physiol, 274, F650–7. Stone, D. K., Seldin, D. W., Kokko, J. P., et al. (1983). Mineralocorticoid modulation of rabbit medullary collecting duct acidification. A sodium-independent effect. J Clin Invest, 72, 77–83. Verlander, J. W., Hassell, K. A., Royaux, I. E., et al. (2003). Deoxycorticosterone upregulates PDS (Slc26a4) in mouse kidney: role of pendrin in mineralocorticoid-induced hypertension. Hypertension, 42, 356–62. Wagner, C. A., Devuyst, O., Belge, H., et al. (2011a). The rhesus protein RhCG: a new perspective in ammonium transport and distal urinary acidification. Kidney Int, 79, 154–61. Wagner, C. A., Devuyst, O., Bourgeois, S., et al. (2009). Regulated acid-base transport in the collecting duct. Pflugers Arch, 458, 137–56. Wagner, C. A., Finberg, K. E., Breton, S., et al. (2004). Renal vacuolar H+-ATPase. Physiol Rev, 84, 1263–314. Wagner, C. A., Giebisch, G., Lang, F., et al. (1998). Angiotensin II stimulates vesicular H+-ATPase in rat proximal tubular cells. Proc Natl Acad Sci U S A, 95, 9665–8. Wagner, C. A., Mohebbi, N., Capasso, G., et al. (2011b). The anion exchanger pendrin (SLC26A4) and renal acid-base homeostasis. Cell Physiol Biochem, 28, 497–504. Walsh, S. B., Shirley, D. G., Wrong, O. M., et al. (2007). Urinary acidification assessed by simultaneous furosemide and fludrocortisone treatment: an alternative to ammonium chloride. Kidney Int, 71, 1310–6. Walter, R., Helmle-Kolb, C., Forgo, J., et al. (1995). Stimulation of Na+/H+ exchange activity by endothelin in opossum kidney cells. Pflugers Arch, 430, 137–44. Weiner, I. D. and Hamm, L. L. (2007). Molecular mechanisms of renal ammonia transport. Annu Rev Physiol, 69, 317–40. Welsh-Bacic, D., Nowik, M., Kaissling, B., et al. (2011). Proliferation of acid-secretory cells in the kidney during adaptive remodelling of the collecting duct. PLoS One, 6, e25240. Wesson, D. E. and Dolson, G. M. (1997). Endothelin-1 increases rat distal tubule acidification in vivo. Am J Physiol, 273, F586–94. Wesson, D. E., Simoni, J., and Green, D. F. (1998). Reduced extracellular pH increases endothelin-1 secretion by human renal microvascular endothelial cells. J Clin Invest, 101, 578–83. Winter, C., Kampik, N. B., Vedovelli, L., et al. (2011). Aldosterone stimulates vacuolar H(+)-ATPase activity in renal acid-secretory intercalated cells mainly via a protein kinase C-dependent pathway. Am J Physiol Cell Physiol, 301, C1251–61. Winter, C., Schulz, N., Giebisch, G., et al. (2004). Nongenomic stimulation of vacuolar H+-ATPases in intercalated renal tubule cells by aldosterone. Proc Nat Acad Sci USA, 101, 2636–41. Yamaguchi, Y., Yonemura, S., and Takada, S. (2006). Grainyhead-related transcription factor is required for duct maturation in the salivary gland and the kidney of the mouse. Development, 133, 4737–48.

CHAPTER 25

Phosphate homeostasis Heini Murer, Jürg Biber, and Carsten A. Wagner Renal handling of phosphate Inorganic phosphate ions (H2PO4−/ HPO42−) (abbreviated as Pi) are almost freely filtered in the glomeruli. Thus Pi concentration of fluid in Bowman’s space equals approximately the free phosphate concentration of plasma. From primary urine Pi is reabsorbed along the proximal tubules by a saturable process, thereby maximal rates vary considerably in response to phosphate intake and levels of different phosphaturic and antiphosphaturic factors. For individuals in phosphate balance, the daily urinary excretion of phosphate equals the net amount absorbed from the intestinal tract and usually represents 10–20% of the amount filtered (fractional excretion). In response to extremes of phosphate intake, the kidneys may excrete close to 100% or close to 0% of the filtered load. The transport maximum for phosphate (TmP) therefore is a variable rather than a constant parameter. The preferred description of the overall renal handling of phosphate is by the renal threshold of phosphate (TmP/glomerular filtration rate (GFR)); its normal range lies between 0.77 and 1.4 mmol/L (Walton and Bijvoet, 1975). Above a GFR of 40 mL/min TmP varies proportionally with GFR, so TmP/GFR is constant and is a reliable index of the tubular reabsorptive capacity. With advanced renal insufficiency (GFR < 40 mL/min), TmP is further decreased (e.g. due to secondary hyperparathyroidism) and fractional excretion of phosphate is increased. As the decrease in TmP is less than the decrease in GFR, TmP/GFR will rise and hyperphosphataemia results. Reabsorption of filtered Pi occurs along the entire proximal tubule. Under normal conditions, reabsorption of Pi in proximal tubules shows both axial and internephron heterogeneity:  highest rates are usually observed in S1 segments of juxtamedullary nephrons. Whether distal tubular segments contribute significantly to renal handling of Pi is still controversial and possible molecular mechanisms that eventually may be involved in a distal tubular reabsorption of Pi have not been characterized.

Proximal tubular Pi reabsorption As demonstrated with isolated tubules and isolated brush border membrane vesicles, proximal tubular reabsorption of Pi across the luminal (brush border) membrane is dependent on the presence of sodium ions (Na/Pi cotransport) (Ullrich and Murer, 1982). Mechanistically, Na/Pi cotransport represents a secondary active, transcellular process that is driven by Na+/K+-ATPases localized at the basolateral membrane. The intracellular concentration of Pi (~1 mM) is below the thermodynamic equilibrium when taking into account the inwardly directed Na+ gradient and the transmembrane potential of around −60 mV. Therefore, altered rates of apical

Na+/Pi cotransport are mostly achieved by a change of the number of (functional) transporter units. Exit of Pi occurs through the basolateral membrane by a mechanism that is poorly understood. No evidence for a paracellular pathway for Pi has yet been obtained. Among the known and characterized mammalian Na+/Pi cotransporters, three have been localized at the apical membrane of proximal tubular cells: two members of the SLC34 family, namely SLC34A1 (NaPi-IIa) and SLC34A3 (NaPi-IIc) and one member of the SLC20 family, namely SLC20A2 (PiT-2) (Biber et al., 2008; Picard et al., 2010). Although the presence of PiT-1 (Glvr-1) mRNA has been detected in mouse kidney (Tenenhouse et al., 1998), the localization of the PiT-1 protein remains to be determined. Transport, by both NaPi-IIa and NaPi-IIc, is dependent on the presence of Na+ ions and displays an apparent affinity constant for Pi of typically < 0.1 mM (divalent HPO42− ions are preferentially transported) and an apparent affinity constant for Na+ ions in the range of 40–60 mM. Arsenate is the only other substrate known to be transported by the type II Na+/Pi cotransporters. Transport activity of NaPi-IIa is electrogenic, whereas for NaPi-IIc it is electroneutral. NaPi-IIa translocates one net positive charge per transport cycle and thus transport rates increase with membrane hyperpolarization. On the other hand, electroneutral transport by NaPi-IIc is insensitive to membrane potential. This functional distinction is reflected in their respective Na+:  Pi stoichiometries:  3 Na+:  1 HPO42− for NaPi-IIa and 2 Na+:1 HPO42− for NaPi-IIc. Loading of NaPi-IIa and NaPi-IIc proteins with substrates is proposed to be ordered. By biophysical analysis it was established that two Na+ ions bind sequentially and cooperatively before one phosphate ion is bound. A third Na+ binding precedes reorientation of the fully loaded carrier. For NaPi-IIc, one of the Na+ ions, which confers electrogenicity to NaPi-IIa, can still interact with the protein but is not cotransported. Transport by NaPi-IIa and NaPi-IIc is blocked by the competitive inhibitor phosphonoformic acid or foscarnet (PFA) with a reported inhibition constant of approximately 0.4–0.6 mM. PFA itself is not transported (Virkki et al., 2007). Both N- and C-termini, of type II Na/Pi-cotransporters are cytoplasmically oriented and 12 transmembrane domains (TMDs) are predicted by a current model of the secondary structure. It was established that the functional unit of NaPi-IIa is a monomer. Yet, by a yeast two-hybrid split ubiquitin approach and by freeze-fracture analysis of Xenopus oocyte membranes containing the expressed NaPi-IIa proteins, evidence has been obtained that dimerization of NaPi-IIa proteins may occur (Forster et al., 2002, 2006). PiT-2 (SLC20A2) was originally identified as a retroviral receptor (Ram-1), and later was shown to be a Na+-coupled Pi transporter (Collins and Ghishan, 2004). In contrast to SLC34 proteins, SLC20

226

Section 2  

fluid, electrolyte, and renal tubular disorders

proteins preferentially transport monovalent Pi (H2PO4−) with a 2:1 Na+: Pi stoichiometry, thus transport is electrogenic. The apparent affinities for Pi and Na+ are typically ≤ 100 µM and approximately 50 mM, respectively (Bottger et al., 2006; Ravera et al., 2007; Virkki et al., 2007). To date, no specific inhibitors for SLC20 transporters have been reported; PFA is a poor inhibitor of SLC20 transporters (Villa-Bellosta et al., 2007). Similarly as for SLC34 transporters, 12 TMDs have been predicted, but with extracellular N- and C-terminal tails (Farrell et al., 2009).

How do different Na/Pi-cotransporters contribute to renal Pi reabsorption? Knowledge about the relative roles of the different proximal tubular apical Na/Pi-cotransporters in renal handling of Pi has been obtained from the use of knockout mouse models (Miyamoto et al., 2011) and from analysis of hereditary human diseases with hypophosphataemia (Alizadeh and Reilly, 2010; Amatschek et al., 2010). Collectively, current information indicates that the relative roles of known renal Na/Pi-cotransporters (specifically of NaPi-IIa and NaPi-IIc) differ between man and mice. Whether PiT-2, localized in proximal tubules, significantly contributes to renal Pi reabsorption is currently not known and remains to be investigated. In adult mice, NaPi-IIa appears to be the dominant Na/Pi cotransporter (Beck et al., 1998; Tenenhouse, 2005). The phenotype of NaPi-IIa knockout mice was described as hypophosphataemic, hyperphosphaturic, and Na+/Pi cotransport in isolated renal membrane vesicles was reduced by approximately 70%. Most of the residual Na+/Pi-cotransport was attributed to NaPi-IIc that is upregulated in these mice. In contrast, NaPi-IIc knockout mice do not show any phenotype related to Pi homeostasis. In these mice, renal Pi handling was normal, indicating that NaPi-IIc does not play a significant role in adult mice (Segawa et al., 2009; Miyamoto et al., 2011). In contrast to mice, in humans NaPi-IIc appears to play a more important role. Various missense mutations and large deletions in the SLC34A3 gene have been ascribed as the genetic cause of patients with hereditary hypophosphataemic rickets with hypercalciuria (HHRH) (Berwitz et al., 2006; Lorenz-Depiereux et al., 2006; Tencza et al., 2009; Amatschek et al., 2010; Miyamoto et al., 2011). These findings indicate that, in adult humans, the functional role of NaPi-IIc is more prominent compared to adult mice. In contrast, the role of the NaPi-IIa transporter in humans appears to be less important and remains somewhat controversial. Despite genetic variants that have been found within the SLC34A1 gene, these polymorphisms within the NaPi-IIa gene seem not to be linked to renal phosphate anomalies (Lapointe et al., 2006). Of interest, in two siblings of a family with hypophosphataemic rickets with renal phosphate wasting, a duplication of a short amino acid stretch in NaPi-IIa was reported that was associated with hyperphosphaturia and a general Fanconi syndrome (Magen et al., 2010).

n Na+

H2PO4–/HPO4=

HPO4=

H2PO4– K+ Na+ S1

S2

S3

NaPi-IIa NaPi-IIc PiT-2

Fig. 25.1  Basis scheme of transepithelial transport of phosphate in proximal tubules and distribution of Na/Pi cotransporters along the proximal tubular segment.

Na+/Pi cotransporters are achieved by regulated endocytosis on the one hand and by insertion of de novo synthesized proteins on the other hand. Currently, no direct modifications of known Na+/Pi cotransporters are known that could change their transport characteristics (e.g. altered Km values for Pi ions or Na+ ions). Known factors (hormones and metabolic factors) that regulate renal reabsorption of Pi by altering the amount of Na+/Pi cotransporters are listed in Table 25.1. Of note, the time course of changes in the amount of the different Na/Pi cotransporters (NaPi-IIa, NaPi-IIc, and PiT-2) varies significantly, indicating that different cellular mechanisms are involved or that certain factors act indirectly and are part of a regulatory network.

Parathyroid hormone Besides its actions on calcium metabolism, PTH modulates the amount of apical Na+/Pi cotransporters (Bacic et al., 2006; Lee and Partridge, 2009). PTH receptors are localized both at the luminal

Table 25.1  Hormonal and non-hormonal factors affecting renal excretion of phosphate Excretion increased by

Excretion decreased by

Parathyroid hormone

Growth hormone

Dopamine

Insulin-like growth factor

Phosphatonins (FGF-23;

1,25(OH)2D3

sFRP-4; MEPE)

Phosphate depletion

Glucocorticoids

metabolic alkalosis

Atrial natriuretic peptide

Volume contraction

Regulation of proximal tubular Pi reabsorption

Phosphate loading

Hypocalcemia

Regulation of proximal tubular reabsorption of Pi and consequently of urinary Pi excretion is primarily achieved by an alteration of the number of Na+/Pi cotransporters residing in the apical membrane of proximal tubular cells (Fig. 25.1). Alterations in the number of

Carbonic anhydrase inhibitors

Metabolic acidosis Estrogen Diuretics

chapter 25 

and basolateral membrane of proximal tubular cells. Upon activation, apically localized PTH receptors activate phospholipase C (PLC)/protein kinase C (PKC) pathway, whereas basolaterally localized PTH receptors activate a cyclic adenosine monophosphate (cAMP)/protein kinase A  (PKA) pathway (Forster et  al., 2002, 2006; Gensure et  al., 2005). By the use of different PTH analogues it was suggested that regulation of the NaPi-IIa protein occurs primarily via the cAMP/PKA pathway (Nagai et al., 2011). In rodents, PTH leads to a rapid (within minutes) downregulation of the NaPi-IIa protein, while downregulation of NaPi-IIc protein is slower and is observed only after a few hours. In contrast, the regulation of PiT2 by PTH is less clear (Picard et  al., 2010). A decrease of the amount of NaPi-IIa proteins occurs via enhanced clathrin-mediated endocytosis and internalized NaPi-IIa proteins accumulate transiently in early endosomes in the subapical compartment. Interestingly, internalized NaPi-IIa proteins are not recycled, but are sorted out of early endosomes and routed to the lysosomes for degradation (Bacic et al., 2004, 2006; Forster et al., 2006). A change of the amount of NaPi-IIc protein by PTH occurs by a slower rate and, although associated with clathrin-coated vesicles, NaPi-IIc protein does not accumulate in lysosomes (Segawa et al., 2007; Lanzano et al., 2011). NaPi-IIa interacts with the NHERF1 protein (Na+/H+ exchange regulatory factor 1). This interaction is via the C-terminal amino acid motif TRL of the NaPi-IIa protein with the first (of two) PDZ domains of NHERF1 (Biber et al., 2004). The importance of NHERF1 for apical localization of NaPi-IIa proteins was revealed by the use of NHERF1 knockout mice, which demonstrate phosphaturia and hypophosphataemia due to reduced apical abundance of NaPi-IIa proteins (Weinman et al., 2005). Stimulation of either PKA or PKC signalling pathways by PTH results in a hyperphosphorylation of NHERF1 at residues Ser77 and Thr95 that are located within the first PDZ domain of NHERF1. Thus, it is conceivable that phosphorylation of Ser77 and Thr95 destabilizes the interaction of NaPi-IIa with NHERF1 and therefore allow higher lateral mobility along the microvilli and internalization of NaPi-IIa at intermicrovillar clefts (Weinman et al., 2010); no such mechanism has been demonstrated for regulation of NaPi-IIc by PTH.

Dopamine In kidney, dopamine is synthesized from dihydroxyphenylalanine (DOPA) and acts via para-/autocrine mechanisms (Glahn et  al., 1993; Friedlander, 1998). Activation of apically localized dopamine receptors D1 results, similar to PTH, in an internalization of the NaPi-IIa transporter (Bacic et  al., 2005). Dopamine-induced activation of signalling cascades involving PKA and PKC results in phosphorylation of the first PDZ domain of NHERF1, resulting in a dissociation of the NaPi-IIa/NHERF1 complex (Weinman et al., 2010). A possible role of dopamine in the adaptive response of renal Pi reabsorption (see below) has been suggested recently as the dopamine concentration in urine was increased after feeding mice with a diet of high Pi content (Weinman et al., 2011).

Phosphatonins The term phosphatonin was introduced for factors that induce hypophosphataemia and an increase of renal Pi wasting but do not result in adequate production of 1,25 dihydroxyvitamin vitamin D3, for example, in patients with oncogenic osteomalacia (TIO) or

phosphate homeostasis

with autosomal dominant hypophosphataemic rickets (ADHR). The following phosphatonins have been identified (Berndt and Kumar, 2009).

Fibroblast growth factor 23 A positional cloning approach identified fibroblast growth factor 23 (FGF23) in ADHR patients (White et al., 2000) and later FGF23 was identified as the causal factor in tumour-induced osteomalacia (Shimada et al., 2001; Berndt and Kumar, 2009). Consequently, FGF23 has emerged as a major phosphaturic hormone (Prie et al., 2009; Bergwitz and Jüppner, 2010; Farrow and White, 2010). Bone cells, osteoblasts, and osteocytes seem to be the major source of circulating levels of FGF23 and synthesis and excretion of FGF23 is regulated by the plasma levels of Pi and vitamin D 3 (Cheng and Hulley, 2010). The renal effects of FGF23 are twofold:  on the one hand, FGF23 decreases reabsorption of Pi and on the other hand influences the level of vitamin D3 by inhibition of 1,25α-hydroxylases and by stimulation of 24,25-hydroxylases (Bergwitz and Jüppner, 2010; Long and Kharitonenkov, 2011). FGF23 decreases the abundance of proximal tubular Na/Pi cotransporters NaPi-IIa, NaPi-IIc, and PiT2 as demonstrated with FGF23 null mice, overexpression and injection of FGF23, and by incubation of isolated proximal tubules with FGF23 (Farrow and White, 2010; Gattineni and Baum, 2010; Tomoe et al., 2010). FGF23 regulates the abundance of apical Na/Pi cotransporters by activation of the fibroblast growth factor receptor FGFR1c. Interestingly, reduction of 1,25α-hydroxylase by FGF23 occurs via the FGF receptor FGFR3 and -4 (Gattineni and Baum 2010; Gattineni et al., 2011). Intracellular mechanisms of FGF23 leading to reduction of Na+/Pi cotransporters are poorly understood, but may be via the mitogen-activated protein kinase (MAPK) cascade involving phosphorylation of extracellular signal-regulated kinase ERK1/2 (Farrow and White, 2010). A possible role of prostaglandins in FGF23 regulation of Na/Pi cotransporters has also been discussed (Gattineni and Baum, 2010). Full bioactivity of FGF23 via the FGFR1c receptor requires the presence of the membrane associated isoform of Klotho. Co-activation of the FGFR1 by Klotho appears to be obligatory as Klotho null mice display the same phenotype as FGF23 null mice (Farrow and White, 2010). Of interest, in kidney, the primary site of expression of Klotho is the distal tubule (Kuro-o 2011), but expression of Klotho in proximal tubules has also been demonstrated (Huang and Moe 2011). The extracellular domain of Klotho is found in circulation, as a secreted isoform or as a proteolytic cleavage product. Klotho exhibits glycosidase-like activities and was shown to modify sugar residues, for example, of the calcium channel TRPV5 and thereby influencing the turnover of this channel (Cha et al., 2008). A similar action of circulating Klotho was postulated for the regulation of the NaPi-IIa protein (Huang and Moe, 2011).Thus, Klotho may be a phosphaturic factor in the absence of FGF23.

Secreted frizzled-related protein-4 In patients with tumour associated osteomalacia and renal Pi wasting, Secreted frizzled-related protein-4 (sFRP-4) is highly overexpressed (De Beur et al., 2002; Berndt and Kumar, 2009). Infusion of sFRP-4 into rats induces phosphaturia due to a reduced abundance of the type lla Na+/Pi cotransporter. Similar observations were made with parathyroidectomized rats, indicating that the action of sFRP-4 is independent of PTH. In agreement with sFRP-4

227

228

Section 2  

fluid, electrolyte, and renal tubular disorders

being an antagonist of the WNT pathway, infusion of sFRP-4 into rats induced phosphorylation of β-catenin (Berndt et  al., 2006). However, the role of sFRP-4 in phosphate homeostasis has been questioned based on genetic deletion in mice (Christov et al., 2011).

Matrix extracellular phosphoglycoprotein Bone cells are the major source of matrix extracellular phosphoglycoprotein (MEPE) and expression of MEPE is increased in XLH patients and in Hyp mice, as well as in patients with oncogenic osteomalacia (Berndt and Kumar, 2009; Friedlander, 2010). Infusion of MEPE results in a reduction of the type IIa Na+/Pi cotransporter and a correlation of increased levels of MEPE with phosphaturia has been demonstrated by a micropuncture study on proximal tubules after infusion of MEPE (Shirley et al., 2010). The cellular pathway involved in the regulation of Pi reabsorption by MEPE remains to be elucidated.

Fibroblast growth factor 7 FGF7 is overexpressed in oncogenic osteomalacia, yet its direct effect on renal Na+/Pi cotransporters has not yet been demonstrated in in vivo situations. Nor is it known if FGF7 alters the metabolisms of vitamin D3 (Gattineni and Baum, 2010).

Dietary intake of Pi In addition to the above mentioned factors, the abundance of proximal tubular Na+/Pi cotransporters and consequently phosphate excretion are influenced by dietary intake of Pi (Biber et al., 2008). The effects provoked by ingestion of different amounts of Pi can be subdivided into (sub-)acute (minutes, hours) and chronic (days) effects. Notably, the time courses of the changes in the abundance of the different Na+/Pi cotransporters (NaPi-IIa, NaPi-IIc, and PiT2) differ markedly (Biber et al., 2008). By mechanisms that are not yet understood, intake of a low-Pi or high-Pi diet alters, independently of PTH and other known factors, urinary excretion of Pi within a few hours that is paralleled primarily by an alteration of the amount of the NaPi-IIa cotransporter. A high-Pi diet acutely downregulates NaPi-IIa proteins by a mechanism similar as that described for PTH-mediated downregulation; that is, NaPi-IIa proteins are internalized and are degraded in lysosomes (Biber et al., 2008). The abundance of the NaPi-IIc and the PiT2 cotransporter is also influenced by altered dietary contents of Pi, yet the time courses of changes of NaPi-IIc and PiT2 abundances are slower and require several hours. The signal(s) that trigger(s) the alterations in the amount of NaPi lla transporters by altered intake of phosphate is (are) not known. Two possibilities have been postulated and discussed (Bergwitz and Jüppner, 2011): (1) proximal tubular cells may sense changes of luminal concentration of Pi (by an as yet unknown Pi sensor mechanism) directly and/or, (2) a respective signal is generated in the small intestine. With regard to the latter possibility, it is of interest that direct application of a Pi bolus into the duodenum of rats induced an increase of the fractional excretion of Pi already after 15 minutes that did not correlate with PTH or FGF23 and is not dependent on the innervation of the kidney. Based on this observation, a factor was postulated that eventually is released from the intestinal mucosa due to the altered amount of Pi within the intestinal lumen (Kumar 2009). This factor remains to be determined. In contrast, another study showed that Pi, when directly applied into the gut, alters PTH levels within minutes (Martin et  al., 2005). Taken together, there is evidence that the gut may sense the amount of ingested Pi and may release (a) factor(s) influencing directly or indirectly renal excretion of Pi.

Intake of different Pi diets also alters renal vitamin D3 metabolism, for example, a low-Pi diet stimulates both transcription and activity of the 1,25α-hydroxylase activity (Dusso et  al., 2005). Compared to alterations of the NaPi-IIa cotransporter, this effect is slower and observed after approximately 1 day and is (in contrast to the response on NaPi-IIa protein) dependent on an intact pituitary gland (Tenenhouse et al., 1988).

Hypokalaemia Phosphaturia associated with chronic hypokalaemia is explained by reduced Na+/Pi cotransport activity in isolated brush border membranes. Hypokalaemia results in a decrease of the mRNAs of all Na+/Pi cotransporters, but, paradoxically, only the amounts of NaPi-IIc and PiT-2 proteins were decreased while the abundance of NaPi-IIa proteins is increased (Breusegem et al., 2009). As hypokalaemia provokes metabolic alkalosis, alterations of apical Na+/Pi cotransporters could, mechanistically, be similar to pathways involved in acid/base induced changes of Pi reabsorption.

Acid/base changes Altered renal excretion of Pi under acidotic or alkalotic conditions respectively can be explained either by a change in extracellular Pi concentrations and altered filtered load of Pi or by alterations of the amount of renal Na+/Pi cotransporters. Respiratory alkalosis causes a redistribution of phosphate into cells, resulting in hypophosphataemia, whilst metabolic acidosis increases bone release of Pi. In rats, metabolic acidosis induces acutely (within a few hours) retrieval of NaPi-IIa cotransporter protein and after chronic metabolic acidosis reduces the amount of NaPi-IIa mRNA. In contrast, in another study in mice and rats, although reduced amount of Na/Pi cotransporter mRNAs were observed after metabolic acidosis, the amount of NaPi-IIa and NaPi-IIc proteins was unchanged (Nowik et al., 2008). A direct interaction between protons and the transporter may explain the reduction of phosphate absorption due to reduced activity of normally expressed transporter proteins.

Oestrogen In rats, oestrogen provokes phosphaturia due to a reduction in the amount of NaPi-IIa cotransporters and this effect was not correlated with levels of PTH (Faroqui et  al., 2008; Guttmann-Rubinstein et  al., 2010). As oestrogen treatment may induce production of dopamine and/or upregulation of dopamine receptors, an indirect effect of oestrogen cannot be ruled out.

Glucocorticoids Several reports indicate that glucocorticoids, independent of PTH, regulate renal excretion of phosphate excretion (Levi et al., 1995 and references therein) due to changes of the abundance of the NaPi-IIa protein (Loffing et  al., 1998). This effect appears to be associated with an altered composition of membrane lipids, such as, for example, glucosylceramide (Levi et al., 1995).

Volume expansion Extracellular fluid volume expansion or contraction induces phosphaturia or decreases excretion of Pi. Increased phosphaturia induced by volume expansion may be explained by different mechanisms: (a) volume expansion leads to inhibition of proximal tubular reabsorption of sodium and water, and consequently to a dilution of the luminal Pi concentration; (b) volume expansion decreases the serum concentration of calcium and thus increases

chapter 25 

secretion of PTH; and (c)  the volume expansion-associated increase of phosphaturia is accompanied by a reduced rate of Na+/Pi cotransport in isolated brush border membrane vesicles (Liput et al., 1989).

Genetic alterations leading to altered renal handling of Pi Genetic defects that alter renal Pi handling can be localized either in NaPi cotransporter genes or in genes coding for factors/cofactors that regulate proximal tubular reabsorption of Pi. As previously discussed, HHRH has been associated with mutations in the NaPi-IIc gene (Bergwitz et  al., 2006; Tencza et  al., 2009), whereas implication on the development of hypophosphataemia of mutants found in the NaPi-IIa gene remains controversial (Lapointe et al., 2006). Several mutations in the FGF23 gene itself or in genes coding for proteins involved in the synthesis, processing, and degradation of FGF23 were identified that cause disturbances in renal Pi reabsorption. Mutations leading to elevated levels of FGF23: mutations within the proteolytic site (RXXR) of FGF23 prevent the degradation of FGF23 and thus result in elevated levels of FGF23. Such mutations are causal for autosomal dominant hypophosphataemic rickets. Elevated levels of FGF23 are achieved also by mutations in PHEX, a gene with homology to endopeptidases, as observed in patients with X-linked hypophosphataemia. However, if FGF23 is a substrate for PHEX remains controversial. Furthermore, high levels of FGF23 have been attributed to mutations in the dentin matrix acidic phosphoprotein 1 (DMP1) (Strom and Juppner, 2008). Mutations leading to reduced levels of FGF23: patients with mutations in the GALINT3 gene are hyperphosphataemic due to low levels of circulating FGF23. GALNT3 is glycosyl transferase that glycosylates FGF23 thereby making FGF23 more resistant to proteolytic cleavage (Ichikawa et al., 2009).

References Alizadeh Naderi, A. S. and Reilly, R. F. (2010). Hereditary disorders of renal phosphate wasting. Nat Rev Nephrol, 6, 657–65. Amatschek, S., Haller, M., and Oberbauer, R. (2010). Renal phosphate handling in human-what can we learn from hereditary hypophosphataemias? Eur J Clin Invest, 40, 552–60. Bacic, D., Capuano, P., Baum, M., et al. (2005). Activation of dopamine D1-like receptors induces acute internalization of the renal Na+/ phosphate cotransporter NaPi-IIa in mouse kidney and OK cells. Am J Physiol, 288, F740–7. Bacic, D., Lehir, M., Biber, J., et al. (2006). The renal Na+/phosphate cotransporter NaPi-IIa is internalized via the receptor-mediated endocytic route in response to parathyroid hormone. Kidney Int, 69, 495–503. Bacic, D., Wagner, C. A., Hernando, N., et al. (2004). Novel aspects in regulated expression of the renal type IIa Na/Pi cotransporter. Kidney Int, 66, S5–S12. Beck, L., Karaplis, A. C., Amizuka, N., et al. (1998). Targeted inactivation of Npt2 in mice leads to severe renal phosphate wasting, hypercalciuria, and skeletal abnormalities. Proc Natl Acad Sci U S A, 95, 5372–7. Bergwitz, C. and Jüppner, H. (2010). Regulation of phosphate homeostasis by PTH, vitamin D, and FGF23. Annu Rev Med, 61, 91–104. Bergwitz, C. and Jüppner, H. (2011). Phosphate sensing. Adv Chronic Kidney Dis, 18, 132–44. Bergwitz, C., Roslin, N. M., Tieder, M., et al. (2006). SLC34A3 mutations in patients with hereditary hypophosphatemic rickets with hypercalciuria predict a key role for the sodium-phosphate cotransporter NaPi-IIc in maintaining phosphate homeostasis. Am J Hum Genet, 78, 179–92.

phosphate homeostasis

Berndt, T. J., Bielesz, B., Craig, T. A., et al. (2006). Secreted frizzled-related protein-4 reduces sodium-phosphate co-transporter abundance and activity in proximal tubule cells. Pflugers Arch, 451, 579–87. Berndt, T. and Kumar, R. (2009). Novel mechanisms in the regulation of phosphorous homeostasis. Physiology, 24, 17–25. Biber, J., Gisler, S. M., Hernando, N., et al. (2004). PDZ interactions and proximal tubular phosphate reabsorption. Am J Physiol, 287, F871–F875. Biber, J., Hernando, N., Forster, I., et al. (2008). Regulation of phosphate transport in proximal tubules. Pflugers Arch, 458, 39–52. Bottger, P., Hede, S. E., Grunnet, M., et al. (2006). Characterization of transport mechanisms and determinants critical for Na+-dependent Pi symport of the PiT family paralogs human PiT1 and PiT2. Am J Physiol, 291, C1377–C87. Breusegem, S. Y., Takahashi, H., Giral-Arnal, H., et al. (2009). Differential regulation of the renal sodium-phosphate cotransporters NaPi-IIa, NaPi-IIc, and PiT-2 in dietary potassium deficiency. Am J Physiol, 297, F350–61. Cha, S. K., Ortega, B., Kurosu, H., et al. (2008). Removal of sialic acid involving Klotho causes cell-surface retention of TRPV5 channel via binding to galectin-1. Proc Natl Acad Sci U S A, 105, 9805–10. Christov, M., Koren, S., Yuan, Q., et al. (2011). Genetic ablation of sfrp4 in mice does not affect serum phosphate homeostasis. Endocrinology, 152, 2031–6. Collins, J. F. and Ghishan, F. K. (2004). The SLC20 family of proteins: dual functions as sodium-phosphate cotransporters and viral receptors. Pflügers Arch, 447, 647–52. De Beur, S. M., Finnegan, R. B., Vassiliadis, J., et al. (2002). Tumors associated with oncogenic osteomalacia express genes important in bone and mineral metabolism. J Bone Miner Res, 17, 1102–10. Dusso, A. S., Brown, A. J., and Slatopolsky, E. (2005). Vitamin D. Am J Physiol, 289, F8–28. Faroqui, S., Levi, M., Soleimani, M., et al. (2008). Estrogen downregulates the proximal tubule type IIa sodium phosphate cotransporter causing phosphate wasting and hypophosphatemia. Kidney Int, 73, 1141–50. Farrell, K. B., Tusnady, G. E., Eiden, M. V. (2009). New structural arrangement of the extracellular regions of the phosphate transporter SLC20A1, the receptor for gibbon ape leukemia virus. J Biol Chem, 284, 29979–87. Farrow, E. G. and White, K. E. (2010). Recent advances in renal phosphate handling. Nat Rev Nephrol, 6, 207–17. Forster, I. C., Hernando, N., Biber, J., et al. (2006). Proximal tubular handling of phosphate: A molecular perspective. Kidney Int, 70, 1548–59. Forster, I. C., Köhler, K., Biber, J., et al. (2002). Forging the link between structure and function of electrogenic transporters: the renal type IIa Na/Pi cotransporter as a case study. Biophys Mol Biol, 80, 69–108. Friedlander G. (1998). Autocrine/paracrine control of renal phosphate transport. Kidney Int Suppl, 65, S18–S23. Friedlander, G. (2010). Welcome to MEPE in the renal proximal tubule. Nephrol Dial Transplant, 25, 3135–6. Gattineni, J. and Baum, M. (2010). Regulation of phosphate transport by fibroblast growth factor 23 (FGF23): implications for disorders of phosphate metabolism. Pediatr Nephrol, 25, 591–601. Gattineni, J., Twombley, K., Goetz, R., et al. (2011). Regulation of serum 1,25 (OH)2 vitamin D3 levels by fibroblast growth factor 23 is mediated by FGF receptors 3 and 4. Am J Physiol, 301, F371–7. Gensure, R. C., Gardella, T. J., and Jüppner, H. (2005). Parathyroid hormone and parathyroid hormone-related peptide, and their receptors. Biochem Biophys Res Commun, 328, 666–78. Glahn, R. P., Onsgard, M. J., Tyce, G. M., et al. (1993). Autocrine/paracrine regulation of renal Na+-phosphate cotransport by dopamine. Am J Physiol, 264, F618–F622. Guttmann-Rubinstein, L., Lichtstein, D., Ilani, A., et al. (2010). Evidence of a parathyroid hormone-independent chronic effect of estrogen on renal phosphate handling and sodium-dependent phosphate cotransporter type IIa expression. Horm Metab Res, 42, 230–6. Heng, F. and Hulley, P. (2010). The osteocyte- a novel endocrine regulator of body phosphate homeostasis. Matuitas, 67, 327–38. Huang, C. L. and Moe, O. W. (2011). Klotho: a novel regulator of calcium and phosphorus homeostasis. Pflugers Arch, 462, 185–93.

229

230

Section 2  

fluid, electrolyte, and renal tubular disorders

Kumar, R. (2009). Phosphate sensing. Curr Opin Nephrol Hypertens, 18, 281–4. Kuro-o, M. (2011). Klotho and the aging process. Korean J Intern Med, 26, 113–22. Lapointe, J. -Y., Tessier, J., Paquette, Y., et al. (2006). NPT2a gene variation in calcium nephrolithiasis with renal phosphate leak. Kidney Int, 69, 2261–67. Lanzano, L., Lei, T., Okamura, K., et al. (2011). Differential modulation of the molecular dynamics of the type IIa and IIc sodium phosphate cotransporters by parathyroid hormone. Am J Physiol, 301, C850–61. Lee, M. and Partridge, N. C. (2009). Parathyroid hormone signaling in bone and kidney. Curr Opin Nephrol Hypertens, 18, 298–302. Levi, M., Shayman, J. A., Abe, A., et al. (1995). Dexamethasone modulates rat renal brush border membrane phosphate transporter mRNA and protein abundance and glycosphingolipid composition. J Clin Invest, 96, 207–16. Liput, J., Rose, M., Galya, C., et al. (1989). Inhibition by volume expansion of phosphate uptake by the renal proximal tubule brush border membrane. Biochem Pharmacol, 38, 321–5. Long, Y. C. and Kharitonenkov, A. (2011). Hormone-like fibroblast growth factors and metabolic regulation. Biochim Biophys Acta, 1812, 791–5. Loffing, J., Lötscher, M., Kaissling, B., et al. (1998). Renal Na/H exchanger NHE-3 and Na-PO4 cotransporter NaPi-2 protein expression in glucocorticoid excess and deficient states. J Am Soc Nephrol, 9, 1560–67. Lorenz-Depiereux, B., Benet-Pages, A., Eckstein, G., et al. (2006). Hereditary hypophosphatemic rickets with hypercalciuria is caused by mutations in the sodium-phosphate cotransporter gene SLC34A3. Am J Hum Genet, 78, 193–201. Ichikawa, S., Sorenson, A. H., Austin, A. M., et al. (2009). Ablation of the Galnt3 gene leads to low-circulating intact fibroblast growth factor 23 (Fgf23) concentrations and hyperphosphatemia despite increased Fgf23 expression. Endocrinology, 150, 2543–50. Magen, D., Berger, L., Coady, M. J., et al. (2010). A loss-of-function mutation in NaPi-IIa and renal Fanconi’s syndrome. N Engl J Med, 362, 1102–9. Martin, D. R., Ritter, C. S., Slatopolsky, E., et al. (2005). Acute regulation of parathyroid hormone by dietary phosphate. Am J Physiol Endocrinol Metab, 289, E729–34. Miyamoto, K., Haito-Sugino, S., Kuwahara, S., et al. (2011). Sodium-dependent phosphate cotransporters: lessons from gene knockout and mutation studies. J Pharm Sci, 100, 3719–30 Nagai, S., Okazaki, M., Segawa, H., et al. (2011). Acute down-regulation of sodium-dependent phosphate transporter NPT2a involves predominantly the cAMP/PKA pathway as revealed by signaling-selective parathyroid hormone analogs. J Biol Chem, 286, 1618–26. Nowik, M., Picard, N., Stange, G., et al. (2008). Renal phosphaturia during metabolic acidosis revisited: molecular mechanisms for decreased renal phosphate reabsorption. Pflugers Arch, 457, 539–49. Picard, N., Capuano, P., Stange, G., et al. (2010). Acute parathyroid hormone differentially regulates renal brush border membrane phosphate cotransporters. Pflugers Arch, 460, 677–87. Prie, D., Urena, T.P., and Friedlander, G. (2009). Latest findings in phosphate homeostasis. Kidney Int, 75, 882–9.

Ravera, S., Virkki, L. V., Murer, H., et al. (2007). Deciphering PiT transport kinetics and substrate specificity using electrophysiology and flux measurements. Am J Physiol, 293, C606–20. Segawa, H., Onitsuka, A., Kuwahata, M., et al. (2009). Type IIc sodium-dependent phosphate transporter regulates calcium metabolism. J Am Soc Nephrol, 20, 104–13. Segawa, H., Yamanaka, S., Onitsuka, A., et al. (2007). Parathyroid hormone-dependent endocytosis of renal type IIc Na-Pi cotransporter. Am J Physiol, 292, F395–403. Shimada, T., Mizutani, S., Muto, T., et al. (2001). Cloning and characteerization of FGF23 as a causative factor of tumor-induced osteomalacia. Proc Natl Acad Sci U S A, 98, 6500–05. Shirley, D. G., Faria, N. J., Unwin, R. J., et al. (2010). Direct micropuncture evidence that matrix extracellular phosphoglycoprotein inhibits proximal tubular phosphate reabsorption. Nephrol Dial Transplant, 25, 3191–5. Strom, T. M. and Juppner, H. (2008). PHEX, FGF23, DMP1 and beyond. Curr Opin Nephrol Hypertens, 17, 357–62. Tencza, A. L., Ichikawa, S., Dang, A., et al. (2009).Hypophosphatemic rickets with hypercalciuria due to mutation in SLC34A3/type IIc sodium-phosphate cotransporter: presentation as hypercalciuria and nephrolithiasis. J Clin Endocrinol Metab, 94, 4433–8. Tenenhouse, H. S. (2005). Regulation of phosphorus homeostasis by the type lla Na/phosphate cotransporter. Annu Rev Nutr, 25, 197–214. Tenenhouse, H. S., Klugerman, A. H., Gurd, W., et al. (1988). Pituitary involvement in renal adaptation to phosphate deprivation. Am J Physiol, 255, R373–8. Tenenhouse, H. S., Roy, S., Martel, J., et al. (1998). Differential expression, abundance, and regulation of Na-phosphate cotransporter genes in murine kidney. Am J Physiol, 275, F 527–34. Tomoe, Y., Segawa, H., Shiozawa, K., et al. (2010). Phosphaturic action of fibroblast growth factor 23 in Npt2 null mice. Am J Physiol, 298, F1341–50. Ullrich, K. J. and Murer, H. (1982). Sulphate and phosphate transport in the renal proximal tubule. Philos Trans R Soc Lond B Biol Sci, 299, 549–58. Villa-Bellosta, R., Bogaert, Y. E., Levi, M., et al. (2007). Characterization of phosphate transport in rat vascular smooth muscle cells. Implications for vascular calcification. Arterioscler Thromb Vasc Biol, 27, 1030–36. Virkki, L. V., Biber, J., Murer, H., et al. (2007). Phosphate transporters: a tale of two solute carrier families. Am J Physiol, 293, F643–54. Walton, R. J. and Bijvoet, O. L. M. (1975). Nomogram for renal threshold phosphate concentrations. Lancet, 2, 309–10. Weinman, E. J., Biswas, R., Steplock, D., et al. (2011). Increased renal dopamine and acute renal adaptation to a high-phosphate diet. Am J Physiol, 300, F1123–9. Weinman, E. J., Cunningham, R., Wade, J. B., et al. (2005). The role of NHERF-1 in the regulation of renal proximal tubule sodium-hydrogen exchanger 3 and sodium-dependent phosphate cotransporter 2a. J Physiol, 567, 27–32. Weinman, E. J., Steplock, D., Zhang, Y., et al. (2010). Cooperativity between the phosphorylation of Thr95 and Ser77 of NHERF-1 in the hormonal regulation of renal phosphate transport. J Biol Chem, 285, 25134–8. White, K. E., Evans, W. E., O’Riordan, J. L. H., et al. (2000). Autosomal dominant hypophosphatemic rickets is associated with mutations in FGF23. Nat Genet, 26, 345–48.

CHAPTER 26

Calcium homeostasis Francesco Trepiccione and Giovambattista Capasso Introduction Calcium (Ca2+) is the fifth most copious element in the human body, with approximately 1000 g present in adults. It plays an important role in skeletal mineralization and in a wide variety of biological functions. Dietary Ca2+ intake is essential for the body. Recommended dietary Ca2+ intake is 1000–1500 mg/day, depending on age (McCabe et al., 2004). Ca2+ homeostasis is regulated by three key mechanisms: intestinal and renal reabsorption, and bone turnover. These, in turn, are adjusted by a set of interrelated hormones, including parathyroid hormone (PTH), 1,25-dihydroxyvitamin D3 (1,25(OH)2D3), ionized Ca2+ itself, and their matching receptors in the gut, kidney, and bone.

Calcium distribution Most of the total body Ca2+ (about 99%) is confined to the skeleton as Ca2+-phosphate complexes, primarily as hydroxyapatite, where it guarantees skeletal strength and, at the same time, a constantly exchangeable store for the body (Wang et al., 2006). Ca2+ regulates a range of crucial functions, including extra- and intracellular signalling, muscle contraction, and nerve impulse conduction (Bootman et al., 2001). Total serum Ca2+ ranges from 2.2 to 2.6 mmol/L (8.8–10.4 mg/dL) in healthy subjects. It comprises free ions (51%), protein-bound complexes (40%), and ionic complexes (9%). To prevent Ca2+ toxicity, the concentration of serum ionized Ca2+ is closely maintained within the physiological range of 1.10–1.35  mmol/L (4.4–5.4 mg/dL). Non-ionized Ca2+ is bound to an array of various proteins and anions in both the extra- and intracellular pools. The main Ca2+ binding proteins are albumin and globulin in serum, and calmodulin and other Ca2+-binding proteins in the cell. The major ionic complexes in serum are calcium phosphate, calcium carbonate, and calcium oxalate.

Calcium homeostasis Ca2+ homeostasis is largely regulated through an integrated hormonal system that controls Ca2+ transport in the gut, kidney, and bone (Fig. 26.1). It involves two major Ca2+ -regulating hormones and their receptors: PTH and the PTH receptor protein (PTHrP) (Potts and Gardella, 2007), 1,25(OH)2D3 and the vitamin D receptor (VDR) (Jurutka et al., 2001), as well as serum ionized Ca2+ and the Ca2+ -sensing receptor (CaSR) (Brown, 2007). Serum Ca2+ homeostasis is set to keep extracellular ionized Ca2+ levels in the physiological range. A decrease in serum Ca2+ inactivates the CaSR in the parathyroid glands, causing an increase in PTH secretion.

PTH stimulates renal and bone Ca2+ reabsorption. In addition, PTH stimulates the synthesis of 1,25(OH)2D3 (active vitamin D) in the kidney, which promotes intestinal Ca2+ absorption, and in a feedback loop inhibition of PTH secretion. The decrease in serum Ca2+ inactivates directly the CaSR in the kidney, leading to additional Ca2+ reabsorption and enhances the renal action of PTH. This integrated hormonal response re-establishes serum Ca2+ and shuts off the negative feedback loop; in contrast, an increase in Ca2+ level increases Ca2+ excretion and bone storage (Peacock, 2010).

Renal calcium handling The kidney is the main organ that controls Ca2+ excretion. Every day, roughly 8 g of Ca2+ is filtered at the glomerulus, of which < 2% is excreted into the urine. Ca2+ is reabsorbed throughout the nephron:  the principal sites are the proximal tubule, the thick ascending limb, and the distal tubule (Fig. 26.2).

The proximal tubule Along the proximal tubule Ca2+ transport is, in essence, an iso-osmotic process, energetically passive, proceeding through the paracellular pathway. Nevertheless, renal micropuncture experiments, performed under experimental conditions in which the driving force for passive Ca2+ movement has been eliminated, demonstrate that 10–15% of the reabsorption is active, implicating a cellular pathway for this process (Ullrich et al., 1977; Petrazzuolo et al., 2010). Total proximal tubule reabsorption accounts for about 65% of total Ca2+ filtered at the glomerulus.

The loop of Henle In the thin descending and ascending limbs of Henle, the permeability for Ca2+ is very low, and basically we can infer that significant net Ca2+ transport does not occur in these segments. Because these thin limbs of Henle do not transport Ca2+, the thick ascending limb of the loop of Henle (TALH) is responsible for the Ca2+ reabsorption between the bend of the loop and the start of the distal convoluted tubule (DCT). Approximately 25% of the Ca2+ filtered at the glomerulus is reabsorbed in Henle’s loop (Suki, 1979). Evidence has been provided indicating that Ca2+ transport is driven by the electrochemical gradient due to the recycling of potassium ions through the luminal membrane, compatible with a passive absorptive process (Bourdeau and Burg, 1979). These results have been confirmed by other investigators who provided evidence that PTH stimulates passive Ca2+ transport by increasing the electrical driving force and permeability for the paracellular pathway (Wittner et  al., 1993).

232

Section 2  

fluid, electrolyte, and renal tubular disorders PTH secretion by Chief cells of parathyroid glands

Intestinal Ca2+ absorption

Bone Ca2+ reabsorption Plasma Ca2+ 8.5–10.6 mg/dL 2.2–2.6 mM Inhibition of osteoblasts activity Indirect activation of osteoclasts

Mainly active transport Potential role in paracellular transport Transcellular transport Renal Ca2+ reabsorption Tubular Ca2+ reabsorption Phosphorus excretion Active vit D synthesis

1,25–(OH)2–VitD3

Fig. 26.1  Hormonal regulation pathway of calcium homeostasis. Ca2+ homeostasis is finely tuned by a complex hormonal system including PTH and vitamin D action on target organs, namely the gut, the kidney, and the bone. Parathyroid glands can sense the circulating level of Ca2+ through the Ca2+ sensing receptor and so modulate PTH secretion. Low Ca2+ levels stimulate PTH release. PTH promotes Ca2+ reabsorption in the kidney and in the bone, some evidence indicate it is active also in the intestinal Ca2+ absorption. PTH promotes active vitamin D synthesis in the kidney. Vitamin D increase active and, potentially, passive Ca2+ absorption in the gut. It potentiates also renal and bone reabsorption and modulates in a negative feedback PTH release.

However there are reports indicating an active Ca2+ transport component in cortical TALH segments (Imai, 1978; Friedman, 1988), although immunohistochemical studies on mouse and rat kidney sections have not detected the presence of proteins identified as Ca2+ transporters in the TALH (Loffing et al., 2001). Interestingly, a new protein, named paracellin 1 (PCLN-1), expressed in human TALH tight junctions, probably plays a decisive role in the control of passive Ca2+, and also Mg2+ reabsorption, since mutations of PCLN-1 are present in patients with the hypomagnesaemia with hypercalciuria syndrome (HHS) (Blanchard et al., 2001). Overall, the TALH certainly plays a significant role in the process of Ca2+ reabsorption, mainly due to paracellular Ca2+ transport. However, the contribution of active Ca2+ transport deserves further investigation (Hoenderop et  al., 2002a; Motoyama and Friedman, 2002).

Furosemide-induced calciuria Loop diuretics reduce the lumen-positive transepithelial voltage and consequently diminish paracellular transport of Ca2+ and Mg2+ (Hebert, 1999). Patients with Bartter syndrome, caused by NKCC2 (Na+-K+-2Cl- co-transporter) mutations, manifest similar electrolyte disturbances to patients given loop diuretics such as furosemide (Simon and Lifton, 1966). Both acute and chronic (Rizzo et al., 2004) furosemide administration stimulate hypercalciuria associated with upregulation of the main Ca2+ transport proteins downstream in the DCT. Co-administration of chlorothiazide decreases furosemide-induced hypercalciuria, given acutely or chronically.

Immunofluorescent staining studies reveal increased apical membrane protein abundance of the transient receptor potential vanilloid subtype 5 (TRPV5) channel, and intracellular Ca2+ binding protein calbindin-D28k, along the DCT, even when both diuretics are given. Increased abundance of Ca2+ transport proteins in the DCT is an increased solute load-dependent effect in response to increased Ca2+ delivery, and serves as a compensatory adjustment in downstream nephron segments (Lee et al., 2007).

The distal tubule The DCT and connecting tubule (CNT) account for about 15% of total renal Ca2+ transport. Along these segments Ca2+ reabsorption is inversely related to Na+ reabsorption. In the DCT and CNT, the transepithelial potential difference opposes Ca2+ reabsorption, and the paracellular permeability of Ca2+ ions is very low. Thus, Ca2+ reabsorption is, mainly an active transcellular process regulated by PTH and 1,25(OH)2D3 (Costanzo et al., 2000). Along the DCT, active Ca2+ reabsorption is restricted to the late distal convoluted tubule (DCT-2). This segment shares similarities with the CNT; both segments express the Ca2+-reabsorptive protein machinery, namely the TRPV5 channel, calbindin-D28k, and the basolateral Na+/Ca2+ exchanger (NCX1), and the plasma membrane Ca2+-ATPase type 1b (PMCA1b) (Hoenderop et al., 1999; Loffing et al., 2001). DCT2 expresses the apical sodium channel, ENaC, as CNT, but DCT-2 solely expresses the Na+-Cl- cotransporter, NCC. Transepithelial transport of Ca2+ is a three-step process:  Ca2+ crosses the apical membrane though TRPV5 (the

chapter 26 

calcium homeostasis

Calcium reabsorption machinery along the DCT

8% TRPV5

ATP

Ca2+ Calbindin 28κ 65%

PMCA 1b Ca2+

Ca2+

Na+ NCX1

25%

1%

~1%

Fig. 26.2  Renal calcium handling. The kidney can reabsorb the majority of the filtered Ca2+. About 65% of the filtered amount of Ca2+ takes a paracellular route along the proximal tubule. TAL reabsorption accounts for about 25% of the total filtered amount. This transport is mainly paracellular, but some evidences suggest also a transcellular component. The distal convolute tubule finely regulates Ca2+ by transcellular reabsorption. In the insert on the right it is illustrated the Ca2+ reabsorption machinery. Finally some studies report about 1% of filtered Ca2+ is reabsorbed along the collecting duct.

gatekeeper), is intracellularly buffered, mainly by calbindin-D28k, and then diffuses into the interstitial space via NCX1 and PMCA1b Bindels, 2010) (Fig. 26.2).

Apical entry of calcium through TRPV5 The apical Ca2+ influx channel involved in transcellular Ca2+ reabsorption was identified by functional expression cloning using a cDNA library from rabbit primary CNT and the cortical collecting duct (Hoenderop et al., 2009). Injection of total mRNA from this isolate into Xenopus laevis oocytes induces Ca2+ uptake two to three times above background. Subsequently this was recognized to be due to a new epithelial Ca2+ channel, ECaC1, and later renamed TRPV5, and a member of the TRP channel superfamily (Hoenderop et al., 2009). TRPV5-null mice excrete 10-fold more Ca2+ than their wild-type littermates. Active Ca2+ reabsorption in DCT2 and CNT segments is severely impaired in TRPV5 knockout (KO) mice, in line with its postulated gatekeeper function (Hoenderop et al., 2003).

Intracellular buffering by calbindin-D28k

Calbindin-D28k dynamically controls TRPV5-mediated Ca2+ influx by physical interaction with the channel at the plasma membrane. Calbindin-D28k moves towards the plasma membrane, where it is directly associated with TRPV5 at low intracellular Ca2+ concentrations (Lambers et al., 2006). Here it can buffer Ca2+ that enters the renal epithelial cell, thereby counteracting local accumulation

of cytosolic-free Ca2+ and inactivation of the channel. After Ca2+ binding, calbindin-D28k facilitates the transport of Ca2+ to the basolateral membrane and operates as a dynamic Ca2+ buffer. However, calbindin-D28k KO mice have no abnormalities in renal Ca2+ handling; so it is possible that other Ca2+ binding proteins can replace its buffering role. Calbindin-D9k has been proposed to have this function in a DCT cell line, but calbindin-D9k KO mice also do not exhibit any obvious phenotypic abnormalities (Kutuzova et al., 2006). The role of renal calbindin-D9k in compensating for impaired calbindin-D28k function (and vice versa) needs further investigations (Schlatter, 2006).

The thiazide-induced calcium-sparing effect Thiazide diuretics, coupled with their natriuretic effect, are able to reduce urinary Ca2+ excretion. This feature is also seen in Gitelman syndrome patients. Thiazide-induced Ca2+ reabsorption takes place mainly along the proximal tubule and is driven by the increase in proximal Na+ reabsorption as a result of thiazide-induced contraction of the extracellular circulating volume (Bindels, 2010). An additional mechanism that has been proposed is a direct effect on Ca2+ reabsorption in the DCT-2 (Costanzo and Weiner, 1976). This hypothesis seems unlikely, because of the persistence of this anticalciuric effect of thiazides in TRPV5 null mice, and the parallel time-course of decreased Ca2+ excretion with the increased Na+ reabsorption in the proximal tubule (Nijenhuis et al., 2005).

233

234

Section 2  

fluid, electrolyte, and renal tubular disorders

Intestinal calcium absorption Dietary Ca2+ intake is essential for systemic Ca2+ homeostasis. Ca2+ absorption accounts for about 30% of total Ca2+ intake. Ca2+ absorption occurs in the small intestine by active (low dietary intake) and passive (high dietary intake) transport mechanisms. Active transcellular Ca2+ absorption is located largely in the duodenum and upper jejunum, whereas passive paracellular Ca2+ absorption occurs throughout the entire length of the intestine (Bronner et al., 1986; Christakos et al., 2011). Both transport systems are controlled by the circulating 1,25(OH)2D3 (Fig. 26.3).

Active calcium absorption Transcellular Ca2+ absorption requires Ca2+ flux through the apical Ca2+ channel, transient receptor potential vanilloid type 6 (TRPV6), Ca2+ buffering in the enterocyte by calbindin-D9k and basolateral extrusion through PMCA1b (Wasserman, 1964) (Fig. 26.3). TRPV6 and calbindin-D9k co-localize in the intestine and their expression is induced at weaning, by low dietary Ca2+ intake and by 1,25(OH)2D3 (Song et al., 2003); expression is reduced in vitamin D receptor (VDR) KO mice (Van Cromphaut et  al., 2003). However, recent evidences from calbindin-D9k KO and TRPV6 KO mice have challenged this traditional model (Kutuzova et al., 2006; Benn et al., 2008). Calbindin-D9k KO mice show active intestinal Ca2+ transport similar to WT mice in response to a low Ca2+ diet or 1,25(OH)2D3, suggesting a compensatory or alternative model for intestinal Ca2+ absorption. TRPV6 KO mice also maintain normal serum Ca2+ levels; in response to 1,25(OH)2D3 intestinal Ca2+ transport is similar in WT and TRPV6 KO mice (Kutuzova et al., 2008; Thyagarajan et al., 2009). However, these KO mice fail to maintain Ca2+ homeostasis when challenged with a low-Ca2+ diet and they develop bones abnormalities (Thyagarajan et  al., 2009; Lieben et al., 2010). In contrast, a mouse model of TRPV6 Active & Passive transport along the gut

overexpression develops hypercalcaemia, hypercalciuria, and soft tissue calcification. A vesicular-mediated transcellular route for Ca2+ absorption has been identified in chick enterocytes after stimulation with active vitamin D. Intracellular vesicles containing calbindin-D28k seem to shuttle Ca2+ through the cell. In duodenal chick enterocytes a rapid Ca2+ efflux pathway has been described in response to acute stimulation with vitamin D. This mechanism is called transcaltachia and seems to be mediated by a membrane-associated rapid response steroid binding protein (MARRS) (Fleet and Schoch, 2010).

Passive calcium absorption Paracellular Ca2+ absorption occurs throughout the entire intestine, especially in the distal segments (Fig. 26.3). The effectiveness of this pathway depends on the lumen-to-interstitium electrochemical gradient and the integrity of the intercellular tight junction complexes, and may be regulated by vitamin D (Fujita et al., 2008). In fact, VDR KO mice show low levels of claudin-2 and claudin-12. In addition, 1,25(OH)2D3 has been shown to induce the expression of claudin-2 and claudin-12 in vitro in an intestinal epithelial cell line, resulting in facilitated paracellular Ca2+ conductance (Fujita et al., 2008). The vitamin D-dependent paracellular pathway could have a role in ameliorating the phenotype of the TRPV6 KO and the TRPV6/calbindin-D9k double KO mice (Christakos, 2012).

Calcium and bone metabolism Bone is the major Ca2+ storage of the body. Osteoblasts and osteoclasts connect bone turnover to systemic Ca2+ homeostasis. However, how this mechanism directly contributes to serum Ca2+ homeostasis has not been completely clarified. Bone reabsorption and formation may be the main pathways, but these processes are quite slow for quick responses to changes in serum Ca2+ (Fig. 26.4). Active & Passive transport TRPV6

Endocytosis

Transcaltachia

TRPV6

TRPV6

Active Passive

Ca2+

Ca2+

Ca2+

Ca2+

2+

Ca

Ca2–

Calbindin9k

Claudin 2–12

Ca2+ Ca2+

Ca2+ Ca2+

PMCA1b NCX Na+

ATP

MARRS Ca2+

Ca2+ Ca2+

Ca2+

Ca2+

Fig. 26.3  Calcium handling by the gut. Calcium absorption along the gut requires both an active (transcelluar) and passive (paracellular) transport. During physiological feeding condition, active transport takes place, mainly, in the duodenum and in the first part of the jejunum, while passive transport occurs along the whole intestine (left side of the picture). On the right side, it is represented the molecular mechanism involved in the cellular Ca2+ absorption and release.

Osteon and osteoblast function Mineralized matrix

Mineralizing matrix

Interstitial space

Ca channels NCX 1,3

2+

Na+

2+

Ca

Ca

Ca ATP

2+

PMCA

NHE3 ? H+

Na+

Mineralized matrix

Osteoclast function Demineralizing matrix

Interstitial space

αvβ3

H+ ATPase

+

H

H2CO3

NBC-1

HCO3

–

Na+ H+

HCO3

–

AE2 CIC5

–

CI

–

HCO3

–

CI

TRPV5 Ca

2+

Transcytosis Ca2+ Ca2+ 2+ Ca

αvβ3

Fig. 26.4  Bone calcium handling: osteoblast and osteoclast function. (Upper) The upper part of the figure describes the bone forming unit, the osteon. Osteoblast contributes to both bone matrix deposition and its mineralization. Ca2+ release in the mineralizing area and proton removal are key steps for bone mineralization. NCX is crucial for Ca2+ release in the mineralizing matrix, while some evidences suggest a potential role for NHE in keeping mineral matrix alkalinization. (Lower) The lower part of the figure represents the osteoclast-induced bone matrix demineralization. Mature osteoclasts seal up to the mineral matrix through αvβ3 integrin. Acid secretion in the demineralizing area is mediated by H+ATPase and CLC5 channel. This acid secretion is buffered by basolateral HCO3− extrusion through AE2 or NBC1 proteins. Matrix acidification promotes the dissolvance of Ca2+ containing crystals. Ca2+ and proteins from the matrix are reabsorbed, mainly, by endocytosis.

236

Section 2  

fluid, electrolyte, and renal tubular disorders

Osteoblasts produce both the protein and mineral component of the extracellular matrix of the osteon, the bone-forming unit. Osteoblasts are permeable to circulating Ca 2+ levels through several membrane Ca2+ channels. L-type voltage-sensitive Ca2+ channels (VSCCs) regulate the Ca2+ entry in the osteoblasts in response to multiple hormonal stimuli (Bergh et al., 2006). The two main Ca 2+ extruding mechanisms expressed in the osteoblasts are the plasma membrane Ca 2+-ATPase (PMCA) present on the membrane facing the interstitial fluid, and the Na +/Ca 2+ exchanger (NCX) on the membrane facing bone matrix (Stains et  al., 2002). This polar distribution suggests that PMCA is mainly involved in intracellular Ca2+ regulation. Many intracellular proteins in the osteoblast can buffer the large amount of Ca2+ being transported continuously through the cells. Calbindin-D28k is not crucial for this functions and it may be replaced by other Ca 2+-buffering proteins (Turnbull et al., 2004). Therefore, Ca2+ efflux into the osteoid seems to be mainly mediated by the NCX type 1 and 3 (Stains et al., 2002). Phosphate accumulation in the matrix is another critical step for generating hydroxyapatite crystals. Pyrophosphate is the major source of phosphate. Its synthesis, transport and degradation to phosphate anions is finely regulated (Orimo, 2010). Lastly, removing protons from osteoid is crucial for the mineralization process. To date this acid-removing mechanism has not been clarified, but a potential role for the Na+/H+ exchanger (NHE) has been hypothesized (Redhead, 1988). Osteoclasts provide bone demineralization and Ca2+ release into plasma. The key event for Ca2+ removal from bone is the acidification of the mineral matrix. This is achieved by proton secretion via a membrane H+-ATPase. Electroneutrality is maintained by concomitant secretion of Cl−, mainly through the CLC5 channel. Intracellular H2CO3 is the source for H+ secretion and the bicarbonate produced from its dissociation is extruded on the basolateral side via a chloride-bicarbonate exchanger (AE-2) or the sodium-coupled bicarbonate co-transporter (NBC1). Ca2+ and degraded matrix proteins are translocated across the cell by vacuolar transcytosis (Nesbitt and Horton, 1997; Salo et al., 1997). This mechanism ensures removal of large amounts of Ca2+ from the mineral matrix to the extracellular space.

Direct calcium regulation of bone turnover The link between bone storage and circulating Ca2+ levels is mediated by several Ca2+-regulating hormones (see below). Recent evidence hypothesizes a direct effect of serum Ca2+ in the regulation of osteoblast and osteoclast activity, and so in bone turnover (Blair et al., 2011). Osteoblast function seems, in part, to be regulated by the Ca2+ sensing receptor. Mice with conditional KO of the CaSR in osteoblasts show a severe deficit in mineralization of the skeleton. These mice have impaired post-natal growth and skeletal development. They suffer from rib and long bone fractures, and die within 3 weeks of birth (Chang et al., 2008). These findings suggest a pivotal role for the CaSR in the regulation of osteoblasts and the mineralization process. However, it is not completely clear if osteoclasts do express the CaSR. The finding that a low amount of CaSR is found in isolated osteoclasts might be related to contamination and dilution from other bone cell types during the isolation process; but the low sensitivity of the osteoclast to Ca2+ stimulation (5–20 mmol/L), is far from the typical for the CaSR (c.1.5 mmol/L).

Regulation by Klotho Klotho is a beta-glucoronidase with multiple renal and extrarenal functions such as ageing, oxidative stress, and mineral metabolism. Klotho deficiency is associated with slight hypercalcaemia, bone demineralization, and hypercalciuria (Hu et al., 2010). The original explanation for these findings was a possible inhibitory effect of Klotho on 1,25(OH)2D3 synthesis; the consequent hypervitaminosis D being responsible for the hypercalcaemia and hypercalciuria. However, a weak point of this theory was the reversal of hypercalcaemia and the bone phenotype, but not of the hypercalciuria, with the normalization of serum 1,25(OH)2D3 levels (Imura et al., 2007). These data prompted a study of Klotho regulation of Ca2+ homeostasis as a primary modulator of TRPV5 activity. Indeed, Klotho mediates an increase in cellular surface expression of TRPV5 in DCT and CNT (Chang et  al., 2005). Klotho null mice have a primary renal Ca2+ leak that contributes to a secondary increase in 1,25(OH)2D3 synthesis with its consequences. In addition, Klotho increases the activity of the Na+-K+-ATPase α1-subunit; the increased Na+ gradient created by increased Na+-K+-ATPase activity might drive the transepithelial transport of Ca2+ through the basolateral membrane via the Na+/Ca2+ exchanger (NCX)-1 (Imura et al., 2007). These studies suggest a fundamental role for Klotho in the regulation of Ca2+ balance.

Regulation by the Ca2+-sensing receptor The extracellular CaSR allows all the tissues involved in Ca 2+ homeostasis to monitor the blood Ca2+ level (Brown et al., 1993). When it senses even minor variations in extracellular Ca2+ from its normal level, the CaSR directly or indirectly modulates various homeostatic tissues to normalize extracellular Ca2+. Key tissues expressing the CaSR include the PTH-secreting parathyroid glands, calcitonin (CT)-secreting thyroidal C cells, the intestine, bone, and kidneys. In non-parathyroid tissue, the CaSR determines how much Ca2+ moves into or out of the body through the intestine and kidneys, and how Ca2+ moves between the extracellular fluid and bone. These Ca2+ fluxes are regulated by PTH and CT, as well as by 1,25(OH)2D3 (Chattopadhyay et al., 1996). In the parathyroid glands, the CaSR represents the molecular mechanism by which parathyroid cells detect changes in blood ionized Ca2+ concentration, modulate PTH secretion accordingly, and so maintain serum Ca2+ levels within a narrow physiological range. Interestingly, in the kidney the CaSR regulates renal Ca2+ excretion and influences the transepithelial movement of water and other electrolytes.

CaSR in the proximal tubule The CaSR is expressed in the subapical region of proximal tubular cells (Riccardi et al., 1998) where it is involved in the regulation of PTH-mediated phosphate (Pi) excretion (Ba et al., 2003). Studies carried out in proximal tubule-derived cell lines also suggest that 1α -hydroxylase activity is inhibited in the presence of high Ca2+ (Maiti et al., 2008). Recent investigations performed on a murine model in which the full-length CaSR has been ablated have shown that the CaSR reduces the response to 1,25(OH)2D3 independent of the actions of PTH (Egbuna et al., 2009). Thus, the CaSR exerts tight control on circulating 1,25(OH)2D3 both at the level of its synthesis (in the

chapter 26 

proximal tubule) and in modulating its effects (specifically on Ca2+ reabsorption by the distal tubule). Conversely, 1,25(OH)2D3, PTH and dietary phosphate modulate both the CaSR gene and CaSR protein expression in the proximal tubule, suggesting the existence of a local feedback loop for the regulation of Ca2+ and Pi excretion independent of systemic changes in calciotropic hormones. In addition, luminal Ca2+ concentration acting via the CaSR can also modulate sodium-dependent proton secretion and water reabsorption along the proximal tubule: increasing luminal Ca2+ or using a calcimimetic agent leads to increased water reabsorption (Capasso et al., 2013).

CaSR in the loop of Henle The CaSR is localized at the basolateral side of TALH cells, where it directly modulates both paracellular and transcellular NaCl and divalent cations transport. Basolateral, but not urinary (luminal), increases in serum Ca2+ (or Mg2+) concentrations reduce their own reabsorption (Quamme, 1982). During hypercalcaemia activation of the basolateral CaSR inhibits ROMK channels (Wang et al., 1997). Since apical K+ recycling in the TAL is the rate-limiting step for Na+-K+-2Cl− co-transport and paracellular divalent cations reabsorption (Greger et al., 2001), CaSR activation induces Ca2+ loss. A  novel role in the direct regulation of paracellular Ca2+ reabsorption along the TAL has also been demonstrated (Loupy et al., 2012). Gain-of-function mutations of CaSR cause a Bartter-like phenotype (Vargas-Poussou et al., 2002). CaSR affinity for Ca2+ and Mg2+ in the TAL is also influenced by pH: as pH increases, the larger is its affinity for divalent cations (Quinn et al., 2004). This is yet another level of complexity whereby the CaSR activity is also influenced by extracellular pH.

CaSR in the distal convoluted tubule The CaSR co-localizes with TRPV5 at the apical membrane and in subapical vesicles of DCT and CNT cells (Topala et al., 2009). In cell lines over-expressing TRPV5 and CaSR, the activation of this receptor increases the activity of TRPV5, activating Ca2+ reabsorption. CaSR also controls basolateral expression of NCX and PMCA, thus promoting the Ca2+ efflux pathway (Hoenderop et al., 2004; Topala et al., 2009). The role of the CaSR in calcium reabsorption is evident in Ca2+ overload along the DCT, for example, following administration of a loop diuretic.

CaSR in the collecting duct The level of urine Ca2+ concentration is known to influence acid secretion and water reabsorption. This mechanism protects the kidney from forming stones. CaSR expression in principal and intercalated cells of the collecting duct is involved in this protection. TRPV5 KO mice have hypercalciuria, but they do not form kidney stones (Renkema et al., 2009), probably because they exhibit marked urinary acidification with a low urine pH and increased urine flow rate. Mice lacking both TRPV5 and the B1 subunit of the H+-ATPase develop severe nephrocalcinosis and die in the first 3 months of life, implying that the inability to acidify their urine leads to renal stones formation in the presence of hypercalciuria. Outer medullary collecting ducts dissected from TRPV5 null mice, when exposed to CaSR agonists, up-regulate H+-ATPase and downregulate aquaporin 2 expression (Renkema et al., 2009).

calcium homeostasis

Hypercalciuria-induced polyuria via the CaSR is even more apparent in the inner medullary collecting duct (IMCD). In the IMCD the CaSR is expressed in principal cells and co-localizes with aquaporin 2 in the apical membrane. Isolated IMCD does not respond to vasopressin (DDAVP) when also exposed to a high Ca2+ concentration (Sands et al., 1997).

Regulation by 1,25-(OH)2D3

Vitamin D3 (cholecalciferol) is essential for body Ca2+ homeostasis. Humans can absorb vitamin D3 from the diet and synthesize it in the skin from its precursor 7-dehydrocholesterol in response to sunlight. Biologically active vitamin D3 needs a double hydroxylation process occurring first in the liver and then in the kidney. In the liver 25-hydroxylation and in the kidney 1α-hydroxylation contribute to the synthesis of the biologically active 1,25-(OH)2D3. Mitochondria of renal proximal tubular cells produce the final 1,25-(OH)2D3, depending on the circulating Ca2+ level and PTH. When there is adequate dietary Ca2+ intake and a normal plasma Ca2+ concentration, 1α-OHase activity is low. However, when dietary Ca2+ intake is low and serum or plasma Ca2+ concentration decreased, the activity of this enzyme increases, promoting 1,25(OH)2D3 synthesis and a rise in serum Ca2+ level. The physiological effect of 1,25-(OH)2D3 is mediated by interaction with the nuclear vitamin D receptor (VDR) (Haussler et al., 1998). Notably, VDR expression is not confined to the tissues involved in Ca2+ homeostasis, specifically intestine, kidney, and bone, but it is present in many cell types and has other functions (Pike et al., 2010).

The effect of 1,25-(OH)2D3 on the intestine

Studies from VDR null mice show that 1,25-(OH)2D3 is mainly involved in the active Ca2+ transport in the intestine. After weaning, VDR null mice have reduced (40%) Ca2+ absorption and develop hypocalcaemia. This phenotype is rescued by either selectively reintroducing the VDR in only the intestine or by a high calcium/lactose diet (Lieben et al., 2011). Transcellular Ca2+ transport in the small intestine is stimulated by 1,25-(OH)2D3 via a genomic action. Active vitamin D promotes the transcription of the Ca2+ transport proteins TRPV5, TRPV6, the calbindins, NCX1, and PMCA1b (Lieben et al., 2011). A potential role for 1,25-(OH)2D3 in modulating passive Ca2+ absorption in the intestine comes from the evidence that active vitamin D increases the expression of claudins 2 and 12 (Fujita et al., 2008).

The effect of 1,25-(OH)2D3 on the kidney

In the kidney 1,25-(OH)2D3 promotes Ca2+ reabsorption in the DCT. Several studies have tried to elucidate whether such an action is the consequence of a direct effect of 1,25-(OH)2D3 or if it is mediated predominantly by a concomitant PTH increase. Early studies of parathyroidectomized dogs supplemented with high doses of active vitamin D showed a direct effect of 1,25-(OH)2D3 on renal Ca2+ handling (Puschett et al., 1972). These observations have been confirmed by data showing that VDR KO mice have inappropriate hypercalciuria both in normo- and hypocalcaemic conditions. In addition, mice lacking the active form of vitamin D3 (Cyp27b1 KO mice) have lower levels of TRPV5, Calbindin-D9k, Calbindin-D28k, and NCX1 in the kidney. This phenotype is rescued by 1,25(OH)2D3 administration (Hoenderop et al., 2002b).

237

238

Section 2  

fluid, electrolyte, and renal tubular disorders

The effect of 1,25-(OH)2D3 on bone

Vitamin D deficiency leads to rickets. Skeletal abnormalities start after weaning and are strongly related to concomitant changes in Ca2+ and phosphate handling. High dietary Ca2+ intake (Dardenne et  al., 2003)  and selective introduction of VDR to the intestine of VDR null mice (Xue and Fleet, 2009)  can rescue the vitamin D-induced skeleton changes. However, since the recent identification of the VDR in chondrocytes, osteoblasts, and osteoclasts, a strong research interest is growing in the role of 1,25-(OH)2D3 in the regulation of bone metabolism and thereby of Ca2+ homeostasis. Briefly, 1,25-(OH)2D3 appears to inhibit osteoblast differentiation and thus bone matrix mineralization (Shi et al., 2007); these findings are confirmed in an experimental model of hypervitaminosis D (St-Arnaud et al., 2000). In the osteoblast 1,25-(OH)2D3 promotes the synthesis of the receptor activator of nuclear factor kappa-B ligand (RANKL) and inhibits the transcription of the osteoprotegerin, activating, ultimately, the osteoclast. Both osteoblast inhibition and osteoclast indirect activation contribute to the 1,25-(OH)2D3-related increase in serum Ca2+ level (Lieben et al., 2011).

Regulation by parathyroid hormone The parathyroid glands are the main organ finely tuning the blood Ca2+ level. Changes in blood Ca2+ are sensed by the CaSR expressed in the chief cells of the parathyroid and PTH secretion is adjusted accordingly; PTH is released in response to a low blood Ca2+ level. PTH is a ligand for the PTH/PTHrP receptor, an interaction that promotes Ca2+ reabsorption in the kidney, the bone and intestine. In the proximal tubule, PTH stimulates the activity of 1α-OHase to increase the circulating level of active vitamin D (1,25-(OH)2D3). Although, the hypercalcaemic effect of PTH is related to its effect on vitamin D activity, experimental models have assessed the direct action of PTH in each of the main tissues involved in Ca2+ homeostasis.

The effect of PTH on the kidney PTH/PTHrP receptor mRNA has been identified in several cell types of the rat kidney, including glomerular podocytes, convoluted and straight parts of the proximal tubule, cortical thick ascending limb and DCT, though not in the thin limb of Henle’s loop or collecting duct. Immunohistochemical localization confirms its presence on the basolateral membrane (Lupp et al., 2010). In the proximal tubule, PTH regulates vitamin D synthesis and phosphate transport, while Ca2+ reabsorption in this segment is mainly paracellular and seems to be independent of hormonal regulation (Friedman, 2000). Along the medullary and cortical part of the TAL, PTH enhances both transcellular and paracellular Ca2+ absorption (Friedman and Gesek, 1995), although the entry pathway for Ca2+ at this site has not been identified. PTH promotes active Ca2+ reabsorption in DCT by inducing the expression of the main Ca2+ regulating proteins expressed in the DCT. PTH administration to parathyroidectomized rats can restore the reduced expression of TRPV5, calbindin-D28k and NCX1 along the DCT2 and CNT (van Abel et al., 2005). In addition, PTH can stimulate the PMCA activity by increasing its affinity for Ca2+. This effect has not been observed after stimulation with vitamin D (Tsukamoto et al., 1992). The PTH signal is mediated by a process involving PKA and PKC-dependent pathways (Friedman et  al., 1996)  affecting Ca2+

channel activation, sorting (Bacskai and Friedman, 1990) and cellular hyperpolarization (Friedman and Gesek, 1994).

The effect of PTH on the intestine and bone Along the intestine PTH/PTHrP receptor is expressed on both the apical and basolateral side of the epithelial cells, but not in the globet cells. Enterocytes increase Ca2+ absorption when cultured and stimulated with PTH (Gentili et al., 2003), showing that in the gut active vitamin D is not the only regulator of Ca2+ absorption (Nemere and Larsson, 2002). Bone is one of the main targets for PTH in the regulation of Ca2+ homeostasis. PTHrP is expressed in osteoblasts, while it is still uncertain whether the same receptor is present in osteoclasts. PTH stimulates osteoblasts to secrete RANKL and inhibits osteoprotegerin (OPG) a target decoy for RANKL. Osteoclasts are activated by PTH stimulation through the interaction of RANKL with RANK. OPG can interfere with this system, directly binding RANKL and so preventing the activation of osteoclasts. PTH normally induces both bone formation and bone reabsorption, increasing the total bone turnover. At the molecular level this is achieved by a fine tuning of the RANKL–OPG–RANK axis. Pathologically, PTH can induce severe bone reabsorption (primary hyperparathyroidism) or when given therapeutically new bone formation (intermittent PTH supplement therapy for osteoporosis) (Aslan et al., 2012).

Calcium regulation by calcitonin CT is produced by the thyroidal C cells and is a serum Ca2+-lowering hormone released in response to a hypercalcaemic stimulus. Its effect derives primarily from inhibition of osteoclast-mediated bone reabsorption. CT receptors are expressed in osteoclasts and their activation leads to cellular detachment from the mineral surface and inhibition of acid secretion. CT is used in the treatment of malignant hypercalcaemia, in osteoporosis, and Paget disease (Civitelli et al., 1988). CT KO mice are more sensitive to hypercalcaemic stimuli, confirming its counteracting role when compared with PTH and active vitamin D.  In addition, CT KO mice have higher bone density, indicating a potential action of CT in regulating bone formation (Hoff et al., 2002). CT has a direct effect on modulating renal Na+ and K+ excretion. Recent evidence shows that it can also directly influence renal Ca2+ excretion. The molecular mechanism of this process is not well understood, but experiments performed on TRPV5 KO mice seem to exclude any effect of CT on this Ca2+ channel (Hsu et al., 2010).

Calcium regulation by sex hormones Humans show gender differences in Ca2+ handling. Nephrolithiasis is less common in women than men before 50 years of age, but this gender difference almost disappears over 50 years of age. Therefore, the menopause is typically associated with increased urinary Ca2+ excretion. These findings have been confirmed in rodents:  male mice excrete more Ca2+ than females during the fertile period. Orchidectomy induces hypocalciuria in male mice and this is rescued by testosterone supplementation. Testosterone-associated urinary Ca2+ excretion is mediated by decreased expression of TRPV5, NCX1, PMCA1b and calbindin-D28k (Hsu et al., 2010b). Oestrogens have the opposite effect on the Ca2+ reabsorption along the DCT (van Abel et al., 2002). Oestrogens modulate the

chapter 26 

expression of TRPV5 directly in a vitamin D-independent manner, as shown in ovariectomized 1α-OHase KO mice. Oestrogens also affect duodenal Ca2+ transport via TRPV6. Oestrogen receptor KO mice have lower expression of both TRPV5 and TRPV6, and this is reversed by oestrogen supplementation (Van Cromphaut et al., 2003). Whether these changes are associated with decreased Ca2+ reabsorption has not been established. However, a recent clinical trial has shown that oestrogen supplements in postmenopausal women increase the risk of kidney stones (Maalouf et al., 2010), although this may be confounded by the additional use of oral calcium supplements.

Calcium regulation by thyroid hormone There is a lot of evidence that thyroid hormone status influences Ca2+ metabolism (Capasso et al., 1987; Cross et al., 1990). Thyrotoxicosis can be accompanied by hypercalcaemia. The severity of thyrotoxicosis directly correlates with bone demineralization and altered biochemical markers of bone turnover (El Hadidy et  al., 2011). Hyperthyroid rats have lower Ca2+ transport rates both at apical and basolateral membranes of enterocytes; the opposite is true in hypothyroid rats (Kumar and Prasad, 2003). Similar findings are described in the kidney (Kumar and Prasad, 2002).

Calcium regulation by magnesium and urinary pH Magnesium can affect urinary Ca2+ excretion (Chesley and Tepper, 1958). How Mg2+ induces hypercalciuria is largely unknown. Systemic changes in pH can also influence Ca2+ homeostasis. Metabolic acidosis induces hypercalciuria, at least in part by increasing the filtered load of Ca2+ from increased free (ionized) calcium (Rizzo et al., 2000), and by decreasing renal Ca2+ reabsorption (Moe and Huang, 2006; Nijenhuis etal., 2006). This effect is also observed during a high dietary protein intake (Amanzadeh et al., 2003). Conversely, urinary alkalinization by potassium citrate or bicarbonate decreases urinary Ca2+ (Sebastian et al., 1994). TRPV5 and TRPV6 function is modulated by variations in the urine pH within the physiological range (Bindels et al., 1994). An acid urine pH increases calciuria by inhibiting TRPV5/6-mediated Ca2+ reabsorption in the DCT. One study (Bonny et al., 2008) has shown that Ca2+ uptake by TRPV5 is directly inhibited by Mg2+ and a low pH. These findings may explain the interaction of Mg2+ with Ca2+ and urinary pH; moreover, they are consistent with the observation that urinary alkalinization and Mg2+ supplementation reduce kidney stones formation.

Calcium regulation by ciclosporin and tacrolimus Abnormalities in mineral metabolism are common complications of organ transplantation. Treatment with ciclosporin and tacrolimus is associated with increased bone turnover and hypercalciuria, leading to osteoporosis (Stempfle et al., 2002). Tacrolimus significantly increases urinary Ca2+ excretion through downregulation of both mRNA and protein expression of TRPV5 and calbindin-D28k (Nijenhuis et al., 2004). Ciclosporin can affect Ca2+ transport along the DCT by reducing the expression of the VDR and consequently inducing vitamin D resistance in this segment. This mechanism is

calcium homeostasis

supported by the inability of the kidney to retain Ca2+, even in presence of elevated circulating levels of vitamin D (Lee et al., 2011).

Calcium homeostasis in hypertension Essential hypertension is associated with hypercalciuria. Hypertensive patients have relative hypercalciuria in the presence of enhanced basal parathyroid function (McCarron et al., 1980). In addition, hypertensive patients have on average a 20% increase in Ca2+ excretion at any given level of urinary sodium (Strazzullo et  al., 1983). Animal models of experimental hypertension confirm this association. The spontaneous hypertensive rat (SHR) has lower renal abundance of PMCA1b and an increase in expression of Calbindin-D28k (Kamijo et  al., 1996). The Milan hypertensive strain (MHS) of rat is also hypercalciuric. Recent data demonstrate that all the Ca2+ transport proteins are upregulated at both mRNA and protein levels along the distal tubules of MHS animals when compared with the normotensive strain (Petrazzuolo et al., 2010). Taken together, these data indicate that hypercalciuria in experimental hypertension results from decreased tubular reabsorption of Ca2+ along the DCT.

References Amanzadeh, J., Gitomer, W. L., Zerwekh, J. E., et al. (2003). Effect of high protein diet on stone-forming propensity and bone loss in rats. Kidney Int, 64, 2142–9. Aslan, D., Andersen, M. D., Gede, L. B., et al. (2012). Mechanisms for the bone anabolic effect of parathyroid hormone treatment in humans. Scand J Clin Lab Invest, 72(1), 14–22. Ba, J., Brown, D., and Friedman, P. A. (2003). Calcium-sensing receptor regulation of PTH-inhibitable proximal tubule phosphate transport. Am J Physiol Renal Physiol, 285, F1233–43. Bacskai, B. J. and Friedman, P. A. (1990). Activation of latent Ca2+ channels in renal epithelial cells by parathyroid hormone. Nature, 347, 388–91. Benn, B. S., Ajibade, D., Porta, A., et al. (2008). Active intestinal calcium transport in the absence of transient receptor potential vanilloid type 6 and calbindin-D9k. Endocrinology, 149, 3196–205. Bergh, J. J., Shao, Y., Puente, E., et al. (2006). Osteoblast Ca2+ permeability and voltage-sensitive Ca2+ channel expression is temporally regulated by 1,25-dihydroxyvitamin D3. Am J Physiol Cell Physiol, 290, C822–31. Bindels, R. J., Hartog, A., Abrahamse, S. L., et al. (1994). Effects of pH on apical calcium entry and active calcium transport in rabbit cortical collecting system. Am J Physiol, 266, F620–7. Bindels, R. J. (2010). 2009 Homer W. Smith Award: minerals in motion: from new ion transporters to new concepts. J Am Soc Nephrol, 21, 1263–9. Blair, H. C., Robinson, L. J., Huang, C. L., et al. (2011). Calcium and bone disease. Biofactors, 37, 159–67. Blanchard, A., Jeunemaitre, X., Coudol, P., et al. (2001). Paracellin-1 is critical for magnesium and calcium reabsorption in the human thick ascending limb of Henle. Kidney Int, 59, 2206–15. Bonny, O., Rubin, A., Huang, C. L., et al. (2008). Mechanism of urinary calcium regulation by urinary magnesium and pH. J Am Soc Nephrol, 19, 1530–7. Bootman, M. D., Collins, T. J., Peppiatt, C. M., et al. (2001). Calcium signalling—an overview. Semin Cell Dev Biol, 12, 3–10. Bourdeau, J. E. and Burg, M. B. (1979). Voltage dependence of calcium transport in the thick ascending limb of Henle’s loop. Am J Physiol, 236, F357–64. Bronner, F., Pansu, D., and Stein, W. D. (1986). An analysis of intestinal calcium transport across the rat intestine. Am J Physiol, 250, G561–9. Brown, E. M. (2007). The calcium-sensing receptor: physiology, pathophysiology and CaR-based therapeutics. Subcell Biochem, 45, 139–67.

239

240

Section 2  

fluid, electrolyte, and renal tubular disorders

Brown, E. M., Gamba, G., Riccardi, D., et al. (1993). Cloning and characterization of an extracellular Ca2+-sensing receptor from bovine parathyroid. Nature, 366, 575–80. Capasso, G., De Santo, N.G., and Kinne, R. (1987). Thyroid hormones and renal transport: cellular and biochemical aspects. Kidney Int, 32, 443–51. Capasso, G., Geibel, P. J., Damiano, S., et al. (2013). The calcium sensing receptor modulates fluid reabsorption and acid secretion in the proximal tubule. Kidney Int, 84(2), 277–84. Chang, Q., Hoefs, S., and van der Kemp, A. W, et al. (2005). The beta-glucuronidase klotho hydrolyzes and activates the TRPV5 channel. Science, 310, 490–3. Chang, W., Tu, C., Chen, T. H., et al. (2008). The extracellular calcium-sensing receptor (CaSR) is a critical modulator of skeletal development. Sci Signal, 1, ra1. Chattopadhyay, N., Baum, M., Bai, M., et al. (1996). Ontogeny of the extracellular calcium-sensing receptor in rat kidney. Am J Physiol, 271, F736–43. Chesley, L. C. and Tepper, I. (1958). Some effects of magnesium loading upon renal excretion of magnesium and certain other electrolytes. J Clin Invest, 37, 1362–72. Christakos, S. (2012). Mechanism of action of 1,25-dihydroxyvitamin D(3) on intestinal calcium absorption. Rev Endocr Metab Disord, 13(1), 39–44. Christakos, S., Dhawan, P., Porta, A., et al. 2011). Vitamin D and intestinal calcium absorption. Mol Cell Endocrinol, 347(1–2), 25–9. Civitelli, R., Gonnelli, S., Zacchei, F., et al. (1988). Bone turnover in postmenopausal osteoporosis. Effect of calcitonin treatment. J Clin Invest, 82, 1268–74. Costanzo, L. S. and Weiner, I. M. (1976). Relationship between clearances of Ca and Na: effect of distal diuretics and PTH. Am J Physiol, 230, 67–73. Costanzo, L. S., Windhager, E. E., and Ellison, D. H. (2000). Calcium and sodium transport by the distal convoluted tubule of the rat. 1978. J Am Soc Nephrol, 11, 1562–80. Cross, H. S., Debiec, H., and Peterlik, M. (1990). Mechanism and regulation of intestinal phosphate absorption. Miner Electrolyte Metab, 16, 115–24. Dardenne, O., Prud’homme, J., Hacking, S. A., et al. (2003). Correction of the abnormal mineral ion homeostasis with a high-calcium, high-phosphorus, high-lactose diet rescues the PDDR phenotype of mice deficient for the 25-hydroxyvitamin D-1alpha-hydroxylase (CYP27B1). Bone, 32, 332–40. Egbuna, O., Quinn, S., Kantham, L., et al. (2009). The full-length calcium-sensing receptor dampens the calcemic response to 1alpha,25(OH)2 vitamin D3 in vivo independently of parathyroid hormone. Am J Physiol Renal Physiol, 297, F720–8. El Hadidy, E. H. M, Ghonaim, M., El Gawad, S. S., et al. (2011). Impact of severity, duration, and etiology of hyperthyroidism on bone turnover markers and bone mineral density in men. BMC Endocr Disord, 11, 15. Fleet, J. C. and Schoch, R. D. (2010). Molecular mechanisms for regulation of intestinal calcium absorption by vitamin D and other factors. Crit Rev Clin Lab Sci, 47, 181–95. Friedman, P. A., Coutermarsh, B. A., Kennedy, S. M., et al. (1996). Parathyroid hormone stimulation of calcium transport is mediated by dual signaling mechanisms involving protein kinase A and protein kinase C. Endocrinology, 137, 13–20. Friedman, P. A. (1988). Basal and hormone-activated calcium absorption in mouse renal thick ascending limbs. Am J Physiol, 254, F62–70. Friedman, P. A. (2000). Mechanisms of renal calcium transport. Exp Nephrol, 8, 343–50. Friedman, P.A. and Gesek, F. A. (1994). Hormone-responsive Ca2+ entry in distal convoluted tubules. J Am Soc Nephrol, 4, 1396–404. Friedman, P. A. and Gesek, F. A. (1995). Cellular calcium transport in renal epithelia: measurement, mechanisms, and regulation. Physiol Rev, 75, 429–71. Fujita, H., Sugimoto, K., Inatomi, S., et al. (2008). Tight junction proteins claudin-2 and -12 are critical for vitamin D-dependent Ca2+ absorption between enterocytes. Mol Biol Cell, 19, 1912–21. Gentili, C., Morelli, S., de Boland, A. R. (2003). Characterization of PTH/ PTHrP receptor in rat duodenum: effects of ageing. J Cell Biochem, 88, 1157–67.

Greger, R., Schlatter, E., Hebert, S. C. (2001). Milestones in nephrology: presence of luminal K+, a prerequisite for active NaCl transport in the cortical thick ascending limb of Henle’s loop of rabbit kidney. J Am Soc Nephrol, 12, 1788–93. Haussler, M. R., Whitfield, G. K., Haussler, C. A., et al. (1998). The nuclear vitamin D receptor: biological and molecular regulatory properties revealed. J Bone Miner Res, 13, 325–49. Hebert, S. C. (1999). Molecular mechanisms. Semin Nephrol, 19, 504–23. Hoenderop, J. G., Chon, H., Gkika, D., et al. (2004). Regulation of gene expression by dietary Ca2+ in kidneys of 25-hydroxyvitamin D3-1 alpha-hydroxylase knockout mice. Kidney Int, 65, 531–9. Hoenderop, J. G., Dardenne, O., Van Abel, M., et al. (2002a). Modulation of renal Ca2+ transport protein genes by dietary Ca2+ and 1,25-dihydroxyvitamin D3 in 25-hydroxyvitamin D3-1alpha-hydroxylase knockout mice. FASEB J, 16, 1398–406. Hoenderop, J. G., Nilius, B., Bindels, R. J. (2002b). Molecular mechanism of active Ca2+ reabsorption in the distal nephron. Annu Rev Physiol, 64, 529–49. Hoenderop, J. G., van der Kemp, A. W., Hartog, A., et al. (1999). Molecular identification of the apical Ca2+ channel in 1, 25-dihydroxyvitamin D3-responsive epithelia. J Biol Chem, 274, 8375–8. Hoenderop, J. G., van Leeuwen, J. P., van der Eerden, B. C., et al. (2003). Renal Ca2+ wasting, hyperabsorption, and reduced bone thickness in mice lacking TRPV5. J Clin Invest, 112, 1906–14. Hoff, A. O., Catala-Lehnen, P., Thomas, P. M., et al. (2002). Increased bone mass is an unexpected phenotype associated with deletion of the calcitonin gene. J Clin Invest, 110, 1849–57. Hsu, Y. J., Dimke, H., Hoenderop, J. G., et al. (2010a). Calcitonin-stimulated renal Ca2+ reabsorption occurs independently of TRPV5. Nephrol Dial Transplant, 25, 1428–35. Hsu, Y. J., Dimke, H., Schoeber, J. P., et al. (2010b). Testosterone increases urinary calcium excretion and inhibits expression of renal calcium transport proteins. Kidney Int, 77, 601–8. Hu, M. C., Kuro-o, M., and Moe, O. W. (2010). Klotho and kidney disease. J Nephrol, 23 Suppl 16, S136–44. Imai, M. (1978). Calcium transport across the rabbit thick ascending limb of Henle’s loop perfused in vitro. Pflugers Arch, 374, 255–63. Imura, A., Tsuji, Y., Murata, M., et al. (2007). Alpha-Klotho as a regulator of calcium homeostasis. Science, 316, 1615–8. Jurutka, P. W., Whitfield, G. K., Hsieh, J. C., et al. (2001). Molecular nature of the vitamin D receptor and its role in regulation of gene expression. Rev Endocr Metab Disord, 2, 203–16. Kamijo, T., Gonzalez, J. M., Jost, L. J., et al. (1996). Renal abnormality of calcium handling in spontaneously hypertensive rats. Kidney Int Suppl, 55, S166–8. Kumar, V. and Prasad, R. (2002). Molecular basis of renal handling of calcium in response to thyroid hormone status of rat. Biochim Biophys Acta, 1586, 331–43. Kumar, V. and Prasad, R. (2003). Thyroid hormones stimulate calcium transport systems in rat intestine. Biochim Biophys Acta, 1639, 185–94. Kutuzova, G. D., Akhter, S., Christakos, S., et al. (2006). Calbindin D(9k) knockout mice are indistinguishable from wild-type mice in phenotype and serum calcium level. Proc Natl Acad Sci U S A, 103, 12377–81. Kutuzova, G. D., Sundersingh, F., Vaughan, J., et al. (2008). TRPV6 is not required for 1alpha,25-dihydroxyvitamin D3-induced intestinal calcium absorption in vivo. Proc Natl Acad Sci U S A, 105, 19655–9. Lambers, T. T., Mahieu, F., Oancea, E., et al. (2006). Calbindin-D28K dynamically controls TRPV5-mediated Ca2+ transport. EMBO J, 25, 2978–88. Lee, C. T., Chen, H. C., Lai, L. W., et al. (2007). Effects of furosemide on renal calcium handling. Am J Physiol Renal Physiol, 293, F1231–7. Lee, C. T., Ng, H. Y., Lien, Y. H., et al. (2011). Effects of cyclosporine, tacrolimus and rapamycin on renal calcium transport and vitamin D metabolism. Am J Nephrol, 34, 87–94. Lieben, L., Benn, B. S., Ajibade, D., et al. (2010). Trpv6 mediates intestinal calcium absorption during calcium restriction and contributes to bone homeostasis. Bone, 47, 301–8.

chapter 26 

Lieben, L., Carmeliet, G., and Masuyama, R. (2011). Calcemic actions of vitamin D: effects on the intestine, kidney and bone. Best Pract Res Clin Endocrinol Metab, 25, 561–72. Loffing, J., Loffing-Cueni, D., Valderrabano, V., et al. (2001). Distribution of transcellular calcium and sodium transport pathways along mouse distal nephron. Am J Physiol Renal Physiol, 281, F1021–7. Loupy, A., Ramakrishnan, S. K., Wootla, B., et al. (2012). PTH-independent regulation of blood calcium concentration by the calcium-sensing receptor. J Clin Invest, 122, 3355–67. Lupp, A., Klenk, C., Rocken, C., et al. (2010). Immunohistochemical identification of the PTHR1 parathyroid hormone receptor in normal and neoplastic human tissues. Eur J Endocrinol, 162, 979–86. Maiti, A., Hait, N. C., and Beckman, M. J. (2008). Extracellular calcium-sensing receptor activation induces vitamin D receptor levels in proximal kidney HK-2G cells by a mechanism that requires phosphorylation of p38alpha MAPK. J Biol Chem, 283, 175–83. McCabe, L. D., Martin, B. R., McCabe, G. P., et al. (2004). Dairy intakes affect bone density in the elderly. Am J Clin Nutr, 80, 1066–74. McCarron, D. A., Pingree, P. A., Rubin, R. J., et al. (1980). Enhanced parathyroid function in essential hypertension: a homeostatic response to a urinary calcium leak. Hypertension, 2, 162–8. Moe, O. W. and Huang, C. L. (2006). Hypercalciuria from acid load: renal mechanisms. J Nephrol, 19 Suppl 9, S53–61. Motoyama, H. I. and Friedman, P. A. (2002). Calcium-sensing receptor regulation of PTH-dependent calcium absorption by mouse cortical ascending limbs. Am J Physiol Renal Physiol, 283, F399–406. Maalouf, N. M., Sato, A. H., Welch, B. J., et al. (2010). Postmenopausal hormone use and the risk of nephrolithiasis: results from the Women’s Health Initiative hormone therapy trials. Arch Intern Med, 170, 1678–85. Nemere, I. and Larsson, D. (2002). Does PTH have a direct effect on intestine? J Cell Biochem, 86, 29–34. Nesbitt, S. A. and Horton, M. A. (1997). Trafficking of matrix collagens through bone-resorbing osteoclasts. Science, 276, 266–9. Nijenhuis, T., Hoenderop, J. G., and Bindels, R. J. (2004). Downregulation of Ca2+ and Mg2+ transport proteins in the kidney explains tacrolimus (FK506)-induced hypercalciuria and hypomagnesemia. J Am Soc Nephrol, 15, 549–57. Nijenhuis, T., Renkema, K. Y., Hoenderop, J. G., et al. (2006). Acid-base status determines the renal expression of Ca2+ and Mg2+ transport proteins. J Am Soc Nephrol, 17, 617–26. Nijenhuis, T., Vallon, V., van der Kemp, A. W., et al. (2005). Enhanced passive Ca2+ reabsorption and reduced Mg2+ channel abundance explains thiazide-induced hypocalciuria and hypomagnesemia. J Clin Invest, 115, 1651–8. Orimo, H. (2010). The mechanism of mineralization and the role of alkaline phosphatase in health and disease. J Nihon Med Sch, 77, 4–12. Peacock, M. (2010). Calcium metabolism in health and disease. Clin J Am Soc Nephrol, 5 Suppl 1, S23–30. Petrazzuolo, O., Trepiccione, F., Zacchia, M., et al. (2010). Hypertension and renal calcium transport. J Nephrol, 23 Suppl 16, S112–17. Pike, J. W. and Meyer, M. B. (2010). The vitamin D receptor: new paradigms for the regulation of gene expression by 1,25-dihydroxyvitamin D(3). Endocrinol Metab Clin North Am, 39, 255–69. Potts, J. T. and Gardella, T. J. (2007). Progress, paradox, and potential: parathyroid hormone research over five decades. Ann N Y Acad Sci, 1117, 196–208. Puschett, J. B., Moranz, J., and Kurnick, W. S. (1972). Evidence for a direct action of cholecalciferol and 25-hydroxycholecalciferol on the renal transport of phosphate, sodium, and calcium. J Clin Invest, 51, 373–85. Quamme, G. A. (1982). Effect of hypercalcemia on renal tubular handling of calcium and magnesium. Can J Physiol Pharmacol, 60, 1275–80. Quinn, S. J., Bai, M., and Brown, E. M. (2004). pH sensing by the calcium-sensing receptor. J Biol Chem, 279, 37241–9. Redhead, C. R. (1988). Ionic regulation of intracellular pH in rat calvarial osteoblasts. J Physiol, 401, 455–68. Renkema, K. Y., Velic, A., Dijkman, H. B., et al. (2009). The calcium-sensing receptor promotes urinary acidification to prevent nephrolithiasis. J Am Soc Nephrol, 20, 1705–13.

calcium homeostasis

Riccardi, D., Hall, A. E., Chattopadhyay, N., et al. (1998). Localization of the extracellular Ca2+/polyvalent cation-sensing protein in rat kidney. Am J Physiol, 274, F611–22. Rizzo, M., Capasso, G., Bleich, M., et al. (2000). Effect of chronic metabolic acidosis on calbindin expression along the rat distal tubule. J Am Soc Nephrol, 11, 203–10. Rizzo, M., Metafora, S., Morelli, F., et al. (2004). Chronic administration of bumetanide upregulates calbindin D28k mRNA and protein abundance in rat distal convoluted tubules. Nephron Physiol, 97, 16–22. Salo, J., Lehenkari, P., Mulari, M., et al. (1997). Removal of osteoclast bone resorption products by transcytosis. Science, 276, 270–3. Sands, J. M., Naruse, M., Baum, M., et al. (1997). Apical extracellular calcium/polyvalent cation-sensing receptor regulates vasopressin-elicited water permeability in rat kidney inner medullary collecting duct. J Clin Invest, 99, 1399–405. Schlatter, E. (2006). Who wins the competition: TRPV5 or calbindin-D28K? J Am Soc Nephrol, 17, 2954–6. Sebastian, A., Harris, S. T., Ottaway, J. H., et al. (1994). Improved mineral balance and skeletal metabolism in postmenopausal women treated with potassium bicarbonate. N Engl J Med, 330, 1776–81. Shi, Y. C., Worton, L., Esteban, L., et al. (2007). Effects of continuous activation of vitamin D and Wnt response pathways on osteoblastic proliferation and differentiation. Bone, 41, 87–96. Simon, D. B. and Lifton, R. P. (1996). The molecular basis of inherited hypokalemic alkalosis: Bartter’s and Gitelman’s syndromes. Am J Physiol, 271, F961–6. Song, Y., Peng, X., Porta, A., et al. (2003). Calcium transporter 1 and epithelial calcium channel messenger ribonucleic acid are differentially regulated by 1,25 dihydroxyvitamin D3 in the intestine and kidney of mice. Endocrinology, 144, 3885–94. Stains, J. P., Weber, J. A., and Gay, C. V. (2002). Expression of Na+/Ca2+ exchanger isoforms (NCX1 and NCX3) and plasma membrane Ca2+ ATPase during osteoblast differentiation. J Cell Biochem, 84, 625–35. St-Arnaud, R., Arabian, A., Travers, R., et al. (2000). Deficient mineralization of intramembranous bone in vitamin D-24-hydroxylase-ablated mice is due to elevated 1,25-dihydroxyvitamin D and not to the absence of 24,25-dihydroxyvitamin D. Endocrinology, 141, 2658–66. Stempfle, H. U., Werner, C., Siebert, U., et al. (2002). The role of tacrolimus (FK506)-based immunosuppression on bone mineral density and bone turnover after cardiac transplantation: a prospective, longitudinal, randomized, double-blind trial with calcitriol. Transplantation, 73, 547–52. Strazzullo, P., Nunziata, V., Cirillo, M., et al. (1983). Abnormalities of calcium metabolism in essential hypertension. Clin Sci (Lond), 65, 137–41. Suki, W. N. (1979). Calcium transport in the nephron. Am J Physiol, 237, F1–6. Thyagarajan, B., Benn, B. S., Christakos, S., et al. (2009). Phospholipase C-mediated regulation of transient receptor potential vanilloid 6 channels: implications in active intestinal Ca2+ transport. Mol Pharmacol, 75, 608–16. Topala, C. N., Schoeber, J. P., Searchfield, L. E., et al. (2009). Activation of the Ca2+-sensing receptor stimulates the activity of the epithelial Ca2+ channel TRPV5. Cell Calcium, 45, 331–9. Tsukamoto, Y., Saka, S., and Saitoh, M. (1992). Parathyroid hormone stimulates ATP-dependent calcium pump activity by a different mode in proximal and distal tubules of the rat. Biochim Biophys Acta, 1103, 163–71. Turnbull, C. I., Looi, K., Mangum, J. E., et al. (2004). Calbindin independence of calcium transport in developing teeth contradicts the calcium ferry dogma. J Biol Chem, 279, 55850–4. Ullrich, K. J., Capasso, G., Rumrich, G., et al. (1977). Coupling between proximal tubular transport processes. Studies with ouabain, SITS and HCO3-free solutions. Pflugers Arch, 368, 245–52. Van Abel, M., Hoenderop, J. G., Dardenne, O., et al. (2002). 1,25-dihydroxyvitamin D(3)-independent stimulatory effect of estrogen on the expression of ECaC1 in the kidney. J Am Soc Nephrol, 13, 2102–9. Van Abel, M., Hoenderop, J. G., van der Kemp, A. W., et al. (2005). Coordinated control of renal Ca2+ transport proteins by parathyroid hormone. Kidney Int, 68, 1708–21.

241

242

Section 2  

fluid, electrolyte, and renal tubular disorders

Van Cromphaut, S. J., Rummens, K., Stockmans, I., et al. (2003). Intestinal calcium transporter genes are upregulated by estrogens and the reproductive cycle through vitamin D receptor-independent mechanisms. J Bone Miner Res, 18, 1725–36. Vargas-Poussou, R., Huang, C., Hulin, P., et al. (2002). Functional characterization of a calcium-sensing receptor mutation in severe autosomal dominant hypocalcemia with a Bartter-like syndrome. J Am Soc Nephrol, 13, 2259–66. Wang, L., Nancollas, G. H., Henneman, Z. J., et al. (2006). Nanosized particles in bone and dissolution insensitivity of bone mineral. Biointerphases, 1, 106–11.

Wang, W., Lu, M., Balazy, M., et al. (1997). Phospholipase A2 is involved in mediating the effect of extracellular Ca2+ on apical K+ channels in rat TAL. Am J Physiol, 273, F421–9. Wasserman, R. H. (1964). Vitamin d and the intestinal absorption of calcium. N Y State J Med, 64, 1329–32. Wittner, M., Mandon, B., Roinel, N., et al. (1993). Hormonal stimulation of Ca2+ and Mg2+ transport in the cortical thick ascending limb of Henle’s loop of the mouse: evidence for a change in the paracellular pathway permeability. Pflugers Arch, 423, 387–96. Xue, Y. and Fleet, J. C. (2009). Intestinal vitamin D receptor is required for normal calcium and bone metabolism in mice. Gastroenterology, 136,1317–22.

CHAPTER 27

Magnesium homeostasis Pascal Houillier Introduction Magnesium (Mg) is the fourth most abundant cation in vertebrates and the second most abundant intracellular cation. Magnesium is critical for a number of biological processes, the most important being the release of chemical energy: adenosine triphosphate (ATP) is formed by Mg-dependent oxidative phosphorylation. Magnesium is also involved in many enzymatic reactions within the cell: enzymes that use ATP use it as a metal chelate, MgATP. Magnesium is also required for glycolysis, DNA transcription, and protein synthesis. Finally, free magnesium (Mg2+) plays a role in ion currents and membrane voltage stabilization.

Distribution in the body Bone and soft tissue contain about 66% and 33% of body magnesium, respectively (1000 mmol, or 24 g in an adult human) (Rude, 1996; Ahmad and Sutton, 2000)  (Fig. 27.1). Bony magnesium is mainly deposited on the bone crystal surfaces and is therefore rapidly exchangeable. Extracellular magnesium represents only 1% of the total body content. The normal plasma magnesium concentration is 0.70–1 mmol/L (1.7–2.4 mg/dL), and comprises 70–80% of free magnesium, the remainder being bound to proteins or complexed with bicarbonate, phosphate, and citrate (Le Grimellec et al., 1975). The concentration of magnesium within the cells is 5–20 mmol/L. Most of intracellular magnesium is bound to metalloenzymes and phosphate, and intracellular magnesium is compartmentalized with the highest concentration in microsomes and mitochondria. Within the cytosol, most (80%) of the magnesium is complexed to ATP. Only 1–5% of cellular Mg is ionized (free). The smaller pool of free magnesium in the cell is in equilibrium with the bound magnesium. This equilibrium can buffer the changes in free magnesium concentration under conditions of magnesium excess or depletion. As a consequence, the cytosolic concentration of magnesium is maintained in the optimal range for many enzymatic reactions (Gunther, 1981), and the difference in concentration between the extracellular fluid and the cytosol is small. Nevertheless, magnesium can enter the cells down an electrical gradient owing to the relative intracellular electronegativity. Conversely, the exit of magnesium out of the cell is against the electrical gradient and, is therefore necessarily active.

Balance of magnesium (For reviews, see Rude 1996; Ahmad and Sutton 2000; Quamme and de Rouffignac 2000.) The daily requirement for magnesium is estimated to be 8–16  mmol (16–32 mEq) (Jones et  al., 1967).

Magnesium is widely distributed in food, the main sources being meat, green vegetables, and grains. Magnesium is believed to be absorbed equally in jejunum and ileum through both active transport and facilitated diffusion (Brannan et al., 1976). The fraction of dietary magnesium absorbed is poorly defined (20–60%) and the role of calcium (Ca2+) and vitamin D metabolites in its regulation is disputed. Substances such as phosphate and cellulose phosphate, which chelate magnesium, decrease its intestinal absorption. Under normal circumstances, only a small amount of endogenous magnesium is secreted across the intestinal epithelium. In the steady state, the net amount of magnesium absorbed by the gastrointestinal tract is eliminated in the urine. The concentration of magnesium in the extracellular fluid is determined by the kidney, and the contribution of bone and cellular pool to the maintenance of extracellular magnesium concentration in states of deficiency is unclear.

Renal handling of magnesium Renal magnesium handling is a filtration-reabsorption process. Seventy per cent of blood magnesium is in the free ionized form and an additional 10% is complexed to low-molecular-weight anions such as citrate, phosphate, and bicarbonate (Le Grimellec et al., 1975). The remainder is bound to plasma proteins, mainly albumin. Consequently, about 70–80% of blood magnesium is ultrafilterable.

Proximal tubule Of the magnesium filtered load, 10–15% is reabsorbed in the proximal convoluted tubule (de Rouffignac et al., 1973, 1991; Le Grimellec et al., 1973; Quamme et al., 1978). Proximal magnesium reabsorption is therefore substantially less than sodium or calcium reabsorption. The available data consistently indicate a relatively low permeability of the proximal tubule to magnesium: the permeability to magnesium has been calculated to be 1.1 × 10−5 cm/s in this segment. Of note, magnesium reabsorption in the proximal tubule of very young rats is proportionally much higher than in adult rats, reaching 70% of the filtered load, a value close to that of sodium and calcium reabsorption (Lelievre-Pegorier et al., 1983); the molecular basis of this difference is unclear. From the available data, it seems that proximal reabsorption of magnesium is unaffected by extracellular fluid volume expansion (Poujeol et al., 1976) or peptide hormones such as parathyroid hormone, calcitonin, or glucagon (de Rouffignac, 1990).

Loop of Henle Micropuncture experiments indicate that a large part (60–70%) of filtered magnesium is reabsorbed in the loop of Henle

244

Section 2  

fluid, electrolyte, and renal tubular disorders Magnesium intake (8–16 mmol/day in adults) EC fluid 1% Bone

Cells

Magnesium excretion (kidney = 3–6 mmol/day; stool= 5–10 mmol/day)

Fig. 27.1  Distribution of magnesium in the human body (normal).

(Fig. 27.2) (Brunette et al., 1974; Le Grimellec et al., 1973, 1974b; Bailly et al., 1984; Wong et al., 1986a; de Rouffignac et al., 1991). A low but significant fraction of filtered magnesium can be reabsorbed together with water in the descending limb in the concentrating kidney. However, evidence indicates that the bulk of reabsorption in the loop of Henle occurs in the thick ascending

Na+,2 Cl– NKCC2 K+ ROMK

ClC-Kb

Mg, Ca Na, NH4

+

–

Fig. 27.2  Model of magnesium transport in the thick ascending limb of the loop of Henle (TALH). The epithelial cells of the cortical part of the TALH reabsorb NaCl via the apical cotransporter NKCC2. Most of the potassium that enters the cell recycles back to the lumen via the apical channel ROMK, thereby hyperpolarizing the apical membrane, while most of the chloride leaves the cell across the chloride channel CLCKB in the basolateral membrane, resulting in a depolarization of the membrane. The difference in voltage of the two membranes accounts for the lumen positive transepithelial potential difference, the driving force for the paracellular diffusion of magnesium. Alterations in claudin-16 and claudin-19 can cause a severe decrease in the paracellular pathway permeability, but the full molecular basis for the permeability to Mg of the paracellular pathway is remains elusive.

limb (TALH). Magnesium is reabsorbed in the cortical but not in the medullary part of the TALH of the rat and the mouse kidney (Wittner et al., 1988, 1993, 1996, 1997; Bailly et al., 1990; Di Stefano et al., 1989, 1990, 1992, 1993; Mandon et al., 1993; Wittner et al., 1991), whereas it is reabsorbed in both segments of the rabbit kidney. Theoretically, magnesium can cross the epithelium through a transcellular or a paracellular pathway, or both. The bulk of evidence indicates that the paracellular pathway is predominant, if not the exclusive, pathway for magnesium transport in the cortical TALH. Sodium chloride absorption in the TALH generates a lumen-positive transepithelial voltage (see Chapter 21), which is the driving force for magnesium reabsorption: magnesium flux is linearly related to the transepithelial voltage in this segment (Shareghi and Agus, 1982; Hebert and Andreoli 1984, 1986; Greger, 1985; Di Stefano et al., 1988). Accordingly, when the voltage is reduced to zero by the presence of furosemide in the lumen, transepithelial magnesium transport is not significantly different from zero (Shareghi and Agus, 1982). The reason why the medullary TALH, which is also characterized by a lumen-positive transepithelial voltage, does not transport a significant amount of magnesium is not obvious. Although a lumen-positive transepithelial voltage is required for magnesium to be reabsorbed in the cortical TAL, this condition is not sufficient and the epithelium must also express a significant permeability to magnesium. Two related proteins clearly play a significant role in the process of magnesium reabsorption. Both claudin-16 and claudin-19 are expressed in the tight junction multiprotein complex. That claudin-16 may be a paracellular Mg2+ pore was originally suggested, based on the finding that mutations in CLDN16, the gene encoding claudin-16, were responsible for a severe impairment in Mg2+ reabsorption (Blanchard et  al., 2001, Simon et  al., 1999). When claudin-19 was found more recently to be co-expressed with claudin-16 in the tight junction, and mutations in CLDN19 gene found to be responsible for a similar disturbance in Mg2+ reabsorption, the hypothesis was extended to claudin-19 (Konrad et  al., 2006). However, an alternative hypothesis has been proposed, based on the finding that in vitro transfection of claudin-16 in epithelial cell lines provokes only a small increase in transepithelial magnesium permeability (Hou et al., 2005). The hypothesis is that the mutated claudin-16 has an indirect effect on magnesium reabsorption by decreasing the paracellular permeability to Na+, thereby reducing the transepithelial voltage and driving force for magnesium reabsorption (Himmerkus et al., 2008). The mechanism by which claudin-19 affects magnesium reabsorption is also unclear (Angelow et  al., 2007; Hou et  al., 2008). However, the effects of claudin-16 and of claudin-19, when co-expressed in LLC-PK1 cells, has been found to be additive (Hou et al., 2008): claudin-16 increases the permeability to Na+, whereas claudin-19 decreases the permeability to Cl−, thus increasing the permeability ratio, a prerequisite for the large diffusion potential in the TALH. Recent data suggest that a third claudin, claudin-14, may interact with claudin-16 in the mouse TALH and decrease the cation selectivity of the claudin-16-claudin-19 heteromeric channel (Gong et al., 2012). It is likely that one reason why the function(s) of the various claudins expressed in the tight junction of the TALH remain(s) elusive is that most of the experiments have not yet been performed on native tissue.

chapter 27 

Distal convoluted tubule The distal tubule reabsorbs 5–10% of filtered magnesium (de Rouffignac and Quamme, 1994), which is > 50% of the load delivered to this segment. It is very likely that this transport occurs, at least in part, in the distal convoluted tubule (DCT) Fig. 27.3 (de Rouffignac and Quamme, 1994). Because of the characteristics of the distal tubule, magnesium transport is transcellular and active (Quamme, 1997). According to the current model of transcellular Mg transport in the distal tubule, free magnesium enters the cell across the apical membrane through the apical channel TRPM6, a member of the melastatin-like subfamily of the transient receptor protein (TRP) channel family (Voets et al., 2004). Because the luminal and cytosolic concentrations of free magnesium are of the same order of magnitude, the driving force for apical magnesium entry is likely to be the cytosol-negative membrane voltage. For this purpose, the shaker-related voltage-gated K+ channel Kv1.1, which localizes with TRPM6 along the apical membrane of the DCT cell, is believed to play an important role in maintaining the polarity of the apical membrane (Glaudemans et al., 2009). The exit out of the cell across the basolateral membrane is necessarily active, since it operates uphill against the electrochemical gradient. However, the molecular identity of the transporters involved in this process remains unknown, although CNNM2, a basolateral protein expressed in the DCT, has been proposed to be one of these transporters (Stuiver et al., 2011). The reabsorption of NaCl by the DCT cell seems to be necessary to sustain magnesium reabsorption, since conditions characterized Cl–Mg++

3 Na+

K+

Ca++ Ca++ 3 Na+

Na+ 2 K+ HNF-1β

Kv1.1 TRPM6

K+

ATP ATP

Mg++

KCNJ10

ClC-Kb

ATP

2 K+

Na+ CaBP

NCC

Na+ Cl–

TRPV5

Ca++

Fig. 27.3  Model of magnesium transport in the distal convoluted tubule (DCT). Magnesium absorption in the DCT occurs exclusively through the transcellular pathway. Magnesium enters the cell through the apical channel TRPM6; however, since the luminal and cytosolic concentrations of free Mg are similar, Mg absorption is not sustained by a concentration gradient. The apical potassium channel Kv1.1 is able to hyperpolarize the apical membrane and provide the driving force for magnesium diffusion into the cell. The mechanism(s) of basolateral magnesium exit out the cell is still unsettled; Mg is believed to leave the cell via a Mg/Na exchanger or a Mg-ATPase. Mg absorption is sustained by NaCl absorption, for which the apical electroneutral Na-Cl cotransporter NCC and the basolateral Na,K-ATPase and the potassium channel KCNJ10 are mandatory (see text for details).

magnesium homeostasis

by a defect in NaCl transport in the DCT are commonly associated with renal loss of magnesium (see Chapter 40). In this process, the ATP-sensitive inward rectifier potassium channel 10 (Kir4.1), encoded by the KCNJ10 gene, may allow potassium ions to recycle across the basolateral membrane and is likely to be important in sustaining basolateral sodium pump activity. Comparisons of fluid delivered at the end of the distal tubule and final urine suggest that little quantitative transport of magnesium takes place beyond the DCT and connecting tubule (CNT).

Determinants of renal tubular magnesium transport Hormones No hormone has been identified as the regulator of blood magnesium concentration, despite the fact that many hormones have been shown to affect the renal magnesium transport, which is the main determinant of plasma magnesium concentration. Parathyroid hormone (PTH) increases magnesium absorption both in the loop of Henle and in the DCT (Bailly et al., 1984). In the cortical TALH of the mouse, PTH increases the NaCl absorption, and subsequently the transepithelial voltage, but also the paracellular permeability for magnesium (Wittner et al., 1993), suggesting direct hormonal control of the function of proteins expressed in the tight junction. PTH also stimulates magnesium absorption in the distal convoluted tubule (Bailly et al., 1985, Harris et al., 1979), but the molecular mechanisms remain uncertain; in particular, TRPM6 gene expression in the kidney is not affected by PTH (Groenestege et al., 2006). As with PTH, calcitonin increases magnesium reabsorption in the cortical TALH and DCT. These effects are reproduced by arginine vasopressin, glucagon, and ß-adrenergic agonists, although the physiological role of these three hormones in magnesium homeostasis is unclear. Insulin also increases magnesium absorption, at least in the loop of Henle. Finally, the epidermal growth factor (EGF) was recently identified as a hormone directly regulating the activity and the trafficking of the TRPM6 channel in DCT cells (Groenestege et al., 2007; Thebault et  al., 2009). Similarly, the transcription factor HNF1ß also seems to be important for magnesium reabsorption in the distal nephron, possibly because it stimulates the expression of the FXYD2 protein, the γ-subunit of Na+,K+-ATPase (Adalat et al., 2009). Prostaglandin E2 (PGE2), the major arachidonate metabolite synthesized in mammalian kidney, inhibits magnesium absorption in the renal tubule and increases urinary magnesium excretion (Schneider et  al., 1973; Roman et  al., 1984). The change in urinary magnesium excretion being paralleled by similar changes in sodium excretion, the hypothesis has been made that PGE 2 mainly acts in the TALH, primarily by decreasing NaCl absorption and thereby the driving force for magnesium reabsorption (Bailly, 1998). Because PGE2 receptors are also expressed in the DCT cells, PGE2 could act in this segment, but its effect has not been assessed, even though PGE2 increases magnesium uptake in immortalized mouse DCT cells (Dai et al., 1998). The mineralocorticoid receptor is present in DCT cells and chronic aldosterone administration results in renal magnesium wasting (Massry et al., 1967; Massry and Coburn, 1973). This effect has been explained by extracellular volume expansion induced by

245

246

Section 2  

fluid, electrolyte, and renal tubular disorders

the hormone and an attendant decrease in magnesium reabsorption. On the other hand, mineralocorticoids have been proposed to enhance the peptide hormone response of DCT cells and to indirectly increase magnesium absorption in this segment. In contrast to calcium, little information exists on the effect of calcitriol, the most active metabolite of vitamin D, on renal tubular magnesium transport, despite the demonstration of expression of the vitamin D receptor in the distal nephron. TRPM6 gene expression is not affected by 1,25-dihydroxyvitamin D (Groenestege et al., 2006). In contrast, it is increased in the presence of 17ßestradiol (Groenestege et al., 2006).

Non-hormonal factors The factors influencing magnesium transport in the proximal tubule are poorly defined. As already mentioned, the transport of magnesium in the proximal tubule is not altered by inhibition of NaCl and water reabsorption (Poujeol et  al., 1976). However, many factors have been shown to affect magnesium reabsorption in the loop of Henle. Magnesium absorption increases when magnesium concentration increases in the lumen (Quamme and Dirks, 1980; Wong et  al., 1983)  and decreases when peritubular magnesium concentration increases; an effect that is most likely to be mediated by activation of the Ca2+/polycation-sensing receptor (CaSR) expressed in the basolateral membrane of the TALH cells. Hypercalcaemia decreases magnesium absorption in the loop of Henle (Le Grimellec et al., 1974a; Quamme, 1982). The effects of peritubular magnesium might explain the apparent maximal transport, or Tm, for magnesium in the kidney, and the decrease in urinary magnesium excretion that rapidly occurs in cases of magnesium depletion. It is noteworthy that a common variant (G allele of rs17251221) in the CaSR gene is associated with a higher serum magnesium concentration (O’Seaghdha et al., 2010); but whether it is associated with an increase in renal tubular magnesium absorption in unknown. The absorption of magnesium in the DCT also increases in response to magnesium depletion. Systemic metabolic acidosis is associated with renal magnesium wasting (Martin and Jones, 1961; Lennon and Piering, 1970; Houillier et al., 1996); by contrast, bicarbonate infusion leads to a decrease in urinary magnesium excretion (Wong et al., 1986b). The acid–base status probably affects magnesium absorption both in the TALH and in the distal tubule (Shapiro et al., 1987; Houillier et al., 1996). In the latter segment, the luminal pH probably affects apical magnesium uptake by the magnesium channel TRPM6 (Dai et al., 1997).

Age and gender Magnesium absorption in the TALH is higher in adult than young animals, whereas no such change has been described for NaCl transport. No difference has been observed between males and females (Wittner et al., 1997). Whether such an age-related change in transport exists in humans is unknown.

Adaptation to a low-magnesium diet A decrease in the magnesium content of the diet is quickly followed by a fall in urinary magnesium excretion, without an initial change in blood magnesium concentration, indicating of an increase in renal tubular magnesium reabsorption (Shafik and Quamme, 1989). Micropuncture studies have shown that the adaptive increase

in magnesium absorption takes place in the loop of Henle (Shafik and Quamme, 1989). Further studies have established that dietary restriction in magnesium is associated with increased magnesium (and calcium, but not sodium or chloride) reabsorption across the epithelium of the cortical TALH (Wittner et al., 2000). This is consistent with an adaptive increase in the paracellular pathway permeability to magnesium, but the molecular mechanisms underlying this change remain unknown. Similarly, dietary magnesium restriction is associated with an increase in magnesium reabsorption in the DCT (Quamme et al., 1980, Shafik and Quamme, 1989). Consistently, expression of the TRPM6 gene and protein level increase during magnesium restriction (Groenestege et al., 2006), allowing the appropriate decrease in urinary magnesium excretion.

References Adalat, S., Woolf, A. S., Johnstone, K. A., et al. (2009). HNF1B mutations associate with hypomagnesemia and renal magnesium wasting. J Am Soc Nephrol, 20, 1123–31. Ahmad, A., and Sutton, R. A. (2000). Disorders of magnesium metabolism. In D. W. Seldin and G. Giebisch (eds.) The Kidney: Physiology and Pathophysiology, pp. 1731–48. Philadelphia, PA: Lippincott Williams & Wilkins. Angelow, S., El-Husseini, R., Kanzawa, S. A., et al. (2007). Renal localization and function of the tight junction protein, claudin-19. Am J Physiol Renal Physiol, 293, F166–77. Bailly, C. (1998). Transducing pathways involved in the control of NaCl reabsorption in the thick ascending limb of Henle’s loop. Kidney Int Suppl, 65, S29–35. Bailly, C., Roinel, N., and Amiel, C. (1984). PTH-like glucagon stimulation of Ca and Mg reabsorption in Henle’s loop of the rat. Am J Physiol, 246, F205–12. Bailly, C., Roinel, N., and Amiel, C. (1985). Stimulation by glucagon and PTH of Ca and Mg reabsorption in the superficial distal tubule of the rat kidney. Pflugers Arch, 403, 28–34. Bailly, C., Imbert-Teboul, M., Roinel, N., et al. (1990). Isoproterenol increases Ca, Mg, and NaCl reabsorption in mouse thick ascending limb. Am J Physiol, 258, F1224–31. Blanchard, A., Jeunemaitre, X., Coudol, P., et al. (2001). Paracellin-1 is critical for magnesium and calcium reabsorption in the human thick ascending limb of Henle. Kidney Int, 59, 2206–15. Brannan, P., Vergne-Marini, P., and Pak, C. (1976). Magnesium absorption in the human small intestine. J Clin Invest, 57, 1412–7. Brunette, M. G., Vigneault, N., and Carriere, S. (1974). Micropuncture study of magnesium transport along the nephron in the young rat. Am J Physiol, 227, 891–6. Dai, L. J., Friedman, P. A., and Quamme, G. A. (1997). Acid-base changes alter Mg2+ uptake in mouse distal convoluted tubule cells. Am J Physiol, 272, F759–66. Dai, L. J., Bapty, B., Ritchie, G., et al. (1998). PGE2 stimulates Mg2+ uptake in mouse distal convoluted tubule cells. Am J Physiol, 275, F833–9. De Rouffignac, C. (1990). The urinary concentrating mechanisms. In R. Kinne (ed.) Urinary Concentrating Mechanisms, pp. 31–102. Basel: Karger. De Rouffignac, C. and Quamme, G. (1994). Renal magnesium handling and its hormonal control. Physiol Rev, 74, 305–22. De Rouffignac, C., Elalouf, J. M., and Roinel, N. (1991). Glucagon inhibits water and NaCl transports in the proximal convoluted tubule of the rat kidney. Pflugers Arch, 419, 472–7. De Rouffignac, C., Morel, F., Moss, N., et al. (1973). Micropuncture study of water and electrolyte movements along the loop of Henle in psammomys with special reference to magnesium, calcium and phosphorus. Pflugers Arch, 344, 309–26. De Rouffignac, C., Di Stefano, A., Wittner, M., et al. (1991). Consequences of differential effects of ADH and other peptide

chapter 27 

hormones on thick ascending limb of mammalian kidney. Am J Physiol, 260, R1023–35. Di Stefano, A., Wittner, M., Gebler, B., et al. (1988). Increased Ca++ or mg++ concentration reduces relative tight-junction permeability to Na+ in the cortical thick ascending limb of Henle’s loop of rabbit kidney. Ren Physiol Biochem, 11, 70–9. Di Stefano, A., Greger, R., de Rouffignac, C., et al. (1992). Active NaCl transport in the cortical thick ascending limb of Henle’s loop of the mouse does not require the presence of bicarbonate. Pflugers Arch, 420, 290–6. Di Stefano, A., Roinel, N., de Rouffignac, C., et al. (1993). Transepithelial Ca2+ and Mg2+ transport in the cortical thick ascending limb of Henle’s loop of the mouse is a voltage-dependent process. Ren Physiol Biochem, 16, 157–66. Di Stefano, A., Wittner, M., Nitschke, R., et al. (1990). Effects of parathyroid hormone and calcitonin on Na+, Cl-, K+, Mg2+ and Ca2+ transport in cortical and medullary thick ascending limbs of mouse kidney. Pflugers Arch, 417, 161–7. Di Stefano, A., Wittner, M., Nitschke, R., et al. (1989). Effects of glucagon on Na+, Cl-, K+, Mg2+ and Ca2+ transports in cortical and medullary thick ascending limbs of mouse kidney. Pflugers Arch, 414, 640–6. Glaudemans, B., van der Wijst, J., Scola, R. H., et al. (2009). A missense mutation in the Kv1.1 voltage-gated potassium channel-encoding gene KCNA1 is linked to human autosomal dominant hypomagnesemia. J Clin Invest, 119, 936–42. Gong, Y., Renigunta, V., Himmerkus, N., et al. (2012). Claudin-14 regulates renal Ca++ transport in response to CaSR signalling via a novel microRNA pathway. EMBO J, 31, 1999–2012. Greger, R. (1985). Ion transport mechanisms in thick ascending limb of Henle’s loop of mammalian nephron. Physiol Rev, 65, 760–97. Groenestege, W. M., Hoenderop, J. G., van den Heuvel, L., et al. (2006). The epithelial Mg2+ channel transient receptor potential melastatin 6 is regulated by dietary Mg2+ content and estrogens. J Am Soc Nephrol, 17, 1035–43. Groenestege, W. M., Thebault, S., van der Wijst, J., et al. (2007). Impaired basolateral sorting of pro-EGF causes isolated recessive renal hypomagnesemia. J Clin Invest, 117, 2260–7. Gunther, T. (1981). Biochemistry and pathobiochemistry of magnesium. Artery, 9, 167–81. Harris, C. A., Burnatowska, M. A., Seely, J. F., et al. (1979). Effects of parathyroid hormone on electrolyte transport in the hamster nephron. Am J Physiol, 236, F342–8. Hebert, S. C. and Andreoli, T. E. (1984). Control of NaCl transport in the thick ascending limb. Am J Physiol, 246, F745–56. Hebert, S. C. and Andreoli, T. E. (1986). Ionic conductance pathways in the mouse medullary thick ascending limb of Henle. The paracellular pathway and electrogenic Cl- absorption. J Gen Physiol, 87, 567–90. Himmerkus, N., Shan, Q., Goerke, B., et al. (2008). Salt and acid-base metabolism in claudin-16 knockdown mice: impact for the pathophysiology of FHHNC patients. Am J Physiol Renal Physiol, 295, F1641–7. Hou, J., Paul, D. L., and Goodenough, D. A. (2005). Paracellin-1 and the modulation of ion selectivity of tight junctions. J Cell Sci, 118, 5109–18. Hou, J., Renigunta, A., Konrad, M., et al. (2008). Claudin-16 and claudin-19 interact and form a cation-selective tight junction complex. J Clin Invest, 118, 619–28. Houillier, P., Normand, M., Froissart, M., et al. (1996). Calciuric response to an acute acid load in healthy subjects and hypercalciuric calcium stone formers. Kidney Int, 50, 987–97. Jones, J., Manalo, R., and Flink, E. B. (1967). Magnesium requirements in adults. Am J Clin Nutr, 20, 632–5. Konrad, M., Schaller, A., Seelow, D., et al. (2006). Mutations in the tight-junction gene claudin 19 (CLDN19) are associated with renal magnesium wasting, renal failure, and severe ocular involvement. Am J Hum Genet, 79, 949–57. Le Grimellec, C., Roinel, N., and Morel, F. (1973). Simultaneous Mg, Ca, P, K, Na and Cl analysis in rat tubular fluid. II. During acute Mg plasma loading. Pflugers Arch, 340, 197–210.

magnesium homeostasis

Le Grimellec, C., Roinel, N., and Morel, F. (1974a). Simultaneous Mg, Ca, P, K, Na and Cl analysis in rat tubular fluid. 3. During acute Ca plasma loading. Pflugers Arch, 346, 171–88. Le Grimellec, C., Roinel, N., and Morel, F. (1974b). Simultaneous Mg, Ca, P, K and Cl analysis in rat tubular fluid. IV. During acute phosphate plasma loading. Pflugers Arch, 346, 189–204. Le Grimellec, C., Poujeol, P., and de Rouffignac, C. (1975). 3H-inulin and electrolyte concentrations in Bowman’s capsule in rat kidney. Comparison with artificial ultrafiltration. Pflugers Arch, 354, 117–31. Lelievre-Pegorier, M., Merlet-Benichou, C., Roinel, N., et al. (1983). Developmental pattern of water and electrolyte transport in rat superficial nephrons. Am J Physiol, 245, F15–21. Lennon, E. J. and Piering, W. F. (1970). A comparison of the effects of glucose ingestion and NH4Cl acidosis on urinary calcium and magnesium excretion in man. J Clin Invest, 49, 1458–65. Mandon, B., Siga, E., Chabardes, D., et al. (1993). Insulin stimulates Na+, Cl−, Ca2+, and Mg2+ transports in TAL of mouse nephron: cross-potentiation with AVP. Am J Physiol, 265, F361–9. Martin, H. E. and Jones, R. (1961). The effect of ammonium chloride and sodium bicarbonate on the urinary excretion of magnesium, calcium, and phosphate. Am Heart J, 62, 206–10. Massry, S. G. and Coburn, J. W. (1973). The hormonal and non-hormonal control of renal excretion of calcium and magnesium. Nephron, 10, 66–112. Massry, S. G., Coburn, J. W., Chapman, L. W., et al. (1967). The acute effect of adrenal steroids on the interrelationship between the renal excretion of sodium, calcium, and magnesium. J Lab Clin Med, 70, 563–70. O’Seaghdha, C. M., Yang, Q., Glazer, N. L., et al. (2010). Common variants in the calcium-sensing receptor gene are associated with total serum calcium levels. Hum Mol Genet, 19, 4296–303. Poujeol, P., Chabardes, D., Roinel, N., et al. (1976). Influence of extracellular fluid volume expansion on magnesium, calcium and phosphate handling along the rat nephron. Pflugers Arch, 365, 203–11. Quamme, G. A. (1982). Effect of hypercalcemia on renal tubular handling of calcium and magnesium. Can J Physiol Pharmacol, 60, 1275–8050. Quamme, G. A. (1997). Renal magnesium handling: new insights in understanding old problems. Kidney Int, 52, 1180–95. Quamme, G. A. and Dirks, J. H. (1980). Intraluminal and contraluminal magnesium on magnesium and calcium transfer in the rat nephron. Am J Physiol, 238, F187–98. Quamme, G. A. and de Rouffignac, C. (2000). Renal magnesium handling. In D. W. Seldin, and G. Giebisch (eds.) The Kidney: Physiology and Pathophysiology, pp. 1711–29. Philadelphia, PA; Lippincott Williams & Wilkins. Quamme, G. A., Carney, S. L., Wong, N. L., et al. (1980). Effect of parathyroid hormone on renal calcium and magnesium reabsorption in magnesium deficient rats. Pflugers Arch, 386, 59–65. Quamme, G. A., Wong, N. L., Dirks, J. H., et al. (1978). Magnesium handling in the dog kidney: a micropuncture study. Pflugers Arch, 377, 95–9. Roman, R. J., Skelton, M., and Lechene, C. (1984). Prostaglandin-vasopressin interactions on the renal handling of calcium and magnesium. J Pharmacol Exp Ther, 230, 295–301. Rude, R. K. (1996) Magnesium disorders. In J. P. Kokko, and R. L. Tannen (eds.) Fluids and Electrolytes, pp. 421–48. Philadelphia, PA: W. B. Saunders. Schneider, E. G., Strandhoy, J. W., Willis, L. R., et al. (1973). Relationship between proximal sodium reabsorption and excretion of calcium, magnesium and phosphate. Kidney Int, 4, 369–76. Shafik, I. and Quamme, G. (1989). Early adaptation of renal magnesium reabsorption in response to magnesium restriction. Am J Physiol, 257, F974–7. Shapiro, R. J., Yong, C. K., and Quamme, G. A. (1987). Influence of chronic dietary acid on renal tubular handling of magnesium. Pflugers Arch, 409, 492–8. Shareghi, G. R. and Agus, Z. S. (1982). Magnesium transport in the cortical thick ascending limb of Henle’s loop of the rabbit. J Clin Invest, 69, 759–69.

247

248

Section 2  

fluid, electrolyte, and renal tubular disorders

Simon, D. B., Lu, Y., Choate, K. A., et al. (1999). Paracellin-1, a renal tight junction protein required for paracellular Mg2+ resorption. Science, 285, 103–6. Stuiver, M., Lainez, S., Will, C., et al. (2011). CNNM2, encoding a basolateral protein required for renal Mg2+ handling, is mutated in dominant hypomagnesemia. Am J Hum Genet, 88, 333–43. Thebault, S., Alexander, R. T., Tiel Groenestege, W. M., et al. (2009). EGF increases TRPM6 activity and surface expression. J Am Soc Nephrol, 20, 78–85. Voets, T., Nilius, B., Hoefs, S., et al. (2004). TRPM6 forms the Mg2+ influx channel involved in intestinal and renal Mg2+ absorption. J Biol Chem, 279, 19–25. Wittner, M., Desfleurs, E., Pajaud, S., et al. (1996). Calcium and magnesium: low passive permeability and tubular secretion in the mouse medullary thick ascending limb of Henle’s loop (MTAL). J Membr Biol, 153, 27–35. Wittner, M., Desfleurs, E., Pajaud, S., et al. (1997). Calcium and magnesium transport in the cortical thick ascending limb of Henle’s loop: influence of age and gender. Pflugers Arch, 434, 451–6. Wittner, M., di Stefano, A., Mandon, B., et al. (1991). Stimulation of NaCl reabsorption by antidiuretic hormone in the cortical thick ascending limb of Henle’s loop of the mouse. Pflugers Arch, 419, 212–14.

Wittner, M., di Stefano, A., Wangemann, P., et al. (1988). Differential effects of ADH on sodium, chloride, potassium, calcium and magnesium transport in cortical and medullary thick ascending limbs of mouse nephron. Pflugers Arch, 412, 516–23. Wittner, M., Jounier, S., Deschenes, G., et al. (2000). Cellular adaptation of the mouse cortical thick ascending limb of Henle’s loop (CTAL) to dietary magnesium restriction: enhanced transepithelial Mg2+ and Ca2+ transport. Pflugers Arch, 439, 765–71. Wittner, M., Mandon, B., Roinel, N., et al. (1993). Hormonal stimulation of Ca2+ and Mg2+ transport in the cortical thick ascending limb of Henle’s loop of the mouse: evidence for a change in the paracellular pathway permeability. Pflugers Arch, 423, 387–96. Wong, N. L., Quamme, G. A., and Dirks, J. H. (1986a). Effects of acid-base disturbances on renal handling of magnesium in the dog. Clin Sci (Lond), 70, 277–84. Wong, N. L., Whiting, S. J., Mizgala, C. L., et al. (1986b). Electrolyte handling by the superficial nephron of the rabbit. Am J Physiol, 250, F590–5. Wong, N. L. M., Dirks, J. H., and Quamme, G. A. (1983). Tubular reabsorptive capacity for magnesium in the dog kidney. Am J Physiol, 244, F78–83.

CHAPTER 28

Approach to the patient with hyponatraemia Ewout J. Hoorn and Robert Zietse Introduction and epidemiology Hyponatraemia counts as the most common electrolyte disorder in hospitalized patients. Its precise epidemiology, however, depends on the serum sodium used to define hyponatraemia, which varies from study to study. The prevalence of hyponatraemia in hospitalized patients is 15–30% when defined as a serum sodium < 136 mmol/L (which is the lower level of normal in most laboratories) and 2–3% when defined as a serum sodium < 125 mmol/L (Hoorn et al., 2006; Upadhyay et al., 2006). A decrease in serum sodium is usually associated with a decrease in serum osmolality (hypo-osmolality), which would normally suppress vasopressin (antidiuretic hormone) and cause a water diuresis. However, in the majority of patients with hyponatraemia, vasopressin levels are either detectable or increased (Anderson et al., 1985). There may be several reasons why vasopressin is present despite hypo-osmolality (Box 28.1). Therefore, hyponatraemia should prompt the question why vasopressin is present despite hypo-osmolality and what the source is of the electrolyte-free water that was retained. Hyponatraemia can be classified according to the time over which it developed, the presence of symptoms, the tonicity, and volume status (Table 28.1). Each of these classifications has their uses and limitations, depending on the clinical context. Ultimately, the clinical setting should dictate which classification is most useful to guide management. It is important to emphasize that these classifications are not mutually exclusive. For example, a patient can have acute and symptomatic hyponatraemia that is further characterized by hypotonicity and euvolaemia. This immediately provides useful information for management, because the presence of cerebral oedema is likely (acute, hypotonic, symptomatic) and the patient therefore requires emergency treatment with hypertonic saline.

Clinical features There are three reasons why hyponatraemia is clinically important. First, hyponatraemia can be an early or even first sign of important underlying disease, such as adrenal insufficiency or lung cancer (van der Hoek et  al., 2009; Hoorn et  al., 2011a). Second, hyponatraemia can be complicated by two types of neurological disorders, including cerebral oedema and the osmotic demyelination syndrome (Fig. 28.1) (Arieff, 1986; Sterns et al., 1986). Third, hyponatraemia is invariably associated with increased morbidity and mortality rates in hospitalized patients (Wald et al., 2010), although it remains unclear whether these associations are mainly

due to the underlying disease, direct effects of hyponatraemia, or a combination of both (Chawla et al., 2011; Hoorn and Zietse, 2011). Symptoms associated with hyponatraemia in general are listed in Box 28.2. Acute hyponatraemia is usually more symptomatic than chronic hyponatraemia. Acute hyponatraemia (decrease to a serum sodium of ≤125 mmol/L in ≤ 48 hours) can cause cerebral oedema, because brain cells have insufficient time to adapt to their hypotonic environment. Severe symptoms such as seizures or coma are usually observed in acute hyponatraemia and reflect the presence of cerebral oedema. Milder symptoms such as nausea and vomiting, however, can also be the first signs of an increase in intracranial pressure due to cerebral oedema. In recent years, it has become clear that even patients with chronic hyponatraemia (present > 48 hours), when analysed more closely, also exhibit symptoms. These are usually more subtle neurocognitive or neuromotor symptoms, including gait disturbances, falls, and concentration deficits (Renneboog et al., 2006). Recently, hyponatraemia has been associated with osteoporosis and fractures, suggesting that hyponatraemia can also affect other organs besides the brain (Verbalis et al., 2010; Hoorn et al., 2011b). These indirect effects of hyponatraemia may also contribute to its association with morbidity and mortality. The osmotic demyelination syndrome is a complication of too rapid correction of chronic hyponatraemia and is therefore usually iatrogenic. Osmotic demyelination often has a ‘biphasic’ course; during the first phase patients are typically asymptomatic (2–6 days), while during the second phase patients develop neurological symptoms, including dysarthria, dysphagia, paresis, confusion, coma, and seizures. These symptoms may be irreversible and even progress to a ‘locked-in’ syndrome or death.

Investigations Because hyponatraemia is so common in hospitalized patients, milder forms of hyponatraemia (usually serum sodium >130  mmol/L) may not require additional investigations, especially if obvious explanations are present (hyperglycaemia, postoperative state) and hyponatraemia is expected to be transient. In all other situations (usually serum sodium 48 h) • Or asymptomatic or only mild symptoms Aim is to correct hyponatraemia and prevent osmotic demyelination

Start hypertonic saline (rise 1–2 mmol/L/h until symptoms abate and not exceeding 8–12 mmol/L/day)

Begin diagnostic evaluation and continue or start therapy

Fig. 28.2  Algorithm for hyponatraemia.

question whether pseudohyponatraemia is present. Another means to identify pseudohyponatraemia is to perform the measurement in an undiluted sample (direct potentiometry), for example using a blood gas analyser. If a low serum sodium concentration is found Box 28.3  Essential and optional investigations in hyponatraemia

All patients with hyponatraemia ◆ Serum glucose ◆ Serum potassium ◆ Serum creatinine ◆ Serum osmolality ◆ Urine sodium ◆ Urine osmolality.

Useful in certain settings ◆ Serum urea ◆ Serum uric acid ◆ Serum cortisol ◆ Serum thyroid stimulating hormone ◆ Fractional excretion of sodium, uric acid, and urea ◆ Urine chloride ◆ Urine potassium.

with indirect potentiometry, but a normal serum sodium with direct potentiometry, this is again suggestive of pseudohyponatraemia and should warrant a search for the underlying cause.

Hyperglycaemia-induced hyponatraemia It is always important to analyse serum glucose in a patient with hyponatraemia. Glucose is an effective osmole and will therefore attract water from the intracellular compartment. Hyperglycaemia can, therefore, result in dilutional hyponatraemia. In fact, this relationship can sometimes be predicted with formulae, with serum sodium decreasing approximately 1.6–2.4  mmol/L for every 5.5 mmol/L increase in glycaemia (Hillier et al., 1999). However, several factors will affect this relationship, including oral intake, intravenous fluids, and ongoing osmotic diuresis. Still, in cases of severe hyperglycaemia, the degree of hyponatraemia can sometimes be completely attributed to hyperglycaemia and does not require therapy other than for hyperglycaemia. It is also important to differentiate hyperglycaemia-induced hyponatraemia from pseudohyponatraemia. In hyperglycaemia-induced hyponatraemia, the serum sodium is truly decreased, although the water shift is opposite to hypotonic forms of hyponatraemia. Because glucose is an effective osmole, it will contribute to serum osmolality and so hyperglycaemia-induced hyponatraemia is usually a hypertonic form of hyponatraemia. Because this will attract water from cells, it does not pose a risk of cerebral oedema. Nevertheless, it is important to regularly calculate the effective osmolality during the correction of hyperglycaemia, because a rapid decrease in effective osmolality can still cause cerebral oedema, for example,

251

252

Section 2  

fluid, electrolyte, and renal tubular disorders

if serum glucose decreases more than serum sodium rises (Hoorn et al., 2007).

Assessment of the extracellular fluid volume The assessment of the extracellular fluid volume should always be performed in patients with hyponatraemia, because it offers a simple approach to detect cases with obvious hyper- or hypovolaemia. However, a caveat is that the assessment of the extracellular fluid volume in less clear-cut cases has a low sensitivity and specificity (Chung et al., 1987). This is especially true for the differentiation between hypovolaemia and euvolaemia. Therefore, the assessment of the extracellular fluid volume should not be the first and main determinant in the differentiation of hyponatraemia (Hoorn et al., 2005). This approach differs from what many of the ‘traditional’ algorithms for hyponatraemia propose, which still use the assessment of volume status as their first determinant. A systematic review on hypovolaemia concluded that more objective parameters such as capillary refill time, postural hypotension, and postural tachycardia have a higher sensitivity for diagnosing hypovolaemia, especially when these indices are combined (McGee et al., 1999).

Urine sodium and urine osmolality The urine sodium and osmolality are very useful parameters in the differentiation of hyponatraemia, because they provide information on the underlying pathophysiology (Kamel et al., 1990). The urine sodium concentration can be considered a measure of the renin–angiotensin–aldosterone system, while the urine osmolality is a measure of vasopressin activity. A low urine sodium concentration usually indicates renal sodium retention by angiotensin II and aldosterone. Similarly, a high urine osmolality nearly always indicates active water reabsorption by the kidneys due to vasopressin or a vasopressin-like effect. These two concepts can be used to analyse the underlying cause of hyponatraemia (Table 28.2). For example, low urine sodium in combination with high urine osmolality suggests that both aldosterone and vasopressin are acting on the renal Table 28.2  Clinical settings of low or high urine sodium and osmolality Urine sodiuma

Urine osmolalitya

Clinical setting

Low

High

Hypovolaemia with extrarenal sodium loss Low effective arterial blood volume (heart failure, liver cirrhosis, nephrotic syndrome) SIADH with low dietary salt intake Recent discontinuation of diuretics

Low

Low

Primary polydipsia and beer potomania Low solute intake (‘tea and toast’) Water diuresis during treatment of hyponatraemia

High

High

Active use of diuretics SIADH Adrenal insufficiency Renal or cerebral salt wasting Chronic kidney disease

a Although depending on other influencing factors, the threshold between low and high is approximately 30–40 mmol/L for urine sodium and between 100 and 280 for urine osmolality.

tubules. Clinical settings in which this can occur include true hypovolaemia or a low effective arterial blood volume (e.g. during heart failure or liver cirrhosis). The combination of finding a high urine sodium and a high urine osmolality is more challenging. This can reflect renal sodium loss with secondary vasopressin release due to hypovolaemia (e.g. with diuretic use). Conversely, this combination may be present in the syndrome of inappropriate antidiuretic hormone secretion (SIADH). During SIADH, urine osmolality is high due to the ‘inappropriate’ secretion of vasopressin, whereas urine sodium is normally excreted and therefore reflects dietary intake (and perhaps some degree of volume expansion). It is important to interpret urine sodium and osmolality in the context of the clinical setting. For example, urine sodium can also be low if the patient consumes a low sodium diet or when there is a water diuresis (when urine osmolality should also be low). The use of diuretics will also affect the urine sodium excretion, increasing it during active use, but decreasing it after recent discontinuation.

Urea and uric acid in hyponatraemia The value of urea and uric acid in the differentiation of hyponatraemia is mainly based on their correlation with volume status. Expansion of the extracellular fluid volume will inhibit the renal tubular reabsorption of urea and uric acid. Therefore, a low serum concentration and high fractional excretion of uric acid and urea can be observed during, for example, hyponatraemia secondary to polydipsia or SIADH. Besides extracellular fluid volume expansion, the vasopressin V1 receptor may also contribute to this effect, because the induction of hyponatraemia with desmopressin (which only stimulates the V2 receptor) resulted in a significantly smaller fall in serum uric acid (Decaux et al., 1996). Recently, a fractional uric acid excretion > 13% was identified as the most sensitive parameter to diagnose SIADH in patients who were also using diuretics (Fenske et al., 2008).

Functional tests in hyponatraemia In more complex cases of hyponatraemia, functional tests can be useful, including assessing the response to isotonic saline, a water load, or a vasopressin receptor antagonist. Assessing the response to isotonic saline can be used to differentiate hypovolaemic from euvolaemic hyponatraemia. The serum sodium concentration is expected to rise during the infusion of isotonic saline in hypovolaemic hyponatraemia, whereas no response or even a deterioration of hyponatraemia can occur in euvolaemic hyponatraemia (Steele et al., 1997). A water loading test can be used to differentiate true SIADH causing hyponatraemia from a reset osmostat, which is a subform of SIADH in which osmolality is regulated at a lower set-point (Fig. 28.3). During a water-loading test, 20 mL/kg water is administered and a 4-hour observation period is used to assess if this water is retained or normally excreted. Normal water excretion is defined as > 80% of the water load, which would suggest a reset osmostat. An important warning is that if the patient has true SIADH, part of the water load will be retained and hyponatraemia may worsen; therefore, close monitoring during the test is required. Vasopressin receptor antagonists, which have recently been introduced as a new treatment for hyponatraemia (see later), can also be applied diagnostically. Namely, a patient with true SIADH is expected to respond to a vasopressin receptor antagonist with the excretion of more dilute urine and therefore a rise in serum sodium. If this response is absent, this could suggest the

chapter 28 

is unknown. Increased fluid intake, for example because thirst is stimulated by elevated angiotensin II, is also believed to play a role. Others have proposed that the potassium depletion associated with thiazides reduces the set-point for baroreceptor-mediated vasopressin release. Finally, there also appears to be an individual predisposition to develop hyponatraemia with a thiazide, because it is reproducible during re-challenge (Friedman et al., 1989). The latter observation suggests that patients who develop thiazide-induced hyponatraemia, may have a contraindication for future use. When considering thiazide-induced hyponatraemia, it is also important to be aware of the many combination preparations that include a thiazide. These are mainly antihypertensive drugs that may also contain thiazide-like compounds, such as indapamide.

12

Plasma AVP (pg/mL)

Type C 8

Type A

4 Type B Type D 0

0

120

130

140

approach to the patient with hyponatraemia

150

Plasma sodium (mmol/L)

Fig. 28.3  Types of the syndrome of inappropriate antidiuretic hormone secretion (SIADH). Patterns of plasma levels of arginine vasopressin (AVP) compared with plasma sodium levels in patients with the syndrome of inappropriate antidiuretic hormone secretion. Type A is characterized by unregulated secretion of AVP and can occur during ectopic production of AVP by neoplasia. Type B is characterized by an elevated basal secretion of AVP, despite normal regulation by osmolality. Type C is characterized by a ‘reset osmostat,’ meaning that the serum sodium is regulated at a lower concentration due to a lower osmotic threshold for vasopressin release. Type D is characterized by undetectable AVP despite the presence of SIADH; this can be observed in nephrogenic syndrome of inappropriate antidiuresis. The shaded area represents normal values of plasma AVP. From: Robertson GL. Regulation of arginine vasopressin in the syndrome of inappropriate antidiuresis. Am J Med 2006; 119: Suppl 1: S36–42.

presence of an activating mutation of the vasopressin-2 receptor, the so-called nephrogenic syndrome of inappropriate antidiuresis (see later) (Decaux et  al., 2007). Vasopressin is rarely measured in hyponatraemia because it is a difficult assay and because its activity is reflected by the urine osmolality. Still, measurement of vasopressin is sometimes useful in difficult cases of SIADH. The development of copeptin, a glycopeptide derived from same precursor peptide as vasopressin, has shown some diagnostic value in hyponatraemia, especially for primary polydipsia and volume depletion (Fenske et al., 2009).

Aetiology and pathogenesis Diuretic-induced hyponatraemia Diuretics are a common cause of hyponatraemia. This side effect is mainly observed with thiazide diuretics and, to a lesser extent, with potassium-sparing diuretics. Loop diuretics alone rarely produce hyponatraemia and can even be used as therapy for hyponatraemia, for example, in SIADH (Sonnenblick et al., 1993). The pathogenesis of thiazide-induced hyponatraemia remains incompletely understood and several mechanisms may contribute. The original theory that renal sodium loss causes hypovolaemia with secondary vasopressin release may be true for patients who use a combination of diuretics. However, patients who are treated with thiazides only and develop hyponatraemia, gain weight (Friedman et al., 1989). This suggests thiazide-induced hyponatraemia is primarily a water-retaining disorder. Whether this is mainly due to more central release of vasopressin or is primarily a renal effect

Syndrome of inappropriate antidiuretic hormone secretion In SIADH, vasopressin secretion is considered ‘inappropriate,’ because neither hypertonicity nor hypovolaemia stimulates vasopressin release. Four patterns of abnormal vasopressin secretion are recognized (Fig. 28.3). Antidiuresis causes progressive hyponatraemia until a renal defence mechanism called ‘vasopressin escape’ is activated. During escape from antidiuresis, the vasopressin V2 receptor and the water channel aquaporin-2 are downregulated, preventing further water reabsorption (Verbalis, 2006). The causes of SIADH are generally related to malignant diseases, pulmonary diseases, central nervous system disease, drugs, and miscellaneous causes (Table 28.3). A number of commonly used drugs are associated with SIADH, including antidepressants, antiepileptics, antipsychotics, and the vasopressin analogues desmopressin and oxytocin. A number of other drugs rarely cause hyponatraemia (reviewed in Liamis et al., 2008). The essential and supplemental criteria for the diagnosis of SIADH remain essentially the same as those originally proposed by Bartter and Schwartz in 1957 (Box 28.4) (Schwartz et  al., 1957). Importantly, although SIADH is common, it should still be regarded as a diagnosis of exclusion, in which diuretic use, pituitary, adrenal, and thyroid insufficiency must be excluded. Although a urine osmolality > 100 mOsm/kg of water is used in the definition of SIADH, it usually exceeds the serum osmolality. A number of the miscellaneous causes of SIADH merit emphasis. For example, nausea, pain, and stress are non-specific, but strong stimuli for vasopressin release, and often contribute to postoperative hyponatraemia, especially when patients are also receiving hypotonic intravenous fluids (Chung et al., 1986). A rarer cause of SIADH is the so-called nephrogenic syndrome of inappropriate antidiuresis. In this genetic disorder, there is a gain-of-function mutation of the vasopressin V2 receptor causing constitutive activation of the V2R-AQP2 cascade, resulting in increased renal water reabsorption (Feldman et al., 2005). This diagnosis should be considered in a patient with chronic SIADH of unknown origin in whom vasopressin is undetectable and who does not respond to a vasopressin receptor antagonist. Another genetic susceptibility for hyponatraemia was identified in individuals with a polymorphism in the TRPV4 gene, which encodes a calcium channel believed to be involved in ‘osmosensing’ (Tian et al., 2009). Finally, exercise-associated hyponatraemia, which is relatively common among marathon runners, is considered a form of SIADH, because besides over-hydration there is evidence for non-osmotic release of vasopressin, possibly

253

254

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 28.3  Causes of the syndrome of inappropriate antidiuretic hormone secretion (SIADH) Malignant diseases

Pulmonary disorders

Disorders of the central nervous system

Drugs

Carcinoma Lung (small cell, mesothelioma) Oropharynx Gastro-intestinal tract (stomach, duodenum, pancreas) Genitourinary tract (ureter, bladder, prostate, endometrium) Endocrine thymoma Lymphomas Sarcomas Ewing sarcoma

Infections Bacterial pneumonia Viral pneumonia Pulmonary abscess Tuberculosis Aspergillosis Asthma Cystic fibrosis Respiratory failure associated with positive pressure

Infections Drugs that stimulate release of AVP or enhance its action Encephalitis Chlorpropamide Meningitis SSRIs Brain abscess TCAs Rocky Mountain spotted fever Clofibrate AIDS Carbamazepine Malaria Vincristine Bleeding and masses Nicotine Subdural haematoma Antipsychotic drugs Subarachnoid haemorrhage Ifosfamide Cerebrovascular accident Cyclo-phosphamide Brain tumours NSAIDs Head trauma MDMA (‘ecstasy’) Hydrocephalus AVP analogues Cavernous sinus thrombosis Desmopressin Other Oxytocin Multiple sclerosis Vasopressin Guillain–Barré syndrome Shy–Drager syndrome Delirium tremens Acute intermittent porphyria

Other causes Hereditary (gain of function mutations of the vasopressin V2 receptor) Idiopathic Transient Exercise-associated hyponatraemia General anaesthesia Nausea Pain Stress

AIDS = acquired immunodeficiency syndrome; AVP = arginine vasopressin; MDMA = methylenedioxymethamphetamine; NSAIDs = non-steroidal anti-inflammatory drugs

mediated through cytokines (Almond et  al., 2005; Siegel et  al., 2007). Specific guidelines for exercise-associated hyponatraemia have been developed (Hew-Butler et al., 2008).

Endocrine causes of hyponatraemia Endocrine causes of hyponatraemia include primary and secondary adrenal insufficiency and hypothyroidism. The mechanism and presentation of hyponatraemia due to primary or secondary adrenal insufficiency differ. Secondary adrenal insufficiency is characterized by hypocortisolism. Because cortisol normally suppresses vasopressin, hypocortisolism will increase central vasopressin release. Hypocortisolism, therefore, causes hyponatraemia that resembles SIADH. In primary adrenal insufficiency (Addison disease), however, there is a deficiency of both glucocorticoids and mineralocorticoids. In this setting, hyponatraemia is not only caused by hypocortisolism, but also by hypoaldosteronism, which leads to renal sodium loss. However, additional signs, including metabolic acidosis, hyperkalaemia, hypercalcaemia, and orthostatic hypotension, typically accompany mineralocorticoid deficiency. Although hyponatraemia is usually part of a more elaborate constellation of physical and biochemical findings, it can be the only or first presentation of adrenal insufficiency (Smith et al., 2004). Therefore, physicians should have an index of suspicion for adrenal insufficiency as a cause of hyponatraemia and have a low threshold for diagnostic testing using a random cortisol or stimulation test with adrenocorticotrophic hormone (Soule, 1999). Hypothyroidism can

cause hyponatraemia, especially in patients with myxoedema (Curtis, 1956). Hyponatraemia in this context may be due to a decrease in cardiac output and glomerular filtration rate. A recent study showed that for every 10 mU/L rise in thyroid-stimulating hormone, serum sodium decreased 0.14 mmol/L (Warner et al., 2006). This suggests that in most cases, hypothyroidism has a very limited effect on serum sodium.

Cerebral salt wasting Cerebral salt wasting (CSW) is an incompletely understood disorder (Singh et al., 2002). The best evidence for its existence comes from patients with subarachnoid haemorrhage in whom polyuria with a natriuresis can be observed (Berendes et al., 1997). It is thought that, if uncorrected, the loss of water and salt causes hypovolaemia with the subsequent release of vasopressin and therefore hyponatraemia. Brain natriuretic peptide has been implicated as the cause of the massive natriuresis. The difficulty is that most of the neurological disorders associated with cerebral salt wasting can also cause other forms of hyponatraemia, including SIADH and secondary adrenal insufficiency. Furthermore, biochemically, CSW and SIADH are remarkably similar, further complicating their differentiation (Table 28.4). Studies have indicated that SIADH is, in fact, more common than CSW following subarachnoid haemorrhage and CSW may therefore be overdiagnosed (Sherlock et al., 2006). Still, its recognition is important, because treatment is opposite to SIADH, since it relies on fluid resuscitation, rather than fluid restriction.

chapter 28 

Box 28.4  Diagnostic criteria for the syndrome of inappropriate antidiuretic hormone secretion (SIADH)

Essential criteria ◆ Decreased effective serum osmolality (< 270 mOsm/kg of water) ◆ Urine osmolality > 100 mOsm/kg of water during hypotonicity ◆ Clinical euvolaemia ◆ Urinary sodium > 40 mmol/L with normal dietary salt intake ◆ Absence of adrenal, thyroid, pituitary, or renal insufficiency or diuretic use.

Supplemental criteria ◆ Serum uric acid < 0.24 mmol/L (< 4 mg/dL) ◆ Serum urea < 3.6 mmol/L (< 10 mg/dL) ◆ Failure to correct hyponatraemia after 0.9% saline infusion ◆ Fractional sodium excretion > 1%; fractional urea excretion > 55% ◆ Correction of hyponatraemia through fluid restriction ◆ Plasma vasopressin level inappropriately elevated relative to plasma osmolality ◆ Abnormal result on test of water load (< 80% excretion of 20 mL water/kg body weight over a period of 4 hours and/or failure to dilute urine osmolality to < 100 mOsm/kg of water). Data are adapted from Schwartz et al. (1957) and Janicic and Verbalis (2003).

approach to the patient with hyponatraemia

in nephrotic syndrome it is due to loss of plasma oncotic pressure. A low effective arterial blood volume will activate both the renin–angiotensin–aldosterone system and the vasopressin axis. The release of vasopressin during a low effective arterial blood volume is mediated through baroreceptors. The activation of these systems explains why urine sodium is typically low and why urine osmolality is high (Table 28.2). Hyponatraemia usually develops in more advanced stages of heart failure (New York Heart Association classes III and IV) and liver cirrhosis (Child–Pugh B and C) and has a prevalence of approximately 20–30%. Hyponatraemia occurs less commonly in nephrotic syndrome and only develops when it is associated with severe intravascular volume depletion (usually serum albumin < 20 g/L). The development of hyponatraemia has been recognized as a poor prognostic sign in heart failure and liver cirrhosis, and has even emerged as an independent predictor for mortality (Gheorghiade et al., 2007; Kim et al., 2009). Interestingly, hyponatraemia was recently also found to be a predictor of long-term mortality, and admission for heart failure after hospital discharge in survivors of acute ST-elevation myocardial infarction (Goldberg et al., 2006). The explanation for these associations is probably not so much a direct effect of hyponatraemia, but rather hyponatraemia as a marker of the extent of the so-called neurohumoral response, and therefore the degree of decompensation. That hyponatraemia is a central feature of the neurohumoral response has clearly been demonstrated in heart and liver failure, in which hyponatraemia correlates with the activity of the renin–angiotensin and prostaglandin systems (Dzau et al., 1984). Of interest, hyponatraemia is also sometimes observed in pulmonary embolism and pulmonary hypertension (Forfia et al., 2008; Scherz et al., 2010); this is likely to be due to right-sided heart failure.

Polydipsia and low solute intake Heart failure, liver cirrhosis, and nephrotic syndrome Heart failure, liver cirrhosis, and nephrotic syndrome are oedema-forming disorders in which an increase in total body water exceeds the increase in total body sodium (see Chapter 30). Despite hypervolaemia, the effective arterial blood volume is decreased in all three disorders, although for different reasons. In heart failure this is due to a low cardiac output, in liver cirrhosis it is due to systemic vasodilatation and arteriovenous fistulae, and

Table 28.4  Biochemical and haemodynamic parameters in the syndrome of inappropriate antidiuretic hormone secretion (SIADH) and cerebral salt wasting (CSW) SIADH

CSW

Serum sodium

Low

Low

Serum urea

Normal–low

Normal–elevated

Serum uric acid

Low

Low

Urine volume

Low

High

Urine sodium

> 40 mmol/L

>> 40 mmol/L

Blood pressure

Normal

Normal—postural drop

Central venous pressure

Normal

Low

Data are adapted from Sherlock et al. (2006).

Hyponatraemia due to primary polydipsia, beer potomania, and low solute intake (‘tea and toast’) are similar in the sense that vasopressin activity is usually absent and urine osmolality, therefore, low (Table 28.2). In primary or psychogenic polydipsia patients consume quantities of water that exceed the water excretory capacity of the kidneys (15–20 L/day); it is most commonly seen during an acute psychosis in patients with schizophrenia. One study, however, suggested that in hyponatraemia due to polydipsia, water intake alone was usually not sufficient to explain the degree of hyponatraemia; it was proposed that apparent loss of solutes (possibly through a renal route) played a significant contributory role (Musch et al., 2003). In beer potomonia and subjects with low solute intake, fluid intake is usually less than in primary polydipsia, but solute excretion becomes the rate-limiting step for electrolyte-free water excretion (Thaler et al., 1998). In these settings, smaller amounts of fluid (e.g. 2–5 L/day) can already cause significant hyponatraemia. This mechanism may also play a role in hyponatraemia in anorexia nervosa, which has an estimated prevalence of 20% (Miller et al., 2005).

Extrarenal sodium loss with water retention Hyponatraemia due to extrarenal sodium loss can be observed with gastrointestinal losses, burn wounds or ‘third spacing’ due to pancreatitis, bowel obstruction, or muscle trauma. Although the fluids lost are hypotonic, hyponatraemia can develop when there is hypovolaemia with ongoing intake or administration of hypotonic fluids. These patients have a total body sodium deficit that

255

256

Section 2  

fluid, electrolyte, and renal tubular disorders

exceeds their water deficit. In this setting, extracellular fluid volume contraction causes the release of vasopressin. This is mediated by baroreceptors located in the aortic arch, carotid sinus, cardiac atria, and pulmonary venous system. These patients typically have a very low urine sodium concentration and high urine osmolality (Table 28.2). One exception is vomiting, in which urine sodium is higher because alkalosis-induced bicarbonaturia causes natriuresis. A better parameter in this setting is urinary chloride, which is typically very low, because hydrogen chloride is lost with vomiting.

Renal insufficiency Several causes of hyponatraemia are also associated with acute or chronic kidney disease, including extrarenal sodium loss, heart failure, liver cirrhosis, hepatorenal syndrome, renal artery stenosis, and salt-losing nephropathy. In these disorders, hyponatraemia is usually caused by baroreceptor-mediated vasopressin release, while renal insufficiency is often of prerenal origin. Therefore, the calculation of the fractional sodium excretion is useful, which is usually low (< 1%) except in salt-losing nephropathy. A more separate and rare entity is the hyponatraemic hypertensive syndrome, in which unilateral renal artery stenosis causes significant hypertension, hyponatraemia, hypokalaemia, polydipsia, and polyuria. The pathogenesis of hyponatraemia is ascribed to volume-mediated vasopressin release and polydipsia stimulated by high angiotensin II levels; hypokalaemia is also factor. This syndrome is most often seen in asthenic elderly women who smoke, but it has also been reported in children with fibromuscular dysplasia (Agarwal et al., 1999). Chronic kidney disease can also play a causal role in the pathogenesis of hyponatraemia. When the glomerular filtration rate is very low, or if patients are already undergoing renal replacement therapy, free water clearance is limited or absent. In this setting, hyponatraemia can easily develop if fluid restriction is not adhered to. A recent study showed that hyponatraemia in this setting was associated with poor outcome, because a lower pre-dialysis serum sodium was associated with an increased risk of death (Waikar et  al., 2011). Hyponatraemia is also relatively common among patients undergoing peritoneal dialysis and may be related to hyperglycaemia, depending on the composition of dialysis fluid used, or a catabolic state with potassium depletion (Zevallos et al., 2001; Zanger, 2010).

Treatment and outcome General principles The treatment of hyponatraemia relies on the following principles: acute hyponatraemia should be treated immediately regardless of the cause, whereas treatment should be directed towards the underlying cause in chronic hyponatraemia, while avoiding rapid or over-correction. These opposite strategies are related to the fact that brain cells start adapting to the hypotonic environment within 1 or 2 days by extruding intracellular electrolytes and organic solutes, including myoinositol, phosphocreatine, and amino acids (Fig. 28.4). Because the time at which hyponatraemia developed is frequently unknown, the decision whether hyponatraemia is acute or chronic often depends on the assessment of symptoms, but must assumed to be chronic, if onset and duration are unclear. Although severe neurological symptoms such as seizures and coma should always point in the direction of acute hyponatraemia, more subtle symptoms can occur in both acute and chronic

hyponatraemia (Box 28.2). Another challenge is that ‘acute on chronic’ hyponatraemia may occur, for example, when hypotonic fluids are administered to a chronically hyponatraemic patient. Therefore, a degree of uncertainty often remains when treating hyponatraemia, emphasizing the importance of careful follow-up to assess the serum sodium and evolution of symptoms. The current recommendations for correction rates are based on consensus and expert opinion, rather than evidence from randomized trials (Box 28.5) (Verbalis et al., 2007). Recently, however, a European guideline on the management of hyponatraemia was published (Spasovski et al., 2014). In general, the recommended correction rates have become slightly more conservative over the years. This probably stems from the recognition that osmotic demyelination can already occur with correction rates of 10–12  mmol/L/ day, whereas cerebral oedema due to hyponatraemia can usually be treated effectively by raising the serum sodium by as little as 4–6 mmol/L. The susceptibility to cerebral oedema or osmotic demyelination is higher in certain patient groups (Table 28.5). Elderly women taking thiazides and hypoxaemic patients have a higher risk for both conditions.

Treatment of acute hyponatraemia Acute hyponatraemia is most commonly seen in primary polydipsia, exercise-associated hyponatraemia, the use of drugs such as 3,4-methylenedioxymethamphetamine (‘Ecstasy’), desmopressin, oxytocin, and thiazides, and excessive administration of hypotonic intravenous fluids (Arieff, 1986; Hsu et al., 2005). All these settings are usually characterized by the intake or administration of a large amount of electrolyte-free water in a short period of time, with vasopressin acting simultaneously to prevent excretion. Hypertonic saline remains the treatment of choice for acute hyponatraemia. By introducing a hypertonic solution into the extracellular space, water will be attracted from the intracellular space. This will reduce the cell swelling associated with acute hyponatraemia and is effective in treating cerebral oedema. Several formulae are available to help predict the rise in serum sodium when therapy with hypertonic saline is commenced. Although each formula has its strengths and limitations, we favour the Adrogué–Madias formula, because of its relative simplicity and its validation in clinical studies (Fig. 28.5) (Adrogue and Madias, 2000; Liamis et al., 2006). The Adrogué–Madias formula predicts what the rise in serum sodium will be when 1 L of a given solution is administered to a patient. It requires information on the amount of sodium present in the solution of choice, the serum sodium concentration of the patient, and an estimate of the patient’s total body water. A simpler approach was recently proposed as initial emergency therapy for acute hyponatraemia, namely a bolus infusion of 3% sodium chloride (100 mL or 2 mL/kg, repeated up to two times) (Moritz and Ayus, 2010). Notably, whatever approach is used, the serum sodium concentration should be measured frequently during therapy with hypertonic saline (preferably every 2–4 hours).

Auto-correction and over-correction It is essential to be aware of the possibility of auto-correction or over-correction during the treatment of hyponatraemia. Auto-correction usually occurs when the stimulus for vasopressin release suddenly abates, which is then followed by the rapid excretion of a dilute urine. During this process, the serum sodium concentration can rise quickly with a consequent risk of osmotic

chapter 28 

approach to the patient with hyponatraemia

Fig. 28.4  Effects of hyponatraemia on the brain and adaptive responses. Within minutes after the development of hypotonicity, water gain causes swelling of the brain and a decrease in osmolality of the brain. Partial restoration of brain volume occurs within a few hours as a result of cellular loss of electrolytes (rapid adaptation). The normalization of brain volume is completed within several days through loss of organic osmolytes from brain cells (slow adaptation). Low osmolality in the brain persists despite the normalization of brain volume. Proper correction of hypotonicity re-establishes normal osmolality without risking damage to the brain. Overly aggressive correction of hyponatraemia can lead to irreversible brain damage. From: Adrogué HJ, Madias NE. Hypnatraemia. N Engl J Med 2000; 342: 1581–9.

demyelination. Common examples include the treatment of hypovolaemic hyponatraemia, discontinuation of desmopressin (DDAVP), or the treatment of adrenal insufficiency with steroids. Conversely, over-correction usually occurs during treatment with hypertonic saline when the actual rise in serum sodium

Box 28.5  General principles of the correction rate of hyponatraemia ◆ The correction rate should be ≤ 10 mmol/L during the first 24 hours and ≤ 18 mmol/L during the first 48 hours of treatment. ◆ The maximum correction rates represent limits and should therefore not be the goal of treatment. ◆ Acute and/or symptomatic hyponatraemia may initially be corrected faster with 1–2 mmol/L/hour. ◆ If hyponatraemia is definitely chronic or if there are risk factors for the osmotic demyelination syndrome (see Table 28.5), slower rates of correction should be applied (≤ 8 mmol/L/day).

concentration exceeds the predicted rise (Mohmand et al., 2007). Auto-correction and overcorrection should be anticipated during the treatment of hyponatraemia by regularly monitoring the serum sodium concentration, urine osmolality, and urine output. If urine production increases and urine tonicity decreases, this suggests the onset of a water diuresis with the likelihood of a rapid rise in serum sodium. If the maximum correction rate is exceeded during auto-correction or over-correction, measures should be taken to curtail the rise in serum sodium concentration. Besides discontinuing the active therapy for hyponatraemia, additional measures include the infusion of hypotonic solutions or the administration of DDAVP (Perianayagam et al., 2008). In fact, some have proposed co-administering DDAVP during the correction of hyponatraemia to achieve a more controlled rise in serum sodium, especially in those patients with risk factors for osmotic demyelination (Table 28.5) (Sterns et al., 2010). In experimental animals, re-induction of hyponatraemia after rapid over-correction of hyponatraemia reduces mortality (Gankam-Kengne et  al., 2009). Other experimental manoeuvres that have been reported to improve the outcome of osmotic demyelination in rodents include

257

258

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 28.5  Risk factors for neurologic complications of hyponatraemia Cerebral oedema

Postoperative state Use of desmopressin Thiazide diuretics Children Psychogenic polydipsia Hypoxia

Osmotic demyelination syndrome (during correction of hyponatraemia) Alcohol abuse Thiazide diuretics Malnourishment Hypokalaemia Hypoxia Burn victims Steroids for adrenal insufficiency Liver cirrhosis

Data are adapted from Lauriat and Berl (1997).

the administration of myoinositol, which increases the uptake of this osmolyte in brain cells (Silver et al., 2006), and minocycline, which decreases the permeability of the blood–brain barrier (Gankam-Kengne et al., 2010).

Treatment of chronic hyponatraemia Because immediate treatment of cerebral oedema is not an issue during chronic hyponatraemia, therapy can be directed towards the underlying cause. The treatment modalities for the different causes of hyponatraemia are shown in Table 28.6. Some of the treatments are straightforward, such as discontinuation of the offending drug, or treatment with steroids or thyroid hormone in hyponatraemia due to hypocortisolism or hypothyroidism. The remaining treatments are less targeted in the sense that they do not suppress the stimulus for hyponatraemia. These treatments are directed to restricting the intake of electrolyte-free water or promoting its excretion and include fluid restriction, loop diuretics, urea, and demeclocycline. Fluid restriction and loop diuretics are most commonly used in the treatment of hyponatraemia due to SIADH, heart failure or liver cirrhosis, while urea and demeclocycline are

Volume (L) =

Change in serum Na+ =

Desired ∆[Na]s ∆[Na]s (with 1 L) (infusate Na+ + infusate K+)–serum Na+ total body water + 1

Figure 28.5  Adrogué–Madias formula for predicting the effect on the serum sodium concentration with the administration of a given infusate. The upper formula can be used to predict the rise in serum sodium (Δ[Na]s) when 1 L of a given infusate is administered. The formula relies on subtracting the sodium concentration of the infusate ([Na]inf) from the serum sodium concentration in the patient ([Na]1) and dividing this number by total body water (TBW) plus one (because TBW will increase with 1 L when 1 L of the infusate is administered). The estimated TBW is calculated as a fraction of body weight. The fraction is 0.6 in children; 0.6 and 0.5 in nonelderly men and women, respectively; and 0.5 and 0.45 in elderly men and women, respectively. When potassium is added to the infusate, this concentration should be included in the formula ([Na + K]inf). The lower formula can be used to calculate the volume of the infusate necessary to achieve the desired rise in serum sodium (Desired Δ[Na]s) by dividing this number with the calculated change in serum sodium concentration in the upper part of the formula (Δ[Na]s). From: Adrogué HJ, Madias NE. Hyponatraemia. N Engl J Med 2000; 342: 1581–9.

Table 28.6  Treatment modalities for hyponatraemia Treatment

Indication

Hypertonic saline

Acute and/or symptomatic hyponatraemia

Isotonic saline

Hyponatraemia associated with hypovolaemia

Vasopressin-receptor antagonists

SIADH, heart failure, liver cirrhosisa

Loop diuretics

SIADH, primary polydipsia, heart failure, liver cirrhosis

Urea

SIADH

Water restriction

SIADH, heart failure, liver cirrhosis, primary polydipsia

Dietary intake of salt and protein

Hyponatraemia due to low solute intake

Cause-directed therapy

E.g. thyroid hormone in hypothyroidism, steroids in adrenal insufficiency or discontinuation of the offending drug

a In Europe, vasopressin-receptor antagonists have only been registered for hyponatraemia

secondary to SIADH. CSW = cerebral salt wasting; SIADH = syndrome of inappropriate antidiuretic hormone secretion.

usually restricted to the treatment of SIADH. The recommended degree of fluid restriction should be determined by relating the urine sodium and potassium concentrations (which determine tonicity) to the serum sodium concentration (Fig. 28.6). Loop diuretics inhibit the generation of a concentration gradient in the renal medulla and promote the excretion of sodium and water. Urea causes an osmotic diuresis, which also promotes the excretion of electrolyte-free water. Demeclocycline is an antibiotic with nephrogenic diabetes insipidus as a side effect; this effect can be exploited during hyponatraemia to induce a water diuresis. Because of significant side effects and potential overcorrection, however, demeclocycline is not recommended. In many patients, especially the elderly, low solute intake plays a contributory role in the development of hyponatraemia. Therefore, fluid restriction or loop diuretics may be combined with increased dietary intake of sodium and protein (or alternatively sodium chloride tablets). It is also important to emphasize that hyponatraemia may resolve when the underlying disorder has been treated, for example, in SIADH due to infection: in this setting supportive treatment may be sufficient to prevent a further fall in the serum sodium. However, there will be causes of chronic hyponatraemia that remain difficult to treat and in which more targeted therapy would be desirable. Recently, a

Urinary sodium + urinary potassium Serum sodium >1 ~1 50% within 1–2 hours) indicates that the V2 vasopressin receptor and its downstream cascade function normally. This proves that the reason for the low urine osmolality is insufficient endogenous AVP secretion. A response of < 10% indicates that there is a resistance to the effects of vasopressin, confirming the diagnosis of nephrogenic DI. However, indeterminate increases between 10% and 50 % occur regularly. This may be due to washout of the medullary concentration gradient or downregulation of aquaporin-2 expression in the intramedullary collecting duct, resulting from chronic vasopressin deficiency. If urinary osmolality is high, the body is efficiently trying to conserve water by maximizing the reabsorption of water in the collecting duct. There is a wide variation in the urine osmolality that can be maximally achieved and this maximum decreases with a reduction in the number of functioning nephrons due to age or renal failure. In general, a urine osmolality of 800 mOsmol/kg H2O is considered an appropriate response to hypernatraemia, indicative of reduced water intake or extrarenal loss. An intermediate urine osmolality (300–800 mOsmol/kg) is compatible with an osmotic diuresis. However, in some patients with DI, either central or nephrogenic, the defect is partial. Such a defect can also result in intermediate urinary osmolalities. Alternatively, severe volume depletion can lead to intermediate urine osmolalities in patients with central DI, as a result of residual water permeability of the collecting duct.

Urinary sodium concentration In the absence of diuretics, sodium output in the urine is primarily useful for the estimation of volume status. Hypovolaemia leads to increased sodium reabsorption along the nephron, resulting in a urine sodium concentration of 20–30 mmol/L or lower. When hypernatraemia is due to (ongoing) osmotic diuresis or a positive sodium balance, urine sodium concentration will be (considerably) higher than 30 mmol/L.

Urinary sodium plus potassium concentrations For hypernatraemia to develop, water must be lost in excess of sodium and potassium salts. Urine contains organic solutes such as urea that have no effect on water balance, but that do affect the osmolality of the urine. Quantitatively, it is more appropriate to compare 2 × (Urinary [Na] + [K]‌) to plasma osmolality. When a

chapter 29 

approach to the patient with hypernatremia

Table 29.4  Diagnostic approach to hypernatraemia based urine parameters Reduced water intake

Diabetes insipidus

Osmotic diuresis

Extrarenal water loss

Sodium gain

Osmolality

Uosm maximal

Uosm < Posm

Uosm > Posm

Uosmmaximal

Uosmmaximal

Urine sodium

< 25 mmol/L

< 25 mmol/L

> 25 mmol/L

< 25 mmol/L

> 25 mmol/L

Urine output

Oliguria

Polyuria

Polyuria

Oliguria

Normal

patient excretes urine in which 2 × (Urinary [Na] + [K]) is lower than plasma osmolality, net (free) water is lost in the urine.

Electrolyte-free water excretion When calculating the urinary excretion of water and solutes, it has been advocated to divide the excreted volume in two parts, one part isotonic fluid loss and one part loss of pure water (Rose, 1986). The latter is called electrolyte-free water and affects net water balance. However, when a patient excretes hypertonic urine, the amount of electrolyte-free water that is lost becomes negative, making interpretation a challenge. Also, this approach does not take the fluid input into account.

Tonicity balance Although urine osmolality is important to determine whether vasopressin is acting or not, it is not particularly useful in quantitative terms. To determine the effects infusion fluids have had or will have, it is more appropriate to calculate the input and output of water and sodium + potassium separately (Fig. 29.2). This approach has been called the tonicity balance (Carlotti et al., 2001). Similar to the calculation of electrolyte free water, it requires the availability of data on both input and output. This generally limits the approach to dysnatraemias that occur during hospitalization. In patients with severe catabolism, the balance of other solutes also needs to be taken into account (Halperin and Bohn, 2002).

Hypernatremia due to a negative water and positive sodium balance Na+ + K+ 557 + 80 mmoL

SNa 138 mmol/L (begin) +384 mmoL

Na+ + K+ 176 + 77 mmoL

Na H2O Water 3.6 L

–0.9 L SNa 155 mmol/L (end)

Water 4.5 L

Fig. 29.2  Tonicity balance in hypernatraemia. The large darker rectangles represent total body water with the serum sodium concentration measured at the beginning and end of the observation shown on top and bottom of this rectangle, respectively. The quantities of sodium (Na+) plus potassium (K+) infused and excreted are shown in the two flanking rectangles, and the volumes of water (H2O) infused and excreted are depicted below. The patient was a 50-year-old female (body weight 75 kg) who was admitted with respiratory insufficiency due to pneumonia. Hypernatremia developed in 4 days and was attributed to a combination of a negative water balance and a positive sodium balance due to the infusion of isotonic fluids and renal water loss from hyperglycaemia, hypercalcaemia, and hypokalaemia. From: Hoorn EJ et al. Nephrol Dial Translant 2008; 23:1562–1568.

Water deprivation test Although the increase in urine osmolality during fluid restriction certainly has merit in the work-up of patients with polyuria, it should not be performed in patients who are overtly hypernatraemic.

Aetiology and pathogenesis Causes of hypernatraemia can be categorized as reduced water intake, renal or extrarenal water loss, and sodium gain (Table 29.1).

Unavailability of, or access to, water In general, in the developed world, water is readily available. This does not hold true for all countries, now that water is becoming a precious commodity in many places. Even if water is available in principle, reduced mobility of the patient can prevent sufficient access. In patients with reduced levels of consciousness, fluid intake can become inadequate, leading to hypertonic dehydration. This can also be the case in the elderly who have inadequate social support or are admitted to understaffed nursing homes. In these patients fluid intake is further challenged by a decrease in thirst sensation that occurs with increasing age (Phillips et  al., 1991). Mortality is high in elderly patients presenting with severe hypernatraemia, with the state of consciousness being the single most important prognostic factor (Chassagne et al., 2006). In general, increased fluid intake can prevent the disorder, but the 2003 heat wave in France has shown that over-zealous drinking (possibly combined with volume depletion) may lead to hyponatraemia in these patients (Kettaneh et al., 2010). Breastfed neonates are completely dependent on their mother for adequate hydration. Neonatal hypernatraemic dehydration is present in 2% of hospitalized neonates (Moritz et al., 2005). It can present with lethargy, fever, and jaundice (Boskabadi et al., 2010). It results from inadequate feeding and is associated with a lower than normal number of feeds and with breastfeeding problems in the mother. This potentially devastating disorder can be prevented by frequent weighing (Unal et al., 2008).

Defective thirst mechanism Thirst is an important physiological mechanism in the defence against dehydration. It can be stimulated by both increases in effective osmolality of the ECV and intravascular hypovolaemia. In elderly patients, water intake is often not increased in the face of increased fluid loss, which leads to hypernatraemia (Molaschi et al., 1997). This can be the result of a decreased sensitivity of the thirst mechanism (Adeleye et al., 2002). Thirst perception appears to be modulated by sodium balance, with a high salt diet (accompanied by low endogenous angiotensin II levels) increasing the osmotic threshold at which thirst occurs (Gordon et al., 1997). Low levels of

265

266

Section 2  

fluid, electrolyte, and renal tubular disorders

angiotensin II have been implicated in reducing thirst perception in the elderly (Yamamoto et al., 1988).

Central diabetes insipidus (See also Chapter 32.) Vasopressin is synthesized in the hypothalamus and is transported down the axons of the supraoptical-hypophyseal tract to be stored in and released from the pituitary. When this tract or the pituitary is damaged, vasopressin release can be diminished. Several forms of central DI can be distinguished.

Hereditary An autosomal dominant form of central DI can result from mutations in the gene that encodes neurophysin II (Stephen et al., 2012). In this disease, loss of neurophysin II, a carrier protein for vasopressin, results in a progressive decrease in vasopressin levels. At birth patients are asymptomatic, but progressive polyuria develops with hypernatraemia later in childhood (Arima and Oiso, 2010) (Table 29.5).

Post-traumatic DI occurs in a substantial percentage of patients presenting with severe head injury (blunt, but especially penetrating) (Hadjizacharia et al., 2008). Independent risk factors are a Glasgow coma score ≤ 8, cerebral oedema, and an Abbreviated (head) Injury Score ≥ 3.

Tumours and infections Tumours that disrupt the supraoptical-hypophyseal tract can lead to central DI. This complication can occur in a variety of primary or secondary malignancies of the brain, as has been reviewed elsewhere (Verbalis, 2003). Several types of infections affecting the brain and or the meninges have been associated with central DI, including tuberculosis (Bajpai et  al., 2008), influenza (Kobayashi et  al., 2011), and histiocytosis (Schmitt et al., 1993).

Post-surgical Disturbances in water excretion are a common cause of morbidity following trans-sphenoidal surgery. In a large series, the incidence of central DI in the postoperative phase was 18.3% (Nemergut et al., 2005). However, only 2% of patients required long-term treatment with DDAVP. Hyponatraemia, resulting from vasopressin release Table 29.5  Forms of diabetes insipidus Central diabetes insipidus

Nephrogenic diabetes insipidus

Hereditary Post-traumatic Tumours and infections Postsurgical Ethanol consumption Idiopathic Adipsogenic

Mutations of the vasopressin V2 receptor Mutations of aquaporin-2 Medullary renal disease. Hypercalcaemia or hypokalaemia Drugs: Lithium Amphotericin B Demeclocycline Foscarnet V2-receptor antagonists

from the damaged pituitary, is observed in 25% of patients (Olson et  al., 1997), and in 5% of patients the clinical course follows a biphasic or triphasic pattern, where central DI is interspersed with a period of SIADH (Hensen et al., 1999; Hoorn and Zietse, 2010).

Adipsogenic In some patients with central DI, reduced vasopressin release from the hypophysis is accompanied by the absence of thirst. This entity is known as adipsic DI and it is due to lesions in the hypothalamus, such as a craniopharyngioma (Crowley et al., 2007). Whereas thirst and subsequent drinking usually prevent overt hypernatraemia in patients with central DI, patients with the adipsic form can be markedly hypernatraemic. In most cases, the hormonal response to non-osmotic stimuli is intact (Smith et al., 2002). Following successful surgery, thirst sensation can recover (Sinha et al., 2011).

Nephrogenic diabetes insipidus (See also Chapter 32.)

Hereditary Congenital defects in either the vasopressin receptor (V2R) or aquaporin-2 (AQP2) can lead to nephrogenic DI (Knoers, 1993). As the V2R gene is located on the X chromosome, the disease follows an X-linked recessive inheritance pattern. It accounts for the vast majority of patients with congenital nephrogenic DI (~ 90%) and is by nature a disease affecting males only. Female carriers are usually asymptomatic, but in some patients skewed inactivation of the X chromosome, favouring the mutated gene, can lead to partial disease (Faerch et al., 2010). In most cases, AQP2-linked nephrogenic DI follows an autosomal recessive inheritance pattern (Nossent et al., 2008). In these patients, decreased expression of aquaporin leads to the concentrating defect, with osmosensing and vasopressin secretion being intact. As a result of renal water loss, circulating vasopressin levels and V2R gene expression may be increased. Activation of endothelial V2 receptors can lead to increase secretion of von Willebrand factor resulting in an increased risk for thromboembolism (Nossent et al., 2010). Hereditary disorders that do not directly affect water transport in the collecting duct may have marked effects on the urinary concentration mechanism leading to hypernatraemia with hypotonic urine (Bockenhauer et al., 2010). In tubular disorders, such as Bartter syndrome and nephronophthisis, the reabsorption of sodium is diminished, resulting in isosthenuria (i.e. urine isotonic to plasma). Such patients may present with episodes of hypernatraemia unresponsive to DDAVP and the erroneous diagnosis of congenital nephrogenic DI. Severe hypokalaemia, that is frequently present, may further reduce renal water reabsorption.

Acquired The most frequent cause of nephrogenic DI encountered in clinical practice is the use of drugs that inhibit the renal concentrating ability. A wide variety of drugs have been associated with nephrogenic DI (Table 29.5) and these have been extensively reviewed elsewhere (Garofeanu et al., 2005). Here, we will only discus the two most clinically relevant: lithium and amphotericin B. By far the most prominent agent that causes DI is lithium, which is used to treat bipolar disorders, with a reported incidence of 40% of patients receiving the drug (Grunfeld and

chapter 29 

Rossier, 2009). Lithium enters the tubular epithelial cell through the epithelial sodium channel (ENaC) and decreases aquaporin phosphorylation, thereby inhibiting translocation and insertion into the luminal membrane (Nielsen et al., 2008). Amiloride, an agent used as a potassium-sparing diuretic, because it blocks the ENaC, prevents the entry of lithium in the collecting duct epithelial cell and improves the renal concentrating ability (Kortenoeven et al., 2009). This putative effect of amiloride has been demonstrated in humans, where the drug increased both concentrating ability and AQP2 excretion in patients on lithium (Bedford et al., 2008). Amphotericin B, decreases the ability of the kidney to concentrate the urine, due to both structural damage to the apical plasma membrane and to decreased AQP2 abundance (Zietse et al., 2009). Although hypokalaemia is frequently present in such patients and may contribute, the concentration defect also occurs when plasma potassium concentration is normal. In some patients, drug-induced DI can be associated with other tubular defects (Hoorn and Zietse, 2007).

Hypokalaemia In rats, hypokalaemia has been shown to reduce the number of aquaporin water channels in the luminal membrane of the collecting duct (Marples et al., 1996). Although hypokalaemia is frequently associated with polyuria, it rarely leads to hypernatraemia. At present it has not been shown whether hypokalaemia-induced polyuria results from nephrogenic DI or from interstitial damage due to prolonged hypokalaemia.

Hypercalcaemia In the kidney, the calcium sensing receptor (CaSR) is present both in the thick ascending limb (TAL) of the loop of Henle (basolateral membrane) and in the collecting duct (luminal membrane). In the collecting duct, the CaSR co-localizes with AQP2. In vitro data have shown that water permeability decreases at higher luminal calcium concentrations (Earm et  al., 1998). Although these findings are compatible with nephrogenic DI, hypercalcaemia-induced polyuria is more likely to result from activation of the CaSR in the TAL, resulting in a loop diuretic-like effect (Hoorn et al., 2009).

Gestational diabetes insipidus In pregnant patients presenting with hypernatraemia, the existence of diabetes mellitus should be excluded, because this is considerably more frequent than DI and can lead to grave consequences (Lee et al., 2010). Gestational DI is a rare disorder occurring in the third trimester of pregnancy (Sherer et al., 2003). It results from the circulating enzyme vasopressinase, which degrades circulating vasopressin, leading to non-renal DI. As vasopressinase is produced in the placenta, hypernatraemia is transient in this disorder (Aleksandrov et al., 2010). Treatment with vasopressin is not effective, but as the synthetic DDAVP is not cleaved by vasopressinase, DDAVP will readily correct the polyuria (Ananthakrishnan, 2009).

Osmotic diuresis If plasma osmolality is greater than urine osmolality and the daily osmole excretion rate is > 1000 mOsmol/day an osmotic diuresis is present. The next step is to determine which solute is responsible for the observed increase in osmole excretion. Hyperglycaemic hyperosmolarity is the best clinical example of a disorder in which osmotic diuresis causes hypernatraemia

approach to the patient with hypernatremia

(Kitabchi and Nyenwe, 2006). When hyperglycaemia causes the amount of filtered glucose to exceed the reabsorptive capacity of the proximal tubule, significant glycosuria results. As glucose acts as an effective osmole in the tubular fluid, it prevents the reabsorption of water in the collecting duct. This reabsorption is further impaired by a relatively low osmolality of the renal medulla (which normally acts as the driving force for water reabsorption). During osmotic diuresis, urine contains a relatively low concentration of electrolytes (sodium plus potassium in the range of 50–80 mmol/L) and therefore the body loses more water than electrolytes, that is, loss of free water (Halperin and Bohn, 2002). Apart from glucose, increased excretion of urea can also induce an osmotic diuresis, especially in the ICU. This can result from hyperalimentation, increased catabolism, gastrointestinal bleeding, or recovery from transient renal failure (Lindner et al., 2012).

Diuretics In rare cases, loop diuretics can lead to hypernatraemia by interfering with the renal concentrating mechanism. The effect of loop diuretics is short-lived and hypernatraemia only occurs when water loss is not replaced by oral intake.

Gastrointestinal water loss Prolonged diarrhoea and/or vomiting can lead to the loss of water from the body (Hartling et al., 2006). In secretory diarrhoeas, such as cholera, the fluid that is lost consists of Na and K salts and is isosmotic to plasma. Such fluid loss will readily lead to volume depletion, but will not cause hypernatraemia. In osmotic diarrhoeas, however, the main osmoles lost are organic compounds such as lactulose. As water is lost in relative excess of Na and K, hypernatraemia may develop. During prolonged vomiting, as occurs in oncology patients during chemotherapy, oral water intake is impaired, whereas insensible water loss continues, leading to hypernatraemia (Berk and Rana, 2006).

Cutaneous water loss Excessive sweating could cause the body to lose large amounts of water that, if not replaced, would lead to hypernatraemia. Usually, however, this water loss leads to thirst and replenishment of lost fluids. Indeed, in collapsed marathon runners hyponatraemia is at least as likely to occur as hypernatraemia (Kratz et al., 2005). Burn injury affects the skin integrity and its protection against fluid loss is lost (Pruitt, 1978). Therefore, severely burned patients need careful and extensive fluid resuscitation. This is even more important, because the presence of hypernatraemia in patients with burns adversely affects skin grafts (Namdar et al., 2011). All states that increase the insensible loss of water through the skin (and from the respiratory tract) put the patient at risk of developing a hyperosmolar state. Preterm infants are especially at risk (Wada et al., 2008), as are patients with an increased body temperature and patients in hot and dry environments. In all these circumstances hypernatraemia will only develop if the patient is unable to drink freely.

Gain of sodium In patients incapable of regulating their own fluid intake, hypernatraemia can develop for a variety of reasons (Lindner et  al., 2009). In the ICU, uncorrected loss of water is certainly responsible for many instances of hypernatraemia. However, in a substantial number of cases the (hypervolaemic) hypernatraemia that develops

267

268

Section 2  

fluid, electrolyte, and renal tubular disorders

in the ICU results from a positive sodium balance (Hoorn et al., 2008). These patients commonly receive isotonic intravenous fluids, but pass relatively hypotonic urine. This indicates that hypernatraemia in the ICU is the result of inadequate fluid management and is, for the most part, preventable. As hypernatraemia in ICU patients is associated with adverse outcome, its incidence has been proposed as an indicator of the quality of medical care in the intensive care environment (Polderman et al., 1999).

Treatment and outcome In general, inducing a positive net water balance treats hypernatraemia. In patients in whom the gastrointestinal tract is functional, this is best achieved with oral water intake. Intravenous water administration with glucose 5% solutions (dextrose; D5W) is limited by the amount of glucose that can be metabolized. If this maximal rate (estimated at 0.3 L/hour) is exceeded, hyperglycaemia may develop, leading to osmotic diuresis and further electrolyte free water loss. The therapeutic approach to the patients should include the steps listed in Table 29.6. Assess potential dangers to the patient. At presentation there are several questions one should ask before embarking on a specific treatment. Question 1: is it acute or is it chronic? In acute hypertonicity, the brain has a lower osmolality than plasma, which leads to shrinkage and ultimately (in rare cases) cerebral haemorrhage. In acute hypernatraemia the brain has not had the time to respond to the hypertonic state by increasing the amount of intracellular osmoles. Although it is uncertain when this correction is complete, it is estimated that it takes at least 48 hours for the brain to accumulate enough osmoles to achieve a new balance. Most patients with hypernatraemia present in the emergency department and the time-course is usually uncertain. Unless a patient has severe symptoms and a history that is compatible with the acute development of hypertonicity, it is prudent to assume that the hypernatraemia is chronic. The main therapeutic implication of the distinction between acute and chronic is that in acute/symptomatic hypernatraemia plasma osmolality should be lowered rapidly to avoid cerebral complications. In chronic hypernatraemia, however, overly ambitious lowering of plasma osmolality would lead to cerebral oedema and death due to brain herniation. Similar to recommendations in hyponatraemia, patients with hypernatraemia should be treated rapidly only if severely symptomatic due to the rapid (< 48 hours) development of hypernatraemia.

Table 29.6  Steps in treating hypernatraemia Assess dangers: Is it acute or is it chronic? Is (profound) hypovolemia present? What has caused hypertonicity? Correct ECV if circulation is threatened Set treatment target for [Na] Select strategy to achieve target Check progress frequently

Question 2: is (profound) hypovolaemia present? Frequently, water depletion is accompanied by the loss of sodium and or potassium salts, leading to true volume depletion. Plasma osmolality will only give an estimate on the amount of water that is lost relative to solute content, but will not indicate the amount of ‘isotonic fluid’ that is lost. Question 3: what has caused the hypertonicity? A major distinction should be ‘is this sodium gain’ (virtually only seen in patients admitted to the hospital/ICU) or ‘is this net water loss’ (nearly always). If considerable sodium gain is present and the patient is fluid overloaded, treatment with large amounts of hypotonic fluid may lead to pulmonary oedema. In such patients, natriuresis should be induced with (loop) diuretics, as discussed below. In patients with hypernatraemia due to water loss, urine output should be measured to determine if (persistent) polyuria is present. As the expected urine flow would be low, in these patients (inappropriate) polyuria can be defined as a urine output > 30 mL/hour. It is important to take the effect of osmolytes other than sodium on plasma tonicity into account. Especially, the contribution of glucose to measured osmolality should be assessed, since the treatment of hypernatraemia with concurrent hyperglycaemia poses specific challenges.

Correct ECV depletion if necessary When a patient presents with circulatory shock, the first step in treatment should be treating with isotonic fluids, such as 0.9% saline, to restore ECV. The goal of this treatment is to restore tissue perfusion. As this ‘isotonic’ saline is hypotonic to the patient it will, depending on the volume and constitution of the urine, also modestly lower plasma osmolality. When the circulation has been restored more hypotonic solutions can be employed.

Set the [Na] target In symptomatic patients, the first aim is to induce a shift of water towards the brain, by rapidly lowering plasma osmolality. Plasma sodium should be lowered by 1–2 mmol/L/h until symptoms disappear. Although there are no data from controlled trials, the overall correction rate should probably not exceed 8 mmol/L per 24 hours.

Start treatment to achieve the target Replace water loss Calculating the water deficit can provide an estimate of the amount of water that has been lost:

(

)

Water deficit = 0.6 × lean body weight × ( plasma [ Na ] / 140 ) − 1 Using this formula, estimation is made of the amount of water that is required to return plasma sodium concentration to 140 mmol/L. Following this, a period must be selected in which to restore water balance. The effect a litre of a given infusate will have on plasma sodium concentration can be calculated using the formula developed by Adrogué and Madias (2000):

chapter 29 

Change in serum Na+ =

(infusate Na+ + infusate K+)–serum Na+ total body water + 1

Fig. 29.3 

In this approaches, estimation must be made of total body water. Usually the total body water is assumed to be 60% of lean body weight, which is not likely to be true in severely water-depleted patients. In cachectic and/or elderly patients, this percentage can easily be as low as 40%. Using these calculations may lead to over-estimation of water deficit with the subsequent danger of over-correction. Both calculations are based on water loss and do not account for water that is lost concurrently with solutes, leading to volume depletion. In many patients dehydration is accompanied by true volume depletion. When estimating the correction rates with these formulae, no correction is made for ongoing (renal) water loss. The effect a given infusate will have on plasma [Na] can be estimated by calculating a tonicity balance, where both input and output of sodium, potassium and water are calculated.

Stop ongoing renal water loss The appropriate homeostatic response to hypertonicity is to increase the release of vasopressin from the pituitary, thereby increasing urinary osmolality and decreasing urine flow. In patients with urine flows that are inappropriately high (> 30 mL/h), the distinction should be made if there is a water diuresis (UOsm < POsm) or a solute diuresis (UOsm > POsm). In patients with a water diuresis, vasopressin may be given in the form of DDAVP to correct a possible vasopressin deficiency. The advantage of DDAVP is that it is not subject to degradation by vasopressinase and can be used in pregnancy-related DI. If DDAVP does not reduce urine output, nephrogenic DI may be present. In patients with a solute diuresis, the nature of the excreted solutes must be determined. The substances most frequently responsible for osmotic diuresis are glucose and urea.

Check progress Calculation of correction rates is a useful basis and starting point for treatment, but these are frequently inaccurate. Many of the caveats previously described can affect the ability of even the best formulae to predict the response to treatment. This may be especially true in the ICU environment (Lindner et al., 2008), where the difference between predicted and measured sodium concentrations was as high as 15 mmol/L in some patients. This indicates that the response to treatment should always be followed closely (repeat measurements every 2–4 hours) and treatment should be adjusted in response to the actual measurements of plasma sodium.

Treatment in specific situations Hypernatraemia in diabetes mellitus In patients with diabetes, hyperglycaemia leads to a shift of water from the intracellular to the extracellular compartment. This dilution causes a reduction in the plasma sodium concentration. As

approach to the patient with hypernatremia

hyperglycaemia leads to osmotic diuresis, water is lost in the urine and plasma sodium may rise. Both glucose and sodium contribute to the effective osmolality in such patients. When glycaemia is corrected, plasma sodium concentration should rise, as water returns to the intracellular compartment. Effective osmolality (formula in p. 263) should be calculated at various time points and the goal of therapy is to prevent a significant fall is effective osmolality during the initial treatment to prevent cerebral oedema. This is especially important in children, because the brain is relatively large compared with the size of the skull at a younger age (Hoorn et al., 2007).

Hypernatraemia in the ICU In the ICU, patients can lose water in a variety of ways and the most obvious treatment is to ‘just add water’ (Sterns, 1999). However, in a substantial number of patients sodium gain is present. In such patients, giving a water load can lead to fluid overload and pulmonary oedema. Excess sodium should be removed using (loop) diuretics. As diuretic-induced urine output is hypotonic to the patient, water should be given to lower the plasma osmolality. In patients with acute kidney injury and concomitant hypernatraemia, continuous renal replacement therapies can be used to gradually correct hypertonicity (Ostermann et al., 2010).

References Adeleye, O., Faulkner, M., Adeola, T., et al. (2002). Hypernatremia in the elderly. J Natl Med Assoc, 94, 701–5. Adrogué, H. J. and Madias, N. E. (2000). Hypernatremia. N Engl J Med, 342, 1493–9. Aleksandrov, N., Audibert, F., Bedard, M. J., et al. (2010). Gestational diabetes insipidus: a review of an underdiagnosed condition. J Obstet Gynaecol Can, 32, 225–31. Alshayeb, H. M., Showkat, A., Babar, F., et al. (2011). Severe hypernatremia correction rate and mortality in hospitalized patients. Am J Med Sci, 341, 356–60. Ananthakrishnan, S. (2009). Diabetes insipidus in pregnancy: etiology, evaluation, and management. Endocr Pract, 15, 377–82. Arima, H. and Oiso, Y. (2010). Mechanisms underlying progressive polyuria in familial neurohypophysial diabetes insipidus. J Neuroendocrinol, 22(7), 754–7. Bajpai, A., Kabra, M., and Menon, P. S. (2008). Central diabetes insipidus: clinical profile and factors indicating organic etiology in children. Indian Pediatr, 45, 463–8. Balanescu, S., Kopp, P., Gaskill, M. B., et al. (2011). Correlation of plasma copeptin and vasopressin concentrations in hypo-, iso-, and hyperosmolar States. J Clin Endocrinol Metab, 96, 1046–52. Barnette, A. R., Myers, B. J., Berg, C. S., et al. (2010). Sodium intake and intraventricular hemorrhage in the preterm infant. Ann Neurol, 67, 817–23. Bedford, J. J., Weggery, S., Ellis, G., et al. (2008). Lithium-induced nephrogenic diabetes insipidus: renal effects of amiloride. Clin J Am Soc Nephrol, 3, 1324–31. Berk, L. and Rana, S. (2006). Hypovolemia and dehydration in the oncology patient. J Support Oncol, 4, 447–54. Bettinelli, A., Ciarmatori, S., Cesareo, L., et al. (2000). Phenotypic variability in Bartter syndrome type I. Pediatr Nephrol, 14, 940–5. Bockenhauer, D., Van’t Hoff, W., Dattani, M., et al. (2010). Secondary nephrogenic diabetes insipidus as a complication of inherited renal diseases. Nephron Physiol, 116, 23–9. Boskabadi, H., Maamouri, G., Ebrahimi, M., et al. (2010). Neonatal hypernatremia and dehydration in infants receiving inadequate breastfeeding. Asia Pac J Clin Nutr, 19, 301–7.

269

270

Section 2  

fluid, electrolyte, and renal tubular disorders

Bourdel-Marchasson, I., Proux, S., Dehail, P., et al. (2004). One-year incidence of hyperosmolar states and prognosis in a geriatric acute care unit. Gerontology, 50, 171–6. Bruce, J. M., Harrington, C. J., Foster, S., et al. (2009). Common blood laboratory values are associated with cognition among older inpatients referred for neuropsychological testing. Clin Neuropsychol, 23, 909–25. Carlotti, A. P., Bohn, D., Mallie, J. P., et al. (2001). Tonicity balance, and not electrolyte-free water calculations, more accurately guides therapy for acute changes in natremia. Intensive Care Med, 27, 921–4. Chassagne, P., Druesne, L., Capet, C., et al. (2006). Clinical presentation of hypernatremia in elderly patients: a case control study. J Am Geriatr Soc, 54, 1225–30. Crowley, R. K., Sherlock, M., Agha, A., et al. (2007). Clinical insights into adipsic diabetes insipidus: a large case series. Clin Endocrinol, 66, 475–82. Darmon, M., Timsit, J. F., Francais, A., et al. (2010). Association between hypernatraemia acquired in the ICU and mortality: a cohort study. Nephrol Dial Transplant, 25, 2510–15. Dmitrieva, N. I. and Burg, M. B. (2011). Increased insensible water loss contributes to aging related dehydration. PloS One, 6, e20691. Earm, J. H., Christensen, B. M., Frokiaer, J., et al. (1998). Decreased aquaporin-2 expression and apical plasma membrane delivery in kidney collecting ducts of polyuric hypercalcemic rats. J Am Soc Nephrol, 9, 2181–93. Faerch, M., Corydon, T. J., Rittig, S., et al. (2010). Skewed X-chromosome inactivation causing diagnostic misinterpretation in congenital nephrogenic diabetes insipidus. Scand J Urol Nephrol, 44, 324–30. Fenske, W., Quinkler, M., Lorenz, D., et al. (2011). Copeptin in the differential diagnosis of the polydipsia-polyuria syndrome—revisiting the direct and indirect water deprivation tests. J Clin Endocrinol Metab, 96, 1506–15. Finberg, L. and Harrison, H. E. (1955). Hypernatremia in infants; an evaluation of the clinical and biochemical findings accompanying this state. Pediatrics, 16, 1–14. Funk, G. C., Lindner, G., Druml, W., et al. (2010). Incidence and prognosis of dysnatremias present on ICU admission. Intensive Care Med, 36, 304–11. Garofeanu, C. G., Weir, M., Rosas-Arellano, M. P., et al. (2005). Causes of reversible nephrogenic diabetes insipidus: a systematic review. Am J Kidney Dis, 45, 626–37. Gordon, M. S., Majzoub, J. A., Williams, G. H., et al. (1997). Sodium balance modulates thirst in normal man. Endocr Res, 23, 377–92. Grunfeld, J. P. and Rossier, B. C. (2009). Lithium nephrotoxicity revisited. Nat Rev Nephrol, 5, 270–6. Hadjizacharia, P., Beale, E. O., Inaba, K., et al. (2008). Acute diabetes insipidus in severe head injury: a prospective study. J Am Coll Surg, 207, 477–84. Halperin, M. L. and Bohn, D. (2002). Clinical approach to disorders of salt and water balance. Emphasis on integrative physiology. Crit Care Clin, 18, 249–72. Hartling, L., Bellemare, S., Wiebe, N., et al. (2006). Oral versus intravenous rehydration for treating dehydration due to gastroenteritis in children. Cochrane Database Syst Rev, 3, CD004390. Hensen, J., Henig, A., Fahlbusch, R., et al. (1999). Prevalence, predictors and patterns of postoperative polyuria and hyponatraemia in the immediate course after transsphenoidal surgery for pituitary adenomas. Clin Endocrinol, 50, 431–9. Hoorn, E. J., Betjes, M. G., Weigel, J., et al. (2008). Hypernatraemia in critically ill patients: too little water and too much salt. Nephrol Dial Transplant, 23, 1562–8. Hoorn, E. J., Carlotti, A. P., Costa, L. A., et al. (2007). Preventing a drop in effective plasma osmolality to minimize the likelihood of cerebral edema during treatment of children with diabetic ketoacidosis. J Pediatr, 150, 467–73. Hoorn, E. J. and Zietse, R. (2007). Combined renal tubular acidosis and diabetes insipidus in hematological disease. Nat Clin Pract Nephrol, 3, 171–5.

Hoorn, E. J. and Zietse, R. (2010). Water balance disorders after neurosurgery: the triphasic response revisited. NDT Plus, 3, 42–4. Hoorn, E. J., Zillikens, M. C., Pols, H. A., et al. (2009). Osmomediated natriuresis in humans: the role of vasopressin and tubular calcium sensing. Nephrology, Dial, Transplant, 24, 3326–33. Kettaneh, A., Fardet, L., Mario, N., et al. (2010). The 2003 heat wave in France: hydration status changes in older inpatients. Eur J Epidemiol, 25, 517–24. Kitabchi, A. E. and Nyenwe, E. A. (2006). Hyperglycemic crises in diabetes mellitus: diabetic ketoacidosis and hyperglycemic hyperosmolar state. Endocrinol Metab Clin North Am, 35, 725–51, viii. Knoers, N. (1993). Nephrogenic diabetes insipidus. In R. A. Pagon, T. D. Bird, C. R. Dolan, et al. (eds.) GeneReviews. [Online] Seattle, WA: University of Washington. Kobayashi, T., Miwa, T., and Odawara, M. (2011). A case of central diabetes insipidus following probable type A/H1N1 influenza infection. Endocr J, 58(10), 913–8. Konetzny, G., Bucher, H. U., and Arlettaz, R. (2009). Prevention of hypernatraemic dehydration in breastfed newborn infants by daily weighing. Eur J Pediatr, 168, 815–8. Kortenoeven, M. L., Li, Y., Shaw, S., et al. (2009). Amiloride blocks lithium entry through the sodium channel thereby attenuating the resultant nephrogenic diabetes insipidus. Kidney Int, 76, 44–53. Kratz, A., Siegel, A. J., Verbalis, J. G., et al. (2005). Sodium status of collapsed marathon runners. Arch Pathol Lab Med, 129, 227–30. Lang, T., Prinsloo, P., Broughton, A. F., et al. (2002). Effect of low protein concentration on serum sodium measurement: pseudohypernatraemia and pseudonormonatraemia! Ann Clin Biochem, 39, 66–7. Lee, I. W., Su, M. T., Kuo, P. L., et al. (2010). Gestational diabetes and central pontine myelinolysis with quadriplegia: a case report. J Matern Fetal Neonatal Med, 23, 728–31. Leung, C., Chang, W. C., and Yeh, S. J. (2009). Hypernatremic dehydration due to concentrated infant formula: report of two cases. Pediatr Neonatol, 50, 70–3. Lim, W. H., Lien, R., Chiang, M. C., et al. (2011). Hypernatremia and grade III/IV intraventricular hemorrhage among extremely low birth weight infants. J Perinatol, 31, 193–8. Lindner, G., Funk, G. C., Lassnigg, A., et al. (2010). Intensive care-acquired hypernatremia after major cardiothoracic surgery is associated with increased mortality. Intensive Care Med, 36, 1718–23. Lindner, G., Kneidinger, N., Holzinger, U., et al. (2009). Tonicity balance in patients with hypernatremia acquired in the intensive care unit. Am J Kidney Dis, 54, 674–9. Lindner, G., Schwarz, C., and Funk, G. C. (2011). Osmotic diuresis due to urea as the cause of hypernatraemia in critically ill patients. Nephrol Dial Transplant, 27(3), 962–7. Lindner, G., Schwarz, C., Kneidinger, N., et al. (2008). Can we really predict the change in serum sodium levels? An analysis of currently proposed formulae in hypernatraemic patients. Nephrol Dial Transplant, 23, 3501–8. Marples, D., Frokiaer, J., Dorup, J., et al. (1996). Hypokalemia-induced downregulation of aquaporin-2 water channel expression in rat kidney medulla and cortex. J Clin Invest, 97, 1960–8. McDonnell, C. M., Pedreira, C. C., Vadamalayan, B., et al. (2005). Diabetic ketoacidosis, hyperosmolarity and hypernatremia: are high-carbohydrate drinks worsening initial presentation? Pediatr Diabet, 6, 90–4. McGee, S., Abernethy, W. B., 3rd, and Simel, D. L. (1999). The rational clinical examination. Is this patient hypovolemic? JAMA, 281, 1022–9. Molaschi, M., Ponzetto, M., Massaia, M., et al. (1997). Hypernatremic dehydration in the elderly on admission to hospital. J Nutr Health Aging, 1, 156–60. Moritz, M. L., Manole, M. D., Bogen, D. L., et al. (2005). Breastfeeding-associated hypernatremia: are we missing the diagnosis? Pediatrics, 116, e343–7.

chapter 29 

Namdar, T., Stollwerck, P. L., Stang, F. H., et al. (2011). Impact of hypernatremia on burn wound healing: results of an exploratory, retrospective study. Ostomy Wound Manage, 57, 30–4. Nemergut, E. C., Zuo, Z., Jane, J. A., Jr, et al. (2005). Predictors of diabetes insipidus after transsphenoidal surgery: a review of 881 patients. J Neurosurg, 103, 448–54. Nielsen, J., Kwon, T. H., Christensen, B. M., et al. (2008). Dysregulation of renal aquaporins and epithelial sodium channel in lithium-induced nephrogenic diabetes insipidus. Semin Nephrol, 28, 227–44. Nossent, A. Y., Ellenbroek, J. H., Frolich, M., et al. (2010). Plasma levels of von Willebrand factor, von Willebrand factor propeptide and factor VIII in carriers and patients with nephrogenic diabetes insipidus. Thromb Res, 125, 554–6. Nossent, A. Y., Vos, H. L., Rosendaal, F. R., et al. (2008). Aquaporin 2 gene variations, risk of venous thrombosis and plasma levels of von Willebrand factor and factor VIII. Haematologica, 93, 959–60. O’Connor, K. A., Cotter, P. E., Kingston, M., et al. (2006). The pattern of plasma sodium abnormalities in an acute elderly care ward: a cross-sectional study. Irish J Med Sci, 175, 28–31. Olson, B. R., Gumowski, J., Rubino, D., et al. (1997). Pathophysiology of hyponatremia after transsphenoidal pituitary surgery. J Neurosurg, 87, 499–507. Ostermann, M., Dickie, H., Tovey, L., et al. (2010). Management of sodium disorders during continuous haemofiltration. Crit Care, 14, 418. Palevsky, P. M., Bhagrath, R., and Greenberg, A. (1996). Hypernatremia in hospitalized patients. Ann Internal Med, 124, 197–203. Passare, G., Viitanen, M., Torring, O., et al. (2004). Sodium and potassium disturbances in the elderly: prevalence and association with drug use. Clin Drug Invest, 24, 535–44. Phillips, P. A., Bretherton, M., Johnston, C. I., et al. (1991). Reduced osmotic thirst in healthy elderly men. Am J Physiol, 261, R166–71. Polderman, K. H., Schreuder, W. O., Strack Van Schijndel, R. J., et al. (1999). Hypernatremia in the intensive care unit: an indicator of quality of care? Crit Care Med, 27, 1105–8. Pruitt, B. A., Jr. (1978). Fluid and electrolyte replacement in the burned patient. Surg Clin North Am, 58, 1291–312. Robertson, G., Carrihill, M., Hatherill, M., et al. (2007). Relationship between fluid management, changes in serum sodium and outcome in

approach to the patient with hypernatremia

hypernatraemia associated with gastroenteritis. J Paediatr Child Health, 43, 291–6. Robertson, G. L., Mahr, E. A., Athar, S., et al. (1973). Development and clinical application of a new method for the radioimmunoassay of arginine vasopressin in human plasma. J Clin Invest, 52, 2340–52. Rose, B. D. (1986). New approach to disturbances in the plasma sodium concentration. Am J Med, 81, 1033–40. Schmitt, S., Wichmann, W., Martin, E., et al. (1993). Pituitary stalk thickening with diabetes insipidus preceding typical manifestations of Langerhans cell histiocytosis in children. Eur J Pediatr, 152, 399–401. Sherer, D. M., Cutler, J., Santoso, P., et al. (2003). Severe hypernatremia after cesarean delivery secondary to transient diabetes insipidus of pregnancy. Obstet Gynecol, 102, 1166–8. Sinha, A., Ball, S., Jenkins, A., et al. (2011). Objective assessment of thirst recovery in patients with adipsic diabetes insipidus. Pituitary, 14(4), 307–11. Smith, D., McKenna, K., Moore, K., et al. (2002). Baroregulation of vasopressin release in adipsic diabetes insipidus. J Clin Endocrinol Metab, 87, 4564–8. Stephen, M. D., Fenwick, R. G., and Brosnan, P. G. (2012). Polyuria and polydipsia in a young child: diagnostic considerations and identification of novel mutation causing familial neurohypophyseal diabetes insipidus. Pituitary, 15 Suppl 1, S1–5. Sterns, R. H. (1999). Hypernatremia in the intensive care unit: instant quality—just add water. Crit Care Med, 27, 1041–2. Unal, S., Arhan, E., Kara, N., et al. (2008). Breast-feeding-associated hypernatremia: retrospective analysis of 169 term newborns. Pediatr Int, 50, 29–34. Verbalis, J. G. (2003). Diabetes insipidus. Rev Endocr Metab Disord, 4, 177–85. Wada, M., Kusuda, S., Takahashi, N., et al. (2008). Fluid and electrolyte balance in extremely preterm infants 3 kg. One study reported that even 7 L of excess ECF excess was not detectable by physical exam (Ferraro et al., 1949). Change after recumbency is a commonly noted feature of oedema. Lesser degrees of oedema may only be noticed in the feet and ankles at the end of the day, with apparent resolution upon

waking, in the morning. While nocturia could explain some of that resolution, it is more likely that the distribution of oedema has shifted overnight from the feet and ankles to the trunk.

Palpation Oedema may feel firm, even resistant to pressure. A brawny texture is felt in lymphoedema. Oedema of the nephrotic syndrome may feel soft and sponge-like. The characteristic of ‘pitting’ is a time-honoured physical finding of oedema (Fig. 30.1). Pressure by pushing with a fingertip on the oedematous area may leave a dent, or ‘pit’ that lasts for more than a few seconds. Pitting does not occur in lymphoedema. In subjects with hypoproteinaemia, the fingertip can create a pit more easily than for oedemas from other causes, and that pit resolves more quickly than the pit of oedema that occurs in other oedematous states such as heart failure (Henry and Altmann, 1978).

Physiology and pathophysiology One can easily understand that oedema formation will depend on imbalances in the forces that determine transcapillary exchange of fluids in the microcirculation. Thus, an excess of arteriolar or venous intracapillary pressure, a deficiency of intracapillary oncotic pressure, or a reduction of lymphatic fluid reabsorption could lead to oedema (Fig. 30.2). These imbalances do not immediately lead to oedema. First, minor imbalances in transcapillary forces may only lead to minor degrees of tissue expansion by ECF. Second, there appear to be compensatory effects that protect against oedema formation. In the case of elevation of venous pressures, those must reach 12 mmHg or more to cause oedema (Aukland, 1984). This depends on the initial transit of fluid from inside the capillaries to the interstitial, extracellular space, which leads to an increase in the extracellular tissue pressure, and a countervailing force on the exit of fluid from inside to outside of the capillary. The same is true for oedemas that may occur when there is hypoproteinaemia. Thus, minor degrees of hypoproteinaemia will not lead to oedema; one needs to reach levels of serum albumin near or below 2.5 g/dL for there is oedema solely on the basis of hypoproteinaemia (Kurnick, 1948). Finally, for any given elevation in venous pressure or reduction in serum proteins, the occurrence of oedema will depend on the total ECF volume as determined by dietary sodium intake. These primary changes in the Starling forces are well-known correlates of oedema. It is worth considering the effect of primary

chapter 30 

Fig. 30.1  Oedema. Photograph showing an indentation or ‘pit’ in the leg of a subject with oedema.

changes in renal sodium transport that cause enhanced sodium reabsorption, and which may result in hypertension. None of these cause oedema. The phenomenon of aldosterone escape is the best studied of these, in which a primary excess of aldosterone may cause hypertension but does not cause oedema. Hormonal or haemodynamic counter-regulatory mechanisms prevent primary aldosterone excess from leading to oedema, the so-called aldosterone escape (Hall et al., 1984). So, too, does enhanced renal sodium reabsorption as occurs in Gordon or Liddle syndromes lead to hypertension but not oedema. This underlines the concept of oedema as one that occurs because of the renal response to persistent arterial underfilling with activation of multiple sodium retentive mechanisms (Anand and Chugh, 1997).

Causes and pathophysiology of oedema The causes of oedema may be inferred from its pathophysiology. Thus, heart failure, either left or right sided, liver failure, and renal disease are the ‘usual suspects’. One’s experience with one or the other cause will vary depending on one’s time and place of practice. Oedema in heart failure is the result of persistent underfilling of the arterial circulation, with renal counter-regulatory sodium reabsorption (Anand and Chugh, 1997). Indeed, in heart failure, the degree of sympathetic and/or renin–angiotensin system activation correlates with the stage of heart failure (Fitzpatrick et al., 1985). This mechanism is valid for left or right ventricular failure. The latter adds the additional element of the effect of elevated central venous pressures to reduce the glomerular filtration rate (GFR) and to promote renal sodium retention (Firth et al., 1988).

Arteriole

Pc Po

Po Pc

Venule

Capillaries

Fig. 30.2  Starling forces. A simplified illustration of the forces that affect transcapillary fluid exchange; the lymphatic circulation is omitted for clarity. The intracapillary capillary hydrostatic pressure (Pc) is higher than the intracapillary oncotic pressure (Po) at the arteriolar side, but Pc is lower than Po at the venular side of the microcirculation. Arteriolar filtration is thus balanced by venular reabsorption. Oedema may result from higher Pc or from lower Po.

approach to the patient with oedema

The oedema of liver failure also results from underfilling, in this case from splanchnic arterial vasodilation and or the simple haemodynamic effect of portal hypertension and limitation to venous return. Secondary activation of the renin–angiotensin–aldosterone system is a prime mover of the sodium retention in cirrhosis as it is in heart failure (Schrier, 2007). To these mechanisms is added the effect of hypoalbuminaemia, a frequent occurrence in cirrhosis. The salt and water retention of severe oliguric renal failure is obvious. In less-severe renal failure, reduction of the GFR is accompanied by elevation of the fractional excretion of sodium, such that in the steady state, sodium retention may not be apparent. But as GFR falls, in progressive decrements, there is a delay in tubular adaptation to that decline, such that progressive sodium retention can result in a cumulative fashion. There is impaired natriuresis in subjects with renal disease, such that there may be accumulation of a sodium load (Johnson et al., 2002). Multiple mediators are implicated in the hypertension of renal failure, including the renin–angiotensin–aldosterone system, nitric oxide deficiency, circulating inhibitors of sodium-potassium (Na-K)-ATPase, and many others. These may contribute to sodium retention and oedema of subjects with renal failure. Much as one overactive sodium transporter leads to hypertension but not oedema, it is likely that several mechanisms coexist to cause oedema in renal failure, rather than is a single one. The association of oedema with hypoproteinaemia of the nephrotic syndrome is well known. Controversy persists as to the mechanisms of nephrotic oedema, specifically the accuracy of the classical construct of proteinuria → hypoproteinaemia → reduced plasma oncotic pressure → accumulation of interstitial oedema and transient underfilling of the circulation → reactive renal sodium retention (Palmer and Alpern, 1997; Schrier and Fassett, 1998). Primary renal sodium retention with an overfilled circulation has also been advocated as a cause of nephrotic oedema. Perhaps the clearest recent explanation of these competing ideas is that of Rodriguez-Iturbe et  al, who state that when the nephrotic syndrome is accompanied by intrarenal inflammation, that inflammation can cause primary renal sodium retention (Rodriguez-Iturbe et al., 2002). This would hold for renal diseases such as diabetic or lupus nephropathy. In those cases, there is ‘overfilling,’ because the sodium retention and oedema are caused by renal sodium retention as well as by the hypoproteinaemia. But in non-inflammatory nephrotic syndrome, particularly minimal change disease, the simpler classical mechanism obtains, starting with proteinuria as stated above. These latter subjects would then be ‘underfilled,’ at least transiently. The oedema of venous diseases is straightforward, being caused by the mechanical effect of elevated venous pressure that is transmitted to the capillaries, leading to transudation of fluid according to the Starling mechanism (Fig. 30.2). The cause of oedema that may be caused by certain medications is less clear. Arteriolar vasodilation is said to a cause of oedema in some subjects who are using minoxidil or amlodipine, but confirmation is lacking. The oedema caused by thiazolidinediones could arise from primary renal sodium retention (Guan et  al., 2005; Panchapakesan et al., 2009)

Testing for and imaging oedema The chest X-ray findings of pulmonary oedema are well known, and ultrasound or computed tomography scan easily detect ascites.

273

274

Section 2  

fluid, electrolyte, and renal tubular disorders

Generally, oedema may be seen on X-ray imaging as tissue expansion, which is hypodense when compared to most tissue densities. This would not allow differentiation of inflammatory from non-inflammatory oedema. Generally, the most important part of testing and investigations of oedematous patients are the history and physical examination.

Differential diagnosis by history and physical examination Findings in left-sided heart failure Oedema in subjects with heart failure will depend on the severity of their heart disease, being generally absent for lesser degrees of heart failure. Classically, left heart failure will cause peripheral oedema only after it has caused pulmonary oedema. The presence of pulmonary oedema correlates with a body fluid excess of about 10% of the total body weight. A lower-than-normal blood pressure and a narrow pulse pressure may be present. There may be a high jugular venous pressure and also a hepatojugular reflux. One may feel the cardiac impulse in two interspaces, rather than just one, and there may be a third heart sound, along with rales or crepitations heard on listening over the lung bases. Oedema in subjects with heart failure is typically apparent in the dependant legs, its extent ascending upwards as the heart failure worsens. It may be perceived as being worse at the end of the day, when the oedema fluid has accumulated in the feet and legs, and ‘better’ in the morning, when that excess fluid has redistributed due to recumbency. There is, however, a lack of good correlation of the presence of oedema and the severity of heart failure (Stevenson and Perloff, 1989), perhaps in part because of variability in dietary sodium intake; subjects adherent to a low sodium diet may not form oedema, even when their ejection fractions are < 20% or when their pulmonary wedge pressures are > 22 mmHg.

Findings in right-sided heart failure

disease, heart failure may add to the tendency to fluid overload, and a single simple cause of oedema may no longer be apparent. When one organ dysfunction is combined with the other, lesser degrees of individual organ damage may cause oedema in combination that would not do so by themselves. But oedema may not be a major feature of kidney disease. This was recognized over a century ago (Richet, 1993). Kidney diseases with no or minimal proteinuria, such as polycystic kidney disease or interstitial nephritis, may show substantial azotaemia, but little or no oedema.

Findings in liver disease Oedema due to liver disease generally occurs only in those with severe liver disease that is clinically apparent. Jaundice, temporal wasting, and palmar erythema will be present; ascites typically precedes the formation of peripheral oedema. Yet, these patients may have superimposed heart or kidney disease, which may cause oedema not due to the liver disease.

Findings in starvation Starvation alone should be evident by weight loss, and cachexia. Oedema of starvation is now quite rare in the developed world, but it was well documented during World War II. In one report from that time, the median total protein level in the serum was 4.5–5 g/ dL in subjects with oedema, whereas non-oedematous subjects had total serum protein levels > 6.5 g/dL (Kurnick, 1948). Protein-losing enteropathy could cause a similar picture.

Findings in venous disease Venous disease that causes oedema is usually accompanied by varicosities that are easily seen. In the absence of that finding, it is risky to ascribe oedema to venous disease alone. Moreover, the oedema of venous disease is usually asymmetric, a cardinal sign that is not the case for other diseases.

Oedema in subjects with right heart failure is the syndrome of cor pulmonale. It depends on severe lung disease, usually obstructive airways disease, although it could also be caused by pulmonary embolism. In classical cor pulmonale, there will be findings of emphysema or chronic bronchitis, and a loud pulmonic second heart sound. These patients will have jugular vein distension and positive hepatojugular reflux with hepatic enlargement accompanying peripheral oedema. Ascites may also be present. Here, as in left-sided failure, the primary stimulus to renal salt and water retention is arterial underfilling, which in the case of cor pulmonale has been attributed to hypercarbia and arteriolar dilation. Elevated venous pressures may also contribute, via their effect to lower the GFR and increase renal sodium retention.

Treatment of oedema

Findings in kidney disease

Diet

Oedema in subjects with kidney diseases is classically worse in the morning. In subjects with nephrotic syndrome, this may be because the bodily fluid overload is most evident in the loose connective tissue around the eyes, where it can easily accumulate after recumbency. Oedema of the nephrotic syndrome may also be dependent, but will ‘pit’ more easily than will the pressure-dependent oedema of heart failure. In subjects with oedema that is solely due to kidney disease, findings of heart failure will be absent. With any kidney

Since ECF volume varies directly with sodium intake, dietary sodium restriction cannot be ignored. For patients with mild ECF volume expansion, a ‘no added salt’ diet may be appropriate, which is approximately 4 g of sodium/day, or approximately 160 mmol of sodium. This involves not adding salt during cooking and having no use of salt at the dining table. It is essential to avoid foods like potato chips, salted peanuts, or processed foods such as manufactured and cured meats. For more severe oedema, a very low sodium

Because oedema is a condition of sodium excess, restricting dietary sodium intake and using diuretics is the cornerstone of treatment. For generalized oedema, reversal of the underlying disorder will resolve the renal sodium retention. The latter is either an exaggerated response to a low effective arterial volume or a response of kidney tubules to damage. Pulmonary oedema is the only form of generalized oedema that is life threatening and requires immediate intravenous therapy within minutes. In all other oedematous states, removal of the excess fluid can proceed more slowly, over days to weeks. This is particularly true in patients with cirrhosis, who are at risk for hepatic coma or hepatorenal syndrome if they have a rapid diuresis.

chapter 30 

chloride (NaCl) diet (2 g/day) should be prescribed. Water restriction per se is not needed, unless there is hyponatraemia. Dietary compliance with sodium restriction can be checked by testing a 24-hour urine collection for sodium.

Medications Prescription or over-the-counter medication may predispose to sodium retention or interfere with diuretic efficacy. Non-steroidal anti-inflammatory drugs (NSAIDs) promote renal sodium retention by reducing GFR and interfere with the efficacy of diuretics by competitively inhibiting the transport system of these diuretics at the proximal tubule, decreasing their concentration at their intratubular site of action. Vasodilators (minoxidil, hydralazine) can stimulate the sympathetic nervous system and renin-angiotensin system and thiazolidinediones promote ECF volume expansion by enhancing distal tubular sodium reabsorption. Dihydropyridine calcium channel blockers may induce dependant oedema through arteriolar vasodilation and increased intracapillary pressure.

Mechanical treatments Elevation of the legs by placing them to above the level of the heart for 10–15 minutes, three to four times a day, stimulates interstitial fluid re-entry into the circulatory system, probably by reducing the venous pressures (Fig. 30.2). This can be useful in combination with sodium restriction in mild oedema. Sitting for long periods will increase swelling in the feet and ankles. Standing and/or walking at least every hour or two will help stimulate blood flow and reduce oedema formation. Use of compression elastic stockings will compress the leg vessels, promoting circulation and decreasing pooling of fluid due to gravity. In the case of oedema in preeclampsia, bed-rest is recommended. It will improve renal blood flow, reduce activity of the sympathetic nervous system and renin–angiotensin–aldosterone system, and thus mobilize oedema fluid from interstitial to intravascular space. Massaging the legs can help to stimulate the release of excess fluids, but should be avoided if the patient has blood clots in the veins. A successful natriuretic response to head-out water immersion has been seen in children suffering from severe nephrotic syndrome caused by minimal change disease (Rascher et  al., 1986). This technique can be tried in patients who have the hepatorenal syndrome. Repeated sessions of head-out water immersion could improve the sodium and water excretion and the renal function in hepatorenal syndrome (Yersin et al., 1995). Aquapheresis, also known as ultrafiltration (UF), is a technique for removing excess fluid from the body. It involves the placement of a catheter in the bloodstream that continuously runs the patient’s blood through a filter. Excess plasma water and electrolytes are removed from the blood through this filter, and the blood is then returned to the patient. This can be very useful in the treatment of refractory fluid overload in subjects with heart failure (Costanzo et al., 2007). But a recent randomized trial showed that diuretics alone were better and caused fewer side effects compared to UF in heart failure patients who had renal impairment (Bart et  al., 2012). In subjects with heart failure whose serum creatinine is > 200 µmol/L, haemodialysis is needed rather than mere UF. In refractory oedema due to severe nephrotic syndromes, isolated UF and hemofiltration have been tried with some success. In cirrhotic patients, re-infusion of ascites and paracentesis with albumin infusion may help in managing ascites refractory to diuretics.

approach to the patient with oedema

Hemofiltration is also useful to control ascitic fluid and oedema (Davenport, 2001). The oedematous patient with severe renal failure may need to start regular dialysis treatments, which are very effective in treating fluid overload.

Diuretics Once the decision to initiate diuretic therapy has been made, the initial choice of drug and dosage depends on the underlying cause of oedema and its severity (Brater, 1998).

Refilling phenomena When diuretics are given, the fluid that is lost initially comes from the intravascular space. This results in a reduction in capillary hydraulic pressure; the plasma volume will be refilled by the mobilization of extracapillary oedema fluid into the intravascular space according to the familiar Starling forces. The speed and magnitude of this refilling is variable. In patients with generalized oedema due to heart failure, nephrotic syndrome, or primary kidney sodium retention, the oedema fluid can be easily and quickly mobilized, since most capillary beds are involved. Thus, in patients with generalized oedema a diuresis of 1–2 L of oedema fluid or more in 24 hours can usually be achieved without an adverse reduction in plasma volume. But in other patients, the decrease in effective circulating arterial volume is sufficient to significantly impair tissue perfusion. This most often occurs in two settings: when there is a low effective arterial circulating volume, as in heart failure with hypotension and after rapid fluid removal in cirrhosis (Pockros and Reynolds, 1986). When a cirrhotic patient has ascites but no peripheral oedema, refilling of the intravascular volume will depend on the peritoneal capillary bed alone, rather than on the capillaries and interstitial fluid of the whole body. In these cirrhotics, the maximum amount of fluid that can be mobilized is about 500 mL/day. If the diuresis proceeds more quickly than that, the ascitic fluid will be unable to completely replenish the plasma volume, resulting in azotaemia and possible precipitation of the hepatorenal syndrome. But in cirrhotics who also have peripheral oedema, the rate of diuresis can be more than 1 kg/day.

Pharmacology of diuretics (See Chapter 33.) For moderate to massive oedema (> 10% excess ECF volume), loop diuretics are most used, because they are the most potent diuretics and they act quickly. They are very useful for oedema caused by heart failure, cirrhosis, or nephrotic syndrome, or when renal failure is present. Loop diuretics can be given orally or intravenously. The onset of diuresis is earlier and the peak diuresis is greater with intravenous therapy because of greater and more rapid bioavailability. There are no significant differences in efficacy between the different loop diuretics (furosemide, bumetanide, and torsemide) if they are given at equipotent doses (40 mg, 1 mg, and 10 mg, respectively). Each is active at all levels of renal function, and act quickly, even following oral administration. They have a sigmoidal dose–response curve characterized by a threshold drug dose to induce a diuresis, an ascending portion of the curve in which increased diuretic dose causes more and more sodium excretion, and a plateau at which further elevations in diuretic dose do not add to the diuresis

275

Section 2  

fluid, electrolyte, and renal tubular disorders

A Dose Bioavailability Tubular secretory capacity Rate of absorption Time course of delivery

y

Maximal response

Effi cie nc

Sodium excretion rate

276

Altered dose–response relationship Braking phenomenon

Threshold

B

Diuretic excretion rate

Fig. 30.3  Diuretic dose–response curve. Urinary sodium excretion according to loop diuretic use in normal conditions and when there is generalized oedema. The diuretic threshold depends on the dose, bioavailability, and tubular secretory capacity and the plateau on the braking phenomenon. CHF = congestive heart failure; HC = hepatic cirrhosis; NS = nephrotic syndrome; CRI = chronic renal insufficiency. Adapted from NKF KDOQI clinical practice guidelines on HTA in CKD 2002 (guideline 12, Fig 56, reproduced with permission).

(Fig. 30.3). In subjects with normal renal function, a diuresis begins with 10 mg of intravenous furosemide given as a bolus with the maximal effect being seen with 40 mg. Going above this maximum will produce little or no further diuresis but may increase the risk of side effects. The doses that give this plateau are higher when either acute or chronic renal failure is present. For intravenous administration of furosemide the plateau dose could increase from 40 mg in subjects with normal renal function to 200 mg when the GFR is < 20 mL/min (Table 30.1) (Brater, 1998; Wilcox, 2002). For bumetanide, these doses range from 1 mg for subjects with normal renal function to 10 mg in subjects with severe renal failure (GFR < 20 mL/min). The plasma half-life of loop diuretics is fairly short. Thus, in oedematous patients with normal renal function it is better to give these diuretics in multiple daily doses to counter the post-diuretic sodium retention. Kidney or liver disease needs to be considered. The half-life of furosemide is increased in renal failure. The half-lives of bumetanide and torsemide are increased when cirrhosis is present. But the efficacy of loop diuretics in patients with liver disease may be reduced due to competition at the secretion site at the proximal tubule between loop diuretics and bile salts.

Loop diuretics can be given orally, although seriously ill patients in the hospital or resistant to oral treatment may need them intravenously for a quicker response. The data on the maximally effective loop diuretic dose are based upon intravenous bolus therapy in patients with a relatively normal GFR and vary with the underlying disease (Brater, 1998). The doses required in congestive heart failure, cirrhosis, or nephrotic syndrome are shown in Table 30.1. The intravenous equivalent dose for furosemide, but not for the other loop diuretics, is one-half the oral dose, because of its decreased oral availability. The effective diuretic dose is typically higher in patients with New  York Heart Association (NYHA) stage III or IV heart failure, advanced cirrhosis, or renal failure. In these settings, decreased renal perfusion (and therefore decreased drug delivery to the kidney), diminished drug secretion into the lumen (due to the retention of competing anions in renal failure), and enhanced activity of sodium-retaining forces (such as the renin–angiotensin–aldosterone system and sympathetic nervous activity) combine to diminish the diuretic effect. The diuretic dose–response curve is shifted to the right and also shows a diminished maximal plateau (Fig. 30.3). Selected hospitalized patients may benefit from a continuous intravenous infusion of a loop diuretic after an initial bolus dose, which can produce a greater diuresis than intravenous boluses (Table 30.2). It may also be safer, and less apt to cause ototoxicity. The enhanced diuresis with a continuous infusion compared with bolus therapy is related to continuous inhibition of sodium chloride reabsorption in the loop of Henle. In contrast, bolus therapy is first associated with initially higher and then followed by lower rates of diuretic excretion; as a result, sodium excretion may be at near maximal levels for the first 2 hours but then gradually fall until the next dose is given. If the patient has received one or more intravenous boluses within the previous few hours, then an infusion can be started without a loading dose. Otherwise, an intravenous loading dose of 40–80 mg of furosemide is given over 5 minutes. After the loading dose, the starting infusion rate with furosemide varies with the level of renal function. When the renal function is normal, an initial furosemide infusion rate of 5 mg/hour could be used. If the diuresis is not sustained, a second bolus can be given followed by a higher infusion rate of 10 mg/hour. But when there is substantial reduction in kidney function, it is suggested that after

Table 30.1  Ceiling doses for loop diuretics in mg/day Medical conditions

Furosemide (mg)

Bumetanide (mg)

Torsemide (mg)

Intravenous

Oral

Intravenous

Oral

Intravenous

Oral

GFR > 50 mL/min

20–40

40–80

1–2

1–2

10–20

10–20

GFR 20–50 mL/min

120

240

3

3

50

50

GFR < 20 mL/min

200

400

10

10

100

100

Nephrotic syndrome 40–60

80–120

2–3

2–3

20–50

20–50

Cirrhosis with oedema

40

40

1

1

10

10

Congestive heart failure

40–80

40–80

1–2

1–2

10–20

10–20

Adapted from Brater (1998) and Wilcox (2002).

chapter 30 

approach to the patient with oedema

Table 30.2  Doses for continuous intravenous loop diuretic infusion Diuretics

Intravenous infusion (mg/h)a

Intravenous bolus dose (mg) GFR > 75 mL/min

GFR 25–75 mL/min

GFR < 25 mL/min

Furosemide

40

5 then 10

10 then 20

20 then 40

Bumetanide

1

0.5

0.5 then 1

1 then 2

Torsemide

20

5

5 then 10

10 then 20

a Before increasing the infusion speed, one needs first to inject a new IV bolus.

Adapted from Brater (1998) and Wilcox (2003).

the intravenous loading dose of 40–80 mg of furosemide, the initial furosemide infusion rate should be 20 mg/hour. If the diuresis is not sustained, a second bolus can be given followed by a higher infusion rate of 40 mg/hour. The equivalent bumetanide and torsemide dose is 1 mg/hour, increasing to 2 mg/hour for bumetanide and 10 mg/hour, increasing to 20 mg/hour for torsemide (Table 30.2). The vigorous diuresis produced by loop diuretics makes these especially useful for rapid reduction of oedematous fluid. However, other diuretics exist. Thiazide diuretics decrease active re-absorption of sodium and chloride ions by inhibiting the sodium/chloride cotransporter in the distal convoluted tubule. They also increase potassium ion loss. These were originally synthesized as carbonic anhydrase inhibitors. Thiazide diuretics, like loop diuretics, are secreted into the tubular fluid by proximal tubule cells, then act at the distal convoluted tubule. While some thiazides have carbonic anhydrase inhibitory activity, the major site of action is to reduce tubular sodium reabsorption. Hydrochlorothiazide is the prototype for this class of drug. chlorthalidone, indapamide, and metolazone are long acting congeners. These drugs do not have the thiazide structure but are referred to as ‘thiazide-like’. The clinically available thiazides and thiazide-like agents have the same mechanism of action. They differ only in their plasma half-lives. Thiazides can be used to treat oedema associated with congestive heart, cirrhosis, renal insufficiency, and the nephrotic syndrome. Oral doses of hydrochlorothiazide are well absorbed and reach peak effect in about 4 hours, and have a 6–12-hour duration of action. It is excreted unchanged in the urine with a half-life of 3–5 hours. The usual adult dose for oedema treatment is 25–100 mg per day, depending on patient response. Hydrochlorothiazide is not metabolized but is eliminated rapidly by the kidneys. At least 61% of the oral dose is eliminated unchanged within 24 hours. The likelihood of side effects with the thiazides increases with the dose of the drug. These include hypokalaemia, hyponatraemia, hypochloraemia, hypercalcaemia, hypomagnesaemia, and hyperuricaemia. The average fall in serum potassium with use of thiazides is 0.6 mmol/L (Morgan and Davidson, 1980); a decrease in glucose tolerance can also be observed. Warning signs of fluid and electrolyte imbalance include dry mouth, thirst, weakness, lethargy, drowsiness, restlessness, muscle pains or cramps, muscular fatigue, hypotension, oliguria, tachycardia, nausea, vomiting, seizures, or confusion. Hypokalaemia may be prevented or treated with potassium-rich foods, potassium supplements, or a potassium-sparing diuretic. ‘Dilutional hyponatraemia’ (see Chapter  28) most commonly

occurs during hot weather in patients with chronic congestive heart failure or hepatic disease. Treatment includes withdrawal of the diuretic, fluid restriction, and potassium and/or magnesium supplementation. Administration of sodium chloride is usually not required, except in rare instances when the hyponatraemia is life-threatening. Caution is necessary when using thiazides in patients with hyperuricaemia or a history of gout. The routine use of thiazide diuretics in an otherwise healthy pregnant woman with oedema is not appropriate. Thiazides cross the placenta and possible risks include fetal or neonatal jaundice, thrombocytopaenia, and possibly other adverse reactions that have occurred in the adult. Metolazone is a thiazide-like diuretic used primarily to treat congestive heart failure and sometimes used together with loop diuretics. Approximately 65% of the amount ingested becomes available in the bloodstream. Its half-life is approximately 14 hours, similar to indapamide but considerably longer than hydrochlorothiazide. Metolazone is around ten times as potent as hydrochlorothiazide. Although most thiazide diuretics lose their effectiveness in renal failure, metolazone remains active even when the GFR is < 30–40 mL/min. This gives it a considerable advantage over other thiazide diuretics, since renal and heart failure often coexist and contribute to fluid retention. For hypokalaemia resulting from thiazide or loop diuretics, the potassium loss can be mitigated with potassium supplementation (potassium salts or foods rich in potassium) or the use of potassium-sparing diuretics. There are two types of potassium-sparing diuretics: the renal epithelial sodium channel inhibitors amiloride and triamterene or the aldosterone antagonists spironolactone and eplerenone (see Chapter 21). Amiloride and triamterene block the epithelial Na+ channel. As a result, the driving force for K+ secretion is eliminated and K+ secretion is reduced; their diuretic effect is modest. Aldosterone interacts with a cytoplasmic mineralocorticoid receptor to enhance the expression of the Na-K-ATPase and the Na+ (and K+) channel in the distal tubule; spironolactone and eplerenone block aldosterone by binding to the mineralocorticoid receptor. Their onset and duration of action are determined by the kinetics of the aldosterone response in the target tissue. Overall, spironolactone has a slow onset of action, requiring several days before its full effect is achieved. It decreases the reabsorption of sodium and water, while decreasing the secretion of potassium. This diuretic is useful for oedema treatment in cirrhosis or in combination therapy with loop diuretics in patients with refractory oedema, for instance in congestive heart failure. For nephrotic patients it can be used when treatment of the underlying disease, restriction of fluid and sodium

277

278

Section 2  

fluid, electrolyte, and renal tubular disorders

intake, and the use of other diuretics do not provide an adequate response. Eplerenone has a lower affinity for the mineralocorticoid receptor compared with spironolactone. Nonetheless, it also blocks aldosterone-induced gene expression. However, eplerenone has little affinity for androgen or progesterone receptors. Therefore, it will not cause steroid hormone-like effects of spironolactone such as gynaecomastia, or hair growth. This diuretic, like spironolactone, has been shown to improve outcomes in patients with heart failure. The approach to diuretic therapy will vary according to the condition causing oedema.

Specific oedematous conditions Cirrhosis For patients with cirrhosis, the aldosterone antagonist spironolactone is the preferred initial regimen, based on the role for aldosterone in the fluid retention of cirrhosis, and also to avoid hypokalaemia. One can start at 25 mg/day but this dose may need to increase to 200 mg/day for optimal benefit. Spironolactone is combined with a loop diuretic when there is peripheral oedema in addition to the ascites. As mentioned above, the diuresis should not exceed 0.5 kg/day in the absence of oedema. When there is leg oedema, one can achieve a weight loss of 1 kg/day. In patients suffering from cirrhosis with large ascites and marked abdominal distension, paracentesis can be done in addition to low sodium diet and diuretics. Up to 10 L of fluid may be drained during the procedure. If fluid drainage is more than 5 L, patients should receive intravenous serum albumin right afterwards to prevent low blood pressure. As mentioned previously, head-out water immersion could be useful in some patients.

Heart failure For patients with heart failure, the rate of diuresis is usually not a limiting issue, but one must monitor for signs of hypoperfusion (e.g. a rise in serum creatinine). In NYHA class II congestive heart failure, in addition to use of an angiotensin-converting enzyme (ACE) inhibitor and low sodium diet, thiazide diuretics may be enough to resolve oedema and ECF overload. But loop diuretics are typically used as first-line therapy in more severe heart failure (NYHA classes III and IV) along with the use of ACE inhibitors. This combination is needed as shown by Anand et al. (1990). The dose–response curves of these loop diuretics are shifted rightward because of delayed gut absorption and diminished nephron response (Brater et al., 1984) (Fig. 30.3). Rather than increasing the dose, it is sometimes more useful to give the loop diuretic several times a day to improve sodium excretion. This is in part due to increased tubular sodium reabsorption at different levels of the renal tubule (Loon et al., 1989; Rose, 1991; Wald et al., 1991) and also because of the half-life of loop diuretics such as furosemide. If worsening heart failure develops, there is even greater activation of the sympathetic nervous system and renin angiotensin aldosterone system. Escalation of the total daily dose of loop diuretics is then needed to overcome these stimuli to sodium retention. In congestive heart failure hyponatraemia may occur because of non-osmotic arginine vasopressin (AVP) release. In these cases, vasopressin V2 receptor antagonists have been used and may attenuate water retention (Gheorghiade et  al., 2007). However, long-term mortality and heart failure-related morbidity are

not improved by use of vasopressin antagonists in heart failure (Konstam et al., 2007).

Nephrotic syndrome For patients with nephrotic syndrome, diuretic treatment is needed to treat oedema unless and until immunosuppression and/or renin–angiotensin system antagonism are effective. In the presence of normal renal function, diuretics will be needed such as furosemide (initial oral dose 1 mg/kg/day) given in two or three separate doses. The combination with spironolactone or amiloride will help when fluid retention is severe, but should be avoided in subjects with hyperkalaemia or hypotension (Deschenes et al., 2004). Thiazide diuretics may be added in some cases, because they act on a different site in the nephron (Garin, 1987; Fliser et al., 1994). However, such combined diuretic use may cause volume depletion, which should be carefully monitored for by assessment of symptoms, weight, heart rate, upright blood pressure, and laboratory testing. Use of albumin has been much discussed for the treatment of oedema in the nephrotic syndrome. Its use is based on the premise that raising the serum albumin will ‘pull’ fluid from the extravascular to the intravascular space. Albumin may also increase diuretic delivery to the kidney by keeping furosemide within the vascular space, decreasing its catabolism and facilitating its secretion in the tubule lumen. Intravenous albumin at 1 g/kg can be given, followed by intravenous furosemide. But hypertension can occur in almost half of the patients treated, and respiratory and cardiac failure can develop (Dorhout Mees, 1996; Reid et  al., 1996). Moreover it is reported that albumin may delay the response of nephrotic syndrome to steroids and may even induce more frequent nephrotic relapses, perhaps by causing severe glomerular epithelial damage (Yoshimura et al., 1992). For children affected by nephrotic syndrome, the use of albumin for severe oedema may be analysed according to the two hypotheses proposed for the pathogenesis of nephrotic oedema. According to the underfill hypothesis, severe hypoalbuminaemia decreases intravascular oncotic pressure, leading to circulatory volume depletion and subsequent salt and water retention. In this condition, albumin infusion could be useful. On the other hand, the overfill mechanism proposes a primary renal defect in sodium excretion leading to salt and water retention and thereby hypervolaemia and oedema. These different mechanisms for oedema would demand different therapies, and Kapur et al recently reported this in a series of 30 children with nephrotic syndrome (Kapur et al., 2009). They showed that the fractional excretion of sodium could be used, along with other indicators of intravascular volume, to differentiate nephrotics who were volume contracted from those who were volume expanded. When the fractional excretion of sodium was < 0.2%, that is, when the nephrotic subject was volume contracted, intravenous albumin could be used, whereas it was not used in those with volume expansion. In the latter patients, the use of diuretics alone was effective and safe. One randomized trial did show an increased urine volume with co-administration of albumin and furosemide, as compared with albumin administration alone or furosemide use alone, in 10 nephrotic patients with normal renal function. However, the 24-hour urinary sodium levels were not different between those on furosemide alone compared to those on furosemide in combination with albumin (208 vs 206 mmol) (Ghafari et al., 2011).

chapter 30 

This confirms what was published more than10 years ago when Fliser et al showed that adding albumin to furosemide produced only a modest increase in sodium excretion compared with furosemide alone (Fliser et al., 1999), roughly equivalent to the amount of sodium contained in the colloid solution, suggesting that volume expansion was a likely explanation for the enhanced natriuresis. A similar lack of efficacy of furosemide plus albumin infusion has been shown in hypoalbuminaemic patients with cirrhosis (Chalasani et al., 2001)

Renal insufficiency Although the oral bioavailability of loop diuretics is the same in renal insufficiency as in normal subjects, the dose of loop diuretic sufficient to cause a diuresis must be higher in subjects with lower GFR. That is because the diuretic gets to its tubular site of action through tubular secretion, and that tubular function is impaired in renal insufficiency. The largest single dose of a loop diuretic giving the maximum natriuretic response in subjects with severe renal insufficiency is an intravenous bolus of 200 mg of furosemide or the equivalent of bumetanide or torsemide (Table 30.1). Some patients may require these large doses several times a day. Only side effects are gained by using larger doses. The maximum response is the excretion of about 20% of the filtered sodium. For furosemide, the usual maximal oral dose (the plateau dose) is twice the intravenous dose; this is not the case for bumetanide and torsemide for which the intravenous and oral doses are similar. In addition, the absorption of furosemide is not always 50% and can vary from a patient to another (between 10% and 100%); if there is some resistance in oedema treatment with this drug, one should use higher doses or change to another loop diuretic. In patients who have poor responses to intermittent doses of a loop diuretic, a continuous intravenous infusion can also be tried. For the latter, an initial loading dose is necessary to decrease the time needed to achieve drug concentrations at the intratubular site of action. A  continuous infusion of bumetanide (Wilcox, 2002) enables a greater loss of sodium than the same total dose given as divided intravenous injections. The rate of the continuous injection is governed by the patient’s renal function (Table 30.2). The addition of thiazides to loop diuretics can be considered in oedematous subjects with renal failure, even though thiazides are commonly held to be ineffective when the GFR is < 50 mL/minute. That is because they act at a different site in the nephron, at the NaCl transporter in the distal tubule, thus potentiating the effect of loop diuretics that act on the Na-K-2Cl cotransporter just proximal to that site.

Idiopathic oedema For patients with idiopathic oedema, if a diuretic needs to be prescribed, spironolactone is the drug of choice, because of the secondary hyperaldosteronism that can be found in these patients. But many patients with idiopathic oedema are actually already taking diuretics when first seen and may have ‘diuretic-induced oedema’. So, the initial approach is to try to discontinue the diuretic treatment for at least 2–3 weeks, after warning the patient that the oedema will probably worsen initially and reassuring her (it more usually affects women) that the diuretic can always be restarted. If the oedema does not improve after 4 weeks, spironolactone can then be initiated at a dose of 50–100 mg daily. When this oedema

approach to the patient with oedema

develops during the menstrual period, high capillary permeability is a possible mechanism and ephedrine could be tried (Edwards and Hudson, 1991).

Resistant oedema For patients with resistant oedema from any cause, a complete review and assessment are needed. First, a high salt intake should be looked for because it can prevent net fluid loss even if there is an adequate urine output. To estimate NaCl intake, a 24-hour urine should be collected. A  sodium value > 100  mmol/day suggests that non-compliance with sodium restriction is responsible for the apparent resistance to therapy. Use of other medication should be considered; NSAIDs should be discontinued, since diminished synthesis of vasodilator and natriuretic prostaglandins can impair diuretic responsiveness. Posture is another factor that can influence diuretic responsiveness. Patients with heart failure, for example, are unable to increase cardiac output appropriately when upright. As a result, renal perfusion is reduced and urinary diuretic delivery diminished. Intermittent supine rest could improve the renal blood flow and the sodium and water excretion (Karnad and Abraham, 1986; Ring-Larsen et al., 1986) Two additional factors can contribute to oedema that is refractory to usual oral or intravenous loop diuretic therapy:  either decreased diuretic secretion into the tubular lumen or increased tubular sodium reabsorption. Loop diuretics are highly (≥ 95%) protein bound. As a result, they primarily enter the tubular lumen by secretion in the proximal tubule, not by glomerular filtration. Diuretic resistance can result from decreased renal perfusion in acute kidney injury from any cause, from heart failure due to the reduced cardiac output, or from cirrhosis due to renal vasoconstriction. In renal failure, acute or chronic, uraemic metabolites and metabolic acidosis may inhibit tubular secretion that may also cause diuretic resistance. Decreased diuretic secretion may occur in two ways in the nephrotic syndrome. With hypoalbuminaemia, there is less delivery of the albumin-bound loop diuretic to the renal tubular epithelial cell for secretion and subsequent action. In addition, binding of the diuretic by filtered albumin within the tubular lumen may reduce loop diuretic efficacy; resistance may be overcome by increasing the dose of the drug. A variety of other factors can account for persistent fluid retention in subjects prescribed diuretics, including variability of furosemide absorption at the intestinal site (substitution of bumetanide may obviate this problem), or delayed intestinal absorption of any diuretics when there is intestinal wall oedema (the use of the intravenous route could bypass this problem). It is worth again noting that diuretics have a dose–response curve, with no natriuresis seen until a threshold rate of drug excretion is attained. If, for example, a patient does not respond to 40 mg of furosemide, the dose may not have exceeded this threshold. Thus, the single dose should be increased to 60 or 80 mg, rather than giving the same dose twice a day. Loop diuretics have short half-lives. In the situation of active sodium retention (i.e. cardiac failure), loop diuretics must be given more frequently. Indeed, as the effect of loop diuretics wanes, the kidneys immediately begin to reabsorb sodium, which attenuates the diuretic effect. This process is called post-diuretic sodium chloride retention or diuretic braking phenomenon.

279

280

Section 2  

fluid, electrolyte, and renal tubular disorders

Combination therapy

Care of patients treated by diuretics

Loop diuretics may be combined with an agent that works elsewhere in the renal tubule. When diuretics that work at different sites are used together, the response may be synergistic, ithat is, greater than the addition of the single responses to the individual diuretics. Combination therapy may include hydrochlorothiazide or metolazone. Thiazide administration should precede the loop diuretic by 2–5 hours, since the peak effect of the latter is at 4–6 hours. Combination therapy must be given with careful monitoring, since a marked diuresis can occur in which daily sodium and potassium losses can be > 300 and 200 mmol, respectively (Oster et al., 1983; Fliser et al., 1994). This could lead to an excessive loss of potassium in the urine, leading to hypokalaemia and depletion of body potassium. These patients should take potassium supplements and/or to eat foods high in potassium. In certain instances, diuresis can be improved by adding a potassium-sparing diuretic, that is, one that does not cause depletion of potassium. Those diuretics include spironolactone, eplerenone, triamterene, and amiloride. Adding one of them may preclude the need for potassium supplements. But neither potassium-sparing diuretics nor potassium supplements are advisable in subjects with reduced kidney function. One can consider adding acetazolamide, which acts to reduce proximal tubular sodium and bicarbonate reabsorption, but that could induce a renal tubular acidosis. To combat diuretic resistance, we thus advise restriction of sodium intake, use of escalating doses of loop diuretics up to ceiling levels, and adroit use of a second diuretic acting at a downstream site. In nephrotic syndrome, measures to reduce the proteinuria can also be very useful. UF is a last resort.

Hypokalaemia may develop in half of all patients treated with thiazide diuretics, and require supplementation with potassium. Hypokalaemia is less common and less severe with use of loop diuretics (Morgan and Davidson, 1980), yet almost always occurs when loop diuretics are used with thiazides or metolazone. Hypotension may occur if diuretic use causes volume depletion. Weight must be measured and should decrease until the patients become free of oedema. When changing the type or dose of diuretic, one must monitor the urine volume, the serum electrolytes (for hypokalaemia or hyponatraemia), and look for signs of a decline in tissue perfusion, such as weakness, asthenia, upright postural hypotension with dizziness, lethargy due to decreased cerebral blood flow, tachycardia, and unexplained rise in blood urea nitrogen and creatinine concentration. As long as these parameters remain constant, it can be assumed that diuretic therapy has not led to a significant impairment in perfusion to the kidneys or to other organs.

Ultrafiltration Some oedematous patients with advanced heart failure or renal failure do not respond to any of the above modalities. In them, one could consider dialysis or UF. Peritoneal dialysis was first used for congestive heart failure in 1959 (Girard et  al., 1959). Silverstein et al first proposed extracorporeal haemofiltration for fluid removal (Silverstein et al., 1974). Since then, this procedure has been much studied, either using haemodialysis with the UF component or by isolated haemofiltration in severe congestive heart failure (Agostoni et  al., 1994; Dormans et  al., 1996; Marenzi et  al., 2001). Grapsa et al. (2004) reported that when patients with congestive heart failure have intractable oedema in spite of intravenous inotropes and diuretics, treatment by haemodialysis with UF yields subjective improvement, but a high death rate persists (Grapsa et al., 2004). A  peripherally inserted UF device (Aquadex System 100)  was recently approved by the US Food and Drug Administration for therapy in heart failure allowing UF at very low flows via a peripheral intravenous catheter with only < 40 mL of extracorporeal blood at any given time (Costanzo et al., 2005). In decompensated heart failure, UF with this method produces greater weight and fluid loss than intravenous diuretics and reduces cost. But, the rate of fluid removal can exceed 500–1000 mL/hour in this setting and careful monitoring is required to prevent volume depletion. A recent randomized trial showed that diuretics alone were better than UF in heart failure patients who had renal impairment (Bart et  al., 2012). In subjects with heart failure whose serum creatinine is > 200 µmol/L, haemodialysis is needed, rather than just UF.

References Agostoni, P., Marenzi, G., Lauri, G., et al. (1994). Sustained improvement in functional capacity after removal of body fluid with isolated ultrafiltration in chronic cardiac insufficiency: failure of furosemide to provide the same result. Am J Med, 96(3), 191–9. Anand, I. S. and Chugh, S. S. (1997). Mechanisms and management of renal dysfunction in heart failure. Curr Opin Cardiol, 12(3), 251–8. Anand, I. S., Kalra, G. S., Ferrari, R., et al. (1990). Enalapril as initial and sole treatment in severe chronic heart failure with sodium retention. Int J Cardiol, 28(3), 341–6. Andreae, L. and Fine, L. G. (1997). Unravelling dropsy: from Marcello Malpighi’s discovery of the capillaries (1661) to Stephen Hales’ production of oedema in an experimental model (1733). Am J Nephrol, 17(3–4), 359–68. Aukland, K. (1984). Distribution of body fluids: local mechanisms guarding interstitial fluid volume. J Physiol (Paris), 79(6), 395–400. Bart, B. A., Goldsmith, S. R., Lee, K. L., et al. (2012). Ultrafiltration in decompensated heart failure with cardiorenal syndrome. N Engl J Med, 367(24), 2296–304. Brater, D. C., Day, B., Burdette, A., et al. (1984). Bumetanide and furosemide in heart failure. Kidney Int, 26(2), 183–9. Brater, D. C. (1998). Diuretic therapy. N Engl J Med, 339(6), 387–95. Bright, R. (1836). Cases and observations illustrative of renal disease accompanied with the secretion of albumenous urine. Guys Hosp Rep, 1, 338–79. Chalasani, N., Gorski, J. C., Horlander, J. C., Sr, et al. (2001). Effects of albumin/furosemide mixtures on responses to furosemide in hypoalbuminemic patients. J Am Soc Nephrol, 12(5), 1010–16. Costanzo, M. R., Guglin, M. E., Saltzberg, M. T., et al. (2007). Ultrafiltration versus intravenous diuretics for patients hospitalized for acute decompensated heart failure. J Am Coll Cardiol, 49(6), 675–83. Costanzo, M. R., Saltzberg, M., O’Sullivan, J., et al. (2005). Early ultrafiltration in patients with decompensated heart failure and diuretic resistance. J Am Coll Cardiol, 46(11), 2047–51. Davenport, A. (2001). Ultrafiltration in diuretic-resistant volume overload in nephrotic syndrome and patients with ascites due to chronic liver disease. Cardiology, 96(3–4), 190–5. Deschenes, G., Guigonis, V., and Doucet, A. (2004). Molecular mechanism of edema formation in nephrotic syndrome. Arch Pediatr, 11(9), 1084–94. Dorhout Mees, E. J. (1996). Does it make sense to administer albumin to the patient with nephrotic oedema? Nephrol Dial Transplant, 11(7), 1224–6.

chapter 30 

Dormans, T. P., Huige, R. M., and Gerlag, P. G. (1996). Chronic intermittent haemofiltration and haemodialysis in end stage chronic heart failure with oedema refractory to high dose frusemide. Heart, 75(4), 349–51. Edwards, B. D. and Hudson, W. A. (1991). A novel treatment for idiopathic oedema of women. Nephron, 58(3), 369–70. Ferraro, L. R., Friedman, M. M., and Morelli, H. E. (1949). Extracellular fluid in cardiac edema and ascites. Arch Intern Med (Chic), 83(3), 292–7. Firth, J. D., Raine, A. E., and Ledingham, J. G. (1988). Raised venous pressure: a direct cause of renal sodium retention in oedema? Lancet, 1(8593), 1033–5. Fitzpatrick, M. A., Nicholls, M. G., Ikram, H., et al. (1985). Stability and inter-relationships of hormone, haemodynamic and electrolyte levels in heart failure in man. Clin Exp Pharmacol Physiol, 12(2), 145–54. Fliser, D., Schroter, M., Neubeck, M., et al. (1994). Coadministration of thiazides increases the efficacy of loop diuretics even in patients with advanced renal failure. Kidney Int, 46(2), 482–8. Fliser, D., Zurbruggen, I., Mutschler, E., et al. (1999). Coadministration of albumin and furosemide in patients with the nephrotic syndrome. Kidney Int, 55(2), 629–34. Garin, E. H. (1987). A comparison of combinations of diuretics in nephrotic edema. Am J Dis Child, 141(7), 769–71. Ghafari, A., Mehdizadeh, A., Alavi-Darazam, I., et al. (2011). Co-administration of albumin-furosemide in patients with the nephrotic syndrome. Saudi J Kidney Dis Transpl, 22(3), 471–5. Gheorghiade, M., Konstam, M. A., Burnett, J. C., Jr, et al. (2007). Short-term clinical effects of tolvaptan, an oral vasopressin antagonist, in patients hospitalized for heart failure: the EVEREST Clinical Status Trials. JAMA, 297(12), 1332–43. Girard, M., Traeger, J., Fries, D., et al. (1959). Peritoneal glucose dialysis in the treatment of refractory cardiac edema; report on 3 new cases. J Med Lyon, 40(948), 585–92. Grapsa, E., Alexopoulos, G. P., Margari, Z., et al. (2004). Ultrafiltration in the treatment of severe congestive heart failure. Int Urol Nephrol, 36(2), 269–72. Guan, Y., Hao, C., Cha, D. R., et al. (2005). Thiazolidinediones expand body fluid volume through PPARgamma stimulation of ENaC-mediated renal salt absorption. Nat Med, 11(8), 861–6. Hall, J. E., Granger, J. P., Smith, M. J., Jr, et al. (1984). Role of renal hemodynamics and arterial pressure in aldosterone ‘escape’. Hypertension, 6(2 Pt 2), I183–92. Henry, J. A. and Altmann, P. (1978). Assessment of hypoproteinaemic oedema: a simple physical sign. Br Med J, 1(6117), 890–1. Johnson, R. J., Herrera-Acosta, J., Schreiner, G. F., et al. (2002). Subtle acquired renal injury as a mechanism of salt-sensitive hypertension. N Engl J Med, 346(12), 913–23. Kapur, G., Valentini, R. P., Imam, A. A., et al. (2009). Treatment of severe edema in children with nephrotic syndrome with diuretics alone—a prospective study. Clin J Am Soc Nephrol, 4(5), 907–13. Karnad, D. R. and Abraham, P. (1986). Diuretic treatment in decompensated cirrhosis and congestive heart failure: effect of posture. Br Med J (Clin Res Ed), 293(6545), 508. Konstam, M. A., Gheorghiade, M., Burnett, J. C., Jr, et al. (2007). Effects of oral tolvaptan in patients hospitalized for worsening heart failure: the EVEREST Outcome Trial. JAMA, 297(12), 1319–31. Kurnick, N. B. (1948). War edema in the civilian population of Saipan. Ann Intern Med, 28(4), 782–91.

approach to the patient with oedema

Loon, N. R., Wilcox, C. S., and Unwin, R. J. (1989). Mechanism of impaired natriuretic response to furosemide during prolonged therapy. Kidney Int, 36(4), 682–9. Marenzi, G., Lauri, G., Grazi, M., et al. (2001). Circulatory response to fluid overload removal by extracorporeal ultrafiltration in refractory congestive heart failure. J Am Coll Cardiol, 38(4), 963–8. Morgan, D. B. and Davidson, C. (1980). Hypokalaemia and diuretics: an analysis of publications. Br Med J, 280(6218), 905–8. Oster, J. R., Epstein, M., and Smoller, S. (1983). Combined therapy with thiazide-type and loop diuretic agents for resistant sodium retention. Ann Intern Med, 99(3), 405–6. Palmer, B. F. and Alpern, R. J. (1997). Pathogenesis of edema formation in the nephrotic syndrome. Kidney Int Suppl, 59, S21–7. Panchapakesan, U., Pollock, C., and Saad, S. (2009). Review article: importance of the kidney proximal tubular cells in thiazolidinedione-mediated sodium and water uptake. Nephrology, 14(3), 298–301. Pockros, P. J. and Reynolds, T. B. (1986). Rapid diuresis in patients with ascites from chronic liver disease: the importance of peripheral edema. Gastroenterology, 90(6), 1827–33. Rascher, W., Tulassay, T., Seyberth, H. W., et al. (1986). Diuretic and hormonal responses to head-out water immersion in nephrotic syndrome. J Pediatr, 109(4), 609–14. Reid, C. J., Marsh, M. J., Murdoch, I. M., et al. (1996). Nephrotic syndrome in childhood complicated by life threatening pulmonary oedema. BMJ, 312(7022), 36–8. Richet, G. (1993). Edema and uremia from 1827 to 1905: the first faltering steps of renal pathophysiology. Kidney Int, 43(6), 1385–96. Ring-Larsen, H., Henriksen, J. H., Wilken, C., et al. (1986). Diuretic treatment in decompensated cirrhosis and congestive heart failure: effect of posture. Br Med J (Clin Res Ed), 292(6532), 1351–3. Rodriguez-Iturbe, B., Herrera-Acosta, J., and Johnson, R. J. (2002). Interstitial inflammation, sodium retention, and the pathogenesis of nephrotic edema: a unifying hypothesis. Kidney Int, 62(4), 1379–84. Rose, B. D. (1991). Diuretics. Kidney Int, 39(2), 336–52. Schrier, R. W. (2007). Decreased effective blood volume in edematous disorders: what does this mean? J Am Soc Nephrol, 18(7), 2028–31. Schrier, R. W. and Fassett, R. G. (1998). A critique of the overfill hypothesis of sodium and water retention in the nephrotic syndrome. Kidney Int, 53(5), 1111–17. Silverstein, M. E., Ford, C. A., Lysaght, M. J., et al. (1974). Treatment of severe fluid overload by ultrafiltration. N Engl J Med, 291(15), 747–51. Stevenson, L. W. and Perloff, J. K. (1989). The limited reliability of physical signs for estimating hemodynamics in chronic heart failure. JAMA, 261(6), 884–8. Touwaide, A. and De Santo, N. G. (1999). Edema in the corpus hippocraticum. Am J Nephrol, 19(2), 155–8. Wald, H., Scherzer, P., and Popovtzer, M. M. (1991). Na,K-ATPase in isolated nephron segments in rats with experimental heart failure. Circ Res, 68(4), 1051–8. Wilcox, C. S. (2002). New insights into diuretic use in patients with chronic renal disease. J Am Soc Nephrol, 13(3), 798–805. Yersin, B., Burnier, M., and Magnenat, P. (1995). Improvement of renal failure with repeated head-out water immersions in patients with hepatorenal syndrome associated with alcoholic hepatitis. Am J Nephrol, 15(3), 260–5. Yoshimura, A., Ideura, T., Iwasaki, S., et al. (1992). Aggravation of minimal change nephrotic syndrome by administration of human albumin. Clin Nephrol, 37(3), 109–14.

281

CHAPTER 31

Approach to the patient with salt-wasting tubulopathies Detlef Bockenhauer and Robert Kleta Introduction Sodium is the main ion constituent of the extracellular and intravascular fluid compartments, and it is through control of sodium reabsorption that the kidneys maintain volume homoeostasis and systemic blood pressure (see Chapter 21). The amount of sodium that is first filtered by the glomerulus and then reabsorbed in the tubule is quite staggering: assuming a glomerular filtration rate of 100 mL/min and a serum sodium concentration of 140 mmol/L, an average-sized person filters about 20,000 mmol of sodium per day, equivalent to the amount in 1.2 kg of cooking salt (Kleta and Bockenhauer, 2006). In the steady state, the amount of sodium excreted is equal to the amount ingested. An average Western diet contains about 8–10 g of salt per day; a low-salt diet would usually be around 2 g per day. Under physiological conditions, the tubules reabsorb about 99% of filtered sodium and without the kidneys’ sophisticated ability to preserve salt (and water), human life would not exist. This enormous task is accomplished by a combination of distinct and sequentially oriented sodium or sodium-coupled transport systems along the nephron and the concerted and parallel action of some of these systems within the kidney (Fig. 31.1). Based on the molecular identity of the involved transporters, we typically distinguish four segments important for sodium and water reabsorption: 1. The proximal tubule (PT) 2. Thick ascending limb (TAL) of Henle’s loop, including macula densa 3. Distal convoluted tubule (DCT) and connecting tubule 4. Collecting duct (CD). Disorders affecting the individual segments lead to specific biochemical ‘fingerprints’, which can guide the diagnosis of these disorders (Table 31.1). Diuretics, or more correctly saluretics, typically target specific transporters involved in tubular sodium reabsorption and many of these disorders can be conceptualized by comparing them with the effects of the different saluretics. An overview of salt-wasting tubulopathies and their ‘salient’ clinical features is given in Table 31.2, including the underlying molecular basis and the corresponding diuretic (if any).

Disorders affecting salt transport in the proximal tubule Salt wasting in the PT is typically in the form of a generalized PT dysfunction, that is, renal Fanconi syndrome, which is discussed

separately (see Chapter  41) and will only be mentioned briefly (Fig. 31.2). Because of the molecular link of sodium transport with the various other transport pathways (see Chapter 20), patients typically exhibit multiple biochemical abnormalities, including metabolic acidosis, hypophosphataemia, glycosuria, and aminoaciduria. There are only a few disorders affecting specific sodium transporters in the PT and these are not salt-wasting disorders per se. The main transporter for sodium in the PT is sodium-hydrogen exchanger 3(NHE3) and so far no human disease has been associated with mutations in NHE3. Interestingly, mice deleted for the encoding gene Slc9a3 survive and exhibit only mild metabolic acidosis (Schultheis et al., 1998). Mutations have been identified in proximal sodium-phosphate transporters. Mutations in the predominant form NaPi-IIa (SLC34A1) have recently been associated with a variant of an incomplete renal Fanconi syndrome (Magen et al., 2010), whereas mutations in NaPi-IIc (SLC34A3) are associated with hypophosphataemic rickets (Bergwitz et al., 2006; Lorenz-Depiereux et al., 2006; Ichikawa et al., 2006). Diseases have also been linked with sodium-glucose transporters: mutations in SGLT1 (SLC5A1) lead to glucose/galactose malabsorption with severe infantile-onset diarrhoea, highlighting the importance of this transporter in the intestine (Turk et al., 1991). In contrast, mutations in SGLT2 (SLC5A2) lead to isolated renal glycosuria, in general without clinical symptoms (van den Heuvel et al., 2002). Whilst the loss of sodium and glucose (up to 150 g of glucose per day) may have been relevant in Palaeolithic times, it could actually be beneficial in the context of a current Western diet, potentially protecting against hypertension and diabetes (Francis et al., 2004); indeed, drugs that inhibit this transporter have been approved recently for control of hyperglycaemia in diabetes.

Thick ascending limb of Henle’s loop: Bartter syndromes Impaired salt transport in the TAL leads to Bartter syndromes. Currently, we recognize four different forms, based on molecular genetics (Table 31.2) (Simon et al., 1996a, 1996b, 1997; Birkenhager et al., 2001). A separate disease, familial hypocalcaemic hypercalciuria, due to dominant (activating) mutations in the calcium-sensing receptor (CaSR) can occasionally cause a biochemical urinary constellation resembling Bartter syndrome, and is sometimes referred to as Bartter type 5 (Watanabe et al., 2002). Similarly, a Bartter-like phenotype can be observed in some patients with mitochondrial

CHAPTER 31 

salt-wasting tubulopathies

Filtration 100% Fanconi syndromes

Gitelman syndrome

70–80%

5–10%

Pseudohypoaldosteronism I

Batter syndromes

2–5%

10–15%

0–1%

Fig. 31.1  A model of salt reabsorption along the nephron. The vast majority of filtered sodium is reabsorbed in proximal tubule. Segment specific human disorders affecting salt reabsorption are indicated.

cytopathies (Goto et  al., 1990; Emma et  al., 2006). However, as these are diseases with a phenotype beyond Bartter syndrome and have a separate molecular basis, only the four main variants, specifically affecting salt-transport in the TAL will be discussed here.

Pathophysiology and aetiology Bartter syndrome type 1 is caused by mutations in NKCC2 (SLC12A1) (Simon et al., 1996a). This transporter is the target of loop diuretics and the symptoms are best compared with chronic furosemide administration (Reinalter et  al., 2004). NKCC2 can only function if it is ‘fully loaded’, that is, if it has bound 1 sodium, 1 potassium and 2 chloride ions. The concentration of these ions in the tubular lumen is roughly the same as in plasma. Hence, the availability of potassium becomes the rate-limiting step for NKCC2 function. This critical availability of potassium is highlighted by Bartter syndrome type 2, which is caused by recessive mutations in ROMK (KCNJ1) (Simon et al., 1996b). ROMK is a potassium channel expressed in the apical membrane of TAL (Fig. 31.3) and ensures an adequate supply of potassium by recycling the potassium taken up by NKCC2 back into the tubular lumen. This may appear unnecessarily complicated at first sight:  why transport potassium first into the cell only to have it leak back out through

this potassium channel? However, there is an evolutionary reason for this arrangement: NKCC2 is electroneutral (it transports two cations and two anions). By having potassium leak back out through ROMK, the two transport molecules establish a voltage gradient (outside positive) across the apical membrane and, ultimately, across the epithelial cell layer (Fig. 31.4). This transepithelial potential then drives the paracellular reabsorption of calcium, magnesium and sodium mediated by claudins (Haisch et al., 2011). For this reason, patients with Bartter syndromes type 1 and 2 have hypercalciuria and hypermagnesuria:  the failure to establish the transepithelial potential leads to losses of calcium and magnesium in the urine. Whereas types 1 and 2 are due to defects in transport molecules on the apical membrane, types 3 and 4 concern the basolateral membrane. As detailed above, potassium taken up by NKCC2 recycles back into the tubular lumen. Sodium can exit the cell via the basolateral Na-K-ATPase, while chloride exits via the chloride channel CLCNKB, leading to net reabsorption of NaCl in this segment. Consequently, defects in CLCNKB impair salt reabsorption in the TAL and mutations in this channel have been found to underlie Bartter syndrome type 3 (Simon et al., 1997). CLCNKB is expressed not only in the TAL, but also in the DCT where it

Table 31.1  Biochemical ‘fingerprints’ of salt wasting tubulopathies based on affected tubular segment Segment

PT

TAL

DCT

CD

Plasma Na

Normal

Normal

Normal

Low–normal

K

Low

Low (high neonatally Bartter type 2)

Low

High

HCO3

Low

High

High

Low

Mg

Normal

Low–normal

Low

Normal

PO4

Low

Normal

Normal

Normal

Urine K

High

High

High

Low

Ca

High

High (or normal in Bartter types 3, 4)

Low

Low

Tubular proteins

High

Normal

Normal

Normal

283

284

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 31.2  Genetics, salient clinical features, and pharmacotyping of salt-wasting tubulopathies Syndrome (inheritance)

Gene (protein)

Typical clinical features

Targeting diuretic

Bartter type 1 (AR)

SLC12A1 (NKCC2)

Antenatal presentation with polyhydramnios and prematurity Nephrocalcinosis Urinary concentrating defect

Loop diuretics (furosemide)

Bartter type 2 (AR)

KCNJ1 (ROMK)

Antenatal presentation with polyhydramnios and prematurity Nephrocalcinosis Neonatal hyperkalaemia Urinary concentrating defect

Bartter type 3 (AR)

CLCNKB (CLCNKB)

Childhood presentation Most severe electrolyte abnormalities, often hypomagnesaemia

Bartter type 4 (AR)

BSND (Barttin)

Antenatal presentation with polyhydramnios and prematurity Sensorineural deafness

Gitelman (AR)

SLC12A3 NCC

Hypocalciuria Hypomagnesaemia

Thiazides

EAST (AR)

KCNJ10 Kir 4.1

Infantile-onset epilepsy; ataxia; sensorineural deafness; hypocalciuria Hypomagnesaemia

Thiazides

Pseudohypoaldosteronism type 1A (AR)

SCNN1A, B, G ENaC

Hypovolaemia, hyperkalaemia, acidosis, hyponatraemia, severe neonatal presentation

Amiloride

Pseudohypoaldosteronism type 1B (AD)

MLR

Hypovolaemia, hyperkalaemia, acidosis, hyponatraemia, acidosis

Spironolactone, eplerone

mediates chloride exit; therefore, there is some phenotypic overlap with Gitelman syndrome (see below). For proper surface membrane expression, CLCNKB needs a subunit, Barttin (encoded by BSND), and mutations in Barttin lead to Bartter syndrome type 4 (Birkenhager et al., 2001).

Clinical features Despite the common category of ‘Bartter syndrome’, there is some clinical distinction between the four forms. An overview of the different forms is given in Table 31.2.

Hyperaldosteronism All four forms are characterized by elevated renin and aldosterone levels, resulting in hypokalaemic hypochloraemic metabolic alkalosis (Table 31.1). This is associated with altered tubuloglomerular feedback in the macula densa. The macula densa senses tubular reabsorption of sodium-chloride: decreased reabsorption is interpreted as decreased delivery, that is, volume depletion (Schnermann et al., 2003). To enhance intravascular volume and glomerular filtration, macula densa cells produce locally prostaglandins, especially PGE2, which is secreted and then sensed by juxtaglomerular cells, which in turn produce and secrete renin (Yang et al., 2000; Peti-Peterdi et al., 2003). Prostaglandins are produced by cyclooxygenases, especially COX-2, explaining the success of treatment with COX-2-inhibitors (see ‘Treatment’) (Komhoff et al., 2000; Reinalter et al., 2002). In Bartter syndromes, this tubuloglomerular feedback

is short-circuited: because the macula densa is part of the TAL and sodium-chloride reabsorption is defective, the patients produce high levels of prostaglandins, renin, and aldosterone. This excessive production of prostaglandins is likely to be the cause of the juxtaglomerular hyperplasia, which was a defining feature in the first description of this syndrome (Bartter et al., 1962). It is the consequent hyperaldosteronism and its action on the collecting duct that leads to the typical biochemical picture, because sodium is reabsorbed in exchange for potassium and protons in this nephron segment (see Fig. 31.5).

Transient neonatal hyperkalaemia A specific feature of Bartter syndrome type 2 is the presence of hyperkalaemia in early postnatal life, often associated with hyponatraemia and acidosis (Finer et  al., 2003; Brochard et  al., 2009). The reason for this is the expression of ROMK in the principal cells of the CD, where it mediates potassium secretion. The biochemical picture in these patients mimics pseudohypoaldosteronism (see later) and can lead to misdiagnosis (Greenberg et al., 1995; Nozu et al., 2007). Usually, this picture disappears at the end of the first couple of weeks, as alternative pathways for potassium secretion mature and the typical biochemical features of Bartter syndrome evolve (Bailey et al., 2006).

Age of onset In the original publication of the syndrome by Frederic Bartter and colleagues in 1962, two patients are described who first came

CHAPTER 31 

Lumen

salt-wasting tubulopathies

Blood Urine 3 Na+

NHE3

Na+

Blood

Na-K-ATPase H+

Na+

2K+

Na+ X

K+

K+

Na+

Na+

3Na+

NCC

2K+

Cl–

Na-K-ATPase K+ KCNJ10

Mg++

Cl–

TRPM6

CLCNKB

Ca++?

Fig. 31.2  Diagram of an epithelial renal proximal tubular cell. Sodium is reabsorbed via apical sodium exchangers and cotransporters. The electrochemical gradient for sodium uptake is provided by the basolateral Na-K-ATPase, which also provides a basolateral sodium exit pathway. This gradient is then utilized for uptake of various ions and small molecules (indicated by X) by luminal transport systems. Paracellular calcium reabsorption follows sodium passively, but the molecular identity of this transport pathway is yet unknown.

Fig. 31.4  Diagram of an epithelial cell in the distal convoluted tubule. Sodium is reabsorbed via the apical sodium-chloride cotransporter NCC, mutations in which cause Gitelman syndrome. The electrochemical gradient for sodium uptake is provided by the basolateral Na-K-ATPase, which also provides a basolateral sodium exit pathway. Activity is dependent on provision of potassium via the basolateral potassium channel KCNJ10, mutations in which cause EAST syndrome. Chloride exits the cell via the basolateral chloride channel CLCNKB.

to medical attention at 5 and 12 years of age, respectively (Bartter et al., 1962). This childhood-onset form is now often referred to as ‘classic’ Bartter syndrome, to distinguish it from the antenatal form that presents with severe polyhydramnios, often requiring repeated amniocentesis to relieve the pressure and typically resulting in premature birth. Molecular genetics has shown that the

antenatal form is usually associated with mutations in SLC12A1, KCNJ1, or BSND, whereas classic Bartter syndrome is caused by mutations in CLCNKB (Peters et al., 2002). The likely explanation for this distinction is that the encoded proteins NKCC2, ROMK, and Barttin are each necessary and required for salt-reabsorption in the TAL. In contrast, CLCNKB has a very close homologue, CLCNKA (Jentsch et al., 2005; Briet et al., 2006). There is some controversy about expression of CLCNKA in the TAL and, unfortunately, due to the close homology there are no antibodies that

Blood

Lumen

Na+

3 Na+

NKCC2

Na-K-ATPase

K+ 2 C1– + + +

– – K+ ROMK

Na+ C1–

Na+

K+

Blood

+ +

+ CLCNKB/A Na

ROMK

Barttin

Mg++ Ca++

2 K+

Lumen

–

3Na+

K+ ENaC α,β,γ

Na-K-ATPase

2K+ +

CLDN16 CLDN19

Fig. 31.3  Diagram of an epithelial cell in the thick ascending limb of Henle’s loop. Sodium is reabsorbed via the apical sodium-potassium-2 chloride-cotransporter NKCC2, mutations in which cause Bartter syndrome type 1. Activity is dependent on provision of potassium via the apical potassium channel ROMK, mutations in which cause Bartter syndrome type 2. Together, these two molecules contribute to a transepithelial voltage gradient (lumen-positive, indicated by the number of + signs), which drives paracellular reabsorption of sodium, calcium and magnesium. The electrochemical gradient (indicated by font size for intracellular potassium and extracellular sodium and by + and − signs) for sodium uptake is provided by the basolateral Na-K-ATPase, which also provides a basolateral sodium exit pathway. Chloride exits the cell via the basolateral chloride channel CLCNKB, mutations in which cause Bartter syndrome type 3. Additional chloride exit is likely provided by CLCNKA (see text). Both channels require a subunit, Barttin, mutations in which cause Bartter syndrome type 4.

Na+

ENaC

Aldosterone MLR

Fig. 31.5  Diagram of a principal cell in the collecting duct. Sodium is reabsorbed via the apical sodium channel ENaC, mutations (deactivating) in which cause the recessive form of pseudohypoaldosteronism type 1. This, in turn enhances potassium secretion through the apical potassium channel ROMK (and proton secretion via the proton pump in adjacent intercalated cells). Expression of ENaC is regulated via the mineralocorticoid receptor MLR, mutations in which cause the dominant form of pseudohypoaldosteronism type 1. The electrochemical gradient for sodium uptake is provided by the basolateral Na-K-ATPase, which also provides a basolateral sodium exit pathway.

285

286

Section 2  

fluid, electrolyte, and renal tubular disorders

can reliably distinguish between the two variants (Kramer et al., 2008). Yet, based on the clinical picture, some expression of CLCNKA in TAL is likely: because CLCNKA can provide an alternative basolateral exit pathway, the phenotype is milder with later onset of symptoms. Barttin is a required subunit for both channels, so if this protein is not functional, there is no chloride exit and the phenotype is severe with antenatal onset. Evidence for this hypothesis was provided by the discovery of a patient with severe antenatal onset who had inherited recessive mutations in both CLCNKB and CLCNKA, mimicking the phenotype of Bartter type 4 (Schlingmann et al., 2004). As logical as the molecular explanation may seem, in clinical reality, separation between the antenatal and classical forms is not as clear-cut: some patients with types 1 and 2 present in child- or even adulthood, whereas occasional patients with type 3 have a severe antenatal presentation (Konrad et al., 2000; Brochard et al., 2009). Partly, this can be explained by differences in the genotype, as some mutations in SLC12A1 or KCNJ1 may retain some functionality of the protein (genotype–phenotype correlation), but in others there is no clear explanation (Pressler et  al., 2006; Brum et al., 2007; Brochard et al., 2009).

Hypercalciuria and nephrocalcinosis As detailed above, the combined action of NKCC2 and ROMK contribute to the generation of a transepithelial voltage gradient driving paracellular reabsorption of cations, including calcium. Consequently, Bartter syndrome type 1 and 2 are associated with hypercalciuria and nephrocalcinosis (Peters et al., 2002; Brochard et  al., 2009). Patients with type 3 and 4 have intact NKCC2 and ROMK and so are still able to generate the voltage gradient and reabsorb calcium in the TAL. Hence, these forms typically exhibit neither hypercalciuria nor nephrocalcinosis. In fact, mutations in CLCNKB can sometimes mimic Gitelman syndrome, which is associated with hypocalciuria (Jeck et al., 2000; Zelikovic et al., 2003).

Hypomagnesaemia About 60% of filtered magnesium is reabsorbed paracellularly in the TAL. Consequently, one would expect that patients with Bartter type 1 and 2 have severe hypermagnesuria and hypomagnesaemia, in addition to the observed hypercalciuria, as seen in patients with a defect in this paracellular pathway (familial hypomagnesaemia with hypercalciuria and nephrocalcinosis; see Chapter 40) (Simon et al., 1999; Konrad et al., 2006). However, this is usually not the case. Most patients with Bartter syndrome type 1 and 2 have, in fact, normal plasma magnesium levels. Likely, there is enhanced magnesium reabsorption in the DCT to compensate for the loss in the TAL, but why this is possible in Bartter syndrome and not in familial hypomagnesaemia with hypercalciuria and nephrocalcinosis is unclear. However, consistent with this hypothesis of a compensatory increase of magnesium transport in the DCT is the observation that patients with Bartter syndrome type 3 and 4 often have hypomagnesaemia (Jeck et al., 2000; Konrad et al., 2000; Peters et al., 2002). Since CLCNKB, together with Barttin, is also expressed in the DCT, these patients have impaired transport in both segments, leading to uncompensated magnesium losses.

Polyuria Polyuria is a common feature of Bartter syndromes, especially in the antenatal forms (Bartter types 1 and 2), where fetal water and salt excretion exceeds the absorptive capacity of the placenta,

leading to polyhydramnios. The TAL is water-impermeable and salt reabsorption leads to urinary dilution and establishes the medullary concentration gradient that drives water extraction in the CD (Bockenhauer, 2008). Consequently, a defect in salt transport in the TAL abolishes urinary dilution, as well as concentration, and patients typically exhibit isosthenuria (similar to plasma or serum). But, again, there are exceptions to the rule: some patients with Bartter syndrome exhibit a phenotype consistent with diabetes insipidus, so that the osmolality of urine is well below that of plasma (Bockenhauer et al., 2008, 2010). Occasionally, this can actually lead to a misdiagnosis of nephrogenic diabetes insipidus (Bettinelli et al., 2000). At first sight, it appears quite remarkable that patients with a genetic ‘knockout’ of the TAL, the urinary diluting segment, are able to dilute their urine so well. This feat is likely achieved by a dramatic compensatory increase of salt transport in the DCT, which is also water impermeable. Indeed, KCNJ1 knockout mice demonstrate marked hypertrophy of the DCT, as a morphological correlate of the increased transport activity (Cantone et al., 2008; Wagner et al., 2008). The apparent ‘aquaporin-deficient’ diabetes insipidus is more difficult to explain, but is likely to be related to the electrolyte abnormalities (hypokalaemia, hypercalciuria), since it also occurs in other disorders with this biochemical picture, and urinary concentration was normalized in one case once the biochemistry was corrected completely (Bockenhauer et al., 2010).

Deafness Sensorineural deafness affecting all frequencies, necessitating cochlear implants early in life, is a unique feature of Bartter type 4. CLCNKB and CLCNKA together with the Barttin subunit are expressed in the stria vascularis of the inner ear and contribute to the generation of the high potassium concentration of the endolymph (Kramer et al., 2008). If only CLCNKB is mutated, CLCNKA can compensate in the inner ear and so patients with Bartter type 3 have no hearing problem. However, with Barttin mutations, both channels are non-functional and these patients suffer from deafness, in addition to Bartter syndrome.

Treatment Salt supplementation There is no curative treatment for Bartter syndrome (kidney transplantation is not seen as a therapeutic option) and management of these patients is entirely symptomatic. In those with an antenatal presentation the problems of prematurity often dominate, at least in the beginning. Salt supplementation is used to ameliorate electrolyte abnormalities, but limited by palatability and side effects, such as vomiting and gastric irritation. Moreover, with the fractional excretion of potassium often exceeding 100% in patients with Bartter syndrome, normalization of plasma potassium levels is virtually impossible to achieve. Bolus administration, especially when given intravenously, may normalize the plasma level in the short term, but will usually result in exaggerated swings of plasma potassium, whereas repeated doses distributed over the day will result in more even plasma levels. In infants, mixing of supplements in the milk formula can increase tolerability and provide a steady administration. Sodium supplements are usually helpful. The importance of salt replacement is emphasized by the marked craving for salt, which is apparent in most patients.

CHAPTER 31 

Non-steroidal anti-inflammatory drugs Non-steroidal anti-inflammatory drugs (NSAIDs) are a mainstay of the treatment of Bartter syndrome, at least during early childhood, when problems with salt and fluid intake and failure to thrive can be severe. Their beneficial effect is due to the suppression of prostaglandin synthesis, since the elevated prostaglandin levels are a defining feature of the disease, especially in the antenatal variant, which is historically sometimes also called hyperprostaglandin E syndrome (Seyberth et al., 1985). The prostaglandins likely have systemic effects, as well, which is why commencement of a NSAID typically results in a marked improvement of the patient’s general condition with subsequent catch-up growth (Dillon et al., 1979). Even antenatal treatment by administration of indomethacin to the mother has been reported, though it is unclear if the improved outcome compared with the affected sibling was a result of drug treatment or just a reflection of the variability often seen in severity of the phenotype (Konrad et al., 1999). Typically, indomethacin is used (1–4 mg/kg/day, divided in four doses), but other non-steroidals can also be used, for example, ibuprofen (20–30 mg/kg/day, divided in three doses). Treatment can be limited by potentially severe side effects, especially gastric ulcers and bleeding (Dillon et al., 1979). Selective COX-2 inhibitors constitute an attractive alternative and their successful use has been reported (Kleta et al., 2000; Reinalter et al., 2002). However, recognition that these drugs are associated with an increased cardiovascular morbidity and the withdrawal of rofecoxib have dampened the initial enthusiasm by some, and the use of these drugs should be considered in each case individually (Dogne et al., 2006). NSAIDs are well recognized to have side effects on the kidney, as well and indeed, chronic NSAID administration has been suspected as causative in cases of Bartter syndrome who developed chronic kidney disease (Kim et al., 2000). Yet, causality is difficult to prove: in patients without Bartter syndrome who take NSAIDs because of pain or rheumatologic conditions, renal prostaglandin levels before NSAID use are presumably normal, and the subsequent depression of prostaglandins can impair renal perfusion and aldosterone production (Zawada et al., 1982). In contrast, patients with Bartter syndrome have elevated prostaglandin levels and NSAIDs are used to decrease these towards the normal range. Since the persistent activation of the renin–angiotensin system in Bartter syndrome has been postulated to contribute to chronic kidney disease, and even glomerulosclerosis, suppression of the system by a NSAID could actually be beneficial for the kidney (Su et al., 2000). NSAID treatment should be instituted after volume repletion to avoid potentially fatal nephrotoxicity (historically, indomethacin has been used to perform non-surgical nephrectomies).

Distal convoluted tubule: Gitelman and EAST syndrome Pathophysiology and aetiology The key transport molecule in the DCT is the sodium-chloride cotransporter NCC (SLC12A3), which can be inhibited by thiazides. The biochemical consequences of inherited defects of salt transport in the DCT are similar to chronic thiazide administration: hypokalaemic hypochloraemic alkalosis with hypomagnesaemia and hypocalciuria (Table 31.1). Gitelman syndrome is caused by loss-offunction mutations in NCC itself, whereas EAST syndrome (also called SeSAME syndrome) is caused by loss-of-function mutations in the basolateral potassium channel KCNJ10 (Simon et al., 1996c;

salt-wasting tubulopathies

Bockenhauer et al., 2009; Scholl et al., 2009). Similar to the interplay between NKCC2 and ROMK (see ‘Bartter syndrome’), KCNJ10 recycles potassium to provide a steady supply of potassium for the basolateral Na-K-ATPase, which establishes the electrochemical gradient for reabsorption of not only sodium-chloride, but also for magnesium in the DCT (Bleich et al., 2009). Thus, the hypokalaemic hypochloraemic alkalosis is due to the aldosterone-mediated compensatory increase in sodium reabsorption in the CD, as in Bartter syndrome, but the hypomagnesaemia is characteristic for the impaired sodium transport in the DCT. The hypocalciuria is more complex: animal data suggest, that the decreased intravascular volume induced by the impaired sodium reabsorption leads to enhanced proximal sodium transport with calcium following passively (Nijenhuis et al., 2005). While this should apply equally to Bartter syndrome, the impaired calcium reabsorption in the TAL in Bartter type 1 and 2 apparently outweighs this proximal effect (see ‘Hypercalciuria’). Clinically, Gitelman syndrome can be difficult to distinguish from Bartter syndrome type 3: as detailed above (Bartter syndrome, pathophysiology), CLCNKB is expressed in both the TAL and DCT, and signs and symptoms can overlap. Indeed, some patients with CLCNKB gene mutations may initially present with features of Bartter syndrome and later fit better with the clinical phenotype of Gitelman syndrome (Jeck et al., 2000).

Clinical features: Gitelman syndrome Gitelman syndrome is often seen as the mild variant of Bartter syndrome: patients typically do not present until adolescence or even adulthood (Knoers and Levtchenko, 2008). Often, the diagnosis is made incidentally, when blood tests were obtained routinely for an unrelated problem. However, more severe cases with early onset and growth failure have been described (Riveira-Munoz et  al., 2007). Moreover, even in those cases with late onset, the disease can have profound effects on quality of life (Cruz et al., 2001). Fatigue, palpitations, cramps, tetany, muscle weakness, and aches are frequently reported symptoms. In one study, about 50% of patients had a prolonged QT time (Foglia et al., 2004) and sudden cardiac arrest has been reported in isolated cases (Riveira-Munoz et  al., 2007; Scognamiglio et  al., 2007). Chondrocalcinosis is another potential complication attributed to chronic hypocalciuria and/ or hypomagnesaemia, but this usually does not appear until later in adulthood (Cobeta-Garcia et al., 1998; Monnens et al., 1998; Ea et al., 2005).

Clinical features: EAST syndrome While the biochemical features of EAST syndrome are indistinguishable from Gitelman syndrome, this severe disorder is dominated by the neurological manifestations, which are independent of the plasma and urine biochemistries (Bockenhauer et al., 2009; Reichold et al., 2010; Bandulik et al., 2011). EAST is an acronym for epilepsy, ataxia, sensorineural deafness, and tubulopathy, and it is the infantile-onset seizures and ataxia that affect patients the most. Electroencephalograms in general are non-diagnostic. Sensorineural deafness is mild in comparison with Bartter type 4 and is present in all EAST patients. Ataxia is severe and debilitating, affecting movement and speech from childhood. Intellectual abilities may not necessarily be compromised, but can be difficult to assess, because the ataxia may affect speech and writing, and the impaired expressive abilities result in the mistaken label of ‘mental retardation’.

287

288

Section 2  

fluid, electrolyte, and renal tubular disorders

Treatment Treatment is entirely symptomatic and consists mainly of electrolyte (potassium, sodium-chloride, and magnesium) supplementation. Indomethacin is rarely used, although a beneficial effect has occasionally been reported, but in an era before molecular diagnosis was possible (Liaw et al., 1999). These cases may, in fact, have had Bartter type 3, rather than Gitelman syndrome. As in Bartter syndrome, normalization of plasma values is difficult to achieve and magnesium supplementation is limited especially by diarrhoea, which may worsen the hypokalaemic alkalosis. Distribution of supplements over several smaller doses during the day can limit side effects and will likely provide more steady plasma levels. Drugs associated with prolonged QT interval or known to cause hypokalaemia (beta-adrenergic mimetics) should be avoided.

Collecting duct: pseudohypoaldosteronism type 1 Pathophysiology and aetiology Impaired sodium reabsorption in the CD results in pseudohypoaldosteronism type 1 (PHA1) (Cheek and Perry, 1958). Although < 5% of total filtered sodium is reabsorbed in the CD, loss of sodium transport in this segment is associated with the clinically most severe form of salt wasting. Presumably, this is due to the absence of a further distal segment that can compensate by taking up the unreabsorbed sodium. Sodium is reabsorbed in the CD via the epithelial sodium channel (ENaC) (Fig. 31.5) and the autosomal recessive form of PHA1 is caused by loss-of-function mutations in one of the channel subunits (Table 31.2) (Chang et al., 1996). Expression of this channel is regulated by the mineralocorticoid receptor (MLR) and the dominant form is caused by loss-of-function mutations in MLR (Geller et  al., 1998). ENaC is blocked by amiloride and MLR by spironolactone: consequently, the biochemical picture of PHA1 is similar to chronic use of these diuretics and characterized by severe hyperkalaemic acidosis and moderate to borderline hyponatraemia (see Table 31.1). Secondary forms of PHA1 are recognized (apart from the use of potassium-sparing diuretics) and can occur in urinary obstruction, acute kidney injury, or with pyelonephritis in infants (Uribarri et al., 1982; Rodriguez-Soriano et al., 1983; Pumberger et al., 1998; Schoen et al., 2002; Asano et al., 2006; Kashimada et al., 2008; Rogers, 2008).

Clinical features In the recessive form, affected infants present in the early neonatal period with severe hypovolaemia and commonly circulatory shock (Dillon et al., 1980; Savage et al., 1982). Occasionally, presentation can be antenatal with polyhydramnios (Wong and Levine, 1998). Plasma aldosterone levels are markedly elevated, consistent with tubular unresponsiveness to the hormone. Since ENaC is expressed also in the lung and skin, children with recessive PHA1 typically have impaired fluid clearance from the lungs, leading to increased respiratory infections and even cystic fibrosis-like symptoms (Kerem et al., 1999; Schaedel et al., 1999; Riepe, 2009). Moreover, as in cystic fibrosis, affected patients have an increased sodium concentration in sweat, since sodium reabsorption from sweat is also mediated by ENaC (Hummler and Horisberger, 1999). Skin rashes are often seen, probably related to the impaired sodium clearance

from sweat (Martin et al., 2005). Conversely, patients with cystic fibrosis can present in early childhood with hypokalaemia mimicking Bartter syndrome, due to the cutaneous salt losses with consequent hyperaldosteronism (Kleta et al., 1999). The dominant form of PHA1 is usually milder and typically recognized in infancy and early childhood with failure to thrive. The biochemical abnormalities are usually only mild (Riepe, 2009).

Treatment Treatment consists mainly of salt supplementation in the form of NaCl and NaHCO3. In the dominant form, supplementation can usually stop once children self-regulate their diet, presumably as they crave the necessary salt and eat accordingly. In the recessive form, supplementation is life-long and NaCl doses of > 50  mmol/kg/day are often required. Crisis with circulatory collapse is often precipitated by intercurrent infections, especially diarrhoea, and in our own practice we have used a portacath in some children to ensure easy venous access during these crises. The use of sodium-potassium exchange resins can also provide a more sustained means of giving sodium and removing potassium (Rosenberg et al., 1980; Saule et al., 1984; Porter et al., 2003).

References Asano, T., Abe, M., Asai, M., et al. (2006). Urinary tract malformation and infection with hyperkalemia and decreased fractional excretion of potassium in an infant. J Nippon Med Sch, 73(5), 289–91. Bailey, M. A., Cantone, A., Yan, Q., et al. (2006). Maxi-K channels contribute to urinary potassium excretion in the ROMK-deficient mouse model of Type II Bartter’s syndrome and in adaptation to a high-K diet. Kidney Int, 70(1), 51–9. Bandulik, S., Schmidt, K., Bockenhauer, D., et al. (2011). The salt-wasting phenotype of EAST syndrome, a disease with multifaceted symptoms linked to the KCNJ10 K+ channel. Pflugers Arch, 461(4), 423–35. Bartter, F. C., Pronove, P., Gill, J. R., Jr, et al. (1962). Hyperplasia of the juxtaglomerular complex with hyperaldosteronism and hypokalemic alkalosis. A new syndrome. Am J Med, 33, 811–28. Bergwitz, C., Roslin, N. M., Tieder, M., et al. (2006). SLC34A3 mutations in patients with hereditary hypophosphatemic rickets with hypercalciuria predict a key role for the sodium-phosphate cotransporter NaPi-IIc in maintaining phosphate homeostasis. Am J Hum Genet, 78(2), 179–92. Bettinelli, A., Ciarmatori, S., Cesareo, L., et al. (2000). Phenotypic variability in Bartter syndrome type I. Pediatr Nephrol, 14(10–11), 940–5. Birkenhager, R., Otto, E., Schürmann, M. J., et al. (2001). Mutation of BSND causes Bartter syndrome with sensorineural deafness and kidney failure. Nat Genet, 29(3), 310–4. Bleich, M. (2009). Membrane physiology—bridging the gap between medical disciplines. N Engl J Med, 360(19), 2012–14. Bockenhauer, D. (2008). Diabetes insipidus. In D. F. Geary and F. Schaefer (eds.) Comprehensive Pediatric Nephrology, pp. 489–98. Philadelphia, PA: Mosby Elsevier. Bockenhauer, D., Cruwys, M., Kleta, R., et al. (2008). Antenatal Bartter’s syndrome: why is this not a lethal condition? Q J Med, 101(12), 927–42. Bockenhauer, D., Feather, S., Stanescu, H. C., et al. (2009). Epilepsy, ataxia, sensorineural deafness, tubulopathy, and KCNJ10 mutations. N Engl J Med, 360(19), 1960–70. Bockenhauer, D., van’t Hoff, W., Dattani, M., et al. (2010). Secondary nephrogenic diabetes insipidus as a complication of inherited renal diseases. Nephron Physiol, 116(4), 23–9. Briet, M., Vargas-Poussou, R., Lourdel, S., et al. (2006). How Bartter’s and Gitelman’s syndromes, and Dent’s disease have provided important insights into the function of three renal chloride channels: ClC-Ka/b and ClC-5. Nephron Physiol, 103(1), 7–13.

CHAPTER 31 

Brochard, K., Boyer, O., Blanchard, A., et al. (2009). Phenotype-genotype correlation in antenatal and neonatal variants of Bartter syndrome. Nephrol Dial Transplant, 24(5), 1455–64. Brum, S., Rueff, J., Santos, J. R., et al. (2007). Unusual adult-onset manifestation of an attenuated Bartter’s syndrome type IV renal phenotype caused by a mutation in BSND. Nephrol Dial Transplant, 22(1), 288–9. Cantone, A., Yang, X., Yan, Q., et al. (2008). Mouse model of type II Bartter’s syndrome. I. Upregulation of thiazide-sensitive Na-Cl cotransport activity. Am J Physiol Renal Physiol, 294(6), F1366–72. Chang, S. S., Grunder, S., Hanukoglu, A., et al. (1996). Mutations in subunits of the epithelial sodium channel cause salt wasting with hyperkalaemic acidosis, pseudohypoaldosteronism type 1. Nat Genet, 12(3), 248–53. Cheek, D. B. and Perry, J. W. (1958). A salt wasting syndrome in infancy. Arch Dis Child, 33(169), 252–6. Cobeta-Garcia, J. C., Gascón, A., Iglesias, E., et al. (1998). Chondrocalcinosis and Gitelman’s syndrome. A new association? Ann Rheum Dis, 57(12), 748–9. Cruz, D. N., Shaer, A. J., Bia, M. J., et al. (2001). Gitelman’s syndrome revisited: an evaluation of symptoms and health-related quality of life. Kidney Int, 59(2), 710–7. Dillon, M. J., Leonard, J. V., Buckler, J. M., et al. (1980). Pseudohypoaldosteronism. Arch Dis Child, 55(6), 427–34. Dillon, M. J., Shah, V., and Mitchell, M. D. (1979). Bartter’s syndrome: 10 cases in childhood. Results of long-term indomethacin therapy. Q J Med, 48(191), 429–46. Dogne, J. M., Hanson, J., Supuran, C., et al. (2006). Coxibs and cardiovascular side-effects: from light to shadow. Curr Pharm Des, 12(8), 971–5. Ea, H. K., Blanchard, A., Dougados, M., et al. (2005). Chondrocalcinosis secondary to hypomagnesemia in Gitelman’s syndrome. J Rheumatol, 32(9), 1840–2. Emma, F., Pizzini, C., Tessa, A., et al. (2006). ‘Bartter-like’ phenotype in Kearns-Sayre syndrome. Pediatr Nephrol, 21(3), 355–60. Finer, G., Shalev, H., Birk, O. S., et al. (2003). Transient neonatal hyperkalemia in the antenatal (ROMK defective) Bartter syndrome. J Pediatr, 142(3), 318–23. Foglia, P. E., Bettinelli, A., Tosetto, C., et al. (2004). Cardiac work up in primary renal hypokalaemia-hypomagnesaemia (Gitelman syndrome). Nephrol Dial Transplant, 19(6), 1398–402. Francis, J., Zhang, J., Farhi, A., et al. (2004). A novel SGLT2 mutation in a patient with autosomal recessive renal glucosuria. Nephrol Dial Transplant, 19(11), 2893–5. Geller, D. S., Rodriguez-Soriano, J., Vallo Boado, A., et al. (1998). Mutations in the mineralocorticoid receptor gene cause autosomal dominant pseudohypoaldosteronism type I. Nat Genet, 19(3), 279–81. Goto, Y., Itami, N., Kajii, N., et al. (1990). Renal tubular involvement mimicking Bartter syndrome in a patient with Kearns–Sayre syndrome. J Pediatr, 116(6), 904–10. Greenberg, D., Abramson, O., and Phillip, M. (1995). Fetal pseudohypoaldosteronism: another cause of hydramnios. Acta Paediatr, 84(5), 582–4. Haisch, L., Almeida, J. R., Abreu da Silva, P. R., et al. (2011). The role of tight junctions in paracellular ion transport in the renal tubule: lessons learned from a rare inherited tubular disorder. Am J Kidney Dis, 57(2), 320–30. Hummler, E. and Horisberger, J. D. (1999). Genetic disorders of membrane transport. V. The epithelial sodium channel and its implication in human diseases. Am J Physiol, 276(3 Pt 1), G567–71. Ichikawa, S., Sorenson, A. H., Imel, E. A., et al. (2006). Intronic deletions in the SLC34A3 gene cause hereditary hypophosphatemic rickets with hypercalciuria. J Clin Endocrinol Metab, 91(10), 4022–7. Jeck, N., Konrad, M., Peters, M., et al. (2000). Mutations in the chloride channel gene, CLCNKB, leading to a mixed Bartter-Gitelman phenotype. Pediatr Res, 48(6), 754–8. Jentsch, T. J., Poët, M., Fuhrmann, J. C., et al. (2005). Physiological functions of CLC Cl- channels gleaned from human genetic disease and mouse models. Annu Rev Physiol, 67, 779–807.

salt-wasting tubulopathies

Kashimada, K., Omori, T., Takizawa, F., et al. (2008). Two cases of transient pseudohypoaldosteronism due to group B streptococcus pyelonephritis. Pediatr Nephrol, 23(9), 1569–70. Kerem, E., Bistritzer, T., Hanukoglu, A., et al. (1999). Pulmonary epithelial sodium-channel dysfunction and excess airway liquid in pseudohypoaldosteronism. N Engl J Med, 341(3), 156–62. Kim, J. Y., Kim, G. A., Song, J. H., et al. (2000). A case of living-related kidney transplantation in Bartter’s syndrome. Yonsei Med J, 41(5), 662–5. Kleta, R. and Bockenhauer, D. (2006). Bartter syndromes and other salt-losing tubulopathies. Nephron Physiol, 104(2), p73–80. Kleta, R., Basoglu, C., and E. Kuwertz-Broking, E. (2000). New treatment options for Bartter’s syndrome. N Engl J Med, 343(9), 661–2. Kleta, R., Brune, T., and Harms, E. (1999). Cystic fibrosis and metabolic alkalosis in children—revisited. Miner Electrolyte Metab, 25(3), 210. Knoers, N. V. and Levtchenko, E. N. (2008). Gitelman syndrome. Orphanet J Rare Dis, 3, 22. Komhoff, M., Jeck, N. D., Seyberth, H. W., et al. (2000). Cyclooxygenase-2 expression is associated with the renal macula densa of patients with Bartter-like syndrome. Kidney Int, 58(6), 2420–4. Konrad, M., Leonhardt, A., Hensen, P., et al. (1999). Prenatal and postnatal management of hyperprostaglandin E syndrome after genetic diagnosis from amniocytes. Pediatrics, 103(3), 678–83. Konrad, M., Schaller, A., Seelow, D., et al. (2006). Mutations in the tight-junction gene claudin 19 (CLDN19) are associated with renal magnesium wasting, renal failure, and severe ocular involvement. Am J Hum Genet, 79(5), 949–57. Konrad, M., Vollmer, M., Lemmink, H. H., et al. (2000). Mutations in the chloride channel gene CLCNKB as a cause of classic Bartter syndrome. J Am Soc Nephrol, 11(8), 1449–59. Kramer, B. K., Bergler, T., Stoelcker, B., et al. (2008). Mechanisms of disease: the kidney-specific chloride channels ClCKA and ClCKB, the Barttin subunit, and their clinical relevance. Nat Clin Pract Nephrol, 4(1), 38–46. Liaw, L. C., Banerjee, K., and Coulthard, M. G. (1999). Dose related growth response to indometacin in Gitelman syndrome. Arch Dis Child, 81(6), 508–10. Lorenz-Depiereux, B., Benet-Pages, A., Eckstein, G., et al. (2006). Hereditary hypophosphatemic rickets with hypercalciuria is caused by mutations in the sodium-phosphate cotransporter gene SLC34A3. Am J Hum Genet, 78(2), 193–201. Magen, D., Berger, L., Coady, M. J., et al. (2010). A loss-of-function mutation in NaPi-IIa and renal Fanconi’s syndrome. N Engl J Med, 362(12), 1102–9. Martin, J. M., Calduch, L., Monteagudo, C., et al. (2005). Clinico-pathological analysis of the cutaneous lesions of a patient with type I pseudohypoaldosteronism. J Eur Acad Dermatol Venereol, 19(3), 377–9. Monnens, L., Bindels, R., and Grunfeld, J. P. (1998). Gitelman syndrome comes of age. Nephrol Dial Transplant, 13(7), 1617–9. Nijenhuis, T., Vallon, V., van der Kemp, A. W., et al. (2005). Enhanced passive Ca2+ reabsorption and reduced Mg2+ channel abundance explains thiazide-induced hypocalciuria and hypomagnesemia. J Clin Invest, 115(6), 1651–8. Nozu, K., Fu, X. J., Kaito, H., et al. (2007). A novel mutation in KCNJ1 in a Bartter syndrome case diagnosed as pseudohypoaldosteronism. Pediatr Nephrol, 22(8), 1219–23. Peters, M., Jeck, N., Reinalter, S., et al. (2002). Clinical presentation of genetically defined patients with hypokalemic salt-losing tubulopathies. Am J Med, 112(3), 183–90. Peti-Peterdi, J., Komlosi, P., Fuson, A. L., et al. (2003). Luminal NaCl delivery regulates basolateral PGE2 release from macula densa cells. J Clin Invest, 112(1), 76–82. Porter, J., Kershaw, M., Kirk, J., et al. (2003). The use of sodium resonium in pseudohypoaldosteronism. Arch Dis Child, 88(12), 1138–9. Pressler, C. A., Heinzinger, J., Jeck, N., et al. (2006). Late-onset manifestation of antenatal Bartter syndrome as a result of residual function of the

289

290

Section 2  

fluid, electrolyte, and renal tubular disorders

mutated renal Na+-K+-2Cl- co-transporter. J Am Soc Nephrol, 17(8), 2136–42. Pumberger, W., Frigo, E., and Geissler, W. (1998). Transient pseudohypoaldosteronism in obstructive renal disease. Eur J Pediatr Surg, 8(3), 174–7. Reichold, M., Zdebik, A. A., Lieberer, E., et al. (2010). KCNJ10 gene mutations causing EAST syndrome (epilepsy, ataxia, sensorineural deafness, and tubulopathy) disrupt channel function. Proc Natl Acad Sci U S A, 107(32), 14490–5. Reinalter, S. C., Jeck, N., Brochhausen, C., et al. (2002). Role of cyclooxygenase-2 in hyperprostaglandin E syndrome/antenatal Bartter syndrome. Kidney Int, 62(1), 253–60. Reinalter, S. C., Jeck, N., Peters, M., et al. (2004). Pharmacotyping of hypokalaemic salt-losing tubular disorders. Acta Physiol Scand, 181(4), 513–21. Riepe, F. G. (2009). Clinical and molecular features of type 1 pseudohypoaldosteronism. Horm Res, 72(1), 1–9. Riveira-Munoz, E., Chang, Q., Godefroid, N., et al. (2007). Transcriptional and functional analyses of SLC12A3 mutations: new clues for the pathogenesis of Gitelman syndrome. J Am Soc Nephrol, 18(4), 1271–83. Rodriguez-Soriano, J., Vallo, A., Oliveros, R., et al. (1983). Transient pseudohypoaldosteronism secondary to obstructive uropathy in infancy. J Pediatr, 103(3), 375–80. Rogers, D. (2008). Final diagnosis: transient pseudohypoaldosteronism (TPH) caused by UTI without concordant obstructive uropathy. Clin Pediatr (Phila), 47(4), 405–8. Rosenberg, S., Franks, R. C., and Ulick, S. (1980). Mineralocorticoid unresponsiveness with severe neonatal hyponatremia and hyperkalemia. J Clin Endocrinol Metab, 50(2), 401–4. Saule, H., Dorr, H. G., and Sippell, W. G. (1984). Pseudohypoaldosteronism in a child with Down syndrome. Long-term management of salt loss by ion exchange resin administration. Eur J Pediatr, 142(4), 286–9. Savage, M. O., Jefferson, I. G., Dillon, M. J., et al. (1982). Pseudohypoaldosteronism: severe salt wasting in infancy caused by generalized mineralocorticoid unresponsiveness. J Pediatr, 101(2), 239–42. Schaedel, C., Marthinsen, L., Kristoffersson, A. C., et al. (1999). Lung symptoms in pseudohypoaldosteronism type 1 are associated with deficiency of the alpha-subunit of the epithelial sodium channel. J Pediatr, 135(6), 739–45. Schlingmann, K. P., Konrad, M., Jeck, N., et al. (2004). Salt wasting and deafness resulting from mutations in two chloride channels. N Engl J Med, 350(13), 1314–9. Schnermann, J. (2003). Homer W. Smith Award lecture. The juxtaglomerular apparatus: from anatomical peculiarity to physiological relevance. J Am Soc Nephrol, 14(6), 1681–94. Schoen, E. J., Bhatia, S., Ray, G. T., et al. (2002). Transient pseudohypoaldosteronism with hyponatremia-hyperkalemia in infant urinary tract infection. J Urol, 167(2 Pt 1), 680–2. Scholl, U. I., Choi, M., Liu, T., et al. (2009). Seizures, sensorineural deafness, ataxia, mental retardation, and electrolyte imbalance (SeSAME syndrome) caused by mutations in KCNJ10. Proc Natl Acad Sci U S A, 106(14), 5842–7.

Schultheis, P. J., Clarke, L. L., Meneton, P., et al. (1998). Renal and intestinal absorptive defects in mice lacking the NHE3 Na+/H+ exchanger. Nat Genet, 19(3), 282–5. Scognamiglio, R., Negut, C., and Calo, L. A. (2007). Aborted sudden cardiac death in two patients with Bartter’s/Gitelman’s syndromes. Clin Nephrol, 67(3), 193–7. Seyberth, H. W., Rascher, W., Schweer, H., et al. (1985). Congenital hypokalemia with hypercalciuria in preterm infants: a hyperprostaglandinuric tubular syndrome different from Bartter syndrome. J Pediatr, 107(5), 694–701. Simon, D. B., Bindra, R. S., Mansfield, T. A., et al. (1997). Mutations in the chloride channel gene, CLCNKB, cause Bartter’s syndrome type III. Nat Genet, 17(2), 171–8. Simon, D. B., Karet, F. E., Hamdan, J. M., et al. (1996a). Bartter’s syndrome, hypokalaemic alkalosis with hypercalciuria, is caused by mutations in the Na-K-2Cl cotransporter NKCC2. Nat Genet, 13(2), 183–8. Simon, D. B., Karet, F. E., Rodriguez-Soriano, J., et al. (1996b). Genetic heterogeneity of Bartter’s syndrome revealed by mutations in the K+ channel, ROMK. Nat Genet, 14(2), 152–6. Simon, D. B., Lu, Y., Choate, K. A., et al. (1999). Paracellin-1, a renal tight junction protein required for paracellular Mg2+ resorption. Science, 285(5424), 103–6. Simon, D. B., Nelson-Williams, C., Bia, M. J., et al. (1996c). Gitelman’s variant of Bartter’s syndrome, inherited hypokalaemic alkalosis, is caused by mutations in the thiazide-sensitive Na-Cl cotransporter. Nat Genet, 12(1), 24–30. Su, I. H., Frank, R., Gauthier, B. G., et al. (2000). Bartter syndrome and focal segmental glomerulosclerosis: a possible link between two diseases. Pediatr Nephrol, 14(10–11), 970–2. Turk, E., Zabel, B., Mundlos, S., et al. (1991). Glucose/galactose malabsorption caused by a defect in the Na+/glucose cotransporter. Nature, 350(6316), 354–6. Uribarri, J., Oh, M. S., Butt, K. M., et al. (1982). Pseudohypoaldosteronism following kidney transplantation. Nephron, 31(4), 368–70. Van den Heuvel, L. P., Assink, K., Willemsen, M., et al. (2002). Autosomal recessive renal glucosuria attributable to a mutation in the sodium glucose cotransporter (SGLT2). Hum Genet, 111(6), 544–7. Wagner, C. A., Loffing-Cueni, D., Yan, Q., et al. (2008). Mouse model of type II Bartter’s syndrome. II. Altered expression of renal sodium- and water-transporting proteins. Am J Physiol Renal Physiol, 294(6), F1373–80. Watanabe, S., Fukumoto, S., Chang, H., et al. (2002). Association between activating mutations of calcium-sensing receptor and Bartter’s syndrome. Lancet, 360(9334), 692–4. Wong, G. P. and Levine, D. (1998). Congenital pseudohypoaldosteronism presenting in utero with acute polyhydramnios. J Matern Fetal Med, 7(2), 76–8. Yang, T., Endo, Y., Huang, Y. G., et al. (2000). Renin expression in COX-2-knockout mice on normal or low-salt diets. Am J Physiol Renal Physiol, 279(5), F819–25. Zawada, E. T., Jr. (1982). Renal consequences of nonsteroidal antiinflammatory drugs. Postgrad Med, 71(5), 223–30. Zelikovic, I., Szargel, R., Hawash, A., et al. (2003). A novel mutation in the chloride channel gene, CLCNKB, as a cause of Gitelman and Bartter syndromes. Kidney Int, 63(1), 24–32.

CHAPTER 32

Approach to the patient with polyuria Daniel G. Bichet Hypotonic and hypertonic urine in polyuric states In the absence of a glucose-induced osmotic diuresis in uncontrolled diabetes mellitus, a hypertonic polyuric state, there are three major causes of hypotonic polyuria, each due to a defect in water balance, leading to the excretion of large volumes of dilute urine (urine osmolality usually < 250 mOsm/kg): primary polydipsia, central diabetes insipidus, and nephrogenic diabetes insipidus (NDI) (Fig. 32.1). Simple, inexpensive blood and urine measurements, together with clinical characteristics and magnetic resonance imaging (MRI) could distinguish between these three aetiologies (Chanson and Salenave, 2011).

Polyuria and nocturia, nocturnal polyuria in enuretic children Polyuria could be constant during the day, but also present at night:  the urine is normally most concentrated in the morning due to lack of fluid ingestion overnight and increased vasopressin secretion during the late sleep period (Trudel and Bourque, 2011). Neurons in the suprachiasmatic nucleus, the brain biological clock, send axonal projections towards the supraoptic nucleus, one of the hypothalamic nuclei producing vasopressin (Burbach et al., 2001), providing a possible anatomical substrate for the circadian modulation, an osmoregulatory gain during the late sleep period (Trudel and Bourque, 2011). As a result, the first manifestation of a mild to moderate loss of concentrating ability is often nocturia. However, nocturia is not diagnostic of a defect in concentrating ability since it can also be caused by other factors such as drinking before going to bed or, in men, by prostatic hypertrophy, which is characterized by urinary frequency rather than polyuria. Psychogenic polydipsic patients tend to ingest large amounts of fluid during the day but not at night, therefore nocturia is rarely seen in primary polydipsic patients (Barlow and de Wardener, 1959). The pattern of nocturnal polyuria in enuretic children is similar to that observed in acute sleep deprivation and enuresis in children might be related to the failure of sleep to cause a reflex reduction in arterial pressure and urine production (Denton, 2012; Mahler et al., 2012).

Plasma sodium and osmolality Plasma sodium (Na+) and osmolality are maintained within normal limits (136–143 mEq/L for plasma Na+; 275–290 mOsmol/kg for

plasma osmolality) by a thirst–arginine vasopressin (AVP)–renal axis (Bourque, 2008; Lechner et al., 2011). Thirst and AVP release, both stimulated by increased osmolality, is a ‘double-negative’ feedback system (Leaf, 1979). Even when the AVP component of this ‘double-negative’ regulatory feedback system is lost, the thirst mechanism still preserves the plasma Na+ and osmolality within the normal range, but at the expense of pronounced polydipsia and polyuria. Thus, the plasma Na+ concentration or osmolality of an untreated patient with diabetes insipidus with unlimited access to water may be slightly greater than the mean normal value, and a decrease in plasma Na+ and osmolality might be observed in primary polydipsic patients, but these small increases have no diagnostic significance (Babey et  al., 2011). Polyuric patients should be asked about their thirst and their way to quench it: cold water will quench thirst more effectively in severely polyuric and dehydrated patients, irrespective of their aetiology (central versus nephrogenic). Primary polydipsic patients may tend to absorb large quantities of water ice-cold or not. Glucose-induced osmotic diuresis is more frequent than any cause of non-osmotic polyuria. High plasma glucose levels with polyuria could also be observed in brain-dead patients with diabetes insipidus receiving glucose infusions at a rate exceeding 500 mL/h, which corresponds to the maximum (25 g/h) capacity for glucose metabolism. The polyuria observed in post-obstructive diuresis is appropriate, representing an attempt to excrete the fluid retained during the period of obstruction (Bichet, 2011).

Mammals are osmoregulators—the cellular perception of tonicity Mammals are osmoregulators: they have evolved mechanisms that maintain extracellular fluid (ECF) osmolality near a stable value. Yet, although mammals strive to maintain a constant ECF osmolality, values measured in an individual can fluctuate around the set-point owing to intermittent changes in the rates of water intake and water loss (through evaporation or diuresis) and to variations in the rates of Na+ intake and excretion (natriuresis). In humans, for example, 40 minutes of strenuous exercise in the heat (Saat et al., 2005; Edwards et al., 2007), or 24 hours of water deprivation (Shirreffs et al., 2004) causes plasma osmolality to rise by more than 10 mOsm/kg. In a dehydrated individual, drinking the equivalent of two large glasses of water (~850 mL) lowers osmolality by approximately 6 mOsm/kg within 30 minutes (Geelen et al., 1996). Similarly, ingestion of 13 g of salt increases plasma osmolality by

292

Section 2  

fluid, electrolyte, and renal tubular disorders Polyuria (3L/day in adults; 2L/m2 in children)

U osm > 300 mOsm/kg H2O: Hypertonic urine – Glucose – Mannitol, Maltose – Glycine

U osm > 300 mOsm/kg H2O: Hypotonic urine – Primary polydipsia – Central diabetes insipidus – Nephrogenic diabetes insipidus

Fig. 32.1  Hypotonic and hypertonic polyuric states.

approximately 5 mOsm/kg within 30 minutes (Andersen et  al., 2000). Although osmotic perturbations larger than these can be deleterious to health, changes in the 1–3% range play an integral part in the control of body-fluid homeostasis. Differences between the ECF osmolality and the desired set-point induce proportional homeostatic responses according to the principle of negative feedback (Bourque, 2008). ECF hyperosmolality stimulates the sensation of thirst, to promote water intake, and the release of vasopressin to enhance water reabsorption in the kidney. By contrast,

ECF hypo-osmolality suppresses basal vasopressin secretion in rats and humans (Claybaugh et al., 2000). As summarized elegantly by Bourque (2008) early studies provided clear evidence that ‘cellular dehydration’ (i.e. cell shrinking) was required for thirst and vasopressin release to be stimulated during ECF hyperosmolality:  these responses could be induced by infusions of concentrated solutions containing membrane-impermeable solutes, which extract water from cells, but not by infusions of solutes that readily equilibrate across the cell membrane (such as urea). Verney coined the term osmoreceptor to designate the specialized sensory elements. He further showed that these were present in the brain and postulated that they might comprise ‘tiny osmometers’ and ‘stretch receptors’ that would allow osmotic stimuli to be ‘transmuted into electrical’ signals (Verney, 1947). Osmoreceptors are, therefore, defined functionally as neurons that are endowed with an intrinsic ability to detect changes in ECF osmolality and it is now known that both cerebral and peripheral osmoreceptors contribute to the body fluid balance (Fig. 32.2). Although magnocellular neurons are themselves osmosensitive, they require input, by glutamatergic afferents, from the lamina terminalis to respond fully to osmotic challenges (Fig. 32.3). Neurons in the lamina terminalis are also osmosensitive and because the subfornical organ (SFO) and the organum vasculosum of the

(a) Hypertonic

Control

–55 mV

VP Antidiuresis Ang II

Set point

–60 mV

+ 10 mosmoL/kg

Basal VP

+ 10 mosmoL/kg

1 min

Set point

Diuresis –65 mV

VP

Hypotonic TRENS in Neurosciences

Fig. 32.2  (Left) Cell autonomous osmoreception in vasopressin neurons. Changes in osmolality cause inversely proportional changes in soma volume. Shrinkage activates non-selective cation channels (NSCCs) and the ensuing depolarization increases action potential firing rate and vasopressin (VP) release from axon terminals in the neurohypophysis. Increased VP levels in blood enhance water reabsorption by the kidney (antidiuresis) to restore extracellular fluid osmolality towards the set point. Hypotonic stimuli inhibit NSCCs. The resulting hyperpolarization and inhibition of firing reduces VP release and promotes diuresis. (Upper right) Whole-cell current clamp recordings from isolated MNCs and averaged data from multiple cells show that the depolarizing and action potential firing responses induced by a hypertonic stimulus are significantly enhanced in the presence of 100 nM Ang II. (Lower right) Hypothetical events mediating central Ang II enhancement of osmosensory gain. Ang II released by afferent nerve terminals (e.g. during hypovolaemia) binds to AT1 receptor (AT1R) coupled to G proteins such as Gq or/and G12/13. Activated G proteins signal through phospholipase C (PLC) and protein kinase C (PKC) to activate a RhoA-specific guanine nucleotide exchange factor (RhoA–GEF), such as p115RhoGEF or LARG (leukaemia-associated Rho guanine–nucleotide exchange factor) Activation of RhoA–GEF converts inactive cytosolic RhoA (RhoA–GDP) into active, membrane-associated RhoA–GTP by promoting the exchange of GDP to GTP. Activated RhoA induces actin polymerization and increases submembrane F-actin density to enhance the mechanical gating of non-specific cation channels. With permission from Masha Prager-Khoutorsky and Charles W. Bourque: Osmosensation in vasopressin neurons: changing actin density to optimize function; Trends Neurosci. 2010 Feb; 33(2):76–83

chapter 32 

SFO

MnPO

PVN

OVLT

SON

Pit iME

OC

Fig. 32.3  Schematic representation of the osmoregulatory pathway of the hypothalamus (sagittal section of midline of ventral brain around the third ventricle in mice). Neurons (lightly filled circles) in the lamina terminalis (OVLT), median preoptic nucleus (MnPO), and subfornical organ (SFO)—that are responsive to plasma hyptertonicity send efferent axonal projections (grey lines) to magnocellular neurons of the paraventricular (PVN) and supraoptic nuclei (SON). The OVLT is one of the brain circumventricular organs and is a key osmosensing site in the mammalian brain (vide infra). The processes (dark lines) of these magnocellular neurons form the hypothalamo-neurohypophysial pathway that courses in the median eminence to reach the posterior pituitary, where neurosecretion of vasopressin and oxytocin occurs. Modified from Wilson et al., Visualization of functionally activated circuitry in the brain. PNAS, 99: 3252–7, 2002.

lamina terminalis (OVLT) lie outside the blood–brain barrier, they can integrate this information with endocrine signals borne by circulating hormones, such as angiotensin II (Ang-II), relaxin, and atrial natriuretic peptide (ANP). While circulating Ang-II and relaxin excite both oxytocin and vasopressin magnocellular neurons, ANP inhibits vasopressin neurons. The non-osmotic pathways are more physiologically described now as ‘osmoregulatory gain,’ since Ang II amplifies osmosensory transduction by enhancing the proportional relationship between osmolality, receptor potential, and action potential firing in rat supraoptic nucleus neurons (Zhang and Bourque, 2008). Modifications in osmoregulatory gain induced by angiotensin explain why the changes in the slope and threshold of the relationship between plasma osmolality and vasopressin secretion are potentiated by hypovolaemia or hypotension and are attenuated by hypervolaemia or hypertension (Robertson and Athar, 1976).

Tonicity information is relayed by central osmoreceptor neurons expressing TRPV1 and peripheral osmoreceptor neurons expressing TRPV4 The osmotic regulation of the release of AVP from the posterior pituitary is primarily dependent, under normal circumstances, on tonicity information relayed by central osmoreceptor neurons expressing TRPV1 (Bourque, 2008) and peripheral osmoreceptor neurons expressing TRPV4 (Lechner et al., 2011). The cellular basis for osmoreceptor potentials has been characterized using patch-clamp recordings and morphometric analysis in magnocellular cells isolated from the supraoptic nucleus of the adult rat. In these cells, stretch-inactivating cationic channels transduce osmotically evoked changes in cell volume into functionally relevant changes in membrane potential. In addition, magnocellular neurons also operate as intrinsic Na+ detectors. The N-terminal variant of the transient receptor potential cation

approach to the patient with polyuria

channel subfamily V member 1 (TRPV1) is an osmotically activated channel expressed in the magnocellular cells producing vasopressin (Sharif Naeini et al., 2006) and in the circumventricular organs, the OVLT, and the SFO (Ciura and Bourque, 2006). Since osmoregulation still operates in Trpv1−/− mice, other osmosensitive neurons or pathways must be able to compensate for loss of central osmoreceptor function (Ciura and Bourque, 2006; Sharif Naeini et al., 2006; Taylor et al., 2008). Afferent neurons expressing the osmotically activated ion channel, TRPV4 in the thoracic dorsal root ganglia that innervate hepatic blood vessels and detect physiological hypo-osmotic shifts in blood osmolality have recently been identified (Lechner et al., 2011). In mice lacking the osmotically activated ion channel, TRPV4, hepatic sensory neurons no longer exhibit osmosensitive inward currents and activation of peripheral osmoreceptors in vivo is abolished. In a large cohort of human liver transplantees, who presumably have denervated livers, plasma osmolality is significantly elevated compared with healthy controls, suggesting the presence of an inhibitory vasopressin effect of hyponatraemia, perceived in the portal vein from hepatic afferents (Lechner et al., 2011). TRPV1 (expressed in central neurons) and TRPV4 (expressed in peripheral neurons) thus appear to play entirely complementary roles in osmoreception. Lechner et al. (2011) have thus identified the primary afferent neurons that constitute the afferent arc of a well-characterized reflex in man and more recently also in rodents (McHugh et al., 2010). This reflex engages the sympathetic nervous system to raise blood pressure and stimulate metabolism (Tank et al., 2003; Boschmann et al., 2007). Of clinical interest, it has already been demonstrated that orthostatic hypotension and postprandial hypotension respond to water drinking (Jordan et al., 2000; Schroeder et al., 2002; Shannon et  al., 2002). Moreover, water drinking in man can prevent neutrally mediated syncope during blood donation or after prolonged standing (Claydon et al., 2006). Finally, water drinking is also associated with weight loss in overweight individuals (Stookey et al., 2008). Other peripheral sensory neurons expressing other mechanosensitive proteins may also be involved in osmosensitivity (Coste et al., 2010).

Quantification of polyuria—volume, osmolality, Na+, K+, and Ca2+ Polyuria is arbitrarily defined as a urine output exceeding 3 L/day in adults and 2 L/m2 in children (Bichet, 2011). It is important to obtain a 24-hour urine collection with measurements of volume, osmolality, Na+, potassium (K+), and calcium (Ca2+) to quantify precisely polyuric symptoms, since polyuria is difficult to measure in young infants and may even be confused with congenital chloride diarrhoea in patients referred with a suspected diagnosis of Bartter syndrome (Choi et al., 2009). Volume loss from the urinary tract versus the gastrointestinal tract may not be immediately discriminated in infants! Conversely, increased urinary frequency in men with prostatic hypertrophy might be confused with polyuria. The maximal attainable urine volume in normal individuals on a regular diet is > 10 L/day (e.g. 10 L of urine at a urine osmolality of 60 mOsm/kg of H2O to excrete a 600 mOsm solute load). If the solute load to be excreted is increased due to increased protein intake (generating urea which accounts for two-thirds of urine solutes) or increased Na+/K+ intake, 1200 mOsm will need to be excreted, representing 20 L of urine with a urine osmolality of 60 mOsm/kg.

293

Section 2  

fluid, electrolyte, and renal tubular disorders

Once plasma Na+, osmolality, and glucose measurements are obtained, the following steps should guide the investigation and treatment of polyuric states: measure 24-hour urine volume and record urine osmolality, Na+, K+, Ca2+, and glucose. A low-salt, low-protein diet will diminish (see earlier) urine output in both central and NDI, but will not change maximal urine osmolality (Earley and Orloff, 1962). In central diabetes insipidus, the effectiveness of vasopressin replacement by desmopressin makes a low salt, low protein diet irrelevant for the treatment of this condition. The approach to a polyuric patient will vary according to the age at presentation. Polyuric conditions in infants < 1 year of age are true emergency conditions, since young infants are unable to express their thirst and may suffer from severe dehydration and volume contraction. Repeated measures of plasma electrolyte, creatinine and urine volume and content may also be challenging in young infants.

Congenital (i.e. present at birth) and early polyuric states Polyuria in an infant with polyhydramnios during the pregnancy leading to her/his birth and prematurity The triad:  polyuria/polyhydramnios/prematurity is a tell-tale sign of Bartter syndrome with abnormal conservation of water but also of Na+, K+, chloride (Cl−), and Ca2+. Bartter syndrome (OMIM 601678, 241200, 607364, and 62522)  refers to a group of autosomal recessive disorders caused by inactivating mutations in one of four genes (SLC12A1, KCNJ1, CLCNKB, CLCNKA and CLCNKB in combination, or BSND) that encode the membrane proteins of the thick ascending limb of the loop of Henle (Puricelli et al., 2010; Bonnardeaux and Bichet, 2012). Since 30% of the filtered sodium chloride is reabsorbed in the thick ascending loop of Henle it is evident that the loss of function of these membrane transporters will induce alterations in the counter-current system (Fig. 32.5). In the

Outer and inner medullary collecting duct

Syntaxin 4

Gi +

Endocytic retrieval

V2 receptor

AQP3 Recycling vesicle

AVP

PDEs

Endocytic vesicle associated PKA c c

Microtubule Luminal

Gαs

Adenylyl cyclase

Actin AP c AK filment c A motor ct in fila m en t

AQP2

ATP AKAP

AQP2 H2O

Exocytic insertion

AKAP

294

c c

C2

C1

cAMP

Microtubule motor

Basolateral AQP4

Fig. 32.4  Schematic representation of the effect of vasopressin (AVP) to increase water permeability in the principal cells of the collecting duct and representation of two types of ‘pure’ nephrogenic diabetes insipidus: loss of function of the AVPR2 protein or loss of function of the AQP2 protein. AVP is bound to the V2 receptor (a G-protein-linked receptor) on the basolateral membrane. The basic process of G-protein-coupled receptor signalling consists of three steps: a hepta-helical receptor that detects a ligand (in this case, AVP) in the extracellular milieu, a G-protein (G∝s) that dissociates into α subunits bound to GTP and βγ subunits after interaction with the ligand-bound receptor, and an effector (in this case, adenylyl cyclase) that interacts with dissociated G-protein subunits to generate small-molecule second messengers. AVP activates adenylyl cyclase, increasing the intracellular concentration of cAMP. The topology of adenylyl cyclase is characterized by two tandem repeats of six hydrophobic transmembrane domains separated by a large cytoplasmic loop and terminates in a large intracellular tail. The dimeric structure (C1 and C2) of the catalytic domains is represented. Conversion of ATP to cAMP takes place at the dimer interface. Two aspartate residues (in C1) coordinate two metal co-factors (Mg2+ or Mn2+ represented here as two small black circles), which enable the catalytic function of the enzyme. Adenosine is shown as an open circle and the three phosphate groups (ATP) are shown as smaller open circles. Protein kinase A (PKA) is the target of the generated cAMP. The binding of cAMP to the regulatory subunits of PKA induces a conformational change, causing these subunits to dissociate from the catalytic subunits. These activated subunits (C) as shown here are anchored to an aquaporin-2 (AQP2)-containing endocytic vesicle via an A-kinase anchoring protein. The local concentration and distribution of the cAMP gradient is limited by phosphodiesterases (PDEs). Cytoplasmic vesicles carrying the water channels (represented as homotetrameric complexes) are fused to the luminal membrane in response to AVP, thereby increasing the water permeability of this membrane. The dissociation of the A-kinase anchoring protein from the endocytic vesicle is not represented. Microtubules and actin filaments are necessary for vesicle movement towards the membrane. When AVP is not available, AQP2 water channels are retrieved by an endocytic process, and water permeability returns to its original low rate. Aquaporin-3 (AQP3) and aquaporin-4 (AQP4) water channels are expressed constitutively at the basolateral membrane.

chapter 32 

large experience of Hans Seyberth and colleagues (Peters et al., 2002), polyuria was the leading symptom postnatally in 19 of 32 patients.

approach to the patient with polyuria

Very early (first week of life) polyuric states are usually nephrogenic but we and others have observed autosomal recessive central diabetes insipidus patients with early polyuria, dehydration episodes responding to DDAVP with specific mutations of the AVP gene (Bichet et al., 1998; Willcutts et al., 1999; Abu Libdeh et al., 2009; Christensen et al., 2013). For complex (water + sodium + calcium) and pure (water only) early polyuric states a rapid molecular diagnosis cuts the ‘diagnostic odyssey’ that often involves false diagnostic leads and ineffective treatment (Fig. 32.6). We are recommending the sequencing of the NDI and Bartter genes in all the affected patients. The genes involved are, with a few exceptions, relatively small and easy to sequence. This genomic information is key to the routine care of patients with congenital polyuria and, as in other genetic diseases, reduces health costs and provides psychological benefits to patients and families (Green and Guyer, 2011) (Fig. 32.6).

‘Pure’ polyuria, that is, loss of water only but normal conservation of Na+, K+, Cl−, and Ca2+ in the first week of life Most of the cases are ‘pure’ NDI secondary to mutations in the vasopressin V2 receptor gene (AVPR2, X-linked, OMIM 304800)  or in the aquaporin-2 water channel gene (AQP2, autosomal recessive, OMIM 222000)  and dominant (OMIM 125800) (Fig. 32.4). The intensity of the polyuric manifestations will depend on the severity of the mutation identified: in both AVPR2 and AQP2 mutations, severe phenotypes (U Osm < 200 mOsm/kg) are observed with specific mutations and less severe phenotypes (U Osm~300 mOsm/kg) are observed with mild mutations (Bockenhauer et al., 2009; Guyon et al., 2009).

Thick ascending loop of Henle Blood

Lumen Lumen positive voltage

+

–

Na+ 2CI–

~

K+ Bartter’s type I

NKCC2 Bartter’s type IV

ROMK

CIC-Ka

CIC-Ka and CIC-Kb Digenic disorder

Barttin CI–

Cldn 16 FHHN

3Na+ CI–

K+ Bartter’s type II

2K+

CIC-Kb

Bartter’s type III

Ca2+, Mg2+, K+, NH4+

OMIM 248250 Cell +8 mV

Fig. 32.5  Schematic representation of transepithelial salt resorption in a cell of the thick ascending limb (TAL) of the loop of Henle. Thirty per cent of the filtered sodium chloride (NaCl) is reabsorbed in the TAL and most of the energy for concentration and dilution of the urine derives from active NaCl transport in the TAL. Filtered NaCl is reabsorbed through NKCC2, which uses the sodium gradient across the membrane to transport chloride and potassium into cell. The potassium ions are recycled (100%) through the apical membrane by the potassium channel ROMK. Sodium leaves the cell actively through the basolateral Na-K-ATPase. Chloride diffuses passively through two basolateral channels, ClC-Ka and ClC-Kb. Both of these chloride channels must bind to the β subunit of barttin to be transported to the cell surface. Four types of Bartter syndrome (types I, II, III, and IV) are attributable to recessive mutations in the genes that encode the NKCC2 cotransporter, the potassium channel (ROMK), one of the chloride channels (CIC-Kb), and barttin, respectively. A fifth type of Bartter syndrome has also been shown to be a digenic disorder that is attributable to loss-of-function mutations in the genes that encode the chloride channels CIC-Ka and CIC-Kb. As a result of these different molecular alterations, NaCl is lost into the urine, positive lumen voltage is abolished, and calcium (Ca2+), magnesium (Mg2+), potassium (K+), and ammonium (NH4+) cannot be reabsorbed in the paracellular space. In the absence of mutations, the recycling of potassium maintains a lumen-positive gradient (+8 mV). Claudin 16 (CLDN16) is necessary for the paracellular transport of calcium and magnesium. Modified from Bichet DG, Fujiwara TM: Reabsorption of sodium chloride—lessons from the chloride channels. N Engl J Med, 350:1281–1283, 2004.

295

296

Section 2  

fluid, electrolyte, and renal tubular disorders

A Young Patient Thirsty Since Birth

Email received on Mar 21, 2010: • My son was born in 2005. Ever since that date he has had near constant thirst. He would not take to breast feeding because the milk came out too slow. When he was of age to eat food he would scream for drink instead. • By age 2 we were starting to get quite concerned with his size. He was only 21 Ibs. He was also still contantly thirsty and urinating frequently. He would often wake up at night screaming frantically for a drink. • My son’s pediatrician currently thinks that his excessive thirst is behavioural. This is becoming very difficult since he is now finding more and more ways to sneak a drink secretly. Response: • Mrs., I do not think that the increased thirst and polyuria observed in your son are behavioural • We would appreciate receiving blood from your son and from you. • Email sent, April 19th: AVPR2 mutation V88M, mother is a carrier. • Urine flow: 6mL/min; UOsm: 55 pre and 57 post dDAVP.

Fig. 32.6  A young patient thirsty since birth.

Other nephrogenic disorders with polyuric manifestations

inhibitory effect of intracellular lithium on water transport (Batlle et al., 1985).

Here, polyuria is appearing later in life, usually after the first year. Polyuria will be observed with variable severity in Bardet–Biedl syndrome (Marion et  al., 2011), nephronophthisis (Hildebrandt et al., 2009), cystinosis, familial hypernatremia with hypervolemia and nephrocalcinosis, and the syndrome of apparent mineralocorticoid excess (Bockenhauer et al., 2010).

Diabetes insipidus and pregnancy

Acquired nephrogenic polyuric disorders Acquired NDI is much more common than congenital NDI, but it is rarely as severe. The ability to produce hypertonic urine is usually preserved even though there is inadequate concentrating ability of the nephron. Polyuria and polydipsia are therefore moderate (3–4 L/day). Among the more common causes of acquired NDI, lithium administration has become the most frequent cause; 54% of 1105 unselected patients on chronic lithium therapy developed NDI (Boton et al., 1987). Nineteen per cent of these patients had polyuria, as defined by a 24-hour urine output exceeding 3 L. The dysregulation of aquaporin-2 expression is the result of cytotoxic accumulation of lithium which enters via the epithelial sodium channel (ENaC) on the apical membrane and leads to the inhibition of signalling pathways that involve glycogen synthase kinase type 3 beta (Grunfeld and Rossier, 2009). The concentration of lithium in urine of patients on well-controlled lithium therapy (i.e. 10–40 mmol/L) is sufficient to exert this effect. For patients on long-term lithium therapy, amiloride has been proposed to prevent the uptake of lithium in the collecting ducts, thus preventing the

Pregnancy in a patient known to have diabetes insipidus An isolated deficiency of vasopressin without a concomitant loss of hormones in the anterior pituitary does not result in altered fertility, and with the exception of polyuria and polydipsia, gestation, delivery, and lactation are uncomplicated (Amico, 1985). Patients may require increasing dosages of DDAVP. The increased thirst may be due to a resetting of the thirst osmostat (Davison et al., 1988). Increased polyuria also occurs during pregnancy in patients with partial NDI (Iwasaki et  al., 1991). These patients may be obligatory carriers of the NDI gene (Forssman, 1945) or may be homozygotes, compound heterozygotes, or may have dominant AQP2 mutations.

Syndromes of diabetes insipidus that begin during gestation and remit after delivery Pregnancy may be associated with several different forms of diabetes insipidus, including central, nephrogenic, and vasopressinase-mediated forms (Hiett and Barton, 1990; Iwasaki et al., 1991; Brewster and Hayslett, 2005; Lindheimer, 2005).

Diagnostic work-up of polyuric states Excepting the context of brain trauma, brain surgery, or long-term lithium administration where the diagnosis of polyuria is obvious, a logical approach to the patient with polyuria is to search for

chapter 32 

arguments supporting known causes of polyuric states. Such arguments may be: (a) morphological (brain MRI), including the presence of a hypothalamic tumour or mass related to a granulomatous or inflammatory process; (B) hormonal, suggesting that the posterior pituitary involvement is not isolated but rather associated with other signs of anterior pituitary deficits; (C) systemic with the presence of a generalized inflammatory process or pituitary metastasis; or (D) hereditary with other members of the family affected with central or NDI. An abrupt onset of polyuria in an adult would suggest acquired central diabetes insipidus. MRI of the hypothalamic structures and of the posterior pituitary should be obtained to assess the posterior pituitary normal ‘bright spot,’ a possible surrogate of the posterior pituitary vasopressin content, and any accompanying lesions. Clinical and biochemical indices of associated anterior pituitary/hormone deficiency should also be obtained (Maghnie et al., 2000) since additional deficits in anterior pituitary hormones were documented in 61% of patients, a median of 0.6 years after the onset of diabetes insipidus. The most frequent abnormality was growth hormone deficiency (59%) followed by hypothyroidism (28%), hypogonadism (24%), and adrenal insufficiency (22%). Seventy-five per cent of the patients with Langerhans cell histiocytosis had an anterior pituitary hormone deficiency that was first detected a median of 3.5 years after the onset of diabetes insipidus. In this context, the dehydration test is rarely necessary and only recommended for patients with isolated polyuria, a normal pituitary stalk, and hypothalamic region on MRI and with no familial history of polyuria. If plasma osmolality and/or Na+ concentration under conditions of ad libitum fluid intake are > 295 mOsm/kg and 143  mmol/L, respectively, the diagnosis of primary polydipsia is excluded (Robertson, 1981). Water restriction tests are described in Bichet (2015). If severe polyuric symptoms and signs are documented, water should be restricted only to 2–4 hours during daytime in infants, plasma Na+ should be available every 2 hours during testing and should not exceed 145–148 mmol/L in children and adults, since a maximal endogenous vasopressin stimulation (> 3.5 pg/mL) should occur at this level with a maximal urine osmolality response (> 800 mOsm/kg). If delays of > 60 minutes are encountered to obtain plasma Na+ or urine osmolalities during dehydration tests, these tests should be done in other institutions where almost immediate laboratory reports are obtained after blood samplings.

References Abu Libdeh, A., Levy-Khademi, F., Abdulhadi-Atwan, M., et al. (2009). Autosomal recessive familial neurohypophyseal diabetes insipidus: onset in early infancy. Eur J Endocrinol, 162, 221–6. Amico, J. A. (1985). Diabetes insipidus and pregnancy. In P. Czernichow and A. G. Robinson (eds.) Frontiers of Hormone Research, pp. 266–77. Basel: Karger. Andersen, L. J., Jensen, T. U., Bestle, M. H., et al. (2000). Gastrointestinal osmoreceptors and renal sodium excretion in humans. Am J Physiol Regul Integr Comp Physiol, 278, R287–94. Babey, M., Kopp, P., and Robertson, G. L. (2011). Familial forms of diabetes insipidus: clinical and molecular characteristics. Nat Rev Endocrinol, 7, 701–14. Barlow, E. D. and de Wardener, H. E. (1959). Compulsive water drinking. Q J Med New Series, 28, 235–58. Batlle, D. C., von Riotte, A. B., Gaviria, M., et al. (1985). Amelioration of polyuria by amiloride in patients receiving long-term lithium therapy. N Engl J Med, 312, 408–14.

approach to the patient with polyuria

Bichet, D. G. (2011). Clinical Manifestations and Causes of Nephrogenic Diabetes Insipidus. [Online] Bichet, D. G. (2015). Diagnostic of polyuria and diabetes insipidus. [Online] Bichet, D. G., Arthus, M. -F., Lonergan, M., et al. (1998). Hereditary central diabetes insipidus: autosomal dominant and autosomal recessive phenotypes due to mutations in the prepro-AVP-NPII gene. J Am Soc Nephrol, 9, 386A. Bockenhauer, D., Carpentier, E., Rochdi, D., et al. (2009). Vasopressin type 2 receptor V88M mutation: molecular basis of partial and complete nephrogenic diabetes insipidus. Nephron Physiol, 114, 1–10. Bockenhauer, D., van’t Hoff, W., Dattani, M., et al. (2010). Secondary nephrogenic diabetes insipidus as a complication of inherited renal diseases. Nephron Physiol, 116, 23–9. Bonnardeaux, A. and Bichet, D.G. (2012). Inherited disorders of the renal tubule. In M.W. Taal, P.A. Marsden, K. Skorecki, et al. (eds.) Brenner & Rector’s The Kidney, pp. 1584–625. Philadelphia, PA: Elsevier Saunders. Boschmann, M., Steiniger, J., Franke, G., et al. (2007). Water drinking induces thermogenesis through osmosensitive mechanisms. J Clin Endocrinol Metab, 92, 3334–7. Boton, R., Gaviria, M., and Batlle, D. C. (1987). Prevalence, pathogenesis, and treatment of renal dysfunction associated with chronic lithium therapy. Am J Kidney Dis, 10, 329–45. Bourque, C. W. (2008). Central mechanisms of osmosensation and systemic osmoregulation. Nat Rev Neurosci, 9, 519–31. Brewster, U. C., and Hayslett, J. P. (2005). Diabetes insipidus in the third trimester of pregnancy. Obstet Gynecol, 105, 1173–6. Burbach, J. P., Luckman, S. M., Murphy, D., et al. (2001). Gene regulation in the magnocellular hypothalamo-neurohypophysial system. Physiol Rev, 81, 1197–267. Chanson, P. and Salenave, S. (2011). Treatment of neurogenic diabetes insipidus. Ann Endocrinol (Paris), 72, 496–9. Choi, M., Scholl, U.I., Ji, W., et al. (2009). Genetic diagnosis by whole exome capture and massively parallel DNA sequencing. Proc Natl Acad Sci U S A, 106, 19096–101. Christensen, J. H., Kvistgaaard, H., Knudsen, J., et al. (2013). A novel deletion partly removing the Avp gene causes autosomal recessive inheritance of early onset neurohypophyseal diabetes insipidus. Clin Genet, 83(1), 44–52. Ciura, S. and Bourque, C. W. (2006). Transient receptor potential vanilloid 1 is required for intrinsic osmoreception in organum vasculosum lamina terminalis neurons and for normal thirst responses to systemic hyperosmolality. J Neurosci, 26, 9069–75. Claybaugh, J. R., Sato, A. K., Crosswhite, L. K., et al. (2000). Effects of time of day, gender, and menstrual cycle phase on the human response to a water load. Am J Physiol Regul Integr Comp Physiol, 279, R966–73. Claydon, V. E., Schroeder, C., Norcliffe, L. J., et al. (2006). Water drinking improves orthostatic tolerance in patients with posturally related syncope. Clin Sci (Lond), 110, 343–52. Coste, B., Mathur, J., Schmidt, M., et al. (2010). Piezo1 and Piezo2 are essential components of distinct mechanically activated cation channels. Science, 330, 55–60. Davison, J. M., Shiells, E. A., Philips, P. R., et al. (1988). Serial evaluation of vasopressin release and thirst in human pregnancy. Role of human chorionic gonadotrophin in the osmoregulatory changes of gestation. J Clin Invest, 81, 798–806. Denton, K. M. (2012). In the arms of Morpheus. Am J Physiol Renal Physiol, 302(2). F234–5. Earley, L. E. and Orloff, J. (1962). The mechanism of antidiuresis associated with the administration of hydrochlorothiazide to patients with vasopressin-resistant diabetes insipidus. J Clin Invest, 41, 1988–97. Edwards, A. M., Mann, M. E., Marfell-Jones, M. J., et al. (2007). Influence of moderate dehydration on soccer performance: physiological responses to 45 min of outdoor match-play and the immediate subsequent

297

298

Section 2  

fluid, electrolyte, and renal tubular disorders

performance of sport-specific and mental concentration tests. Br J Sports Med, 41, 385–91. Forssman, J. (1945). On hereditary diabetes insipidus, with special regard to a sex-linked form. Acta Med Scand, 159, 1–196. Geelen, G., Greenleaf, J. E., and Keil, L. C. (1996). Drinking-induced plasma vasopressin and norepinephrine changes in dehydrated humans. J Clin Endocrinol Metab, 81, 2131–5. Green, E. D. and Guyer, M. S. (2011). Charting a course for genomic medicine from base pairs to bedside. Nature, 470, 204–13. Grunfeld, J. P. and Rossier, B. C. (2009). Lithium nephrotoxicity revisited. Nat Rev Nephrol, 5, 270–6. Guyon, C., Lussier, Y., Bissonnette, P., et al. (2009). Characterization of D150E and G196D aquaporin-2 mutations responsible for nephrogenic diabetes insipidus: importance of a mild phenotype. Am J Physiol Renal Physiol, 297, F489–98. Hiett, A. K.,and Barton, J. R. (1990). Diabetes insipidus associated with craniopharyngioma in pregnancy. Obstet Gynecol, 76, 982–4. Hildebrandt, F., Attanasio, M., and Otto, E. (2009). Nephronophthisis: disease mechanisms of a ciliopathy. J Am Soc Nephrol, 20, 23–35. Iwasaki, Y., Oiso, Y., Kondo, K., et al. (1991). Aggravation of subclinical diabetes insipidus during pregnancy. N Engl J Med, 324, 522–6. Jordan, J., Shannon, J. R., Black, B. K., et al. (2000). The pressor response to water drinking in humans: a sympathetic reflex? Circulation, 101, 504–9. Leaf, A. (1979). Nephrology forum: neurogenic diabetes insipidus. Kidney Int, 15, 572–80. Lechner, S. G., Markworth, S., Poole, K., et al. (2011). The molecular and cellular identity of peripheral osmoreceptors. Neuron, 69, 332–44. Lindheimer, M. D. (2005). Polyuria and pregnancy: its cause, its danger. Obstet Gynecol, 105, 1171–2. Maghnie, M., Cosi, G., Genovese, E., et al. (2000). Central diabetes insipidus in children and young adults. N Engl J Med, 343, 998–1007. Mahler, B.T., Kamperis, K., Schroeder, M., et al. (2012). Sleep deprivation induces excess diuresis and natriuresis in healthy children. Am J Physiol Renal Physiol, 302(2), F236–43. Marion, V., Schlicht, D., Mockel, A., et al. (2011). Bardet-Biedl syndrome highlights the major role of the primary cilium in efficient water reabsorption. Kidney Int, 79, 1013–25. McHugh, J., Keller, N.R., Appalsamy, M., et al. (2010). Portal osmopressor mechanism linked to transient receptor potential vanilloid 4 and blood pressure control. Hypertension, 55, 1438–43. Peters, M., Jeck, N., Reinalter, S., et al. (2002). Clinical presentation of genetically defined patients with hypokalemic salt-losing tubulopathies. Am J Med, 112, 183–90.

Puricelli, E., Bettinelli, A., Borsa, N., et al. (2010). Long-term follow-up of patients with Bartter syndrome type I and II. Nephrol Dial Transplant, 25, 2976–81. Robertson, G. L. (1981). Diseases of the posterior pituitary. In D. Felig, J. D. Baxter, E. Broadus, et al. (eds.) Endocrinology and Metabolism, pp. 251–277. New York: McGraw-Hill. Robertson, G. L. and Athar, S. (1976). The interaction of blood osmolality and blood volume in regulating plasma vasopressin in man. J Clin Endocrinol Metab, 42, 613–20. Saat, M., Sirisinghe, R. G., Singh, R., et al. (2005). Effects of short-term exercise in the heat on thermoregulation, blood parameters, sweat secretion and sweat composition of tropic-dwelling subjects. J Physiol Anthropol Appl Human Sci, 24, 541–9. Schroeder, C., Bush, V. E., Norcliffe, L. J., et al. (2002). Water drinking acutely improves orthostatic tolerance in healthy subjects. Circulation, 106, 2806–11. Shannon, J. R., Diedrich, A., Biaggioni, I., et al. (2002). Water drinking as a treatment for orthostatic syndromes. Am J Med, 112, 355–60. Sharif Naeini, R., Witty, M. F., Seguela, P., et al. (2006). An N-terminal variant of Trpv1 channel is required for osmosensory transduction. Nat Neurosci, 9, 93–8. Shirreffs, S. M., Merson, S. J., Fraser, S. M., et al. (2004). The effects of fluid restriction on hydration status and subjective feelings in man. Br J Nutr, 91, 951–8. Stookey, J. D., Constant, F., Popkin, B. M., et al. (2008). Drinking water is associated with weight loss in overweight dieting women independent of diet and activity. Obesity (Silver Spring), 16, 2481–8. Tank, J., Schroeder, C., Stoffels, M., et al. (2003). Pressor effect of water drinking in tetraplegic patients may be a spinal reflex. Hypertension, 41, 1234–9. Taylor, A. C., McCarthy, J. J., and Stocker, S. D. (2008). Mice lacking the transient receptor vanilloid potential 1 channel display normal thirst responses and central Fos activation to hypernatremia. Am J Physiol Regul Integr Comp Physiol, 294, R1285–93. Trudel, E. and Bourque, C. W. (2011). Central clock excites vasopressin neurons by waking osmosensory afferents during late sleep. Nat Neurosci, 13, 467–74. Verney, E. (1947). The antidiuretic hormone and the factors which determine its release. Proc R Soc London Ser B, 135, 25–6. Willcutts, M. D., Felner, E., and White, P. C. (1999). Autosomal recessive familial neurohypophyseal diabetes insipidus with continued secretion of mutant weakly active vasopressin. Hum Mol Genet, 8, 1303–7. Zhang, Z. and Bourque, C. W. (2008). Amplification of transducer gain by angiotensin II-mediated enhancement of cortical actin density in osmosensory neurons. J Neurosci, 28, 9536–44.

CHAPTER 33

Clinical use of diuretics David H. Ellison and Arohan R. Subramanya Introduction Oedema is usually a manifestation of expanded extracellular fluid (ECF) volume most typically caused by heart failure (HF), hepatic cirrhosis, nephrotic syndrome, or kidney dysfunction; it can also result from local factors or lymphatic obstruction. Surprisingly, primary renal NaCl retention does not lead to oedema, but instead to hypertension, because ‘pressure natriuresis’ occurs, preventing substantial ECF volume expansion. In contrast, when NaCl is retained because the effective arterial blood volume is reduced, oedema results. Regardless of its cause, symptomatic oedema often requires treatment with diuretics. Diuretics now comprise two classes, the natriuretics and the aquaretics, although diuretic treatment of oedema typically relies primarily on natriuretic diuretics. In addition to their use for oedema, diuretic drugs are indicated for a wide variety of non-oedematous disorders. Treatment of hypertension, nephrolithiasis (see Chapter 30), and hyponatraemia (see Chapter 28) are discussed elsewhere. This chapter will focus on renal mechanisms of diuretic action and diuretic therapy of oedema. The molecular targets of diuretic drugs are predominantly Na+ transport pathways at the apical (luminal) surface of kidney tubule cells. When coupled with the basolateral Na/K-ATPase, these pathways permit the vectorial transport of sodium. A rational classification of diuretic drugs (see Table 33.1) is based on the primary nephron site of action.

Osmotic diuretics Osmotic diuretics are substances that are freely filtered at the glomerulus, but are poorly reabsorbed. Inhibition of NaCl reabsorption by these drugs depends on the osmotic pressure exerted by the drug molecules in solution, not on interaction with specific transport proteins. Mannitol is the prototypical osmotic diuretic (Better et  al., 1997). Because the relationship between the magnitude of diuretic effect and concentration of osmotic diuretic in solution is linear, all osmotic diuretics are small molecules. Other agents considered in this class include urea, sorbitol, and glycerol. Although osmotic agents do not act directly on transport pathways, ion transport is affected. Following mannitol infusion, sodium, potassium, calcium, magnesium, bicarbonate and chloride excretion rates increase (Table 33.1). Sodium and water reabsorption rates are reduced by 27% and 12%, respectively (Seely and Dirks, 1969). Magnesium and calcium reabsorption is also reduced along the proximal tubule and loop of Henle. The mechanisms by which mannitol produces a diuresis include (a)  increasing the luminal osmotic pressure along the proximal tubule and loop of Henle, thereby retarding the passive reabsorption

of water; and (b) increasing the renal plasma flow (RPF), thereby washing out medullary tonicity. Mannitol is freely filtered at the glomerulus and its presence in tubule fluid minimizes passive water reabsorption. When an osmotic diuretic is administered, the osmotic force of the non-reabsorbable solute in the lumen opposes the osmotic force produced by sodium reabsorption, and sodium reabsorption eventually stops. Perhaps surprisingly, mannitol has a greater effect on inhibiting Na and water reabsorption in the loop of Henle than in the proximal tubule. Further downstream, in the collecting duct, mannitol also can reduce sodium and water reabsorption (Buerkert et al., 1981). During the administration of mannitol, its molecules diffuse from the blood stream into the interstitial space. In the interstitial space, the increased osmotic pressure draws water from the cells to increase extracellular fluid (ECF) volume, increasing renal plasma flow (Buerkert et al., 1981). Renal cortical and medullary blood flow rates increase following mannitol infusion (Buerkert et al., 1981). Single nephron glomerular filtration rate (GFR) increases in cortex, but decreases in medulla (Gennari and Kassirer, 1974) via unknown mechanisms. The net effect of mannitol on total kidney GFR has been variable, but most studies indicate that the overall effect is to increase GFR (Blantz, 1974). The combination of enhanced renal plasma flow and reduced medullary GFR washes out the medullary osmotic gradient by reducing papillary sodium and urea content. Experimental studies indicate that the osmotic effect of mannitol to increase water movement from intracellular to extracellular space leads to a decrease in haematocrit and in blood viscosity. This fact contributes to a decrease in renal vascular resistance and increase in renal plasma flow (RPF). In addition, both prostacyclin (PGI2) (Johnston et al., 1981) and atrial natriuretic peptide (Yamasaki et al., 1988) may participate in the effect of mannitol on RPF Following infusion, mannitol distributes in ECF with a volume of distribution of approximately 16 L (Anderson et al., 1988); its excretion is almost entirely by glomerular filtration (Weiner, 1990). Of the filtered load, < 10% is reabsorbed by the renal tubule, and a similar quantity is metabolized, probably in the liver. With normal GFR, plasma half-life is approximately 2.2 hours. Marked accumulation of mannitol in patients can lead to reversible acute kidney injury (AKI) with vasoconstriction and tubular vacuolization (Dorman and Sondheimer, 1990; Visweswaran et al., 1997).

Proximal tubule diuretics (carbonic anhydrase inhibitors) Carbonic anhydrase inhibitors increase urinary sodium excretion somewhat, but have a limited therapeutic role as diuretics, because

300

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 33.1  Physiological classification of diuretic drugs Natriuretics

Aquaretics

Proximal diuretics

Loop diuretics

DCT diuretics

Cortical collecting diuretics

Vaptans

Carbonic anhydrase inhibitors Acetazolamide

Na-K-2Cl transporter (NKCC2) inhibitors Furosemide Bumetanide Torsemide Ethacrynic acida

Na-Cl cotransporter (NCC) inhibitors Hydrochlorothiazide Metolazone Chlorthalidone Indapamideb others

Na channel blockers (ENaC inhibitors) Amiloride Triamterene Aldosterone antagonists Spironolactone Eplerenone

Vasopressin receptor blockers Conivaptan Tolvaptan

Classification based on site and mechanism of action. a Mechanism of ethacrynic acid differs from that of other loop diuretics. b Indapamide and metolazone may have other actions, as well.

DCT = distal convoluted tubule.

they are only weakly natriuretic during chronic use. Bicarbonate excretion rises by 25–30%, producing an alkaline diuresis (Table 33.1). Carbonic anhydrase inhibitors also increase potassium excretion, likely indirectly. The effect of carbonic anhydrase inhibitors on the proximal tubule ion transport facilitates an increase in tubular fluid flow rate and sodium and bicarbonate delivery to the distal nephron, where the lumen negative voltage (Malnic et  al., 1966) and urine flow rate increase (Good and Wright, 1979). The biochemical, morphological, and functional properties of carbonic anhydrase have been reviewed (Pastorekova et al., 2004). Normally the proximal tubule reabsorbs 80% of the filtered load of bicarbonate and 60% of the filtered load of sodium chloride via cellular mechanisms depicted in Fig. 33.1. Carbonic anhydrase inhibitors act primarily in this segment, yet their natriuretic potency is relatively weak. Several factors explain this observation. First, proximal sodium reabsorption is mediated by carbonic anhydrase-independent as well as carbonic anhydrase-dependent pathways. Second, the increased sodium delivered to distal nephron segments is largely reabsorbed. Third, carbonic anhydrase inhibitors generate a hyperchloraemic metabolic acidosis, further reducing the effects of subsequent doses of carbonic anhydrase inhibitor. Finally, metabolic acidosis increases the Ki for bicarbonate absorption by membrane impermeant carbonic anhydrase inhibitors by a factor of 100 to 500, suggesting that metabolic acidosis is associated with changes in the physical properties of the carbonic anhydrase protein (Shuichi and Schwartz. 1998). For these reasons, carbonic anhydrase inhibitors alone are rarely used as diuretic agents chronically; they do, however, play an important role in short-term treatment of diuretic resistance. Systemic administration of carbonic anhydrase inhibitors reduces GFR by as much as 30%. Single nephron glomerular filtration rate (SNGFR) was 23% lower during acetazolamide infusion, partly because increased solute delivery to the macula densa activates the tubuloglomerular feedback (TGF) mechanism (Skott et  al., 1989), although other factors are likely to contribute (Hashimoto et al., 2004). Acetazolamide is well absorbed from the gastrointestinal (GI) tract. More than 90% of the drug is plasma protein bound. The highest concentrations are found in tissues that contain large amounts of carbonic anhydrase (e.g. renal cortex, red blood cells). Renal effects are noticeable within 30 minutes and are usually maximal

at 2 hours. Acetazolamide is not metabolized, but is excreted rapidly by glomerular filtration and proximal tubular secretion. The half-life is approximately 5 hours and renal excretion is essentially complete in 24 hours (Weiner, 1990). In comparison, methazolamide is absorbed more slowly from the GI tract, and its duration of action is long, with a half-life of approximately 14 hours. Generally, carbonic anhydrase inhibitors are well tolerated with infrequent serious adverse effects. Side effects of carbonic anhydrase inhibitors may arise from the continued excretion of electrolytes. Significant hypokalaemia and metabolic acidosis may develop. In elderly patients with glaucoma treated with acetazolamide (250–1000 mg/day), metabolic acidosis was a frequent finding (Heller et al., 1985). Even though carbonic anhydrase inhibitors do not increase urinary calcium excretion, they do increase the risk for nephrocalcinosis and nephrolithiasis, owing to their effects on urine pH and citrate excretion. Premature infants treated with furosemide and acetazolamide are particularly susceptible to nephrocalcinosis, presumably due to the combined effect of an alkaline urine and hypercalciuria (Stafstrom et al., 1992).

Loop diuretics The loop diuretics inhibit sodium and chloride transport along the loop of Henle and macula densa (Fig. 33.1). Although these drugs also impair ion transport by proximal and distal tubules under some conditions, these effects probably contribute little to their action clinically. The loop diuretics available include furosemide, bumetanide, torsemide, and ethacrynic acid. Loop diuretics increase water, Na+, K+, Cl−, phosphate, magnesium, and calcium excretion rates (Table 33.1). The loop diuretic dose–response relationship is sigmoidal (Fig. 33.2), which has labelled them as ‘threshold’ drugs (Brater, 1997). Loop diuretics have the highest natriuretic and chloriuretic efficacy of any class of diuretics; they are sometimes called ‘high ceiling’ diuretics, for this reason. Loop diuretics can increase Na+ and Cl− excretion up to 25% of the filtered load. If administered during water loading, solute-free water clearance (CH 2 O ) decreases and osmolar clearance increases, although the urine always remains dilute. During water restriction, loop diuretics impair the reabsorption of solute-free water (THC2 O ) . During maximal loop diuretic action, the urinary Na+ concentration is usually between 75 and 100 mmol/L (Puschett and Goldberg, 1968). Because urinary K+ concentrations during

chapter 33 

Tubular lumen Na+ –

CI

Na+

DCT Diuretic

Carbonic anhydrase inhibitors Acetazolamide

Tubular lumen

HCO3

Na+

Na+ NHE3

+

H2O

H

H2CO3 H2O

K+

CIC-KB

Distal convoluted tubule diuretics Hydrochlorothiazide Chlorthalidone Metolazone

~

Tubular lumen

K+

+

Na 3HCO–3

OH– CA Inhibitors

CI–

~

Interstitium

PCT

Filtrate

Interstitium

DCT

NCC

CA IV CO2

CA Inhibitors

Principal cell (CCD)

Interstitium

–30mV

+

CA II

CO2

clinical use of diuretics

Na

Amiloride

Collecting duct diuretics Spironolactone Amiloride Triamterene

Tubular lumen

K+

Na+ 2CI–+ K

Na+

NKCC

Loop diuretic

Na+ Ca++M++

K+

MR

Aldo

Aldo

Aldosterone Antagonists

Interstitium

TAL

+10mV

K+

+

Triamterene

Loop diuretics Furosemide Torsemide Bumetanide

~

Circulation

Na+ ENaC

~

ROMK

K+

CI–

Fig. 33.1  Mechanisms and sites of diuretic action. The figure shows a cartoon of the nephron, with segments identified. Diuretics, classified as in Table 33.1, are shown. Functional models of diuretic actions are also shown, for each site of action.

furosemide-induced natriuresis remain relatively low, electrolyte free water (CH 2 O � ) excretion increases (Puschett and Goldberg, 1968). This effect of loop diuretics has been exploited to treat hyponatraemia, when combined with normal or hypertonic saline (Hantman et al., 1973; Decaux et al., 1981).

Na+ and Cl− transport The predominant effect of loop diuretic drugs is to inhibit the electroneutral Na-K-2Cl cotransporter at the apical surface of thick ascending limb (TAL) cells (Fig. 33.1). This transporter is a member of the cation chloride cotransporter family (Hebert et  al., 2004; Gamba, 2005); it is referred to as the Na-K-2Cl cotransporter, second isoform, (NKCC2), and is encoded by the gene SLC12A1. This protein uses the electrochemical gradient favouring Na+ entry across the apical membrane to move Cl− into the cell along with K+, while K+ diffuses back into the luminal fluid via a K+ channel; thus, net reabsorption across this segment is primarily NaCl. The combination of K+ movement across the apical membrane and Cl− movement across the basolateral cell membrane generates a transepithelial voltage oriented in the

lumen-positive direction (Greger and Schlatter, 1983), which drives absorption of Na+, Ca2+ and Mg2+ via the paracellular pathway. It should be noted, however, that both the transcellular and the paracellular components of Na+ transport are inhibited by loop diuretics, the former directly and the latter indirectly. The thick ascending limb is virtually impermeable to water. The combination of solute absorption and water impermeability determines the role of the thick ascending limb as the primary diluting segment of the kidney. The predominant effect of the loop diuretics, furosemide, bumetanide, and torsemide, is to inhibit NKCC2 directly; the mechanisms of action of ethacrynic acid are not as clear. These drugs, however, have other important actions. Thick ascending limb cells have been shown to produce prostaglandin E2 following stimulation with furosemide (Miyanoshita et al., 1989), perhaps via inhibition of prostaglandin dehydrogenase (Abe et al., 1977; Wright et al., 1976). Blockade of cyclooxygenase reduces the effects of furosemide to inhibit loop segment chloride transport in rats (Kirchner, 1985; Kirchner et al., 1986), and this effect appears to be important clinically, since non-steroidal anti-inflammatory drugs (NSAIDs)

301

Section 2  

fluid, electrolyte, and renal tubular disorders

Intravenous

20

16 14

Normal CKD HF

Secretory defect

Decreased maximal response

12 10 8 6 4 2 0 Oral loop diuretic dose log scale

Rate of diuretic excretion, mg/min

18

FENa, %

302

Rate of maximal diuretic efficiency

Oral Time, hours

Fig. 33.2  Dose–response curve for loop diuretics. (A) Panel A shows the fractional Na excretion (FENa) as a function of loop diuretic concentration. Compared with normal patients, patients with chronic renal failure (CKF) show a rightward shift in the curve, owing to impaired diuretic secretion, but the maximal response is preserved. Note that when sodium excretion is expressed in absolute terms (not shown), the maximal diuretic effect in CKD is reduced. In contrast, patients with oedema demonstrate a rightward and downward shift in natriuretic effect, even when expressed as FENa. (B) Panel B compares the response to intravenous and oral doses of loop diuretics. The natriuretic threshold is shown. When given intravenously, peak diuretic concentrations are reached rapidly, but levels may decline more rapidly than during oral administration. As natriuresis depends on the time above the natriuretic threshold, and as the threshold is impacted by disease (A), the relationship between oral and intravenous efficacy is complex.

cause loop diuretic resistance. Increases in renal prostaglandins may also contribute to the haemodynamic effects of loop diuretics described below.

Ca2+ and Mg2+ transport Loop diuretics increase the excretion of the divalent cations, calcium and magnesium, owing to their effects to reduce the transepithelial voltage. This stops passive paracellular calcium and magnesium absorption.

Renin secretion Loop diuretics strongly stimulate renin secretion. Although a component of this effect results from ECF volume contraction (see below), loop diuretics also stimulate renin secretion directly, by inhibiting Na-K-2Cl cotransport, because NaCl entry into macula densa cells, via NKCC2, modulates renin release (Schnermann and Briggs, 2012). When loop diuretics are present, the cells cannot sense luminal NaCl and renin secretion cannot be suppressed. Interestingly, loop diuretics also may stimulate renin secretion by inhibiting NKCC1, the secretory form of the three ion co-transport mechanism present in the basolateral membrane (Schnermann and Briggs, 2012). Prostaglandin production also participates in regulating renin secretion. Cyclooxygenase, COX-2, is expressed by macula densa cells and by interstitial cells in the kidney (Harris et al., 1994; Guan et al., 1997; Khan et al., 1998; Komhoff et al., 2000). This isoform is often found only after induction by inflammatory cytokines. Blockade of prostaglandin synthesis, either by non-specific cyclooxygenase inhibitors (Frölich et al., 1976) or by specific COX-2 blockers (Harding et al., 1997; Kammerl et al., 2001) reduces both the loop

diuretic-induced natriuresis and the renin secretory response. These results have been corroborated in humans (Kammerl et al., 2001). GFR and RPF flow tend to be preserved during loop diuretic administration (Hook et  al., 1966), although GFR and RPF can decline if ECF volume contraction is severe. Loop diuretics reduce renal vascular resistance and increase RPF under experimental conditions (Ludens et al., 1968; Dluhy et al., 1970), probably resulting from the diuretic-induced vasodilatory prostaglandins (discussed earlier). Another factor that may contribute to the tendency of loop diuretics to maintain GFR and RPF despite volume contraction is their effect on TGF. As noted earlier, the sensing mechanism that activates TGF involves NaCl transport across the apical membrane by the loop diuretic sensitive Na-K-2Cl cotransporter (Schnermann and Briggs, 2008). Under normal conditions, when the luminal concentration of NaCl reaching the macula densa rises, GFR decreases via TGF. Loop diuretic drugs block TGF by interfering with the sensing step of TGF (Wright and Schnermann, 1974). In the absence of effects on the macula densa, loop diuretics would be expected to suppress GFR and RPF by increasing distal NaCl delivery and activating the TGF system. Instead, TGF blockade permits GFR and RPF to be maintained. Acute intravenous administration of loop diuretics increases venous capacitance (Dikshit et al., 1973), perhaps via prostaglandins (Bourland et  al., 1977; Mukherjee et  al., 1981), although direct effects in peripheral vascular beds may participate as well (Schmieder et  al., 1987). Although venodilation and improvements in cardiac haemodynamics frequently result from intravenous therapy with loop diuretics, the haemodynamic response to intravenous loop diuretics may be more complex (Ellison, 1997b).

chapter 33 

Johnston et al. reported that low dose furosemide increased venous capacitance, but that higher doses did not (Johnston et al., 1984). These investigators suggested that furosemide-induced renin secretion leads to angiotensin II-induced vasoconstriction, an effect that might overwhelm the prostaglandin-mediated vasodilatory effects. In two series, 1–1.5 mg/kg furosemide boluses administered to patients with chronic HF, resulted in transient deteriorations in haemodynamics (during the first hour), with declines in stroke volume index, increases in left ventricular filling pressure (Francis et al., 1985), and exacerbation of HF symptoms. These changes may be related to activation of both the sympathetic nervous system and the renin–angiotensin system by the diuretic drug. Evidence for a role of the renin–angiotensin system in the furosemide-induced deterioration in systemic haemodynamics includes the temporal association between its activation and haemodynamic deterioration (Francis et al., 1985), and the ability of angiotensin-converting enzyme inhibitors (ACEIs) to prevent much of the pressor effect (Goldsmith et al., 1989). The effects of renal denervation on sympathetic responses to furosemide were studied. These results confirm that the effects are mediated both by direct renal nerve traffic and indirectly by activation of the renin–angiotensin axis (Fitch and Weiss, 2000; Fitch et al., 2000). The three loop diuretics that are used most commonly, furosemide, bumetanide, and torsemide, are absorbed quickly after oral administration, reaching peak concentrations within 0.5–2 hours. Furosemide absorption is slower than its elimination in normal subjects; thus the time to reach peak serum level is slower for furosemide than for bumetanide and torsemide. This

clinical use of diuretics

phenomenon is called ‘absorption-limited kinetics’, as the rate of absorption is often slower than the rate of elimination (Brater, 1997). The bioavailability of loop diuretics varies from 50% to 90% (Table 33.2); furosemide bioavailability is approximately 50% (Shankar and Brater, 2003); when a patient is switched from intravenous to oral furosemide, it is therefore customary to double the dose to compensate for its poor bioavailability (Brater, 1997); in practice, however, there are many other variables that affect furosemide efficacy, and a fixed intravenous/oral conversion cannot be given (Brater, 1983). The half-lives of the loop diuretics available vary, but all are relatively short (ranging from approximately 1 hour for bumetanide to 3–4 hours for torsemide). The half-lives of muzolimine, xipamide, and ozolinone, which are not so widely available, are longer (6–15 hours). Loop diuretics are organic anions that circulate tightly bound to albumin (> 95%), thus their volume of distribution is small except during extreme hypoalbuminaemia (Inoue et  al., 1987). Approximately 50% of an administered dose of furosemide is excreted unchanged into the urine. The remainder appears to be eliminated by glucuronidation, probably by the kidney. Torsemide and bumetanide are eliminated both by hepatic processes and through renal excretion. The differences in metabolic fate mean that the half-life of furosemide is altered by kidney failure more than the half-lives of torsemide and bumetanide. Loop diuretics gain access to the tubular fluid almost exclusively by proximal secretion. The -uptake is mediated by the organic anion transporters OAT1 and OAT3, whereas the apically located multidrug resistance-associated protein 4 (Mrp-4) mediates secretion into the tubular fluid. Mice

Table 33.2  Effects of diuretics on electrolyte excretion Diuretic

Na

Cl

K

Pi

Ca

Mg

Osmotic diuretics (Wesson and Anslow, 1948; Wesson, 1967; Seely Dirks, 1969; Eknoyan et al., 1970; Benabe and Martinez-Maldonado, 1986)

⇑ (10–25%)

⇑ (15–30%)

⇑ (6%)

⇑ (5–10%)

⇑ (10–20%)

⇑ (> 20%)

Carbonic anhydrase inhibitors (Puschett and Goldberg, 1968; Eknoyan et al., 1970; Cogan et al., 1979)

⇑ (6%)

⇑ (4%)

⇑ (60%)

⇑ (> 20%)

⇑ or ⇔ (< 5%)

⇑ (< 5%)

Loop diuretics (Earley and Friedler, 1964; Suki et al., 1965; Puschett and Goldberg, 1968; Eknoyan et al., 1970; Duarte et al., 1971; Hropot et al., 1985)

⇑ (30%)

⇑ (40%)

⇑ (60–100%)

⇑ (> 20%)

⇑ (> 20%)

⇑ (> 20%)

DCT diuretics (Demartini et al., 1962; Suki et al., 1965; Eknoyan et al., 1970; Hropot et al., 1985)

⇑ (6–11%)

⇑ (10%)

⇑ (200%)

⇑ (> 20%)

⇓

⇑ (5–10%)

Na channel blockers (Eknoyan et al., 1970; Duarte et al., 1971; Hropot et al., 1985)

⇑ (3–5%)

⇑ (6%)

⇓ (8%)

⇔

⇔

⇓

Collecting duct diuretics (Eknoyan et al., 1970)

⇑ (3%)

⇑ (6%)

⇓

⇔

⇔

⇓

Figures indicate approximate maximal fractional excretions of ions following acute diuretic administration in maximally effective doses. ⇑ indicates that the drug increases excretion; ß indicates that the drug decreases excretion; ⇔ indicates that the drug has little of no direct effect on excretion. During chronic treatment, effects often wane (Na excretion), may increase (K excretion during DCT diuretic treatment), or may reverse as with uric acid (not shown).

303

304

Section 2  

fluid, electrolyte, and renal tubular disorders

lacking OAT1, OAT3, or Mrp-4 are remarkably resistant to both loop and thiazide diuretics, illustrating the functional importance of these proteins (Eraly et al., 2006; Hasegawa et al., 2007; Vallon et al., 2008). The most common adverse effects of loop diuretics result from their primary actions. Loop diuretics are frequently administered to treat oedematous expansion of the ECF volume. As noted above, oedema usually results from a decrease in the ‘effective’ arterial blood volume, but this volume is impossible to measure precisely. Overzealous diuretic usage or intercurrent complicating illnesses can lead to excessive contraction of the intravascular volume with orthostatic hypotension, renal dysfunction, and sympathetic overactivity. Patients suffering from HF are typically treated with both diuretics and ACEIs or angiotensin receptor blockers (ARBs); this combination is especially likely to worsen renal function in certain circumstances. AKI in such patients often responds to reduced diuretic doses and liberalization of dietary NaCl intake, permitting continued administration of the ACEIs/ARBs (Packer et al., 1987; Packer, 1989). Other patients at increased risk for renal dysfunction during diuretic therapy include the elderly (Smith and Steele, 1983), patients with pre-existing renal insufficiency (Kaufman and Levit, 1985), patients with right-sided HF or pericardial disease, and patients taking NSAIDs. In a case–control study of NSAID use and renal failure, diuretic users had a 2.77 relative risk of AKI, compared with those who did not use the drugs (Huerta et al., 2005). Disorders of Na+ and K+ concentration are among the most frequent adverse effects of loop diuretics. Hyponatraemia is less common with loop diuretics than with distal convoluted tubule diuretics (see later), but can occur. It is usually multifactorial, but involves both depletion of the ‘effective’ arterial volume and impairment of free water clearance. Additional factors that may contribute include the non-osmotic release of arginine vasopressin (Bichet et  al., 1982), hypokalaemia, and hypomagnesaemia (Dyckner and Webster, 1982). Conversely, loop diuretics have been used to treat hyponatraemia when combined with hypertonic saline in the setting of the syndrome of inappropriate antidiuretic hormone (ADH) secretion (Hantman et al., 1973; Schrier, 1978). The combination of loop diuretics and ACEIs has been reported to ameliorate hyponatraemia in the setting of congestive HF (Dzau and Hollenberg, 1984). The value of adding a loop diuretic to treatment with a vasopressin V2 receptor antagonist for hyponatraemic syndrome of inappropriate ADH secretion has been suggested (Shimizu, 2003; Kazama et al., 2005). Hypokalaemia occurs commonly during therapy with loop diuretics, although the magnitude is smaller than that induced by distal convoluted tubule diuretic (loop diuretics, 0.3 mmol/L versus DCT diuretics, 0.5–0.9 mmol/L) (Ram et al., 1981; Palmer, 1997). Loop diuretics increase the delivery of potassium to the distal tubule, because they block potassium reabsorption via the Na-K2Cl cotransporter. In rats, under control conditions, approximately half the excreted potassium was delivered to the ‘early’ distal tubule. During furosemide infusion, the delivery of potassium to the ‘early’ distal tubule rose to 28% of the filtered load (Hropot et al., 1985). Thus, a component of the effect of loop diuretics on potassium excretion reflects their ability to block potassium reabsorption by the thick ascending limb, but during chronic diuretic therapy the degree of potassium wasting correlates best with ECF volume contraction and serum aldosterone levels (Wilcox et  al., 1984). This

suggests that when used chronically loop diuretics stimulate potassium excretion primarily because they increase mineralocorticoid hormones and increase distal Na+ and water delivery into the aldosterone-sensitive distal nephron (ASDN). Metabolic alkalosis is very common during chronic treatment with loop diuretics. Loop diuretics cause metabolic alkalosis via several mechanisms. First, they increase the excretion of urine that is bicarbonate free but contains Na+ and Cl−. This leads to contraction of the ECF around a fixed amount of bicarbonate buffer; a phenomenon known as ‘contraction alkalosis’. Second, loop diuretics directly inhibit transport of Na+ and Cl− into thick ascending limb cells, which may stimulate H+ secretion via Na+/H+ exchange (Good et al., 1984; Oberleithner et al., 1984; Good, 1985). Third, loop diuretics stimulate the renin–angiotensin–aldosterone system, increasing Na+ reabsorption along the ASDN, which renders the tubule lumen more negative and increases H+-ATPase activity (O’Neil et al., 1977). Aldosterone also activates the vacuolar H+-ATPase in the outer medullary collecting tubule directly (Stone et al., 1983; Winter et al., 2004). Hypokalaemia itself also contributes to metabolic alkalosis by increasing ammonium production (Tannen, 1970), stimulating bicarbonate reabsorption by proximal tubules (Soleimani et al., 1987; Soleimani and Aronson, 1989), and increasing the activity of the H+/K+ ATPase in the distal nephron (Wingo and Straub, 1989; Okusa et al., 1992). Some of these effects may be offset because loop diuretics also strongly increase the expression and activity of pendrin, a chloride/bicarbonate exchanger expressed by type B intercalated cells (Quentin et al., 2004; Na et al., 2007). Ototoxicity with deafness is the most common toxic effect of loop diuretics unrelated to their effects on the kidney. It appears likely that all loop diuretics cause ototoxicity, because ototoxicity can occur during use of chemically dissimilar drugs such as furosemide and ethacrynic acid (Maher and Schreiner, 1965; Nochy et  al., 1976). The stria vascularis, which is responsible for maintaining endolymphatic potential and ion balance, appears to be a primary target for toxicity (Ikeda et al., 1997). A characteristic finding in loop diuretic ototoxicity is strial oedema, because an isoform of the Na-K-2Cl cotransporter in expressed there (Mizuta et  al., 1997). Loop diuretics cause loss of outer hair cells in the basal turn of the cochlea, rupture of endothelial layers, cystic formation in the stria vascularis, and marginal cell oedema in the stria vascularis (Ryback, 1993). Ototoxicity appears to be related to the peak serum concentration of loop diuretic and therefore tends to occur during rapid drug infusion of high doses. For this reason, this complication is most common in patients with uraemia (Star, 1997). It has been recommended that furosemide infusion be no more rapid than 4 mg/ minute (Wigand and Heidland, 1971). In addition to those with renal failure, infants, patients with cirrhosis, and patients receiving aminoglycosides or cis-platinum may be at increased risk for ototoxicity (Star, 1997).

Distal convoluted tubule diuretics (thiazides) The distal convoluted tubule diuretics represent a distinct and important class of diuretics; many are analogues of 1,2,4-benzothiadiazine-1,1-dioxide, but other structurally related diuretics, including the quinazolinones (such as metolazone) and

chapter 33 

substituted benzopehenone sulphonamide (such as chlorthalidone) also appear to share the same mechanism of action. Although the term ‘thiazide diuretics’ is frequently used to describe this class, a more accurate descriptor is distal convoluted tubule (DCT) diuretics. Acute administration of these drugs increases Na, K, Cl, HCO3, phosphate, and urate excretion (Table 33.1). The increases in HCO3, phosphate, and urate excretion are probably related primarily to carbonic anhydrase inhibition, and not to inhibition of the Na-Cl cotransporter (see below). As such, the effects of DCT diuretics to increase HCO3, phosphate, and urate excretion may vary, depending on the carbonic anhydrase inhibiting potency. Chronically, as ECF volume contraction occurs, uric acid excretion declines and hyperuricaemia can occur (Toto, 1997). Further, bicarbonate excretion ceases, and continuing losses of chloride without bicarbonate, coupled with ECF volume contraction, may lead to metabolic alkalosis. In contrast to loop and proximally acting diuretics, DCT diuretics reduce urinary calcium excretion (see below). DCT diuretics inhibit the clearance of solute free water when administered during water diuresis, because their site of action is a portion of the renal diluting segment. In contrast to loop diuretics, however, DCT diuretics do not limit water retention during antidiuresis.

Na+ and water transport in the proximal tubule Most DCT diuretics retain some carbonic anhydrase inhibiting activity (Goldfarb et al., 1991). Although this effect occurs during acute treatment (as during intravenous chlorothiazide administration), it probably contributes relatively little to overall natriuresis during chronic use (Kunau et al., 1975; Walter and Shirley, 1986). Yet this effect may play a role in the tendency for DCT diuretics to reduce the GFR by activating TGF (Okusa et al., 1989). The relative carbonic anhydrase inhibiting potency of some commonly used DCT diuretics (shown in parentheses) is chlorthalidone (67) > benthiazide (50) > polythiazide (40) > chlorothiazide (14) > hydrochlorothiazide (1)  > bendroflumethiazide (0.07) (Friedman and Hebert, 1997, pp. 75–111).

NaCl absorption in the distal nephron As the name indicates, the predominant site at which DCT diuretics inhibit ion transport is the DCT. The predominant action of these drugs is to inhibit the thiazide-sensitive Na-Cl cotransporter (NCC) encoded by SLC12A3 (Fig. 33.1). Like the loop diuretics, DCT diuretics are organic anions that bind to and inhibit NCC directly, although the specific site on which the drugs inhibit NCC continues to be debated (Tran et al., 1990; Monroy et al., 2000; Moreno et al., 2006). There is some evidence that thiazides inhibit solute transport in medullary collecting tubules of rats (Wilson et al., 1983) and in the cortical collecting ducts when animals are NaCl deprived (Terada and Knepper, 1990). Most recently, Leviel and colleagues found that thiazides inhibit a sodium-dependent chloride-bicarbonate exchanger (NDCBE) in the collecting duct of animals exposed to low NaCl diets (Leviel et al., 2010). There is now evidence that this pathway does contribute to the net effect of DCT diuretics (Soleimani et al., 2012).

Ca2+ and Mg2+ transport When administered chronically, DCT diuretics reduce calcium excretion, but the mechanisms remain controversial. Acute

clinical use of diuretics

administration of DCT diuretics has a variable effect on calcium excretion (Eknoyan et al., 1970; Popovtzer et al., 1975), probably reflecting the carbonic anhydrase inhibition along the proximal tubule. During chronic treatment, DCT diuretics reduce calcium excretion, and the effect likely occurs at several levels. First, the filtered calcium load may decrease slightly owing to ECF volume depletion and a decline in GFR. Second, proximal calcium reabsorption increases, as contraction of the ECF volume stimulates proximal Na+ reabsorption, thereby stimulating calcium reabsorption secondarily. Third, DCT diuretics increase renal calcium reabsorption along the distal nephron (Costanzo and Windhager, 1978). Stimulation of distal calcium reabsorption is accompanied by an increase in intracellular calcium activity, suggesting that a primary effect is to increase apical calcium entry (Gesek and Friedman, 1992). The drugs, however, may also enhance basolateral calcium uptake, as DCT cells express the Na/Ca exchanger and a Ca-ATPase at the basolateral cell membrane. The Na/Ca exchanger is electrogenic, and when the intracellular Na+ concentration declines, the electrochemical driving force favouring calcium movement from cell to interstitium increases. The roles of proximal and distal processes in effects of DCT diuretics on urinary calcium excretion continue to be debated (Reilly and Huang, 2011). Lee and colleagues confirmed an acute effect of DCT diuretics on distal calcium uptake, but also found that a large portion of the chronic effects of DCT diuretic required ECF volume depletion. They speculated that acute exposure to DCT diuretics activates calcium channels (TrpV5) along the distal tubule. During chronic exposure, however, they noted that ECF volume contraction reduces distal NaCl delivery, thereby reducing this effect (Lee et al., 2004). Nijenhuis and colleagues found that DCT diuretic reduced calcium excretion, even when TrpV5 was deleted genetically, suggesting a predominant role for enhanced proximal reabsorption (Nijenhuis et  al., 2005). They suggested that distal processes contribute little to the overall effect. Yet there are compelling data indicating that effects along the DCT do play a role. In humans ECF volume is not altered substantially during chronic treatment. In animals with hypocalciuria resulting from deletion of salt transporting genes or regulators along the distal nephron, salt loading does not correct the hypocalciuria (McCormick et al., 2011). Similarly, in humans with Gitelman syndrome in which the thiazide-sensitive Na-Cl cotransporter is dysfunctional, saline infusion increases NaCl excretion, without correcting the hypocalciuria (Cheng et al., 2007). Finally, in mice with genetic disruption of parvalbumin, which disrupts NCC activity without causing ECF volume depletion, hypocalciuria is also observed (Belge et al., 2007). Overall, it appears that both proximal and distal processes contribute to the hypocalciuric effect of DCT diuretics (Reilly and Huang, 2011). DCT diuretics enhance magnesium excretion, but the effects are generally much less profound than their effects on calcium excretion. This is in contrast to the effects of genetic NCC deletion or inactivity, as occurs in Gitelman syndrome, where hypomagnesaemia is a cardinal feature. Acute thiazide infusions have little effect on magnesium excretion (Duarte, 1968; Eknoyan et al., 1970; Quamme et al., 1975), whereas chronic administration may cause hypomagnesaemia (Hollifield, 1989; Douban et al., 1996; Quamme, 1997). One important pathway for magnesium reabsorption across the apical membrane of DCT cells is the transient receptor potential, TrpM6 (Schlingmann et al., 2002). The predominant mechanism

305

306

Section 2  

fluid, electrolyte, and renal tubular disorders

for DCT diuretic-induced magnesium wasting involves destruction of DCT cells, from apoptosis induced by DCT diuretics (Loffing et al., 1996, 2004). Several groups have reported that inactivation of NCC reduces the abundance of Trpm6, which would be expected to impair magnesium reabsorption (Nijenhuis et al., 2004, 2005). DCT diuretics increase renal vascular resistance and decrease the GFR when given acutely. Okusa et al. (1989) showed that intravenous chlorothiazide reduced the GFR by 16% only when TGF was intact. During chronic treatment with DCT diuretics, mild contraction of the ECF volume develops, thereby increasing solute and water reabsorption by the proximal tubule. This effect reduces Na+ delivery to the macula densa. This would be expected to return GFR toward baseline values during chronic treatment with DCT diuretics (Earley and Orloff, 1962; Walter and Shirley, 1986). Thus, when used chronically, DCT diuretics lead to a state of mild ECF volume contraction, increased fractional proximal reabsorption, and relatively preserved GFR (Earley and Orloff, 1962; Walter and Shirley, 1986). When administered acutely, the effect of DCT diuretics on renin secretion is variable (McGuffin and Gunnells, 1969). If urinary NaCl losses are replaced, these drugs tend to suppress renin secretion (Brown et al., 1966), probably by increasing NaCl delivery to the macula densa (Okusa et al., 1989). In contrast, during chronic administration, renin secretion increases both because solute delivery to the macula densa declines (Walter and Shirley, 1986) and because volume depletion activates the vascular mechanism for renin secretion. Like the loop diuretics, DCT diuretics are organic anions that circulate in a highly protein bound state. As a result, the predominant route of entry into tubular fluid is by secretion via the organic anion secretory pathway in the proximal tubule (Brater, 1997). DCT diuretics are rapidly absorbed across the gut, reaching peak concentrations within 1.5–4 hours (Brater, 1997). The amount of administered drug that reaches the urine varies greatly (for a review see Brater, 1997), as does the half-life. Shorter acting DCT diuretics include bendroflumethiazide, hydrochlorothiazide, tizolemide, and trichlormethiazide. Medium-acting DCT diuretics include chlorothiazide, hydroflumethiazide, indapamide, and mefruside. Long-acting DCT diuretics include chlorthalidone, metolazone, and polythiazide (Brater, 1997). These differences in half-life may have implications with regard to the efficacy of these drugs for the treatment of hypertension (Flack et  al., 2011), and perhaps with regard to the incidence of hypokalaemia, which may be more common in patients taking the longer-acting drugs such as chlorthalidone (Dhalla et al., 2013). DCT diuretics are used most commonly to treat essential hypertension; their use in this situation is beyond the scope of this chapter. DCT diuretics have become drugs of choice to prevent the recurrence of kidney stones in patients with idiopathic hypercalciuria. In several controlled and many uncontrolled studies, the recurrence rate for calcium stones has been reduced by up to 80% (Yendt and Cohanim, 1978; Laerum and Larsen, 1984; Ettinger et al., 1988). Relatively high doses of DCT diuretics are often employed for the treatment of nephrolithiasis (Breslau, 1997). Some studies suggest that the hypocalciuric effect of DCT diuretics wanes during chronic use, in the setting of absorptive hypercalciuria (Preminger and Pak, 1987). The observation that Gitelman syndrome, an inherited disorder of NCC inactivity, may present during adulthood with hypocalciuria suggests that compensatory mechanisms may not exist for

the effects of DCT diuretics on calcium transport (Ellison, 2000). Reilly and colleagues reviewed DCT diuretic use for nephrolithiasis and suggested doses of indapamide at 2.5 mg/day, chlorthalidone at 25–50 mg/day, or HCTZ at 25 mg twice daily or 50 mg/daily (Reilly et al., 2010). The ability of DCT diuretics to reduce urinary calcium excretion suggests that these drugs may prevent bone loss. Some (Ray et al., 1989; Felson et al., 1999), but not all (Heidrich et al., 1991; Cauley et  al., 1993), epidemiological studies suggest that DCT diuretics reduce the risk of hip fracture and osteoporosis. A  randomized controlled study confirmed that DCT diuretics reduce bone loss in women (Reid et al., 2000). DCT diuretics are also employed to treat nephrogenic diabetes insipidus, causing a paradoxical decrease in urinary volume and flow rate. This action of DCT diuretics results from the combination of mild ECF volume contraction, owing to diuretic-induced natriuresis and suppression of GFR, due largely to diuretic-induced activation of TGF. The DCT, like the thick ascending limb, is nearly impermeable to water (Coleman et al., 1997). Solute reabsorption by NCC, therefore, contributes directly to urinary dilution. The central role of ECF volume contraction in the efficacy of DCT diuretics in diabetes insipidus was highlighted by the observation that dietary salt restriction is necessary to reduce urinary volume effectively (Earley and Orloff, 1962; Janjua et al., 2001). DCT diuretics may also increase the ADH-independent water permeability of the medullary collecting tubule (Cesar and Magaldi, 1999). DCT diuretic treatment increased the abundance of aquaporin-2, NCC and the alpha subunit of the epithelial Na channel (Kim et al., 2004), when administered to rats with lithium-induced nephrogenic diabetes insipidus. It was suggested that the upregulation of the abundance of the renal Na and water transporters might explain the antidiuretic effectiveness of DCT diuretics. Electrolyte disorders, such as hypokalaemia, hyponatraemia, and hypomagnesaemia are common side effects of DCT diuretics. A  measurable decline in serum K+ concentration is nearly universal in patients given DCT diuretics, but most patients do not become frankly hypokalaemic. In the ALLHAT trial, mean serum K+ concentrations declined from 4.3 to 4.0 and 4.1 mmol/L, after 2 and 4  years of treatment (ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group, 2002b). The clinical significance of diuretic-induced hypokalaemia continues to be debated. Unlike the loop diuretics, DCT diuretics do not influence K+ transport directly (Velázquez and Wright, 1986), but instead increase tubule fluid flow and Na+ concentration in the CNT and collecting duct. In addition, DCT diuretic-induced ECF volume contraction activates the renin–angiotensin–aldosterone system, further stimulating K+ secretion. Evidence for the central role of aldosterone in diuretic-induced hypokalaemia includes the observation that hypokalaemia is more common during treatment with long-acting DCT diuretics, such as chlorthalidone, than with shorter-acting DCT diuretics, such as hydrochlorothiazide, or with the very short-acting loop diuretics (Ram et al., 1981; Dhalla et al., 2013). Another reason that DCT diuretics may produce more K+ wasting than loop diuretics is the difference in effect on calcium transport. As discussed above, loop diuretics inhibit calcium transport by the thick ascending limb, increasing distal calcium delivery; in contrast, DCT diuretics stimulate calcium transport, reducing calcium delivery to sites of potassium secretion. Okusa and colleagues (1990) showed that high luminal concentrations of calcium

chapter 33 

inhibit the functional activity of epithelial sodium channels (ENaC) in the distal nephron, thereby inhibiting potassium secretion. DCT diuretics also increase urinary magnesium excretion and can lead to hypomagnesaemia, which may also contribute to hypokalaemia (Rude, 1989; Dorup et al., 1993; Huang and Kuo, 2007). Some studies suggest that maintenance magnesium therapy can prevent or attenuate the development of hypokalaemia (Dorup et al., 1993), but this has not been supported universally. Diuretics have been reported to contribute to more than one half of all hospitalizations for serious hyponatraemia. Hyponatraemia is especially common during treatment with DCT diuretics, compared with other classes of diuretics, and the disorder is potentially life-threatening (Ashraf et al., 1981). A recent case–control study suggested that hyponatraemia during thiazide treatment is more common than generally appreciated, but that in most cases, it does not prove morbid (Leung et al., 2011); on the other hand, some studies do suggest an association with mortality (Liamis et al., 2013). Several factors contribute to DCT diuretic-induced hyponatraemia. First, DCT diuretics inhibit solute transport in the terminal portion of the ‘diluting segment’. Second, DCT diuretics reduce the GFR, limiting solute delivery to the diluting segment and impairing solute-free water clearance. Third, DCT diuretics lead to ECF volume contraction, which increases proximal tubule solute and water reabsorption. Fourth, hyponatraemia has been correlated with the development of hypokalaemia (Fichman et  al., 1971). Finally, susceptible patients may be stimulated to consume water during therapy with DCT diuretics. One report suggests that patients who are predisposed to develop hyponatraemia during treatment with DCT diuretics will demonstrate an acute decline in serum sodium concentration in response to a single dose of the drug (Friedman et al., 1989). Other studies suggest that risk factors for DCT diuretic-induced hyponatraemia include older age, lower body mass, and concomitant administration of selective serotonin reuptake inhibitors or benzodiazepines (Chow et al., 2003; Neafsey, 2004; Liamis et al., 2013). DCT diuretics frequently cause mild metabolic alkalosis. The mechanisms are similar to those described above for loop diuretics, except that DCT diuretics do not stimulate Na/H exchange in the TAL. Glucose intolerance has been a recognized complication of DCT diuretic use since the 1950s and appears to be dose related (Carlsen et al., 1990; Harper et al., 1995). In the ALLHAT trial there was a 1.8% increase in new onset diabetes at 4 years of treatment with chlorthalidone versus patients treated with calcium channel blockers (ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group, 2002a). This difference did not translate into adverse clinical outcomes in the diuretic group, but has generated a great deal of discussion. The pathogenesis of DCT diuretic-induced glucose intolerance remains unclear, but several factors have been suggested to contribute. First, diuretic-induced hypokalaemia may decrease insulin secretion by the pancreas, via effects on the membrane voltage of pancreatic beta cells. When hypokalaemia was prevented by oral potassium supplementation, the insulin response to hyperglycaemia normalized, suggesting an important role for hypokalaemia (Helderman et al., 1983). Hypokalaemia may also interfere with insulin-mediated glucose uptake by muscle, but most patients demonstrate relatively normal insulin sensitivity (Toto, 1997). ECF volume depletion may

clinical use of diuretics

stimulate catecholamine secretion, but volume depletion during therapy with DCT diuretics is usually very mild. It has also been suggested that DCT diuretics directly activate calcium-activated potassium channels that are expressed by pancreatic beta cells (Pickkers et  al., 1996). Activation of these channels is known to inhibit insulin secretion. Inhibiting the renin–angiotensin–aldosterone axis appears to reduce the development of new diabetes (Scheen, 2004). Drugs that inhibit this pathway might attenuate the effects of diuretics to impair glucose homeostasis, but this has not been tested directly. Other factors may contribute to glucose intolerance as well, including drug-specific factors (Ellison and Loffing, 2009). DCT diuretics increase levels of total cholesterol, total triglyceride, and LDL cholesterol, and reduce HDL (Toto, 1997; Wilcox, 1999). Definitive information on the mechanisms by which DCT diuretics alter lipid metabolism is not available, but many of the mechanisms that affect glucose homeostasis have been suggested to contribute. Hyperlipidaemia, like hyperglycaemia, is a dose-related side effect, and one that wanes with chronic diuretic use. In the ALLHAT study, treatment with chlorthalidone resulted in total cholesterol 2.2 mg/dL (0.06  mmol/L) higher than did treatment with ACEIs (ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group, 2002a). In several large clinical studies, the effect of low-dose DCT diuretic treatment on serum LDL was not significant (Grimm et al., 1996). Furthermore, treatment of hypertension with DCT diuretics reduces the risk of stroke, coronary heart disease, congestive HF, and cardiovascular mortality.

Cortical collecting duct diuretics (distal potassium-sparing diuretics) Diuretic drugs that act primarily in the cortical collecting duct (CCD) or the CNT (potassium-sparing diuretics) comprise three pharmacologically distinct groups:  aldosterone antagonists (spironolactone and eplerenone), pteridines (triamterene), and pyrazinoylguanidines (amiloride). The site of action for all diuretics of this class is the CNT and the CCD, where they interfere with sodium reabsorption and indirectly potassium secretion. The recently introduced vasopressin V2-receptor antagonists (the ‘vaptans’) also act in the collecting duct and could be categorized as diuretics (Decaux et  al., 2008). Because vasopressin-receptor antagonists are primarily used for the treatment of hyponatraemia secondary to the syndrome of inappropriate ADH secretion, HF, or liver cirrhosis, these compounds are discussed in other chapters. The diuretic activity of amiloride, triamterene and aldosterone antagonists is weak acutely. Because these drugs are relatively weak natriuretic agents, they are used most commonly in combination with thiazides or loop diuretics, in combination or as a single preparation, to restrict potassium losses. In certain conditions, however, potassium-sparing diuretics are used as first-line agents (see below). Mineralocorticoid blocking drugs have become standard parts of the treatment of patients with systolic dysfunction HF, where these drugs reduce mortality of patients with severe (Pitt et al., 1999) or mild HF (Pitt et al., 2001; Zannad et al., 2011). Amiloride, triamterene and spironolactone are weak natriuretic agents when given acutely (Table 33.1). Additionally, these agents decrease hydrogen ion secretion by the late distal tubule and collecting ducts. A  common mechanism is likely to be involved in mediating the effects of all three diuretic agents on hydrogen ion secretion, as they all reduce the lumen-negative voltage in the distal

307

308

Section 2  

fluid, electrolyte, and renal tubular disorders

nephron and thus decrease the electrochemical gradient favouring hydrogen ion secretion. Clearance studies in rats have demonstrated that amiloride decreases calcium excretion (Costanzo and Weiner, 1976). In these studies, amiloride produced both a decrease in the calcium clearance/sodium clearance ratio (CCa/CNa), as well as a decrease in the fractional excretion of calcium. Amiloride is believed to stimulate calcium absorption through its ability to block sodium channels, thereby hyperpolarizing the apical membrane (Friedman PA, Gesek, 1995). Hyperpolarization of the apical membrane may stimulate TrpV5, as discussed above. Amiloride has also been reported to reduce magnesium excretion (Bundy et al., 1995; Douban et al., 1996) and to prevent the development of hypomagnesaemia during therapy with a DCT diuretic (Dyckner et al., 1988). The site of action of potassium-sparing diuretics is the CNT and collecting  duct. Although these segments reabsorb only a small percentage of the filtered Na+ load, two characteristics render this segment important in the physiology of diuretic action. First, these segments are the primary site of action of the mineralocorticoid, aldosterone, a hormone that controls Na+ reabsorption and K+ secretion. Second, virtually all of the potassium that is excreted is due to the secretion of potassium by the connecting and collecting tubules. The apical membrane of cells in these segments expresses separate channels that permit selective conductive transport of sodium and potassium (Fig. 33.1). The low intracellular sodium concentration as a result of the basolateral Na/K-ATPase generates a favourable electrochemical gradient for sodium entry through sodium channels. Thus, sodium conductance depolarizes the apical membrane, resulting in a lumen-negative transepithelial potential difference, which tends to favour K+ secretion. Amiloride-sensitive sodium conductance is a function of the ENaC. The amount of sodium and potassium present in the final urine is tightly controlled by aldosterone action on ENaC. The cellular mechanisms that are responsible for these events have been extensively studied and reviewed (Thomas et al., 2008).

Mineralocorticoid receptor blockers Spironolactone (Fig. 33.1) is an analogue of aldosterone that is extensively metabolized (Karim, 1978; Shackleton et  al., 1986), having the principal effect of blocking aldosterone action (Fanestil, 1988; Menard, 2004). Spironolactone is converted by deacylation to 7a-thiospironolactone or by diethioacetylation to canrenone (Fanestil, 1988). In the kidney, spironolactone and its metabolites enter target cells from the peritubular side, bind to cytosolic mineralocorticoid receptors, and act as competitive inhibitors of the endogenous hormone. Spironolactone induces a mild increase in sodium excretion (1–2%) and a decrease in potassium and hydrogen ion excretion (Kagawa, 1960; Liddle, 1966). Its effect depends on the presence of aldosterone (Coppage and Liddle, 1960; Botero-Velez et al., 1994). In cortical collecting tubules perfused in vitro, spironolactone added to the bath solution reduced the aldosterone induced lumen negative transepithelial voltage (Gross and Kokko, 1977). By blocking sodium absorption in the collecting tubule, a decrease in lumen negative voltage reduces the driving force for passive sodium and hydrogen ion secretion (Gross and Kokko, 1977). Spironolactone commonly causes troubling oestrogenic side effects. Eplerenone was developed as a competitive aldosterone

antagonist that is more selective for mineralocorticoid receptors and therefore less likely to cause troubling side effects. Eplerenone was derived from spironolactone; in humans, it appears to be 50–75% as potent in inhibiting mineralocorticoid receptors (Weinberger et al., 2002; Brown, 2003). This structural modification significantly enhances the relative affinity of the drug for mineralocorticoid receptors over other steroid receptors.

Amiloride and triamterene Amiloride and triamterene are both organic cations that act via the same mechanism (Fig. 33.1). Their actions on sodium and potassium transport, unlike spironolactone, are not dependent on aldosterone. Systemically administered amiloride produced a small increase in sodium excretion and a much larger decrease in potassium excretion (Duarte et  al., 1971; Giebisch, 1978). Amiloride decreases potassium secretion by blocking ENaC in the apical membrane of CNT and collecting tubule cells (Koeppen et  al., 1983; O’Neil et al., 1984), thereby decreasing the electrochemical gradient for potassium secretion, although in higher concentrations (>100 µmol/L), amiloride can inhibit Na+/H+ exchange along the proximal tubule. As used clinically, however, amiloride interacts specifically with ENaC (Garty and Benos, 1988; Garty, 1994). The molecular mechanism by which amiloride blocks ENaC remains incompletely defined, but it appears that the drug occludes the pore of the sodium channel, ENaC (Garty, 1994; Garty and Palmer, 1997). Clearance and free-flow micropuncture studies using triamterene demonstrated results similar to studies with amiloride (Hropot et al., 1985), although its mechanism of action is not as clearly defined (Busch et al., 1996). The bioavailability of spironolactone is approximately 90%. The drug is rapidly metabolized in the liver into a number of metabolites (see Karim, 1978; Shackleton et al., 1986). Spironolactone and its metabolites are extensively bound to plasma protein (98%). In normal volunteers, taking spironolactone (100 mg/ day) for 15  days, the mean half-lives for spironolactone, canrenone, 7a-thiomethylspironolactone and 6b-hydroxy-7athiomethylspironolactone were 1.4, 16.5, 13.8, and 15 hours, respectively. Thus, although unmetabolized spironolactone is present in serum, it has a rapid elimination time. The onset of its physiological action, however, is extremely slow for spironolactone, with peak response sometimes occurring 48 hours or more after the first dose; effects gradually wane over a period of 48–72 hours. Spironolactone is used in cirrhotic patients to induce a natriuresis. In these patients, pharmacokinetic studies indicate that the half-lives of spironolactone and its metabolites are increased (Table 33.3). The half-lives for spironolactone, canrenone, 7a-thiomethylspironolactone and 6b-hydroxy-7athiomethylspironolactone are 9, 58, 24, and 126 hours respectively (Sungaila et al., 1992). Eplerenone is rapidly absorbed, with peak serum levels at 1.5 hours (Brown, 2003). Its volume of distribution is 43–90 L, with approximately 50% protein bound. It is cleared primarily via the CYP4503A4 system to inactive metabolites with an elimination half-life of 4–6 hours (Brown, 2003). This is in contrast to spironolactone, where the half-life of the parent compound is short, but the half-life of metabolites is very long. The maximal plasma concentration and area under the curve are increased in people > 65 years of age and with kidney failure; eplerenone is not removed by haemodialysis (Brown, 2003).

chapter 33 

clinical use of diuretics

Table 33.3  Pharmacokinetics of loop diuretics Elimination half-life (hours) Bioavailability, % oral dose absorbed

Healthy

Kidney disease

Liver disease

Heart failure

1.5–2

2.8

2.5

2.7

Furosemide

50% (range, 10–100%)

Bumetanide

80–100%

1

1.6

2.3

1.3

Torsemide

80–100%

3-4

4-5

8

6

Data from Shankar, S.S. (2003). Am J Physiol, 284, F11–F21.

CCT diuretics can be used for the treatment of hypertension, primary aldosteronism, and secondary aldosteronism; they are also used to limit the kaliuretic effects of loop or DCT diuretics, and sometimes primarily to treat hypokalaemia due to renal potassium loss of various causes. Spironolactone (or eplerenone) is the treatment of choice in patients with primary aldosteronism due to bilateral adrenal hyperplasia (Brown et al., 1972; Ganguly, 1998). The drug is also appropriate for cirrhosis with ascites (see below). A  third use of spironolactone (or eplerenone) is in systolic HF, where mineralocorticoid antagonists have been shown to reduce morbidity and mortality (Zannad et  al., 2011; Markowitz et  al., 2012). Finally, there is growing interest in using spironolactone to treat resistant hypertension, even when demonstrable hyperaldosteronism is not present (Calhoun et al., 2008). Triamterene or amiloride is generally used in combination with potassium-wasting diuretics (thiazide or loop diuretics), especially when maintenance of normal serum potassium concentrations is clinically important. In addition, amiloride (or triamterene) has also been used as initial therapy in potassium wasting states such as primary hyperaldosteronism (Ganguly and Weinberger, 1981; Griffing et al., 1982), Liddle syndrome (Botero-Velez et al., 1994), Bartter syndrome, or Gitelman syndrome (Okusa et  al., 1987), although use in the latter situation has been disputed (Seyberth et al., 2011). Amiloride is recommended to treat lithium-induced nephrogenic diabetes insipidus (Batlle et al., 1985), where it blocks the pathway by which lithium gains entry into cells (Christensen et al., 2011). A small placebo-controlled cross-over trial (Bedford et al., 2008) and an animal study (Kortenoeven et al., 2009) confirmed these effects. The most serious adverse reaction encountered during therapy with the CCD diuretics is hyperkalaemia. Serum K+ should be monitored periodically, even when the drugs are administered with a potassium-wasting diuretic. Patients at highest risk are those with low GFR, patients with concurrent medication predisposing to hyperkalaemia, and individuals who take potassium supplements concurrently (Chapagain and Ashman, 2012). This problem has become more important, because of the wide use of aldosterone blocking drugs, together with ACEIs, ARBs, and beta blockers in patients with HF (Juurlink et al., 2004). Another group at risk for hyperkalaemia are the elderly receiving chronic treatment with spironolactone and who are intermittently treated with trimethoprim-sulfamethoxazole for a urinary tract infection (Antoniou et  al., 2011a, 2011b). Surprisingly, however, another recent population-based study showed that despite a marked increase in the use of spironolactone, no increase was seen in

hospital admissions for hyperkalaemia and that rates of outpatient hyperkalaemia actually fell; the authors ascribed these findings to more careful monitoring (Wei et al., 2010). In patients with cirrhosis and ascites treated with spironolactone, hyperchloraemic metabolic acidosis can develop independent of changes in renal function (Gabow et al., 1979). Gynaecomastia may occur in men, especially as the dose is increased (Rose et al., 1977), but even at low doses (The Randomized Aldactone Evaluation Study, 1996); decreased libido and impotence have also been reported. Women may develop menstrual irregularities, hirsutism, or breast swelling and tenderness. Spironolactone induced agranulocytosis has also been reported (Whitling et al., 1997). Triamterene and amiloride may also cause hyperkalaemia. The risk of hyperkalaemia is highest in patients with limited renal function (e.g. renal insufficiency, diabetes mellitus, and elderly patients). Additional complications included elevated serum blood urea nitrogen and uric acid, glucose intolerance, and gastrointestinal disturbances. Triamterene induces crystalluria or cylinduria (Fairley et al., 1986), which may contribute to or initiate formation of renal stones (Carr et al., 1990), and may cause AKI when combined with non-steroidal anti-inflammatory agents (Favre et  al., 1982; Weinberg et al., 1985). Cortical collecting duct diuretics are contraindicated in patients with hyperkalaemia, individuals taking potassium supplements in any form, and in patients with severe renal failure with progressive oliguria.

Clinical use of diuretics General concepts Determinants of maximal diuresis The change in urinary flow seen during the administration of a diuretic depends on many factors, including its mechanism of action, dose, kinetics of entry into the bloodstream, and delivery to its site of action. The maximal diuretic effect of these drugs is determined, largely by the transport protein and nephron site of action. For example, loop diuretics have a higher ‘ceiling’ action than DCT diuretics. This observation results from the fact that loop diuretics inhibit a transport pathway responsible for reabsorbing up to 25% of the filtered sodium load, while DCT diuretics inhibit a pathway responsible for reabsorbing only 5–10%. Similarly, mineralocorticoid antagonists have a mild natriuretic effect due to the fact that they suppress a pathway responsible for reabsorbing only 3% of the filtered Na load. There are, of course, exceptions to this rule. The carbonic anhydrase inhibitors, which reduce proximal tubule reabsorption are only weakly natriuretic, due to adaptive changes in the loop of Henle and DCT (Lorenz et al., 1999). Within classes, drugs

309

Section 2  

fluid, electrolyte, and renal tubular disorders

vary in potency (effect/amount of drug required). For example, a lower dose of bumetanide is required to elicit the same effect as furosemide, even though the maximal natriuretic effect of each is similar. Diuretic efficacy is also dependent on the kinetics of drug entry into the bloodstream. The dynamics of drug absorption may be perturbed in certain clinical situations, and this might result in a diminished effect. This is exemplified by the pharmacokinetics of furosemide. In normal individuals, the rate of furosemide absorption from the GI tract is not rapid, and a reservoir of drug can persist long after the diuretic is administered (Brater, 1997). This reservoir makes the effective half-life of the drug longer than the actual plasma half-life. In certain oedematous states, however, absorption from the gut may be slowed, so that furosemide absorption never reaches the diuretic threshold, rendering it ineffective (Vasko et al., 1985). To compensate for this, giving a high dose of the drug, or switching to a different diuretic with better absorption, such as torsemide or bumetanide, may facilitate a brisker diuresis (Murray et al., 2001). Another approach is to switch to an intravenous loop diuretic preparation, which is, of course, 100% bioavailable. The effectiveness of a diuretic is also dependent on its rate of delivery to its site of action. In the cases of loop and DCT diuretics, these sites are at the luminal surface of the tubules. Brater established that diuretics such as furosemide have an excretion rate of maximal efficiency, that is, a rate of diuretic delivery that is associated with a maximal natriuretic response (Kaojarern et al., 1982). This concept helps to explain why an orally administered dose of furosemide can be more effective than an equivalent single intravenous dose in individuals with normal GI absorption (Fig. 33.2). When a dose of furosemide is given as an intravenous bolus, the rate of diuretic excretion is very high early on in the time-course, substantially greater than the rate of maximal efficiency. This rate tapers down over time, but the curve quickly dips below the maximal efficiency rate. In contrast, oral administration of the same dose of diuretic reaches the bloodstream more gradually, because of the processes described above regarding absorption. Thus, a reservoir

of furosemide in the GI tract may keep the circulating level above the natriuretic threshold and close to the rate of maximal diuretic efficiency for a longer period. As described above, most diuretics reach their sites of action via tubular secretion, primarily along the proximal tubule. These transport processes are relatively non-specific, and a single transporter type can facilitate the movement of a variety of similarly charged molecules into the tubular lumen. Accordingly, any exogenous or endogenous substance that competes with a diuretic for one of these transport processes can potentially limit the efficient arrival of that diuretic to its site of action. For instance, cimetidine, an organic cation, has been shown to inhibit the tubular secretion of amiloride (Somogyi et al., 1989). Other substances, such as NSAIDs, probenecid, penicillins, and uraemic anions all compete with loop and thiazide diuretics for tubular secretion (Rose et al., 1976; Brater, 1978). In certain disease states, competition between different drugs or endogenous substances for transport to the tubular lumen may lead to diuretic resistance. The prototypical example of such a condition is chronic kidney disease (CKD), in which diuretic delivery to the urine is impaired (Brater, 1978). In CKD, impaired drug delivery shifts the diuretic dose–response curve to the right, and a higher dose is required to achieve a diuretic effect (Fig. 33.2). This could potentially unmask the competitive effects of two different pharmacologic agents on an organic ion transport process, since a slight decrease in the rate of transport of the diuretic to the urinary space could make the tubular diuretic concentration fall beneath its threshold of effectiveness.

Diuretic adaptation and resistance Typically, the brisk increase in urinary solute and water excretion that is seen just after diuretic therapy is initiated wanes over several days of treatment (Fig. 33.3). This phenomenon occurs because certain renal and systemic adaptations take place in response to diuretic therapy. These adaptations are essential for chronic diuretic use, since continued depletion of the ECF volume would prove harmful; yet when they occur before the achievement of the desired

24 hours 140 120

The braking phenomenon

100 80 60 40 20

p

di ur et ic-

Post-diuretic NaCI retention

Lo o

di ur et icp Lo o

p Lo o

Lo o

di ur et ic-

6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6 6

p

0

di ur et ic-

Urinary Na excretion, mmoL/6 hours

310

Fig. 33.3  Effects of intermittent diuretic dosing. The figure shows urinary Na+ excretion, in 6-hour intervals, at baseline, and after dosing with a loop diuretic. Note that urinary Na+ excretion at baseline is near to dietary intake (dashed line). After each dose of diuretic, Na+ excretion rises transiently, and then declines below baseline. The phenomena of ‘Post-diuretic NaCl retention’ and ‘Braking’ are shown.

chapter 33 

ECF volume, they are causes of diuretic resistance. These adaptations can be classified into immediate, short-term, and chronic (Okusa and Ellison, 2008). Immediate adaptation refers to the instantaneous secondary changes in sodium transport along the nephron during diuretic-induced natriuresis. An example of an immediate diuretic adaptation would be the increased sodium reabsorption along the loop of Henle during acetazolamide use. As discussed, this effect is a major factor that limits the natriuretic effectiveness of carbonic anhydrase inhibitors. Coadministration of a loop diuretic with a carbonic anhydrase inhibitor can limit sodium reabsorption along the TAL and counteract the immediate diuretic adaptations. Short-term adaptation refers to stimulation of sodium reabsorption along the nephron once the diuretic concentration declines beneath the natriuretic threshold (Fig. 33.3). This phenomenon is often referred to as ‘post-diuretic NaCl retention,’ and has been attributed to three factors. First, short-term changes occur in response to an acute decrease in ECF volume. These effects are both renal and systemic and involve the activation of the renin–angiotensin–aldosterone axis and sympathetic nervous system, changes in GFR, and suppression of atrial natriuretic peptide secretion (reviewed in Ellison and Wilcox, 2008). The net effect of these responses is to enhance renal NaCl retention in an effort to increase ECF volume. Second, the decline of a diuretic drug concentration to a level beneath the natriuretic threshold induces rebound effects at its direct site of action. For example, in the case of loop diuretics, the number of NKCC2 cotransporters expressed at the apical surface of the TAL increases in response to a reduction in intracellular chloride concentration (Gagnon et al., 2004). While a loop diuretic is present in the lumen of the TAL, NaCl transport is inhibited despite any increase, but once the loop diuretic concentration declines, the increased transport capacity is unmasked. Third, post-diuretic NaCl retention occurs as a consequence of changes in sodium chloride reabsorption along nephron segments downstream of the diuretic’s molecular site of action. For example, the number of thiazide-sensitive cotransporters in the DCT increases within 60 minutes of loop diuretic administration (Chen et al., 1990). This effect is likely to be a consequence of the increase in luminal sodium chloride concentration, which activates molecular mechanisms that stimulate NCC synthesis and delivery to the DCT apical surface. Chronic adaptation, often termed the ‘braking phenomenon’ (Fig. 33.3), refers to the decline in natriuresis following each dose of diuretic, when the diuretic is administered repetitively. The braking phenomenon is likely due to a combination of factors. These factors include those that contribute to post-diuretic NaCl retention, such as the chronic intermittent stimulation of the sympathetic nervous and renin–angiotensin–aldosterone systems from ECF volume contraction. But other, more long-term changes also take place. One of the most significant of these is the effect of chronic diuretic therapy to induce structural changes in the epithelium lining the nephron. Specifically, chronic diuretic therapy can lead to both hypertrophy and hyperplasia of sodium chloride reabsorbing cells (Kaissling et al., 1985; Ellison et al., 1989). These effects act together to enhance the sodium chloride reabsorbing capacity of the nephron. For instance, loop diuretic infusions of 7 days increase the number and size of distal convoluted cells substantially (Kaissling et al., 1985; Ellison et al., 1989). Accordingly, the same treatment increases the number of active thiazide-sensitive NaCl

clinical use of diuretics

cotransporters in the DCT (Kaissling and Stanton, 1988; Stanton and Kaisslin, 1988; Ellison et  al., 1989). These changes increase the sodium chloride transport capacity (Ellison et al., 1989); this can undermine the therapeutic effectiveness of loop diuretics and contribute to diuretic resistance. Since chronic loop diuretic therapy increases the fraction of thiazide sensitive NaCl reabsorption, combining a low-dose thiazide with a loop diuretic can be a highly effective approach to counteracting resistance (see below).

Approach to the treatment of oedema The treatment of generalized oedema consists of four key interventions: optimizing treatment of the underlying disorder, dietary sodium restriction, measures to mobilize fluid from oedematous tissues, and diuretic drug therapy. The focus here will be on the contribution of diuretics to this approach.

Oral diuretic therapy When frank oedema is present, diuretics are usually necessary to reach therapeutic goals, even though dietary salt restriction is an essential element of the treatment regimen. In general, the goal of diuretic therapy in patients with ECF volume overload is to facilitate an efficient negative NaCl balance without compromising EABV substantially. The rate of fluid removal is dictated both by the urgency of need and safety. In some patients with HF, fluid readily moves from the interstitium to the intravascular compartment, and up to 2 L of oedema fluid can be removed per day without complications; in the outpatient setting, however, the goal is much less. In other situations, the rate of refilling can be slower, as in cirrhosis without peripheral oedema, where a negative fluid balance on the order of up to only 750 mL/day can be safely achieved without depleting intravascular volume (Pockros and Reynolds, 1986). Thus, in all outpatient and most inpatient situations where diuretic therapy is required, gentle but consistent fluid removal is recommended. In the outpatient setting, the goal of therapy should be to find the minimum dose of diuretic that consistently ensures a natriuretic response. Loop diuretics are typically the initial treatment of choice for patients who present with significant generalized oedema, even though DCT diuretics may also be effective, when oedema is mild. Patients with normal GFR who are naïve to the effects of a loop diuretic can develop a natriuresis with as little as 10 mg of furosemide per day. In contrast, those with CKD typically require a higher initial dose to achieve natriuresis (Brater, 1998). In either case, the clinician can inquire about the urine output within hours after taking an oral dose of loop diuretic to gauge the therapeutic threshold (Ellison and Wilcox, 2008). In addition, patients on diuretic therapy should weigh themselves on a daily basis. If the patient does not perceive a significant difference in urine output or if the patient’s weight does not change within a few days of starting diuretic therapy, it is unlikely that the prescribed diuretic dose is generating a negative fluid balance. The choice of oral diuretic is not dictated by strong evidence. Furosemide has been used as first-line treatment for many years. It is effective and inexpensive, but it does suffer from limited bioavailability; furthermore, the bioavailability varies between patients, between days, and with food (Table 33.3) (Brater, 1997). In contrast, both bumetanide and torsemide exhibit excellent and more consistent bioavailability. Bumetanide has a very short half-life,

311

312

Section 2  

fluid, electrolyte, and renal tubular disorders

whereas torsemide’s half-life is longer (Table 33.3). Based purely on pharmacokinetic parameters, torsemide’s diuretic profile should make it the most effective among the three; unfortunately, the data to support this speculation are limited. Nevertheless, a comprehensive review comparing the three primary loop diuretics noted that several studies suggested better outcomes with torsemide compared with furosemide (Wargo and Banta, 2009). They found less evidence supporting the superiority of one or other loop diuretic in other clinical conditions. The cost differential between the three diuretics has declined, since all are now available generically, but bumetanide remains more expensive than furosemide. Typical starting doses for HF are 20–40 mg twice daily for furosemide, 0.5–1.0 mg twice daily for bumetanide, and 5–10 mg daily for torsemide. For CKD, it is often necessary to start with a higher dose, for reasons discussed above.

Intravenous diuretics When a patient is hospitalized for oedema, it is often useful to use loop diuretics intravenously to obviate problems associated with limited bioavailability and guarantee efficacy. Typically, the loop diuretics are administered as bolus doses, but in some cases continuous infusions have been recommended. When given intravenously, differences in bioavailability become irrelevant, but important pharmacokinetic differences do persist. Importantly, the apparent half-life of furosemide becomes shorter, because ‘absorption limited kinetics’, resulting from the GI depot discussed above, is no longer relevant. Furthermore, as shown in Table 33.3, the half-life of bumetanide remains shorter, leaving more time for post-diuretic NaCl retention, when it is administered as a bolus; once again, torsemide has the longest half-life and the most favourable pharmacokinetic profile. Based on these differences, there seems little rational basis to switch from furosemide to bumetanide when given intravenously, although torsemide may still be more effective. Many smaller trials have suggested benefits of continuous infusions. In one prospective randomized crossover trial that studied modes of diuretic administration in patients with HF, continuous furosemide infusion preceded by a loading dose produced a greater diuresis and natriuresis than a 24-hour dose equivalent of furosemide given in boluses intermittently (Lahav et al., 1992). No significant differences in side effects were noted between the two groups. Similar findings were reported from a study of patients with CKD in which bumetanide was administered either by bolus or infusion (Rudy et al., 1991). In this case, side effects were also reduced by the continuous infusion. The effectiveness of continuous loop diuretic infusion (Table 33.5) likely results from the fact that a constant supply of diuretic is being maintained in the blood stream. This serves to clamp urinary diuretic levels at a concentration above the diuretic threshold, close to the concentration of maximal diuretic efficiency (Fig. 33.2). Moreover, continuous therapy has the benefit of minimizing the adaptive effect of post-diuretic NaCl retention, and should facilitate negative fluid balance more effectively (Ellison, 1997a, pp. 209–32). In contrast, a recent well-designed randomized controlled trial of furosemide in HF (Felker et al., 2011) compared bolus versus continuous infusion, and lower versus higher doses, of loop diuretic for patients hospitalized with acute decompensated HF. The results showed no significant difference in efficacy or safety endpoints for bolus versus continuous infusion. Patients assigned to intravenous bolus therapy were more likely to require an increased dose at

48 hours; as a result, the total dose of furosemide over 72 hours was higher in the bolus group compared with the continuous infusion group, a difference that was not quite significant (592 vs 480 mg, P = 0.06). In this study, the higher-dose furosemide regimen (2.5 × the daily home dose) produced greater net fluid loss, weight loss, and relief from dyspnoea, but also more frequent, though transient, worsening of renal function. There was an almost significant trend toward greater improvement in patients’ global assessment of symptoms in the high-dose group (P = 0.06); the mean change in the serum creatinine was < 0.1 mg/dL (9 micromol/L) in both groups. It should be emphasized that the continuous infusion protocol used in this study did not include a bolus diuretic dose at the beginning of treatment, as recommended by Brater and others (Brate, 1998; Ellison and Wilcox, 2008). In post hoc studies derived from the same dataset (Shah et  al., 2012), several additional insights emerged. Although the comparison between bolus and continuous infusion was neutral overall, when baseline diuretic dose was taken into consideration, there was an interaction effect. In this case, those who presented on a lower baseline diuretic dose responded more favourably to continuous infusion (in terms of weight loss), whereas those on a higher basal dose (> 120 mg furosemide) responded better to bolus administration. These differences are likely to result from the induction of the adaptive processes described above in those individuals whose home diuretic doses were higher. While these data suggest that the initial approach to diuretic treatment can include bolus diuretics, several caveats emerge. First, the HF trial did not examine patients who were truly resistant to loop diuretics, only those who presented to the hospital with decompensated HF. For those who fail initial bolus therapy, it may still be reasonable to try a continuous approach. Second, the trial did not examine patients with substantial CKD, a situation in which continuous infusions may be more effective, or at least safer. Third, the maximum dose, up to 600 mg/day, may not have achieved the maximal recommended dose during continuous infusion (960 mg/day or 40 mg/hour (Breater, 1998)), especially because the protocol did not necessarily deliver the maximal dose to all resistant individuals. When switching back from intravenous to oral furosemide, however, it is often recommended to double the dose, because the average bioavailability of furosemide is approximately 50%, but this is only a guideline, since the actual relative efficacy cannot be predicted. Clearly, since bumetanide and torsemide are more completely absorbed, their oral and intravenous doses should be closer, under most circumstances.

Combination diuretic therapy Diuretic resistance can often be treated with two classes of diuretic used simultaneously. Controlled trials (Chemtob et al., 1989) suggest little or no benefit from giving two agents of the same class (e.g. ethacrynic acid and furosemide). In contrast, adding a proximal tubule diuretic or a distal convoluted tubule diuretic to a loop diuretic is often dramatically effective. Distal convoluted tubule diuretics added to loop diuretics are synergistic (the combination is more effective than the sum of the effects of each drug alone) (Brater, 1985; Heller et al., 1985; Loon et al., 1989; Ellison, 1991; Knauf et al., 1995; Knauf and Mutschler, 1997). Distal convoluted tubule diuretics do not alter the pharmacokinetics or the bioavailability of loop diuretics. The addition

chapter 33 

of a distal convoluted tubule diuretic to a loop diuretic enhances NaCl excretion via several mechanisms (for a review, see Ellison, 1999). The most important mechanism is probably by inhibiting NaCl transport along the distal tubule, where tubular Na+ and Cl− uptake is stimulated by the loop diuretic. During prolonged loop diuretic use, distal nephron cells become hypertrophic and hyperplastic (Kaissling et  al., 1985; Kaissling and Stanto, 1988; Ellison et al., 1989) and there is an increase in the abundance of Na/K-ATPase pumps (Scherzer et al., 1987; Barlet-Bas et al., 1990), NCC (Abdallah et al., 2001), and the capacity to reabsorb Na+ and Cl− (Stanton and Kaisslin, 1988; Ellison et al., 1989). Thus, when microperfused with a standard NaCl load, distal tubules from animals treated chronically with loop diuretics reabsorb Na+ and Cl− more rapidly than tubules from control animals (Ellison et al., 1989). Because distal convoluted tubule diuretics inhibit NCC even under these stimulated conditions, the effect of these diuretics will be greatly magnified in patients in whom high doses of loop diuretics have led to hypertrophy and hyperplasia. Wilcox and co-workers (Loon et al., 1989) showed that the natriuretic effect of chlorothiazide in humans was enhanced following treatment with furosemide for 1 month, suggesting that daily oral furosemide treatment, even in modest doses, may induce adaptive distal changes. When a second class of diuretic is added, the dose of loop diuretic should not be altered, because the shape of the loop diuretic dose–response curve is not affected by addition of other classes of diuretic. Thus, the loop diuretic should be given in an effective or ceiling dose (Table 33.4). The choice of distal convoluted tubule diuretic is arbitrary. Many clinicians choose metolazone because its half-life is longer than some classic thiazide diuretics, but direct comparisons between metolazone and classic thiazides have shown little difference in natriuretic potency during combination use (Garin, 1987; Channer et al., 1994; Fliser et al., 1994). Distal convoluted tubule diuretics may be added in full doses (Table 33.6) when a rapid and robust response is needed, but this is likely to lead to complications and an extremely close follow-up is mandatory. Patients should be monitored closely when combination therapy is begun, because fluid and electrolyte depletion, sometimes massive, occurs commonly. Serious side effects have been noted frequently (Jentzer et al., 2010). One reasonable approach is to establish a therapeutic target weight and start with a low dose of DCT diuretic. The dose can then be escalated if necessary until the

clinical use of diuretics

clinical goals are achieved. When the target weight is attained, the distal convoluted tubule diuretic can often be prescribed only three times weekly or the dose adjusted to maintain the ECF volume at the desired level. Another approach to combination therapy may be a short fixed-dose course. Comparison was made of adding a thiazide-type diuretic to furosemide for either a fixed 3-day period or adjusting the dose to achieve targeted volume losses during 5–7 days. Both regimens were equally effective in reducing ECF volume and symptoms. Surprisingly, natriuresis and diuresis continued even after the thiazide-type diuretic was discontinued during the fixed regimen (Channer et al., 1994). For outpatients requiring combined therapy, one approach is to add a modest dose of distal convoluted tubule diuretic, such as 2.5–5 mg/day of metolazone, for 3  days only. Higher doses or longer time periods are effective, but may increase risk. Because distal convoluted tubule diuretics are absorbed more slowly than loop diuretics (peak levels at 1.5–4.0 hours for distal convoluted tubule diuretics compared with 0.5–2.0 hours for loop diuretics), it is sensible to recommend that the distal convoluted tubule diuretic be taken 0.5–1 hour prior to the loop diuretic, although this suggestion has not been tested. Cortical collecting duct diuretics, such as amiloride, spironolactone, or eplerenone, can be added to a regimen of loop diuretics, but their natriuretic effects are generally less dramatic than those of distal convoluted tubule diuretics (Levy, 1977; Ramsay et al., 1980). The combination of spironolactone and loop diuretics has not been shown to be synergistic, but can prevent hypokalaemia, while maintaining renal Na+ excretion. One situation in which cortical collecting duct diuretics may be preferred agents in combination is in patients with cirrhosis. A combination of furosemide and spironolactone or eplerenone is now considered the preferred regimen for cirrhotic ascites (Runyon, 2004), where some guidelines suggest maintaining a ratio of 40 mg furosemide/100 mg spironolactone. Potassium-sparing distal diuretics also reduce Mg2+ excretion, making hypomagnesaemia less likely than when combined with loop diuretics. Mineralocorticoid receptor blockade reduces mortality of patients with systolic dysfunction, whether severe (spironolactone) HF (173) or mild (eplerenone) (Zannad et al., 2011). While this effect has been attributed to direct cardiac or vascular effects (Pitt et  al., 1999; Rossignol et  al., 2011), renal effects are also

Table 33.4  Ceiling doses of loop diuretics Furosemide (mg)

Bumetanide (mg)

Torsemide (mg)

IV

PO

IV

PO

IV

PO

GFR 20–50 mL/min

80

80–160

2–3

2–3

50

50

GFR < 20 mL/min

200

240

8–10

8–10

100

100

Severe AKI

500

NA

12

NA

Nephrotic syndrome

120

240

3

3

50

50

Cirrhosis

40–80

80–160

1

1–2

10–20

10–20

Heart failure

40–80

160–240

2–3

2–3

20–50

50

Note: ceiling dose indicates the dose that produces the maximal increase in fractional sodium excretion. Larger doses may increase net daily natriuresis by increasing the duration of natriuresis without increasing the maximal rate. GFR = glomerular filtration rate.

313

314

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 33.5  Continuous diuretic infusion Diuretic

Loading dose

GFR < 25 mL/min

GFR 25–75 mL/min

GFR >75 mL/min

mg

mg/hour

mg/hour

mg/hour

Furosemide

40

20–40

10–20

10

Torsemide

20

10–20

5–10

5

Bumetanide

1

1–2

0.5–1

0.5

Data adapted from Brater, D. C. (1998). Diuretic therapy. N Engl J Med. 339, 387–395.

likely to participate. Barr et  al. (1995) randomized 42 patients with New York Heart Association class II–III congestive HF to either 50–100 mg/day of spironolactone or placebo added to a regimen of loop diuretics and ACEIs. Spironolactone increased Na+ excretion, urinary Na/K ratio, and serum Mg2+ concentration, and reduced ventricular arrhythmias. Others have reported similar results (Dehlström and Karlsson, 1993; Van Vliet et al., 1993). Nevertheless, hyperkalaemia is a concern when adding spironolactone to ACEI therapy, especially in those patients with renal insufficiency. In one study, potentially life-threatening hyperkalaemia during spironolactone treatment was found to be predicted by renal insufficiency, diabetes, older age, dehydration, and concomitant use of other medications that may cause hyperkalaemia (Schepkens et al., 2001). Yet results of some other large trials suggest that spironolactone or eplerenone is often tolerated with GFR>60 ml/min/1.73 m2 (Eschalier et al., 2013). Combination diuretic therapy is often indicated for hospitalized patients in an intensive care unit who need urgent diuresis in the setting of obligate fluid and solute loads. Two intravenous drugs are available to supplement loop diuretics: chlorothiazide (500–1000 mg once or twice daily) and acetazolamide (250–375 mg up to four times daily). Chlorothiazide has relatively potent carbonic anhydrase-inhibiting capacity in the proximal tubule and also blocks NCC in the distal convoluted tubule. It has a longer half-life than some other thiazides. Both chlorothiazide and acetazolamide can act synergistically with loop diuretics. Acetazolamide is especially useful when metabolic alkalosis complicates the treatment of oedema, since this may make it difficult to correct hypokalaemia or to wean a patient from a ventilator (Miller and Berns, 1977). The use of acetazolamide can correct alkalosis without the need to Table 33.6  Combination diuretic therapy (to add to a ceiling dose of a loop diuretic) Distal convoluted tubule diuretics: Metolazone 2.5–10 mg P.O. dailya Hydrochlorothiazide (or equivalent) 25–100 mg orally daily Chlorothiazide 500–1000 mg intravenously Proximal tubule diuretics: Acetazolamide 250–375 mg daily or up to 500 mg intravenously Collecting duct diuretics:Spironolactone 100–200 mg daily Amiloride 5–10 mg daily a Metolazone may be given for a limited period of time (3–5 days) or the response should be monitored very closely, once ECF volume has declined to the target level. Only in patients who remain volume expanded should full doses be continued indefinitely.

administer saline. In other situations, combination diuretic therapy may be targeted at the underlying disease process.

Ultrafiltration During the past 20 years, there has been ongoing interest in using mechanical processes to reduce ECF volume, when pharmacological therapy proves insufficient. Plasma ultrafiltration, with or without accompanying haemodialysis, may be used to remove extracellular fluid. Agostoni and colleagues (Agostoni et al., 1994; Marenzi et al., 2001) randomized patients with congestive HF to equal volume removal by ultrafiltration or furosemide. The extracellular fluid volume remained contracted following ultrafiltration, but rebounded to baseline after the intravenous diuretic treatment was discontinued. The extracellular fluid volume rebound following loop diuretic usage was associated with a brisk rise in plasma renin and angiotensin II levels. These observations led to the development of methods for fluid removal without the need for central lines. Positive outcomes were suggested by smaller studies (Costanzo et al., 2005, 2007), but a more recent large randomized controlled trial for patients with acute decompensated HF found that diuretic approaches were just as successful, with fewer adverse effects than ultrafiltration (Bart et al., 2012). Another recent retrospective analysis reached similar conclusions that ultrafiltration led to worsening renal function (Dev et al., 2012). Until contradictory information becomes available, therefore, ultrafiltration will generally be reserved for situations in which patients need dialysis, as well as fluid removal.

Diuretics in special situations Acute kidney injury Oliguric AKI generally implies that urine output is < 400–500 mL/24 hours. It is associated with a markedly worse prognosis than AKI without oliguria (Anderson et al., 1977). Loop diuretic therapy has been proposed to serve as a potential treatment for AKI, supported by studies suggesting that loop diuretics increase the degree of oxygenation of the renal medulla (Heyman et al., 1994). In addition, since volume overload is commonly an indication for dialysis in patients with AKI, it was thought that loop diuretics might improve outcomes by minimizing the number of patients that require acute dialysis. Finally, it was also proposed that in many cases, loop diuretics could increase urinary flow and wash casts out of the kidney tubules. Multiple small randomized controlled trials used loop diuretic therapy as an intervention to treat AKI (Brown et al., 1981; Allison and Shilliday, 1993). The results from these studies were negative; in each case, although loop diuretics were able to increase the urine

chapter 33 

output above the defined oliguric threshold, they did not improve patient mortality or reduce the need for dialysis. It is important to note however, that these trials were small and statistically underpowered. More recently, Mehta and colleagues (2002) conducted a large-scale multicenter retrospective analysis of the outcomes of all patients hospitalized in intensive care units with AKI who were seen in nephrology consultation over a 6-year period. In this study, diuretic treatment was associated with an increased risk of death and lack of recovery of renal function. Although these findings suggest that high-dose furosemide therapy might be harmful to patients with AKI, it is important to note that these observational studies are subject to confounding-by-indication. Indeed, a recent meta-analysis of nine acute renal failure trials encompassing 849 patients was unable to replicate the association between loop diuretic therapy and higher patient mortality (Ho and Sheridan, 2006). There has also been interest, however, in the effects of fluid balance on mortality and morbidity in intensive care units. This led to a study testing ‘conservative versus liberal’ approaches to fluid balance in the intensive care unit (the FACTT) (Wiedemann et  al., 2006). While this study did not assess AKI specifically, it provided the opportunity to test the effects of fluid strategies in which treatment approaches were determined randomly. A post hoc analysis of FACTT patients who developed AKI found substantially higher mortality in patients randomized to more liberal fluid administration. The conservative arm of this study involved much more aggressive use of diuretics, leading the authors to conclude that diuretic use in patients with AKI in the intensive care unit is not likely to be harmful, and may be beneficial. This suggestion warrants further evaluation (Schrier, 2009, 2010). In the meantime, it seems reasonable to consider the use of loop diuretics to maintain ECF volume in patients with AKI safe, and potentially beneficial. Their use simply to increase urine output, however, cannot be supported.

Cirrhosis A reasonable initial daily negative fluid balance in a cirrhotic patient with ascites should total approximately 250–750 mL/24 hours. Given the overactivity of the renin–angiotensin system in cirrhotic ascites, aldosterone receptor antagonists are the first-line diuretic of choice. Spironolactone is typically prescribed initially at a dose of up to 100 mg orally per day. If the patient does not appear to respond to aldosterone receptor antagonist monotherapy, a loop diuretic such as furosemide may be added, usually starting at 20–40 mg orally per day. The American Society for the Study of Liver Disease recommends that the ratio of furosemide to spironolactone be 40 mg/100 mg (Runyon, 2004). In the setting of tense ascites requiring large volume paracentesis, diuretic therapy may need to be adjusted to account for any fluid shifts that might occur following the bulk removal of peritoneal fluid. Aquaretic therapy may eventually be useful to facilitate a water diuresis in the cirrhotic patient with oedema and hyponatraemia, and studies are currently being conducted to confirm the safety and efficacy of this novel treatment modality.

Nephrotic syndrome Loop diuretics are the treatment of choice for oedema in the nephrotic syndrome, due to the fact that other diuretic classes are less capable of facilitating a clinically significant natriuretic effect. Massive proteinuria, the hallmark of the nephrotic syndrome,

clinical use of diuretics

diminishes loop diuretic efficacy. When Brater and colleagues measured the diuretic efficiency of furosemide in nephrotic rats, sodium reabsorption was decreased relative to urinary furosemide excretion, compared with non-nephrotic controls (Voelker et al., 1989). This finding illustrates that, compared with their efficacy in some oedematous disorders; loop diuretics are less capable of provoking a natriuresis in the nephrotic syndrome. The authors suggested that this observation may be due to the fact that a large fraction of the furosemide that enters the loop of Henle during diuretic therapy remains bound to albumin and is, therefore, unable to inhibit Na-K-2Cl cotransport. Yet work by the same group later showed that albumin binding to loop diuretics in the tubule lumen is not a major contributor to diuretic resistance:  agents that reduce diuretic binding to albumin had no substantial effect on diuretic efficacy in nephrotic patients (Agarwal et al., 2000). Hypoalbuminaemia also may act to diminish the effectiveness of loop diuretics. Once loop diuretics are absorbed into the bloodstream, they become largely bound to albumin. A  low serum albumin level diminishes the total blood concentration of loop diuretic and increases its volume of distribution. Therefore, the renal circulation will convey less diuretic to the nephron, and less will be extruded by basolateral-to-apical proximal tubule organic anion transport into the tubule lumen for delivery to the TAL. This scenario provides the rationale for infusing albumin together with loop diuretics to patients with substantial hypoalbuminaemia, a suggestion that has received some support in the literature (Mattana et al., 1996; Blendis and Wong, 1999; Fliser et al., 1999; Gentilini et al., 1999; Brater et al., 2001). Yet there is little evidence that such an approach is useful, if the serum albumin concentration exceeds 2 g/dL (Brater et al., 2001).

References Abdallah, J. G., Schrier, R. W., Edelstein, C., et al. (2001). Loop diuretic infusion increases thiazide-sensitive Na(+)/Cl(-)-cotransporter abundance: role of aldosterone. J Am Soc Nephrol, 12(7), 1335–41. Abe, K., Yasuima, M., Cheiba, L., et al. (1977). Effect of furosemide on urinary excretion of prostaglandin E in normal volunteers and patients with essential hypertension. Prostaglandins, 14, 513–21. Agarwal, R., Gorski, J. C., Sundblad, K., et al. (2000). Urinary protein binding does not affect response to furosemide in patients with nephrotic syndrome. J Am Soc Nephrol, 11(6), 1100–5. Agostoni, P., Marenzi, G., Lauri, G., et al. (1994). Sustained improvement in functional capacity after removal of body fluid with isolated ultrafilration in chronic cardiac insufficiency: failure of furosemide to provide the same result. Am J Med, 96, 191–9. ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group (2002b). Major outcomes in moderately hypercholesterolemic, hypertensive patients randomized to pravastatin vs usual care: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT-LLT). JAMA, 288(23), 2998–3007. ALLHAT Officers and Coordinators for the ALLHAT Collaborative Research Group (2002a). Major outcomes in high-risk hypertensive patients randomized to angiotensin-converting enzyme inhibitor or calcium channel blocker vs diuretic: the Antihypertensive and Lipid-Lowering Treatment to Prevent Heart Attack Trial (ALLHAT). JAMA, 288(23), 2981–97. Allison, M. E. M. and Shilliday, I. (1993). Loop diuretic therapy in acute and chronic renal failure. J Cardiovasc Pharmacol, 22 Suppl 3, S59–S70. Anderson, P., Boreus, L., Gordon, E., et al. (1988). Use of mannitol during neurosurgery: interpatient variability in the plasma and CSF levels. Eur J Clin Pharmacol, 35, 643–9. Anderson, R. J., Linas, S., Berns, A. S., et al. (1977). Nonoliguric acute renal failure. N Engl J Med, 296, 1134–8.

315

316

Section 2  

fluid, electrolyte, and renal tubular disorders

Antoniou, T., Gomes, T., Mamdani, M. M., et al. (2011a). Trimethoprim/ sulfamethoxazole-induced phenytoin toxicity in the elderly: a population-based study. Br J Clin Pharmacol, 71(4), 544–9. Antoniou, T., Gomes, T., Mamdani, M. M., et al. (2011b). Trimethoprim-sulfamethoxazole induced hyperkalaemia in elderly patients receiving spironolactone: nested case-control study. BMJ, 343,d5228. Ashraf, N., Locksley, R., and Arieff, A. (1981). Thiazide-induced hyponatremia associated with death or neurologic damage in outpatients. Am J Med, 70, 1163–8. Barlet-Bas, C., Khadouri, C., Marsy, S., et al. (1990). Enhanced intracellular sodium concentration in kidney cells recruits a latent pool of Na-K-ATPase whose size is modulated by corticosteroids. Proc Natl Acad Sci U S A, 265, 7799–803. Barr, C. S., Lang, C. C., Hanson, J., et al. (1995). Effects of adding spironolactone to an angiotensin-converting enzyme inhibitor in chronic congestive heart failure secondary to coronary artery disease. Am J Cardiol, 76, 1259–65. Bart, B. A., Goldsmith, S. R., Lee, K. L., et al. (2012). Ultrafiltration in decompensated heart failure with cardiorenal syndrome. N Engl J Med, 367(24), 2296–304. Batlle, D. C., von Riotte, A. B., Gaviria, M., et al. (1985). Amelioration of polyuria by amiloride in patients receiving long-term lithium therapy. N Engl J Med, 312(7), 408–14. Bedford, J. J., Leader, J. P., Jing, R., et al. (2008). Amiloride restores renal medullary osmolytes in lithium-induced nephrogenic diabetes insipidus. Am J Physiol Renal Physiol, 294(4), F812–20. Belge, H., Gailly, P., Schwaller, B., et al. (2007). Renal expression of parvalbumin is critical for NaCl handling and response to diuretics. Proc Natl Acad Sci U S A, 104(37), 14849–54. Benabe, J. E. and Martinez-Maldonado, M. (1986). Effects on divalent ion excretion. In G. Eknoyan and M. Martinez-Maldonado (eds.) The Physiological Basis of Diuretic Therapy in Clinical Medicine, pp. 109–24. Orlando, FL: Grune & Stratton, Inc. Better, O. S., Rubinstein, I., and Winaver, J. M., et al. (1997). Mannitol therapy revisited (1940-1997). Kidney Int, 51, 886–94. Bichet, D. G., Van Putten, V. J., and Schrier, R. W. (1982). Potential role of increased sympathetic activity in impaired sodium and water excretion in cirrhosis. N Engl J Med, 307, 1552–7. Blantz, R. C. (1974). Effect of mannitol on glomerular ultrafiltration in the hydropenic rat. J Clin Invest, 54, 1135–43. Blendis, L. and Wong, F. (1999). Intravenous albumin with diuretics: protean lessons to be learnt? J Hepatol, 30(4), 727–30. Botero-Velez, M., Curtis JJ, Warnock DG. (1994). Brief report: Liddle’s syndrome revisited – a disorder of sodium reabsorption in the distal tubule. N Engl J Med, 174, 178–81. Bourland, W. A., Day, D. K., and Williamson, H. E. (1977). The role of the kidney in the early nondiuretic action of furosemide to reduce elevated left atrial pressure in the hypervolemic dog. J Pharmacol Exp Ther, 202, 221–9. Brater, D. C. (1978). Effects of probenecid on furosemide response. Clin Pharmacol Ther, 24, 548–54. Brater, D. C. (1983). Pharmacodynamic considerations in the use of diuretics. Ann Rev Pharmacol Toxicol, 23, 45–62. Brater, D. C. (1997). Diuretic pharmacokinetics and pharmacodynamics. In D. W. Seldin and G. Giebisch (eds.) Diuretic Agents: Clinical Physiology and Pharmacology, pp. 189–208. San Diego, CA: Academic Press. Brater, D. C. (1998). Diuretic therapy. N Engl J Med, 339, 387–95. Brater, D. C., Chalasani, N., Gorski, J. C., et al. (2001). Effect of albumin-furosemide mixtures on response to furosemide in cirrhotic patients with ascites. Trans Am Clin Climatol Assoc, 112, 108–15. Brater, D. C., Pressley, R. H., and Anderson, S. A. (1985). Mechanisms of the synergistic combination of metolazone and bumetanide. J Pharmacol Exp Ther, 233, 70–3. Breslau, N. A. (1997). Use of diuretics in disorders of calcium metabolism. In D. W. Seldin and G. Giebisch (eds.) Diuretic Agents: Clinical Physiology and Pharmacology, pp. 495–512. San Diego, CA: Academic Press.

Brown, C. B., Ogg, C. S., and Cameron, J. S. (1981). High dose frusemide in acute renal failure: a controlled trial. Clin Nephrol, 15, 90–6. Brown, J. J., Davies, D. L., Ferriss, J. B., et al. (1972). Comparison of surgery and prolonged spironolactone therapy in patients with hypertension, aldosterone excess, and low plasma renin. Br Med J, 2, 729–0. Brown, N. J. (2003). Eplerenone: cardiovascular protection. Circulation, 107(19), 2512–8. Brown, T. C., Davis, J. O., and Johnston, C. I. (1966). Acute response in plasma renin and aldosterone secretion to diuretics. Am J Physiol, 211, 437–41. Buerkert, J., Martin, D., Prasad, J., et al. (1981). Role of deep nephrons and the terminal collecting duct in a mannitol-induced diuresis. Am J Physiol, 240, F411–F22. Bundy, J. T., Connito, D., Mahoney, M. D., et al. (1995). Treatment of idiopathic renal magesium wasting with amiloride. Am J Nephrol, 15, 75–7. Busch, A. E., Suessbrich, H., Kunzelmann, K., et al. (1996). Blockade of epithelial Na+ channels by triamterene-underlying mechanisms and molecular basis. Pflügers Arch, 432, 760–6. Calhoun, D. A., Jones, D., Textor, S., et al. (2008). Resistant hypertension: diagnosis, evaluation, and treatment: a scientific statement from the American Heart Association Professional Education Committee of the Council for High Blood Pressure Research. Circulation, 117(25), e510–26. Carlsen, J. E., Kober, L., Torp-Pedersen, C., et al. (1990). Relation between dose of bendrofluazide, antihypertensive effect, and adverse biochemical effects. BMJ, 300, 975–8. Carr, M. C., Prien, E. L., Jr., and Babayan, R. K. (1990). Triamterene nephrolithiasis: renewed attention is warranted. J Urol, 144, 1339–40. Cauley, J. A., Cummings, S. R., Seeley, D. G., et al. (1993). Effects of thiazide diuretic therapy on bone mass, fractures, and falls. Ann Intern Med, 118, 666–73. Cesar, K. R. and Magaldi, A. J. (1999). Thiazide induces water absorption in the inner medullary collecting duct of normal and Brattleboro rats. Am J Physiol, 277(5 Pt 2), F756–60. Channer, K. S., McLean, K. A., Lawson-Matthew, P., et al. (1994). Combination diuretic treatment in severe heart failure: a randomised controlled trial. Br Heart J, 71, 146–50. Chapagain, A. and Ashman, N. (2012). Hyperkalaemia in the age of aldosterone antagonism. QJM, 11, 1049–57. Chemtob, S., Doray, J. –L., Laudignon, N., et al. (1989). Alternating sequential dosing with furosemide and ethacrynic acid in drug tolerance in the newborn. Am J Dis Child, 143, 850–4. Chen, Z. F., Vaughn, D. A., Beaumont, K., et al. (1990). Effects of diuretic treatment and of dietary sodium on renal binding of 3H-metolazone. J Am Soc Nephrol, 1, 91–8. Cheng, C. J., Shiang, J. C., Hsu, Y. J., et al. (2007). Hypocalciuria in patients with Gitelman syndrome: role of blood volume. Am J Kidney Dis, 49(5), 693–700. Chow, K. M., Szeto, C. C., Wong, T. Y., et al. (2003). Risk factors for thiazide-induced hyponatraemia. QJM, 96(12), 911–17. Christensen, B. M., Zuber, A. M., Loffing, J., et al. (2011). AlphaENaC-mediated lithium absorption promotes nephrogenic diabetes insipidus. J Am Soc Nephrol, 22(2), 253–61. Cogan, M. G., Maddox, D. A., Warnock, D. G., et al. (1979). Effect of acetazolamide on bicarbonate reabsorption in the proximal tubule of the rat. Am J Physiol, 237, F447–F54. Coleman, R. A., Knepper, M. A., and Wade, J. B. (1997). Rat renal connecting cells express AQP2. J Am Soc Nephrol, 7, 16A. Coppage, W. S. and Liddle, G. W. (1960). Mode of action and clinical usefulness of aldosterone antagonists. Ann NY Acad Sci, 88, 815–20. Costanzo, L. S. and Weiner, I. M. (1976). Relationship between clerarances of Ca and Na: effect of distal diuretics and PTH. Am J Physiol, 230, 67–73. Costanzo, L.S. and Windhager, E. E. (1978). Calcium and sodium transport by the distal convoluted tubule of the rat. Am J Physiol, 235, F492–F506. Costanzo, M. R., Guglin, M. E., Saltzberg, M. T., et al. (2007). Ultrafiltration versus intravenous diuretics for patients hospitalized for acute decompensated heart failure. J Am Coll Cardiol, 49(6), 675–83.

chapter 33 

Costanzo, M. R., Saltzberg, M., O’Sullivan, J., et al. (2005). Early ultrafiltration in patients with decompensated heart failure and diuretic resistance. J Am Coll Cardiol, 46(11), 2047–51. Decaux, G., Soupart, A., and Vassart, G. (2008). Non-peptide arginine-vasopressin antagonists: the vaptans. Lancet, 371(9624), 1624–32. Decaux, G., Waterlot, Y., Genette, F., et al. (1981). Treatment of the syndrome of inappropriate antidiuretic hormone with furosemide. N Engl J Med, 304, 329–30. Dehlström, U. and Karlsson, E. (1993). Captopril and spironolactone therapy for refractory congestive heart failure. Am J Cardiol, 71, 29A–33A. Demartini, F. E., Wheaton, E. A., Healy, L. A., et al. (1962). Effect of chlorothiazide on the renal excretion of uric acid. Am J Med, 32, 572–7. Dev, S., Shirolkar, S. C., Stevens, S. R., et al. (2012). Reduction in body weight but worsening renal function with late ultrafiltration for treatment of acute decompensated heart failure. Cardiology, 123(3), 145–53. Dhalla, I. A., Gomes, T., Yao, Z., et al. (2013). Chlorthalidone versus hydrochlorothiazide for the treatment of hypertension in older adults: a population-based cohort study. Ann Internal Med, 158(6), 447–55. Dikshit K, Vyden JK, Forrester JS, et al. (1973). Renal and extrarenal hemodynamic effects of furosemide in congestive heart failure after acute myocardial infarction. N Engl J Med, 288, 1087–90. Dluhy, R. G., Wolf, G. L., and Lauler, D. P. (1970). Vasodilator properties of ethacrynic acide in the perfused dog kidney. Clin Sci, 38, 347–57. Dorman, H. R., Sondheimer, J. H., and Cadnapaphornchai P. (1990). Mannitol-induced acute renal failure. Medicine, 69, 153–9. Dorup, I., Skajaa, K., and Thybo, N. K. (1993). Oral magnesium supplementation restores the concentrations of magnesium, potassium and sodium-potassium pumps in skeletal muscle of patients receiving diuretic treatment. J Intern Med, 233, 117–23. Douban, S., Brodsky, M. A., and Whang, D. D. (1996). Significance of magnesium in congestive heart failure. Am Heart J, 132, 664–71. Duarte, C. G. (1968). Effects of chlorothiazide and amipramizide (MK 870) on the renal excretion of calcium, phosphate and magnesium. Metabolism, 17(5), 420–9. Duarte, C. G., Chomety, G., and Giebisch, G. (1971). Effect of amiloride, ouabain, and furosemide on distal tubular function in the rat. Am J Physiol, 221, 632–9. Dyckner, T. and Webster, P. O. (1982). Magnesium treatment of diuretic-induced hyponatremia with a preliminary report on a new aldosterone antagonist. J Am Coll Nutr, 1, 149–53. Dyckner, T., Wester, P. –O., and Widman, L. (1988). Amiloride prevents thiazide-induced intracellular potassium and magnesium losses. Acta Med Scand, 224, 25–30. Dzau, V. J. and Hollenberg, N. K. (1984). Renal response to captopril in severe heart failure: role of furosemide in natriuresis and reversal of hyponatremia. Ann Intern Med, 100, 777–82. Earley, L. E. and Friedler, R. M. (1964). Renal tubular effects of ethacrynic acid. J Clin Invest, 43, 1495. Earley, L. E. and Orloff, J. (1962). The mechanism of antidiuresis associated with the administration of hydrochlorothiazide to patients with vasopressin-resistant diabetes insipidus. J Clin Invest, 41, 1988–97. Eknoyan, G., Suki, W. N., and Martinez-Maldonado, M. (1970). Effect of diuretics on urinary excretion of phosphate, calcium, and magnesium in thyroparathyroidectomized dogs. J Lab Clin Med, 76, 257–66. Ellison, D. H. (1991). The physiologic basis of diuretic synergism: Its role in treating diuretic resistance. Ann Intern Med, 114, 886–94. Ellison, D. H. (1997a). Adaptation to Diuretic Drugs. Diuretic Agents: Clinical Physiology and Pharmacology. San Diego, CA: Academic Press. Ellison, D. H. (1997b). Intensive diuretic therapy: high doses, combinations, and constant infusions. In D. W. Seldin and G. Giebisch (eds.) Diuretic Agents: Clinical Physiology and Pharmacology, pp. 281–300. San Diego, CA: Academic Press. Ellison, D. H. (1999). Diuretic resistance: physiology and therapeutics. Semin Nephrol, 19(6), 581–97. Ellison, D. H. (2000). Divalent cation transport by the distal nephron: insights from Bartter’s and Gitelman’s syndromes. Am J Physiol Renal Physiol, 279(4), F616–25.

clinical use of diuretics

Ellison, D. H. and Loffing, J. (2009). Thiazide effects and adverse effects: insights from molecular genetics. Hypertension, 54(2), 196–202. Ellison, D. H., Velazquez, H., and Wright, F. S. (1989). Adaptation of the distal convoluted tubule of the rat. Structural and functional effects of dietary salt intake and chronic diuretic infusion. J Clin Invest, 83(1), 113–26. Ellison, D. H. and Wilcox, C. S. (2008). Diuretics. In B. M. Brenner (ed.) Brenner and Rector’s The Kidney, pp. 1646–78. Philadelphia, PA: Lippincott Williams & Wilkins. Eraly, S. A., Vallon, V., Vaughn, D. A., et al. (2006). Decreased renal organic anion secretion and plasma accumulation of endogenous organic anions in OAT1 knock-out mice. J Biol Chem, 281(8), 5072–83. Eschalier, R., McMurray, J. J., Swedberg, K., et al. (2013). Safety and efficacy of eplerenone in patients at high-risk for hyperkalemia and/or worsening renal function: analyses of EMPHASIS-HF study subgroups. J Am Coll Cardiol, 62(17), 1585–93. Ettinger, B., Citron, J. T., Livermore, B., et al. (1988). Chlorthalidone reduces calcium oxalate calculous recurrence but magnesium hydroxide does not. J Urol, 139(4), 679–84. Fairley, K. F., Woo, K. T., Birch, D. F., et al. (1986). Triamtrene-induced crystalluria and cylinduria: clinical and experimental studies. Clin Nephrol, 26, 169–73. Fanestil, D. D. (1988). Mechanism of action of aldosterone blockers. Sem Nephrol, 8, 249–63. Favre, L., Glasson, P., and Vallotton, M. B. (1982). Reversible acute renal failure from combined triamterene and indomethacin: a study in healthy subjects. Ann Intern Med, 96, 317–20. Felker, G. M., Lee, K. L., Bull, D. A., et al. (2011). Diuretic strategies in patients with acute decompensated heart failure. N Engl J Med, 364(9), 797–805. Felson, D. T., Sloutskis, D., Anderson, J. J., et al. (1991). Thiazide diuretics and the risk of hip fracture. Results from the Framingham study. JAMA, 265, 370–3. Fichman, M. P., Vorherr, H., Kleeman, C. R., et al. (1971). Diuretic-induced hyponatremia. Ann Intern Med, 75, 853–63. Fitch, G. K. and Weiss, M. L. (2000). Activation of renal afferent pathways following furosemide treatment. II. Effect of angiotensin blockade. Brain Res, 861(2), 377–89. Fitch, G. K., Patel, K. P., and Weiss, M. L. (2000). Activation of renal afferent pathways following furosemide treatment. I. Effects of survival time and renal denervation. Brain Res, 861(2), 363–76. Flack, J. M., Sica, D. A., and Nesbitt, S. (2011). Chlorthalidone versus hydrochlorothiazide as the preferred diuretic: is there a verdict yet? Hypertension, 57(4), 665–6. Fliser, D., Schröter, M., Neubeck, M., et al. (1994). Coadministration of thiazides increases the efficacy of loop diuretics even in patients with advanced renal failure. Kidney Int, 46, 482–8. Fliser, D., Zurbruggen, I., Mutschler, E., et al. (1999). Coadministration of albumin and furosemide in patients with the nephrotic syndrome. Kidney Int, 55(2), 629–34. Francis, G. S., Siegel, R. M., Goldsmith, S. R., et al. (1985). Acute vasoconstrictor response to intravenous furosemide in patients with chronic congestive heart failure. Ann Intern Med, 103, 1–6. Friedman, E., Shadel, M., Halkin, H., et al. (1989). Thiazide-induced hyponatremia: reproducibility by single dose rechallenge and an analysis of pathogenesis. Ann Intern Med, 110, 24–30. Friedman, P. A. and Gesek, F. A. (1995). Stimulation of calcium transport by amiloride in mouse distal convoluted tubule cells. Kidney Int, 48, 1427–34. Friedman, P. A. and Hebert, S. C. (1997). Site and Mechanism Of Diuretic Action. San Diego, CA: Academic Press. Frölich, J. C., Hollifield, J. W., Dormois, J. C., et al. (1976). Suppression of plasma renin activity by indomethacin in man. Circ Res, 39, 447–52. Gabow, P. A., Moore, S., and Schrier, R. W. (1979). Spironolactone-induced hyperchloremic acidosis in cirrhosis. Ann Intern Med, 90, 338–40. Gagnon, E., Bergeron, M. J., Brunet, G. M., et al. (2004). Molecular mechanisms of Cl- transport by the renal Na(+)-K(+)-Cl- cotransporter.

317

318

Section 2  

fluid, electrolyte, and renal tubular disorders

Identification of an intracellular locus that may form part of a high affinity Cl(-)-binding site. J Biol Chem, 279(7), 5648–54. Gamba, G. (2005). Molecular physiology and pathophysiology of electroneutral cation-chloride cotransporters. Physiol Rev, 85(2), 423–93. Ganguly, A. (1998). Primary aldosteronism. N Engl J Med, 339(25), 1828–34. Ganguly, A. and Weinberger, M. H. (1981). Triamteren-thiazide combination:alternative therapy for primary aldosteronism. Clin Pharmacol Ther, 30, 246–50. Garin, E. H. (1987). A comparison of combinations of diuretics in nephrotic edema. Am J Dis Child, 141, 769–71. Garty, H. (1994). Molecular properties of epithelial, amiloride-blockable Na+ channels. FASEB J, 8, 522–8. Garty, H. and Benos, D. (1988). Characteristics and regulatory mechanisms of the amiloride-blockable Na+ channel. Physiol Rev, 68, 309–73. Garty, H. and Palmer, L. G. (1997). Epithelial sodium channels: function, structure, and regulation. Physiol Rev, 77, 359–96. Gennari, F. J. and and Kassirer, J. P. (1974). Osmotic diuresis. N Engl J Med, 291(14), 714–20. Gentilini, P., Casini-Raggi, V., Di Fiore, G., et al. (1999). Albumin improves the response to diuretics in patients with cirrhosis and ascites: results of a randomized, controlled trial. J Hepatol, 30(4), 639–45. Gesek, F. A. and Friedman, P. A. (1992). Mechanism of calcium transport stimulated by chlorothiazide in mouse distal convoluted tubule cells. J Clin Invest, 90(2), 429–38. Giebisch, G. (1978). Amiloride effects on distal nephron function. In R. W. Straub and L. Bolis (eds.) Cell Membrane Receptors for Drugs and Hormones: A Multidisciplinary Approach, pp. 337–42. New York: Raven Press. Goldfarb, D. S., Chan, A. J., Hernandez, D., et al. (1991). Effect of thiazides on colonic NaCl absorption: role of carbonic anhydrase. Am J Physiol Renal Fluid Electrolyte Physiol, 261, F452–F8. Goldsmith, S. R., Francis, G., and Cohn, J. N. (1989). Attenuation of the pressor response to intravenous furosemide by angiotensin converting enzyme inhibition in congestive heart failure. Am J Cardiol, 64, 1382–5. Good, D. W. (1985). Sodium-dependent bicarbonate absorption by cortical thick ascending limb of rat kidney. Am J Physiol, 248, F821–F9. Good, D. W. and Wright, F. S. (1979). Luminal influences on potassium secretion: sodium concentration and fluid flow rate. Am J Physiol, 236, F192–F205. Good, D. W., Knepper, M. A., Burg, M. B. (1984). Ammonia and bicarbonate transport by thick ascending limb of rat kidney. Am J Physiol, 247, F35–F44. Greger, R. and Schlatter, E. (1983). Cellular mechanism of the action of loop diuretics on the thick ascending limb of Henle’s loop. Klinische Wochenschrift, 61, 1019–27. Griffing, G. T., Cole, A. G., Aurecchia, S. A., et al. (1982). Amiloride for primary hyperaldosteronism. Clin Pharmacol Ther, 31, 56–61. Grimm, R. H., Jr., Flack, J. M., andGranditis, G. A. (1996). Treatment of Mild Hypertension Study (TOMHS) Research Group. Long-term effects on plasma lipids of diet and drugs to treat hypertension. JAMA, 275, 1549–56. Gross, J. B. and Kokko, J. P. (1977). Effects of aldosterone and potassium-sparing diuretics on electrical potential differences across the distal nephron. J Clin Invest, 59, 82–9. Guan, Y., Chang, M., Cho, W., et al. (1997). Cloning, expression, and regulation of rabbit cyclooxygenase-2 in renal medullary interstitial cells. Am J Physiol, 273(1 Pt 2), F18–26. Hantman, D., Rossier, B., Zohlman, R., et al. (1973). Rapid correction of hyponatremia in the syndrom of inappropriate secretion of antidiuretic hormone: An alternative treatment to hypertonic saline. Ann Intern Med, 78, 870–5. Harding, P., Sigmon, D. H., Alfie, M. E., et al.(1997). Cyclooxygenase-2 mediates increased renal renin content induced by low-sodium diet. Hypertension, 29, 297–302. Harper R., Ennis, C. N., Heaney, A. P., et al. (1995). A comparison of the effects of low- and conventional-dose thiazide diuretic on insulin action in hypertensive patients with NIDDM. Diabetologia, 38, 853–9.

Harris, R. C., McKanna, J. A., Akai, Y., et al. (1994). Cyclooxygenase-2 is associated with the macula densa of rat kidney and increases with salt restriction. J Clin Invest, 94(6), 2504–10. Hasegawa, M., Kusuhara, H., Adachi, M., et al. (2007). Multidrug resistance-associated protein 4 is involved in the urinary excretion of hydrochlorothiazide and furosemide. J Am Soc Nephrol, 18(1), 37–45. Hashimoto, S., Huang, Y. G., Castrop, H., et al. (2004). Effect of carbonic anhydrase inhibition on GFR and renal hemodynamics in adenosine-1 receptor-deficient mice. Pflugers Arch, 448(6), 621–8. Hebert, S. C., Mount, D. B., and Gamba, G. (2004). Molecular physiology of cation-coupled Cl- cotransport: the SLC12 family. Pflugers Arch, 447(5), 580–93. Heidrich, F. E., Stergachis, A., and Gross, K. M. (1991). Diuretic drug use and the risk for hip fracture. Ann Intern Med, 115, 1–6. Helderman, J. H., Elahi, D., Andersen, D. K., et al. (1983). Prevention of the glucose intolerance of thiazide diuretics by maintenance of body potassium. Diabetes, 32(2), 106–11. Heller, I., Halevy, J., Cohen, J., et al. (1985). Significant metabolic acidosis induced by acetazolamide. Not a rare complication. Arch Intern Med, 145, 1815–17. Heyman, S. N., Rosen, S., Epstein, F. H., et al. (1994). Loop diuretics reduce hypoxic damage to proximal tubules of the isolated perfused rat kidney. Kidney Int, 45(4), 981–5. Ho, K. M. and Sheridan, D. J. (2006). Meta-analysis of frusemide to prevent or treat acute renal failure. BMJ, 333(7565), 420. Hollifield, J. W. (1989). Thiazide treatment of systemic hypertension: Effects on serum magnesium and ventricular ectopy. Am J Cardiol, 63, 22G–5G. Hook, J. B., Blatt, A. H., Brody, M. J., et al. (1966). Effects of several saluretic-diuretic agents on renal hemodynamics. J Pharmacol Exp Ther, 154, 667–73. Hropot, M., Fowler, N. B., Karlmark, B., et al. (1985). Tubular action of diuretics: distal effects on electrolyte transport and acidification. Kidney Int, 28, 477–89. Huang, C. L. and Kuo, E. (2007). Mechanism of hypokalemia in magnesium deficiency. J Am Soc Nephrol, 18(10), 2649–52. Huerta, C., Castellsague, J., Varas-Lorenzo, C., et al. (2005). Nonsteroidal anti-inflammatory drugs and risk of ARF in the general population. Am J Kidney Dis, 45(3), 531–9. Ikeda, K., Oshima, T., Hidaka, H., et al. (1997). Molecular and clinical implications of loop diuretic ototoxicity. Hear Res, 107, 1–8. Inoue, M., Okajima, K., Itoh, K., et al. (1987). Mechanism of furosemide resistance in analbuminemic rats and hypoalbuminemic patients. Kidney Int, 32, 198–203. Janjua, N. R., Jonassen, T. E., Langhoff, S., et al. (2001). Role of sodium depletion in acute antidiuretic effect of bendroflumethiazide in rats with nephrogenic diabetes insipidus. J Pharmacol Exp Ther, 299(1), 307–13. Jentzer, J. C., DeWald, T. A., and Hernandez, A. F. (2010). Combination of loop diuretics with thiazide-type diuretics in heart failure. J Am Coll Cardiol, 56(19), 1527–34. Johnston, G. D., Nicholls, D. P., and Leahey, W.J. (1984). The dose-response characteristics of the acute non-diuretic peripheral vascular effects of frusemide in normal subjects. Br J Clin Pharmacol, 18, 75–81. Johnston, P. A., Bernard, D. B., Perrin, N. S., et al. (1981). Prostaglandins mediate the vasodilatory effect of mannitol in the hypoperfused rat kidney. J Clin Invest, 68, 127–33. Juurlink, D. N., Mamdani, M. M., Lee, D. S., et al. (2004). Rates of hyperkalemia after publication of the Randomized Aldactone Evaluation Study. N Engl J Med, 351(6), 543–51. Kagawa, C. M. (1960). Blocking the renal electrolyte effects of mineralocorticoids with an orally active steroidal spirolactone. Endocrinology, 65, 125–32. Kaissling, B. and Stanton, B. A. (1988). Adaptation of distal tubule and collecting duct to increased sodium delivery. I. Ultrastructure. Am J Physiol, 255, F1256–F68. Kaissling, B., Bachmann, S., and Kriz, W. (1985). Structural adaptation of the distal convoluted tubule to prolonged furosemide treatment. Am J Physiol, 248, F374–F81.

chapter 33 

Kammerl, M. C., Nusing, R. M., Richthammer, W., et al.(2001). Inhibition of COX-2 counteracts the effects of diuretics in rats. Kidney Int, 60(5), 1684–91. Kammerl, M. C., Nusing, R. M., Schweda, F., et al. (2001). Low sodium and furosemide-induced stimulation of the renin system in man is mediated by cyclooxygenase 2. Clin Pharmacol Ther, 70(5), 468–74. Kaojarern, S., Day, B., and Brater, D. C. (1982). The time course of delivery of furosemide into urine: an independent determinant of overall response. Kidney Int, 22, 69–74. Karim, A. (1978). Spironolactone: disposition, metabolism, pharmacodynamics and bioavailability. Drug Metab Rev, 8, 151–88. Kaufman, A. M. and Levitt, M. F. (1985). The effect of diuretics on systemic and renal hemodynamics in patients with renal insufficiency. Am J Kidney Dis, 5, A71–A8. Kazama, I., Hatano, R., Michimata, M., et al. (2005). BSC1 inhibition complements effects of vasopressin V receptor antagonist on hyponatremia in SIADH rats. Kidney Int, 67(5), 1855–67. Khan, K. N., Venturini, C. M., Bunch, R. T., et al. (1998). Interspecies differences in renal localization of cyclooxygenase isoforms: implications in nonsteroidal antiinflammatory drug-related nephrotoxicity. Toxicol Pathol, 26(5), 612–20. Kim, G. H., Lee, J. W., Oh, Y. K., et al. (2004). Antidiuretic effect of hydrochlorothiazide in lithium-induced nephrogenic diabetes insipidus is associated with upregulation of aquaporin-2, Na-Cl co-transporter, and epithelial sodium channel. J Am Soc Nephrol, 15(11), 2836–43. Kirchner, K. A. (1985). Prostaglandin inhibitors alter loop segment chloride uptake during furosemide diuresis. Am J Physiol, 248, F698–F704. Kirchner, K. A., Martin, C. J., and Bower, J. D. (1986). Prostaglandin E2 but not I2 restores furosemide response in indomethacin-treated rats. Am J Physiol, 250, F980–F985. Knauf, H. and Mutschler, E. (1995). Diuretic effectiveness of hydrochlorothiazide and furosemide alone and in combination in chronic renal failure. J Cardiovasc Pharmacol, 26, 394–400. Knauf, H. and Mutschler, E. (1997). Sequential nephron blockade breaks resistance to diuretics in edematous states. J Cardiovasc Pharmacol, 29(3), 367–72. Koeppen, B. M., Biagi, B. A., and Giebisch, G. (1983). Intracellular microelectrode charcterization of the rabbit cortical collecting duct. Am J Physiol, 244, F35–F47. Komhoff, M., Jeck, N. D., Seyberth, H. W., et al. (2000). Cyclooxygenase-2 expression is associated with the renal macula densa of patients with bartter-like syndrome. Kidney Int, 58(6), 2420–4. Kortenoeven, M. L., Li, Y., Shaw, S., et al. (2009). Amiloride blocks lithium entry through the sodium channel thereby attenuating the resultant nephrogenic diabetes insipidus. Kidney Int, 76(1), 44–53. Kunau, R. T., Jr., Weller, D. R., and Webb, H. L. (1975). Clarification of the site of action of chlorothiazide in the rat nephron. J Clin Invest, 56(2), 401–7. Laerum, E. and Larsen, S. (1984). Thiazide prophylaxis of urolithiasis. A double-blind study in general practice. Acta Med Scand, 215(4), 383–9. Lahav, M., Regev, A., Ra’anani, P., et al. (1992). Intermittent administration of furosemide vs continuous infusion preceded by a loading dose for congestive heart failure. Chest, 102, 725–31. Lee, C. T., Shang, S., Lai, L. W., et al. (2004). Effect of thiazide on renal gene expression of apical calcium channels and calbindins. Am J Physiol Renal Physiol, 287(6), F1164–70. Leung, A. A., Wright, A., Pazo, V., et al. (2011). Risk of thiazide-induced hyponatremia in patients with hypertension. Am J Med, 124(11), 1064–72. Leviel, F., Hubner, C. A., Houillier, P., et al. (2010). The Na+-dependent chloride-bicarbonate exchanger SLC4A8 mediates an electroneutral Na+ reabsorption process in the renal cortical collecting ducts of mice. J Clin Invest, 120(5), 1627–35. Levy, B. (1977). The efficacy and safety of furosemide and a combination of spironolactone and hydrochlorothiazide in congestive heart failure. J Clin Pharmacol, 17(7), 420–30.

clinical use of diuretics

Liamis, G., Rodenburg, E. M., Hofman, A., et al. (2013). Electrolyte disorders in community subjects: prevalence and risk factors. Am J Med, 126(3), 256–63. Liddle, G. W. (1966). Aldosterone antagonists and triamterene. Ann NY Acad Sci, 134, 466–9. Loffing, J., Loffing-Cueni, D., Hegyi, I., et al. (1996). Thiazide treatment of rats provokes apoptosis in distal tubule cells. Kidney Int, 50(4), 1180–90. Loffing, J., Vallon, V., Loffing-Cueni, D., et al. (2004). Altered renal distal tubule structure and renal Na(+) and Ca(2+) handling in a mouse model for Gitelman’s syndrome. J Am Soc Nephrol, 15(9), 2276–88. Loon, N. R., Wilcox, C. S., and Unwin, R. J. (1989). Mechanism of impaired natriuretic response to furosemide during prolonged therapy. Kidney Int, 36, 682–9. Lorenz, J. N., Schultheis, P. J., Traynor, T., et al. (1999). Micropuncture analysis of single-nephron function in NHE3-deficient mice. Am J Physiol, 277(3 Pt 2), F447–53. Ludens, J. H., Hook, J. B., Brody, M. J., et al. (1968). Enhancement of renal blood flow by furosemide. J Pharmacol Exp Ther, 163, 456–60. Maher, J. F. and Schreiner, G. F. (1965). Studies on ethacrynic acid in patients with refractory edema. Ann Intern Med, 62, 15–29. Malnic, G., Klose, R. M., and Giebisch, G. (1966). Microperfusion study of distal tubular potassium and sodium transfer in rat kidney. Am J Physiol, 211, 548–59. Marenzi, G., Lauri, G., Grazi, M., et al. (2001). Circulatory response to fluid overload removal by extracorporeal ultrafiltration in refractory congestive heart failure. J Am Coll Cardiol, 38(4), 963–8. Markowitz, M., Messineo, F., and Coplan, N. L. (2012). Aldosterone receptor antagonists in cardiovascular disease: a review of the recent literature and insight into potential future indications. Clin Cardiol, 35(10), 605–9. Mattana, J., Patel, A., Ilunga, C., et al. (1996). Furosemide-albumin complexes in refractory nephrotic syndrome and chronic renal failure [letter]. Nephron, 73. 73(1.1), 122–3. McCormick, J. A., Mutig, K., Nelson, J. H., et al. (2011). A SPAK isoform switch modulates renal salt transport and blood pressure. Cell Metab, 14, 352–64. McGuffin, W. L., Jr. and Gunnells, J. C. (1969). Intravenously administered chlorothiazide in diagnostic evaluation of hypertensive disease. Arch Intern Med, 123, 124–30. Mehta, R. L., Pascual, M. T., Soroko, S, et al. (2002). Diuretics, mortality, and nonrecovery of renal function in acute renal failure. JAMA, 288(20), 2547–53. Menard, J. (2004). The 45-year story of the development of an anti-aldosterone more specific than spironolactone. Mol Cell Endocrinol, 217(1–2), 45–52. Miller, P.D. and Berns, A. S. (1977). Acute metabolic alkalosis perpetuating hypercarbia: a role for acetazolamide in chronic obstructive pulmonary disease. JAMA, 238, 2400–1. Miyanoshita, A., Terada, M., and Endou, H. (1989). Furosemide directly stimulates prostaglandin E2 production in the thick ascending limb of Henle’s loop. J Pharmacol Exp Ther, 251, 1155–9. Mizuta, K., Adachi, M., and Iwasa, K. H. (1997). Ultrastructural localization of the Na-K-Cl cotransporter in the lateral wall of the rabbit cochlear duct. Hear Res, 106(1–2), 154–62. Monroy, A., Plata, C., Hebert, S. C., et al. (2000). Characterization of the thiazide-sensitive Na(+)-Cl(-) cotransporter: a new model for ions and diuretics interaction. Am J Physiol Renal Physiol, 279(1), F161–9. Moreno, E., San Cristobal, P., Rivera, M., et al. (2006). Affinity defining domains in the Na-Cl cotransporter: different location for Cl- and thiazide binding. J Biol Chem, 281(25), 17266–75. Mukherjee, S. K., Katz, M. A., Michael, U. F., et al. (1981). Mechanisms of hemodynamic actions of furosemide: differentiation of vascular and renal effects on blood pressure in functionally anephric hypertensive patients. Am Heart J, 101, 313–18. Murray, M. D., Deer, M. M., Ferguson, J. A., et al. (2001). Open-label randomized trial of torsemide compared with furosemide therapy for patients with heart failure. Am J Med, 111(7), 513–20.

319

320

Section 2  

fluid, electrolyte, and renal tubular disorders

Na, K. Y., Kim, G. H., Joo, K. W., et al. (2007). Chronic furosemide or hydrochlorothiazide administration increases H+-ATPase B1 subunit abundance in rat kidney. Am J Physiol Renal Physiol, 292(6), F1701–9. Neafsey, P. J. (2004). Thiazides and selective serotonin reuptake inhibitors can induce hyponatremia. Home Healthc Nurse, 22(11), 788–90. Nijenhuis, T., Hoenderop, J. G., and Bindels, R. J. (2004). Downregulation of Ca(2+) and Mg(2+) transport proteins in the kidney explains tacrolimus (FK506)-induced hypercalciuria and hypomagnesemia. J Am Soc Nephrol, 15(3), 549–57. Nijenhuis, T., Vallon, V., van der Kemp, A. W., et al. (2005). Enhanced passive Ca2+ reabsorption and reduced Mg2+ channel abundance explains thiazide-induced hypocalciuria and hypomagnesemia. J Clin Invest, 115(6), 1651–8. Nochy, D., Callard, P., Bellon, B., et al. (1976). Association of overt glomerulonephritis and liver disease: a study of 34 patients. Clin Nephrol, 6, 422–7. O’Neil, R. G. and Helman, S. I. (1977). Transport characteristics of renal collecting tubules: influences of DOCA and diet. Am J Physiol, 233, F544–F58. O’Neil, R. G. and Sansom, S. C. (1984). Characterization of apical cell membrane Na+ and K+ conductances of cortical collecting duct using microelectrode techniques. Am J Physiol, 247, F14–F24. Oberleithner, H., Lang, F., Messner, G., et al. (1984). Mechanism of hydrogen ion transport in the diluting segment of frog kidney. Pflügers Arch, 402, 272–80. Okusa, M. D., Erik, A., Persson, G., et al. (1989). Chlorothiazide effect on feedback-mediated control of glomerular filtration rate. Am J Physiol, 257, F137–F44. Okusa, M. D., Unwin, R. J., Velázquez, H., et al. (1992). Active potassium absorption by the renal distal tubule. Am J Physiol Renal, Fluid Electrolyte Physiol, 262, F488–F93. Okusa, M. D., Velazquez, H., Ellison, D. H., et al. (1990). Luminal calcium regulates potassium transport by the renal distal tubule. Am J Physiol, 258(2 Pt 2), F423–8. Okusa, M.D. and Bia, M. J. (1987). Bartter’s syndrome. In P. P. Foa and M. P. Cohen (eds.) Endocrinology and Metabolism, pp. 231–63. New York: Springer-Verlag. Okusa, M.D. and Ellison, D. H. (2008). Physiology and pathophysiology of diuretic action. In R. J. Alpern and S. C. Hebert (eds.) The Kidney: Physiology and Pathophysiology (4th ed.), pp. 1051–94. Amsterdam: Elsevier. Packer, M. (1989). Identification of risk factors predisposing to the development of functional renal insufficiency during treatment with converting-enzyme inhibitors in chronic heart failure. Cardiology, 76 (Suppl 2), 50–5. Packer, M., Lee, W. H., Medina, N., et al. (1987). Functional renal insufficiencey during long-term therapy with captopril and enalapril in severe congestive heart failure. Ann Intern Med, 106, 346–54. Palmer, B. F. (1997). Potassium disturbances associated with the use of diuretics. In D. W. Seldin and G. Giebisch (eds.) Diuretic Agents: Clinical Physiology and Pharmacology, pp. 571–83. San Diego, CA: Academic Press. Pastorekova, S., Parkkila, S., Pastorek, J., et al. (2004). Carbonic anhydrases: current state of the art, therapeutic applications and future prospects. J Enzyme Inhib Med Chem, 19(3), 199–229. Pickkers, P., Schachter, M., Hughes, A. D., et al. (1996). Thiazide-induced hyperglycaemia: a role for calcium-activated potassium channels? Diabetologia, 39, 861–4. Pitt, B., Williams, G., Remme, W., et al. (2001). The EPHESUS trial: eplerenone in patients with heart failure due to systolic dysfunction complicating acute myocardial infarction. Eplerenone Post-AMI Heart Failure Efficacy and Survival Study. Cardiovasc Drugs Ther, 15(1), 79–87. Pitt, B., Zannad, F., Remme, W. J., et al. (1999). The effect of spironolactone on morbidity and mortality in patients with severe heart failure. Randomized Aldactone Evaluation Study Investigators. N Engl J Med, 341(10), 709–17. Pockros, P. J. and Reynolds, T. B. (1986). Rapid diuresis in patients with ascites from chronic liver disease: the importance of peripheral edema. Gastroenterology, 90(6), 1827–33.

Popovtzer, M. M., Subryan, V. L., Alfrey, A. C., et al. (1975). The acute effect of chlorothiazide on serum-ionized calcium. Evidence for a parathyroid hormone-dependent mechanism. J Clin Invest, 55, 1295–302. Preminger, G. M. and Pak, C. Y. C. (1987). Eventual attenuation of hypocalciuric response to hydrochlorothiazide in absorptive hypercalciuria. J Urol, 137, 1104–9. Puschett, J. B. and Goldberg, M. (1968). The acute effects of furosemide on acid and electrolyte excretion in man. J Lab Clin Med, 71, 666–77. Quamme, G. A. (1997). Renal magnesium handling: New insights in undertsanding old problems. Kidney Int, 52, 1180–95. Quamme, G. A., Wong, N. L., Sutton, R. A., et al. (1975). Interrelationship of chlorothiazide and parathyroid hormone: a micropuncture study. Am J Physiol, 229(1), 200–5. Quentin, F., Chambrey, R., Trinh-Trang-Tan, M. M., et al. (2004). The Cl-/ HCO3- exchanger pendrin in the rat kidney is regulated in response to chronic alterations in chloride balance. Am J Physiol Renal Physiol, 287(6), F1179–88. Ram, C. V., Garrett, B. N., and Kaplan, N. M. (1981). Moderate sodium restriction and various diuretics in the treatment of hypertension. Arch Intern Med, 141(8), 1015–19. Ramsay, L. E., Silas, J. H., and Freestone, S. (1980). Diuretic treatment of resistant hypertension. Br Med J, 281(6248), 1101–3. Ray, W. A., Griffin, M. R., Downey, W., et al. (1989). Long-term use of thiazide diuretics and risk of hip fracture. Lancet, 1, 687–90. Reid, I. R., Ames, R. W., Orr-Walker, B. J., et al. (2000). Hydrochlorothiazide reduces loss of cortical bone in normal postmenopausal women: a randomized controlled trial. Am J Med, 109(5), 362–70. Reilly, R. F. and Huang, C. L. (2011). The mechanism of hypocalciuria with NaCl cotransporter inhibition. Nat Rev Nephrol, 7(11), 669–74. Reilly, R. F., Peixoto, A. J., and Desir, G. V. (2010). The evidence-based use of thiazide diuretics in hypertension and nephrolithiasis. Clin J Am Soc Nephrol, 5(10), 1893–903. Rose, H. J., Pruitt, A. W., and McNay, J. L. (1976). Effect of experimental azotemia on renal clearance of furosemide in the dog. J Pharmacol Exp Ther, 196, 238–47. Rose, L. I., Underwood, R. H., Newmark, S. R., et al. (1977). Pathophysiology of spironolactone-induced gynecomastia. Ann Intern Med, 87, 398–403. Rossignol, P., Menard, J., Fay, R., et al. (2011). Eplerenone survival benefits in heart failure patients post-myocardial infarction are independent from its diuretic and potassium-sparing effects. Insights from an EPHESUS (Eplerenone Post-Acute Myocardial Infarction Heart Failure Efficacy and Survival Study) substudy. J Am Coll Cardiol, 58(19), 1958–66. Rude, R. K. (1989). Physiology of magnesium metabolism and the important role of magnesium in potassium deficiency. Am J Cardiol, 63, 31G–4G. Rudy, D. W., Voelker, J. R., Greene, P. K., et al. (1991). Loop diuretics for chronic renal insufficiency: a continuous infusion is more efficacious than bolus therapy. Ann Intern Med, 115, 360–6. Runyon, B. A. (2004). Practice Guidelines Committee AAftSoLD. Management of adult patients with ascites due to cirrhosis. Hepatology, 39(3), 841–56. Ryback, L. P. (1993). Ototoxicity of loop diuretics. Otolaryngol Clin N Am, 26, 829–44. Scheen, A. J. (2004). Renin-angiotensin system inhibition prevents type 2 diabetes mellitus. Part 2. Overview of physiological and biochemical mechanisms. Diabetes Metab, 30(6), 498–505. Schepkens, H., Vanholder, R., Billiouw, J. M., et al. (2001). Life-threatening hyperkalemia during combined therapy with angiotensin- converting enzyme inhibitors and spironolactone: an analysis of 25 cases. Am J Med, 110(6), 438–41. Scherzer, P., Wald, H., and Popovtzer, M. M. (1987). Enhanced glomerular filtration and Na+-K+-ATPase with furosemide administration. Am J Physiol, 252, F910–F915. Schlingmann, K. P., Weber, S., Peters, M., et al. (2002). Hypomagnesemia with secondary hypocalcemia is caused by mutations in TRPM6, a new member of the TRPM gene family. Nat Genet, 31(2), 166–70.

chapter 33 

Schmieder, R. E., Messerli, F. H., deCarvalho, J. G. R., et al. (1987). Immediate hemodynamic response to furosemide in patients undergoing chronic hemodialysis. Am J Kidney Dis, 9, 55–9. Schnermann, J. and Briggs, J. P. (2008). Tubuloglomerular feedback: mechanistic insights from gene-manipulated mice. Kidney Int, 74(4), 418–26. Schnermann, J. and Briggs, J. P. (2012). Synthesis and secretion of renin in mice with induced genetic mutations. Kidney Int, 81, 529–38. Schrier, R. W. (1978). New treatments for hyponatremia. N Engl J Med, 298, 214–15. Schrier, R. W. (2009). AKI: fluid overload and mortality. Nat Rev Nephrol, 5(9),485. Schrier, R. W. (2010). Fluid administration in critically ill patients with acute kidney injury. Clin J Am Soc Nephrol, 5(4), 733–9. Seely, J. F. and Dirks, J. H. (1969). Micropuncture study of hypertonic mannitol diuresis in the proximal and distal tubule of the dog kidney. J Clin Invest, 48, 2330–9. Seyberth, H.W. and Schlingmann, K. P. (2011). Bartter- and Gitelman-like syndromes: salt-losing tubulopathies with loop or DCT defects. Pediatr Nephrol, 26(10), 1789–802. Shackleton, C. R., Wong, N. L. M., and Sutton, R. A. (1986). Distal (potassium-sparing) diuretics. In J. H. Dirks and R. A. L. Sutton (eds.) Diuretics: Physiology, Pharmacology and Clinical Use (1st ed.), pp. 117–34. Philadelphia, PA: W. B. Saunders. Shah, R. V., McNulty, S., O’Connor, C. M., et al. (2012). Effect of admission oral diuretic dose on response to continuous versus bolus intravenous diuretics in acute heart failure: an analysis from diuretic optimization strategies in acute heart failure. Am Heart J, 164(6), 862–8. Shankar, S. S. and Brater, D. C. (2003). Loop diuretics: from the Na-K-2Cl transporter to clinical use. Am J Physiol Renal Physiol, 284(1), F11–21. Shimizu, K. (2003). Combined effects of vasopressin V2 receptor antagonist and loop diuretic in humans. Clin Nephrol, 59(3), 164–73. Shuichi, T. and Schwartz, G. J. (1998). HCO3- absorption in rabbit outer medullary collecting duct: role of luminal carbonic anhydrase. Am J Physiol, 274, F139–F47. Skott, P., Hommel, E., Bruun, N. E., et al. (1989). The acute effect of acetazolamide on glomerular filtration rate and proximal tubular reabsorption of sodium and water in normal man. Scand J Clin Lab Invest, 49(6), 583–7. Smith, W. E. and Steele, T. H. (1983). Avoiding diuretic related complications in older patients. Geriatrics, 38, 117–19. Soleimani, M. and Aronson, P. S. (1989). Ionic mechanism of Na+-HCO3cotransport in rabbit renal basolateral membrane vesicles. J Biol Chem, 264, 18302–8. Soleimani, M., Barone, S., Xu, J., et al. (2012). Double knockout of pendrin and Na-Cl cotransporter (NCC) causes severe salt wasting, volume depletion, and renal failure. Proc Natl Acad Sci U S A, 109(33), 13368–73. Soleimani, M., Grassl, S. M., and Aronson, P. S. (1987). Stoichiometry of Na+-HCO3- cotransport in basolateral membrane vesicles isolated from rabbit renal cortex. J Clin Invest, 79, 1276–80. Somogyi, A. A., Hovens, C. M., Muirhead, M. R., et al. (1989). Renal tubular secretion of amiloride and its inhibition by cimetidine in humans and in an animal model. Drug Metab Dispos, 17(2), 190–6. Stafstrom, C. E., Gilmore, H. E., and Kurtin, P. S. (1992). Nephrocalcinosis complicating medical treatment of posthemorrhagic hydrocephalus. Pediatr Neurol, 8, 179–82. Stanton, B. A. and Kaissling, B. (1988). Adaptation of distal tubule and collecting duct to increased sodium delivery. II. Na+ and K+ transport. Am J Physiol, 255, F1269–F75. Star, R. A. (1997). Ototoxicity. San Diego, CA: Academic Press. Stone, D. K., Seldin, D. W., Kokko, J. P., et al. (1983). Mineralocorticoid modulation of rabbit medullary collecting duct acidification. J Clin Invest, 72, 77–83. Suki, W., Rector, F. C., Jr., and Seldin, D. W. (1965). The site of action of furosemide and other sulfonamide diuretics in the dog. J Clin Invest, 44, 1458–69.

clinical use of diuretics

Sungaila, I., Bartle, W. R., Walker, S. E., et al. (1992). Spironolactone pharmacokinetics and pharmacodynamics in patients with cirrhotic ascites. Gastroenterology, 102, 1680–5. Tannen, R. L. (1970). The effect of uncomplicated potassium depletion on urine acidification. J Clin Invest, 49, 813–27. Terada, Y. and Knepper, M. A. (1990). Thiazide-sensitive NaCl absorption in rat cortical collecting duct. Am J Physiol Renal Fluid Electrolyte Physiol, 259, F519–F28. The Randomized Aldactone Evaluation Study (1996). Effectiveness of spironolactone added to an angiotensin-converting enzyme inhibitor and a loop diuretic for severe chronic congestive heart failure (The Randomized Aldactone Evaluation Study [RALES]). Am J Cardiol, 78, 902–7. Thomas, W., McEneaney, V., and Harvey, B. J. (2008). Aldosterone-induced signalling and cation transport in the distal nephron. Steroids, 73(9–10), 979–84. Toto, R. A. (1997). Metabolic derangements associated with diuretic use: Insulin resistance, dyslipidemia, hyperuricemia, and anti-adronergic effects. In D. W. Seldin and G. Giebisch (eds.) Diuretic Agents: Clinical Physiology and Pharmacology, pp. 621–36. San Diego, CA: Academic Press. Tran, J. M., Farrell, M. A., and Fanestil, D. D. (1990). Effect of ions on binding of the thiazide-type diuretic metolazone to kidney membrane. Am J Physiol, 258, F908–F15. Vallon, V., Rieg, T., Ahn, S. Y., et al. (2008). Overlapping in vitro and in vivo specificities of the organic anion transporters OAT1 and OAT3 for loop and thiazide diuretics. Am J Physiol Renal Physiol, 294(4), F867–73. Van Vliet, A. A., Donker, A. J. M., Nauta, J. J. P., et al. (1993). Spironolactone in congestive heart failure refractory to high-dose loop diuretic and low-dose angiotensin-converting enzyme inhibitor. Am J Cardiol, 71, 21A–8A. Vasko, M. R., Brown-Cartwright, D., Knochel, J. P., et al. (1985). Furosemide absorption altered in decompensated congestive heart failure. Ann Intern Med, 102, 314–18. Velázquez, H. and Wright, F. S. (1986). Effects of diuretic drugs on Na, Cl, and K transport by rat renal distal tubule. Am J Physiol, 250, F1013–F23. Visweswaran, P., Massin, E. K., and Dubose, T. D. J. (1997). Mannitol-induced acute renal failure. J Am Soc Nephrol, 8(6), 1028–33. Voelker, J. R., Jameson, D. M., and Brater, D. C. (1989). In vitro evidence that urine composition affects the fraction of active furosemide in the nephrotic syndrome. J Pharmacol Exp Ther, 250, 772–8. Walter, S. J. and Shirley, D. G. (1986). The effect of chronic hydrochlorothiazide administration on renal function in the rat. Clin Sci (Lond), 70(4), 379–87. Wargo, K. A. and Banta, W. M. (2009). A comprehensive review of the loop diuretics: should furosemide be first line? Ann Pharmacother, 43(11), 1836–47. Wei, L., Struthers, A. D., Fahey, T., et al. (2010). Spironolactone use and renal toxicity: population based longitudinal analysis. BMJ, 340, c1768. Weinberg, M. S., Quigg, R. J., Salant, D. J., et al. (1985). Anuric renal failure precipitated by indomethacin and triamterene. Nephron, 40, 216–18. Weinberger, M. H., Roniker, B., Krause, S. L., et al (2002). Eplerenone, a selective aldosterone blocker, in mild-to-moderate hypertension. Am J Hypertens, 15(8), 709–16. Weiner, I. M. (1990). Diuretics and other agents employed in the mobilization of edema fluid. In A. G. Gilman, T. W. Rall, A. S. Nies, et al. (eds.) The Pharmacological Basis of Therapuetics (8th ed.), pp. 713–42. New York: Pergamon Press. Wesson, L. G. (1967). Magnesium, calcium and phosphate excretion during osmotic diuresis in the dog. J Lab Clin Med, 60, 422–32. Wesson, L. G., Jr., and Anslow, W. P. (1948). Excretion of sodium and water during osmotic diuresis in the dog. Am J Physiol, 153, 465–74. Whitling, A. M., Pergola, P. E., Sang, J. L., et al. (1997). Spironolactone-induced agranulocytosis. Ann Pharmacother, 31, 582–5.

321

322

Section 2  

fluid, electrolyte, and renal tubular disorders

Wiedemann, H. P., Wheeler, A. P., Bernard, G. R., et al. (2006). Comparison of two fluid-management strategies in acute lung injury. N Engl J Med, 354(24), 2564–75. Wigand, M. E. and Heidland, A. (1971). Ototoxic side effects of high doses of furosemide in patients with uremia. Postgrad Med J, 47, 54–6. Wilcox, C. S. (1999). Metabolic and adverse effects of diuretics. Semin Nephrol, 19(6), 557–68. Wilcox, C. S., Mitch, W. E., Kelly, R. A., et al. (1984). Factors affecting potassium balance during frusemide administration. Clin Sci, 67, 195–203. Wilson, D. R., Honrath, U., and Sonnenberg, H. (1983). Thiazide diuretic effect on medullary collecting duct function in the rat. Kidney Int, 23, 711–6. Wingo, C. S. and Straub, S. G. (1989). Active proton secretion and potassium absorption in the rabbit outer medullary collecting duct. Functional evidence for proton-potassium-activated adenosine triphosphatase. J Clin Invest, 84, 361–5.

Winter, C., Schulz, N., Giebisch, G., et al. (2004). Nongenomic stimulation of vacuolar H+-ATPases in intercalated renal tubule cells by aldosterone. Proc Natl Acad Sci U S A, 101(8), 2636–41. Wright, F. S. and Schnermann, J. (1974). Interference with feedback control of glomerular filtration rate by furosemide, triflocin, and cyanide. J Clin Invest, 53, 1695–708. Wright, J. T., Corder, C. N., and Taylor, R. (1976). Studies on rat kidney 15-hydroxyprostaglandin dehydrogenase. Biochem Pharmacol, 25, 1669–73. Yamasaki, Y., Nishiuchi, T., Kojima, A., et al. (1988). Effects of an oral water load and intravenous administration o fisotonic glucose, hypertonic saline, mannitol and furosemide on the release of atrial natriuretic peptide in men. Acta Endocrinol (Copenh), 119, 269–78. Yendt, E. R. and Cohanim, M. (1978). Prevention of calcium stones with thiazides. Kidney Int, 13, 397–409. Zannad, F., McMurray, J. J., Krum, H., et al. (2011). Eplerenone in patients with systolic heart failure and mild symptoms. N Engl J Med, 364(1), 11–21.

CHAPTER 34

Approach to the patient with hypo-/hyperkalaemia Charles S. Wingo and I. David Weiner Hypokalaemia Hypokalaemia is a common electrolyte disorder that may have a wide range of presentations. Patients may have no symptoms, exhibit neuromuscular symptoms ranging from weakness to frank paralysis, exhibit polyuria, have impairment of glucose control if they have diabetes mellitus, have exacerbation of their hypertension if hypertensive, or present with sudden death. The frequency of hypokalaemia (serum potassium concentration (SK) < 3.5 mmol/L) largely depends on the patient population. Hypokalaemia is present in < 1% of healthy adults not receiving pharmacologic agents. However, as many as 20% of hospitalized patients on a general internal medicine service (Widmer et  al., 1995)  and as many as 50% of patients treated with either loop or thiazide-type diuretics (Bloomfield et al., 1986) will exhibit hypokalaemia. Hypokalaemia frequently occurs either as a consequence or as a complication of other diseases. Thiazide and loop diuretics directly increase renal potassium excretion (Table 34.1) and result in hypokalaemia, but with excessive diuresis they also can present with prerenal azotaemia, which predisposes to hyperkalaemia. Carbonic anhydrase inhibitions, used for refractory glaucoma and for both prophylaxis and treatment of ‘mountain sickness,’ directly increase renal potassium excretion, but the ensuing potassium depletion and metabolic acidosis can prevent continued potassium loss (Maren et al., 1954). Individuals with secondary hyperaldosteronism, whether due to congestive heart failure, hepatic insufficiency, or nephrotic syndrome, frequently exhibit hypokalaemia. Patients that have increased NaCl delivery to the collecting duct and distal nephron, whether resulting from high dietary NaCl intake or from the use of diuretics are at high risk for hypokalaemia.

Classification of hypokalaemia Treating hypokalaemia should begin by identifying the cause. Broadly, hypokalaemia may be due to pseudohypokalaemia, redistribution, extrarenal potassium loss, or renal potassium loss. Pseudohypokalaemia is a rare laboratory artefact and in its absence, one should determine whether the hypokalaemia is associated with normal or decreased total body potassium content. In the latter case, the hypokalaemia corrects with sufficient replacement of the potassium deficits, which may be substantial. Hypokalaemia with normal total body potassium stores reflects potassium redistribution from the extracellular to the intracellular space. In such individuals, even ample potassium replacement therapy usually fails to fully correct the SK. Individuals with hypokalaemia and normal

cellular potassium will exhibit an increase in potassium excretion commensurate with intake that is usually due to enhanced renal potassium clearance. Thus, the assessment of persistent hypokalaemia must include not only ongoing potassium losses, but also potassium intake as well.

Pseudohypokalaemia Pseudohypokalaemia is a rare laboratory artefact from abnormal cellular uptake of potassium in the blood sample between collection and measurement, which is seen in acute myelogenous leukaemia and when blood is stored for prolonged periods at room temperature, and results in low plasma potassium measurements (Sodi et  al., 2009). However, acute myelogenous leukaemia can also cause ‘true’ hypokalaemia through inappropriate renal loss of potassium (Perry et al., 1983).

Redistribution The intracellular fluid contains > 98% of total body potassium, with the predominant stores in skeletal muscle (see Chapter 23). The hormones insulin, catecholamines, and aldosterone stimulate cellular redistribution-induced hypokalaemia. Insulin stimulates active cellular potassium uptake and can lead to hypokalaemia (Unwin et al., 2011). This frequently occurs acutely with treatment of diabetic ketoacidosis. Decreased end-organ responsiveness to insulin in adult-onset diabetes may contribute to the frequently observed hyperkalaemia. Potassium redistribution is frequently due to catecholamines and sympathomimetic agents, including β2-adrenergic agonists, dopamine, and dobutamine. These agents directly stimulate cellular potassium uptake and increase insulin release, which indirectly stimulate potassium uptake (DeFronzo, 1992; Kamel et  al., 1996). Sympathomimetic-induced redistribution leading to hypokalaemia is important in acute myocardial ischaemia and the treatment of severe asthmatic attacks. Myocardial ischaemia commonly increases sympathetic tone, whether as a direct result of the ischaemia, decreased cardiac output, or pain and anxiety from the ischaemia. The resulting hypokalaemia increases the risk of ventricular arrhythmias and sudden cardiac death. Theophylline, used to treat asthma, can lead to potassium redistribution and hypokalaemia and impair respiratory muscle function with development of CO 2 retention. Beta-agonist therapy of pregnant women with premature labour can provoke hypokalaemia.

324

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 34.1  Causes of renal potassium loss 1. Drugs A. Diuretics i. Thiazide diuretics ii. Loop diuretics iii. Osmotic diuretics B. Antibiotics i. Penicillin and penicillin analogues ii. Amphotericin B iii. Aminoglycosides C. Other drugs i. Cisplatin ii. Ifosphamide iii. Carbonic anhydrase inhibition 2. Hormones A. Aldosterone

of weakness, which may progress to paralysis, associated with redistribution-induced hypokalaemia. Attacks occur typically upon awakening from sleep or after resting, and may be precipitated by a large carbohydrate or salt meal, or by alcohol intake. Genetic abnormalities underlie the aetiology of most individuals with hypokalaemic periodic paralysis. Most hereditary cases exhibit an autosomal dominant distribution but an X-linked recessive form occurs and sporadic cases, presumably reflecting de novo mutations, have also been identified. Certain cases are due to a genetic defect in dihydropyridine-sensitive calcium channel (Ptacek et al., 1994) and others are due to defects in specific sodium channels (Jurkat-Rott et al., 2000).The attacks are best treated with oral potassium chloride if the patient is able to ingest it; otherwise intravenous potassium should be used. Because the weaknesses is due to potassium redistribution, upon resolution of the attack, SK levels increase and, if over-aggressive potassium supplementation is continued, may result in significant increases in the SK, potentially to dangerous levels. Carbonic anhydrase inhibitors (acetazolamide 250 mg four times daily), β-blockers or spironolactone may prevent attacks. Similar attacks with muscle weakness, hypokalaemia and thyrotoxicosis are observed in individuals of Asian descent (Knochel, 1992).

i. Primary

Other drugs and agents

ii. Secondary

Chloroquine poisoning (Clemessy et  al. 1995), severe verapamil overdose (Minella and Schulman, 1991; Oe et  al., 1998), and barium intoxication (Knochel, 1992; Rosa et al., 1992) have been reported to cause hypokalaemia. Clinical studies show little evidence of hypokalaemia with calcium channel antagonists (Freed et al., 1991), but several of these agents can accentuate the effect of catecholamines to induce hypokalaemia (Mimran et al., 1993a, 1993b). ‘Pa Ping paralysis’ from barium poisoning has been reported in Chinese patients ingesting food or wine with high barium concentrations (Bowen et al., 2010) and simulates familial hypokalaemic periodic paralysis. Ionized barium is a potent potassium channel blocker that impairs cellular potassium exit and repolarization of excitable tissue. Hypokalaemia with cardiac arrhythmias, skeletal muscle paralysis, and depolarization of excitable tissues is a predictable consequence of barium poisoning (Knochel, 1992).

B. Glucocorticoid-remediable hypertension C. Glucocorticoid-excess states 3. Magnesium deficiency 4. Intrinsic renal transport defects A. Bartter syndrome B. Gitelman syndrome C. Liddle syndrome 5. Bicarbonaturia A. Distal renal tubular acidosis B. Treatment of proximal renal tubular acidosis C. Correction phase of metabolic alkalosis 6. Acquired tubular transport defects A. Recovery from acute tubular necrosis B. Lysozymuria associated with leukaemia

States of rapid cellular proliferation such as acute leukaemia, high-grade lymphomas, granulocyte-macrophage colony-stimulating factor treatment of refractory anaemia, the initial treatment of pernicious anaemia with vitamin B12 (Lawson et  al., 1970), or acute anabolic states can result in hypokalaemia from cellular potassium uptake. This can cause acute hypokalaemia that may lead to arrhythmias and sudden death (Lawson et al., 1972).

Hypokalaemic periodic paralysis and thyrotoxicosis A rare, but dramatic, cause of potassium redistribution and hypokalaemia is hypokalaemic periodic paralysis (Kamel et  al., 1996; Knochel, 1992). Affected individuals have normal SK levels between attacks, but experience intermittent acute episodes

Non-renal potassium loss Gastrointestinal and sweat potassium losses can occasionally be sufficiently large to result in hypokalaemia. Normally, these sources of net fluid and potassium loss are small, but prolonged exertion in hot, dry environments, or chronic diarrhoea, can lead to severe potassium loss and hypokalaemia (Knochel et al., 1972). Hypokalaemia is a predictable consequence of prolonged loss of gastric contents, from vomiting or nasogastric suctioning. Most potassium loss is indirect due to the concomitant metabolic alkalosis, which increases renal potassium excretion (Kassirer and Schwartz, 1966). Diarrhoea, whether infectious or due to laxative abuse, can cause profound gastrointestinal potassium loss. The presence of hypokalaemia with non-anion (normal) gap metabolic acidosis should raise the possibility of diarrhoea as the aetiology of the hypokalaemia. Patients with acquired immune deficiency syndrome can develop refractory diarrhoea and hypokalaemia. Patients with laxative abuse frequently deny this condition, as do diuretic abusers, and both conditions can present a particularly difficult diagnostic

chapter 34 

challenge. Calculation of the urine anion gap (as an indirect measure of urinary ammonium excretion) may be helpful in identifying a diarrhoea-induced (with increased urinary ammonium and negative anion gap) aetiology of hypokalaemic metabolic acidosis.

Renal potassium loss The most common cause of hypokalaemia is increased renal potassium excretion, usually from drugs or, in rare conditions, intrinsic renal defects. Diuretics are the most common cause and Table 34.5 (later in the chapter) summarizes drugs frequently provoking hypokalaemia.

Drugs Many medicines increase renal potassium excretion, including diuretics, certain antibiotics, and anti-neoplastic agents, and toxins. Both thiazide and loop diuretics increase urinary potassium excretion; (Siegel et al., 1992; Stacpoole et al., 1992), but loop diuretics generally have a shorter pharmacologic half-life, enabling adaptive renal potassium conservation. All diuretics, except the potassium-sparing diuretics, increase potassium excretion increasing collecting duct luminal flow rate and luminal sodium delivery and high dietary sodium chloride intake exacerbates the kaliuretic effects of diuretics. Some antibiotics, anti-neoplastic drugs, and toxins can increase urinary potassium excretion by several mechanisms. High-dose penicillin and some penicillin analogues, such as carbenicillin, oxacillin and ampicillin, increase distal tubular delivery of a non-reabsorbable anion, which increases urinary potassium excretion (Gill et al., 1977). Polyene antibiotics, particularly amphotericin B, create cation channels in the apical membrane of collecting duct cells, which increases potassium secretion and results in impaired potassium conservation (Kamel et al., 1996). Cisplatin may induce hypokalaemia via an increase in renal potassium excretion (Jones and Chesney, 1995), and ifosfamide causes a Fanconi-like syndrome with hypokalaemia in up to 4% of patients who receive this drug (Ho et al., 1995). Toluene inhalation, which often results from ‘glue sniffing’, can also cause hypokalaemia, presumably by increasing renal potassium excretion (Mujais and Katz, 1992). Aminoglycoside antibiotics can cause hypokalaemia either in the presence or absence of overt nephrotoxicity. Potassium supplementation protects against experimental aminoglycoside nephrotoxicity (Thompson et  al., 1990)  and potassium depletion enhances aminoglycoside nephrotoxicity (Dobyan et al., 1982; Cronin and Thompson, 1991). Most antibiotics do not cause hypokalaemia, and trimethoprim and pentamidine can cause hyperkalaemia.

Hormones Endogenous hormones are important causes of hypokalaemia. Aldosterone is an important hormone that regulates total body potassium homeostasis, and excess aldosterone activity frequently leads to hypokalaemia. Hyperaldosteronism can be either primary or secondary. Primary hyperaldosteronism results in hypertension (Holland, 1995), in part due to the sodium-retaining effects of aldosterone and partly through direct effects of aldosterone on vascular endothelium and on vascular smooth muscle cells, and through central nervous system-induced mechanisms. In addition, the associated hypokalaemia may also contribute by sensitizing the vasculature to neurohumoral regulators of blood pressure. An aldosterone-producing

approach to the patient with hypo-/hyperkalaemia adrenal adenoma (APA) is a potentially surgically curable cause of primary hyperaldosteronism, but this condition should be distinguished from bilateral adrenal hyperplasia, which is not amenable to surgical correction. With the current use of the aldosterone:renin ratio (ARR) as a screening tool for the identification of primary hyperaldosteronism, an APA is now recognized as being present in only a minority of patients with primary hyperaldosteronism (Weiner and Wingo, 2010). Angiotensin II stimulates adrenal gland aldosterone synthesis, and conditions that increase plasma angiotensin II concentration typically cause secondary hyperaldosteronism. This may occur in a variety of conditions that stimulate renin secretion, including intravascular volume depletion, congestive heart failure, acute and chronic liver dysfunction, and nephrotic syndrome. Activation of the renin–angiotensin–aldosterone system is a consistent finding in malignant hypertension (Holten and Peterson, 1955), renovascular hypertension (Simon et  al., 1972), and renin-secreting tumours (Brown et al., 1973). Certain genomic defects lead to excessive aldosterone production. In glucocorticoid-remediable aldosteronism, an adrenocorticotropin (ACTH)-regulated gene promoter is linked to the coding sequence of the aldosterone synthase gene, the rate-limiting enzyme for aldosterone synthesis (Lifton et al., 1992). Consequently, aldosterone synthase is regulated by ACTH rather than angiotensin-II, and excessive aldosterone production ensues. Congenital adrenal hyperplasia from either 11β-hydroxylase or 17β-hydroxylase enzyme deficiency, results in, excessive hypothalamic corticotropin-releasing hormone (CRH) secretion and persistent adrenal synthesis of 11-desoxycorticosterone, a potent mineralocorticoid (White et  al., 1987). Phenotypically 17β-hydroxylase deficiency inhibits sex hormone metabolism and, leads to incomplete development of sexual characteristics, whereas 11β-hydroxylase deficiency results in increased androgen production, leading to early virilization of males and females. Rarely glucocorticoids function as mineralocorticoids, causing hypokalaemia and hypertension. The glucocorticoid, cortisol, has a high affinity for the mineralocorticoid receptor, but in selectively mineralocorticoid-responsive cells normally is metabolized intracellularly by the enzyme 11β-hydroxysteroid dehydrogenase type 2(11β-HSDH2) which converts cortisol to cortisone, and cortisone does not bind to the mineralocorticoid receptor (Funder et al. 1988). The importance of this enzyme is illustrated by such rare conditions as the syndrome of apparent mineralocorticoid excess, which results from absence of 11β-HSDH2 activity (Mune et  al. 1995; Ferrari et  al. 1996a, 1996b). Children with this syndrome exhibit early onset severe hypertension, a high incidence of cerebral infarction, and electrolyte features of mineralocorticoid excess, including hypokalaemia (Oberfield et al., 1979; New et al., 1986). Certain natural products and drugs such as glycerrhetinic acid and carbenoxolone inhibit 11β-HSDH2, allowing cortisol to exert mineralocorticoid-like effects (Farese et al. 1991).

Magnesium depletion Magnesium depletion is found in many clinical circumstances associated with potassium depletion (see Chapter  27). Between 10% and 40% of individuals with potassium deletion also exhibit magnesium depletion (Watson and O’Kell, 1980; Whang et  al., 1992). Simultaneous magnesium and potassium depletion are observed frequently with diuretic administration, diabetic ketoacidosis, chronic alcoholism, and with the recovery from acute tubular

325

326

Section 2  

fluid, electrolyte, and renal tubular disorders

necrosis. This is also true in certain cases of aminoglycoside toxicity and cisplatin toxicity, hypokalaemia associated with lysozymuria in acute leukaemia, and in individuals with Gitelman syndrome (see below). Hypomagnesaemia is frequently observed with hypokalaemia and may contribute to its development. In humans, magnesium depletion is associated with hypokalaemia (Kelepouris and Agus, 1998)  and may induce renal potassium wasting (Shils, 1969). In addition, dietary magnesium depletion causes a selective potassium loss from cardiac as well as skeletal muscle, and intracellular potassium deficiency may not be restored by potassium administration alone in the presence of magnesium deficiency (Whang, 1987; Rodriguez et al., 1989). These observations suggest co-administration of magnesium may further improve potassium handling when both are deficient. Potassium loss in magnesium deficiency may also be related to a relative increase in the activity of renal potassium channels and, hence, continued urinary potassium excretion (Hille, 1992). Magnesium oxide 250–500 mg by mouth four times daily is the preferred method of replacement.

Renal transport defects Genetic diseases of hypokalaemia are rare, but have advanced our understanding of renal physiology. In 1962, Bartter described the association of hypokalaemia, hypomagnesaemia, hyper-reninaemia, and metabolic alkalosis (Bartter et al., 1962). Further phenotypic refinement led to the recognition of two syndromes: Bartter syndrome and Gitelman syndrome (Gitelman et al., 1966). Patients with Bartter syndrome feature hypercalciuria and present generally at an early age with severe volume depletion and evidence of ‘failure to thrive’. This condition most commonly results from defects in either the renal Na-K-2Cl cotransporter gene, NKCC2 (Simon et al., 1996a), the Kir1.1 potassium channel (ROMK) or a basolateral Cl− channel (ClC-1). These three proteins are necessary for sodium reabsorption in the thick ascending limb of the loop of Henle (Simon et al., 1996b). Gitelman syndrome features hypocalciuria, hypomagnesaemia, milder clinical manifestations, and it generally presents at a later age. This syndrome most commonly is due to mutations affecting the thiazide-sensitive NaCl cotransporter (Simon et al., 1996c). Hypotension and intravascular volume depletion due to renal sodium-wasting are common features in both. Hypokalaemia results, in part, from renal defects in regulation of potassium excretion resulting from failure of sodium absorption in the loop of Henle or distal convoluted tubule, which results in increased distal sodium delivery and luminal flow rates, and increased renal potassium excretion at normal SK values. Concomitant intravascular volume depletion and secondary hyperaldosteronism further exacerbates the hypokalaemia. Treatment of the hypokalaemia frequently involves oral administration of large amounts of potassium chloride, but some degree of hypokalaemia frequently persists. Liddle syndrome is associated with hypertension, hypokalaemia, metabolic alkalosis, and suppressed renin and aldosterone levels (Liddle et al., 1963). Defects in the collecting duct epithelial sodium channel, ENaC, are responsible for this condition, which leads to excessive sodium reabsorption and presumed volume expansion, hypertension, and suppression of renin and aldosterone (Schild et al., 1995). Mutations in the β and γ, but not the α, subunits have been reported (Bubien et al., 1996; Gao et al., 2001).

Bicarbonaturia and other poorly reabsorbable anions Bicarbonaturia and increased excretion of poorly reabsorbable anions stimulate potassium secretion in part by reducing luminal chloride ion concentration which facilitates KCl secretion by the distal nephron and collecting duct (Ellison et  al., 1985; Wingo, 1989). This occurs with intravenous infusion of the NaHCO3, Na2HPO4, Na2SO4, MgSO4, and semi-synthetic penicillin such as carbenicillin and related compounds. The obligatory loss of cations, to preserve electroneutrality, results in renal potassium wasting. Bicarbonaturia can result from metabolic alkalosis, distal renal tubular acidosis (RTA) or treatment of proximal RTA. In each case, increased distal tubular luminal bicarbonate delivery increases potassium secretion (Malnic et al., 1971).

Diagnosis of hypokalaemia Evaluation of the patient with hypokalaemia should begin with a thorough history and physical examination. Fig. 34.1 provides a logical algorithm. One should first consider and exclude pseudohypokalaemia due to potassium uptake by abnormal leucocytes and consider hormones, drugs or conditions that result in redistribution of potassium from the extra- to the intracellular space. If none of these possibilities is present, then the hypokalaemia likely represents total body potassium depletion resulting from potassium loss via the kidney, gastrointestinal tract, or skin. Diuretics are a frequent cause of hypokalaemia due to renal loss of potassium. Hypomagnesaemia-induced hypokalaemia may causes renal potassium wasting, and can occur with aggressive diuresis. Gastrointestinal potassium loss occurs from diarrhoea, vomiting, nasogastric suction, or a gastrointestinal fistula. Many cases are apparent from the history and the clinical setting, but less obvious causes include surreptitious vomiting and laxative abuse, which are frequently diagnostic challenges. Erosion of the dental enamel, metabolic alkalosis, and low urinary chloride content are all features of chronic vomiting and clues to its diagnosis. In patients with self-induced diarrhoea from catharctics, the history of laxative use may be difficult to obtain. Habitual use of anthracene laxatives, such as senna, cascara, and aloe, leads to melanosis coli (Wittoesch et al., 1958), and the diagnosis can be supported by sigmoidoscopy. Phenolphthalein is a cathartic and has been previously used in laxatives that turns pink or purple in the presence of a strong alkali. The development of a pink-purple colour to the stool after the application of NaOH or KOH suggests the diagnosis. Excessive potassium loss from the skin can result from prolonged exposure in hot environments where sweat loss is high, and this diagnosis should be apparent from the history. Type 1 (distal) and II (proximal) RTA (see Chapter  36), diabetic ketoacidosis, and ureterosigmoidostomy also cause renal potassium loss. These conditions may present with frank acidosis (Wrong and Davies, 1959). Bartter and Gitelman syndromes are rare genetic disorders that exhibit hypokalaemia and usually exhibit metabolic alkalosis with a normal or low blood pressure. However, a much more frequent cause for this constellation of findings is surreptitious diuretic abuse. A  urine screen for diuretics is an important component of the evaluation of the patient with possible Bartter or Gitelman syndrome in order to exclude surreptitious diuretic use. Bartter and Gitelman syndrome can be differentiated from each other by assessment of urinary calcium excretion, which is high with Bartter syndrome and suppressed in Gitelman syndrome.

chapter 34 

approach to the patient with hypo-/hyperkalaemia

Diagnostic evaluation of hypokalemia (K < 3.5 mEq/L) WBC count 50,000

Is acute redistribution likely from recent insulin, beta-adrenergic agonists or theophylline use? Less common causes include hyperaldosteronism, acute anabolic stimulus or history suggestive of hypokalemic periodic paralysis?

K after rapid separation of plasma and storage at 4ºC Normal

Low

Pseudohypokalemia

Yes

No Skin, GI or renal potassium loss

< 20 mEq/24 hr

Urine K?

Redistribution or hypokalemic periodic paralysis

> 20 mEq/24 hr

Skin or GI K loss

Renal K loss Recent diuretic use?

Prolonged exertion in hot environment?

No

Diarrhea or GI fistula?

Yes No

No

Yes Probable diuretic-induced hypokalemia

Yes

Skin loss GI K loss

CHF, hepatic insufficiency, nephrotic syndrome or renal artery stenosis?

Presence of cathartics or osmotic agents (high magnesium or phosphate concentration) in stool? Yes

Yes

No

Hypomagnesemia? No Yes

Surreptitious cathartic use Inadequate dietary potassium intake or diuretic use recently discontinued

Probable secondary hyperaldosteronism No Serum bicarbonate?

Hyomagnesemia-induced hypokalemia Either normal or high Blood pressure? Low

Nasogastric suction, surreptitious vomiting or diuretic use, or, more rarely, Bartters or Gittleman’s syndromes

Low

Renal tubular acidosis, ureterosignoidostomy High

Possible primary hyperaldosteronism or, rarely, other mineralocorticoid excess states

Fig. 34.1  A method for the evaluation of hypokalaemia. Modified from Weiner and Wingo (1997).

Poor potassium intake in combination with elevated sodium chloride intake and primary hyperaldosteronism are two likely diagnoses in patients with persistently low or borderline low SK concentrations and hypertension, particularly if metabolic alkalosis is present (Hilden and Krogsgaard, 1958; Wrong, 1961). The absolute plasma aldosterone concentration in combination with the plasma

aldosterone to plasma renin activity ratio has been used to differentiate these possibilities. In patients with conditions other than primary hyperaldosteronism, the plasma aldosterone level is normal and the ARR is normal, usually 10:1. In typical primary hyperaldosteronism, the plasma aldosterone level is at least > 10 ng/mL and the ARR is elevated, typically 50:1, or greater (depending on units).

327

328

Section 2  

fluid, electrolyte, and renal tubular disorders

However, assays reporting renin concentration, rather than activity, will have different normal renin values and consequently different values for the ARR (higher ratio and threshold for renin activity). For complete details of the evaluation and treatment of primary hyperaldosteronism refer to textbook chapters that deal with the diagnosis of hyperaldosteronism (Weiner and Wingo, 2010). Rare causes of hypokalaemia include liquorice-like compounds, which inhibit metabolism of cortisol to cortisone by 11β-OH-steroid dehydrogenase type 2, and exhibit a clinical picture of mineralocorticoid excess (Blachley and Knochel, 1980).

Treatment of hypokalaemia The primary short-term risks of hypokalaemia is cardiovascular, and the most important effect in the short term is to predispose to cardiac arrhythmias. However, the primary risk of too rapid potassium replacement is the development of hyperkalaemia, with resultant ventricular fibrillation. Thus, the risks associated with hypokalaemia must be balanced against the risks of therapy when determining the appropriate approach to the patient. Whenever possible, replacement therapy should be administered orally, which allows endogenous gastrointestinal potassium sensors to monitor potassium repletion therapy (Morita et al., 2000; Lee et al., 2007; Oh et al., 2011). Occasionally, incorrect therapy of hypokalaemia (e.g. using D5W-containing solutions, see below) can lead to paradoxical worsening of the hypokalaemia. Situations that require emergent therapy are rare but may include the patient with severe hypokalaemia that requires emergent surgery, and the concern is heightened if the patient has known coronary artery disease or is receiving digitalis. Some retrospective studies have suggested that the incidence of intraoperative complications attributable to hypokalaemia is low (Hirsch et al., 1988; Wong et  al., 1993), and that occurrence of complex ventricular arrhythmias appears to correlate better with a history of long-term digoxin therapy or congestive heart failure than with plasma potassium levels (Hirsch et al., 1988). A second generally accepted indication for emergent therapy is patient with an acute myocardial infarction and significant ventricular ectopy. In such cases, administration of 5–10 mmol of KCl over 15–20 minutes and repeated as needed may be used to increase SK above 3.0 mEq/L. Continuous monitoring of the serum level and the electrocardiogram (ECG) are necessary to reduce the risk of hyperkalaemia. Finally, hypokalaemia is frequently associated with some degree of skeletal muscle weakness and with severe hypokalaemia frank paralysis can ensue with respiratory compromise which requires urgent treatment. The choice of parenteral versus oral therapy usually depends on the ability of the patient to take oral medicine and a normally functioning gastrointestinal tract (Weiner and Wingo, 1997). If the patient is unable to take oral potassium safely, KCl may need to be given intravenously. When given intravenously, KCl replacement can be given safely at a rate of 10 mmol/h, typically for an individual dose of 40 mmol. One study has found that 20 mmol/h of KCl causes the SK to increase by an average of 0.25 mmol/L per hour (Kruse and Carlson, 1990). If more rapid replacement is necessary, then 40 mmol/h can be administered through a central catheter provided continuous ECG monitoring is in use. However, such rates are rarely needed and oral replacement therapy is safer and is the preferred route of administration. The choice of parenteral fluids used for potassium administration can affect the response. Intravenous D-glucose administration increases serum insulin levels, which can stimulate cellular

potassium uptake. As a result, administering KCl in D5W can paradoxically lower SK (Kunin et al., 1962). Thus, parenteral KCl should be provided in normal saline. If large concentrations of KCl are necessary and are added to the parenteral fluid, then KCl may be administered in half normal saline to avoid administration of a hypertonic solution, but the potassium concentration in the replacement fluid should not exceed 40 mmol/L. Hypokalaemia usually can be treated successfully with oral therapy. If diuretic therapy is required, for example, in the treatment of hypertension or heart failure, concomitant use of a potassium-sparing diuretic, such as amiloride or triamterene, should be considered and the dietary sodium and potassium content reassessed. KCl is the preferred potassium salt in most patients, except for patients with metabolic acidosis, because it minimizes renal potassium losses. In metabolic acidosis with hypokalaemia, potassium citrate is preferred. Hypomagnesaemia can lead to renal potassium wasting, and refractoriness to potassium replacement (Kamel et  al., 1996). Correction of the hypokalaemia may not occur until the hypomagnesaemia is corrected (Shils, 1969). Patients with diuretic-induced hypokalaemia who are refractory to oral potassium chloride administration should be tested for hypomagnesaemia, and magnesium replacement therapy begun if indicated. The coexistence of other electrolyte abnormalities, particularly hypophosphataemia, should be also sought.

Hyperkalaemia Hyperkalaemia, when severe, has predictable effects on cardiac electrical conduction which make this condition a potentially lethal disorder; however, from a clinical perspective many cases of hyperkalaemia are asymptomatic. The assessment of hyperkalaemia includes exclusion of laboratory error and pseudohyperkalaemia, determination of the urgency for treatment, and institution of appropriate therapy. Long-term treatment requires identification of the aetiology and prevention of recurrence.

Classification of hyperkalaemia Hyperkalaemia reflects impaired potassium clearance relative to potassium intake or an altered distribution between intra- and extracellular potassium, but chronic stable hyperkalaemia without a change in potassium intake indicates renal adaptation albeit at an abnormal plasma potassium concentration. To evaluate a patient with hyperkalaemia, one should consider four broad groups of aetiologies: pseudohyperkalaemia and laboratory artefacts, excessive intake, redistribution, and impaired renal potassium clearance. A careful history and physical examination in combination with selected laboratory tests is sufficient to differentiate most cases.

Laboratory artefacts and pseudohyperkalaemia The method of collection and sample handling can significantly affect the estimate of the patient’s true blood potassium concentration. Pseudohyperkalaemia refers to reported potassium values that do not accurately reflect the potassium activity in the patient’s blood and usually extracellular space. Frequently, potassium concentration is measured in blood that has been allowed to clot and centrifuged to obtain the serum. Potassium release from any of the cellular elements of blood can artificially elevate the SK concentration. The most common cause of pseudohyperkalaemia is haemolysis, which is usually noted by the laboratory due to a pink tinge to

chapter 34 

the serum resulting from release of haemoglobin from the damaged red blood cells. Small needle size and excessive aspiration are frequent causes of haemolysis. Haemolysis that does not produce visible colour change to the serum should not significantly increase SK concentration. Centrifuging the specimen before the clot has formed completely can increase the susceptibility of red blood cells to membrane damage during centrifugation. This can lead to leakage of potassium from erythrocytes and to development of pseudohyperkalaemia. In addition, exercising muscle releases potassium. Ischaemia, as with an excessively tight tourniquet, can increase SK in some cases by > 2 mEq/L (Skinner and Adelaide, 1961). Even minimal exercise, such as ‘fist squeezing’ during the phlebotomy procedure, can result in sufficient skeletal muscle potassium release to invalidate the potassium measurement. Release of potassium from leucocytes and platelets can also cause pseudohyperkalaemia. Leucocytosis (Bronson et al., 1966; Bellevue et al., 1975; Lichtman and Rowe, 1982), > 70,000/cm3, or thrombocytosis (Hartmann and Mellinkoff, 1955; Harman et al., 1958; Paice et al., 1983), > 1,000,000/cm3, can frequently lead to pseudohyperkalaemia. The resulting change in SK generally is proportional to the severity of the leucocytosis or thrombocytosis, and can occur with less severe platelet or leucocyte values. With platelet counts between 500,000/cm3 and 1,000,000/cm3, 34% of patients exhibit pseudohyperkalaemia (Graber et al. 1988). Pseudohyperkalaemia should also be suspected if there is a family history of hyperkalaemia, or if conditions associated with significant leucocytosis or thrombocytosis are present. Rarely, pseudohyperkalaemia has been reported in association with rheumatoid arthritis (Ralston et al., 1988) and mononucleosis (Ho-Yen and Pennington, 1980). Occasional families have abnormal red blood cell membrane potassium permeability, which leads to excessive potassium leakage rates and pseudohyperkalaemia (Stewart et al., 1979; James and Stansbie, 1987; Dagher et al., 1989). Recognizing pseudo-hyperkalaemia is important, because it is purely a laboratory artefact and does not require specific therapy. Inappropriate treatment of pseudohyperkalaemia can result in serious hypokalaemia and increase the risk of hypokalaemia-related complications. Pseudohyperkalaemia can be excluded by simultaneously measuring plasma and SK concentrations. Plasma potassium can be measured by obtaining a heparinized blood specimen, using a ‘green-top’, or heparinized, tube. If the SK is abnormal and exceeds the plasma potassium by > 0.3 mmol/L, pseudohyperkalaemia is strongly suspected, and subsequent potassium measurements should be determined using plasma samples.

Excess intake or potassium release The normal kidney can excrete hundreds of mmoles of potassium per day (Rabelink et al., 1990), and early studies demonstrated the ability of the normal kidney to adapt chronically to even greater potassium loads (Schwartz, 1955). Thus, excessive potassium ingestion is an infrequent cause of hyperkalaemia in the absence of other contributing factors. However, if renal potassium excretion is impaired, whether through drugs, renal insufficiency, or other causes, then excess potassium intake can produce hyperkalaemia. It is important to recognize that the renal mechanisms which enable excretion of large amounts of potassium are much more sensitive to oral potassium intake than to intravenous potassium administration possibly due to gastrointestinal potassium sensors. Thus, the risk of acute hyperkalaemia with potassium supplementation is much greater with intravenous compared to oral potassium administration.

approach to the patient with hypo-/hyperkalaemia

Potassium intake Although abnormal renal potassium clearance is generally necessary for the development of persistent hyperkalaemia, excessive potassium intake is often an aggravating factor. Essentially all foods contain potassium, although the relative amounts of potassium differ greatly (Table 34.2). Common causes of hyperkalaemia are potassium supplements and salt substitutes. For example, as many as 4% of patients receiving potassium chloride supplements develop hyperkalaemia (Lawson, 1974). Salt substitutes frequently contain 10–13 mmol/g, which is equivalent to 283 mmol/tablespoon Table 34.2  Foods rich in potassium Greatest potassium content (> 1000 mg (25 mmol)/100 g:

Sun-dried tomatoesLima beans, mature Molasses Wheat bran, crude

Very large potassium content (>500 mg (12.5 mmol)/100 g):

Plums, dried (Prunes) Dates Peanuts, raw Cashews, raw Dried figs Oat Bran, raw Spinach, raw Walnuts, black dried

Large potassium content (>250 mg (6.2 mmol)/100 g)

Plantains

Vegetables

Avocados Potatoes Sweet potatoes Winter squash Beets Carrots Broccoli Cauliflower Tomatoes Summer Squash

Fruits

Bananas Purple Passion fruit Kiwis Cantaloupe Mangos Oranges Grapefruit

Meats

Lamb Lean Pork Steak Veal

329

330

Section 2  

fluid, electrolyte, and renal tubular disorders

(Sopko and Freeman, 1977). Many enteral nutrition products contain 40 mmol/L KCl or more; and administration of 100 mL/h of such products can result in a potassium intake of approximately 100 mmol/day. Some studies estimate that 50% of all cases of hyperkalaemia are related to potassium supplements (Shapiro et al., 1971; Paice et al., 1983; Shemer et al., 1983; Borra et al., 1988). Oral alimentation solutions are another source of dietary potassium (Table 34.3). Patients with renal failure who are receiving complete nutritional support through enteral nutritional supplements frequently develop hyperkalaemia. For intravenous hyperalimentation fluids, the potassium content recommended in normal individuals usually should be reduced when administered to patients with renal insufficiency, and frequent monitoring of the serum/plasma potassium concentration is recommended. Medicines are another source of potassium. Potassium supplements may provoke hyperkalaemia, especially if administered intravenously. They are ordered frequently for patients receiving diuretics who may also be receiving other drugs that predispose to hyperkalaemia. Potassium supplements may be prescribed with increasing frequency for conditions other than hypokalaemia, due to the recognition that potassium supplementation decreases blood pressure (Smith et al., 1992; Whelton et al., 1997) and may improve mineral balance and skeletal calcium metabolism in post-menopausal women (Sebastian et al. 1994). This has the potential for an increased incidence of hyperkalaemia, particularly in patients with chronic kidney disease (CKD) or if the SK is not monitored. Penicillin and citrate salts are other medicines that supply potassium. Penicillin G is supplied as a potassium salt, supplying 1.7 mmol/1,000,000 units, but may be supplied as sodium penicillin G.  Citrate therapy is a common method to supply alkali to patients with a variety of conditions including advanced renal insufficiency, RTA, and nephrolithiasis. Citrate can be supplied either as a sodium salt, potassium salt, or as a sodium-potassium salt (Table 34.4). Because some citrate preparations can provide large amounts of potassium, significant hyperkalaemia may develop if one does not monitor SK and note the method of alkali replacement therapy.

Tissue necrosis Tissue necrosis can lead to hyperkalaemia, depending on the mass and the rapidity of the cell lysis. Common examples include rhabdomyolysis, ischaemic extremities or bowel and haemorrhage, particularly retroperitoneal haemorrhage. Rhabdomyolysis can result from crush injury, seizures, electrical shock, cocaine ingestion, sepsis, Table 34.3  Potassium content of common enteral supplements Calories/cc

Potassium (mmol/L)

Sodium (mmol/L)

Osmolality (mOsm/kg)

Ensure®

1.06

42

35

520

Ensure Plus®

1.50

43

40

680

Glucerna®

1.00

40

40

355

Osmolite®

1.06

40

40

300

Pulmocare®

1.50

50

57

475

Suplena®

2.00

29

35

780

Ultracal®

1.06

41

40

310

Vivonex TEN®

1.00

24

27

630

Table 34.4  Potassium content of various citrate-based alkali preparations Potassium (mmol/mL)

Sodium (mmol/mL)

Citrate/citric acid (mmol/mL)

Polycitra-K®

2

–

2

Bicitra® (Shohl’s solution)

–

1

1

Polycitra®

1

1

2

ischaemia, blunt or penetrating trauma, and excessive exertion. Of note, potassium deficiency can cause or predispose to rhabdomyolysis, resulting in potassium liberation into the extracellular fluid compartment, which results in either normokalaemia or hyperkalaemia, despite total body potassium deficiency. Rhabdomyolysis is a cause of acute kidney injury that impairs renal potassium excretion potentially lead to lethal hyperkalaemia. Large amounts of potassium and nucleic acids are liberated in the treatment of rapidly proliferating lymphomas. Without prior hydration and volume expansion to preserve urine output, hyperkalaemia and uric acid acute kidney injury may ensue (Arrambide and Toto, 1993).

Redistribution Several common clinical conditions are known to cause redistribution. These include membrane-depolarizing anaesthetics, and extracellular hypertonicity if due to ‘effective osmoles’. In addition, less common causes of hyperkalaemia include drugs and conditions that affect membrane voltage.

Acid–base disturbances Patients without endogenous renal function exhibit little change in SK with acute or sustained NaHCO3 infusion and substantial change in acid–base status, which suggests that the effect of acid–base disturbances on SK is principally due to effects on renal potassium clearance (Toussaint and Vereerstraeten, 1962). In general, metabolic acidosis due to ‘organic acids’, such as β-hydroxybutyric acid or lactic acid, has little direct effect on SK concentration. In diabetic ketoacidosis, hyperkalaemia is generally due to the lack of insulin-stimulated cellular potassium uptake and to the presence of glucose as an ineffective extracellular osmole (discussed in more detail below) and not the concomitant metabolic acidosis.

Hyperosmolality Hyperosmolality can cause hyperkalaemia as a predictable effect of potassium redistribution from intracellular compartment into the extracellular space. Hyperosmolality, when caused by ‘effective osmoles’, can increase SK by 1–2 mmol/L (Goldfarb et al., 1975; Makoff et al. 1970, 1971; Viberti, 1978). Hyperglycaemia, if occurring in the absence of either sufficient insulin or tissue responsiveness to insulin, and mannitol, often used in neurosurgical patients, are common causes of hyperosmolality. Hypertonicity causes cell shrinkage, leading to stimulation of net cellular potassium efflux. Hyperosmolality from solutes that rapidly cross plasma membranes, such as urea, does not alter cellular volume and does not cause hyperkalaemia. In diabetic ketoacidosis the hyperglycaemia and attendant hyperosmolality may also contribute to the hyperkalaemia in addition to the insulinopenic state.

chapter 34 

approach to the patient with hypo-/hyperkalaemia

Drugs

Renal potassium excretion

Drugs may interfere with the hormonal systems that regulate the distribution of potassium between the intra- and extracellular fluid compartments. Typically, they do so through predictable effects on hormone systems that regulate potassium homeostasis and renal transport mechanisms know to be involved in renal potassium excretion. Aldosterone, insulin, and β-adrenergic agonists are known to affect the transcellular potassium distribution. The normal response to increased potassium intake is increased aldosterone synthesis, and aldosterone stimulates cellular potassium uptake. Many drugs in common use affect aldosterone synthesis or action and aldosterone’s action has a major effect on the SK. Aldosterone synthesis is regulated in large part by renin-stimulated angiotensin II production acting through the angiotensin II, type 1 (AT1) receptor. Thus, drugs that (a) decrease renin secretion, such as β-adrenergic antagonists and atrial natriuretic peptide (ANP) analogues; (b) block renin action (direct renin inhibitors); (c) inhibit angiotensin converting enzyme (ACE); (d) block AT1 receptors (ARBs); (e) decrease adrenal aldosterone synthesis; or (f) inhibit aldosterone action are frequent causes of hyperkalaemia (Atlas and Maack, 1992). Heparin is a dose-dependent inhibitor adrenal aldosterone production. It can cause hyperkalaemia in a small percentage (~ 7%) of patients, particularly when given intravenously at high doses (Oster et al., 1995). Low-molecular-weight heparin has less effect on aldosterone synthesis and is less likely to cause hyperkalaemia. Spironolactone and eplerenone, clinically important mineralocorticoid receptor antagonists, inhibit aldosterone action at the cellular level. Increased use of spironolactone and eplerenone, in response to evidence that they decrease mortality in many patients with congestive heart failure, has resulted in a substantial increase in the number of patients being admitted with hyperkalaemia. Additionally, the progesterone agonist drospirenone used in some birth control pills can cause hyperkalaemia (Cremer et al., 2010). Cationic amino acids, such as arginine and lysine, can cause hyperkalaemia, probably by exchanging with cellular potassium. Toxic levels of cardiac glycosides, such as the medicine digoxin, or from poisoning with related compounds, can lead to hyperkalaemia as a predictable effect of inhibiting Na+,K+-ATPase and thereby decreasing cellular K+ uptake (Weizenberg et al., 1985; Wenger et al., 1985). Fluoride intoxication can rarely cause of death due to hyperkalaemia (Baltazar et al., 1980; Bradberry and Vale, 1995).

The normal kidney’s ability to excrete potassium is sufficiently large that chronic hyperkalaemia is rarely observed unless either renal function or the renal mechanism of active potassium secretion is impaired (Schwartz, 1955). For example, daily ingestion of 400 mmol of KCl, approximately four- to ten-fold greater than the usual daily intake, increases SK by < 1 mmol/L if renal function is normal and potassium excretion mechanisms are intact (Rabelink et al., 1990). Persistent hyperkalaemia which is not due to laboratory artefact or redistribution thus almost always involves alteration in renal potassium clearance. Essentially all regulation of renal potassium excretion occurs in the renal collecting duct, and it proximal extension the initial collecting tubule. This is the site of action of many medicines which affect collecting duct potassium secretion (Table 34.5). Many drugs affect potassium clearance by chronically inhibiting potassium secretion, by either inhibiting aldosterone synthesis production or action, or by inhibiting the cellular processes necessary for collecting duct potassium secretion (see below). Mineralocorticoids play an important role regulating extracellular potassium concentration through both effects on cellular potassium redistribution and by enhancing the maximum capacity for potassium secretion by the aldosterone-sensitive distal nephron. A major mechanism regulating aldosterone production is angiotensin II-stimulation of adrenal cortical cells through the AT1 receptor. Accordingly, any drug that affects angiotensin II production and/or activation of AT1 receptor can contribute to the development of hyperkalaemia. Medicines such as direct renin inhibitors, ACE inhibitors and ARBs directly or indirectly inhibit the action of the AT1 receptor and can provoke hyperkalaemia. Heparin can Table 34.5  Drugs that frequently cause hyperkalaemia Common

 Amiloride  Triamterene   Spironolactone & eplerenone   Potassium-sparing diuretics  Ciclosporin

Effects on membrane voltage Rarely, drugs and certain conditions can cause membrane depolarization that can lead to hyperkalaemia. The most common example is the skeletal muscle relaxant succinylcholine (Sterns et al., 1981). Hyperkalaemic periodic paralysis represents a rare form of period paralysis associated with weakness frequently provoked by exercise. Defects in the gating of skeletal muscle sodium channels has been identified as a cause of this condition (Lehmann-Horn et al., 1991). Genetic analysis has revealed mutations and polymorphisms at the SCN4A locus for this disorder and the closely related condition paramyotonia congenita (McClatchey et al., 1992). These two, rare genetic muscle disorders share common features of myotonia and episodic weakness. In hyperkalaemic periodic paralysis, patient symptoms and signs are worsened by increased SK In paramyotonia congenita muscle cooling exacerbates symptoms (Plassart et al., 1994). Exercise frequently provokes attacks in hyperkalaemia periodic paralysis and β-adrenergic agonists have been reported to improve attacks (Wang and Clausen, 1976; Bendheim et al., 1985).

  Non-steroidal anti-inflammatory drugs

  Tacrolimus (FK506)  Heparin  Angiotensin-converting enzyme inhibitors   Angiotensin receptor blockers  Pentamidine  Sulfamethoxazole-trimethoprim (high-dose therapy) Less common

  Cationic amino acids, arginine, lysine   β-Adrenergic antagonists  Succinylcholine   Digitalis poisoning   Fluoride poisoning   Lithium toxicity

331

332

Section 2  

fluid, electrolyte, and renal tubular disorders

inhibit aldosterone synthase, the rate-limiting enzyme for aldosterone synthesis, and thereby, in a dose-dependent mechanism can contribute to hyperkalaemia. Without prior adaptation, sodium reabsorption in the collecting duct is essential for maximum rates of collecting duct potassium secretion. This occurs because potassium secretion normally requires sodium absorption. Several medications inhibit collecting duct sodium reabsorption. The potassium-sparing diuretics amiloride and triamterene are specific inhibitors of the apical sodium channel, ENaC (Hartmann and Mellinkoff, 1955), as are the antibiotics trimethoprim and pentamidine (Velazquez et  al., 1993; Schlanger et  al., 1994; Kleyman et al., 1995) (Lachaal and Venuto, 1989; Briceland and Bailie, 1991; O’Brien et al., 1997). High-dose trimethoprim, typically used with sulfamethoxazole to treat Pneumocystis jirovecii (formerly carinii) pneumonia, increases the SK an average of 1.1 mmol/L (Greenberg et  al., 1993b), and can cause life-threatening hyperkalaemia if not recognized (Greenberg et al., 1993a, 1993b; Hsu and Wordell, 1995; Marinella, 1995). Those with renal insufficiency and the elderly may develop hyperkalaemia even with conventional doses of sulfamethoxazole-trimethoprim (Perazella and Mahnensmith, 1996; Perlmutter et  al., 1996). Lithium therapy has been reported to cause hyperkalaemia (Mercado and Michelis, 1977), which may be related to its effect on collecting duct function (Eiam-Ong et al., 1993). Severe intravascular volume depletion can cause hyperkalaemia if more proximal renal epithelial sodium reabsorption is nearly complete. Arachidonic acid metabolites play an important role in collecting duct potassium secretion. This is in part due to their action to decrease renin release, thereby decreasing aldosterone production (Larsson et al., 1974). Arachidonic acid metabolites also regulate potassium channels; and non-steroidal anti-inflammatory drugs (NSAIDs) reduce their production and decrease potassium channel activity (Ling et al., 1992; Macica et al., 1996), thereby decreasing potassium secretion and, potentially, causing hyperkalaemia. Renal potassium excretion requires intact basolateral Na+,K+-ATPase function, both to enable continued sodium reabsorption and to enable renal collecting duct cell potassium uptake and secretion. Digoxin, and its analogues, inhibit Na+,K+-ATPase, and through this action can cause hyperkalaemia (Citrin et  al., 1972; Reza et al., 1974). In addition, Na+,K+-ATPase is critical for cellular potassium uptake and the maintenance of high intracellular potassium content in non-renal cells. Accordingly, digoxin, and other digitalis glycosides, can cause hyperkalaemia in predisposed patients, such as those with advanced CKD (Papadakis et al., 1985). Digoxin-induced hyperkalaemia most commonly occurs in conditions of super-therapeutic digoxin levels, but can contribute to hyperkalaemia in patients predisposed because of other co-morbid conditions or medications. Another class of pharmacologic agents with similar actions are bufadienolides, naturally occurring cardioactive steroids that have digoxin-like effects, which have been associated with several deaths (Brubacher et al., 1995). Calcineurin inhibitors are widely used immunosuppressive medications, and have multiple effects that can contribute to hyperkalaemia. They inhibit basolateral Na+,-K+-ATPase and they can inhibit apical sodium reabsorption, and may have effects on the apical potassium channels involved in collecting duct potassium secretion. Accordingly, calcineurin inhibitors can, particularly in patients with impaired renal function, contribute to hyperkalaemia.

In addition to specific drugs that affect specific components of potassium secretion, decreased renal function, whether due to acute kidney injury (AKI) or chronic kidney disease (CKD) is frequently observed in individuals with hyperkalaemia. Generally, basal rates of potassium excretion are well preserved as long as the patient does not have severe reduction in urinary output (oliguria < 400 mL/day).

Obstructive uropathy Obstructive uropathy frequently presents with hyperkalaemia. This occurs through multiple mechanisms. First, there is decreased flow through renal tubules, leading to decreased potassium secretion (Batlle et al., 1981; Pelleya et al., 1983; Perez et al., 1983). Chronic obstruction also induces a degree of collecting duct dysfunction involving altered expression of proteins involved in sodium reabsorption and potassium secretion (Batlle et al., 1981; Pelleya et al., 1983). Patients with obstructive uropathy-induced hyperkalaemia may remain hyperkalaemic for several weeks following relief of the obstruction, which reflects the delayed resolution of the collecting duct dysfunction.

Intrinsic renal parenchymal disease Both AKI and CKD are associated with hyperkalaemia (Acker et al., 1998). Diabetic nephropathy and interstitial nephritis are frequent pathological lesions associated with involvement of the medullary interstitium that predisposes to hyperkalaemia. In some patients with modest CKD, the degree of hyperkalaemia is disproportionate to the degree of intrinsic renal disease and hypoaldosteronism secondary to subnormal plasma renin activity has been proposed to explain the acidosis and hyperkalaemia (Schambelan et al., 1972; Perez et al. 1977). Intravascular volume expansion, which may not be detectable by usual physical examination techniques, likely contributes to the hypoaldosteronism; volume expansion increases atrial natriuretic peptide (ANP) release that inhibits aldosterone synthesis (Williams, 2005). The aldosterone analogue, fludrocortisone may improve the hyperkalaemia, but is not recommended as a general therapeutic approach because it stimulates fluid retention, hypertension, and can accelerate the progression of CKD. In this syndrome, frequently referred to as hyperkalaemic RTA, diuretic and alkali therapy generally achieves better results (Rastogi et al., 1985).

Genetic hyperkalaemic syndromes Two rare syndromes exhibit severe, persistent hyperkalaemia, but with widely divergent effects on sodium balance and blood pressure. The first, pseudohypoaldosteronism type 1, is characterized by severe renal sodium chloride wasting, dehydration, hypotension, metabolic acidosis, and hyperkalaemia with normal or elevated plasma aldosterone levels (Cheek and Perry, 1958; Bosson et al., 1986). These patients usually present in infancy and require large sodium intake. There may be both autosomal dominant and autosomal recessive patterns of inheritance (Bosson et al., 1986). Defects in the mineralocorticoid receptor (Bosson et al., 1986) and mutations of a subunit of the epithelial sodium channel (Chang et al., 1996) have been reported. The second syndrome is referred to as pseudohypoaldosteronism type 2 or Gordon syndrome. In 1986, Gordon described 28 patients with refractory hypertension, hyperkalaemia, and normal renal function as assessed by GFR. Both renin and aldosterone values were suppressed, but these patients responded to sodium restriction and diuretic therapy (Gordon, 1986a, 1986b; Gordon and

chapter 34 

Hodsman, 1986). In this syndrome patients frequently present during adolescence or as young adults with severe hypertension. Other features included short statue, intellectual impairment, and muscle weakness. This condition is known to be due to genetic abnormalities in specific intracellular signalling proteins (WNK kinases), resulting in overexpression of the apical sodium chloride cotransporter (NCC) in the distal convoluted tubule. Overexpression causes excessive NaCl reabsorption, leading to intravascular volume expansion and resultant hypertension. Excessive sodium chloride reabsorption also decreases sodium delivery to the collecting duct, which is proposed to blunt collecting duct potassium secretion and lead to the hyperkalaemia. Thiazide diuretics, which are inhibitors of NCC, are a highly effective therapy. (For a more detailed review of WNK kinases, see Hoorn et al. 2011).

Diagnosis of hyperkalaemia The evaluation of hyperkalaemia should begin with the exclusion of pseudohyperkalaemia or a laboratory artefact. The severity of the hyperkalaemia and the risk of cardiotoxicity should be assessed with an ECG. Electrocardiographic changes of hyperkalaemia are potentially an ominous sign and warrant urgent corrective therapy. Comparison with a previous ECG should be done if one is available. If ambiguous, a repeat ECG immediately after calcium gluconate infusion may be informative. Acute cellular potassium release, either from tissue necrosis, rhabdomyolysis, and membrane-depolarizing states, such as succinyl choline and hyperkalaemia periodic paralysis, are usually apparent from the clinical setting. Strenuous exercise frequently leads to mild hyperkalaemia resulting from skeletal muscle potassium release. This will typically resolve spontaneously and in the absence of ECG changes, usually does not require specific therapy. After the acute stabilization of the patient, diagnosis of the factors which contributed to the development of hyperkalaemia should be preceded. As discussed above, hyperkalaemia, if not due to pseudohyperkalaemia or acute redistribution, almost always reflects contributions of impaired renal potassium excretion, either from drugs or decreased renal function, and potentially involving multiple medicines in association with decreased renal function. Excessive potassium dietary intake is also an important contributing cause to consider. Common causes of hyperkalaemia and drugs that impair renal potassium excretion and frequently offending agents are listed in Table 34.5.

Treatment of hyperkalaemia If true hyperkalaemia is present, the potassium content of intravenous fluids and enteral intake should be assessed, and all medicines should be reviewed. With mild hyperkalaemia, these simple measures may be sufficient. For more serious hyperkalaemia, specific treatment is necessary. Therapies for hyperkalaemia include (a)  minimize the cardiac effects of hyperkalaemia, (b)  induce potassium uptake by cells resulting in a decrease in plasma potassium, and (c) remove potassium from the body.

Stabilize membrane potential and antagonize cardiac effects Intravenous calcium administration specifically antagonizes the effects of hyperkalaemia on the myocardial conduction system and on myocardial repolarization (Schwartz, 1978). Intravenous calcium is the most rapid way to treat hyperkalaemia, and is effective even in normocalcaemic patients. Effects can be seen on the ECG

approach to the patient with hypo-/hyperkalaemia within seconds following the administration, and last for 20–30 minutes. A second dose can be given if no effect is seen. Because of the rapid onset of its effect, intravenous calcium administration should be the initial, and temporizing, treatment for individuals with ECG abnormalities related to hyperkalaemia. Rare exceptions include instances of hyperkalaemia induced by cardiac glycosides, or related toxins, in which digoxin-specific antibody treatment is more specific (see below). In some instances, continuous infusion may be helpful if more definitive therapies are delayed. Several precautions should be observed when using intravenous calcium. First, calcium should not be administered in solutions containing NaHCO3 because CaCO3 precipitation can occur. Second, hypercalcaemia which occurs during rapid calcium infusion can potentiate the myocardial toxicity of digitalis. Although it has been traditionally taught that calcium infusion should not be used in cases of digoxin overdose, more recent evidence indicates that treatment with intravenous calcium was not associated with a greater mortality or potentially fatal dysrhythmias when used for patients with life-threatening hyperkalaemia and digoxin toxicity (Levine et al., 2011). Although sodium salts, including NaHCO3, have some of the same effects to stabilize membrane potential and antagonize the effects of hyperkalaemia on the cardiac conduction system, they are not as consistent in their action as intravenous calcium. They may also worsen hyperkalaemia if administered as hypertonic solutions due to their effects on serum osmolality. Their use is not routinely recommended in the acute setting to oliguric patients unless deemed necessary for coexisting metabolic acidosis.

Accelerate cellular potassium uptake The second most rapid method to treat hyperkalaemia is to alter potassium distribution by increasing cellular uptake. Insulin is the most consistent and frequently used hormone to promote cellular potassium uptake, but β2-adrenergic agonist can be used also. Insulin rapidly stimulates cellular potassium uptake by extrarenal cells, primarily hepatocytes and myocytes (Andres et  al., 1962). Insulin, 10 U, should be administered intravenously to ensure rapid and consistent bioavailability, and will begin to affect SK levels within 10–20 minutes, with effects lasting for 4–6 hours (Clausen and Hansen, 1977). Glucose is generally co-administered to avoid hypoglycaemia, but should not be given to hyperglycaemic individuals. Glucose-induced hyperglycaemia can lead to further increases in the potassium concentration due to hypertonicity-induced potassium redistribution. If glucose is not administered, frequent rechecks of the serum glucose level should be performed because of the possibility of insulin-induced hypoglycaemia. In patients with CKD, delayed insulin clearance resulting from the CKD, can lead to more persistent effects of insulin on glucose levels; these patients should also have frequent reassessment of the serum glucose in order to detect possible insulin-induced hypoglycaemia. β2-agonist administration is a second method to stimulate cellular potassium uptake and treat hyperkalaemia. Intravenous albuterol (salbutamol), 0.5 mg, rapidly stimulates prompt potassium uptake, and can decrease SK by approximately 1 mmol/L (Montoliu et al., 1987). Although this agent is not approved for intravenous use in the United States, nebulized β2-agonists can be used. Albuterol, when administered by nebulizer at a dose of 10 or 20 mg decreases SK by 0.6 and 1.0 mmol/L, respectively, with an immediate onset of action and maximal effect at 90–120 minutes (Allon et al.,1989).

333

334

Section 2  

fluid, electrolyte, and renal tubular disorders

The primary limitations of β2-agonist therapy are tachycardia when given intravenous (Montoliu et al., 1987) and lack of response in 20–33% of patients when given by nebulizer (Allon et  al., 1989; Liou et  al. 1994). In addition, albuterol may decrease potassium removal during subsequent haemodialysis (Allon and Shanklin 1995). Of note, the dose of albuterol required is two- to eightfold greater that usually given by nebulizer and 50–100 times the dose administered by metered dose inhalers (Greenberg 1998). In severe hyperkalaemia, combined therapy with insulin and albuterol may be more effective than either alone (Allon and Copkney, 1990). Importantly, these medications have only temporary effects on SK. Within 4–6 hours they are removed from the body by normal metabolism, and their effect is no longer present. Accordingly, their use is indicated only when needed as a temporizing approach before more definitive measures, typically to increase potassium removal, are instituted. Bicarbonate administration cannot be justified in the absence of renal function as a primary treatment for hyperkalaemia, but may be beneficial if there is a coexistent metabolic acidosis. Although early studies demonstrated a substantial benefit of bicarbonate therapy, most of these patients had significant acidosis and residual renal function, in which case bicarbonate administration increases renal potassium excretion. However, in patients without endogenous renal function the changes in SK with intravenous bicarbonate are small and inconsistent (Blumberg et al., 1988; Allon and Shanklin, 1996; Kim 1996). Generally, sodium bicarbonate therapy should be reserved for those patients with intact renal function, metabolic acidosis, and either intravascular volume contraction or normal intravascular volume. In such patients administration of 5% dextrose solutions with the addition of 150 mmol/L sodium bicarbonate (3 amps of sodium bicarbonate in 1L D5W) may correct the acidosis, promote a kaliuresis, and correct the hyperkalaemia.

Enhance potassium removal Removal of potassium from the body is the definitive treatment for hyperkalaemia requires. Table 34.6 summarizes the available options for potassium removal. In many patients, renal potassium elimination may be adequate for treatment of hyperkalaemia. With chronic, mild hyperkalaemia stimulation of renal potassium excretion with either loop or thiazide diuretics may suffice. Diuretics are usually less effective for acute hyperkalaemia because the rate of potassium excretion usually will not be adequate, and most patients with hyperkalaemia

have underlying renal insufficiency as a contributing factor (Acker et al., 1998). If a rapidly reversible cause of renal failure is present, such as obstructive uropathy, treatment of the underlying condition and close assessment of the potassium level in association with continuous ECG observation may be adequate. A second mode of potassium elimination is with the resin, sodium polystyrene sulphonate. This resin exchanges sodium for potassium in the gastrointestinal tract, and allows potassium elimination in the stool. In general, 1 g of sodium polystyrene sulphonate removes approximately 0.5–1.0 mmol of potassium in exchange for 2–3 mmol of sodium. This agent can be administered either orally or per rectum as a retention enema. The rate of potassium removal is relatively slow, requiring approximately 4 hours for full effect. When given orally, sodium polystyrene sulphonate is generally administered with 20% sorbitol to avoid constipation. If given as an enema, sorbitol should usually be omitted because several case reports suggest an association between rectal administration of sodium polystyrene sulphonate with 20% sorbitol and subsequent colonic perforation (Lillemoe et  al., 1987; Gerstman et al., 1992; Rashid and Hamilton, 1997). Animal models suggest that the sorbitol is responsible for the colonic perforation, possibly due to mucosal dehydration related to fluid loss into the colon lumen (Lillemoe et al., 1987). Dialysis should be considered for potassium removal when renal function in absent and hyperkalaemia is persistent or severe despite medical therapy. In this setting, and if vascular access is immediately available (arteriovenous fistula, haemodialysis catheter, etc.) and haemodialysis is also readily available, this modality provides rapid potassium removal but dialysis does not obviate the need for medical treatment until dialysis has commenced. With severe hyperkalaemia, there is an urgency to reduce the plasma potassium concentration, but precipitous reduction can precipitate cardiac arrhythmias (Feig et al., 1981). Thus, the use of a 0 or 1 mmol/L K+ dialysate generally should be avoided to prevent precipitating hypokalaemia. Depending on the patient, their history of cardiac disease and the degree of hyperkalaemia a 3 mmol/L K+ dialysate during the first 1–2 hours of dialysis followed during the remaining time with a 2 mmol/L K+ dialysate, is likely to be both safe and effective. Continuous dialytic modalities, such as peritoneal dialysis or chronic venovenous haemodialysis are effective for chronic hyperkalaemia, but do not remove potassium sufficiently quickly for use in life-threatening hyperkalaemia.

Table 34.6  Treatment of hyperkalaemia Mechanism

Therapy

Dose

Onset

Duration

Antagonize membrane depolarization

Calcium

Calcium gluconate, 10% solution, 10 mL intravenously (IV) over 10 minutes

1–3 minutes

30–60 minutes

Increase cellular potassium uptake

Insulin

Regular insulin, 10U IV; add dextrose, 50%, 50 mL IV if plasma glucose < 250 mg/dL

30 minutes

4–6 hours

β2-adrenergic agonist

Nebulized albuterol (salbutamol), 10 mg

30 minutes

2–4 hours

Sodium polystyrene sulphonate

Kayexalate, 60 g orally, in 20% sorbitol, or Kayexalate, 60 g per retention enema, without sorbitol

1–2 hours

4–6 hours

Haemodialysis

Blood flow and dialysate flow as tolerated; avoid excessive K gradient (e.g. 1 or 0 mmol/L [K+] dialysate) which can cause cardiac ectopy and hypokalaemia).

Immediate

Until dialysis completed

Remove potassium

chapter 34 

It is important for the clinician to recognize that although dialysis is the most rapid method to treat most cases of hyperkalaemia, other treatments should not be delayed while waiting to initiate dialysis. In many instances, dialysis initiation can be substantially delayed such as during non-routine dialysis hours or if vascular access for haemodialysis is not already present. Thus, whereas haemodialysis removes more potassium than either restoration of renal function after chronic obstruction or with administration of sodium polystyrene sulphonate, the time required to institute therapy is frequently greater, and the patient may progress to life-threatening hyperkalaemia if other therapies are not instituted while awaiting dialysis. Specific therapies may be quite valuable and depend on the causes of hyperkalaemia. In cases of digitalis or related toxicity, digoxin-specific antibody Fab fragments administration may be beneficial (Smith et  al., 1982; Marchlinski et  al., 1991). However, impaired excretion of the Fab fragments occurs in patients with concomitant renal insufficiency. In these patients, because of delayed clearance of the Fab fragments, there can be delayed release of digoxin and recurrence of digoxin toxicity. Repeat doses of the Fab fragments may be necessary in this condition. Because serum levels are not interpretable in the presence of antidigoxin antibodies, the decision as to whether to provide additional doses must be based on clinical indications of recurrence of digoxin toxicity. Thus, substantial adjustment may be required by the clinical setting. Relief of urinary tract obstruction may effectively treat the associated hyperkalaemia, but the rate of potassium excretion may be variable, which requires frequent measurement of plasma potassium. Preparation of this chapter supported by funding from the NIH (R01 DK082680 to CSW, and DK045788 to IDW) and from the Department of Veterans Affairs (I01 BX001472-01 to CSW, and I01-BX00818 to IDW).

References Acker, C. G., Johnson, J. P., Palevsky, P. M., et al. (1998). Hyperkalemia in hospitalized patients. Arch Intern Med, 158, 917–24. Allon, M. and Copkney, C. (1990). Albuterol and insulin for treatment of hyperkalemia in hemodialysis patients. Kidney Int, 38, 869–72. Allon, M., Dunlay, R., and Copkney, C. (1989). Nebulized albuterol for acute hyperkalemia in patients on hemodialysis. Ann Intern Med, 110(6), 426–9. Allon, M. and Shanklin, N. (1995). Effect of albuterol treatment on subsequent dialytic potassium removal. Am J Kidney Dis, 26(4), 607–13. Allon, M. and Shanklin, N. (1996). Effect of bicarbonate administration on plasma potassium in dialysis patients: interactions with insulin and albuterol. Am J Kidney Dis, 28(4), 508–14. Andres, R., Baltazan, M. A., and Cader, G. (1962). Effect of insulin on carbohydrate metabolism and on potassium in the forearm of man. J Clin Invest, 41(1), 108–12. Arrambide, K. and Toto, R. D. (1993). Tumor lysis syndrome. Semin Nephrol, 13, 273–80. Atlas, S. A. and Maack, T. (1992). Atrial natriuretic factor. In E. E. Windhager (ed.) Handbook of Physiology, pp. 1577–673. New York: Oxford University Press. Baltazar, R. F., Mower, M. M., Reider, R., et al. (1980). Acute fluoride poisoning leading to fatal hyperkalemia. Chest, 78(4), 660–3. Bartter, F. C., Pronove, P., Gill, J. R., Jr., et al. (1962). Hyperplasia of the juxtaglomerular complex with hyperaldosteronism and hypokalemic alkalosis. A new syndrome. Am J Med, 33, 811–28. Batlle, D. C., Arruda, J. A. L., and Kurtzman, N. A. (1981). Hyperkalemic distal renal tubule acidosis associated with obstructed uropathy. N Engl J Med, 304, 373–80. Bellevue, R., Dosik, H., Spergel, G., et al. (1975). Pseudohyperkalemia and extreme leukocytosis. J Lab Clin Med, 85, 660–4.

approach to the patient with hypo-/hyperkalaemia Bendheim, P. E., Reale, E. O., et al. (1985). Beta-adrenergic treatment of hyperkalemic periodic paralysis. Neurology, 35(5), 746–9. Blachley, J. D. and Knochel, J. P. (1980). Tobacco chewer’s hypokalemia: licorice revisited. N Engl J Med, 302, 784–5. Bloomfield, R. L., Wilson, D. J., and Buckalew, V. M., Jr. (1986). The incidence of diuretic-induced hypokalemia in two distinct clinic settings. J Clin Hypertens, 2(4), 331–8. Blumberg, A., Weidmann, P., Shaw, S., et al. (1988). Effect of various therapeutic approaches on plasma potassium and major regulation factors in terminal renal failure. Am J Med, 85, 507–512. Borra, S., Shaker, R., and Kleinfeld, M. (1988). Hyperkalemia in an adult hospitalized population. Mt Sinai J Med, 55(3), 226–9. Bosson, D., Kuhnle, U., Mees, N., et al. (1986). Generalized unresponsiveness to mineralocorticoid hormones: familial recessive pseudohypoaldosteronism due to aldosterone-receptor deficiency. Acta Endocrinol Suppl (Copenh), 279:376-80. Bowen, L. N., Subramony, S. H., Cheng, J., et al. (2010). Elementary, my dear Dr. Allen: the case of barium toxicity and Pa Ping. Neurology, 74(19), 1546–9. Bradberry, S. M. and Vale, J. A. (1995). Disturbances of potassium homeostasis in poisoning. J Toxicol Clin Toxicol, 33(4), 295–310. Briceland, L. L. and Bailie, G. R. (1991). Pentamidine-associated nephrotoxicity and hyperkalemia in patients with AIDS. DICP, 25(11), 1171–4. Bronson, W. R., DeVita, V. T., Carbone, P. P., et al. (1966). Pseudohyperkalemia due to release of potassium from white blood cells during clotting. N Engl J Med, 274(7), 369–75. Brown, J. J., Fraser, R., Lever, A. F., et al. (1973). Hypertension and secondary hyperaldosteronism associated with a renin-secreting renal juxtaglomerular-cell tumour. Lancet, 2, 1228–32. Brubacher, J., Hoffman, R. S., Bania, T., et al. (1995). Deaths associated with a purported aphrodisiac—New York City, February 1993-May 1995. MMWR Morb Mortal Wkly Rep, 44(46), 853–5, 861. Bubien, J. K., Ismailov, I. I., Berdiev, B. K., et al. (1996). Liddle’s disease: abnormal regulation of amiloride-sensitive Na+ channels by beta-subunit mutation. Am J Physiol, 270(1 Pt 1), C208–13. Chang, S. S., Grunder, S., Hanukoglu, A., et al. (1996). Mutations in subunits of the epithelial sodium channel cause salt wasting with hyperkalaemic acidosis, pseudohypoaldosteronism type 1. Nat Genet, 12, 248–53. Cheek, D. B. and Perry, J. W. (1958). A salt wasting syndrome in infancy. Arch Dis Child, 33, 252–6. Citrin, D., Stevenson, I. H., and O’Malley, K. (1972). Massive digoxin overdose: observations on hyperkalaemia and plasma digoxin levels. Scott Med J, 17(8), 275–7. Clausen, T. and Hansen, O. (1977). Active Na-K transport and the rate of ouabain binding. The effect of insulin and other stimuli on skeletal muscle and adipocytes. J Physiol (Lond), 270(2), 415–30. Clemessy, J. L., Favier, C., Borron, S. W., et al. (1995). Hypokalaemia related to acute chloroquine ingestion. Lancet, 346(8979), 877–80. Cremer, M., Phan-Weston, S., and Jacobs, A. (2010). Recent innovations in oral contraception. Semin Reprod Med, 28(2), 140–6. Cronin, R. E. and Thompson, J. R. (1991). Role of potassium in the pathogenesis of acute renal failure. Miner Electrolyte Metab, 17(2), 100–5. Dagher, G., Vantyghem, M. C., Doise, B., et al. (1989). Altered erythrocyte cation permeability in familial pseudohyperkalaemia. Clin Sci, 77(2), 213–6. DeFronzo, R. A. (1992). Clinical disorders of hyperkalaemia. In D. W. Seldin and G. Giebisch (eds.) The Kidney (2nd ed.), pp. 2279–337. New York: Raven Press, Ltd. Dobyan, D. C., Cronin, R. E., and Bulger, R. E. (1982). Effect of potassium depletion on tubular morphology in gentamicin-induced acute renal failure in dogs. Lab Invest, 47(6), 586–94. Eiam-Ong, S., Dafnis, E., Spohn, M., Kurtzman, N. A., et al. (1993). H-K-ATPase in distal renal tubular acidosis: urinary tract obstruction, lithium, and amiloride. Am J Physiol, 265, F875–80. Ellison, D. H., Velazquez, H., and Wright, F. S. (1985). Stimulation of distal potassium secretion by low luminal chloride in the presence of barium. Am J Physiol, 248, F638–49.

335

336

Section 2  

fluid, electrolyte, and renal tubular disorders

Farese, R. V., Jr., Biglieri, E. G., Shackleton, C. H., et al. (1991). Licorice-induced hypermineralocorticoidism. N Engl J Med, 325, 1223–7. Feig, P. U., Shook, A., and Stearns, R. H. (1981). Effect of potassium removal during hemodialysis on the plasma potassium concentration. Nephron, 27(1), 25–30. Ferrari, P., Obeyesekere, V. R., Li, K., et al. 1996a). The 11 beta-hydroxysteroid dehydrogenase type II enzyme: biochemical consequences of the congenital R337C mutation. Steroids, 61(4), 197–200. Ferrari, P., Obeyesekere, V. R., Li, K., et al. (1996b). Point mutations abolish 11 beta-hydroxysteroid dehydrogenase type II activity in three families with the congenital syndrome of apparent mineralocorticoid excess. Mol Cell Endocrinol, 119(1), 21–4. Freed, M. I., Rastegar, A., and Bia, M. J. (1991). Effects of calcium channel blockers on potassium homeostasis. Yale J Biol Med, 64(2), 177–86. Funder, J. W., Pearce, P. T., Smith, R., et al. (1988). Mineralocorticoid action: target tissue specificity is enzyme, not receptor, mediated. Science, 242, 583–5. Gao, P. J., Zhang, K. X., Zhu, D. L., et al. (2001). Diagnosis of Liddle syndrome by genetic analysis of beta and gamma subunits of epithelial sodium channel—a report of five affected family members. J Hypertens, 19(5), 885–9. Gerstman, B. B., Kirkman, R., and Platt, R. (1992). Intestinal necrosis associated with postoperative orally administered sodium polystyrene sulfonate in sorbitol. Am J Kidney Dis, 20(2), 159–61. Gill, M. A., DuBe, J. E., and Young, W. W. (1977). Hypokalemic, metabolic alkalosis induced by high-dose ampicillin sodium. Am J Hosp Pharm, 34, 528–31. Gitelman, H. J., Graham, J. B., and Welt, L. G. (1966). A new familial disorder characterized by hypokalemia and hypomagnesemia.Trans Assoc Am Physicians, 79, 221–35. Goldfarb, S., Strunk, B., Singer, I., et al. (1975). Paradoxical glucose-induced hyperkalemia. Combined aldosterone-insulin deficiency. Am J Med, 59(5), 744–50. Gordon, R. D. (1986a). The syndrome of hypertension and hyperkalaemia with normal GFR. A unique pathophysiological mechanism for hypertension? Clin Exp Pharmacol Physiol, 13(4), 329–33. Gordon, R. D. (1986b). The syndrome of hypertension and hyperkalemia with normal glomerular filtration rate: Gordon’s syndrome. Aust N Z J Med, 16(2), 183–4. Gordon, R. D. and Hodsman, G. P. (1986). The syndrome of hypertension and hyperkalaemia without renal failure: long term correction by thiazide diuretic. Scott Med J, 31(1), 43–4. Graber, M., Subramani, K., Corish, D., et al. (1988). Thrombocytosis elevates serum potassium. Am J Kidney Dis, 12, 116–20. Greenberg, A. (1998). Hyperkalemia: treatment options. Semin Nephrol, 18(1), 46–57. Greenberg, S., Reiser, I. W., and Chou, S. Y. (1993). Hyperkalemia with high-dose trimethoprim-sulfamethoxazole therapy. Am J Kidney Dis, 22(4), 603–6. Greenberg, S., Reiser, I. W., Chou, S. Y., et al. (1993). Trimethoprim-sulfamethoxazole induces reversible hyperkalemia. Ann Intern Med, 119, 291–5. Harman, R. C., Auditore, J. V., and Jackson, D. P. (1958). Studies on thrombocytosis: I. Hyperkalemia due to release of potassium from platelets during clotting. J Clin Invest, 37, 699–704. Hartmann, R. C. and Mellinkoff, S. M. (1955). The relationship of platelets to the serum potassium concentration. J Clin Invest, 34, 938–42. Hilden, T. and Krogsgaard, A. R. (1958). Low serum potassium level in severe hypertension. Am J Med Sci, 236, 487–91. Hille, B. (1992). Ionic Channels of Excitable Membranes (2nd ed.). Sunderland, MA: Sinauer Associates Inc. Hirsch, I. A., Tomlinson, D. L., Slogoff, S., et al. (1987). Familial pseudohyperkalaemia: inhibition of erythrocyte K+ efflux at 4°C by quinine. (1988). The overstated risk of preoperative hypokalemia. Anesth Analg, 67(2), 131–6. Ho, P. T., Zimmerman, K., Wexler, L. H., et al. (1995). A prospective evaluation of ifosfamide-related nephrotoxicity in children and young adults. Cancer, 76(12), 2557–64.

Ho-Yen, D. O. and Pennington, C. R. (1980). Pseudohyperkalaemia and infectious mononucleosis. Postgrad Med J, 56, 435–6. Holland, O. B. (1995). Primary aldosteronism. Semin Nephrol, 15(2), 116–25. Holten, C. and Peterson, V. P. (1955). Malignant hypertension with increased secretion of aldosterone and depletion of potassium. Br Med J, 2, 419. Hoorn, E. J., Nelson, J. H., McCormick, J. A., et al. (2011). The WNK kinase network regulating sodium, potassium, and blood pressure. J Am Soc Nephrol, 22(4), 605–14. Hsu, I. and Wordell, C. J. (1995). Hyperkalemia and high-dose trimethoprim/sulfamethoxazole. Ann Pharmacother, 29(4), 427–9. James, D. R. and Stansbie, D. (1987). Familial pseudohyperkalaemia: inhibition of erythrocyte K+ efflux at 4°C by quinine. Clin Sci, 73(5), 557–60. Jones, D. P. and Chesney, R. W. (1995). Renal toxicity of cancer chemotherapeutic agents in children: ifosfamide and cisplatin. Curr Opin Pediatr, 7(2), 208–13. Jurkat-Rott, K., Mitrovic, N., Hang, C., et al. (2000). Voltage-sensor sodium channel mutations cause hypokalemic periodic paralysis type 2 by enhanced inactivation and reduced current. Proc Natl Acad Sci U S A, 97(17), 9549–54. Kamel, K. S., Halperin, M. L., Faber, M. D., et al. (1996). Disorders of potassium balance. In B. M. Brenner (ed.) The Kidney (5th ed.), pp. 999–1037. Philadelphia, PA: W. B. Saunders Co. Kassirer, J. P. and Schwartz, W. B. (1966). Correction of metabolic alkalosis in man without repair of potassium deficiency. Am J Med, 40, 19–26. Kelepouris, E. and Agus, Z. S. (1998). Hypomagnesemia: renal magnesium handling. Semin Nephrol, 18(1), 58–73. Kim, H. J. (1996). Combined effect of bicarbonate and insulin with glucose in acute therapy of hyperkalemia in end-stage renal disease patients. Nephron, 72(3), 476–82. Kleyman, T. R., Roberts, C., and Ling, B. N. (1995). A mechanism for pentamidine-induced hyperkalemia: Inhibition of distal nephron sodium transport. Ann Intern Med, 122, 103–6. Knochel, J. P. (1992). Potassium gradients and neuromuscular function. In D. W. Seldin and G. Giebisch (eds.) The Kidney (2nd ed.) pp. 2191–208. New York: Raven Press, Ltd. Knochel, J. P., Dotin, L. N., and Hamburger, R. J. (1972). Pathophysiology of intense physical conditioning in a hot climate. I. Mechanisms of potassium depletion. J Clin Invest, 51(2), 242–55. Kruse, J. A. and Carlson, R. W. (1990). Rapid correction of hypokalemia using concentrated intravenous potassium chloride infusions. Arch Internal Med, 150(3), 613–17. Kunin, A. S., Surawicz, B., and Sims, E. A. H. (1962). Decrease in serum potassium concentration and appearance of cardiac arrhythmias during infusion of potassium with glucose in potassium depleted patients. N Engl J Med, 266, 288. Lachaal, M. and Venuto, R. C. (1989). Nephrotoxicity and hyperkalemia in patients with acquired immunodeficiency syndrome treated with pentamidine. Am J Med, 87(3), 260–3. Larsson, C., Weber, P., and Anggard, E. (1974). Arachidonic acid increases and indomethacin decreases plasma renin activity in the rabbit. Eur J Pharmacol, 28(2), 391–4. Lawson, D. H. (1974). Adverse reactions to potassium chloride. QJM, 43(171), 433–40. Lawson, D. H., Murray, R. M., and Parker, J. L. (1972). Early mortality in the megaloblastic anaemias. QJM, 41, 1–14. Lawson, D. H., Murray, R. M., Parker, J. L., et al. (1970). Hypokalaemia in megaloblastic anaemias. Lancet, 2, 588–90. Lee, F. N., Oh, G., McDonough, A. A., et al. (2007). Evidence for gut factor in K+ homeostasis. Am J Physiol Renal Physiol, 293(2), F541–7. Lehmann-Horn, F., Iaizzo, P. A., and Franke, C. (1991). Altered gating and conductance of Na+ channels in hyperkalemic periodic paralysis. Pflugers Arch, 418, 297–9. Levine, M., Nikkanen, H., and Pallin, D. J. (2011). The effects of intravenous calcium in patients with digoxin toxicity. J Emerg Med, 40(1), 41–6.

chapter 34 

Lichtman, M. A. and Rowe, J. M. (1982). Hyperleukocytic leukemias: rheological, clinical, and therapeutic considerations. Blood, 60(2), 279–83. Liddle, G. W., Bledsoe, T., and Coppage, W. S. (1963). A familial renal disorder simulating primary aldosteronism but with negligible aldosterone secretion. Trans Assoc Am Physicians, 76, 199–213. Lifton, R. P., Dluhy, R. G., Powers, M., et al. (1992). A chimaeric 11b-hydroxylase/aldosterone synthase gene causes glucocorticoid-remediable aldosteronism and human hypertension. Nature, 355, 262–5. Lillemoe, K. D., Romolo, J. L., Hamilton, S. R., et al. (1987). Intestinal necrosis due to sodium polystyrene (Kayexalate) in sorbitol enemas: clinical and experimental support for the hypothesis. Surgery, 101, 267–72. Ling, B. N., Webster, C. L., and Eaton, D. C. (1992). Eicosanoids modulate apical Ca(2+)-dependent K+ channels in cultured rabbit principal cells. Am J Physiol, 263(1 Pt 2), F116–26. Liou, H. H., Chiang, S. S., Wu, S. C., et al. (1994). Hypokalemic effects of intravenous infusion or nebulization of salbutamol in patients with chronic renal failure: comparative study. Am J Kidney Dis, 23(2), 266–71. Macica, C. M., Yang, Y., Hebert, S. C., et al. (1996). Arachidonic acid inhibits activity of cloned renal K+ channel, ROMK1. Am J Physiol, 271(40), F588–94. Makoff, D. L., Da Silva, J. A., and Rosenbaum, B. J. (1971). On the mechanism of hyperkalaemia due to hyperosmotic expansion with saline or mannitol. Clin Sci, 41(5), 383–93. Makoff, D. L., Silva, J. A., Rosenbaum, B. J., et al. (1970). Hypertonic expansion: acid-base and electrolyte changes. Am J Physiol, 218(4), 1201–7. Malnic, G., DeMello Aires, M., and Giebisch, G. (1971). Potassium transport across renal distal tubules during acid-base disturbances. Am J Physiol, 221, 1192–208. Marchlinski, F. E., Hook, B. G., and Callans, D. J. (1991). Which cardiac disturbances should be treated with digoxin immune Fab (ovine) antibody? Am J Emerg Med, 9(2 Suppl 1), 24–8. Maren, T. H., Mayer, E. A., and Wadsworth, B. C. (1954). Carbonic anhydrase inhibition. Bull Johns Hopkins Hosp, 95(4), 199–255. Marinella, M. A. (1995). Case report: reversible hyperkalemia associated with trimethoprim-sulfamethoxazole. Am J Med Sci, 310, 115–17. McClatchey, A. I., Trofatter, J., McKenna-Yasek, D., et al. (1992). Dinucleotide repeat polymorphisms at the SCN4A locus suggest allelic heterogeneity of hyperkalemic periodic paralysis and paramyotonia congenita [published erratum appears in Am J Hum Genet 1992; 51(4):942]. Am J Hum Genet, 50(5), 896–901. Mercado, R. and Michelis, M. F. (1977). Severe sodium depletion syndrome during lithium carbonate therapy. Arch Internal Med, 137(12), 1731–3. Mimran, A., Mourad, G., and Ribstein, J. (1993a). Early systemic and renal responses to nephrectomy in normotensive kidney donors. Nephrol Dial Transplant, 8(5), 448–53. Mimran, A., Ribstein, J., and Sissmann, J. (1993b). Effects of calcium antagonists on adrenaline-induced hypokalaemia. Drugs, 46(Suppl 2), 103–7. Minella, R. A. and Schulman, D. S. (1991). Fatal verapamil toxicity and hypokalemia. Am Heart J, 121(6 Pt 1), 1810–2. Montoliu, J., Lens, X. M., and Revert, L. (1987). Potassium-lowering effect of albuterol for hyperkalemia in renal failure. Arch Internal Med, 147, 713–7. Morita, H., Fujiki, N., Miyahara, T., et al. (2000). Hepatoportal bumetanide-sensitive K(+)-sensor mechanism controls urinary K(+) excretion. Am J Physiol Regul Integr Comp Physiol, 278(5), R1134–9. Mujais, S. K. and Katz, A. I. (1992). Potassium deficiency. In D. W. Seldin and G. Giebisch (eds.) The Kidney: Physiology and Pathophysiology, pp. 2249–78. New York: Raven Press. Mune, T., Rogerson, F. M., Nikkila, H., et al. (1995). Human hypertension caused by mutations in the kidney isozyme of 11 beta-hydroxysteroid dehydrogenase. Nat Genet, 10(4), 394–9. New, M. I., Stoner, E., and DiMartino-Nardi, J. (1986). Apparent mineralocorticoid excess causing hypertension and hypokalemia in children. Clin Exp Hypertens A, 8(4–5), 751–72. O’Brien, J. G., Dong, B. J., Coleman, R. L., et al. (1997). A 5-year retrospective review of adverse drug reactions and their risk factors in human

approach to the patient with hypo-/hyperkalaemia immunodeficiency virus-infected patients who were receiving intravenous pentamidine therapy for Pneumocystis carinii pneumonia. Clin Infect Dis, 24(5), 854–9. Oberfield, S. E., Levine, L. S., Carey, R. M., et al. (1979). Pseudohypoaldosteronism: multiple target organ unresponsiveness to mineralocorticoid hormones. J Clin Endocrinol Metab, 48, 228–34. Oe, H., Taniura, T., and Ohgitani, N. (1998). A case of severe verapamil overdose. Jpn Circ J, 62(1), 72–6. Oh, K. S., Oh, Y. T., Kim, S. W., et al. (2011). Gut sensing of dietary K+ intake increases renal K+ excretion. Am J Physiol Regul Integr Comp Physiol, 301(2), R421–9. Oster, J. R., Perez, G. O., and Vaamonde, C. A. (1978). Relationship between blood pH and potassium and phosphorus during acute metabolic acidosis. Am J Physiol, 235, F345–51. Oster, J. R., Singer, I., and Fishman, L. M. (1995). Heparin-induced aldosterone suppression and hyperkalemia. Am J Med, 98(6), 575–86. Paice, B., Gray, J. M., McBride, D., et al. (1983). Hyperkalaemia in patients in hospital. Br Med J (Clin Res Ed), 286(6372), 1189–92. Papadakis, M. A., Wexman, M. P., Fraser, C., et al. (1985). Hyperkalemia complicating digoxin toxicity in a patient with renal failure. Am J Kidney Dis, 5(1), 64–6. Pelleya, R., Oster, J. R., and Perez, G. O. (1983). Hyporeninemic hypoaldosteronism, sodium wasting and mineralocorticoid-resistant hyperkalemia in two patients with obstructive uropathy. Am J Nephrol, 3(4), 223–7. Perazella, M. A. and Mahnensmith, R. L. (1996). Trimethoprim-sulfamethoxazole: hyperkalemia is an important complication regardless of dose. Clin Nephrol, 46(3), 187–92. Perez, G. O., Lespier, L., Jacobi, J., et al. (1977). Hyporeninemia and hypoaldosteronism in diabetes mellitus. Arch Intern Med, 137(7), 852–55. Perez, G. O., Oster, J. R., and Vaamonde, C. A. (1981). Serum potassium concentration in acidemic states. Nephron, 27, 233–43. Perez, G. O., Pelleya, R., Oster, J. R., et al. (1983). Blunted kaliuresis after an acute potassium load in patients with chronic renal failure. Kidney Int, 24, 656–62. Perlmutter, E. P., Sweeney, D., Herskovits, G., et al. (1996). Case report: severe hyperkalemia in a geriatric patient receiving standard doses of trimethoprim-sulfamethoxazole. Am J Med Sci, 311, 84–5. Perry, M. C., Bauer, J. H., and Farhangi, M. (1983). Hypokalemia in acute myelogenous leukemia. South Med J, 76(8), 958–61. Plassart, E., Reboul, J., Rime, C. S., et al. (1994). Mutations in the muscle sodium channel gene (SCN4A) in 13 French families with hyperkalemic periodic paralysis and paramyotonia congenita: phenotype to genotype correlations and demonstration of the predominance of two mutations. Eur J Hum Genet, 2(2), 110–24. Ptacek, L. J., Tawil, R., Griggs, R. C., et al. (1994). Dihydropyridine receptor mutations cause hypokalemic periodic paralysis. Cell, 77, 863–8. Rabelink, T. J., Koomans, H. A., Hene, R. J., et al. (1990). Early and late adjustment to potassium loading in humans. Kidney Int, 38, 942–7. Ralston, S. H., Lough, M., and Sturrock, R. D. (1988). Rheumatoid arthritis: an unrecognised cause of pseudohyperkalaemia. Br Med J, 297(6647), 523–4. Rashid, A. and Hamilton, S. R. (1997). Necrosis of the gastrointestinal tract in uremic patients as a result of sodium polystyrene sulfonate (Kayexalate) in sorbitol: an underrecognized condition. Am J Surg Pathol, 21(1), 60–9. Rastogi, S., Bayliss, J. M., Nascimento, L., et al. (1985). Hyperkalemic renal tubular acidosis: effect of furosemide in humans and in rats. Kidney Int, 28(5), 801–7. Reza, M. J., Kovick, R. B., Shine, K. I., and et al. (1974). Massive intravenous digoxin overdosage. N Engl J Med, 291(15), 777–8. Rodriguez, M., Solanki, D. L., and Whang, R. (1989). Refractory potassium repletion due to cisplatin-induced magnesium depletion. Arch Internal Med, 149(11), 2592–4. Rosa, R. M., Williams, M. E., and Epstein, F. H. (1992). Extrarenal potassium metabolism. In D. W. Seldin and G. H. Giebisch (eds.) The Kidney: Physiology and Pathophysiology (2nd ed.), pp. 2165–90. New York: Raven Press Ltd.

337

338

Section 2  

fluid, electrolyte, and renal tubular disorders

Schambelan, M., Stockigt, J. R., and Biglieri, E. G. (1972). Isolated hypoaldosteronism in adults. A renin-deficiency syndrome. N Engl J Med, 287(12), 573–8. Schild, L., Canessa, C. M., Shimkets, R. A., et al. (1995). A mutation in the epithelial sodium channel causing Liddle disease increases channel activity in the Xenopus laevis oocyte expression system. Proc Natl Acad Sci U S A, 92(12), 5699–703. Schlanger, L. E., Kleyman, T. R., and Ling, B. N. (1994). K+-sparing diuretic actions of trimethoprim: Inhibition of Na+ channels in A6 distal nephron cells. Kidney Int, 45(4), 1070–6. Schwartz, A. B. (1978). Potassium-related cardiac arrhythmias and their treatment. Angiology, 29(3), 194–205. Schwartz, W. B. (1955). Potassium and the kidney. N Engl J Med, 253(14), 601–3. Sebastian, A., Harris, S. T., Ottaway, J. H., et al. (1994). Improved mineral balance and skeletal metabolism in postmenopausal women treated with potassium bicarbonate. N Engl J Med, 330, 1776–81. Shapiro, S., Slone, D., Lewis, G. P., et al. (1971). Fatal drug reactions among medical inpatients. JAMA, 216(3), 467–72. Shemer, J., Modan, M., Ezra, D., et al. (1983). Incidence of hyperkalemia in hospitalized patients. Isr J Med Sci, 19(7), 659–61. Shils, M. E. (1969). Experimental human magnesium depletion. Medicine, 48, 61–5. Siegel, D., Hulley, S. B., Black, D. M., et al. (1992). Diuretics, serum and intracellular electrolyte levels, and ventricular arrhythmias in hypertensive men. JAMA, 267, 1083–9. Simon, D. B., Karet, F. E., Hamdan, J. M., et al. (1996a). Bartter’s syndrome, hypokalaemic alkalosis with hypercalciuria, is caused by mutations in the Na-K-2Cl cotransporter NKCC2. Nat Genet, 13(2), 183–8. Simon, D. B., Karet, F. E., Rodriguez-Soriano, J., et al. (1996b). Genetic heterogeneity of Bartter’s syndrome revealed by mutations in the K+ channel, ROMK. Nat Genet, 14(2), 152–6. Simon, D. B., Nelson-Williams, C., Bia, M. J., et al. (1996c). Gitelman’s variant of Bartter’s syndrome, inherited hypokalaemic alkalosis, is caused by mutations in the thiazide-sensitive Na-Cl cotransporter. Nat Genet, 12(1), 24–30. Simon, N., Franklin, S. S., Bleifer, K. H., et al. (1972). Clinical characteristics of renovascular hypertension. JAMA, 220, 1209–18. Skinner, S. L. and Adelaide, M. B. (1961). A cause of erroneous potassium levels. Lancet, 1, 478–82. Smith, S. R., Klotman, P. E., and Svetkey, L. P. (1992). Potassium chloride lowers blood pressure and causes natriuresis in older patients with hypertension. J Am Soc Nephrol, 2(8), 1302–9. Smith, T. W., Butler, V. P. J., Haber, E., et al. (1982). Treatment of life-threatening digitalis intoxication with digoxin-specific Fab antibody fragments: experience in 26 cases. N Engl J Med, 307(22), 1357–62. Sodi, R., Davison, A. S., Holmes, E., et al. (2009). The phenomenon of seasonal pseudohypokalemia: effects of ambient temperature, plasma glucose and role for sodium-potassium-exchanging-ATPase. Clin Biochem, 42(9), 813–18. Sopko, J. A. and Freeman, R. M. (1977). Salt substitutes as a source of potassium. JAMA, 238, 608–10. Stacpoole, P. W., Wright, E. C., Baumgartner, T. G., et al. (1992). A controlled clinical trial of dichloroacetate for treatment of lactic acidosis in adults. N Engl J Med, 327, 1564–9. Sterns, R. H., Cox, M., Feig, P. U., et al. (1981). Internal potassium balance and the control of the plasma potassium concentration. Medicine, 60(5), 339–54. Stewart, G. W., Corrall, R. J. M., Fyffe, J. A., et al. (1979). Familial pseudohyperkalaemia. A new syndrome. Lancet, 2(8135), 175–7.

Thompson, J. R., Simonsen, R., Spindler, M. A., et al. (1990). Protective effect of KCl loading in gentamicin nephrotoxicity. Am J Kidney Dis, 15, 583–91. Toussaint, C. and Vereerstraeten, P. (1962). Effects of blood pH changes on potassium excretion in the dog. Am J Physiol, 202, 768–72. Unwin, R. J., Luft, F. C., and Shirley, D. G. (2011). Pathophysiology and management of hypokalemia: a clinical perspective. Nat Rev Nephrol, 7(2), 75–84. Velazquez, H., Perazella, M. A., Wright, F. S., et al. (1993). Renal mechanism of trimethoprim-induced hyperkalemia. Ann Intern Med, 119(4), 296–301. Viberti, G. C. (1978). Glucose-induced hyperkalaemia: a hazard for diabetics? Lancet, 1(8066), 690–1. Wang, P. and Clausen, T. (1976). Treatment of attacks in hyperkalemic familial periodic paralysis by inhalation of salbutamol. Lancet, 1(97953), 221–3. Watson, K. R. and O’Kell, R. T. (1980). Lack of relationship between Mg2+ and K+ concentrations in serum. Clin Chem, 26(3), 520–1. Weiner, I. D. and Wingo, C. S. (1997). Hypokalemia—consequences, causes, and correction. J Am Soc Nephrol, 8(7), 1179–88. Weiner, I. D. and Wingo, C. S. (2010). Endocrine causes of hypertension—aldosterone. In J. Freehally, R. J. Johnson, and J. Floege (eds.) Comprehensive Clinical Nephrology (3rd ed.), pp. 469–76. St. Louis, MO: Saunders. Weizenberg, A., Class, R. N., and Surawicz, B. (1985). Effects of hyperkalemia on the electrocardiogram of patients receiving digitalis. Am J Cardiol, 55(8), 968–73. Wenger, T. L., Butler, V. P. J., Haber, E., et al. (1985). Treatment of 63 severely digitalis-toxic patients with digoxin-specific antibody fragments. J Am Coll Cardiol, 5(5 Suppl A), 118A–23A. Whang, R. (1987). Magnesium deficiency: pathogenesis, prevalence, and clinical implications. Am J Med, 82(3A), 24–9. Whang, R., Whang, D. D., and Ryan, M. P. (1992). Refractory potassium repletion. A consequence of magnesium deficiency. Arch Internal Med, 152(1), 40–5. Whelton, P. K., He, J., Cutler, J. A., et al. (1997). Effects of oral potassium on blood pressure. Meta-analysis of randomized controlled clinical trials. JAMA, 277(20), 1624–32. White, P. C., New, M. I., and Dupont, B. (1987). Congenital adrenal hyperplasia. II. N Engl J Med, 316, 1580–6. Widmer, P., Maibach, R., Kunzi, U. P., et al. (1995). Diuretic-related hypokalaemia: the role of diuretics, potassium supplements, glucocorticoids and beta 2-adrenoceptor agonists. Results from the comprehensive hospital drug monitoring programme, berne (CHDM). Eur J Clin Pharmacol, 49(1-2), 31–6. Williams, G. H. (2005). Aldosterone biosynthesis, regulation, and classical mechanism of action. Heart Fail Rev, 10(1), 7–13. Wingo, C. S. (1989). Reversible Cl dependent K flux across the rabbit cortical collecting tubule. Am J Physiol, 256, F697–F704. Wittoesch, J. H., Jackman, R. J., and McDonald, J. R. (1958). Melanosis coli: general review and a study of 887 cases. Dis Colon Rectum, 1, 172. Wong, K. C., Schafer, P. G., and Schultz, J. R. (1993). Hypokalemia and anesthetic implications. Anesth Analg, 77(6), 1238–60. Wrong, O. (1961). Incidence of hypokalemia in severe hypertension. Br Med J, 2, 419–21. Wrong, O. and Davies, H. E. F. (1959). The excretion of acid in renal disease. QJM, 28, 259.

CHAPTER 35

Approach to the patient with metabolic acidosis or alkalosis Mitchell L. Halperin and Kamel S. Kamel Introduction The concentration of hydrogen ions (H+) in all body compartments is maintained at a very low level, because H+ may bind very avidly to histidine residues in proteins. Binding of H+ to proteins changes their charge to a more positive valence, which might alter their shape, and possibly their functions. Since most proteins are enzymes, transporters, contractile elements, and structural compounds, a change in their function could pose a major threat to survival. The concentration of H+ in body fluids is exceedingly tiny (in the nmol/L range) and, moreover, is maintained within a very narrow range. In the extracellular fluid (ECF) compartment, the [H+] is 40 ± 2 nmol/L, while in the intracellular fluid (ICF) compartment, the [H+] is approximately 80 nmol/L. These are all the more impressive, because an enormous quantity of H+ is produced and removed by metabolism each day. In more detail, acids are obligatory intermediates of carbohydrate, fat, and protein metabolism. For example, since adults typically consume and oxidize 1500 mmol of glucose per day, at least 3000 mmol (3,000,000,000 nmol) of H+ are produced, as pyruvic and/or L-lactic acids in glycolysis. Their oxidation to carbon dioxide (CO2) and water (H2O) removes both the H+ and conjugate bases almost as quickly as they are formed. However, in an adult eating a typical Western diet, a net of 70 mmol (70,000,000 nmol) of H+ are added daily to the body. This implies that there are very effective control mechanisms that minimize fluctuations in [H+].

Acid–base balance An analysis of acid–base balance must consider not only acid balance, but also the balance for bases or alkali (see Chapter 24).

Acid balance There are three major components to consider in the physiology of acid balance (Halperin et al., 1987). First, H+ are produced during the metabolism of sulphur-containing amino acids. Second, H+ are removed from the body largely because they react with bicarbonate ions (HCO3−), forming CO2 and H2O. The CO2 is eliminated via the lungs. The net result of these reactions is a deficit of HCO3 in the body that is equal to the H+ gain. Since H+ cannot be excreted with the sulphate anions, and sulphate anion cannot be metabolized to regenerate HCO3−, the third component of the process to achieve

acid balance is to generate new HCO3− to replace the HCO3− lost in titrating these H+. This is done through metabolism of the amino acid glutamine in the cells of the proximal convoluted tubule (PCT) to yield ammonium (NH4+) and α-ketoglutarate anion. Metabolism of α-ketoglutarate anion to neutral end-products (CO2 or glucose) yields HCO3− that are added to the body. Nevertheless, for a net gain of HCO3−, NH4+ must be made into end products of metabolism by being excreted in the urine (Fig. 35.1). H+ are also produced when dietary phosphates enter the body as monovalent inorganic phosphate (H2PO4). One does not need urinary NH4+ excretion to restore acid balance, because at usual urine pH of about 6, filtered HPO42−anions are excreted in the urine as H2PO4−.

Base-balance The diet also provides alkaline salts; the best example is the ingestion of citrus fruits that contain potassium (K+) plus citrate anions. Metabolism of these citrate anions occurs rapidly in the liver, the net result is the production of HCO3−. Removal of this HCO3− load is achieved by production of a variety of new endogenous organic acids (e.g. citric acid) (Fig. 35.2). The H+ of these acids titrate HCO3− and base balance is maintained by excreting the conjugate base of these acids, for example citrate3−, in the urine as their sodium (Na+), K+, and/or calcium (Ca2+) salts (Cheema-Dhadli et al., 2002). This disposal of alkali with the excretion of organic anions serves also to minimize the risk of forming calcium-containing kidney stones.

Metabolic acidosis Metabolic acidosis is a process that causes a fall in the concentration of HCO3− in plasma (PHCO3) and a rise in the concentration of H+. Metabolic acidosis represents a diagnostic category with many different causes (Table 35.1). The risks for the patient are those due to the underlying disorder that caused the metabolic acidosis, the ill effects due to the binding of H+ to intracellular proteins in vital organs (e.g. the brain and the heart), and possible dangers associated with the anions that accompanied the H+ load (e.g. chelation of ionized calcium (Ca2+) by citrate anions in a patient with metabolic acidosis due to ingestion of a large amount of citric acid) (DeMars et al., 2001).

340

Section 2  

fluid, electrolyte, and renal tubular disorders

Diet

2 1

ECF 2 HCO–3

2 H+

Sulphur-AA

2 CO2 + 2 H2O 2 HCO–3

SO42– SO42–

Urine

Glutamine

3 2 NH4+

2 NH4+

Kidney

Fig. 35.1 H+ balance during the metabolism of sulphur-containing amino acids. Renal events are represented in the shaded area. When sulphur-containing amino acids are converted to SO42− anions in the liver, 2 H+ are produced (site 1). These H+ will react with HCO3− and this produces a deficit of HCO3− in the body (site 2). To achieve H+ and HCO3− balance, an equivalent quantity of new HCO3− must be regenerated and this occurs when the kidney excretes 2 NH4+ per SO42− anion (site 3).

a quantitative estimate of the ECF volume is needed to assess its content of HCO3−. To obtain a quantitative assessment of the ECF volume we use the haematocrit (Napolova et al., 2003), whereas others use the concentration of total proteins in plasma (Love and Phillips, 1969). The assumptions are that the patient does not have a pre-existing anaemia or a low total plasma protein concentration. The haematocrit is the ratio of total red blood cell (RBC) volume (2 L in a normal adult) to blood volume (5 L in a normal adult) (Equation 35.2). For example, if the initial haematocrit is 60%, this implies that the plasma volume is contracted by more than 50%: Normal : 0.40 = 2 L RBC volume / 5 L Blood volume (3 L plasma + 2 L RBC ) Patient : 0.60 = 2 L RBC volume / 3.3 L blood volume (1.3 L plasma + 2 L RBC )

Our goal is to provide a logical bedside approach to the management of the patient with metabolic acidosis (Kamel and Halperin, 2006). The first step is to determine if an emergency is present and to anticipate and prevent threats that may develop during therapy (Fig. 35.3). The concepts that provide the underpinning of our clinical approach are outlined, followed by a discussion of specific causes of metabolic acidosis.

Clinical approach to the patient with metabolic acidosis Concept: the PHCO3 is the ratio of the content of HCO3− in the ECF compartment to the ECF volume (Equation 35.1).

−  HCO3  =

Content of HCO3−in the ECF compartment ECF volume



(35.1)

It is important to distinguish between acidaemia, or a lower plasma pH, and acidosis. Acidaemia may not be present in a patient who has metabolic acidosis. The PHCO3 may not be appreciably low, even though there is a marked decrease in the content of HCO3− in the ECF compartment, if there is a large decrease in the in the ECF volume. Clinical examples of this have been described in a patient with severe diarrhoea (Zalunardo et  al., 2004), and in some patients with diabetic ketoacidosis (DKA) (Napolova et al., 2003). To make a diagnosis of metabolic acidosis in this setting,

Diet

ECF 1

HCO–3

CO2

H+

K++ OA–

2 Glucose

–

K+ OA– Urine

OA 3

OA–

PHCO3 = less renal OA– reabsorption

Fig. 35.2  Overview of base balance. Step 1 is the production of HCO3− from dietary K+ salts of organic anions. In step 2, organic acids were produced and their H+ titrated these new added HCO3−. In step 3, the renal component of the process, is shown in the shaded rectangle. The organic anions (OA−) are made into end-products of metabolism by being excreted in the urine.

(35.2) Concept: H+ must be removed by the bicarbonate buffer system (BBS) to avoid their binding to intracellular proteins.

Buffering of H+ load should not only diminish the degree of acidaemia, but also, more importantly, minimize the binding of H+ to proteins in cells of vital organs (i.e. brain and heart). Binding of H+ to proteins could change their charge, shape, and possibly their functions. To minimize binding of H+ to proteins, H+ removal should be carried out by the BBS, the bulk of which is in the ICF compartment and the interstitial space of skeletal muscle (Vasuvattakul et al., 1992). As shown in Equation 35.3, the BBS is driven by ‘pull’ (i.e. by a lower PCO2 primarily in the interstitial space and ICF of skeletal muscle). H + + HCO3 − → H2CO3 → H2O + CO2 (35.3) Acidaemia stimulates the respiratory centre, which leads to a fall in the arterial PCO2. While the arterial PCO2 sets the lower limit on the PCO2 in capillaries, it does not guarantee that the capillary PCO2 in skeletal muscle will be low enough to ensure effective buffering of H+ by the BBS. Because the free-flowing brachial venous PCO2 reflects the capillary PCO2 in skeletal muscles, it should provide a means of assessing the effectiveness of the BBS in patients with metabolic acidosis (Fig. 35.3) (Gowrishankar et al., 2007). The capillary PCO2 in skeletal muscles will be higher if the rate of blood flow to muscles is low for example as a result of decreased effective arterial blood volume (EABV). In this setting, if muscle oxygen consumption remains unchanged, more oxygen will be extracted from, and more CO2 will be added to, each litre of blood. The higher PCO2 in muscle capillaries will diminish the effectiveness of their BBS to remove extra H+; hence, acidaemia may become more pronounced, with the risk that more H+ will be titrated by intracellular proteins, including critical enzymes in vital organs (Fig. 35.5). At usual rates of blood flow and metabolic work at rest, the brachial venous PCO2 is approximately 6 mmHg > arterial PCO2 (Geers and Gros, 2000). If the blood flow rate to muscles is low, their venous PCO2 will be > 6 mmHg greater than arterial PCO2. Enough saline should be administered to increase the blood flow

chapter 35 

metabolic acidosis or alkalosis

Table 35.1  Causes of metabolic acidosis Acid gain: with the retention of new anions in plasma 1. L-lactic acidosis A. Due predominantly to overproduction of L-lactic acid i. Hypoxic lactic acidosis: Inadequate deliver of O2 (cardiogenic shock, shunting of blood past organs, e.g. sepsis, or excessive demand for oxygen, e.g. seizures) ii. Increased production of L-lactic acid in absence of hypoxia: Overproduction of NADH and accumulation of pyruvate in the liver (e.g. metabolism of ethanol plus a deficiency of thiamine) Decreased pyruvate dehydrogenase activity (e.g. thiamine deficiency, inborn errors of metabolism) Compromised mitochondrial electron transport system (e.g. cyanide, riboflavin deficiency, inborn errors affecting the electron transport system) Excessive degree of uncoupling of oxidative phosphorylation (e.g. phenformin or metformin) B. Due predominantly to reduced removal of L-lactate: liver failure (e.g. severe acute viral hepatitis, shock liver, drugs) C. Due to a combination of reduced removal and overproduction of L-lactic acid Antiretroviral drugs (inhibition of mitochondrial electron transport plus hepatic stenosis) Metastatic tumours (especially large tumours with hypoxic areas plus liver involvement) 2. Ketoacidosis (diabetic ketoacidosis, alcoholic ketoacidosis, hypoglycaemic ketoacidosis including starvation (ketoacidosis), ketoacidosis due to a large supply of short-chain fatty acids (i.e. acetic acid from fermentation of poorly absorbed carbohydrate plus inhibition of acetyl-CoA carboxylase) 3. Renal insufficiency (metabolism of dietary sulphur-containing amino acids and decreased renal excretion of NH4+) 4. Metabolism of toxic alcohols (e.g. formic acid from metabolism of methanol, glycolic acid, and oxalic acid from metabolism of ethylene glycol) 5. D-lactic acidosis 6. Pyroglutamic acidosis Loss of NaHCO3 1. Direct loss of NaHCO3 A. Via the GI tract (e.g. diarrhoea, ileus, fistula) B. Via the urine (proximal renal tubular acidosis or low carbonic anhydrase II activity) 2. Indirect loss of NaHCO3 (low urinary NH4+ secretion): A. Low glomerular filtration rate B. Renal tubular acidosis i. Low availability of NH3 (urine pH ~ 5): Problem in PCT causing low ammoniagenesis: alkaline ICF pH in PCT (e.g. hyperkalaemia) and/or problem in the renal medullary NH3 transfer ii. Defect in net distal H+ secretion (urine pH often ~7): H+ ATPase defect or alkaline α-intercalated cells (a number of autoimmune disorders or disorders with hyper γ-globulinaemia, e.g. Sjögren syndrome) H+ back-leak (e.g. amphotericin B) HCO3 secretion in the collecting ducts (e.g. a molecular defect in the Cl−/HCO3− exchanger leading to its mis-targeting to the luminal membrane of α-intercalated cells (e.g. some patients with Southeast Asian ovalocytosis) iii. Problem with both distal H+ secretion and medullary NH3 transfer (urine pH ~6): Diseases involving the renal interstitial compartment (e.g. sickle cell disease)

rate to muscle to achieve a brachial venous PCO2 that is no more than 6 mmHg greater than arterial PCO2. Initial steps in the clinical approach to the patient with metabolic acidosis: 1. Determine if emergencies are present, anticipate and prevent dangers that may arise during therapy (Fig. 35.3).

2. Determine whether H+ have been buffered adequately by the BBS in skeletal muscle. 3. Determine whether the basis of the metabolic acidosis is due to added acid and/or a deficit of base (NaHCO3) (Figs 35.4 and 35.6).

341

342

Section 2  

fluid, electrolyte, and renal tubular disorders

Metabolic Acidosis Is there a major threat for this patient?

Yes

Proceed to the right side after this evaluation.

No Is buffering adequate by the BBS?

When will this emergency occur?

Before therapy • Haemodynamic emergency • Cardiac arrhythmia • Respiratory failure • Toxins • Metabolic/nutrition issues

Ventilation

During therapy • Shift K+ into cells • Pulmonary oedema • Cerebral oedema in children with DKA • Thiamine deficiency • Rapid correction of chronic hyponatraemia

• Assess with the: arterial PCO2

In muscles • Assess with the: brachial venous PCO2

Fig. 35.3  Initial steps in the clinical approach to a patient with metabolic acidosis.

Metabolic acidosis due to added acids Concept: addition of H+ can be detected by the appearance of new anions. These new anions may remain in the body, and/or be excreted in the urine or diarrhoea fluid.

Acid–base balance is maintained if the new anions are metabolized to neutral end products, or if they are excreted in the urine along with H+ or NH4+. On the other hand, there is a net gain of H+ in the body if these anions are retained in the body or are excreted as their Na+ or K+ salts. The accumulation of new anions in plasma can be detected from a calculation of the anion gap in plasma (PAnion gap) (Emmett and Narins, 1977; Emmett, 2006; Kraut and Madias, 2007) (Equation 35.4). The major cation in plasma is Na+, and the major anions are Cl− and HCO3−. The term PAnion gap is used for the difference between the plasma concentration of Na+ (PNa) and the concentrations of Cl− (PCl) and HCO3− (PHCO3), which reflect the usual excess of the other unmeasured anions in plasma over that of the other unmeasured cations in plasma. This difference is largely due to the net anionic valence on plasma proteins, principally plasma albumin (PAlbumin). If the difference is larger than the ‘normal’ value of the PAnion gap, then other anions are present in plasma. Of note, however, because of difference in laboratory methods (e.g., measurement of PCl), there is a large difference in the mean value for the PAnion gap reported by clinical laboratories. Furthermore, regardless of the laboratory method used, there is a wide range within the normal values of the PAnion gap. The clinician should know what are the normal values PAnion gap for his/her clinical laboratory. Nevertheless, it is difficult to know what is the individual patient base line PAnion gap considering the wide range of normal values. When using this calculation to detect the presence of new anions in plasma, one must adjust the baseline value of the PAnion gap for the PAlbumin. As a rough estimate, the the PAnion gap rises or falls by approximately 2.5 mEq/L for every 10g/L rise or fall in the PAlbumin.

PAnion gap = PNa − ( PCl + PHCO3 ) (35.4) Another approach to detect new anions in plasma was recommended by Stewart (1981); it is based on calculation of strong ions difference. This approach is rather complex and offers only a minor advantage over the PAnion gap in that it includes a correction for the net negative charge on PAlbumin.

Use of delta AG/delta HCO3−

It is widely held that the rise in the concentration of new anions, as reflected by a higher value for the PAnion gap, should be equal to the fall in the PHCO3. This relationship is used to detect the presence of coexisting metabolic alkalosis (the rise in PAnion gap is larger than the fall in PHCO3) and/or the presence of both an ‘acid over-production type’ and a ‘NaHCO3 loss’ type of metabolic acidosis (the rise in PAnion gap is smaller than the fall in PHCO3). There are several pitfalls in using the PAnion gap and this relationship that must be recognized (Halperin and Kamel, 2010). Failure to adjust for changes in the ECF volume: this is an important step, as illustrated in the following paragraphs. Clinical example: a patient with diabetic ketoacidosis (DKA) has a fall in the PHCO3 from 25 to 10 mmol/L and the expected rise in his PAnion gap of 15 mmol/L. This patient had a normal ECF volume of 10 L before DKA developed, but as a result of the glucose-induced osmotic diuresis, his current ECF volume is 7 L. Although, the fall in the PHCO3 and the rise in the concentration of ketoacid anions in plasma (as reflected by the rise in PAnion gap ) are equal, the deficit of HCO3− exceeds the amount of ketoacids added to the ECF compartment. The deficit of HCO3− is 180 mmoles ([10 L × 25 mmol/L] − [7 L × 10 mmol/L]), but the quantity of extra new ketoacid anions in the ECF is only 105 mmoles (7 L × 15 mmol/L). Thus, there is another important explanation for the deficit of HCO3− when ketoacids were added, which is that some of the ketoacid anions were excreted in the urine with K+ and/or Na+ (indirect form of NaHCO3 loss) and

chapter 35 

Skeletal muscle BBS cannot function; hence more H+ bind to proteins Low blood flow rate = high venous PCO2 and this amputates the BBS

CO2

[H+]

metabolic acidosis or alkalosis

The BBS in brain cells fails so more H+ bind to proteins in the brain

HCO–3

HCO–3 + H+

CO2

PTN • H+

PTN • H+

Low blood flow rate = high venous PCO2 and this amputates the BBS CO2

CO2

Fig. 35.4  Consequences of failure of BBS in a patient with metabolic acidemia and a low effective arterial blood volume (EABV). The oval represents a skeletal muscle cell and the circle depicts a cell in the brain. The goal is to remove, as many H+ as possible by the BBS in skeletal muscle in a patient with acidaemia to minimize H+ binding to its proteins. This requires a low PCO2 in muscle ICF and interstitial space. If the patient has a contracted EABV, the PCO2 in ICF and interstitial space in muscle (which is reflected by muscle capillary blood PCO2) may not be low enough for effective buffering of H+ by its BBS. As a result, the degree of acidemia may become more pronounced and more H+ should bind to proteins in the intracellular fluids in other organs, including the brain.

is not reflected by the increase in PAnion gap The PAnion gap did not reveal the actual quantity of H+ that were added during DKA and the fall in PHCO3 did not reflect the actual magnitude of the deficit of HCO3−. However, on re-expansion of the ECF volume with saline, the degree of deficit of HCO3− will become evident. In addition, the fall in the PAnion gap will not be matched by the rise in the PHCO3. Failure to correct for the net negative valence attributable to PAlbumin:  when calculating the PAnion gap, one must adjust for changes in the concentration of albumin as it is the most abundant unmeasured anion in plasma. We emphasize that adjustments should be made for a fall or an increase in the PAlbumin.

Detect new anions in the urine The presence of new anions in the urine can be detected with the calculation of the urine anion gap (UAnion gap) (Equation 35.5) (Halperin et al., 1992). For this calculation, the concentration of NH4+ in the urine (UNH4) is estimated from the urine osmolal gap (UOsm gap), as discussed in the next section. The nature of these new anions may sometimes be deduced by comparing their filtered load to their excretion rate. For example, when there is a very large quantity of new anions excreted in the urine compared with the rise in the PAnion gap, suspect that this anion is secreted in PCT, for example, the hippurate anion from metabolism of toluene (Carlisle et al., 1991), or is freely filtered and poorly reabsorbed by the PCT, for example, the absorption of ketoacid anions may be inhibited by salicylate anions. On the other hand, a very low excretion of new anions suggests that they were avidly reabsorbed in the PCT, for example, L-lactate−

Urine anion gap = (U Na + U K + U NH4 ) − U Cl . (35.5)

Detect toxic alcohols The presence of alcohols in plasma can be detected by finding a large increase in the plasma osmolal gap (POsmolal gap) (Worthley et al., 1987; Kraut and Xing, 2011) (Equation 35.6). This large increase in POsmolal gap occurs because alcohols are uncharged compounds, have a low molecular weight, and because large quantities have been ingested. = Measured POsm − P Osmolal gap

(2 (PNa ) + PGlucose + PUrea , all in mmol units)

(35.6)

Clinical approach to patients with metabolic acidosis due to added acids The steps to follow are illustrated in Fig. 35.5. The causes of metabolic acidosis are listed in Table 35.1.

Hyperchloraemic metabolic acidosis This type of metabolic acidosis is due to the loss of NaHCO3. In this type of metabolic acidosis, very few new anions are present in plasma, hence the term ‘non-anion gap metabolic acidosis’. As the fall in the PHCO3 is matched by a rise in the PCl, this type of metabolic acidosis is also called hyperchloremic metabolic acidosis. There are two major groups of causes for this type of metabolic acidosis (Table 35.1), one is the direct loss of NaHCO3 and the other is an indirect loss of NaHCO3. The direct loss of NaHCO3 may be via the gastrointestinal (GI) tract (e.g. patients with diarrhoea) or the urine in patients at the start of a disease process that causes proximal renal tubular acidosis (pRTA) (see Chapter 36). The indirect loss of NaHCO3 may be due to a low rate of excretion of NH4+ that is insufficient to match the daily rate of production of sulphuric acid produced from the metabolism of sulphur-containing amino acids, (e.g. in patients with chronic renal failure, patients with distal renal tubular acidosis (dRTA)) (Fig. 35.1). Indirect loss of NaHCO3 may also be due to an over-production of an acid (e.g. hippuric acid formed during the metabolism of toluene) with the excretion of its conjugate base (hippurate anions) in the urine at a rate that exceeds the rate of excretion of NH4+ (Carlisle et al., 1991). Concept: in a patient with chronic metabolic acidosis, the expected rate of excretion of NH4+ should be > 200 mmol/day (Simpson, 1983).

To generate new HCO3−, glutamine, must be metabolized in the cells of the PCT to yield NH4+ and the α-ketoglutarate anion. Metabolism of α-ketoglutarate to neutral end-products (CO2 or glucose) produces HCO3− ions that are added to the body. Nevertheless, for a net gain of HCO3−, NH4+ must be made into an end-product of metabolism by being excreted in the urine. Concept:  a low rate of excretion of NH4+ could be due a decreased production of NH4+ or a decreased NH4+ transfer into the urine. The later could be due to a decreased renal medullary interstitial NH3 and/ or a decreased net H+ secretion in the distal nephron.

343

344

Section 2  

fluid, electrolyte, and renal tubular disorders Diagnostic steps Metabolic Acidosis Severe acidaemia over a short period of time? YES

NO New anions in plasma?

• Hypoxic lactic acidosis • Ingestion of ethanol in patients with thiamine deficiency • Ingestion of an acid

NO

YES Is the ECFV very low?

Is NH+4 excretion high?

NO • RTA: see Fig 35.8

NO

YES • Renal insufficiency

High POsm gap? NO

YES

Cl

• Hippurate • Ketoacid anions

Very low GFR?

• Diabetic KA • Alcoholic KA

What anion is excreted with NH+4 ?

Not Cl

NO

YES

YES

• Methanol • Ethylene glycol

• Diarrhoea

• D-lactic acidosis • Non-hypoxic L-lactic acidosis Pyroglutamic acidosis

Fig. 35.5  Steps in the clinical approach to the diagnosis of the cause of metabolic acidosis.

Production of NH4+

Several factors influence the rate of production of NH4+; it is important to recognize that there is a 1–2-day lag period before acidaemia stimulates renal ammoniagenesis. Hypokalaemia also stimulates ammoniagenesis, because it is associated with an intracellular acidosis (Tannen, 1980). The opposite is true for hyperkalaemia. There is an upper limit on the rate of NH4+ production in cells of the PCT set by the rate of regeneration of ATP in these cells (Halperin et al., 1984). ATP is utilized in PCT cells primarily to provide the energy for the reabsorption of filtered

Na+. Patients with a low GFR filter less Na+ and they have a lower rate of reabsorption of Na+ in the PCT, and hence a lower rate of production of NH4+. Patients with isolated proximal renal tubular acidosis (pRTA) have a lower rate of NH4+ production, it was suggested that an alkaline PCT cell pH may be its underlying pathophysiology (Halperin et al., 1989).

Transfer of NH3 into the urine

(See Eladari and Chambrey, 2010; Weiner and Verlander, 2013.) NH4+ produced in cells of the PCT is secreted into its lumen, at

Metabolic acidosis Normal plasma anion gap Is the rate of NH4+ excretion high?

YES

NO

– accompany

Did Cl NH4+ in the urine?

YES • GI loss of NaHCO3

Is the GFR very low?

NO • High acid production and excretion of its conjugate base in urine

YES • Chronic renal insufficiency

NO • Distal or proximal RTA

Fig. 35.6  Initial steps in the clinical approach to a patient with metabolic acidosis and a normal anion gap in plasma.

chapter 35 

least in part, by replacing H+ on the sodium/hydrogen exchanger-3 (NHE-3), making it a Na+/NH4+ exchanger. Reabsorption of NH4+ in the medullary thick ascending limb of the loop of Henle occurs when NH4+ replaces K+ on the Na+/K+/2Cl−cotransporter. This provides the ‘single effect’ for the medullary recycling of NH4+ required for the establishment of a high concentration of NH4+ in the medullary interstitium. Excretion of NH4+ in the urine requires the transfer of NH3 from the medullary interstitial compartment across the membrane of the collecting duct, which occurs non-erythroid Rh glycoproteins, Rhbg and Rhcg, that function as gas channels, plus the secretion of H+ in the lumen of the medullary collecting duct (MCD). H+ secretion in the late distal nephron is mediated primarily by an H+-ATPase, but it may also occur via an H+/K+-ATPase.

Assess the rate of excretion of ammonium in the urine If a direct assay for urine NH4+ is not available, the calculation of the UOsmolal gap provides the best indirect estimate of the UNH4, because it detects all NH4+ salts in the urine (Dyck et al., 1990) (Equation 35.7). We no longer use the urine net charge (or urine anion gap) for this purpose (Kamel and Halperin, 2006). We use the UNH4/ UCreatinine ratio in a spot urine sample to assess the rate of excretion of NH4+. In a patient with chronic metabolic acidosis and the usual rate of excretion of creatinine of 10 mmol/day, the expected renal response is a UNH4/UCreatinine ratio in a spot urine sample of > 20. U

Osmolal gap

Calculated U

= Measured UOsm − calculated U osm

= 2 (U Na + U K ) + U Urea + Osmolality U

Concentration of

Glucose NH +4

, in mM units

in the urine = U

Osmolal gap

/2

(35.7)

Determine why the rate of excretion of ammonium is low The urine pH is not a reliable indicator for the rate of excretion of NH4+ (Richardson and Halperin, 1987). On the other hand, the basis of a low rate of excretion NH4+ may be deduced from the urine pH. A urine pH that is approximately 5 suggests that the defect is a decreased availability of NH3 in the medullary interstitial compartment—low urinary pH due to reduced urinary NH4+ content and buffering capacity. A urine pH that is > 7.0 suggests that its basis is a defect in diminished net distal H+ secretion in the late distal nephron—raised urinary pH due to reduced H+ secretion (Kamel et al., 1997).

Assess distal H+ secretion Hydrogen ion secretion in the distal nephron can be evaluated using the PCO2 in alkaline urine (UPCO2) during bicarbonate loading (Halperin et al., 1974). A UPCO2 that is < 70 mmHg in a second-voided alkaline urine implies that H+ secretion in the distal nephron is likely to be impaired. In patients with low net distal H+ secretion, the UPCO2 can be high if there is a lesion causing a back-leak of H+ from the lumen of the collecting ducts (e.g. use of amphotericin B) or distal secretion of HCO3−, as may occur in some rare patients with South Asian ovalocytosis (SAO) who have a second mutation in the HCO3−/Cl− anion exchanger (AE1) that may lead to its mis-targeting to the luminal membrane of the α-intercalated cell (Kaitwatcharachai et al., 1999). In this setting, the secretion of HCO3− causes the luminal pH to increase, liberating H+ from H2PO4−, which raises the urine PCO2 in alkaline urine to > 70 mm Hg (Kaitwatcharachai et al., 1999). A caveat with

metabolic acidosis or alkalosis

this test is that the UPCO2 is also influenced by the renal medullary concentrating ability and therefore the presence of nephrocalcinosis (see Chapter 36).

Assess PCT cell pH In patients with metabolic acidosis associated with a low capacity to reabsorb filtered HCO3− (e.g. disorders with defects in H+ secretion in the PCT; pRTA), the fractional excretion of HCO3− after infusing NaHCO3 may be measured to confirm this diagnosis (Kamel et al., 1997). In our opinion, this is not needed in most cases, because the results are often not clear (e.g. in a patient with a contracted ECF volume or low PK) and the test can impose a danger (e.g. worsening hypokalaemia in a patient with an already low PK). These patients will be detected clinically by failure to correct their metabolic acidosis, despite being given large amounts of NaHCO3.

Rate of citrate excretion This is a marker of pH in cells of the PCT (Simpson, 1983). The rate of excretion of citrate in children and adults consuming their usual diet is approximately 400 mg (~ 2.1 mmol)/day. The rate of excretion of citrate is very low during most forms of metabolic acidosis, a notable exception is in patients with isolated pRTA, suggesting that its basis is an alkaline PCT cell pH (Halperin et al., 1989).

General comments about alkali therapy in patients with metabolic acidosis The issue of the use of NaHCO3 in the treatment of patients with metabolic acidosis is controversial. In patients with hyperchloraemic metabolic acidosis, administration of NaHCO3 may be necessary, since there are no anions in the ECF that can be metabolized to produce HCO3−. It is interesting to note that some patients with severe diarrhoea due to cholera developed pulmonary oedema when given an amount of saline that was not sufficient to restore their effective arterial blood volume (EABV) (Greenough et  al., 1976). Paradoxically, pulmonary oedema in these patients could be averted by the administration of NaHCO3. Perhaps, the explanation is that these patients developed worsening acidaemia with the administration of saline, which lacks HCO3− and potential HCO3−, and hence may have developed constriction of peripheral veins and an acute increase in central blood volume. In a patient who has metabolic acidosis due to a low rate of excretion of NH4+, metabolic acidosis will persist unless NaHCO3 is given. In this case, one must give enough NaHCO3 to titrate H+ that have accumulated, then maintain the patient on enough alkali to match the daily rate of production of H+ from metabolism of sulphur containing amino acids (usually 20–40 mmol/day in an adult). In patients with acute metabolic acidosis due to added organic acids, there are arguments for and against the use of NaHCO3 (Stacpoole, 1986; Narins and Cohen, 1987; Forsythe and Schmidt, 2000; Sabatini and Kurtzman, 2009; Kraut and Madias, 2012) with strong feelings on both sides, but there is a lack of convincing clinical data.

Arguments for the use of NaHCO3

First, it seems intuitively obvious that at some point during acidaemia, too many H+ may bind to intracellular proteins and compromise vital cellular functions. For example, in vitro studies showed that myocardial contractility and binding of adrenaline to its receptors are decreased when the fall in pH is large (Mitchell et al., 1972; Huang et al., 1995); these adverse effects may be reversed by

345

346

Section 2  

fluid, electrolyte, and renal tubular disorders

lowering the concentration of H+. Notwithstanding, the administration of NaHCO3 does not seem to enhance the contractility of the ischaemic myocardium in dogs in vivo (Mazer et al., 1996). The administration NaHCO3 appeared to be beneficial in the setting of hypoxic L-lactic acidosis in rats, induced by ventilation with a hypoxic gas mixture (Halperin et al., 1996). The survival period in these rats was extended, even though NaHCO3 led to an enhanced rate of production of L-lactic acid. If these data apply to humans with hypoxia-induced L-lactic acidosis, the administration of NaHCO3 might be viewed as a temporary measure to allow for more direct interventions to deal with the underlying cause for the metabolic acidosis to be employed. Of note, a large dose of NaHCO3 was administered in these rats, which is not likely to be feasible in the clinical setting, because of the risk of pulmonary oedema.

Arguments against the use of NaHCO3

When NaHCO3 is administered, HCO3− reacts with H+ and CO2 is produced. Some have argued that this represents an important deleterious effect of HCO3−, because this CO2 can enter cells and cause a paradoxical acidification of the ICF. This is a circular argument, since if the source of the H+ that is titrated by the administered HCO3− is H+ bound to intracellular proteins, the ICF should be alkalinized and not acidified. Alternatively, if the source of these H+ is from the production of ATP in the process of anaerobic glycolysis, stimulation of L-lactic acid production, and ATP generation, by alkali could be beneficial (Halperin et al., 1994). However, a possible danger would be if some of the CO2 produced were retained due to impaired lung function or inadequate mechanical ventilation. Another argument against the use of NaHCO3 to treat patients with a severe degree of metabolic acidosis is that the administration of alkali to rats pretreated with HCl failed to yield an acute and significant restoration of non-bicarbonate buffers (Kamel, 1996). This, however, could have been because these rats were on fixed ventilation and there may have been a rise in tissue PCO2. A fall in concentration of calcium (Ca2+) in plasma can occur with the administration of NaHCO3, because the addition of HCO3− leads to rapid production of carbonate (CO32−) and the precipitation of CaCO3. It has been suggested that a fall in the concentration of ionized calcium in the myocardial interstitial fluid compartment can depress myocardial contractility; but if this were a problem, it should have been obvious in situations where carbonate-bicarbonate (carb-bicarb) buffer (i.e. CO32−) was used. Hypokalaemia could be another problem if the administered NaHCO3 causes an acute shift of K+ into cells, especially if there is already total body K+ depletion, as seen in patients with DKA, dRTA, secretory diarrhoea, and in glue sniffers. If PK is low, the use of NaHCO3 should be delayed until the PK has been raised toward the normal range. Volume overload is only a problem in patients with metabolic acidosis, which has an expanded ECF volume (e.g. patients with renal failure, patients with cardiogenic shock, or those in whom large amounts of alkali are used, e.g. hypoxic lactic acidosis). Hypernatraemia may occur if a large volume of a hypertonic NaHCO3 solution is given. In patients with metabolic acidosis due to overproduction of H+ and anions that can be metabolized to produce HCO3− (e.g. L- or D-lactate, or ß-hydroxybutyrate− (ß-HB−)), and are given a large infusion of NaHCO3, the final PHCO3 can be higher than normal if these anions are retained and metabolized to HCO3−. The

clinical significance of ‘rebound metabolic alkalosis’ is primarily when patients are being weaned from mechanical ventilation, since alkalaemia depresses ventilation, and the impact of the increase in filtered HCO3− on the renal excretion of K+, which can exacerbate hypokalaemia.

Recommendations One must individualize the decision for each patient, balancing potential beneficial versus adverse effects (see later discussion of individual causes of metabolic acidosis). If the decision is made to administer NaHCO3, other issues include how much to give and how fast it should be given. It is important to recognize that as the PHCO3 falls, even more of the added H+ are bound to intracellular proteins (Fig. 35.4). Therefore, when NaHCO3 is given, the hope is that much of it will titrate these intracellular H+, and that they will disappear as CO2 and water, and the increment in the PHCO3 will be small. Moreover, CO2 removal by the lungs must keep up with the increment in CO2 production. A decision needs to be made on the initial target PHCO3 when a patient has an extremely low baseline PHCO3 (e.g. 3  mmol/L). A reasonable target in this setting is either to double the PHCO3 by aiming for an absolute value of 5–6 mmol/L. If this rise in PHCO3 was achieved, and the arterial PCO2 remained unchanged, its pH will rise by 0.3 units. Nevertheless, the amount needed to achieve this goal may be rather large, depending on the ongoing rate of H+ production and amount of added H+ that was buffered by intracellular proteins.

Specific causes for metabolic acidosis The list of causes of metabolic acidosis is provided in Table 35.1.

Ketoacidosis (See Schreiber et al., 1994; Halperin et al., 2002.) The process of production of ketoacids in the liver can be divided into two major steps; first, the formation of acetyl-CoA and second, the conversion of acetyl-CoA to ketoacids (Halperin et al., 2010). There are three substrates from which acetyl-CoA can be made rapidly enough in in hepatic mitochondria to lead to an appreciable rate of formation of ketoacids. The major physiologic function of the metabolic process involving ketoacids is to supply the brain with a water-soluble, fat-derived fuel when its major fuel in the fed state, glucose, is in short supply. The only important physiologic substrate for hepatic ketogenesis is free fatty acids derived from storage fat. In prolonged fasting the PGlucose is low; hence there is a low concentration of insulin in blood delivered to the liver. In the patient with DKA, there is lack of insulin due to damage of ß-cells of the pancreas. In either case, the relative lack of insulin provides the signal to activate the enzyme lipase, which catalyses the release of free fatty acids from triglycerides in adipose tissues. The second substrate for ketoacid formation is ethanol. To permit the liver to remove the maximum quantity of ethanol that is produced from fermentation in the colon, to avoid a disturbance in cerebral function, ketoacids must be the final product of its metabolism. This biochemistry however, may lead to a serious degree of ketoacidosis when a large quantity of ethanol is ingested and insulin levels are low. The third substrate for ketoacids production in the liver is a group of short-chain organic acids, the most abundant of which is acetic acid, that are produced during the fermentation of poorly

chapter 35 

absorbed carbohydrates (fibre or fructose) by bacteria in the colon (Davids et al., 2004). A general rule of regulation of coupled oxidative phosphorylation is that the availability of adenosine diphosphate (ADP), which depends on the rate of utilization of adenosine triphosphate (ATP) to perform biological work, sets an upper limit on the rate of fuel oxidation (Flatt, 1972). Biological work can be mechanical, electrical (ion pumping) and/or biosynthesis. While there are two major fates for acetyl-CoA, formation of ketoacids becomes its major removal pathway when these two pathways are inhibited. Fatty acids synthesis is inhibited because insulin is required for the conversion of acetyl-CoA to fatty acids by activating the enzyme acetyl-CoA carboxylase. Oxidation of long chain fatty acids (e.g. palmitate) in hepatic mitochondria produces acetyl-CoA and converts nicotinamide adenine dinucleotide (NAD+) to its more reduced form, NADH. Hence, one limiting factor for the rate of hepatic ketoacid formation could be the availability of mitochondial NAD+. NADH is converted back to NAD+ during coupled oxidative phosphorylation, which converts ADP into ATP. The availability of ADP depends on the rate of utilization of ATP to perform biologic work. Because the liver, unlike muscle, does not perform mechanical work, and since in the absence of protein ingestion there is not a large enough supply of amino acids to have high rates of gluconeogenesis, the rate of oxidative phosphorylation in the liver would be quite low during conditions associated with ketoacidosis. When the oxidation of acetyl-CoA in the citric acid cycle and its conversion to long chain fatty acids are inhibited, acetyl-CoA is converted to ketoacids.Two molecules of acetyl CoA condense to form acetoacetyl CoA. Acetoacetyl CoA is metabolized to acetoacetic acid in the HMG-CoA pathway. The major ketoacid that is produced by the liver is ß-hydroxybuutyric acid, which is formed from acetoacetic acid in a reaction driven by a high mitochondrial ratio of NADH/NAD+. Nevertheless, the liver needs to produce a high enough quantity of ketoacids for consumption by the brain and the kidney. The observed rates of production of ketoacids during prolonged fasting (~ 1500 mmol/day) would suggest that there are other ways for the liver to bypass the limitation by availability of ADP. One such process is uncoupled oxidative phosphorylation, in which H+ re-enter mitochondria via un-coupler proteins and hence are not linked to the conversion of ADP to ATP.

Removal of ketoacids There are two major sites of ketoacid removal, the brain and the kidneys. The brain oxidizes approximately 750  mmol of ketoacids per day; almost half the quantity of ketoacids produced when ketogenesis is most rapid during prolonged fasting. If the rate of generation of ADP declines in the brain, because of less biological work (e.g. due to coma, intake of sedatives or ethanol, or effects of anaesthesia), fewer ketoacids can be oxidized and the degree of acidaemia will become more severe. The kidneys remove approximately 400 mmol of ketoacids per day. If renal work (largely the reabsorption of filtered Na+) is at its usual rate, the kidneys will oxidize approximately 250 mmol of ketoacids per day. Because more ketoacids are filtered than reabsorbed, approximately 150 mmol of ketoacid anions are excreted per day during the ketoacidosis of prolonged fasting. Since most of these anions are excreted along with NH4+, acid-base balance is maintained. H+ will accumulate if ketoacid anions are excreted in the urine with a cation other than NH4+ (or H+).

metabolic acidosis or alkalosis

In DKA, the filtered load of Na+ declines (due to prerenal failure secondary to the loss of Na+ in the glucose-induced osmotic diuresis). Accordingly, renal removal of ß-HB− and H+ declines, because the rates of ß-HB− oxidation and NH4+ excretion are both reduced. From an energy point of view, oxidation of ß-HB− and glutamine are equivalent in terms of ADP utilization. To summarize, unless there is a much larger degree of uncoupling of oxidative phosphorylation during DKA, the rate of production of ketoacids is not substantially higher than in subjects with ketosis of prolonged fasting. The reason a severe degree of acidaemia develops in a patient with DKA is likely to be due to factors that compromise the rate of removal of generated ketoacids

Diabetic ketoacidosis DKA is the metabolic consequence of insufficient actions of insulin, and it is characterized by the accumulation of glucose and ketoacids in the body. The precipitating illness and the complications of this metabolic disturbance can be life threatening. DKA may be the first presentation of undiagnosed type 1 diabetes mellitus in children. In patients with known type 1 diabetes mellitus, the precipitating causes include gastroenteritis, pancreatitis, infections, and conditions where counter-regulatory hormones may be present in excess (e.g. thyrotoxicosis, surgery, stress, pregnancy, and hyperadrenocorticism). Failure to take insulin can be an important aetiological factor in patients with repeated episodes of DKA. The clinical manifestations of DKA are the expected consequences of the major biochemical changes, hyperglycaemia, glucosuria, and ketoacidosis. Early symptoms represent exacerbations of the classic features of diabetes mellitus that is poorly controlled: thirst, polydipsia, polyuria, weakness, lethargy, and malaise. Hyperglycaemia causes an osmotic diuresis with loss of Na+ and water, resulting in ECF volume contraction, low blood pressure, postural hypotension, and tachycardia. In contrast, a higher PGlucose can also lead to a higher ECF volume, by drawing water out of cells in patients with end stage renal disease. Metabolic acidosis results in an increased rate and depth of breathing (air hunger, Kussmaul respiration). The conversion of acetoacetic acid to acetone imparts the characteristic fruity odour to the breath. Not all the clinical findings, however, are completely explained in terms of these biochemical aberrations. The state of consciousness does not correlate well with the concentration of ketoacids in blood. A much better correlation was found between the level of consciousness and the plasma hyperosmolality. This, however, may also reflect a larger degree of osmotic diuresis and natruresis with a very contracted EABV and hence, possibly a higher PCO2 in cells of the brain which results in more H+ bind intracellular proteins. Another feature of DKA that remains unexplained is hypothermia, even in the presence of infection. This together with the fact that leucocytosis is a common finding in these patients may diminish one’s suspicion of an underlying infection. Anorexia, nausea, vomiting, and abdominal pain are frequent non-specific GI complaints, especially in children. These symptoms, together with abdominal tenderness, decreased bowel sounds, guarding, and leucocytosis, may be severe, mimicking an acute abdominal emergency. Rebound tenderness is usually absent. The cause for the abdominal pain is not entirely clear, but in some cases it may be related to hypertriglyceridaemia and pancreatitis. Signs and symptoms of the disorder that precipitated DKA may dominate the clinical picture.

347

348

Section 2  

fluid, electrolyte, and renal tubular disorders

Laboratory evaluation Hyperglycaemia, ketonaemia, glucosuria, and ketonuria are the four hallmarks of the laboratory diagnosis of DKA. Hyperglycaemia:  the degree of hyperglycaemia varies markedly—the PGlucose usually exceeds 250 mg/dL (14 mmol/L). Higher PGlucose values are seen if there is a large reduction in the GFR due to marked intravascular volume depletion (usually with oliguria), or if the patient has consumed a large quantity of carbohydrates, for example, in the form of sweetened soft drinks to quench thirst (usually with polyuria). Ketoacids: in DKA, serum ketones are usually strongly positive in a dilution of 1 in 8. However, only acetoacetate and acetone yield a positive reaction with the nitroprusside test (Acetest) used for clinical screening for ketoacids. If there is an increase in the NADH/ NAD+ ratio, as occurs with hypoxia or due to metabolism of ethanol, ketoacids will be predominantly in the form of ß-HB acid, which is not detected by Acetest; a specific enzymatic analysis is necessary to measure ß-HB acid. Sodium: owing to glucose-induced osmotic diuresis, patients with DKA have a large deficit of Na+ (usually 3 to 9 mmol/kg body weight). Plasma Na+ concentration: much attention is given to the possibility that glucose will draw water out of cells and thereby lower the PNa by dilution. This, however, occurs only when the addition of glucose to the body is as a solution that is hyperosmolar to plasma. In contrast, when glucose is added as a solution that has an osmolality similar to or lower than that of plasma, there is no shift of water from cells. In this circumstance, the PNa will be lower than seen with hypertonic glucose addition for an identical rise in PGlucose (Davids et al., 2002). Therefore, calculation of the expected fall in PNa for a given rise in PGlucose or of the expected rise in PNa with a fall in PGlucose, based on a shift of water should not be done, because the assumptions made may not be valid. Thus, it is incorrect based on these calculations to justify the use of hypotonic saline to avoid the development of hypernatremia during therapy, since this may increase the risk for the development of cerebral oedema in children with DKA. In our view, the plasma effective osmolality (PEffective Osm) (Equation 35.8) must not be permitted to fall in the first 15 hours of treatment.

( ) + PGlucose (in mmol / L) (35.8)

Effective Posm = 2 PNa

Potassium:  despite a deficit of K+ that is usually in the range of 4–6 mmol/kg body weight, the PK is usually increased to the mid-5 range, because K+ has shifted from the ICF to the ECF compartment due primarily to the lack of action of insulin. PHCO3: in patients with DKA, the PHCO3 is low, because many H+ were added to the ECF along with ß-HB− anions. There is also an indirect loss of Na+ and HCO3− early in the course of DKA, because there is a lag period before there is a large increase in the rate of NH4+ excretion. As a result, ketoacid anions are excreted in the urine with Na+ and/or K+. PCO2:  acidaemia in arterial blood stimulates the respiratory centre and leads to a predictable degree of decrease in the arterial PCO2 (a fall of approximately 1.2 mm Hg (0.16 kPa) in the arterial PCO2 for every 1  mmol/L reduction in PHCO3) (Pierce et al., 1970). As discussed above, while the arterial PCO2 sets a lower limit on the PCO2 in capillaries, it does not guarantee that

the PCO2 in skeletal muscle capillary blood, which reflects the PCO2 in their ICF and the interstitial space, will be low enough to ensure effective buffering of H+ by the BBS. Since most patients with DKA have a decreased EABV, the rate of blood flow to muscles will be low and hence their capillary PCO2 will be higher, which diminishes the effectiveness of their BBS to remove extra H+. As a result, the degree of acidaemia may become more pronounced and more H+ may be titrated by proteins in the ECF and intracellular fluids in other organs, including the brain. Owing to autoregulaion of cerebral blood flow, it is likely that the PCO2 in brain capillary blood will change minimally with all but a severe degree of contraction of EABV. Hence the BBS in the brain will continue to titrate this H+ load. Nevertheless, considering its limited quantity of BBS, and that the brain will receive a relatively larger proportion of blood flow, there is a risk that more H+ will bind to intracellular proteins in the brain. If cerebral autoregulation fails, because of a severe degree of intravascular volume depletion, the PCO2 in capillary blood in the brain should rise and its BBS will fail and an even larger H+ load will bind to proteins in brain cells. GFR:  since patients with DKA often have a very low EABV, their GFR will be reduced and the concentrations of creatinine in plasma (Pcreatinine) will be elevated. There may be errors in the measurement of creatinine, depending on the method used. Higher Pcreatinine values are reported with the picric acid method, if the level of acetoacetate in plasma is elevated, whereas lower Pcreatinine values are reported with severe hyperglycaemia, if the enzymatic assay for creatinine is performed on the Kodak analyser.

Treatment of the patient with DKA DKA is a medical emergency. Mortality is influenced by a number of factors, including precipitating causes, the age of the patient, the level of consciousness, and the severity of the biochemical abnormalities. In children, the leading cause of morbidity and mortality is the development of cerebral oedema (see later in chapter). Other causes of death are infection, vascular thrombosis, and shock. Early diagnosis, a better design of therapy, and dealing with the underlying causes of DKA may reduce mortality. Treat a haemodynamic emergency if present A true haemodynamic emergency is uncommon in children with DKA. A large bolus of intravenous saline can be a risk factor for the development of cerebral oedema, because it increases the hydrostatic pressure and diminishes the colloid osmotic pressure in capillaries in the blood–brain barrier (BBB). If a haemodynamic emergency is present, enough saline should be given to restore haemodynamic stability. In the absence of a haemodynamic emergency, we use the brachial venous PCO2 as a guide to the rate of infusion of saline. Enough saline should be administered to lower the brachial venous PCO2 to a value that is no more than 6 mmHg above the arterial PCO2 (Geers and Gros, 2000). This is important to allow effective buffering of H+ by the BBS to occur in muscle and decrease binding of H+ to intracellular proteins in vital organs. Avoid a large fall in the PEffective osm To prevent this fall in the PEffective osm, the effective osmolality of the infusate should be equal to that of the urine in this polyuric state. A  solution of isotonic saline with 40  mmol KCl per litre (when addition of K+ is needed) has an effective osmolality of close to 400

chapter 35 

mOsm/L. Hypotonic saline should not be used to treat this hyperosmolar state, because this can be dangerous, since it may increase the risk of developing cerebral oedema (see later in this chapter). Replace the Na+ deficit A guide to the total amount of Na + that is needed is obtained by estimating the deficit of Na+ in the ECF compartment on presentation from the PNa and a quantitative assessment of the ECF volume (using the haematocrit). Hence, over-expansion of the ECF volume, which is a common occurrence during therapy in these patients, can be avoided. Stop ketoacid production Insulin plays a central role in arresting ketogenesis, but this is usually not an urgent aspect of therapy, because the maximum possible rate of ketogenesis is only approximately 1 mmol/min (Flatt, 1972). In our view, the only emergency action of insulin needed is its effect to decrease the PK by accelerating a shift of K+ into cells in a patient with a significantly abnormal ECG. The hypoglycaemic effects of insulin are minimal early in therapy. Rather, the PGlucose will fall initially as a result of re-expansion of the ECF volume (dilution) and glucosuria. Six to 8 hours after therapy begins, insulin will lower PGlucose by increasing the rate of oxidation of glucose, because competing fat fuels are no longer available, and by promoting the synthesis of glycogen. A bolus of insulin should not be used in children, because it may lead to brain cell swelling (Carlotti et  al., 2003). Insulin therapy has potentially detrimental side effects that should be anticipated and avoided. The major ones are hypokalaemia (at 1–3 hours) and hypoglycaemia (at 6–10 hours). The former risk will be discussed below; the latter is minimized by infusing glucose when the PGlucose falls to approximately 250 mg/dL (~ 15 mmol/L). K+ therapy KCl should be added to each litre of fluid infused, once insulin is given, if the PK is < 5 mmol/L. If the initial PK is < 4 mmol/L, the patient is profoundly K+ depleted. Because the PK will fall after insulin administration, do not administer any insulin for the first 1–2 hours until PK is raised to around 4  mmol/L with aggressive replacement of K+. These patients are particularly at risk for the development of severe hypokalaemia later during therapy as the rate of excretion of K+ in the urine increases significantly (Carlotti, 2013). NaHCO3 therapy Severe acidaemia may be associated with decreased cardiac contractility, diminished responses to both endogenous and administered catecholamines, and a predisposition to cardiac arrhythmias (Mitchell et al., 1972). In addition, severe acidaemia may interfere with binding of insulin to its receptor, and hence may diminish its action to slow the rate of production of ketoacids (Sonne et al., 1981). Most patients with DKA do not require administration of NaHCO3, because administered insulin will slow the rate of production of ketoacids, and HCO3− will be produced when ketoacids are oxidized. The consensus of opinion is not to administer NaHCO3 to adult patients with DKA unless the plasma pH is close to 6.90. We suggest that this decision in patients with DKA should be individualized and is not based solely on an arbitrary blood pH value. Therapy with NaHCO3 should be considered in the initial treatment of a subset of patients who are at risk for developing

metabolic acidosis or alkalosis

a more severe degree of acidaemia, particularly those who are haemodynamically unstable, before a dangerous fall in plasma pH develops (Kamel and Halperin, 2015). In patients with a very low PHCO3, a quantitatively small additional H+ load will produce a proportionately larger fall in the PHCO3 and plasma pH. The rate of ketoacid oxidation will be diminished if there is a lower rate of work in the brain (e.g. coma, intake of sedatives, including ethanol) and the kidneys (e.g. patients with a very low GFR). The target of NaHCO3 therapy would be at least to avoid a significant fall in the PHCO3. To achieve this goal, the rate of infusion of NaHCO3 should match the expected rate of production of ketoacids by the liver. Based on data from subjects with starvation ketosis, this is approximately 60 mmol/hour (Kamel et al., 1998). We think that this a reasonable start, which can be re-evaluated based on serial measurements of the PHCO3. We would give NaHCO3 at a rate of 1 mmol/min as an infusion in a solution with a similar tonicity to the calculated PEffective osm. In a multicentre, case-controlled retrospective study in paediatric patients with DKA, Glaser et al found that patients and who were treated with NaHCO3 had a significantly greater risk of developing cerebral oedema (Glaser et al., 2001). In our opinion, NaHCO3 should not be administered to children with DKA, unless acidaemia is very severe and haemodynamic instability is unresponsive to the usual manoeuvres to restore blood pressure. Phosphate therapy Patients with DKA are catabolic; they have large deficits of phosphate and of K+. Because the plasma phosphate levels decline markedly once insulin acts, there is a rationale to administer phosphate to patients with DKA. On the other hand, there is no evidence that this alters the course of recovery of patients with DKA and there is the danger of development of hypocalcaemia due to precipitation of calcium with the administered phosphate. Search for and treat an underlying illness Always look for an underlying illness (e.g. an infection) that initiated this metabolic emergency and for complications that may arise during therapy such as venous thrombosis or aspiration pneumonia.

Cerebral oedema during therapy in children with DKA The incidence of cerebral oedema (CE) during therapy for DKA in children remains unacceptably high (0.5–1% of admissions in excellent medical centres) (Sperling, 2006). It is most important to recognize that CE is usually discovered 3–13 hours after treatment begins. This suggests that important risk factors for CE develop during therapy, and hence the current treatment may not be ideal (Brown, 2004). Understanding the pathophysiology of brain cell swelling may provide a framework to address this problem (Carlotti et al., 2003): ◆ Brain

cells swell when there is a higher effective osmolality in brain cells as compared with that in plasma in capillaries near the BBB.

◆ The

interstitial compartment of the ECF of the brain will expand if there is a higher capillary hydrostatic pressure, a lower plasma colloid osmotic pressure, and/or an increase in capillary permeability.

Based on these considerations, we suggest that the following may be risk factors for the development of CE:

349

350

Section 2  

fluid, electrolyte, and renal tubular disorders

An increase in number of effective osmoles in brain cells The Na+/H+ exchanger 1 (NHE-1), which is normally inactive in cell membranes, becomes activated by a high H+ concentration in the ICF and a high insulin concentration in plasma. Following an intravenous bolus of insulin in the presence of a severe degree of acidaemia due to accumulation of monocarboxylic acids, NHE-1 in brain cell membranes could become activated (Van der Meulen, Klip, and Grinstein, 1987). This will lead to a gain of Na+ and/or K+ in, and a loss of H+ from, the ICF compartment. This increases the number of effective osmoles in cells, because the bulk of H+ exported from the cell were bound to ICF proteins (Kamel and Halperin, 2015). A fall in the PEffective Osm This could occur if there is a rapid fall in PGlucose and/or a gain of electrolyte-free water (EFW). A rapid fall in the PGlucose A major factor, which leads to a rapid fall in the PGlucose, is glucosuria due to a rise in GFR following EABV expansion. When the concentrations of ketoacids in plasma decline, glucose becomes the primary brain fuel. Furthermore, more glucose may be oxidized in skeletal muscle, because there are fewer circulating FFA owing to actions of insulin to inhibit hormone-sensitive lipase in adipocytes. Another metabolic pathway for the removal of glucose is its conversion to glycogen in the liver and/or skeletal muscle. Gain of EFW There are several possible sources of EFW in this setting, including the administration of hypotonic saline, and/or D5W to prevent neuroglucopaenia when the PGlucose falls. Another two sources of EFW that are less obvious are: ◆ Gastric

emptying: patients with DKA often consume large volumes of fluid to quench their thirst. This ingested fluid may be retained in the stomach, because hyperglycaemia slows stomach emptying. This, however, will represent a gain of water when absorbed, if water has been ingested or after glucose is metabolized, if fruit juice or sugar-containing soft drinks have been consumed (Carlotti et al., 2009). Rapid absorption of a large volume of water may result in an appreciable fall in arterial PEffective Osm, to which the brain is exposed and may not be detected by the measurement of venous PEffective Osm.

◆ Desalination

of administered isotonic saline:  large volumes of saline are usually administered to patients with DKA. As the excretion of glucose diminishes, and in the presence of vasopressin, this salt load may be excreted in a hypertonic form in the urine, thus generating retained EFW.

Increase ECFV in the brain A large bolus of saline may lead to an increase the interstitial volume of the brain ECF compartment and lead to CE, since it causes an increase in the capillary hydrostatic pressure and a decrease in colloid osmotic pressure. There is evidence to suggest that the BBB may be leaky in patients with DKA.

Clinical implications We suggest the following modifications to the current management of children with DKA (Kamel and Halperin, 2015): 1. Do not administer an intravenous bolus of insulin. 2. Prevent a fall in PEffective Osm. The PEffective Osm must not be permitted to fall in the first 15 hours of treatment. The goal of fluid

therapy should be to raise the PNa by ½ of the fall in PGlucose in mmol/L. To prevent a fall in the PEffective osm, the effective osmolality of the infusate should be equal to that of the urine in this polyuric state. A solution of isotonic saline with 40 mmol KCl per litre has an effective osmolality of close to 400 mOsm/L. As children with DKA often present with near-normal PNa, a degree of hypernatraemia will develop; this, however, is needed to prevent a fall in the PEffective osmolality 3. Use of D10-0.9 % NaCl instead of D5W to minimize the amount of EFW retained after glucose is metabolized. 4. Monitor for signs of gastric emptying. This is suggested by the absence of a large fall in PGlucose when glucosuria is large, a large increase in the urine flow rate with little fall in the PGlucose, or a sudden fall in the PEffective osm. The later occurs if water without sugar was ingested. 5. Avoid overzealous saline administration: a large bolus of saline should be given only if there is a haemodynamic emergency. The goal of saline therapy should be to maintain haemodynamic stability and, as discussed, above to increase blood flow rate to muscle sufficiently to have a brachial venous PCO2 that is close to 6 mmHg (0.8 kPa) higher than arterial PCO2.

Alcoholic ketoacidosis Alcoholic ketoacidosis (AKA) is seen following binge drinking of large amounts of ethanol, complicated by vomiting (usually due to alcohol-induced gastritis) and poor food intake (Halperin et  al., 1983; Wrenn et  al., 1991; Kamel, 1997). The lack of food intake and the EABV depletion due to vomiting (causing the release of α-adrenergics) lead to suppression of insulin secretion (Porte, 1969). Ethanol is metabolized in the cytosol of hepatocytes to produce acetic acid. NAD+ is made into NADH in the metabolism of ethanol by alcohol dehydrogenase and acetaldhyde dehydrogenase. Acetic acid enters the mitochondria and is converted to acetyl-CoA. Oxidation of acetyl-CoA in the citric acid cycle converts NAD+ to NADH. NADH is converted back to NAD+ during coupled oxidative phosphorylation, which converts ADP into ATP. The availability of ADP depends on the rate of utilization of ATP to perform biologic work. As acestyl CoA accumulates, two molecules of acetyl CoA condense to form acetoacetyl CoA. Acetoacetyl CoA is metabolized to acetoacetic acid in the HMG-CoA pathway. Because of the high mitochondrial ratio of NADH/NAD+ due to metabolism of ethanol, the majority of acetoacetic acid is converted to β-hydroxybutyric acid. Therefore, the rate of ethanol metabolism and ketoacids production may be diminished due to decreased availability of NAD+. Nevertheless, a severe degree of ketoacidosis may develop, as the rate of ketoacids by the brain is diminished because of the sedative effect of alcohol (Flatt, 1972; Schreiber et al., 1994). Establishing the diagnosis of AKA may not be straightforward. One reason for this is that there are frequently coexisting acid–base disturbances that may result in the blood pH being normal or even alkalaemic in a substantial number of patients. Metabolic alkalosis commonly occurs as a result of the vomiting, and respiratory alkalosis may occur due to stimulation of ventilation by alcohol withdrawal or because of aspiration pneumonia. Another difficulty in making the diagnosis in patients with alcoholic ketoacidosis is that the NAD+/NADH ratio in the liver is often even more reduced due in part to the metabolism of ethanol. In this setting, more of the ketoacids produced will be in the form of ß-HB acid and hence the

chapter 35 

nitroprusside screening test for ketones may be falsely low (see earlier). The PGlucose is usually only modestly high, rather than being markedly elevated as in DKA. At times it is difficult to distinguish AKA from methanol or ethylene glycol poisoning as the primary cause of acidosis as all of them can cause an elevated value for the plasma POsmolal gap, a high PAnion gap, and a near-normal PGlucose. If the ECF volume is not markedly contracted, one should suspect methanol or ethylene glycol intoxication. A direct assay for methanol and ethylene glycol is needed to establish the diagnosis. Treatment of AKA is usually straightforward. Isotonic saline is required to correct the marked degree of EABV depletion. If the PGlucose is low, a small quantity of glucose should be added to raise the PGlucose to the high-normal range. Reducing α-adrenergic activity with the expansion of the EABV and the higher PGlucose should stimulate insulin secretion and thereby diminish the rate of ketoacid production. Treatment with NaHCO3 is rarely required, because the degree of acidaemia is usually mild and the net production of ketoacids can be reversed quickly with appropriate intravenous fluid therapy. Thiamine must be given with the initial therapy. Attention must also be paid to K+ and phosphate depletion, which are common in this disorder.

L-lactic acidosis A rise in the concentration of L-lactate− and H+ in plasma can be caused by an increased rate of production and/or a decreased rate of removal. Although both of these pathways are involved in most cases, usually one will predominate (Luft, 2001). In virtually every condition of an increased production of L-lactate− from glucose, there is a rise in the concentration of ADP in cells (Halperin et al., 2010). When work is being done, and there is a problem converting ADP to ATP in mitochondria (e.g. hypoxia, uncoupling of oxidative phosphorylation, inhibition of electron transport), a high ADP concentration in the cytosol will drive the glycolytic pathway, which regenerates ATP (Equation 35.9). Since the rate of glycolysis can greatly exceed the rate of oxidation of pyruvate in mitochondria, accumulation of pyruvate and the rise in NADH/NAD+ drives the conversion of pyruvate anions into L-lactate− in a reaction catalysed by the enzyme lactate dehydrogenase (LDH). While approximately 32 mmol of ATP are regenerated from oxidation of 1 mmol of glucose, only 2 mmol of ATP are regenerated from 1 mmol of glucose in anaerobic glycolysis. Nevertheless, it provides the cell with a rapid way to regenerate some ATP. But the price to pay is high, as 1 mmol of L-lactic acid is produced per 1 mmol of ATP formed in this process. Although it seems ‘obvious’ that H+ are produced during flux through the glycolytic pathway, H+ are actually formed during the hydrolysis of ATP when work is performed; this is the initial step that generates ADP and augments the rate of glycolysis:

Work + n ATP → n (ADP + Pi ) + n H 5−

4−

+

Glucose + n ( ADP + Pi ) → n L − lactate− + n ATP 4 − 5−

Sum: Work + Glucose + n ( ADP + Pi ) → 5−



−

4−

n L − lactate + n ATP + n H

+

(35.9)

(n equals any whole number). H+ can accumulate very quickly when the concentration of ADP rises in cells. This marked rise in the concentration of H+ will cause one of the key enzymes in glycolysis, phosphofrucotokinase-1, to lose all of its activity. While this minimizes the drop in intracellular

metabolic acidosis or alkalosis

pH, there is a huge price to pay, since this may lead to an energy crisis, especially in cells of vital organs (e.g. brain). A high rate of glycolysis may occur in the presence of an adequate supply of oxygen if there is a defect in the electron transport system or if there is a very high rate of uncoupling of oxidative phosphorylation, because in these conditions ADP cannot be converted back to ATP quickly enough. The major clinical scenario causing L-lactic acidosis is cardiogenic shock in which inadequate delivery of oxygen to the tissues impairs rapid regeneration of ATP. This type of L-lactic acidosis is known as type A L-lactic acidosis; all other causes are lumped together as type B L-lactic acidosis. We don’t find this classification helpful, since cardiogenic shock is such an obvious clinical diagnosis. Furthermore, it ignores the fact that among patients with type B L-lactic acidosis are those in which the underlying pathophysiology is also due to overproduction of L-lactic acid for reasons other than hypoxia.

Clinical settings with L-lactic acid overproduction Inadequate delivery of O2 The commonest clinical setting for rapid overproduction of L-lactic acidosis is cardiogenic shock. Other examples of conditions that lead to an inadequate delivery of O2 to tissues include acute airway obstruction, haemorrhagic shock, and carbon monoxide poisoning. In patients with sepsis, there can be circulatory disturbances that lead to tissue hypoxia (both decreased delivery of oxygen and impaired extraction of oxygen). In addition to an energy crisis due to failure to regenerate ATP, when L-Lactic acidosis is associated with a decreased EABV and a high capillary PCO2, the BBS will fail to remove enough H+; hence more H+ will bind to intracellular proteins in vital organs (e.g. the brain), which may further impair their functions. The aim of therapy is to increase blood flow and delivery of oxygen to vital organs by whatever means are necessary—no other therapy will save the patient if the cardiac output cannot be significantly improved. Therefore, the crucial issue in therapy is to improve the ability of the patient to regenerate ATP in vital organs, rather than correction of the metabolic acidosis per se. Measures to improve haemodynamics to restore adequate cardiac output and tissue perfusion (e.g. ionotropic agents) are critical, as are means to ensure that blood has an adequate content of oxygen. The use of NaHCO3 during severe hypoxia may be of no value, because of the large magnitude of the H+ load. Nevertheless, NaHCO3 may gain some time to improve myocardial function in cases in which hypoxia is marginal and potentially reversible, but this issue remains controversial. The Na+ load accompanying the HCO3− poses a major limit to this type of therapy in patients with cardiogenic shock and pulmonary oedema. Excessive demand for oxygen L-lactic acidosis due to excessive demand for oxygen occurs during seizures or extreme exercise. Another example is the mini-seizures causing L-lactic acidosis in some patients given isoniazid, a drug commonly used to treat tuberculosis. This may be due to the rapid development of vitamin B6 (pyridoxine) deficiency, because of the formation of an isoniazid-vitamin B6 complex. Pyridoxine is a cofactor for the reaction catalysed by the enzyme glutamic acid decarboxylase, in which glutamate is converted to the inhibitory neurotransmitter γ-amino butyric acid (GABA). Therefore, a deficiency of GABA could result in increased excitability and increased muscle twitching, and at times seizures (Chin et al., 1979). Patients on chronic haemodialysis are at increased risk, because they tend

351

352

Section 2  

fluid, electrolyte, and renal tubular disorders

to be deficient in vitamin B6 due to removal of this vitamin by haemodialysis (Siskind et al., 1993).

Clinical settings with increased production of L-lactic acid in absence of hypoxia Ethanol intoxication The degree of L-lactic acidosis is usually mild (~ 5  mmol/L), because it reflects the higher NADH/NAD+ ratio due to ongoing production of NADH from ethanol metabolism, which is largely restricted to the liver. A more severe degree of L-lactic acidosis suggests that there is L-lactic acid overproduction caused by hypoxia (e.g. shock following GI bleeding), thiamine deficiency, seizures (alcohol withdrawal, delirium tremens, and or a central nervous system (CNS) lesion), and/or L-lactic acid under-utilization due to severe liver disease from an acute alcoholic hepatitis superimposed on chronic liver disease (e.g. fatty liver, cirrhosis). Thiamine deficiency and ethanol intoxication A severe degree of lactic acidosis may develop rapidly in these patients (Shull and Rapoport, 2010). Thiamine (vitamin B1) is a key cofactor for pyruvate dehydrogenase (PDH). The site of L-lactic acid production is likely to be the liver, because it is the site where there is accumulation of pyruvate owing to diminished activity of PDH and a high NADH/NAD+ ratio (due to metabolism of ethanol). Nevertheless, for a severe degree of L-lactic acidosis to develop there must be a high flux in glycolysis. This occurs when the rate of hydrolysis of ATP to ADP and inorganic phosphate to perform work exceeds the rate of regeneration of ATP from ADP in oxidative phosphorylation. As pyruvate accumulates, it gets converted to L-lactate in a reaction catalysed by lactate dehydrogenase, in which NADH is converted to NAD+. The conversion of NADH to NAD+ in the cytosol limits its availability for mitochondrial oxidative phosphorylation and the regeneration of ATP. Thus the concentration of ADP will rise and anaerobic glycolysis will be stimulated in the liver to make ATP. While acidaemia may be severe, damage to the brain is the major concern in these patients. The brain must regenerate ATP as fast as it is being used to perform biologic work. Ketoacids are the preferred fuel if they are present, because they are derived from storage fat and hence proteins from lean body mass are spared as a source of glucose for the brain during prolonged starvation. After successful treatment of the disorder, and as ketoacids become unavailable, the brain must regenerate most of its ATP from the oxidation of glucose, which will be limited by the diminished activity of PDH due to a lack of thiamine. Perhaps of greater significance is the likelihood of an increased demand for ATP regeneration in this setting (e.g. due to delirium tremens or the use of salicylates that may uncouple oxidative phosphorylation in the brain). Hence the concentration of ADP will rise and anaerobic glycolysis will be stimulated in the brain to make ATP. As a result, there will be a sudden rise in the production of H+ and L-lactate anions in areas of the brain where the metabolic rate is more rapid and/or ones that have the lowest reserve of thiamine. Treatment is obviously to administer thiamine early in the course of therapy before the ketoacids concentration in plasma falls. Riboflavin deficiency and the use of tricyclic antidepressants The active metabolites formed from vitamin B2 (riboflavin), flavine mononucleotide (FMN) and flavine adenine dinucleotide (FAD),

are components of the mitochondrial electron transport system, which is the principal pathway to regenerate ATP (Luzzati et al., 1999). Riboflavin deficiency may cause L-lactic acidosis. Riboflavin must be activated via an ATP-dependent kinase to produce FMN and FAD. Tricyclic antidepressant drugs (e.g. amitriptyline and imipramine) inhibit this kinase (Pinto et al., 1981). The activity of this kinase is also decreased in hypothyroidism and L-lactic acidosis may be seen in patients with myxoedema crisis. Uncoupling of oxidative phosphorylation In coupled oxidative phosphorylation, H+ are pumped out from the mitochondrial matrix through the inner mitochondrial membrane using the energy derived from the oxidation of fuels. These H+ re-enter the mitochondrial matrix via special H+ channels in the inner mitochondrial membrane that are linked to the conversion of ADP to ATP. In contrast, if H+ re-enter the mitochondrial matrix through another H+ channel that is not linked to the conversion of ADP to ATP, this is called uncoupled oxidative phosphorylation (Hanstein, 1976). Phenformin is a biguanide drug that is no longer in use because it caused a high incidence of L-lactic acidosis in patients with type 2 diabetes mellitus. This drug has a large hydrophobic end, which allows it to cross the lipid-rich mitochondrial membrane rapidly. In doing so, it brought H+ into the mitochondria quickly unlinked to the conversion of ADP to ATP. Metformin is another biguanide drug, but since it does not have a large hydrophobic tail, it is only a very weak uncoupler of oxidative phosphorylation, and is rarely (in the absence of severe renal impairment in which the drug can accumulate sufficiently to be) a cause of L-lactic acidosis (Lalau, 2010; Salpeter et al., 2010). Acetyl salicylic acid is also an uncoupler of oxidative phosphorylation (Miyahara and Karler, 1965).

Clinical settings with primary slow removal of L-lactic acid This type of L-lactic acidosis does not have the same urgency as the type with primary overproduction of L-lactic acid, because it is not caused by a problem in regenerating ATP. In addition, the rate of H+ accumulation is usually much slower. A chronic steady state of L-lactic acidosis is often present and the causes are a low rate of removal of L-lactic acid usually related to problems with the liver due to hepatitis, replacement of normal liver cells (e.g. by tumour cells or large fat deposits), or destruction of liver from prior hypoxia (e.g. ‘shock liver’). In patients with a malignancy and hepatic metastases, the mechanisms that contribute to the L-lactic acidosis are the replacement of a substantial number of liver cells with tumour cells to impair L-lactic acid removal or production of metabolites by tumour cells such as the amino acid tryptophan, which may inhibit the conversion of pyruvate to glucose in the liver, and/or the fact that ischaemic tumour cells will produce more L-lactic acid. Administration of NaHCO3 to these patients may have detrimental long-term effects, because the load of alkali may increase L-lactic acid production from glucose by de-inhibiting an important rate-controlling enzyme in glycolysis in malignant cells (phosphofructokinase-1). A  considerable amount of lean body mass may be lost if the source of pyruvate is glucose that is made from amino acids (gluconeogenesis) (Fields et al., 1981). Antiretroviral drugs L-lactic acidosis has been reported in patients with HIV infection treated with various anti-retroviral agents. The agent most

chapter 35 

frequently associated with L-lactic acidosis is zidovudine, but didanosine, stavudine, lamivudine, and indinavir have also been implicated. There are two possible mechanisms whereby anti-retroviral agents may cause L-lactic acidosis. First, they block the electron transport system. This may lead to mitochondrial myopathy manifested by ragged-red fibres and mitochondrial DNA depletion (Brinkman et al., 1999). Second, these drugs may lead to replacement of liver parenchyma with storage fat (steatosis) (Coghlan et al., 2001). This view is supported by the fact that in some of these patients the consumption of a small dose of ethanol, which results in a higher NADH concentration in hepatocytes and the diversion of pyruvate to L-lactate, led to a significant increase in the severity of L-lactic acidosis (Gopinath et al., 1992). Methanol intoxication Methanol is methyl alcohol (molecular weight 32); it is used as antifreeze, an additive to gasoline, and a solvent in the manufacture of various drugs. Methanol itself is not toxic, but its metabolic product, formaldehyde, is the major cause of toxicity (Oh et al., 2005), because it rapidly binds to tissue proteins. Methanol is converted to formaldehyde by alcohol dehydrogenase in the liver, but a high concentration of methanol is required for rapid rates of oxidation. Formaldehyde is rapidly converted to formic acid by aldehyde dehydrogenase. In each step, NAD+ is converted to NADH. The metabolic acidosis of methanol poisoning is associated with an increased PAnion gap due to the accumulation of formate and L-lactate anions; the L-lactate level greatly exceeds that of formate. The L-lactic acidosis results from inhibition of cytochrome oxidase by formate and also from the conversion of pyruvate to L-lactate in the liver due to an increased NADH/NAD+ ratio caused by methanol metabolism. HCO3− is regenerated when lactate and formate anions are metabolized to neutral end products; folic acid is a cofactor in formate metabolism. Early on, symptoms of intoxication (e.g. ataxia, and slurred speech) dominate the clinical picture. Later, when methanol is converted to formaldehyde by retinol dehydrogenase, blurred vision and blindness may develop. Abdominal pain, malaise, headache, and vomiting are other findings. Fundoscopic examination may reveal papilloedema. Methanol intoxication should always be considered in the differential diagnosis of metabolic acidosis with an increased PAnion gap, particularly if the ECF volume is not very contracted. This diagnosis should be suspected when there is an elevated POsmolal gap (see earlier) and confirmed by a direct assay for methanol in the blood. Ethylene glycol intoxication Ethylene glycol (molecular weight of 62) is widely used as an antifreeze, as a solvent in the manufacture of paint and plastics, and in the formulation of printers’ inks, stamp pad inks, and inks for ballpoint pens. Ethylene glycol is converted to glycoaldehyde by alcohol dehydrogenase in the liver, the affinity of this enzyme for ethylene glycol is close to 100 times lower than for ethanol; thus, the rate of metabolism of ethylene glycol is rapid only when its concentration is high. Glycoaldehyde is further metabolized to glycolic acid by hepatic aldehyde dehydrogenase, which is the major acid that accumulates in ethylene glycol poisoning (Oh et al., 2005). One per cent or less of glycolic acid is converted to oxalic acid, mainly by the action of the enzyme lactae dehydrogenase. Virtually all oxalate

metabolic acidosis or alkalosis

produced is precipitated as calcium oxalate, contributing to acute renal failure and hypocalcaemia. The major end product of glycolic acid metabolism is glycine via transamination with alanine; vitamin B6 is a cofactor. Ethylene glycol itself causes CNS symptoms such as inebriation, ataxia, and slurred speech. At this stage, the POsmolal gap is high. After a latent period of about 4–12 hours, patients develop nausea, vomiting, hyperventilation, elevated blood pressure, tachycardia, tetany, and convulsions. At this point, metabolic acidosis with an increased PAnion gap is present. The tetany is most likely caused by hypocalcaemia, which is thought to be the result of deposition of calcium oxalate crystals. Renal failure is common and usually develops 36–48 hours after the ingestion of ethylene glycol; glycoaldehyde appears to be the main toxin. The principles of treatment of methanol or ethylene glycol poisoning are virtually identical. They include administration of ethanol to achieve blood concentrations of about 20 mmol/L to reduce metabolism, and removal of these toxic alcohols and their metabolites by haemodialysis. One could administer fomepizole, an inhibitor of hepatic alcohol dehydrogenase, instead of ethanol. The major difference in treating ethylene glycol poisoning is that when acute oliguric renal failure is present, ECF volume overload or pulmonary oedema may limit the amount of NaHCO3 that can be administered, so early institution of dialysis may be important. Salicylate intoxication The major issue with an overdose of aspirin is the toxicity related to the effect of salicylate anions in cells (Oh et al., 2005; Halperin et al., 2010). This may result from direct toxic effects of salicylate on cell functions. It is also possible that this organic acid uncouples oxidative phosphorylation (Miyahara and Karler, 1965). If an increased consumption of O2 and production of CO2 occurs near the respiratory centre, this could stimulate alveolar ventilation and may explain the respiratory alkalosis that is commonly seen in these patients. A modest degree of uncoupling of oxidative phosphorylation can increase the production of ketoacids in the liver. In severe intoxications, the degree of uncoupling of oxidative phosphorylation may be excessive. If this compromises the rate of conversion of ADP to ATP, anaerobic glycolysis is stimulated and a severe degree of L-lactic acidosis develops. Hypoglycaemia is common in patients with salicylate intoxication, which likely reflects increased utilization of glucose by the brain (due to uncoupling of oxidative phosphorylation) and/or impaired gluconeogenesis (Oh et  al., 2005). Toxicity caused by the monovalent salicylate anion occurs when its concentration is 3–5 mmol/L; if the PAnion gap is elevated by a much greater amount, accumulation of ketoacid or L-lactate anions is likely to be present. Treatment is initially aimed at preventing the accumulation of salicylates in brain cells. Salicylic acid (H•SA) is a weak organic acid that is transported across cell membranes in its undissociated form. To appreciate the effect of acidaemia on salicylate toxicity, consider the following example in which the total salicylate concentration in the ECF is 7 mmol/L. Because of its low pK (~ 3.5), only a very tiny fraction is in the H•SA form at normal blood pH value of 7.40 (i.e. H•SA = 0.3 μmol/L). H•SA diffuses across cell membranes until its concentration is equal inside and outside cells. In the cell, the concentration of total salicylate depends on the ICF pH. As ICF pH is normally close to 7.1, the pK of salicylic

353

354

Section 2  

fluid, electrolyte, and renal tubular disorders

acid/salicylate is ~3.5, at H•SA concentration of 0.3 umol/L, the concentration of salicylate anion in cells will be close to 3.5 mmol/L. If the pH in ECF drops to 7.1, the concentration of H•SA will rise from 0.3 to 0.6 μmol/L. Because H•SA diffuses across cell membranes to achieve equilibrium, the intracellular total salicylate concentration will rise from 3.5 to 6.0 mmol/L. Hence, alkalinizing the ECF tends to decrease salicylate accumulation in cells. By the same token, alkalinizing the urine may reduce salicylate reabsorption by PCT cells and markedly enhance its excretion. Dialysis should be instituted if the salicylate blood level exceeds 6  mmol/L (90 mg/dL). If the salicylate blood level exceeds 4 mmol/L (60 mg/dL), dialysis should be considered, particularly if further absorption is anticipated. In patients with an unexplained decreased level of consciousness, dialysis should be started, even at lower levels of salicylate in blood, because of the poor prognosis. Haemodialysis is more efficient for the removal of salicylate, but peritoneal dialysis may be considered if there will be a long delay before haemodialysis can be initiated. In the absence of severe toxicity, the therapeutic efforts in salicylate intoxication are to decrease the concentration of H•SA in blood and to promote the urinary excretion of salicylate with the administration of NaHCO3. Notwithstanding, aggressive therapy with NaHCO3 should be avoided, because patients may become very alkalaemic due to coexistent respiratory alkalosis. Furthermore, patients with salicylate intoxication may have increased capillary permeability and be at risk for pulmonary oedema and cerebral oedema following excessive fluid administration. Acetazolamide, a carbonic anhydrase inhibitor, may be useful in the therapy for salicylate intoxication to increase excretion of salicylate anions in the urine. Its mechanism of action is controversial. The traditional view is that acetazolamide increases the excretion of the salicylate anions by raising the pH in the lumen of the PCT, thereby decreasing the concentration of the undissociated H•SA. However, acetazolamide causes an acid disequilibrium pH (by slowing the conversion of H2CO2 to CO2 and H2O) in the lumen of the PCT, yet it still promotes salicylate excretion. Caution is needed because acetazolamide competes with salicylate anions for binding to plasma albumin, which may increase the free concentration of H•SA in blood. In addition, acetazolamide may induce acidaemia by increasing the excretion of HCO3− in the urine. There is some experimental evidence in humans to suggest that 250 mg of acetazolamide has a tubular effect that lasts for approximately 16 hours (Bayoumi et al., 1993). Therefore, one could use a low dose of acetazolamide instead of alkali therapy in a patient with a high blood pH (i.e. > 7.5) and a modestly elevated blood level of salicylate. Glue sniffing It was thought that glue sniffing causes dRTA, but the high rate of excretion of NH4+ in response to the metabolic acidosis in many of these patients means that they do not have a major defect in renal acid excretion. Patients who sniff glue for its intoxicating properties absorb a significant quantity of toluene (methylbenzene). Toluene is metabolized via a series of reactions in the liver to hippuric acid (Carlisle et  al., 1991). Despite the production of the hippurate anion, the PAnion gap is not significantly elevated, because the kidneys, via filtration and especially via tubular secretion, efficiently excrete hippurate. As a result, there is development of a hyperchloraemic metabolic acidosis. Together with the hippurate anion, Na+ and K+ are excreted in the urine, leading to ECF volume

contraction and hypokalaemia. The presence of NH4+ and hippurate in the urine could be detected by the presence of an appreciable UOsmolal gap ( see above). When the inhalation of toluene stops, ultimately the production of hippuric acid will be diminished, but there can be a lag of 1–3 days, because of the large volume of distribution of toluene. With regard to treatment, hypokalaemia and ECF volume contraction need to be corrected with the administration of KCl and saline. If metabolic acidosis is severe, consider giving NaHCO3, because there is no appreciable amount of hippurate anions present in the body that can be metabolized to HCO3−. The major caveat to the use of NaHCO3 is coexisting K+ depletion, which could be severe, leading to the risk of a cardiac arrhythmia. Thus the PK must be raised first to the mid-3 mmol/L range before NaHCO3 is given.

D-lactic acidosis Certain bacteria in the GI tract may convert carbohydrate (cellulose and fructose) into organic acids (Halperin and Kamel, 1996; Uribarri et al., 1998). Three factors that make this possible are: slow GI transit (blind loops, obstruction), change of the normal gut flora (usually prior antibiotic therapy), and the supply of carbohydrate substrate to these bacteria (foods containing cellulose or fructose). The most prevalent organic acid that is produced in this process is D-lactic acid. Although humans lack the enzyme D-lactate dehydrogenase, metabolism of D-lactate occurs via the enzyme D-2-hydroxyacid dehydrogenase. Therefore, humans metabolize this D-isomer more slowly than L-lactate, but since the rate of production of these organic acids is not rapid, the degree of acidaemia is usually not severe, unless there is also a defect in the excretion of NH4+. There are three additional points that should be noted with respect to D-lactic acidosis. First, the usual clinical laboratory test for lactate is specific for the L-lactate isomer and so the laboratory measurement for lactate will not be elevated in this setting. Second, GI bacteria produce amines, mercaptans, and other compounds that may cause the clinical signs and symptoms related to CNS dysfunction (personality changes, gait changes, confusion, etc.). Third, some of the D-lactate anions may be lost in the GI tract or in the urine (if the GFR is not too low) and the rise in the PAnion gap may not be as high as expected for the fall in the PHCO3. Treatment should be directed at the GI problem. The oral intake of fructose and complex carbohydrates should be decreased. Antacids and oral NaHCO3 should be avoided, because they may lead to a higher intestinal luminal pH and a higher rate of production of organic acids and other toxic products of fermentation. Insulin may be helpful by lowering the rate of oxidation of fatty acids and hence permitting a higher rate of oxidation of these organic acids. Poorly absorbed antibiotics (e.g. vancomycin) can be used to change the bacterial flora in the GI tract.

Pyroglutamic acidosis Pyroglutamic acidosis (PGA) was previously thought to represent a rare inborn error of metabolism in the glutathione synthesis pathway (a defect in 5-oxoprolinase or in glutathione synthase). A number of cases of PGA have been reported, particularly in critically ill patients (Dempsey et al., 2000; Mizock and Mecher, 2000). Of interest, the majority of these cases occur in women, the reason for this is not clear. Glutathione (GSH) is made up of three amino acids: glutamate, cysteine, and glycine. It is the sulfhydryl moiety of cysteine that endows this compound with its ability to detoxify

chapter 35 

metabolic acidosis or alkalosis

reactive oxygen species (ROS). In this process, the reduced form of glutathione (GSH) is converted to its oxidized form (GS-SG). Key to understanding the basis for the accumulation of PGA is that GSH inhibits the enzyme γ-glutamylcysteine (γ-GC) synthase, which catalyses the first step in the cycle that converts glutamate to γ-GC (Fig. 35.7). Hence, when ROS accumulate, as in a patient with sepsis, the concentration of GSH declines, and its inhibitory effect on γ-GC synthase is removed. This results in accelerated formation of γ-glutamyl phosphate. If the patient is cysteine deficient, γ-GC will not be formed, instead, γ-glutamyl phosphate will be transformed to PGA by the enzyme cyclotransferase (Emmett, 2014). A number of drugs have been identified as causes of PGA accumulation. N-acetyl-p-benzoquinonimide (NAPBQ I), a highly reactive metabolite of acetaminophen, depletes GSH. Thus, the feedback inhibition of the enzyme γ-GC synthase is removed, and PGA develops as above (Fenves et al., 2006). Certain drugs (e.g. the antibiotic flucloxacillin and the anticonvulsant, vigabatrin) inhibit 5-oxoprolinase, which converts PGA to glutamate (Fig. 35.6). The major danger in patients with PGA related to sepsis is tissue damage by ROS. While NaHCO3 may be needed, treatment of sepsis is the key issue. Drugs that may cause PGA should be discontinued. N-acetyl cysteine should be given in patients with acetaminophen overdose and is also likely beneficial in patients with PGA related to sepsis. Malnutrition is thought to be a risk factor for the development of PGA and so nutritional support must be provided. Riboflavin deficiency may be important in this regard, because FMN and FAD are cofactors in the reaction catalysed by glutathione reductase, which mediates the conversion of oxidized GS-SG to GSH.

Diet alkali: the alkali load is derived mainly from dietary fruit and vegetables (e.g. K+ plus organic anions (OA−)), and it is normally eliminated as K+ plus OA− in the urine if the K+ can be excreted. Once acidaemia develops, the excretion of alkali, as potential HCO3−, declines and the dietary HCO3− load may be retained. In this setting, these HCO3− ions would titrate many of the H+ from H2SO4 produced from the metabolism of sulphur-containing amino acids, and thereby diminish the net H+ load that must be eliminated each day by the excretion of NH4+. Because patients with renal insufficiency are usually placed on a low K+ diet, they eat less alkali and, as a result, are more likely to become acidaemic. The PAnion gap usually does not rise appreciably until the GFR has fallen to < 20 mL/min. The high PAnion gap in patients with renal insufficiency does not represent the production of an unusually high amount of new acids; rather it is due to the low GFR and accumulation of SO42− and phosphate anions. Experimental evidence from studies in rats strongly suggests that acidaemia is a catabolic signal in uraemia, although evidence from human data is less robust (Weinstein et al., 2004). NaHCO3 supplementation has also been shown to slow the progression of chronic kidney disease (Yaqoob, 2010). It is now recommend that acidaemia in patients with chronic kidney disease should be corrected. After the PHCO3 is corrected, the dose of NaHCO3 needed to maintain PHCO3 in the normal range is usually < 30–40 mmol/day (i.e. enough to titrate the acid load produced from the metabolism of dietary sulphur-containing amino acids). This salt load should not represent a problem to most patients with chronic kidney disease.

Chronic renal failure

Diarrhoea

When the GFR is markedly reduced, the following changes can be expected. NH4+ excretion: the excretion of NH4+ declines markedly when there is a low availability of ADP in PCT cells (lower filtered load of Na+, so less renal work). The quantity of H+ retained can be as high as 30–40  mmol/day, if dietary protein intake is not reduced and excretion of NH4+ is very low. Flucoxacillin

Secretion of HCO3− by the pancreas and small intestine

This secretion is stimulated by the load of H+ from the stomach and is approximately 100 mmol/day. Hence, there is only a modest daily net deficit of NaHCO3 if most of these upper GI secretions are lost.

Glutamate –

Vigabatrin

There are two major sites where HCO3− are added to the lumen of the GI tract and are possible sites for its loss (Field, 2003).

ATP

5-Oxoprolinase Pyroglutamic acid

γ -gluyamyl-cyclotransferase γ -Glutamyl-phosphate Cysteine CysteinyI-gIycine

γ -Glutamyl-cysteine Glycine

Glutathione-SH Inactivated Acetaminophen → NAPBQI

GS-SG (oxidized form)

Fig. 35.7  Production of pyroglutamic acid. When there are low levels of reduced glutathione (e.g. due to depletion by NAPBQI, a highly reactive metabolite of acetaminophen, or when ROS accumulate, as in a patient with sepsis), its inhibitory effect on γ-GC synthase is removed. This results in accelerated formation of γ-glutamyl phosphate. If the patient is cysteine deficient, γ-GC will not be formed, instead, γ-glutamyl phosphate will be transformed to PGA by the enzyme cyclotransferase. In addition, if 5-oxyprolinase is inhibited (e.g. by flucloxacillin, vigabatrin), pyroglutamic acid will accumulate.

355

356

Section 2  

fluid, electrolyte, and renal tubular disorders

Therefore, metabolic acidosis in these patients is likely to be mild, unless the duration of these losses is prolonged and/or there is also another disorder that diminishes the rate of excretion of NH4+ in the urine. Metabolic acidosis may also develop if the fluid rich in NaHCO3 is retained in the lumen of the intestine (e.g. ileus).

Secretion of HCO3− by the late small intestine and the colon

Two luminal transport mechanisms are involved in this process, an NHE and a Cl−/HCO3− anion exchanger (AE). Normally, the maximum transport capacity of NHE is less than that of AE. If there is a large delivery of Na+ and Cl− to the colon, due to osmotic or secretory diarrhoea, the net effect, as more Cl− is exchanged with HCO3− than Na+ is exchanged for H+, is a loss of NaHCO3 in stool, which may lead to the development of acidaemia. On the other hand, if there is a large loss of NaCl in stool and a large decrease in the ECF volume, a significant degree of acidaemia is less likely to occur. In some patients who have diarrhoea, there may be overproduction of organic acids in the colon, such as D-lactic acid. The degree of metabolic acidosis may be more severe, if there is also a low rate of excretion of NH4+ due to a low the GFR resulting from a contracted EABV. For treatment, one must first identify and treat emergencies that may be present on admission (e.g. haemodynamic instability), as well as anticipate and avoid those that might develop with therapy (e.g. hypokalaemia). It is interesting to note that some patients with severe diarrhoea due to cholera developed pulmonary oedema when given an amount of saline that is not sufficient to restore their EABV (Greenough et  al., 1976). Paradoxically, pulmonary oedema in these patients was cured with the administration of NaHCO3. The explanation may be that these patients developed worsening acidaemia with the administration of saline, which results in venoconstriction and an acute increase in central blood volume. Thus, administration of NaHCO3 (or Na+ with an anion that can be metabolized to produce HCO3−, such as lactate) may be necessary in patients with diarrhoea, since there are no retained anions that can be metabolized to produce HCO3−. Enhancing the absorption of NaCl secreted by the intestinal tract diminishes the volume of diarrhoea fluid. This can be achieved by giving oral rehydration therapy (ORT) with an equimolar solution of glucose and NaCl. In more modern versions of this solution, a form of potential HCO3− is added by, for example, replacing 25 to 50  mmol of Cl− with citrate anions that can be metabolized to HCO3−.

Proximal renal tubular acidosis (See Chapter 36.) Proximal RTA may occur as an isolated defect or as part of more generalized PCT cell dysfunction (Fanconi syndrome with glucosuria, phosphaturia, uricosuria, citraturia, and aminoaciduria) (Haque et al., 2012). The major causes of pRTA in adults are paraproteinaemias (e.g. patients with multiple myeloma) and use of carbonic anhydrase inhibitors (e.g. acetazolamide, the antiepileptic drug topiramate). In contrast, cystinosis and the use of ifosfamide are the most common causes of pRTA in children. Hereditary isolated pRTA is a rare autosomal recessive disease that can present with ocular abnormalities such as band keratopathy, cataract and glaucoma. Mutations in the gene encoding for the Na (HCO3)32− cotransporter (NBC1) have been identified in these families (Igarashi et al., 2002).

The initial mechanism for the acidosis in patients with pRTA is the loss of HCO3− in the urine. Once a steady state supervenes, chronic metabolic acidosis is sustained, because the rate of NH4+ excretion is lower than expected for the presence of acidaemia (Kamel et al., 1997). Because the high rate of excretion of citrate decreases the urinary concentration of ionized Ca2+, calcium stones are not usually seen in patients with pRTA. Patients on acetazolamide or topiramate may develop calcium phosphate stones. From a therapeutic standpoint, the acidaemia in these patients is usually mild and complications due to the acidosis are minor, which argues against alkali therapy in adults. In addition, if NaHCO3 is given, the PHCO3 rises temporarily, and urinary excretion HCO3− will also increase. A large increase in delivery of Na+ and HCO3− to the cortical distal nephron can augment the secretion of K+, resulting in hypokalaemia and also increases the risk of forming calcium phosphate kidney stones. In contrast, alkali therapy is useful in children to prevent growth retardation.

Distal renal tubular acidosis (See Chapter 36.) The hallmark of distal RTA is a low rate of excretion of NH4+ in a patient with chronic metabolic acidosis, a normal value for the PAnion gap, and a GFR that is not markedly reduced (Carlisle et al., 1991). Having defined these components, the next step is to find out why the rate of excretion of NH4+ may be lower than expected in this setting. We use the urinary pH at this point to separate patients into three categories; those with a primary problem with NH3 availability (urine pH close to 5), those with a defect in net distal H+ secretion (urine pH close to 7), and those in which there is a structural lesion in the renal medulla that compromises both medullary NH3 availability and distal H+ secretion (the urine pH is usually close to 6).

Subtypes of disorders causing low NH4+ excretion

Low availability of NH3 The usual causes of a low production of NH4+ are an alkaline PCT cell pH due to hyperkalaemia, and a low availability of ADP in PCT cells due to a very low GFR (Fig. 35.8). Less common causes include an alkaline PCT cell pH due to a genetic disorder or a disease process that causes a defect in the exit step for HCO3−, decreased availability of glutamine due to malnutrition, and/or high levels of other fuels that PCT cells can oxidize in place of glutamine to regenerate ATP (e.g. during total parenteral nutrition). Low net distal H+ secretion Autoimmune disorders (such as Sjögren syndrome, systemic lupus erythematosus, hyper-γ-globulinaemia) are the most common causes of dRTA with a high urinary pH in adults. RTA in patients with Sjögren syndrome seems to be due to a defect in H+ secretion in the distal nephron. In some of these patients, absence of the H+-ATPase pump in intercalated cells of the collecting tubule was found on immunocytochemical analysis of tissue obtained by renal biopsy (Cohen et al., 1992). It is not known how the immune injury leads to the loss of H+-ATPase activity. It has also been suggested that the defect may be due to autoantibodies against carbonic anhydrase II or the basolateral Cl−/HCO−3 anion exchanger, as well as decreased number of intercalated cells. Amphotericin B may cause a low net H+ secretion in the distal nephron, because of a back-leak of H+ into α-intercalated cells. In rare patients with Southeast Asian

chapter 35 

metabolic acidosis or alkalosis

Hyperchloraemic metabolic acidosis (HCMA) and low excretion of NH+4 What is the urine pH?

~5.0

~6.0

• Alkaline PCT cell pH (e.g. hyperkalaemia, isolated pRTA) • Low NH3 transfer

~7.0

• Low NH3 and low net distal H+ secretion • Medullary interstitial disease

What is Pco2 in alkaine urine?

~40 mmHg

~70 mmHg

Is citrate excretion low?

• Distal HCO–3 secretion (SAO) • Backleak of H+ (amphotericin B)

YES

• Low distal H+ ATPase (e.g. Sjögren syndrome)

NO • Alkaline cell pH in PCT & MCD (e.g. CAII deficiency)

Fig. 35.8  Steps in clinical approach to the patient with metabolic acidosis and a low rate of excretion of ammonium.

ovalocytosis (SAO), a second mutation in the anion exchanger AE1, results in some of the AE1 being targeted abnormally to the luminal membrane of α-intercalated cells. Alkali therapy is usually needed in patients with dRTA and a urinary pH > 6.5, as they are unable to excrete enough NH4+ to regenerate the HCO3− consumed by the dietary acid load. Bicarbonaturia, however, should be minimized, because it might predispose to excessive renal K+ loss and CaHPO4 kidney stone formation (see earlier for pRTA). Therefore, the dose of NaHCO3 should be as small as possible and distributed throughout the day. After the PHCO3 is corrected, the dose of NaHCO3 needed to maintain PHCO3 in the normal range is usually less than 30 to 40  mmol/day (i.e. enough to titrate the acid load produced from the metabolism of dietary sulphur-containing amino acids). Lesions involving both distal H+ secretion and NH3 availability The list of causes of disorders that affect the renal medullary interstitial compartment is long and includes infections, drugs, infiltrations, precipitations, inflammatory disorders, and sickle cell anaemia, among others. Because of the medullary interstitial disorder, these patients may also have a reduced urinary concentrating ability. Hyperkalaemia may be present if the disease process also involves the distal cortical nephron. Nevertheless, when the PK returns to the normal range, the acidaemia persists (see below). Administration of alkali is needed to correct the acidaemia. The issues concerning alkali therapy were discussed above in the subgroup with diminished net H+ secretion and apply in this setting too.

Distal RTA with hyperkalaemia The term type IV RTA is used to describe the constellation of findings of hyperkalaemia and metabolic acidosis due to a low rate of excretion of NH4+. Nevertheless, there are two distinct ways that hyperkalaemia and a low rate of excretion of NH4+ may coexist (Halperin et al., 2010).

Hyperkalaemia is responsible for the low rate of excretion of NH4+ Hyperkalaemia may cause a low rate of NH4+ excretion by inhibiting either its production (hyperkalaemia is associate with an alkaline PCT cell pH) and/or the transfer of NH3 in the loop of Henle (K+ competes with NH4+ on the Na+/K+/2Cl− cotransporter). In these patients, the urinary pH is low (~ 5). If this is the case, the patient should have a sufficient increase in the rate of excretion of NH4+ to correct the metabolic acidemia after the PK returns to the normal range. Hyperkalaemia is not the major reason for the low rate of excretion of NH4+ In this subgroup of patients the low rate of excretion of NH4+ is not causally linked to the hyperkalaemia, as is indicated by the fact that the rate of excretion of NH4 + remains low after the PK has returned to the normal range. In general, the basis of the low rate of excretion of NH4 + in this subgroup is a combination of low NH3 availability and low distal H+ secretion; hence, the urine pH would be close to 6. For chronic hyperkalaemia to be present, the disease process must involve the late cortical distal nephron, the major site of K+ secretion.

Metabolic alkalosis Metabolic alkalosis is an electrolyte disorder that is accompanied by changes in acid–base parameters in plasma, namely an elevated PHCO3 and plasma pH. Most patients with metabolic alkalosis have a deficit of NaCl, KCl, and/or HCl (Fig. 35.9), each of which leads to a higher PHCO3. The following concepts are central to our understanding of why metabolic alkalosis develops (Halperin et al., 2010). They also provide the basis for our clinical approach to this diagnostic category, and to the design of optimal therapy. Concept:  the concentration of HCO3− is the ratio of the content of HCO3− in the ECF compartment (numerator) and the ECF volume (denominator).

357

358

Section 2  

fluid, electrolyte, and renal tubular disorders Type of Cl-salt deficit What is the initial loss?

HCl

NaCl Why did the PHCO3 rise

Why did the PHCO3 rise

ECFV contraction

Causes KCl deficit

• Contraction Alkalosis

• Associated with low pH in PCT cells: – High NH4+ & Cl– excretion – Low OA– and/or HCO3– excretion

Gain HCO3– • Swap of Cl– for HCO3– in stomach (e.g. vomiting) • High delivery of NaCl to colon with low AE activity (e.g. Cl– diarrhoea)

Fig. 35.9  Pathophysiology of metabolic alkalosis caused by a deficiency of chloride salts.

A rise in the concentration of HCO3− might be due to an increase in its numerator (addition of HCO3−) and/or a decrease in its denominator (diminished ECF volume). A quantitative estimate of the ECF volume is critical to determine the quantity of HCO3− in the ECF compartment and to determine the basis of the metabolic alkalosis. Concept:  electroneutrality must be present in every body compartment and in the urine.

This implies that terms such as ‘Cl− depletion alkalosis’ are misleading; deficits must be defined as HCl, KCl, and/or NaCl. Concept:  knowing the balances for Na+, K+, and Cl− allows one to decide why the PHCO3 has risen and what changes have occurred in the composition of the ECF and ICF compartments.

Although balance data are not available in most patients, a quantitative assessment of ECF volume using the haematocrit can help to reach a tentative conclusion about the contribution of individual deficits of the different Cl−-containing compounds to the development of metabolic alkalosis. Concept: critical to the understanding of the pathophysiology of metabolic alkalosis is that there is no tubular maximum for HCO3− reabsorption in the kidney.

Angiotensin II and the usual pH in PCT cells are the two major physiologic stimuli for NaHCO3 reabsorption in this nephron segment. Both of these stimuli must be removed for NaHCO 3 to be excreted. Contrary to the widely held impression, there is no renal tubular maximum for the reabsorption of HCO 3−. Rather HCO 3− ions are retained unless their reabsorption is inhibited (low angiotensin II (AII), because of expansion of the EABV and/or an alkaline PCT cell pH or a high plasma HCO 3 − which diminishes HCO 3 − reabsorption by PCT) (Rubin et  al., 1994). NaHCO 3 loading will not cause metabolic alkalosis, because it expands the EABV and raises the PHCO3. Nevertheless, NaHCO3 may be retained when there is

a significant decrease in its filtered load due to an appreciable fall in the GFR. A disorder which results in a deficit of NaCl or HCl, which can cause a higher PHCO3, may also lead to a secondary deficit in KCl and hypokalaemia. A deficit of K+ is associated with an acidified PCT cell pH and can initiate and sustain a high PHCO3 as a result of renal new HCO3− generation (higher rate of excretion of NH4+), reduced excretion of dietary HCO3− in the form of organic anions, and enhanced reabsorption of HCO3− in the PCT. Concept: alkalaemia suppresses the respiratory centre and this leads to hypoventilation

Diagnostic tests Quantitative estimate of the ECF volume: it is critical to have a quantitative estimate of the ECF volume to determine the content of HCO3− in the ECF compartment and why there was a rise in the PHCO3. We use the haematocrit for this purpose (see ‘Metabolic acidosis’). Balance data for Na+, K+, and Cl−: although these data are not available in most patients, they can be inferred if one knows the new ECF volume and the PNa, PCl and PHCO3. One cannot know the balances for K+ from these calculations, but one can deduce their rough magnitude by comparing the differences in the content of Na+ with that of Cl− and HCO3− in the ECF compartment. Arterial PCO2: hypoventilation due to metabolic acidosis raise the arterial PCO2 by 0.7 mmHg for every 1 mmol /L rise in PHCO3 (Adrogue and Madias, 2010).

Clinical approach A list of causes of metabolic alkalosis is provided in Table 35.2. Four aspects of the clinical picture in a patient with metabolic alkalosis merit careful attention and include the medical history (e.g. vomiting, diuretic use), the presence of hypertension, the EABV status, and the PK. Our clinical approach to a patient with metabolic alkalosis is outlined in Fig. 35.10. The first step is to rule out the common causes of metabolic alkalosis such as vomiting and use of diuretics. Although this may be evident from the history, some

chapter 35 

Table 35.2  Causes of metabolic alkalosis Causes usually associated with a contracted ‘effective’ arterial blood volume Low UCl Loss of gastric secretions (e.g. vomiting, nasogastric suction) Remote use of diuretics Delivery of Na+ to CDN with non-reabsorbable anions plus a reason for Na+ avidity Post-hypercapnic states Loss of HCl via lower GI tract (e.g. congenital disorder with Cl− loss in diarrhoea, acquired forms of DRA) High UCl Recent diuretic use Endogenous diuretic effect (occupancy of the Ca-SR in the thick ascending limb of the loop of Henle, inborn errors affecting transporters of Na+ and/ or Cl− in the nephron, e.g. Bartter syndrome or Gitelman syndrome) Causes associated with an expanded ECF volume and possibly hypertension Disorders with primary enhanced mineralocorticoid activity causing hypokalaemia Primary hyperaldosteronism Primary hyper-reninaemic hyperaldosteronism (e.g. renal artery stenosis, malignant hypertension, renin-producing tumour) Disorders with cortisol acting as a mineralocorticoid (e.g. apparent mineralocorticoid excess syndrome, liquorice ingestion, ACTH producing tumour) Disorders with constitutively active ENaC in the CDN (e.g. Liddle syndrome) Large reduction in GFR plus a source of NaHCO3 CaSR = calcium-sensing receptor; CDN = cortical distal nephron (including the late distal convoluted tubule, the connecting tubule and the cortical collecting duct); DRA = downregulated Cl/HCO3 exchanger in adenoma/adenocarcinoma.

patients may deny the intake of diuretics or self-induced vomiting; examining the urinary electrolytes is particularly helpful if you suspect these diagnoses. An excellent initial test is to examine the concentration of Cl− in the urine (UCl) (Fig. 35.10). A very low UCl is expected when there is a deficit of HCl and/or NaCl, but the recent intake of diuretics causes the excretion of Na+ and Cl− in the urine. The UNa may be high if there is a recent episode of vomiting. If the UCl is not low, assessment of EABV and blood pressure helps separate patients with disorders of high epithelial sodium channel (ENaC) activity in the distal cortical nephron (the EABV is not low, presence of hypertension) from those with Bartter’s or Gitelman syndromes (EABV is low, absence of hypertension). Serial measurements of UCl in spot urine samples are helpful to separate patients with Bartter or Gitelman syndromes (persistently high UCl) from those with diuretic abuse (intermittently high UCl).

Common causes for metabolic alkalosis Vomiting or nasogastric suction The diagnosis is obvious if the patient has a history of prolonged vomiting or nasogastric suction. The difficulty arises if the patient denies vomiting. Nevertheless, there are several helpful clues to

metabolic acidosis or alkalosis

make the diagnosis, for example, the patient is particularly concerned with body image, has a profession where weight control is a very important factor (ballet dancer, fashion model, or beautician), has an eating disorder, and/or has a psychiatric disorder that might lead to self-induced vomiting. The physical examination may also provide some helpful clues. These may include a calloused lesion on the back of the finger or knuckles, which are often inserted into the mouth to induce vomiting, and erosion of dental enamel from repeated exposure to HCl. The EABV is often contracted. Hypokalaemia is always present and the deficit of KCl is a major factor in the pathophysiology of the metabolic alkalosis in these patients (Kassirer and Schwartz, 1966; Halperin and Scheich, 1994). Alkalaemia suppresses the respiratory centre and this leads to hypoventilation. A primary respiratory acidosis may be present if respiratory muscle weakness results from hypokalaemia. On the other hand, a primary respiratory alkalosis may be present if, for example, the patient develops aspiration pneumonia. The urinary electrolytes are very helpful when this diagnosis is suspected—the key finding is an extremely low UCl. If there has been recent vomiting, the UNa may be high due to bicarbonaturia (the urine pH will be > 7.0), which increases the excretion of Na+ (Kamel et al., 1990).

Diuretics The key findings in patients with metabolic alkalosis due to the use of diuretics are low EABV, hypokalaemia, and intermittently high concentrations of Na+ and Cl− in the urine (when the diuretic acts) (Fig. 35.10). Hypokalaemia is more likely to occur in patients who have a low intake of K+. A large deficit of NaCl is most commonly seen in patients who have a low intake of NaCl (e.g. in the elderly). The use of diuretics might be denied at times, especially in patients concerned about their body image or those seeking medical attention. To help distinguish these patients from those with Bartter or Gitelman syndromes (which are genetic, rare causes of hypokalemia and metabolic alkalosis) (Simon and Lifton. 1998), measure the urinary electrolytes using multiple random spot urine samples. If there is doubt, an assay for diuretics in the urine may be helpful, but make sure that the assay is performed in a urine sample that contains a high concentration of Na+ and Cl−. Some cationic agents (e.g. drugs such as gentamicin or cis platin or cationic proteins) may bind to the calcium sensing receptor in the thick ascending limb of the loop of Henle and lead to a picture that mimic Bartter syndrome (see Chapter 35).

Less common causes of metabolic alkalosis Conditions with high mineralocorticoid activity Clinical clues to conditions with high mineralocorticoid activity as the cause of metabolic alkalosis usually include hypokalaemia and hypertension. Hypokalaemia is of major importance in causing metabolic alkalosis in these patients. The specific disorders are listed in Table 35.2. Because of the high mineralocorticoid activity or a constitutively active ENaC, principal cells of the cortical distal nephron (CDN) are poised to reabsorb Na+. Initially Na+ and Cl− are retained, and hence the ECF volume will be expanded. Subsequently, K+ will be lost in the urine with Cl−, if reabsorption of Na+ in CDN is electrogenic (i.e. without Cl−) and if principal cells have open K+ channels in their luminal membrane −. Hypokalaemia is associated with

359

360

Section 2  

fluid, electrolyte, and renal tubular disorders Chronic Metabolic Alkalosis Is the UCl very low?

NO

YES

Is the patient on a diuretic?

Is the urine alkaline?

YES

NO

NO

• Recent vomiting

• Remote diuretics • Remote vomiting

YES

Does the patient have hypertension?

• Recent diuretics (may be denied)

NO

YES • High mineralocorticoid activity See Table 35-2

• Bartter syndrome • Gitelman syndrome • Ca-SR ligand is present (e.g. hypercalcemia, drugs such as gentamicin, cationic proteins)

Fig. 35.10  Steps in the clinical approach to a patient with metabolic alkalosis. Ca-SR, calcium sensing receptor.

an acidified PCT cell, which results in the excretion of more NH4+ and also the retention of dietary alkali. Overall, the body continues to have a surplus of Na+, but some of the retained Cl− are excreted in the urine (with NH4+) and are replaced in the body with HCO3−. The ICF compartment has a deficit in K+. The cations that are retained in the ICF are likely to be Na+ and possibly some H+ (Gowrishankar et al., 1996).

Metabolic alkalosis associated with the milk-alkali syndrome Although milk and absorbable alkali to treat duodenal ulcers are not used much nowadays, this form of metabolic alkalosis still continues to occur, but the setting has changed (Beall etal., 2006; Felsenfeld et al., 2006). Its cardinal features are still a dietary source of alkali and absence of suppression of the stimuli for the PCT to retain this alkali load. The intake of calcium supplements, commonly in the form of CaCO3 tablets, is now a common cause of hypercalcaemia, particularly in elderly women. Traditional Betel nut chewing with chalk is another example. Hypercalcaemia develops primarily because more calcium is absorbed in the GI tract (especially if the intake of calcium exceeds that of dietary phosphate). When more Ca2+ binds to its receptor in the medullary thick ascending limb of the loop of Henle, it acts as a loop diuretic that leads to an excessive excretion of NaCl and KCl. Both the high levels of AII due to decreased EABV and the intracellular acidosis in PCT cells associated with hypokalaemia lead to the retention of ingested alkali (Lin et al., 2002). The combination of a contracted EABV and direct effects of hypercalcaemia can also cause a marked reduction in the GFR, which itself further reduces the filtration and excretion of HCO3−. Therapy consists of stopping the intake of calcium and alkali, and replacing the deficits of NaCl and KCl.

Metabolic alkalosis associated with a post-hypercapnic state In the course of chronic hypercapnia, the high PCO2 causes acidosis in the cells of the PCT. This leads to an enhanced excretion of NH4Cl in the urine and stimulates reabsorption of HCO3, and results in increased PHCO3 (Schwartz et al., 1961). If the patient has a contracted EABV when the hypercapnia resolves, NaHCO3 will still be retained because of the high AII levels maintaining reabsorption of NaHCO3 by PCT cells. Expansion of the EABV will lower AII levels and cause the excretion of the excess NaHCO3.

Metabolic alkalosis associated with the intake of non-reabsorbed anions If a patient has a contracted EABV and takes a Na+ salt with an anion that cannot be reabsorbed by the kidney (e.g. Na+ carbenicillinate), the patient may develop hypokalaemia and metabolic alkalosis. In the cortical distal nephron, the actions of aldosterone cause Na+ to be reabsorbed in conjunction with K+ secretion, because of the low delivery of Cl− and hence hypokalaemia develops. The rise in PHCO3 is the result of the NaCl and the KCl deficits.

Metabolic alkalosis associated with magnesium depletion Patients with Mg2+ depletion may have hypokalaemia and metabolic alkalosis. The usual clinical settings for this deficiency include malabsorption, chronic alcoholism, chronic use of proton pump inhibitors, use of loop diuretics, or the administration of drugs that may bind the calcium sensing receptor in the loop of Henle (e.g. cisplatin, or aminoglycosides). These patients must be distinguished from those with primary hyperaldosteronism and those with Bartter or Gitelman syndrome who may also have Mg2+ deficiency.

chapter 35 

References Adrogue, H. J. and Madias, N. E. (2010). Secondary responses to altered acid-base status: the rules of engagement. J Am Soc Nephrol, 21, 920–3. Bayoumi, A., Campbell, P., Schreiber, M., et al. (1993). Towards a more rational treatment of aspirin (ASA) overdose. J Am Soc Nephrol, 3, 293. Beall, D., Henslee, H., Webb, H., et al. (2006). Milk-alkali syndrome: a historical review and description of the modern version of the syndrome. Am J Med Sci, 331, 233–42. Brinkman, K., Smeitink, J. A., Romijn, J. A., et al. (1999). Mitochondrial toxicity induced by nucleoside-analogue reverse-transcriptase inhibitors is a key factor in the pathogenesis of antiretroviral-therapy-related lipodystrophy. Lancet, 354, 1112–15. Brown, T. B. (2004). Cerebral oedema in childhood diabetic ketoacidosis: Is treatment a factor? Emerg Med J, 21, 141–4. Carlisle, E. J. F., Donnelly, S. M., and Halperin, M. L. (1991). Renal tubular acidosis (RTA): Recognize The Ammonium defect and pHorget the urine pH. Pediatr Nephrol, 5, 242–8. Carlisle, E. J. F., Donnelly, S. M., Vasuvattakul, S., et al. (1991). Glue-sniffing and distal renal tubular acidosis: sticking to the facts. J Am Soc Nephrol, 1, 1019–27. Carlotti, A., Bohn, D., and Halperin, M. (2003). Importance of timing of risk factors for cerebral oedema during therapy for diabetic ketoacidosis. Arch Dis Child, 88, 170–3. Carlotti, A. C. P., St George-Hyslop, C., Guerguerian, A. M., et al. (2009). Occult risk factor for the development of cerebral edema in children with diabetic ketoacidosis: possible role for stomach emptying. Pediatr Diabetes, 10, 522–33. Cheema-Dhadli, S., Lin, S.-H., and Halperin, M. L. (2002). Mechanisms used to dispose of a progressively increasing alkali load in the rat. Am J Physiol, 282, F1049–F55. Chin, L., Sievers, M. L., Herrier, R. N., et al. (1979). Convulsions as the etiology of lactic acidosis in acute isoniazid toxicity in dogs. Toxicol Appl Pharmacol, 49, 377–84. Coghlan, M. E., Sommadossi, J. P., Jhala, N. C., et al. (2001). Symptomatic lactic acidosis in hospitalized antiretroviral-treated patients with human immunodeficiency virus infection: a report of 12 cases. Clin Infect Dis, 33, 1914–21. Cohen, E. P., Bastani, B., Cohen, M. R., et al. (1992). Absence of H(+)-ATPase in cortical collecting tubules of a patient with Sjogren’s syndrome and distal renal tubular acidosis. J Am Soc Nephrol, 3, 264–71. Davids, M. R., Lin, S. –H., Edoute, Y., et al. (2002). Hyponatraemia and hyperglycaemia during laproscopic surgery. QJM, 95, 321–30. Davids, M. R., Segal, A. S., Brunengraber, H., et al. (2004). An unusual cause for ketoacidosis. QJM, 97, 365–76. DeMars, C., Hollister, K., Tomassoni, A., et al. (2001). Citric acidosis: a life-threatening cause of metabolic acidosis. Ann Emerg Med, 38, 588–91. Dempsey, G. A., Lyall, H. J., Corke, C. F., et al. (2000). Pyroglutamic acidemia: a cause of high anion gap metabolic acidosis. Crit Care Med, 28, 1803-7. Dyck, R., Asthana, S., Kalra, J., et al. (1990). A modification of the urine osmolal gap: an improved method for estimating urine ammonium. Am J Nephrol, 10, 359–62. Eladari, D. and Chambrey, R. (2010). Ammonium transport in the kidney. J Nephrol, 23 Suppl 16, S28–34. Emmett, M. (2006). Anion-gap interpretation: the old and the new. Nat Clin Pract Nephrol, 2, 4–5. Emmett, M. and Narins, R. G. (1977). Clinical use of the anion gap. Medicine, 56, 38–54. Felsenfeld, A. J. and Levine, B. S. (2006). Milk alkali syndrome and the dynamics of calcium homeostasis. Clin J Am Soc Nephrol, 1, 641–54. Fenves, A. Z., Kirkpatrick, H. M., 3rd, Patel, V. V., et al. (2006). Increased anion gap metabolic acidosis as a result of 5-oxoproline (pyroglutamic acid): a role for acetaminophen. Clin J Am Soc Nephrol, 1, 441–7.

metabolic acidosis or alkalosis

Field, M. (2003). Intestinal ion transport and the pathophysiology of diarrhea. J Clin Invest, 111, 931–43. Fields, A. L. A., Wolman, S. L., and Halperin, M. L. (1981). Chronic lactic acidosis in a patient with cancer: therapy and metabolic consequences. Cancer, 47, 2026–9. Flatt, J. P. (1972). On the maximal possible rate of ketogenesis. Diabetes, 21, 50–3. Forsythe, S. M. and Schmidt, G. A. (2000). Sodium bicarbonate for the treatment of lactic acidosis. Chest, 117, 260–7. Geers, C. and Gros, G. (2000). Carbon dioxide transport and carbonic anhydrase in blood and muscle. Physiol Rev, 80, 681–715. Glaser, N., Barnett, P., McCaslin, I., et al. (2001). Risk factors for cerebral edema in children with diabetic ketoacidosis. N Engl J Med, 344, 264–9. Gopinath, R., Hutcheson, M., Cheema-Dhadli, S., et al. (1992). Chronic lactic acidosis in a patient with acquired immunodeficiency syndrome and mitochondrial myopathy: biochemical studies. J Am Soc Nephrol, 3, 1212–19. Gowrishankar, M. G., Chen, C. –B., Mallie, J. P., et al. (1996). What is the impact of potassium excretion on the intracellular fluid volume: importance of urine anions. Kidney Int, 50, 1490–5. Gowrishankar, M., Kamel, K. S., and Halperin, M. L. (2007). Buffering of a H+ load; a ‘brain-protein-centered’ view. J Am Soc Nephrol, 18, 2278–80. Greenough, W. B., III, Hirschhorn, N., Gordon, R. S., Jr., et al. (1976). Pulmonary oedema associated with acidosis in patients with cholera. Trop Georg Med, 28, 86–90. Halperin, F. A., Cheema-Dhadli, S., Chen, C. B., et al. (1996). Alkali therapy extends the period of survival during hypoxia: studies in rats. Am J Physiol, 271(2 Pt 2), R381–7. Halperin, M. L., Cherney, D. Z. I., and Kamel, K. S. (2002). Ketoacidosis. In T. D. DuBose, Jr and L. L. Hamm (eds.) Acid-Base and Electrolyte Disorders: A Companion to Brenner and Rector’s, The Kidney, pp. 67–82. Philadelphia: W. B. Saunders. Halperin, M. L., Goldstein, M. B., Haig, A., et al. (1974). Studies on the pathogenesis of type I (distal) renal tubular acidosis as revealed by the urinary PCO2 tensions. J Clin Invest, 53, 669–77. Halperin, M., Goldstein, M., Jungas, R., et al. (1984). Biochemistry and physiology of ammonium excretion. In D. Seldin and G. Giebisch (eds.) Physiology and Pathology of Electrolyte Metabolism, pp. 1471–89. New York: Raven Press. Halperin, M. L., Halperin, F. A., Cheema Dhadli, S., et al. (1994). Rationale for the use of sodium bicarbonate in a patient with lactic acidosis due to a poor cardiac output. Nephron, 66, 258–61. Halperin, M. L., Hammeke, M., Josse, R. G., et al. (1983). Metabolic acidosis in the alcoholic: a pathophysiologic approach. Metabolism, 32, 308–15. Halperin, M. L., Jungas, R. L., Cheema-Dhadli, S., et al. (1987). Disposal of the daily acid load: an integrated function of the liver, lungs and kidneys. Trends in Biol Sci, 12, 197–9. Halperin, M. L. and Kamel, K. S. (1996). Turning sugar into acids in the gastrointestinal tract. Kidney Int, 49, 1–8. Halperin, M. L. and Kamel, K. S. (2010). Some observations on the clinical approach to metabolic acidosis. J Am Soc Nephrol, 21, 894–7. Halperin, M. L., Kamel, K. S., Ethier, J. H., et al. (1989). What is the underlying defect in patients with isolated, proximal renal tubular acidosis? Am J Nephrol, 9, 265–8. Halperin, M. L., Kamel, K., and Goldstein, M. B. (2010). Fluid, Electrolytes and Acid-Base Physiology: A Problem Based Approach (5th ed.). Philadelphia, PA: Saunders Elsevier. Halperin, M. L. and Scheich, A. (1994). Should we continue to recommend that a deficit of KCl be treated with NaCl? A fresh look at chloride-depletion metabolic alkalosis. Nephron, 67, 263–9. Halperin, M. L., Vasuvattakul, S., and Bayoumi, A. (1992). A modified classification of metabolic acidosis: a pathophysiologic approach. Nephron, 60, 129–33. Hanstein, W. G. (1976). Uncoupling of oxidative phosphorylation. Biochim Biophys Acta, 456, 129–48. Haque, S. K., Ariceta, G., and Batlle, D. (2012). Proximal renal tubular acidosis: a not so rare disorder of multiple etiologies. Nephrol Dial Transplant, 27, 4273–87.

361

362

Section 2  

fluid, electrolyte, and renal tubular disorders

Huang, Y. G., Wong, K. C., Yip, W. H., et al. (1995). Cardiovascular responses to graded doses of 3 catecholamines during lactic and hydrochloric acidosis in dogs. Br J Anaesth, 74, 583–90. Igarashi, T., Sekine, T., Inatomi, J., et al. (2002). Unraveling the molecular pathogenesis of isolated proximal renal tubular acidosis. J Am Soc Nephrol, 13, 2171–7. Kaitwatcharachai, C., Vasuvattakul, S., Yenchitsomanus, P., et al. (1999). Distal renal tubular acidosis in a patient with Southeast Asian ovalocytosis: possible interpretations of a high urine PCO2. Am J Kidney Dis, 33, 1147–52. Kamel, K. S. (1996). The acute impact of NaHCO3 in treatment of metabolic acidosis on back-titration of non-bicarbonate buffers: a quantitative analysis. Clin Nephrol, 45, 51–5. Kamel, K. S. (1997). Alcoholic ketoacidosis: emphasis on biochemical, metabolic and quantitative aspects. Geriatr Nephrol Urol, 6, 159–68. Kamel, K. S., Briceno, L. F., Santos, M. I., et al. (1997). A new classification for renal defects in net acid excretion. Am J Kidney Dis, 29, 126–36. Kamel, K. S., Ethier, J. H., Bear, R., et al. (1990). Urine electrolytes and osmolality: when and how to use them. Am J Nephrol, 10, 89–102. Kamel, K. S. and Halperin, M. L. (2006). An improved approach to the patient with metabolic acidosis: a need for four amendments. J Nephrol, 19 Suppl 9, S76–85. Kamel, K. S. and Halperin, M. L. (2015). Acid-base problems in diabetic ketoacidosis. N Engl J Med, 372, 546–54. Kamel, K. S., Lin, S.-H., Cheema-Dhadli, S., et al. (1998). Prolonged total fasting: a feast for the integrative physiologist. Kidney Int, 53, 531–9. Kassirer, J. P. and Schwartz, W. B. (1966). The response of normal man to selective depletion of hydrochloric acid. Am J Med, 40, 10–18. Kraut, J. and Madias, N. (2007). Serum anion gap: its uses and limitations in clinical medicine. Clin J Am Soc Nephrol, 2007, 162–74. Kraut, J. A. and Madias, N. E. (2012). Treatment of acute metabolic acidosis: a pathophysiologic approach. Nat Rev Nephrol, 8, 589–601. Kraut, J. A. and Xing, S. X. (2011). Approach to the evaluation of a patient with an increased serum osmolal gap and high-anion-gap metabolic acidosis. Am J Kidney Dis, 58, 480–4. Lalau, J. D. (2010). Lactic acidosis induced by metformin: incidence, management and prevention. Drug Saf, 33, 727–40. Lin, S. H., Lin, Y. F., Cheema-Dhadli, S., et al. (2002). Hypercalcaemia and metabolic alkalosis with betel nut chewing: emphasis on its integrative pathophysiology. Nephrol Dial Transplant, 17, 708–14. Love, A. H. G. and Phillips, R. A. (1969). Measurement of dehydration in cholera. J Infect Dis, 119, 39–42. Luft, F. C. (2001). Lactic acidosis update for critical care clinicians. J Am Soc Nephrol, 12 Suppl 17, S15–19. Luzzati, R., Del Bravo, P., Di Perri, G., et al. (1999). Riboflavine and severe lactic acidosis. Lancet, 353, 901–2. Mazer, C. D., Naser, B., and Kamel, K. S. (1996). Effect of alkali therapy with NaHCO3 or Tham on cardiac contractility. Am J Physiol, 270, R955–R62. Mitchell, J. H., Wildenthal K., and Johnson, R. L., Jr. (1972). The effects of acid-base disturbances on cardiovascular and pulmonary function. Kidney Int, 1, 375–89. Miyahara, J. T. and Karler, R. (1965). Effect of salicylate on oxidative phosphorylation and respiration of mitochondrial fragments. Biochem J, 97, 194–8. Mizock, B. A. and Mecher, C. (2000). Pyruglutamic acid and high anion gap: looking through the keyhole? Crit Care Med, 28, 2140–1. Napolova, O., Urbach, S., Davids, M. R., et al. (2003). How to assess the degree of extracellular fluid volume contraction in a patient with a severe degree of hyperglycemia. Nephrol Dial Transplant, 18, 2674–7. Narins, R. G. and Cohen, J. J. (1987). Bicarbonate therapy for organic acidosis: the case for its continued use. Ann Internal Med, 106, 615–8. Oh, M. and Halperin, M. L. (2005). Toxin-induced metabolic acidosis. In F. J. Gennari, H. J. Adrogue, J. H. Galla, et al. (eds.) Acid-Base Disorders and their Treatment, pp. 377–409. Boston, MA: Marcel Decker Inc. Pierce, N. F., Fedson, D. S., Brigham, K. L., et al. (1970). The ventilatory response to acute base deficit in humans. Time course during

development and correction of metabolic acidosis. Ann Internal Med, 72, 633–40. Pinto, J., Huang, Y. P., and Rivlin, R. S. (1981). Inhibition of riboflavin metabolism in rat tissues by chlorpromazine, imipramine, and amitriptyline. J Clin Invest, 67, 1500–6. Porte DJ. (1969). Sympathetic regulation of insulin secretion. Arch Intern Med, 123, 252–60. Richardson, R. M. A. and Halperin, M. L. (1987). The urine pH: a potentially misleading diagnostic test in patients with hyperchloremic metabolic acidosis. Am J Kidney Dis, 10, 140–3. Rubin, S. I., Sonnenberg, B., Zettle, R., et al. (1994). Metabolic alkalosis mimicking the acute sequestration of HCl in rats: bucking the alkaline tide. Clin Invest Med, 17, 515–21. Sabatini, S. and Kurtzman, N. A. (2009). Bicarbonate therapy in severe metabolic acidosis. J Am Soc Nephrol, 20, 692–5. Salpeter, S. R., Greyber, E., Pasternak, G. A, et al. (2010). Posthumous EE. Risk of fatal and nonfatal lactic acidosis with metformin use in type 2 diabetes mellitus. Cochrane Database Syst Rev, 1, CD002967. Schreiber, M., Kamel, K. S., Cheema-Dhadli, S., et al. (1994). Ketoacidosis: an integrative view. Diabetes Rev, 2, 98–114. Schwartz, W. B., Hays, R. M., Polak, A., et al. (1961). Effects of chronic hypercapnia on electrolyte and acid-base equilibrium. II. Recovery, with special reference to the influence of chloride intake. J Clin Invest, 40, 1238–49. Shull, P. D. and Rapoport, J. (2010). Life-threatening reversible acidosis caused by alcohol abuse. Nat Rev Nephrol, 6, 555–9. Simon, D. B. and Lifton, R. P. (1998). Ion transporter mutations in Gitelman’s and Bartter’s syndromes. Curr Opin Nephrol Hypertens, 7, 43–7. Simpson, D. (1983). Citrate excretion: a window on renal metabolism. Am J Physiol, 244, F223–F34. Siskind, M. S., Thienemann, D., and Kirlin, L. (1993). Isoniazid-induced neurotoxicity in chronic dialysis patients: report of three cases and a review of the literature. Nephron, 64, 303–6. Sonne, O., Gliemann, J., and Linde, S. (1981). Effect of pH on binding kinetics and biological effect of insulin in rat adipocytes. J Biol Chem, 256, 6250–4. Sperling, M. A. (2006). Cerebral edema in diabetic ketoacidosis; an underestimated complication? Pediatr Diabetes, 7, 73–4. Stacpoole, P. W. (1986). Lactic acidosis: the case against bicarbonate therapy. Ann Internal Med, 105, 276–9. Stewart, P. A. (1981). How to Understand Acid-Base. A Quantitative Acid-Base Primer for Biology and Medicine. New York: Elsevier. Tannen, R. L. (1980). Effect of potassium on renal acidification and acid-base homeostasis. Sem Nephrol, 7, 263–73. Uribarri, J., Oh, M. S., and Carroll, H. J. (1998). D-lactic acidosis. A review of clinical presentation, biochemical features, and pathophysiologic mechanisms. Medicine, 77, 73–82. Van der Meulen, J. A., Klip, A., and Grinstein, S. (1987). Possible mechanism for cerebral oedema in diabetic ketoacidosis. Lancet, 2, 306–8. Vasuvattakul, S., Wasrner, L. C., and Halperin, M. L. (1992). Quantitative role of the intracellular bicarbonate buffer system in response to an acute acid load. Am J Physiol, 262, R305–R9. Weiner, I. D. and Verlander, J. W. (2013). Renal ammonia metabolism and transport. Comp Physiol, 3, 201–20. Weinstein, J., Schreiber, M., and Kamel, K. S. (2004). Should mild metabolic acidosis in CKD patients be treated? Semin Dial, 17, 351–64. Worthley, L. I., Guerin, M., and Pain, R. W. (1987). For calculating osmolality, the simplest formula is the best. Anaesth Intensive Care, 15, 199–202. Wrenn, K. D., Slovis, C. M., Minion, G. E., et al. (1991). The syndrome of alcoholic ketoacidosis. Am J Med, 91, 119–28. Yaqoob, M. M. (2010). Acidosis and progression of chronic kidney disease. Curr Opin Nephrol Hypertens, 19, 489–92. Zalunardo, N., Lemaire, M., Davids, M. R., et al. (2004). Acidosis in a patient with cholesra: a need to redefine concepts. QJM, 97, 681–96.

CHAPTER 36

Approach to the patient with renal tubular acidosis Stephen B. Walsh Introduction Classification The first description of what would later be termed renal tubular acidosis (RTA) was by an English paediatrician, Reginald Lightwood, whose report was an abstract from a meeting (Lightwood, 1936) in 1936 that did not mention acidosis, but a ‘calcium infarction of the kidneys’ in infants who died of a dehydrating, salt-wasting disease. Later the same year, Butler, Wilson, and Farber of the Massachusetts General Hospital described ‘Dehydration and acidosis with calcification of renal tubules’ (Butler et  al., 1936). The paper was chiefly concerned with infants, but the appendices described two adults, one of whom had recently had a parathyroid adenoma removed for primary hyperparathyroidism, the first report of RTA and nephrocalcinosis secondary to hypercalcaemic renal damage. Subsequently, Lightwood, Payne, and Black added acidosis and a urinary acidification defect to the syndrome suffered by the sick infants with nephrocalcinosis (Lightwood et al., 1953). In this paper, and others appearing at the same time, including those from Stapleton (Stapleton, 1949, 1954), it was acknowledged that hypercalciuria was often a feature of these patients. New cases of the so-called Lightwood syndrome failed to appear through the 1950s and 1960s, and it was also reported that the urinary acidification defect tended to disappear, as affected children grew older (Buchanan and Komrower, 1958). This finding raised the possibility that its cause had been removed from the environment: vitamin D or calcium excess, or mercury in teething powders was suggested. The Lightwood syndrome came to be regarded as an unexplained historic accident of the 1940s that no longer existed, and the interest of tubular physicians tended to shift more to older children, where renal acidosis with a urinary acidification defect seemed to be a commoner and more persistent problem. Albright, a colleague of Butler’s at the Massachusetts General Hospital described clearly what we would now term ‘distal renal tubular acidosis’ (dRTA) in his book with Reifenstein (Albright and Reifenstein, 1948, p. 393), detailing the resultant bone disease and nephrocalcinosis and the role of alkali treatment, and labelling it as a form of ‘tubular insufficiency without glomerular insufficiency’; he attributed the acidosis to renal tubular inability to make ammonia and an acid urine. Albright did not use the term ‘renal tubular acidosis’, the term was probably first used by Pines and Mudge in 1951 (although one

of the adults that they reported had a ureterocolic anastomosis, and so a tubular defect was an unlikely explanation for their acidosis) (Pines and Mudge, 1951) Wrong and Davies introduced the concept of an acidification defect in the absence of a systemic acidosis, the so-called incomplete syndrome, and also quantified the threshold pH that patients with a dRTA were unable to acidify their urine below; 5.3 (Wrong and Davies, 1959). Although acidosis was recognized in the Fanconi syndrome, it was not initially described as a form of RTA, as it was rare and features other than the acidosis dominated the clinical picture. The acidosis was attributed to a loss of base in the urine, either bicarbonate, or a base that would have been converted to bicarbonate, if it were retained. The terms ‘proximal renal tubular acidosis’ and ‘distal renal tubular acidosis’ were coined by Rodriguez-Soriano and colleagues in their classification published in 1969 (Rodriguez-Soriano et al., 1969). The nomenclature was confused somewhat when Curtis Morris introduced a numerical system for their classification (Morris and McSherry, 1972). Morris labelled distal or classical RTA as ‘type 1’ RTA, proximal RTA (pRTA) as ‘type 2’, and designated ‘type 3’ as a paediatric form of RTA that was a mixture of types 1 and 2, which most nephrologists regarded as a severe form of type 1, requiring rather more alkali than most type 1 patients. Morris’ ‘type 3’ disappeared from the clinical spectrum, later to be replaced by the different cause of carbonic anhydrase deficiency, but its existence meant that hyperkalaemic renal tubular acidosis had to be designated ‘type 4’.

Distal renal tubular acidosis (dRTA, type 1 RTA) Aetiology Distal RTA is caused by a failure of the acid-secreting α-intercalated cell in the cortical collecting duct. It has a number of causes, primary and secondary:

Primary Mutations of the anion exchanger AE1, normally present on the basolateral surface of the α-intercalated cell (Bruce et  al., 1997). These may be autosomal dominant or recessive. Mutations of the a4 or B1 subunit of the vH+-ATPase, normally present on the apical surface of the α-intercalated cell (Karet et al., 1999). These are all autosomal recessive.

364

Section 2  

fluid, electrolyte, and renal tubular disorders

Secondary In association with autoimmune disease, notably Sjögren syndrome (Shearn and Tu, 1968), but also systemic lupus erythematosus (Tu and Shearn, 1967), rheumatoid arthritis (Pasternack et al., 1970), autoimmune thyroid disease (Mason and Golding, 1970), primary biliary cirrhosis (Golding and Mason, 1971) and hypergammaglobulinaemia (McCurdy et al., 1967). Due to a number of toxins, such as ifosfamide (Boddy et al., 1996), toluene (Batlle et al., 1988), lithium carbonate (Boton et al., 1987), and amphotericin B (McCurdy et  al., 1968). Nephrocalcinosis (Butler et  al., 1936)  (which can cause, as well as be caused by, dRTA: it can be secondary to medullary sponge kidney (MSK), vitamin D intoxication and sarcoidosis). It has also been attributed to other conditions, but the evidence in these cases is less secure: ◆ Wilson

disease (Hoppe et al., 1993)

◆ Sickle

cell disease (Goossens et al., 1972)

◆ Renal

transplantation (Wilson and Siddiqui, 1973)

◆ Chronic ◆ Liver

urinary tract obstruction (Van der Heijden et al., 1985)

cirrhosis (Shear et al., 1969).

Epidemiology As a rare syndrome, there is little published literature on the epidemiology of dRTA. Little is known about the prevalence of the mutations of the a4 or B1 subunits of the apical proton pump (vH+ATPase), other than that they are rare. The autosomal dominant AE1 mutations, which are generally found in Western populations, are also rare. Autosomal recessive AE1 mutations, found in tropical populations, predominantly in Southeast Asia, are more common there. dRTA is endemic in Northeast Thailand (Nilwarangkur et al., 1990), present in as much as 2.8% of the population (Nimmannit et al., 1996), predominantly in the Lao-Thai people of that region, in whom just one of the disease-causing AE1 mutations approaches polymorphic frequency (Yenchitsomanus et al., 2003). Since this hereditary tropical dRTA is prevalent in malarious areas, and because AE1 is also a major constituent of the red cell membrane, some have hypothesized that it may confer some protection against malaria (Wrong et al., 2002; Walsh et al., 2009; Khositseth et al., 2012). Acquired causes of dRTA are also rare, and the literature on the epidemiology of acquired dRTA is even sparser than for hereditary dRTA. The commonest cause of acquired dRTA is autoimmune disease, predominantly Sjögren syndrome. The reported prevalence of urinary acidification defects (i.e. dRTA) is quite high:  33% of patients with primary Sjögren in one series (Siamopoulos et  al., 1992), the majority of these were ‘incomplete dRTA’—in other words, had no systemic acidosis, but needed urinary acidification testing to reveal their acidification defect. Earlier studies showed similar high incidences of dRTA: Talal et al. found 50% of Sjögren patients had dRTA, of which half were incomplete dRTA (Talal et al., 1968); Shearn and Tu found incomplete dRTA in 30% of a series of 10 Sjögren patients without metabolic acidosis (Shearn and Tu, 1968). Pertovaara et al. found dRTA in 18 of 55 (32.7%) patients with primary Sjögren (Pertovaara et al., 1999). Ren and co-workers found RTA in 73% (95 of 130 patients) (Ren et  al., 2008); they found more complete dRTA than incomplete (66 vs 25), and they may have overdiagnosed complete dRTA according to their methodology.

As bone demineralization/osteoporosis is a part of dRTA, some investigators have looked for occult urinary acidification defects in patients with ‘primary’ osteoporosis. Weger et al found incomplete dRTA in 9 of 48 subjects (19%), even after 11 subjects with dRTA were excluded, since they had an obvious cause for their acidification defect (Weger et al., 1999). The same group found incomplete dRTA in 10 of 46 subjects (22%) investigated for ‘primary’ osteoporosis by using the both the furosemide (see ‘Investigations’) and ammonium chloride tests (Weger et al., 2000). Furthermore, Deutschmann and co-workers found incomplete dRTA in 12 of 285 (4.2%) women and 15 of 92 (16.3%) men with osteoporosis (Deutschmann et al., 2002).

Clinical features dRTA is the classical and commonest form, and involves a failure of acid secretion in the distal nephron, more specifically by the acid secreting α-intercalated cells of the distal convoluted tubule and the cortical collecting duct. It is characterized by an inability to acidify the urine to a pH of < 5.3, which can lead to: ◆ variable

metabolic acidosis

◆ osteomalacia

acidosis

and/or bone demineralization subsequent to the

◆ nephrocalcinosis

(calcium salt deposition in renal parenchyma)

◆ renal

calculi (calcium phosphate stones due to high calcium, low citrate, alkaline urine).

◆ variable

hypokalaemia due to urinary potassium wasting

◆ variable

salt-losing nephropathy due to urinary sodium wasting.

The metabolic acidosis can be very variable, ranging from very severe (serum bicarbonate < 10  mmol/L) to completely normal. DRTA with a normal serum bicarbonate is termed ‘incomplete dRTA’; the other clinical features may still be present

Investigations The diagnosis of dRTA is often difficult, because the routine biochemical tests done in patients with kidney stones or bone disease are usually fairly unremarkable (Table 36.1). Hypokalaemia is often present (Sebastian et al., 1971) and may be severe enough to cause paralysis (Zimhony et  al., 1995)  or cardiac arrhythmias (Palkar et  al., 2011); this is usually in the setting of a coincident problem worsening the hypokalaemia (e.g. a diarrhoeal illness). There

Table 36.1  Distal renal tubular acidosis: investigations Routine serum biochemistry

Hypokalaemia Low serum bicarbonate

Urine biochemistry

Hypocitraturia Hypercalciuria

Stone biochemistry

Calcium phosphate stones

Radiology

AXR: radiolucent stones, nephrocalcinosis CT KUB: radiolucent stones, nephrocalcinosis DXA: reduced bone mineral density

Dynamic test

Inability to acidify urine < pH 5.3 in urinary acidification test (ammonium chloride or furosemide plus fludrocortisone test)

chapter 36 

approach to the patient with renal tubular acidosis

may be a severe metabolic acidosis (with the serum bicarbonate < 10 mmol/L), but there may be no systemic acidosis at all. Urinary biochemistry will reveal hypocitraturia; this is a strong clue that dRTA may be present. Hypercalciuria is also usually present. An inappropriately high urine pH (> 5.3) in the face of a metabolic acidosis is diagnostic of dRTA; however, the urine pH must be measured immediately with a glass electrode pH meter (dipstick urine pHs are unreliable) since delay will allow diffusive loss of CO2 from urine, raising the pH. This usually means that an accurate urine pH measurement is often difficult in routine clinical practice. Analysis of any stones will show the calculi are composed mainly or exclusively of calcium phosphate, rather than the commoner calcium oxalate. Radiologically, the main features are of calcium deposition. A plain abdominal film may detect visible calculi at any point between the renal pelvis and the bladder. It may also reveal medullary nephrocalcinosis:  deposition of calcium phosphate in the medullary parenchyma. Both of these radiological signs will also be visible on computed tomography (CT), because CT is very good at visualizing calcium. Dual X-ray absorptiometry (DXA) scanning can show reduced bone mineral density, although this may be attenuated by alkali treatment (Domrongkitchaiporn et al., 2002). As mentioned previously, the serum bicarbonate may be normal (‘incomplete dRTA’), which makes a definitive diagnosis of dRTA by urine pH measurement alone difficult, since you cannot determine whether the urine pH is inappropriately high. The solution to this problem is to do a urinary acidification test. This can be done either by the administration of oral ammonium chloride capsules (100 mg/kg) to produce a metabolic acidosis, followed by hourly urine pH measurements for at least 6 hours:  a failure to acidify the urine to a pH < 5.3 is diagnostic of dRTA (Wrong and Davies, 1959)  Ammonium chloride capsules can be difficult to procure, and often provoke nausea and vomiting, which can invalidate the test. An alternative is to give furosemide with fludrocortisone (40 mg and 1 mg respectively) and measure urine pH every hour afterwards for at least 4 hours (Walsh et al., 2007). Again, failure to acidify the urine to a pH of < 5.3 is diagnostic of dRTA. In this case, no acidosis is produced; the stimulus to urinary acidification is due to the increased delivery of sodium ions to the cortical collecting duct and enhanced absorption there by principal cells. The addition of the mineralocorticoid fludrocortisone is a more consistent stimulus to collecting duct acidification that the original furosemide (alone) test mentioned earlier.

1997). Mutations in SLC4A1 can cause autosomal dominant (AD) dRTA (Bruce et  al., 1997; Karet et  al., 1998; Weber et  al., 2000; Sritippayawan et  al., 2003), as well as autosomal recessive (AR) dRTA (Tanphaichitr et al., 1998; Vasuvattakul et al., 1999; Bruce et  al., 2000; Ribeiro et  al., 2000). The phenotypes of hereditary AE1-associated AD and AR dRTA are determined by SLC4A1 point mutations or deletions that affect AE1 folding and tertiary structure, without significantly changing the anion transport properties of the protein. These structural alterations do not seem to affect the dimerization of these mutant proteins as hetero- or homodimers. They do, however, affect the trafficking of the mutant homodimers, from the ER and the trans-Golgi network to the cell membrane. The trafficking of the mutant/wild type (wt) heterodimer determines the dominant or recessive nature of the phenotype. In AD dRTA the mutant kAE1 in the heterodimer exerts a ‘dominant negative’ effect, inducing a trafficking defect on the dimerized wt kAE1. In AR dRTA, the wt kAE1 in the heterodimer corrects the defective trafficking of the mutant kAE1, resulting in a ‘dominant positive’ effect and thus normal cell surface expression of kAE1 in the heterozygote. In two cases (R901X and G609R) the dominant mutants can exit the ER and is partially mis-sorted to the apical membrane; this and the reduced basolateral expression of AE1 lead to the dRTA phenotype.

Aetiology and pathogenesis As already mentioned, dRTA is due to failure of the acid-secreting α-intercalated cell in the cortical collecting duct. Acid secretion is driven by the hydration of carbon dioxide, catalysed by carbonic anhydrase 2. This produces carbonic acid, which rapidly dissociates to form a free proton and a bicarbonate ion. The proton is extruded apically by the apical proton pump, vH+-ATPase, while the bicarbonate ion is reclaimed systemically by being transported by the basolateral anion exchanger 1(AE1, band 3, SLC4A1). The hereditary forms of this disease can be caused by a defect in vH+-ATPase or by a defect in the basolateral anion exchanger, AE1.

AE1 The role of AE1 in hereditary dRTA was first postulated by Wrong, Unwin, and Tanner in 1996 (Wrong et  al., 1996), and this was confirmed a year later by Bruce and co-workers (Bruce et  al.,

vH+-ATPase Mutations of two genes encoding two subunits (B1 and a4) of the vH+-ATPase have been reported, mainly by Karet and co-workers, to cause AR dRTA (Karet et al., 1999; Smith et al., 2000; Stover et al., 2002; Ruf et al., 2003). The respective genes are ATP6V1B1, which encodes the B1 subunit of the catalytic V1 domain and ATP6V0A4, which encodes the a4 subunit in the Vo domain of the vH+-ATPase. Bilateral hearing loss is an early feature of ATP6B1 mutations, due to expression of the ATP6B1 subunit in the cochlea, where it is responsible for the acidification of endolymph (Karet et al., 1999). Although the initial report of the ATP6VOA4 mutations claimed that ATP6VOA4 was expressed only in the α-intercalated cell, and that hearing was preserved (Smith et al., 2000), it was subsequently demonstrated by the same group that affected patients developed late onset deafness, and that there is expression of ATP6VOA4 in the cochlea (Stover et al., 2002). Carbonic anhydrase II mutations cause a syndrome incorporating both dRTA and pRTA. Because of this, some authors have reclassified it as ‘type 3 renal tubular acidosis’ since the original type 3 RTA is no longer seen. Carbonic anhydrase catalyses the hydration of CO2 into carbonic acid from which H+ and HCO3- dissociate. Carbonic anhydrase II (CAII) is the enzyme that performs this function in the α-intercalated cell, where it is necessary for the generation of protons to acidify the urine; and also in the brush border of the proximal tubule cell, where it is crucial for the reabsorption of bicarbonate. This syndrome also includes osteopetrosis and cerebral calcification with subsequent cognitive impairment (Sly et al., 1985; Hu et  al, 1992)  and is very rare, with most cases reported from the Maghreb region of North Africa (Fathallah et al., 1997).

Sjögren syndrome Sjögren syndrome is an autoimmune disorder that has been estimated to affect as much as 1–2% of the adult female population (Kabasakal et al., 2006). It is characterized by the ‘sicca syndrome,’ which is due to an infiltration of inflammatory cells into the lacrimal and salivary glands, resulting in a lack of secretory ability. Indeed,

365

366

Section 2  

fluid, electrolyte, and renal tubular disorders

Sjögren has been dubbed an ‘inflammatory epitheliitis’ (Tzioufas et al., 2012). This is a useful term, because epithelial cells appear to be central targets for a cell-mediated autoimmune response in the lacrimal and salivary glands, as well as the renal tubulointerstitial compartment. In the salivary glands the plasma cells are predominantly T cells (CD4+) when mild and predominantly B cells when severe (Tzioufas et al., 2012). The prevalence of renal involvement in primary Sjögren syndrome (pSS) has been estimated at approximately 30%. The predominant lesion is a tubulointerstitial nephritis (TIN) again related to cellular autoimmunity. It is characterized by a focal or diffuse lymphoplasmocellular infiltrate of mononuclear cells with variable tubular atrophy, often with interstitial fibrosis. The histopathological appearance of the TIN is strikingly similar to that seen in the salivary gland. In the kidney too the majority of invading cells are CD4+ (Skopouli, 2001). The functional consequence of this are: 1. Distal tubular dysfunction, most commonly dRTA and less commonly nephrogenic diabetes insipidus 2. Reduced renal excretory function, with a reduced glomerular filtration rate (GFR) 3. Proximal tubular dysfunction with the Fanconi syndrome has been reported infrequently (Walker et al., 1971). Glomerular involvement may also occur in pSS. This is an unusual complication, occurring late in the disease, if at all (Skopouli, 2001), and appears to be related to humoral autoimmunity. Palpable purpura, low C4, and circulating cryoglobulin are correlated with the development of a glomerulonephritis in Sjögren (Skopouli et al., 2000), suggesting that the pathogenesis involves the deposition of monoclonal immunoglobulin (Ig)-M and polyclonal IgA and IgG in glomeruli.

Treatment and outcome General The mainstay of treatment for dRTA is supportive and is primarily treatment of metabolic acidosis with sodium bicarbonate replacement. Correction of the metabolic acidosis may have a number of beneficial effects. In children, it can reverse osteopenia and restore normal bone growth, even in those with severely stunted skeletal development (McSherry et  al., 1978; Caldas et  al., 1992). It may reduce the risk of forming stones or nephrocalcinosis (Coe et al., 1980; Preminger et al., 1985), and theoretically might reduce urinary potassium loss. The aim of alkali supplementation is to normalize the serum bicarbonate (i.e. > 22 mmol/L). As there is no bicarbonate wasting in dRTA (as opposed to pRTA), the amount of oral sodium bicarbonate needed to do this is relatively modest (around 500 mg to 1g twice daily). Potassium citrate may be substituted for those with problematic calcium phosphate stone disease (as it will help to increase the urinary citrate concentration) or problematic hypokalaemia. Citrate is metabolized to bicarbonate and is an oral alkali supplement in its own right. Many patients end up requiring oral potassium supplements, especially if oral potassium citrate isn’t tolerated or available.

Specific For secondary dRTA, treatment of the underlying disease is the key to treatment. This has been studied in pSS. There are no clinical trials for treating renal pSS, but there are numerous case reports. The main agents that are described in the literature are corticosteroids

alone (Bailey and Swainson, 1986; Akposso et al., 2000; Komatsu et al., 2003; Kawamoto et al., 2005; Kobayashi et al., 2006), or in combination with azathioprine (Kaufman et  al., 2008)  or cyclophosphamide (Goules et al., 2000; Mukai et al., 2001), and a few reports on the successful use of rituximab (Maripuri et al., 2009). Resolution of the functional renal lesions has been repeatedly described, although a return of normal renal acidification has not been formally tested (Kaufman et al., 2008).

Proximal renal tubular acidosis (pRTA, type 2 RTA) Aetiology pRTA is caused by a failure of bicarbonate reabsorption along the proximal tubule. It is, in all but two specific hereditary cases, associated with generalized proximal tubular dysfunction, causing the renal Fanconi syndrome. It has a several causes:

Hereditary ◆ AR

NBCe1 (bicarbonate transporter) loss-of-function mutation (Igarashi et al., 1999)

◆ Carbonic

anhydrase II mutation

◆ Sodium

phosphate transporter (Na-PiIIa) loss of function mutation (Magen et al., 2010)

◆ Cystinosis

(Drablos, 1951)

◆ Tyrosinaemia ◆ Hereditary

(Gentz et al., 1965)

fructose intolerance (Morris et al., 1968)

◆ Galactosaemia ◆ Wilson

(Golberg et al., 1956)

disease (Bearn et al., 1957)

◆ Lowe

syndrome (Bockenhauer et al., 2008)/Dent disease (Wrong et al., 1994)

◆ Fanconi–Bickle ◆ Mitochondrial

syndrome (Manz et al., 1987)

cytopathies (Niaudet and Rotig, 1997).

Acquired ◆ Multiple

myeloma/light chain disease (Maldonado et al., 1975)

◆ Drugs

(tenofovir, ifosfamide (Burk et  al., 1990), carbonic anhydrase inhibitors e.g. acetazolamide, topiramate (Mirza et al., 2009))

◆ Heavy

metals (lead, cadmium (Kazantzis et al., 1963), mercury)

◆ Paroxysmal

nocturnal haemoglobinuria (Riley et al., 1977).

Epidemiology pRTA is rarer than dRTA, and there is even less documentary evidence on its epidemiology. However, the prevalence of some of the acquired causes can be inferred. Myeloma is often thought of as one of the commonest causes of renal Fanconi syndrome, but published series of these patients in the literature are small: the largest series was published by the Mayo clinic and it comprised 32 patients over 34 years (Ma et al., 2004). Tenofovir is an antiretroviral agent that can cause Fanconi syndrome and acute kidney injury (AKI). It was described in 22 patients by Woodward et  al. which represented approximately 1.5% of patients exposed to the agent at the referring centre (Woodward et al., 2009).

chapter 36 

Clinical features

approach to the patient with renal tubular acidosis Table 36.2  Proximal renal tubular acidosis: investigations

The key feature of pRTA is a failure of proximal tubular cells to reabsorb bicarbonate, which leads to bicarbonaturia. It is important to realize that there are other, less efficient bicarbonate reclaiming mechanisms in the more distal tubule (in the loop of Henle and the collecting duct). Therefore, as bicarbonaturia causes the serum bicarbonate to fall, the filtered load of bicarbonate also falls, until it reaches a level at which all of the filtered load can be reclaimed by these less efficient mechanisms. This tends to happen at a serum bicarbonate concentration of approximately14  mmol/L. This has two consequences:  first, it means that the acidosis of pRTA is never particularly severe, as it can be in dRTA; second, it means that bicarbonaturia is not a consistent feature of pRTA, since it will naturally resolve as bicarbonate is lost, and, conversely, will reappear if bicarbonate is replaced (the basis of the bicarbonate infusion test used to confirm the diagnosis). With the exception of isolated mutations of NBCe1 or carbonic anhydrase II mutations, pRTA is always accompanied by signs of generalized proximal tubular dysfunction, which is known as the renal Fanconi syndrome and comprises the pentad of:

Serum biochemistry

Metabolic acidosis: mild (bicarbonate > 12 mmol/L) if present Hypokalaemia: possible if bicarbonaturia ongoing Possible hypophosphataemia Possible hypouricaemia

Urine biochemistry

Urine pH: high if serum bicarbonate >14 mmol/L Glycosuria Phosphaturia (TmP/GFR 12  mmol/L) and there may be hypophosphataemia and occasionally hypouricaemia, if there is also a renal Fanconi syndrome (Table 36.2). Urinary examination may be more revealing. The urine dipstick may reveal glycosuria, although this is not invariable in the Fanconi syndrome, but if it is present in a non-diabetic, it is a strong clue that proximal tubular dysfunction is present. The urine dipstick is usually only a test for albuminuria, rather than low molecular weight proteinuria and is therefore often negative. A laboratory protein/creatinine ratio will detect tubular proteinuria; thus a negative urine dipstick test for protein and a positive protein/creatinine ratio should raise suspicion of the presence of

tubular proteinuria. Urinary phosphate excretion (usually measured as a FEPO4 or TmP/GFR) and urinary urate concentration can be elevated in the Fanconi syndrome. Aminoaciduria is usually difficult and expensive to measure (by chromatography), and is not done routinely. Direct measurement of tubular protein excretion is more straightforward, but is sometimes not locally available. If it is, measurement of retinol binding protein or β2-microglobulin excretion may be helpful in detecting proximal tubular dysfunction. Urinary pH can vary, as indicated above. It will be alkaline when there is bicarbonaturia, but appropriately acid when the bicarbonate threshold is reached and acid excretion challenged with ammonium chloride. When on bicarbonate replacement therapy, there will be bicarbonaturia and alkaline urine. It is unlikely that there will be much difficulty differentiating pRTA from dRTA, given that Fanconi syndrome almost invariably accompanies pRTA. However, the nature of the urinary acidification defect can be tested if there is any uncertainty. In dRTA, the urinary pH will never be below 5.3, even if the patient is challenged with an oral acid load or furosemide and fludrocortisone. In pRTA, once the serum bicarbonate has fallen low enough for all of the filtered bicarbonate to be reabsorbed, the urine can be acidified by the α-intercalated cells normally, and the urine pH may fall below 5.3. A challenge with intravenous sodium bicarbonate (0.5–1 mmol/ kg/h) to raise the serum bicarbonate to 18–20 mmol/L will cause bicarbonate to reappear in the urine and the urine pH will rise rapidly to approximately 7.5. Furthermore, the measured urinary bicarbonate will increase, and a calculation of the fractional excretion of bicarbonate can be done, which will rise to 15% or more (Rose and Post, 2001). Radiography is likely to be unhelpful in pRTA and Fanconi syndrome, since apart from Dent and Lowe syndromes, nephrocalcinosis and nephrolithiasis are not typical features. Bone density may be decreased on DEXA scanning if phosphaturia has been significant for some time.

367

368

Section 2  

fluid, electrolyte, and renal tubular disorders

Aetiology Hereditary Mutations of the bicarbonate transporter SLC4A4 result in an isolated pRTA without renal Fanconi syndrome, but associated with short stature and ocular abnormalities, presumably reflecting the distribution of this transporter in ocular tissue and possibly bone (Igarashi et al., 2001).

Myeloma Almost certainly the most common cause of pRTA is secondary to a monoclonal gammopathy, which like other forms of myeloma-related kidney disease may otherwise be latent (Maldonado et al., 1975). Filtered light chains are endocytosed by proximal tubular cells by the megalin/cubulin pathway for disposal (Sanders et  al., 1988). Some varieties of light chain appear to be more toxic than others, which may be due to the variable domain of these ‘toxic’ light chains being more resistant to lysosomal proteases, and therefore cleared less efficiently (Leboulleux t al., 1995). Accumulation of fragments of these variable domains and subsequent crystallization is apparently responsible for the development of the Fanconi syndrome, at least in a mouse model of myeloma-associated Fanconi syndrome (Decourt et al., 1999). Other acquired causes are increasing in relevance as causes of pRTA, for example, the antiretroviral Tenofovir. Tenofovir is a nucleotide analogue reverse transcriptase inhibitor that inhibits viral replication by being incorporated into the growing DNA strand and stopping elongation. However, they also inhibit DNA polymerase γ, the enzyme involved in mitochondrial DNA (mtDNA) synthesis. Mitochondrial toxicity is almost certainly the mechanism of Tenofovir-associated proximal tubular injury; biopsy specimens show abnormal mitochondria (Woodward et al., 2009; Herlitz et al., 2010), and Tenofovir exposure is associated with mtDNA depletion (Cote et al., 2006). In fact, the proximal tubule is very vulnerable to mitochondrial dysfunction, because it has limited anaerobic respiratory capacity (Bagnasco et al., 1985). Indeed, Fanconi syndrome is a common phenotype in diseases that are caused by mtDNA mutations (Niaudet and Rotig, 1997; Martin-Hernandez et al., 2005); other drugs that are mitochondrial toxins (e.g. ifosphamide, valproate, aminoglycosides) also cause Fanconi syndrome (Izzedine et al., 2003).

Correction of hypophosphataemia, if present, will reverse incipient or overt osteomalacia. Replacement of phosphate (e.g. phosphate Sandoz 2 tablets BD) reverses biochemical abnormalities (often including raised alkaline phosphatase, secondary hyperparathyroidism and inappropriately low serum calcitriol levels) (Clarke et al., 1995), symptoms (e.g. bone pain) and increases bone mineral density. Replacement of vitamin D (not calcitriol) may also be necessary (Clarke et al., 1995). Caution is needed in not prescribing too much of either supplement: phosphate can stimulate PTH secretion and vitamin D-induced hypercalcaemia may lead to nephrocalcinosis.

Hyperkalaemic renal tubular acidosis (hypoaldosteronism, type 4 RTA) Introduction A number of hyperkalaemic conditions that are variably associated with a mild metabolic acidosis were classified by Morris and McSherry (1972) as ‘type 4 renal tubular acidosis, and so merit some consideration here. They are entirely different from dRTA and pRTA, and are never part of the same differential diagnosis. They are typified by real or apparent hypoaldosteronism, and the cardinal feature is hyperkalaemia.

Aetiology ◆ Hypoaldosteronism ◆ Diabetes ◆ Adrenal

and renal impairment

insufficiency

◆ Non-steroidal ◆ Calcineurin ◆ ACE

anti-inflammatory agent-associated TIN

inhibitors

inhibitors

◆ Angiotensin

receptor blockers

◆ Heparin ◆ Pseudohypoaldosteronism ◆ Congenital

◆ Aldosterone ◆ Potassium

type 2 (Gordon syndrome)

isolated hypoaldosteronism resistance

sparing diuretics (e.g. spironolactone, amiloride)

Treatment and outcome

◆ Trimethoprim

The aim of treatment of pRTA and the renal Fanconi syndrome is chiefly protection of the skeleton. This is achieved in two ways, by correction of the metabolic acidosis and also by replacement of any phosphate losses, if these are prominent. Correction of the acidosis is desirable, because it will reverse osteomalacia/rickets (Brenner et  al., 1982)  and restore normal bone growth in growth-delayed children (1981). However, achieving a normal serum bicarbonate concentration is much more difficult than in dRTA, since oral bicarbonate replacement will result in bicarbonaturia and higher bicarbonate requirements. The bicarbonaturia will also stimulate a kaliuresis and thus tend to cause hypokalaemia, as discussed previously. Therefore, the dose of sodium bicarbonate required to normalize the serum bicarbonate concentration is much higher than in dRTA, and may be as much as 4–6 g twice daily. If hypokalaemia is problematic, potassium replacement may be necessary, in which case, some of the alkali load could be given as potassium citrate or bicarbonate.

◆ Pseudohypoaldosteronism

and pentamidine (also inhibit ENaC) type 1.

Clinical features Hyperkalaemia This is usually mild, unless combined with another factor that tends to increase the serum potassium (e.g. ACE inhibitors).

Metabolic acidosis This is variable, mild when present and is a hyperchloraemic (i.e. normal anion gap) acidosis. It is thought to be due to reduced ammonium excretion.

Investigations Serum biochemistry may reveal a mild hyperkalaemia and a mild metabolic acidosis. Also, the plasma renin activity (or renin concentration) and serum aldosterone level should be measured. The following should be considered in diagnosis:

chapter 36 

approach to the patient with renal tubular acidosis

plasma renin activity/renin concentration and serum aldosterone:

Batlle, D. C., Sabatini, S., and Kurtzman, N. A. (1988). On the mechanism of toluene-induced renal tubular acidosis. Nephron, 49(3), 210–18. Bearn, A. G., Yu, T. F., and Gutman, A. B. (1957). Renal function in Wilson’s disease. J Clin Invest, 36(7), 1107–14. Bockenhauer, D., Bokenkamp, A., van’t Hoff, W., et al. (2008). Renal phenotype in Lowe syndrome: a selective proximal tubular dysfunction. Clin J Am Soc Nephrol, 3(5), 1430–6. Boddy, A. V., English, M., Pearson, A. D., et al. (1996). Ifosfamide nephrotoxicity: limited influence of metabolism and mode of administration during repeated therapy in paediatrics. Eur J Cancer, 32A(7), 1179–84. Boton, R., Gaviria, M., and Batlle, D. C. (1987). Prevalence, pathogenesis, and treatment of renal dysfunction associated with chronic lithium therapy. Am J Kidney Dis, 10(5), 329–45. Brenner, R. J., Spring, D. B., Sebastian, A., et al. (1982). Incidence of radiographically evident bone disease, nephrocalcinosis, and nephrolithiasis in various types of renal tubular acidosis. N Engl J Med, 307(4), 217–21. Bruce, L. J., Cope, D. L., Jones, G. K., et al. (1997). Familial distal renal tubular acidosis is associated with mutations in the red cell anion exchanger (Band 3, AE1) gene. J Clin Invest, 100(7), 1693–707. Bruce, L. J., Wrong, O., Toye, A. M., et al. (2000). Band 3 mutations, renal tubular acidosis and South-East Asian ovalocytosis in Malaysia and Papua New Guinea: loss of up to 95% band 3 transport in red cells. Biochem J, 350 Pt 1, 41–51. Buchanan, E. U. and Komrower, G. M. (1958). The prognosis of idiopathic renal acidosis in infancy with observations on urine acidification and ammonia production in children. Arch Dis Child, 33(172), 532–5. Burk, C. D., Restaino, I., Kaplan, B. S., et al. (1990). Ifosfamide-induced renal tubular dysfunction and rickets in children with Wilms tumor. J Pediatr, 117(2 Pt 1), 331–5. Butler, A. M., Wilson, J. L., and Farber, S. (1936). Dehydration and acidosis with calcification at renal tubules. J Pediatr, 8, 489–99. Caldas, A., Broyer, M., Dechaux, M., et al. (1992). Primary distal tubular acidosis in childhood: clinical study and long-term follow-up of 28 patients. J Pediatr, 121(2), 233–41. Clarke, B. L., Wynne, A. G., Wilson, D. M., et al. (1995). Osteomalacia associated with adult Fanconi’s syndrome: clinical and diagnostic features. Clin Endocrinol, 43(4), 479–90. Coe, F. L. and Parks, J. H. (1980). Stone disease in hereditary distal renal tubular acidosis. Ann Internal Med, 93(1), 60–1. Cote, H. C., Magil, A. B., Harris, M., et al. (2006). Exploring mitochondrial nephrotoxicity as a potential mechanism of kidney dysfunction among HIV-infected patients on highly active antiretroviral therapy. Antivir Ther, 11(1), 79–86. Decourt, C., Rocca, A., Bridoux, F., et al. (1999). Mutational analysis in murine models for myeloma-associated Fanconi’s syndrome or cast myeloma nephropathy. Blood, 94(10), 3559–66. Deutschmann, H. A., Weger, M., Weger, W., et al. (2002). Search for occult secondary osteoporosis: impact of identified possible risk factors on bone mineral density. J Intern Med, 252(5), 389–97. Domrongkitchaiporn, S., Pongskul, C., Sirikulchayanonta, V., et al. (2002). Bone histology and bone mineral density after correction of acidosis in distal renal tubular acidosis. Kidney Int, 62(6), 2160–6. Drablos, A. (1951). The de Toni-Fanconi syndrome with cystinosis. Acta Paediatr, 40(5), 438–49. Fathallah, D. M., Bejaoui, M., Lepaslier, D., et al. (1997). Carbonic anhydrase II (CA II) deficiency in Maghrebian patients: evidence for founder effect and genomic recombination at the CA II locus. Hum Genet, 99(5), 634–7. Gentz, J., Jagenburg, R., and Zetterstroem, R. (1965). Tyrosinemia. J Pediatr, 66, 670–96. Golberg, L., Holzel, A., Komrower, G. M., et al. (1956). A clinical and biochemical study of galactosaemia; a possible explanation of the nature of the biochemical lesion. Arch Dis Child, 31(158), 254–64. Golding, P. L. and Mason, A. S. (1971). Renal tubular acidosis and autoimmune liver disease. Gut, 12(2), 153–7. Goossens, J. P., Statius van Eps, L. W., Schouten, H., et al. (1972). Incomplete renal tubular acidosis in sickle cell disease. Clin Chim Acta, 41, 149–56.

◆ Low

◆ Diabetes ◆ Adrenal

and renal impairment

insufficiency

◆ Non-steroidal ◆ PHA

anti-inflammatory agent associated with TIN

2/Gordon syndrome

◆ Calcineurin

inhibitors

◆ High

plasma renin activity/renin concentration and low serum aldosterone:

◆ ACE

inhibitors

◆ ARBs ◆ Heparin ◆ Congenital

isolated hypoaldosteronism

◆ High

plasma renin activity/renin concentration and serum aldosterone:

◆ Potassium-sparing

diuretics (e.g. spironolactone, amiloride)

◆ Trimethoprim

and pentamidine (also inhibit epithelial sodium channels (ENaC))

◆ Pseudohypoaldosteronism

type 1.

Aetiology The hallmark of hyperaldosteronism is a hypokalaemic metabolic alkalosis, and the hyperkalaemia and acidosis seen in type 4 dRTA is a mirror image of this. Aldosterone causes increased expression of ENaC on the apical surface of principal cells in the cortical collecting duct (CCD). ENaC reabsorbs sodium from tubular fluid, which depolarizes the apical membrane of the principal cell leading to a negative transepithelial electrical potential difference along the CCD. This negative potential favours the secretion of cations (either protons or potassium) from the α-intercalated cell. This process is increased when there is more aldosterone acting on the CCD, resulting in hypokalaemia (from potassium losses in the urine) and alkalosis (from increased secretion of hydrogen ions, which can form ammonium ions in the urine). In conditions of reduced aldosterone action, there is less lumen electronegativity, so less driving force for potassium secretion (causing a tendency to hyperkalaemia) and less proton secretion (causing a tendency toward acidosis and reduced ammonium excretion).

Treatment and outcome Hypoaldosteronism (but not aldosterone resistance) respond well to fludrocortisone (0.05–0.2 mg/day for adrenal insufficiency, 0.2–1 mg/day for other causes). However, it will worsen hypertension and fluid overload, so if these are present, a loop or thiazide diuretic can be used instead.

References Akposso, K., Martinant de Preneuf, H., Larousserie, F., et al. (2000). [Rapidly progressive acute renal failure. A rare complication of primary Sjogren syndrome]. Presse Med, 29(30), 1647–9. Albright, F. and Reifenstein, E. C. (1948). The Parathyroid Glands and Metabolic Bone Disease. Baltimore, MD: Williams & Wilkins. Bagnasco, S., Good, D., Balaban, R., et al. (1985). Lactate production in isolated segments of the rat nephron. Am J Physiol, 248(4 Pt 2), F522–6. Bailey, R.R. and Swainson, C. P. (1986). Renal involvement in Sjogren’s syndrome. N Z Med J, 99(807), 579–80.

369

370

Section 2  

fluid, electrolyte, and renal tubular disorders

Goules, A., Masouridi, S., Tzioufas, A. G., et al. (2000). Clinically significant and biopsy-documented renal involvement in primary Sjogren syndrome. Medicine, 79(4), 241–9. Herlitz, L. C., Mohan, S., Stokes, M. B., et al. (2010). Tenofovir nephrotoxicity: acute tubular necrosis with distinctive clinical, pathological, and mitochondrial abnormalities. Kidney Int, 78(11), 1171–7. Hoppe, B., Neuhaus, T., Superti-Furga, A., et al. (1993). Hypercalciuria and nephrocalcinosis, a feature of Wilson’s disease. Nephron, 65(3), 460–2. Hu, P. Y., Roth, D. E., Skaggs, L. A., et al. (1992). A splice junction mutation in intron 2 of the carbonic anhydrase II gene of osteopetrosis patients from Arabic countries. Hum Mutat, 1(4), 288–92. Igarashi, T., Inatomi, J., Sekine, T., et al. (1999). Mutations in SLC4A4 cause permanent isolated proximal renal tubular acidosis with ocular abnormalities. Nat Genet, 23(3), 264–6. Igarashi, T., Inatomi, J., Sekine, T., et al. (2001). Novel nonsense mutation in the Na+/HCO3- cotransporter gene (SLC4A4) in a patient with permanent isolated proximal renal tubular acidosis and bilateral glaucoma. J Am Soc Nephrol, 12(4), 713–18. Izzedine, H., Launay-Vacher, V., Isnard-Bagnis, C., et al. (2003). Drug-induced Fanconi’s syndrome. Am J Kidney Dis, 41(2), 292–309. Kabasakal, Y., Kitapcioglu, G., Turk, T., et al. (2006). The prevalence of Sjogren’s syndrome in adult women. Scand J Rheumatol, 35(5), 379–83. Karet, F. E., Finberg, K. E., Nelson, R. D., et al. (1999). Mutations in the gene encoding B1 subunit of H+-ATPase cause renal tubular acidosis with sensorineural deafness. Nat Genet, 21(1), 84–90. Karet, F. E., Gainza, F. J., Gyory, A. Z., et al. (1998). Mutations in the chloride-bicarbonate exchanger gene AE1 cause autosomal dominant but not autosomal recessive distal renal tubular acidosis. Proc Natl Acad Sci U S A, 95(11), 6337–42. Kaufman, I., Schwartz, D., Caspi, D., et al. (2008). Sjogren’s syndrome—not just Sicca: renal involvement in Sjogren’s syndrome. Scand J Rheumatol, 37(3), 213–18. Kawamoto, S., Ichinose, M., Ito, Y., et al. (2005). [Interstitial pneumonia and nephritis with Sjogren’s syndrome: successful treatment with corticosteroid therapy]. Nihon Jinzo Gakkai Shi, 47(4), 451–7. Kazantzis, G., Flynn, F. V., Spowage, J. S., et al. (1963). Renal tubular malfunction and pulmonary emphysema in cadmium pigment workers. QJM, 32, 165–92. Khositseth, S., Bruce, L. J., Walsh, S. B., et al. (2012). Tropical distal renal tubular acidosis: clinical and epidemiological studies in 78 patients. QJM, 105(9), 861–77. Kobayashi, T., Muto, S., Nemoto, J., et al. (2006). Fanconi’s syndrome and distal (type 1) renal tubular acidosis in a patient with primary Sjogren’s syndrome with monoclonal gammopathy of undetermined significance. Clin Nephrol, 65(6), 427–32. Komatsu, H., Hara, S., Kikuchi, M., et al. (2003). [Two cases of interstitial nephritis with primary Sjogren’s syndrome successfully treated by steroid therapy]. Nihon Jinzo Gakkai Shi, 45(4), 398–404. Leboulleux, M., Lelongt, B., Mougenot, B., et al. (1995). Protease resistance and binding of Ig light chains in myeloma-associated tubulopathies. Kidney Int, 48(1), 72–9. Lemann, J., Jr., Adams, N. D., Wilz, D. R., et al. (2000). Acid and mineral balances and bone in familial proximal renal tubular acidosis. Kidney Int, 58(3), 1267–77. Lightwood, R. (1936). Communication No. 1. Arch Dis Child, 10, 205. Lightwood, R., Payne, W. W., and Black, J. A. (1953). Infantile renal acidosis. Pediatrics, 12(6), 628–44. Ma, C. X., Lacy, M. Q., Rompala, J. F., et al. (2004). Acquired Fanconi syndrome is an indolent disorder in the absence of overt multiple myeloma. Blood, 104(1), 40–2. Magen, D., Berger, L., Coady, M. J., et al. (2010). A loss-of-function mutation in NaPi-IIa and renal Fanconi’s syndrome. N Engl J Med, 362(12), 1102–9. Maldonado, J. E., Velosa, J. A., Kyle, R. A., et al. (1975). Fanconi syndrome in adults. A manifestation of a latent form of myeloma. Am J Med, 58(3), 354–64. Manz, F., Bickel, H., Brodehl, J., et al. (1987). Fanconi-Bickel syndrome. Pediatr Nephrol, 1(3), 509–18.

Maripuri, S., Grande, J. P., Osborn, T. G., et al. (2009). Renal involvement in primary Sjogren’s syndrome: a clinicopathologic study. Clin J Am Soc Nephrol, 4(9), 1423–31. Martin-Hernandez, E., Garcia-Silva, M. T., Vara, J., et al. (2005). Renal pathology in children with mitochondrial diseases. Pediatr Nephrol, 20(9), 1299–305. Mason, A. M. and Golding, P. L. (1970). Renal tubular acidosis and autoimmune thyroid disease. Lancet, 2(7683), 1104–7. McCurdy, D. K., Cornwell, G. G., 3rd, and DePratti, V. J. (1967). Hyperglobulinemic renal tubular acidosis. Report of two cases. Ann Internal Med, 67(1), 110–17. McCurdy, D. K., Frederic, M., and Elkinton, J. R. (1968). Renal tubular acidosis due to amphotericin B. N Engl J Med, 278(3), 124–30. McSherry, E. (1981). Renal tubular acidosis in childhood. Kidney Int, 20(6), 799–809. McSherry, E. and Morris, R. C., Jr. (1978). Attainment and maintenance of normal stature with alkali therapy in infants and children with classic renal tubular acidosis. J Clin Invest, 61(2), 509–27. Mirza, N., Marson, A. G., and Pirmohamed, M. (2009). Effect of topiramate on acid-base balance: extent, mechanism and effects. Br J Clin Pharmacol, 68(5), 655–61. Morris, R. C., Jr. (1968). An experimental renal acidification defect in patients with hereditary fructose intolerance. II. Its distinction from classic renal tubular acidosis; its resemblance to the renal acidification defect associated with the Fanconi syndrome of children with cystinosis. J Clin Invest, 47(7), 1648–63. Morris, R. C., Jr. and McSherry, E. (1972). Symposium on acid-base homeostasis. Renal acidosis. Kidney Int, 1(5), 322–40. Mukai, M., Shibata, T., Honda, H., et al. (2001). [A case of Sjogren’s syndrome presenting with hypokalemic myopathy due to renal tubular acidosis]. Nihon Jinzo Gakkai Shi, 43(2), 69–75. Niaudet, P. and Rotig, A. (1997). The kidney in mitochondrial cytopathies. Kidney Int, 51(4), 1000–7. Nilwarangkur, S., Nimmannit, S., Chaovakul, V., et al. (1990). Endemic primary distal renal tubular acidosis in Thailand. QJM, 74(275), 289–301. Nimmannit, S., Malasit, P., Susaengrat, W., et al. (1996). Prevalence of endemic distal renal tubular acidosis and renal stone in the northeast of Thailand. Nephron, 72(4), 604–10. Palkar, A. V., Pillai, S., and Rajadhyaksha, G. C. (2011). Hypokalemic quadriparesis in Sjogren syndrome. Indian J Nephrol, 21(3), 191–3. Pasternack, A., Martio, J., Nissila, M., et al. (1970). Renal acidification and hypergammaglobulinemia. A study of rheumatoid arthritis. Acta Med Scand, 187(1–2), 123–7. Pertovaara, M., Korpela, M., Kouri, T., et al. (1999). The occurrence of renal involvement in primary Sjogren’s syndrome: a study of 78 patients. Rheumatology, 38(11), 1113–20. Pines, K. L. and Mudge, G. H. (1951). Renal tubular acidosis with osteomalacia; report of 3 cases. Am J Med, 11(3), 302–11. Preminger, G. M., Sakhaee, K., Skurla, C., et al. (1985). Prevention of recurrent calcium stone formation with potassium citrate therapy in patients with distal renal tubular acidosis. J Urol, 134(1), 20–3. Ren, H., Wang, W. M., Chen, X. N., et al. (2008). Renal involvement and followup of 130 patients with primary Sjogren’s syndrome. J Rheumatol, 35(2), 278–84. Ribeiro, M. L., Alloisio, N., Almeida, H., et al. (2000). Severe hereditary spherocytosis and distal renal tubular acidosis associated with the total absence of band 3. Blood, 96(4), 1602–4. Riley, A. L., Ryan, L. M., and Roth, D. A. (1977). Renal proximal tubular dysfunction and paroxysmal nocturnal hemoglobinuria. Am J Med, 62(1), 125–9. Rodriguez-Soriano, J. and Edelmann, C. M., Jr. (1969). Renal tubular acidosis. Annu Rev Med, 20, 363–82. Rose, B. D. and Post, T. W. (2001). Clinical Physiology of Acid-Base and Electrolyte Disorders (5th ed.). New York: McGraw-Hill, Medical Pub. Division. Ruf, R., Rensing, C., Topaloglu, R., et al. (2003). Confirmation of the ATP6B1 gene as responsible for distal renal tubular acidosis. Pediatr Nephrol, 18(2), 105–9.

chapter 36 

approach to the patient with renal tubular acidosis

Sanders, P. W., Herrera, G. A., Lott, R. L., et al. (1988). Morphologic alterations of the proximal tubules in light chain-related renal disease. Kidney Int, 33(4), 881–9. Sebastian, A., McSherry, E., and Morris, R. C., Jr. (1971). Renal potassium wasting in renal tubular acidosis (RTA): its occurrence in types 1 and 2 RTA despite sustained correction of systemic acidosis. J Clin Invest, 50(3), 667–78. Shear, L., Bonkowsky, H. L., and Gabuzda, G. J. (1969). Renal tubular acidosis in cirrhosis. A determinant of susceptibility to recurrent hepatic precoma. N Engl J Med, 280(1), 1–7. Shearn, M. A. and Tu, W. H. (1968). Latent renal tubular acidosis in Sjogren’s syndrome. Ann Rheum Dis, 27(1), 27–32. Siamopoulos, K. C., Elisaf, M., Drosos, A. A., et al. (1992). Renal tubular acidosis in primary Sjogren’s syndrome. Clin Rheumatol, 11(2), 226–30. Skopouli, F. N. (2001). Kidney injury in Sjogren’s syndrome. Nephrol Dial Transplant, 16 Suppl 6, 63–4. Skopouli, F. N., Dafni, U., Ioannidis, J. P., et al. (2000). Clinical evolution, and morbidity and mortality of primary Sjogren’s syndrome. Semin Arthritis Rheum, 29(5), 296–304. Sly, W. S., Whyte, M. P., Sundaram, V., et al. (1985). Carbonic anhydrase II deficiency in 12 families with the autosomal recessive syndrome of osteopetrosis with renal tubular acidosis and cerebral calcification. N Engl J Med, 313(3), 139–45. Smith, A. N., Skaug, J., Choate, K. A., et al. (2000). Mutations in ATP6N1B, encoding a new kidney vacuolar proton pump 116-kD subunit, cause recessive distal renal tubular acidosis with preserved hearing. Nat Genet, 26(1), 71–5. Sritippayawan, S., Kirdpon, S., Vasuvattakul, S., et al. (2003). A de novo R589C mutation of anion exchanger 1 causing distal renal tubular acidosis. Pediatric Nephrol, 18(7), 644–8. Stapleton, T. (1949). Idiopathic renal acidosis in an infant with excessive loss of bicarbonate in the urine. Lancet, 1(6556), 683–5. Stapleton, T. (1954). The response of idiopathic renal acidosis to oral sodium lactate. Acta Paediatr, 43(1), 49–63. Stover, E. H., Borthwick, K. J., Bavalia, C., et al. (2002). Novel ATP6V1B1 and ATP6V0A4 mutations in autosomal recessive distal renal tubular acidosis with new evidence for hearing loss. J Med Genet, 39(11), 796–803. Talal, N., Zisman, E., and Schur, P. H. (1968). Renal tubular acidosis, glomerulonephritis and immunologic factors in Sjogren’s syndrome. Arthritis Rheum, 11(6), 774–86. Tanphaichitr, V. S., Sumboonnanonda, A., Ideguchi, H., et al. (1998). Novel AE1 mutations in recessive distal renal tubular acidosis. Loss-of-function is rescued by glycophorin A. J Clin Invest, 102(12), 2173–9. Tu, W. H. and Shearn, M. A. (1967). Systemic lupus erythematosus and latent renal tubular dysfunction. Ann Internal Med, 67(1), 100–9. Tzioufas, A. G., Kapsogeorgou, E. K., and Moutsopoulos, H. M. (2012). Pathogenesis of Sjogren’s syndrome: what we know and what we should learn. J Autoimmun, 39(1–2), 4–8.

Van der Heijden, A. J., Versteegh, F. G., Wolff, E. D., et al. (1985). Acute tubular dysfunction in infants with obstructive uropathy. Acta Paediatr Scand, 74(4), 589–94. Vasuvattakul, S., Yenchitsomanus, P. T., Vachuanichsanong, P., et al. (1999). Autosomal recessive distal renal tubular acidosis associated with Southeast Asian ovalocytosis. Kidney Int, 56(5), 1674–82. Walker, B. R., Alexander, F., and Tannenbaum, P. J. (1971). Fanconi syndrome with renal tubular acidosis and light chain proteinuria. Nephron, 8(1), 103–7. Walsh, S. B., Shirley, D. G., Wrong, O. M., et al. (2007). Urinary acidification assessed by simultaneous furosemide and fludrocortisone treatment: an alternative to ammonium chloride. Kidney Int, 71(12), 1310–16. Walsh, S., Borgese, F., Gabillat, N., et al. (2009). Southeast Asian AE1 associated renal tubular acidosis: cation leak is a class effect. Biochem Biophys Res Commun, 382(4), 668–72. Weber, S., Soergel, M., Jeck, N., et al. (2000). Atypical distal renal tubular acidosis confirmed by mutation analysis. Pediatric Nephrol, 15(3–4), 201–4. Weger, M., Deutschmann, H., Weger, W., et al. (1999). Incomplete renal tubular acidosis in ‘primary’ osteoporosis. Osteoporos Int, 10(4), 325–9. Weger, W., Kotanko, P., Weger, M., et al. (2000). Prevalence and characterization of renal tubular acidosis in patients with osteopenia and osteoporosis and in non-porotic controls. Nephrol Dial Transplant, 15(7), 975–80. Wilson, D. R. and Siddiqui, A. A. (1973). Renal tubular acidosis after kidney transplantation. Natural history and significance. Ann Internal Med, 79(3), 352–61. Woodward, C. L., Hall, A. M., Williams, I. G., et al. (2009). Tenofovir-associated renal and bone toxicity. HIV Med, 10(8), 482–7. Wrong, O. and Davies, H. E. (1959). The excretion of acid in renal disease. QJM, 28(110), 259–313. Wrong, O. M., Norden, A. G., and Feest, T. G. (1994). Dent’s disease; a familial proximal renal tubular syndrome with low-molecular-weight proteinuria, hypercalciuria, nephrocalcinosis, metabolic bone disease, progressive renal failure and a marked male predominance. QJM, 87(8), 473–93. Wrong, O., Bruce, L. J., Unwin, R. J., et al. (2002). Band 3 mutations, distal renal tubular acidosis, and Southeast Asian ovalocytosis. Kidney Int, 62(1), 10–9. Wrong, O., Unwin, R., Cohen, E., et al. (1996). Unravelling of the molecular mechanisms of kidney stones. Report of a Meeting of Physicians and Scientists. Lancet, 348(9041), 1561–5. Yenchitsomanus, P. T., Sawasdee, N., Paemanee, A., et al. (2003). Anion exchanger 1 mutations associated with distal renal tubular acidosis in the Thai population. J Hum Genet, 48(9), 451–6. Zimhony, O., Sthoeger, Z., Ben David, D., et al. (1995). Sjogren’s syndrome presenting as hypokalemic paralysis due to distal renal tubular acidosis. J Rheumatol, 22(12), 2366–8.

371

CHAPTER 37

Approach to the patient with hypercalcaemia Dennis Joseph and Theresa A. Guise Introduction Definition of hypercalcaemia The normal range of total serum calcium concentration (0.25  × mg/dL  =  mmol/L), corrected for albumin, is 2.2–2.6  mmol/L (8.7–10.2 mg/dL) (Kratz et al., 2011). Hypercalcaemia is defined as an increased calcium concentration in the blood, usually > 2 standard deviations (SDs) above the normal mean in a given laboratory. Forty per cent of serum calcium is bound to protein, primarily albumin. Hence, pseudohypercalcaemia (with normal levels of ionized calcium) may be present in the setting of hyperalbuminemia (due to dehydration) or a paraproteinemia (due to multiple myeloma). Alternatively, hypercalcaemia can be present in hypoalbuminemia with normal total serum calcium levels, but elevated ionized calcium levels. Alterations in acid–base balance also affect the concentration of ionized calcium; ionized calcium increases by 0.05 mmol/L (0.2 mg/dL) when pH decreases by 0.1. Hypercalcaemia results from an abnormality in the calcium flux between the extracellular fluid and the main calcium regulatory compartments: bone, gastrointestinal tract (GIT), and kidney. Thus, excessive bone resorption, increased gastrointestinal calcium absorption, reduced renal excretion of calcium, or a combination of these, are responsible for hypercalcaemia.

Symptoms and signs Mild hypercalcaemia (calcium < 3  mmol/L (12 mg/dL)) may be asymptomatic and is usually discovered incidentally on routine blood tests. However, marked symptoms occur with acute increases in serum calcium levels > 3  mmol/L. The severity of symptoms depends on the degree of hypercalcaemia, the rapidity of rise in serum calcium concentration and co-morbidities present in the patient (Stewart, 2005). The common symptoms associated with hypercalcaemia are listed in Table 37.1.

ingested calcium is absorbed through the small and large intestine. Active transport occurs via transient receptor potential vanilloid type 6 (TRPV6) and an intracellular protein known as calbindin, and is principally regulated by 1,25-dihydroxyvitamin D (Barger-Lux et al., 1989). Additionally, passive transport of calcium occurs paracellularly in proportion to the calcium intake. The GIT is also a site of constant calcium secretion approximating about 150 mg daily, resulting in a net calcium intake of 150 mg per day. The newly acquired 150 mg of dietary calcium enters the blood stream and is filtered by the kidney, where 98% of calcium is reabsorbed in the proximal convoluted tubule, while the remaining excess calcium is eventually excreted in the urine. The plasma calcium concentration also remains in a steady state in dynamic equilibrium with the calcium pool in the bones. Thus, the kidneys, bone, and GIT are the main regulators of plasma (or extracellular) calcium homeostasis, which is tightly controlled by the hormones parathyroid hormone (PTH) and 1,25-dihydroxyvitamin D. The calcium-sensing receptor (CaSR) is a G protein-coupled receptor that senses extracellular levels of calcium ions (Brown et al., 1993). It is present in the parathyroid gland, distal nephron of the kidney, and in bone. In the parathyroid gland, the CaSR controls calcium homeostasis by regulating the release of PTH. PTH acts to increase the concentration of calcium in the blood by increasing bone resorption and distal tubular reabsorption of calcium. In addition, it indirectly enhances gastrointestinal calcium absorption by increasing the activity of 1α-hydroxylase enzyme, which converts 25-hydroxyvitamin D (25-OH vitamin D) to 1,25-dihydroxyvitamin D, the active form of vitamin D.  PTH enhances the release of calcium from bone by indirectly activating osteoclasts through osteoblast expression of receptor activator of nuclear factor kappa B ligand (RANKL). Calcitonin (a polypeptide hormone produced by the parafollicular cells (C cells) of the thyroid gland) acts to decrease calcium concentration by directly inhibiting bone resorption and increasing renal excretion of calcium. The physiological role of calcitonin in humans remains uncertain.

Pathophysiology Normal calcium metabolism

Causes of hypercalcaemia

The adult human body contains about 1100 g of calcium (Barrett et al., 2011). Ninety-nine per cent of this is present in the skeletal framework formed by bones and teeth in the form of calcium phosphate or hydroxyapatite (Ca5[PO4]3[OH]). The daily dietary intake of elemental calcium in a healthy adult consuming a Western diet is roughly 1 g. About 30% (300 mg) of the

Hypercalcaemia occurs when there is a mismatch between the entry of calcium into plasma (from bone or gut) and its removal through bone deposition and renal excretion. The most common causes of hypercalcaemia are primary hyperparathyroidism and malignancy, which together account for > 90% of cases of hypercalcaemia. These conditions can easily be differentiated by measurement of intact

chapter 37 

Table 37.1  Clinical features of hypercalcaemia Neurologic and psychiatric

Lethargy, drowsiness Confusion, disorientation Disturbed sleep, nightmares Irritability, depression Hypotonia, decreased deep tendon reflexes Stupor, coma

Gastrointestinal

Anorexia, vomiting Constipation Peptic ulceration Acute pancreatitis

Cardiovascular

Arrhythmias Synergism with digoxin Hypertension

Renal

Polyuria, polydipsia Hypercalciuria Nephrocalcinosis Impaired glomerular filtration

serum PTH levels, which are elevated or inappropriately normal in primary hyperparathyroidism, and suppressed in hypercalcaemia of malignancy. Other causes of hypercalcaemia associated with increased PTH levels include tertiary hyperparathyroidism, usually due to chronic renal failure, lithium-induced hypercalcaemia, and familial hypocalciuric hypercalcaemia (FHH). Suppressed levels of PTH are seen in hypercalcaemia due to chronic granulomatous diseases, including sarcoidosis, acute kidney injury, immobilization, thyrotoxicosis, thiazide diuretic use, and rarer causes such as hypervitaminosis D, milk-alkali syndrome, hypervitaminosis A, and adrenal insufficiency.

Conditions with elevated PTH levels Primary hyperparathyroidism The routine measurement of serum calcium with the use of automated multichannel biochemical screening has led to a marked rise in the incidence of primary hyperparathyroidism. Most cases occur after age 45 (Wermers et al., 2006) and women are twice as likely to be affected as men, likely due to increased bone resorption occurring at the menopause, which unmasks primary hyperparathyroidism. The causes of primary hyperparathyroidism are poorly understood. The better recognized causes include a history of ionizing radiation to the neck (Beard et al., 1989), chronically low calcium intake (Paik et al., 2012), chronic lithium exposure, and rare genetic abnormalities such as the multiple endocrine neoplasia syndromes (Marx et al., 2002). The majority of cases of primary hyperparathyroidism are due to a single adenoma, while the rest are due to multiple gland hyperplasia. Parathyroid carcinoma is a very rare cause of primary hyperparathyroidism (Wynne et  al., 1992). The diagnosis of primary

approach to the patient with hypercalcaemia

hyperparathyroidism is made when PTH concentrations are high or inappropriately normal in the setting of high serum calcium, when other causes of hypercalcaemia with elevated PTH (especially FHH) have been excluded. Other laboratory findings may include a urinary calcium to creatinine clearance ratio > 0.02, and reduced serum phosphate levels. Surgical resection is the definitive treatment for primary hyperparathyroidism. However, since many patients are asymptomatic and may not progress to symptomatic primary hyperparathyroidism, periodic monitoring of calcium levels is an option. Indications for parathyroidectomy are serum calcium concentration of 0.25 mmol/L (1.0 mg/dL) or more above the upper limit of normal, a creatinine clearance < 60 mL/min, recurrent nephrolithiasis, bone density at the hip, lumbar spine, or distal 1/3 radius that is more than 2.5 SDs below peak bone mass (T score < −2.5), and/or previous fragility fracture and age < 50  years (Bilezikian et al., 2009). Calcimimetic agents (such as cinacalcet) that interact with the CaSR to lower PTH, may be used for primary hyperparathyroidism when surgery is not possible or high risk (Peacock et al., 2005). Lithium-induced hypercalcaemia Hypercalcaemia induced by lithium is a subtype of primary hyperparathyroidism and results from lithium’s effect of decreasing parathyroid gland sensitivity to calcium, shifting the set-point of the Ca-PTH curve to the right (Mallette et al., 1989). This condition is frequently irreversible, even with discontinuation of lithium in those who have been on chronic therapy for more than ten years, and may require parathyroidectomy.

Tertiary hyperparathyroidism Stimuli such as calcitriol deficiency and hyperphosphatemia, as in chronic kidney disease, or long-term administration of phosphate and vitamin D preparations in X-linked hypophosphatemic rickets (Makitie et al., 2003), or chronic hypocalcemia, cause increased mitotic activity in parathyroid cells, leading to parathyroid gland hyperplasia and secondary hyperparathyroidism. With prolonged secondary hyperparathyroidism, the parathyroid glands can autonomously produce PTH, a condition known as tertiary hyperparathyroidism. Medical management options in this condition include calcimimetics such as cinacalcet. The main indications for parathyroidectomy for tertiary hyperparathyroidism in dialysis patients include severe refractory hypercalcaemia, progressive hyperparathyroid bone disease, intractable pruritus, or progressive extraskeletal calcification or calciphylaxis.

Familial hypocalciuric hypercalcaemia FHH is a benign cause of hypercalcaemia characterized by mild hypercalcaemia and hypocalciuria. It is a genetically heterogeneous autosomal dominant disorder with three variants: types 1, 2, and 3. FHH1 is caused by inactivating mutations in the gene for the CaSR located on the long arm of chromosome 3 (Pollak et al., 1993). The CaSR mutation decreases the sensitivity of the parathyroids and kidney to calcium, resulting in mildly elevated or inappropriately normal PTH levels and increased tubular reabsorption of calcium. FHH2 and FHH2 are caused by mutations in GNA11 and AP2S1 respectively (Nesbit et al., 2013a, Nesbit et al., 2013b). Most patients with FHH have urinary calcium to creatinine clearance ratios < 0.01. Distinguishing FHH from primary hyperparathyroidism can be difficult because there is considerable overlap

373

374

Section 2  

fluid, electrolyte, and renal tubular disorders

in the urinary calcium to creatinine clearance ratios in these two conditions; however a calcium to creatinine clearance ratio > 0.02 makes FHH very unlikely (Fuleihan Gel, 2002). The presence of family history, lifelong mild hypercalcaemia or mutational analysis can provide important clues to differentiate FHH from primary hyperparathyroidism. FHH has a benign natural history and patients and their families should be counseled on the benign nature of this condition and, consequently, the importance of avoiding parathyroid surgery.

action of locally acting osteolytic factors released in conjunction with tumour deposits adjacent to bone. LOH accounts for about 20% of the patients with malignancy associated hypercalcaemia and is frequently encountered in patients with multiple myeloma, lymphoma, leukaemia, and breast cancer (Stewart et  al., 1980; Kinder and Stewart, 2002). Patients with LOH have widespread osteolytic bone metastasis and the tumour cells within the skeletal space secrete several cytokines in a paracrine manner that induce osteoclastogenesis.

Conditions with suppressed PTH levels

Treatment of hypercalcaemia in malignancy In addition to vigorous hydration with isotonic saline solution to increase the urinary excretion of calcium, the primary treatment strategy involves treatment of the underlying tumour. Bisphosphonates have emerged as the cornerstone of the pharmacological treatment of hypercalcaemia of malignancy. Intravenous bisphosphonates, such as pamidronate, ibandronate, and zoledronate have proven to be effective in treating malignancy associated hypercalcaemia through their ability to reduce osteoclastic bone resorption by inhibiting osteoclastic bone resorption. Denosumab is a human immunoglobulin G2 monoclonal antibody against RANKL. This drug mimics the endogenous effects of osteoprotegerin (OPG), a protein produced by osteoblasts, which acts as an alternative receptor for RANKL, thereby modulating the RANK/RANKL-induced osteoclast activity. By inhibiting osteoclastic bone resorption, it is also effective in malignancy-associated hypercalcaemia (Bech and de Boer, 2012; Boikos and Hammers, 2012). Haemodialysis or peritoneal dialysis with a low calcium dialysate provides an effective strategy for patients where other treatment modalities have failed and also for individuals with renal failure or cardiac disease who cannot tolerate large fluid infusions or bisphosphonates (Koo et al., 1996).

Hypercalcaemia of malignancy Up to 10–30% of all patients with advanced cancer can have hypercalcaemia during the course of their disease (Horwitz and Stewart, 2003)  and it carries a poor prognosis. It is seen in both solid tumours and haematologic malignancies. There are several basic mechanisms of cancer-associated hypercalcaemia. The most common cause is tumour cell secretion of parathyroid hormone-related peptide (PTHrP), also known as humoral hypercalcaemia of malignancy (HHM) (Stewart, 2005). Direct osteolytic activity at sites of skeletal metastases is responsible in 20% of cases. Rarely, certain lymphomas and ovarian tumours have been described in association with tumour secretion of 1,25-dihydroxyvitamin D (Seymour et al., 1994; Hibi et al., 2008). Ectopic secretion of PTH by non-parathyroid tumours is exceedingly rare and reported in only a handful of patients (Yoshimoto et al., 1989; Nussbaum et al., 1990; Strewler et al., 1993; Rizzoli et al., 1994; Nielsen et al., 1996; Iguchi et al., 1998). Parathyroid hormone-related peptide PTHrP and PTH have a 70% structural homology for the first 13 amino acids of the amino-terminal portion that accounts for the biological activity of the peptide. PTH and PTHrP bind to a common PTH/PTHrP receptor and share similar biologic activities (Abou-Samra et  al., 1992), such as activating adenylate cyclase in renal and bone systems, increasing renal tubular reabsorption of calcium and osteoclastic bone resorption, and reducing renal phosphate uptake. It has a multifunctional role in cancer, including mediating hypercalcaemia, promoting the development and progression of osteolytic bone metastasis, regulating the growth of cancer cells (Luparello et al., 1993; Luparello et al., 1995; Li et al., 1996), and functioning as a cell survival factor (Chen et al., 2002). HHM occurs most commonly with squamous cell carcinoma, but can occur with other solid tumours (e.g. renal, breast, ovarian, and urothelial carcinomas) and leukaemia. 1,25-dihydroxyvitamin D A major mediator of hypercalcaemia in Hodgkin’s disease, non-Hodgkin’s lymphoma, and other hematological malignancies, is extrarenal secretion of 1,25-dihydroxyvitamin D. Normally, immune cells of the lymphocyte and macrophage lineage produce small amounts of 1,25-dihydroxyvitamin D where it acts as a local cytokine (Edfeldt et  al., 2010; Nelson et  al., 2010). The mechanisms responsible for hypercalcaemia in the setting of elevated 1,25-dihydroxyvitamin D levels are increased intestinal absorption of calcium and osteoclastic bone resorption. Local osteolytic hypercalcaemia Local osteolytic hypercalcaemia (LOH) is by definition, a syndrome of malignancy-associated hypercalcaemia resulting from the direct

Granulomatous disease Hypercalcaemia has been described in association with most granulomatous disorders. Among them, sarcoidosis (Adams et al., 1983; Insogna et al., 1988), tuberculosis (Gkonos et al., 1984; Cadranel et al., 1990), and histoplasmosis (Murray and Heim, 1985), are probably the most common. Hypercalcaemia of granulomatous disease is mediated through extrarenal secretion of 1,25-dihydroxyvitamin D.  Activated mononuclear cells (particularly macrophages) in granulomas are resistant to the normal feedback control of calcitriol production, probably as a result of interferon-gamma (Dusso et al., 1997). In addition to vigorous hydration, specific treatment options include treatment of the underlying granulomatous disorder (such as glucocorticoid therapy in the setting of sarcoidosis) and restriction of dietary calcium and vitamin D intake.

Miscellaneous causes Rarely hypercalcaemia occurs due to increased stimulation of bone resorption (as with thyrotoxicosis (Iqbal et al., 2003), immobilization (Stewart et al., 1982), or vitamin A toxicity (Villablanca et al., 1993)), enhanced active calcium reabsorption in the distal tubule (as with thiazide diuretics), and increased oral calcium intake in the setting of metabolic alkalosis (as with the milk—alkali syndrome (Beall and Scofield, 1995)).

Diagnostic approach Hypercalcaemia should be confirmed by measurement of total serum calcium. If a paraproteinaemia or abnormal albumin levels

chapter 37 

approach to the patient with hypercalcaemia

Elevated serum calcium

Confirm on repeat measurement. Rule out pseudo-hypercalcemia with measurement of ionized calcium level

Measure intact PTH level Suppressed

Elevated/inappropriately normal Measure urinary calcium to creatinine clearance ratio

Consider FHH

Measure 1,25 vitamin D and 25-OH vitamin D Elevated 1,25 Vit D

0.02

Evaluate for lymphoma or granulomatous diseases Check supplements

Suppressed/normal 1,25 and 25-OH Vit D Measure PTHrP levels Measure urinary calcium to creatinine clearance ratio Elevated

Measure SPEP, serum free light chains and bone resorption marker

Elevated Normal

Decreased

Evaluate for humoral hypercalcemia of malignancy

Evaluate for causes of decreased urinary calcium excretion or thiazide use

Increased SPEP/free light chains: Investigate for multiple myeloma Increased bone resorption: Evaluate for thyrotoxicosis, vitamin A intoxication or hypercalcemia of immobilization

Fig. 37.1  Algorithm for dxiagnosis.

are suspected, ionized calcium levels should be obtained to rule out pseudohypercalcaemia. Alternatively, total serum calcium values can be adjusted for hypoalbuminemia using the following formula: corrected calcium (mg/dL) = measured calcium (mg/dL) + (0.8 × [4.0 – albumin (mg/dL]) (Pelosof and Gerber, 2010). This first step in the work-up of hypercalcaemia is to obtain an intact serum PTH level. Conditions associated with increased PTH levels include primary and tertiary hyperparathyroidism, and FHH. In addition to obtaining a family history, urinary calcium to creatinine clearance ratio should be measured to distinguish these. Mutational analysis of the CASR, GNA11, and AP2S1 genes should be considered in patients with urinary calcium to creatinine clearance ratio less than 0.02 to identify FHH and avoid unnecessary parathyroid surgery. Tertiary hyperparathyroidism is usually associated with PTH levels that exceed 800 pg/mL (88 pmol/L). It is important to note that inappropriately normal PTH levels in the setting of hypercalcaemia also indicate primary hyperparathyroidism.

It is not unusual to have coexistent primary hyperparathyroidism with an additional cause of hypercalcaemia. Hence, the finding of elevated or inappropriately normal PTH levels should not preclude additional work-up, if clinical suspicion for other conditions is high. The appropriate parathyroid response to hypercalcaemia is demonstration of suppressed PTH levels. In this setting, PTHrP and 1,25 dihydroxyvitamin D levels should be measured to identify the mediator of hypercalcaemia. Up to 80% of patients with HHM have an elevated intact PTHrP level measured by two-site immunoradiometric assays (IRMAs) (Burtis et al., 1990). In some cancer cells, tumour-specific processing of PTHrP occurs, resulting in the secretion of biologically active amino-terminal fragments of PTHrP. These active amino-terminal fragments may not be detected in the IRMAs that measure intact PTHrP, but could be measured by radioimmunoassays for the amino-terminal end of PTHrP (Rankin et al., 1997). However, it should be noted that

375

376

Section 2  

fluid, electrolyte, and renal tubular disorders

the diagnosis of HHM is often made on clinical grounds and clinical judgment should over-rule any discrepancies in PTHrP assay. If 1,25 dihydroxyvitamin D levels are elevated, the patient should be evaluated for lymphoma, hematological malignancies, or granulomatous disorders. Hypercalcaemia can also occur in patients with markedly elevated levels of 25-OH vitamin D, and this should also be measured. If PTHrP, 25-OH vitamin D, and 1,25 dihydroxyvitamin D levels are normal, a 24-hour urinary calcium to creatinine clearance ratio should be obtained to rule out decreased urinary excretion of calcium. Also, known as the fractional excretion of calcium, this is calculated using the formula: (urinary calcium × serum creatinine)/(urinary creatinine × serum calcium). An additional step would be to rule out conditions causing increased bone turn-over such as multiple myeloma with serum protein electrophoresis and serum free light chain assay, and measurement of bone turnover markers such as urinary excretion of cross-linked N-telopeptides of type I collagen or serum C-terminal collagen crosslink. Additional investigations and focused history should be done to identify the cause of increased bone turnover—such as thyrotoxicosis, vitamin A toxicity, or immobilization. The algorithm in Fig. 37.1 illustrates the diagnostic approach to hypercalcaemia. Though supportive measures such as hydration are the initial step in the treatment of hypercalcaemia, recognizing the pathophysiology responsible for an elevated serum calcium level and treatment of the underlying cause are key to managing hypercalcaemia.

References Abou-Samra, A. B., Juppner, H., Force, T., et al. (1992). Expression cloning of a common receptor for parathyroid hormone and parathyroid hormone-related peptide from rat osteoblast-like cells: a single receptor stimulates intracellular accumulation of both cAMP and inositol trisphosphates and increases intracellular free calcium. Proc Natl Acad Sci U S A, 89, 2732–6. Adams, J. S., Sharma, O. P., Gacad, M. A., et al. (1983). Metabolism of 25-hydroxyvitamin D3 by cultured pulmonary alveolar macrophages in sarcoidosis. J Clin Invest, 72, 1856–60. Barger-Lux, M. J., Heaney, R. P., and Recker, R. R. (1989). Time course of calcium absorption in humans: evidence for a colonic component. Calcif Tissue Int, 44, 308–11. Barrett, K. E., Barman, S. M., Boitano, S., et al. (2011). Hormonal control of calcium and phosphate metabolism and the physiology of bone. In K. E. Barrett, S. M. Barman, S. Boitano, et al. (eds.) Ganong’s Review of Medical Physiology (23rd ed.), pp. 363–75. New York: McGraw-Hill. Beall, D. P. and Scofield, R. H. (1995). Milk-alkali syndrome associated with calcium carbonate consumption. Report of 7 patients with parathyroid hormone levels and an estimate of prevalence among patients hospitalized with hypercalcaemia. Medicine, 74, 89–96. Beard, C. M., Heath, H., 3rd, O’Fallon, W. M., et al. (1989). Therapeutic radiation and hyperparathyroidism. A case-control study in Rochester, Minn. Arch Intern Med, 149, 1887–90. Bech, A. and De Boer, H. (2012). Denosumab for tumour-induced hypercalcaemia complicated by renal failure. Ann Intern Med, 156, 906–7. Bilezikian, J. P., Khan, A. A., and Potts, J. T., Jr. (2009). Guidelines for the management of asymptomatic primary hyperparathyroidism: summary statement from the third international workshop. J Clin Endocrinol Metab, 94, 335–9. Boikos, S. A. and Hammers, H. J. (2012). Denosumab for the treatment of bisphosphonate-refractory hypercalcaemia. J Clin Oncol, 30, e299. Brown, E. M., Gamba, G., Riccardi, D., et al. (1993). Cloning and characterization of an extracellular Ca(2+)-sensing receptor from bovine parathyroid. Nature, 366, 575–80.

Burtis, W. J., Brady, T. G., Orloff, J. J., et al. (1990). Immunochemical characterization of circulating parathyroid hormone-related protein in patients with humoral hypercalcaemia of cancer. N Engl J Med, 322, 1106–12. Cadranel, J., Garabedian, M., Milleron, B., et al. (1990). 1,25(OH)2D2 production by T lymphocytes and alveolar macrophages recovered by lavage from normocalcemic patients with tuberculosis. J Clin Invest, 85, 1588–93. Chen, H. L., Demiralp, B., Schneider, A., et al. (2002). Parathyroid hormone and parathyroid hormone-related protein exert both pro- and anti-apoptotic effects in mesenchymal cells. J Biol Chem, 277, 19374–81. Dusso, A. S., Kamimura, S., Gallieni, M., et al. (1997). gamma-Interferon-induced resistance to 1,25-(OH)2 D3 in human monocytes and macrophages: a mechanism for the hypercalcaemia of various granulomatoses. J Clin Endocrinol Metab, 82, 2222–32. Edfeldt, K., Liu, P. T., Chun, R., et al. (2010). T-cell cytokines differentially control human monocyte antimicrobial responses by regulating vitamin D metabolism. Proc Natl Acad Sci U S A, 107, 22593–8. Fuleihan Gel, H. (2002). Familial benign hypocalciuric hypercalcaemia. J Bone Miner Res, 17 Suppl 2, N51–6. Gkonos, P. J., London, R., and Hendler, E. D. (1984). Hypercalcaemia and elevated 1,25-dihydroxyvitamin D levels in a patient with end-stage renal disease and active tuberculosis. N Engl J Med, 311, 1683–5. Hibi, M., Hara, F., Tomishige, H., et al. (2008). 1,25-dihydroxyvitamin D-mediated hypercalcaemia in ovarian dysgerminoma. Pediatr Hematol Oncol, 25, 73–8. Horwitz, M. J. and Stewart, A. F. (2003). Humoral Hypercalcaemia of Malignancy. Washington, DC: American Society for Bone and Mineral Research. Iguchi, H., Miyagi, C., Tomita, K., et al. (1998). Hypercalcaemia caused by ectopic production of parathyroid hormone in a patient with papillary adenocarcinoma of the thyroid gland. J Clin Endocrinol Metab, 83, 2653–7. Insogna, K. L., Dreyer, B. E., Mitnick, M., et al. (1988). Enhanced production rate of 1,25-dihydroxyvitamin D in sarcoidosis. J Clin Endocrinol Metab, 66, 72–5. Iqbal, A. A., Burgess, E. H., Gallina, D. L., et al. (2003). Hypercalcaemia in hyperthyroidism: patterns of serum calcium, parathyroid hormone, and 1,25-dihydroxyvitamin D3 levels during management of thyrotoxicosis. Endocrine Pract, 9, 517–21. Kinder, B. K. and Stewart, A. F. (2002). Hypercalcaemia. Curr Probl Surg, 39, 349–448. Koo, W. S., Jeon, D. S., Ahn, S. J., et al. (1996). Calcium-free hemodialysis for the management of hypercalcaemia. Nephron, 72, 424–8. Kratz, A., Pesce, M. A., Basner, R. C., et al. (2011). Appendix: laboratory values of clinical importance. In D. L. Longo, A. S. Fauci, D. L. Kasper, et al. (eds.) Harrison’s Principles of Internal Medicine (18th ed.), pp. 3835–410. New York: McGraw-Hill. Li, H., Seitz, P. K., Selvanayagam, P., et al. (1996). Effect of endogenously produced parathyroid hormone-related peptide on growth of a human hepatoma cell line (Hep G2). Endocrinology, 137, 2367–74. Luparello, C., Burtis, W. J., Raue, F., et al. (1995). Parathyroid hormone-related peptide and 8701-BC breast cancer cell growth and invasion in vitro: evidence for growth-inhibiting and invasion-promoting effects. Molecular and cellular endocrinology, 111, 225–32. Luparello, C., Ginty, A. F., Gallagher, J. A., et al. (1993). Transforming growth factor-beta 1, beta 2, and beta 3, urokinase and parathyroid hormone-related peptide expression in 8701-BC breast cancer cells and clones. Differentiation, 55, 73–80. Makitie, O., Kooh, S. W., and Sochett, E. (2003). Prolonged high-dose phosphate treatment: a risk factor for tertiary hyperparathyroidism in X-linked hypophosphatemic rickets. Clin Endocrinol, 58, 163–8. Mallette, L. E., Khouri, K., Zengotita, H., et al. (1989). Lithium treatment increases intact and midregion parathyroid hormone and parathyroid volume. J Clin Endocrinol Metab, 68, 654–60. Marx, S. J., Simonds, W. F., Agarwal, S. K., et al. (2002). Hyperparathyroidism in hereditary syndromes: special expressions and special managements. J Bone Miner Res, 17 Suppl 2, N37–43.

chapter 37 

Murray, J. J. and Heim, C. R. (1985). Hypercalcaemia in disseminated histoplasmosis. Aggravation by vitamin D. Am J Med, 78, 881–4. Nelson, C. D., Reinhardt, T. A., Beitz, D. C., et al. (2010). In vivo activation of the intracrine vitamin D pathway in innate immune cells and mammary tissue during a bacterial infection. PloS One, 5, e15469. Nesbit, M. A., Hannan, F. M., Howles, S. A., et al. (2013a). Mutations affecting G-protein subunit alpha11 in hypercalcaemia and hypocalcemia. N Engl J Med, 368, 2476–86. Nesbit, M. A., Hannan, F. M., Howles, S. A., et al. (2013b). Mutations in AP2S1 cause familial hypocalciuric hypercalcaemia type 3. Nat Genet, 45, 93–7. Nielsen, P. K., Rasmussen, A. K., Feldt-Rasmussen, U., et al. (1996). Ectopic production of intact parathyroid hormone by a squamous cell lung carcinoma in vivo and in vitro. J Clin Endocrinol Metab, 81, 3793–6. Nussbaum, S. R., Gaz, R. D., and Arnold, A. (1990). Hypercalcaemia and ectopic secretion of parathyroid hormone by an ovarian carcinoma with rearrangement of the gene for parathyroid hormone. N Engl J Med, 323, 1324–8. Paik, J. M., Curhan, G. C., and Taylor, E. N. (2012). Calcium intake and risk of primary hyperparathyroidism in women: prospective cohort study. BMJ, 345, e6390. Peacock, M., Bilezikian, J. P., Klassen, P. S., et al. (2005). Cinacalcet hydrochloride maintains long-term normocalcemia in patients with primary hyperparathyroidism. J Clin Endocrinol Metab, 90, 135–41. Pelosof, L. C. and Gerber, D. E. (2010). Paraneoplastic syndromes: an approach to diagnosis and treatment. Mayo Clin Proc, 85, 838–54. Pollak, M. R., Brown, E. M., Chou, Y. H., et al. (1993). Mutations in the human Ca(2+)-sensing receptor gene cause familial hypocalciuric hypercalcaemia and neonatal severe hyperparathyroidism. Cell, 75, 1297–303.

approach to the patient with hypercalcaemia

Rankin, W., Grill, V., and Martin, T. J. 1997. Parathyroid hormone-related protein and hypercalcaemia. Cancer, 80, 1564–71. Rizzoli, R., Pache, J. C., Didierjean, L., et al. (1994). A thymoma as a cause of true ectopic hyperparathyroidism. J Clin Endocrinol Metab, 79, 912–5. Seymour, J. F., Gagel, R. F., Hagemeister, F. B., et al. (1994). Calcitriol production in hypercalcemic and normocalcemic patients with non-Hodgkin lymphoma. Ann Intern Med, 121, 633–40. Stewart, A. F. (2005). Clinical practice. Hypercalcaemia associated with cancer. N Engl J Med, 352, 373–9. Stewart, A. F., Adler, M., Byers, C. M., et al. (1982). Calcium homeostasis in immobilization: an example of resorptive hypercalciuria. N Engl J Med, 306, 1136–40. Stewart, A. F., Horst, R., Deftos, L. J., et al. (1980). Biochemical evaluation of patients with cancer-associated hypercalcaemia: evidence for humoral and nonhumoral groups. N Engl J Med, 303, 1377–83. Strewler, G. J., Budayr, A. A., Clark, O. H., et al. (1993). Production of parathyroid hormone by a malignant nonparathyroid tumour in a hypercalcemic patient. J Clin Endocrinol Metab, 76, 1373–5. Villablanca, J. G., Khan, A. A., Avramis, V. I., et al. 1993. Hypercalcaemia: a dose-limiting toxicity associated with 13-cis-retinoic acid. J Pediatr Hematol Oncol, 15, 410–15. Wermers, R. A., Khosla, S., Atkinson, E. J., et al. (2006). Incidence of primary hyperparathyroidism in Rochester, Minnesota, 1993–2001: an update on the changing epidemiology of the disease. J Bone Miner Res, 21, 171–7. Wynne, A. G., Van Heerden, J., Carney, J. A., et al. (1992). Parathyroid carcinoma: clinical and pathologic features in 43 patients. Medicine, 71, 197–205. Yoshimoto, K., Yamasaki, R., Sakai, H., et al. (1989). Ectopic production of parathyroid hormone by small cell lung cancer in a patient with hypercalcaemia. J Clin Endocrinol Metab, 68, 976–81.

377

CHAPTER 38

Approach to the patient with hypocalcaemia Agnès Linglart and Anne-Sophie Lambert Introduction Calcium is so essential for normal cellular functions—nerve impulse conduction, muscle fibre contraction, blood coagulation, cell differentiation, control of secretory mechanisms, and post-receptor second messenger transduction—that evolution has narrowed the tolerated range for extracellular ionized calcium to 0.5  mmol/L (2.25–2.75 mmol/L), and developed a complex system to regulate and maintain calcium levels, as well as bone mineralization, with few redundant mechanisms. Discovery of hypo- or hypercalcaemia always reflects a pathological process challenging the fine balance of calcium absorption, parathyroid hormone (PTH) secretion and action, vitamin D production and action, cellular compartmentalization of calcium ions, and renal function. Therefore, when faced with a patient with hypo/hypercalcaemia, newborn or elderly, we must consider two things: (1) therapy to restore the calcium level to normal and (2) investigations to determine the cause of hypo/ hypercalcaemia. Dietary calcium is mainly absorbed in the duodenum and proximal jejunum through a passive paracellular mechanism and an active transcellular process involving the calcium channel TRPV6, calbindins, and membrane exchangers such as Na+/ Ca2+ (NCX1) and PMCA1b. The dihydroxy form of vitamin D, 1,25-dihydroxyvitamin D (1,25-(OH)2D), is the most powerful stimulus to calcium absorption, enhancing expression of channels and transporters (Hoenderop et al., 2005). In the kidney, calcium reabsorption depends on the additional action of the calcium sensing receptor (CaSR), as well as PTH. Vitamin D (25-(OH)D) is generated through a series of enzymatic steps from cholesterol, or it is absorbed from the diet, and then further hydroxylated to 1,25-(OH)2D, which in turn is able to bind the nuclear receptor VDR (vitamin D receptor) to activate various target genes, including PTH, FGF23, a phosphaturic factor, and CYP24, encoding the enzyme 24-hydroxylase, which inactivates 1,25-(OH)2D. Activation of PTH1R, the PTH/parathyroid hormone-related protein receptor, increases serum calcium levels through the stimulation of 1,25-(OH)2D synthesis and tubular reabsorption of calcium by the kidney (see Chapter 26). Ionized calcium is the main ligand for the CaSR: on binding calcium, CaSR inhibits PTH secretion by parathyroid cells and inhibits calcium reabsorption in the loop of Henle (thick ascending limb) (Bergwitz and Juppner, 2010). In children, normal growth and bone mass accrual require a large amount of mineral, so determinants of blood calcium level are tightly controlled to ensure normal skeletal growth. Indeed,

periods of rapid growth can reveal hypocalcaemia, if normal calcium handling is disturbed (impaired digestive absorption, bone resorption, or renal reabsorption) (Gilsanz and Nelson, 2003). In summary, molecular defects, severe and prolonged environmental injuries, or dysregulation of organs and tissues involved in calcium balance, can all result in hypocalcaemia.

Is it really hypocalcaemia? Measurement of calcium Only 1% of the total body calcium circulates between fluids and tissues, 99% being trapped in hydroxyapatite. In physiological conditions, the extracellular calcium (2.2–2.60  mmol/L in serum) is 45–50% ionized calcium, 40% protein-bound calcium (mainly albumin), and 10% in diffusible complexes. Thus, the total calcium value can be influenced by metabolic disorders or dysproteinaemias. In theory, the measurement of ionized calcium is the most accurate way of evaluating serum calcium. However, since it requires collection and handling under anaerobic conditions with immediate measurement, total calcium is often used as a surrogate measure. Examples of patients requiring ionized calcium to be measured are the critically ill; especially patients receiving citrated blood, those with advanced renal failure, and neonates. In disorders affecting the albumin concentration, formulas can be devised to correct the measured total calcium (deduct 0.25 mmol/L from total calcium for each 10 g/L decrease in albumin). While diseases associated with acidosis or alkalosis can affect free calcium levels (acidosis decreases protein-bound calcium and increases free calcium), ionized calcium should not be corrected for pH using formulae. The formulae devised to estimate ionized calcium or to correct total calcium have no real advantage over uncorrected calcium; if needed, ionized calcium should be measured stringently, as above, and at the patient’s pH.

Definition of hypocalcaemia Hypocalcaemia is defined as a serum calcium below the normal range, that is, < 2.2 mmol/L or < 1.15 mmol/L of total calcium or ionized calcium, respectively. Calcium values do not vary with age and this definition applies to children and adults. Hypomagnesaemia should be excluded in patients with hypocalcaemia, since its clinical features can be similar: it may lead to hypocalcaemia, but its causes and treatment are different.

chapter 38 

Clinical evaluation Diagnosis of hypocalcaemia Symptoms correlate with the severity of hypocalcaemia, ranging from asymptomatic to acute and life threatening. The evolution of hypocalcaemia (acute versus chronic) affects its clinical presentation, with better tolerance of chronic and slowly evolving hypocalcaemia. Hypocalcaemia is more symptomatic in children, especially during early life and adolescence, because of the increased need for calcium for growth.

Symptoms of hypocalcaemia ◆ Mainly

neuromuscular irritability manifesting as paraesthesiae, cramps, tetany, Chvostek’s sign (spasm of the circumoral muscles in response to a gentle tap of the facial nerve) and Trousseau’s sign (carpal spasm in response to the inflation of a blood pressure cuff to 20 mmHg above the patient’s systolic pressure), seizures of various types, laryngospasm, prolonged QT interval on electrocardiogram (ECG), and dysrhythmias.

◆

Memory loss, difficulty thinking, problems at school, poor physical and mental performance; cerebral calcification of the basal ganglia (Fig. 38.1) may be present in patients with long-standing hypocalcaemia.

◆ In

neonates, the presence of hypocalcaemia is often revealed by tremor or seizures. Weight gain is not usually affected, but development of cognitive function is often delayed.

approach to the patient with hypocalcaemia

◆ Except

for ectopic calcification, all signs and symptoms resolve with the restoration of a normal, or almost normal, calcium level.

In critically ill patients, attention should be focused on acute causes, such as pancreatitis, rhabdomyolysis, tumour lysis syndrome, or therapies with calcium chelators or resorption inhibitors. In non-critically ill patients, consider previous neck surgery or irradiation, autoimmune disease, digestive malabsorption, alcoholism, renal or liver disease, or an iatrogenic cause. In young adults and children, rickets suggests severe vitamin D deficiency or resistance; other symptoms such as candidiasis (Fig. 38.1), dysmorphic features, or chondrodysplasia (Fig. 38.1) will suggest a form of primary hypoparathyroidism due to a defect in parathyroid gland development or function, or PTH resistance. In neonates, maternal hypercalcaemia (due to primary hyperparathyroidism) or maternal vitamin D deficiency can explain most hypocalcaemic episodes (Thakker, 2003; Mallet et al., 2010; Holt, 2012).

Laboratory investigation Biochemistry The careful investigation of hypocalcaemia aims to evaluate: ◆ severity ◆ total ◆ if

and consequences of hypocalcaemia

serum calcium

possible, ionized calcium with albumin

◆ serum

phosphate

(A)

(C)

(B)

Fig. 38.1  (A) Cerebral calcification of lenticular and caudate nuclei in a 13-year-old girl with recently diagnosed hypoparathyroidism (unknown cause, no 22 q1.1 deletion, no GCMB, PTH or CaSR mutation). (B) Candidiasis affecting one nail in an 8-year-old boy with hypoparathyroidism and autoimmune polyglandular syndrome type 1 (APECED). (C) Brachymetacarpy (3rd, 4th, and 5th digit) and brachymetatarsy (3rd and 4th digits) in a patient with pseudohypoparathyroidism type IA and a mutation in the GNAS gene.

379

Table 38.1  Main causes of hypocalcaemia and their typical biochemical pattern; genes or genomic regions involved in hypocalcaemia are in parentheses Cause

Calcium and phosphate

Urinary calcium excretion

PTH

25-(OH)D and 1,25-(OH)2D

Comments

Hypoparathyroidism Defect in parathyroid gland embryogenesis DiGeorge syndrome (22q11 deletion or TBX1) Sanjad–Sakati syndrome (TBCE) Kenny–Caffey (TBCE, FAM111A) Hypoparathyroidism, deafness and renal dysplasia, HDR (GATA3) Mitochondrial diseases like Kearns– Sayre syndrome Isolated hypoparathyroidism (GCMB, X-linked, autosomal recessive) Defect in PTH production or secretion Isolated hypoparathyroidism (PTH) Autosomal dominant hypoparathyroidism (activating mutation of the calcium sensing receptor CaSR, G11) Destruction of the parathyroid glands Auto-immune polyendocrinopathy type 1 or APECED (AIRE) Auto-immune hypoparathyroidism (anti-CaSR antibodies) Surgery Infiltration (neoplasia, granulomas) or irradiation

Low calcium High phosphate Low calcium High phosphate Low calcium High phosphate Low calcium High phosphate

Low at diagnosis Low at diagnosis Elevated Low at diagnosis

Low Low Low Low

Normal or low 25-(OH)D Low-normal or low 1,25-(OH)2D Normal or low 25-(OH)D; lownormal or low 1,25-(OH)2D Normal or low 25-(OH)D; lownormal or low 1,25-(OH)2D Normal or low 25-(OH)D Low-normal or low 1,25-(OH)2D

Low alkaline phosphatases Hungry bone syndrome may occur

Pseudohypoparathyroidism Type 1A, 1B or 1C (GNAS) Acrodysostosis (PRKAR1A)

Low calcium High phosphate Low or low-normal calcium High phosphate

Low Low

Elevated Elevated

Normal or low 25-(OH)D; lownormal or low 1,25 (OH)2D Normal or low 25-(OH)D; lownormal or low 1,25-(OH)2D

Low alkaline phosphatase

Vitamin D deficiency or resistance Insufficient vitamin D intake or sunlight exposure, digestive malabsorption, deficient steroid metabolism (antiepileptics, liver diseases) Vitamin D-resistant rickets type I (CYP24A1 mutations) Vitamin D-resistant rickets type II (vitamin D receptor mutations) Oncogenic osteomalacia induced by FGF23 secreting tumours

Low calcium Low phosphate Low calcium Low phosphate Low calcium Low phosphate Low calcium Low phosphate

Low Low Low Low

Elevated Elevated Elevated Elevated

Low 25-(OH) D; low-normal or 1,25-(OH)2D Normal or high 25-(OH)D; low 1,25-(OH)2D Normal or high 25-(OH)D; high 1,25-(OH)2D Normal 25-(OH)D; 1,25-(OH)2D

High alkaline phosphatase High alkaline phosphatase High alkaline phosphatase Alopecia in 50% of the cases High alkaline phosphatase, osteomalacia

Renal failure

Low calcium High phosphate

Low

Elevated

Normal 25-(OH)D; 1,25-(OH)2D

Elevated creatinine High alkaline phosphatase

(Continued)

chapter 38 

approach to the patient with hypocalcaemia

Table 38.1 Continued Cause

Calcium and phosphate

Urinary calcium excretion

PTH

25-(OH)D and 1,25-(OH)2D

Comments

Iatrogenic Phosphate supplements Bisphosphonates and inhibitors of bone resorption

Low calcium High phosphate Low calcium Low phosphate

Low Low

Elevated Elevated

Normal 25-(OH) D; normal or high 1,25-(OH)2D Normal 25-(OH) D; normal or high 1,25-(OH)2D

Low bone resorption markers

Miscellaneous Acute pancreatitis Cell lysis (rhabdomyolysis, tumour) Toxic shock syndrome

◆ QT

interval on ECG

◆ differential ◆ serum

diagnosis

magnesium

◆ aetiology ◆ serum

creatinine

◆ serum

PTH—even a few hours after the initial treatment of acute hypocalcaemia, PTH levels are still reliable in determining its cause

◆ serum

(bone) alkaline phosphatase

◆ serum

25-(OH)D should be measured before vitamin D administration—in urgent cases, a blood sample should be drawn and serum frozen, or kept at 4°C and protected from light until measured

◆ in

rare situations, a 1,25-(OH)2D level is required as part of the work-up of hypocalcaemia (to distinguish vitamin D-resistant rickets (VDRR), VDRR type I with low levels of 1,25-(OH)2D from VDRR type II with high levels of 1,25-(OH)2D)—sampling conditions are similar to those for 25-(OH)D above

◆ other

measurements depend on the clinical context, for example, pancreatic enzymes or liver function tests.

Typical biochemical features of the most common causes of hypocalcaemia are set out in Table 38.1.

Interpretation of the laboratory tests First, hypomagnesaemia must be excluded. PTH is the most important determinant of the serum calcium level. Hypocalcaemia may be divided into hypocalcaemia caused by PTH insufficiency or hypoparathyroidism (low serum calcium and PTH levels) or hypocalcaemia from other causes with a normal parathyroid gland response, usually described as secondary hyperparathyroidism (low serum calcium and high PTH levels). The latter is typically due to renal insufficiency, vitamin D deficiency (or resistance), or PTH resistance (Table 38.1).

2011). Hypocalcaemia may be associated with chronic renal failure, acute diseases (e.g. pancreatitis, rhabdomyolysis, or toxic shock syndrome) or use of therapies such as bisphosphonates, plicamycin (mithramycin), or calcitonin. Conditions leading to severe malabsorption of both calcium and vitamin D may also cause hypocalcaemia. Isolated hypoparathyroidism revealed by hypocalcaemia can occur in adults, when an autoimmune disease affecting the parathyroids should be suspected; if not due to this, genetic investigation should be carried out to test for DiGeorge syndrome or activating mutations of the CaSR (autosomal dominant hypoparathyroidism or ADH), and extended to family members (Pollak et al., 1994). In children, vitamin D deficiency, with or without rickets, is the most likely cause of hypocalcaemia. Apart from this, hypocalcaemia is often due to a genetic defect in the secretion or action of PTH. Hypoparathyroidism in childhood occurs in DiGeorge syndrome, autoimmune polyendocrinopathy-candiasis-ectodermal dystrophy (APECED) syndrome, Kearns–Sayre syndrome, and autosomal dominant hypocalcaemia (ADH); pseudohypoparathyroidism is due to genetic or epigenetic defects at the imprinted GNAS locus (Linglart et  al., 2013). In rare cases, hypocalcaemia and rickets diagnosed between 1 and 3 years of age may be due to vitamin D resistance from loss of function mutations of the renal 1α-hydroxylase (CY27B1) (Glorieux and St-Arnaud, 1998) or of the vitamin D3 receptor gene (VDR) (Hochberg et al., 1992). In neonates, hypocalcaemia is often related to intrauterine growth retardation or to maternal conditions such as maternal vitamin D deficiency or maternal primary hyperparathyroidism and hypercalcaemia. Transient hypoparathyroidism, (hypocalcaemia and a low level of PTH) may occur because of immaturity of the parathyroid glands. In the absence of maternal disease, hypoparathyroidism due to lack of parathyroid gland development should be investigated, for example, DiGeorge syndrome, GCMB (autosomal dominant or recessive), GATA3 or TBCE mutations, or activating mutations of the CaSR (ADH) (Carpenter, 2003).

Causes of hypocalcaemia

Management of hypocalcaemia

In adults, hypocalcaemia is often the consequence of an acquired disease. Although usually transient in most patients, hypocalcaemia following thyroid surgery may persist in some patients following complete removal of the parathyroid glands (Bilezikian et al.,

In emergency situations such as neuromuscular spasms, seizures, or cardiac dysrhythmias, calcium gluconate 10% should be given intravenously (Table 38.2). If there is no intravenous access, rectal

Acute or severe hypocalcaemia

381

382

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 38.2  Treatment of hypocalcaemia Therapy

Molecule

Situations

Protocol of administration

Comments

Parenteral infusion of calcium

Calcium gluconate 10%

Life-threatening emergency

Intravenous 0.5 mL/kg up to 10 mL, slowly over 10–15 minutes to avoid bradycardia Intrarectal

Repeatable once, under cardiac monitoring Repeatable once

Oral calcium supplements

Calcium carbonate or calcium citrate

Symptomatic hypocalcaemia or total calcium < 1.75 mmol/L Intravenous diluted in 5% dextrose Neonates or 0.9% sodium chloride 80 mg/kg/day Children 50 mg/kg/day or 1000 mg/m2/day Adults 50 mL/h of the solution: 100 mL of calcium gluconate 10% (10 ampules) in 1 L of 5% dextrose or 0.9% sodium chloride

Verify serum calcium every 12 hours Stop the intravenous when calcium ≥ 2.2 mmol/L

Relay of intravenous infusion of calcium gluconate Asymptomatic hypocalcaemia or total calcium > 1.75 mmol/L Neonates

Maximum 2000 mg/day in adults Divided in 2 or 3 doses daily May be removed in children after a few months if calcium requirements are achieved through diet

Children Adults 25-(OH) vitamin D

Cholecalciferol Vitamin D deficiency (vitamin D3) Ergocalciferol (vitamin D2)

Active vitamin D analogues

Alfacalcidol Calcitriol

Hypoparathyroidism Acute hypocalcaemia Neonates Children Adolescence Adults

administration is possible. Whatever the age, symptomatic (e.g. neurological symptoms, cardiac dysrhythmias) and/or profound (e.g. total calcium < 1.75  mmol/L without hypoalbuminaemia) hypocalcaemia require intravenous calcium infusion. Various protocols and doses are available (Table 38.2). Infusion should start with 1000 mg/m2 body surface area per 24 hours. The duration of infusion depends on the severity and duration of hypocalcaemia: chronic and severe hypocalcaemia requiring 3–6 days; hungry bone syndrome requiring up to several weeks; acute hypocalcaemia requiring only a few hours. Intravenous infusion should be stopped when total calcium reaches 2.2 mmol/L, since it will usually drop by 0.2  mmol/L in the following hours. In most cases, intravenous infusion is followed by oral calcium supplements.

Promote breastfeeding; 200 mg/day of supplements are possible 5-years-old: 500 mg/day; 10-years-old: 1000 mg/day 1000–1500 mg/day; 2000 mg if proton-pump inhibitors associated

Many protocols exist and cannot be For store replenishment all described here. Half-life 20–45 days 100,000 IU orally repeated after 1 week

Alfacalcidol: 2 to 6 mcg/day; calcitriol: 1 to 3 mcg/day Alfacalcidol: 1 to 1.5 mcg/day; calcitriol: 0.5 to 0.75 mcg/day Alfacalcidol: ~ 1 mcg/day; calcitriol: ~ 0.5 mcg/day Alfacalcidol: 1 to 1.5 mcg/day; calcitriol: 0.5 to 0.75 mcg/day

Hypomagnesaemia should be corrected or hypocalcaemia will not resolve. Vitamin D therapy is almost always given with intravenous calcium, except for hypocalcaemia in severe illnesses such as pancreatitis or rhabdomyolysis. The type of vitamin D given (e.g. D2, D3, 25-(OH)D3, 1,25(OH)2D3) depends on the origin of the hypocalcaemia and dosage depends on the severity and duration of hypocalcaemia (Table 38.2) (Mallet et al., 2010; Holt, 2012; Kelly and Levine, 2013).

Therapy for chronic hypocalcaemia In most situations, chronic hypocalcaemia is treated with calcium supplements and vitamin D analogues (see Table 38.2 for doses and protocols) (Thakker, 2003). In children, calcium supplements are

chapter 38 

often stopped if calcium requirements can be achieved through a normal diet. Patients with vitamin D deficiency should receive an acute oral load of cholecalciferol (vitamin D3) or, if not available, ergocalciferol (vitamin D2) over a few weeks as a daily dose of vitamin D. As the hydroxylation of calciferol may take 2–3 days, a short course of calcitriol or alfacalcidol may also be given. Patients with digestive malabsorption, impaired liver function, or alcoholic hepatitis should receive parenteral vitamin D (fat-soluble preparations). Most cases of hypocalcaemia are associated with defective 1,25-(OH)2D synthesis and will resolve with the administration of 1α-hydroxylated vitamin D and oral calcium supplementation (Thakker et al., 1998). Alfacalcidol (1α-hydroxycholecalciferol) and calcitriol (1,25 dihydroxycholecalciferol) are commonly used (Table 38.2). Due to their different half-lives, alfacalcidol is given once a day and calcitriol twice a day. In patients with poor liver function calcitriol is preferred, because it does not require 25-hydroxylation (enzymatic conversion in the liver). The treatment goals depend on the underlying disease. In children and adults with hypoparathyroidism therapy should restore low-normal calcium levels (2–2.2  mmol/L total calcium). Normalization of calcium is to be avoided, because it can cause hypercalciuria in the absence of PTH-stimulated calcium reabsorption in the distal renal tubule. In ADH, it may be particularly difficult to increase serum calcium without causing hypercalciuria (Lienhardt et al., 2001). In children, 1α-hydroxylated vitamin D analogues are mandatory because of the high calcium requirement for skeletal growth. The highest doses are used in infancy, early childhood, and during adolescence. After puberty, calcium requirements fall and patients with hypoparathyroidism can maintain adequate, though subnormal, calcium levels with vitamin D and calcium supplements. It is noteworthy that patients with pseudohypoparathyroidism usually do not present with hypercalciuria following vitamin D analogue therapy; treatment should aim to decrease their PTH level and maintain normal serum calcium levels (Linglart et al., 2013). Patients affected with VDRR type I are easily controlled with low doses of alfacalcidol or calcitriol with calcium supplements. Treatment should restore normal calcium and phosphate levels within days, PTH levels within weeks, alkaline phosphatase levels within months, and rickets, leg bowing, and growth over years. In both children and adults, administration of alfacalcidol or calcitriol is closely monitored by serum and urine calcium measurements to avoid hypercalciuria, nephrocalcinosis, and nephrolithiasis (Bilezikian et al., 2011; Lienhardt and Linglart, 2012; Mitchell et al., 2012). In contrast, VDRR type II is the only cause of hypocalcaemia that cannot be corrected by giving 1α-hydroxylated vitamin D. In these extremely rare patients, calcium is given intravenously for months or years, together with high doses of oral calcium (Hochberg et al., 1992; Tiosano and Gepstein, 2012). Recombinant PTH (recPTH 1-34 or recPTH 1-84) has been successfully given to children and adults with hypoparathyroidism (Linglart et al., 2011; Cusano et al., 2013). Although it is not approved by the regulatory authorities (and so is off licence) recPTH may be an alternative to current therapy in patients with refractory hypoparathyroidism who experience life-threatening complications of their disease (Linglart et al., 2011; Winer et al., 2012). In our experience, doses of recPTH1-34 vary between 0.1 and 0.6 micrograms/kg/day to ensure a normal calcium level.

approach to the patient with hypocalcaemia

References Bergwitz, C. and Juppner, H. (2010). Regulation of phosphate homeostasis by PTH, vitamin D, and FGF23. Annu Rev Med, 61, 91–104. Bilezikian, J. P., Khan, A., Potts, J. T., Jr., et al. (2011). Hypoparathyroidism in the adult: epidemiology, diagnosis, pathophysiology, target-organ involvement, treatment, and challenges for future research. J Bone Miner Res, 26, 2317–37. Carpenter, T. (2003). Neonatal hypocalcaemia. In M. J. Favus (ed.) Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism, pp. 286–7. Washington, DC: American Society of Bone and Mineral Research. Cusano, N. E., Rubin, M. R., McMahon, D. J., et al. (2013). Therapy of hypoparathyroidism with PTH(1-84): a prospective four-year investigation of efficacy and safety. J Clin Endocrinol Metab, 98, 137–44. Gilsanz, V. and Nelson, D. (2003). Childhood and adolescence. In M. J. Favus (ed.) Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism, pp. 55–63. Washington, DC: American Society of Bone and Mineral Research. Glorieux, F. and St-Arnaud, R. (1998). Molecular cloning of (25-OH D)-1 alpha-hydroxylase: an approach to the understanding of vitamin D pseudo-deficiency. Recent Prog Horm Res, 53, 341–90. Hochberg, Z., Tiosano, D., and Even, L. (1992). Calcium therapy for calcitriol-resistant rickets. J Pediatr, 121, 803–8. Hoenderop, J. G., Nilius, B., and Bindels, R. J. (2005). Calcium absorption across epithelia. Physiol Rev, 85, 373–422. Holt, E. (2012). Diagnosis and treatment of the patient with abnormal calcium. In A. Licata and V. Lerma (eds.) Diseases of the Parathyroid Glands, pp. 53–68. New York: Springer. Kelly, A. and Levine, M. A. (2013). Hypocalcaemia in the critically ill patient. J Intensive Care Med, 28,166–77. Lienhardt, A., Bai, M., Lagarde, J. P., et al. (2001). Activating mutations of the calcium-sensing receptor: management of hypocalcaemia. J Clin Endocrinol Metab, 86, 5313–23. Lienhardt, A. and Linglart, A. (2012). Hypoparathyroidism in children. In A. Licata and A. Lerma (eds.) Diseases of the Parathyroids Glands, pp. 299–310. New York: Springer. Linglart, A., Maupetit-Mehouas, S., and Silve, C. (2013). GNAS-related loss-of-function disorders and the role of imprinting. Horm Res Paediatr, 79, 119–29. Linglart, A., Rothenbuhler, A., Gueorgieva, I., et al. (2011). Long-term results of continuous subcutaneous recombinant PTH (1-34) infusion in children with refractory hypoparathyroidism. J Clin Endocrinol Metab, 96, 3308–12. Mallet, E., Garabedian, M., Lienhardt, A., et al. (2010). Fiches pratiques: diagnostic et premiers traitementts des troubles du métabolisme phosphocalcique chez l’enfant. In E. Mallet, M. Garabedian, A. Lienhardt, et al. (eds.) Métabolisme phosphocalcique et osseux de l’enfant, pp. 24–34. France: Lavoisier. Mitchell, D. M., Regan, S., Cooley, M. R., et al. (2012). Long-term follow-up of patients with hypoparathyroidism. J Clin Endocrinol Metab, 97, 4507–14. Pollak, M. R., Brown, E. M., Estep, H. L., et al. (1994). Autosomal dominant hypocalcaemia caused by a Ca2+-sensing receptor gene mutation. Nat Genet, 8, 303–7. Thakker, R. (2003). Hypocalcaemia, pathogenesis, diffrential diagnosis, and managment. In M. J. Favus (ed.) Primer on the Metabolic Bone Diseases and Disorders of Mineral Metabolism, pp. 271–4. Washington, DC: American Society of Bone and Mineral Research. Thomas, M. K., Lloyd-Jones, D. M., Thadhani, R. I., et al. (1998). Hypovitaminosis D in medical inpatients. N Engl J Med, 338, 777–83. Tiosano, D. and Gepstein, V. (2012). Vitamin D action: lessons learned from hereditary 1,25-dihydroxyvitamin-D-resistant rickets patients. Curr Opin Endocrinol Diabetes Obes, 19, 452–9. Winer, K. K., Zhang, B., Shrader, J. A., et al. (2012). Synthetic human parathyroid hormone 1-34 replacement therapy: a randomized crossover trial comparing pump versus injections in the treatment of chronic hypoparathyroidism. J Clin Endocrinol Metab, 97, 391–9.

383

CHAPTER 39

Approach to the patient with hypo-/hyperphosphataemia Judith Blaine, Hector Giral, Sabina Jelen, and Moshe Levi Introduction Phosphate is the most abundant anion in the human body and has an indispensable role in numerous biological functions, including energy metabolism, bone formation, signal transduction, and as a constituent of phospholipids and nucleic acids. Total body phosphorus content in an average adult is 600–700 g (8.5–10g/kg body weight), distributed in the skeleton (85%) mostly in the form of hydroxyapatite (Ca10(PO4)6(OH)2), in soft tissues (14%) and only 1% in extracellular fluid, including serum levels. Despite constituting only a small part of the total store, serum phosphate (Pi) levels are subject to a dynamic fine tuning involving several hormones to maintain a normal range from 0.81 to 1.45 mmol/L (2.5–4.5 mg/ dL) in adulthood and higher levels during infancy and childhood (Table 39.1) (Levine and Kleeman, 1994). Phosphate homeostasis is regulated by a complex interplay of different tissues modulating the renal tubular reabsorption, the intestinal absorption, and the bone resorption.

Hypophosphataemia Hypophosphatemia refers to serum Pi concentrations of < 0.81  mmol/L (2.5 mg/dL). Hypophosphatemia usually results from one or a combination of the following factors (Fig. 39.1): (1) increased excretion of Pi in the urine, (2) decreased dietary Pi intake, (3) decreased GI absorption of Pi, or (4) translocation of Pi from the extracellular to the intracellular space. The major causes of hypophosphatemia are shown in Table 39.2.

Increased excretion of phosphorus in the urine Several pathophysiologic conditions increase excretion of Pi in the urine. Some of these are characterized by elevated levels of circulating parathyroid hormone (PTH) or fibroblast growth factor 23 (FGF23). Because PTH and FGF23 decrease Pi reabsorption by the kidney (see Chapter 25), elevations of the hormones increase urinary excretion (Table 39.2). Decreased tubular reabsorption of Pi may also occur without increased levels of PTH or FGF23 and may be due to changes in the reabsorption of salt and water or to renal tubular defects specific for the reabsorption of certain solutes or Pi. Hypophosphataemia may also occur in the diuretic phase of acute tubular necrosis or in post-obstructive diuresis, presumably due to a combination of high levels of PTH and decreased tubular reabsorption of salt and water.

Inherited disorders—familial hypophosphatemia In spite of the low incidence of inherited hypophosphataemic disorders (a few familial cases in most of the syndromes), the identification of the genes inducing these syndromes has dramatically increased our understanding of Pi homeostasis. The emerging importance of SLC34A3 (NaPi-IIc) (Bergwitz et al., 2006; Ichikawa et  al., 2006; Lorenz-Depiereux et  al., 2006)  and FGF23 (Jonsson et  al., 2003; Shimada et  al., 2004)  as mediators of Pi regulation, and the identification of the kidney–intestine–bone hormonal axis managing the complex signalling interplay of Pi homeostasis are some of the breakthroughs. Rickets is the common denominator in most inherited hypophosphataemic syndromes during infancy and childhood. Rickets is a disease of the growth plate, and hence only children are affected, while osteomalacia is present in both children and adult hypophosphataemic patients (Greenbaum, 2011) In an attempt to describe with clarity the variety of disorders involving hypophosphataemia we will classify them based on the pathophysiological mechanisms leading to the clinical features. In this chapter, we will emphasize the disorders associated with FGF23, NaPi-IIa, and NaPi-IIc in regulation of general Pi homeostasis and refer to other sources to discuss the syndromes related to PTH and vitamin D malfunction, which are complex disorders that alter the homeostasis of other ions, including calcium.

Hypophosphatemia caused by defective renal Na/Pi transporters activity The most obvious target causing hereditary hypophosphataemic disorders are genetic mutations inducing loss of function of the renal sodium-dependent Pi cotransporters (Na/Pi), with subsequent renal Pi wasting as the cause of the hypophosphataemia.

Hereditary hypophosphataemic rickets with hypercalciuria Hereditary hypophosphataemic rickets with hypercalciuria (HHRH) (OMIM 241530) is an autosomal recessive inherited disorder characterized by early childhood onset presenting rickets, short stature, bone pain, and muscle weakness. First identified in a consanguineous Bedouin kindred in 1985 (Tieder et al., 1985), it was not until 2006 that genetic mutations were identified in the SLC34A3 gene (NaPi-IIc) (Bergwitz et al., 2006; Ichikawa et al., 2006; Lorenz-Depiereux et al., 2006), one of the Na/Pi transporters mediating reabsorption in the renal proximal tubule. Homozygous individuals display the full

chapter 39 

consists of Pi salts supplementation since the administration of vitamin D analogues could cause worsening hypercalciuria and associated nephrolithiasis (Tieder et al., 1992).

Table 39.1  Serum phosphate levels range reference in normophosphataemia, hypophosphataemia, and hyperphosphataemia Age

Other syndromes

Serum phosphate range mmol/L (mg/dL)

Normal range Infancy

0–5

1.45–2.67 (4.5–8.3)

Childhood

6–17

1.19–1.81 (3.7–5.6)

Adulthood

> 18

0.81–1.45 (2.5–4.5)

Hypophosphataemia in adults Adult

Mild

0.65–0.81 (2–2.5)

Moderate

0.32–0.65 (1–2)

Severe

< 0.32 (< 1)

Hyperphosphataemia in adults Adult

Moderate Severe

1.48–1.93 (4.6–6) > 1.93 (>6)

From Marwaha et al. (2010).

clinical spectrum of symptoms and show frequently compound heterozygous SLC34A3 mutations, meaning that they carry different mutations in the maternal and paternal alleles. Heterozygous carriers often show some of the symptoms, mostly hypercalciuria with milder hypophosphataemia and elevation of 1,25-dihydroxyvitamin D (1,25(OH)2D) (Jaureguiberry et al., 2008). Laboratory findings consist of hypophosphataemia due increased renal Pi clearance, leading to rickets or osteomalacia. Compensatory upregulation of circulating 1,25(OH)2D levels enhances intestinal absorption of calcium, which then causes hypercalciuria despite normal serum calcium levels (Tieder et al., 1985). PTH and FGF23 levels are generally normal or low (Ichikawa et al., 2006). Nephrolithiasis seems to be more common in HHRH than originally described, presumably induced by the increased urinary calcium and Pi excretion (Alizadeh Naderi and Reilly, 2010). Treatment of different hereditary hypophosphataemic conditions depends on the underlying genetic defect. In the case of HHRH, where the patients show elevated levels of 1,25(OH)2D, treatment

Serum Pi

While NaPi-IIa has a dominant role in Pi homeostasis in the mouse (as demonstrated by the NaPi-IIa−/− mouse (Beck et  al., 1998)), mutations identified in hypophosphataemic patients have been associated more often with NaPi-IIc, rather than NaPi-IIa. These findings have raised a controversy about the relative importance of each of these transporters in human Pi homeostasis. Correlation of hypophosphataemic syndromes and mutations in SLC34A1 (NaPi-IIa) have not been reliably established until recently. Although different heterozygous NaPi-IIa variations have been described in patients with urolithiasis or osteoporosis, and persistent idiopathic hypophosphataemia (Prie et al., 2002), and in a large cohort of hypercalciuric stone-forming kindred (Lapointe et  al., 2006), in both cases these genetic variations do not seem to be the cause of the abnormalities in these patients (Bastepe and Juppner, 2008). Magen et al. described two siblings with autosomal recessive Fanconi syndrome, severe renal Pi wasting, and hypophosphataemic rickets, carrying genetic mutation in NaPi-IIa (Magen et al., 2010), in the form of an in-frame duplication of 21 base pairs that induce loss of function (Alizadeh Naderi and Reilly, 2010). It is interesting to note that mutations in the main intestinal Na/Pi transporter, SLC34A2 or NaPi-IIb, have been reported as the cause of pulmonary alveolar microlithiasis (PAM) (Corut et  al., 2006; Huqun et al., 2007), but these patients do not exhibit hypophosphataemia or any other symptoms of unregulated Pi homeostasis. Moreover, NaPi-IIb deficiency is embryonically lethal in knockout mice (Ohi et al., 2011), but obviously not in humans, which raises questions about species specific mechanisms involved in Pi homeostasis, at least in the embryo. Finally, mutations in the sodium/hydrogen exchanger regulatory factor 1 (NHERF1) have been implicated in the occurrence of nephrolithiasis and osteoporosis (Karim et al., 2008). NHERF1 is a PDZ protein involved in the stabilization of NaPi-IIa, and presumably also NaPi-IIc, in the apical membrane of proximal tubules (Shenolikar et al., 2002; Villa-Bellosta et al., 2008). The loss of function of NHERF1 would imply decreased activity of the Na/Pi transporters with impaired renal Pi reabsorption (Levi and Breusegem, 2008; Giral et al., 2011)

Familial hypophosphataemia caused by increased FGF23 signalling pathway

Dietary Intake

Intestinal Absorption

hypo-/hyperphosphataemia

Cells

Urinary excretion

Fig. 39.1  The major determinants of serum inorganic phosphate (Pi) concentration. Serum phosphate concentration is determined by dietary intake of Pi, shift of Pi into and out of cells, intestinal absorption and urinary excretion of Pi.

Several inherited hypophosphataemic disorders have been associated with abnormally elevated circulating levels of FGF23, occurring due to its impaired degradation or increased secretion from osteocytes. FGF23 was first identified in the serum of patients with tumour-induced osteomalacia (TIO) as a circulating phosphaturic factor that regulates Pi reabsorption and 1,25(OH)2D synthesis in the renal proximal tubule (Shimada et al., 2001; White et al., 2001). (For a more detailed discussion of the FGF23 actions see Chapter 25 or the excellent reviews by Rowe (2004) and Juppner (2007).)

X-linked hypophosphataemic rickets The most common form of hereditary rickets is X-linked hypophosphataemic rickets (XLH) (OMIM 307800)  with an

385

386

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 39.2  Causes of hypophosphataemia Increased urinary excretion

Decreased gastrointestinal absorption

Miscellaneous causes/increased translocation

A. Primary hyperparathyroidism

A. Abnormalities of vitamin D metabolism 1. Vitamin D-deficient rickets 2. Familial a. Vitamin D-dependent rickets b. X-linked hypophosphataemia

A. Leukaemia, lymphoma

B. Secondary hyperparathyroidism

B. Malabsorption

B. D  iabetes mellitus: during treatment for ketoacidosis

C. Renal tubular defects (Fanconi syndrome

C. Malnutrition-starvation

C. Severe respiratory alkalosis

D. Diuretic phase of acute tubular necrosis

D. Recovery phase of malnutrition

E. Post-obstructive diuresis

E. Alcohol withdrawal

F. Extracellular fluid volume expansion

F. Toxic shock syndrome

G. Familial 1. X-linked hypophosphataemia 2. Autosomal dominant hypophosphataemic rickets 3. Autosomal recessive hypophosphataemic rickets 1; autosomal recessive hypophosphataemic rickets 2 4. Mutations in NaPi-IIa

G. Severe burns

H. Acquired 1. Oncogenic hypophosphataemic osteomalacia or tumour-induced osteomalacia 2. McCune–Albright syndrome/fibrous dysplasia 3. Post-transplant hypophosphataemia

incidence of 1:20,000 individuals. XLH syndrome is caused by loss of function mutations in the PHEX gene (Phosphate-regulating gene with Homologies to Endopeptidases on the X-chromosome), a membrane-bound member of the neutral endopeptidases expressed mainly by osteoblasts in bone and odontoblasts in teeth. Several investigations have proved that FGF23 is not the direct substrate for PHEX (Liu et al., 2003; Benet-Pages et al., 2004; Sitara et al., 2004) as was first believed, implying the existence of an unidentified intermediate substrate that controls FGF23 expression (Pettifor, 2008). It has been proposed that PHEX loss of function may cause overexpression of FGF23 through a role in osteocyte maturation (see later for DMP1 in ARHR1) (Farrow and White, 2010). Thus, increased circulating levels of intact FGF23 (the active form) would induce hypophosphataemia by two mechanisms: (1) Na/Pi transporter inactivation with subsequent renal Pi wasting, and (2) reduction of circulating 1,25(OH)2D levels that decrease intestinal Pi absorption.

hormone causing phosphaturia and reduced 1,25(OH)2D levels in a similar way to XLH. However, the clinical spectrum of ADHR is highly variable, possibly due to individual adaptability of patients to control FGF23 levels, as discussed below.

Autosomal dominant hypophosphataemic rickets

DMP1 is a key regulatory protein required for bone growth and development. DMP1 has been also proposed to suppress production or secretion of FGF23 by an unclear mechanism. Inactivating mutations in DMP1 presumably induce increased circulating levels of FGF23 that cause the hypophosphataemic symptoms. As already mentioned, PHEX loss of function may also cause overexpression of FGF23 through an effect on osteocyte maturation (cf. DMP1 in ARHR1). ENPP1 is a cell surface enzyme that generates inorganic pyrophosphate (PPi), a physiological inhibitor of hydroxyapatite crystal

Autosomal dominant hypophosphataemic rickets (ADHR) (OMIM 193100) is a very rare disease that is caused by mutations in the cleavage site of FGF23 (RXXR), increasing its resistance to proteolysis and therefore the half-life of the intact FGF23 hormone. It is important to note that two different commercial assays to measure FGF23 levels are currently available: one detects the full active form (intact peptide) and the other the C-terminal region (which would include active and inactive forms) (Juppner, 2007). In theory, increased stability of FGF23 would induce higher levels of the

Autosomal recessive hypophosphataemic rickets Mutations in two different genes have been found in autosomal recessive hypophosphataemic rickets (ARHR) resulting in two subtypes: ◆ ARHR1

(OMIM 241520) patients present with mutations in the gene that encodes the Dentin Matrix Protein 1 (DMP1), a member of the SIBLING (small integrin-binding ligand N-linked glycoprotein) family (Bastepe and Juppner, 2008).

◆ ARHR2

(OMIM 613312) was described recently as a syndrome induced by loss of function mutations in the Ectonucleotide Py rophosphatase/Phosphodiesterase 1 (ENPP1) gene (Levy-Litan et al., 2010; Lorenz-Depiereux et al., 2010).

chapter 39 

deposition and osteoblasts differentiation. Loss of function mutations of ENPP1 are the cause of generalized arterial calcification of infancy (GACI) syndrome. However, several ARHR patients were recently identified with ENPP1 mutations in which a hypophosphataemic phenotype seems to be protective against the calcifications observed in CAGI patients. It is believed that an increased level of FGF23 induced by an unknown mechanism also causes the hypophosphatemia observed in ARHR2 patients.

Related acquired disorders—tumour-induced osteomalacia and McCune–Albright syndrome Although TIO and McCune–Albright syndrome (MAS) are not genetic disorders, we include them in this section because they show many similar features with the inherited syndromes associated to FGF23 gain of function. TIO is a rare paraneoplastic syndrome caused by small endocrine tumours that secrete high levels of the phosphaturic hormones or ‘phosphatonins’ including FGF23, MEPE and frizzled related protein-4 (FRP-4). FGF23 is the most extensively studied. However, serum levels of FGF23 are found to be elevated in most but not all TIO patients. The length of time from onset of symptoms until diagnosis is often long, resulting in multiple fractures, height loss and bone pain, and muscle weakness. Curiously, there are also a high percentage of patients that show symptoms without presenting with a localized tumour. Although the disease has been described mainly in adult subjects, paediatric patients have also been reported. The diagnosis is based on the impaired renal Pi reabsorption, making it necessary to differentially diagnose from the inherited disorders (XLH, ADHR, and AHRH). Once inherited and other acquired disorders have been eliminated, it is important to locate the tumour causing the disease. The treatment of choice is the resection of the tumour, which results in rapid improvement in most cases. When the tumour cannot be localized, treatment is similar to inherited hypophosphataemias, including Pi supplementation and calcitriol (for review, see Chong et al., 2011). MAS is a rare disease that results from gain-of-function mutations occurring during early development in the gene encoding the α subunit of a stimulatory G protein, GNAS. The mutations in the GNAS gene may affect many different tissues, but when they occur in bone induce fibrous dysplasia. These patients develop hypophosphatemia induced by renal Pi wasting due to elevated circulating FGF23, presumably produced by the bone cells with mutations in the GNAS gene by an unknown mechanism.

Klotho Recently, a case of hypophosphataemic rickets and hyperparathyroidism was associated with a de novo translocation of the α-Klotho gene (Brownstein et  al., 2008). Klotho is a protein found in both transmembrane and soluble forms that is required for FGF23 signalling through binding with the FGF receptors (FGFR). The mutation of Klotho in this patient causes increased circulating levels of α-Klotho that result in marked elevation of plasma FGF23 and PTH levels. The induced hypophosphataemia is more severe than in usual XLH patients and could be caused by the action of the elevated phosphaturic hormones, FGF23 and PTH, but also from the direct action of Klotho on the renal NaPi transporters.

hypo-/hyperphosphataemia

Clinical features and laboratory findings XLH, ADHR, and ARHR syndromes show a similarity in clinical presentation, including growth retardation, skeletal deformities of the lower and upper extremities, and metaphyseal widening (Alizadeh Naderi and Reilly, 2010). Although it is considered that elevated FGF23 plays a central role in hypophosphataemia and disturbed vitamin D metabolism, it is unclear whether the bone disease is a direct or indirect effect of FGF23. Affected patients develop rickets, a disorder in which abnormal mineralization of bone and growth plate cartilage results in diminished bone strength, deformity, short stature, and bone pain. Rapidly growing bones of the lower extremities generally show the most striking abnormalities. As a consequence of the specific genes affected, there are differences in clinical presentation for each disorder. For example, XLH manifests frequently during late infancy and therapy can improve, but usually not completely resolve, the symptoms (Petersen et al., 1992; Makitie et al., 2003). However, clinical manifestations vary in severity:  while males manifest the full spectrum of symptoms, females show a more diverse response from asymptomatic hypophosphataemia to a severe syndrome identical to that present in males (Graham et al., 1959). In ADHR, the incomplete penetrance leads to a more variable age of clinical onset. Young ADHR patients present with similar clinical features to XLH, but adolescent or adult female patients present with milder symptoms such as muscle weakness, bone pain and fractures, indicative of a late onset (Econs and McEnery, 1997). Remittance of symptoms has even been described in a number of older subjects (Econs and McEnery, 1997). In contrast to ADHR, affected individuals with ARHR present late during childhood and even into adulthood. Several striking features are helpful in differentiating TIO from the other hypophosphataemic syndromes: clinical features include severe muscle weakness, marked bone demineralization and severe osteomalacia. In addition, 1,25(OH)2D concentrations are markedly suppressed. Typical laboratory findings in TIO include hypophosphataemia and low or inappropriately normal circulating 1,25(OH)2D, normal serum calcium and 25(OH)D, and high or inappropriately normal FGF23 levels. Values in the normal range of 1,25(OH)2D or FGF23 are inappropriate in the context of hypophosphataemia, because the physiological homeostatic mechanisms would try to compensate by increasing 1,25(OH)2D (increasing intestinal absorption) and decreasing FGF23 levels (reduced Pi excretion). Hypophosphataemia develops within the first few months of life and is the first indicator in young infants with a family history of hypophosphataemia. Definitive evidence of renal Pi wasting is critical for diagnosis of heritable hypophosphataemic disorders, thereby ruling out the nutritional form of rickets mediated by intestinal malabsorption. Renal Pi wasting can be determined by measurement of the percentage tubular reabsorption of Pi (TRPi) or measuring the serum Pi threshold (tubular threshold maximum corrected for glomerular filtration rate, or TmPi/GFR). Both of these values are far below the normal range in inherited hypophosphataemic patients. The diagnosis normally includes FGF23 measurements since elevated circulating levels are the cause of the underlying pathogenesis. A recent report found that FGF23 values were surprisingly not consistently elevated in patients with ADHR and that serum concentrations fluctuated between normal and elevated values, depending

387

388

Section 2  

fluid, electrolyte, and renal tubular disorders

on whether or not the individual subject was hypophosphataemic (symptomatic) (Imel et al., 2007; Sun et al., 2012). Although FGF23 concentrations are not invariably increased in all the patients, an inverse relationship between FGF23 and the degree of hypophosphataemia has been found. It has been proposed that individual genetic and environmental background could help to find alternative pathways to bypass the otherwise increased FGF23 activity. In ADHR, for example, adjustment of the expression and secretion of the more stable mutated FGF23 from osteoblasts could in some cases compensate for the extended half-life of the hormone, thereby maintaining Pi homeostasis and resulting in patients with asymptomatic or milder clinical presentation. XLH, as the most common familial hypophosphataemia, has the best established medical management, but similar treatment strategies could be adopted in the other hypophosphataemic conditions associated with elevated FGF23 levels such as ADHR and AHRH. Treatment is started in the first year of life with high doses of elemental phosphorus daily and 1,25(OH)2D supplementation to avoid development of secondary hyperparathyroidism. 1,25(OH)2D has to be increased later in childhood and needs to be adjusted to avoid hypercalcaemia and hypercalciuria. However, a balanced treatment is difficult to achieve and often patients develop nephrocalcinosis, and in later stages tertiary hyperparathyroidism with hypercalcaemia, hypertension, and kidney damage (Verge et al., 1991; Alon et al., 2003; Tournis et al., 2011). A more detailed discussion about the classical treatment and dosages can be found elsewhere (Bastepe and Juppner, 2008; Carpenter et  al., 2011). Future treatments may also include administration of FGF23 antibody (Yamazaki et al., 2008; Aono et al., 2009), although a recent study identified complications related to this approach (Shalhoub et al., 2012).

Familial hypophosphataemias caused by impaired PTH or vitamin D signalling pathway Several other hypophosphataemic inherited disorders have been associated with the two classical hormones controlling Pi and calcium homeostasis, PTH and vitamin D.  Several syndromes listed in Table 39.3 are associated with hyperparathyroidism from disturbances of calcium homeostasis inducing subsequent hypophosphataemia. Although the patients affected by these syndromes usually present with hypophosphataemia, the clinical features are associated with important abnormalities of calcium metabolism. The diseases associated with these pathways are discussed more extensively in Chapters 37 and 38.

Post-transplant hypophosphataemia Post-transplant hypophosphataemia, a common disorder, is well described in the literature. Although described mainly in patients following renal transplantation (Gyory et al., 1969; Moorhead et al., 1974; Ward et al., 1977; Graf et al., 1979; Better, 1980; Garabedian et al., 1980; Olgaard et al., 1980; Rosenbaum et al., 1981; Sakhaee et al., 1985; Parfitt et al., 1986; Pabico and McKenna, 1988; Steiner et al., 1993; Levi, 2001), post-transplant hypophosphataemia also occurs in patients undergoing bone marrow transplantation (Raanani et  al., 1995; Crook et  al., 1996). In all reports, the decrease in serum Pi concentration was associated with an increase in urinary Pi excretion and a significant decrease in the measured or derived TmPi/GFR (Walton and Bijvoet, 1975). In addition to the

Table 39.3  Familial hypophosphataemic disorders classified by pathophysiological cause Disease Defective renal Na/Pi transporter activity Hereditary hypophosphataemic rickets with hypercalciuria

HHRH

Autosomal recessive Fanconi syndrome and hypophosphataemic rickets

ARFS

Nephrolithiasis/osteoporosis, hypophosphataemic 2 Increased FGF23 signalling pathway X-linked hypophosphataemic rickets

XLH

Autosomal-dominant hypophosphataemic rickets

ADHR

Autosomal-recessive hypophosphataemic rickets

ARHR1 ARHR2

Hypophosphataemic rickets and hyperparathyroidism Activated PTH signalling pathway Jansen-type metaphyseal chondrodysplasia

JMC

Multiple endocrine neoplasia type 1

MEN1

Hyperparathyroidism-jaw tumour syndrome

HPTJT

Familial hypocalciuric hypercalcaemia

FHH

Neonatal severe primary hyperparathyroidism

NSHPT

Decreased vitamin D signalling pathway Vitamin D-dependent rickets type 1

VDDR1

Vitamin D-dependent rickets type 1

VDDR2

impairment in renal tubular Pi reabsorption, evidence indicates that intestinal Pi absorption is impaired in transplant patients (Farrington et al., 1979; Rosental et al., 1982; Massari, 1997; Levi, 2001). The mechanism for post-transplant hypophosphataemia has not been fully elucidated, but it is linked to disordered regulation of renal tubular reabsorption of Pi. Although a role for PTH in mediating this defect is still uncertain, recent studies suggest a role for FGF23 (Sanchez Fructuoso et al., 2012a, 2012b).

Decrease in gastrointestinal absorption of phosphorus Abnormalities of vitamin D metabolism Vitamin D and its metabolites play an important role in Pi homeostasis (Gray et al., 1977). Vitamin D promotes the intestinal absorption of calcium and Pi, and is necessary to maintain the normal mineralization of bone.

Vitamin D deficient rickets Diets deficient in vitamin D lead to the metabolic disorder known as rickets when it occurs in children and osteomalacia when it appears in adults (Nemere and Norman, 1987). Vitamin D deficiency in childhood results in severe deformities of bone, because of rapid growth.

Vitamin D-dependent (-resistant) rickets These are recessively inherited forms of vitamin D refractory rickets. The conditions are characterized by hypophosphataemia, hypocalcaemia, elevated levels of serum alkaline phosphatase, and, in

chapter 39 

hypo-/hyperphosphataemia

Table 39.4  Serum FGF23, PTH, vitamin D, and calcium levels in hypophosphataemic disorders Disease

Gene

FGF23

PTH

1,25(OH)2D

Calcium

Defective renal Na/Pi transporter activity HHRH

SLC34A3

Low

Low

High

High

ARFS

SLC34A1

Low/normal

n.d.

Low

Normal

NHERF1

Normal

Normal

High

Normal

Increased FGF23 signalling pathway XLH

PHEX

High/normala

Normal

Low/normala

Normal

ADHR

FGF23

High/normal

Normal

Low

Normal

ARHR1

DMP1

High

Normal

Low/normal

Normal

ENPP1

High/normala

Normal

Low/normala

Normal

KL

High

High

High

High

ARHR2

a Normal refers here to values in normal range but inappropriately high or low in the context of hypophosphataemia.

some instances, generalized aminoaciduria and severe bone lesions. Currently, two main forms of vitamin D-dependent rickets have been characterized. The serum concentrations of 1,25(OH)2D serves to differentiate the two types of vitamin D-dependent rickets Table 39.4. Type I vitamin D-dependent rickets is associated with reduced calcitriol levels. It is caused by a mutation in the gene converting 25(OH)D to 1,25(OH)2D1, the renal 1α -hydroxylase enzyme (Fu et al., 1997). This condition responds to very large doses of vitamin D2 and D3 (× 100–300 the normal requirement of physiological doses), or to 0.5 to 1.0 micrograms per day of 1,25(OH)2D. Type II vitamin D-dependent rickets is characterized by end-organ resistance to1,25(OH)2D. Plasma levels of 1,25(OH)2D are elevated. This finding, in association with radiographic and biochemical signs of rickets, and implies resistance to 1,25(OH)2D in its target tissues.

Clinical and biochemical manifestations of hypophosphataemia The manifestations of hypophosphataemia are presented in Table 39.5. The clinical manifestations of hypophosphatemia and severe Pi depletion are related to disturbances in cellular energy and metabolism. Patients with mild degrees of hypophosphataemia are usually asymptomatic. However, if hypophosphataemia is severe, that is, if serum Pi levels are < 0.48 mmol/L (< 1.5 mg per dL), a series of haematological, neurological, and metabolic disorders may develop. In general, the patients become anorectic and weak, and mild bone pain may be present if the hypophosphataemia persists for several months (Table 39.5).

Treatment of hypophosphataemia There are several general principles that apply to the treatment of hypophosphataemic patients. As with any predominantly intracellular ion (e.g. potassium), the state of total body phosphorus stores, as well as the magnitude of phosphorus losses, cannot be readily assessed by measurement of the concentrations in serum. In fact, under conditions in which a rapid shift of Pi has resulted from glucose infusion or hyperalimentation, total body stores of phosphorus may be normal, although with diminished intake and renal losses, there may be severe Pi depletion. Furthermore, the volume

of distribution of Pi may vary widely, reflecting in part the intensity and duration of the underlying cause. In clinical situations in which hypophosphataemia is to be expected (e.g. glucose infusion or hyperalimentation in the alcoholic or nutritionally compromised patient, and during treatment of diabetic ketoacidosis), careful monitoring of the concentration of serum Pi is crucial. In these situations, addition of Pi supplementation to prevent the development of severe hypophosphataemia may prove very helpful. It is now generally recommended that hyperalimentation solutions contain a Pi concentration of 12–15 mmol/L or 37–46.5 mg/dL, to provide an appropriate amount of phosphorus in patients where renal impairment is absent (Lentz et  al., 1978). Phosphorus supplementation during glucose infusion or during the treatment of diabetic ketoacidosis is usually withheld until the serum Pi levels decrease to < 0.32 mmol/L (1 mg/dL). Phosphorus may be given orally to these patients and others with mild asymptomatic hypophosphataemia in the form of skimmed milk, which contains 0.9 mg/mL, Neutra-Phos (3.3 mg/mL), or phosphorus soda (129 mg/ mL). However, intestinal absorption is quite variable, and diarrhoea often complicates the oral administration of phosphate-containing compounds. For these reasons, parenteral administration is usually recommended in the hospitalized patient. If oral therapy is permissible, Fleet Phospho-Soda may be given at a dosage of 60 mmol daily in three doses (21 mmol/5 mL or 643 mg/5 mL). In patients with severe Pi depletion, it is difficult to determine the magnitude of the total deficit in Pi and to calculate a precise initial dose. It is usually prudent to proceed with caution and repair the deficit slowly. The most frequently recommended regimen is 0.08 mmol/ kg body weight (2.5 mg/kg body weight) given over 6 hours for severe, but uncomplicated, hypophosphataemia and 0.16  mmol/ kg body weight (5 mg/kg of body weight) in symptomatic patients (Lentz et al., 1978). Parenteral administration should be discontinued when the serum Pi concentration is > 0.65 mmol/L (> 2 mg/dL).

Hyperphosphataemia In adults, hyperphosphataemia occurs at serum Pi levels > 1.6  mmol/L, whereas in children and adolescents, serum phosphorus levels > 2.24  mmol/L are considered abnormal. Hyperphosphataemia is a common clinical problem, particularly

389

390

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 39.5  Clinical and biochemical manifestations of severe hypophosphataemia Cardiovascular and Carbohydrate skeletal muscle metabolism

Haematological alterations

Neurological manifestations

Skeletal abnormalities

Biochemical manifestations

A. Decreased cardiac output

A. Hyperinsulinaemia

A.  Red blood cells 1. Decreased adenosine triphosphate (ATP) content 2.  Decreased 2,3-DPG 3. Decreased P 50 4. Increased oxygen affinity 5.  Decreased lifespan 6. Haemolysis 7. Spherocytosis

A. Anorexia

A.  Bone pain

A. Low A. Hypercalciuria parathyroid hormone levels

B. Muscle weakness

B. Decreased glucose metabolism

B. Leucocytes 1. Decreased phagocytosis 2. Decreased chemotaxis 3. Decreased bactericidal activity

B. Irritability C. Confusion

B. Radiolucent areas (X-ray)

B. Increased 1,25(OH)2D3

C. Platelets 1. Impaired clot retraction 2. Thrombocytopenia 3. Decreased ATP content 4. Megakaryocytosis 5.  Decreased lifespan

D. Paraesthesias

C. Pseudofractures C. Increased C. Hypermagnesuria creatinine D. Hypophosphaturia phosphokinase

E. Dysarthria

D. Rickets or osteomalacia

C. Decreased transmembrane resting potential

D. Rhabdomyolysis

D. Increased aldolase

Renal manifestations

B. Hypomagnesaemia

E. Decreased glomerular filtration rate

F. Ataxia

F. Decreased Tm for bicarbonate

G. Seizures

G. Decreased renal gluconeogenesis

H. Coma

H. Decreased titratable acid excretion

in those with significantly reduced renal function, because the kidneys play a key role in eliminating phosphorus from the body. A number of mechanisms result in hyperphosphataemia (Table 39.6), which are discussed in more detail below.

Impaired renal excretion Since the kidney is the major organ via which phosphorus is removed from the body, elevated serum Pi occurs in patients with acute or chronic renal failure. Serum Pi levels are almost always elevated in patients with an estimated glomerular filtration rate (eGFR) 2.56 mmol/L) causes symptoms by inducing hypocalcaemia, which results from calcium precipitating with phosphorus. Hypocalcaemia often causes muscle cramps, but can also result in altered mental status, seizures, arrhythmias, and hypotension (Shiber and Mattu, 2002). Other manifestations of hyperphosphataemia include anorexia, nausea, and vomiting. Calcium phosphate deposition in the kidney can cause renal failure. Acute Pi nephropathy may occur after ingestion of oral sodium phosphate purgatives, which are used in preparation for colonoscopy (Markowitz and Perazella, 2009). Criteria for diagnosis of acute Pi nephropathy are: (1) AKI after exposure to oral sodium phosphate purgatives, (2) renal biopsy showing widespread tubular and interstitial deposition of calcium phosphate crystals, (3) no prior evidence of hypercalcaemia or conditions known to cause hypercalcaemia, and (4)  no evidence of any other renal disease. Those at greatest risk for acute Pi nephropathy are patients who are > 60 years of age, those with pre-existing renal disease, those with hypertension, concomitant use of angiotensin-converting enzyme inhibitors, angiotensin receptor blockers or diuretics, and patients who are female. The reported incidence of acute Pi nephropathy after ingestion of oral sodium phosphate varies widely among studies (from 0.5% to 6.3% of patients). This variability is due in part to the variation in the definition of AKI used. An elevated calcium × phosphorus product (product > 60) is one of the risk factors for development of calcific uraemic arteriolopathy (CUA), although CUA may be seen in patients with relatively normal levels of serum calcium and Pi. CUA is most often seen in patients with CKD, particularly those requiring renal replacement therapy (Hayden et al., 2008). It is characterized by arteriolar calcification in the artery media and thrombosis within subcutaneous fat tissue. Clinically CUA leads to painful skin lesions and non-healing skin ulcerations. The prevalence of CUA is estimated to be between 1% and 5% of dialysis patients. Treatment is aimed at controlling mineral metabolism parameters and meticulous wound care. Sodium thiosulfate is also sometimes used as an antioxidant and anti-inflammatory agent. CUA can be a life-threatening disease, since many patients succumb to infection and sepsis.

Laboratory findings Laboratory work-up of hyperphosphataemia includes creatinine, electrolytes (including calcium and Pi), albumin, PTH, 25(OH)D and 1,25(OH)2D, and a complete (full) blood count (CBC). The creatinine and electrolytes give an indication of renal function. The calcium (corrected for albumin), Pi and PTH allow assessment of whether there is hypoparathyroidism or resistance to PTH, or whether secondary hyperparathyroidism is present; 25(OH)D and 1,25(OH)2D give an indication of whether vitamin D intoxication is present. This is important, because 1,25(OH)2D increases Pi absorption in the gut, which can lead to hyperphosphataemia. The CBC gives an indication of whether a haematological malignancy is

present. Renal biopsy is generally reserved only for suspected cases of acute Pi nephropathy. Imaging is not indicated unless it is part of the workup of AKI or CKD.

Treatment and outcomes Shift or release Hyperphosphataemia due to a shift from intracellular to extracellular pools or due to release from intracellular stores is treated by treating the underlying condition. It is also important to preserve renal function to ensure adequate excretion of Pi.

Impaired renal function Dietary restriction of Pi is a key component of treatment in patients with impaired renal function, especially in those with CKD and an eGFR < 30 mL/min/1.73 m2. In most cases of CKD with an eGFR < 30 mL/min/1.73 m2 dietary restriction alone does not suffice and oral Pi binders must be used. There are several different formulations of these binders, which are discussed in more detail below.

Phosphate binders Calcium-based phosphate binders Calcium carbonate has been used for decades in dialysis patients whereas calcium acetate is a newer formulation. There have been a few trials comparing the efficacy of calcium carbonate with that of calcium acetate, but these are limited by small numbers of participants and relatively high dropout rates. Pflanz et al. conducted a randomized crossover trial in dialysis patients in which participants received equimolar doses of calcium carbonate or calcium acetate for 6 weeks each (Pflanz et al., 1994). Of the 31 patients originally enrolled, only 23 completed both arms of the study. Patients in the acetate group had significantly lower serum Pi, calcium × phosphorus product, and intact PTH (iPTH) after completion of treatment than those in the carbonate group, but they also had a significantly higher serum calcium level. Another trial by Janssen et  al. compared the efficacy of aluminium hydroxide (discussed below), calcium carbonate, and calcium acetate in 53 dialysis patients (Janssen et al., 1996). In this trial the per gram dose of elemental calcium administered was the same in the carbonate and acetate groups, but the number of capsules of calcium acetate taken daily was less than the number of tablets of calcium carbonate. Serum Pi levels were equivalent in the carbonate and acetate groups, but significantly lower in the aluminium hydroxide group. Interestingly, in this trial the incidence of hypercalcaemia was significantly less in the acetate than the carbonate group.

Sevelamer Sevelamer is a resin that binds phosphate. It is currently marketed as sevelamer carbonate, but most of the trials conducted used the original formulation of sevelamer hydrochloride. The most well-known trial comparing the efficacy of sevelamer with that of calcium-based Pi binders is the Dialysis Clinical Outcomes Revisited (DCOR) trial (Suki et al., 2007). There were 2103 dialysis patients enrolled in the study, of which 1068 patients completed treatment. Patients were randomized to the sevelamer arm or the calcium arm (70% received calcium acetate and 30% received calcium carbonate). There was no significant difference between the groups in all cause or cause-specific mortality. However, a subgroup analysis showed a significantly lower mortality rate in

chapter 39 

the sevelamer group in patients over the age of 65. Two recent meta-analyses of trials comparing calcium-based Pi binders and sevelamer have also shown no statistically significant differences in cardiovascular mortality between patients taking either type of binder (Tonelli et al., 2007; Jamal et al., 2009). Tonelli et al. examined 14 publications of randomized trials involving 3193 dialysis patients. In comparison with calcium-containing Pi binders, sevelamer use was associated with slightly lower serum calcium levels, similar to slightly higher serum Pi levels, and a similar calcium × phosphorus product. There was no significant difference in all-cause mortality, cardiovascular mortality or the frequency of symptomatic bone disease. There was a trend towards decreased all-cause mortality in subjects taking non-calcium based versus calcium-based binders, but this was not statistically significant. Five trials involving 469 patients reported coronary artery calcification scores. There was no difference in coronary artery calcification between patients taking calcium versus non-calcium containing binders. In summary, there is little hard evidence that sevelamer is superior to calcium-based Pi binders in decreasing the occurrence of clinically relevant endpoints such as cardiovascular mortality or bone disease.

Lanthanum Lanthanum carbonate is a Pi binder that does not contain calcium or aluminium. Studies to date with this binder have not been adequately powered to show differences in mortality (Tonelli et al., 2010). Results of trials performed so far show no difference in cardiovascular complications in subjects taking lanthanum versus calcium-containing Pi binders. In the largest study, dialysis patients were randomized to receive lanthanum (N = 682) or their usual Pi binder (N = 677). Over a 2-year follow-up period serum Pi levels were similar between the two groups and the lanthanum group had better serum calcium levels and iPTH values than the usual Pi binder group. It should be noted, however, that the dropout rate among subjects in the lanthanum group was high (71%) over the study period.

Aluminium hydroxide While aluminium hydroxide is highly effective in lowering serum Pi levels, its use is complicated by aluminium retention and aluminium-induced encephalopathy in patients with renal failure (McDermott et al., 1978; Salusky et al., 1991). Thus, aluminium-containing Pi binders should never be used for more than a few days. In our practice, aluminium hydroxide is only used for very short periods (< 1 week) in patients with very high serum Pi levels that are refractory to other Pi binders; repeat courses of aluminium hydroxide are never given.

Guidelines Kidney Disease: Improving Global Outcomes (KDIGO) is a global non-profit organization whose mission is to improve the treatment and outcomes of CKD. KDIGO recommendations for monitoring calcium, Pi, and intact PTH are given in Table 39.8 (see ) For renal transplant patients, KDIGO guidelines recommend measuring serum calcium and Pi at least once weekly in the immediate post-transplant period until values stabilize.

hypo-/hyperphosphataemia

Table 39.8  Frequency of measurement of calcium, phosphorus, and iPTH for different stages of CKD CKD stage

Parameters

Frequency of measurement

3

Ca, Phos

6–12 months

3

iPTH

Depends on disease progression

4

Ca, Phos

3–6 months

4

iPTH

6–12 months

5 (including dialysis)

Ca, Phos

1–3 months

5 (including dialysis)

iPTH

3–6 months

Ca = calcium; iPTH = intact PTH; Phos = phosphorus.

In terms of treatment, KDIGO guidelines suggest maintaining serum Pi in the normal range for patients with CKD stages 3–5 and lowering serum Pi towards the normal range for patients on dialysis. Recommendations also state that calcium-based Pi binder use should be restricted in patients with persistent or recurrent hypercalcaemia, those with a low iPTH, adynamic bone disease, or evidence of vascular calcification. (A full set of KDIGO guidelines can be found at )

National Kidney Foundation Kidney Disease Outcomes Quality Initiative In the United States, the National Kidney Foundation (NKF) has issued its own set of guidelines that were last published in 2003. A  new set of guidelines is to be published soon. In general, the Kidney Disease Outcomes Quality Initiative (KDOQI) guidelines offer similar recommendations to those of KDIGO. (A full set of the NKF KDOQI guidelines can be found at .)

Outcomes Acute transient hyperphosphataemia due to AKI generally has no long-term sequelae. In patients with CKD, chronic hyperphosphataemia is believed to contribute to cardiovascular morbidity and mortality (Block et al., 2004; Tuttle and Short, 2009). Patients that develop acute Pi nephropathy usually do not recover their baseline renal function (Markowitz and Perazella, 2009). In one study in Iceland, of 15 patients diagnosed with acute Pi nephropathy after a mean follow up of 26.6 months, one patient reached end-stage renal disease, one patient died with progressive renal failure, and the remaining patients had CKD with a mean serum creatinine of 184.4 µmol/L (Pálmadóttir et al., 2010).

References Abu-Alfa, A. K. and Younes, A. (2010). Tumor lysis syndrome and acute kidney injury: evaluation, prevention, and management. Am J Kidney Dis, 55(5, Suppl 3), S1–S13. Alizadeh Naderi, A. S. and Reilly, R. F. (2010). Hereditary disorders of renal phosphate wasting. Nat Rev Nephrol, 6(11), 657–65. Alon, U. S., Monzavi, R., Lilien, M., et al. (2003). Hypertension in hypophosphatemic rickets—role of secondary hyperparathyroidism. Pediatr Nephrol, 18(2), 155–8.

393

394

Section 2  

fluid, electrolyte, and renal tubular disorders

Aono, Y., Yamazaki, Y., Yasutake, J., et al. (2009). Therapeutic effects of anti-FGF23 antibodies in hypophosphatemic rickets/osteomalacia. J Bone Miner Res, 24(11), 1879–88. Bastepe, M. and Juppner, H. (2008). Inherited hypophosphatemic disorders in children and the evolving mechanisms of phosphate regulation. Rev Endocr Metab Disord, 9(2), 171–80. Beck, L., Karaplis, A. C., Amizuka, N., et al. (1998). Targeted inactivation of Npt2 in mice leads to severe renal phosphate wasting, hypercalciuria, and skeletal abnormalities. Proc Natl Acad Sci U S A, 95(9), 5372–7. Benet-Pages, A., Lorenz-Depiereux, B., Zischka, H., et al. (2004). FGF23 is processed by proprotein convertases but not by PHEX. Bone, 35(2), 455–62. Bergwitz, C., Roslin, N. M., Tieder, M., et al. (2006). SLC34A3 mutations in patients with hereditary hypophosphatemic rickets with hypercalciuria predict a key role for the sodium-phosphate cotransporter NaPi-IIc in maintaining phosphate homeostasis. Am J Hum Genet, 78(2), 179–92. Better, O. S. (1980). Tubular dysfunction following kidney transplantation. Nephron, 25(5), 209–13. Block, G. A., Klassen, P. S., Lazarus, J. M., et al. (2004). Mineral metabolism, mortality, and morbidity in maintenance hemodialysis. J Am Soc Nephrol, 15(8), 2208–18. Brown, E. M. (2009). Anti-parathyroid and anti-calcium sensing receptor antibodies in autoimmune hypoparathyroidism. Endocrinol Metab Clin North Am, 38(2), 437–45. Brownstein, C. A., Adler, F., Nelson-Williams, C., et al. (2008). A translocation causing increased alpha-klotho level results in hypophosphatemic rickets and hyperparathyroidism. Proc Natl Acad Sci U S A, 105(9), 3455–60. Carpenter, T. O., Imel, E. A., Holm, I. A., et al. (2011). A clinician’s guide to X-linked hypophosphatemia. J Bone Miner Res, 26(7), 1381–8. Chatzizisis, Y. S., Misirli, G., Hatzitolios, A. I., et al. (2008). The syndrome of rhabdomyolysis: Complications and treatment. Eur J Intern Med, 19(8), 568–74. Chefetz, I. and Sprecher, E. (2009). Familial tumoral calcinosis and the role of O-glycosylation in the maintenance of phosphate homeostasis. Biochim Biophys Acta, 1792(9), 847–52. Chong, W. H., Molinolo, A. A., Chen, C. C., et al. (2011). Tumor-induced osteomalacia. Endocr Relat Cancer, 18(3), R53–77. Corut, A., Senyigit, A., Ugur, S. A., et al. (2006). Mutations in SLC34A2 cause pulmonary alveolar microlithiasis and are possibly associated with testicular microlithiasis. Am J Hum Genet, 79(4), 650–6. Crook, M., Swaminathan, R., and Schey, S. (1996). Hypophosphataemia in patients undergoing bone marrow transplantation. Leuk Lymphoma, 22(3–4), 335–7. Econs, M. J. and McEnery, P. T. (1997). Autosomal dominant hypophosphatemic rickets/osteomalacia: clinical characterization of a novel renal phosphate-wasting disorder. J Clin Endocrinol Metab, 82(2), 674–81. Farrington, K., Varghese, Z., Newman, S. P., et al. (1979). Dissociation of absorptions of calcium and phosphate after successful cadaveric renal transplantation. Br Med J, 1(6165), 712–4. Farrow, E. G. and White, K. E. (2010). Recent advances in renal phosphate handling. Nat Rev Nephrol, 6(4), 207–17. Fu, G. K., Lin, D., Zhang, M. Y., et al. (1997). Cloning of human 25-hydroxyvitamin D-1 alpha-hydroxylase and mutations causing vitamin D-dependent rickets type 1. Mol Endocrinol, 11(13), 1961–70. Garabedian, M., Silve, C., Lévy, D., et al. (1980). Chronic hypophosphatemia in kidney transplanted children and young adults. Adv Exp Med Biol, 128, 249–54. Giral, H., Lanzano, L., Caldas, Y., et al. (2011). Role of PDZK1 protein in apical membrane expression of renal sodium-coupled phosphate transporters. J Biol Chem, 286(17), 15032–42. Graf, H., Kovarik, J., Stummvoll, H. K., et al. (1979). Handling of phosphate by the transplanted kidney. Proc Eur Dial Transplant Assoc, 16, 624–9. Graham, J. B., McFalls, V. W., and Winters, R. W. (1959). Familial hypophosphatemia with vitamin D-resistant rickets. II: three additional kindreds of the sex-linked dominant type with a genetic analysis of four such families. Am J Hum Genet, 11, 311–32.

Gray, R. W., Wilz, D. R., Caldas, A. E., et al. (1977). The importance of phosphate in regulating plasma 1,25-(OH)2-vitamin D levels in humans: studies in healthy subjects in calcium-stone formers and in patients with primary hyperparathyroidism. J Clin Endocrinol Metab, 45(2), 299–306. Greenbaum, L. A. (2011). Rickets and hypervitaminosis D. In R. M. Kliegman, R. E. Behrman, H. B. Jenson, et al. (eds.) Nelson Textbook of Pediatrics (19th ed.), pp. 261–3. Philadelphia, PA: Saunders. Gyory, A. Z., Stewart, J. H., George, C. R., et al. (1969). Renal tubular acidosis, acidosis due to hyperkalaemia, hypercalcaemia, disordered citrate metabolism and other tubular dysfunctions following human renal transplantation. Q J Med, 38(150), 231–54. Hayden, M., Goldsmith, D., Sowers, J. R., et al. (2008). Calciphylaxis: calcific uremic arteriolopathy and the emerging role of sodium thiosulfate. Int Urol Nephrol, 40(2), 443–51. Huqun, Izumi, S., Miyazawa, H., et al. (2007). Mutations in the SLC34A2 gene are associated with pulmonary alveolar microlithiasis. Am J Respir Crit Care Med, 175(3), 263–68. Ichikawa, S., Sorenson, A. H., Imel, E. A., et al. (2006). Intronic deletions in the SLC34A3 gene cause hereditary hypophosphatemic rickets with hypercalciuria. J Clin Endocrinol Metab, 91(10), 4022–7. Imel, E. A., Hui, S. L., and Econs, M. J. (2007). FGF23 concentrations vary with disease status in autosomal dominant hypophosphatemic rickets. J Bone Miner Res, 22(4), 520–6. Jamal, S. A., Fitchett, D., Lok, C. E., et al. (2009). The effects of calcium-based versus non-calcium-based phosphate binders on mortality among patients with chronic kidney disease: a meta-analysis. Nephrol Dial Transplant, 24(10), 3168–74. Janssen, M. J., van der Kuy, A., ter Wee, P. M., et al. (1996). Aluminum hydroxide, calcium carbonate and calcium acetate in chronic intermittent hemodialysis patients. Clin Nephrol, 45(2), 111–19. Jaureguiberry, G., Carpenter, T. O., Forman, S., et al. (2008). A novel missense mutation in SLC34A3 that causes HHRH identifies threonine 137 as an important determinant of sodium-phosphate co-transport in NaPi-IIc. Am J Physiol Renal Physiol, 295(2), F371–9. Jonsson, K. B., Zahradnik, R., Larsson, T., et al. (2003). Fibroblast growth factor 23 in oncogenic osteomalacia and X-linked hypophosphatemia. N Engl J Med, 348(17), 1656–63. Juppner, H. (2007). Novel regulators of phosphate homeostasis and bone metabolism. Ther Apher Dial, 11 Suppl 1, S3–S22. Karim, Z., Gerard, B., Bakouh, N., et al. (2008). NHERF1 mutations and responsiveness of renal parathyroid hormone. N Engl J Med, 359(11), 1128–35. Kebler, R., McDonald, F. D., and Cadnapaphornchai, P. (1985). Dynamic changes in serum phosphorus levels in diabetic ketoacidosis. Am J Med, 79(5), 571–6. Kestenbaum, B., Sampson, J. N., Rudser, K. D., et al. (2005). Serum phosphate levels and mortality risk among people with chronic kidney disease. J Am Soc Nephrol, 16(2), 520–8. Lapointe, J. Y., Tessier, J., Paquette, Y., et al. (2006). NPT2a gene variation in calcium nephrolithiasis with renal phosphate leak. Kidney Int, 69(12), 2261–7. Lentz, R. D., Brown, D. M., and Kjellstrand, C. M. (1978). Treatment of severe hypophosphatemia. Ann Internal Med, 89(6), 941–4. Levi, M. (2001). Post-transplant hypophosphatemia. Kidney Int, 59(6), 2377–87. Levi, M. and Breusegem, S. (2008). Renal phosphate-transporter regulatory proteins and nephrolithiasis. N Engl J Med, 359(11), 1171–3. Levine, B.S. and Kleeman, C. R. (1994). Hypophosphatemia and hyperphosphatemia: clinical and pathophysiologic aspects. In M. H. Maxwell and C. R. Kleeman (eds.) Clinical Disorders of Fluid and Electrolyte Metabolism, pp. 1040–5. New York: McGraw Hill. Levy-Litan, V., Hershkovitz, E., Avizov, L., et al. (2010). Autosomal-recessive hypophosphatemic rickets is associated with an inactivation mutation in the ENPP1 gene. Am J Hum Genet, 86(2), 273–8.

chapter 39 

Liu, S., Guo, R., Simpson, L. G., et al. (2003). Regulation of fibroblastic growth factor 23 expression but not degradation by PHEX. J Biol Chem, 278(39), 37419–26. Lorenz-Depiereux, B., Benet-Pages, A., Eckstein, G., et al. (2006). Hereditary hypophosphatemic rickets with hypercalciuria is caused by mutations in the sodium-phosphate cotransporter gene SLC34A3. Am J Hum Genet, 78(2), 193–201. Lorenz-Depiereux, B., Schnabel, D., Tiosano, D., et al. (2010). Loss-of-function ENPP1 mutations cause both generalized arterial calcification of infancy and autosomal-recessive hypophosphatemic rickets. Am J Hum Genet, 86(2), 267–72. Lynch, K. E., Lynch, R., Curhan, G. C., et al. (2011). Prescribed dietary phosphate restriction and survival among hemodialysis patients. Clin J Am Soc Nephrol, 6(3), 620–9. Magen, D., Berger, L., Coady, M. J., et al. (2010). A loss-of-function mutation in NaPi-IIa and renal Fanconi’s syndrome. N Engl J Med, 362(12), 1102–9. Makitie, O., Doria, A., Kooh, S. W., et al. (2003). Early treatment improves growth and biochemical and radiographic outcome in X-linked hypophosphatemic rickets. J Clin Endocrinol Metab, 88(8), 3591–7. Mantovani, G. (2011). Pseudohypoparathyroidism: diagnosis and treatment. J Clin Endocrinol Metab, 96(10), 3020–30. Markowitz, G. S. and Perazella, M. A. (2009). Acute phosphate nephropathy. Kidney Int, 76(10), 1027–34. Marwaha, R. K., Khadgawat, R., Tandon, N., et al. (2010). Reference intervals of serum calcium, ionized calcium, phosphate and alkaline phosphatase in healthy Indian school children and adolescents. Clin Biochem, 43(15), 1216–19. Massari, P. U. (1997). Disorders of bone and mineral metabolism after renal transplantation. Kidney Int, 52(5), 1412–21. McDermott, J. R., Smith, A. I., Ward, M. K., et al. (1978). Brain-aluminium concentration in dialysis encephalopathy. Lancet, 311(8070), 901–4. Moorhead, J. F., Wills, M. R., Ahmed, K. Y., et al. (1974). Hypophosphataemic osteomalacia after cadaveric renal transplantation. Lancet, 1(7860), 694–7. Nemere, I. and Norman, A. W. (1987). The rapid, hormonally stimulated transport of calcium (transcaltachia). J Bone Miner Res, 2(3), 167–9. O’Connor, L. R., Klein, K. L., and Bethune, J. E. (1977). Hyperphosphatemia in lactic acidosis. N Engl J Med, 297(13), 707–9. Ohi, A., Hanabusa, E., Ueda, O., et al. (2011). Inorganic phosphate homeostasis in sodium-dependent phosphate co-transporter Npt2b+/- mice. Am J Physiol Renal Physiol, 301(5), F1105–13. Olgaard, K., Madsen, S., Lund, B., et al. (1980). Pathogenesis of hypophosphatemia in kidney necrograft recipients: a controlled trial. Adv Exp Med Biol, 128, 255–61. Pabico, R. C. and McKenna, B. A. (1988). Metabolic problems in renal transplant patients. Persistent hyperparathyroidism and hypophosphatemia: effects of intravenous calcium infusion. Transplant Proc, 20(1 Suppl 1), 438–42. Pálmadóttir, V. K., Gudmundsson, H., Hardarson, S., et al. (2010). Incidence and Outcome of Acute Phosphate Nephropathy in Iceland. PLoS ONE, 5(10), e13484. Parfitt, A. M., Kleerekoper, M., and Criz, C. (1986). Reduced phosphate reabsorption unrelated to parathyroid hormone after renal transplantation: implications for the pathogenesis of hyperparathyroidism in chronic renal failure. Miner Electrolyte Metab, 12(5–6), 356–62. Petersen, D. J., Boniface, A. M., Schranck, F. W., et al. (1992). X-linked hypophosphatemic rickets: a study (with literature review) of linear growth response to calcitriol and phosphate therapy. J Bone Miner Res, 7(6), 583–97. Pettifor, J. M. (2008). What’s new in hypophosphataemic rickets? Eur J Pediatr, 167(5), 493–9. Pflanz, S., Henderson, I. S., McElduff, N., et al. (1994). Calcium acetate versus calcium carbonate as phosphate-binding agents in chronic haemodialysis. Nephrol Dial Transplant, 9(8), 1121–4. Prie, D., Huart, V., Bakouh, N., et al. (2002). Nephrolithiasis and osteoporosis associated with hypophosphatemia caused by mutations in

hypo-/hyperphosphataemia

the type 2a sodium-phosphate cotransporter. N Engl J Med, 347(13), 983–91. Raanani, P., Berkowicz, M., Harden, I., et al. (1995). Severe hypophosphataemia in autograft recipients during accelerated leucocyte recovery. Br J Haematol, 91(4), 1031–2. Rosenbaum, R. W., Hruska, K. A., Korkor, A., et al. (1981). Decreased phosphate reabsorption after renal transplantation: evidence for a mechanism independent of calcium and parathyroid hormone. Kidney Int, 19(4), 568–78. Rosental, R., Babarykin, D., Fomina, O., et al. (1982). [Hypophosphatemia after successful transplantation of the kidney. Clinico-experimental study]. Z Urol Nephrol, 75(6), 393–9. Rowe, P. S. (2004). The wrickkened pathways of FGF23, MEPE and PHEX. Crit Rev Oral Biol Med, 15(5), 264–81. Sakhaee, K., Brinker, K., Helderman, J. H., et al. (1985). Disturbances in mineral metabolism after successful renal transplantation. Miner Electrolyte Metab, 11(3), 167–72. Salusky, I. B., Foley, J., Nelson, P., et al. (1991). Aluminum accumulation during treatment with aluminum hydroxide and dialysis in children and young adults with chronic renal disease. N Engl J Med, 324(8), 527–31. Sanchez Fructuoso, A. I., Maestro, M. L., Calvo, N., et al. (2012). Role of fibroblast growth factor 23 (FGF23) in the metabolism of phosphorus and calcium immediately after kidney transplantation. Transplant Proc, 44(9), 2551–4. Sanchez Fructuoso, A. I., Maestro, M. L., Pérez-Flores, I., et al. (2012). Serum level of fibroblast growth factor 23 in maintenance renal transplant patients. Nephrol Dial Transplant, 27(11), 4227–35. Shalhoub, V., Shatzen, E. M., Ward, S. C., et al. (2012). FGF23 neutralization improves chronic kidney disease-associated hyperparathyroidism yet increases mortality. J Clin Invest, 122(7), 2543–53. Shenolikar, S., Voltz, J. W., Minkoff, C. M., et al. (2002). Targeted disruption of the mouse NHERF-1 gene promotes internalization of proximal tubule sodium-phosphate cotransporter type IIa and renal phosphate wasting. Proc Natl Acad Sci U S A, 99(17), 11470–5. Shiber, J. R. and Mattu, A. (2002). Serum phosphate abnormalities in the emergency department. J Emerg Med, 23(4), 395–400. Shimada, T., Hasegawa, H., Yamazaki, Y., et al. (2004). FGF-23 is a potent regulator of vitamin D metabolism and phosphate homeostasis. J Bone Miner Res, 19(3), 429–35. Shimada, T., Mizutani, S., Muto, T., et al. (2001). Cloning and characterization of FGF23 as a causative factor of tumor-induced osteomalacia. Proc Natl Acad Sci U S A, 98(11), 6500–5. Sitara, D., Razzaque, M. S., Hesse, M., et al. (2004). Homozygous ablation of fibroblast growth factor-23 results in hyperphosphatemia and impaired skeletogenesis, and reverses hypophosphatemia in Phex-deficient mice. Matrix Biol, 23(7), 421–32. Steiner, R. W., Ziegler, M., and Dialysis Clinical Outcomes Revisited Investigators. (1993). Effect of daily oral vitamin D and calcium therapy, hypophosphatemia, and endogenous 1-25 dihydroxycholecalciferol on parathyroid hormone and phosphate wasting in renal transplant recipients. Transplantation, 56(4), 843–6. Stowell, K. M. (2008). Malignant hyperthermia: a pharmacogenetic disorder. Pharmacogenomics, 9(11), 1657–72. Suki, W. N., Zabaneh, R., Cangiano, J. L., et al. (2007). Effects of sevelamer and calcium-based phosphate binders on mortality in hemodialysis patients. Kidney Int, 72(9), 1130–7. Sun, Y., Wang, O., Xia, W., et al. (2012). FGF23 analysis of a Chinese family with autosomal dominant hypophosphatemic rickets. J Bone Miner Metab, 30(1), 78–84. Tieder, M., Arie, R., Bab, I., et al. (1992). A new kindred with hereditary hypophosphatemic rickets with hypercalciuria: implications for correct diagnosis and treatment. Nephron, 62(2), 176–81. Tieder, M., Modai, D., Samuel, R., et al. (1985). Hereditary hypophosphatemic rickets with hypercalciuria. N Engl J Med, 312(10), 611–17.

395

396

Section 2  

fluid, electrolyte, and renal tubular disorders

Tonelli, M., Pannu, N., and Manns, B. (2010). Oral phosphate binders in patients with kidney failure. N Engl J Med, 362(14), 1312–24. Tonelli, M., Wiebe, N., Culleton, B., et al. (2007). Systematic review of the clinical efficacy and safety of sevelamer in dialysis patients. Nephrol Dial Transplant, 22(10), 2856–66. Tournis, S., Georgoulas, T., Zafeiris, C., et al. (2011). Tertiary hyperparathyroidism in a patient with X-linked hypophosphatemic rickets. J Musculoskelet Neuronal Interact, 11(3), 266–9. Tuttle, K. R. and Short, R. A. (2009). Longitudinal relationships among coronary artery calcification, serum phosphorus, and kidney function. Clin J Am Soc Nephrol, 4(12), 1968–73. Verge, C. F., Lam, A., Simpson, J. M., et al. (1991). Effects of therapy in X-linked hypophosphatemic rickets. N Engl J Med, 325(26), 1843–8. Villa-Bellosta, R., Barac-Nieto, M., Breusegem, S. Y., et al. (2008). Interactions of the growth-related, type IIc renal sodium/phosphate cotransporter with PDZ proteins. Kidney Int, 73(4), 456–64.

Walton, R. J. and Bijvoet, O. L. (1975). Nomogram for derivation of renal threshold phosphate concentration. Lancet, 2(7929), 309–10. Ward, H. N., Pabico, R. C., McKenna, B. A., et al. (1977). The renal handling of phosphate by renal transplant patients: correlation with serum parathyroid hormone (SPTH), cyclic 3’,5’-adenosine monophosphate (cAMP) urinary excretion, and allograft function. Adv Exp Med Biol, 81, 173–81. Wen, H. Y., Schumacher, H. R., Jr, and Zhang, L. Y. (2010). Parathyroid disease. Rheumat Dis Clin North Am, 36(4), 647–64. White, K. E., Jonsson, K. B., Carn, G., et al. (2001). The autosomal dominant hypophosphatemic rickets (ADHR) gene is a secreted polypeptide overexpressed by tumors that cause phosphate wasting. J Clin Endocrinol Metab, 86(2), 497–500. Yamazaki, Y., Tamada, T., Kasai, N., et al. (2008). Anti-FGF23 neutralizing antibodies show the physiological role and structural features of FGF23. J Bone Miner Res, 23(9), 1509–18.

CHAPTER 40

Approach to the patient with hypomagnesaemia Martin Konrad and Karl P. Schlingmann Introduction Magnesium plays an important role in many different cellular processes. Magnesium homeostasis depends on the balanced regulation of intestinal absorption and renal excretion. Renal magnesium handling is a pure filtration-reabsorption process, as there is little evidence of tubular secretion (Fig. 40.1). About 80% of the total plasma magnesium (0.65–1.2 mmol/L) is filtered through the glomeruli; of this amount, 15–20% is reabsorbed by the proximal tubule. The thick ascending limb of Henle’s loop (TALH) plays a major role in reclaiming filtered magnesium (55–70%), whereas only 5–10% is reabsorbed in the distal convoluted tubule (DCT). The DCT mediates the selective regulation of magnesium reabsorption and plays an important role in determining the final urinary excretion (Quamme, 1997). Only 3–5% of the filtered load normally appears in the urine.

Physiology of renal tubular magnesium reabsorption Proximal tubule In adults, the proximal tubular magnesium reabsorption rate (15–20%) is considerably less than the fractional reabsorption of sodium and calcium (de Rouffignac and Quamme, 1994). In contrast, at neonatal age, the proximal tubule reabsorbs about 70% of the filtered magnesium, which is the same as the fractional reabsorption of sodium and calcium (Lelievre-Pegorier et  al., 1983). This difference clearly indicates that the permeability of the proximal tubule changes during development, so that more magnesium is delivered to the loop of Henle in the adult. This maturation in segmental handling of magnesium should be taken into consideration when renal magnesium handling in very young children is assessed. Different hormones affect magnesium reabsorption in the proximal tubule by influencing salt and water transport (de Rouffignac, 1995). Reabsorption is load-dependent, that is, transport is greater with elevated luminal magnesium concentrations. Extracellular volume expansion or anything that retards salt and water transport, results in diminished fluid absorption and greater magnesium delivery to the loop of Henle and the DCT. The increase in distal delivery is normally reclaimed in these nephron segments, but may be large enough to cause an increase in urinary magnesium excretion and hypermagnesiuria (Quamme, 1989).

Loop of Henle The cortical thick ascending limb of the loop of Henle (cTALH) reabsorbs the predominant portion of the filtered magnesium, amounting up to 70%. In this nephron segment, transepithelial magnesium reabsorption is passive moving from lumen to the interstitial space through the paracellular pathway (Fig. 40.2). The driving force for magnesium reabsorption is the positive luminal transepithelial voltage generated by K+ recycling across the apical membrane (Fig. 40.2). Any influence that alters transepithelial voltage or the permeability of the paracellular pathway will alter magnesium reabsorption in the cTALH (de Rouffignac, 1995). The voltage in the TALH depends on apical ROMK potassium channel activity and Na+-K+-2Cl- cotransport that control current flow across the apical membrane. Sodium exits by the basolateral Na+,K+-ATPase and chloride through basolateral ClC-Ka and ClC-Kb members of the chloride channel family (ClC). For proper function, ClC-Ka and ClC-Kb channels require the co-expression of barttin, an essential βsubunit of these channels (Estévez et al., 2001). Changes in their transport rates will affect the transepithelial voltage and thus magnesium absorption. The permeability of the paracellular pathway also plays an important role in determining transepithelial magnesium transport. Paracellular magnesium movement is influenced by electrostatic charges of proteins comprising this route (de Rouffignac and Quamme, 1994). Moreover, there appears to be selectivity of the pathway to divalent cations (Quamme, 1989). Members of the claudin family of tight junction proteins, claudin-16 and claudin-19, have been identified in the TALH and are involved in controlling magnesium and calcium permeability of the paracellular pathway (Simon et al., 1999; Konrad et al., 2006). This notion is supported by the phenotype resulting from mutations in the encoding genes CLDN16 and CLDN19. Affected individuals present with massive calcium and magnesium wasting due to defective reabsorption in the cTALH (Simon et al., 1999; Konrad et al., 2006).

Distal convoluted tubule Despite a net reabsorption of only 5–10% of the filtered magnesium, the DCT plays an essential role in determining the final urinary magnesium excretion, since little or no magnesium is reabsorbed beyond this segment (Quamme, 1997). Magnesium transport within the DCT is transcellular and active in nature (Fig. 40.3). Magnesium enters the cell through selective ion channels (TRPM6) across the apical membrane, driven by the transmembrane negative

fluid, electrolyte, and renal tubular disorders

Distal convoluted tubule 5–10%

15–20 % Proximal tubule

Thick ascending limb

100%

55–70 % Collecting duct

Section 2  

Henle`s loop

398

3–5%

Fig. 40.1  Segmental magnesium reabsorption along the nephron.

electrical potential (Dai et al., 2001). Apical magnesium entry is the rate-limiting step in reabsorption and many of the hormonal and non-hormonal controls act at this site. Cellular magnesium is actively extruded at the basolateral membrane, possibly by a sodium-dependent exchange mechanism, which is still unresolved at the molecular level (de Rouffignac and Quamme, 1994). Magnesium transport in the DCT is dependent on luminal magnesium concentration and apical transmembrane voltage (Dai et al., 2001). (For more details, see Chapter 27.)

Clinical assessment of magnesium deficiency Although magnesium is an abundant cation in the whole body, > 99% is located either intracellularly or in the skeleton. The < 1% of total magnesium present in the body fluids is the most assessable for clinical testing, and the total serum magnesium concentration is the most widely used measure of magnesium status; although its limitations in reflecting magnesium deficiency are well recognized (Elin, 1994). The reference range for normal total serum magnesium concentration is a subject of ongoing debate, but concentrations of 0.7–1.1 mmol/L are widely accepted. Because the measurement of serum magnesium concentration does not necessarily reflect the true total body magnesium content, it has been suggested that measurement of ionized serum magnesium or intracellular magnesium concentrations might provide more precise information on magnesium status. However, the relevance of such measurements to body magnesium stores has been questioned, because the ionized serum magnesium and intracellular magnesium did not correlate with tissue magnesium and the correlation with the results of magnesium retention tests was contradictory (Arnold et  al., 1995; Hébert et al., 1997; Hashimoto et al., 2000). The use of stable magnesium isotopes and muscle 31P-nuclear magnetic resonance spectroscopy represent promising new methods for non-invasive estimation of body and/or tissue magnesium pools. However, they are not particularly suitable for routine measurements.

Hypomagnesaemia develops late in the course of magnesium deficiency and intracellular magnesium depletion may be present despite normal serum magnesium levels. Due to the kidney’s ability to sensitively adapt its magnesium transport rate to imminent deficiency, the urinary magnesium excretion rate is important in the assessment of the magnesium status. In hypomagnesaemic patients, urinary magnesium excretion rates help to distinguish renal magnesium wasting from extra-renal losses. In the presence of hypomagnesaemia, the 24-hour magnesium excretion is expected to decrease below 1 mmol (Sutton and Domrongkitchaiporn, 1993). Magnesium/creatinine ratios and fractional magnesium excretion rates have also been advocated as indicators of evolving magnesium deficiency (Elisaf et al., 1997; Tang et al., 2000). However, the interpretation of these results seems to be limited due to intra- and interindividual variability (Nicoll et al., 1991; Djurhuus et al., 1995). In patients at risk for magnesium deficiency, but with normal serum magnesium levels, the magnesium status can be further evaluated by determining the amount of magnesium excreted in the urine following an intravenous infusion of magnesium. This procedure has been described as ‘parenteral magnesium loading test’ and is still the gold standard for the evaluation of the body magnesium status (Elin, 1994; Hébert et al., 1997). Normal subjects excrete at least 80% of an intravenous magnesium load within 24 hours, whereas patients with magnesium deficiency excrete much less. The magnesium loading test, however, requires normal renal

Thick ascending limb pro-urine

Blood

Mg2+ Ca2+ 3 Na+

NKCC2

2 K+

ATPase

Na+ 2 Cl– K+

CLC-Kb Cl– ROMK

Barttin

K+

Cl– CLC-Ka

Mg2+

Ca2+

Claudin-16/–19

+ 8 mV

Fig. 40.2  Magnesium reabsorption in the cortical thick ascending limb of Henle’s loop (TALH). The membrane proteins influencing magnesium reabsorption are indicated. Magnesium reabsorption is passive and occurs through the paracellular pathway. The driving force is the lumen-positive transcellular voltage which is generated by the transcellular reabsorption of NaCl and the potassium recycling back to the tubular fluid via ROMK.

chapter 40 

Distal convoluted tubule pro-urine

Blood CLC-Kb K+

Cl–

Kir4.1 Cl– Na+

ATPase 2K+

3 Na+

NCCT

+ Kv1.1

γ-subunit

Transcription

K+

Mg2+

?

Na+

HNF1β EGFR

Mg2+ TRPM6

+ EGF

?

CNNM2

approach to the patient with hypomagnesaemia

handling of magnesium. If excess magnesium is being excreted by the kidneys due to diuresis, the magnesium load test may yield an inappropriate negative result. Conversely, if renal function is impaired and less blood is being filtered, this test could give a false-positive result.

Pathophysiology of renal magnesium handling Inherited magnesium-wasting disorders Hereditary hypomagnesaemia comprises a still growing number of rare genetically determined disorders primarily or secondarily affecting renal magnesium handling. In recent years, numerous genetic defects in genes encoding components of the renal tubular salt and electrolyte transport machinery or regulating factors have been described (Table 40.1). The spectrum ranges from the most frequent variant, the Gitelman syndrome with a primary defect in salt reabsorption in the DCT to extremely rare disorders discovered in single patients or families only. In the clinical setting, the assessment of additional biochemical parameters in serum and urine together with extrarenal finding and the mode of inheritance may help to confine the possibly underlying genetic defects (Table 40.2).

Hypomagnesaemia associated with abnormal renal salt handling

Na+

Fig. 40.3  Magnesium reabsorption in the distal convoluted tubule. The key proteins influencing magnesium reabsorption are indicated. Magnesium transport through DCT cells is active and mainly depends on the negative plasma membrane potential because the chemical gradient is very low. The Kv1.1 potassium channel determines the transmembrane voltage that allows magnesium entry through TRPM6. The molecular identity of the basolateral extrusion mechanism is still unknown but this mechanism seems to depend on a sodium gradient, which results from the coordinated action of NCCT, Na+-K+-ATPase and Kir4.1. The paracrine action of EGF regulates apical magnesium transport via TRPM6. HNF1B increases the transcription of the γ subunit of the Na+-K+-ATPase. The exact role of CNNM2 on magnesium reabsorption is still unknown.

Tubular salt reabsorption affects the membrane potential of tubular epithelial cells and is involved in the generation of the transepithelial potential, both of which are a prerequisite for the processes of magnesium reabsorption along the different segments of the nephron. Salt-wasting disorders with hypokalaemia and metabolic alkalosis, also known as Bartter-like syndromes, impair tubular reabsorption of sodium chloride in different parts of the distal nephron. The renal conservation of magnesium is secondarily affected to a varying extent according to the nephron segment affected in each of the Bartter-like syndromes or EAST syndrome (for details see Chapter 31).

Table 40.1  Inherited disorders of magnesium handling Disorder

OMIM #

Inheritance

Gene

Protein

Primary salt-wasting disorders Classic Bartter syndrome Gitelman syndrome EAST/SeSAME syndrome

607364 263800 612780

AR AR AR

CLCNKB SLC12A3 KCNJ10

ClC-Kb, chloride channel NCCT, NaCl cotransporter Kir4.1, potassium channel

Familial hypomagnesaemia with hypercalciuria/ nephrocalcinosis

248250 248190

AR AR

CLDN16 CLDN19

Claudin-16, tight junction Claudin-19, tight junction

Hypomagnesaemia/secondary hypocalcaemia

602014

AR

TRPM6

TRPM6, cation channel

Isolated dominant hypomagnesaemia

154020 176260 613882

AD AD AD

FXYD2 KCNA1 CNNM2

Gamma subunit Na/K/ATPase Kv1.1, potassium channel Cyclin M2

HNF1B nephropathy

137920

AD

HNF1B

HNF1beta, transcription factor

Isolated recessive hypomagnesaemia

611718

AR

EGF

Pro-EGF, epidermal growth factor

Hypomagnesaemia/metabolic syndrome

50005

maternal

MTTI

Mitochondrial tRNA

399

400

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 40.2  Clinical and biochemical characteristics in inherited magnesium disorders Disorder

Serum Mg2+

Serum Ca2+

Serum K+

Blood pH

Urine Mg2+

Urine Ca2+

Other findings

↔↘ ↓ ↓

↓↑ ↔ ↔

↓↓ ↓ ↓

↑ ↑ ↑

↔↗ ↑ ↑

↔ ↓ ↓

Failure to thrive, polyuria Chondrocalcinosis Epilepsy, ataxia, deafness

Familial hypomagnesaemia with hypercalciuria/ nephrocalcinosis

↓

↔

↔

↔

↑↑

↑↑

Nephrocalcinosis, renal failure Ocular abnormalities

Hypomagnesaemia/secondary hypocalcaemia

↓↓

↓

↔

↔

↔↗

↔↗

Mental retardation?

Isolated dominant hypomagnesaemia Related to FXYD2 defects Related to KCNA1 defects Related to CNNM2 defects

↓↓ ↓↓ ↓↓

↔ ↔ ↔

↔ ↔ ↔

↔ ? ↔

↔↗ ↔↗ ↔↗

↓ ↔ ?

None Episodic ataxia, myokymia Unknown

HNF1B nephropathy

↓

↔

?

?

↔↗

↔↘

Cystic kidneys, diabetes (MODY5)

Isolated recessive hypomagnesaemia

↓↓

↔

↔

↔

↔↗

↔

Mental retardation

Primary salt-wasting disorders Classic Bartter syndrome Gitelman syndrome EAST/SeSAME syndrome

Familial hypomagnesaemia with hypercalciuria and nephrocalcinosis Familial hypomagnesaemia with hypercalciuria and nephrocalcinosis (FHHNC) is an autosomal recessive disorder caused by mutations in two different members of the claudin family of tight junction proteins, namely claudin-16 and claudin-19 (Simon et al., 1999; Konrad et al., 2006). Since its first description, > 100 different patients have been reported, allowing a comprehensive characterization of the clinical spectrum of this disorder and discrimination from other magnesium-losing tubular diseases (Praga et  al., 1995; Benigno et al., 2000; Weber et al., 2001; Wolf et al., 2002; Godron et al., 2012). Due to excessive renal magnesium and calcium wasting, affected individuals develop the characteristic triad of hypomagnesaemia, hypercalciuria, and nephrocalcinosis that gave the disease its name. The majority of patients present during early childhood with recurrent urinary tract infections, polyuria/polydipsia, nephrolithiasis, and/or failure to thrive. Clinical signs of severe hypomagnesaemia such as seizures and muscular tetany are less common. Additional biochemical abnormalities include elevated serum intact parathyroid hormone (iPTH levels) before the onset of chronic renal failure, incomplete distal tubule acidosis, hypocitraturia, and hyperuricaemia present in most patients (Weber et al., 2000). The prognosis of FHHNC patients is rather poor with a high risk of progressive renal failure early during adolescence. A considerable number of patients already exhibit a markedly reduced GFR (< 60 mL/min/1.73 m2) at the time of diagnosis. Hypomagnesaemia may completely disappear with the decline of GFR due to reduction in filtered magnesium limiting urinary magnesium excretion. Whereas the renal phenotype is almost identical in carriers of CLDN16 and CLDN19 mutations, ocular involvement, including severe myopia, nystagmus, or macular coloboma, is observed only in patients with CLDN19 mutations (Nicholson et al., 1995; Praga et al., 1995; Konrad et al., 2006; Haisch et al., 2011; Godron et al., 2012).

In addition to oral magnesium supplementation, therapy aims at the reduction of calcium excretion to prevent the progression of nephrocalcinosis and stone formation, because the degree of renal calcification has been correlated with progression of chronic renal failure (Praga et  al., 1995). However, therapeutic strategies do not seem to significantly influence the progression of renal failure. Supportive therapy is important for the protection of kidney function and should include provision of sufficient fluids and effective treatment of stone formation and bacterial colonization. As expected, renal transplantation is performed without evidence of recurrence because the primary defect resides in the kidney. Based on clinical observations and clearance studies, it had been postulated that the primary defect in FHHNC was related to disturbed magnesium and calcium reabsorption in the TALH (Rodríguez-Soriano and Vallo, 1994). In 1999, Simon et al. identified a new gene (CLDN16, formerly PCLN1), which is mutated in patients with FHHNC (Simon et al., 1999) CLDN16 codes for claudin-16, a member of the claudin family, which is important for the formation and function of tight junctions. The individual composition of tight junction strands with different claudins confers the characteristic properties of different epithelia regarding paracellular permeability and/or transepithelial resistance. In this context, a crucial role has been attributed to the first extracellular domain of the claudin proteins, which is extremely variable in number and position of charged amino acid residues (Colegio et al., 2003). Individual charges have been shown to influence paracellular ion selectivity, suggesting that claudins positioned on opposing cells forming the paracellular pathway provide charge-selective pores within the tight junction complex. The majority of mutations reported in FHHNC are simple missense mutations affecting the transmembrane domains and the extracellular loops, with a particular clustering in the first extracellular loop that contains the ion selectivity filter. Within this domain, patients originating from Germany or Eastern European countries exhibit a common mutation (L151F) due to a founder

chapter 40 

effect (Weber et  al., 2001). Defects in CLDN16 have also been shown to underlie the development of a chronic interstitial nephritis in Japanese cattle that rapidly develop chronic renal failure shortly after birth (Ohba et al., 2000). Interestingly, affected animals typically show hypocalcaemia but no hypomagnesaemia, which might be explained by advanced renal failure present at the time of examination. The fact that, in contrast to the point mutations identified in human FHHNC, large deletions of CLDN16 are responsible for the disease in cattle, might explain the more severe phenotype with early-onset renal failure. However, Cldn16 knockout mice do not display renal failure during the first months of life (Will et  al., 2010). In FHHNC patients, progressive renal failure is generally thought to more likely be a consequence of massive urinary calcium wasting and nephrocalcinosis. A  study of a large cohort of FHHNC patients showed that the presence of CLDN16 mutations leading to a complete loss of function on both alleles display a younger age at manifestation, as well as a more rapid decline in renal function compared with patients with at least one allele with residual claudin-16 function (Konrad et al., 2008). These findings support the theory that a complete lack of claudin-16 is associated with a more severe phenotype, whereas some residual function delays the progression of renal failure. There is evidence from family analyses that carriers of heterozygous CLDN16 mutations may also present with clinical symptoms. Two independent studies reported a high incidence of hypercalciuria, nephrolithiasis, and nephrocalcinosis in first-degree relatives of FHHNC patients (Praga et al., 1995; Weber et al., 2001). A subsequent study also found a tendency towards mild hypomagnesaemia in family members with heterozygous CLDN16 mutations (Blanchard et  al., 2001). Thus, one might speculate that CLDN16 mutations are involved in idiopathic hypercalciuric stone formation. In addition to mutations causing FHHNC, a homozygous CLDN16 mutation (T303G) affecting the C-terminal PDZ domain has been identified in two families with isolated hypercalciuria and nephrocalcinosis without disturbances in renal magnesium handling (Müller et  al., 2003). Interestingly, the hypercalciuria disappeared during follow-up and urinary calcium levels reached normal values beyond puberty. Transient transfection of MDCK cells with the CLDN16 (T303G) mutant revealed mistargeting of the mutant claudin-16 to the apical membrane. It still remains to be determined why this type of targeting defect is associated with transient isolated hypercalciuria without increased magnesium excretion. Molecular genetic studies FHHNC patients with severe ocular involvement resulted in identification of mutations in a second member of the claudin family, claudin-19 (encoded by CLDN19) (Konrad et  al., 2008). Claudin-19 is expressed together with claudin-16, predominantly in the TALH. Tight-junction strands in this part of the renal tubule also express claudin-10 and claudin-18 (Hou et al., 2009). These other claudins are able to maintain the barrier function of the tight junction complex in the absence of claudin-16 and -19; however, claudin-16 and -19-depleted tight junctions displayed a loss in cation permselectivity (Hou et  al., 2009). Unfortunately, it remains an unanswered question whether claudin-16 and -19 directly take part in the formation of a paracellular pore structure selective for magnesium and calcium, or if they are simply involved in generating the cation selectivity of the tight junction complex required for maintaining the lumen-positive

approach to the patient with hypomagnesaemia

potential difference in the TALH (Hou and Goodenough, 2010). In this context, it is interesting to note that claudin-16 and claudin-19 deficient mice also display increased renal losses of sodium, as well as of potassium, in addition to the disturbance in renal magnesium and calcium handling (Hou et al., 2009). Patients with claudin-19 defects exhibit a renal phenotype indistinguishable from patients with defective claudin-16 function (Konrad et al., 2008). However, the ocular phenotype observed in patients with claudin-19 defects is much more severe. The molecular basis for this phenomenon is the expression of claudin-19 in different layers of the retina (Konrad et al., 2008). Patients show abnormal development of the optic disc, leading to severe visual impairment and the development of horizontal nystagmus.

Hypomagnesaemia with secondary hypocalcaemia Hypomagnesaemia with secondary hypocalcaemia (HSH) is an autosomal recessive disorder caused by mutations in the TRPM6 gene coding for TRPM6, a member of the transient receptor potential cation channel family (Schlingmann et  al., 2002; Walder et al., 2002). Since its first description in 1968 (Paunier et  al., 1968), at least 50 HSH kindreds have been described (Walder et  al., 1997, 2002; Schlingmann et  al., 2005; Jalkanen et  al., 2006). Almost all patients present in early infancy with generalized seizures refractory to anticonvulsant treatment or other symptoms of increased neuromuscular excitability such as muscle spasms or tetany. Laboratory evaluation at initial presentation reveals dramatically reduced serum magnesium levels of around 0.2  mmol/L. Hypomagnesaemia is accompanied by hypoparathyroidism with barely detectable iPTH levels, and consecutive hypocalcaemia with total serum calcium levels around 1.6  mmol/L. The unexpected finding of hypoparathyroidism is thought to result from an inhibition of iPTH synthesis and secretion in the presence of extreme hypomagnesaemia (Anast et al., 1972; Cole and Quamme, 2000; Vetter and Lohse, 2002). In addition, iPTH-induced release of calcium from bone is substantially impaired in hypomagnesaemia, because magnesium depletion interferes with the generation of cAMP in response to iPTH (Cole and Quamme, 2000). The hypocalcaemia is resistant to treatment with calcium or vitamin D. Treatment of HSH consists of immediate administration of magnesium (equivalent to 25–50 mg magnesium sulfate) per kilogram body weight, up to a maximum of 8 mmol magnesium (equivalent to 2 g magnesium sulfate) (Schlingmann et al., 2004, 2005). Administration of magnesium alone rapidly leads to relief of clinical symptoms, normocalcaemia, and normalization of iPTH levels. Acute parenteral therapy is followed by lifelong high-dose oral magnesium supplementation (Shalev et al., 1998). In the majority of patients, organic magnesium salts such as aspartate or citrate are used. Daily requirements of up to 4 mmol/kg of body weight per day (16 times the recommended daily allowance) have been described (Cole et al., 2000), although an average daily requirement of around 1 mmol/kg of body weight per day seems to be sufficient in most HSH patients. Adolescent patients usually tolerate oral magnesium to a lesser extent than infants and younger children, who on average receive larger amounts per kilogram of body weight. While gastrointestinal complaints are clearly aggravated with increasing amounts of oral magnesium in the same patient, the susceptibility to diarrhoea shows marked inter-individual variability.

401

402

Section 2  

fluid, electrolyte, and renal tubular disorders

The laxative effect of magnesium salts can result in pronounced gastrointestinal side effects that occasionally necessitate considering alternative routes of administration such as intravenous or subcutaneous injections. Splitting of oral doses can reduce fluctuations of serum magnesium levels and peak urinary excretion and also alleviate diarrhoea. Additional intramuscular magnesium injections might be necessary to reduce oral intake. A regimen consisting of daily intramuscular injections given over a 20-year period has been described (Cole et al., 2000). Ultimately, the authors used continuous nocturnal administration via nasogastric tube as a therapeutic alternative to improve quality of life. In another HSH patient, hypomagnesaemic seizures only ceased after implantation of a subcutaneous pump system that provided continuous magnesium infusion (Aries et al., 2000). Delay in diagnosis may lead to neurological deficits or may even be fatal, since seizures are refractory to anticonvulsive treatment. Several HSH patients with severe mental retardation after long-lasting seizures have been reported. In contrast to all other known forms of hereditary hypomagnesaemia, pathophysiologic studies in affected patients using radioactive magnesium isotopes pointed to a primary defect in intestinal magnesium absorption (Lombeck et al., 1975). The presence of an additional renal magnesium loss in HSH was controversial until magnesium-loading studies clearly demonstrated a renal magnesium leak (Matzkin et  al., 1989). With rising serum magnesium levels during substitution, renal magnesium loss, which is barely detectable at initial presentation, becomes evident, demonstrating a decreased renal threshold for magnesium (Walder et al., 2002). Despite remaining hypomagnesaemic with serum levels around 0.6 mmol/L, HSH patients display inappropriately high fractional excretion of magnesium in the range of 2–4% (Schlingmann et al., 2002); assuming intact renal magnesium conservation, fractional excretions rate would be expected to drop below 1% in the presence of hypomagnesaemia (Ahmad and Swaminathan, 2000). A positional candidate gene approach allowed the identification of mutations in the TRPM6 gene as the underlying defect in HSH (Schlingmann et al., 2002; Walder et al., 2002). TRPM6 codes for a member of the transient receptor potential (TRP) family of cation channels. All TRP channels share the common feature of six transmembrane domains with a putative pore-forming region between the fifth and sixth transmembrane domain and intracellular N- and C-termini. Four TRP protein subunits assemble to form a functional channel complex. The TRPM subfamily comprises eight members that exhibit a significant diversity in domain structure, as well as cation selectivity and activation mechanisms. Three members—TRPM2, TRPM6, and TRPM7—are set apart from all other known ion channels because they harbour enzyme domains in their respective C termini and represent prototypes of an intriguing new family of enzyme-coupled ion channels. TRPM6 and TRPM7 contain protein kinase domains with sequence similarity to elongation factor 2 (eEF-2) serine/threonine kinases (Dorovkov and Ryazanov, 2004; Ryazanova et al., 2004). The functional characterization of TRPM7 demonstrated permeability for various cations, including calcium and magnesium (Monteilh-Zoller et  al., 2003). TRPM7 gating was shown to be regulated by intracellular magnesium and magnesium-nucleotide complexes (Nadler et al., 2001). Targeted deletion of TRPM7 in cell lines results in intracellular magnesium depletion and growth arrest

(Schmitz et al., 2003). These data point to the essential role of ubiquitously expressed TRPM7 for cellular magnesium homeostasis. TRPM6 is highly homologous to the TRPM7 ion channel. To date, only one group has succeeded in the functional expression of TRPM6 in a mammalian cell line (Voets et al., 2004). The authors were able to show channel properties similar to those observed for TRPM7. In contrast, another group demonstrated that heteromultimerization with TRPM7 is essential for correct membrane targeting of TRPM6 (Chubanov et al., 2004). In this study, TRPM7-induced currents were significantly increased by co-expression of TRPM6. Heteromultimerization has been described previously for other members of the TRP family (Hofmann et al., 2002). The detection of TRPM6 expression in the DCT, together with functional studies in HSH patients who clearly demonstrated a renal magnesium leak, points to an important role of TRPM6 for active transcellular magnesium reabsorption in the DCT (Schlingmann et al., 2002; Voets et al., 2004). Whether TRPM6 alone or in cooperation with TRPM7 constitutes the apical magnesium channel in DCT cells remains to be clarified in future studies. TRPM6 mutations in HSH patients comprise the following classes of mutations:  stop mutations and frame shift mutations, both leading to premature stops of translation, as well as splice site mutations, which impede proper mRNA synthesis and presumably lead to absence of the corresponding exon and large exon deletions (Schlingmann et al., 2002; Walder et al., 2002). Thus far most mutations described in HSH result in truncated TRPM6 proteins. Only a few missense mutations have been identified (Schlingmann et al., 2002; Jalkanen et al., 2005; Chubanov et al., 2007). Functional characterization of these mutations revealed a complete loss of function of the TRPM6 protein (Chubanov et al., 2004, 2007). Obviously, complete lack of TRPM6 ion channel function is required for the development of the typical HSH phenotype. It is intriguing to speculate whether minor changes in TRPM6 function by single point mutations might result in a less severe clinical picture or even in subclinical magnesium deficiency. How does an impairment of TRPM6 protein function impede epithelial magnesium transport in the intestine and kidney? The observation that in HSH patients the administration of high oral doses of magnesium are successful in achieving at least subnormal serum magnesium levels supports the existing evidence of two independent transport systems for magnesium in the gastrointestinal tract (Schweigel and Martens, 2000). TRPM6 probably represents a molecular component of active transcellular magnesium transport. An increased intraluminal magnesium concentration achieved by increased oral intake would allow compensation for the defect of the active transcellular pathway by increasing absorption via the passive paracellular route (Kerstan and Quamme, 2002; Schlingmann et al., 2002).

Isolated dominant hypomagnesaemia FXYD2 (γ subunit) Isolated dominant hypomagnesaemia (IDH) was first linked to a mutation in the FXYD2 gene on chromosome 11q23 which codes for a γ subunit of the Na+,K+-ATPase (Meij et al., 2000). Only two related families with a FXYD2 mutation have been described so far (Geven et al., 1987b; Meij et al., 2000). The index patients of both families presented with generalized seizures at ages 7 and 13 years, respectively. Serum magnesium levels in the two patients at that time were around 0.4 mmol/L. One index patient was treated for

chapter 40 

seizures of unknown origin with antiepileptic drugs until serum magnesium levels were evaluated adolescence. At that time severe mental retardation was evident. Serum magnesium measurements performed in members of both families revealed low serum magnesium levels around 0.5  mmol/L in several apparently healthy individuals. Detailed haplotype analyses identified a common haplotype segregating in the two families, which suggested a common ancestor. Indeed, the mutational screening of the FXYD2 gene demonstrated the identical mutation G41R in all affected individuals of both family branches (Meij et al., 2000). A 28Mg-retention study in one index patient pointed to a primary renal defect (Geven et al., 1987b). The intestinal absorption of magnesium was preserved and even stimulated in compensation for the increased renal losses. Urinary magnesium measurements in affected family members revealed daily magnesium excretions of around 5  mmol per day despite profound hypomagnesaemia (Geven et al., 1987b). In addition, urinary calcium excretions were found to be low in all hypomagnesaemic family members, a finding reminiscent of patients presenting with Gitelman syndrome; however, in contrast to Gitelman syndrome, no other associated biochemical findings were reported. The γ subunit encoded by FXYD2 is a member of a family of small single transmembrane proteins that share the common amino acid motif F-X-Y-D. It comprises two isoforms (γ-a and γ-b) that are differentially expressed in the kidney. The γ-a isoform is present predominantly in the proximal tubule, whereas expression of the γ-b isoform predominates in the distal nephron, especially in the DCT and connecting tubule (Arystarkhova et al., 2002b). The ubiquitous Na+,K+-ATPase is a dimeric enzyme invariably consisting of one α and one β subunit. FXYD proteins constitute a third or γ subunit that represents a tissue-specific regulator of Na+,K+-ATPase and produces distinct effects on the affinity of the Na+,K+-ATPase for sodium, potassium, and ATP. The FXYD2 γ subunit increases the apparent affinity of Na+,K+-ATPase for ATP, while decreasing its sodium affinity (Arystarkhova et al., 2002a). Thus, it might provide a mechanism for balancing energy utilization and maintaining appropriate Na and K gradients across the cell membrane. Expression studies of the mutant G41R-γ subunit in mammalian renal tubule cells revealed a dominant negative effect of the mutation leading to retention of the γ subunit within the cell. Whereas initial data pointed to retention of the entire Na+,K+-ATPase complex in intracellular compartments, subsequent data suggest an isolated trafficking defect of the mutant γ subunit (Blostein et al., 2003). The mutant γ subunit is retarded in the Golgi complex, which points to disturbed post-translational processing. The assumption of a dominant negative effect was first substantiated by the observation that individuals with a large heterozygous deletion of chromosome 11q that includes the FXYD2 gene exhibit normal serum magnesium levels (Meij et al., 2003). Meanwhile, it could be shown that wild type γ subunits oligomerize within the cell before trafficking to the plasma membrane. The G41R-mutant was shown to also oligomerize with itself and the wild type γ subunit and thus prevent proper routing of wild type γ subunits to the plasma membrane and incorporation into functional ATPase complexes (Cairo et al., 2008). Urinary magnesium wasting together with the expression pattern of the FXYD2 gene indicate a defect of active transcellular magnesium reabsorption in the DCT in affected individuals. Meij et al. have suggested that diminished intracellular potassium might depolarize the apical membrane, resulting in a decrease in

approach to the patient with hypomagnesaemia

magnesium uptake (Meij et al., 2000). Alternatively, an increase in intracellular sodium could impair basolateral magnesium transport, which is presumably achieved by a sodium-coupled exchange mechanism. Another explanation could be that the γ subunit is involved not only in Na+,K+-ATPase function, but also is an essential component of a yet unidentified ATP-dependent transport system specific for magnesium. An interesting finding in IDH linked to the FXYD2-G41R mutation is hypocalciuria. In patients with the FXYD2 mutant, there is no evidence for renal salt wasting and no stimulation of the renin–angiotensin–aldosterone system (RAAS). Therefore, the mechanism leading to hypocalciuria in IDH related to FXYD2 mutations remains unclear. One could speculate that, similar to Gitelman syndrome, a defect in Na+,K+-ATPase function and energy metabolism might lead to a cell apoptosis in the early DCT responsible for magnesium reabsorption, while later parts of the distal nephron remain intact.

KCNA1 Genetic heterogeneity in isolated dominant hypomagnesaemia was demonstrated by the identification of a dominant-negative missense mutation in KCNA1 encoding the voltage-gated potassium channel Kv1.1 (Glaudemans et al., 2009). The phenotype of affected individuals originating from a large Brazilian family included recurrent episodes of muscle cramps, tetanic episodes and tremor, and muscle weakness, starting in infancy. Laboratory analyses revealed isolated hypomagnesaemia < 0.4 mmol/L, while urinary magnesium excretion was found to be inappropriately elevated, pointing to a renal magnesium leak. No alterations in renal calcium handling were observed. Interestingly, KCNA1 mutations had been identified before in patients with episodic ataxia with myokymia (OMIM 160120), a neurologic disorder characterized by periodical appearance of incoordination and imbalance, as well as myokymia, an involuntary, spontaneous, and localized trembling of muscles (Browne et  al., 1994). In addition to muscle cramps and tetany attributed to magnesium deficiency, these symptoms were also present in members of the above mentioned Brazilian kindred with hypomagnesaemia. By using a genome-wide single nucleotide polymorphism-based linkage strategy followed by subsequent conventional sequencing of candidate genes identified a heterozygous mutation in the KCNA1 gene co-segregating with the disease (Glaudemans et al., 2009). The mutation leads to a non-conservative amino acid exchange (N255D) in the encoded Kv1.1 potassium channel. Functional voltage-gated potassium channels of the KCNA family are composed of homoor heterotetramers in assembly with other Kv channel subunits. Co-expression of the mutant N255D-Kv1.1 with wild type channel subunits in HEK293 cells resulted in a significant reduction in current amplitudes, compatible with a dominant-negative effect of the mutant. The dominant-negative effect seems to be the result of an impaired gating of the potassium channel tetramer, since trafficking to the plasma membrane is preserved (van der Wijst et al., 2010). Kv1.1 expression was demonstrated in the kidney to be co-localized with TRPM6 at the apical membrane of the DCT. Glaudemans et al. propose a model in which Kv1.1 permits hyperpolarization of the DCT apical cell membrane potential as a prerequisite for TRPM6-mediated magnesium entry (Fig. 40.3). Thus these authors, for the first time, linked magnesium reabsorption in the DCT to potassium secretion, and identified a new dependency

403

404

Section 2  

fluid, electrolyte, and renal tubular disorders

between renal magnesium and potassium handling at the molecular level (Glaudemans et al., 2009).

it remains a subject of further study as to how exactly CNNM2 is involved in basolateral magnesium transport processes.

CNNM2

Hepatocyte nuclear factor 1B nephropathy

Another form of isolated dominant hypomagnesaemia was recently described in which mutations in CNNM2 were discovered (Stuiver et al., 2011). CNNM2 (or ACDP2) had been identified before by differential gene expression microarray analysis of genes upregulated in face of hypomagnesaemia in mice (Goytain and Quamme, 2005). The authors had demonstrated that the CNNM2 protein is able to induce the transport of different divalent cations, including magnesium (Goytain and Quamme, 2005). Furthermore, the CNNM2 gene locus had been linked to serum magnesium levels in a genome-wide association study (Meyer et al., 2010). Stuiver and colleagues screened CNNM2 as a candidate gene in patients with unresolved magnesium wasting disorders and identified heterozygous mutations in two unrelated families with isolated dominant hypomagnesaemia (Stuiver et al., 2011). In affected individuals, clinical symptoms and age at manifestation seem to be highly variable with symptoms ranging from convulsive episodes in early childhood to muscle weakness, vertigo and headaches during adolescence. Other heterozygous carriers from both families remained asymptomatic. Serum magnesium levels in affected individuals were in the range of 0.4–0.5 mmol/L and failed to normalize under oral magnesium supplementation. All other serum electrolytes were found to be normal. Since data on urinary calcium excretions vary between the described index patients, it remains unclear if the finding of hypocalciuria seen in a number of other inherited magnesium wasting disorders (see above) is also a concomitant feature in patients with CNNM2 mutations. The CNNM2 gene codes for CNNM2 or Cyclin M2, a transmembrane protein that is expressed in kidney at the basolateral membrane of TALH and DCT, but also in other organs especially the brain (Stuiver et al., 2011). CNNM2 is one of four members of a protein family with a conserved domain structure (for which this family was previously named ancient conserved domain proteins or ACDP family) with sequence similarities to the bacterial CorC protein involved in bacterial magnesium transport (Gibson et al., 1991). Whereas a truncating frame-shift mutation was identified in one of the described families, affected individuals of the second family carry a missense mutation, leading to a non-conservative amino acid exchange of the CNNM2 protein (T568I) (Stuiver et al., 2011). Functional characterization of the mutant T568I-CNNM2 demonstrated that protein trafficking was preserved in HEK293 cells; however, patch clamp analyses revealed a significant reduction in magnesium-sensitive, inwardly-rectifying sodium currents. In contrast to previous experiments in Xenopus oocytes by Goytain and Quamme (2005), the authors did not observe significant magnesium currents with overexpression of CNNM2 in HEK293 cells (Stuiver et al., 2011). The localization data, together with the functional studies, lead to the assumption that CNNM2 might represent a basolateral magnesium sensing mechanism in renal tubular cells, rather than being a molecular component of the yet uncharacterized basolateral magnesium extrusion machinery itself. Since Stuiver and colleagues identified a truncating mutation, as well as a missense mutation in their families they only speculate on a reduced amount of functional protein as a putative mechanism for the dominant mode of inheritance (Stuiver et al., 2011). Therefore,

Hepatocyte nuclear factor 1B (HNF1B) is a transcription factor critical for the development of kidney and pancreas. Heterozygous mutations in HNF1B were first been implicated in a subtype of maturity-onset diabetes of the young (MODY5) before an association with developmental renal disease was reported. The renal phenotype is highly variable, including enlarged hyperechogenic kidneys, multicystic kidney disease, renal agenesis, renal hypoplasia, cystic dysplasia, as well as hyperuricaemic nephropathy. The association of diabetes with renal cysts leads to the term ‘renal cysts and diabetes (RCAD) syndrome’. HNF1B mutations are present in heterozygous state, either inherited or de novo, and comprise point mutations, as well as whole-gene deletions (Heidet et  al., 2010). Interestingly, around 50% of affected individuals present with hypomagnesaemia due to renal magnesium wasting (Heidet et al., 2010; Adalat et al., 2009). The degree of hypomagnesaemia is usually mild to moderate (~ 0.65  mmol/L) and the defect in renal magnesium conservation is accompanied by the occurrence of hypocalciuria. The HNF1B gene encodes a transcription factor regulating the expression of numerous renal genes, including the FXYD2 gene which contains several HNF1B -binding sites in its promoter region (Adalat et al., 2009). In accordance with the phenotype and in silico data, Adalat and colleagues could show that HNF1B was able to stimulate the expression of FXYD2 in vitro. Therefore, defective FXYD2 transcription represents a putative mechanism explaining renal magnesium wasting in patients with HNF1B mutations.

Isolated recessive hypomagnesaemia Geven and colleagues initially reported a form of isolated hypomagnesaemia in a consanguineous family, indicating autosomal recessive inheritance (Geven et al., 1987a). Two affected girls presented with generalized convulsions during infancy. Unfortunately, late diagnosis resulted in neurodevelopmental deficits in both patients. A  thorough clinical and laboratory workup at 4 and 8  years of age, respectively, revealed serum magnesium levels of 0.5–0.6 mmol/L with no other associated electrolyte abnormalities. A  28Mg-retention study in one patient pointed to a primary renal defect, while intestinal magnesium uptake was preserved (Geven et al., 1987a). Both patients exhibited renal magnesium excretions of 3–6 mmol per day, despite hypomagnesaemia, which confirmed renal magnesium wasting, whereas calcium excretion rates were in the normal range. Using a homozygosity mapping strategy, Groenestege and colleagues identified a candidate interval on chromosome 4q (Groenestege et al., 2007). Subsequent screening of candidate genes within this region resulted in the identification of a homozygous missense mutation in the EGF gene, leading to a non-conservative amino acid exchange in the encoded pro-EGF protein (pro-epidermal growth factor) in the two sisters (Groenestege et  al., 2007). Pro-EGF is a small peptide hormone expressed in various tissues, including the kidney (predominantly in the DCT). Pro-EGF is a membrane protein that is inserted in both the luminal and basolateral membrane of polarized epithelia. After membrane insertion, it is processed by unknown proteases into active EGF peptide. EGF activates specialized EGF receptors (EGFRs),

chapter 40 

which are expressed in the basolateral membrane. This activation leads to an increase in TRPM6 trafficking to the luminal membrane and increased magnesium reabsorption (Thebault et al., 2009). The mutation described in IRH (P1070L) disrupts the basolateral sorting motif in pro-EGF (Groenestege et  al., 2007). Therefore, the activation of EGFRs in the basolateral membrane is compromised, which finally leads to reduced magnesium reabsorption. Despite acting in a paracrine fashion in the DCT, the authors speculate on a role for EGF as a selectively acting magnesiotropic hormone (Groenestege et al., 2007).

Mitochondrial hypomagnesaemia In 2004, a mutation in the mitochondrial isoleucine tRNA gene, tRNAIle, or MTTI, was discovered in a large Caucasian kindred (Wilson et  al., 2004). Extensive clinical evaluation of this family was initiated after the discovery of hypomagnesaemia in the index patient. Pedigree analysis was compatible with mitochondrial inheritance as the phenotype was exclusively transmitted by affected females. The phenotype includes hypomagnesaemia, hypercholesterolaemia, and hypertension. Among the adults on the maternal lineage, the majority of offspring exhibit at least one of the mentioned symptoms; approximately half of the individuals show a combination of two or more symptoms, and around one-sixth had all three features. Serum magnesium levels of family members in the maternal lineage greatly vary, ranging from 0.3 to 1.0 mmol/L, with approximately 50% of individuals being hypomagnesaemic (< 0.75 mmol/L). Hypomagnesaemic individuals showed higher fractional excretions rates (median around 7.5%) than their normomagnesaemic relatives in the maternal lineage (median around 3%), which clearly pointed to renal magnesium wasting. Hypomagnesaemia was accompanied by decreased urinary calcium excretion, a finding pointing to the DCT as the affected tubular segment. The mitochondrial mutation observed in the examined family affects the tRNAIle gene MTTI (Wilson et al., 2004). The observed nucleotide exchange occurs at the T nucleotide directly adjacent to the anticodon triplet. This position is highly conserved among species and critical for codon-anticodon recognition. The functional consequences of the tRNA defect for mitochondrial function remain to be elucidated in detail. As ATP consumption along the tubule is highest in the DCT, the authors speculate on an impaired energy metabolism of DCT cells as a consequence of the mitochondrial defect, which in turn could lead to disturbed transcellular magnesium reabsorption.

Acquired renal magnesium wasting disorders Cisplatin and carboplatin The cytostatic agent cisplatin and the newer antineoplastic drug, carboplatin, are widely used in various protocols for the therapy of solid tumours (Lajer and Daugaard, 1999). Among diverse side effects, nephrotoxicity receives most attention as the major dose-limiting factor (Yao et  al., 2007). Carboplatin has been reported to have less severe side effects than cisplatin (English et al., 1999; Boulikas and Vougiouka, 2004; Carrick et al., 2004). Hypomagnesaemia following renal magnesium wasting is regularly observed in patients treated with cisplatin (Lajer and Daugaard, 1999; Goren, 2003; Hodgkinson et  al., 2006). The incidence of

approach to the patient with hypomagnesaemia

magnesium deficiency is > 30%, but increases to > 70% have been observed with longer cisplatin usage and greater accumulated doses (Hodgkinson et al., 2006). Cisplatin exerts direct damage renal tubular cells and causes proximal, as well as distal, tubular dysfunction (Yao et al., 2007). Acute cisplatin toxicity involves defective mitochondrial function, decreased ATPase activity, altered cell cation content, and altered solute transport. In the distal tubule, the reabsorption of salt and particularly of magnesium are compromised. In rodents, cisplatin leads to an impairment of proximal tubular function, whereas in dogs and humans morphological changes appear predominantly in the DCT and the collecting duct (Mavichak et al., 1985; Swainson et al., 1985; Magil et al., 1986). Thus, for cisplatin toxicity the interpolation from animal studies may be misleading, but the evidence from both clinical and experimental studies indicate that the drug acts on distal tubular magnesium transport. Since there is no magnesium reabsorption within the cortical collecting ducts, it is likely that actions within the DCT are responsible for the renal magnesium leak. Using micropuncture, Mavichak et  al. showed that magnesium reabsorption was diminished in the distal tubule of rats receiving cisplatin (Mavichak et al., 1985). The molecular mechanisms for the acute, apparently selective effects on magnesium remain undefined. It would be of interest to determine if amiloride retains its magnesium-conserving actions in these patients (Quamme, 1997). The effects of cisplatin may persist for months or years later—long after the inorganic platinum has disappeared from the renal tissue (Bianchetti et  al., 1991; Markmann et  al., 1991). Whatever cellular mechanisms are involved they must include genetic alterations of magnesium transport. The fact that cisplatin exerts its cytotoxicity binding cellular DNA may be relevant (Lajer and Daugaard, 1999; Boulikas and Vougiouka, 2004). The chronic alterations in distal tubular ion transport comprise a clinical picture reminiscent of Gitelman syndrome (Panichpisal et al., 2006). Morphological studies in humans with cisplatin nephrotoxicity could demonstrate focal tubular necrosis predominantly in the DCT, a feature also observed in NCCT-deficient mice, suggestive of DCT cell apoptosis (Arany and Safirstein, 2003; Loffing et al., 2004). Therefore, fluid repletion, and especially magnesium supplementation with sufficient doses, should be promoted routinely to prevent acute and chronic cisplatin nephrotoxicity, and the risk of adverse effects from hypomagnesaemia.

Aminoglycosides Aminoglycosides, such as gentamicin, induce renal impairment in up to 35% of patients, dependent on the dose and duration of administration. In addition, aminoglycosides cause hypermagnesiuria and hypomagnesaemia (Shah and Kirschenbaum, 1991). As many as 25% of patients receiving gentamicin will exhibit hypomagnesaemia (Shah and Kirschenbaum, 1991). The hypermagnesiuric response occurs soon after the onset of therapy; it is dose dependent and readily reversible on withdrawal. As with adults, neonates also display an immediate increase of calcium and magnesium excretion after gentamicin infusion (Elliott et  al., 2000; Giapros et al., 2004). Magnesium wasting is associated with hypokalaemia and hypercalciuria that may also lead to diminished plasma calcium concentrations (Keating et al., 1977). This would suggest that aminoglycosides affect renal magnesium and calcium transport in the tubular segments where both are reabsorbed. Experimental studies with animals support this notion (Garland et al., 1994). The

405

406

Section 2  

fluid, electrolyte, and renal tubular disorders

cellular mechanisms are not completely understood, but hypermagnesiuria and hypercalciuria are observed in the absence of histopathological changes (Weinberg et al., 1983). Because gentamicin is a polyvalent cation, it was postulated that it may have effects on the CaSR (Kang et al., 2000; Dai et al., 2001; McLarnon et al., 2002; Ward et al., 2005). Activation of this receptor by polyvalent cations inhibits the passive reabsorption of magnesium and calcium in the TALH and active hormone-mediated transport in the DCT, leading to renal magnesium and calcium wasting.

Calcineurin inhibitors The calcineurin inhibitors ciclosporin and tacrolimus (FK506) are widely prescribed as immunosuppressants to organ transplant recipients and in numerous immunological disorders. On this therapy patients are at high risk of developing renal injury and hypertension. Tubular dysfunction with subsequent disturbance of mineral metabolism is another common side effect. Both drugs commonly lead to renal magnesium wasting and hypomagnesaemia (Rob et  al., 1996). These drugs also cause modest hypercalcaemia with hypercalciuria and hypokalaemia (Rob et al., 1996). The hypomagnesaemic effect is probably attenuated by the fall in GFR and reduction in filtered magnesium, but this defect appears to be specific for magnesium (Wong and Dirks, 1988). Calcineurin inhibitor therapy is associated with an inappropriately high fractional excretion rate of magnesium, suggesting impaired passive reabsorption in the TALH or active magnesium transport in the DCT (Lote et al., 2000). Chang and colleagues have reported that ciclosporin reduces claudin-16 expression in the TALH (Chang et  al., 2007). Furthermore, ciclosporin and tacrolimus have been shown to inhibit PTH-stimulated magnesium uptake in a mouse DCT (mDCT) cell line (Kim et al., 2006). Accordingly, Ledeganck and colleagues could show that ciclosporin decreases the renal expression of TRPM6, TRPM7, NCC, and EGF (Ledeganck et al., 2011). In contrast to the study by Chang et  al. (2007), these authors did not find a decrease in claudin-16 or claudin-19 expression in their rat model (Ledeganck et al., 2011). In a previous animal study, Nijenhuis et al. had demonstrated that tacrolimus induces a decrease in TRPV5, calbindin-D28k, and TRPM6 at the mRNA level (Nijenhuis et al., 2004). They could also show a decrease in TRPV5 and calbindin-D28k at the protein level. These effects appeared to be specific, because no morphologic features of tubular toxicity were observed. Finally, Ledeganck and co-workers could demonstrate that the EGF-mediated increase in TRPM6 abundance was abrogated by ciclosporin (Ledeganck et al., 2011). As mentioned above, ciclosporin, beyond its effect on magnesium transport in the DCT, downregulates the sodium chloride cotransporter NCCT in the DCT, resulting in renal salt wasting (Ledeganck et al., 2011). With respect to magnesium, it is interesting to note that TRPM6 expression in the intestine was not changed on tacrolimus administration. It is not known whether these drugs act through calcineurin, which is the intracellular receptor for these agents. It is speculated that FK506-binding proteins, which are known to bind and regulate the calcium-permeable transient receptor potential-like (TRPL) cation channels, might be involved because tacrolimus disrupts this binding (Mervaala et al., 1999). By analogy, consider that certain FKBPs might also regulate TRPV5 or TRPM6 expression or activity. Hypomagnesaemia has been implicated as a contributor to the nephrotoxicity and arterial hypertension associated with

calcineurin inhibitors. Mervaala and colleagues could demonstrate that the adverse effects of ciclosporin in spontaneously hypertensive rats largely depend on dietary sodium and that these adverse effects can be prevented by magnesium supplementation (Mervaala et  al., 1999). Magnesium supplementation also had a beneficial effect on ciclosporin nephrotoxicity in a rat model used by Miura and colleagues (Miura et al., 2002).

EGF receptor antibodies The EGF hormone axis has been implicated in renal magnesium handling by the discovery of a homozygous mutation in the EGF gene in a family with isolated recessive hypomagnesaemia (see above) (Groenestege et al., 2007). The way for these findings was paved by the observation that anticancer treatments with monoclonal antibodies against the EGF receptor (EGFR) resulted in renal magnesium wasting and hypomagnesaemia (Tejpar et  al., 2007). Of note, patients treated with EGFR targeting antibodies (cetuximab, panitumumab) for colorectal cancer usually receive combination therapy with platinum compounds, potentially aggravating the effects on serum magnesium levels. A  significant number of patients receiving such a chemotherapeutic regimen show decreasing serum magnesium concentrations over time (Tejpar et  al., 2007; Cao et  al., 2010). Twenty-four-hour urine collections, as well as magnesium loading tests in single patients, demonstrated a defect in renal magnesium conservation. Together with the genetic findings in isolated recessive hypomagnesaemia due to a pro-EGF mutation, these findings imply a selective effect of EGFR targeting on transcellular magnesium transport in the DCT. There, TRPM6 mediated magnesium uptake into DCT cells is stimulated by basolaterally secreted EGF via its receptor (EGFR) (Groenestege et al., 2007). The initial report of Tejpar and colleagues describes oral as well as intravenous magnesium supplementation in patients with different degrees of hypomagnesaemia after cetuximab treatment, but with limited success in correcting serum magnesium levels (Tejpar et al., 2007). Subsequent studies investigated the development of hypomagnesaemia under EGFR targeting therapy in relation to the antitumour effect and outcome of patients (Vincenzi et al., 2008, 2011). They found that the development of hypomagnesaemia is correlated with the tumour response rate and outcome. Patients with a reduction of serum magnesium > 50% showed a better tumour response rate and a better overall survival (Vincenzi et  al., 2011). The authors, therefore, consider serum magnesium as an easily measurable biomarker for efficacy in colorectal cancer patients treated with cetuximab.

Proton-pump inhibitors Over the last 20 years, proton-pump inhibitors (PPIs) used for the reduction of gastric acidity have emerged as one of the most widely prescribed class of drug worldwide (Cundy and Mackay, 2011). Due to the large number of patients, even rare side effects could become apparent that remain undiscovered during initial clinical trials. Hypomagnesaemia, clinically apparent as muscle cramps, tetany, nausea, vomiting, but also cerebral convulsions, has been observed in a small, but significant number of patients receiving PPIs. It is conceivable that in a substantially larger number of patients, the relationship between low serum magnesium levels and the use of PPIs remains unrecognized. A recent review summarizing the data from previous publications reveals severely lowered serum magnesium levels < 0.4  mmol/L with concomitant hypocalcaemia, a

chapter 40 

laboratory constellation reminiscent of HSH due to TRPM6 defects (Cundy and Mackay, 2011). A previous report on hypomagnesaemia following PPI treatment had already described suppressed parathyroid hormone levels during phases of severe hypomagnesaemia as the probable cause of hypocalcaemia (as in HSH) (Epstein et al., 2006). Although a number of patients also receive diuretics, this finding does not explain the extraordinary degree of magnesium deficiency observed in patients receiving PPIs. What is the mechanism underlying hypomagnesaemia in PPI users? Data regarding urinary magnesium excretions in hypomagnesaemic PPI are contradictory. Fractional magnesium excretions were reported to be low in face of profound hypomagnesaemia, pointing to an intact tubular magnesium reabsorption. However, as observed in HSH patients, a renal magnesium leak might only become apparent if serum magnesium levels reach a certain threshold. An alternative explanation could involve a disturbed intestinal reabsorption of magnesium. Unfortunately, the molecular link between proton pump inhibition and hypomagnesaemia remains unclear so far. In relation to the severe degree of hypomagnesaemia, possible molecular mechanisms include an inhibition of TRPM6, leading to a combined intestinal and renal defect, but also a disturbance of ATPases or ATPase-subunits other than gastric H+-K+-ATPase involved in epithelial magnesium transport. In any case, it is recommended to monitor serum magnesium levels in patients receiving PPIs, particularly those with concomitant cardiac disease and risk for arrhythmia. Moreover, attention should be drawn to the medication list of patients presenting with hypomagnesaemia and secondary hypocalcaemia due to apparent hypoparathyroidism.

Miscellaneous agents A number of antibiotics, anti-tuberculosis therapies, and antiviral drugs may result in renal magnesium wasting (Shah and Kirschenbaum, 1991; Ahmad and Swaminathan, 2000). The cellular basis of this effect on magnesium reabsorption is still largely unknown. Many are associated with general cytotoxicity. Amphotericin B may lead to acquired distal renal tubular acidosis, which in turn can reduce renal magnesium reabsorption (Ahmad and Swaminathan, 2000). Pamidronate used in the treatment of tumour-associated hypercalcaemia has been reported to cause transient hypomagnesaemia (Elisaf et al., 1998). The cellular mechanisms are difficult to predict, since this drug is used in patients with hypercalcaemia, which may aggravate renal magnesium wasting.

Metabolic acidosis It has long been known that systemic acidosis is associated with renal magnesium wasting. Acute metabolic acidosis produced by infusion of NH4Cl or HCl leads to significant increases in urinary magnesium excretion (Quamme, 1997; Ahmad and Swaminathan, 2000). Chronic acidosis also leads to urinary magnesium wasting which, as with the acidosis itself, may be partially corrected by the administration of bicarbonate (Quamme, 1997). In contrast to metabolic acidosis, acute and chronic metabolic alkalosis consistently lead to a fall in urinary magnesium excretion (Quamme, 1997). The cellular basis for the acid–base effects on magnesium transport appears to be diverse. Di Stefano et  al. perfused isolated mouse cTAL segments harvested from mice maintained on alkali in the drinking water (20 mM sodium bicarbonate) for 3 days (Di

approach to the patient with hypomagnesaemia

Stefano et al., 1995). They showed that alkalosis doubled magnesium absorption from control levels without a change in transepithelial voltage. They interpreted these data to indicate that alkalosis changes the permeability of the paracellular pathway, so that magnesium moves passively through the pathway to a greater degree, resulting in more magnesium absorption. The effects of metabolic acidosis were not determined in this study (Di Stefano et al., 1995). Nevertheless, these data suggest that pH influences the paracellular pathway, perhaps through alteration of claudins. Dai et al. have used an mDCT cell line to determine the effects of extracellular pH changes on cellular magnesium uptake (Dai et al., 2001). The results of these experiments showed that acute alkalosis markedly enhances magnesium uptake, whereas acidosis diminishes transport. Metabolic acidosis of any aetiology would be expected to lead to diminished magnesium reabsorption in the distal tubule. Diabetic subjects frequently present with hypomagnesaemia and cellular magnesium deficiency (Husmann et  al., 1997; Khan et  al., 1999; Takaya et  al., 2003). This is, in part, due to urinary magnesium wasting (Tosiello, 1996). In combination, insulin deficiency and ketoacidosis of uncontrolled diabetes diminish magnesium transport in the TALH (Guerrero-Romero et al., 2002; Barbagallo et al., 2003) and DCT (Dai et al., 2001), resulting in exacerbation of renal magnesium wasting.

Phosphate restriction and phosphate depletion One of the hallmarks of hypophosphataemia and cellular phosphate depletion is the striking increase in urinary calcium and magnesium excretion (Kelepouris and Agus, 1998; Ahmad and Swaminathan, 2000; Dai et al., 2001). The hypermagnesiuria may be sufficiently large to lead to overt hypomagnesaemia (Coburn and Massry, 1970). The increase in divalent ion excretion in both human disease and experimental animal models occurs within hours of dietary phosphate restriction. Three mechanisms have been proposed to account for the increased renal excretion: (1) mobilization of calcium and magnesium from bone, (2) suppression of PTH secretion, and (3) disturbed tubular transport (Dai et al., 2001). It is evident from clearance experiments that the urinary excretion of divalent cations in phosphate-depleted human subjects is inappropriate for the plasma concentration, supporting the notion of defective tubular transport (Coburn and Massry, 1970). Dai et al. have shown that cellular phosphate depletion leads to diminished magnesium uptake in mDCT cells (Dai et al., 1997b). This observation supports the notion that the DCT may be involved, in part, in decreased magnesium absorption and increased magnesium excretion associated with hypophosphataemia. The reasons why cellular phosphate depletion leads to diminished magnesium reabsorption are not known.

Cellular potassium depletion Hypokalaemia and cellular potassium depletion are associated with diminished magnesium absorption within TALH and DCT that may lead to increased magnesium excretion (Quamme, 1997). The increase in urinary excretion of divalent cations may be explained by the effect of potassium depletion on salt absorption in the TALH. Chloride conservation is impaired in potassium-depleted rats, which may be related to a change in basolateral Na+,K+-ATPase activity, resulting in impaired apical NKCC2-mediated sodium chloride co-transport.

407

408

Section 2  

fluid, electrolyte, and renal tubular disorders

To date, there is no direct evidence for changes in magnesium absorption in the TALH with potassium depletion. However, as magnesium and calcium are absorbed by passive mechanisms, it is probable that impaired salt transport can lead to diminished divalent cation absorption in this segment. Studies using isolated mDCT cells suggest that potassium depletion may have additional effects on magnesium transport in the DCT (Dai et al., 1997a). Cellular potassium depletion results in the inhibition of magnesium uptake into mDCT cells, as determined by microfluorescence. The exact mechanism for the disturbed magnesium entry is not known, but might involve a diminished potassium excretion via Kv1.1 (KCNA1), as suggested by the discovery of mutant Kv1.1 in a dominantly inherited form of hypomagnesaemia (see above) (Glaudemans et al., 2009). The authors propose a model in which Kv1.1 activity is critical for the establishment and maintenance of the DCT cell membrane potential as a prerequisite for magnesium entry via TRPM6. Further studies are required to fully explain the defective magnesium transport in the DCT associated with cellular potassium depletion. Experimental and clinical data suggest a close association of serum magnesium, potassium, and phosphate levels. Crook et al. reported a twofold increase in the prevalence of hypophosphataemia (plasma phosphate < 0.8  mmol/L) and a sixfold increase in hypokalaemia (plasma potassium < 3.5 mmol/L) in patients with hypomagnesaemia (plasma magnesium < 0.70  mmol/L) (Crook, 1994). A  trilogy consisting of hypomagnesaemia, hypophosphataemia, and hypokalaemia was also found in 8% of patients with hypomagnesaemia and 17% of patients with severe hypomagnesaemia (plasma magnesium < 0.50 mmol/L). The evidence suggests that hypokalaemia and hypophosphataemia may have profound effects on tubular magnesium transport. Many of the syndromes associated with potassium depletion and phosphate depletion are complicated by concurrent alterations in acid–base balance (Crook, 1994). There is evidence that acid–base changes have different effects on magnesium transport relative to potassium or phosphate depletion, such that the three disturbances may act in an additive manner to compromise renal magnesium conservation (Dai et al., 2001).

References Adalat, S., Woolf, A. S., Johnstone, K. A., et al. (2009). HNF1B mutations associate with hypomagnesemia and renal magnesium wasting. J Am Soc Nephrol, 20(5), 1123–31. Ahmad, A. and Swaminathan, R. (2000). Disorders of magnesium metabolism. In R. Alpern, M. Caplan, and O. Moe (eds.) The Kidney: Physiology and Pathophysiology, pp. 1732–48. New York: Raven Press. Anast, C. S., Mohs, J. M., Kaplan, S. L., et al. (1972). Evidence for parathyroid failure in magnesium deficiency. Science, 177(4049), 606–8. Arany, I. and Safirstein, R. L. (2003). Cisplatin nephrotoxicity. Semin Nephrol, 23(5), 460–4. Aries, P. M., Schubert, M., Müller-Wieland, D., et al. (2000). [Subcutaneous magnesium pump in a patient with combined magnesium transport defect]. Dtsch Med Wochenschr, 125(33), 970–2. Arnold, A., Tovey, J., Mangat, P., et al. (1995). Magnesium deficiency in critically ill patients. Anaesthesia, 50(3), 203–5. Arystarkhova, E., Donnet, C., Asinovski, N. K., et al. (2002b). Differential regulation of renal Na,K-ATPase by splice variants of the gamma subunit. J Biol Chem, 277(12), 10162–72. Arystarkhova, E., Wetzel, R. K., and Sweadner, K. J. (2002a). Distribution and oligomeric association of splice forms of Na+-K+-ATPase regulatory

gamma-subunit in rat kidney. Am J Physiol Renal Physiol, 282(3), F393–407. Barbagallo, M., Dominguez, L. J., Galioto, A., et al. (2003). Role of magnesium in insulin action, diabetes and cardio-metabolic syndrome X. Mol Aspects Med, 24(1–3), 39–52. Benigno, V., Canonica, C. S., Bettinelli, A., et al. (2000). Hypomagnesaemia-hypercalciuria-nephrocalcinosis: a report of nine cases and a review. Nephrol Dial Transplant, 15(5), 605–10. Bianchetti, M. G., Kanaka, C., Ridolfi-Lüthy, A., et al. (1991). Persisting renotubular sequelae after cisplatin in children and adolescents. Am J Nephrol, 11(2), 127–30. Blanchard, A., Jeunemaitre, X., Coudol, P., et al. (2001). Paracellin-1 is critical for magnesium and calcium reabsorption in the human thick ascending limb of Henle. Kidney Int, 59(6), 2206–15. Blostein, R., Pu, H. X., Scanzano, R., et al. (2003). Structure/function studies of the gamma subunit of the Na,K-ATPase. Ann N Y Acad Sci, 986, 420–7. Boulikas, T. and Vougiouka, M. (2004). Recent clinical trials using cisplatin, carboplatin and their combination chemotherapy drugs (review). Oncol Rep, 11(3), 559–95. Browne, D. L., Gancher, S. T., Nutt, J. G., et al. (1994). Episodic ataxia/ myokymia syndrome is associated with point mutations in the human potassium channel gene, KCNA1. Nat Genet, 8(2), 136–40. Cairo, E. R., Friedrich, T., Swarts, H. G., et al. (2008). Impaired routing of wild type FXYD2 after oligomerisation with FXYD2-G41R might explain the dominant nature of renal hypomagnesemia. Biochim Biophys Acta, 1778(2), 398–404. Cao, Y., Liao, C., Tan, A., et al. (2010). Meta-analysis of incidence and risk of hypomagnesemia with cetuximab for advanced cancer. Chemotherapy, 56(6), 459–65. Carrick, S., Ghersi, D., Wilcken, N., et al. (2004). Platinum containing regimens for metastatic breast cancer. Cochrane Database Syst Rev, 2, CD003374. Chang, C. T., Hung, C. C., Tian, Y. C., et al. (2007). Ciclosporin reduces paracellin-1 expression and magnesium transport in thick ascending limb cells. Nephrol Dial Transplant 22(4), 1033–40. Chubanov, V., Schlingmann, K. P., Wäring, J., et al. (2007). Hypomagnesemia with secondary hypocalcemia due to a missense mutation in the putative pore-forming region of TRPM6. J Biol Chem, 282(10), 7656–67. Chubanov, V., Waldegger, S., Mederos y Schnitzler, M., et al. (2004). Disruption of TRPM6/TRPM7 complex formation by a mutation in the TRPM6 gene causes hypomagnesemia with secondary hypocalcemia. Proc Natl Acad Sci U S A, 101(9), 2894–9. Coburn, J. W. and Massry, S. G. (1970). Changes in serum and urinary calcium during phosphate depletion: studies on mechanisms. J Clin Invest, 49(6), 1073–87. Cole, D. E. and Quamme, G. A. (2000). Inherited disorders of renal magnesium handling. J Am Soc Nephrol, 11(10), 1937–47. Cole, D. E., Kooh, S. W., and Vieth, R. (2000). Primary infantile hypomagnesaemia: outcome after 21 years and treatment with continuous nocturnal nasogastric magnesium infusion. Eur J Pediatr, 159(1–2), 38–43. Colegio, O. R., Van Itallie, C., Rahner, C., et al. (2003). Claudin extracellular domains determine paracellular charge selectivity and resistance but not tight junction fibril architecture. Am J Physiol Cell Physiol, 284(6), C1346–54. Crook, M. A. (1994). Hypophosphataemia and hypokalaemia in patients with hypomagnesaemia. Br J Biomed Sci, 51(1), 24–7. Cundy, T. and Mackay, J. (2011). Proton pump inhibitors and severe hypomagnesaemia. Curr Opin Gastroenterol, 27(2), 180–5. Dai, L. J., Friedman, P. A., and Quamme, G. A. (1997a). Cellular mechanisms of chlorothiazide and cellular potassium depletion on Mg2+ uptake in mouse distal convoluted tubule cells. Kidney Int, 51(4), 1008–17. Dai, L. J., Friedman, P. A., and Quamme, G. A. (1997b). Phosphate depletion diminishes Mg2+ uptake in mouse distal convoluted tubule cells. Kidney Int, 51(6), 1710–18.

chapter 40 

Dai, L. J., Ritchie, G., Kerstan, D., et al. (2001). Magnesium transport in the renal distal convoluted tubule. Physiol Rev, 81(1), 51–84. De Rouffignac, C. (1995). Multihormonal regulation of nephron epithelia: achieved through combinational mode? Am J Physiol, 269(4 Pt 2), R739–48. De Rouffignac, C. and Quamme, G. (1994). Renal magnesium handling and its hormonal control. Physiol Rev, 74(2), 305–22. Di Stefano, A., Desfleur, E., Moine, G., et al. (1995). Regulation of Ca2+ and Mg2+ reabsorption in the mouse cortical thick ascending limb of Henle’s loop (cTAL) by low magnesium diet and metabolic alkalosis. In The 2nd European Kidney Research Forum,1995, Abstract P 015. Djurhuus, M. S., Gram, J., Petersen, P. H., et al. (1995). Biological variation of serum and urinary magnesium in apparently healthy males. Scand J Clin Lab Invest, 55(6), 549–58. Dorovkov, M. V. and Ryazanov, A. G. (2004). Phosphorylation of annexin I by TRPM7 channel-kinase. J Biol Chem, 279(49), 50643–6. Elin, R. J. (1994). Magnesium: the fifth but forgotten electrolyte. Am J Clin Pathol, 102(5), 616–22. Elisaf, M., Kalaitzidis, R., and Siamopoulos, K. C. (1998). Multiple electrolyte abnormalities after pamidronate administration. Nephron, 79(3), 337–9. Elisaf, M., Panteli, K., Theodorou, J., et al. (1997). Fractional excretion of magnesium in normal subjects and in patients with hypomagnesemia. Magnes Res, 10(4), 315–20. Elliott, C., Newman, N., and Madan, A. (2000). Gentamicin effects on urinary electrolyte excretion in healthy subjects. Clin Pharmacol Ther, 67(1), 16–21. English, M. W., Skinner, R., Pearson, A. D., et al. (1999). Dose-related nephrotoxicity of carboplatin in children. Br J Cancer, 81(2), 336–41. Epstein, M., McGrath, S., and Law, F. (2006). Proton-pump inhibitors and hypomagnesemic hypoparathyroidism. N Engl J Med, 355(17), 1834–6. Estévez, R., Boettger, T., Stein, V., et al. (2001). Barttin is a Cl- channel beta-subunit crucial for renal Cl- reabsorption and inner ear K+ secretion. Nature, 414(6863), 558–61. Garland, H. O., Birdsey, T. J., Davidge, C. G., et al. (1994). Effects of gentamicin, neomycin and tobramycin on renal calcium and magnesium handling in two rat strains. Clin Exp Pharmacol Physiol, 21(2), 109–15. Geven, W. B., Monnens, L. A., Willems, J. L, et al. (1987a). Isolated autosomal recessive renal magnesium loss in two sisters. Clin Genet, 32(6), 398–402. Geven, W. B., Monnens, L. A., Willems, H. L., et al. (1987b). Renal magnesium wasting in two families with autosomal dominant inheritance. Kidney Int, 31(5), 1140–4. Giapros, V. I., Cholevas, V. I., Andronikou, S. K. (2004). Acute effects of gentamicin on urinary electrolyte excretion in neonates. Pediatr Nephrol 19(3), 322–5. Gibson, M. M., Bagga, D. A., Miller, C. G., et al. (1991). Magnesium transport in Salmonella typhimurium: the influence of new mutations conferring Co2+ resistance on the CorA Mg2+ transport system. Mol Microbiol, 5(11), 2753–62. Glaudemans, B., van der Wijst, J., Scola, R. H., et al. (2009). A missense mutation in the Kv1.1 voltage-gated potassium channel-encoding gene KCNA1 is linked to human autosomal dominant hypomagnesemia. J Clin Invest, 119(4), 936–42. Godron, A., Harambat, J., Boccio, V., et al. (2012). Familial hypomagnesemia with hypercalciuria and nephrocalcinosis: phenotype-genotype correlation and outcome in 32 patients with CLDN16 or CLDN19 mutations. Clin J Am Soc Nephrol, 7(5), 801–9. Goren, M. P. (2003). Cisplatin nephrotoxicity affects magnesium and calcium metabolism. Med Pediatr Oncol, 41(3), 186–9. Goytain, A. and Quamme, G. A. (2005). Functional characterization of ACDP2 (ancient conserved domain protein), a divalent metal transporter. Physiol Genomics, 22(3), 382–9. Groenestege, W. M., Thébault, S., van der Wijst, J., et al. (2007). Impaired basolateral sorting of pro-EGF causes isolated recessive renal hypomagnesemia. J Clin Invest, 117(8), 2260–7.

approach to the patient with hypomagnesaemia

Guerrero-Romero, F. and Rodríguez-Morán, M. (2002). Low serum magnesium levels and metabolic syndrome. Acta Diabetol, 39(4), 209–13. Haisch, L., Almeida, J. R., Abreu da Silva, P. R., et al. (2011). The role of tight junctions in paracellular ion transport in the renal tubule: lessons learned from a rare inherited tubular disorder. Am J Kidney Dis, 57(2), 320–30. Hashimoto, Y., Nishimura, Y., Maeda, H., et al. (2000). Assessment of magnesium status in patients with bronchial asthma. J Asthma, 37(6), 489–96. Hébert, P., Mehta, N., Wang, J., et al. (1997). Functional magnesium deficiency in critically ill patients identified using a magnesium-loading test. Crit Care Med, 25(5), 749–55. Heidet, L., Decramer, S., Pawtowski, A., et al. (2010). Spectrum of HNF1B mutations in a large cohort of patients who harbor renal diseases. Clin J Am Soc Nephrol, 5(6), 1079–90. Hodgkinson, E., Neville-Webbe, H. L., and Coleman, R. E. (2006). Magnesium depletion in patients receiving cisplatin-based chemotherapy. Clin Oncol (R Coll Radiol), 18(9), 710–8. Hofmann, T., Schaefer, M., Schultz, G., et al. (2002). Subunit composition of mammalian transient receptor potential channels in living cells. Proc Natl Acad Sci U S A, 99(11), 7461–6. Hou, J. and Goodenough, D. A. (2010). Claudin-16 and claudin-19 function in the thick ascending limb. Curr Opin Nephrol Hypertens, 19(5), 483–8. Hou, J., Renigunta, A., Gomes, A. S., et al. (2009). Claudin-16 and claudin-19 interaction is required for their assembly into tight junctions and for renal reabsorption of magnesium. Proc Natl Acad Sci U S A, 106(36), 15350–5. Husmann, M. J., Fuchs, P., Truttmann, A. C., et al. (1997). Extracellular magnesium depletion in pediatric patients with insulin-dependent diabetes mellitus. Miner Electrolyte Metab, 23(2), 121–4. Jalkanen, R., Pronicka, E., Tyynismaa, H., et al. (2006). Genetic background of HSH in three Polish families and a patient with an X;9 translocation. Eur J Hum Genet, 14(1), 55–62. Kang, H. S., Kerstan, D., Dai, L., et al. (2000). Aminoglycosides inhibit hormone-stimulated Mg2+ uptake in mouse distal convoluted tubule cells. Can J Physiol Pharmacol, 78(8), 595–602. Keating, M. J., Sethi, M. R., Bodey, G. P., et al. (1977). Hypocalcemia with hypoparathyroidism and renal tubular dysfunction associated with aminoglycoside therapy. Cancer, 39(4), 1410–4. Kelepouris, E. and Agus, Z. S. (1998). Hypomagnesemia: renal magnesium handling. Semin Nephrol, 18(1), 58–73. Kerstan, D. and Quamme, G. (2002). Physiology and pathophysiology of intestinal absorption of magnesium. In S. G. Massry, H. Morii, and Y. Nishizawa (eds.) Calcium in Internal Medicine, pp. 171–83. London: Springer-Verlag. Khan, L. A., Alam, A. M., Ali, L., et al. (1999). Serum and urinary magnesium in young diabetic subjects in Bangladesh. Am J Clin Nutr, 69(1), 70–3. Kim, S. J., Kang, H. S., Jeong, C. W., et al. (2006). Immunosuppressants inhibit hormone-stimulated Mg2+ uptake in mouse distal convoluted tubule cells. Biochem Biophys Res Commun, 341(3), 742–8. Konrad, M., Hou, J., Weber, S., et al. (2008). CLDN16 genotype predicts renal decline in familial hypomagnesemia with hypercalciuria and nephrocalcinosis. J Am Soc Nephrol, 19(1), 171–81. Konrad, M., Schaller, A., Seelow, D., et al. (2006). Mutations in the tight-junction gene claudin 19 (CLDN19) are associated with renal magnesium wasting, renal failure, and severe ocular involvement. Am J Hum Genet, 79(5), 949–57. Lajer, H. and Daugaard, G. (1999). Cisplatin and hypomagnesemia. Cancer Treat Rev, 25(1), 47–58. Ledeganck, K. J., Boulet, G. A., Horvath, C. A., et al. (2011). Expression of renal distal tubule transporters TRPM6 and NCC in a rat model of cyclosporine nephrotoxicity and effect of EGF treatment. Am J Physiol Renal Physiol, 301(3), F486–93. Lelievre-Pegorier, M., Merlet-Benichou, C., Roinel, N., et al. (1983). Developmental pattern of water and electrolyte transport in rat superficial nephrons. Am J Physiol, 245(1), F15–21.

409

410

Section 2  

fluid, electrolyte, and renal tubular disorders

Loffing, J., Vallon, V., Loffing-Cueni, D., et al. (2004). Altered renal distal tubule structure and renal Na+ and Ca2+ handling in a mouse model for Gitelman's syndrome. J Am Soc Nephrol, 15(9), 2276–88. Lombeck, I., Ritzl, F., Schnippering, H. G., et al. (1975). Primary hypomagnesemia. I. Absorption Studies. Z Kinderheilkd, 118(4), 249–58. Lote, C. J., Thewles, A., Wood, J. A., et al. (2000). The hypomagnesaemic action of FK506: urinary excretion of magnesium and calcium and the role of parathyroid hormone. Clin Sci (Lond), 99(4), 285–92. Magil, A. B., Mavichak, V., Wong, N. L., et al. (1986). Long-term morphological and biochemical observations in cisplatin-induced hypomagnesemia in rats. Nephron, 43(3), 223–30. Markmann, M., Rothman, R., Reichman, B., et al. (1991). Persistent hypomagnesemia following cisplatin chemotherapy in patients with ovarian cancer. J Cancer Res Clin Oncol, 117(2), 89–90. Matzkin, H., Lotan, D., and Boichis, H. (1989). Primary hypomagnesemia with a probable double magnesium transport defect. Nephron 52(1), 83–6. Mavichak, V., Wong, N. L., Quamme, G. A., et al. (1985). Studies on the pathogenesis of cisplatin-induced hypomagnesemia in rats. Kidney Int, 28(6), 914–21. McLarnon, S., Holden, D., Ward, D., et al. (2002). Aminoglycoside antibiotics induce pH-sensitive activation of the calcium-sensing receptor. Biochem Biophys Res Commun, 297(1), 71–7. Meij, I. C., Koenderink, J. B., De Jong, J. C., et al. (2003). Dominant isolated renal magnesium loss is caused by misrouting of the Na+,K+-ATPase gamma-subunit. Ann N Y Acad Sci, 986, 437–43. Meij, I. C., Koenderink, J. B., van Bokhoven, H., et al. (2000). Dominant isolated renal magnesium loss is caused by misrouting of the Na+,K+-ATPase gamma-subunit. Nat Genet, 26(3), 265–6. Mervaala, E. M., Müller, D. N., Park, J. K., et al. (1999). Monocyte infiltration and adhesion molecules in a rat model of high human renin hypertension. Hypertension, 33(1 Pt 2), 389–95. Meyer, T. E., Verwoert, G. C., Hwang, S. J., et al. (2010). Genome-wide association studies of serum magnesium, potassium, and sodium concentrations identify six Loci influencing serum magnesium levels. PLoS Genet, 6(8). Miura, K., Nakatani, T., Asai, T., et al. (2002). Role of hypomagnesemia in chronic cyclosporine nephropathy. Transplantation, 73(3), 340–7. Monteilh-Zoller, M. K., Hermosura, M. C., Nadler, M. J., et al. (2003). TRPM7 provides an ion channel mechanism for cellular entry of trace metal ions. J Gen Physiol, 121(1), 49–60. Müller, D., Kausalya, P. J., Claverie-Martin, F., et al. (2003). A novel claudin 16 mutation associated with childhood hypercalciuria abolishes binding to ZO-1 and results in lysosomal mistargeting. Am J Hum Genet, 73(6), 1293–301. Nadler, M. J., Hermosura, M. C., Inabe, K., et al. (2001). LTRPC7 is a Mg.ATP-regulated divalent cation channel required for cell viability. Nature, 411(6837), 590–5. Nicholson, J. C., Jones, C. L., Powell, H. R., et al. (1995). Familial hypomagnesaemia—hypercalciuria leading to end-stage renal failure. Pediatr Nephrol, 9(1), 74–6. Nicoll, G. W., Struthers, A. D., and Fraser, C. G. (1991). Biological variation of urinary magnesium. Clin Chem, 37(10 Pt 1), 1794–5. Nijenhuis, T., Hoenderop, J. G., and Bindels, R. J. (2004). Downregulation of Ca2+ and Mg2+ transport proteins in the kidney explains tacrolimus (FK506)-induced hypercalciuria and hypomagnesemia. J Am Soc Nephrol, 15(3), 549–57. Ohba, Y., Kitagawa, H., Kitoh, K., et al. (2000). A deletion of the paracellin-1 gene is responsible for renal tubular dysplasia in cattle. Genomics, 68(3), 229–36. Panichpisal, K., Angulo-Pernett, F., Selhi, S., et al. (2006). Gitelman-like syndrome after cisplatin therapy: a case report and literature review. BMC Nephrol, 7, 10. Paunier, L., Radde, I. C., Kooh, S. W., et al. (1968). Primary hypomagnesemia with secondary hypocalcemia in an infant. Pediatrics, 41(2), 385–402.

Praga, M., Vara, J., González-Parra, E., et al. (1995). Familial hypomagnesemia with hypercalciuria and nephrocalcinosis. Kidney Int, 47(5), 1419–25. Quamme, G. A. (1989). Control of magnesium transport in the thick ascending limb. Am J Physiol, 256(2 Pt 2), F197–210. Quamme, G. A. (1997). Renal magnesium handling: new insights in understanding old problems. Kidney Int, 52(5), 1180–95. Rob, P. M., Lebeau, A., Nobiling, R., et al. (1996). Magnesium metabolism: basic aspects and implications of ciclosporine toxicity in rats. Nephron, 72(1), 59–66. Rodríguez-Soriano, J. and Vallo, A. (1994). Pathophysiology of the renal acidification defect present in the syndrome of familial hypomagnesaemia-hypercalciuria. Pediatr Nephrol, 8(4), 431–5. Ryazanova, L. V., Dorovkov, M. V., Ansari, A., et al. (2004). Characterization of the protein kinase activity of TRPM7/ChaK1, a protein kinase fused to the transient receptor potential ion channel. J Biol Chem, 279(5), 3708–16. Schlingmann, K. P., Konrad,M., and Seyberth, H. W. (2004). Genetics of hereditary disorders of magnesium homeostasis. Pediatr Nephrol, 19(1), 13–25. Schlingmann, K. P., Sassen, M. C., Weber, S., et al. (2005). Novel TRPM6 mutations in 21 families with primary hypomagnesemia and secondary hypocalcemia. J Am Soc Nephrol, 16(10), 3061–9. Schlingmann, K. P., Weber, S., Peters, M., et al. (2002). Hypomagnesemia with secondary hypocalcemia is caused by mutations in TRPM6, a new member of the TRPM gene family. Nat Genet, 31(2), 166–70. Schmitz, C., Perraud, A. L., Johnson, C. O., et al. (2003). Regulation of vertebrate cellular Mg2+ homeostasis by TRPM7. Cell, 114(2), 191–200. Schweigel, M. and Martens, H. (2000). Magnesium transport in the gastrointestinal tract. Front Biosci, 5, D666–77. Shah, G. M. and Kirschenbaum, M. A. (1991). Renal magnesium wasting associated with therapeutic agents. Miner Electrolyte Metab, 17(1), 58–64. Shalev, H., Phillip, M., Galil, A., et al. (1998). Clinical presentation and outcome in primary familial hypomagnesaemia. Arch Dis Child, 78(2), 127–30. Simon, D. B., Lu, Y., Choate, K. A., et al. (1999). Paracellin-1, a renal tight junction protein required for paracellular Mg2+ resorption. Science, 285(5424), 103–6. Stuiver, M., Lainez, S., Will, C., et al. (2011). CNNM2, encoding a basolateral protein required for renal Mg2+ handling, is mutated in dominant hypomagnesemia. Am J Hum Genet, 88(3), 333–43. Sutton, R. A. and Domrongkitchaiporn, S. (1993). Abnormal renal magnesium handling. Miner Electrolyte Metab, 19(4–5), 232–40. Swainson, C. P., Colls, B. M., and Fitzharris, B. M. (1985).Cis-platinum and distal renal tubule toxicity. N Z Med J, 98(779), 375–8. Takaya, J., Higashino, H., Kotera, F., et al. (2003). Intracellular magnesium of platelets in children with diabetes and obesity. Metabolism, 52(4), 468–71. Tang, N. L., Cran, Y. K., Hui, E., et al. (2000). Application of urine magnesium/creatinine ratio as an indicator for insufficient magnesium intake. Clin Biochem, 33(8), 675–8. Tejpar, S., Piessevaux, H., Claes, K., et al. (2007). Magnesium wasting associated with epidermal-growth-factor receptor-targeting antibodies in colorectal cancer: a prospective study. Lancet Oncol, 8(5), 387–94. Thebault, S., Alexander, R. T., Tiel Groenestege, W. M., et al. (2009). EGF increases TRPM6 activity and surface expression. J Am Soc Nephrol, 20(1), 78–85. Tosiello, L. (1996). Hypomagnesemia and diabetes mellitus. A review of clinical implications. Arch Intern Med, 156(11), 1143–8. Van der Wijst, J., Glaudemans, B., Venselaar, H., et al. (2010). Functional analysis of the Kv1.1 N255D mutation associated with autosomal dominant hypomagnesemia. J Biol Chem, 285(1), 171–8. Vetter, T. and Lohse, M. J. (2002). Magnesium and the parathyroid. Curr Opin Nephrol Hypertens, 11(4), 403–10. Vincenzi, B., Galluzzo, S., Santini, D., et al. (2011). Early magnesium modifications as a surrogate marker of efficacy of cetuximab-based anticancer

chapter 40 

treatment in KRAS wild-type advanced colorectal cancer patients. Ann Oncol, 22(5), 1141–6. Vincenzi, B., Santini, D., Galluzzo, S., et al. (2008). Early magnesium reduction in advanced colorectal cancer patients treated with cetuximab plus irinotecan as predictive factor of efficacy and outcome. Clin Cancer Res, 14(13), 4219–24. Voets, T., Nilius, B., Hoefs, S., et al. (2004). TRPM6 forms the Mg2+ influx channel involved in intestinal and renal Mg2+ absorption. J Biol Chem, 279(1), 19–25. Walder, R. Y., Landau, D., Meyer, P., et al. (2002). Mutation of TRPM6 causes familial hypomagnesemia with secondary hypocalcemia. Nat Genet, 31(2), 171–4. Walder, R. Y., Shalev, H., Brennan, T. M., et al. (1997). Familial hypomagnesemia maps to chromosome 9q, not to the X chromosome: genetic linkage mapping and analysis of a balanced translocation breakpoint. Hum Mol Genet, 6(9), 1491–7. Ward, D. T., McLarnon, S. J., Riccardi, D. (2002). Aminoglycosides increase intracellular calcium levels and ERK activity in proximal tubular OK cells expressing the extracellular calcium-sensing receptor. J Am Soc Nephrol, 13(6), 1481–9. Weber, S., Hoffmann, K., Jeck, N., et al. (2000). Familial hypomagnesaemia with hypercalciuria and nephrocalcinosis maps to chromosome 3q27 and is associated with mutations in the PCLN-1 gene. Eur J Hum Genet, 8(6), 414–22.

approach to the patient with hypomagnesaemia

Weber, S., Schneider, L., Peters, M., et al. (2001). Novel paracellin-1 mutations in 25 families with familial hypomagnesemia with hypercalciuria and nephrocalcinosis. J Am Soc Nephrol, 12(9), 1872–81. Weinberg, J. M., Harding, P. G., and Humes, H. D. (1983). Alterations in renal cortex cation homeostasis during mercuric chloride and gentamicin nephrotoxicity. Exp Mol Pathol, 39(1), 43–60. Will, C., Breiderhoff, T., Thumfart, J., et al. (2010). Targeted deletion of murine Cldn16 identifies extra- and intrarenal compensatory mechanisms of Ca2+ and Mg2+ wasting. Am J Physiol Renal Physiol, 298(5), F1152–61. Wilson, F. H., Hariri, A., Farhi, A., et al. (2004). A cluster of metabolic defects caused by mutation in a mitochondrial tRNA. Science, 306(5699), 1190–4. Wolf, M. T., Dötsch, J., Konrad, M., et al. (2002). Follow-up of five patients with FHHNC due to mutations in the Paracellin-1 gene. Pediatr Nephrol, 17(8), 602–8. Wong, N. L. and Dirks, J. H. (1988). Cyclosporin-induced hypomagnesaemia and renal magnesium wasting in rats. Clin Sci (Lond), 75(5), 509–14. Yao, X., Panichpisal, K., Kurtzman, N., et al. (2007). Cisplatin nephrotoxicity: a review. Am J Med Sci, 334(2), 115–24.

411

CHAPTER 41

Approach to the patient with renal Fanconi syndrome, glycosuria, or aminoaciduria Detlef Bockenhauer and Robert Kleta Introduction The proximal tubule reabsorbs the bulk of the glomerular filtrate. Up to 80% of filtered salt and water is returned back into the circulation in this tubular segment (Gottschalk, 1961; Lassiter et al., 1961; Bockenhauer et al., 2008). Moreover, several solutes, such as phosphate, glucose, low-molecular weight proteins (LWMPs), and amino acids are exclusively reabsorbed in the proximal tubule. An entire orchestra of specialized apical and basolateral transporters, as well as paracellular molecules, mediate this reabsorption. Defects in proximal tubular function can be isolated (e.g. isolated renal glycosuria, aminoacidurias, or hypophosphataemic rickets) or generalized. In the latter case it is called the Fanconi–Debre–de Toni syndrome, based on the initial clinical descriptions (Fanconi, 1931, 1936, de Toni, 1933; Debre et al., 1934). However, in clinical practice it is usually referred to as just the ‘renal Fanconi syndrome’. Severity of proximal tubular dysfunction can vary: a complete loss of function with intact glomerular filtration rate (GFR) and no compensation elsewhere is probably not compatible with life, since it would result in a urine output of > 100 L per day, assuming a daily filtrate volume of 150 L.

Clinical manifestations of renal Fanconi syndrome The clinical presentation of renal Fanconi syndrome depends on the underlying cause, severity, and age at presentation. Onset of symptoms is usually insidious and non-specific. Polyuria/polydipsia is typically present. Not uncommonly the diagnosis is incidental, prompted by blood or urine tests obtained for another indication. Some patients may experience bone pain and metabolic bone disease. In children, rickets is a common presenting symptom, as originally described by Fanconi, Debre, and de Toni (Fanconi, 1931, 1936, de Toni, 1933; Debre et al., 1934). Several factors contribute to the development of rickets: (a) the renal phosphate losses with secondary hypophosphataemia and (b) renal calcium losses, and (c) deficiency of 1,25(OH) vitamin D. Vitamin D is fat-soluble and circulates in the plasma by means of a carrier protein, vitamin D binding protein (DBP). DBP is a LMWP and the complex of DBP and vitamin D is filtered and then reabsorbed in the proximal tubule (Negri, 2006). Following uptake in the epithelial cell of the

proximal tubule, vitamin D can be converted into its most active form by 1α-hydroxylation. Consequently, there is a deficiency of 1α-hydroxylated vitamin D with impaired proximal tubular function.

Biochemical abnormalities Once suspected, the diagnosis of renal Fanconi syndrome is easily confirmed. Since many solutes, including LMWP, phosphate, amino acids, and glucose are exclusively reabsorbed in the proximal tubule, wasting all these substances in the urine clearly establishes the diagnosis. In addition, the biochemical profile includes a hypokalaemic, hyperchloraemic metabolic acidosis with hypophosphataemia and hypercalciuria. Markers of disturbed calcium/phosphate homeostasis, such as parathyroid hormone (PTH) and alkaline phosphatase are typically elevated. Moreover, plasma levels of vitamin D, especially 1,25-OH vitamin D are depressed (see above).

Proteinuria Proteinuria consists mainly of LMWP, that is, those proteins that are physiologically filtered and need to be reabsorbed in the proximal tubule via the endocytic receptors megalin and cubilin (Christensen and Gburek, 2004). Under physiologic circumstances > 99% of filtered proteins are reabsorbed this way. Determination of LMWP, for example, by measurement of urinary α1-microglobulin, β2-microglobulin, or retinol-binding protein (RBP) in relation to urine creatinine, is an exquisitely sensitive marker of proximal tubular function. Median RBP to creatinine ratios were shown to be more than 1000-fold higher in patients with proximal tubular disorders, such as Dent disease compared with normal controls (Norden et  al., 2000). Of note, LMWP proteinuria is typically missed or significantly underestimated by traditional urine dipstick testing. While albumin is usually considered a marker of glomerular function, it is important to remember that some albumin is also filtered and reabsorbed (Birn et al., 2000; Hryciw et al., 2006; Comper et al., 2008). The magnitude of physiologic albumin filtration is debated, but some animal experiments suggest it may actually be in the nephrotic range (Gekle, 2007; Pollock and Poronnik, 2007; Russo et al., 2007). Therefore, it is not surprising that albuminuria is also a typical feature of renal Fanconi syndrome, although plasma levels

chapter 41 

of albumin are usually normal (Norden et al., 2001). In an inherited form of renal Fanconi syndrome without glomerular failure, total proteinuria was reported to be approximately 850 mg/day/m2 (Tolaymat et al., 1992). Moreover, two siblings with a homozygous frameshift mutation in CUBN, encoding a key endocytic receptor in the proximal tubule, Cubulin, were recently described who had proteinuria up to 2 g per day (Ovunc et al., 2011).

Aminoaciduria Amino acids, like LMWP, are exclusively reabsorbed in the proximal tubule. Approximately 50 g (450 mmol) of free amino acids are filtered daily into the primary urine by human kidney glomeruli and > 99% are reabsorbed under physiologic conditions (Camargo et al., 2012). Distinct transporters are expressed on the apical and basolateral side of the proximal tubular epithelium to facilitate transcellular amino acid transport (Camargo et al., 2008). Consequently, generalized aminoaciduria is a typical feature of renal Fanconi syndrome, although usually not as dramatically increased as the LMWP (Norden et al., 2004).

Organic aciduria Organic acids are also exclusively reabsorbed in the proximal tubule and excretion is increased in renal Fanconi syndrome (Cogan, 1982); transport is mediated by specialized transport molecules, including the urate transporter SLC22A12 and SLC2A9, involved in renal hypouricaemia (Enomoto et  al., 2002; Matsuo et al., 2008; Vitart et al., 2008). Consequently, patients with renal Fanconi syndrome typically have hyperuricosuria with hypouricaemia. Other organic acid transporters include the sodium dicarboxylate transporter NaDC1, involved in citrate reabsorption, leading to hypercitraturia with impaired proximal tubular function (Unwin et al., 2004). Several of these transporters are also involved in renal handling of drugs, such as probenecid, furosemide, or penicillin, potentially altering pharmacokinetics in renal Fanconi syndrome (Roch-Ramel, 1998). The increased excretion of lactate may have diagnostic implications in mitochondrial cytopathies: these are typically characterized by elevated plasma lactate levels. However, in those with associated renal Fanconi syndrome (see ‘Pathogenesis’), the leakage of lactate in the urine can result in misleading normal plasma levels (Niaudet and Rotig, 1996).

Glycosuria Under physiologic conditions, virtually all filtered glucose is reabsorbed in the proximal tubule. Thus, glycosuria is another marker of proximal tubular dysfunction. Glycosuria can be isolated (see ‘Isolated renal glycosuria’), or seen in the context of renal Fanconi syndrome. Obviously, a concomitant blood glucose level is necessary to determine whether the glycosuria is secondary to an increased filtered load (hyperglycaemia) or impaired tubular reabsorption (normoglycaemia). It is important to remember, that dipsticks only pick up a urine glucose concentration > 5 mmol/L. Hence, glycosuria, especially in the context of polyuria with dilute urine, can be missed. A  formal biochemical measurement is preferable and a 24-hour urine collection for glucose contents the gold standard. Normally, < 1.5 mmol (300 mg)/day/1.73 m2 are excreted (Elsas and Rosenberg, 1969). Assuming a blood glucose level of 5 mmol/L and a GFR of 100 mL/min/1.73 m2 means that under physiologic conditions > 99.5% of filtered glucose is reabsorbed.

renal fanconi syndrome

Phosphaturia Renal phosphate wasting with secondary hypophosphataemia is another hallmark of proximal tubule dysfunction with consequent clinical signs and symptoms, that is, rickets or bone disease. This can be isolated, as in hypophosphataemic rickets or in the context of generalized proximal tubule dysfunction, such as in renal Fanconi syndrome. By convention and tradition, urine phosphate handling is usually assessed as the tubular reabsorption (TRP), which is the complement to the fractional excretion of phosphate (FEP). It is calculated as: TRP (%) = 100 – FEP (%). If 10% of filtered phosphate is excreted (FEP 10%), then 90% of filtered phosphate must have been reabsorbed (TRP 90%). By definition, a TRP > 70% is considered normal, however this can be misleading: after a large phosphate load, a TRP < 70% may be physiologic. Conversely, with severe hypophosphataemia and a decreased filtered load, a TRP of > 70% may be pathologic. To account for the filtered load, urinary phosphate excretion is best assessed using the tubular threshold concentration for phosphate excretion, corrected for glomerular filtration rate (TmP/GFR) (Walton and Bijvoet, 1975). It is calculated as follows: TmP/GFR = Phosphate plasma – Phosphate urine / Creatinine urine × Creatinine plasma Normal values are age dependent and listed in Table 41.1.

Bicarbonaturia and metabolic acidosis Under physiologic conditions the vast majority of filtered bicarbonate is reabsorbed in the proximal tubule, mainly facilitated via the sodium-hydrogen exchanger NHE3 and by carbonic anhydrase CAII (Aronson, 2002; Bobulescu and Moe, 2009). Consequently, bicarbonaturia, and proximal renal tubular acidosis (type 2 renal tubular acidosis), is another obligatory symptom in renal Fanconi syndrome (Rothstein et  al., 1990). However, in clinical practice urinary bicarbonate is rarely measured and even if it is, in steady state it may be normal: the functionally impaired proximal tubule has a decreased bicarbonate threshold and the initial bicarbonate wasting leads to a reduction in plasma bicarbonate levels until the filtered load matches the bicarbonate threshold. Table 41.1  Normal age-specific values for TmP/GFR Age

mmol/L

< 1 month

1.48–3.43

1–3 months

1.48–3.30

4–6 months

1.15–2.60

7 months–2 years

1.10–2.70

2–4 years

1.04–2.79

4–6 years

1.05–2.60

6–8 years

1.26–2.35

8–10 years

1.10–2.31

10–12 years

1.15–2.58

12–15 years

1.18–2.09

> 15 years

0.80–1.35

Derived from Kruse et al. (1982), Bistarakis et al. (1986), and Shaw et al. (1990).

413

414

Section 2  

fluid, electrolyte, and renal tubular disorders

Hypercalciuria Micropuncture studies have shown that approximately 70% of filtered calcium is reabsorbed in the proximal tubule (Hoenderop et al., 2005). The molecular pathway remains to be elucidated, but absorption is presumed to be passive and in parallel with sodium and water reabsorption (Suki, 1979). Consequently, impaired proximal tubular reabsorption of sodium and water also affects calcium transport, and hypercalciuria is another characteristic of renal Fanconi syndrome. Nephrocalcinosis and stone formation can ensue, but interestingly are rather uncommon. Presumably, the polyuria inherent in Fanconi syndrome is protective against these complications. Moreover, the impaired reabsorption and therefore increased luminal concentration of citrate (see ‘Organic aciduria’) is also protective (Pajor, 1999). Nephrocalcinosis/lithiasis is seen more commonly with Dent disease and Lowe syndrome (Thakker, 2000; Bockenhauer et al., 2008; Bokenkamp et al., 2009; Claverie-Martin et al., 2011). The reason for the preponderance to stone formation in these disorders is unclear, but argues for a special role of the underlying molecules, CLCN5 and OCRL, respectively, for calcium transport.

Hypokalaemia Hypokalaemia due to renal potassium wasting is another typical feature of renal Fanconi syndrome. There are two potential mechanisms that contribute to potassium wasting: (a) decreased proximal tubular reabsorption. Like most ions, potassium is predominantly reabsorbed in the proximal tubule and impaired proximal tubular function will increase potassium delivery to the distal tubule (Wright and Giebisch, 1978). (b)  Aldosterone–mediated distal potassium secretion. The volume depletion occurring as a result of impaired proximal sodium reabsorption can lead to activation of the renin–angiotensin–aldosterone system and potassium secretion in the collecting duct (Sebastian et al., 1971).

Pathogenesis The aetiology of renal Fanconi syndrome varies with the age of onset, severity of symptoms, and exposure to certain toxins or medications (see Table 41.2 and Table 41.3). The variety of causes suggests that multiple pathways can disturb proximal tubular function, yet they can be grouped according to some common mechanisms. However, the mechanism of several causes of renal Fanconi syndrome remains to be elucidated, including those associated with valproic acid (Lande et  al., 1993; Zaki and Springate, 2002)  and aristolochic acid (Chinese herb nephropathy) (Yang et  al., 2002; Kazama et al., 2004; Lee et al., 2004; Hong et al., 2006).

Disruption of cellular energy production This appears to be the key mechanism leading to generalized proximal tubular dysfunction. Consistent with this notion, renal Fanconi syndrome is often associated with mitochondrial cytopathies, including defined phenotypes, such as Pearson, Kearns–Sayre, Leigh, and MELAS syndromes (Van Biervliet et  al., 1977; Sperl et al., 1988; Superti-Furga et al., 1993; Wendel et al., 1995; Niaudet and Rotig, 1996; Mourmans et al., 1997; Kuwertz-Broking, 2000; Neiberger et al., 2002; Rotig, 2003). Several acquired forms probably also act by perturbing mitochondrial function. One of the most common causes of renal Fanconi syndrome in adults nowadays is exposure to antiretroviral medications, especially tenofovir

(Verhelst et al., 2002; Earle et al., 2004; Malik et al., 2005; Hussain et  al., 2006; Izzedine et  al., 2006; Woodward et  al., 2009; Hall et al., 2011). In one study, 1.5% of patients exposed to this drug developed renal Fanconi syndrome associated with marked ultrastructural changes of the mitochondria (Woodward et al., 2009). Potentially, genetic variations in genes encoding transporters such as MRP2 (ABCC2) affect intracellular concentrations of tenofovir, where it may affect mitochondrial DNA repair/synthesis (Izzedine et  al., 2006). Aminoglycosides, especially gentamicin, have also been associated with renal Fanconi syndrome (Melnick et  al., 1994; Gainza et  al., 1997; Alexandridis et  al., 2003; Ghiculescu and Kubler, 2006). Aminoglycosides bind to bacterial ribosomes, reducing fidelity of transcription, leading to errors in bacterial protein synthesis (Spahn and Prescott, 1996). The bacterial ancestry of mitochondria makes them particularly susceptible to aminoglycoside toxicity (Bockenhauer et al., 2009). Renal Fanconi syndrome associated with heavy metal intoxication (Thevenod, 2003; Barbier et al., 2005; Gonick, 2008; Johri et al., 2010; Sirac et al., 2011), paraquat (Gil et al., 2005) and suramin (Rago et al., 1974) is presumably also due to mitochondrial dysfunction. An interesting aspect of the cellular energy production pathway as a cause of renal Fanconi syndrome was provided by the recent discovery of proximal tubular dysfunction associated with bi-allelic mutations in SLC34A1, encoding the renal phosphate transporter NaPi-IIa (Magen et al., 2010). It has been speculated that the intracellular deficiency of phosphate may impair proximal tubular ATP generation. However, further clinical details from more affected individuals are needed to convincingly demonstrate generalized proximal tubular dysfunction. Moreover, it is unclear, why renal Fanconi syndrome does not develop in other phosphaturic disorders, such as hypophosphataemic rickets.

Impaired endocytosis and intracellular trafficking Impaired proximal tubular endocytosis is not only a key symptom of renal Fanconi syndrome, but may also contribute to its development. At least, this is suggested by two other genetic forms of partial proximal tubular dysfunction, Dent disease (Devuyst et al., 2005; Wang et  al., 2005; Guggino, 2007)  and Lowe syndrome (Erdmann et al., 2007; Ooms et al., 2009; Cui et al., 2010). Dent disease can be caused by mutations in either CLCN5, a chloride/ proton antiporter (also called Dent 1 disease) or OCRL, an inositol polyphosphate-5-phosphatase (Dent 2 disease) (Lloyd et al., 1996; Hoopes et al., 2005). Interestingly, OCRL was initially identified as the molecular basis of the oculo-cerebral-renal syndrome of Lowe (Attree et al., 1992). It is still unclear why some patients with OCRL mutations appear to have an isolated renal problem (Dent 2), whereas others have the full-blown Lowe syndrome. Careful clinical studies, however, reveal a spectrum of severity with OCRL mutations without a clear distinction between the two diagnostic categories. There is evidence of extrarenal manifestations also in some Dent 2 patients, including elevated muscle enzymes, mild to moderate mental impairment, and stunted growth (Utsch et  al., 2006; Bokenkamp et al., 2009). Both of these proteins are involved in proximal tubular endocytosis and recently it was shown that CLCN5 is involved not only in endocytosis, but also in exocytic trafficking of proximal tubular transporters (Lin et al., 2011). Similarly, OCRL appears to play an important role in the regulation of membrane trafficking (Erdmann et al., 2007; Mao et al., 2009; Cui et al., 2010). However,

chapter 41 

renal fanconi syndrome

Table 41.2  Aetiology of renal Fanconi syndromes and associated clinical features by age of onset Age at onset

Disorder (OMIM #)

Gene (inheritance)

Associated features

Biochemical/diagnostic test

Neonatal

Galactosaemia (230400)

GALT (AR)

Liver dysfunction, jaundice, sepsis, encephalopathy

Red cell galactose 1-phosphate; enzyme tests

Mitochondrial disorders

Multiple (AR or mitochondrial)

Usually multisystem dysfunction (brain, muscle, liver, heart, kidney)

Plasma/cerebrospinal fluid: lactate/ pyruvate (may be normal due to urinary losses), muscle enzymology

Tyrosinaemia type 1 (276700)

FAH (AR)

FTT, hepatic enlargement and dysfunction

Plasma amino acids, urine organic acids (succinyl acetone)

ARC syndrome (208085)

VPS33B (AR)

Arthrogryposis, cholestasis, giant platelets, FTT

Clinical picture, platelet morphology

Fructosaemia (229600)

ALDOB (AR)

Rapid onset of vomiting after fructose ingestion, hypoglycaemia, hepatomegaly

Hepatic fructose-1-phosphate aldolase B activity (liver biopsy)

Cystinosis (219800)

CTNS (AR)

FTT, vomiting, corneal cystine crystals (may be absent, if age < 18 months)

Leucocyte cystine concentration

Fanconi–Bickel syndrome (227810)

SLC2A2 (AR)

FTT, hepatomegaly, hypoglycaemia, glycosuria, galactosuria

Molecular, glycogen storage (liver biopsy)

Lowe syndrome (309000)

OCRL (X-linked)

Cataracts, hypotonia, developmental delay

Molecular

Dent disease (300009)

CLCN5 (X-linked)

Nephrocalcinosis, Phosphate Molecular and bicarbonate wasting often absent

Wilson disease (277900)

ATP7B (AR)

Hepatic and neurological disease, Kayser–Fleischer rings

Isolated incomplete renal Fanconi syndrome (613388)

SLC34A1 (AR)

Infancy

Childhood

Isolated renal Fanconi syndrome with ? kidney failure (134600) (AD)

Rickets, glomerular kidney failure during adolescence

Isolated renal Fanconi syndrome without kidney failure (615605)

Rickets

EHHADH (AD)

Plasma copper, coeruloplasmin, liver biopsy

AR = autosomal recessive; AD = autosomal dominant; FTT = failure to thrive.

even though Dent disease and Lowe syndrome are typically cited as causes of renal Fanconi syndrome, it is debatable whether they always do cause a generalized proximal tubular dysfunction (Kleta, 2008). In a detailed clinical assessment of 16 patients with Lowe syndrome, we showed that there was a spectrum of severity with respect to renal involvement:  while all had hypercalciuria and LMWP, and almost all aminoaciduria, the majority had no clear evidence of phosphate or bicarbonate wasting and none had appreciable glycosuria (Bockenhauer et al., 2008). The clinical features in Dent disease and proven CLCN5 mutations can be even more restricted: while they all have tubular proteinuria, some may have no other recognisable tubular abnormality, not even hypercalciuria, which was considered an invariant feature of this disease (Ludwig et al., 2006). Therefore, if renal Fanconi syndrome is defined exclusively as a generalized proximal tubular dysfunction affecting all transport pathways, Dent disease and Lowe syndrome do not strictly or consistently fulfil this diagnostic criterion.

Unspecified renal Fanconi syndromes Renal Fanconi syndrome can also occur with non-specific damage to the proximal tubule, for instance from ischaemia (Ashworth and Molitoris, 1999). The high-transport activity in this tubular segment makes it particularly susceptible to damage from deprivation of metabolic fuel and oxygen. Renal Fanconi syndrome can also occur in monoclonal gammopathies (Kleta et al., 2004), thought to be a result of tubular obstruction from aggregated light chains (myeloma casts) and/or intracellular crystal formation due to incomplete lysosomal digestion of specific light chains (Lajoie et al., 2000; Messiaen et al., 2000; Kobayashi et al., 2006; Vanmassenhove et al., 2010). Similarly, the development of renal Fanconi syndrome in Fanconi–Bickel syndrome (systemic glycogen storage disorder due to autosomal recessive GLUT2—a passive basolateral glucose transporter—deficiency) is not understood.

415

416

Section 2  

fluid, electrolyte, and renal tubular disorders

Table 41.3  Acquired causes of the renal Fanconi syndrome Anti-cancer drugs Ifosfamide (Burk et al., 1990; Pratt et al., 1991; Rossi et al., 1992, 1999a, 1999b; Rossi and Ehrich, 1993; Negro et al., 1998; Ciarimboli et al., 2011) Streptozocin (Sadoff, 1970; Kintzel, 2001) Antibiotics Aminoglycoside (Melnick et al., 1994; Gainza et al., 1997; Alexandridis et al., 2003; Ghiculescu and Kubler, 2006; Zietse et al., 2009) Expired tetracyclines (Cleveland et al., 1965; Guggenbichler and Schabel, 1979; Montoliu et al., 1981; Zietse et al., 2009) Antiretrovirals Adefovir/cidofovir/tenofovir (Vittecoq et al., 1997; Verhelst et al., 2002; Malik et al., 2005; Hussain et al., 2006; Izzedine et al., 2006; Woodward et al., 2009; Jhaveri et al., 2010; Girgis et al., 2011; Hall et al., 2011) Dideoxyinosine (Crowther et al., 1993; Izzedine et al., 2005) Heavy metals Lead poisoning (Chisolm, 1968; Barbier et al., 2005) Cadmium (Kazantzis et al., 1963; Thevenod, 2003; Barbier et al., 2005; Rago et al., 2010) Others

acidosis with sodium and/or potassium bicarbonate. Often this is prescribed in the form of citrate:  each Mol of citrate is converted by the liver into 2 Mol of bicarbonate and it is thought that in addition to correcting the acidosis it may decrease the risk of nephrolithiasis by increasing urinary citrate (Nicar et al., 1984). Dosage is titrated to normalize plasma bicarbonate levels. Supplementation with potassium chloride in addition to potassium bicarbonate or citrate may be necessary to normalize plasma potassium levels. Sodium chloride may be needed in those patients with evidence of volume depletion. Phosphate supplementation is given to increase plasma phosphate levels, although normalization may not always be achievable, as dosage is limited by gastrointestinal side effects, especially diarrhoea, which may actually worsen the salt and fluid losses. A key complication of renal Fanconi syndrome is rickets (see above) and besides phosphate supplementation, the primary treatment is 1α-hydroxylated vitamin D. This can be in the form of 1,25(calcitriol) or 1-hydroxylated (1α-calcidol) vitamin D, although the latter is often preferred due to the longer half-life. Dosage should be titrated to achieve normalization of plasma calcium, PTH, and alkaline phosphatase. Constipation is a common complication, especially in younger children and a consequence of the renal fluid losses and should be treated by adequate fluid supplementation and laxatives, if needed.

Sodium valproate (Lande et al., 1993; Zaki and Springate, 2002)

Isolated glycosuria

Aristolochic acid (Chinese herb nephropathy) (Yang et al., 2002; Lee et al., 2004; Hong et al., 2006)

Isolated renal glycosuria is defined as renal glucose excretion above the normal range, that is, > 1.5 mmol/1.73 m2 per day (Elsas and Rosenberg, 1969), in the absence of other proximal tubular defects, pregnancy, or hyperglycaemia. Therefore, isolated renal glycosuria is caused by a specific defect in renal glucose transport. Glucose reabsorption in the proximal tubule is mediated by three key transporters: on the apical side are expressed at least two secondary active sodium-glucose cotransporters:  the high-affinity, low-capacity transporter SGLT1 (SLC5A1) and the low-affinity, high-capacity transporter SGLT2 (SLC5A2). Transport on the basolateral side is facilitated by GLUT2 (SLC2A2) (Brown, 2003). GLUT2 is also expressed in the liver and recessive mutations in GLUT2 cause a glycogen storage disorder called Fanconi–Bickel syndrome (Santer et  al., 1997). SGLT1 is also expressed in enterocytes and recessive mutations in this transporter cause glucose/galactose malabsorption (Turk et al., 1991). In contrast, SGLT2 is functionally expressed exclusively in the proximal tubule and mutations in this transporter are the basis of isolated renal glycosuria (van den Heuvel et al., 2002; Santer et al., 2003; Calado et al., 2004, 2006, 2008; Francis et  al., 2004; Kleta et  al., 2004). Inheritance can be dominant or recessive and there is some genotype/phenotype correlation: patients with only one mutant allele typically have milder (< 10 g/day/1.73 m2) or absent glycosuria, whereas those with two mutant alleles usually excrete more (Santer and Calado, 2010). Despite the definition of isolated renal glycosuria, other proximal tubular abnormalities can sometimes be seen in patients with defined SGLT2 mutations: some have accompanying hypercalciuria (Schneider et al., 1992; Scholl-Burgi et al., 2004) and others aminoaciduria (Gotzsche, 1977; Sankarasubbaiyan et  al., 2001). The reasons for this are not clear. However, since calcium reabsorption in the proximal tubule is passive and parallels sodium reabsorption, it can be speculated that the decreased sodium-glucose transport also impairs calcium re-uptake. Regarding the aminoaciduria,

Toluene/glue sniffing (Streicher et al., 1981) Fumaric acid (Fliegner and Spiegel, 1992; Raschka and Koch, 1999; Haring et al., 2011) Suramin (Rago et al., 1994) Paraquat (Vaziri et al., 1979; Gil et al., 2005) L-Lysine (Lo et al., 1996)

Treatment Treatment of the underlying cause Treatment of renal Fanconi syndrome should be aimed primarily at the underlying problem. Potentially causative toxins, such as heavy metals or aristolochic acid need to be identified and stopped. Potentially causative medications, such as aminoglycosides, valproate or tenofovir should be weaned off in close collaboration with the prescribing specialities and replaced by alternative drugs. However, for most of the genetic forms of renal Fanconi syndrome, there is no specific treatment available, with the notable exception of cystinosis (specific treatment with cysteamine) (Kleta and Gahl, 2004; Kleta et al., 2005), and potentially those mitochondrial cytopathies that respond to treatment with ubiquinone (Montini et al., 2008).

Supportive treatment In those patients awaiting normalization of proximal tubular function after removal of the underlying cause and in those forms of renal Fanconi syndrome without specific treatment, supportive treatment is needed. This includes treatment of the metabolic

chapter 41 

it has been hypothesized that glycosuria causes dissipation of the electric gradient of sodium-dependent amino acid transporters (Santer and Calado, 2010). While there may be some complications of isolated glycosuria, such as an increased incidence of urinary tract infections (De Marchi et al., 1983; De Paoli et al., 1984), episodes of dehydration/volume depletion (Calado et al., 2008), and ketosis during starvation (Oemar et al., 1987), it is usually considered a benign condition and not a disease and so no treatment is needed. Indeed, currently, with obesity and hypertension being some of the most serious threats to public health, renal sodium and glucose losses may be beneficial (109). Pharmacological blockade of SGLT2 is now a treatment option for type 2 diabetes (Oku et al., 1999; Santer and Calado, 2010). The role of another gene linked to glucose transport in the kidney, SLC16A12, has not been investigated yet (Kloeckener-Gruissem et al., 2008).

Aminoacidurias Virtually all filtered amino acids are reabsorbed in the proximal tubule by specialized transporters. The modern classification (solute carriers SLC) is based on the molecular identity of these transporters; prior to this a complicated system based on functional properties was used with letters and symbols indicating amino acid specificity, chemical properties of transported amino acids (acidic, neutral, basic) and whether transport was sodium-dependent or not (Camargo et al., 2012). For instance, the transporter involved in Hartnup disorder, now called SLC6A19, was previously called B(0) AT1, with the first letter indicating broad amino acid specificity the capital B indicating that transport was sodium-dependent, and the 0 indicating that it was selective for neutral amino acids (Kleta et al., 2004). The functional information contained in these names actually gives diagnostic clues, since aminoacidurias are characterized by the loss of specific amino acids in the urine (see below). Some transporters consist of two subunits, so-called heterodimeric amino acid transporters. For instance, the transporter involved in cystinuria consists of two subunits, now called SLC7A9 and SLC3A1 (Chillaron et al., 2010). SLC7A9, previously called b(0,+) AT, constitutes the actual transporter subunit with broad amino acid specificity (b), sodium-independence (lower case b) and selectivity for neutral (0) and basic (+) amino acids. SLC3A1 encodes a subunit necessary for proper trafficking of the transporter and is called rBAT for ‘related to BAT’ (Camargo et al., 2008).

Cystinuria (OMIM #220100) Cystinuria is characterized by an excessive excretion of the dibasic amino acids cystine, lysine, and ornithine. The key manifestation of cystinuria is urolithiasis and specific clinical and molecular features of cystinuria are described in Chapter 203.

Lysinuric protein intolerance (OMIM #222700) Lysinuric protein intolerance is a rare disorder with severe though non-specific clinical manifestations. The diagnosis is prompted by the recognition of excess urinary excretion of lysine, arginine and ornithine and the urinary amino acid findings are similar to cystinuria. However, in contrast to cystinuria, urinary cystine is only mildly elevated and urolithiasis is not a recognized feature of lysinuric protein intolerance (Sebastio et al., 2011). Whereas the transporter underlying cystinuria mediates apical uptake of dibasic amino acids, it is the basolateral export of these amino acids that is impaired in lysinuric protein intolerance. The underlying

renal fanconi syndrome

basolateral transporter is also heteromeric and consists of two subunits, SLC7A7 (y+-LAT1), the actual catalytic unit, and SLC3A2, as the associated heavy chain (Verrey et al., 2000). So far, only mutations in SLC7A7 have been identified in Lysinuric protein intolerance (Borsani et al., 1999; Torrents et al., 1999; Sebastio et al., 2011). This transporter is also expressed in enterocytes, leading to not only renal losses of dibasic amino acids, but also to impaired intestinal uptake (Sebastio et al., 2011). It is also expressed in the lungs, potentially explaining the pulmonary complications of the disease (see below) (Rotoli et al., 2005). Affected patients typically come to medical attention during infancy with non-specific symptoms, such as failure to thrive and developmental delay. Breastfeeding and infant formulas delay the onset, probably due to their low protein content, and so symptoms become apparent with the introduction of solids. Patients usually experience vomiting and diarrhoea after ingesting protein (hence the name). Later, symptoms as diverse as hepato-splenomegaly, interstitial lung disease with alveolar proteinosis, osteopenia, and chronic kidney disease and bone marrow abnormalities can develop (Parenti et al., 1995). The pathogenesis of these complications is thought to relate to unbalanced metabolism of specific amino acids, especially arginine (Palacin et al., 2004). This amino acid is involved in the urea cycle, potentially explaining the hyperammonaemia. However, ammonia is only raised directly after protein-rich meals, so that timing of diagnostic blood draws is critical. Arginine is also a precursor for nitric oxide (NO) and the manifold functions of NO may explain some of the divergent symptoms of Lysinuric protein intolerance (Sebastio et al., 2011).

Hartnup disorder (OMIM #234500) Hartnup disorder is named after the family in which this defect of neutral amino acid transport was first described in London, United Kingdom (Baron et  al., 1956). Interestingly, affected patients nowadays are typically asymptomatic and it was only the protein-restricted diet after World War II that precipitated the pellagra-like skin rash, cerebellar ataxia and psychosis-like symptoms that characterize this disorder and prompted its recognition. Hartnup disorder is caused by mutations in the transporter for neutral amino acids SLC6A19 (see above), which is expressed on the apical side of proximal tubular epithelial cells (Kleta et al., 2004). Biochemically, the disorder is characterized by excess excretion of neutral amino acids. Proline and glycine excretions are not increased. Plasma amino acid levels are usually in the normal range. Symptoms are thought to relate to a deficiency in tryptophan, which only manifests with a protein-restricted diet. Tryptophan is a precursor for niacin (pellagra-like rash) and serotonin (neurological symptoms). Consequently, treatment consists of a protein-rich diet. The skin rash can also be alleviated by niacin supplementation.

Iminoglycinuria (OMIM #242600) Iminoglycinuria is not linked to clinical manifestations, apart from an excess urinary excretion of glycine and the imino acids proline and hydroxyproline. Iminoglycinuria had previously been linked to mental retardation, deafness and visual impairment, but this probably just reflected an ascertainment bias, as the respective patient populations were screened by urinary amino acid determinations (Fraser et al., 1968; Rosenberg et al., 1968). The incidence of iminoglycinuria is estimated to be 1:10,000. Clinical heterogeneity of this disorder was suggested by the additional presence of an

417

418

Section 2  

fluid, electrolyte, and renal tubular disorders

intestinal uptake defect for imino acids in some patients (Rosenberg et al., 1968). The molecular basis of iminoglycinuria remains to be definitively resolved. Multiallelic mutations and polymorphisms in four genes, SLC36A2 (PAT-2), SLC6A20 (SIT-1), SLC6A19 (B0AT1), and SLC6A18 (B0AT3) have been suggested (Broer et al., 2008), but this remains controversial (Broer et al., 2012).

Dicarboxylic aminoaciduria (OMIM 222730) Dicarboxylic aminoaciduria refers to the excess excretion of the dicarboxylic amino acids aspartate and glutamate in the urine. It is an autosomal recessive condition caused by mutations in SLC1A1 (EAAC1/EAAT3) (Bailey et al., 2011). It has an estimated incidence of 1:35,000 (Auray-Blais et al., 2007). Like iminoglycinuria, it has been associated with intellectual impairment, but this may again reflect an association bias (Swarna et al., 1989, 2004), as dicarboxylic aminoaciduria has also been described in otherwise asymptomatic individuals (Melancon et al., 1977).

References Alexandridis, G., Liberopoulos, E., and Elisaf, M. (2003). Aminoglycoside-induced reversible tubular dysfunction. Pharmacology, 67(3), 118–20. Aronson, P. S. (2002). Ion exchangers mediating NaCl transport in the renal proximal tubule. Cell Biochem Biophys, 36(2–3), 147–53. Ashworth, S. L. and Molitoris, B. A. (1999). Pathophysiology and functional significance of apical membrane disruption during ischemia. Curr Opin Nephrol Hypertens, 8(4), 449–58. Attree, O., Olivos, I. M., Okabe, I., et al. (1992). The Lowe’s oculocerebrorenal syndrome gene encodes a protein highly homologous to inositol polyphosphate-5-phosphatase. Nature, 358(6383), 239–42. Auray-Blais, C., Cyr, D., and Drouin, R. (2007). Quebec neonatal mass urinary screening programme: from micromolecules to macromolecules. J Inherit Metab Dis, 30(4), 515–21. Bailey, C.G., Ryan, R. M., Thoeng, A. D., et al. (2011). Loss-of-function mutations in the glutamate transporter SLC1A1 cause human dicarboxylic aminoaciduria. J Clin Invest, 121(1), 446–53. Barbier, O., Jacquillet, G., Tauc, M., et al. (2005). Effect of heavy metals on, and handling by, the kidney. Nephron Physiol, 99(4), 105–10. Baron, D. N., Dent, C. E., Harris, H., et al. (1956). Hereditary pellagra-like skin rash with temporary cerebellar ataxia, constant renal amino-aciduria, and other bizarre biochemical features. Lancet, 271(6940), 421–8. Birn, H., Fyfe, J. C., Jacobsen, C., et al. (2000). Cubilin is an albumin binding protein important for renal tubular albumin reabsorption. J Clin Invest, 105(10), 1353–61. Bistarakis, L., Voskaki, I., Lambadaridis, J., et al. (1986). Renal handling of phosphate in the first six months of life. Arch Dis Child, 61(7), 677–81. Bobulescu, I. A. and Moe, O. W. (2009). Luminal Na+/H + exchange in the proximal tubule. Pflugers Archiv, 458(1), 5–21. Bockenhauer, D., Bokenkamp, A., van't Hoff, W., et al. (2008). Renal phenotype in Lowe Syndrome: a selective proximal tubular dysfunction. Clin J Am Soc Nephrol, 3(5), 1430–6. Bockenhauer, D., Cruwys, M., Kleta, R. et al. (2008). Antenatal Bartter’s syndrome: why is this not a lethal condition? QJM, 101(12), 927–42. Bockenhauer, D., Hug, M. J., and Kleta, R. (2009). Cystic fibrosis, aminoglycoside treatment and acute renal failure: the not so gentle micin. Pediatr Nephrol, 24(5), 925–8. Bokenkamp, A., Böckenhauer, D., Cheong, H. I., et al. (2009). Dent-2 disease: a mild variant of Lowe syndrome. J Pediatr, 155(1), 94–9. Borsani, G., Bassi, M. T., Sperandeo, M. P., et al. (1999). SLC7A7, encoding a putative permease-related protein, is mutated in patients with lysinuric protein intolerance. Nat Genet, 21(3), 297–301.

Broer, S., Bailey, C. G., Kowalczuk, S., et al. (2008). Iminoglycinuria and hyperglycinuria are discrete human phenotypes resulting from complex mutations in proline and glycine transporters. J Clin Invest, 118(12), 3881–92. Brown, G. K. (2000). Glucose transporters: structure, function and consequences of deficiency. J Inherit Metab Dis, 23(3), 237–46. Burk, C. D., Restaino, I., Kaplan, B. S., et al. (1990). Ifosfamide-induced renal tubular dysfunction and rickets in children with Wilms tumor. J Pediatr, 117(2 Pt 1), 331–5. Calado, J., Loeffler, J., Sakallioglu, O., et al. (2006). Familial renal glucosuria: SLC5A2 mutation analysis and evidence of salt-wasting. Kidney Int, 69(5), 852–5. Calado, J., Soto, K., Clemente, C., et al. (2004). Novel compound heterozygous mutations in SLC5A2 are responsible for autosomal recessive renal glucosuria. Hum Genet, 114(3), 314–16. Calado, J., Sznajer, Y., Metzger, D., et al. (2008). Twenty-one additional cases of familial renal glucosuria: absence of genetic heterogeneity, high prevalence of private mutations and further evidence of volume depletion. Nephrol Dial Transplant, 23(12), 3874–9. Camargo, S. M., Bockenhauer, D., and Kleta, R. (2008). Aminoacidurias: clinical and molecular aspects. Kidney Int, 73(8), 918–25. Camargo, S. M., Makrides, V., Kleta, R., et al. (2012). Kidney transport of amino acids and oligopeptides and aminoacidurias. In R. J. Alpern, O. W. Moe, and M. J. Caplan (eds.) Seldin and Giebisch’s The Kidney: Physiology and Pathophysiology, pp. 2405–23. San Diego, CA: Elsevier/Academic Press. Chillaron, J., Font-Llitjós, M., Fort, J., et al. (2010). Pathophysiology and treatment of cystinuria. Nat Rev Nephrol, 6(7), 424–34. Chisolm, J. J., Jr. (1968). The use of chelating agents in the treatment of acute and chronic lead intoxication in childhood. J Pediatr, 73(1), 1–38. Christensen, E. I. and Gburek, J. (2004). Protein reabsorption in renal proximal tubule-function and dysfunction in kidney pathophysiology. Pediatr Nephrol, 19(7), 714–21. Ciarimboli, G., Holle, S. K., Vollenbröcker, B., et al. (2011). New clues for nephrotoxicity induced by ifosfamide: preferential renal uptake via the human organic cation transporter 2. Mol Pharmaceut, 8(1), 270–9. Claverie-Martin, F., Ramos-Trujillo, E., and Garcia-Nieto, V. (2011). Dent’s disease: clinical features and molecular basis. Pediatr Nephrol, 26(5), 693–704. Cleveland, W. W., Adams, W. C., Mann, J. B., et al. (1965). Acquired Fanconi syndrome following degraded tetracycline. J Pediatr, 66, 333–42. Cogan, M.G. (1982). Disorders of proximal nephron function. Am J Med, 72(2), 275–88. Comper, W. D., Haraldsson, B., and Deen, W. M. (2008). Resolved: normal glomeruli filter nephrotic levels of albumin. J Am Soc Nephrol, 19(3), 427–32. Crowther, M. A., Callaghan, W., Hodsman, A. B., et al. (1993). Dideoxyinosine-associated nephrotoxicity. Aids, 7(1), 131–2. Cui, S., Guerriero, C. J., Szalinski, C. M., et al. (2010). OCRL1 function in renal epithelial membrane traffic. Am J Physiol Renal Physiol, 298(2), F335–45. De Marchi, S., Cecchin, E., Basile, A., et al. (1983). Is renal glycosuria a benign condition? Proc Eur Dial Transplant Assoc, 20, 681–5. De Paoli, P., Battistin, S., Jus, A., et al. (1984). Immunological characterization of renal glycosuria patients. Clin Exp Immunol, 56(2), 289–94. De Toni, G. (1933). Remarks on the relationship between renal rickets (renal dwarfism) and renal diabetes. Acta Pediatrica, 16, 479–84. Debre, R., Cleret, F., and Messimy, R. (1934). Rachitisme tardif coexistant avec une Nephrite chronique et une Glycosurie. Archive de Medicine des Enfants, 37, 597–606. Devuyst, O., Jouret, F., Auzanneau, C., et al. (2005). Chloride channels and endocytosis: new insights from Dent’s disease and ClC-5 knockout mice. Nephron Physiol, 99(3), 69–73. Earle, K. E., Seneviratne, T., Shaker, J., et al. (2004). Fanconi’s syndrome in HIV+ adults: report of three cases and literature review. J Bone Miner Res, 19(5), 714–21.

chapter 41 

Elsas, L. J. and Rosenberg, L. E. (1969). Familial renal glycosuria: a genetic reappraisal of hexose transport by kidney and intestine. J Clin Invest, 48(10), 1845–54. Enomoto, A., Kimura, H., Chairoungdua, A., et al. (2002). Molecular identification of a renal urate anion exchanger that regulates blood urate levels. Nature, 417(6887), 447–52. Erdmann, K. S., Mao, Y., McCrea, H. J., et al. (2007). A role of the Lowe syndrome protein OCRL in early steps of the endocytic pathway. Dev Cell, 13(3), 377–90. Fanconi, G. (1931). Die nicht diabetischen Glykosurien und Hyperglyklaemien des aelteren Kindes. Jahrbuch fuer Kinderheilkunde, 133, 257–300. Fanconi, G. (1936). Der fruehinfantile nephrotisch-glykosurische Zwergwuchs mit hypophosphataemischer Rachitis. Jahrbuch fuer Kinderheilkunde, 147, 299–338. Fliegner, L. and Spiegel, P. (1992). [Osteomalacia as an apparently rare side effect of oral fumaric acid therapy. Secondary DeToni-Debre Fanconi syndrome in the adult]. Der Hautarzt, 43(9), 554–60. Francis, J., Zhang, J., Farhi, A., et al. (2004). A novel SGLT2 mutation in a patient with autosomal recessive renal glucosuria. Nephrol Dial Transplant, 19(11), 2893–5. Fraser, G. R., Friedmann, A. I., Patton, V. M., et al. (1968). Iminoglycinuria—a ‘harmless’ inborn error of metabolism? Humangenetik, 6(4), 362–7. Gainza, F. J., Minguela, J. I., and Lampreabe, I. (1997). Aminoglycoside-associated Fanconi’s syndrome: an underrecognized entity. Nephron, 77(2), 205–11. Gekle, M. (2007). Renal albumin handling: a look at the dark side of the filter. Kidney Int, 71(6), 479–81. Ghiculescu, R. A. and Kubler, P. A. (2006). Aminoglycoside-associated Fanconi syndrome. Am J Kidney Dis, 48(6), e89–93. Gil, H. W., Yang, J. O., Lee, E. Y., et al. (2005). Paraquat-induced Fanconi syndrome. Nephrology, 10(5), 430–2. Girgis, C. M., Wong, T., Ngu, M. C., et al. (2011). Hypophosphataemic osteomalacia in patients on adefovir dipivoxil. J Clin Gastroenterol, 45(5), 468–73. Gonick, H. C. (2008). Nephrotoxicity of cadmium & lead. Indian J Med Res, 128(4), 335–52. Gottschalk, C. W. (1961). Fifth Bowditch lecture. Micropuncture studies of tubular function in the mammalian kidney. Physiologist, 4(1), 35–55. Gotzsche, O. (1977). Renal glucosuria and aminoaciduria. Acta Med Scand, 202(1–2), 65–7. Guggenbichler, J. P. and Schabel, F. (1979). [Tetracyclin intoxication versus idiopathic pancreatitis: report of a case with multiple organ involvement (author’s transl.)]. Padiatr Padol, 14(3), 267–72. Guggino, S. E. (2007). Mechanisms of disease: what can mouse models tell us about the molecular processes underlying Dent disease? Nat Clin Pract Nephrol, 3(8), 449–55. Hall, A. M., Hendry, B. M., Nitsch, D., et al. (2011). Tenofovir-associated kidney toxicity in HIV-infected patients: a review of the evidence. Am J Kidney Dis, 57(5), 773–80. Haring, N., Mähr, H. S., Mündle, M., et al. (2011). Early detection of renal damage caused by fumaric acid ester therapy by determination of urinary beta2-microglobulin. Br J Dermatol, 164(3), 648–51. Hoenderop, J. G., Nilius, B., and Bindels, R. J. (2005). Calcium absorption across epithelia. Physiol Rev, 85(1), 373–422. Hong, Y. T., Fu, L. S., Chung, L. H., et al. (2006). Fanconi’s syndrome, interstitial fibrosis and renal failure by aristolochic acid in Chinese herbs. Pediatr Nephrol, 21(4), 577–9. Hoopes, R. R., Jr., Shrimpton, A. E., Knohl, S. J., et al. (2005). Dent disease with mutations in OCRL1. Am J Hum Genet, 76(2), 260–7. Hryciw, D. H., Ekberg, J., Pollock, C. A., et al. (2006). ClC-5: a chloride channel with multiple roles in renal tubular albumin uptake. Int J Biochem Cell Biol, 38(7), 1036–42. Hussain, S., Khayat, A., Tolaymat, A., et al. (2006). Nephrotoxicity in a child with perinatal HIV on tenofovir, didanosine and lopinavir/ritonavir. Pediatr Nephrol, 21(7), 1034–6.

renal fanconi syndrome

Izzedine, H., Hulot, J. S., Villard, E., et al. Association between ABCC2 gene haplotypes and tenofovir-induced proximal tubulopathy. J Infect Dis, 194(11), 1481–91. Izzedine, H., Launay-Vacher, V., and Deray, G. (2005). Fanconi syndrome associated with didanosine therapy. Aids, 19(8), 844–5. Jhaveri, M. A., Mawad, H. W., Thornton, A. C., et al. (2010). Tenofovir-associated severe bone pain: I cannot walk! J Int Assoc Physicians AIDS Care, 9(5), 328–34. Kazama, I., Matsubara, M., Michimata, M., et al. (2004). Adult onset Fanconi syndrome: extensive tubulo-interstitial lesions and glomerulopathy in the early stage of Chinese herbs nephropathy. Clin Exp Nephrol, 8(3), 283–7. Kazantzis, G., Flynn, F. V., Spowage, J. S., et al. (1963). Renal tubular malfunction and pulmonary emphysema in cadmium pigment workers. QJM, 32, 165–92. Kintzel, P. E. (2001). Anticancer drug-induced kidney disorders. Drug Saf, 24(1), 19–38. Kleta, R. (2008). Fanconi or not Fanconi? Lowe syndrome revisited. Clin J Am Soc Nephrol, 3(5), 1244–5. Kleta, R. and Gahl, W. A. (2004). Pharmacological treatment of nephropathic cystinosis with cysteamine. Expert Opin Pharmacother, 5(11), 2255–62. Kleta, R., Blair, S. C., Bernardini, I., et al. (2004). Keratopathy of multiple myeloma masquerading as corneal crystals of ocular cystinosis. Mayo Clinic Proc, 79(3), 410–2. Kleta, R., Kaskel, F., Dohil, R., et al. (2005). First NIH/Office of Rare Diseases Conference on Cystinosis: past, present, and future. Pediatr Nephrol, 20(4), 452–4. Kleta, R., Romeo, E., Ristic, Z., et al. (2004). Mutations in SLC6A19, encoding B0AT1, cause Hartnup disorder. Nat Genet, 36(9), 999–1002. Kleta, R., Stuart, C., Gill, F. A., et al. (2004). Renal glucosuria due to SGLT2 mutations. Mol Genet Metab, 82(1), 56–8. Kloeckener-Gruissem, B., Vandekerckhove, K., Nürnberg, G., et al. (2008). Mutation of solute carrier SLC16A12 associates with a syndrome combining juvenile cataract with microcornea and renal glucosuria. Am J Hum Genet, 82(3), 772–9. Kobayashi, T., Muto, S., Nemoto, J., et al. (2006). Fanconi’s syndrome and distal (type 1) renal tubular acidosis in a patient with primary Sjogren’s syndrome with monoclonal gammopathy of undetermined significance. Clin Nephrol, 65(6), 427–32. Kruse, K., Kracht, U., and Gopfert, G. (1982). Renal threshold phosphate concentration (TmPO4/GFR). Arch Dis Child, 57(3), 217–23. Kuwertz-Broking, E., Koch, H. G., Marquardt, T., et al. (2000). Renal Fanconi syndrome: first sign of partial respiratory chain complex IV deficiency. Pediatr Nephrol, 14(6), 495–8. Lajoie, G., Leung, R., and Bargman, J. M. (2000). Clinical, biochemical, and pathological features in a patient with plasma cell dyscrasia and Fanconi syndrome. Ultrastruct Pathol, 24(4), 221–6. Lande, M. B., Kim, M. S., Bartlett, C., et al. (1993). Reversible Fanconi syndrome associated with valproate therapy. J Pediatr, 123(2), 320–2. Lassiter, W. E., Gottschalk, C. W., and Mylle, M. (1961). Micropuncture study of net transtubular movement of water and urea in nondiuretic mammalian kidney. Am J Physiol, 200, 1139–47. Lee, S., Lee, T., Lee, B., et al. (2004). Fanconi’s syndrome and subsequent progressive renal failure caused by a Chinese herb containing aristolochic acid. Nephrology, 9(3), 126–9. Lin, Z., Jin, S., Duan, X., et al. (2011). Chloride channel (Clc)-5 is necessary for exocytic trafficking of Na+/H+ exchanger 3 (NHE3). J Biol Chem, 286(26), 22833–45. Lloyd, S. E., Pearce, S. H., Fisher, S. E., et al. (1996). A common molecular basis for three inherited kidney stone diseases. Nature, 379(6564), 445–9. Lo, J. C., Chertow, G. M., Rennke, H., et al. (1996). Fanconi’s syndrome and tubulointerstitial nephritis in association with L-lysine ingestion. Am J Kidney Dis, 28(4), 614–17. Ludwig, M., Utsch, B., Balluch, B., et al. (2006). Hypercalciuria in patients with CLCN5 mutations. Pediatr Nephrol, 21(9), 1241–50.

419

420

Section 2  

fluid, electrolyte, and renal tubular disorders

Magen, D., Berger, L., Coady, M. J., et al. A loss-of-function mutation in NaPi-IIa and renal Fanconi’s syndrome. N Engl J Med, 362(12), 1102–9. Magen, D., Sprecher, E., Zelikovic, I., et al. (2005). A novel missense mutation in SLC5A2 encoding SGLT2 underlies autosomal-recessive renal glucosuria and aminoaciduria. Kidney Int, 67(1), 34–41. Malik, A., Abraham, P., and Malik, N. (2005). Acute renal failure and Fanconi syndrome in an AIDS patient on tenofovir treatment—case report and review of literature. J Infect, 51(2), E61–5. Mao, Y., Balkin, D.M., Zoncu, R., et al. (2009). A PH domain within OCRL bridges clathrin-mediated membrane trafficking to phosphoinositide metabolism. EMBO J, 28(13), 1831–42. Matsuo, H., Chiba, T., Nagamori, S., et al. (2008). Mutations in glucose transporter 9 gene SLC2A9 cause renal hypouricemia. Am J Hum Genet, 83(6), 744–51. Melancon, S.B., Dallaire, L., Lemieux, B., et al. (1977). Dicarboxylic aminoaciduria: an inborn error of amino acid conservation. J Pediatr, 91(3), 422–7. Melnick, J. Z., Baum, M., and Thompson, J. R. (1994). Aminoglycoside-induced Fanconi’s syndrome. Am J Kidney Dis, 23(1), 118–22. Messiaen, T., Deret, S., Mougenot, B., et al. (2000). Adult Fanconi syndrome secondary to light chain gammopathy. Clinicopathologic heterogeneity and unusual features in 11 patients. Medicine, 79(3), 135–54. Montini, G., Malaventura, C., and Salviati, L. (2008). Early coenzyme Q10 supplementation in primary coenzyme Q10 deficiency. N Engl J Med, 358(26), 2849–50. Montoliu, J., Carrera, M., Darnell, A., et al. (1981). Lactic acidosis and Fanconi’s syndrome due to degraded tetracycline. Br Med J, 283(6306), 1576–7. Mourmans, J., Wendel, U., Bentlage, H. A., et al. (1997). Clinical heterogeneity in respiratory chain complex III deficiency in childhood. J Neurol Sci, 149(1), 111–17. Negri, A. L. (2006). Proximal tubule endocytic apparatus as the specific renal uptake mechanism for vitamin D-binding protein/25-(OH)D3 complex. Nephrology, 11(6), 510–15. Negro, A., Regolisti, G., Perazzoli, F., et al. (1998). Ifosfamide-induced renal Fanconi syndrome with associated nephrogenic diabetes insipidus in an adult patient. Nephrol Dial Transplant, 13(6), 1547–9. Neiberger, R. E., George, J. C., Perkins, L. A., et al. (2002). Renal manifestations of congenital lactic acidosis. Am J Kidney Dis, 39(1), 12–23. Niaudet, P. and Rotig, A. (1996). Renal involvement in mitochondrial cytopathies. Pediatr Nephrol, 10(3), 368–73. Nicar, M.J., Peterson, R., and Pak, C. Y. (1984). Use of potassium citrate as potassium supplement during thiazide therapy of calcium nephrolithiasis. J Urol, 131(3), 430–3. Norden, A. G., Lapsley, M., Lee, P. J., et al. (2001). Glomerular protein sieving and implications for renal failure in Fanconi syndrome. Kidney Int, 60(5), 1885–92. Norden, A. G., Sharratt, P., Cutillas, P. R., et al. (2004). Quantitative amino acid and proteomic analysis: very low excretion of polypeptides >750 Da in normal urine. Kidney Int, 66(5), 1994–2003. Norden, A.G., Scheinman, S. J., Deschodt-Lanckman, M. M., et al. (200). Tubular proteinuria defined by a study of Dent’s (CLCN5 mutation) and other tubular diseases. Kidney Int, 57(1), 240–9. Oemar, B. S., Byrd, D. J., and Brodehl, J. (1987). Complete absence of tubular glucose reabsorption: a new type of renal glucosuria (type 0). Clin Nephrol, 27(3), 156–60. Oku, A., Ueta, K., Arakawa, K., et al. (1999). T-1095, an inhibitor of renal Na+-glucose cotransporters, may provide a novel approach to treating diabetes. Diabetes, 48(9), 1794–800. Ooms, L. M., Horan, K. A., Rahman, P., et al. (2009). The role of the inositol polyphosphate 5-phosphatases in cellular function and human disease. Biochem J, 419(1), 29–49. Ovunc, B., Otto, E. A., Vega-Warner, V., et al. (2011). Exome sequencing reveals cubilin mutation as a single-gene cause of proteinuria. J Am Soc Nephrol, 22(10), 1815–20.

Pajor, A. M. (1999). Citrate transport by the kidney and intestine. Semin Nephrol, 19(2), 195–200. Palacin, M., Bertran, J., Chillarón, J., et al. (2004). Lysinuric protein intolerance: mechanisms of pathophysiology. Mol Genet Metab, 81 Suppl 1, S27–37. Parenti, G., Sebastio, G., Strisciuglio, P., et al. (1995). Lysinuric protein intolerance characterized by bone marrow abnormalities and severe clinical course. J Pediatr, 126(2), 246–51. Pollock, C. A. and Poronnik, P. (2007). Albumin transport and processing by the proximal tubule: physiology and pathophysiology. Curr Opin Nephrol Hypertens, 16(4), 359–64. Pratt, C.B., Meyer, W. H., Jenkins, J. J., et al. (1991). Ifosfamide, Fanconi’s syndrome, and rickets. J Clin Oncol, 9(8), 1495–9. Rago, N., Jacquillet, G., and Unwin, R. (2010). Heavy metal poisoning: the effects of cadmium on the kidney. Biometals, 23(5), 783–92. Rago, R. P., Miles, J. M., Sufit, R. L., et al. (1994). Suramin-induced weakness from hypophosphatemia and mitochondrial myopathy. Association of suramin with mitochondrial toxicity in humans. Cancer, 73(7), 1954–9. Raschka, C. and Koch, H. J. (1999). Longterm treatment of psoriasis using fumaric acid preparations can be associated with severe proximal tubular damage. Hum Exper Toxicol, 18(12), 738–9. Roch-Ramel, F. (1998). Renal transport of organic anions. Curr Opin Nephrol Hypertens, 7(5), 517–24. Rosenberg, L. E., Durant, J. L., and Elsas, L. J. (1968). Familial iminoglycinuria. An inborn error of renal tubular transport. N Engl J Med, 278(26), 1407–13. Rossi, R. and Ehrich, J. H. (1993). Partial and complete de Toni-Debre-Fanconi syndrome after ifosfamide chemotherapy of childhood malignancy. Eur J Clin Pharmacol, 44 Suppl 1, S43–5. Rossi, R., Helmchen, U., and Schellong, G. (1992). Tubular function and histological findings in ifosfamide-induced renal Fanconi syndrome—a report of two cases. Eur J Pediatr, 151(5), 384–7. Rossi, R., Kleta, R., and Ehrich, J. H. (1999a). Renal involvement in children with malignancies. Pediatr Nephrol, 13(2), 153–62. Rossi, R., Pleyer, J., Schäfers, P., et al. (1999b). Development of ifosfamide-induced nephrotoxicity: prospective follow-up in 75 patients. Med Pediatr Oncol, 32(3), 177–82. Rothstein, M., Obialo, C., and Hruska, K. A. (1990). Renal tubular acidosis. Endocrinol Metab Clin North Am, 19(4), 869–87. Rotig, A. (2003). Renal disease and mitochondrial genetics. J Nephrol, 16(2), 286–92. Rotoli, B.M., Bussolati, O., Sala, R., et al. (2005). The transport of cationic amino acids in human airway cells: expression of system y+L activity and transepithelial delivery of NOS inhibitors. FASEB J, 19(7), 810–2. Russo, L. M., Sandoval, R. M., McKee, M., et al. (2007). The normal kidney filters nephrotic levels of albumin retrieved by proximal tubule cells: retrieval is disrupted in nephrotic states. Kidney Int, 71(6), 504–13. Sadoff, L. (1970). Nephrotoxicity of streptozotocin (NSC-85998). Cancer Chemother Rep, 54(6), 457–9. Sankarasubbaiyan, S., Cooper, C., and Heilig, C. W. (2001). Identification of a novel form of renal glucosuria with overexcretion of arginine, carnosine, and taurine. Am J Kidney Dis, 37(5), 1039–43. Santer, R. and Calado, J. (2010). Familial renal glucosuria and SGLT2: from a mendelian trait to a therapeutic target. Clin J Am Soc Nephrol, 5(1), 133–41. Santer, R., Kinner, M., Lassen, C. L., et al. (2003). Molecular analysis of the SGLT2 gene in patients with renal glucosuria. J Am Soc Nephrol, 14(11), 2873–82. Santer, R., Schneppenheim, R., Dombrowski, A., et al. (1997). Mutations in GLUT2, the gene for the liver-type glucose transporter, in patients with Fanconi-Bickel syndrome. Nat Genet, 17(3), 324–6. Schneider, D., Gauthier, B., and Trachtman, H. (1992). Hypercalciuria in children with renal glycosuria: evidence of dual renal tubular reabsorptive defects. J Pediatr, 121(5 Pt 1), 715–19. Scholl-Burgi, S., Santer, R., and Ehrich, J. H. (2004). Long-term outcome of renal glucosuria type 0: the original patient and his natural history. Nephrol Dial Transplant, 19(9), 2394–6.

chapter 41 

Sebastian, A., McSherry, E., and Morris, R. C. Jr. (1971). On the mechanism of renal potassium wasting in renal tubular acidosis associated with the Fanconi syndrome (type 2 RTA). J Clin Invest, 50(1), 231–43. Sebastio, G., Sperandeo, M. P., and Andria, G. (2011). Lysinuric protein intolerance: reviewing concepts on a multisystem disease. Am J Med Genet C Semin Med Genet, 157(1), 54–62. Shaw, N. J., Wheeldon, J., and Brocklebank, J. T. (1990). Indices of intact serum parathyroid hormone and renal excretion of calcium, phosphate, and magnesium. Arch Dis Child, 65(11), 1208–11. Sirac, C., Bridoux, F., Essig, M., et al. (2011). Toward understanding renal Fanconi syndrome: step by step advances through experimental models. Contrib Nephrol, 169, 247–61. Spahn, C. M. and Prescott, C. D. (1996). Throwing a spanner in the works: antibiotics and the translation apparatus. J Mol Med, 74(8), 423–39. Sperl, W., Ruitenbeek, W., Trijbels, J. M., et al. (1988). Mitochondrial myopathy with lactic acidaemia, Fanconi-De Toni-Debre syndrome and a disturbed succinate: cytochrome c oxidoreductase activity. Eur J Pediatr, 147(4), 418–21. Streicher, H. Z., Gabow, P. A., Moss, A. H., et al. (1981). Syndromes of toluene sniffing in adults. Ann Internal Med, 94(6), 758–62. Suki, W. N. (1979). Calcium transport in the nephron. Am J Physiol, 237(1), F1–6. Superti-Furga, A., Schoenle, E., Tuchschmid, P., et al. (1993). Pearson bone marrow-pancreas syndrome with insulin-dependent diabetes, progressive renal tubulopathy, organic aciduria and elevated fetal haemoglobin caused by deletion and duplication of mitochondrial DNA. Eur J Pediatr, 152(1), 44–50. Swarna, M., Jyothy, A., Usha Rani, P., et al. (2004). Amino acid disorders in mental retardation: a two-decade study from Andhra Pradesh. Biochem Genet, 42(3–4), 85–98. Swarna, M., Rao, D. N., and Reddy, P. P. (1989). Dicarboxylic aminoaciduria associated with mental retardation. Hum Genet, 82(3), 299–300. Thakker, R. V. (2000). Pathogenesis of Dent’s disease and related syndromes of X-linked nephrolithiasis. Kidney Int, 57(3), 787–93. Thevenod, F. (2003). Nephrotoxicity and the proximal tubule. Insights from cadmium. Nephron Physiol, 93(4), 87–93. Tolaymat, A., Sakarcan, A., and Neiberger, R. (1992). Idiopathic Fanconi syndrome in a family. Part I. Clinical aspects. J Am Soc Nephrol, 2(8), 1310–17. Torrents, D., Mykkänen, J., Pineda, M., et al. (1999). Identification of SLC7A7, encoding y+LAT-1, as the lysinuric protein intolerance gene. Nat Genet, 21(3), 293–6. Turk, E., Zabel, B., Mundlos, S., et al. (1991). Glucose/galactose malabsorption caused by a defect in the Na+/glucose cotransporter. Nature, 350(6316), 354–6. Unwin, R. J., Capasso, G., and Shirley, D. G. (2004). An overview of divalent cation and citrate handling by the kidney. Nephron Physiol, 98(2), 15–20.

renal fanconi syndrome

Utsch, B., Bökenkamp, A., Benz, M. R., et al. (2006). Novel OCRL1 mutations in patients with the phenotype of Dent disease. Am J Kidney Dis, 48(6), 942–56. Van Biervliet, J. P., Bruinvis, L., Ketting, D., et al. (1977). Hereditary mitochondrial myopathy with lactic acidemia, a De Toni-Fanconi-Debre syndrome, and a defective respiratory chain in voluntary striated muscles. Pediatr Res, 11(10 Pt 2), 1088–93. Van den Heuvel, L. P., Assink, K., Willemsen, M., et al. (2002). Autosomal recessive renal glucosuria attributable to a mutation in the sodium glucose cotransporter (SGLT2). Hum Genet, 111(6), 544–7. Vanmassenhove, J., Sallée, M., Guilpain, P., et al. (2010). Fanconi syndrome in lymphoma patients: report of the first case series. Nephrol Dial Transplant, 25(8), 2516–20. Vaziri, N. D., Ness, R. L., Fairshter, R. D., et al. (1979). Nephrotoxicity of paraquat in man. Arch Intern Med, 139(2), 172–4. Verhelst, D., Monge, M., Meynard, J. L., et al. (2002). Fanconi syndrome and renal failure induced by tenofovir: a first case report. Am J Kidney Dis, 40(6), 1331–3. Verrey, F., Meier, C., Rossier, G., et al. (2000). Glycoprotein-associated amino acid exchangers: broadening the range of transport specificity. Pflugers Archiv, 440(4), 503–12. Vitart, V., Rudan, I., Hayward, C., et al. (2008). SLC2A9 is a newly identified urate transporter influencing serum urate concentration, urate excretion and gout. Nat Genet, 40(4), 437–42. Vittecoq, D., et al. (1997). Fanconi syndrome associated with cidofovir therapy. Antimicrob Agents Chemother, 41(8), 1846. Walton, R. J. and Bijvoet, O. L. (1975). Nomogram for derivation of renal threshold phosphate concentration. Lancet, 2(7929), 309–10. Wang, Y., Cai, H., Cebotaru, L., et al. (2005). ClC-5: role in endocytosis in the proximal tubule. Am J Physiol Renal Physiol, 289(4), F850–62. Wendel, U., Ruitenbeek, W., Bentlage, H. A., et al. (1995). Neonatal De Toni-Debre-Fanconi syndrome due to a defect in complex III of the respiratory chain. Eur J Pediatr, 154(11), 915–18. Woodward, C. L., Hall, A. M., Williams, I. G., et al. (2009). Tenofovir-associated renal and bone toxicity. HIV Med, 10(8), 482–7. Wright, F. S. and Giebisch, G. (1978). Renal potassium transport: contributions of individual nephron segments and populations. Am J Physiol, 235(6), F515–27. Yang, S. S., Chu, P., Lin, Y. F., et al. (2002). Aristolochic acid-induced Fanconi’s syndrome and nephropathy presenting as hypokalemic paralysis. Am J Kidney Dis, 39(3), E14. Zaki, E. L. and Springate, J. E. (2002). Renal injury from valproic acid: case report and literature review. Pediatr Neurol, 27(4), 318–19. Zietse, R., Zoutendijk, R., and Hoorn, E. J. (2009). Fluid, electrolyte and acid-base disorders associated with antibiotic therapy. Nat Rev Nephrol, 5(4), 193–202.

421

SECTION 3

The patient with glomerular disease

42 The glomerulus and the concept of glomerulonephritis  425 Alexander Woywodt and Diana Chiu

43 The renal glomerulus: the structural basis of ultrafiltration  436 Marlies Elger and Wilhelm Kriz

44 Function of the normal glomerulus  451 Jean-Claude Dussaule, Martin Flamant, and Christos Chatziantoniou

45 Mechanisms of glomerular injury: overview  459 Neil Turner

46 The patient with haematuria  463 John Neary and Neil Turner

47 Loin pain haematuria syndrome  469 John Neary

48 Nutcracker syndrome and phenomenon  473 John Neary and Neil Turner

49 Exercise-related pseudonephritis  476 Neil Turner

50 Proteinuria 478 Neil Turner and Stewart Cameron

51 Postural proteinuria (benign orthostatic proteinuria)  485 Neil Turner

52 Nephrotic syndrome  487 Premil Rajakrishna, Stewart Cameron, and Neil Turner

53 Pathophysiology of oedema in nephrotic syndrome  496 Neil Turner and Premil Rajakrishna

54 Idiopathic nephrotic syndrome: overview  499 Patrick Niaudet and Alain Meyrier

55 Minimal change disease: clinical features and diagnosis  501 Patrick Niaudet and Alain Meyrier

56 Minimal change disease: treatment and outcome  506 Patrick Niaudet and Alain Meyrier

57 Primary focal segmental glomerulosclerosis: clinical features and diagnosis  515 Alain Meyrier and Patrick Niaudet

58 Primary focal segmental glomerulosclerosis: treatment and outcome  525 Alain Meyrier and Patrick Niaudet

59 Pathogenesis of proteinuria in minimal change disease and focal segmental glomerulosclerosis  533 Patrick Niaudet and Alain Meyrier

60 Membranous glomerulonephritis: overview  537 Daniel C. Cattran and Heather N. Reich

61 Membranous glomerulonephritis: clinical features and diagnosis  539 Daniel C. Cattran and Heather N. Reich

424

62 Membranous glomerulonephritis: treatment and outcome  544 Daniel C. Cattran and Heather N. Reich

63 Secondary membranous glomerulonephritis  557 Daniel C. Cattran and Heather N. Reich

64 Membranous glomerulonephritis: pathogenesis  560 Daniel C. Cattran and Heather N. Reich

65 Immunoglobulin A nephropathy: overview  565 Kar Neng Lai and Sydney C. W. Tang

66 Immunoglobulin A nephropathy: clinical features  566 Kar Neng Lai and Sydney C. W. Tang

67 Immunoglobulin A nephropathy: diagnosis  572 Kar Neng Lai and Sydney C. W. Tang

68 Immunoglobulin A nephropathy: treatment and outcome  577 Kar Neng Lai and Sydney C. W. Tang

69 Immunoglobulin A nephropathy: pathogenesis  586 Kar Neng Lai and Sydney C. W. Tang

70 Crescentic (rapidly progressive) glomerulonephritis  592 Neil Turner

71 Antiglomerular basement membrane disease: overview  598 Zhao Cui, Neil Turner, and Ming-hui Zhao

72 Antiglomerular basement membrane disease: clinical features and diagnosis  599 Zhao Cui, Neil Turner, and Ming-hui Zhao

73 Antiglomerular basement membrane disease: treatment and outcome  606 Zhao Cui, Neil Turner, and Ming-hui Zhao

74 Antiglomerular basement membrane disease: pathogenesis  609 Zhao Cui, Neil Turner, and Ming-hui Zhao

75 Alport post-transplant antiglomerular basement membrane disease  619 Zhao Cui, Neil Turner, and Ming-hui Zhao

76 Post-infectious glomerulonephritis: overview  621 Bernardo Rodriguez-Iturbe and Mark Haas

77 Post-streptococcal glomerulonephritis  623 Bernardo Rodríguez-Iturbe and Mark Haas

78 Immunoglobulin A-dominant post-infectious glomerulonephritis  633 Bernardo Rodriguez-Iturbe and Mark Haas

79 Glomerulonephritis associated with endocarditis, deep-seated infections, and shunt nephritis  636 Bernardo Rodriguez-Iturbe and Mark Haas

80 Membranoproliferative glomerulonephritis and C3 glomerulopathy  641 Daniel P. Gale and Terry Cook

81 Fibrillary and immunotactoid glomerulopathy  649 Stephen M. Korbet, Melvin M. Schwartz, and Edmund J. Lewis

82 Drug-induced and toxic glomerulopathies  656 Alexander Woywodt and Diana Chiu

CHAPTER 42

The glomerulus and the concept of glomerulonephritis Alexander Woywodt and Diana Chiu Introduction Glomerular diseases are the pathological processes most often found on native renal biopsy (Jennette et al., 2007). Their clinical presentation is varied, ranging from mild proteinuria detected as a chance finding in an entirely asymptomatic patient to rapidly progressive renal failure in the context of life-threatening systemic disease. Some patterns of glomerular disease on biopsy are characteristically associated with, but not specific to, each of these clinical presentations. These six clinical syndromes and common correlates on renal biopsy provide a convenient and appropriate first approach to glomerular disease (Table 42.1).

The evolution of the concept of glomerular disease The concept that the kidneys produce urine emerged relatively late in antiquity. Egyptian medicine around 2000 BC first toyed with the idea that urine must be produced somewhere in the bladder area (Stratta et al., 1999). It took as long as another one and a half millennia until Hippocrates of Kos (460–377 BC) observed urine abnormalities, such has floating bubbles, in patients of whom we must assume had some form of glomerular disease (Eknoyan, 1988). Only much later did Pliny (23–79 AD) and Galen (130–200 AD) recognize and state much more explicitly the role of the kidneys in producing urine (Stratta et al., 1999). It is clear from his books that Pliny saw and recognized patients with blood in the urine, with peripheral oedema and also advocated various potions for treatment (De Santo et al., 1989). Rufus of Ephesus, in the first century AD, also described haematuria in the first ever textbook on diseases of the kidneys (Eknoyan, 2002) and we have to assume that at least some of the cases described had glomerular disease. Not much progress was made in Europe in subsequent centuries and the medieval understanding of the kidney and urine essentially relied on concepts and ideas developed in antiquity. What progress there was relied very much on the contribution of Arabic physicians, such as Rhazes (865–925 AD) (Changizi Ashtiyani and Cyrus, 2010), who described haematuria that we have to assume was glomerular in origin (Changizi Ashtiyani and Cyrus, 2010). Similarly, Avicenna (980–1037 AD) had a keen interest in the kidney and possessed a good concept of the role of the kidneys, akin to the ideas of Pliny and Galen, although much of his interest was on urological problems (Changizi Ashtiyani et al., 2011). We must also credit these Arabic physician scholars for the taking their interest in urine

analysis into clinical practice, as is demonstrated by textbook illustrations from the period (Fig. 42.1) (Eknoyan, 1994). This focus on urine analysis continued to be a main theme in medieval medicine and made its way from the consulting rooms of Arabic physicians to clinical practice all over Europe. Driven in particular by the Salernitan School of Medicine (800–1400 AD), the visual inspection of urine (uroscopy) thus gained popularity throughout medieval Europe as evidenced by the widespread use of urine charts (Fig. 42.2). It is quite remarkable just how detailed these urine charts were in terms of different shades of colour and their presumed diagnostic significance (Diskin, 2008). It must be said, however, that uroscopy was viewed largely as a way to deduce changes in the composition of body fluids (humores), not as a diagnostic tool in kidney disease. Therefore, although some urine samples shown in Fig. 42.2 are suggestive of glomerulonephritis (GN), the link was clearly not understood. It has also been suggested that uroscopy was both over-used and also abused by charlatans and impostors (Stratta et  al., 1999). Interestingly, the technique of urine collection was also thought to be important:  Ismail of Jurjani, an eleventh-century Persian physician, first recommended the 24-hour urine collection we request today (Armstrong, 2006). Oedema was also occasionally observed and the term ‘dropsy’ was in widespread use as an umbrella term for an assortment of underlying diseases of the heart, liver, and kidney (Stratta et al., 1999). A more detailed concept of glomerular disease (George, 2003) now required both substantial progress in terms of anatomy and histology of the kidney and a more detailed way of investigating abnormal urine. In 1664, Dekkers in Leiden first added acetic acid to urine and observed a milk-like precipitant (Cameron, 2003), although he failed to relate his observation to the peripheral oedema of the patient. Two years later, made possible by Galileo Galilei’s invention of the lens, Marcello Malpighi (1628–1694) published his discovery of the glomerulus (Mezzogiorno, 1997; George, 2006). However, it was not until 1764, when Domenico Cotugno (1736–1822) (Fig. 42.3) in Naples demonstrated ‘albumin’ in the urine of a 28-year-old soldier with oedema and gave us the term albuminuria (Schena, 1994a). His report was clearly noted by fellow scholars (Schena, 1994b) although it did not lead to an understanding of the link between oedema and proteinuria. Indeed, well into the nineteenth century John Blackall (Jennette, 2007), physician to the Devon and Exeter Hospital, published a comprehensive work on dropsy (Blackall, 1814; Fine and English, 1994), which failed to grasp the connection.

426

Section 3  

the patient with glomerular disease

Table 42.1  Clinical glomerular syndromes and some, but not all, histologic patterns of glomerular injury that can cause each syndrome. See Chapter 45 for a pathophysiological approach and diagram Clinical presentation

Key features

Common patterns of histologic injury and diagnoses on renal biopsy

Asymptomatic proteinuria and haematuria and chronic glomerulonephritis

Hypertension, invisible glomerular haematuria with proteinuria and mild renal impairment, dense renal parenchyma on ultrasound

Alport syndrome and thin basement membrane disease, IgA nephropathy

Macroscopic haematuria

Visible, painless, glomerular haematuria, often coinciding with upper respiratory tract infection, invisible haematuria and proteinuria in between the attacks with or without mild renal impairment

IgA nephropathy

Nephritic syndrome

Abrupt onset of oedema and hypertension, glomerular haematuria and red cell casts, proteinuria (usually < 1.5 g/day)

Endocapillary-proliferative GN, IgA nephropathy

Nephrotic syndrome

Oedema, proteinuria (adult) > 3.5 g/day, hypoalbuminuria, hyperlipidaemia, with or without renal impairment

Amyloidosis, diabetic nephropathy, focal segmental glomerulosclerosis (FSGS), membranous GN, minimal change disease and its variants, membranoproliferative GN, pre-eclampsia, IgA nephropathy

Rapidly progressive glomerulonephritis

May or may not have extrarenal symptoms (skin, lung, joints); rapidly declining renal function (days, weeks), glomerular haematuria with red cell casts and proteinuria.

Focal necrotizing and crescentic glomerulonephritis, endocapillary proliferative glomerulonephritis, fibrillary glomerulonephritis, membranoproliferative GN

Fig. 42.1  Doctor performing urine analysis. There is a group of patients holding their little baskets, each containing a matula (the vessel in which urine is collected), awaiting their turns with the physician. The matula became a symbol of medical powers in general and physicians would make a ritual of holding it to the light before giving a diagnosis. It was also used as a billboard in some European cities (Armstrong, 2006). From Avicenna (?980–1037), Canon (US National Library of Medicine Images from the History of Medicine Collection, image in the public domain ).

However, these observations set the scene for Richard Bright when he became interested in the subject in 1816 (Cameron, 2003). His crucial role was perhaps not so much to make a singular pioneering discovery, but rather to put previous and scattered observations together into a unifying concept that included clinical, biochemical,

and morphological characteristics. It was through his work that the term ‘nephritis’ came to us, although clearly Bright’s generic term encompassed a much wider spectrum of renal diseases that we would associate with the term today (Stratta et al., 1999). Bright’s ideas rapidly gained popularity (Stratta et al., 1999) and for almost a

Chapter 42 

the glomerulus and glomerulonephritis

Fig. 42.2  Medieval urine wheel from Fasciculus Medicinae, a fifteenth-century collection of medical texts owned by Johannes de Ketham. From Sciencephoto Ltd, London, with permission.

century nephritis was to bear the eponym ‘Bright’s disease’. A reassessment of his original specimens (Weller and Nester, 1972) demonstrated just how incredibly accurate his descriptions of kidney disease actually were. Bright also showed that low protein levels in serum were associated with urine protein leak and understood that kidney damage leads to retention of urea (Stratta et al., 1999). The following years saw the crucial transition from vitalism to experimental physiology, pioneered by Claude Bernard (Arunachalam and Woywodt, 2010) and with it the discovery, by Carl Ludwig (1816–1895), of glomerular filtration and tubular re-absorption (Davis et al., 1996). Subsequently, Frerichs in his monograph on Bright’s disease published in 1851, was the first to emphasize the importance of ‘anatomical lesions’, from which one could deduce the underlying pathological processes (Schwarz and Ritz, 1997). Remarkably, Frerichs also predicted some degree of interplay between interstitial scarring and glomerulosclerosis, a concept that rings very familiar today (Schwarz and Ritz, 1997). The second half of the nineteenth century also saw the term ‘glomerulonephritis’ emerge, coined by Swiss-German pathologist Edwin Klebs in his

textbook of pathology (Ritz et al., 1994). Distinct entities of GN were then described from the beginning of the twentieth century (Stratta et  al., 1999):  Heinrich Reichel (1876–1943) in Vienna (Reichel, 1905)  and later British nephrologist Arthur Osman (1893–1972) (Cameron, 1997)  accurately described post-streptococcal GN (Rodríguez-Iturbe and Batsford, 2007)  although William Charles Wells (1757–1817). had, a century earlier, already observed that haematuria follows scarlet fever in some cases (Wells, 1812). The next important milestone towards a modern understanding of GN was reached when physician Franz Volhard (1872–1950) (Fig. 42.4) and pathologist Karl Theodor Fahr (1877–1945) published, in 1914, a new classification for glomerular disease (Fogazzi and Ritz, 1998; and see Fig. 70.1). Their seminal contribution was firstly the use of a clinico-pathological approach as we know it today, which is also reflected in the accurate and at the same time astonishingly beautiful illustrations in their textbook (Fogazzi and Ritz, 1998). In addition, they were the first to distinguish nephrotic and nephritic syndrome in the post-Bright, but still very much pre-biopsy, era of nephrology (Luft and Dietz, 1993).

427

428

Section 3  

the patient with glomerular disease

Fig. 42.3  Domenico Cotugno (1736–1822). Apart from his role in discovering albuminuria, Cotugno is also notable for having been appointed as Professor of Medicine at the age of only 19 at the famous Medical School of Salerno in Italy. From Schena (1994), with permission.

Fig. 42.4  Franz Volhard lecturing to students. On the board are Volhard’s famous circle diagrams separating the various forms of Bright’s disease, with and without hypertension. Volhard points to the group with essential ‘red’ hypertension. From Wolf (2000), with permission.

Around the same time of Volhard and Fahr’s textbook, examination of the urine sediment saw a revival, pioneered by Thomas Addis (1881–1949) in Stanford, originally a Scottish haematologist (Blagg, 2009). He also standardized the technique for quantitative evaluation of the urine sediment—the ‘Addis count’. Remarkably, the origin of proteinuria remained to some degree controversial until well into the 1930s (Waldherr and Ritz, 1999). It took until the 1940s when Edmund Randerath (1881–1949) in Germany deduced, from elegant experiments in salamanders, that the glomerulus was indeed the source of proteinuria (Waldherr and Ritz, 1999). Important clinical observations were also made during that time, such as that of a peculiar syndrome of pulmonary haemorrhage and GN by Ernest Goodpasture (1886–1960) in 1919 (Goodpasture, 2009), followed by the first reports of renal involvement in what we now know as small vessel vasculitis by Wohlwill in 1923 (Matteson, 2002), Klinger in 1931 (Klinger, 1931), and Wegener in 1936 (Woywodt et  al., 2006). However, in general the understanding and classification of GN remained patchy and essentially confined the ideas of Volhard and Fahr well into the 1950s, not least because histological diagnoses could only be made at autopsy. Further progress towards our current understanding of glomerular disease now required the introduction, in the 1950s, of percutaneous renal biopsy (Cameron and Hicks, 1997). This milestone development is usually associated with the names of Poul Iversen (1889–1966) and Claus Brun (b. 1910) in Copenhagen (Cameron and Hicks, 1997). However, Nils Alwall (1904–1986), who is more commonly known for his seminal contribution to dialysis, performed renal biopsies as early as 1944 (Alwall, 1952). However he did not pursue this approach further, chiefly due to an early patient death (Cameron and Hicks, 1997). Crucial advances in laboratory medicine then allowed for further progress in the analysis of biopsy material and a more and more diverse nomenclature of what had until recently carried the umbrella term of Bright’s disease. Immunofluorescence became available in the 1960s, leading to the discovery of immunoglobulin (Ig)-A nephropathy by Jean Berger (1930–2011) and Nicole Hinglais at Hôpital Necker in Paris (Berger and Hinglais, 1968; Woo et al., 2009). By 1963, antibodies directed against class-specific epitopes of the Ig light chains were available commercially. Finally, electron microscopy appeared in the 1950s and was increasingly used by the mid 1960s (Carlson, 1961). These developments culminated in the 1961 Ciba Symposium on Renal Biopsy in London. It is probably fair to say that this event was in itself a milestone towards our current understanding of GN. In the spirit of Volhard and Fahr it was also a celebration of a very successful collaboration between physician and pathologist as narrated elsewhere by Robert Heptinstall, himself one of the doyens of renal pathology in the twentieth century (Heptinstall, 1990). By 1961, this collaboration had established as a standard the diagnosis of GN with renal biopsy, and of the interpretation of biopsy specimens with light microscopy, immunofluorescence, and electron microscopy. These developments also nurtured the establishment of nephrology as a specialty in its own right, with, among others, the first European Congress of Nephrology being held in Geneva in 1960. From here on, developments in our understanding of GN are beyond the scope of this chapter, since they are part of our present concepts of the disease, rather than its history.

Chapter 42 

Classification of glomerular disease The overview presented here may be considered alongside Chapter 18 describing the appearances of the renal biopsy, and Chapter 45 describing major pathophysiological mechanisms behind the major presentations. For didactic reasons, we will differentiate between primarily non-inflammatory and inflammatory glomerular disease, that is, GN. However, this scheme is to some degree arbitrary and artificial. Many non-inflammatory glomerular diseases, while not primarily mediated by cells and effectors of the immune system, are still propagated and have their progression determined by secondary inflammatory processes (Abbate et al., 1998). Alpha-1-antitrypsin deficiency, for example, although primarily an inherited disease not characterized by immune activation, features deposition of immunoglobulins and complement (Heidet and Gubler, 2009). Similarly, diabetic nephropathy, although primarily very clearly a metabolic non-inflammatory disorder, is characterized by abnormal cytokine profiles (Navarro-Gonzalez

the glomerulus and glomerulonephritis

and Mora-Fernandez, 2008). Moreover, immunosuppression has been used successfully in animal models of the disease (Utimura et al., 2003).

Classification of non-inflammatory glomerular disease Inherited non-inflammatory glomerular disease The group of inherited glomerular diseases includes a long list of categories, individual diseases, and syndromes, some of whom are very rare (Kashtan and Gubler, 2009). Table 42.2 provides an overview. Many other inherited diseases of the kidney, such as the nephronophthisis group (see Chapter 316), spare the glomerulus. When observed in these disorders, glomerular changes and proteinuria are usually regarded as secondary to scarring and loss of nephrons. Some knowledge even of the rare inherited glomerular diseases is important, as this will often enable the clinician to make a clinical diagnosis and also guide genetic testing. Crucially, a spot diagnosis will usually be facilitated not by the renal features

Table 42.2  Examples of inherited non immune-mediated glomerular disease (See also Section 15 and Kashtan and Gubler, 2009) Category

Main site of injury

Genetics and pathogenetic mechanisms

Associated extrarenal disease

Typical clinical presentation

Alport syndrome (type IV collagen) (Kashtan, 1999) (Chapter 322)

Glomerular basement membrane

COL4A1 and COL4A2 at 13q34 COL4A3 and COL4A4 at 2q35–37 COL4A5 and COL4A6 on chromosome X.

Sensorineural deafness Microscopic haematuria, and anterior lenticonus; followed by proteinuria, leiomyomas (Kashtan, 1999) hypertension and slow progression to ESRD

Nail patella syndrome (Kashtan and Gubler, 2009; Lemley, 2009) (Chapter 326)

Glomerular basement membrane

Mutation in LMX1B transcription factor at Sensorineural deafness, 9q34 (autosomal dominant) (Bongers et al., dystrophic nails, hypoplasia 2005); this seems to affect COL4A4 and of patellae/elbows COL4A3 (Morello et al., 2001; Lemley, 2009)

Microscopic haematuria, proteinuria and renal impairment, progressive

Thin basement membrane disease (TBMN) (Tryggvason and Patrakka, 2006) (Chapter 325)

Glomerular basement membrane

Disorder of collagen IV trimer α3:α4:α5; autosomal-dominant with mutations in COL4A5, COL4A3, or COL4A4 (Tryggvason and Patrakka, 2006)

None

Microscopic haematuria, minimal proteinuria and normal renal function.

Denys–Drash/Frasier syndrome (Chapter 329)

Podocyte

WT1 gene at 11p13 (Baird et al., 1992)

Wilms tumour, pseudohermaphroditism

Nephrotic syndrome, progressing to ESRD

Laminin deficiency (Pierson syndrome) (Chapter 320)

Podocyte

Mutation in LAMB2, the gene encoding the β2 chain of laminin, at 3p14–p22 (Zenker et al., 2005)

Cataract, microcoria and other eye abnormalities (Choi et al., 2008)

Congenital nephrotic syndrome (Kashtan and Gubler, 2009)

Steroid-resistant nephrotic syndrome (SRNS) (Caridi et al., 2010; Hildebrandt, 2010) (Chapter 327)

Podocyte

SRNS1 (Finnish type and adult variants): Nephrin at 19q13.1 (Godefroid and Dahan, 2010) SRNS2: (recessive): podocin at 1q25–q31 (Boute et al., 2000) SRNS3: PLCE1 gene at 10q23 (Boyer et al., 2010) SRNS4: CD2AP gene at 6p12 (Kim et al., 2003) (and others, see Chapter 327)

NPHS1: pyloric stenosis in some families

Severe steroid-resistant nephrotic syndrome

Disorders of GBM collagen and its transcription (Kashtan and Gubler, 2009) (see Chapter 30)

Podocytopathies (see Chapter 327)

(continued)

429

430

Section 3  

the patient with glomerular disease

Table 42.2 Continued Category

Main site of injury

Genetics and pathogenetic mechanisms

Associated extrarenal disease

Typical clinical presentation

Hereditary amyloidosis (Kashtan and Gubler, 2009) (Chapter 152)

Glomerulus, blood vessels

Protein mutations: transthyretin, gelsolin, the apolipoproteins, lysozyme, fibrinogen Periodic fever syndromes: familial Mediterranean fever (FMF), Muckle–Wells syndrome, hyper IgD syndrome (Kashtan and Gubler, 2009)

Highly variable, including Nephrotic syndrome with vascular involvement, ocular renal impairment and often amyloid; neuropathy and progression to ESRD autonomic failure

Fabry disease (Muckle Wells syndrome, 2007) (Chapter 335)

Glomerulus, vascular endothelium, tubulointerstitial cells

X-linked lysosomal α-galactosidase deficiency

Neuropathy, cardiovascular disease, angiokeratoma of the skin, corneal opacities

Progressive CKD with vascular/cardiac disease and severe neuropathy (Clarke, 2007)

Other storage and deposition diseases: alpha-1-antitrypsin deficiency, Alagille syndrome (Krantz et al., 1987), and others.

Glomerulus, vascular endothelium, tubulointerstitial cells

For review see Kashtan and Gubler (2009)

Often multisystem disease (Kashtan and Gubler, 2009)

Proteinuria, renal impairment and occasionally ESRD (Alagille syndrome) (Kashtan and Gubler, 2009)

Storage and deposition diseases

CD2AP = CD2-associated protein; CKD = chronic kidney disease; ESRD = end-stage renal disease; IgD = immunoglobulin D; LAMB-2 = laminin beta2 chain; LMX1B = LIM homeobox transcription factor 1 beta; OMIM = Online Mendelian Inheritance in Man; PLCE-1 = 1-phosphatidylinositol-4,5-bisphosphate phosphodiesterase epsilon-1; TRPC-6 = transient receptor potential cation channel, subfamily C, member 6; WT-1 = Wilms tumour 1.

of the disease, but by its typical extrarenal manifestations. Alport syndrome (see Chapter  322), the nail patella syndrome (see Chapter 326), and Fabry disease (see Chapter 336) serve as good examples. The topic is also of importance because the number of young patients with inherited disorders graduating from paediatric to adult renal care continues to increase, not least due to improved management and better survival (Watson et al., 2011) (see Chapter 322). Among the inherited glomerular diseases, the disorders of glomerular type IV collagen (Table 42.2), that is, Alport syndrome (see Chapter 322) and the thin basement membrane disease (see Chapter 325), are most commonly encountered in adult nephrology. Their genetic basis, clinical manifestations, and diagnosis have been well understood for quite some time although progress has now been made with regard to mutation analysis. Other diseases have been added to the spectrum, such as the nail–patella syndrome (caused by mutations in the gene encoding the LMX1B transcription factor, see Chapter 326) (Table 42.2). They feature significant, and characteristic, extrarenal manifestations, namely hearing impairment in Alport syndrome, and hypoplasia of patellae and elbows in the nail–patella syndrome (Table 42.2). Exciting progress has also occurred regarding a growing group of genetic defects of the podocyte and the term ‘podocytopathies’ has been coined (Barisoni et al., 2007) (see Chapter 327). These disorders, while mostly rare, are nonetheless of importance, particularly in paediatric and adolescent nephrology. Their importance stems from the fact that they usually present with severe nephrotic syndrome that is resistant to steroids and other immunosuppressant drugs (Wiggins, 2007). These disorders have in common a genetic defect of one of the proteins of the podocyte’s ultrastructure, resulting in proteinuria, nephrotic syndrome, and progressive renal impairment, often in childhood or adolescence (Wiggins, 2007). Taxonomy and nomenclature of these diseases are still very much

in the making (Barisoni et al., 2007) but their characterization has already improved our understanding of the glomerular slit membrane (Welsh and Saleem, 2010) with many of the mutated proteins playing a crucial role in its structural integrity and function (see Chapter 327). These disorders also serve as an example of a disease in which mutations determine age of onset and treatment response (Hildebrandt, 2010). Some of the proteins that are implicated, such as nephrin and podocin, are well conserved in evolution, which underscores their pivotal role in glomerular filtration (Weavers et al., 2009). Nephrin appears to have a particularly central role in the slit diaphragm and interactions with many other proteins have now been described (Welsh and Saleem, 2010). It is highly likely that some more podocytopathies are going to emerge in the near future, together with more crucial proteins of the slit diaphragm (Michaud and Kennedy, 2007). The next subcategory of inherited glomerular diseases is that of the storage and deposition disorders, which includes hereditary amyloidosis (see Chapter 152), Fabry disease (see Chapter 335), and finally some even rarer syndromes, such as the Alagille syndrome (Kashtan and Gubler, 2009). Some authors differentiate, within this group, between diseases with primary glomerular involvement, such as Fabry disease, and others with secondary glomerular involvement, such as hereditary amyloidosis (see Chapter 152). Notable progress has occurred within the group of hereditary amyloidosis and a number of individual disease entities are now well characterized (Table 42.2). Much progress has also been made with regard to Fabry disease, and particularly its treatment with enzyme replacement (see Chapter 338).

Acquired non-inflammatory glomerular disease This group encompasses glomerular diseases due to drugs and medication, metabolic diseases, deposition disorders, and those due to vascular disease (Table 42.3). Differentiating these from the

Chapter 42 

the glomerulus and glomerulonephritis

Table 42.3  Acquired non-immune-mediated glomerular disease Acquired non-immune-mediated glomerular disease

Main site of injury Pathogenetic mechanisms

Typical clinical presentation

Drug-induced (Izzedine et al., 2006) (Chapter 82 and Chapter 362)

Endothelial cell, podocyte, GBM

Mostly unclear and idiosyncratic although some concept are beginning to emerge, e.g. proteinuria due to mTOR inhibitors (Inoki et al., 2011) and VEGF inhibitors (George et al., 2007) 

New-onset proteinuria with or without renal impairment in conjunction with drug treatment

AA amyloid (Chapter 152)

Diffuse deposition in glomerulus, interstitium, blood vessels

Extracellular tissue deposition of fibrils that are composed of fragments of serum amyloid A (SAA) protein in conjunction with an chronic inflammatory condition (Gillmore et al., 2001)

Nephrotic syndrome with renal impairment and the presence of a predisposing rheumatic or a chronic inflammatory disease extrarenal abnormalities (hepatosplenomegaly, macroglossia, restrictive cardiomyopathy) (Gertz and Kyle, 1991)

AL amyloid (Chapter 152)

Diffuse deposition in glomerulus, interstitium, blood vessels

Clonal proliferation of plasma cells, leading to production of monoclonal immuno-globulins and tissue deposition of organized immunoglobulin fibrils (Sanchorawala, 2006) 

Nephrotic syndrome with renal impairment in conjunction with a serum/urinary paraprotein; extrarenal abnormalities (hepatosplenomegaly, macroglossia, restrictive cardiomyopathy) (Kyle and Greipp, 1983)

Diabetic nephropathy (Chapter 149)

Endothelium, mesangial cell, podocyte

Multifactorial, including glomerular hyperfiltration, direct effects of hyperglycaemia and advanced glycation end products, vascular injury and podocyte changes

Proteinuria or nephrotic syndrome with progressive renal impairment in a patient with long-standing diabetes, usually with concurrent evidence of other end-organ damage

Monoclonal immunoglobulin deposition diseases (MIDD): light chain deposition disease, heavy chain deposition disease (Chapter 150)

Diffuse deposition in glomerulus, interstitium, blood vessels

Clonal proliferation of plasma cells, leading to production of monoclonal immunoglobulins and tissue deposition of non-organized immunoglobulins(Ronco et al., 2006) 

Proteinuria/nephrotic syndrome with renal impairment in a patient with monoclonal Ig in serum and/or urine and evidence of plasma cell proliferation on bone marrow biopsy

Radiation nephritis (Luxton, 1961) (Chapter 91)

Endothelium, glomerulus

Cellular injury due to irradiation

Proteinuria (Luxton, 1961) (occasionally nephrotic syndrome (Jennette and Ordonez, 1983)) and renal impairment (can be progressive)

Hypertensive nephropathy (Chapter 211)

Endothelial cell

Ischaemic injury, leading to glomerulosclerosis, with nephron loss, hyperfiltration, and further injury (Harvey et al., 1992)

Slowly progressive renal impairment and mild-to-moderate proteinuria in a patient with long-standing hypertension (Fogo et al., 1997)

Pre-eclampsia and HELLP syndrome (Chapter 296)

Endothelial cell

Imbalance between pro- and anti-angiogenic factors (Maynard and Karumanchi, 2011), leading to microangiopathy

New onset of hypertension and proteinuria after 20 weeks of gestation (Sibai, 2003)

Thrombotic microangiopathy (Chapter 174)

Endothelial cell

Glomerular platelet-rich microthrombi, fibrin deposition, and ischaemia (Benz and Amann, 2010)

AKI, laboratory evidence of MAHA, and dysfunction of other organ systems (Forzley and Clark, 2009)

Others (antiphospholipid syndrome, embolism and cholesterol embolism (Chapters 164, 212))

Endothelial cell

Macro- and microthrombi and ischaemia

AKI with (embolism) or without (cholesterol embolism) flank pain; typical laboratory features

Metabolic/deposition diseases

Vascular disorders

AKI = acute kidney injury; GBM = glomerular basement membrane; HELLP syndrome = haemolytic anaemia; elevated liver enzymes, low platelets syndrome; Ig = immunoglobulin; MAHA = microangiopathic haemolytic anaemia; mTOR = mammalian target of rapamycin; VEGF = vascular endothelial growth factor.

inflammatory glomerulonephritides is in some cases arbitrary. The monoclonal immunoglobulin deposition diseases (MIDDs) and AL amyloidosis, for example, are regarded as deposition diseases by some authors while others view them as glomerulonephritides. The taxonomy of immunotactoid and fibrillary glomerulopathy is similarly controversial. Drug-induced glomerular disease is comparatively rare and much less well appreciated than nephrotoxicity occurring in

the tubulointerstitium (see Chapter 84). A long list of drugs has been implicated, although causality is often difficult to ascertain. Diabetic nephropathy (see Chapter 149) is the prime example of a metabolic disorder causing glomerular disease. Diabetic nephropathy is now the leading cause of end-stage renal failure in many developed countries and its early recognition and effective management determines the fate of a large proportion of patients seen in adult nephrology. The remainder of the acquired metabolic and

431

432

Section 3  

the patient with glomerular disease

deposition diseases and AL amyloid (see Chapter 152) have some clinical importance, although this is perhaps still surpassed by their scientific relevance (Table 42.3). Nephrologists and haematologists have made some progress in the management of AL amyloid and the MIDDs, although their management remains challenging for the clinician. In comparison, radiation nephritis (see Chapter 91) is exceedingly rare and poorly understood. Finally, a number of vascular disorders (Table 42.3) are clinically important and some of them are of substantial scientific interest as well. In particular, pre-eclampsia (see Chapter 296) and thrombotic microangiopathy (see Chapter 174) come to mind (Table 42.3), with interesting new insight into the pathogenesis occurring during the last decade.

Classification of inflammatory glomerular diseases (glomerulonephritis) Attempts to classify this group of diseases have been made ever since German-Swiss pathologist Edwin Klebs first coined the term ‘glomerulonephritis’ in his textbook published in 1870 (Klebs, 1870). Between 1881 and 1893, American physician Francis Delafield (1841–1915) worked on a complex classification of ‘Bright’s disease’ that differentiated between acute and chronic forms, and subdivided each into different subcategories (Campbell et al., 2003).

Osler also emphasized the course of the disease over time as a criterion for classification and, making reference to Delafield, used the terms acute and chronic nephritis (Stratta et al., 1999). Volhard and Fahr, in their landmark textbook published in 1914, made another attempt at dividing the umbrella term Bright’s disease into different clinical patterns of glomerular disease, namely ‘degenerative/ nephrotic’, ‘focal/diffuse’, and ‘atherosclerotic’ (Luft and Dietz, 1993). Thomas Addis (1881–1949) in 1925 also used the term ‘chronic latent nephritis’ (Addis, 1925). In 1929, W. T. Longcope (1877–1953) (Longcope, 1929) proposed two types of GN, namely ‘focal glomerular nephritis’ and ‘diffuse glomerular nephritis’. Calvin Ellis (1826–1883) in 1942 also differentiated between two types of GN, one with preceding infection; type 1 featured haematuria, and a high percentage of recovery while type 2 was ‘lipoid nephrosis’ with a more insidious onset and infrequent recovery (Ellis, 1942). The widespread introduction of renal biopsy in the 1950s and 1960s led to more patterns of histologic injury (Table 42.1) being defined from the combined use of light and electron microscopy, and immunofluorescence microscopy. Immunological terms, referring to patterns of complement and immunoglobulin deposition, thus made their appearance in the classification in the 1960, such

Table 42.4  Patterns of glomerular injury observed by light microscopy and examples of glomerulonephritides and other glomerular disorders that can cause each pattern of injury Patterns are described in more detail in Chapter 18 Pattern of glomerular injury

Glomerulonephritides that cause this pattern of injury

Other glomerular diseases that can cause this pattern of injury

No abnormality by light microscopy

Minimal change disease (Chapter 55) and variants, mild/early glomerular disease (e.g. IgA nephropathy)

Thin basement membrane disease (Chapter 325) and early Alport syndrome (Chapter 321) Early/mild amyloidosis (Chapter 152)

Thick capillary walls without hypercellularity or mesangial expansion

Membranous GN (with thickened glomerular basement membrane) Pre-eclampsia (with endothelial swelling) (Chapter 296) (Chapter 60) Thrombotic microangiopathy (with expanded Fibrillary GN (with predominant capillary wall deposits) (Chapter 81) sub-endothelial zone) (Chapter 174)

Thick capillary walls with mesangial expansion but little or no hypercellularity

Secondary membranous GN with mesangial deposits (Chapter 63) Fibrillary GN (Chapter 81) Dense deposit disease (type II membranoproliferative GN) (Chapter 80)

Focal segmental glomerulosclerosis FSGS (primary or secondary) (Chapter 57) (FSGS) without hypercellularity Chronic sclerotic disease of focal GN; secondary FSGS (Box 57.1)

Diabetic nephropathy (Chapter 149) Amyloidosis (Chapter 152) Monoclonal immunoglobulin deposition disease (Chapter 150) Alport syndrome (Chapter 321)

Mesangial or endocapillary hypercellularity

Mesangioproliferative GN; IgA nephropathy (Chapter 65) and others Endocapillary proliferative/post-infectious GN (Chapter 623) Membranoproliferative GN type I–III (Chapter 80)

Extracapillary hypercellularity

Crescentic GN (Chapter 70) Collapsing variant FSGS (Chapter 57) and HIV nephropathy (Chapter 186)

Drug-induced collapsing variant FSGS (bisphosphonates) (Chapter 82)

Membranoproliferative lobular or nodular pattern

Membranoproliferative GN type I–III (Chapter 80) Fibrillary or Immunotactoid GN (Chapter 81)

Diabetic nephropathy (Chapter 149) Monoclonal immunoglobulin deposition disease (Chapter 150) Thrombotic microangiopathy (particularly healing) (Chapter 174)

Diffuse global glomerular sclerosis

End-stage glomerular disease

End-stage vascular disease End-stage tubulointerstitial disease

Adapted from Johnson et al. (2000).

Chapter 42 

as that of IgA nephropathy (Berger and Hinglais, 1968) in 1968. In some conditions, additional criteria for classification emerged on the basis of serological markers and extrarenal disease as with the discovery of antibodies against the cytoplasm of neutrophils in 1982 (Davies et al., 1982). In parallel, individual diseases were named on the basis of their aetiology, such as post-streptococcal GN (Rodríguez-Iturbe and Batsford, 2007). It is worthwhile to remember that this was a long, and often chaotic, process, essentially reflecting the tortuous development of nephrology and renal pathology in the nineteenth and twentieth century, and not conscious process of layering. As of today, a universally accepted and unifying diagnostic classification of GN that integrates clinical features, morphology on renal biopsy, and other immunological findings, does not exist. To the physician, the clinical syndromes associated with GN (Table 42.1) remain a good starting point to narrow the differential diagnosis, to guide investigations and diagnostic thinking, and to begin a process that will eventually lead to a specific diagnosis (Jennette et al., 2007). In comparison to the clinical syndromes associated with tubulointerstitial and vascular lesions, these syndromes are quite specific and predict a lesion of the glomerular capillary wall with reasonable certainty (Jennette et al., 2007). Of note, even the nomenclature of glomerular disease lacks consensus. Some authors, for example, have referred to minimal change disease (see Chapter 53) as a GN while others prefer the term glomerulopathy, on the account that inflammatory lesions are lacking. Another good example is the term of membranoproliferative GN (MPGN) (Appel et al., 2005) (see Chapter 80), which contrasts with that of mesangiocapillary GN preferred by others (D’Amico and Ferrario, 1992). In the absence of a unifying classification and nomenclature of GN, this book relies on a morphological categorization based on patterns on renal biopsy, except where aetiological factors are clearly identified (e.g. HIV-associated nephropathy), an associated multisystem disease is defined (e.g. lupus nephritis), or the immunopathogenesis is well characterized (e.g. antiglomerular basement membrane (anti-GBM) disease). Table 42.4 lists common patterns of glomerular injury by light microscopy and the glomerulonephritides that can cause these patterns. For the sake of completeness, glomerular diseases other than GN are also listed as a differential diagnosis. This is, of course, not a classification by aetiology as many histological patterns have a variety of aetiologies. Some have therefore emphasized that renal biopsy yields a histological pattern, not a distinct diagnosis (Johnson et al., 2000 and see Chapter 18). In MPGN (see Chapter 61), for example, further investigations and tests, such as obtaining a family history, hepatitis C serology, alpha-1-antitrypsin levels, etc., will allow for a diagnosis based on aetiology. Another good example is focal necrotizing/crescentic GN (see Chapter 70) where only comprehensive testing for antineutrophil cytoplasmic antibodies, anti-GBM, etc. allows for a clinically useful diagnosis. Conversely, one aetiology may produce different patterns of histological damage: hepatitis B, for example, is capable of causing as diverse patterns of injury as membranous GN, MPGN, and even polyarteritis nodosa (see Chapter 185) (Johnson and Couser, 1990).

References Abbate, M., Zoja, C., Corna, D., et al. (1998). In progressive nephropathies, overload of tubular cells with filtered proteins translates glomerular

the glomerulus and glomerulonephritis

permeability dysfunction into cellular signals of interstitial inflammation. J Am Soc Nephrol, 9, 1213–24. Addis, T. (1925). A clinical classification of Bright’s diseases. J Am Med Assoc, 85, 163–7. Alwall, N. (1952). Aspiration biopsy of the kidney, including i.a. a report of a case of amyloidosis diagnosed through aspiration biopsy of the kidney in 1944 and investigated at an autopsy in 1950. Acta Med Scand, 143, 430–5. Appel, G. B., Cook, H. T., Hageman, G., et al. (2005). Membranoproliferative glomerulonephritis type II (dense deposit disease): an update. J Am Soc Nephrol, 16, 1392–403. Armstrong, J. A. (2006). Urinalysis in Western culture: a brief history. Kidney Int, 71, 384–7. Arunachalam, C. and Woywodt, A. (2010). Turbid urine and beef-eating rabbits: Claude Bernard (1813–78)—a founder of modern physiology. NDT Plus, 3, 335–7. Baird, P. N., Santos, A., Groves, N., et al. (1992). Constitutional mutations in the WT1 gene in patients with Denys-Drash syndrome. Hum Mol Genet, 1, 301–5. Barisoni, L., Schnaper, H. W., and Kopp, J. B. (2007). A proposed taxonomy for the podocytopathies: a reassessment of the primary nephrotic diseases. Clin J Am Soc Nephrol, 2, 529–42. Barisoni, L., Schnaper, H. W., and Kopp, J. B. (2007). A proposed taxonomy for the podocytopathies: a reassessment of the primary nephrotic diseases. Clin J Am Soc Nephrol, 2, 529–42. Benz, K. and Amann, K. (2010). Thrombotic microangiopathy: new insights. Curr Opin Nephrol Hypertens, 19, 242–7. Berger, J. and Hinglais, N. (1968). [Intercapillary deposits of IgA-IgG]. J Urol Nephrol (Paris), 74, 694–5. Blackall, J. (1814). Observations on the Nature and Cure of Dropsies etc.etc. London: Longman, Hurst, Rees, Orme and Brown. Blagg, C. R. (2009). Thomas Addis, 1881–1949, clinical scientist, hematologist and pioneering nephrologist: a brief biography. J Nephrol, 22 Suppl 14, 115–19. Bongers, E. M. H. F., Huysmans, F. T., Levtchenko, E., et al. (2005). Genotype-phenotype studies in nail-patella syndrome show that LMX1B mutation location is involved in the risk of developing nephropathy. Eur J Hum Genet, 13, 935–46. Boute, N., Gribouval, O., Roselli, S., et al. (2000). NPHS2, encoding the glomerular protein podocin, is mutated in autosomal recessive steroid-resistant nephrotic syndrome. Nat Genet, 24, 349–54. Boyer, O., Benoit, G., Gribouval, O., et al. (2010). Mutational analysis of the PLCE1 gene in steroid resistant nephrotic syndrome. J Med Genet, 47, 445–52. Cameron, J. S. (1997). Arthur Arnold Osman (1893-1972): a forgotten pioneer of nephrology. Nephrol Dial Transplant, 12, 1526–30. Cameron, J. S. (2003). Milk or albumin? The history of proteinuria before Richard Bright. Nephrol Dial Transplant, 18, 1281–5. Cameron, J. S. and Hicks, J. (1997). The introduction of renal biopsy into nephrology from 1901 to 1961: a paradigm of the forming of nephrology by technology. Am J Nephrol, 17, 347–58. Campbell, J. L. and Eknoyan, G. (2003). Francis Delafield (1841–1915): the original contributions of an American investigator to diseases of the kidney. J Nephrol, 16, 779–84. Caridi, G., Trivelli, A., Sanna-Cherchi, S., et al. (2010). Familial forms of nephrotic syndrome. Pediatr Nephrol, 25, 241–52. Carlson, H. E. (1961). Electron microscopy of the kidney in health and disease. South Med J, 54, 985–92. Changizi Ashtiyani, S. and Cyrus, A. (2010). Rhazes, a genius physician in diagnosis and treatment of kidney calculi in medical history. Iran J Kidney Dis, 4, 106–10. Changizi Ashtiyani, S., Shamsi, M., Cyrus, A., et al. (2011). A critical review of the works of pioneer physicians on kidney diseases in ancient Iran: Avicenna, Rhazes, Al-akhawayni, and Jorjani. Iran J Kidney Dis, 5, 300–8. Choi, H. J., Lee, B. H., Kang, J. H., et al. (2008). Variable phenotype of Pierson syndrome. Pediatr Nephrol, 23, 995–1000.

433

434

Section 3  

the patient with glomerular disease

Clarke, J. T. (2007). Narrative review: Fabry disease. Ann Intern Med, 146, 425–33. D’Amico, G. and Ferrario, F. (1992). Mesangiocapillary glomerulonephritis. J Am Soc Nephrol, 2, S159–66. Davies, D. J., Moran, J. E., Niall, J. F., et al. (1982). Segmental necrotising glomerulonephritis with antineutrophil antibody: possible arbovirus aetiology? Br Med J (Clin Res Ed), 285, 606. Davis, J. M., Thurau, K., and Häberle, D. (1996). Carl Ludwig: the discoverer of glomerular filtration. Nephrol Dial Transplant, 11, 717–20. De Santo, N. G., Capasso, G., Giordano, D. R., et al. (1989). Nephrology in the natural history of Pliny the Elder (23–79 A.D.). Am J Nephrol, 9, 252–60. Diskin, C. J. (2008). de Ketham revisited: a modern-day urine wheel. Med J Aust, 189, 658–9. Eknoyan, G. (1988). Origins of nephrology: Hippocrates, the father of clinical nephrology. Am J Nephrol, 8, 498–507. Eknoyan, G. (1994). Arabic medicine and nephrology. Am J Nephrol, 14, 270–8. Eknoyan, G. (2002). Rufus of Ephesus and his ‘Diseases of the Kidneys’. Nephron, 91, 383–90. Ellis, A. (1942). Natural history of Bright’s disease. Lancet, I, 1–7. Fine, L. G. and English, J. A. (1994). John Blackall (1771–1860): failure to see the obvious in dropsical patients with coagulable urine? Am J Nephrol, 14, 371–6. Floege, J., Burns, M. W., Alpers, C. E., et al. (1992). Glomerular cell proliferation and PDGF expression precede glomerulosclerosis in the remnant kidney model. Kidney Int, 41, 297–309. Fogazzi, G. B. and Ritz, E. (1998). Novel classification of glomerulonephritis in the monograph of Franz Volhard and Theodor Fahr. Nephrol Dial Transplant, 13, 2965–7. Fogo, A., Breyer, J. A., Smith, M. C., et al. (1997). Accuracy of the diagnosis of hypertensive nephrosclerosis in African Americans: a report from the African American Study of Kidney Disease (AASK) Trial. AASK Pilot Study Investigators. Kidney Int, 51, 244–52. Forzley, B. R. and Clark, W. F. (2009). TTP/HUS and prognosis: the syndrome and the disease(s). Kidney Int Suppl, S59–61. George, B. A., Zhou, X. J., and Toto, R. (2007). Nephrotic syndrome after bevacizumab: case report and literature review. Am J Kidney Dis, 49, e23–9. George, C. R. (2003). The cellular history of the glomerulus. J Nephrol, 16, 949–57. Gertz, M. A. and Kyle, R. A. (1991). Secondary systemic amyloidosis: response and survival in 64 patients. Medicine, 70, 246–56. Gillmore, J. D., Lovat, L. B., Persey, M. R., et al. (2001). Amyloid load and clinical outcome in AA amyloidosis in relation to circulating concentration of serum amyloid A protein. Lancet, 358, 24–9. Godefroid, N. and Dahan, K. (2010). Expanding the clinical spectrum of congenital nephrotic syndrome caused by NPHS1 mutations. Nephrol Dial Transplant, 25, 2837–9. Goodpasture, E. W. (2009). Landmark publication from The American Journal of the Medical Sciences: the significance of certain pulmonary lesions in relation to the etiology of influenza. Am J Med Sci, 338, 148–51. Harvey, J. M., Howie, A. J., Lee, S. J., et al. (1992). Renal biopsy findings in hypertensive patients with proteinuria. Lancet, 340, 1435–6. Heidet, L. and Gubler, M. -C. (2009). The renal lesions of Alport syndrome. J Am Soc Nephrol, 20, 1210–15. Heptinstall, R. H. (1990). The development of renal pathology. Am J Kidney Dis, 16, 568–73. Hildebrandt, F. (2010). Genetic kidney diseases. Lancet, 375, 1287–95. Hirooka, M., Hirota, M., and Kamada, M. (1988). Renal lesions in Cockayne syndrome. Pediatr Nephrol, 2, 239–43. Inoki, K., Mori, H., Wang, J., et al. (2011). mTORC1 activation in podocytes is a critical step in the development of diabetic nephropathy in mice. J Clin Invest, 121, 2181–96. Izzedine, H., Launay-Vacher, V., Bourry, E., et al. (2006). Drug-induced glomerulopathies. Expert Opin Drug Saf, 5, 95–106.

Jennette, J. C., Olson, J. L., Schwartz, M. M., et al. (2007). Primer on the pathologic diagnosis of renal disease. In J. C. Jennette, J. L. Olson, M. M. Schwartz et al. (eds.) Heptinstall’s Pathology of the Kidney, pp. 97–123. Philadelphia, PA: Lippincott. Jennette, J. C. and Ordonez, N. G. (1983). Radiation nephritis causing nephrotic syndrome. Urology, 22, 631–4. Johnson, R. J. and Couser, W. G. (1990). Hepatitis B infection and renal disease: clinical, immunopathogenetic and therapeutic considerations. Kidney Int, 37, 663–76. Johnson, R. J., Rennke, H., and Feehally, J. (2000). Introduction to glomerular disease: pathogenesis and classification. In R. J. Johnson and J. Feehally (eds.) Comprehensive Clinical Nephrology, pp. 1263–70. Edinburgh: Mosby. Kashtan, C. and Gubler, M. C. (2009). Inherited glomerular disease. In E. D. Avner, W. E. Harmon, P. Niaudet, et al. (eds.) Pediatric Nephrology, pp. 621–41. Berlin: Springer. Kashtan, C. E. (1999). Alport syndrome. An inherited disorder of renal, ocular, and cochlear basement membranes. Medicine, 78, 338–60. Kim, J. M., Wu, H., Green, G., et al. (2003). CD2-associated protein haploinsufficiency is linked to glomerular disease susceptibility. Science, 300, 1298–300. Klebs, T. A. E. (1870). Handbuch der pathologischen Anatomie. Berlin: Hirschwald. Klinger, H. (1931). Grenzformen der periarteriitis nodosa. Frankfurt Pathol, 42, 455. Krantz, I. D., Piccoli, D. A., and Spinner, N. B. (1997). Alagille syndrome. J Med Genet, 34, 152–7. Kyle, R. A. and Greipp, P. R. (1983). Amyloidosis (AL). Clinical and laboratory features in 229 cases. Mayo Clin Proc, 58, 665–83. Longcope, W. T. (1929). The pathogenesis of glomerular nephritis. Bull Johns Hopkins Hosp, 45, 335–60. Luft, F. and Dietz, R. (1993). Franz Volhard in historical perspective. Hypertension, 22, 253–6. Luxton, R. W. (1961). Radiation nephritis. A long-term study of 54 patients. Lancet, 2, 1221–4. Malpass, K. (2011). Peripheral neuropathies: new pathogenetic insights into Charcot-Marie-Tooth disease. Nat Rev Neurol, 7(9), 476. Matteson, E. (2002). Historical perspective of vasculitis: polyarteritis nodosa and microscopic polyangiitis. Curr Rheumatol Rep, 4, 67–74. Maynard, S. E. and Karumanchi, S. A. (2011). Angiogenic factors and preeclampsia. Semin Nephrol, 31, 33–46. Mezzogiorno, A. and Mezzogiorno, V. (1997). Marcello Malpighi (1628–1694). Am J Nephrol, 17, 269–73. Michaud, J. -L. R. and Kennedy, C. R. J. (2007). The podocyte in health and disease: insights from the mouse. Clin Sci, 112, 325–35. Navarro-Gonzalez, J. F. and Mora-Fernandez, C. (2008). The role of inflammatory cytokines in diabetic nephropathy. J Am Soc Nephrol, 19, 433–42. Paul, M. D., Fernandez, D., Pryse-Phillips, W., et al. (1990). Charcot-Marie-Tooth disease and nephropathy in a mother and daughter with a review of the literature. Nephron, 54, 80–5. Reichel, H. (1905). Ueber Nephritis bei Scharlach [About nephritis in scarlet fever]. Z Heil, 6. Ritz, E., Kuster, S., and Zeier, M. (1994). Clinical nephrology in 19th century Germany. Am J Nephrol, 14, 443–7. Rodríguez-Iturbe, B. and Batsford, S. (2007). Pathogenesis of poststreptococcal glomerulonephritis a century after Clemens von Pirquet. Kidney Int, 71, 1094–104. Ronco, P., Plaisier, E., Mougenot, B., et al. (2006). Immunoglobulin light (Heavy)-chain deposition disease: from molecular medicine to pathophysiology-driven therapy. Clin J Am Soc Nephrol, 1, 1342–50. Rovin, B. H., Roncone, D., McKinley, A., et al. (2007). APOE Kyoto mutation in European Americans with lipoprotein glomerulopathy. N Engl J Med, 357, 2522–4. Sanchorawala, V. (2006). Light-chain (AL) amyloidosis: diagnosis and treatment. Clin J Am Soc Nephrol, 1, 1331–41.

Chapter 42 

Schena, F. P. (1994a). Domenico Cotugno and his interest in proteinuria. Am J Nephrol, 14, 325–9. Schena, F. P. (1994b). The role of Domenico Cotugno in the history of proteinuria. Nephrol Dial Transplant, 9, 1344–5. Schwarz, U. and Ritz, E. (1997). Glomerulonephritis and progression—Friedrich Theodor von Frerichs, a forgotten pioneer. Nephrol Dial Transplant, 12, 2776–8. Schwimmer, J. A., Markowitz, G. S., Valeri, A. M., et al. (2003). Secondary focal segmental glomerulosclerosis in non-obese patients with increased muscle mass. Clin Nephrol, 60, 233–41. Sethi, S. (2008). Renal failure with intracapillary thrombi. Lipoprotein glomerulopathy. Kidney Int, 73, 1097–8. Sibai, B. M. (2003). Diagnosis and management of gestational hypertension and preeclampsia. Obstet Gynecol, 102, 181–92. Stratta, P., Canavese, C., Sandri, L., et al. (1999). The concept of ‘glomerulonephritis’. The fascinating history of evolution and emergence of a specialist’s nosology focus on Italy and Torino. Am J Nephrol, 19, 83–91. Strom, E. H., Banfi, G., Krapf, R., et al. (1995). Glomerulopathy associated with predominant fibronectin deposits: a newly recognized hereditary disease. Kidney Int, 48, 163–70. Strom, E. H., Sund, S., Reier-Nilsen, M., et al. (2011). Lecithin: cholesterol acyltransferase (LCAT) deficiency: renal lesions with early graft recurrence. Ultrastruct Pathol, 35, 139–45. Tryggvason, K. and Patrakka, J. (2006). Thin basement membrane nephropathy. J Am Soc Nephrol, 17, 813–22. Utimura, R., Fujihara, C. K., Mattar, A. L., et al. (2003). Mycophenolate mofetil prevents the development of glomerular injury in experimental diabetes. Kidney Int, 63, 209–16. Waldherr, R. and Ritz, E. (1999). Edmund Randerath (1899–1961): experimental proof for the glomerular origin of proteinuria. Kidney Int, 56, 1591–6.

the glomerulus and glomerulonephritis

Watson, A. R., Harden, P., Ferris, M., et al. (2011). Transition from pediatric to adult renal services: a consensus statement by the International Society of Nephrology (ISN) and the International Pediatric Nephrology Association (IPNA). Pediatr Nephrol, 26(10), 1753–7. Weavers, H., Prieto-Sanchez, S., Grawe, F., et al. (2009). The insect nephrocyte is a podocyte-like cell with a filtration slit diaphragm. Nature, 457, 322–6. Weller, R. O. and Nester, B. (1972). Histological reassessment of three kidneys originally described by Richard Bright in 1827–36. Br Med J, 2, 761–3. Wells, W. C. (1812). Observations on the dropsy that succeeds scarlet fever. Trans Soc Imp Med Cir Knowledge, 3, 167–86. Welsh, G. I. and Saleem, M. A. (2010). Nephrin-signature molecule of the glomerular podocyte? J Pathol, 220, 328–37. Wiggins, R. C. (2007). The spectrum of podocytopathies: a unifying view of glomerular diseases. Kidney Int, 71, 1205–14. Wolf, G. (2000). Franz Volhard and his students’ tortuous road to renovascular hypertension. Kidney Int, 57, 2156–66. Woo, K.T., Glassock, R. J., and Lai, K. N. (2009). IgA nephropathy: discovery of a distinct glomerular disorder. In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 1–7. Hong Kong: World Scientific Ebooks. Woywodt, A., Haubitz, M., Haller, H., et al. (2006). Wegener’s granulomatosis. Lancet, 367, 1362–6. Zenker, M., Pierson, M., Jonveaux, P., et al. (2005). Demonstration of two novel LAMB2 mutations in the original Pierson syndrome family reported 42 years ago. Am J Med Genet A, 138, 73–4. Zucchelli, P., Cagnoli, L., Casanova, S., et al. (1983). Focal glomerulosclerosis in patients with unilateral nephrectomy. Kidney Int, 24, 649–55.

435

CHAPTER 43

The renal glomerulus: the structural basis of ultrafiltration Marlies Elger and Wilhelm Kriz Introduction The correct name for the structure to be described is ‘renal corpuscle’; ‘glomerulus’’ strictly refers only to the tuft of glomerular capillaries (glomerular tuft). However, the use of the term glomerulus for the entire corpuscle is widely accepted. A renal corpuscle is made up of a tuft of specialized capillaries supplied by an afferent arteriole, drained by an efferent arteriole, and enclosed in Bowman’s capsule (Figs 43.1 and 43.2). The entire tuft of capillaries is covered by the epithelial cells (podocytes), representing the visceral layer of Bowman’s capsule. At the vascular pole, the visceral layer of Bowman’s capsule transforms into the parietal layer, which is a simple squamous epithelium. At the urinary pole the parietal epithelium abruptly changes into the epithelium of the proximal tubule. The space between both layers of Bowman’s capsule is called the urinary space that opens into the tubule lumen. The glomerular basement membrane (GBM) lies at the interface between the glomerular capillaries and the mesangium at one side and the podocytes at the other. At the vascular pole the GBM transforms into the multilayered basement membrane of the parietal epithelium of Bowman’s capsule (parietal basement membrane (PBM)). Renal corpuscles (glomeruli) are the first components of the nephrons. Thus, the number of nephrons exactly correlates with the number of renal corpuscles—in man, this is about 1  million in each kidney, in rat about 30,000 per kidney. Renal corpuscles are roughly spherical in shape, with a diameter of approximately 200 µm in man, and about 120 µm in rat. In rodents, juxtamedullary renal corpuscles are generally somewhat larger (by about 20%) than midcortical and superficial corpuscles (Tisher and Brenner, 1989; Kriz et al., 1992); however, this does not apply to the human kidney.

The glomerular tuft—a ‘wonder net’ At the entrance to Bowman’s capsule the afferent arteriole divides into several (three to five) primary capillary branches (Figs 43.1 and 43.3) (Yang and Morrison, 1980). Each of these branches gives rise to a capillary network (glomerular lobule) which runs towards the urinary pole, turns back towards the vascular pole, and unites

with tributaries form the other lobules to form the efferent arteriole. The separation into lobules is not strict; near the vascular pole anastomoses are found interconnecting the lobules, but they are not formed between the afferent and efferent portions of the same lobule (Winkler et al., 1991).The efferent arteriole develops deep within the centre of the glomerular tuft from the confluence of tributaries from all lobules. Thus, in contrast to the afferent arteriole, the efferent arteriole has an intraglomerular segment. Upon leaving the glomerulus, the efferent arteriole is closely associated with the extraglomerular mesangium. The intraglomerular segment is completely surrounded by mesangium; a smooth muscle layer is gradually established until the efferent arteriole leaves the extraglomerular mesangium.

Topography of a glomerular lobule—glomerular capillaries are unique The capillary network, together with the mesangium, is surrounded by the GBM followed by the visceral epithelium (podocytes). Deep invaginations of the GBM separate the tuft into lobules, less deep invaginations separate individual capillaries. Since the GBM does not completely encircle the capillary tube, a small portion of the endothelium is directly attached to the mesangium. The mesangial–endothelial interface (juxtamesangial portion) comprises only a small portion of the capillary circumference. The major part of the endothelial tube is in close contact with the GBM (pericapillary or peripheral portion) and the layer of interdigitating foot processes of the visceral epithelium. This part of the capillary wall represents the actual filtering surface (Fig. 43.4). At the points where the capillaries come into contact with the mesangium, the GBM and the podocyte layer deviate from a pericapillary course and cover the mesangium; these points have been called mesangial angles. Therefore, two parts of the GBM and the visceral epithelium can be distinguished:  a pericapillary part and a perimesangial part. The pericapillary part of the GBM is smooth and follows the outline of the capillary, whereas the perimesangial part is irregular in thickness and is frequently wrinkled.

Chapter 43 

EA

structure of the glomerulus

AA

EGM

MD

EA N

G

AA E PO

PE

F

M

GBM

Fig. 43.1  Scanning electron micrograph (400×) showing a vascular cast of two juxtamedullary glomeruli (rat). Each capillary tuft is supplied by an afferent arteriole (AA) which, on the surface of the tuft, immediately divides into several branches. Efferent arterioles (EA) emerge out of the centre of the tuft.

Capillary endothelium—a perforated, highly charged structure The capillary tube is made up of a particular kind of fenestrated endothelial cells (Fig. 43.4). The ‘fenestrae’ actually represent round to oval pores of varying in size from 50 to 100 nm in diameter (Fig. 43.5). Unlike endothelia with diaphragm-bridged fenestrae at other sites of the body (e.g. peritubular capillaries) the endothelial pores in mature glomerular capillaries lack a diaphragm corroborated by the lack of immunreactivity to the glycoprotein PV-1 (Stan et al., 2004). In the glomerulus an endothelium with diaphragm-bridged fenestrae is only found in the direct tributaries to the efferent arteriole. For details, see the work by Elger and colleagues (Elger et al., 1991; Elger et al., 1998). PV-1 is considered an essential molecule to the formation of both stomatal (caveolar) and fenestrial diaphragms (Stan, 2005). Both of them as well as PV-1-positive endothelial cells are abundant during development of glomerular endothelia (Ichimura et al., 2008). Thus, the formation of transendothelial pores lacking diaphragms is preceded by the formation of fenestrae with diaphragms (Satchell et al., 2009). The same is true during the recapillarization of the glomerular tuft, for example, in Thy-1 nephritis. Of note, the formation of glomerular endothelial fenestrae is independent of caveolin-1 (a caveolar protein) indicating that fenestrae do not develop from caveolae (Drab et al., 2001; Sörensson et al., 2002). The cell bodies of endothelial cells are generally located adjacent to the mesangial interface. The peripheral flat ‘porous’ parts of the cells comprise about 60% of the capillary surface. Individual pores are encircled by a network of microfilaments (Vasmant et al., 1984). Clusters of pores are separated by ridges of cytoplasm, containing intermediate filaments and microtubules. Pores occupy about 13% of the capillary surface in the rat (Bulger et al., 1983); in absolute

US

PT

Fig. 43.2  Diagram of a longitudinal section through a renal corpuscle and the juxtaglomerular apparatus (JGA). The capillary tuft consists of a network of specialized capillaries, which are outlined by a fenestrated endothelium (E). At the vascular pole an afferent arteriole (AA) enters and an efferent arteriole (EA) leaves the tuft. The capillary network is surrounded by Bowman’s capsule, comprising two different epithelia: the visceral and the parietal epithelium. The visceral epithelium consisting of highly branched podocytes (PO) directly follows—together with the glomerular basement membrane (GBM)—the surface of the capillaries and the mesangium (M). At the vascular pole, the visceral epithelium and the GBM are reflected into the parietal epithelium (PE) of Bowman’s capsule (and its basement membrane), which passes over into the epithelium of the proximal tubule (PT) at the urinary pole. Mesangial cells (M) are situated in the axes of glomerular lobules. At the vascular pole the glomerular mesangium is continuous with the extraglomerular mesangium (EGM), consisting of cells and matrix. The EGM, together with the terminal portion of the afferent arteriole (containing the granular cells, G), the efferent arteriole, and the macula densa (MD), establish the JGA. All cells that are suggested to be of smooth muscle origin are shown in a dark colour. F = foot processes; N = sympathetic nerve terminals; US = urinary space.

terms, the total area of all pores in a rat glomerulus amounts to about 22 mm2 (Larsson et al., 1980). Under clinical settings, loss of endothelial pores correlates with decrease of glomerular filtration rate, as was shown in diabetes type 1 and type 2 and in pre-eclampsia (Lafayette et al., 1998; Toyoda et al., 2007; Weil et al., 2011; Salmon and Satchell, 2012; Satchell, 2012). A layer of membrane-bound and loosely attached molecules covers the capillary endothelium on its luminal side. Depending on the visualizing technique it amounts to a thickness of about 300–500 nm and covers the entire surface including the endothelial pores (Haraldsson et al., 2009). This entire layer may be called

437

438

Section 3  

the patient with glomerular disease

PO US AA

MD

G

E C EA

EGM C

PE E

PE

cGBM

PO

US

C

M C

mGBM

* Fig. 43.3  Longitudinal section through the glomerular vascular pole showing the juxtaglomerular apparatus with both arterioles (rat). At the entrance into the glomerulus, the afferent arteriole (AA) immediately branches into capillaries (C). The efferent arteriole (EA) usually arises deeper in the tuft and can be identified by the high number of endothelial cells (E) at the exit from the glomerulus. The macula densa (MD) of the thick ascending limb is in contact with the extraglomerular mesangium (EGM) and the glomerular arterioles. The media of the AA contains granular cells (G). M, mesangial cells; PE, parietal epithelium; PO, podocytes; US, urinary space. Transmission electron micrograph (1300×).

glycocalyx. Due to polyanionic glycoproteins it is negatively charged (Horvat et al., 1986; Sawada et al., 1986). The components of the glycocalyx proper (Pries et al., 2000) are covalently bound to the endothelial cell membrane. The thickness of this sublayer amounts to about 50–100 nm and is composed of the membrane-bound proteoglycans such as glypican (with heparan sulphate side chains) and syndecan (with chondroitin and heparan sulphate side chains) (Haraldsson et al., 2008). Attached to components of the glycocalyx proper are secreted proteoglycans, for example, versican and perlecan, hyaluronan (a non-sulphated glycosaminoglycan), as well as adsorbed plasma proteins (including orosomucoid and albumin) (Haraldsson et al., 2008). This layer of more loosely attached components is vulnerable, for example, by haemodynamic factors (Ryan et  al., 1976; Fridén et  al., 2011; Haraldsson et al., 2012) it can be visualized by intravital (Desjardins et al., 1990) and electron microscopic techniques (Rostgaard et al., 2002; Hjalmarsson et al., 2004). Damage to the glomerular endothelial glycocalyx is clinically apparent as albuminuria (Obeidat et al., 2012; Salmon and Satchell, 2012) (see below). Experimentally, degradation of glycosaminoglycans in glomerular capillaries increases the clearance of albumin (Jeansson et al., 2006), and administration of hyaluronidase causes

E

M C

C

Fig. 43.4  Part of a glomerular lobule (rat), showing the arrangement of structures in the glomerular tuft. The capillary (C) is outlined by a flat fenestrated endothelium (E). The podocyte layer (PO) and the glomerular basement membrane (GBM) do not encircle the individual capillary completely, they form a common surface cover around the lobule. In the peripheral portion of the capillary the filtration barrier is formed (see also Fig. 43.5). Two subdomains of the GBM are delineated from each other by mesangial angles (arrows): the pericapillary GBM (cGBM) faced by the podocyte layer and the endothelial layer, and the perimesangial GBM (mGBM) bordered by the podocyte layer and the mesangium. Within the mesangium two types of cells are shown: contractile mesangial cells (M) and a cell (*) which is probably a macrophage that has invaded the mesangium. Note the intimate relationships between the endothelium and the mesangium (arrowheads). US = urinary space. 6100×.

proteinuria (Meuwese et al., 2010). Removal of sialic acid residues results in albuminuria (Gelberg et al., 1996; Bakker et al., 2005). Elution of non-covalently bound components of the glomerular endothelial glycocalyx caused a 12-fold increase in the fractional clearance of albumin (Fridén et al., 2011). Thus, both soluble and bound components of the endothelial glycocalyx determine permeability of the glomerular filtration barrier. Endothelial cells are active participants in processes controlling coagulation, inflammation, and immune processes, and an aberration in the controlling mechanisms may contribute to the development of disease within the glomerulus (Savage, 1998). Renal endothelial cells share an antigen system with cells of the monocyte/macrophage lineage; they express surface antigens of the

Chapter 43 

structure of the glomerulus

US F

3 2 1

P E C

Fig. 43.5  Filtration barrier. The peripheral part of the glomerular capillary wall comprises the fenestrated endothelial layer (E), the glomerular basement membrane, and the interdigitating foot processes (F). The filtration slits between the foot processes are bridged by thin diaphragms (long arrows). Arrowheads point to the endothelial pores. The glomerular basement membrane shows a lamina densa (2) bounded by the lamina rara interna (l) and the lamina rara externa (3). In this picture, tannic acid staining allows discrimination between the alternating foot processes of two neighbouring podocytes: the more densely stained processes belong to one cell, and the others to the neighbouring cell. C = capillary lumen. 60,000×.

class II histocompatibility antigens. Similar to platelets, glomerular endothelial cells contain components of the coagulation pathway and are capable of binding factors IXa and Xa, and of synthesizing, releasing, and binding von Willebrand factor (factor VIII) (Wiggins et  al., 1989). Glomerular endothelial cells synthesize and release endothelin-1 and endothelium-derived relaxing factor (EDRF) (Ott et al., 1993; Herman et al., 1998).

Visceral epithelium—filtration slits are formed by interdigitating foot processes The visceral epithelial layer of Bowman’s capsule consists of highly differentiated cells, the podocytes (Kriz et al., 2013a). The podocytes are attached to the outer surface of glomerular capillaries, that is, to the GBM only by their processes, their cell bodies float within the filtrate in Bowman’s space. In the developing glomerulus, the visceral epithelium consists of simple polygonal cells. In rat, mitotic activity of these cells is completed soon after birth, along with the cessation of the formation of new nephron anlagen (Nagata et al., 1993), and the final number of podocytes is determined. In humans this point is reached during prenatal life. Differentiated podocytes are unable to replicate (Fries et al., 1989; Nagata and Kriz, 1992; Kriz, 2002), thus, in the adult degenerated podocytes cannot be replaced. In response to an extreme stimulation (long-term treatment with FGF-2) the nucleus may enter into mitosis; however, the cells are not able to complete cell division (cytokinesis), usually resulting in mitotic catastrophe or, at best, in binucleated cells (Kriz et al., 1995b). Recent evidence that there may be podocyte progenitors at the junction of the parietal stem cells with proximal tubular cells (see Chapter 344) notwithstanding, there is little evidence for significant replacement in adults. Podocytes have a voluminous cell body, which bulges into the urinary space (Figs 43.6 and 43.7). Long primary processes emerge from the cell body and extend towards the capillaries, to which they affix by their distal portions and their final ramifications the so-called foot processes. The foot processes of neighbouring podocytes regularly interdigitate with each other, leaving between them

F

Fig. 43.6  Scanning electron micrograph (3300×) of rat glomerular capillaries. The urinary side of the capillary is covered by the highly branched podocytes. The interdigitating system of primary (P) and secondary (F) processes lines the entire surface of the glomerular basement membrane and proceeds also beneath the cell bodies (see Fig. 43.7). In between the interdigitating foot processes (F) of neighbouring cells the filtration slits are spared.

meandering slits (filtration slits), which are bridged by a thin proteinaceous membrane (slit membrane or slit diaphragm). Podocytes are polarized epithelial cells with a luminal (apical) and an abluminal (basal) cell membrane. The basal domain corresponds to the sole plates of the foot processes, which are embedded into the GBM up to approximately 60 nm. The border between basal and luminal membranes is represented by the slit diaphragm. The luminal membrane, including the slit diaphragm, is covered by a thick surface coat, which is rich in sialoglycoproteins that are responsible for the high negative surface charge of podocytes. They include podoendin (Huang and Langlois, 1985), SGP115/107 (Mendrick and Rennke, 1988), and podocalyxin (Kerjaschki et al., 1984; Sawada et al., 1986), which via the linker protein NHERF2 (Na+/H+-exchanger regulatory factor 2) and ezrin is attached to the actin cytoskeleton. The surface charge of podocytes contributes to the maintenance of the interdigitating pattern of the foot processes. In response to neutralization of the surface charge by cationic substances (e.g. protamine sulphate, poly-l-lysin), the glomerular epithelium undergoes a series of changes, including retraction of foot processes and formation of adhesive junctions between adjacent foot processes (Seiler et al., 1977). Podocytes display many surface receptors and ion channels that are generally found on the entire surface of podocytes, many of them accumulate close to the slit membrane; the schematic in Fig. 43.9 shows some of them. They include receptors for angiotensin II (AT1 and AT2; Nitschke et al., 2000; Sharma et  al., 2001), noradrenaline (α1; Huber et  al., 1998), acetylcholine (M5; Nitschke et al., 2001), prostaglandin (Bek et al., 1999), ATP (Fischer et  al., 2001), endothelin (ETA; Rebibou et  al.,

439

440

Section 3  

the patient with glomerular disease

(A)

ER

GO

P

C (B) IF

MT

MF

C

Fig. 43.7  (A) Electron micrograph (25,000×) showing part of a podocyte cell body anchored via processes to the glomerular basement membrane. The majority of these processes are foot processes. The others are ridge-like bases of the cell body and distal portions of primary processes (P). Note the prominent Golgi apparatus (GO); also rough endoplasmic reticulum (ER) is fairly abundant. C = capillary lumen. (B) Primary and secondary processes of podocytes showing cytoskeletal elements. Intermediate filaments (IF) and microtubules (MT) are abundant in the primary processes; thick bundles of microfilaments (MF) are located in the foot processes. C = capillary lumen. 46,000×.

1992; Spath et al., 1995), ANP (Zhao et al., 1994), PTH/PTHrP (Jorgensen, 1966; Endlich et al., 2001b), TGFβ (Yamamoto et al., 1998), IL-4/IL-13 (van den Berg et al., 2000), FGF-2 (Ford et al., 1997), C3b-receptor (Kazatchkine et al., 1982), and gp330/megalin (Kerjaschki and Farquharl, 1983). Megalin, a glycoprotein of 330 kDa (Saito et al., 1994) is the major podocyte antigen of rat Heymann nephritis; in man megalin is lacking. In addition to non-specific cation channels (NSCC), the specific Ca2+ channels TRPC5 and TRPC6 (Greka et al., 2012) are currently in discussion as important regulators of foot process dynamics. The cell body contains a prominent nucleus, a well-developed Golgi apparatus, abundant rough and smooth endoplasmic reticulum, prominent lysosomes, and mitochondria (Fig. 43.7A). In contrast to the cell body, the cell processes contain only a few organelles.

The organelles in the cell body indicate a high level of anabolic as well as catabolic activity. In addition to the work necessary to sustain structural integrity of these specialized cells, most components of the GBM are synthesized by the podocytes (Abrahamson, 1987, 2012). The cell contains a well-developed cytoskeleton. In the cell body and the primary processes, microtubules and intermediate filaments (vimentin, desmin) dominate (Fig. 43.7B). Microfilament bundles containing actin, myosin, and α-actinin are found in a highly organized pattern in the foot processes. In addition, in the cell body and the primary processes, microfilaments are seen as a thin layer underlying the cell membrane (Bachmann et al., 1983; Vasmant et al., 1984; Drenckhahn et al., 1988; Kriz et al., 1995a). Bundles of microtubules and intermediate filaments extend from the cell body to the distal end of the primary processes. The microtubules are of mixed polarity, which appears to be essential for process formation of podocytes (Kobayashi et al., 1998). As intracellular transport systems the microtubules are probably involved in the transport of substances (e.g. for the GBM) to the peripheral parts of the cell. Another important function of the microtubules and intermediate filaments may be found in their cytoskeletal properties in relation to mechanical forces. Microtubules resist compression of their long axis and, as bundles interconnected by microtubule-associated proteins, they withstand bending forces. Intermediate filaments, on the other hand, are able to resist tensile forces (Wang et al., 1993). In the foot processes, a prominent actin based contractile apparatus is present consisting of bundles of microfilaments running longitudinally through the processes. At the transition to the primary processes, the microfilament bundles form loops that are connected to the microtubules of the primary processes. Peripherally, the microfilament bundles anchor in the dense cytoplasm associated with the inner aspect of the basal cell membrane (Fig. 43.8). The bases of the foot processes are firmly attached to the GBM, mediated by integrins and dystroglycans (Fig. 43.9). The integrin complex is comprised mainly of α3β1 integrin dimers, which are connected inside the cell to the complex of vinculin, paxillin, and talin, and outside the cell to collagen IV α3, α4, and α5 chains as well as to laminin α5/β2/γ1 (Miner, 1999; Kreidberg and Symons, 2000). In addition, further integrin dimers have been found (Schordan et al., 2010). The dystroglycan complex (Durbeej et al., 1998; Raats et al., 2000; Regele et al., 2000) consists of the cytoplasmic adaptor protein utrophin, of the transmembranous β-dystroglycan, and of the extracellular matrix-binding α-dystroglycan which is a receptor for agrin and laminin α5 chains. In addition to the important mechanical relevance of these connections (which is insufficiently understood), outside-in signalling via these systems quite obviously influences the function of the cytoskeleton which, in cases of any stress to the podocytes, may lead to foot process effacement (FPE) (see below). The filtration slits are the site of convective fluid flow through the visceral epithelium. The total length of the filtration slit in a rat glomerulus amounts to about 50 cm (Kriz et al., 1995a; Mundel and Kriz, 1995). Since the slit has a rather constant width of 30–40 nm, the total area of the slit membrane approximates 20  µm2 × 103, comprising about 10–13% of the peripheral capillary surface. The structure and biochemical composition of the slit membrane are still incompletely understood. Chemically fixed and tannic acid treated tissue reveals a zipper-like structure with a row of ‘pores’ (approximately 4  × 14  nm in size on either side of a central bar

Chapter 43 

Relevance of podocytes

(A)

(B)

structure of the glomerulus

Foot processes with microfilaments y

w

Major processes with microtubules

x

z (C)

(D)

Fig. 43.8  Arrangement of cytoskeletal elements in podocyte processes: (A) as seen in a cross section through a capillary (B) view from above; (C) section of foot processes parallel (along the line w to x) and (D) perpendicular (along the line y to z) to the longitudinal axis of foot processes. Two major processes (one in white, one in yellow) with their foot processes are shown. The actin filaments (red) of foot processes form continuous loops which terminate in the foot process sole plates. At their bends they are in close association with microtubules (green) that run longitudinally in the major processes. After Mundel et al. (1995).

(Rodewald et al., 1974). In quick-frozen tissue, a more homogeneous structure with only a central bar is apparent (Hora et al., 1990). Proteins that establish the slit membrane include nephrin, Neph1, 2, and 3, p-cadherin, and FAT. Presently, it is not clear how these proteins participate in the molecular organization of this structure. The cytoplasmic tails of these proteins allow a dynamic connection of the slit membrane to the cytoskeleton. Proteins that mediate and/or regulate this connection include ZO1, α-, ß-, γ-catenins, podocin, CD2AP, and α-actinin 4. Fig. 43.9 summarizes some main features of the molecular organization of podocyte foot processes (modified after Endlich et al., 2001a).

It is difficult to precisely define the function of podocytes; they are still a mysterious cell type. Generally, they have been considered as a modified type of a pericyte with specialized intercellular junctions, that is, the slit diaphragm; thus, as a supportive cell that, in addition, controls paracellular permeability. The pericyte function, counteracting the distension of the capillary wall, has been seriously challenged in recent time (Kriz et al., 2013b; see below). The most important function is provided by the slit diaphragm that essentially contributes to the barrier function of the glomerular filter. We will discuss this function below together with the possible relevance of the other layers of the glomerular barrier. As seen during glomerular development podocytes have the exclusive commandership in building a glomerulus. In the adult, they are responsible for the maintenance of the complex structure of the glomerular tuft, with vascular endothelial growth factor A (VEGF-A) playing the major role in the regulatory processes (Kriz, 2007). As shown recently in a transgenic model with deletion of β-catenin (Grouls et al., 2012) podocytes develop within the parietal epithelium of Bowman’s capsule followed by formation of a small but complete tuft-like structure consisting of a proper capillary supported by a mesangium, a GBM and covered by an interdigitating foot process layer.

Relevance of podocytes in glomerular pathology Podocytes are terminally differentiated cells incapable of replicating. Thus, lost podocytes cannot be replaced by proliferation of neighbouring undamaged cells. We are born with a certain number of podocytes, roughly 800 per glomerulus in the 2 million nephrons of the two kidneys. So far there is little evidence for much replenishment of this population from progenitors in adults (see Chapter 344). The only way to immediately compensate for lost podocytes consists of cell hypertrophy in order to cover the glomerular tuft with a smaller number of podocytes. This mechanism, however, increases their vulnerability to any challenge. Moreover, podocytes live in a precarious situation, being fixed to the outer aspect of glomerular capillaries only by their processes, their cell bodies float in the filtrate in Bowman’s capsule. This exposes podocytes to the danger of being lost by detachment. Indeed, podocytes are continually excreted as viable cells in the urine, and the rate of excretion dramatically increases in glomerular diseases (Hara et al., 1995, 1998; Nakamura et al., 2000; Vogelmann et al., 2003; Petermann et al., 2004; Yu et al., 2005; Weil et al., 2011). This fully agrees with structural findings in many models of glomerular disease showing that podocytes detach as viable cells from the GBM (Kriz et al., 2013b). This precarious situation leads to a very specific set of reactions of podocytes to challenges that may be interpreted as to serve for a stronger adhesion to the GBM, thus counteracting detachment and loss into the urine. FPE, that is, the retraction of the foot processes into the primary processes, finally into the cell bodies, can be seen as a mechanism to promote the adherence of podocytes to the GBM. Since FPE seems to be inevitably associated with proteinuria we come to the conclusion that changes in the slit diaphragm are secondary to FPE, and that the associated defect in the filtration

441

442

Section 3  

the patient with glomerular disease –

–

Actin

CI–

–

– N

–

M

–

NSCC

Ca2+

AT1

PC

S

–

–

–

– Ez

Z

–

α-Actinin-4

Laminin 521

– – Agrin

Ca2+

Podocin

–

–

ANG II

TRPC5 TRPC6

–

CD

Cas

U

Cat

TPV

FAK ILK

– – TPV

β

Nephrin NEPH 1–3 P-Cadherin FAT 1

α

β3

Dystroglycan

α v βI

α3

Integrine

– Collagen IV (α3, α4, α5)

– –

– –

–

– –

–

Fig. 43.9  Glomerular filtration barrier. Two podocyte foot processes bridged by the slit membrane, the GBM, and the porous capillary endothelium are shown. The surfaces of podocytes and of the endothelium are covered by a negatively-charged glycocalyx containing the sialoprotein podocalyxin (PC). The GBM is mainly composed of collagen IV (α3, α4, and α5), of laminin 11 (α5, β2, and γ1 chains) and the heparan sulphate proteoglycan agrin. The slit membrane represents a porous proteinaceous membrane composed of (as far as is known) Nephrin, Neph1, 2, and 3, P-cadherin, and FAT1. The actin-based cytoskeleton of the foot processes connects to both the GBM and the slit membrane. With regard to the GBM, β1/α3 integrin dimers specifically interconnect the TVP complex (talin, paxillin, vinculin) to laminin 11; the β and α dystroglycans interconnect utrophin to agrin. The slit membrane proteins are joined to the cytoskeleton via various adaptor proteins, including Podocin, Zonula occludens protein 1 (ZO-1; Z), CD2 associated protein (CD), and catenins (Cat). The Ca2+ channels TRPC6 and TRPC5 are concentrated at the slit membrane. Among the many surface receptors only the angiotensin II (ANG II) type 1 receptor (AT1) is shown. Cas= P130Cas; Ez = ezrin; FAK = focal adhesion kinase; ILK = integrin-linked kinase; M = myosin; N = NHERF2 (Na+/H+-exchanger regulatory factor); NSCC = non-selective cation channel; S = synaptopodin. Modified from Endlich et al. (2001a).

barrier leading to albumin leakage is, under stressful situations, not a top priority for the podocytes. The survival of the podocyte is more important and protein leakage is, at least in part, merely an epiphenomenon that occurs in many situations when podocytes under stress make adaptations to resist detachment. These two weaknesses of podocytes (inability to replicate and the danger of being lost by detachment) are responsible that loss of podocytes occupies centre stage in glomerular pathology. The progressing loss of podocytes is essentially responsible for the progression of chronic renal disease to end stage (see Chapter 139).

Glomerular basement membrane—the backbone of the glomerular tuft The GBM represents the skeletal backbone of the glomerular tuft (Fig. 43.4). In transmission electron micrographs of traditionally fixed tissue, the GBM appears as a trilaminar structure made up of a lamina densa bounded by two less dense layers—the lamina rara interna and externa (Fig. 43.5). In humans, the thickness of the GBM ranges from 300 to 370 nm (Steffes et al., 1983); in children it may be considerably thinner (Morita et al., 1988). In rats and other experimental animals, the thickness is between 110 and 190 nm (Rasch, 1979; Bulger et al., 1988).

In accordance with basement membranes at other sites, the major components of the GBM include type IV collagen, heparan sulphate proteoglycans, and laminin (Mohan and Spiro, 1986; Timpl and Dziadek, 1986). Types V and VI collagen, and entactin/nidogen have also been demonstrated. On the other hand, the GBM has many unique properties, notably a distinct spectrum of type IV collagen and laminin isoforms (Couchman et al., 1994; Miner, 1999; Abrahamson, 2012), described in Chapter 320. At early stages of glomerulogenesis, only α1 and α2 chains are detected, but as the glomerular capillaries begin to mature, there is a gradual increase in α3, α4, and α5 chains (Miner et al., 1994). Consistent with the distribution of the isoforms, podocytes—but not endothelial cells—synthesize the α3α4α5 (345) network (St John and Abrahamson, 2001; Abrahamson et al., 2009), whereas the α1 and α2 chains (112 network) likely derive from the glomerular endothelial cells (Miner, 2011) At early stages of glomerulogenesis, Laminin LM-111 and LM-511 are the major laminin components (Miner et al., 1997). However, as the capillaries begin to mature, LM-521 begins to be deposited by the podocytes and endothelial cells, and LM-111 and -511 are eventually eliminated (Miner et  al., 1994; Miner et  al., 1997). Mice lacking β2 develop massive proteinuria and FPE, and die during early postnatal life, known in humans as

Chapter 43 

Pierson syndrome (Hansen and Abrass, 1999; Miner, 1999; Zenker et al., 2004). The major role in mediating the interconnection of the various components of the GBM (as well as the connection to the surface receptors of podocytes and endothelial cells) is played by laminin (and entactin/nidogen). The relative amount of components varies between basement membranes of different sources. The GBM contains more type IV collagen and less laminin than does the renal tubular basement membrane (Brees et al., 1995). Because of covalent cross-linking, the type IV collagen network provides a stronger scaffolding to render the basement membrane more resilient and permanent than a laminin polymer (Yurchenco and Cheng, 1994). The higher content of type IV collagen in the GBM than in the tubular basement membrane may be indicative of a greater tensile strength of the GBM adapted to the high transmural pressure differences at this site.

Filtration barrier—filtration occurs along an extracellular pathway The filtration barrier is composed of (a) the endothelium with large open pores, (b) the dense matrix network of the GBM, and (c) the slit diaphragms between the podocyte foot processes. Compared with the barrier established in capillaries elsewhere in the body, there are at least two outstanding characteristics of the filtration barrier in the glomerulus: the permeability for water, small solutes, and ions is extremely high, while the permeability for plasma proteins the size of albumin and larger is very low. The high hydraulic permeability is easily explained by the fact that filtration occurs along extracellular routes. All components of this route, the endothelial pores (including the glycocalyx), the highly hydrated GBM, and the slit membrane can be expected to be quite permeable for water and small solutes. The hydraulic conductance of the individual layers of the filtration barrier is difficult to examine. In a mathematical model of glomerular filtration the hydraulic resistance of the endothelium was predicted to be small, whereas the GBM and filtration slits each contributed roughly one-half of the total hydraulic resistance of the capillary wall (Drumond and Deen, 1994). Harper argues that the existence of a sub-podocyte space should alter our views on free filtration beyond the slit diaphragm (Salmon et al. 2009). In pathological models as well as in human glomerulopathies such as membranous nephropathy or minimal change nephropathy, FPE leads to a drastic reduction of the overall filtration slit length. This decrease in slit length (or slit frequency) is correlated with a decrease in the ultrafiltration coefficient, Kf, (Kiberd, 1992; Guasch and Myers, 1994). The decrease in total slit membrane area also causes an increase in the average path length for the filtrate through the GBM, thereby explaining the decreased hydraulic permeability in these nephropathies (Drumond et al., 1994). The barrier function for macromolecules is based on the size, shape, and charge of the respective molecule (reviewed in Daniels et al., 1993; Deen et al., 2001); the relevance of each of these parameters is in debate. Our interpretation of the available data is that a size/shape barrier for very large molecules (effective radii of > 4.0 nm) is provided by the slit membrane (Deen et  al., 2001). Since most of plasma

structure of the glomerulus

proteins, including albumin, are negatively charged, their repulsion is mainly charge dependent. The size/shape selectivity for macromolecules of the filtration barrier seems to be established by the slit membrane (Deen et al., 2001). Uncharged macromolecules up to an effective radius of 1.8 nm pass freely through the filter. Larger compounds are more and more restricted (indicated by their fractional clearances which progressively decrease), and are totally restricted at effective radii of > 4.0 nm. The term ‘effective radius’ is an empirical value, measured in artificial membranes, which takes into account the shape of macromolecules and attributes a radius to non-spherical molecules. Plasma albumin has an effective radius of 3.6 nm; without the repulsion due to the negative charge, plasma albumin would pass through the filter in considerable amounts (see below) (Deen et al., 1979). The importance of the slit diaphragm for size selectivity is evidenced by experiments with ferritin (radius 6.1  nm). Whereas anionic ferritin particles accumulate at the level of endothelial fenestrae and the subendothelial space, cationized ferritin penetrates the lamina densa and accumulates beneath the slit diaphragm. The charge barrier may be seen to consist of two components. First, negatively charged molecules are accumulated throughout the entire depth of the filtration barrier, including the surface coats of endothelial and epithelial cells, and the high content of negatively charged heparan sulphate proteoglycans in the GBM. The glycocalyx/surface coat of the endothelium seems to be most effective, establishing an electronegative shield at the entry side (discussed in detail above in conjunction with the endothelium). Polyanionic macromolecules in the plasma, such as albumin, are repelled by these assemblies of negatively charged molecules, but the effect is not complete. A minor portion of such molecules penetrates into the filter and the mechanism how it is prevented from being excreted into the urine is under debate. A recent study by the group of Marcus Moeller (Hausmann et al., 2010) proposes an electrophoretic mechanism for the repulsion of this fraction of negatively charged macromolecules. According to their hypothesis (based on direct measurements in the Necturus kidney) the convective flow of the filtrate through the filter creates a potential difference that increases the negativity of the urinary side of the glomerular filter compared to the capillary side by up to −0.05 mV. Thus, albumin molecules that enter the filter will, on their way through the filter, be exposed to increasingly negatively charged surroundings. Thereby, they will be repelled at various depths in the filter and forced to diffuse back into the capillary. The charm of this hypothesis consists of being independent of any structural pore size. The barrier consist of a strictly filtration dependent potential difference. Thus, without sufficient convective flow of filtrate no barrier will be created. Among other attractive features this might help explain the so-called orthostatic proteinuria that is observed in patients only in upright position but not when supine (Chapter 51): a low perfusion pressure does not generate sufficient flow through the filter to create a charge potential.

Mesangium—maintenance of the structural integrity of the tuft The mesangium occupies the axial region of a glomerular lobule and consists of mesangial cells and the surrounding mesangial matrix (Figs 43.4 and 43.10), first described by (Zimmermann,

443

444

Section 3  

the patient with glomerular disease

M

Fig. 43.10  Glomerular capillary and mesangium. In the juxtacapillary region, long mesangial cell processes extend between opposite mesangial angles, where they are fixed to the GBM (arrows). In the axial region, finger-like processes connect the mesangial cells to the perimesangial glomerular basement membrane (arrowheads). Note bundles of microfilaments in the cell processes. M = mesangial cell. 13,000×.

1929). Since the ultrastructural characterization of the mesangium in the early sixties (Latta et al., 1960; Farquhar et al., 1962), mesangial cells have been in the forefront of glomerular research. They are generally believed to form a supporting framework that maintains the structural integrity of the glomerular tuft. The mesangial matrix consists mostly of basement membrane components (Karkavelas et al., 1988). At variance to the GBM it contains the α1 and α2 chains of type IV collagen, the β1 chain of laminin, considerable amounts of fibronectin, chondroitin sulphate proteoglycan, the small leucine-rich proteoglycans biglycan and decorin and the heparan sulphate proteoglycans perlecan, bamacan, and collagen type XVIII (Border et al., 1989; Couchman et al., 1994; Miner, 1999); in addition, microfibrillar proteins are abundant (Gibson et al., 1989; Sterzel et al., 2000). Mesangial cells are considered to be contractile cells that have a common origin with smooth muscle cells. They are irregular in shape, with numerous cytoplasmic processes filled with prominent assemblies of microfilaments (Fig. 43.10) that contain actin, myosin, and α-actinin (Kreisberg et al., 1985; Drenckhahn et al., 1988). Moreover, the mesangial cells are electrically coupled by gap junctions (Pricam et  al., 1974). Mesangial cells possess a great variety of receptors including for angiotensin II, vasopressin, atrial natriuretic factor, prostaglandins, TGFβ, PDGF-B, EGF, and CTGF (Dworkin et al., 1983; Stockand and Sansom, 1998).

A  more complete list is found in a review by Schlöndorff and Banas (2009). The GBM is the primary effector site of mesangial cell contraction (Sakai and Kriz, 1987; Kriz et al., 1990). Mesangial cells are connected extensively with the GBM, either by direct adherence of mesangial cell processes (focal adhesions) or by microfibrils (Fig. 43.10). These connections, which appear to be mechanically strong, are found throughout the mesangial region. Microfibrils are unbranched non-collagenous, tubular structures about 15 nm thick. They are a major component of the mesangial matrix, as has been shown by transmission electron microscopy after tannic acid staining (Mundel et al., 1988) and by immunocytochemistry using antibodies against several microfibrillar proteins (Gibson et al., 1989). Microfibrils are generally coated by fibronectin (Schwartz et al., 1985). The high content of fibronectin within the mesangium (Madri et  al., 1980)  may be related to the need for firm connections between the different mesangial components: cells to matrix and among the various components of the matrix itself. Fibronectin mediates the connection between actin and extracellular matrix components, including type IV collagen (Burridge et  al., 1988). Important in these connections are α3β1 dimers of integrins, which bind fibronectin to the termini of actin filaments (Cosio et  al., 1990). As a whole, these abundant interconnections between the cells and matrix as well as between the various matrix components establish a strong mechanical cohesion to counteract the expansion of the mesangium.

Parietal epithelium The parietal layer of Bowman’s capsule consists of squamous epithelial cells resting on a basement membrane (Figs 43.2 and 43.3). The flat cells are polygonal, with a central cilium and few microvilli. Parietal cells are filled with bundles of actin filaments running in all directions (Pease, 1968). Within the cells surrounding the vascular pole, the actin filaments are very dense and located within cytoplasmic ridges that run in a circular fashion around the glomerular entrance. The finding of muscarinic receptors on the parietal epithelium (Lebrun et al., 1992) indicates that the contractile tone of these cells is subject to regulation. Recent observations suggest that a niche of glomerular epithelial stem cells resides within the parietal epithelium at the transition to the proximal tubule (Sagrinati et al., 2006; Ronconi et al., 2009). It is an intriguing hypothesis that proliferating stem cells from this locus may transform into podocytes and may reach the tuft by migration via the transition at glomerular vascular pole. Migration of parietal cells onto the tuft via the vascular pole and subsequent transition into podocytes have been shown to occur in the new-born mouse (Appel et al., 2008). However, evidence that such a process may be of any relevance in the adult has so far not been presented (Appel et al., 2008). The PBM is, in contrast to the GBM, composed of several dense layers separated by translucent layers and contains bundles of fibrils (‘microligaments’; Mbassa et al., 1988). Collagen α1, α2, α5 and α6 (type IV) chains prevail in this basement membrane (Miner, 1999). In addition collagen type XIV was found (Lethias et al., 1994). In contrast to the GBM, the predominant proteoglycan of this basement membrane is a chondroitin sulphate proteoglycan, and the laminin isoforms 1 and 10 prevail (Couchman et  al., 1994; Miner, 1999). The transition from the GBM to the basement membrane of Bowman’s capsule borders the glomerular entrance. This

Chapter 43 

transitional region is mechanically connected to the smooth muscle cells of the afferent and efferent arterioles as well as to extraglomerular mesangial cells (see below).

Extraglomerular mesangium—a closure device of the glomerular entrance At the vascular pole of the glomerulus the mesangium passes through the opening of Bowman’s capsule and continues into the extraglomerular mesangium (Barajas et  al., 1989; Barajas, 1997) (Figs 43.2 and 43.3). The extraglomerular mesangium represents a solid complex of cells and matrix that is neither penetrated by blood vessels nor lymphatic capillaries. The extraglomerular mesangium is located in the cone-shaped space between the two glomerular arterioles and the macula densa cells of the thick ascending limb and, laterally, faces the renal interstitium. Extraglomerular mesangial cells are flat and elongated, separating into bunches of long cell processes at their poles (Spanidis and Wunsch, 1979). They are arranged in several layers, parallel to the base of the macula densa. The cells are embedded in a matrix similar in composition as the mesangial matrix; however, microfibrils are comparably rarely found. Affixation of macula densa cells to the extraglomerular mesangium appears to be mediated by β6-integrin, which is known to associate with αv-integrin to form the fibronectin binding heterodimer αvβ6 (Breuss et al., 1993). Although direct evidence is lacking, extraglomerular mesangial cells can be expected to be contractile for several reasons. First, they contain prominent bundles of microfilaments containing F-actin in their processes. Second, like intraglomerular mesangial cells, they have strong structural similarities with arteriolar smooth muscle cells and granular cells, suggesting that they are all of the same origin. Third, these cells are all extensively coupled by gap junctions (Pricam et al., 1974; Taugner et al., 1978). Fourth, high amounts of heat shock protein 25 (HSP25) are present in glomerular and, especially, in extraglomerular mesangial cells. HSP25/27, which is believed to be a mediator of sustained smooth muscle cell contraction, might be a component of the contraction machinery in the glomerular and extraglomerular cells (Müller et al., 1999). The contractile processes of extraglomerular mesangial cells are connected to the basement membrane of Bowman’s capsule and to the walls of both glomerular arterioles. As a whole, the extraglomerular mesangium interconnects all structures of the glomerular entrance. The extraglomerular mesangium can be regarded as a closure device of the glomerular entrance, maintaining its structural integrity against the distending forces exerted on the entrance by the high intra-arteriolar and intraglomerular pressure (Elger et al., 1998). A function of the extraglomerular mesangium for the recruitment of mesangial cells has been proposed. In anti-Thy-1 glomerulonephritis the cellular re- population of the mesangium apparently occurs from the extraglomerular mesangium (Hugo et al., 1997).

Juxtaglomerular apparatus—intersection of local and systemic regulation The juxtaglomerular apparatus (JGA) is situated at the vascular pole of the glomerulus. It comprises: (a) the macula densa, (b) the extraglomerular mesangium (described above), and (c) the terminal portion of the afferent arteriole with its renin-producing granular cells, as well as the beginning of the efferent arteriole (Fig. 43.3).

structure of the glomerulus

The macula densa is a plaque of specialized cells within the thick ascending limb at the site where the latter is affixed to the extraglomerular mesangium of the parent glomerulus. The most obvious structural features are the large, narrowly packed, cell nuclei, which account for the name ‘macula densa’ (Zimmermann, 1933). In contrast to other parts of the thick ascending limb, the cells of the macula densa do not interdigitate with each other but have a polygonal outline. The luminal cell membrane is densely studded by stubby microvilli and bears one cilium. At their bases, the cells display numerous infoldings and folds of the plasma membrane; the latter are anchored to the underlying basement membrane, blending with the matrix of the extraglomerular mesangium (Kriz et al., 1992). The lateral membrane of macula densa cells bears folds and finger-like villi which are frequently connected to those of neighbouring cells by desmosomes. Near the apex, the cells are joined by tight junctions consisting of several parallel junctional strands similar to those in the thick ascending limb throughout. The cells contain the usual cytoplasmic organelles, comprising some small mitochondria, Golgi apparatus, and smooth endoplasmic reticulum; free ribosomes are abundant but rough endoplasmic reticulum is rare. The lateral intercellular spaces are a prominent feature of the macula densa. Electron microscopic studies, and studies on isolated macula densa segments in vitro, have shown that the width of the lateral intercellular spaces varies under different functional conditions (Kaissling et al., 1982; Kirk et al., 1985). In agreement with the suggestion that water flow through the macula densa epithelium is secondary to active sodium reabsorption, compounds such as furosemide, that block sodium transport, as well as high osmolalities of impermeable solutes such as mannitol, are associated with narrowing of the intercellular spaces (Kaissling et al., 1982; Alcorn et al., 1986). The spaces are apparently dilated under most physiological conditions, usually regarded as normal control conditions. The most conspicuous difference in the protein inventory between macula densa cells and any other epithelial cell of the nephron is the high content of nitric oxide synthase I  (Mundel et al., 1992; Persson and Bachmann, 2000) and of cyclooxygenase-2 (Schnermann, 2001) in macula densa cells. The granular cells are assembled in clusters within the terminal portion of the afferent arteriole, replacing smooth muscle cells (Fig. 43.3). Their name refers to the cytoplasmic granules, which are dark, membrane-bound, and irregular in size and shape. Renin, the major secretion product, is stored in these granules. Small granules with crystalline substructure represent protogranules containing both renin prosegment and mature renin. Renin release occurs by exocytosis into the surrounding interstitium (Taugner and Hackenthal, 1989). Granular cells are modified smooth muscle cells. Under conditions requiring enhanced renin synthesis (e.g. volume depletion or stenosis of the renal artery) additional smooth muscle cells located upstream in the wall of the afferent arteriole transform into granular cells. Granular cells are connected to the extraglomerular mesangial cells, to adjacent smooth muscle cells, and to endothelial cells by gap junctions, and are densely innervated by sympathetic nerve terminals (Taugner and Hackenthal, 1989). The structural organization of the JGA suggests a regulatory function. Goormaghtigh (1937) was the first to propose that some component of the distal urine is sensed by the macula densa and this information is used to adjust the tonus of the glomerular

445

446

Section 3  

the patient with glomerular disease

arterioles, thereby producing a change in glomerular blood flow and filtration rate. Moreover, since the JGA is the major site of renin secretion, the function of the JGA is of great systemic relevance. These two functions (regulation of the vascular tone of glomerular arterioles and regulation of renin release from granular cells) seem to be strictly separated from each other. For both mechanisms, it is well established that changes in the chloride concentration of the tubular fluid at the macula densa cause graded releases of mediators that reach their target by diffusion, thus acting in a paracrine fashion (Kurtz, 2011). Note that the extraglomerular mesangium that mediates the contact between the macula densa and the effector cells is not vascularized, so that the build-up of any paracrine agent would not be perturbed by blood flow. With respect to renin release, the most likely paracrine mediators of this process are prostaglandin E2 and nitric oxide (Wilcox et  al., 1992; Peti-Peterdi et  al., 2003; Schweda and Kurtz, 2004). With respect to the vasoconstrictor response purinergic mediators, either ATP or adenosine, as first suggested by Oswald and colleagues in 1980 (Osswald et al., 1980), appear to play the major

PO

E MM GBM M

MM GBM

M C MF

Fig. 43.12  Schematic showing the filtration barrier as well as the centrolobular position of a mesangial cell (M) and its relationships to the glomerular capillaries and to the glomerular basement membrane (GBM). The glomerular capillary consists of a fenestrated endothelium (E). The peripheral portion of the capillary is surrounded by the GBM which, at the mesangial angles (arrow), deviates from the pericapillary course and covers the mesangium. The interdigitating system of the podocyte (PO) foot processes forms the distal layer of the filtration barrier. Connections between mesangial cell processes and the GBM are prominent at mesangial angles, and are also numerous along the perimesangial GBM. Many of these connections are mediated by microfibrils which are a major constituent of the mesangial matrix (MM). Thus, a mechanical firm linkage of the perimesangial GBM to the contractile apparatus of the mesangial cells is established. Modified after Kriz et al. (1992).

Fig. 43.11  Juxtacapillary portion of mesangium showing tongue-like mesangial processes fixed to the glomerular basement membrane (GBM) at the mesangial angles (arrows). Note the rich equipment of mesangial processes with bundles of microfilaments (MF) which are attached to the cell membrane. The mesangial matrix (MM) contains abundant microfibrils (arrowhead). C = capillary lumen. 61,200×.

role (Schnermann and Levine, 2003; Castrop et al., 2004; Thomson et al., 2000). For an up-to-date discussion of the function of the JGA, see the reviews by Schnermann and Levine (2003), Persson et al. (2004), Komlosi et al. (2004), Kurtz (2011), and Schnermann and Briggs (2013). Mesangial cell-to-GBM contacts are found along the entire perimesangial GBM. The intracellular actin filament bundles are arranged in such a way that segments of the GBM located on opposing sides of the mesangium are interconnected. These connections are most prominent at mesangial angles consisting of tongue-like mesangial processes which establish a bridge between both mesangial angles of the GBM (Figs 43.10, 43.11, and 43.12), (Sakai and Kriz, 1987). Evidence for the importance of these connections for the structural integrity of the tuft was obtained by selective destruction of the mesangium by experimental application of antibody against the cell surface antigen Thy 1 (Paul et al., 1984; Kriz et al., 2003). In this

Chapter 43 

situation, the mesangial region as well as the mesangial–endothelial interface is greatly enlarged, leading to a partial ‘unfolding’ of glomerular capillaries associated with capillary ballooning (Lemley et al., 1992), and profound changes in the glomerular haemodynamics (Blantz et al., 1991). The large pressure gradients across the GBM represent the crucial challenge to the glomerular tuft. The distending forces acting on the GBM (across the peripheral and across the perimesangial interface) have to be counterbalanced by inwardly directed forces. This comprises two aspects. First, the tuft as a whole, that is, the folding pattern of the GBM providing space for the capillaries, and second, the width of the capillaries have to be maintained (or adapted to varying situations). The folding pattern of the GBM is supported by the overall centripetal fixation of the GBM to the mesangium (by centripetal contractile forces) (Kriz et al., 1995a). In addition, podocytes are involved; podocyte processes that fill the niches of GBM-infoldings interconnect opposing portions of the GBM from outside, thereby stabilizing the folding pattern of the GBM (Kriz et al., 2003). The maintenance of the capillary width is supported by the GBM in conjunction with the mesangium. The basic supportive structure of a glomerular capillary wall is represented by (A) the GBM which forms an incomplete cylinder that is open towards the mesangium like a tyre to the rim and (B) by a mesangial cell, which bridges this gap by a contractile cell process fixed to the GBM at both sides. Thus, the GBM (which is an elastic structure (Welling et al., 1995)) together with the mesangial cell bridge form a complete cylinder which is able to develop wall tension and thus to resist distension. Traditionally, the podocyte foot processes have been considered as a kind of pericyte processes contributing to the generation of wall tension; however, this view has recently been challenged (Kriz et al., 2013b). Thus, the contractile apparatus of the mesangial cells being effective at the GBM appears to be static in nature, operating by isometric or minute isotonic contractions. Whether mesangial cell contractility also plays a role in the regulation of glomerular haemodynamics is still a matter of debate. Because the mesangial–endothelial interface comprises only a small part of the capillary circumference, a contraction of mesangial cells at this site would lead to minor changes in the capillary diameter.

References Abrahamson, D. (1987). Structure and development of the glomerular capillary wall and basement membrane. Am J Physiol, 253, F783–94. Abrahamson, D. (2012). Role of the podocyte (and glomerular epithelium) in building the GBM. Semin Nephrol, 32, 342–9. Abrahamson, D., Hudson, B. G., Stroganova, L., et al. (2009). Cellular origins of type IV collagen networks in developing glomeruli. J Am Soc Nephrol, 20, 1471–9. Alcorn, D., Anderson, W., and Ryan, G. (1986). Morphological changes in the renal macula densa during natriuresis and diuresis. Renal Physiol, 9, 335–47. Appel, D., Kershaw, D. B., Smeets, B., et al. (2008). Recruitment of podocytes from glomerular parietal epithelial cells. J Am Soc Nephrol, 20, 333–43. Bachmann, S., Kriz, W., Kuhn, C., et al. (1983). Differentiation of cell types of the mammalian kidney by immunofluorescence microscopy using antibodies to intermediate filament proteins and desmoplakins. Histochemistry, 77, 365–94. Bakker, W., Borghuis, T., Harmsen, M. C., et al. (2005). Protease activity of plasma hemopexin. Kidney Int, 68, 603–10.

structure of the glomerulus

Barajas, L. (1997). Cell-specific protein and gene expression in the juxtaglomerular apparatus. Clin Exp Pharmacol Physiol, 24, 520–6. Barajas, L., Salido, E. C., Smolens, P., et al. (1989). Pathology of the juxtaglomerular apparatus including Bartter’s syndrome. In C. Tisher and B. Brenner (eds.) Renal Pathology, pp. 877–912. Philadelphia, PA: Lippincott. Bek, M., Nüsing, R., Kowark, P., et al. (1999). Characterization of prostanoid receptors in podocytes. J Am Soc Nephrol, 10, 2093. Blantz, R., Gabbai, F., and Wilson, C. (1991). Glomerular hemodynamic (GH) response to changes in volume status after lysis of mesangial cells (MC). J Am Soc Nephrol, 1, 661. Border, W., Okuda, S., and Nakamura, T. (1989). Extracellular matrix and glomerular disease. Semin Nephrol, 9, 307–17. Brees, D., Ogle, R., and Williams, J., Jr. (1995). Laminin and fibronectin content of mouse glomerular and tubular basement membrane. Renal Physiol Biochem, 18, 1–11. Breuss, J., Gillett, N., Lu, L., et al. (1993). Restricted distribution of integrin b6 mRNA in primate epithelial tissue. J Histochem Cytochem, 41, 1521–7. Bulger, R., Eknoyan, G., Purcell, D., et al. (1983). Endothelial characteristics of glomerular capillaries in normal, mercuric chloride-induced, and gentamicin-induced acute renal failure in the rat. Clin Invest, 72, 128–41. Bulger, R. and Hebert, S. (1988). Structural-functional relationships in the kidney. In R. Schrier and A. Gottschalk (eds.) Diseases of the Kidney, pp. 3–63. Boston, MA: Little, Brown and Company. Burridge, K., Fath, K., Kelly, T., et al. (1988). Focal adhesion: transmembrane junctions between the extracellular matrix and the cytoskeleton. Ann Rev Cell Biol, 4, 487–525. Castrop, H., Huang, Y., Hashimoto, S., et al. (2004). Impairment of tubuloglomerular feedback regulation of GFR in ecto-5’-nucleotidase/ CD73-deficient mice. Clin Invest, 114, 634–42. Cosio, F., Sedmak, D., and Nahman, N. (1990). Cellular receptors for matrix proteins in normal human kidney and human mesangial cells. Kidney Int, 38, 886–95. Couchman, J., Beavan, L., and McCarthy, K. (1994). Glomerular matrix: synthesis, turnover and role in mesangial expansion. Kidney Int, 45, 328–35. Daniels, B., Deen, W. M., Mayer, G., et al. (1993). Glomerular permeability barrier in the rat. Clin Invest, 92, 929–36. Deen, W., Bohrer, M., and Brenner, B. (1979). Macromolecule transport across glomerular capillaries: application of pore theory. Kidney Int, 16, 353–65. Deen, W., Lazzara, M., and Myers, B. (2001). Structural determinants of glomerular permeability. Am J Physiol Renal Physiol, 281, F579–96. Desjardins, M. and Duling, B. (1990). Heparinase treatment suggests a role for the endothelial cell glycocalyx in regulation of capillary hematocrit. Am J Physiol, 258, H647–54. Drab, M., Verkade, P., Elger, M., et al. (2001). Loss of caveolae, vascular dysfunction, and pulmonary defects in caveolin-1 gene-disrupted mice. Science, 293, 2449–52. Drenckhahn, D. and Franke, R. (1988). Ultrastructural organization of contractile and cytoskeletal proteins in glomerular podocytes of chicken, rat, and man. Lab Invest, 59, 673–82. Drumond, M. and Deen, W. (1994a). Structural determinants of glomerular hydraulic permeability. Am J Physiol, 266, F1–12. Drumond, M., Kristal, B., Myers, B., et al. (1994b). Structural basis for reduced glomerular filtration capacity in nephrotic humans. Clin Invest, 94, 1187–95. Durbeej, M., Henry, M., and Campbell, K. (1998). Dystroglycan in development and disease. Curr Opin Cell Biol, 10, 594–601. Dworkin, L., Ichikawa, I., and Brenner, B. (1983). Hormonal modulation of glomerular function. Am J Physiol, 244, F95–104. Elger, M., Sakai, T., and Kriz, W. (1998). The vascular pole of the renal glomerulus of rat. Adv Anat Embryol Cell Biol, 139, 1–98. Elger, M., Sakai, T., Winkler, D., et al. (1991). Structure of the outflow segment of the efferent arteriole in rat superficial glomeruli. Contrib Nephrol, 95, 22–33.

447

448

Section 3  

the patient with glomerular disease

Endlich, K., Kriz, W., and Witzgall, R. (2001a). Update in podocyte biology. Curr Opin Nephrol Hypertens, 10, 331–40. Endlich, N., Nobiling, R., Kriz, W., et al. (2001b). Expression and signaling of parathyroid hormone-related protein in cultured podocytes. Exp Nephrol, 9, 436–43. Farquhar, M. (1991). The glomerular basement membrane: a selective macromolecular filter. In E. Hay (ed.) Cell Biology of Extracellular Matrix, pp. 365–418. New York: Plenum Press. Farquhar, M. and Palade, G. (1962). Functional evidence for the existence of a third cell type in the renal glomerulus. Phagocytosis of filtration residues by a distinctive ‘third’ cell. J Cell Biol, 13, 55–87. Fischer, K., Saueressig, U., Jacobshagen, C., et al. (2001). Extracellular nucleotides regulate cellular functions of podocytes in culture. Am J Physiol Renal Physiol, 281, F1075–81. Ford, M., Cauchi, J., Greferath, U., et al. (1997). Expression of fibroblast growth factors and their receptors in rat glomeruli. Kidney Int, 51, 1729–38. Fridén, V., Oveland, E., Tenstad, O., et al. (2011). The glomerular endothelial cell coat is essential for glomerular filtration. Kidney Int, 79, 1322–30. Fries, J., Sandstrom, D., Meyer, T., et al. (1989). Glomerular hypertrophy and epithelial cell injury modulate progressive glomerulosclerosis in the rat. Lab Invest, 60, 205–18. Gelberg, H., Healy, L., Whiteley, H., et al. (1996). In vivo enzymatic removal of a26-linked sialic acid from the glomerular filtration barrier results in podocyte charge alteration and glomerular injury. Lab Invest, 74, 907–20. Gibson, M., Kumaratilake, J., and Cleary, E. (1989). The protein components of the 12-nanometer microfibrils of elastic and nonelastic tissues. J Biol Chem, 264, 4590–98. Goormaghtigh, N. (1937). L’appareil neuro-myo-artériel juxta-glomérulaire du rein: ses réactions en pathologie et ses rapports avec le tube urinifère. C R Seances Soc Biol Fil, 124, 293–6. Greka, A. and Mundel, P. (2012). Cell biology and pathology of podocytes. Ann Rev Physiol, 74, 299–323. Grouls, S., Iglesias, D. M., Wentzensen, N., et al. (2012). Lineage specification of parietal epithelial cells requires ß-catenin/Wnt signaling. J Am Soc Nephrol, 23, 63–72. Guasch, A. and Myers, B. (1994). Determinants of glomerular hypofiltration in nephrotic patients with minimal change nephropathy. J Am Soc Nephrol, 4, 1571–81. Hansen, K. and Abrass, C. (1999). Role of laminin isoforms in glomerular structure. Pathology, 67, 84–91. Hara, M., Yamamoto, T., Yanagihara, T., et al. (1995). Urinary excretion of podocalyxin indicates glomerular epithelial cell injuries in glomerulonephritis. Nephron, 69, 397–403. Hara, M., Yanagihara, T., Takada, T., et al. (1998). Urinary excretion of podocytes reflects disease activity in children with glomerulonephritis. Am J Nephrol, 18, 35–41. Haraldsson, B. and Jeansson, M. (2009). Glomerular filtration barrier. Curr Opin Nephrol Hypertens, 18, 331–5. Haraldsson, B. and Nyström, J. (2012). The glomerular endothelium: new insights on function and structure. Curr Opin Nephrol Hypertens, 21, 258–63. Haraldsson, B., Nyström, J., and Deen, W. (2008). Properties of the glomerular barrier and mechanisms of proteinuria. Physiol Rev, 88, 451–87. Hausmann, R., Kuppe, C., Egger, H., et al. (2010). Electrical forces determine glomerular permeability. J Am Soc Nephrol, 21, 2053–8. Herman, W., Emancipator, S., and Simonson, M. (1998). Vascular and glomerular expression of endothelin-1 in normal human kidney. Am J Physiol Renal Physiol, 275, F8–17. Hjalmarsson, C., Johansson, B., and Haraldsson, B. (2004). Electron microscopic evaluation of the endothelial surface layer of glomerular capillaries. Microvasc Res, 67, 9–17. Hora, K., Ohno, S., Oguchi, H., et al. (1990). Three-dimensional study of glomerular slit diaphragm by quick-freezing and deep-etching replica method. Eur J Cell Biol, 53, 402–6.

Horvat, R., Hovorka, A., Dekan, G., et al. (1986). Endothelial cell membranes contain podocalyxin- the major sialoprotein of visceral glomerular epithelial cells. J Cell Biol, 102, 484–91. Huang, T. and Langlois, J. (1985). Podoendin. A new cell surface protein of the podocyte and endothelium. J Exp Med, 162, 245–67. Huber, T., Gloy, J., Henger, A., et al. (1998). Catecholamines modulate podocyte function. J Am Soc Nephrol, 9, 335–45. Hugo, C., Shankland, S. J., Bowen-Pope, D. F., et al. (1997). Extraglomerular origin of the mesangial cell after injury- a new role of the juxtaglomerular apparatus. Clin Invest, 100, 786–94. Ichimura, K., Stan, R., Kurihara, H., et al. (2008). Glomerular endothelial cells form diaphragms during development and pathological conditions. J Am Soc Nephrol, 19(8), 1463–71. Jeansson, M. and Haraldsson, B. (2006). Morphological and functional evidence for an important role of the endothelial cell glycocalyx in the glomerular barrier. Am J Physiol Renal Physiol, 290, F111–16. Jorgensen, F. (1966). The Ultrastructure of the Normal Human Glomerulus. Copenhagen: Ejnar Munksgaard. Kaissling, B. and Kriz, W. (1982). Variability of intercellular spaces between macula densa cells: a transmission electron microscopic study in rabbits and rats. Kidney Int, 22, 9–17. Karkavelas, G. and Kefalides, N. (1988). Comparative ultrastructural localization of collagen types III, IV, VI and laminin in rat uterus and kidney. J Ultrastruct Mol Struct Res, 100, 137–55. Kazatchkine, M., Fearon, D. T., Appay, M. D., et al. (1982). Immunohistochemical study of the human glomerular C3b receptor in normal kidney and in seventy-five cases of renal diseases. Clin Invest, 69, 900–12. Kerjaschki, D. and Farquhar, M. (1983). Immunocytochemical localization of the Heymann antigen (gp 330) in glomerular epithelial cells of normal Lewis rats. J Exp Med, 157, 667–86. Kerjaschki, D., Sharkey, D., and Farquhar, M. (1984). Identification and characterization of podocalyxin-the major sialoprotein of the renal glomerular epithelial cell. J Cell Biol, 98, 1591. Kiberd, B. (1992). The functional and structural changes of the glomerulus throughout the course of murine lupus nephritis. J Am Soc Nephrol, 3, 930–9. Kirk, K., Bell, P., Barfuss, D., et al. (1985). Direct visualization of the isolated and perfused macula densa. Am J Physiol, 248, F890–4. Kobayashi, N., Reiser, J., Kriz, W., et al. (1998). Nonuniform microtubular polarity established by CHO1/MKLP1 motor protein is necessary for process formation of podocytes. J Cell Biol, 143, 1961–70. Komlosi, P., Fintha, A., and Bell, P. (2004). Current mechanisms of macula densa cell signaling. Acta Physiol Scand, 181, 463–9. Kreidberg, J. and Symons, J. (2000). Integrins in kidney development, function, and disease. Am J Physiol Renal Physiol, 279(2), F233–42. Kreisberg, J., Venkatachalam, K., and Troyer, D. (1985). Contractile properties of cultured glomerular mesangial cells. Am J Physiol, 249, F457–63. Kriz, W. (2002). Podocyte is the major culprit accounting for the progression of chronic renal disease. Microsc Res Tech, 57, 189–95. Kriz, W. (2007). Ontogenetic development of the filtration barrier. Nephron Exp Nephrol, 106, e44–50. Kriz, W., Elger, M., Lemley, K., et al. (1990). Mesangial cell—glomerular basement membrane connections counteract glomerular capillary and mesangium expansion. Am J Nephrol, 10, 4–13. Kriz, W., Elger, M., Mundel, P., et al. (1995a). Structure-stabilizing forces in the glomerular tuft. J Am Soc Nephrol, 5, 1731–9. Kriz, W., Hähnel, B., Hosser, H., et al. (2003). Pathways to recovery and loss of nephrons in anti-Thy-1 nephritis. J Am Soc Nephrol, 14, 1904–26. Kriz, W., Hähnel, B., Rosener, S., et al. (1995b). Long-term treatment of rats with FGF-2 results in focal segmental glomerulosclerosis. Kidney Int, 48, 1435–50. Kriz, W. and Kaissling, B. (1992). Structural organization of the mammalian kidney. In D. Seldin and G. Giebisch (eds.) The Kidney: Physiology and Pathophysiology, pp. 707–77. New York: Raven Press. Kriz, W. and Kaissling, B. (2013a). Structural organization of the mammalian kidney. In R. Alpern, M. Caplan, and O. Moe (eds.) Seldin and

Chapter 43 

Giebisch’s The Kidney: Physiology and Pathophysiology, pp. 595–691. Amsterdam: Academic Press Elsevier. Kriz, W., Shirato, I., Nagata, M., et al. (2013b). The podocyte’s response to stress: the enigma of foot process effacement. Am J Physiol Renal Physiol, 304, F333–47. Kurtz, A. (2011). Renin release: sites, mechanisms, and control. Ann Rev Physiol, 73, 377–99. Lafayette, R., Druzin, M., Sibley, R., et al. (1998). Nature of glomerular dysfunction in pre-eclampsia. Kidney Int, 54, 1240–9. Larsson, L. and Maunsbach, A. (1980). The ultrastructural development of the glomerular filtration barrier in the rat kidney: a morphometric analysis. J Ultrastruct Res, 72, 392–406. Latta, H., Maunsbach, A., and Madden, S. (1960). The centrolobular region of the renal glomerulus studied by electron microscopy. J Ultrastruct Res, 4, 455–72. Lebrun, F., Morel, F., Vassent, G., et al. (1992). Cholinergic effects on intracellular free calcium concentration in renal corpuscle: role of parietal sheet. Am J Physiol, 262, F248–55. Lemley, K., Elger, M., Koeppen-Hagemann, I., et al. (1992). The glomerular mesangium: capillary support function and its failure under experimental conditions. Clin Invest, 70, 843–56. Lethias, C., Aubert-Foucher, E., Dublet, B., et al. (1994). Structure, molecular assembly and tissue distribution of facit collagen molecules. Contrib Nephrol, 107, 57–63. Madri, J., Roll, F., Furthmayr, H., et al. (1980). Ultrastructural localization of fibronectin and laminin in the basement membranes of the murine kidney. J Cell Biol, 86, G82–7. Mbassa, G., Elger, M., and Kriz, W. (1988). The ultrastructural organization of the basement membrane of Bowman’s capsule in the rat renal corpuscle. Cell Tissue Res, 253, 151–63. Mendrick, D. and Rennke, H. (1988). Induction of proteinuria in the rat by a monoclonal antibody against SGP-115/107. Kidney Int, 33, 818–30. Meuwese, M., Broekhuizen, L. N., Kuikhoven, M., et al. (2010). Endothelial surface layer degradation by chronic hyaluronidase infusion induces proteinuria in apolipoprotein E-deficient mice. PLOS One, 5, 1–7. Miner, J. (1999). Renal basement membrane components. Kidney Int, 56, 2016–24. Miner, J. (2011). Glomerular basement membrane composition and the filtration barrier. Pediatr Nephrol, 26, 1413–7. Miner, J., Patton, B. L., Lentz, S. I., et al. (1997). The laminin a chanins: expression, development transitions, and chromosomal locations of a1-5, identification of heterotrimeric laminis 8-11, and cloning of a novel a3 isoform. J Cell Biol, 137, 685–701. Miner, J. and Sanes, J. (1994). Collagen IV a3, a4, and a5 chains in rodent basal laminae: Sequence, distribution, association with laminins, and developmental switches. J Cell Biol, 127, 879–91. Mohan, P. and Spiro, R. (1986). Macromolecular organization of basement membranes. J Biol Chem, 261, 4328–36. Morita, M., White, R., and Raafat, F. (1988). Glomerular basement membrane thickness in children. Pediatr Nephrol, 2, 190–5. Müller, E., Burger-Kentischer, A., Neuhofer, W., et al. (1999). Possible involvement of heat shock protein 25 in the angiotensin II-induced glomerular mesangial cell contraction via p.38 MAP kinase. J Cell Physiol, 181, 462–9. Mundel, P., Bachmann, S., Bader, M., et al. (1992). Expression of nitric oxide synthase in kidney macula densa cells. Kidney Int, 42, 1017–9. Mundel, P., Elger, M., Sakai, T., et al. (1988). Microfibrils are a major component of the mesangial matrix in the glomerulus of the rat kidney. Cell Tissue Res, 254, 183–7. Mundel, P. and Kriz, W. (1995). Structure and function of podocytes: an update. Anat Embryol, 192, 385–97. Nagata, M. and Kriz, W. (1992). Glomerular damage after uninephrectomy in young rats. II. Mechanical stress on podocytes as a pathway to sclerosis. Kidney Int, 42, 148–60. Nagata, M., Yamaguchi, Y., and Ito, K. (1993). Loss of mitotic activity and the expression of vimentin in glomerular epithelial cells of developing human kidneys. Anat Embryol, 187, 275–9. Nakamura, T., Ushiyama, C., and Suzuki, S. (2000). Effect of angiotensin-converting enzyme inhibitor, angiotensin II receptor

structure of the glomerulus

antagonist and calcium antagonist on urinary podocytes in patients with IgA nephropathy. Am J Nephrol, 20, 373–9. Nitschke, R., Henger, A., Ricken, S., et al. (2000). Angiotensin II increases the intracellular calcium activity in podocytes of the intact glomerulus. Kidney Int, 57, 41–9. Nitschke, R., Henger, A., Ricken, S., et al. (2001). Acetylcholine increases the intracellular calcium activity in prodocytes in intact rat glomeruli via muscarinic M(5) receptors. J Am Soc Nephrol, 12, 678–87. Obeidat, M., Obeidat, M., and Ballermann, B. (2012). Glomerular endothelium: a porous sieve and formidable barrier. Exp Cell Res, 318, 964–72. Osswald, H., Nabakowski, G., and Hermes, H. (1980). Adenosine as a possible mediator of metabolic control of glomerular filtration rate. Int J Biochem, 12, 263–7. Ott, M., Olson, J., and Ballermann, B. (1993). Phenotypic differences between glomerular capillary (GE) and aortic (AE) endothelial cells in vitro. (Abstract). J Am Soc Nephrol, 4, 564. Paul, L., Rennke, H., Milford, E., et al. (1984). Thy-1.1 in glomeruli of rat kidneys. Kidney Int, 25, 771–7. Pease, D. (1968). Myoid features of renal corpuscles and tubules. J Ultrastruct Res, 23, 304–20. Persson, A. and Bachmann, S. (2000). Constitutive nitric oxide synthesis in the kidney—functions at the juxtaglomerular apparatus. Acta Physiol Scand, 169, 317–24. Persson, A., Ollerstam, A., Liu, R., et al. (2004). Mechanism for macula densa cell release of renin. Acta Physiol Scand, 181, 471–4. Petermann, A., Pippin, J., Krofft, R., et al. (2004). Viable podocytes detach in experimental diabetic nephropathy: potential mechanism underlying glomerulosclerosis. Nephron Exp Nephrol, 98, 114–23. Peti-Peterdi, J., Komlosi, P., Fuson, A. L., et al. (2003). Luminal NaCl delivery regulates basolateral PGE2 release from macla densa cells. Clin Invest, 112, 76–82. Pricam, C., Humbert, F., Perrelet, A., and Orci, L. (1974). Gap junctions in mesangial and lacis cells. J Cell Biol, 63, 349–54. Pries, A., Secomb, T., and Gaehtgens, P. (2000). The endothelial surface layer. Pflugers Archiv, 440, 653–66. Raats, C., van den Born, J., and Berden, J. (2000). Glomerular heparan sulfate alterations: mechanisms and relevance for proteinuria. Kidney Int, 57(2), 385–400. Rasch, R. (1979). Prevention of diabetic glomerulopathy in streptozotocin diabetic rats by insulin treatment. Glomerular basement membrane thickness. Diabetologia, 16, 319–24. Rebibou, J., He, C. J., Delarue, F., et al. (1992). Functional endothelin-1 receptors on human glomerular podocytes and mesangial cells. Nephrol Dial Transplant, 7, 288–92. Regele, H., Fillipovic, E., Langer, B., et al. (2000). Glomerular expression of dystroglycans is reduced in minimal change nephrosis but not in focal segmental glomerulosclerosis. J Am Soc Nephrol, 11(3), 403–12. Rodewald, R. and Karnovsky, M. (1974). Porous substructure of the glomerular slit diophragm in the rat and mouse. J Cell Biol, 60, 423–33. Ronconi, E., Sagrinati, C., Angelotti, M. L., et al. (2009). Regeneration of glomerular podocytes by human renal progenitors. J Am Soc Nephrol, 20, 322–32. Rostgaard, J. and Qvortrup, K. (2002). Sieve plugs in fenestrae of glomerular capillaries—site of the filtration barrier? Cells Tissues Organs, 170, 132–8. Ryan, G., Hein, S., and Karnovsky, M. (1976). Glomerular permeability to proteins. Effects of hemodynamic factors on the distribution of endogenous immunoglobulin G and exogenous catalase in the rat glomerulus. Lab Invest, 34, 415. Sagrinati, C., Netti, G. S., Mazzinghi, B., et al. (2006). Isolation and characterization of multipotent progenitor cells from the Bowman’s capsule of adult human kidneys. J Am Soc Nephrol, 17, 2443–56. Saito, A., Pietromonaco, S., Loo, A., et al. (1994). Cloning and sequencing of gp330/megalin, the major Heymann nephritis antigen. J Am Soc Nephrol, 5, 767.

449

450

Section 3  

the patient with glomerular disease

Sakai, T. and Kriz, W. (1987). The structural relationship between mesangial cells and basement membrane of the renal glomerulus. Anat Embryol, 176, 373–86. Salmon, A. H., Neal, C. R., and Harper, S. J. (2009). New aspects of glomerular filtration barrier structure and function: five layers (at least) not three. Curr Opin Nephrol Hypertens, 18, 197–205. Salmon, A. and Satchell, S. (2012). Endothelial glycocalyx dysfunction in disease: albuminuria and increased microvascular permeability. J Pathol, 226, 562–74. Satchell, S. (2012). The glomerular endothelium emerges as a key player in diabetic nephropathy. Kidney Int, 82, 949–51. Satchell, S. and Braet, F. (2009). Glomerular endothelial cell fenestration: an integral component of the glomerular filtration barrier. Am J Physiol Renal Physiol, 296, F947–56. Savage, C. (1998). Injury mechanisms in vasculitis. Kidney Blood Press Res, 21, 269–70. Sawada, H., Stukenbrok, H., Kerjaschki, D., et al. (1986). Epithelial polyanion (podocalyxin) is found on the sides but not the soles of the foot processes of the glomerular epithelium. Am J Pathol, 125, 309–18. Schlöndorff, D. and Banas, B. (2009). The mesangial cell revisited: No cell is an island. J Am Soc Nephrol, 20, 1179–87. Schnermann, J. (2001). Cyclooxygenase-2 and macula densa control of renin secretion. Nephrol Dial Transplant, 16, 1735–8. Schnermann, J. and Briggs, J. (2013). Tubular control of renin synthesis and secretion. Pflugers Arch, 465, 39–51. Schnermann, J. and Levine, D. (2003). Paracrine factors in tubuloglomerular feedback: adenosine, ATP, and nitric oxide. Ann Rev Physiol, 65, 501–29. Schordan, S., Schordan, E., Endlich, K., et al. (2010). Alphav-integrins mediate the mechanoprotective action of osteopontin in podocytes. Am J Physiol Renal Physiol, 300, F119–32. Schwartz, E., Goldfischer, S., Coltoff-Schiller, B., et al. (1985). Extracellular matrix microfibrils are composed of core proteins coated with fibronectin. J Histochem Cytochem, 33, 268–74. Schweda, F. and Kurtz, A. (2004). Cellular mechanism of renin release. Acta Physiol Scand, 181, 383–90. Seiler, M., Rennke, H., Venkatachalam, M., et al. (1977). Pathogenesis of polycation-induced alteration (fusion) of glomerular epithelium. Lab Invest, 36, 48–61. Sharma, R., Sharma, M., Vamos, S., et al. (2001). Both subtype 1 and 2 receptors of angiotensin II participate in regulation of intracellular calcium in glomerular epithelial cells. J Lab Clin Med, 138, 40–49. Sörensson, J., Fierlbeck, W., Heider, T., et al. (2002). Glomerular endothelial fenestrae in vivo are not formed from caveolae. J Am Soc Nephrol, 13, 2639–47. Spanidis, A. and Wunsch, H. (1979). Rekonstruktion einer Goormaghtigh’schen und einer Epitheloiden Zelle der Kaninchenniere. Dissertation. University of Heidelberg, Heidelberg. Spath, M., Pavenstädt, H., Müller, C., et al. (1995). Regulation of phosphoinositide hydrolysis and cytosolic free calcium induced by endothelin in human glomerular epithelial cells. Nephrol Dial Transplant, 10, 1304. St John, P. and Abrahamson, D. (2001). Glomerular endothelial cells and podocytes jointly synthesize laminin-1 and -11 chains. Kidney Int, 60, 1037–46. Stan, R. (2005). Structure of caveolae. Biochim Biopyhs Acta, 1746, 334–48. Stan, R., Tkachenko, E., and Niesman, I. (2004). PV1 is a key structural component for the formation of the stomatal and fenestral diaphragms. Mol Biol Cell, 15, 3615–30. Steffes, M., Barbosa, J., Basgen, J. M., et al. (1983). Quantitative glomerular morphology of the normal human kidney. Lab Invest, 49, 82–6. Sterzel, R., Hartner, A., Schlötzer-Schrehardt, U., et al. (2000). Elastic fiber proteins in the glomerular mesangium in vivo and in cell culture. Kidney Int, 58, 1588–602.

Stockand, J. and Sansom, S. (1998). Glomerular mesangial cells: Electrophysiology and regulation of contraction. Physiol Rev, 78, 723–44. Taugner, R. and Hackenthal, E. (1989). The Juxtaglomerular Apparatus. Heidelberg: Springer-Verlag. Taugner, R., Schiller, A., Kaissling, B., et al. (1978). Gap junctional coupling between the JGA and the glomerular tuft. Cell Tissue Res, 186, 279–85. Thomson, S., Bao, D., Deng, A., et al. (2000). Adenosine formed by 5’-nucleotidase mediates tubuloglomerular feedback. Clin Invest, 106, 289–98. Timpl, R. and Dziadek, M. (1986). Structure, development, and molecular pathology of basement membranes. Int Rev Exp Pathol, 29, 1–112. Tisher, C. and Brenner, B. (1989). Structure and function of the glomerulus. In C. Tisher and B. Brenner (eds.) Renal Pathology, pp. 92–110. Philadelphia, PA: Lippincott. Toyoda, M., Najafian, B., Kim, Y., et al. (2007). Podocyte detachment and reduced glomerular capillary endothelial fenestration in human type 1 diabetic nephropathy. Diabetes, 56, 2155–60. Van den Berg, J., Aten, J., Chand, M. A., et al. (2000). Interleukin-4 and interleukin-13 act on glomerular visceral epithelial cells. J Am Soc Nephrol, 11, 413–22. Vasmant, D., Maurice, M., and Feldmann, G. (1984). Cytoskeleton ultrastructure of podocytes and glomerular endothelial cells in man and in the rat. Anat Rec, 210, 17–24. Vogelmann, S., Nelson, W., Myers, B., et al. (2003). Urinary excretion of viable podocytes in health and renal disease. Am J Physiol Renal Physiol, 285, F40–8. Wang, J., Yang, A. H., Chen, S. M., et al. (1993). Amelioration of antioxidant enzyme suppression and proteinuria in cyclosporin-treated puromycin nephrosis. Nephron, 65, 418–25. Weil, E., Lemley, K. V., Yee, B., et al. (2011). Podocyte detachment in type 2 diabetic nephropathy. Am J Nephrol, 33, 21–4. Welling, L., Zupka, M., and Welling, D. (1995). Mechanical properties of basement membrane. News Physiol Sci, 10 (1), 30–5. Wiggins, R., Fantone, J., and Phan, S. (1989). Mechanisms of vascular injury. In C. Tisher and B. Brenner (eds.) Renal Pathology, pp. 965–93. Philadelphia, PA: Lippincott. Wilcox, C., Welch, W. J., Murad, F., et al. (1992). Nitric oxide synthase in macula densa regulates glomerular capillary pressure. Proc Natl Acad Sci U S A, 89, 11993–7. Winkler, D., Elger, M., Sakai, T., et al. (1991). Branching and confluence pattern of glomerular arterioles in the rat. Kidney Int, 39, S-2–8. Yamamoto, T., Watanabe, T., Ikegaya, N., et al. (1998). Expression of types I, II, and III TGF-beta receptors in human glomerulonephritis. J Am Soc Nephrol, 9, 2253–61. Yang, G. and Morrison, A. (1980). Three large dissectable rat glomerular models reconstructed from wide-field electron micrographs. Anat Rec, 196, 431–40. Yu, D., Petermann, A., Kunter, U., et al. (2005). Urinary podocyte loss is a more specific marker of ongoing glomerular damage than proteinuria. J Am Soc Nephrol, 16, 1733–41. Yurchenco, P. and Cheng, Y. (1994). Laminin self-assembly: a three-arm interaction hypothesis for the formation of a network in basement membranes. Contrib Nephrol, 107, 47–56. Zenker, M., Aigner, T., Wendler, O., et al. (2004). Human laminin b2 deficiency causes congenital nephrosis with mesangial sclerosis and distinct eye abnormalities. Hum Mol Genet, 13, 2625–32. Zhao, J., Ardaillou, N., Lu, C. Y., et al. (1994). Characterization of C-type natriuretic peptide receptors in human mesangial cells. Kidney Int, 46(3), 717–25. Zimmermann, K. (1929). Ueber den Bau des Glomerulus der menschlichen Niere. Z Mikrosk Anat Forsch, 18, 520–52. Zimmermann, K. (1933). Ueber den Bau des Glomerulus der Saeugerniere. Z Mikrosk Anat Forsch, 32, 176–278.

CHAPTER 44

Function of the normal glomerulus Jean-Claude Dussaule, Martin Flamant, and Christos Chatziantoniou Relationship between glomerular structure and function The detailed structure of the glomerulus is described in Chapter 43. The glomerular barrier is composed of a layer of endothelial cells, the basal membrane, the foot processes interconnected by the slit diaphragm (SD) (Venkatachalam and Renke, 1978; Kanwar et  al., 1991; Jarad and Miner, 2009; Miner, 2011). Despite the lack of open pores, the glomerular barrier can be considered as a semi-permeable membrane due to its structural and biochemical characteristics. Water and soluble molecules of small molecular weight (up to 5 kDa) pass freely depending on the hydrostatic or osmotic pressure. For most authors, the layer of endothelial cells, its cell-coat, the three-dimensional structure of proteoglycans (long and ramified chains of polysaccharides), and the SD are the major physical barriers for the passage of macromolecules (Haraldsson et al., 2008; Satchell and Braet, 2009; Fogo and Kon, 2010; Friden et al., 2011). Recently, a subpodocyte space has been characterized that could contribute to the hydraulic resistance (Salmon et  al., 2009). Proteins that are required for the correct function of the SD include podocin, TRPC6, nephrin, and FAT or proteins that interact with the SD complex (CD2AP, Nck, Zona Occludens-1, synaptopodin) (Chuang and He, 2009). The important role of these proteins in ultrafiltration is evidenced by the proteinuria observed when their expression or structure is altered (see the hereditary nephrotic syndromes observed due to mutations of nephrin or podocin genes (Kestila et al., 1998)). The water permeability of this barrier is the highest observed in the organism since it can reach 180 L/24 hours for both kidneys (60 nL/glomerulus/ min). However, the glomerular barrier is not entirely semi-permeable since macromolecules up to 70 kDa can pass through by diffusion across the gel of the basal membrane, depending on their three-dimensional structure and electrical charge (Smithies, 2003). The negative electrostatic properties of the barrier have already been demonstrated. Moreover, a filtration-dependent electrical potential has been measured in micropuncture experiments in Necturus maculosus (Hausmanm et al., 2010). This electrical potential that is negative within the Bowman space is likely generated during the filtration process in humans (Hausmanm et al., 2012). Using dextrans (Chang et al., 1975), it has been clearly demonstrated that the percentage of the clearance of a macromolecule relative to a

freely filtered substance diminished not only with regards to its size but also to its negative electrical charge (Fig. 44.1). Independently of the charge, proteins or blood microparticles above the limit of 70 kDa are not filtered (Table 44.1). Some of these issues are described further in Chapter 43.

Composition of primitive urine The primitive urine results from the ultrafiltration of plasma and contains low levels of proteins (100–300 mg/L vs 72 g/L in plasma), which are almost completely reabsorbed in the proximal tubule. In contrast, the concentration in the primitive urine of molecules < 5 kDa is close to their plasma concentration. This concentration can be altered if one of the following applies: 1. Urine does not contain blood proteins and thus a correction factor equal to 7% should be applied for the soluble molecules 2. The retention of proteins (mainly negatively charged) by the glomerular barrier creates a shift in the equilibrium between the diffused anions and cations on both sides of the barrier. According to the Gibbs–Donnan law, anion concentration is higher in the urine than in plasma (and vice versa for cations). The highest the charge of electrolytes, the strongest becomes the shift of equilibrium. 3. Small molecules or ions (like calcium) that are usually bound to proteins undergo incomplete ultrafiltration.

Determinants of glomerular filtration Single nephron glomerular filtration rate The pressure leading to the creation of glomerular filtration follows similar principles with the exchanges between vascular and interstitial compartments applied in the rest of the body. In each point of glomerular capillary, the flow towards the Bowman’s space is proportional to the difference of hydrostatic and osmotic pressures according to the Starling’s equation (Maddox et al., 1975):

J = K [(Pcap − Pu) − (πcap − πu)]

where Pcap and πcap are the capillary hydrostatic and osmotic pressures, respectively, and Pu and πu are the hydrostatic and osmotic

Section 3  

the patient with glomerular disease is approximate because the filtration surface (and thus Kf) varies among glomeruli (it is 30–50% higher in juxtamedullary compared to cortical glomeruli).

100 Cationic dextran

80 Filtration (%)

Hydrostatic and osmotic profiles of the glomerulus

60 Anionic dextran

40

Neutral dextran

20 0

1.6

2

2.4

2.8 3.2 3.6 Molecular radius (nm)

4

4.4

Fig. 44.1  Percent of glomerular filtration of dextran molecules depends on their size and their electric charge.

Table 44.1  Relations between molecular weight, the Stokes’s molecular radius, and the percentage of glomerular filtration Substance

Molecular weight (Da)

Molecular radius (nm)

% of filtration

Water

18

0.10

100

Sodium

23

0.14

100

Urea

60

0.16

100

Glucose

180

0.36

100

Inulin

5500

1.48

100

Myoglobin

17,000

1.95

75

Egg albumin

43,500

2.85

22

Haemoglobin

68,000

3.25

3

Human albumin

69,000

3.55

40% at 80 years) (see Chapter 300). There is no actual reference for GFR in the ageing population. The previously proposed notion of a loss of 1 mL/min/1.73 m² per year after the age of 40 years appears excessive. It is however admitted that a value < 60 mL/min/1.73 m² is pathological, independently of age. Changes in the sodium intake between 20 and 1000  mmol/ 24 hours induce small variations of GFR. GFR is increased with the protein intake due to increase renal plasma flow by shifting the equilibrium of filtration along the glomerular capillary. The capacity of the kidneys to increase GFR above normal levels is called renal functional reserve (Bosch, 1995). It can be calculated by measuring the difference of GFR before and after protein overload. This process was proposed as predictive marker of renal hypertrophy following nephrectomy, although a direct link was never clearly established.

Regulation of renal blood flow and glomerular filtration rate The endocrine or paracrine regulation of GFR is a complex process because a variety of local or hormonal factors can interfere and alter the physical parameters of glomerular filtration (Table 44.2).

Table 44.2  Hormones and autacoids modulating GFR and RPF Vasoconstrictors

Vasodilators

Adenosine

Adenosine

Angiotensin II

Adrenomedullin

Antidiuretic hormone

ATP

ATP

Bradykinin

Endothelin

Dopamine

Growth factors (epidermal growth CGRP (calcitonin gene-related peptide) factor (EGF), platelet-derived growth factor (PDGF) VC Eff Art

50

25

function of the normal glomerulus

VC Aff Art

0

Fig. 44.3  In experimental conditions, vasoconstriction of afferent or efferent arterioles without changes of haemodynamic pressure induces a fall of renal blood flow. In the first case (afferent vasoconstriction), glomerular filtration rate decreases while it does not change when efferent arteriole is vasoconstricted because, in this case, glomerular capillary pressure increases.

Neuropeptide Y

Histamine

Leukotrienes LTC4 and LTD4

Insulin and insulin-like growth factor

Platelet activating factor

Natriuretic peptides (atrial, brain, and C-type (ANP, BNP, CNP))

Norepinephrine

Nitric oxide

Thromboxane (TXA2)

Prostaglandins E2 and I2

Vascular endothelial growth factor

PTH (parathyroid hormone)

Vasopressin

PTHrp (PTH-related peptide)

20 hydroxyeicosatetraenoic acid (20 HETE)

Relaxin

Note: adenosine and ATP can be vasodilators in several tissues but they are predominantly vasoconstrictors in renal vessels.

453

454

Section 3  

the patient with glomerular disease

Among these factors, angiotensin II, a vasoconstrictor, nitric oxide, and prostaglandins, vasodilators, and adenosine vasoconstrictor or vasodilator according to the activation of A1 or A2 receptors, play a major role in regulating GFR.

Hypovolaemia Activation of adrenergic system Activation of renin–angiotensin system

Endocrine and paracrine vasoconstrictors Angiotensin II Angiotensin II acts on its cell targets by activating the AT1 or AT2 receptors (Ardaillou et al., 1998). AT1 is the major receptor in the mature kidney and its activation is responsible for most of the vasoconstrictor actions of angiotensin II. Recent studies propose that a part of the effects of angiotensin II is mediated by the transactivation of growth factor receptors (Metha and Griendling, 2007), such as epidermal growth factor and platelet-derived growth factor. AT2 activation antagonizes the AT1 signalling and can create a negative retro-control of the AT1-induced effects. Angiotensin II is catabolized to angiotensin III by cleaving the N-terminal, and then to angiotensin IV, a hexapeptide. Deletion of the C terminal of angiotensin II leads to the formation of angiotensin-(1–7). Juxtaglomerular cells are the only renal cells capable of producing active renin. Renin synthesis and release by these cells is the limiting step of the activation of the renin–angiotensin system (RAS). In addition, the angiotensin-converting enzyme may act either as an ectoenzyme inserted in the cell membrane of endothelial or epithelial cells or as a plasma circulating enzyme. The renal synthesis of angiotensin II is particularly elevated due to the high local renin secretion and to the presence of the angiotensin-converting enzyme in the arteriolar endothelium, glomeruli, and renal tubular proximal epithelium (Ardaillou and Michel, 1999). Angiotensin II can also act on the renal vasculature as a circulating hormone because it can be also synthesized systemically (Crowley and Coffman, 2012). When a pressor dose of angiotensin II is injected in a euvolaemic or dehydrated animal, the afferent and efferent arteriolar resistances are increased whereas RBF and Kf decreased. The effect of angiotensin II is predominant in the mesangial and smooth muscle cells of the efferent arterioles and is mediated by the activation of AT1 receptors leading to an increase of cytosolic calcium. This vasoconstrictor effect has little influence in the glomerular filtration. GFR is decreasing but to a much lesser degree than RBF because of an increased Pcap. This is due to a more pronounced (or to at least equal) vasoconstriction of the efferent compared to the afferent arteriole. Furthermore, when the dose of angiotensin II is sub-pressor (without an increase in systemic blood pressure), GFR is very little altered whereas RBF clearly decreases. Similarly, in conditions of renin release (response to hypovolemia or following induction of renin synthesis), GFR is affected little compared to the major reduction of RBF (Navar 1998) (Fig. 44.4). In experimental conditions, chronic exposition to high levels of angiotensin II leads to a change of phenotype of podocytes that favours the presence of proteinuria (Huby et al., 2009; Palm, 2012) The dissociation of the effects of angiotensin II on RBF and GFR is better seen in some pathophysiological conditions. For instance, administration of angiotensin-converting enzyme inhibitors to heart failure or renovascular hypertension patients can lead to a decrease of GFR while RBF is normalized. This paradox can be explained by considering a predominant action of angiotensin II on efferent arterioles which in these particular pathological conditions is protective for the glomerular function.

Ang II

RAff REff

Pcap

NO PGE2 PGI2 RBF

–

–

KF

Stability of GFR

Fig. 44.4  Angiotensin II, in response to severe hypovolaemia, increases renal resistances and decreases renal blood flow (RBF). In these pathophysiological conditions, endocrine regulation overwhelms local autoregulation. Glomerular filtration rate is only slightly affected by the fall of RBF because angiotensin II favours the increase of glomerular Pcap and stimulates NO and prostaglandin production that counteracts its vasoconstrictor action on smooth muscle and mesangial cells.

When the RAS is not activated (euvolaemic conditions) antagonists of AT1 receptors have little effect on GFR, suggesting that endogenous angiotensin II is not a major regulator of glomerular filtration under normal conditions (Navar et al., 1996). It is difficult to dissociate the renal vascular effects of angiotensin II from those of nitric oxide (NO) and endogenous prostaglandins (PGE2 and PGI2) in clinical conditions in which RAS is activated or during experimental perfusion of angiotensin II. The close and rapid interaction between these vasoactive agents is clearly demonstrated by using cyclooxygenase or NO-synthase inhibitors which exacerbated the angiotensin II-induced decrease of RBF and GFR. Inversely, angiotensin II antagonizes its own action by inducing the synthesis of the above vasodilators (Navar et al., 1996) (Fig. 44.4).

Other vasoconstrictors Endothelin (ET) is a very potent vasoconstrictor synthesized mainly in endothelial cells. Two of the three existing isoforms, ET-1 and ET-3, are present in the kidney. These peptides act locally in the mesangial and arteriolar smooth muscle cells to activate signalling pathways that are similar to those of angiotensin II. ET-1 synthesis, the major renal isoform, is induced by agents increasing intracellular calcium and activating protein kinase C in the endothelium, by transforming growth factor-β and by shear stress. In contrast, NO inhibits ET-1 synthesis. Renal perfusion of ET-1, induces a sustained important decrease of RBF and GFR (Guan and Inscho, 2011), after a short initial increase due to the stimulation of NO. This biphasic action is due to the presence of ETB receptors in endothelial cells stimulating NO release and antagonizing the vasoconstrictor effect of ETA receptors present in smooth muscle cells. During the phase of vasoconstriction, ET-1 increases renal arteriolar resistance and decreases Kf, and its action is more prolonged compared to that of angiotensin II.

Chapter 44 

The role of ET-1 in renal physiopathology is well described (Hocher et  al., 1997), while is more controversial in physiological conditions. Administration of ETA antagonists does not affect RBF or GFR under normal conditions. Experimental studies suggested that ET-1 interacts with angiotensin II during hypovolemia. Angiotensin II induces ET-1 synthesis which in turn can inhibit renin synthesis (Herizi et  al., 1998). Inversely, ET-1 can have an indirect vasodilatory effect through activation of ETB endothelial receptors and subsequent release of NO. This interaction can explain the mechanisms of action of relaxin, involved in the glomerular hyperfiltration during pregnancy, because its effects are inhibited by the antagonists of ETB receptors and the blockers of NO synthesis. ATP and adenosine have paracrine effects on renal circulation (Vallon and Osswald, 2009). ATP is a neuro-modulator present in the renal nerve terminal and a circulating factor released by vascular cells (Jankowski, 2008). ATP when it is bound to P2y receptors of smooth muscle cells is a vasoconstrictor (Navar, 1998). When it is bound to endothelial P2y receptors stimulating NO synthesis, ATP is vasodilatory. Adenosine synthesized in macula densa cells acts as vasoconstrictor through A1 receptor activation leading to decreased cAMP synthesis in the afferent arteriolar smooth muscle cells. Adenosine is mainly involved in the control of tubuloglomerular feedback and can be partly involved in the NaCl–renin interaction. Its local production depends on activity of enzymes under the control of angiotensin II, which leads to a synergistic effect of both vasoconstrictors (Franco et al., 2009). Because at high doses, adenosine can activate A2 vasodilatory receptors in renal vasculature, it may be hypothesized that these A2 receptors buffer A1-induced vasoconstriction of pre- and post-glomerular arterioles (Carlstrom, 2011). The other vasoconstrictors shown in Table 44.2 play little role in the regulation of GFR under physiological conditions. In contrast, the involvement of several of them, such as thromboxane or leukotrienes sulphido-peptides is well established in pathological conditions.

Endocrine and paracrine vasodilators Nitric oxide NO is a major regulator of RBF and GFR under physiological conditions. NO is synthesized in endothelial and epithelial cells from L-arginine and in presence of two NO-synthase isoforms, NOS III and NOS I, respectively (Lamas and Rodriguez-Puyol, 2012). The renal glomerular and arteriolar endothelium is producing NO in response to shear stress and to the action of several vasoactive peptides increasing intracellular calcium (Gabbai and Blantz, 1999). NO secretion raised in macula densa during an increase of NaCl reabsorption. NO acts in the adjacent place to its production site cells, such as smooth muscle cells of afferent and efferent arterioles, mesangial and juxtaglomerular cells (NO is a vasodilator because it induces cGMP production in vascular and glomerular cells). Although its action on juxtaglomerular cells is less understood, most authors agree that NO increase renin synthesis. The non-selective inhibitors of NO-synthase like L-NAME induce an immediate and sustained reduction of RBF and GFR. These effects are due to a higher increase of resistance of the afferent compared to efferent arteriole and a decrease of Kf (Deng and Baylis, 1993). The effect of NO inhibitors on Pcap is variable and depends on the induction or not of hypertension with the used

function of the normal glomerulus

doses. These results show a vasodilator effect of NO in renal vessels and mesangium, and can explain, at least partly, the moderate increase of GFR during hypervolemia inducing an increased tubular NO synthesis. Some authors have proposed that L-arginine, the natural substrate of NO synthases, is involved in the activation of renal functional reserve during protein overload by increasing NO production in renal vessels. In these circumstances, NO could act in synergy to glucagon induced in pancreas by the amino acids (Bosch, 1995). The interactions of NO with angiotensin II have been described in the above paragraph. The involvement of NO as a major player of GFR regulation is also confirmed by its interaction with other factors inducing NO-synthesis such as kinins, insulin, insulin-like growth factor, calcitonin gene-related peptide (CGRP), and parathyroid hormone-related peptide (PTHrp, sharing structural similarities with the parathyroid hormone).

Other vasodilators The vasodilator prostaglandins prostacyclin or PGI2, and PGE2 are the major products synthesized by the arachidonic acid through activation of an enzymatic cascade involving phospholipase A2 and cyclooxygenases (constitutive or inducible). PGE2 acts through activation of its receptors EP2 and EP4 in renal resistance vessels and mesangium. PGI2 and PGE2 increase cAMP synthesis and thus induce vasodilatation (Sugimoto et al., 1994). Their main role is to antagonize the vasoconstrictor action of angiotensin II. Use of cyclooxygenase inhibitors has a detrimental effect and decrease RBF and GFR only in conditions in which RAS is activated (Navar et al., 1992). Perfusion of PGE2 or of a stable analogue of PGI2 in the renal artery increases RBF but has little effect on GFR because the decrease of arteriolar resistance is counterbalanced by the decrease of Kf (Schnermann and Briggs 1981). Kf is decreased because both PGs stimulate renin secretion and thus angiotensin II synthesis. A  similar dissociation between the alterations of RBF and GFR has been also observed with the other vasodilators, such as kinins, PTH, histamine, CGRP, or dopamine (Brenner et al., 1977). The natriuretic factors produced in the heart (such as atrial natriuretic peptide (ANP) and brain natriuretic peptide (BNP)) have active receptors in the renal vessels and glomeruli. When infused at pharmacological doses, ANP or BNP increase RBF and GFR. The increase of Kf is more due to the hydraulic permeability modification than to an increase of filtration surface, because the active receptors of ANP and BNP are mainly expressed in podocytes. The cytoskeletal modifications of podocytes result from the ANP- or BNP-induced increase of cytosolic cGMP concentration. Since the pharmacological doses of ANP or BNP perfusion exceed by far their physiological concentrations, it is not clear if natriuretic peptides are involved in the control of GFR under physiological conditions. It cannot be excluded that these peptides participate in the regulation of GFR during hypervolemia conditions which stimulate their cardiac synthesis (Brenner et al., 1990).

Renal nerve regulation The physiological importance of renal nerve regulation has been reassessed as a result of clinical and experimental data on the effects of renal denervation on blood pressure (Veelken and Schmieder, 2014). Renal innervations consist of efferent sympathetic and afferent sensory nerves (cholinergic innervations are absent). The renal action of afferent sensory nerves is not yet completely understood

455

456

Section 3  

the patient with glomerular disease

while the physiological role of efferent sympathetic renal nerves is well known. Renal sympathetic activity is implied in the response to hypovolemia, including sodium retention, renin secretion, and, to a lesser extent, decrease of RBF and GFR. The sympathetic system innervates cortical vessels including juxtaglomerular apparatus, mesangial cells and macula densa cells among other renal structures (Johns, 1989). The major renal neurotransmitter is norepinephrine. Stimulation of sympathetic nerves induces arterial vasoconstriction and thus decreases RBF and GFR by two mechanisms:  a direct vasoconstriction due to α-1-adrenergic receptors stimulation and an indirect effect of β-adrenergic receptors stimulation in the juxtaglomerular cells increasing renin secretion and angiotensin II synthesis. RBF decrease is more pronounced compared to GFR because of the concomitant vasoconstriction of afferent and efferent arterioles (Fig. 44.4) (Kalaitzidis et al., 2013).

Autoregulation of glomerular filtration rate and renal blood flow Alterations of afferent and efferent arteriole diameters affect RBF. When preglomerular resistance changes in a physiological range, RBF and GFR change to the same direction. When postglomerular resistance increases, RBF decreases but GFR remains almost unchanged due to the increase of glomerular capillary pressure. RBF does not change significantly for variations of mean arterial pressure between 80 and 180 mm Hg. This process, called ‘autoregulation’ of RBF, is observed even after denervation or in isolated perfused kidney thus suggesting the existence of internal adaptive mechanism(s) of haemodynamic stability. Autoregulation is considered to be essential for the normal function of the kidneys: due to the existence and efficiency of RBF autoregulation, GFR can remain stable within a large range of changes of perfusion pressure. Micropuncture experiments performed in a rat strain with easily accessible cortical glomeruli suggested that pre-glomerular resistance is a major factor influencing autoregulation (Just, 2007). RBF autoregulation is controlled by two mechanisms, one, very fast, is called myogenic response, and another one, slower, is called tubuloglomerular feedback. Myogenic response is an intrinsic function of vascular wall to contract in response to external stretching force. Tubuloglomerular feedback is a more complex mechanism specific to the kidney that leads to constriction of the afferent arteriole in response to an increase in sodium chloride concentration in the distal tubule and juxtaglomerular apparatus.

Myogenic control of renal blood flow RBF autoregulation is suppressed by papaverin which inhibits pre-glomerular vessel contractility. Similar inhibition is observed by blockers of voltage-operated type 2 calcium channels indicating the involvement of these channels in the myogenic response. It is noteworthy that the density of calcium channels is high in the pre-glomerular vessels, whereas it is negligible in post-glomerular vessels. The mechanisms leading to the opening of calcium channels during increase of vascular wall stretching are not yet very well known. It appears though that the parietal stretch-induced conformational change of plasma membrane is more important than the action of local paracrine agents. The opening of calcium channels produces calcium influx into the smooth muscle cells leading to cell contraction and to the increase of arteriolar resistance. This increase of vascular resistance maintains stable RBF despite the

increase of arterial pressure (Loutzenhiser et al., 2002). Moreover, this autoregulatory capacity plays a protective role against hypertensive renal damage since increased pre-glomerular resistances in response to high blood pressure prevent the parallel increase of glomerular capillary pressure (Bidani et al., 2009).

Tubuloglomerular feedback The proximity of macula densa with the afferent arteriole suggests a cross-talking between alterations of urinary flow rate or concentration and pre-glomerular resistance. Although several questions still remain unanswered for a complete comprehension of the tubuloglomerular feedback, the general outline of this mechanism is the following. During a sudden increase of perfusion pressure, filtration pressure and GFR are increased instantaneously leading to an increased delivery of sodium concentration to the macula densa (despite the above-described increase of pre-glomerular resistance and the glomerulotubular equilibrium of proximal tubule antagonizing this phenomenon). The increase of NaCl reabsorption by the NaKCC2 co-transporter localized at the apical side of the macula densa cells triggers the release of a vasoconstrictor signal in the afferent arteriole, vasoconstriction leading to the subsequent decrease of RBF, glomerular capillary pressure and GFR (Singh and Thomson, 2010). Recently, experiments on genetically altered mice allowed adenosine to be recognized as the major vasoconstrictor signal by the way of activation of A1 receptors. Studies on deficient mice in NTPDase1 or ecto-5′-nucleotidase demonstrate that adenosine comes from dephosphorylation of ATP released by tubular cells (Oppermann et al., 2008; Schnermann and Briggs, 2008). Angiotensin II, acting on AT1 receptors, is a co-factor of the vasoconstriction induced by adenosine in response to increased NaCl uptake by macula densa cells (Franco et  al., 2009)  while numerous vasodilators such as NO (Carlstrom et al., 2011), carbon monoxide (Ren et al., 2012), prostaglandins (Araujo and Welsh, 2010), and adenosine itself via A2 receptors (Bell and Welsh, 2009) can negatively modulate this tubular signal.

References Araujo, M. and Welch, W. J. (2010). Tubuloglomerular feedback is decreased in COX-1 knockout mice after chronic angiotensin II infusion. Am J Physiol Renal Physiol, 298, F1059–63. Ardaillou, R., Chansel, D., Chatziantoniou, C., et al. (1998). Biology and functions of renal receptors for angiotensin II and its active fragments. Adv Nephrol Necker Hosp, 28, 225–57. Ardaillou, R. and Michel, J. B. (1999). The relative roles of circulating and tissue renin-angiotensin systems. Nephrol Dial Transplant, 14, 283–6. Bell, T. D. and Welch, W. J. (2009). Regulation of renal arteriolar tone by adenosine: novel role for type 2 receptors. Kidney Int, 75, 769–71. Bidani, A. K., Griffin, K. A., Williamson, G., et al. (2009). Protective importance of the myogenic response in the renal circulation. Hypertension, 54, 393–8. Bosch, J. P. (1995). Renal reserve: a functional view of glomerular filtration rate. Semin Nephrol, 15, 381–5. Brenner, B. M., Ballermann, B. J., Gunning, M. E., et al. (1990). Diverse biological actions of atrial natriuretic peptide. Physiol Rev, 70, 665–99. Brenner, B. M., Bohrer, M. P., Baylis, C., et al. (1977). Determinants of glomerular permselectivity: insights derived from observations in vivo. Kidney Int, 12, 229–37. Brenner, B. M., Troy, J. L., and Daugharty, T. M. (1971). The dynamics of glomerular ultrafiltration in the rat. J Clin Invest, 50, 1776–80.

Chapter 44 

Bröchner-Mortensen, J. (1972). A simple method for the determination of glomerular filtration rate. Scand J Clin Lab Invest, 30, 271–4. Carlstrom, M., Wilcox, C. S., and Welch, W. J. (2011). Adenosine A2A receptor activation attenuates tubuloglomerular feedback responses by stimulation of endothelial nitric oxide synthase. Am J Physiol Renal Physiol, 300, F457–64. Chang, R. L., Ueki, I. F., Troy, J. L., et al. (1975). Permselectivity of the glomerular capillary wall to macromolecules. II. Experimental studies in rats using neutral dextran. Biophys J, 15, 887–906. Chuang, P. Y. and He, J. C. (2009). Signaling in regulation of podocyte phenotypes. Nephron Physiol, 111, 9–15. Cockcroft, D. W. and Gault, M. H. (1976). Prediction of creatinine clearance from serum creatinine. Nephron, 16, 31–41. Coresh, J., Astor, B. C., McQuillan, et al. (2002). Calibration and random variation of the serum creatinine assay as critical elements of using equations to estimate glomerular filtration rate. Am J Kidney Dis, 39, 920–9. Crowley, S. D. and Coffman, T. M. (2012). Recent advances involving the renin-angiotensin system. Exp Cell Res, 318, 1049–56. Deen, W. M., Robertson, C. R., and Brenner, B. M. (1972). A model of glomerular ultrafiltration in the rat. Am J Physiol, 223, 1178–83. Deng, A. and Baylis, C. (1993). Locally produced EDRF controls preglomerular resistance and ultrafiltration coefficient. Am J Physiol, 264, F212–15. Fawaz, A. and Badr, K. F. (2006). Measuring filtration function in clinical practice. Curr Opin Nephrol Hypertens, 15, 643–7. Flamant, M., Haymann, J. P., Vidal-Petiot, E., et al. (2012). GFR estimations using the Cokcroft-Gault, MDRD study and CKD-EPI equations in the elderly. Am J Kidney Dis, 60, 847–9. Flamant, M., Vidal-Petiot, E., Metzger, M., et al. (2013). Performance of GFR estimating equations in African Europeans: basis for a lower race-ethnicity factor than in African Americans. Am J Kidney Dis, 62, 182–4. Fogo, A. B. and Kon, V. (2010). The glomerulus—a view from the inside—the endothelial cell. Int J Biochem Cell Biol, 42, 1388–97. Franco, M., Perez-Mendez, O., and Martinez, F. (2009). Interaction of intrarenal adenosine and angiotensin II in kidney vascular resistance. Curr Opin Nephrol Hypertens, 18, 63–7. Friden, V., Oveland, E., Tenstad, O., et al. (2011). The glomerular endothelial cell coat is essential for glomerular filtration. Kidney Int, 79, 1322–30. Froissart, M., Rossert, J., Jacquot, C., et al. (2005). Predictive performance of the modification of diet in renal disease and Cockcroft-Gault equations for estimating renal function. J Am Soc Nephrol, 16, 763–73. Gabbai, F. B. and Blantz, R. C. (1999). Role of nitric oxide in renal hemodynamics. Semin Nephrol, 19, 242–50. Guan, Z. and Inscho, E. W. (2011). Endothelin and the renal vasculature. Contrib Nephrol, 172, 35–49. Haraldsson, B., Nystrom, J., and Deen, W. M. (2008). Properties of the glomerular barrier and mechanisms of proteinuria. Physiol Rev, 88, 451–87. Hausmann, R., Grepl, M., Knecht, V., et al. (2012). The glomerular filtration barrier function: new concepts. Curr Opin Nephrol Hypertens, 21, 441–9. Hausmann, R., Kuppe, C., Egger, H., et al. (2010). Electrical forces determine glomerular permeability. J Am Soc Nephrol, 21, 2053–8. Herizi, A., Jover, B., Bouriquet, N., et al. (1998). Prevention of the cardiovascular and renal effects of angiotensin II by endothelin blockade. Hypertension, 31, 10–14. Hocher, B., Thone-Reineke, C., Rohmeiss, P., et al. (1997). Endothelin-1 transgenic mice develop glomerulosclerosis, interstitial fibrosis, and renal cysts but not hypertension. J Clin Invest, 99, 1380–9. Huby, A. C., Rastaldi, M. P., Caron, K., et al. (2009). Restoration of podocyte structure and improvement of chronic renal disease in transgenic mice overexpressing renin. PLoS One, 4, e6721. Inker, L. A., Schmid, C. H., Tighiouart, H., et al. (2012). Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med, 367, 20–9.

function of the normal glomerulus

Jankowski, M. (2008). Purinergic regulation of glomerular microvasculature and tubular function. J Physiol Pharmacol, 59 Suppl 9, 121–35. Jarad, G. and Miner, J. H. (2009). Update on the glomerular filtration barrier. Curr Opin Nephrol Hypertens, 18, 226–32. Johns, E. J. (1989). Role of angiotensin II and the sympathetic nervous system in the control of renal function. J Hypertens, 7, 695–701. Just, A. (2007). Mechanisms of renal blood flow autoregulation: dynamics and contributions. Am J Physiol Regul Integr Comp Physiol, 292, R1–17. Kalaitzidis, R. G., Karasavvidou, D. and Siamopoulos, K. C. (2013). Renal sympathetic denervation and renal physiology. Curr Clin Pharmacol, 8(3), 189–96. Kanwar, Y. S., Liu, Z. Z., Kashihara, N., et al. (1991). Current status of the structural and functional basis of glomerular filtration and proteinuria. Semin Nephrol, 11, 390–413. Kestila, M., Lenkkeri, U., Mannikko, M., et al. (1998). Positionally cloned gene for a novel glomerular protein—nephrin—is mutated in congenital nephrotic syndrome. Mol Cell, 1, 575–82. Lamas, S. and Rodriguez-Puyol, D. (2012). Endothelial control of vasomotor tone: the kidney perspective. Semin Nephrol, 32, 156–66. Levey, A. S. (1990). Measurement of renal function in chronic renal disease. Kidney Int, 38, 167–84. Levey, A. S., Bosch, J. P., Lewis, J. B., et al. (1999). A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med, 130, 461–70. Levey, A. S., Coresh, J., Greene, T., et al. (2007). Expressing the Modification of Diet in Renal Disease Study equation for estimating glomerular filtration rate with standardized serum creatinine values. Clin Chem, 53, 766–72. Lewis, J., Agodoa, L., Cheek, D., et al. (2001). Comparison of cross-sectional renal function measurements in African Americans with hypertensive nephrosclerosis and of primary formulas to estimate glomerular filtration rate. Am J Kidney Dis, 38, 744–53. Loutzenhiser, R., Bidani, A. and Chilton, L. (2002). Renal myogenic response: kinetic attributes and physiological role. Circ Res, 90, 1316–24. Maddox, D. A., Bennett, C. M., Deen, W. M., et al. (1975). Determinants of glomerular filtration in experimental glomerulonephritis in the rat. J Clin Invest, 55, 305–18. Mariat, C., Alamartine, E., Barthelemy, J. C., et al. (2004). Assessing renal graft function in clinical trials: can tests predicting glomerular filtration rate substitute for a reference method? Kidney Int, 65, 289–97. Mehta, P. K. and Griendling, K. K. (2007). Angiotensin II cell signaling: physiological and pathological effects in the cardiovascular system. Am J Physiol Cell Physiol, 292, C82–97. Miner, J. H. (2011). Glomerular basement membrane composition and the filtration barrier. Pediatr Nephrol, 26, 1413–17. Nankivell, B. J., Gruenewald, S. M., Allen, R. D., et al. (1995). Predicting glomerular filtration rate after kidney transplantation. Transplantation, 59, 1683–9. Navar, L. G. (1998). Integrating multiple paracrine regulators of renal microvascular dynamics. Am J Physiol, 274, F433–44. Navar, L. G., Inscho, E. W., Majid, S. A., et al. (1996). Paracrine regulation of the renal microcirculation. Physiol Rev, 76, 425–536. Olbrich, O., Ferguson, M. H., Robson, J. S., et al. (1950). Simplified procedure for determining the renal clearance of inulin and diodone. Lancet, 2, 565–7. Oppermann, M., Friedman, D. J., Faulhaber-Walter, R., et al. (2008). Tubuloglomerular feedback and renin secretion in NTPDase1/ CD39-deficient mice. Am J Physiol Renal Physiol, 294, F965–70. Palm, F. (2012). The dark side of angiotensin II: direct dynamic regulation of the glomerular filtration barrier permeability to macromolecules. Am J Physiol Renal Physiol, 303, F789. Remuzzi, A., Brenner, B. M., Pata, V., et al. (1992). Three-dimensional reconstructed glomerular capillary network: blood flow distribution and local filtration. Am J Physiol, 263, F562–72.

457

458

Section 3  

the patient with glomerular disease

Ren, Y., D’ambrosio, M. A., Wang, H., et al. (2012). Mechanisms of carbon monoxide attenuation of tubuloglomerular feedback. Hypertension, 59, 1139–44. Salmon, A. H., Neal, C. R., and Harper, S. J. (2009). New aspects of glomerular filtration barrier structure and function: five layers (at least) not three. Curr Opin Nephrol Hypertens, 18, 197–205. Satchell, S. C. and Braet, F. (2009). Glomerular endothelial cell fenestrations: an integral component of the glomerular filtration barrier. Am J Physiol Renal Physiol, 296, F947–56. Savin, V. J. (1983). Ultrafiltration in single isolated human glomeruli. Kidney Int, 24, 748–53. Schnermann, J. and Briggs, J. P. (1981). Participation of renal cortical prostaglandins in the regulation of glomerular filtration rate. Kidney Int, 19, 802–15. Schnermann, J. and Briggs, J. P. (2008). Tubuloglomerular feedback: mechanistic insights from gene-manipulated mice. Kidney Int, 74, 418–26. Singh, P. and Thomson, S. C. (2010). Renal homeostasis and tubuloglomerular feedback. Curr Opin Nephrol Hypertens, 19, 59–64.

Smithies, O. (2003). Why the kidney glomerulus does not clog: a gel permeation/diffusion hypothesis of renal function. Proc Natl Acad Sci U S A, 100, 4108–13. Stevens, L. A., Claybon, M. A., Schmid, C. H., et al. (2011). Evaluation of the Chronic Kidney Disease Epidemiology Collaboration equation for estimating the glomerular filtration rate in multiple ethnicities. Kidney Int, 79, 555–62. Stevens, L. A., Coresh, J., Greene, T., et al. (2006). Assessing kidney function—measured and estimated glomerular filtration rate. N Engl J Med, 354, 2473–83. Sugimoto, Y., Namba, T., Shigemoto, R., et al. (1994). Distinct cellular localization of mRNAs for three subtypes of prostaglandin E receptor in kidney. Am J Physiol, 266, F823–8. Vallon, V. and Osswald, H. (2009). Adenosine receptors and the kidney. Handb Exp Pharmacol, 443–70. Veelken, R. and Schmieder, R. E. (2014). Renal denervation – implications for chronic kidney disease. Nat Rev Nephrol, 10, 305–13. Venkatachalam, M. A. and Rennke, H. G. (1978). The structural and molecular basis of glomerular filtration. Circ Res, 43, 337–47.

CHAPTER 45

Mechanisms of glomerular injury: overview Neil Turner Introduction Seen as a cartoon, the glomerulus is composed of three main cell types plus its glomerular basement membrane (GBM) (Fig. 45.1). The different pathological processes that damage glomeruli usually target one of these, although other cells and structures may be injured in bystander damage, or altered by the response to injury. Processes that may lead to glomerular injury include genetic, toxic, immune, and metabolic. Genetic conditions affecting the podocyte and GBM are summarized in Chapter  327 and Chapter 320 respectively, and toxic diseases in Chapter 82. The intention here is to give an overview of the type of processes leading to glomerular damage. The paradigm is particularly useful for understanding the primary glomerulopathies which are mostly autoimmune mediated, but metabolic, toxic, and genetic diseases can be understood in the same framework (Fig. 45.1). Mechanisms involved in individual diseases (e.g. autoantibodies in membranous nephropathy and anti-GBM disease, tissue injury in vasculitis and lupus) are considered in detail in chapters on individual diseases.

Two extremes define a spectrum The glomerulus has a limited repertoire of responses to injury. It can leak protein; it can leak blood; it can lose filtration function; it can cause hypertension. These are the key features of glomerular disease.

GBM Podocyte Endothelium

Mesangium

Fig. 45.1  Schematic diagram of a segment of a glomerulus. The podocyte (yellow) lies in the urinary space and is disordered in proteinuric diseases. In order to create glomerular haematuria, the red blood cell shown in the central capillary loop has to break through the GBM via defects created by pathological process (usually inflammatory).

It is useful to distinguish the pathological processes that cause proteinuria from those that cause haematuria. This distinction is clear at the initiation of most glomerular diseases, but some blurring occurs as it becomes chronic and the architecture of the glomerulus becomes increasingly abnormal. In particular, diseases that cause haematuria cause proteinuria if they leave scarring or progress. Nevertheless it remains a helpful distinction. Most glomerular disease can be positioned at an approximate point on a spectrum with these two quite different modes of glomerular injury at the extremes (Fig. 45.2). At one end, proteinuria, and at the other end, haematuria; ‘nephrotic’ versus ‘nephritic’ (see Table 42.1, Chapter 42). Some diseases are so protean that they cannot be put at a single point, they may present in a number of different ways.

Diseases that cause proteinuria It has become clear from studies of genetics, in vitro cell biology, and in vivo models that the podocyte is central to proteinuria. The three primary causes of pure nephrotic syndrome are all associated with some type of podocyte pathology: ◆

Minimal change disease (see Chapter 55)—podocyte dysfunction

◆

Membranous nephropathy (see Chapter 61)—podocyte attack by autoantibodies

◆ Focal

segmental glomerulosclerosis (see Chapter 57)—podocyte injury or death.

So are most of the genetic causes of nephrotic syndrome, which almost all involve podocyte genes (see Chapter 327). The major metabolic and systemic diseases causing nephrotic syndrome are associated with alterations to the matrix in the environment of podocytes through deposition of abnormal components and/or architectural changes. It is likely that this will disturb the function of these highly specialized, highly differentiated cells. In amyloidosis (see Chapter 152), deposition of fibrils affects the GBM directly beneath podocytes. In diabetes (see Chapter 149), a variety of abnormal matrix proteins are deposited. It is also possible that podocytes are directly injured by high glucose levels or by abnormally glycosylated proteins. An early change in diabetic nephropathy is increased deposition of normal GBM proteins (including collagen IV 345 network; see Chapter 320), similar to the changes seen in early membranous nephropathy, where the injury is mediated by autoantibodies. The podocyte may be responding to injury in diabetic nephropathy too.

460

Section 3  

the patient with glomerular disease The spectrum of glomerular diseases SLE IgA nephropathy Minimal change nephropathy FSGS

Anti-GBM disease Diabetic nephropathy

Small vessel vasculitis MPGN

Nephrotic

Membranous nephropathy

Amyloidosis

Post-streptococcal glomerulonephritis Nephritic

Mechanism • Injury to podocytes • Changed architecture: • Scanning • Deposition of matrix or other elements

Mechanism • Inflammation • Reactive cell proliferation Haematuria

• Breaks in GBM • Crescent formation

Proteinuria

Fig. 45.2  The spectrum of glomerular diseases. It is useful to contrast the pathological processes causing proteinuria (‘nephrosis’) on the left, from those causing glomerular haematuria (‘nephritis’) on the right. Common causes of acquired nephropathy are shown at the point on the spectrum at which they generally present, but genetic and other diseases can also be fitted into this concept. Further details in text.

Similar alteration of glomerular matrix may occur in post-inflammatory scarring caused by diseases that are burnt out, or active at low level, causing their position on the spectrum to shift from mostly haematuric to more of a mixture (see below). Low-level haematuria may occur in diabetes mellitus and some other diseases usually associated with nephrotic syndrome. This usually appears to be because of GBM fragility caused by deposition of abnormal material. Most rarer causes of nephrotic syndrome can also be explained by these mechanisms.

Modulating proteinuria via the podocyte Proteinuria is caused by podocyte dysfunction, and closely associated with progressive loss of renal function. There is evidence that some of the therapies we use now may be acting directly on podocytes, and understanding this may lead to new therapies. Contrary to the impression that can be created by electron micrographs, it emerges that podocytes are dynamic, active cells that are studded with scores of receptor types enabling them to respond to external mediators. For example Winn et al. (2006) listed 25 heptahelical receptors of 11 types, and there are many other classes. The slit diaphragm is not a static barrier to protein, it is susceptible to a wide range of external influences both chemical and mechanical (Table 45.1). Ransom et al. (2005) produced the first hard evidence that corticosteroids, one of our key immunosuppressive agents and the mainstay of therapy for minimal change nephrotic syndrome, might have direct structure-protective effects on podocytes. Faul et  al. (2008) showed that the immunosuppressive calcineurin

inhibitor tacrolimus also had direct effects on podocytes. This helps to explain why calcineurin inhibitors seem able to reduce proteinuria regardless of the cause (see Chapter 50; and described further in Chapter 52). Both of these drugs had previously been assumed to be acting through the immune system when used to treat proteinuric diseases (Mathieson, 2008). The observations raise the questions not only of whether minimal change disease is an immune disorder at all, but more broadly whether shared pathways in immune cells and podocytes might underlie some of the antiproteinuric effects of other immunomodulating agents. Schießl and Castrop (2013) showed in rats that angiotensin II directly modulates glomerular permeability to albumin largely independently of the perfusion pressure. This action was mediated by angiotensin II type 1 (AT1) receptors and partially attenuated by stimulation of AT2 receptors. Other in vitro and in vivo evidence points the same way and suggests that a direct effect on podocytes may be an important mode of action of agents that block the renin–angiotensin system. The literature includes a large amount of evidence around other mediators too (Table 45.1). Early evidence that targeting B7-1 may alter outcome of some nephrotic diseases (Yu et al., 2013) is exciting. We must hope that further additional therapeutic strategies are possible.

Diseases that cause haematuria Haematuric diseases are characterized by breaks in the GBM. This can be caused by having a fragile GBM for genetic or other reasons, but usually it is caused by inflammatory disruption of the GBM. The majority of haematuric conditions are slowly or rapidly

Chapter 45 

Table 45.1  Agents that alter podocyte phenotype or behaviour in vitro or in vivo. Derangement and differentiation are shorthand terms. ‘Deranged’ phenotype seems likely to be physiologically important, possibly for motility, filter de-clogging, or other purposes, but if sustained may be associated with proteinuria Promote derangement

Promote differentiation Comment and key citations

Unknown factor in minimal change disease Circulating factor in idiopathic FSGS Angiotensin

ACE inhibitors, ARBs

Schießl and Castrop (2012)

Endothelin Insulin Anti-VEGF therapy

VEGF

Stretch/tension TRPC6 signalling TLR4 signalling B7-1 signalling

Stimulated by LPS Abatacept

In patients with post-transplant FSGS, Yu et al. (2014)

Corticosteroids

Ransome et al. (2005)

Calcineurin inhibitors (tacrolimus, ciclosporin)

Faul et al. (2008)

TGF-β

Aldosterone antagonists All-trans retinoic acid PPAR-γ signalling PPAR-α signalling?

Fenofibrate, Ting and Keech (2013)

CDK2 inhibition Vitamin D Any cellular injury Abnormal matrix ACE = angiotensin-converting enzyme; ARB = angiotensin receptor blocker; CKD2 = cyclin-dependent kinase 2; FSGS = focal segmental glomerulosclerosis; LPS = lipopolysaccharide; PPAR = peroxisome proliferator-activated receptor; TGF-β = transforming growth factor beta; VEGF = vascular endothelial growth factor.

destructive diseases associated with infiltration of inflammatory cells and proliferation of endogenous cells of the glomerulus, probably in attempts at repair. When the injury is very recent or very minor (e.g. early glomerulonephritis or Alport syndrome), haematuria may occur alone, without substantial proteinuria.

Crescentic nephritis and rapidly progressive nephritis Crescentic nephritis can occur on a background of any of the ‘nephritic’ conditions. It is nephritic disease at its most severe and

mechanisms of glomerular injury: overview

usually associated with the clinical syndrome of rapidly progressive nephritis (see Chapter 70). It reflects a response to severe injury, not a specific immune phenomenon. Evidence for this comes from its occasional occurrence in non-inflammatory causes of haematuria such as Alport syndrome and amyloidosis.

Nephrotic/nephritic diseases Conditions such as membranoproliferative glomerulonephritis typically present with both haematuria and proteinuria but show some variation. These are inflammatory diseases, so sometimes the proteinuria may be due to scarring and alteration of the milieu for podocytes. However, in some there may be direct damage to the podocyte.

Diseases that cannot be pinned to  one region of the spectrum Immunoglobulin A  nephropathy may present with haematuria alone, but commonly also presents at a late stage by which time urinary abnormalities have progressed as a consequence of glomerular damage and scarring (see Chapter 66). However, it is always associated with some haematuria. Lupus nephritis may present with very different pathologies, so it may present with ‘pure’ nephrotic syndrome (no haematuria; e.g. lupus membranous), or with the most severe, crescentic disease, or with subacute disease when there is both haematuria and proteinuria (see Chapter 161).

Progressive renal disease after glomerular injury Mechanisms to explain the progression of renal disease after glomerular injury, and its close relationship to proteinuria, are summarized in Chapter 136. Three broad mechanisms are proposed to explain this and while not mutually exclusive, they have different implications for therapy and new drug targets: ◆ Haemodynamic—increased

glomerular perfusion pressure or stretch is the primary problem, angiotensin-converting enzyme inhibitors work by reducing this.

◆ Podocyte—podocyte

stress and death is the engine of progression (see Chapter 139) (Zhou and Turner, 2010). If so, potential therapies are listed in Table 45.1.

◆ Toxicity

of proteinuria—proteinuria creates progression by its effects on the tubulointerstitium (see Chapter 137).

These hypotheses are not mutually exclusive, and concepts such as disordered repair and replacement of cells must be important too (see Chapter 140). Proven ways to delay or arrest progression in clinical practice are discussed in Chapter 99.

References Faul, C., Donnelly, M., Merscher-Gomez, S., et al. (2008). The actin cytoskeleton of kidney podocytes is a direct target of the antiproteinuric effect of cyclosporine A. Nat Med, 14, 931–8. Mathieson, P. W. (2008). Proteinuria and immunity—an overstated relationship? N Engl J Med, 359, 2492–94. Ransom, R. F., Lam, N. G., Hallett, M. A., et al. (2005). Glucocorticoids protect and enhance recovery of cultured murine podocytes via actin filament stabilization. Kidney Int, 68, 2473–83.

461

462

Section 3  

the patient with glomerular disease

Schießl, I. M. and Castrop, H. (2013). Angiotensin II AT2 receptor activation attenuates AT1 receptor-induced increases in the glomerular filtration of albumin: a multiphoton microscopy study. Am J Physiol Renal Physiol, 305, F1189–200. Ting, R. D. and Keech, A. (2013). Fenofibrate and renal disease. Clin Lipidol, 8, 669–80 Winn, M. P., Daskalakis, N., Spurney, R. F., et al. (2006). Unexpected role of TRPC6 channel in familial nephrotic syndrome: does it have clinical implications? J Am Soc Nephrol, 17, 378–87.

Yu, C.-C., Fornoni, A., Weins, A., et al. (2013). Abatacept in B7-1–positive proteinuric kidney disease. N Engl J Med, 369, 2416–23. Zhou, Y. S. and Turner, A. N. (2010). New ways of thinking about proteinuria and progression of renal damage. Nephron Exp Nephrol, 16, e1–2.

CHAPTER 46

The patient with haematuria John Neary and Neil Turner The origin of haematuria Haematuria is a common primary care problem and in the majority of cases is not caused by renal disease (Fig. 46.1). It is also a common finding in screening programmes or routine medical examinations. Many algorithms have therefore been developed to aid management at primary care level. Age and whether the haematuria is visible (macroscopic) or invisible (microscopic) are commonly used as a primary sorting method. This is because the great majority of causes of visible, red blood in the urine fall into the province of the urologist, and similarly, with increasing age broadly urological causes are more likely than renal disease. This is an age gradient, not accurately defined by a fixed cut-off age, but US and UK guidelines suggest that for patients over the age of 40 the ‘urological pathway’ should usually be the first choice for investigation of even microscopic (non-visible) haematuria. Although the risks of urological malignancy are much lower with microscopic haematuria, in older patients this is usually the most important diagnosis to exclude initially. Recommended pathways differ according to resource availability, regional differences in incidence of causes, and simply variation in policies. Comparing the United Kingdom with the United States, recommendations are largely similar but the US guidelines specify computed tomography urography (CTU) as the imaging method of choice (Grossfeld et al., 2001a, 2001b).

Epidemiology of haematuria Studies looking at the prevalence of haematuria in the general population are complicated by the fact that results are method dependent, that is, whether dipsticks or microscopy have been used, or whether a single or multiple tests are used, and suggest widely different but relatively high prevalences of between 0.8% and 16%. The figure increases markedly with age (Table 42.1). The most informative study to date comes from screening data on over 1.2 million 16–25-year-olds screened for fitness for military service in Israel (Vivante et al., 2011), and followed for an average of 21 years by linkage to the end-stage renal disease (ESRD) database. 0.3% (3690; 0.4% of males, 0.2% of females) had persistent non-visible haematuria (>5 red blood cells per high power field (HPF), measured if dipstick positive, see below; identified on three separate occasions). Those with proteinuria > 200 mg per 24 hours were excluded. End-stage renal failure (ESRF) occurred in 34 per 100,000 person years in those with haematuria and two per 100,000 patient years without. The main causes of ESRD in this group were Alport syndrome (n = 4), immunoglobulin A (IgA) nephropathy (n = 4), and other glomerular disease (n = 7); 11 were attributed to

other conditions. Only one had cystic renal disease. The lower overall rate of haematuria in this study, and the increased risk associated with haematuria, is likely to be due to the high threshold used. The data from most of the studies in Table 46.1 reflect a belief that screening is valuable in identifying future risk. Its place is better established for proteinuria than haematuria (see Chapter 50), but even that is not an accepted screening test in most countries. The US taskforce (Grossfeld et al., 2001a, 2001b) found no case for haematuria screening. Some have questioned that following the Vivante (2011) study, but to put it in perspective, only 26 of the 565 (4.6%) cases of ESRF in that study came from those with persistent microscopic haematuria, and in 21 years only 26 of the 3690 haematuric individuals, 0.7%, developed ESRF. The test was therefore neither sensitive nor specific. Rate of ESRF did seem to be climbing in a linear manner however, suggesting that the lifetime rate of ESRF in those individuals will be at least three to four times higher. In Japan, screening urinalysis has been mandatory for workers and school-age children since the 1970s, extended in 1983 to include all adults aged 40 years or older (Iseki, 2012). Participants are classified based on the results of a single dipstick test. Results from 100,000 individuals in Okinawa in 1983 (Iseki et al., 1996) show that the incidence of haematuria measured this way rises quite steeply with age, and is significantly higher at all ages in women than men (7% versus 1% in 20s; 14% versus 6% in 70s). In 10 years of follow-up, 193 patients started dialysis. Proteinuria was a 10-fold stronger risk factor for ESRF than haematuria alone, but the combination of both conveyed approximately twice the risk of proteinuria alone.

Testing for haematuria Urinalysis techniques are described in Chapter 6. Dipsticks are the most common method for testing.

Detecting non-visible haematuria Dipsticks depend on the peroxidase activity of haem proteins and employ buffered tetramethylbenzidine and an organic peroxide to create a colour change. The detection limit is 5–20 intact red cells per microlitre or 15–600 micrograms per litre of free haemoglobin. Myoglobin will produce the same changes (Lamb and Price, 2011). Definitions of haematuria have traditionally emanated from microscopy, definitions varying between 1 and 10 red cells per HPF, but typically five. Dipstick testing can pick up blood in urine at significantly lower levels, one to two red blood cells/HPF, thus picking up cases which would not normally meet the criteria. False-positive results can occur with dipsticks if myoglobinuria or haemoglobinuria is present, and can also occur with the presence of semen in the

464

Section 3  

the patient with glomerular disease (A)

(B)

Fig. 46.1  Glomerular red cells in urine—different types of acanthocytes (left) and a red cell cast (right; inset shows a haemoglobin cast).

Table 46.1  Prevalence of haematuria in population studies Study and reference

Prevalence (%)

Children Dodge et al. (1976)

0.6 (girls), 0.3 (boys)

Vehaskari et al. (1979)

1.0–4.0

Armed forces recruits 18–33 years Froom et al. (1984)

5.2

Vivante et al. (2011); see text

0.4 (males), 0.2 (females)

Older adults (> 50 years) Mariani et al. (1989)

13 (males and females)

Mohr et al. (1986)

13 (males),14 (females)

Ritchie et al. (1986)

2.5 (males, age not stated)

Thompson (1987)

4 (males, aged 40–90 years)

Messing et al. (1982)

13 (males)

Britton et al. (1992)

18 (males)

Iseki et al. (1996)

6 (males), 14 (females)

Elderly (> 75 years) Mohr et al. (1986)

13 (males), 9 (females)

Different criteria were used to define the upper limit of ‘normal’ and hence the lower limit of ‘microhaematuria’ in all of these studies, as discussed in the text. Original papers should be consulted for details.

urine or a strongly alkaline urine. If free myoglobin or haemoglobin is present in urine, microscopy on fresh urine will be negative for red cells. False negative findings can be found in patients on high doses of vitamin C (Brigden et al., 1992).

Blood-coloured urine In the setting of macroscopic haematuria, it is useful to spin down the urine and examine the sediment and supernatant. In haematuria, the supernatant should be clear. If the supernatant is still coloured, it would suggest ingestion of foodstuffs such as beetroot/ food colourings, drug use with rifampicin or phenolphthalein, or the presences of porphyrin or urates (for a list of substances that may colour urine see Chapter 6, Table 6.1).

Evaluation of haematuria Exclude simple causes Red urine may be a consequence of some foods, drugs (see Chapter 6). Positivity of stick tests during menstruation is normal. Urinary tract infection commonly causes invisible haematuria (dipstick test for nitrite, leucocytes; microscopy and culture).

Visible haematuria Even a single episode of visible haematuria is significant unless there is an immediately apparent cause (Fig. 46.2); investigations as indicated in Fig. 46.3 are justified, unless the patient is suspected of having an aggressive glomerulonephritis. In those circumstances the quickest investigation is fresh urine microscopy for casts or fragmented and deformed red cells (see Chapter 6), along with scrutiny of the progression of tests of renal function and other signs of intrinsic renal disease. Presence of clots indicates non-glomerular bleeding.

Non-visible haematuria For non-visible haematuria it is important to repeat the test, consider other explanations, then test for infection by culture and microscopy. Non-visible haematuria is often transient and can be

Chapter 46 

Source of bleeding Renal

Glomerular

If age < 40 yrs

the patient with haematuria

Differences if < 40 yrs

IgA Nephropathy

Glomerular

Thin basement membrane

Renal

Alport syndrome Other glomerulonephrits

Interstitial

Intersitial nephritis (non-infective)

Interstitial

Infective

Infection - Bacterial/Viral/Fungal

Infective

Physical

Malignancy or other tumour

Infection - Tuberculosis

Calculi

Malignancy

Physical

Hypercalciuria/nephrocalcinosis

Trauma

Structural

Cysts - PKD Medullary sponge kidney

Vascular/ischamic

Papillary necrosis Sickle cell disease/trait Infarction AV fistula

Ureter

Malignancy

Structural Vascular/ischamic Infarction Malignancy

Calculi

Ureter

Stricture

Bladder

Uncertain

Bacterial cystitis (mainly women) Urethritis, prostatitis Tuberculosis/Schistomiasis Malignancy Calculi Stricture Radiation

UTI in older men and women

Bladder

Malignancy

Uncertain

Exercise haematuria Anticoagulation Facititious haematuria

Fig. 46.2  Causes of haematuria. More common causes are shown in larger font. On the right hand column are shown causes that differ in frequency in older patients. The division at 40 years is arbitrary but is chosen by some guidelines.

related to exercise, trauma, sexual activity, or menstruation. In large population studies it has been shown that up to 40% of people might expect to have dipstick haematuria on at least one occasion, but the number positive is approximately halved if the test is repeated (Froom et al., 1984; Loo et al., 2013). Few bacteriological laboratory microscopists use phase contrast microscopy or are alert to features of renal disease such as red cell morphology and nature of urinary casts, so this is rarely a means of picking out renal haematuria. As above, microscopy may be negative for red cells—this does not make the diagnosis of non-visible haematuria incorrect. Autoanalysers and automated image analysis may improve the diagnostic utility of this step in the future. If non-visible haematuria is persistent, the algorithm in Fig. 46.3 is relevant. Regardless of age, key features suggesting a renal explanation include: ◆ proteinuria—elevated

urinary albumin:creatinine ratio is a useful pointer to renal disease in young patients who do not have comorbid conditions, but less useful in older patients

◆ renal

impairment

◆ hypertension—though

this becomes less informative with age as the incidence of ‘essential’ hypertension is very high.

History should not only seek to exclude reversible causes such as urine infection but also explore possible family history of kidney disease (particularly Alport syndrome/ thin glomerular basement membrane nephropathy; see Chapters 321 and 323), and seek risk

factors for urothelial cancers (age, smoking history, occupational risks—in particular working with dyes, rubber, or chemicals, drug exposure including cyclophosphamide, aristocholic acid in herbal remedies, and analgesics). Travel history should ask about travel to areas where schistosomiasis is endemic (Chapter 181). Physical examination should include blood pressure and look for any physical manifestations of renal or other disease. Testing for urine protein and serum creatinine should be undertaken. Urine cytology no longer appears in the diagnostic pathway in the United Kingdom. While it has reasonably good sensitivity for bladder cancers (80%), it is less good for upper urothelial tract cancer. These sensitivities are not high enough to rule out malignancy so cystoscopy is required; cystoscopy is also required if it is positive. It is no longer a first-line test.

Patients on anticoagulants Anticoagulation has historically been said to be a cause of haematuria, but with careful monitoring of anticoagulation, this should not be the case. Patients on anticoagulation treatment should be investigated with the same concerns as any other patients with haematuria (unless international normalized ratio (INR) is significantly > 3 or there is bleeding from other sites. Patients with glomerulonephritis on anticoagulants may also be at increased risk of acute kidney injury (AKI) related to glomerular bleeding (see below).

465

466

Section 3  

the patient with glomerular disease Visible haematuria (VH) Plasma creatinine/eGFR Exclude transient cause including UTI

Non-visible haematuria (NVH) Exclude transient causes including UTI

Symptomatic non-visible haematuria (s-NVH)

Asymptomatic non-visible haematuria (a-NVH) 2 of 3 dipstick tests positive

Yes

No

Stop

Blood pressure Plasma creatinine/eGFR Send urine for ACR or PCR

≥ 40yrs

< 40yrs

Normal All of: • eGFR ≥60ml/min AND • ACR 10ml/min fall within 5 years) BAUS/RA Guidelines: Initial assessment of haematuria.

Fig. 46.3  UK recommendations for the management of haematuria. Joint guideline from the Renal Association and the British Association of Urological Surgeons. Further commentary and detail is provided in Anderson et al. (2008).

Patients who live in or visit the tropics In some parts of the world, schistosomiasis (see Chapter 181) is the dominant cause of haematuria. Pathways for assessment will recognize this with urine microscopy for ova, or empirical treatment, as initial steps. Recognizing travellers who have picked up schistosomiasis and then returned to non-endemic areas is important but difficult as they may not recall their travel, or its significance. Serum antibody for schistosomal exposure may be useful to exclude the diagnosis in this group.

Management when investigations are negative Visible haematuria with negative investigations Investigations for visible haematuria should usually include cystoscopy and imaging of kidneys, ureters, and bladder. CTU is now the preferred imaging technique if available as it has superior sensitivity and specificity compared to intravenous urography (IVP)

Chapter 46 

(Maher et al., 2004; O’Connor et al., 2008) and one study comparing CTU to IVP demonstrated a sensitivity of 100% versus 61% and a specificity of 97% versus 91% (Gray Sears et al., 2002). Renal ultrasound may be useful in determining size of kidney and any large masses, but is less sensitive than CTU , especially with lesions < 2 cm—sensitivity 26–60% (Warshauer et al., 1998). There are circumstances when CTU cannot be carried out—renal impairment, pregnancy, and hypersensitivity to contrast medium. In these settings, ultrasound or consideration of magnetic resonance urography would be reasonable alternatives. If these investigations have been carried out and are negative after a single episode of visible haematuria, reassessment after a period may be indicated. If they are negative and significant visible haematuria is still occurring, testing for rarer causes of bleeding is indicated. This may include cannulation of ureters to localize bleeding, angiography to identify renal arteriovenous malformations, and possibly consideration of nutcracker phenomenon (see Chapter 48), for example.

Persistent non-visible haematuria with negative investigations Between 19% and 68% of patients with non-visible haematuria remain undiagnosed (Howard and Golin, 1991; Khadra et  al., 2000; Cohen and Brown, 2003). There is much debate as to how these patients should be managed long term, but a reasonable approach is that they should be followed in 6 months then annually. At each visit, checks of blood pressure, serum creatinine, urine protein:creatinine ratio, and dipstick should be made. If the haematuria settles on two consecutive urinalyses and no other features have developed, then they can be discharged. If the patient develops proteinuria, renal impairment, hypertension, or visible haematuria they should be re-evaluated. In patients with microscopic haematuria in the absence of proteinuria or renal impairment, or a family history, the most common abnormalities on renal biopsy in developed world series are normal findings, IgA nephropathy, or thin basement membrane disease. The management of none of these conditions is altered by knowing this; monitoring remains the usual management (Richards et  al., 1994; Sparwasser et  al., 1994; Topham et  al., 1994; McGregor et al., 1998; Hall et al., 2004; Choo et al., 2013). Most nephrologists therefore do not recommend renal biopsy unless there are particular reasons for seeking greater diagnostic certainty.

Acute kidney injury caused by glomerular haematuria The phenomenon of AKI during profuse glomerular bleeding causing macroscopic haematuria was first reported in 1983 (Kincaid-Smith et al., 1983). Since then there have been numerous reports, mostly relating to IgA nephropathy in relatively young patients (reviewed in Moreno et  al., 2012); indeed, IgA nephropathy seems over-represented, it is rarely recognized in other conditions. AKI can be severe requiring temporary dialysis, and while most cases resolve, recovery of glomerular filtration rate (GFR) may be incomplete. Biopsies tend to show many red cell casts and it is presumed to have a mixed obstructive-toxic aetiology (see Chapter 221). Free haem is well known to cause AKI in

the patient with haematuria

haemoglobinuria of various aetiologies, and also in rhabdomyolysis, and this may be part of the aetiology. Possible mechanisms of tubular damage are reviewed by Moreno et al. (2012). Patients on anticoagulants may be at increased risk of developing this complication of glomerular bleeding. A number of reports have identified AKI more severe than glomerular changes would predict in patients on anticoagulants, usually warfarin, attracting the label warfarin-related nephropathy (Brodsky et al., 2009). Patients affected have usually had INR well over the therapeutic range. A later study (Brodsky et al., 2011) looked at creatinine changes retrospectively in 4006 patients developing an INR > 3, raising the possibility that the phenomenon might occur in patients without glomerular disease, but this is speculative. Outcomes in case reports have tended to be poor, with incomplete recovery of GFR.

References Anderson, J., Fawcett, D., Feehally, J., et al. (2008). Joint Consensus Statement on the Initial Assessment of Haematuria. [Online] Brigden, M. L., Edgell, D., McPherson, M., et al. (1992). High incidence of significant urinay ascorbic acid concentrations in a West Coast population—implication for routine analysis. Clin Chem, 38, 426–31. Britton, P. J., Dowell, A. C., Whelan, P., et al. (1992). A community study of bladder cancer screening by the detection of occult urinary bleeding. J Urol, 148, 788–90. Brodsky, S. V., Satoskar, A., Chen, J., et al. (2009). Acute kidney injury during warfarin therapy associated with obstructive tubular red blood cell casts: a report of 9 cases. Am J Kidney Dis, 54, 1121. Brodsky, S. V., Nadasdy, T., Rovin, B. H., et al. (2011). Warfarin-related nephropathy occurs in patients with and without chronic kidney disease and is associated with an increased mortality rate. Kidney Int, 80, 181. Choo, B. S., Hahn, W. H., Cheong, H. I., et al. (2013). A nationwide study of mass urine screening tests on Korean school children and implications for chronic kidney disease management. Clin Exp Nephrol, 17, 205–10. Cohen, R. A. and Brown, R. S. (2003). Clinical practice. Microscopic hematuria. N Engl J Med, 348(23), 2330–8. Froom, P., Ribak, J., and Benbassat, J. (1984). Significance of microhaematuria in young adults. Br Med J (Clin Res Ed), 288, 20–2. Gray Sears, C. L., Ward, J. F., Sears, S. T., et al. (2002). Prospective comparison of computerized tomography and excretory urography in the initial evaluation of asymptomatic microhaematuria. J Urol, 168, 2457. Grossfeld, G. D., Litwin, M. S., Wolf, J. S., Jr., et al. (2001a). Evaluation of asymptomatic microscopic hematuria in adults: the American Urological Association best practice policy—Part I: Definition, detection, prevalence, and etiology. Urology, 57, 599–603. Grossfeld, G. D., Litwin, M. S., Wolf, J. S., Jr., et al. (2001b). Evaluation of asymptomatic microscopic hematuria in adults: the American Urological Association best practice policy. II. Patient evaluation, cytology, voided markers, imaging, cytoscopy, nephrology evaluation, and follow-up. Urology, 57, 604–10. Hall, C. L., Bradley, R., Kerr, A., et al. (2004). Clinical value of renal biopsy in patients with asymptomatic microscopic hematuria with and without low-grade proteinuria. Clin Nephrol, 62(4), 267–72. Howard, R. S. and Golin, A.L. (1991). Long-term follow up of asymptomatic microhematuria. J Urol, 145, 335–6. Iseki, K. (2012). Evidence for asymptomatic microhematuria as a risk factor for the development of ESRD. Am J Kid Dis, 60, 12–14. Iseki, K., Iseki, C., Ikemiya, Y., et al. (1996). Risk of developing end-stage renal disease in a cohort of mass screening. Kidney Int, 49, 800–5. Khadra, M. H., Pickard, R. S., Charlton, M., et al. (2000). A prospective analysis of 1,930 patients with hematuria to evaluate current diagnostic practice. J Urol, 163, 524–7. Kincaid-Smith, P., Bennet, W. M., Dowling, J. P., et al. (1983). Acute renal failure and tubular necrosis associated with hematuria due to glomerulonephritis. Clin Nephrol, 19, 206–10.

467

468

Section 3  

the patient with glomerular disease

Lamb, E. J. and Price, C. P. (2011). Kidney function tests. In C. A. Burtis, E. R. Ashwood, and D. E. Bruns (eds.) Tietz Textbook of Clinical Chemistry and Molecular Diagnostics (5th ed.), pp. 669–707. St. Louis, MO: Elsevier, Loo, R. K., Lieberman, S. F., Slezak, J. M., et al. (2013). Stratifying risk of urinary tract malignant tumors in patients with asymptomatic microscopic hematuria. Mayo Clin Proc, 88(2), 129–38. Maher, M. M., Kalra, M. K., Rizzo, S., et al. (2004). Multidetector CT urography in imaging of the urinary tract in patients with hematuria. Korean J Radiol, 5(1), 1–10. Mariani, A. J., Mariani, M. C., Macchioni, C., et al. (1989). The significance of adult hematuria: 1,000 hematuria evaluations including a risk benefit and cost-effectiveness analysis. J Urol, 141, 350–5. McGregor, D. O., Lynn, K. L., Bailey, R. R., et al. (1998). Clinical audit of the use of renal biopsy in the management of isolated microscopic hematuria. Clin Nephrol, 49, 345–8. Messing, E. M., Young, T. B., Hunt, V. B., et al. (1992). Home screening for hematuria: results of a multiclinic study. J Urol, 148, 289–92. Mohr, D. N., Offord, K. P., Owen, R. A., et al. (1986). Asymptomatic microhematuria and urologic disease: a population-based study. JAMA, 256, 224–9.

Moreno, J. A., Martín-Cleary, C., Gutiérrez, E., et al. (2012). AKI associated with macroscopic glomerular hematuria: clinical and pathophysiologic consequences. Clin J Am Soc Nephrol, 7, 175–84. O’Connor, O. J., McSweeney, S. E., and Maher, M. M. (2008). Imaging of hematuria. Radiol Clin North Am, 46, 113–32. Richards, N. T., Darby, S., Howie, A. J., et al. (1994). Knowledge of renal histology alters patient management in over 40% of cases. Nephrol Dial Transplant, 9, 1255–9. Sparwasser, C., Cimniak, H. U., Treiber, U., et al. (1994). Significance of the evaluation of asymptomatic microscopic haematuria in young men. Br J Urol, 74, 723–9. Sutton, J. M. (1990). Evaluation of hematuria in adults. JAMA, 263, 2475–80. Topham, P. S., Harper, S. J., Furness, P. N., et al. (1994). Glomerular disease as a cause of isolated microscopic haematuria. QJM, 87, 329–35. Vivante, A., Afek, A., Frenkel-Nir, Y., et al. (2011). Persistent asymptomatic isolated microscopic hematuria in Israeli adolescents and young adults and risk for end-stage renal disease. JAMA, 306, 729–36. Warshauer, D. M., McCarthy, S. M., Street, L., et al. (1998). Detection of renal masses: sensitivities and specificities of excretory urography/linear tomography, US, and CT. Radiology, 169, 363–5.

CHAPTER 47

Loin pain haematuria syndrome John Neary Introduction Loin pain haematuria syndrome (LPHS) was first described by Little and colleagues in 1967 (Little et al., 1967) who described three young females with recurrent bouts of flank pain and haematuria which occurred in the absence of identifiable or relevant urinary tract disease. Since then, a number of case series have been published amounting to approximately 300 cases worldwide, meaning it remains a rare condition. However, it is likely that significant under-reporting occurs.

Clinical features The most obvious clinical feature is that of loin pain which is most often unilateral although can progress to bilateral pain. This may or may not be associated with haematuria which is generally microscopic. Renal function is normal, no significant proteinuria is present, and hypertension is usually absent. This is a diagnosis of exclusion, so no other obvious cause of haematuria or pain should be found on urological or nephrological investigation—in particular, no evidence of infection, obstruction, arteriovenous malformations, or malignancy. Patients described in the literature tend to be young (mostly < 30 years), Caucasian (94%), and female (74%). They often have a background of renal calculi (~ 50%) or less commonly immunoglobulin A nephropathy (~ 20%). Patients with a history of calculi should have no evidence of obstruction on imaging at the time of presentation with LPHS. Patients with an underlying glomerular disorder are given the diagnosis of secondary LPHS (Spetie et al., 2006). By the time they are seen by a nephrologist, patients have often seen a number of physicians, had multiple investigations, have frequently developed serious disability due to chronic pain, and may be opiate dependent (Coffman, 2009). Some authors feel there is a large element of somatization involved in the syndrome and point to peaks of pain associated with episodes of parental illness or increased pain with psychological triggers (Lucas et al., 1995; Lall et al., 1997). Many of these patients also meet the criteria for somatoform pain disorder but the same authors acknowledge the fact that the psychological symptoms may be secondary to long-standing pain and frustration at lack of an effective cure (Bass et al., 2007). Another small series reported use of a number of pain scoring systems which suggest that LPHS patients have a predominant or significant neuropathic element to their pain (Smith and Bajwa, 2012).

Investigations Patients with LPHS are often found to have either microscopic or less commonly macroscopic haematuria, with dysmorphic red cells (acanthocytes) on urine microscopy suggesting a glomerular

aetiology. In general, renal biopsy is not often performed due to lack of evidence of kidney disease with normal creatinine and lack of proteinuria. However, if renal biopsy is undertaken, blood is often seen in the renal tubular cross-sections (~7% LPHS vs 2% healthy controls) (Hebert et al., 1996; Spetie et al., 2006). A series of other findings have been reported, but importantly, none of these changes are specific or sensitive and none of them can confirm or rule out the diagnosis. For example, the renal vasculature may show C3 deposition in the arteriolar walls (Naish et  al., 1975; Spitz et  al., 1997). Some early groups had suggested abnormalities in the renal vasculature (Little et al., 1967; Burden et al., 1979; Fletcher et al., 1979), but this does not feature in recent reports. It has been suggested that these earlier angiographic abnormalities were due to contrast-induced spasm (Bergroth et al., 1987). The largest series of biopsies in LPHS reported a wide variation in glomerular basement membrane (GBM) width between thin, thick, and normal basement membrane (~ 1/3 in each group) (Spetie et al., 2006). This partially confirmed an earlier finding of increased prevalence of thin basement membrane in these patients (Hebert et al., 1996). Other investigators report a high prevalence of hypercalciuria and hyperuricosuria in these patients and have suggested that, in combination with glomerular haematuria causing tubular obstruction, this may lead to intratubular microcrystal formation (Praga et al., 1998). It is difficult to find a clearly defined diagnosis of LPHS in the earliest studies, and there remains a large variation in clinical features in reported patients to date. Some patients have a history of renal calculi, some of somatoform disorder, and presence of haematuria is variable. This heterogenous group of patients may account for the variation in investigative findings in the patients studied. However, one group has recently suggested diagnostic criteria for primary LPHS (Spetie et al., 2006), including renal pain which is constant or recurrent over a 6-month period, haematuria present on almost all urinalyses, and absence of obstruction on imaging if past history of calculi.

Aetiology/pathology An important observation is that many of the associated abnormalities can explain the finding of blood in the urine, but cannot easily explain the pain. The great majority of patients with glomerular haematuria do not experience similar pain. It has been suggested that LPHS is part of a general somatoform pain disorder with some renal involvement (Lucas et al., 1995; Lall et al., 1997; Bass et al., 2007). Some of the common longitudinal features of the condition, such as recurrence despite autotransplantation, may seem to support that.

470

Section 3  

the patient with glomerular disease 2

1

Episode of glomerular hypertension induced by: a. Systernic hypertension b. Renal efferent arteriolar vasoconstriction (e.g., exercise) c. Renal afferent arteriolar vasodilation (e.g., vasodilator drugs, hyperthermia) d. Renal tubular obstruction by collecting duct microcrystals

Structurally abnormal GBM, genetically determined

6

5

3 Spontaneous GBM instability 4 Rupture of glomerular capillary wall with haemorrhage into renal tubules

Microcrystal deposition in ducts due to: a. Hyperconcentration of glomerular filtrate in tubules with slowed flow from partial tubular obstruction b. Damage to collecting duct epithelial cells causing loss of protective proteins against calcium nidation (e.g., osteopontin)

7 Tubular obstruction

Glomerular hypertension

9

8

Backleak of glomerular filtrate and local parenchymal oedema

Compression of adjacent tubules

10 Local renal parenchymal oedema leads to global renal parenchymal oedema because of vicious cycle (9 8 7 4 5 9) 12 Increased stretching of renal capsule because of abnormally increased compliance of the capsule or the perirenal adventitia

11 Stretching of renal capsule

14

13 Severe flank pain

Abnormally enhanced pain perception, genetic? Acquired?

Fig. 47.1  Proposed pathogenesis of pain in idiopathic loin pain haematuria syndrome. Note that the process is initiated by glomerular bleeding. From Spetie et al. (2006).

A number of potential aetiologies for the associated haematuria have been proposed including abnormal platelet activation and prostacyclin production, complement activation, clotting abnormalities, renal tubular calcification, oestrogen-containing oral contraceptive pill, or somatization (Weisberg et  al., 1993; Smith and Bajwa, 2012). These are reported in individual case studies or small case series only. The presence of dysmorphic red cells is common and would seem to point to a glomerular source of bleeding. As in others with ‘benign’ haematuria, there is a suggestion that variation in GBM thickness may have a role. This has resulted in a hypothesis that abnormal GBM leads to glomerular bleeding which results in tubular obstruction by red cells. This in turn leads to back-leak and parenchymal oedema, with subsequent glomerular hypertension, which predisposes to further glomerular bleeding and so on (Spetie et al., 2006). However, this process occurs in the great majority of patients with thin GBM nephropathy (see Chapter  325) without causing any symptoms. Fig. 47.1 illustrates a proposed mechanism (Spetie et al., 2006).

It is interesting that of the many case reports of nutcracker syndrome (see Chapter 48), a condition that is generally associated with macroscopic haematuria, pain is rarely a mentioned feature. It has been hypothesized that this cycle may be exacerbated by the presence of hypercalciuria and hyperuricosuria leading to microcrystal formation (Praga et  al., 1998). Tubular obstruction and interstitial oedema then results in renal parenchymal oedema and renal capsular stretching, causing renal pain. Enhanced pain perception in these patients may also play a role.

Treatment A number of pharmacologic treatments, which have been directed towards correction of platelet activation or clotting abnormalities including aspirin and warfarin, have been unsuccessful (Weisberg et al., 1993). Use of renin–angiotensin system inhibition to reduce glomerular hypertension has been proposed and this led to some improvement in symptoms in four of seven patients treated with enalapril (Hebert et al., 1996). In patients with a history of

Chapter 47 

nephrolithiasis, adherence to stone prevention fluid and diet guidelines should be followed, if not doing so already. It should also be noted that a number of cases of LPHS resolve spontaneously as patients get older, with approximately 30% of cases resolving at a mean of 3.5 years (Weisberg et al., 1993). The cornerstone of treatment of LPHS remains analgesia management and for many patients the best results may be achieved through management by a multidisciplinary team including psychiatric or psychological and expert pain management input. A conservative approach, with long-term support and continuity of care with as few practitioners as possible, can have beneficial results, with up to a 38% improvement rate, even in long-standing cases (Bass et al., 2007). A number of surgical treatments have also been attempted although the poor overall results of these are attested to in many case reports: ◆ Intrarenal

capsaicin injections were initially reported to have an approximately 60% improvement in symptoms, but this was only short term (mean duration 2–17 weeks) and the side effects were significant including bladder pain, worsening renal pain, and deterioration in renal function. Most worryingly, a nephrectomy rate of between 20% and 67% was reported and thus use of capsaicin is no longer encouraged (Uzoh et al., 2009).

◆ A 

trial of ureteric bupivacaine infusion (Ahmed et  al., 2010) reported 95% improvement rate at 1 year in a cohort of 17 patients, with 23% reporting complete resolution after one treatment. Several patients required repeat treatments (mean number of treatments = 2.9) and a well-validated pain-scoring method was not used. A larger trial is required.

◆ Denervation

of the kidney, with surgical stripping of the renal pedicle and renal capsulotomy, has been reported with an approximately 30% cure rate, but some groups have reported recurrence of symptoms secondary to re-innervation in approximately 70% cases at a mean of 11 months, development of pain in the contralateral kidney, and, of concern, a large percentage of patients, 38%, subsequently proceeding to nephrectomy (Andrews et al., 1997; Greenwell et al., 2004).

◆ Autotransplantation

of the kidney (with coincident renal denervation) has also been reported with good success in some groups reporting up to 76% of patients pain free at 1 year with follow-up of up to 8.4 years (Chin et al., 1998; Sheil et al., 1998), although other groups have had less success, with up to 75% recurrence and significant morbidity including occasional graft loss or recurrence of pain in contralateral kidney (Gibson et al., 1994; Harney et  al., 1994; Parnham et  al., 1996). Even in the programmes reporting success with autotransplantation, there is significant morbidity with a 7% rate of graft loss due to surgical complications and an 11% rate of nephrectomy following autotransplant (Chin et al., 1998).

There is some debate as to why the denervation seen in autotransplantation seems to have better success than standard denervation, as regrowth of nerves is seen in both groups. It has been proposed that use of interposed polytetrafluroethylene (PTFE) arterial grafts should reduce re-innervation but experience with this technique is limited (Blacklock et al., 1999). The majority of authors would propose that autotransplantation is only an option of last resort (Dube et al., 2006), while others feel

loin pain haematuria syndrome

that due to the high risk of morbidity and/or recurrence of pain, this procedure should not be considered. As negative results are less likely to get published, it is also likely that many unsuccessful cases of surgical treatment have not been reported. Potential other treatments include intrathecal opiates (Prager et al., 1995), paraspinal nerve root stimulation (Goroszeniuk et al., 2009), and finally catheter-based renal denervation (Gambaro et al., 2013), but many of these are isolated case reports or small case series.

Summary In summary, the relatively small numbers of patients with LPHS and the wide heterogeneity in clinical findings may account for the varying degree of success of treatment options. Anecdotal successful reports may be due in part to placebo effect or other confounding variables. In future studies, clearly defined patient categories may need to be identified before trialling treatments in order to identify which treatments may be successful.

References Ahmed, M., Acher, P., and Deane, A. M. (2010). Ureteric bupivicaine infusion for loin pain haematuria syndrome. Ann R Coll Surg Engl, 92(2), 139–41. Andrews, B. T., Jones, N. F., and Browse, N. L. (1997). The use of surgical sympathectomy in the treatment of chronic renal pain. Br J Urol, 80, 6–10. Bass, C. M., Parrott, H., Jack, T., et al. (2007). Severe unexplained loin pain (loin pain haematuria syndrome): management and long-term outcome. QJM, 100, 369–81. Bergroth, V., Konttinen, Y. T., Nordstrom, D., et al. (1987). Loin pain and haematuria syndrome: possible association with intrarenal arterial spasms. Brit Med J, 294(6558), 1657. Blacklock, A. R., Raabe, A. L., and Lam, F. T. (1999). Renal auto-transplantation with interposed PTFE arterial graft: not necessarily a cure for loin pain/haematuria syndrome. J R Coll Surg Edinb, 44(2), 134. Burden, R. P., Dathan, J. R., Etherington, M. D., et al. (1979). The loin pain/ haematuria syndrome. Lancet, 1(8122), 897–900. Chin, J. L., Kloth, D., Pautler, S. E., et al. (1998). Renal autotransplantation for the loin pain hematuria syndrome: long-term follow-up of 26 cases. J Urol, 160, 1232–5. Coffman, K. L (2009). Loin pain hematuria syndrome: a psychiatric and surgical conundrum. Curr Opin Org Transplant, 14, 186–90. Dube, G. K., Hamilton, S. E., Ratner, L. E., et al. (2006). Loin pain hematuria syndrome. Kidney Int, 70, 2152–5. Fletcher, P., Al-Khader, A. A., Parsons, V., et al. (1979). The pathology of intrarenal vascular lesions associated with the loin-pain-haematuria syndrome. Nephron, 24(3), 150–4. Gambaro, G., Fulignati, P., Spinelli, A., et al. (2013). Percutaneous renal sympathetic nerve ablation for loin pain haematuria syndrome. Nephrol Dial Transplant, 28, 2393–95. Gibson, P., Winney, R. J., Masterton, G., et al. (1994). Bilateral nephrectomy and haemodialysis for the treatment of severe loin pain haematuria syndrome. Nephrol Dial Transplant, 9, 1640–41. Goroszeniuk, T., Khan, R., and Kothari, S. (2009). Lumbar sympathetic chain neuromodulation with implanted electrodes for long-term pain relief in loin pain haematuria syndrome. Neuromodulation, 12, 284–291. Greenwell, T. J., Peters, J. L., Neild, G. H., et al. (2004). The outcome of renal denervation for managing loin pain haematuria syndrome. BJU Int, 93(6), 818–21. Harney, J., Rodgers, E., Campbell, E., et al. (1994). Loin pain-hematuria syndrome: how effective is renal autotransplantation in its treatment? Urology, 44, 493–6.

471

472

Section 3  

the patient with glomerular disease

Hebert, L. A., Betts, J. A., Sedmak, D. D., et al. (1996). Loin pain-hematuria syndrome associated with thin glomerular basement membrane disease and hemorrhage into renal tubules. Kidney Int, 49(1), 168–73. Lall, R., Mailis, A., and Rapoport, A. (1997). Hematuria-loin pain syndrome: its existence as a discrete clinicopathological entity cannot be supported. Clin J Pain, 13, 171–7. Little, P. J., Sloper, J. S., and de Wardener, H. E. (1967). A syndrome of loin pain and haematuria associated with disease of peripheral renal arteries. QJM, 36, 253–9. Lucas, P. A., Leaker, B. R., Murphy, M., et al. (1995). Loin pain and haematuria syndrome: a somatoform disorder. QJM, 88, 703–9. Naish, P. F., Aber, G. M., and Boyd, W. N. (1975). C3 deposition in renal arterioles in the loin pain and haematuria syndrome. Br Med J, 3, 746. Parnham, A. P., Low, A., Finch, P., et al. (1996). Recurrent graft pain following renal autotransplantation for loin pain hematuria syndrome. Br J Urol, 78, 25–8. Praga, M., Martinez, M. A., Andres, A., et al. (1998). Association of thin basement membrane nephropathy with hypercalciuria, hyperuricosuria and nephrolithiasis. Kidney Int, 54, 915–20.

Prager, J. P., DeSalles, A., Wilkinson, A., et al. (1995). Pain hematuria syndrome: pain relief with intrathecal morphine. Am J Kidney Dis, 25, 629–31. Sheil, A. G., Chui, A. K., Verran, D. J., et al. (1998). Evaluation of the loin pain/hematuria syndrome treated by renal autotransplantation or radical renal neurectomy. Am J Kidney Dis, 32, 215–20. Smith, H.S. and Bajwa, Z. H. (2012). Loin pain hematuria syndrome-visceral or neuropathic pain syndrome? Clin J Pain, 28, 646–51. Spetie, D. N., Nadasdy, T., Nadasdy, G., et al. (2006). Proposed pathogenesis of idiopathic loin pain-hematuria syndrome. Am J Kidney Dis, 47, 419–27. Spitz, A., Huffman, J. L., and Mendez, R. (1997). Autotransplantation as an effective therapy for the loin pain-hematuria syndrome: case reports and a review of the literature. J Urol, 157(5), 1554–9. Uzoh, C. C., Kumar, V., and Timoney, A. G. (2009). The use of capsaicin in loin pain-haematuria syndrome. BJU Int, 103, 236–9. Weisberg, L. S., Bloom, P. B., Simmons, R. L. et al. (1993). Loin pain hematuria syndrome. Am J Nephrol, 13, 229–37.

CHAPTER 48

Nutcracker syndrome and phenomenon John Neary and Neil Turner Introduction Obstruction of venous return by pressure of the superior mesenteric artery on the left renal vein was noted as a possible cause of varicocele in 1950 (El-Sadr and Mina, 1950). The term nutcracker was introduced over 20 years later. The phenomenon and syndrome are well reviewed by Kurklinsky and Rooke (2010). A distinction between the anatomical nutcracker phenomenon (NCP) and nutcracker syndrome, in which specific symptoms are attributed to NCP, is useful (Kurklinsky and Rooke, 2010). See also Chapters 46, 47, and 51.

Anatomy and incidence of nutcracker phenomenon Figs 48.1 and 48.2 illustrate the NCP. Gonadal and lumbar veins may be distended in association with the narrowing and varicocele and vulval varicosities have been attributed to NCP. Cross-sectional imaging techniques show narrowing and dilatation of the left renal vein. The phenomenon can also be demonstrated by Doppler ultrasound, which can show haemodynamic effects of narrowing. However, the findings on Doppler ultrasound are prone to variation depending on posture, patient anatomy, and operator. Detailed studies are more likely to be performed in patients with symptoms, which complicates assessment of unblinded studies performed using this technique. It is difficult to find neutral studies that look at the incidence of NCP in the absence of symptoms. However, in a useful study of 99 consecutive adult potential renal transplant donors by computed tomography angiography (Grimm et al., 2013), there was an average 35% reduction in diameter of the left renal vein as it crossed the aorta; 27% of patients had substantial (> 50%) compression of the left renal vein including 8% with > 70% narrowing. Dilated gonadal or lumbar veins were also common (16% or 24% respectively) and this finding was unrelated to the degree of renal vein compression. These were healthy, symptom-free individuals but five (5%) had dipstick haematuria; this was not associated with renal vein compression. There was no significant age or sex influence on the degree of NCP in this selected group. The phenomenon has mostly been described in young adults and older children, but there are reports at all ages. It has been suggested that low body mass index (BMI) increases the likelihood of NCP as reduced abdominal fat reduces the angle between superior

mesenteric artery and aorta. However, there is not a good correlation between BMI and abdominal fat, and there is not enough evidence to prove or disprove this hypothesis (discussed by Park and Shin, 2013).

Haematuria Haematuria is the most commonly associated renal feature of NCP. Episodic macroscopic haematuria is the symptom that most commonly leads to the imaging studies in which NCP is identified. Occasional patients have been described in whom bleeding has necessitated blood transfusion, but this is rare. Cystoscopy may reveal left-sided bleeding (illustrated by Matsubara et al., 2013). There are numerous case reports in which surgical intervention or stenting of the left renal vein has been followed by cessation of bleeding episodes (e.g. Vince et  al., 2011; Chen et  al., 2012). However, spontaneous improvement may also occur (e.g. Matsubara et al., 2013). The precise location of bleeding, other than that it seems to come from the kidney itself, is not described. The red cells do not have characteristics of glomerular haematuria. It is interesting that in reported cases autotransplantation or stenting do not seem to halt bleeding immediately but over a period of days to weeks. It is not clear that persistent microscopic haematuria can be attributed to NCP.

Pain Few case reports of NCP with episodic haematuria report pain as a prominent feature, although some do (e.g. Vince et al., 2011). A variety of pain syndromes have been attributed to NCP, with symptoms in the abdomen, flank (presumably left flank), and sometimes buttock. Overt haemorrhage may lead to left renal colic (clot colic). Pelvic pain syndromes have sometimes been attributed to NCP but the case for this is usually weak. It is an unlikely cause of classic loin pain haematuria syndrome as this is more typically associated with glomerular bleeding (see Chapter 47).

Postural proteinuria While the relationship with episodic visible haematuria seems firmly based, a possible relationship between NCP and proteinuria remains uncertain. It was suggested as a cause of postural proteinuria (see Chapter 51) in 1958, and a number of unblinded studies using Doppler ultrasound have found an apparently persuasive

Section 3  

the patient with glomerular disease

AA

SMA

SMA

LRV LRV RH

AA

Fig. 48.1  Computed tomography angiography of a 29-year-old woman with a 4-year history of intermittent visible haematuria. Cystoscopy showed bleeding from the left ureter. There is compression of the left renal vein between the abdominal aorta and the superior mesenteric artery. An endovascular stent was implanted and the haematuria ceased two weeks later. From Song et al. (2013).

Left renal vein

Left kidney

Aorta

Superior mesenteric artery

Vena cava

Aorta

Left kidney

Superior mesenteric artery

Vena cava

474

Left renal vein

Pre-aortic left renal vein

Retro-aortic left renal vein

Meso-aortic (= anterior) entrapment

Retro-aortic (= posterior) entrapment

Fig. 48.2  Anatomy of the left renal vein, showing normal anatomy (left) and the mechanism by which the renal vein can be entrapped: and (right) a variation seen in approximately 3% of the population that can be associated with a posterior entrapment. From Mazzoni et al. (2011).

relationship (Mazzoni et al., 2011). Problems with the interpretation of these studies are discussed in Chapter 51. As outcomes of orthostatic proteinuria appear generally good, and there are usually no symptoms, intervention to alter anatomy cannot be easily justified for this reason. One study found that angiotensin-converting enzyme inhibitors lowered protein excretion in postural proteinuria (Ha and Lee, 2006).

Management Reported treatments have included nephrectomy, autotransplantation, vascular bypass procedures, and more recently intravascular or extravascular stenting (Kurlinsky and Rooke, 2010; Chen et al., 2012). Almost always these treatments are potentially more harmful than the condition itself, and as described above, the relationship of NCP to symptoms is often uncertain. Conservative management is therefore appropriate for the great majority of patients, as is remaining open to an alternative diagnosis. Intervention might be indicated for severe recurrent haematuria, but haematuria may also remit spontaneously (e.g. Matsubara et al., 2013). Interestingly, several reports describe episodes of haemorrhage as reducing not immediately but over days to weeks after stenting. Spontaneous recovery may also occur. Intervention can rarely be justified where it is suspected that nutcracker syndrome might explain proteinuria.

References El-Sadr, A. R. and Mina, E. (1950). Anatomical and surgical aspects in the operative management of varicocele. Urol Cutaneous Rev, 54, 257–62. Ha, T. S. and Lee, E. J. (2006). ACE inhibition can improve orthostatic proteinuria associated with nutcracker syndrome. Pediatr Nephrol, 21, 1765–8. Chen, S., Zhang, H., Tian, L., et al. (2012). Endovascular management of nutracker syndrome after migration of a laparoscopically placed extravascular stent. Am J Kidney Dis, 50, 322–6. Grimm, L. J., Engstrom, B. I., Nelson, R. C., et al. (2013). Incidental detection of nutcracker phenomenon on multidetector CT in an asymptomatic population: prevalence and associated findings. J Comput Assist Tomogr, 37, 415–18.

Chapter 48 

Ha, T. S. and Lee, E. J. (2006). ACE inhibition can improve orthostatic proteinuria associated with nutcracker syndrome. Pediatr Nephrol, 21, 1765–8. Kurklinsky, A. K. and Rooke, T. W (2010). Nutcracker phenomenon and nutcracker syndrome. Mayo Clin Proc, 85, 552–9. Matsubara, T., Ogawa, O., and Yanagita, M. (2013). Physical finding of nutcracker phenomenon. Kidney Int, 83, 335. Mazzoni, M. B., Kottanatu, L. K., Simonetti, G. D., et al. (2011). Renal vein obstruction and orthostatic proteinuria: a review. Nephrol Dial Transplant, 26, 562–5.

nutcracker syndrome and phenomenon

Park, S. J. and Shin, J. I. (2013). Low body mass index in nutcracker phenomenon: an underrecognized condition. Kidney Int, 84, 1287. Song, Y., Wu, J. Y., and Chen, J. H. (2013). Haematuria and the nutcracker syndrome. QJM, 106, 879–80. Vince, H. B., Tomson, C. R., Loveday, E. J., et al. (2011). Nutcracker syndrome presenting as loin pain haematuria syndrome. NDT Plus, 4, 418–20.

475

CHAPTER 49

Exercise-related pseudonephritis Neil Turner Introduction The effects of exercise on urine sediment were recognized well over a century ago but the phenomena still resist complete explanation. The findings can appear extreme and the term pseudonephritis coined by Gardner (1956) is accurate.

Exercise-related haematuria and haemoglobinuria March haematuria, first documented in 1881 (reviewed by Gilligan and Blumgart, 1941), has from the first reports probably been a mix of exercise-induced haemoglobinuria, exercise-related myoglobinuria, and exercise-related haematuria. In the earliest literature, many cases are described in which no blood cells could be seen despite positive tests for globin; some cases are more suggestive of recurrent rhabdomyolysis. Others are more suggestive of immunoglobulin A nephropathy in young adults, while yet others hint at a urological origin in bladder or urethra. Tobal et al. (2008) described an informative series of cases from Uruguay related to prolonged, intense, hand drumming (candombe

Fig. 49.1  ‘The Start at Iffley’ from Tom Brown at Oxford, 1861.

drumming). Apparently rust-coloured urine following carnival is widely recognized. Over 2000 drummers participate; over 6 years, five patients were admitted with acute kidney injury (AKI) following these events, with two of these requiring dialysis. In a series of 45 individuals studied after drumming, rust-coloured urine was noted in 20% and urinary abnormalities on analysis in 64%. All of those with rust-coloured urine had evidence of intravascular haemolysis. Only one patient in the series had detectable myoglobinuria. Of the 20 with rust-coloured urine, six had urine samples collected 48–72 hours after drumming and all had returned to normal. No cases of even mild AKI were seen in their study group, so this must be a rare event. The most common scenario in which these changes are seen now seems to be long-distance running, after which some see visible haematuria and the prevalence of abnormal dipstick tests may be 25% (Siegel et al., 1979; Kallmeyer and Miller, 1993). It is usually asymptomatic and resolves within 72 hours. There is again evidence that this can be related to red cell trauma and lysis (Telford et  al., 2003). However, phase contrast microscopy is reported to show increased excretion of red blood cells which not only have the characteristics of glomerular red cells, but are also commonly

Chapter 49 

accompanied by red cell casts (Barach, 1910; Gardner, 1956; Fassett et al., 1982). It is not at all clear what the explanation for glomerular bleeding is. Others find evidence for a non-glomerular origin in the bladder or urethra (Kallmeyer and Miller, 1993; and others discussed in Kincaid-Smith, 1982), and it is not satisfactorily resolved which origin is more common.

Exercise-related proteinuria Proteinuria is less visually obvious after strenuous exercise but has an equally distinguished history, first described by Leube in 1878 and lucidly described by Collier in 1907. Collier described proteinuria appearing in the urine shortly after exercise and usually gone again the following day, only to recur again following further exercise. He found albumin in the post-exercise urine of 57% of 156 Oxford student rowers in intensive training for college races, with higher proportions in the more successful boats (Fig. 49.1). Relating these findings to prior observations on the longevity of boat race participants he concluded that it was a benign phenomenon, going against then prevalent views on the implications of proteinuria. Others (reviewed by Poortmans, 1985) have since described in detail an early post-exercise peak in protein excretion which drops over hours, the magnitude depending on the intensity of the exercise. The protein is mostly albumin. The mechanism of exercise-induced proteinuria is unknown. Most of the individuals in these studies did not have haematuria or smoky urine; the proteinuria effect seems to be encountered after

exercise-related pseudonephritis

lesser durations of exercise than haematuria, and in types of exertion less likely to involve trauma to red cells.

References Barach, J. H. (1910). Physiological and pathological effects of severe exertion (the Marathon Race) on the circulatory and renal systems. Arch Intern Med, 5, 382–405. Collier, W. (1907). Functional albuminuria in athletes. Br Med J, i, 4–6. Fassett, R. G., Owen, J. E., Fairley, J., et al. (1982). Urinary red-cell morphology during exercise. Br Med J, 285, 1455–7. Gardner, K. D. (1956). ‘Athletic pseudonephritis’—alteration of the urine sediment by athletic competition. JAMA, 161, 1613–17. Gilligan, D. R. and Blumgart, H. L (1941). March hemoglobinuria: studies of the clinical characteristics, blood metabolism and mechanism: with observations on three new cases, and review of literature. Medicine, 20, 341–95. Kallmeyer, J. C. and Miller, N. M. (1993). Urinary changes in ultra long-distance marathon runners. Nephron, 64, 119–21. Kincaid-Smith, P. (1982). Haematuria and exercise-related haematuria. Br Med J, 285, 1595–7. Poortmans, J. R. (1985). Postexercise proteinuria in humans: facts and mechanism. JAMA, 253, 236–40. Siegel, A. J., Hennekens, C. H., Solomon, H. S., and Van Boeckel, B. (1979). Exercise-related hematuria: findings in a group of marathon runners. JAMA, 241, 391–2. Telford, R. D., Sly, G. J., Hahn, A. G., et al. (2003). Footstrike is the major cause of hemolysis during running. J Appl Physiol, 94, 38–42. Tobal, D., Olascoaga, A., Moreira, G., et al. (2008). Rust urine after intense hand drumming is caused by extracorpuscular hemolysis. Clin J Am Soc Nephrol, 3, 1022–7. Von Leube, W. (1878). Über ausscheidung von Eiweiss in harn des gesunden menschen. Virchows Arch Pathol Anat Physiol, 72, 145–7.

477

CHAPTER 50

Proteinuria Neil Turner and Stewart Cameron Introduction Richard Bright’s outstanding clinical and autopsy observations in 1827 demonstrated the association between proteinuria and fatal kidney disease, laying the ground for the specialty of nephrology (see Chapter 42). He built his observations on many going before him, but his insights pushed proteinuria into the forefront of medical thinking in the 1800s. Measuring proteinuria became a regular part of clinical assessment. By the end of the nineteenth century the implications of proteinuria as a risk factor were being considered by insurance companies as well as by doctors (Collier, 1907; Barringer, 1912; Turner, 2013).

The origin of proteinuria Unlike haematuria, which can come from any level in the urinary tract, proteinuria almost always originates in the kidney. Pathological mechanisms of proteinuria mainly affect the glomerulus, and specifically the podocyte, as outlined in Chapter 45. The glomerulus filters only very small amounts of serum albumin, but proteins smaller than 30 kDa filter almost freely. These lower-molecular-weight proteins and the small amounts of albumin that are filtered are mostly reabsorbed in the proximal tubule (Fig. 50.1). Although there has been some debate about the permeability of the glomerular barrier to albumin (discussed further in Chapter 137), the weight of evidence confirms that little albumin is filtered in health (Haraldsson and Tanner, 2014). Internalization of filtered proteins in the proximal tubule is a process that involves the cell surface receptors megalin and cubulin (Amsellem et al., 2010; Haraldsson, 2010). It is then degraded in lysosomes, a process that can be beautifully visualized in vivo (Slattery et  al., 2008). Ovunc et  al. (2011) described an interesting family with autosomal inherited cubulin mutations causing intermittent proteinuria up to 2 g/day. Mutations in cubulin have also been associated with proteinuria and megaloblastic anaemia because of failure to internalize the intrinsic factor–vitamin B12 complex. Polymorphisms in tubulin may account for variance in the albumin:creatinine ratio (ACR) in the general population (Böger et al., 2011). Cubulin interacts with megalin to function. Mice with deletion of megalin survive poorly, but one abnormality is severe rickets from failure of tubular cells to absorb filtered vitamin D in complex with its binding protein. In Dent disease (see Chapter  41), a failure of lysosomal function appears to be responsible for a widespread failure of tubular reabsorption.

What are the proteins in normal urine? Albumin is the most abundant protein in normal urine, as in serum. It makes up a little less than half of average daily urinary protein, about 30 mg. Proteins in normal urine fall into three groups: filtered proteins that have not been reabsorbed such as albumin; proteins actively secreted into urine; and proteins lost from cells in the nephron and urinary tract, particularly via extruded exosomes, small cell-derived vesicles that can be identified in all biological fluids. Many of these proteins are present at very low level and analysis is a challenging biochemical task. The most abundant components of normal urine are serum proteins (albumin, α1-antitrypsin, zinc α2-glycoprotein, α1-microglobulin), but there is also a large group of kidney secreted and structural proteins, of which uromodulin (Tamm–Horsfall protein) is a leading representative. Uromodulin is secreted into the thick ascending limb of the loop of Henle and early distal tubule, and is a major component of urinary casts (Rustecki et al., 1971; Hoyer and Sieler, 1979; Scherberich, 1990). Its gene is implicated in familial juvenile hyperuricaemic nephropathy and in medullary cystic kidney disease (see Chapter 318). By two-dimensional gel electrophoresis the complexity of urinary proteins can be seen, but a typical review was able to confirm identities of only 275 of 1118 spots representing 82 proteins. Many of the uncharacterized species were low-abundance, low-molecular-weight proteins (Candiano et  al., 2010). The proportion of serum proteins increases dramatically in nephrotic syndrome. Modern mass spectrometry pushes the boundaries further, but probably with different selection bias; 868 identified proteins are listed in one study (Liu et al., 2012).

Biomarkers: albumin remains the one to beat Many groups are working to find specific molecules that might be informative about particular disease processes, but few candidates have survived tests of clinical utility. Albumin remains as good a predictor of most outcomes as many putative biomarkers—a fascinating and important observation.

Pathological proteinuria Proteinuria in excess of the usual modest limit can come about by: 1. Glomerular proteinuria: the glomerular filter becomes more permeable to proteins of large molecular size. This is common. 2. Tubular proteinuria: the proximal tubule is damaged so that proteins that are normally reabsorbed (principally of low molecular

Chapter 50 

Proteinuria

Glomerular filtration barrier

> 60kD

< 20kD Tubular metabolism (saturable) Albumin β2 microglobulin lg light chain

67kD 12kD 25kD

Fig. 50.1  Handling of protein in the nephron. Proteins the size of albumin (67 kD) are largely excluded at the glomerular filtration barrier. Proteins smaller than approximately 20 kD pass freely through it into the filtrate. Proteins between these sizes are filtered progressively less well as size increases. Most filtered proteins are internalized into proximal tubular cells by a process involving the cell surface receptors cubulin and megalin, and degraded in lysosomes. Albuminuria is the defining characteristic of increased glomerular permeability (glomerular proteinuria). β2-microglobulin is an example of a freely filtered protein that is found in the urine in increased quantities if the internalization process into proximal tubular cells is disrupted (tubular proteinuria). If immunoglobulin light chains are overproduced they filter fairly freely (overflow proteinuria) and some light chains may be nephrotoxic through aggregation in the tubular lumen or in proximal tubular cells after internalization.

weight) pass into the urine. This usually occurs as part of the Fanconi syndrome of multiple proximal tubular dysfunction (see Chapter 41). This is much less common. 3. Overflow proteinuria: an increase in the plasma concentration of a filterable protein, so that the amount filtered exceeds the reabsorptive capacity of the proximal tubule. Immunoglobulin (Ig) light chains or fragments, and lysozyme in monomyelocytic leukaemia, are the only clinical examples. Most significant proteinuria is glomerular.

Discerning the origin of proteinuria Glomerular origin of proteinuria can be inferred by the quantity, ratio of albumin to other components, or by specific assays for tubular proteins.

Quantity Tubular proteinuria rarely exceeds 1.5–2 g per 24 hours (protein:creatinine ratio (PCR) 150–200 mg/mol) and is usually less than this. Therefore (almost) all patients with higher excretion rates, and all patients with nephrotic syndrome (see Chapter 52), have a glomerular leak, primarily.

Ratio The simplest test is to compare the results of a specific albumin assay test with a total protein result; ACR versus PCR (see ‘Tests’). A ratio of > 0.4 suggests glomerular proteinuria. A ratio of < 0.4 suggests non-albumin proteinuria (Methven et  al., 2012; Smith et al., 2012). A historic way to show tubular or ‘overflow’ protein in urine was to compare the sulphosalicylic acid precipitation test (see ‘Tests’), which precipitates many proteins, with the results of dipstick tests, which are more sensitive to albumin.

proteinuria

Specific assays for tubular proteins Specific assays enable the identification of low-molecular-weight proteins that would normally be internalized and destroyed, but pass through the nephron if tubules are damaged. β2-microglobulin (molecular weight (MW) 12 kDa) is a sensitive indicator of tubular damage, but the protein is unstable in urine of normal pH (5–6.5) and even alkalinization of urine to pH 7 or above immediately on voiding may not stop degradation. Normally, < 0.4 micrograms/L of β2-microglobulin is present in urine. Also useful, but rarely used, are assays for retinol-binding protein (21 kDa) (Tomlinson et al., 1997), α1-microglobulin (30 kDa), and lysozyme (15 kDa, normal excretion < 1 mg/mmol creatinine) (Barratt,  1983). For further details see Chapter 7.

Tests for overflow or exogenous proteins Immunofixation for Ig light chains (Bence Jones protein) remains an important test. Analysis of serum and/or urine for free light chains is more sensitive and also useful in monitoring light chain dyscrasias (see Chapter 150) and their response to treatment. Sodium dodecyl sulphate polyacrylamide gel electrophoresis (SDS-PAGE) gel electrophoresis is not a routine technique but it is very useful for investigating unusual proteinuria. It can clearly show the low-molecular-weight proteins of tubular proteinuria overflow proteins, as well as shining light on rare examples where exogenous proteins have been used to mimic proteinuria. Very occasionally, proteinuria may be factitious (Tojo et al., 1990), egg albumin or other protein being added to the urine. Even more rarely parents may add protein to their child’s urine, as a variety of the Munchausen syndrome by proxy (Meadow, 1977).

Epidemiology of proteinuria Incidence The method used to screen for proteinuria affects both the proportion found to have it, and the strength of the association with outcomes. It is the same for haematuria (see Chapter 46), but proteinuria is a much stronger predictor of poor long-term outcomes, and risk is graded according to severity of proteinuria. This reduces the problem of false classification by a positive/negative test. Table 50.1 summarizes results from a number of surveys internationally. They show a high prevalence of simple positivity with varying cut-offs. Some report detailed results of quantitation in very large numbers, for example, the Okinawa studies (Iseki et al., 1996, 2003).

Prognostic significance of proteinuria in population studies At the first meeting of the Assurance Medical Society in 1894, the importance of measuring albuminuria in assessing insurance risk was discussed. In 1912, Barringer described increased mortality in 396 New York men who had been found to have proteinuria, but otherwise normal health, 10 years previously (Barringer, 1912). He also cited insurance company data from larger numbers of patients that suggested that mortality was more than doubled in individuals with a small amount of albuminuria and urinary casts. Chapter 97 summarizes the risks measured in population studies. It is interesting that the implications of proteinuria in population studies (cardiovascular outcomes more likely than renal) are quantitatively different from the outcomes in patients known

479

480

Section 3  

the patient with glomerular disease

Table 50.1  Prevalence of dipstick or sulphosalicylic acid-positive proteinuria in population studies Study and reference

Prevalence (%)

Schoolchildren 10 studies from 1950 (Vehaskari and Rapola, 1982)

0.6–6.4 (weighted mean 2.2)

1000–50,000 children Lin et al. (2000) dipstick confirmed by SSA

5

Young adults—armed forces recruits McLean (1919)

5.6

Murphy (1944)

3.0

Sinniah (1977)

0.9

The 24-hour excretion of protein is 80 ± 24 mg (mean ± standard deviation (SD)) in normal individuals, so that 128 mg/24 hours (Peterson et al., 1969; Berggård, 1970) represents the mean ± 2SD, a rate of excretion of 103 micrograms/minute total protein. This represents < 11 mg/mmol creatinine (PCR) in spot urine samples (0.1 mg/mg). The normal range is usually taken to be up to 150 mg/day, for PCR up to 15 mg/mmol, significantly above this level. This avoids misclassification, but means that some levels within the normal range may in fact be elevated. Normal levels and different ways of measuring and expressing proteinuria are summarized in Table 50.2. Further laboratory details of tests are shown in Chapter 7.

Dipsticks Total protein

College students, both sexes Diehl and McKinney (1944)

5.3

Burden (1933)

26.0

Lee (1920)

5.0

Adults, both sexes Blatherwick (1942) (> 20 mg/dL)

1.7

Alwall (1973)

1.7

Yudkin (1988) (microalbuminuria)

9.4

Bigazzi (1992) (microalbuminuria > 20–50 mg/L)

10.0

Ritz et al. (1994) (> 20 mg/L)a

3.0

Iseki et al. (1996, 2003) dipstick tests:

Early tests for protein based on heating and/or acidifying urine were superseded for most clinical purposes by dry stick tests (dipsticks) in the 1950s (Turner, 2013). In these, protein detection is based on colour reactivity with tetrabromphenol blue, which is yellow in the absence of protein at pH 3 and green in its presence. Very alkaline urine will interfere with this reaction, and other substances may sometimes interfere with different measurement methods. The lower limit of detection is about 150–300 mg/L. Depending on urine concentration therefore, normal levels of urinary protein may give weakly positive dipstick tests. This reagent is more sensitive to albumin than to some other proteins, so that most light chains (Bence Jones protein) and mucoproteins are less sensitively detected (Lamb and Price, 2011).

Albuminuria

+ or higher

5.1

++ or higher

1.3

+++ or higher

0.3

Adults > 60 years Sawyer (1988)

6.6

Casiglia (1993)

10.0

Adults > 80 years Casiglia (1993)

Testing for proteinuria

16.1

For details of pre-1994 references see Ritz et al. (1994) and Vehaskari and Rapola (1982) SSA = sulphosalicylic acid turbidimetry test, see Table 50.3.

to have renal disease, where renal outcomes are more likely. See, for instance, long-term outcomes of the Modification of Diet in Renal Disease (MDRD) study in which the great majority of participants developed end-stage renal failure (ESRF) (see Chapter  99). The nature of the association of proteinuria with renal outcomes is discussed further below. In the Okinawa study of adult patients screened in 1983, and reviewed by inspecting the end-stage renal disease register at the end of 2000, the risk of ESRF was 0.2% overall, 0.4% if the screening result was +/1, 1.4% if +, 7% if ++, and 15% if it was +++ (Iseki et al., 2003). The risk for patients with haematuria as well as proteinuria was approximately double that for patients with proteinuria (all levels) alone (Iseki et al., 1996).

Dipsticks specific for albumin, based on more specific chemical or on immunodetection methods, are available, but are more

Table 50.2  Approximately equivalent levels of proteinuria estimated by PCR/ACR. For accurate conversions between units: to convert mg/ mmol to mg/g, multiply by 8.8. To convert mg/g to mg/mmol, multiply by 0.113 24-hour excretiona

PCR mg/mmolb (mg/g)

ACR mg/mmolc (mg/g)

Paediatric equivalentd

150 mg

> 15 (150)

> 2.5/3.5 (M/F)

> 4 mg/m2/h;

0.5 g

50 (500)

30 (300)

100 mg/m2/day

1g

100 (1000)

70 (700)

PCR < 20 mg/ mmol

3g

300 (3000)

250 (2500)

> 40 mg/m2/h; 1 g/m2/day

a Normal 24-hour protein excretion for adults varies in different studies and 150 mg is

unequivocally elevated, for example, the mean + 2SD was 128 mg according to Peterson et al. (1969) and Berggård (1970). b Protein:creatinine ratio. c Albumin:creatinine ratio. d First-morning samples are particularly favoured in children as this helps to distinguish

postural proteinuria. The upper limit of normal 24-hour excretion for children defines a large proportion as having postural proteinuria, which appears usually to be associated with normal health (see Chapter 51) (Hladunewich and Schaefer, 2011).

Chapter 50 

expensive. For detection of microalbuminuria an ACR or timed excretion must be tested.

PCR and ACR Measuring protein excretion rates by 24-hour collection is inconvenient and error prone. The correlation between 24-hour protein or albumin excretion and the protein or albumin to creatinine ratio (PCR or ACR) is reported to be high (although the 95% confidence limits are wide), and this has led to ‘spot’ urines being recommended for routine practice in preference to 24-hour collections (Chitalia et al., 2001; Johnson et al., 2012; National Institute for Health and Care Excellence, 2014). Individual test results may vary more between measurements than 24-hour collection results. However, most feel that because ACR or PCR can be checked more easily and frequently, and because 24-hour collections are notoriously difficult to complete, some variability in results is an acceptable price. Clearly major treatment decisions should not be made on moderate changes seen in only one test. Most nephrologists use spot tests most of the time but there continues to be some debate on the question (Ginsberg et al., 1983; Guy et al., 2009; Hebert et al., 2009; Naresh et al., 2013). Point-of-care devices are available that measure both creatinine and either protein or albumin on a single stick. These can give rapid results but are costly for routine use if compared to bulk analysis in a laboratory.

Table 50.3  Methods for measuring urine protein. The sulphosalicylic acid test (0.5 mL of 3% solution to 0.5 mL urine, can be graded manually or read in a densitometer) has been largely superseded but was used in some population studies as the primary method, or to verify dipstick results Method

Description

Detection limit

Comments

Kjeldahl

Remove non-protein nitrogen, digest protein, measure protein nitrogen

10–20 mg/L

Reference and research method

Biuret

Copper reagent, 50 mg/L measures peptide bonds

Requires precipitation of proteins, used for 24-hour measurement in some laboratories

Turbidimetric (trichloracetic acid, sulphosalicylic acid)

Addition of 50–100 mg/L trichloracetic or sulphosalicylic acids alters colloid properties and produces turbidity to be read in densitometer. Benzethomecin also used

Imprecise, different readings for albumin and globulin

Dye-binding

Indicator changes 50–100 mg/L colour in presence of protein (e.g. Coomassie brilliant blue)

Different proteins bind differently; several different dyes in use; used in many laboratories for 24-hour excretion

Nephelometric

Specific antialbumin antibody used

Measures albumin excretion not total protein. Does not detect globulins or any other proteins

Stick tests

Impregnated with 100 mg/L indicator dye which changes colour in the presence of protein

Reacts poorly with globulins. Usual clinical screening test

Lab and other methods including selectivity of proteinuria Laboratory and other methods and their features are shown in Table 50.3. Historically, tests for selectivity of proteinuria were sometimes used to provide evidence that patients with nephrotic syndrome were likely to have minimal change disease and that therefore a trial of steroid therapy would be reasonable without a biopsy. The test has fallen out of use, but evidence for it, and that selectivity may predict remission in membranous nephropathy, is discussed by D’Amico and Bazzi (2003). The test compared the clearance of a larger molecule with that of a smaller: IgG, IgM, or α2-macroglobulin, against that of albumin or transferrin. A clearance of IgG > 20% (0.20) of transferrin or albumin represents ‘non-selective’ proteinuria; < 0.10 is ‘highly selective’ and suggests that a minimal change lesion may be present. The range between 0.10 and 0.20 is of little discriminatory value.

Evaluation of proteinuria Many entirely healthy individuals have protein at ‘trace’ or ‘+’ levels, as described above (and see Table 50.1). Two to five per cent of children, 5% of young adults, and up to 16% of the elderly will show proteinuria on testing of a single sample.

Confirm and quantitate Note whether haematuria is also present. Exclude infection by dipstick for nitrite, leucocytes, and by microscopy and culture. As for non-visible haematuria (see Chapter 46), simply repeating the test makes these figures fall, but quantitation by PCR can distinguish normal from abnormal levels. Chyluria is a rare cause of non-renal proteinuria caused by lymphatic disease. It is usually obvious because urine is cloudy or milky. Leg swelling may be caused by lymphatic disease rather than

proteinuria

proteinuria. Urine contains large numbers of lymphocytes (Cheng et al., 2006).

Exclude transient proteinuria There are several causes of transient proteinuria that usually do not convey the same connotations of renal disease and risk as sustained proteinuria: ◆

Postural proteinuria (see Chapter 51), the most common diagnosis in children and young people. Test a first-in-morning sample.

481

482

Section 3  

the patient with glomerular disease

◆ Post-exercise

proteinuria (see Chapter 49). Occurs within hours of exercise, magnitude depends on exertion.

◆ Transient ◆ Heart

proteinuria during fever (Marks et al., 1970).

failure (Albright et al., 1983).

Factitious and pseudoproteinuria Factitious proteinuria was mentioned above. Electrophoresis of the urine can demonstrate foreign proteins such as ovalbumin. Pseudoproteinuria may be noted in patients receiving infusions of gelatin (MW 30 kDa)-based volume expanders such as ‘gelofusine’, which are readily filtered at the glomerulus (Jones et al., 1999), if the molybdate pyrogallol or possibly other dye-binding methods are used. The presence of blood interferes with protein assays variably, but usually only at quite high levels of blood content.

Assess the degree of proteinuria Nephrotic range proteinuria Proteinuria at nephrotic levels (Table 50.2), > 3 g/day, 40 mg/m2/ day in children, PCR > 300 mg/mmol or 3000 mg/g, almost always comes from glomerular disease. Unless the cause is readily apparent, investigation including renal biopsy is usually indicated. Diagnosis and management of nephrotic syndrome are considered in Chapter 52. For patients with nephrotic range proteinuria who fall short of nephrotic syndrome, many will require the same investigation and management. There may be exceptions when comorbidity makes it unlikely that disease-specific therapies could be tolerated.

Proteinuria at levels of 1–3 g/day A PCR of 100 mg/mmol (1000 mg/g), approximately equivalent to 1 g proteinuria per day, is suggested in the United Kingdom as a suitable threshold for referral to a nephrology service, as above this level diagnosable renal disease becomes more likely. This was also the level selected in the MDRD study to define patients with significant proteinuria, who benefited most from blood pressure control, and for whom renal outcomes were substantially worse. Of course the risk of proteinuria is graded, there is no threshold at 1 g. Because of the duration at risk, younger people with lower levels of proteinuria may be at substantial lifetime risk of ESRF, justifying review at lower levels. Conversely, elderly patients with comorbid conditions may be unlikely to have their management altered significantly by nephrological review at this level of proteinuria. If significant haematuria coexists with proteinuria, renal risk is substantially higher. Nephrological assessment is justified at lower levels of proteinuria. Renal impairment, or even more important, worsening GFR, are indications for further assessment. The diseases causing proteinuria are increasingly likely to be recognizable glomerular conditions as the levels of proteinuria rise. Towards the lower end of this range diagnoses are less likely to alter management, except possibly in relatively young patients. At levels of proteinuria below 2–2.5g consider tubular proteinuria. If a glomerular cause is not apparent measure both ACR and PCR; a ratio of albumin:protein of < 0.4 suggests non-glomerular proteinuria. In all patients consider the possibility of plasma cell dyscrasias. Estimation of urinary light chains by immunofixation is the most widely used test, and though not as sensitive as assays for free Ig light chains it will certainly be positive if the explanation for proteinuria at this level.

Protein at levels of 0.3–1.0 g/day The chances of identifying a specific glomerular disease become substantially lower unless there are pointers or known risk factors (e.g. genetic, infective, and systemic disease), but considerations are the same as above for 1–3 g/day. The coexistence of haematuria greatly increases the probability of identifiable glomerular disease. Lower thresholds should be considered relevant in young patients because they face longer at risk.

Microalbuminuria Microalbuminuria is presumed to be a consequence of derangement of function of the glomerular filtration barrier, although this has not been formally proven in all circumstances. This would explain the early appearance of microalbuminuria in glomerulopathies, as first demonstrated in diabetes (Mogensen and Christensen, 1984; Viberti, 1988)  (see Chapter  149). However the origin of microalbuminuria in many patients with cardiovascular disease (see Chapters 97 and 98) has not been satisfactorily explained, and in theory it could alternatively be caused by reduced capacity of proximal tubules to reabsorb tubular proteins from filtrate. In young patients without comorbid conditions, microalbuminuria is likely to reflect early glomerular disease. It can be a useful test in patients with isolated dipstick haematuria. In older patients it is less specific as a test for primary renal disease.

Assess other renal risk factors These include family history, drug history, hypertension, and low or (even more important) reducing estimated glomerular filtration rate.

Assess and manage cardiovascular risk Patients with proteinuria are at significantly increased risk of cardiovascular disease and death (see Chapter 98). Like the renal risk, this is graded with increasing severity of proteinuria. However, in the general population without an established renal diagnosis, cardiovascular endpoints are more likely than renal (discussed in Chapter 99).

Proteinuria and progression of renal diseases The association between proteinuria and population outcomes was known before 1900. In the late 1970s, it was first appreciated how strongly the outcome of diverse renal diseases was dependent on proteinuria, and that this seemed even more important than the nature of the renal diagnosis (Fig. 50.2) (Cameron, 1979). This experience was extended across other diseases including those in which proteinuria was not an initial feature such as reflux nephropathy (Kincaid-Smith and Becker, 1979), and during the 1980s it was found that the association of poor outcome with proteinuria seemed universal (Williams et al., 1988; Cameron, 1990; D’Amico, 1991). Poorer outcomes in the group of patients with > 1 g of proteinuria were illustrated again in the long-term results of the MDRD study (Peterson et al., 1995; Menon et al., 2009). These observations hold true. Moriyama et al. (2014) show the renal survival curves of patients with IgA nephropathy and different degrees of proteinuria (Fig. 50.3). Gutiérrez et al. (2012) confirm excellent long-term outcomes of proven IgA disease in which there is haematuria with no or minimal proteinuria.

Chapter 50 

References

100 90 80

% Survival

70

NS+

60 50

NS– FSGS MCGN I Membranous (o)

40 30 20 10 0

0

5

10 Years from onset

15

20

Fig. 50.2  Proteinuria predicts outcome more strongly than histological type of glomerulonephritis. 196 patients with three different histological diagnoses (90 FSGS, 40 Mesangiocapillary glomerulonephritis type 1, 66 Membranous nephropathy) were divided into NS+ (proteinuria > 2 g/24 hours) and NS− (proteinuria < 2 g/24 hours). From Cameron (1979).

Renal survival rate

1.0

n

U-Prot (g/day)

0.8

100 mg/m2/24 hours, higher in neonates (Hogg et al., 2000). Using this definition, postural proteinuria is common. Brandt’s small study of 91 healthy children and young people aged 6–19 years found 20% of them qualified for a diagnosis of postural proteinuria using these criteria, others (e.g. Dodge et al., 1976) are closer to 5%. This does raise the question of whether this is simply physiology at this age, rather than pathology. Perhaps the accepted norm is too low. Some studies have suggested that the incidence may be higher in children with obesity and hypertension, or it may simply be that they have slightly higher urine excretion so meet the criteria more easily. Other authors have suggested that gain in body mass may reduce obstruction to the left renal vein (nutcracker phenomenon, see ‘Relationship to nutcracker phenomenon’) when/if this explains proteinuria.

There is a contentious relationship in the literature between nutcracker phenomenon/syndrome (see Chapter  48) and postural proteinuria, a suggestion first raised in 1958 (reviewed by Mazzoni et al., 2011). Nutcracker syndrome is more commonly associated with macroscopic haematuria, but modern imaging methods are no doubt picking up this anatomical feature more frequently. A positive review of published reports of this association (Mazzoni et al., 2011) found that five studies included 229 patients aged 5–17 years. Nutcracker syndrome was identified by Doppler ultrasound in 68%, versus reports of the same venous anatomical features (in different studies) in up to 5% of controls. These studies were mostly undertaken in Asia, and the analyses were not blinded. A reassessment of 13 of their own patients 6 years later was described as showing that both nutcracker syndrome and proteinuria had disappeared in nine patients, in three patients both had persisted, and in one patient the proteinuria had persisted but nutcracker syndrome resolved (Milani et al., 2010). However, in a study of 99 consecutive adult potential renal transplant donors by computed tomography angiography (Grimm et al., 2013), a high incidence of asymptomatic nutcracker phenomenon was found. Twenty-seven per cent of patients had substantial (> 50%) compression of the left renal vein, and the finding of dilated gonadal or lumbar veins was common in these healthy individuals who had neither haematuria nor proteinuria. Lee et al. (1997) studied two girls aged 11 and 12 with marked postural proteinuria (24-hour excretion > 40 mg/m2/hour, 10 times the upper limit of normal) extensively. Both had marked left renal vein obstruction with collateral vein formation. Bilateral ureteral catheterisation was undertaken and showed that only urine from the left kidney contained increased protein. The relationship remains uncertain. Issues around diagnosis of nutcracker syndrome are discussed in Chapter 48. However, as outcomes of orthostatic proteinuria appear generally good, intervention to alter the anatomy cannot be easily justified for this reason.

Mechanism

Management

Diurnal variation of protein excretion has been identified in those with pathological explanations for proteinuria (Wan et al., 1995), and also in those with proteinuria in the normal range (Brandt et  al., 2010). This may be related to postural changes in renal haemodynamics and could also be related to changes in renal vein compression (see ‘Relationship to nutcracker phenomenon’).

Outcomes are reported to be uniformly good, with little suggestion that there is extra long-term risk associated. However, large and very long studies would probably be required to identify such outcomes, given the size and duration of study required to show, for example, adverse long-term outcomes from microscopic haematuria (see Chapter 46). There are a few long-term reports (Levitt,

Postural proteinuria was defined as one of the relatively benign variants of ‘cyclical’ or recurrent proteinuria in studies in the nineteenth century, following widespread testing of Bright’s ominous association of proteinuria with serious renal disease. Pavy (1886) gives a clear account of the disorder in three patients. He tested the effect of delayed rising from bed, and ascertained that it was posture that kept the proteinuria away rather than cold exposure causing it, by comparing the effect of a cold bath.

Epidemiology

486

Section 3  

the patient with glomerular disease

1967; Antoine et al., 1968; McLaine and Drummond, 1970), including a remarkable follow-up of patients seen by over 40 years previously Thomas Addis (Rytand and Spreiter, 1981). If there is extra long-term risk it is likely to be very low. If morning samples have normal levels of protein and there are no other pointers to renal disease (including normal blood pressure and no haematuria), no further investigations are warranted. The simplest monitoring technique is occasional measurement of protein:creatinine or albumin:creatinine ratios in first-in-morning urine samples. Investigation seeking to demonstrate nutcracker phenomenon (see Chapter 48) is not recommended. The results may be misleading and results are highly unlikely to alter management. Where total protein excretion is very high, or there are other pointers to disease, further investigation may be considered. It has been observed that angiotensin-converting enzyme inhibitors can reduce proteinuria in this condition as in proteinuria of other causes (Ha and Lee, 2006).

References Antoine, B., Symvoulidis, A., and Dardenne, M. (1968). La stabilité évolutive des états de proteinurie permanente isolée. Nephron, 6, 526–36. Brandt, J. R., Jacobs, A., Raissy, H. H., et al. (2010). Orthostatic proteinuria and the spectrum of diurnal variability of urinary protein excretion in healthy children. Pediatr Nephrol, 25, 1131–17. Dodge, W. F., West, E. F., Smith, E. H., et al. (1976). Proteinuria and hematuria in schoolchildren: epidemiology and early natural history. J Pediatrics, 88, 327–47. Grimm, L. J., Engstrom, B. I., Nelson, R. C., et al. (2013). Incidental detection of nutcracker phenomenon in an asymptomatic

population: prevalence and associated findings. J Comput Assist Tomogr, 37, 415–18. Hogg, R. J., Portman, R. J., Milliner, D., et al. (2000). Evaluation and management of proteinuria and nephrotic syndrome in children: recommendations from a pediatric nephrology panel established at the National Kidney Foundation conference on proteinuria, albuminuria, risk, assessment, detection, and elimination (PARADE). Pediatrics, 105, 1242. Ha, T. S. and Lee, E. J. (2006). ACE inhibition can improve orthostatic proteinuria associated with nutcracker syndrome. Pediatr Nephrol, 21, 1765–8. Lee, S. J., You, E. S., Lee, J. E., et al. (1997). Left renal vein entrapment syndrome in two girls with orthostatic proteinuria. Pediatr Nephrol, 11, 218–20. Levitt, J. I. (1967). The prognostic significance of proteinuria in young college students. Ann Internal Med, 66, 685–96. McLaine, P. N. and Drummond, K. N. (1970). Benign persistent asymptomatic proteinuria in childhood. Pediatrics, 46, 548–52. Mazzoni, M. B., Kottanatu, L. K., Simonetti, G. D., et al. (2011). Renal vein obstruction and orthostatic proteinuria: a review. Nephrol Dial Transplant, 26, 562–5. Milani, G. P., Mazzoni, M. B., Burdick, L., et al. (2010). Postural proteinuria associated with left renal vein entrapment: a follow-up evaluation. Am J Kidney Dis, 55, e29–31. Pavy, F. W. (1886). A further contribution on cyclical albuminuria; with observations on the effect of various conditions upon the diurnal appearance of albumen. Lancet, 127, 437–8. Rytand, D. A. and Spreiter, S. (1981). Prognosis in postural (orthostatic) proteinuria. Forty to fifty-year follow-up of six patients after diagnosis by Thomas Addis. N Engl J Med, 305, 618–21. Wan, L. L., Yano, S., Hiromura, K., et al. (1995). Effects of posture on creatinine clearance and protein excretion in patients with various renal diseases. Clin Nephrol, 43, 312–17.

CHAPTER 52

Nephrotic syndrome Premil Rajakrishna, Stewart Cameron, and Neil Turner Introduction Richard Bright’s cases of advanced renal disease causing dropsy were not examples of simple nephrotic syndrome, but by drawing attention to the significance of urinary protein he led other physicians to study the problem. Robert Christison in Edinburgh quickly identified patients with episodes of dropsy associated with proteinuria that recovered—later described by physicians as nephrosis, and subsequently as nephrotic syndrome (Cameron and Hicks, 2002; Turner, 2010; see Chapter 42).

Definition Criteria for diagnosis of nephrotic syndrome are inconsistent, and this is understandable (Glassock et al., 2015). The essential elements are that there should be high-level proteinuria with lowered serum albumin. Most definitions also include oedema. However, oedema appears at different levels of proteinuria in different individuals, and correlates poorly with serum albumin. Age and salt intake may explain some of this variation. Setting strict limits on proteinuria or serum albumin for diagnosis is, however, problematic. We commonly use the term nephrotic range proteinuria, usually meaning > 3.5 g/24 hours, (see Chapter 50), but many patients only develop low serum albumin and oedema at levels of proteinuria appreciably higher than this. Others have such severe nephrotic syndrome that serum albumin is very low, and protein excretion may fall as a consequence of this. It may also fall because of hypovolaemia and acute kidney injury (AKI). We also have to make our decisions based on measurements of known variability or unreliability such as spot urinary protein:creatinine ratio (PCR) and 24-hour urine collections. The clinical implications of nephrotic syndrome are broadly proportional to the severity of the protein leak and the oedema, and not defined by whether a particular threshold is reached. The concept of nephrotic syndrome includes the immediate symptomatic consequences, notably oedema, but also the increased risk of infection, thrombosis, hyperlipidaemia, and disturbances of circulating volume.

Clinical features Proteinuria itself is symptomless in most individuals. Frothy urine may be caused by heavy proteinuria, but is not a specific sign. Oedema is the most common presenting feature, but presentation with complications is also common. Typically these are infections, notably spontaneous peritonitis in children with ascites; venous thrombosis or thromboembolism; or manifestations of hyperlipidaemia.

Primary features Nephrotic oedema The five attributes of nephrotic oedema are gradually increasing, gravitational, generalized, pitting, and softness. Nephrotic oedema is noticeable first only around the eyes in the morning, and the ankles in the evening, but with increasing fluid retention there is sustained swelling of ankles and face (Fig. 52.1) which can lead to a misdiagnosis of allergy. If patients are in bed, fluid accumulates as a sacral pad and oedematous elbows. The effects of gravity are less evident in children:  children and sometimes young adults may suffer considerable ascites and facial oedema without ankle oedema. Younger patients also seem to tolerate lower levels of serum albumin before forming detectable oedema. In adults, retention of up to 4 L of salt and water remains undetectable, revealed only by weighing. With increasing oedema, ascites may appear followed by pleural effusions, which are usually bilateral, occasionally unilateral, and usually limpid, but sometimes opaque and chylous. Genital oedema may be distressing, especially in males. The oedema remains soft and pitting even when profound (Fig. 52.2), but if it remains untreated for long periods it may become indurated and pit only with difficulty especially around the ankles. Ankle swelling may be asymmetrical if deep venous thrombosis supervenes. Striae may appear even if no corticosteroids are being given, and the skin may actually split and weep spontaneously. Needlestick punctures may also weep profusely. The liver may be painlessly enlarged, especially in children. Patients with increased portal vein pressure will form ascites.

General features Patients with severe nephrotic syndrome often feel tired and lacking in energy, even if they do not have substantial oedema. The explanation for this is not clear. The jugular venous pressure is usually normal or low, but if raised in association with a low or normal blood pressure in an older adult with nephrotic syndrome this raises suspicion of cardiac amyloidosis as a cause. The nails may show white bands corresponding to periods of hypoalbuminaemia (Fig. 52.3). Where nephrotic syndrome is caused by a systemic disease or is otherwise secondary (e.g. to drugs or malignancy), history and examination may hint at or reveal these. Many nephrotic patients lose muscle and flesh weight as well as oedema. This is probably a mixture of reduced appetite and activity, compounded by the effects of corticosteroid therapy in some.

488

Section 3  

the patient with glomerular disease

(A)

(A)

(B)

(B)

Fig. 52.2  Legs of an adult patient with minimal change disease (A) before and (B) after treatment with prednisolone.

but events are not evenly distributed. Most thrombotic events are reported within the first 6  months of the diagnosis (Mahmoodi et al., 2008, Lionaki et al., 2012). Membranous nephropathy is usually more commonly associated with thrombosis than other causes (Bellomo and Atkins, 1993), though the incidence was relatively

Fig. 52.1  Face of a patient with severe nephrotic syndrome (A) before and (B) after diuretics.

Complications Thrombosis Presentation with a thrombotic complication is rare in children compared to adults, but more serious if it occurs (Deshpande and Griffiths, 2005). Both arterial thrombosis and venous thrombosis are reported but venous thrombosis is much more common in most series (Cameron et al., 1988). Deep vein thrombosis is the commonest. The tendency for thrombosis is approximately related to the severity of nephrotic syndrome,

Fig. 52.3  The white-banded nail of an adult nephrotic patient with membranous nephropathy, representing a period of relapse when there was a severe nephrotic syndrome and profound hypoalbuminaemia. The exact pathogenesis of this type of nail appearance is not known; in some patients the nail becomes diffusely white from the lunula outwards as the disease continues (sometimes called a ‘half and half’ nail).

Chapter 52 

low in 898 patients studied in the Netherlands (Lionaki et  al., 2012) at 7% over 3 years. Pulmonary embolism should be suspected in cases of sudden shortness of breath, although chest infection, pleural effusion, acidosis with renal insufficiency, abdominal distension with ascites, or anaemia may be the cause. Thrombosis of the deep calf veins is common, overt thrombosis occurring in 3–12% of adults in older series (Llach, 1982; Cameron, 1984). Mahmoodi et al. (2008) studied 298 consecutive adults with different diagnoses and found an annualized incidence of thrombosis of about 1%, but heavily biased to the first 6 months when the incidence was 10%. The rate of arterial events was also high in this study. Thrombosis is less frequent in children; < 1% of 4158 paediatric patients in published series showed clinically evident deep vein thrombosis (Andrew and Brooker, 1996). Studies seeking asymptomatic thrombosis find higher rates of occult thrombi. However, active approaches to seeking and preventing thrombosis beyond routine prophylactic measures are controversial (see ‘Management’).

Renal vein thrombosis Clinically apparent renal vein thrombosis is rare, though the literature contains many reports suggesting a high but also extraordinarily variable rate of occult thrombi, 5–50% (Rostoker et al., 1992; Singhal and Birmble, 2006). The higher figures surprise clinicians who have looked with modern techniques and rarely identified renal vein thrombosis. Chronic renal vein thrombosis is usually asymptomatic, but may be detected after a pulmonary embolus is identified. Acute renal vein thrombosis presents with loin pain, haematuria, and elevated lactate dehydrogenase levels. If the patient has associated renal failure, bilateral involvement should be suspected. The latter is rare and often seen when additional risk factors for thrombosis (severe dehydration, antiphospholipids, protein C or S deficiency, etc.) are present. It may be a presenting feature of nephrotic syndrome, with or without pulmonary embolism, when the complex acute clinical picture often creates diagnostic difficulty. Selective renal venography remains the gold standard test for the diagnosis, but it is invasive and rarely undertaken (Singhal and Brimble, 2006). Renal ultrasound may show enlarged kidneys and Doppler ultrasound can identify thrombosis but its sensitivity is not clear. Contrast-enhanced spiral computed tomography is probably now the most commonly employed technique. It does, however, add to risk of AKI (see Chapter 14). It has not been established that seeking symptomless renal venous thrombi is useful (Rostocker et al., 1992), since their prognosis appears to be benign.

Infections Infections remain a significant cause of morbidity and sometimes mortality in nephrotic syndrome, particularly in the developing world. Six of 10 deaths in 389 children with minimal change nephrotic syndrome were from sepsis (International Study of Kidney Disease in Children, 1984). Children with nephrotic syndrome appear more vulnerable to infections than adults but they can be serious in both. Pneumococcal infections are particularly prevalent. In a series of studies of peritonitis in nephrotic children, Streptococcus pneumoniae and Escherichia coli were the most common pathogens (Krensky et al., 1982). Increased incidences of urinary, respiratory, and central nervous system infections are also reported (Uncu et al., 2010).

nephrotic syndrome

The exact mechanisms of immune defects are not well understood, but urinary losses of immunoglobulins and other immune mediator molecules are presumed to contribute. Immunoglobulin levels are lowered but not to levels associated with infection in inherited hypogammaglobulinaemia. Sites of abnormal fluid collection are common infection sites, patients can present with peritonitis, cellulitis, or empyema. Spontaneous peritonitis may occur at presentation or as a later complication. Varicella infection is also reported to be increased, but steroid treatment is likely to be a major contributor to this propensity.

Hyperlipidaemia Dyslipidaemia is a universal finding in nephrotic syndrome, and characterized by often very high plasma cholesterol (> 10 mmol/L), low-density lipoprotein (LDL), triglyceride, and low high-density lipoprotein levels (Crew et  al., 2004). The major clinical impact is seen in those with chronic nephrotic syndrome, and those on long-term steroid therapy complicated by additional risk factors such as hypertension. Lipid abnormalities are proportionate to proteinuria, and remit as nephrotic state remits (Shearer et  al., 2001). Treatments that lower proteinuria correct lipid abnormalities as well. Hyperlipidaemia is thought to be due to a compensatory liver response to reduced plasma oncotic pressure mediated by hepatic apoprotein B gene transcription, but there may be a contribution from defective lipid catabolism as well (Shearer et al., 2001).

Acute kidney injury AKI is unusual but well recognized. It may be related to the nephrotic state, iatrogenic, or related to the underlying disease process. AKI at presentation Apparently ‘idiopathic’ AKI is an occasional but important complication, and can be distinguished from AKI from identifiable causes such as interstitial nephritis, thrombosis, sepsis, or contrast media. Bilateral acute renal vein thrombosis is a rare explanation which needs to be considered (see above). AKI at presentation occurs mostly in older patients of either gender, overwhelmingly (81%) in those with minimal change/ focal segmental glomerulosclerosis (FSGS) histology (Smith and Hayslett, 1992; Waldman et al., 2007). In children with the same diagnosis, AKI is rare and usually follows either sepsis or thrombosis (Cameron et al., 1988; Cavangnaro and Lagomarsino, 2000). Most adult patients present already in AKI; some develop it subsequent to diagnosis or in relapse. One-sixth were judged to be seriously hypovolaemic or in shock, and all had a very low serum albumin. Urine volume is low, containing < 5 mmol/L of sodium and unresponsive to diuretics and/or volume repletion, loaded with protein, and containing red cells and often red cell casts. Thus, renal biopsy is almost always necessary to establish a diagnosis, as this pattern of sediment suggests a proliferative nephritis rather than minimal changes. The role of hypovolaemia and reduced renal perfusion is not clear as AKI is often present before diuretic therapy is initiated, and by the time of diagnosis many patients have a full circulation and hypertension was common in early series. However, renal circulatory disturbance may be a common feature in severe nephrotic syndrome and could predispose to AKI (Koomans et al., 2001; Vande Walle et al., 2004) (see Chapter 53).

489

490

Section 3  

the patient with glomerular disease

Renal biopsy usually shows moderate to severe tubular changes (Venkataseshan et al., 1993). Interstitial oedema is usually present, perhaps indicating increased interstitial pressure (Lowenstein et  al., 1970)  and it has been suggested this could contribute to pathogenesis. Management of these often elderly and severely ill patients follows usual principles (Chapters 228 and 233) but is difficult. They continue to pass large amounts of protein in tiny amounts of urine, have very low serum albumin and sometimes unstable circulation, and are of course uraemic. If not already malnourished, they rapidly become so. Secondary AKI Delayed AKI is commonly related to over-diuresis or a secondary complication such as vomiting or diarrhoea. Otherwise unexplained low blood pressure, tachycardia, cold extremities, restlessness, renal dysfunction should point to this (Vande Walle et al., 1995). The commoner form is a transient mild rise in creatinine levels secondary to intravascular volume depletion due to over-diuresis or severe hypoalbuminaemia. Children and adults with minimal change disease are more vulnerable and correction of volume status reverses it.

Endocrine dysfunction Endocrine dysfunctions such as hypothyroidism and vitamin D deficiencies are noted in nephrotics, but steroid-induced endocrine abnormalities are more commonly encountered (Crew et al., 2004).

Diagnosis Proteinuria and serum albumin There should be heavy proteinuria, for example, PCR > 300 mg/ mmol, or 24-hour protein > 3.5 g. This odd value corresponds to an extension of the criterion for nephrotic proteinuria being > 40 mg/ m2/hour of albumin, applied to a 70-kg man. So 3.5 g should not be used as a diagnostic level for a 50-kg man or woman. The limits only apply to proteinuria that is composed predominantly of albumin, but in clinical practice almost all nephrotic range proteinuria is predominantly albumin. Unusually high levels of overflow protein such as immunoglobulin light chains (overflow proteinuria), or of tubular proteinuria, do not cause nephrotic syndrome (see Chapter 50). These exact limits of even albuminuria should not be over-interpreted. The manifestations of nephrotic syndrome do not start at a given value of albumin or proteinuria, they are graded. However, it is clinically useful that at a certain level of proteinuria a characteristic set of renal diagnoses becomes most likely. Whether or not there is associated haematuria is important in narrowing down the possible underlying cause (see Chapter 45). Serum albumin will be low, usually < 30 g/L and sometimes very low, < 10 g/L. Correlation with the degree of oedema is imperfect, another indication that changes in Starling pressures alone cannot explain the pathogenesis of oedema in nephrotic syndrome (see Chapter 53).

Differential diagnosis Demonstrating nephrotic range proteinuria and hypoalbuminaemia in an oedematous patient confirms the diagnosis of nephrotic syndrome. But in an outpatient setting and in complicated patients

the diagnosis may be challenging. Findings in oedema of different causes are discussed in Chapter 30. Patients with advanced renal impairment commonly present with fluid retention with significant proteinuria and oedema, mimicking nephrotic syndrome at first glance. These patients are typically characterized by excess intravascular fluid with elevated jugular venous pressure and hypertension, not simply peripheral oedema (see Chapter  53). They may show features of uraemia. Estimation of creatinine and albumin and renal ultrasound can quickly prove this. Nephrotic syndrome implies dysfunction of the glomerular filtration barrier affecting the podocyte (see Chapter  45). ‘Pure’ nephrotic syndrome, in which there is no or minimal haematuria, has a characteristic set of causes. If there is significant haematuria the differential diagnosis broadens. Haematuria suggests that the glomerular basement membrane is being breached, sometimes by genetic cause (Alport syndrome), but most commonly by inflammation within the glomerulus. Many inflammatory diseases cause proteinuria, some to nephrotic levels, via podocyte damage which either occurs directly, or through alterations in the glomerular matrix and milieu that lead to podocyte dysfunction (see Chapter 45, Fig. 45.1). The range of causes in a UK centre across several decades is shown in Fig. 52.4. In pregnancy, pre-eclampsia (see Chapter  296) must be added to the list of common causes of nephrotic syndrome.

Urinary sediment Clues given by urine sediment examination under light microscopy can be time saving and save cost, particularly in resource-poor settings. Non-proliferative glomerulopathies (minimal change, membranous, and FSGS) are more likely to have a bland urinary sediment showing hyaline casts and oval fat bodies but few red cell or other cellular casts (see Chapter 6). The term ‘nephrosis’, now abandoned, was used to describe this type of kidney disease. More red cells and cellular casts in urine is indicative of proliferative

Other Henoch-Schnlein purpura Lupus Other proliferative diseases Mesangiocapillary glomerulonephritis Membranous nephropathy FSGS Minimal change

Childhood

Young adult

Middle and old age %

Other Lupus Amyloid

80 Diabetes other proliferative 60 Mesangiocapillary glomerulonephritis 40 Membranous nephropathy 20

FSGS Minimal change

Fig. 52.4  Underlying histological appearances found in renal biopsies from more than 1000 nephrotic patients of all ages seen at Guy’s Hospital, London, 1963–1990. Note that the majority of children under the age of 15 years have minimal change disease, the proportion falling steadily from 2 to 15 years of age. However, minimal change disease remains an important cause of the nephrotic syndrome in adult nephrotics, and overall is the commonest form. Membranous nephropathy, in contrast, becomes steadily more common with age and is the commonest form of nephrotic syndrome in elderly patients.

Chapter 52 

glomerulonephritis, (International Study of Kidney Disease in Children, 1978) (see Chapter 45).

Other blood tests Hyponatraemia is common during diuretic therapy. Hypovolaemia may provoke AKI. Blood count may show haemoconcentration. Immunological investigations (Table 52.1) should be used selectively, based on probability. Few of these tests prove a cause alone, or can completely replace all the information that may come from a renal biopsy.

Renal biopsy Renal biopsy examination with light microscopy and immunofluorescence will allocate patients to a particular histological category, but the final interpretation often needs correlation with clinical and serological data. Table 52.1  Tests to consider for main causes of nephrotic syndrome (Howard et al., 1990; Hofstra et al., 2011) Nephropathy

Test

Minimal change/FSGS

Comment No tests available. Age and race influence likelihood

Membranous

Hepatitis B and C and HIV serology (Antibodies to phospholipase A2 receptor (PLA2R)? Other?)

PLA2R not yet proven as a diagnostic test (Chapter 61) Consider skin-lightening creams—mercury level?

Diabetes mellitus

HbA1c

Check clinical likelihood—duration of diabetes, course (Chapter 149)

Amyloidosis

C-reactive protein (for chronic inflammation) Urinary immunofixation for light chains (Bence Jones protein) Serum free light chains; serum protein electrophoresis

Serum amyloid A component may be appropriate in monitoring for suppression of AA amyloid formation (Chapter 152)

Lupus nephritis

Complement (C3, C4, CH50) ANA, anti-dsDNA

May cause ‘pure’ nephrotic syndrome or mixed nephritic/ nephrotic (see Chapter 162)

Membranoproliferative glomerulonephritis

Complement (C3, C4, CH50) Hepatitis B and C and HIV serology

More detailed complement studies may be indicated depending on subtype (Chapter 80) Consider cryoglobulins

Fibrillary and immunotactoid

Serum free light (Chapter 81) chains; serum protein electrophoresis; urine for Bence Jones protein

nephrotic syndrome

Common exceptions to the need for a renal biopsy are ◆ Diabetes—long-standing

diabetes with entirely typical progression from microalbuminuria to proteinuria over many years, with evidence of microvascular complications affecting other organs, particularly retinopathy or neuropathy. In these circumstances many clinicians do not undertake a renal biopsy.

◆ Children—if

a child presents with nephrotic syndrome between the ages of 1 and 10 years, with normal renal function, normal complement levels without hypertension or haematuria, then the diagnosis is highly likely to be minimal change disease (International Study of Kidney Disease in Children, 1978). If these clinical criteria are met, treatment with steroids may be initiated avoiding a renal biopsy. At a later point, renal biopsy may be indicated if there is a poor response to steroids or clinical profile changes. The evidence behind this approach comes mostly from series in Caucasian populations, in which minimal change disease is the dominant cause. It caused 76% of primary disease as reported by the International Study of Kidney Disease in Children (1978).

◆ Frail

elderly or others with severe comorbid conditions where a pathological diagnosis is very unlikely to alter best management.

Epidemiology Incidence In general, there is male predominance in the occurrence of nephrotic syndrome with a ratio of approximately 2:1. Lupus is the main condition with a contrary ratio, and if pre-eclampsia was included the distribution would look very different. In children, a series of prospective studies in Caucasians give incidence rates between 1.2 and 2.0 cases per 100,000 per year, though the incidence is much higher in retrospective studies of African or Asian patients (Eddy and Symons, 2003; Wong et al., 2007).

Causative disease Primary causes are more common than secondary in both children and adults. The diseases most commonly causing nephrotic syndrome vary with age and by race and/or geographical region. Fig. 52.4 shows the age effect in one centre over nearly three decades. In a US study in adults, FSGS (35%) and membranous glomerulonephritis (33%) were the major primary histological pattern followed by minimal change disease (15%) and membranoproliferative glomerulonephritis (14%). This report is also one of several to have highlighted a trend towards increasing incidence of FSGS over membranous glomerulonephritis (Korbet et al., 1996). Out of the secondary causes of nephrotic syndrome in adults, diabetic nephropathy (50 cases per 100,000 population) leads, followed by lupus and amyloid (Haas et al., 1997).

Ethnicity-related demographics Asian children are at six times higher risk of developing primary nephrotic syndrome than Europeans, while black and Hispanic children report a higher rate of FSGS and steroid-resistant nephrotic syndrome (Niaudet, 2004; Boyer et al., 2007). Africa reports a strikingly low rate of primary steroid-sensitive nephrotic syndrome in

491

the patient with glomerular disease

Management of nephrotic syndrome Management of nephrotic oedema Diuretics and salt restriction Oedema is caused by sodium retention (see Chapter  53) so salt restriction is rational. Patients are advised to avoid salty food, use no added salt (using pepper or chilli or substitute tastes instead), to restrict salt intake below 50–70 mmol/day (about 4 g). Few patients can achieve lower on a modern diet, and given the effectiveness of current diuretics this is usually enough. Diuretics are almost always required. Diuresis initially depletes intravascular volume and promotes gradual fluid shifts from tissue spaces to vascular compartment. Usually, adult nephrotics tolerate up to 2–3 L of fluid loss per day for short periods without critical depletion of intravascular volume. Acute depletion of effective circulatory volume may be detected clinically by tachycardia, reduced peripheral perfusion, postural drop, and restlessness. Ongoing subacute volume depletion may be picked up by noticing otherwise unexplained gradual increase in serum creatinine levels in the absence of the above-mentioned clinical features (Geers et al., 1985). Those with severe hypoalbuminaemia are more prone to this complication (see Chapter 51). Daily weights and fluid balance charts are helpful to titrate the diuretic dosage in severe nephrotic syndrome, avoiding this complication, but most patients can be managed as outpatients, particularly if able to weigh themselves and allowed to adjust diuretic dose responsively. Loop diuretics are usually first line in nephrotic syndrome as there is relative diuretic resistance. Higher doses are needed than used in other oedematous patients with the same level of kidney function. Often up-titration of the loop diuretic dose is required which can be done by doubling the doses at short intervals, even daily. At each up-titration, caution has to be exercised to avoid critical plasma volume depletion (Wilcox, 2002).

Reducing proteinuria Treatment that induces disease remission rapidly mobilizes oedema and diuretics are quickly no longer required. This is best illustrated in minimal change disease with steroids (Fig. 52.5). Where the cause of nephrotic syndrome cannot be reversed, other drugs that reduce proteinuria without altering the natural history of the disease may make oedema easier to control, and reduce risk of some of the general complications of nephrotic syndrome. Least controversially, angiotensin-converting enzyme inhibitors (ACEIs) reduce proteinuria in all glomerular proteinuric diseases. They are indicated in all patients with substantial proteinuria, whether they have nephrotic syndrome or not (see Chapter 50)

Second-line therapies Agents in addition to ACEIs/angiotensin receptor blockers (ARBs) that might further reduce proteinuria have a less certain role in improving outcomes in renal disease (Chapters  45 and 99), but may have a place in the control of severe uncontrolled nephrotic syndrome. Dual therapy with ACEIs and ARBs is one approach, but does increase risk of hyperkalaemia and AKI. Some additional proteinuria lowering has been claimed for a variety of other drugs. Many of these may act at the podocyte. Candidates include aldosterone antagonists, vitamin D, endothelin antagonists, peroxisome proliferator-activated receptor (PPAR)-γ agonists, the PPAR-α agonist fenofibrate, and others. Most of these have relatively small potential incremental benefit, however. Clinically much more effective are the calcineurin inhibitors, tacrolimus or ciclosporin. While

Prednisolone 60 mg/day.

20 mg/day.

10 mg/day.

13 11 9

Resistant oedema Thiazide diuretics and aldosterone antagonists (spironolactone, epleronone) or amiloride may be used as second-line add-on diuretics. Amiloride has theoretical advantages that have not been systemically explored (see Chapter  53). The value of albumin-furosemide infusion to enhance diuretic action and reduce oedema has not been shown to provide additional benefit on average (Fliser et  al., 1999)  (see Chapter  30). Albumin infusion greatly increases proteinuria. Patients with ascites may better respond to intravenous diuretics if oral absorption is suspected to be poor due to gut oedema; but moving to intravenous diuretic therapy is in any case the next step up in therapy for patients not responding to maximum doses of combined oral diuretics. Sodium excretion should be checked in those who appear to be diuretic resistant, as some patients find sodium restriction very difficult. (See also Chapters 30 and 33.)

40 mg/day.

15 Weight kg.

children (e.g. Pakasa and Sumaili, 2009). The pattern of histology from nephrotic children in India, Pakistan, and Turkey seems similar to that from European and North American studies (Kumar et al., 2003; Ozkaya et al., 2004; Kazi and Mubarak, 2007; El Bakkali et al., 2011). However, the wide variability in biopsy indications, time frames, and sample sizes of various studies makes it difficult to draw solid conclusions.

Protein excretion g/24 hr.

Section 3  

Jrine volume L/24 hr.

492

4 3 2 1

3 2 1 0

4

3 12 16 20 24 28 32 36 40 Days after onset of treatment

Fig. 52.5  Treatment of nephrotic syndrome in a child in the 1950s, before effective diuretics were available and soon after the discovery of the effect of corticosteroids in childhood nephrotic syndrome. Weight falls from 15 to 10 kg over about a week as the protein leak responds to steroid therapy. Note that the spontaneous diuresis commences as soon as the level of proteinuria drops below 1 g and before albumin levels could have recovered. From DeWardener (1958), with permission.

Chapter 52 

some of their effect may be related to lowering glomerular filtration rate (GFR), it is possible that they have an additional direct effect on podocytes (see Chapter 136). Their ability to produce a complete remission in minimal change disease may be a pointer to this (see Chapter 45). Progressively increasing the dose will usually achieve some reduction in proteinuria regardless of the cause. Physical therapies are possibly under-used. Head-out water immersion can improve fluid loss in generalized oedema of various causes (see Chapter 30).

Chemical or actual nephrectomy Some patients with very severe nephrotic syndrome cannot achieve adequate quality of life despite all these therapies, requiring diuretic therapy that causes severe postural symptoms or AKI. This is rare, but particularly likely in congenital Finnish nephrotic syndrome, and is sometimes seen in amyloidosis. Even more rarely it may occur in other causes of nephrotic syndrome. Chemical nephrectomy usually refers to adding high dose non-steroidal therapy, usually with indomethacin, to the measures above, including high-dose calcineurin inhibition. This combination reduces GFR significantly. As a last resort, renal tissue can be destroyed by embolization (which will provoke infarction that is likely to be painful and will be pro-inflammatory) or by physical nephrectomy.

General management The optimal dietary protein intake for patients with a persisting nephrotic syndrome remains controversial. Although recommended in the past, it has long been known that a high protein intake (> 1.5 g/kg/24 hours) leads to an increase in urinary protein excretion but without any increase in serum albumin or total plasma protein concentrations. In contrast, reduction in protein intake to 0.8 g/kg/24 hours reduces proteinuria, although with controversial effects on serum albumin concentration (Mansy et al., 1989; Kaysen, 1991) and a risk of protein malnutrition (Guarneiri et al., 1989). Therefore, a dietary protein intake of 0.8 g of protein of high biological value/kg/24 hours, plus 1 g protein per gram of proteinuria, has been advocated. In fact 0.8 g/kg/24 hours represents an average protein intake in Europe, although less than American norms.

Management of hyperlipidaemia Hyperlipidaemia is correlated with proteinuria, and patients with frequent relapses over longer periods, or with persistently poorly controlled proteinuria, are at a high risk of cardiovascular disease (Joven et al., 1990). Lipid control in renal disease is discussed in Chapter 102. It has been established that lipid-lowering therapy is safe in patients with chronic kidney disease in general. There is not good evidence specifically in nephrotic syndrome, but cholesterol and LDL are so high in many patients with persistent nephrotic syndrome, and tolerability of therapy seems high, so few nephrologists disagree with prescribing at least HMG-CoA inhibitors (statins) in these circumstances. These are the only current agents that make a substantial impact on the hyperlipidaemia of nephrotic syndrome, at least as a single agent. They reduce total cholesterol and LDL levels by 20–45% (Massy et al., 1995). Other anti-lipidaemic drugs such as fibrates and nicotinic acids can lower triglycerides effectively but their clinical benefits are less certain. Although dietary modification seems good general advice, its role is unproven.

nephrotic syndrome

Thrombosis and thrombotic risk Treatment of thrombosis Overt thromboembolic events such as pulmonary embolism and deep venous thrombosis are treated the same way as in non-nephrotics; starting with unfractionated or (if renal function is good) low-molecular-weight heparin (Wu et al., 2006) and then warfarin to maintain an international normalized ratio of 2–3. The duration of therapy may be 6–12 months but arguably should be continued as long as the nephrotic state persists. Glassock (2007) describes this as a ‘conundrum’. Most of the risk of nephrotic syndrome is concentrated around the time of diagnosis (see ‘Complications’ above). The risks of anticoagulation are increased in patients with renal disease and there is a clinical impression that this may be particularly true in those with nephrotic syndrome. Warfarin is bound to albumin, the concentration of which may change (Ganeval et al., 1986), and there may be other interfering factors. Higher doses of heparin may be required as it activates antithrombin III, the concentration of which may be diminished in nephrotic patients (Kerlin et al., 2012), although the complex effects on other proteins are likely to interact too. Asymptomatic renal vein or other thrombosis found incidentally is usually treated with anticoagulation, although there are no controlled randomized data available to support this approach. Bilateral acute renal vein thrombosis is treated with standard anticoagulation. For severe cases local thrombolytic therapy rather than systemic fibrinolytic therapy may be used, given the high risk of bleeding with the latter. Thrombectomy may be considered with local therapy.

Management of long-term thrombotic risk There is agreement on the importance of usual prophylactic measures in patients at temporary extra risk, for instance, in hospital, but views on full prophylactic anticoagulation vary widely. Data is lacking to support routine anticoagulation of all nephrotics with low serum proteins. Patients at greatest risk are within 6 months of diagnosis, and have the most severe nephrotic syndrome. Several studies observe that patients with membranous nephropathy are at higher risk. After this early period, risk remains increased, and may be of the order of 1% per annum, approximately eight times higher than the rate in matched controls (Mahmoodi et al., 2008, Lionaki et al., 2012). Very few of these late events are lethal. Lee et  al. (2014) applied a decision analysis strategy to Lionaki et al.’s (2012) series of 898 patients from two North American centres with membranous nephropathy. The model assumed no extra risk for anticoagulation in nephrotic patients, and could have overestimated thrombosis rate as a constant level of risk was assumed was assumed without adjustment for peak onset at the time of diagnosis and lower rate thereafter. With these assumptions it suggested that for patients at low risk of bleeding, benefits seemed certain for those with albumin < 20 g/L. For those at intermediate bleeding risk and albumin < 20 g/L benefits were lower, but still positive. It is not possible to recommend a universal approach. The risks of thrombosis must be balanced with the bleeding risks of anticoagulation, which may be increased in this group. Individual decisions are appropriate. Statins have been associated with reduced thrombotic risk in the general population, and a retrospective study in nephrotic syndrome (Resh et al., 2011) provided equivocal evidence for such an effect.

493

494

Section 3  

the patient with glomerular disease

Management of infection risk Prompt induction of remission of oedema or proteinuria are the most important goals and the decline in death rate from infection in nephrotic children is probably mainly the result of this, and the availability of effective antibiotics. The need for supplementary corticosteroids in those taking these drugs, or in those who have recently stopped them, should be remembered. Any severe infection should prompt discontinuation of cytotoxic therapy. Immunization against pneumococci is recommended. There is a high rate of seroconversion even in children taking high-dose prednisolone (Ulinski et al., 2008). The use of prophylactic penicillin or intravenous immunoglobulin administration is not supported by evidence (Wu et al., 2012) and most guidance does not recommend either of these. In nephrotic children with active disease, varicella/chicken pox is a threat. Those taking high-dose corticosteroids or other immunosuppressive agents within the previous 3 months are at risk of severe progressive disseminated disease. Prophylactic therapy should be given to any non-immune contacts of cases. Where an effective varicella vaccine is available it should be administered to non-immune patients. It is a live vaccine so cannot be given during high-dose steroid or immunosuppressive therapy, but it has been shown to be safe to administer to children in a study that accepted those taking up to 2 mg/kg prednisone (maximum 40 mg) on alternate days (Furth et al., 2003).

Loss of hormones, vitamins, and other molecules A number of plasma proteins important in the transport of metals, hormones, and drugs are of relatively small molecular weight and thus are lost easily into the urine of nephrotic patients. Free protein hormones, especially of low molecular weight, are also lost. Many of these have been studied, but remarkably few have substantial clinical impact. A number of abnormalities of calcium and vitamin D metabolism have been described, in part the result of losses of vitamin D binding protein (molecular weight 59 kDa) and its associated vitamin in the urine (Vaziri, 1993; Harris and Ismail, 1994). Nephrotics with reduced renal function do more readily develop bone disease (Tessitore et al., 1984), and earlier treatment than usual with vitamin D may have a place. A Fanconi syndrome has been described in a small number of nephrotic patients. Some of these had reversible tubular defects suggesting tubular damage from proteinuria (Shioji et al., 1974).

Proteinuria and progression of renal failure This is discussed in Chapter 50 and Chapter 136.

References Andrew, M. and Brooker, L. A. (1996). Hemostatic complications in renal disorders of the young. Pediatr Nephrol, 10, 88–99. Bellomo, R. and Atkins, R. C. (1993). Membranous nephropathy and thromboembolism: is prophylactic anticoagulation warranted? Nephron, 63, 249–54. Boyer, O., Moulder, J. K., and Somers, M. J. (2007). Focal and segmental glomerulosclerosis in children: a longitudinal assessment. Pediatr Nephrol, 22, 1159–66. Cameron, J.S. (1984). Coagulation and thromboembolic complications in the nephrotic syndrome. Adv Nephrol Necker Hosp, 13, 75–114.

Cameron, J. S. and Hicks, J. A. (2002). The origins and development of the concept of a ‘nephrotic syndrome’. Am J Nephrol, 22, 240–7. Cameron, J. S., Ogg, C. S., and Wass, V. J. (1988). Complications of the nephrotic syndrome. In J. S. Cameron and R. J. Glassock (eds.) The Nephrotic Syndrome, pp. 849–920. New York: Marcel Dekker. Crew, R. J., Radhakrishnan, J., and Appel, G. (2004). Complications of the nephrotic syndrome and their treatment. Clin Nephrol, 62, 245–59. Deshpande, P. V. and Griffiths, M. (2005). Pulmonary thrombosis in steroid-sensitive nephrotic syndrome. Pediatr Nephrol, 20, 665–9. De Wardener, H. E. (1958). The Kidney (1st ed.). London: Churchill. Eddy, A. A. and Symons, J. M. (2003). Nephrotic syndrome in childhood. Lancet, 362, 629–39. El Bakkali, L., Rodrigues Pereira, R., Kuik, D. J., et al. (2011). Nephrotic syndrome in The Netherlands: a population-based cohort study and a review of the literature. Pediatr Nephrol, 26, 1241. Fliser, D., Zurbrüggen, I., Mutschler, E., et al. (1999). Coadministration of albumin and furosemide in patients with the nephrotic syndrome. Kidney Int, 55, 629–34. Furth, S. L., Arbus, G. S., Hogg, R., et al. (2003). Varicella vaccination in children with nephrotic syndrome: a report of the Southwest Pediatric Nephrology Study Group. J Pediatr, 142, 145–8. Ganeval, D., Fischer, A. M., Barre, J., et al. (1986). Pharmacokinetics of warfarin in the nephrotic syndrome and effect on vitamin K dependent clotting factors. Clin Nephrol, 25, 75–80. Geers, A. B., Koomans, H. A., Roos, J. C., et al. (1985). Preservation of blood volume during edema removal in nephrotic subjects. Kidney Int, 28, 652–7. Glassock, R. J. (2007). Prophylactic anticoagulation in nephrotic syndrome: a clinical conundrum. J Am Soc Nephrol, 18, 2221–5. Glassock, R. J., Fervenza, F. C., Hebert, L., et al. (2015). Nephrotic syndrome redux, a clinical perspective. Nephrol Dial Transplant, 30, 12–17. Guarneiri, G. F., Toigo, G., Situlin, R., et al. (1989). Nutritional status in patients on long-term low-protein diet or with nephrotic syndrome. Kidney Int, 36 (Suppl. 27), S195–200. Haas, M., Meehan, S. M., Karrison, T. G., et al. (1997). Changing etiologies of unexplained adult nephrotic syndrome: a comparison of renal biopsy findings from 1976–1979 and 1995–1997. Am J Kidney Dis, 30, 621–31. Harris, R. C. and Ismail, N. (1994). Extra renal complications of the nephrotic syndrome. Am J Kidney Dis, 23, 477–497. Hofstra, J. M., Beck, L. H., Jr., Beck, D. M., et al. (2011). Anti-phospholipase A2 receptor antibodies correlate with clinical status in idiopathic membranous nephropathy. Clin J Am Soc Nephrol, 6, 1286–91. Howard, A. D., Moore, J., Jr., Gouge, S. F., et al. (1990). Routine serologic tests in the differential diagnosis of the adult nephrotic syndrome. Am J Kidney Dis, 15, 24–30. International Study of Kidney Disease in Children (1978). Nephrotic syndrome in children: prediction of histopathology from clinical and laboratory characteristics at time of diagnosis. A report of the International Study of Kidney Disease in Children. Kidney Int, 13, 159–65. International Study of Kidney disease in Children (1984). Minimal change nephrotic syndrome in children: deaths during the first 5 to 15 years' observation. Report of the International Study of Kidney Disease in Children. Pediatrics, 73(4), 497–501. Joven, J., Villabona, C., Vilella, E., et al. (1990). Abnormalities of lipoprotein metabolism in patients with the nephrotic syndrome. N Engl J Med, 323, 579–84. Kazi, J. I. and Mubarak, M. (2007). Pattern of glomerulonephritides in adult nephrotic patients—report from SIUT. J Pak Med Assoc, 57, 574. Kerlin, B. A., Ayoob, R., and Smoyer, W. E. (2012) Epidemiology and pathophysiology of nephrotic syndrome-associated thromboembolic disease. Clin J Am Soc Nephrol, 7, 513–20. Koomans, H. A. (2001). Pathophysiology of acute renal failure in idiopathic nephrotic syndrome. Nephrol Dial Transplant, 16, 221–4. Korbet, S. M., Genchi, R. M., Borok, R. Z., et al. (1996). The racial prevalence of glomerular lesions in nephrotic adults. Am J Kidney Dis, 27, 647–51.

Chapter 52 

Krensky, A. M., Ingelfinger, J. R., and Grupe, W. E. (1982). Peritonitis in childhood nephrotic syndrome: 1970–1980. Am J Dis Child, 136, 732–6. Kumar, J., Gulati, S., Sharma, A. P., et al. (2003). Histopathological spectrum of childhood nephrotic syndrome in Indian children. Pediatr Nephrol, 18, 657–60. Lionaki, S., Derebail, V. K., Hogan, S. L., et al. (2012). Venous thromboembolism in patients with membranous nephropathy. Clin J Am Soc Nephrol, 7, 43–51. Lowenstein, J., Schacht, R. G., and Baldwin, D. S. (1970). Renal failure in minimal change nephrotic syndrome. Am J Med, 70, 227–33. Mahmoodi, B. K., ten Kate, M. K., Waanders, F., et al. (2008). High absolute risks and predictors of venous and arterial thromboembolic events in patients with nephrotic syndrome: results from a large retrospective cohort study. Circulation, 117, 224–30. Massy, Z. A., Ma, J. Z, Louis, T. A., et al. (1995). Lipid-lowering therapy in patients with renal disease. Kidney Int, 48, 188–98. Kaysen, G. A. (1991). Hyperlipidemia of the nephrotic syndrome. Kidney Int, 39 (Suppl. 31), S8–15. Lee, T., Biddle, A. K., and Lionaki, S. (2014). Personalized prophylactic anticoagulation decision analysis in patients with membranous nephropathy. Kidney Int, 85(6), 1412–20. Llach, F. (1982). Nephrotic syndrome: hypercoagulability, renal vein thrombosis and other thromboembolic complications. In B. M. Brenner and J. H. Stein (eds.) The Nephrotic Syndrome, pp. 121–44. New York: Churchill- Livingstone. Mansy, H., Goodship, T. H., Tapson, J. S., et al. (1989). Effect of a high protein diet in patients with the nephrotic syndrome. Clin Sci, 77, 445–51. Niaudet, P. (2004). Steroid-sensitive idiopathic nephrotic syndrome in children. In E. Avner, W. Harmon, and P. Niaudet (eds.) Pediatric Nephrology (5th ed.), pp. 557–73. Philadelphia, PA: Lippincott, Williams & Wilkins. Ozkaya, N., Cakar, N., Ekim, M., et al. (2004). Primary nephrotic syndrome during childhood in Turkey. Pediatr Int, 46, 436–8. Pakasa, N. M. and Sumaili, E. K. (2009). The nephrotic syndrome in the Democratic Republic of Congo. N Engl J Med, 354, 1085–6. Resh, M., Mahmoodi, B. K., Navis, G. J., et al. (2011). Statin use in patients with nephrotic syndrome is associated with a lower risk of venous thromboembolism. Thromb Res, 127, 395–9. Rostoker, G., Texier, J. P., Jeandel, B., et al. (1992). Asymptomatic renal-vein thrombosis in adult nephrotic syndrome. Ultrasonography and urinary fibrin–fibrinogen degradation products: a prospective study. Eur J Med, 1, 19–22. Schrier, R. W. and Fassett, R. G. (1998). A critique of the overfill hypothesis of sodium and water retention in the nephrotic syndrome. Kidney Int, 53, 1111–17. Shearer, G. C., Stevenson, F. T., Atkinson, D. N., et al. (2001). Hypoalbuminemia and proteinuria contribute separately to reduced lipoprotein catabolism in the nephrotic syndrome. Kidney Int, 59, 179–89.

nephrotic syndrome

Shioji, R., Sasaki, Y., Saito, H., et al. (1974). Reversible tubular dysfunction associated with chronic renal failure in an adult patient with the nephrotic syndrome. Clin Nephrol, 2, 76–80. Singhal, R. and Brimble, K. S. (2006). Thromboembolic complications in the nephrotic syndrome: pathophysiology and clinical management. Thromb Res, 118, 397–407. Smith, J. D. and Hayslett, J. P. (1992). Reversible renal failure in the nephrotic syndrome. Am J Kidney Dis, 19, 201–3. Tessitore, N., Bonucci, E., D’Angelo, A., et al. (1984). Bone histology and calcium metabolism in patients with nephrotic syndrome and normal or reduced renal function. Nephron, 37, 153–9. Turner, A. N. (2010). Dropsy, Nephrosis, Nephrotic Syndrome. [Online] Ulinski, T., Leroy, S., Dubrel, M., et al. (2008). High serological response to pneumococcal vaccine in nephrotic children at disease onset on high-dose prednisone. Pediatr Nephrol, 23, 1107–13. Uncu, N., Bülbül, M., Yildiz, N., et al. (2010). Primary peritonitis in children with nephrotic syndrome: results of a 5-year multicenter study. Eur J Pediatr, 169, 73–6. Vande Walle, J. G., Donckerwolcke, R. A., van Isselt, J. W., et al. (1995). Volume regulation in children with early relapse of minimal-change nephrosis with or without hypovolaemic symptoms. Lancet, 346, 148–52. Vande Walle, J. G., Mauel, R., Raes, A., et al. (2004). ARF in children with minimal change nephrotic syndrome may be related to functional changes of the glomerular basal membrane. Am J Kidney Dis, 43, 399–404. Vaziri, N. D. (1993). Endocrinological consequences of the nephrotic syndrome. Am J Nephrol, 13, 360–4. Venkataseshan, V. S., Faraggiana, T., Grishman, E., et al. (1993). Renal failure due to tubular obstruction by large protein casts in patients with massive proteinuria. Clin Nephrol, 39, 321–6. Waldman, M., Crew, R. J., Valeri, A., et al. (2007). Adult minimal-change disease: clinical characteristics, treatment, and outcomes. Clin J Am Soc Nephrol, 2, 445–53. Wilcox, C. S. (2002). New insights into diuretic use in patients with chronic renal disease. J Am Soc Nephrol, 13, 798–805. Wong, W. (2007). Idiopathic nephrotic syndrome in New Zealand children, demographic, clinical features, initial management and outcome after twelve-month follow-up: results of a three-year national surveillance study. J Paediatr Child Health, 43(5), 337–41. Wu, C. H., Ko, S. F., Lee, C. H., et al. (2006). Successful outpatient treatment of renal vein thrombosis by low-molecular weight heparins in 3 patients with nephrotic syndrome. Clin Nephrol, 65, 433–40. Wu, H. M., Tang, J. -L., Cao, L., et al. (2012). Interventions for preventing infection in nephrotic syndrome. Cochrane Database Syst Rev, 4, CD003964.

495

CHAPTER 53

Pathophysiology of oedema in nephrotic syndrome Neil Turner and Premil Rajakrishna Introduction

ENaC is overactive in nephrotic syndrome

For many decades it was taught that the drop in colloid osmotic pressure was the primary driver to sodium retention in nephrotic syndrome. This was presumed to lead to fluid egress into the interstitial space, and thus to a reduction of intravascular fluid volume—the ‘underfill’ hypothesis. Problems with this model have been highlighted over a long period. Notably there is not consistent evidence that the circulation is actually underfilled. The contrary ‘overfill’ hypothesis proposes that sodium retention is the primary problem. A major problem with this hypothesis was lack of a mechanism to explain it. There is now a plausible mechanism, but it does not on its own fully explain findings in nephrotic syndrome.

The epithelial sodium channel (ENaC), the amiloride-sensitive sodium channel, is the major route by which sodium is reabsorbed in the collecting duct. In animal models, amiloride inhibited sodium retention in nephrotic syndrome, but mRNA levels for the channel were not affected suggesting that the overactivity was not due to increased production. In fact the mRNA levels of other sodium transporters in the nephron (NHE3, Na/K-ATPase, NCC, NKCC2; see Chapter 21) tended to be reduced, a fact which might help to explain diuretic resistance in nephrotic syndrome (summarized by Svenningsen et al., 2013, 2015).

Key observations on sodium retention

ENaC activity may be controlled by moving preformed channel molecules to the apical cell membrane, a process influenced by aldosterone and vasopressin; or by proteolytic activation of the channel which increases the proportion of time it spends in the ‘open’ configuration. Proteolytic activation occurs in the Golgi apparatus intracellularly, but it emerged that serine proteases such as plasmin could activate the channels on the cell surface, and that plasmin activity could be found in nephrotic urine. Nephrotic urine was shown to contain urokinase-type plasminogen activator which could explain the activation of filtered plasminogen (Passero et al., 2008; Svenningsen et al., 2009).

Evidence for a primary renal origin of sodium retention has been accumulating for over three decades, until in 2008/2009 evidence for a convincing mechanism was presented. Key points are summarized here. For more detailed bibliography see comprehensive reviews by Siddall and Radhakrishnan (2012) and Svenningsen et al. (2013).

Renal sodium retention is a local consequence of proteinuria A landmark finding was made by Ichikawa et al. (1983) when they created a unilateral nephrotic lesion in a rat by treating a single kidney with puromycin aminonucleoside (PAN). PAN produces a podocyte injury mimicking focal segmental glomerulosclerosis (FSGS) (see Chapter 45), including a severe nephrotic syndrome followed later by loss of glomerular filtration rate. In their experiments the kidney with proteinuria also retained sodium, whereas the undamaged kidney handled both protein and sodium normally. They went on to show by in vivo micropuncture studies that the location of increased sodium reabsorption was distal to the distal convoluted tubule. This suggested that the mechanism of sodium retention was local, not related to circulating hormones, intimately related to proteinuria, and that the abnormality was probably located in the collecting duct. A number of subsequent experimental studies using different rodent models of nephrotic syndrome (e.g. Adriamycin® (doxorubicin) in mice, mercuric chloride in rats) confirmed the localization of sodium retention to the collecting duct.

ENaC activity is increased by proteolytic activation

Sodium retention commences before serum albumin drops This explanation helps to explain the sequence of events long observed in acute nephrotic syndrome (Koomans, 2003; Siddall and Radhakrishnan, 2012), which is (1)  proteinuria, (2)  sodium retention, then (3) serum protein levels fall. In other words, sodium retention commences before albumin levels have changed substantially. When remission occurs, the abnormalities switch off in the same sequence, sodium excretion rising before serum albumin levels have shown a rise (Fig. 53.1).

Activation of ENaC alone does not replicate nephrotic syndrome Patients with Liddle syndrome (see Chapter  21) have mutations leading to constitutive activation of ENaC. They retain sodium but become hypertensive, not oedematous. Additional hypotheses to

Chapter 53 

Prednisolone 60 mg/day.

40 mg/day.

20 mg/day.

10 mg/day.

Weight kg.

15 13 11

Jrine volume L/24 hr.

Protein excretion g/24 hr.

9 4 3 2 1

pathophysiology of oedema in nephrotic syndrome gradient. Interstitial albumin concentrations fall virtually to zero if serum albumin falls very low (Koomans, 2003; Siddall and Radhakrishnan, 2012). However, there is a limit to how low interstitial COP can go, so that in very severe nephrotic syndrome, there may indeed be an adverse gradient, and fluid may be lost from the circulation by this mechanism. But in mild and moderate nephrotic syndrome, oedema occurs without reduced circulating volume. The homeostatic mechanisms maintaining the COP gradient are not fully understood. There is a second exception to this. If serum protein concentration falls very rapidly, there is a period of increased movement of fluid into the interstitial space, until homeostatic mechanisms have restored the COP gradient. This is most likely in the very acute, very severe proteinuria that can occur in minimal change disease.

Blood volume is not generally low in nephrotic syndrome

3 2 1 0

4

8 12 16 20 24 28 32 36 40 Days after onset of treatment

Fig. 53.1  Treatment of nephrotic syndrome in a child in the 1950s, before effective diuretics were available and soon after the discovery of the effect of corticosteroids in childhood nephrotic syndrome. Weight falls from 15 to 10 kg over about a week as the protein leak responds to steroid therapy. Note that the spontaneous diuresis commences as soon as the level of proteinuria drops below 1 g and before albumin levels could have recovered. From DeWardener (1958), with permission.

explain this at present centre on capillary permeability or deranged osmotic forces (see below).

Inhibition of ENaC as a therapeutic strategy The effect of amiloride alone in human proteinuric conditions has not been the subject of much published research, although it is known to potentiate the action of loop diuretics. Interestingly it may also have inhibitory effects on urokinase-like plasminogen activator (Svenningsen et al., 2013). But there may be serine protease inhibitors which could inhibit plasmin production in nephrotic urine, and that will be a fascinating drug target.

Key observations on the circulation According to the ‘underfill’ hypothesis, reduced circulatory volume is a drive to sodium retention. Circulatory volume can be low but this is not a general feature of the syndrome; indeed, hypertension can be a feature. This begs the question, why is circulating volume not low, surely Starling forces will lead to egress of fluid into the interstitial space if serum proteins are decreased?

Interstitial colloid osmotic pressure is also low For fluid to leave the circulation by osmotic pressure in nephrotic syndrome, the osmotic gradient between circulation and the interstitium must change. However observations in animal models and in man suggest that interstitial colloid osmotic pressure (COP) falls in parallel with the fall in serum COP, preserving the

Neither direct measurements of blood volume, nor levels of renin, aldosterone, or other hormones suggest that hypovolaemia is a general feature of nephrotic syndrome. When hypovolaemia is identified it is in patients with the most severe hypoproteinaemia. Modest hypervolaemia is a common finding. Blood pressure is normal or increased and tends to fall after steroid-induced recovery in minimal change disease in both adults and children (Koomans, 2003). These observations may help to explain why albumin infusions do not generally improve natriuresis substantially in patients with nephrotic syndrome (Fliser et al., 1999; see Chapter 52).

But blood volume can be low in some circumstances The above paragraphs pointed out that low circulating volume can occur in patients with very severe nephrotic syndrome, particularly if it is of recent and rapid onset. This combination is most frequently found in minimal change disease, and matches the clinical observation that hypovolaemia is clinically most likely to be encountered in this group. It is most likely to cause acute kidney injury in adult patients with minimal change disease (see Chapter 52). The important implication of these observations are that individual patients must be assessed individually, a point made long ago (Schrier and Fassett, 1998).

Extracellular fluid distributes differently in nephrotic syndrome Koomans et  al. (1986) made a revealing comparison between patients with nephrotic syndrome and patients with chronic kidney disease (CKD) who each had extracellular fluid (ECF) volume expansion. Those with CKD had hypertension and increased blood volume with ECF volume about double normal. Those with nephrotic syndrome had greater (threefold) expansion of ECF volume but did not have increased blood pressure. So both groups had ECF volume expansion, but the fluid tended to be distributed more towards the intravascular compartment in patients with CKD, versus the interstitial compartment in patients with nephrotic syndrome. This explains how patients with nephrotic syndrome can tolerate greater ECF accumulation than patients with reduced cardiac or renal function.

497

498

Section 3  

the patient with glomerular disease

Vascular permeability

References

A generalized abnormality of vascular permeability is one of the explanations that could help to explain the differences between nephrotic syndrome and other causes of increased extracellular fluid volume. There are some observations to support this (Rostoker et al., 2000; Siddall and Radhakrishnan, 2012).

De Wardener, H. E. (1958). The Kidney (1st ed.). London: Churchill. Fliser, D., Zurbrüggen, I., Mutschler, E., et al. (1999). Coadministration of albumin and furosemide in patients with the nephrotic syndrome. Kidney Int, 55, 629. Ichikawa, I., Rennke, H. G., Hoyer, J. R., et al. (1983). Role for intrarenal mechanisms in the impaired salt excretion of experimental nephrotic syndrome. J Clin Invest, 71, 91–103. Koomans, H. A. (2003). Pathophysiology of oedema in idiopathic nephrotic syndrome. Nephrol Dial Transplant, 18 Suppl 6, vi30–2. Koomans, H. A., Braam, B., Geers, A. B., et al. (1986). The importance of plasma protein for blood volume and blood pressure homeostasis. Kidney Int, 30, 730–5. Passero, C. J., Mueller, G. M., Rondon-Berrios, H., et al. (2008). Plasmin activates epithelial Na+ channels by cleaving the gamma sub-unit. J Biol Chem, 283, 36586–91. Rostoker, G., Behar, A., and Lagrue, G. (2000). Vascular hyperpermeability in nephrotic edema. Nephron, 85, 194–200. Schrier, R. W. and Fassett, R. G. (1998). A critique of the overfill hypothesis of sodium and water retention in the nephrotic syndrome. Kidney Int, 53, 1111–7. Siddall, E. C. and Radhakrishnan, J. (2012). The pathophysiology of edema formation in the nephrotic syndrome. Kidney Int, 82, 635–42. Svenningsen, P., Andersen, H., Nielsen, L. H., et al. (2015). Urinary serine proteases and activation of EnaC in kidney – implications for physiological renal salt handling and hypertensive disorders with albuminuria. Pflugers Arch, 467, 531–42. Svenningsen, P., Bistrup, C., Friis, U. G., et al. (2009). Plasmin in nephrotic urine activates the epithelial sodium channel. J Am Soc Nephrol, 20, 299–310. Svenningsen, P., Friis, U. G., Versland, J. B., et al. (2013). Mechanisms of renal NaCl retention in proteinuric disease. Acta Physiol, 207, 536–45.

Other abnormalities Atrial natriuretic peptide resistance is reported in nephrotic syndrome, and levels are often low or normal. Arginine vasopressin (antidiuretic hormone) levels are high and may contribute to water retention. Increased sympathetic activity is described in animal models, but these tend to create severe nephrotic syndrome.

Mechanisms in specific diseases There is no reason to think that the mechanism of oedema is different in different diseases that cause nephrotic syndrome. In the case of FSGS there is some evidence for a circulating factor affecting glomerular permeability (though it may simply be a podocyte-toxic factor) (Chapter  57). It has been suggested that this factor might also affect permeability of other membranes, for example, leading to increased protein loss through peritoneal dialysis.

Comparison with other causes of oedema Chapter 30 discusses and compares oedema in other contexts.

CHAPTER 54

Idiopathic nephrotic syndrome: overview Patrick Niaudet and Alain Meyrier Introduction Idiopathic nephrotic syndrome is defined by the combination of massive proteinuria, hypoalbuminaemia, hyperlipidaemia, and oedema and of non-specific histological abnormalities of the glomeruli. Light microscopy may disclose minimal changes (Chapter 55), diffuse mesangial proliferation, or focal and segmental glomerular sclerosis (FSGS) (Chapter 57). On electron microscopy the glomerular capillaries show a fusion of visceral epithelial cell (podocyte) foot processes and with the exception of some variants no significant deposits of immunoglobulins or complement by immunofluorescence. This excludes other idiopathic glomerulopathies, such as membranous glomerulopathy, that are in some publications lumped together with MCD and FSGS under the umbrella denomination of ‘INS’. In a majority of children only minimal changes are seen on light microscopy. These children are referred to as having ‘minimal-change disease’. In adults with idiopathic nephrotic syndrome, lesions of FSGS are more frequent. A question remains:  do minimal change disease (MCD) and FSGS represent two facets of the same disease, or distinct pathophysiologic entities?

The unitary view The unitary view is the most appropriate regarding treatment options. This is especially true of paediatricians. The unitary view is compatible with varied causes and/or pathophysiological mechanisms. Rather than distinguishing FSGS from MCD on a kidney biopsy, the best guide to prognosis and to subsequent response to other drugs is the initial response to glucocorticoids. Patients with FSGS generally suffer a more severe disease, are often resistant to corticosteroids, and are prone to progressing to renal failure. In the early stages FSGS and MCD are histologically indistinguishable. The best illustration of this is the appearance of a kidney biopsy carried out shortly after relapse of nephrotic syndrome following transplantation in a patient whose primary renal disease was FSGS. Despite heavy proteinuria, the glomeruli initially show minimal changes. A number of patients with FSGS respond to steroids whilst some steroid-resistant patients have no sclerotic changes on adequate biopsies. Therefore, some nephrologists believe that, although histological variants of the idiopathic nephrotic syndrome carry prognostic significance, they cannot at present be considered as separate entities. In fact, considering that corticosteroid (CS) responsiveness has a bearing on the response to other treatments, such as

the efficacy of rituximab, one cannot exclude the hypothesis that CS-sensitive as opposed to CS-resistant MCD and FSGS might be due to different pathophysiologic mechanisms.

The pluralistic view More recent data have tended to reinforce the concept that MCD and FSGS are different entities. MCD appears to be a functional podocyte disease whereas FSGS clearly appears to be a structural podocytopathy (Kriz et al., 1994; Barisoni et al., 2007) that develops, amongst other subsets, to the cell and the scar variants of the glomerular lesion and less frequently to a highly cellular and shrinking appearance of the glomerular tuft, ‘collapsing glomerulopathy’. The notion of ‘podocyte dysregulation’ (Barisoni et  al., 1999; Bariety et  al., 2001), the different expression of cyclin-dependent kinase inhibitors in MCD and in FSGS, the role of these cell cycle disturbances leading to podocyte proliferation and maturation (Shankland et al., 2000), are in favour of distinct entities. The finding that intrarenal transcription of cytotoxic T-lymphocyte effectors and transforming growth factor beta 1 is increased in a majority of children with FSGS, contrasting with the rare occurrence of this phenomenon in children with MCD is in favour of the pluralistic view (Strehlau et al., 2002). Moreover the fact that overexpression of interleukin 13 induces minimal change nephropathy and no FSGS in rats (Lai et  al., 2007), whereas in the human high circulating levels of the soluble urokinase receptor are found in FSGS but not in MCD (Wei et al., 2011; Jefferson and Alpers, 2013)  lend support to interpreting MCD and FSGS as podocytopathies originating from different glomerular permeability factors. On these grounds minimal change disease and focal segmental glomerulosclerosis are covered separately. Minimal change disease is considered in chapters 55 and 56, FSGS in chapters 57 and 58, and pathogenesis of these conditions in chapter 59. Matters more specifically paediatric as opposed to those related to adults will be as much as possible distinguished, which does not avoid some repetitions.

References Bariety, J., Bruneval, P., Hill, G., et al. (2001). Posttransplantation relapse of FSGS is characterized by glomerular epithelial cell transdifferentiation. J Am Soc Nephrol, 12, 261–74. Barisoni, L., Kriz, W., Mundel, P., et al. (1999). The dysregulated podocyte phenotype: a novel concept in the pathogenesis of collapsing idiopathic focal segmental glomerulosclerosis and HIV-associated nephropathy. J Am Soc Nephrol, 10, 51–61.

500

Section 3  

the patient with glomerular disease

Barisoni, L., Schnaper, H. W., and Kopp, J. B. (2007). A proposed taxonomy for the podocytopathies: a reassessment of the primary nephrotic diseases. Clin J Am Soc Nephrol, 2, 529–42. Jefferson, J. A. and Alpers, C. E. (2013). Glomerular disease: 'suPAR'exciting times for FSGS. Nat Rev Nephrol, 9, 127–8. Kriz, W., Elger, M., Nagata, M., et al. (1994). The role of podocytes in the development of glomerular sclerosis. Kidney Int Suppl, 45, S64–72. Lai, K. W., Wei, C. L., Tan, L. K., et al. (2007). Overexpression of interleukin-13 induces minimal-change-like nephropathy in rats. J Am Soc Nephrol, 18, 1476–85.

Shankland, S. J., Eitner, F., Hudkins, K. L., et al. (2000). Differential expression of cyclin-dependent kinase inhibitors in human glomerular disease: role in podocyte proliferation and maturation. Kidney Int, 58, 674–83. Strehlau, J., Schachter, A. D., Pavlakis, M., et al. (2002). Activated intrarenal transcription of CTL-effectors and TGF-beta1 in children with focal segmental glomerulosclerosis. Kidney Int, 61, 90–5. Wei, C., El Hindi, S., Li, J., et al. (2011). Circulating urokinase receptor as a cause of focal segmental glomerulosclerosis. Nat Med, 17, 952–60.

CHAPTER 55

Minimal change disease: clinical features and diagnosis Patrick Niaudet and Alain Meyrier Epidemiology The incidence of idiopathic nephrotic syndrome varies with age, race, and geography. The annual incidence in children in the United States has been estimated to be 2–2.7/100,000 (McEnery and Strife, 1982) with a cumulative prevalence of 16/100,000. In the United Kingdom, the incidence of idiopathic nephrotic syndrome is sixfold greater in Asian than in European children (Sharples et  al., 1985). Whereas idiopathic nephrotic syndrome accounts for only 25% of adult cases (Cameron et  al., 1974b), it is by far the most common cause of nephrotic syndrome in children. The International Study of Kidney Disease in Children found minimal change disease (MCD) in 76.6% of children with primary nephrotic syndrome, with a male/female ratio of 2/1 (International Study of Kidney Disease in Children, 1978).

Conditions with a possible aetiologic role Many factors are commonly cited as possible ‘causes’ or temporally associated conditions for MCD (Glassock, 2003). They include viral diseases, allergies, drugs, vaccinations, and some malignancies. The prevalence of positive Epstein–Barr virus DNA detection and recent infection or reactivation is higher in children at onset of idiopathic nephrotic syndrome compared to a population matched for age, gender, and time of sampling (Dossier et al., 2014). It is not easy to understand what final common pathway permits these differing factors to result in the common clinical and pathological outcome of MCD, or how this relates to pathogenesis as outlined below.

Allergy Allergy is associated with up to 30% of cases, which suggests some involvement of type IV reactions in the pathogenesis of MCD. An allergic episode is often followed by a relapse of the nephrotic syndrome. Amongst a list of anecdotal cases, the allergens reported include fungi, poison ivy, ragweed and timothy grass pollen, house dust, medusa stings, bee stings, and cat fur. A food allergen may be responsible for relapsing episodes of steroid-sensitive nephrosis, such as cows’ milk and eggs. Laurent et al. evaluated the effect of an oligoantigenic diet given for 10–15 days to 13 patients with an unsatisfactory response to corticosteroids (Laurent et al., 1987). This diet coincided with an improvement of proteinuria in nine, including complete remission in five.

Drugs A number of drugs may induce a nephrotic syndrome with the histopathological appearance of MCD. The list (see Chapter 82) comprises unrelated drugs such as non-steroidal anti-inflammatory agents, including sulfasalazine and mesalazine, D-penicillamine, lithium, rifampicin, heavy metals (gold, mercury), and trimethadione.

Malignancy The association with malignancies mainly concerns lymphomatous disorders and rarely solid tumours. MCD has been associated with haematologic malignancies, such as Hodgkin lymphoma, non-Hodgkin lymphoma, and leukaemia (Alpers and Cotran, 1986; Dabbs et al., 1986; Audard et al., 2006). MCD has been reported in 0.4% of patients with Hodgkin lymphoma (Dabbs et al., 1986). Audard et al. reported that MCD appeared prior to the discovery of the lymphoma in 8 out of 21 patients (Audard et  al., 2006). In such cases, MCD was either steroid dependent or steroid resistant. The haemopathy is already apparent at the time of onset of MCD or is diagnosed simultaneously in the other cases. In most but not all patients, remission of proteinuria is obtained with the cure of lymphoma. Solid tumours are more commonly associated with membranous nephropathy (Alpers and Cotran, 1986). However, some patients with solid tumours may develop MCD. These include thymoma, renal cell carcinoma, mesothelioma, and bronchogenic, colon, bladder, lung, breast, pancreatic, duodenal, and prostate cancer (Meyrier et al., 1992; Auguet et al., 1998; Glassock, 2003). Removal of the tumour may be followed by remission of the nephrotic syndrome.

Inheritance The familial occurrence of MCD is well known. White found that 3.3% of 1877 patients with idiopathic nephrotic syndrome had affected family members mainly siblings (White, 1973). This inherited trend is different from the genetic forms of focal segmental glomerulosclerosis (FSGS) that are described in Chapter 327.

Histocompatibility antigens A three- to fourfold increased incidence of human leucocyte antigen (HLA)-DR7 in nephrotic children has been reported (Alfiler et al., 1980; de Mouzon-Cambon et al., 1981). Clark et al. found a strong association between HLA-DR7 and the DQB1 gene of HLA-DQW2 and steroid-sensitive nephrosis, and suggested that DR7 and

502

Section 3  

the patient with glomerular disease

DQW2 contribute to disease susceptibility (Clark et  al., 1990). HLA-DR3 has been associated with steroid-resistant nephrosis in children, with a relative risk of 3. The incidence of HLA-DR3-DR7 is increased in steroid-resistant patients with a relative risk of 9.3. An association with HLA-B8 was reported in Europe. Children with atopy and HLA-B12 have a 13-fold increased risk of developing nephrosis.

Clinical features in children The disease is rarely discovered on routine urine analysis, and oedema is the most frequent presenting symptom (see Chapter 52). The onset is usually rapid and even ‘explosive’. Oedema increases gradually and becomes clinically detectable when fluid retention exceeds 3–5% of body weight. It is often initially apparent around the eyes and can be misdiagnosed as allergy. During the day, periorbital oedema decreases whilst it localizes at the lower extremities. In the reclining position, it localizes on the back. It is white, soft, and pitting. Oedema of the scrotum and penis, or labia, may also be observed. Anasarca (severe generalized oedema) may develop. Abdominal pain may result from ascites, severe hypovolaemia, peritonitis, pancreatitis, thrombosis, or steroid-induced gastritis. Shock with abdominal pain and peripheral circulatory failure may follow a sudden fall of plasma albumin and requires emergency treatment. Blood pressure is usually normal, but sometimes elevated. The disease may also be revealed by a complication. Peritonitis due to Streptococcus pneumoniae is a classical mode of onset. Other infections include meningitis, cellulitis, and pneumonia. Deep vein or arterial thromboses and pulmonary embolism may also occur during the first attack or during a relapse (see Chapter 52).

Clinical features in adults The clinical picture in adults is similarly characterized by generalized oedema of sudden onset. Contrary to FSGS in which proteinuria apparently precedes oedema for a period of time of uncertain duration (with the exception of the ‘tip lesion’ variant (Stokes et al., 2004)), oedema sets in in a matter of days. However, hypovolaemic shock and abdominal pain are quite unusual. Blood pressure is moderately elevated in about half of the cases.

Box 55.1  Definitions with regard to the nephrotic syndrome and its response to treatment ◆ Nephrotic range proteinuria: • Adults: proteinuria > 3.5 g per day • Children: > 40 mg/m2 per hour; urinary protein:creatinine ratio > 2 mg/mg or > 200 mg/mmol ◆ Complete remission: • Adults: proteinuria < 0.3–1.0 g per day, normal serum albumin (> 30 g/L), and stable renal function • Children: urinary protein:creatinine ratio < 0.2–0.3 mg/mg or < 30 mg/mmol and normal serum albumin (> 30 g/L) ◆ Partial remission: • Adults: proteinuria 0.3–3.5 g per day and/or ≥ 50% decrease in proteinuria from baseline, and stable renal function • Children: urinary protein:creatinine ratio 0.2–2.0 mg/mg or 30–350 mg/mmol; and serum albumin > 30 g/L ◆ Steroid-dependent nephrotic syndrome: • Two consecutive relapses whilst receiving predniso(lo)ne on alternate days, or within 15 days of its discontinuation ◆ Steroid-resistant nephrotic syndrome: • Children:  lack of remission despite 4–8 weeks of therapy with daily predniso(lo)ne at a dose of 60 mg/m2 or 2 mg/kg (maximum 60 mg) per day • Adults: lack of remission despite 4 months of therapy with daily prednisone at a dose of 1 mg/kg/day (maximum 80 mg/day) ◆ Calcineurin-inhibitor (CNI) dependent nephrotic syndrome: • Remission of steroid-dependent nephrotic syndrome is achieved during therapy with CNIs (ciclosporin or tacrolimus) ◆ CNI-resistant and steroid-resistant nephrotic syndrome: • No response to therapy with predniso(lo)ne as defined above, or to CNI therapy.

Laboratory abnormalities Urine analysis Nephrotic-range proteinuria is defined as > 50 mg/kg/day or 40 mg/hour/ m2 in children and > 3.5 g/24 hours in adults (see Chapter 50). In children, the urinary protein:creatinine ratio or urinary albumin:creatinine ratio are useful (Box 55.1). For these two indices, the nephrotic range is 200–400 mg/mmol. In minimal change, steroid-sensitive nephrotic syndrome, proteinuria consists mainly of albumin and low-molecular-weight proteins, whilst in severe nephrotic syndrome with glomerular lesions and steroid resistance the urine also contains globulins. This can be quantified by means of the selectivity index, that is, the ratio of immunoglobulin (Ig)-G to albumin or transferrin clearance (see Chapter  52). A  favourable index would be < 0.05; a poor index > 0.15–0.20. There is a considerable overlap in results and the test has a limited value, especially in adults, and is rarely performed. Some children with severe steroid-resistant nephrotic syndrome

have both glomerular and tubular proteinuria. Macroscopic haematuria is rare, occurring in 1% of steroid responders and in 3% of non-responders. Persistent microscopic haematuria is more common, and may be observed in up to 30% of patients, with no particular histopathologic or prognostic significance.

Blood chemistries Serum proteins are markedly reduced, to < 50 g/L. Albumin concentration is usually < 20 g/L and may be < 10 g/L. Electrophoresis of plasma proteins shows a typical pattern with low albumin, increased alpha-2 globulins and, to a lesser extent, beta globulins whilst gamma globulins are decreased. IgG is considerably decreased, IgA slightly, and IgM is increased. Amongst other proteins, fibrinogen, and lipoproteins are increased, whilst small molecules such as antithrombin III are lost in the urine and their concentration in plasma is decreased. A  detailed analysis of the

Chapter 55 

minimal change disease: clinical features

hyperlipidaemia of nephrotic syndrome and other complications is to be found in Chapter 52. Serum electrolytes are usually within the normal range. Low plasma sodium may be related to dilution from inappropriate renal water retention. Mild hyponatraemia may be an artefact related to hyperlipidaemia. Serum calcium is low as a result of hypoalbuminaemia. Ionized calcium may be decreased in persistent nephrotic syndrome, due to urinary loss of 25-hydroxyvitamin D3. Blood urea and serum creatinine are often within the normal range, or increased in relation to functional renal insufficiency.

picture. However, the main indication is failure to respond to a 4-week course of prednisone given and taken in adequate dosage. A biopsy may be necessary to allow assessment of nephrotoxicity in patients receiving ciclosporin. Since MCD is much less common in adults, a kidney biopsy is in most cases performed before any treatment and should be repeated in case of resistance to a 4-month course of steroids. In such a case, this repeat biopsy usually discloses lesions of FSGS that had either been overlooked on initial histology, or have appeared since, explaining steroid resistance.

Haematology

Histopathology

Haemoglobin and the haematocrit may be increased in patients with a reduced plasma volume. Microcytic anaemia may be observed in chronic, steroid-resistant nephrotic syndrome, in some cases following urinary loss of transferrin (see Chapter 57). Thrombocytosis is common.

Light microscopy shows three patterns: minimal changes, diffuse mesangial proliferation, or FSGS. Their relative incidence is difficult to determine. Minimal changes are found in the majority of children, and FSGS in only 5–7% of them (Southwest Pediatric Nephrology Study Group, 1985). Mesangial proliferation is reported in a small number (3–5%) of patients. These proportions are different in adults, less than one-third of whom have MCD and a majority focal segmental sclerosis (see Chapter 52).

Hypercoagulability Hypercoagulability is a common feature of all forms of severe nephrotic syndrome. It is worth recalling here that before the first effective therapeutic measures became available, that is, adrenocorticotrophic hormone and glucocorticoids, thromboembolic events represented a major cause of mortality in children with nephrosis.

Renal function Renal function is usually normal, but some patients have a reduction of the glomerular filtration rate (GFR) attributed to hypovolaemia, with return to normal after remission. A reduced GFR may also be found despite normal effective plasma flow (Dorhout et al., 1979; Bohlin and Berg, 1984) with a rapid return to normal after remission. Tubular functions are occasionally altered with glycosuria, aminoaciduria, hypokalaemia, and acidosis.

Acute kidney injury complicating nephrosis Marked oliguria occurs mainly in adults, particularly in middle-aged or older patients (Cameron et al., 1974b). It may also occur in children (Sakarcan et al., 1994). Bilateral renal vein thrombosis may be recognized by sonography. Interstitial nephritis has been reported, usually allergic in response to drugs (see Chapter 101). Acute kidney injury (see Chapter 52) is usually reversible, it may be one of those rare justifications for intravenous infusion of albumin (Fliser et al., 1999; and see Chapter 52). In some cases where glomerular structure is close to normal on initial histology, acute kidney injury may last for as long as a year (Sakarcan et al., 1994) and sometimes be irreversible (Raij et al., 1976).

Kidney biopsy Indications Kidney biopsy is not indicated at onset in a child 1–10 years old with typical symptoms and complete remission obtained by corticosteroids. Biopsy is, however, indicated at onset in circumstances suggesting another type of glomerular disease, including moderate nephrotic syndrome or a long previous course of minor proteinuria, macroscopic haematuria, marked hypertension, and/or persistent renal insufficiency. A  decreased plasma C3 fraction is also an indication for performing a biopsy. Age < 12 months and > 11  years is another indication, even in patients with a typical

Light microscopy Minimal glomerular changes Under light microscopy the glomeruli are mostly normal. Mild changes, including podocyte swelling and vacuolation, a slight increase in mesangial matrix, and mild, focal, mesangial hypercellularity may be seen (Churg et al., 1970; Cameron et al., 1974b). Lipid vacuoles and degenerative changes of proximal tubules are rare. Scattered foci of tubular lesions and interstitial fibrosis may be observed, such as obstruction by hyaline casts, dilatation with epithelial cell thinning, tubular basement membrane thickening, interstitial foam cells, and calcium deposits. Vascular changes are absent in children. In adults they are age related (Cameron et al., 1974b).

Diffuse mesangial proliferation It is not always easy to draw the line between ‘mild’ and ‘marked’ mesangial hypercellularity. A  subset of patients shows a marked increase in mesangial matrix associated with hypercellularity (Churg et al., 1970; Waldherr et al., 1978). However, peripheral capillary walls are normal, and immunofluorescence does not show humps. Electron microscopy shows foot-process effacement (Fig. 55.1). Mesangial hypercellularity has been attributed a prognostic significance. Waldherr et al. found a higher rate of initial steroid resistance and of progression to renal failure (Waldherr et  al., 1978). Other studies failed to confirm these findings (Southwest Pediatric Nephrology Study Group, 1985).

Electron microscopy Ultrastructural changes are constant, mainly involving podocytes and mesangial stalks. Podocyte foot-process effacement is generalized (Fig. 55.2) and closely related to the degree of proteinuria. This flattening of foot processes is due to a reversible rearrangement of the podocyte actin cytoskeleton that affects an elongated disposition. Immunoelectron microscopy has shown that the expression of nephrin is lower than normal in regions where the foot processes are effaced (Huh et al., 2002). Other epithelial changes consist of microvilli formation and numerous protein reabsorption droplets. The glomerular basement membranes are normal. The endothelial cells are often swollen.

503

504

Section 3  

the patient with glomerular disease

Fig. 55.1  Diffuse mesangial proliferation: the glomerular basement membrane is normal but there is an increased number of mesangial cells and amount of mesangial matrix. (Masson’s trichrome, 250×.)

Immunofluorescence patterns IgM-associated nephropathy Cohen et al. proposed that mesangial IgM be considered a separate entity ‘IgM-associated nephropathy’ (Cohen et al., 1978). Patients with IgM nephropathy are less likely to respond to immunosuppressive agents than those with MCD and a higher probability of developing end-stage renal disease (Border, 1988; O’Donoghue et al., 1991). However, Habib et al. found that even if IgM was the immunoglobulin most frequently found in the glomeruli, there was no relationship between these deposits and initial response to therapy or final outcome (Habib et al., 1988). In fact, IgM is a large molecule that can be non-specifically trapped in an injured glomerulus and it is likely that mesangial IgM deposits represent an epiphenomenon. IgA and minimal change disease Some patients with nephrosis show mesangial deposits of IgA (Lai et al., 1986). It is likely that mesangial IgA in patients with minimal changes without cellular proliferation is coincidental (Barbiano di Belgiojoso et al., 1986; Habib et al., 1988). These patients have a favourable response to steroids, which would not be the case in true Berger’s disease.

Fig. 55.2  Minimal change disease (electron microscopy). The glomerular basement membrane is normal; the cytoplasm of the podocytes is vacuolated, with effacement of foot processes and microvilli. (Methenamine silver, 2800×.)

C1q glomerulopathy C1q glomerulopathy refers to a disorder in which mesangial proliferation is associated with mesangial deposits on electron microscopy and prominent C1q deposits on immunofluorescence microscopy (Jennette and Hipp, 1985; Iskandar et al., 1991; Markowitz et al., 2003; Kersnik Levart et al., 2005) and no clinical and laboratory evidence of systemic lupus erythematosus. C1q nephropathy has been thought to be a subgroup of primary FSGS. However, many reports describe different symptoms, histopathologies, therapeutic responses, and prognoses, suggesting that C1q glomerulopathy may be a combination of several disease groups rather than a single disease entity (Mii et al., 2009). C1q glomerulopathy may be associated with either MCD, FSGS, or proliferative glomerulonephritis.

References Alfiler, C. A., Roy, L. P., Doran, T., et al. (1980). HLA-DRw7 and steroid-responsive nephrotic syndrome of childhood. Clin Nephrol, 14, 71–4. Alpers, C. E., and Cotran, R. S. (1986). Neoplasia and glomerular injury. Kidney Int, 30, 465–73. Audard, V., Larousserie, F., Grimbert, P., et al. (2006). Minimal change nephrotic syndrome and classical Hodgkin's lymphoma: report of 21 cases and review of the literature. Kidney Int, 69, 2251–60. Auguet, T., Lorenzo, A., Colomer, E., et al. (1998). Recovery of minimal change nephrotic syndrome and acute renal failure in a patient with renal cell carcinoma. Am J Nephrol, 18, 433–5. Barbiano Di Belgiojoso, G., Mazzucco, G., Casanova, S., et al. (1986). Steroid-sensitive nephrotic syndrome with mesangial IgA deposits: a separate entity? Observation of two cases. Am J Nephrol, 6, 141–5. Bohlin, A. B., and Berg, U. (1984). Renal water handling in minimal change nephrotic syndrome. Int J Pediatr Nephrol, 5, 93–8. Border, W. A. (1988). Distinguishing minimal-change disease from mesangial disorders. Kidney Int, 34, 419–34. Cameron, J. S., Turner, D. R., Ogg, C. S., et al. (1974b). The nephrotic syndrome in adults with 'minimal change' glomerular lesions. QJM, 43, 461–88. Churg, J., Habib, R., and White, R. H. (1970). Pathology of the nephrotic syndrome in children: a report for the International Study of Kidney Disease in Children. Lancet, 760, 1299–302. Clark, A. G., Vaughan, R. W., Stephens, H. A., et al. (1990). Genes encoding the beta-chains of HLA-DR7 and HLA-DQw2 define major susceptibility determinants for idiopathic nephrotic syndrome. Clin Sci (Lond), 78, 391–7. Cohen, A. H., Border, W. A., and Glassock, R. J. (1978). Nehprotic syndrome with glomerular mesangial IgM deposits. Lab Invest, 38, 610–19. Dabbs, D. J., Striker, L. M., Mignon, F., et al. (1986). Glomerular lesions in lymphomas and leukemias. Am J Med, 80, 63–70. De Mouzon-Cambon, A., Bouissou, F., Dutau, G., et al. (1981). HLA-DR7 in children with idiopathic nephrotic syndrome. Correlation with atopy. Tissue Antigens, 17, 518–24. Dorhout, E. J., Roos, J. C., Boer, P., et al. (1979). Observations on edema formation in the nephrotic syndrome in adults with minimal lesions. Am J Med, 67, 378–84. Dossier, C., Sellier-Leclerc, A. L., Rousseau, A., et al. (2014). Prevalence of herpesviruses at onset of idiopathic nephrotic syndrome. Pediatr Nephrol, 29(12), 2325–31. Fliser, D., Zurbruggen, I., Mutschler, E., et al. (1999). Coadministration of albumin and furosemide in patients with the nephrotic syndrome. Kidney Int, 55, 629–34. Glassock, R. J. (2003). Secondary minimal change disease. Nephrol Dial Transplant, 18 Suppl 6, vi52–8. Habib, R., Girardin, E., Gagnadoux, M. F., et al. (1988). Immunopathological findings in idiopathic nephrosis: clinical significance of glomerular “immune deposits”. Pediatr Nephrol, 2, 402–8.

Chapter 55 

Huh, W., Kim, D. J., Kim, M. K., et al. (2002). Expression of nephrin in acquired human glomerular disease. Nephrol Dial Transplant, 17, 478–84. International Study of Kidney Disease in Children (1981). The primary nephrotic syndrome in children. Identification of patients with minimal change nephrotic syndrome from initial response to prednisone. A report of the International Study of Kidney Disease in Children. J Pediatr, 98, 561–4. Iskandar, S. S., Browning, M. C., and Lorentz, W. B. (1991). C1q nephropathy: a pediatric clinicopathologic study. Am J Kidney Dis, 18, 459–65. Jennette, J. C., and Hipp, C. G. (1985). C1q nephropathy: a distinct pathologic entity usually causing nephrotic syndrome. Am J Kidney Dis, 6, 103–10. Kersnik Levart, T., Kenda, R. B., Avgustin Cavic, M., et al. (2005). C1Q nephropathy in children. Pediatr Nephrol, 20, 1756–61. Lai, K. N., Lai, F. M., Chan, K. W., et al. (1986). An overlapping syndrome of IgA nephropathy and lipoid nephrosis. Am J Clin Pathol, 86, 716–23. Laurent, J., Rostoker, G., Robeva, R., et al. (1987). Is adult idiopathic nephrotic syndrome food allergy? Value of oligoantigenic diets. Nephron, 47, 7–11. Markowitz, G. S., Schwimmer, J. A., Stokes, M. B., et al. (2003). C1q nephropathy: a variant of focal segmental glomerulosclerosis. Kidney Int, 64, 1232–40. McEnery, P. T., and Strife, C. F. (1982). Nephrotic syndrome in childhood. Management and treatment in patients with minimal change disease, mesangial proliferation, or focal glomerulosclerosis. Pediatr Clin North Am, 29, 875–94.

minimal change disease: clinical features

Meyrier, A., Delahousse, M., Callard, P., et al. (1992). Minimal change nephrotic syndrome revealing solid tumors. Nephron, 61, 220–3. Mii, A., Shimizu, A., Masuda, Y., et al. (2009). Current status and issues of C1q nephropathy. Clin Exp Nephrol, 13, 263–74. O’Donoghue, D. J., Lawler, W., Hunt, L. P., et al. (1991). IgM-associated primary diffuse mesangial proliferative glomerulonephritis: natural history and prognostic indicators. QJM, 79, 333–50. Raij, L., Keane, W. F., Leonard, A., et al. (1976). Irreversible acute renal failure in idiopathic nephrotic syndrome. Am J Med, 61, 207–14. Sakarcan, A., Timmons, C., and Seikaly, M. G. (1994). Reversible idiopathic acute renal failure in children with primary nephrotic syndrome. J Pediatr, 125, 723–7. Sharples, P. M., Poulton, J., and White, R. H. (1985). Steroid responsive nephrotic syndrome is more common in Asians. Arch Dis Child, 60, 1014–7. Southwest Pediatric Nephrology Study Group (1985). Focal segmental glomerulosclerosis in children with idiopathic nephrotic syndrome. A report of the Southwest Pediatric Nephrology Study Group. Kidney Int, 27, 442–9. Stokes, M. B., Markowitz, G. S., Lin, J., et al. (2004). Glomerular tip lesion: a distinct entity within the minimal change disease/focal segmental glomerulosclerosis spectrum. Kidney Int, 65, 1690–702. Waldherr, R., Gubler, M. C., Levy, M., et al. (1978). The significance of pure diffuse mesangial proliferation in idiopathic nephrotic syndrome. Clin Nephrol, 10, 171–9. White, R. H. (1973). The familial nephrotic syndrome. I. A European survey. Clin Nephrol, 1, 215–19.

505

CHAPTER 56

Minimal change disease: treatment and outcome Patrick Niaudet and Alain Meyrier Introduction The principles of management of minimal change disease (MCD) are similar in children and adults with a key difference being whether or when to undertake a renal biopsy: ◆ Children

with nephrotic syndrome are usually treated with steroids first, and a renal biopsy only undertaken if there is no response to corticosteroids. However, a biopsy should be undertaken if there are atypical features such as macroscopic haematuria, hypertension, or lasting renal function impairment, all of which argue against a diagnosis of minimal change disease.

◆ In

adults, the pre-treatment probability of MCD is lower, and the response to treatment often slower. A kidney biopsy is indicated to decide on treatment.

General management The general management of nephrotic syndrome is described in Chapter 52.

Treatment and outcome in children Specific treatment The majority of children with minimal changes are steroid responsive (White et al., 1970). Steroid responders may relapse, but the majority still responds over the subsequent course (Pollak et al., 1968; Habib and Kleinknecht, 1971). Only 1–3% of patients who are initially steroid sensitive subsequently become steroid resistant and are defined as ‘late non-responders’ (Habib and Kleinknecht, 1971; Siegel et al., 1974; Trainin et al., 1975). Conversely, patients who do not respond to an initial steroid regimen given at an adequate dosage usually remain non-responders although they may respond to a combination of prednisone and calcineurin inhibitors (Niaudet, 1994). Response to corticosteroids is therefore the best prognostic index, not only since non-responders are more exposed to complications of persistent nephrotic syndrome, but also (and mainly) because they may develop end-stage renal disease (ESRD) after several years.

Initial treatment: corticosteroids Steroids should not be started too early as spontaneous remission may occur in 5% of cases within the first 8–15 days. Some of these early spontaneous remissions are definitive, others are not.

Infection must be sought and treated before starting steroids, not only to prevent the risk of overwhelming sepsis during treatment, but also because occult infection may be responsible for steroid resistance (McEnery and Strife, 1982). Metronidazole must not be forgotten for preventing malignant Strongyloides stercoralis infestation in patients from countries where carriage of this worm is endemic. Steroid therapy is started when the diagnosis of idiopathic nephrotic syndrome is most likely in a child or after renal biopsy has been performed. Prednisone remains the reference drug. Prednisolone has the advantage of being soluble in water, making treatment easier in young children, and is the standard drug in many countries. The International Study of Kidney Disease in Children (ISKDC) regimen consists of prednisone, 60 mg/m2/day with a maximum of 80 mg/day, in divided doses for 4 weeks followed by 40 mg/ m2/day with a maximum of 60 mg/day, on alternate days, for 4 weeks. A  response occurs in most cases within 10–15  days (median 11 days). According to the ISKDC (1981), approximately 90% of responders enter into remission within 4 weeks after starting steroids whilst < 10% enter remission after 2–4 more weeks of a daily regimen. Prolongation of daily steroid treatment beyond 4 or 5 weeks increases the risk of side effects. An alternative for patients who are not in remission after 4 weeks consists of three to four pulses of methylprednisolone (1 g/1.73 m2). This additional regimen seems to be associated with few side effects, and probably produces remission more rapidly in the few patients who would have entered the second month of daily therapy (Murnaghan et al., 1984). A working committee recently developed the following Kidney Disease: Improving Global Outcomes (KDIGO) guidelines (Lombel et  al., 2013). Initial prednisone therapy consists of 60 mg/m2or 2 mg/kg per day for 6 weeks (maximum dose of 60 mg/day), followed by alternate-day prednisone of 40 mg/m2 or 1.5 mg/kg (maximum dose of 40 mg/day) for an additional 6 weeks. A longer duration of the initial course of steroids appears to reduce the risk of relapse. A meta-analysis concluded that initial steroid therapy should be of at least 3 months duration in children with idiopathic nephrotic syndrome to reduce the risk of subsequent relapse. In a pooled analysis from six trials, treatment with prednisone for 3–7  months versus a 2-month regimen reduced the risk of relapse at 12–24 months post-therapy (relative risk (RR) 0.70; 95%

Chapter 56 

confidence interval (CI) 0.58–0.84). There was no difference in cumulative steroid dose. Similarly, in a pooled analysis of four trials of 382 children, the risk of relapse was lower with 6 versus 3 months of therapy (RR 0.57; 95% CI 0.45–0.71). There was no difference in cumulative steroid dose. There was an inverse linear relationship between treatment duration and the risk of relapse (Hodson et al., 2000, 2007). A longer duration is more important than the cumulative dose of prednisone in reducing the risk of relapse. This RR decreases by 0.133 (13%) for every additional month of treatment up to 7  months (Hodson et  al., 2000). There are no data showing that treating for > 7 months is beneficial. However, two recent studies have shown no benefit of a prolonged initial course of prednisolone therapy in reducing the rate of relapse. In one trial, 181 children received an initial treatment with a 6-week course of prednisolone 2 mg/kg per day followed by 1.5 mg/kg on alternate days for 6 weeks and were randomized to receive either placebo or decreasing doses of prednisolone for an additional duration of 3 months (Sinha et al., 2015). There was no difference in the mean number of relapses between both groups or in the rate of relapses, although the mean time to first relapse was later in the group receiving prednisolone for 6 months. The second study from Japan included 255 children who received in the control group a daily dose of 60 mg/m2 for 4 weeks followed by 40 mg/m2 on alternate days for 4 weeks or the same initial treatment followed by a progressive decrease for a total duration of 6 months. There were no differences in the mean number of relapses per patient-year between both groups, or in the rate of relapse after 2 years of follow-up (Yoshikawa et al., 2015).

Corticosteroid-responsive minimal change disease in children Most children with  idiopathic nephrotic syndrome respond to  steroid therapy but a majority of  them experience relapses. A  prospective cohort study showed that boys are more likely to  respond initially, more likely to  relapse, and to  be classified as having frequently relapsing nephrotic syndrome (Sureshkumar et al., 2014). About 30% of patients experience only one attack and are definitively cured after a single course of steroids. Persistent remission for 18–24 months after stopping treatment is likely to reflect definitive cure, and the risk of later relapses is low. About 10–20% of patients relapse several months after stopping treatment and are apparently free of disease after three or four episodes, which respond to a standard course of corticosteroids. The remaining 50–60% experience frequent relapses as soon as steroid therapy is stopped or when dosage is decreased. The risk of relapse is greater in children aged < 5 years at onset and in males. These steroid-dependent patients often raise difficult therapeutic problems. Steroid-dependent patients may be treated with repeated courses of prednisone, 60 mg/m2/day, continued 3 days after the urine has become protein free, followed by alternate-day prednisone, 40 mg/m2, for 4 weeks as proposed by the ISKDC (1981). Another option is based on treating relapses with daily prednisone, 40–60 mg/m2, until proteinuria has disappeared for 4–5  days. Thereafter, prednisone is switched to alternate days and the dosage is tapered to 15–20 mg/m2 every other day, according to the steroid threshold, that is, the dosage at which the relapse has occurred. Treatment is then continued for 12–18 months. The first approach allows better

minimal change disease: treatment and outcome

definition in terms of relapses but is associated with more relapses. The latter regimen is associated with fewer steroid side effects as the cumulative dosage is lower. Prolonged courses of alternate-day steroid therapy are often well tolerated by young children and growth velocity is not affected. However, prednisone dosage must be as low as possible in order to reduce side effects. The role of upper respiratory tract infections in exacerbating nephrotic syndrome has been highlighted in all series: 71% of relapses were preceded by such an event in a prospective study, although only 45% of respiratory infections were followed by an exacerbation of proteinuria (MacDonald et al., 1986). Gulati et al. performed a randomized controlled trial and found that the risk of relapse was significantly decreased during upper respiratory tract infections when prednisone was given daily for 7 days rather than on alternate days (Gulati et al., 2011). Leisti and Koskimies studied the degree of adrenocortical suppression and found that adrenocortical suppression increased the risk of relapse (Leisti and Koskimies, 1983). Severe suppression was always associated with a relapse, the longest remission time being 6 months. In patients with moderate suppression, several long, relapse-free intervals were observed, but the risk of relapse was still higher than in episodes with normal adrenocortical function. Cortisol substitution possibly decreased the risk of a relapse after severe adrenocortical suppression. These findings might incite to add adrenocortical substitution to a long-term corticosteroid regimen for entertaining remission in patients with idiopathic nephrotic syndrome.

Alternative treatments An alternative treatment is required in children who relapse on alternate-day prednisone therapy and suffer severe side effects such as growth retardation, behaviour disturbances, cushingoid features, hypertension, cataract, or osteopenia. Such treatment is also indicated in children at risk of toxicity such as diabetes or during puberty, in children with severe relapses accompanied by thrombotic complications or severe hypovolaemia, and in those with poor compliance. Alternative treatments include levamisole, alkylating agents, mycophenolate mofetil (MMF), ciclosporin, and rituximab (RTX).

Levamisole The beneficial effect of levamisole was first described by Tanphaichitr et al. (1980) and was subsequently reported to reduce the risk of relapse in steroid-dependent patients. A significant steroid-sparing effect at a dose of 2.5 mg/kg every other day was demonstrated in two controlled trials (British Association for Paediatric Nephrology, 1991; Dayal et al., 1994). However, the beneficial effect of levamisole is not sustained after stopping treatment. Side effects occasionally include neutropenia, agranulocytosis, vomiting, cutaneous rash, neurological symptoms including insomnia, hyperactivity, and seizures. Levamisole is well tolerated in most children but is not widely available.

Alkylating agents Alkylating agents, such as cyclophosphamide and chlorambucil, have been used for > 50 years to achieve long-lasting remission. Unfortunately they are toxic, and in the long term, remissions may not seem long enough. The efficacy of alkylating agents is illustrated by a meta-analysis that compared alkylating agents with prednisone in maintaining remission. In three trials including 102 patients, oral

507

508

Section 3  

the patient with glomerular disease

cyclophosphamide reduced the risk of relapse at 6–12  months (RR 0.44; 95% CI 0.26–0.73). Similarly, chlorambucil compared with prednisone alone reduced the risk of relapse at 12  months in 32 children (RR 0.13; 95% CI 0.03–0.57) (Hodson et al., 2008). Several studies involving patients with frequently relapsing or steroid-dependent nephrotic syndrome showed that cyclophosphamide resulted in sustained remission in 57–93% of patients at 1 year, 31–66% at 5 years, and 25% at 10 years (Cameron et al., 1974a; Chiu and Drummond, 1974; McDonald et al., 1974; Vester et al., 2005). However, a more recent series reported lower remission rates of 44%, 27%, and 13% at 1, 2, and 5  years after cyclophosphamide therapy (Cammas et al., 2011). In another study of 90 children with a steroid-dependent course, sustained remissions were observed in 31% of patients at 5-year follow-up (Azib et al., 2011). These variations are probably due to differences in the patient populations as steroid-dependent patients have a lower response rate than frequently relapsing patients. The degree of steroid dependency also affects remission rates (Zagury et al., 2011). The Arbeitsgemeinschaft für pädiatrische Nephrologie (1987) reported that treatment for 12 weeks at a daily dose of 2 mg/kg was more effective than an 8-week course, with 67% as compared to 22% remaining in remission after 2 years. However, a randomized trial showed that prolonging the course of cyclophosphamide from 8 to 12 weeks did not further reduce the proportion of children experiencing relapses (Ueda et al., 1990). Cyclophosphamide toxicity includes bone marrow depression, haemorrhagic cystitis, gastrointestinal disturbances, alopecia, and infection (Latta et al., 2001). Leucopoenia is frequently observed, but weekly haematological monitoring may limit its severity and concomitant steroids help blunt marrow depression. Haemorrhagic cystitis rarely occurs. Alopecia, which is variably pronounced, remits a few weeks after stopping treatment. Viral infections can be overwhelming if cyclophosphamide is not stopped in due time. Long-term toxicity includes malignancy, even more rarely pulmonary fibrosis. Gonadal toxicity is well established and the risk of sterility is greater in boys than in girls. The cumulative threshold dose above which oligo/azoospermia may be feared lies between 150 and 250 mg/kg (Penso et al., 1974; Hsu et al., 1979; Trompeter et al., 1981). In females, the cumulative dose associated with sterility is greater, but not well defined. In this and other contexts, early menopause may be a late consequence of alkylating agents. Most authors would prescribe a 12-week course of oral cyclophosphamide at a daily dose of 2 mg/kg. Beneficial results have also been achieved with chlorambucil in steroid-responsive nephrosis (Grupe et  al., 1976; Baluarte et  al., 1978; Williams et al., 1980). Acute and long-term toxic effects are similar to those observed with cyclophosphamide.

Calcineurin inhibitors Ciclosporin Ciclosporin has been shown in a number of studies to reduce the incidence of relapses in 75–90% of patients with steroid-dependent nephrotic syndrome thereby allowing withdrawal of prednisone (Kitano et  al., 1990; Niaudet and Habib, 1994; Gregory et  al., 1996; Inoue et  al., 1999; El-Husseini et  al., 2005; Tanaka et  al., 2006; Cattran et al., 2007; Ishikura et al., 2008, 2010). However, most patients relapse when the drug is withdrawn (Ishikura

et al., 2012). Therefore a prolonged treatment is necessary with an increased risk of nephrotoxicity (Niaudet and Habib, 1994; Hulton et al., 1994). Ciclosporin has been compared to alkylating agents in two randomized trials (Niaudet, 1992; Ponticelli et al., 1993a). The effect of ciclosporin was initially the same as chlorambucil and cyclophosphamide in maintaining remission. However, after ciclosporin was discontinued, it was less effective in maintaining remission at 12 months compared with either alkylating agents and at 24 months for chlorambucil. Because of the concern for nephrotoxicity, the serum creatinine concentration should be monitored regularly in patients who are maintained on a long-term course of ciclosporin. However, serial renal biopsies demonstrate histologic lesions of nephrotoxicity without clinical evidence of renal function impairment (Habib and Niaudet, 1994; Iijima et al., 2002; Kengne-Wafo et al., 2009). Histological lesions most often consist of tubulointerstitial injury, characterized by stripes of interstitial fibrosis containing clusters of atrophic tubules and by lesions of arteriolopathy. Thus, some authors propose to routinely perform a kidney biopsy in asymptomatic patients after 18 months of ciclosporin therapy. Other side effects include hypertension, hyperkalaemia, hypertrichosis, gum hypertrophy, and hypomagnesaemia. The recommended starting ciclosporin dose is 150 mg/m2 per day divided into two oral doses. The dose should be adjusted to maintain trough whole blood levels between 100 and 200 ng/mL, and the level should not exceed 200 ng/mL. In order to limit the risk of nephrotoxicity, once remission is achieved, we recommend decreasing the dose to < 5 mg/kg, if possible. Low-dose alternate-day prednisone in combination with ciclosporin may be a good approach to maintain remission with lower doses of ciclosporin.

Tacrolimus Though data is not so comprehensive, tacrolimus is probably as effective as ciclosporin in maintaining remission in children with steroid-sensitive nephrotic syndrome, Transplantation experience (see Chapter 281) suggests that it is less nephrotoxic but more likely to be associated with diabetes (Dotsch et  al., 2006; Sinha et al., 2006). In a series of five children treated with tacrolimus, two developed type 1 diabetes mellitus, which resolved after stopping tacrolimus therapy (Dittrich et al., 2006). However, one advantage of tacrolimus over ciclosporin is reduced cosmetic side effects (hypertrichosis, gum hypertrophy).

Mycophenolate mofetil MMF inhibits T- and B-cell proliferation. Small studies suggest that MMF is effective in increasing the duration of remission in children with idiopathic nephrotic syndrome; however, relapses often occur after the treatment is discontinued in steroid-dependent children (Novak et al., 2005; Hogg et al., 2006; Afzal et al., 2007). In a small randomized trial comparing MMF to ciclosporin, sustained complete remission was achieved in 7 of 12 patients who received MMF and in 11 of 12 patients treated with ciclosporin suggesting that ciclosporin is more effective than MMF (Dorresteijn et  al., 2008). Side effects of MMF include gastrointestinal disturbances (abdominal pain and diarrhoea) and haematological abnormalities.

Chapter 56 

A ‘Bayesian study’ was conducted in 23 children with steroid-dependent idiopathic nephrotic syndrome having received prior alkylating-agent treatment and two-thirds of them levamisole (Baudouin et al., 2012). They were treated with MMF and prednisone according to a defined schedule (reduction of alternate-day dose to 50% of pre-MMF dose at 3  months, 25% at 6  months). Twenty-three children completed the study. Four relapsed during the first 6 months and two at months 8 and 11.5. In the 19 patients free of relapse during the first 6 months, median prednisone maintenance dose decreased from 25 (10–44) to 9 (7.5–11.2) mg/m2 and cumulative dose from 459 (382–689) to 264 (196–306) mg/m2 before and on MMF respectively. The authors concluded that MMF reduces relapse rate and steroid dose in children with steroid-dependent nephrotic syndrome and should be proposed before ciclosporin. Whilst awaiting further information, MMF appears to be reasonable choice to treat children with steroid-dependent nephrotic syndrome with steroid toxicity. Although two trials comparing ciclosporin and MMF suggest that ciclosporin is more effective in preventing relapses (Dorresteijn et al., 2008; Gellermann et al., 2013), MMF may still be more appropriate as it is not nephrotoxic

Rituximab Several reports originally suggested that RTX, a chimeric anti-CD20 monoclonal antibody that depletes B-cell lymphocytes, may be effective in steroid-dependent or calcineurin inhibitor-dependent patients (Guigonis et al., 2008; Fujinaga et al., 2010; Ravani et al., 2011; Sellier-Leclerc et al., 2011). Larger and now randomised studies have confirmed this. In a case series of 54 children with severe steroid or ciclosporindependent nephrotic syndrome, the administration of RTX allowed the discontinuation of one or more immunosuppressive agents (Guigonis et al., 2008). The response to RTX appears to be better if the patient is in remission at the time of the infusion. Seven patients were nephrotic at the time of RTX treatment. Remission was induced in three of the seven proteinuric patients. RTX was effective in all patients when administered during a proteinuria-free period in association with other immunosuppressive agents. Therefore, RTX should not be administered during a relapse but after remission has been induced by increased doses of steroids. One or more immunosuppressive treatments could be withdrawn in 19 patients, with no relapse of proteinuria and without increasing other drug dosage. When relapses occurred, they were associated with an increase in CD19 cell count. Adverse effects were observed in 45% of cases, but most of them were mild and transient. Ravani et al. randomized 54 children (mean age 11 ± 4 years) with idiopathic nephrotic syndrome dependent on prednisone and calcineurin inhibitors for > 12 months (Ravani et al., 2011). RTX with lower doses of prednisone and calcineurin inhibitors was compared to current therapy alone. Three-month proteinuria was 70% lower in the RTX arm as compared with standard therapy arm (intention-totreat); relapse rates were 18.5% (intervention) and 48.1% (standard arm) (P = 0.029). Probabilities of being drug-free at 3 months were 62.9% and 3.7%, respectively (P < 0.001); 50% of RTX cases were in stable remission without drugs after 9 months. The authors concluded that RTX and lower doses of prednisone and calcineurin inhibitors are non-inferior to standard therapy in maintaining short-term remission in children with idiopathic nephrotic syndrome dependent on both drugs and allow their temporary withdrawal.

minimal change disease: treatment and outcome

Iijama et al. reported the results of a clinical trial involving 48 children with frequently relapsing steroid-dependent nephritic syndrome who were randomly assigned to receive rituximab or placebo once a week for 4 weeks whilst in remission (Iijima et al., 2014). Patients received standard steroid therapy and immunosuppressive agents were stopped 6  months after randomization. At 1 year, median relapse-free duration was longer in the rituximab group compared with controls (267 vs 101  days). Moreover, the relapse rate and the daily prednisone dose were lower although by 19 months, all patients had relapsed. A significant proportion of patients relapse after RTX administration. Most relapses occur after recovery of B lymphocyte counts. Maintenance therapy with MMF has been shown to be effective in preventing relapse after treatment with RTX (Ito et al., 2011). Fujinaga et al. found that ciclosporin was even more effective than MMF for maintaining remission after a single infusion of RTX (Fujinaga et al., 2013). However, RTX may be associated with adverse effects including infusion-related reactions (hypotension, fever, and rigors), serious infections due to leukopenia and/or hypogammaglobulinaemia (Delbe-Bertin et al., 2013; Sellier-Leclerc et al., 2013; Kamei et  al., 2015), and progressive multifocal leucoencephalopathy. In addition, there is one published case report of death due to lung fibrosis (Chaumais et al., 2009) associated with RTX therapy in a child with nephrotic syndrome and another one of severe myocarditis (Sellier-Leclerc et  al., 2013)  in two children with nephrotic syndrome treated with rituximab. Additional severe adverse effects reported in childhood nephrotic syndrome include Pneumocystis jirovecii pneumonia requiring heart transplantation (Guigonis et al., 2008; Sato et al., 2013) and severe immune-mediated ulcerative gastrointestinal disease (Ardelean et al., 2010). The KDIGO Glomerulonephritis Workgroup recommended that rituximab be considered only in children who have continuing relapses despite optimal combinations of prednisone and steroid-sparing agents or in patients who had developed calcineurin inhibitor nephrotoxicity (Lombel et al., 2013). Well-defined randomized controlled studies are mandatory to examine which drug has the best risk/benefit profile and to better define the place of rituximab in the treatment of childhood steroid-sensitive nephrotic syndrome (Boyer and Niaudet, 2013), knowing that rituximab does not cure the disease and repeated courses may be necessary (Razzak, 2014).

Corticosteroid-resistant minimal change disease in children Less than 5% of children with MCD are steroid resistant, either primary non-responders or late non-responders. Hammad et  al. found that patients who are primary non-responders have a low expression of glucocorticoid receptors on peripheral blood mononuclear cells before starting therapy, and they suggest that this low expression may be one of the pathophysiological mechanisms of steroid resistance in these children (Hammad et al., 2013). Patients with steroid-resistant MCD may develop in the long-term lesions of focal segmental glomerular sclerosis and progress to ESRD. Some of these children have a genetic form of idiopathic nephrotic syndrome and are at high risk of progression to ESRD. The treatment of these patients remains difficult. Several therapeutic options have been proposed. The results of these treatments are difficult to analyse, as a significant proportion of the patients

509

510

Section 3  

the patient with glomerular disease

progress to cure gradually, with a slow decrease of proteinuria and in such cases, relapses are rare. Intravenous cyclophosphamide pulses were found to be more effective than oral cyclophosphamide. In this trial, patients receiving cyclophosphamide pulses had more sustained remissions despite lower cumulative dose (Elhence et al., 1994). Another study found that sustained remission is more likely to occur in patients with late steroid résistance (Bajpai et al., 2003). The French Society of Pediatric Nephrology reported the results of a prospective study where 45 children with minimal change steroid-resistant nephrotic syndrome received a combination of ciclosporin and prednisone (Niaudet, 1994). Complete remission was observed in 21 patients and none of them progressed to ESRD. Six patients relapsed after ciclosporin was stopped but they responded to a second course of prednisone. Gulati et al. found that tacrolimus and prednisone are more effective than cyclophosphamide pulses in children with steroid-resistant nephrotic syndrome, including those with MCD (Gulati et al., 2012).

Treatment in adults There are hints that MCD is not the same illness in adults and in children. This is especially true in terms of progression and treatment.

Corticosteroid therapy The first-line treatment of MCD in adults is based on glucocorticoids. However considerable differences in treatment modes between adults and children have been reported in the literature (Wang et al., 1982; Nolasco et al., 1986; Nair et al., 1987; Korbet et  al., 1988a; Fujimoto et  al., 1991; Mak et  al., 1996; Nakayama et al., 2002). Definitions of steroid sensitivity, dependency, resistance, and multiple relapses vary among papers dealing with adult idiopathic nephrotic syndrome, owing to lack of unified agreement regarding treatment protocol. The response to treatment is currently defined as disclosed in Box 56.1. Initial response to corticosteroids does not seem to differ greatly according to age, although it appears to be somewhat lower in adults. The response of MCD to corticosteroids in adults is much slower than in children (Nolasco et al., 1986; Fujimoto et al., 1991; Korbet et al., 1994, 1995, 1999; Rydel et al., 1995; Mak et al., 1996; Nakayama et al., 2002) (Fig. 56.1). Fujimoto et  al. treated 33 patients having adult-onset MCD with prednisolone at 1 mg/kg per day for 4–8 weeks, followed by slow tailing off, for a total duration of > 6 months (Fujimoto et al., 1991). Seventy-six per cent of patients were pushed into remission within 8 weeks, but in the extant cases the longest time to remission was 4  months. Five patients went into remission whilst the steroid dosage was being tapered. Mak et al. treated 40 patients with adult-onset minimal change nephropathy. The remission rate was 46% by the fourth week, 69% by the eighth week, 85% by the sixteenth week, and 87% by the twenty-first week (Mak et al., 1996). These figures clearly show that corticosteroid resistance in adult nephrosis should not be pronounced before 4 months and probably even 6 months of full-dose (1 mg/kg/day) treatment. Another difference stems from the trend toward a diminished frequency of relapses with increasing age, provided idiopathic nephrotic syndrome did not start during childhood.

Box 56.1 Definitions ◆ Nephrotic range proteinuria: • Adults: proteinuria > 3.5 g per day • Children: > 40 mg/m2 per hour; urinary protein:creatinine ratio >2 mg/mg or >200 mg/mmol ◆ Complete remission: • Adults: proteinuria < 0.3–1.0 g per day, normal serum albumin (> 30 g/L), and stable renal function • Children: urinary protein:creatinine ratio < 0.2–0.3 mg/mg or < 30 mg/mmol and normal serum albumin (> 30 g/L) ◆ Partial remission: • Adults: proteinuria 0.3–3.5 g per day and/or ≥ 50% decrease in proteinuria from baseline, and stable renal function • Children: urinary protein:creatinine ratio 0.2–2.0 mg/mg or 30–350 mg/mmol; and serum albumin >30 g/L ◆ Steroid-dependent nephrotic syndrome: • Two consecutive relapses whilst receiving predniso(lo)ne on alternate days, or within 15 days of its discontinuation ◆ Steroid-resistant nephrotic syndrome: • Children:  lack of remission despite 4–8 weeks of therapy with daily predniso(lo)ne at a dose of 60 mg/m2 or 2 mg/kg (maximum 60 mg) per day • Adults: lack of remission despite 4 months of therapy with daily prednisone at a dose of 1 mg/kg/day (maximum 80 mg/day) ◆ Calcineurin-inhibitor (CNI) dependent nephrotic syndrome: • Remission of steroid-dependent nephrotic syndrome is achieved during therapy with CNIs (ciclosporin or tacrolimus) ◆ CNI-resistant and steroid-resistant nephrotic syndrome: • No response to therapy with predniso(lo)ne as defined above, or to CNI therapy. Nolasco et  al. followed 89 patients with adult-onset MCD, 58 of whom responded to corticosteroid treatment:  24% never relapsed, 56% relapsed on a single occasion or infrequently, and only 21% were frequent relapsers (Nolasco et al., 1986). Korbet et al. followed 40 adults with MCD: 34 were treated with prednisone, and 31 (91%) achieved remission, of whom only three suffered multiple relapses (Korbet et al., 1988a). Relapses were infrequent in 99 adult nephrotics with MCD followed by Wang et al. for 3–102 months (Wang et al., 1982). In 85 patients whose urine was protein-free for at least 6 months, four relapsed; of 46 who were protein-free for 24 months, three relapsed; of 37 followed for 36–96 months, only three relapsed. Nair et al. treated 54 adults with MCD. Only 17 had relapses, a rate of 31% at 3 years of follow-up (Nair et al., 1987). Fujimoto et al. also found an incidence of relapse significantly lower in patients > 30 years (Fujimoto et al., 1991). Nakayama et al. analysed retrospectively 62 Japanese adults with minimal change nephrotic syndrome.

Chapter 56 

100

1

Cumulative % of patients with complete remission

90

patient with a haemorrhagic gastrointestinal ulcer. In such cases first-line treatment may be based on a calcineurin inhibitor (Meyrier, 1997).

3

2

80

4

70

Corticosteroid dependency and multiple relapses

5

60

6

50 40 30 20 10 0

0

2

4

8

minimal change disease: treatment and outcome

16

28

Weeks from starting corticosteroid therapy

Fig. 56.1  Comparing the time of response to corticosteroid treatment in children (1) shows that the definition of corticosteroid resistance in adults (2–6) is in the order of 4 months. Many articles in the literature on the treatment of adult nephrosis are biased, as they adopt the same definition of ‘steroid resistance’ as for children, that is, 6–8 weeks of this regimen without remission. (1) International Study of Kidney Disease in Children (1981); (2) Fujimoto et al. (1991); (3) Mak et al. (1996); (4) Nakayama et al. (2002); (5) Nolasco et al. (1986); (6) Korbet et al. (1981). From Nakayama et al. (2002).

Five experienced remission spontaneously. Fifty-three entered complete remission, three partial remission, and one patient showed no response to corticosteroids (Nakayama et al., 2002). Fifty-three patients with complete remission were divided into two groups:  38 early responders who experienced remission completely within 8 weeks after starting treatment and 15 late responders who experienced remission after 8 weeks. Thirty-three patients experienced a relapse; 13 experienced multiple relapses. Fifty-three patients with remission were divided into three groups:  16 patients who experienced relapse within 6  months after the initial response (early relapsers), 17 who experienced relapse after 6  months (late-relapsers), and 20 non-relapsers. Mean age at onset was younger in early relapsers than in late or non-relapsers. Age at onset correlated inversely with relapse rate in 53 patients with remission and correlated positively with timing of the first relapse in 33 relapsers. Thus, the experience of nephrologists treating adult MCD is comparable in Europe, America, and the Far East, with a multiple relapse rate in the order of 10–20%, as opposed to a greater percentage in children. Apart from age, insufficient treatment might also explain some of the relapses observed in adults. In reports published between 1971 and 1988, patients were treated with a short course of steroids. The initial remission rate was comparable with a short versus a long treatment mode, but duration of remission was superior when patients received > 8 weeks of prednisone. A long initial alternate-day corticosteroid regimen followed by slow tapering is effective in obtaining sustained remission in adult MCD.

Other treatment modes Contraindications to corticosteroid therapy In some patients, high-dose corticosteroids are hazardous. This is the occasional case of a patient with morbid obesity and glucose intolerance, of a patient with psychiatric disorders or of a

Steroid dependency is defined by relapses occurring at the time when tapering treatment reaches a threshold dose. When this dose is high, for example, > 20 mg/day of prednisone the patient is exposed to all the long-term complications of steroids. Multiple relapses defined by three or four attacks per year despite an adequate initial treatment put the nephrologist in a similar quandary with respect to the tolerability of available drugs. There are no randomized series to determine the best approach for suppressing or spacing out the nephrotic episodes. A  first option is based on a continuous course of the smallest dose of prednisone that maintains remission, either on a daily or on an alternate-day treatment. The cumulative dosage of steroids, however, portends a risk of toxicity in the long run, in particular of hip osteonecrosis. Calcineurin inhibitors (ciclosporin or tacrolimus), alone or with a very small dose of prednisone, are quite effective but even with the small risk shown by serial biopsies the long term implications for kidney function are concerning minimal risk of renal toxicity as shown by repeat kidney biopsies (Meyrier et al., 1994; Cattran et al., 2007). A third option consists of a 4-month course of cyclophosphamide. This alkylating agent is credited with long-lasting remissions with a protracted ‘treatment holiday’. However, this alkylating agent is characterized by a narrow margin between efficacy and toxicity and in a young male entails a risk of definitive infertility. The efficacy of MMF in multirelapsing or steroid-dependent adult minimal change nephrotic syndrome has been documented in small series (Waldman et al., 2007; Siu et al., 2008; Hogan and Radhakrishnan, 2013). RTX offers a new hope to suppress the relapses or control steroid dependency in adults, although long term studies are still lacking to determine outcome and predictors of response and long term toxicity (Sinha and Bagga, 2013). Munyentwali et al. analysed the efficacy and safety of RTX in 17 adult patients with steroid-dependent MCD over a mean follow-up of 29.5 months (range 5.1–82  months) (Munyentwali et  al., 2013). Seventeen patients with steroid-dependent or frequently relapsing minimal change nephrotic syndrome, unresponsive to several immunosuppressive medications, were treated with RTX. Eleven patients had no relapses after RTX infusion (mean follow-up 26.7  months, range 5.1–82  months) and nine of them were able to come off all other immunosuppressive drugs and steroids during follow-up. Six patients relapsed at least once after a mean time of 11.9 months (mean follow-up 34.5 months, range 16.9–50.1  months), but their immunosuppressive drug treatment could be stopped or markedly reduced during this time. No adverse events were recorded. The authors concluded that RTX is efficient and safe in adult patients suffering from severe steroid-dependent MCD.

Late relapses A relapse occurring years after a first episode of nephrotic syndrome followed by a long period of remission is not unusual. In such a case it is not mandatory to commence a 4-month course of high-dose steroids. A stable remission is often achieved with a short

511

512

Section 3  

the patient with glomerular disease

(in the order of 1-month) treatment with rapid tapering of steroids to a stop over the following 4 weeks.

Corticosteroid resistance An adult with an initial diagnosis of MCD and resistance to a 4-month course of glucocorticoids represents an indication to perform a repeat kidney biopsy. In most cases histology reveals lesions of focal segmental glomerulosclerosis that have been overlooked on the first biopsy or have appeared since. In such a case the treatment is as described for patients with focal segmental glomerulosclerosis. A possible cause of apparent steroid resistance in patients treated with prednisolone is a concomitant treatment with aluminium gels for gastric protection, as (contrary to prednisone) the intestinal absorption of prednisolone is significantly diminished or abolished by these antacids.

Long-term outcome in children About one-third of patients experience only one attack and are definitively cured after the course of corticosteroids. Ten to 20% of patients experience relapses several months after stopping the treatment and a cure takes place after three or four episodes. The remaining 40–50% of patients have either frequent relapses or are steroid dependent. These steroid-dependent patients may have a prolonged course. However, if the patient continues to respond to steroids, the risk of progression to chronic renal failure is minimal. Trompeter et  al. reported the late outcome of 152 children steroid-responsive nephrotic syndrome after a follow-up of 14–19  years:  127 (83%) were in remission, four had hypertension, 10 were still relapsing, and 11 had died (Trompeter et  al., 1985). The duration of the disease was longer in children who had started before the age of 6 years. Wynn et al. found that 15% of 132 patients had a persistent relapsing course with a mean follow-up of 27.5 years (Wynn et al., 1988). Lewis et al. reported on 26 patients over the age of 20 years, of whom five were still relapsing in adulthood (Lewis et al., 1989). Koskimies et al. reported on 94 cases. Twenty-four per cent of steroid responders had no relapse, 22% had infrequent relapses and 54% frequent relapses. More than two-thirds were in remission at time of report (Koskimies et  al., 1982). None of these patients developed renal insufficiency and none died from the disease. Lahdenkari et al. reported a 30-year follow-up of the patients reported previously by Koskimies et al. (Lahdenkari et al., 2005). Of 104 patients, 10% had further relapses in adulthood. Fakhouri et  al. reported on the outcome in adulthood of 102 patients born between 1970 and 1975 (Fakhouri et  al., 2003). Forty-two per cent presented at least one relapse in adulthood. A young age at onset and a high number of relapses during childhood were associated with a higher risk of relapse in adulthood. Similarly, Ruth et al. in a study of 42 patients, found that 14 (33%) relapsed in adulthood (Ruth et al., 2005). The higher relapse rates in these two reports probably reflect patient selection with more steroid-dependent cases compared to Koskimies et al.’s series.

References Afzal, K., Bagga, A., Menon, S., et al. (2007). Treatment with mycophenolate mofetil and prednisolone for steroid-dependent nephrotic syndrome. Pediatr Nephrol, 22, 2059–65. Arbeitsgemeinschaft für Padiatrische Nephrologie (1987). Cyclophosphamide treatment of steroid dependent nephrotic

syndrome: comparison of eight week with 12 week course. Report of Arbeitsgemeinschaft fur Padiatrische Nephrologie. Arch Dis Child, 62, 1102–6. Ardelean, D. S., Gonska, T., Wires, S., et al. (2010). Severe ulcerative colitis after rituximab therapy. Pediatrics, 126, e243–6. Azib, S., Macher, M. A., Kwon, T., et al. (2011). Cyclophosphamide in steroid-dependent nephrotic syndrome. Pediatr Nephrol, 26, 927–32. Bajpai, A., Bagga, A., Hari, P., et al. (2003). Intravenous cyclophosphamide in steroid-resistant nephrotic syndrome. Pediatr Nephrol, 18, 351–6. Baluarte, H. J., Hiner, L., and Gruskin, A. B. (1978). Chlorambucil dosage in frequently relapsing nephrotic syndrome: a controlled clinical trial. J Pediatr, 92, 295–8. Baudouin, V., Alberti, C., Lapeyraque, A. L., et al. (2012). Mycophenolate mofetil for steroid-dependent nephrotic syndrome: a phase II Bayesian trial. Pediatr Nephrol, 27, 389–96. Boyer, O. and Niaudet, P. (2013). Nephrotic syndrome: rituximab in childhood steroid-dependent nephrotic syndrome. Nat Rev Nephrol, 9, 562–3. Cameron, J. S., Chantler, C., Ogg, C. S., et al. (1974a). Long-term stability of remission in nephrotic syndrome after treatment with cyclophosphamide. Br Med J, 4, 7–11. Cammas, B., Harambat, J., Bertholet-Thomas, A., et al. (2011). Long-term effects of cyclophosphamide therapy in steroid-dependent or frequently relapsing idiopathic nephrotic syndrome. Nephrol Dial Transplant, 26, 178–84. Cattran, D. C., Alexopoulos, E., Heering, P., et al. (2007). Cyclosporin in idiopathic glomerular disease associated with the nephrotic syndrome : workshop recommendations. Kidney Int, 72, 1429–47. Chaumais, M. C., Garnier, A., Chalard, F., et al. (2009). Fatal pulmonary fibrosis after rituximab administration. Pediatr Nephrol, 24, 1753–5. Chiu, J. and Drummond, K. N. (1974). Long-term follow-up of cyclophosphamide therapy in frequent relapsing minimal lesion nephrotic syndrome. J Pediatr, 84, 825–30. Dayal, U., Dayal, A. K., Shastry, J. C., et al. (1994). Use of levamisole in maintaining remission in steroid-sensitive nephrotic syndrome in children. Nephron, 66, 408–12. Delbe-Bertin, L., Aoun, B., Tudorache, E., et al. (2013). Does rituximab induce hypogammaglobulinemia in patients with pediatric idiopathic nephrotic syndrome? Pediatr Nephrol, 28, 447–51. Dittrich, K., Knerr, I., Rascher, W., (2006). Transient insulin-dependent diabetes mellitus in children with steroid-dependent idiopathic nephrotic syndrome during tacrolimus treatment. Pediatr Nephrol, 21, 958–61. Dorresteijn, E. M., Kist-Van Holthe, J. E., Levtchenko, E. N., et al. (2008). Mycophenolate mofetil versus cyclosporine for remission maintenance in nephrotic syndrome. Pediatr Nephrol, 23, 2013–20. Dotsch, J., Dittrich, K., Plank, C., et al. (2006). Is tacrolimus for childhood steroid-dependent nephrotic syndrome better than ciclosporin A? Nephrol Dial Transplant, 21, 1761–3. El-Husseini, A., El-Basuony, F., Mahmoud, I., et al. (2005). Long-term effects of cyclosporine in children with idiopathic nephrotic syndrome: a single-centre experience. Nephrol Dial Transplant, 20, 2433–8. Elhence, R., Gulati, S., Kher, V., et al. (1994). Intravenous pulse cyclophosphamide--a new regime for steroid-resistant minimal change nephrotic syndrome. Pediatr Nephrol, 8, 1–3. Fakhouri, F., Bocquet, N., Taupin, P., et al. (2003). Steroid-sensitive nephrotic syndrome: from childhood to adulthood. Am J Kidney Dis, 41, 550–7. Fujimoto, S., Yamamoto, Y., Hisanaga, S., et al. (1991). Minimal change nephrotic syndrome in adults: response to corticosteroid therapy and frequency of relapse. Am J Kidney Dis, 17, 687–92. Fujinaga, S., Hirano, D., Nishizaki, N., et al. (2010). Single infusion of rituximab for persistent steroid-dependent minimal-change nephrotic syndrome after long-term cyclosporine. Pediatr Nephrol, 25, 539–44. Fujinaga, S., Someya, T., Watanabe, T., et al. (2013). Cyclosporine versus mycophenolate mofetil for maintenance of remission of steroid-dependent nephrotic syndrome after a single infusion of rituximab. Eur J Pediatr, 172, 513–8.

Chapter 56 

Gellermann, J., Weber, L., Pape, L., et al. (2013). Mycophenolate mofetil versus cyclosporin A in children with frequently relapsing nephrotic syndrome. J Am Soc Nephrol, 24, 1689–97. Gregory, M. J., Smoyer, W. E., Sedman, A., et al. (1996). Long-term cyclosporine therapy for pediatric nephrotic syndrome: a clinical and histologic analysis. J Am Soc Nephrol, 7, 543–9. Grupe, W. E., Makker, S. P., and Ingelfinger, J. R. (1976). Chlorambucil treatment of frequently relapsing nephrotic syndrome. N Engl J Med, 295, 746–9. Guigonis, V., Dallocchio, A., Baudouin, V., et al. (2008). Rituximab treatment for severe steroid- or cyclosporine-dependent nephrotic syndrome: a multicentric series of 22 cases. Pediatr Nephrol, 23(8), 1269–79. Gulati, A., Sinha, A., Gupta, A., et al. (2012). Treatment with tacrolimus and prednisolone is preferable to intravenous cyclophosphamide as the initial therapy for children with steroid-resistant nephrotic syndrome. Kidney Int, 82, 1130–5. Gulati, A., Sinha, A., Sreenivas, V., et al. (2011). Daily corticosteroids reduce infection-associated relapses in frequently relapsing nephrotic syndrome: a randomized controlled trial. Clin J Am Soc Nephrol, 6, 63–9. Habib, R. and Kleinknecht, C. (1971). The primary nephrotic syndrome of childhood. Classification and clinicopathologic study of 406 cases. Pathol Annu, 6, 417–74. Habib, R. and Niaudet, P. (1994). Comparison between pre- and posttreatment renal biopsies in children receiving ciclosporine for idiopathic nephrosis. Clin Nephrol, 42, 141–6. Hammad, A., Yahia, S., Gouida, M. S., et al. (2013). Low expression of glucocorticoid receptors in children with steroid-resistant nephrotic syndrome. Pediatr Nephrol, 28, 759–63. Hodson, E. M., Knight, J. F., Willis, N. S., et al. (2000). Corticosteroid therapy in nephrotic syndrome: a meta-analysis of randomised controlled trials. Arch Dis Child, 83, 45–51. Hodson, E. M., Willis, N. S., and Craig, J. C. (2007). Corticosteroid therapy for nephrotic syndrome in children. Cochrane Database Syst Rev, CD001533. Hodson, E. M., Willis, N. S., and Craig, J. C. (2008). Non-corticosteroid treatment for nephrotic syndrome in children. Cochrane Database Syst Rev, CD002290. Hogan, J. and Radhakrishnan, J. (2013). The Treatment of Minimal Change Disease in Adults. J Am Soc Nephrol. Hogg, R. J., Fitzgibbons, L., Bruick, J., et al. (2006). Mycophenolate mofetil in children with frequently relapsing nephrotic syndrome: a report from the Southwest Pediatric Nephrology Study Group. Clin J Am Soc Nephrol, 1, 1173–8. Hsu, A. C., Folami, A. O., Bain, J., et al. (1979). Gonadal function in males treated with cyclophosphamide for nephrotic syndrome. Fertil Steril, 31, 173–7. Hulton, S. A., Neuhaus, T. J., Dillon, M. J., et al. (1994). Long-term cyclosporin A treatment of minimal-change nephrotic syndrome of childhood. Pediatr Nephrol, 8, 401–3. Iijima, K., Hamahira, K., Tanaka, R., et al. (2002). Risk factors for cyclosporine-induced tubulointerstitial lesions in children with minimal change nephrotic syndrome. Kidney Int, 61, 1801–5. Iijima, K., Sako, M., Nozu, K., et al. (2014). Rituximab for childhood-onset, complicated, frequently relapsing nephrotic syndrome or steroid-dependent nephrotic syndrome: a multicentre, double-blind, randomised, placebo-controlled trial. Lancet, 384(9950), 1273–81. Inoue, Y., Iijima, K., Nakamura, H., et al. (1999). Two-year cyclosporin treatment in children with steroid-dependent nephrotic syndrome. Pediatr Nephrol, 13, 33–8. International Study of Kidney Disease in Children. 1981. The primary nephrotic syndrome in children. Identification of patients with minimal change nephrotic syndrome from initial response to prednisone. A report of the International Study of Kidney Disease in Children. J Pediatr, 98, 561–4. Ishikura, K., Ikeda, M., Hattori, S., et al. (2008). Effective and safe treatment with cyclosporine in nephrotic children: a prospective, randomized multicenter trial. Kidney Int, 73, 1167–73.

minimal change disease: treatment and outcome

Ishikura, K., Yoshikawa, N., Hattori, S., et al. (2010). Treatment with microemulsified cyclosporine in children with frequently relapsing nephrotic syndrome. Nephrol Dial Transplant, 25, 3956–62. Ishikura, K., Yoshikawa, N., Nakazato, H., et al. (2012). Two-year follow-up of a prospective clinical trial of cyclosporine for frequently relapsing nephrotic syndrome in children. Clin J Am Soc Nephrol, 7, 1576–83. Ito, S., Kamei, K., Ogura, M., et al. (2011). Maintenance therapy with mycophenolate mofetil after rituximab in pediatric patients with steroid-dependent nephrotic syndrome. Pediatr Nephrol, 26, 1823–8. Kamei, K., Takahashi, M., Fuyama, M., et al. (2015). Rituximab-associated agranulocytosis in children with refractory idiopathic nephrotic syndrome: case series and review of literature. Nephrol Dial Transplant, 30, 91–6. Kengne-Wafo, S., Massella, L., Diomedi-Camassei, F., et al. (2009). Risk factors for cyclosporin A nephrotoxicity in children with steroid-dependant nephrotic syndrome. Clin J Am Soc Nephrol, 4, 1409–16. Kitano, Y., Yoshikawa, N., Tanaka, R., et al. (1990). Ciclosporin treatment in children with steroid-dependent nephrotic syndrome. Pediatr Nephrol, 4, 474–7. Korbet, S. M. (1995). Management of idiopathic nephrosis in adults, including steroid-resistant nephrosis. Curr Opin Nephrol Hypertens, 4, 169–76. Korbet, S. M. (1999). Clinical picture and outcome of primary focal segmental glomerulosclerosis. Nephrol Dial Transplant, 14 Suppl 3, 68–73. Korbet, S. M., Schwartz, M. M., and Lewis, E. J. (1988a). Minimal-change glomerulopathy of adulthood. Am J Nephrol, 8, 291–7. Korbet, S. M., Schwartz, M. M., and Lewis, E. J. (1994). Primary focal segmental glomerulosclerosis: clinical course and response to therapy. Am J Kidney Dis, 23, 773–83. Koskimies, O., Vilska, J., Rapola, J., et al. (1982). Long-term outcome of primary nephrotic syndrome. Arch Dis Child, 57, 544–8. Lahdenkari, A. T., Suvanto, M., Kajantie, E., et al. (2005). Clinical features and outcome of childhood minimal change nephrotic syndrome: is genetics involved? Pediatr Nephrol, 20, 1073–80. Latta, K., Von Schnakenburg, C., and Ehrich, J. H. (2001). A meta-analysis of cytotoxic treatment for frequently relapsing nephrotic syndrome in children. Pediatr Nephrol, 16, 271–82. Lewis, M. A., Baildom, E. M., Davis, N., et al. (1989). Nephrotic syndrome: from toddlers to twenties. Lancet, 1, 255–9. Lombel, R. M., Gipson, D. S., and Hodson, E. M. (2013). Treatment of steroid-sensitive nephrotic syndrome: new guidelines from KDIGO. Pediatr Nephrol, 28, 415–26. Macdonald, N. E., Wolfish, N., McLaine, P., et al. (1986). Role of respiratory viruses in exacerbations of primary nephrotic syndrome. J Pediatr, 108, 378–82. Mak, S. K., Short, C. D., and Mallick, N. P. (1996). Long-term outcome of adult-onset minimal-change nephropathy. Nephrol Dial Transplant, 11, 2192–201. McDonald, J., Murphy, A. V., and Arneil, G. C. (1974). Long-term assessment of cyclophosphamide therapy for nephrosis in children. Lancet, 2, 980–2. McEnery, P. T., and Strife, C. F. 1982. Nephrotic syndrome in childhood. Management and treatment in patients with minimal change disease, mesangial proliferation, or focal glomerulosclerosis. Pediatr Clin North Am, 29, 875–94. Meyrier, A. (1997). Treatment of idiopathic nephrotic syndrome with cyclosporine A. J Nephrol, 10, 14–24. Munyentwali, H., Bouachi, K., Audard, V., et al. (2013). Rituximab is an efficient and safe treatment in adults with steroid-dependent minimal change disease. Kidney Int, 83, 511–6. Murnaghan, K., Vasmant, D., and Bensman A. (1984). Pulse methylprednisolone therapy in severe idiopathic childhood nephrotic syndrome. Acta Paediatr Scand, 73, 733–9. Nair, R. B., Date, A., Kirubakaran, M. G., et al. (1987). Minimal-change nephrotic syndrome in adults treated with alternate-day steroids. Nephron, 47, 209–10.

513

514

Section 3  

the patient with glomerular disease

Nakayama, M., Katafuchi, R., Yanase, T., et al. (2002). Steroid responsiveness and frequency of relapse in adult-onset minimal change nephrotic syndrome. Am J Kidney Dis, 39, 503–12. Niaudet, P. (1992). Comparison of cyclosporin and chlorambucil in the treatment of steroid-dependent idiopathic nephrotic syndrome: a multicentre randomized controlled trial. The French Society of Paediatric Nephrology. Pediatr Nephrol, 6, 1–3. Niaudet, P. (1994). Treatment of childhood steroid-resistant idiopathic nephrosis with a combination of cyclosporine and prednisone. French Society of Pediatric Nephrology. J Pediatr, 125, 981–6. Niaudet, P., and Habib, R. (1994). Cyclosporine in the treatment of idiopathic nephrosis. J Am Soc Nephrol, 5, 1049–56. Nolasco, F., Cameron, J. S., Heywood, E. F., et al. (1986). Adult-onset minimal change nephrotic syndrome: a long-term follow-up. Kidney Int, 29, 1215–23. Novak, I., Frank, R., Vento, S., et al. (2005). Efficacy of mycophenolate mofetil in pediatric patients with steroid-dependent nephrotic syndrome. Pediatr Nephrol, 20, 1265–8. Penso, J., Lippe, B., Ehrlich, R., et al. (1974). Testicular function in prepubertal and pubertal male patients treated with cyclophosphamide for nephrotic syndrome. J Pediatr, 84, 831–6. Pollak, V. E., Rosen, S., Pirani, C. L., et al. 1968. Natural history of lipoid nephrosis and of membranous glomerulonephritis. Ann Intern Med, 69, 1171–96. Ponticelli, C., Edefonti, A., Ghio, L., et al. (1993a). Cyclosporin versus cyclophosphamide for patients with steroid-dependent and frequently relapsing idiopathic nephrotic syndrome: a multicentre randomized controlled trial. Nephrol Dial Transplant, 8, 1326–32. Ravani, P., Magnasco, A., Edefonti, A., et al. (2011). Short-term effects of rituximab in children with steroid- and calcineurin-dependent nephrotic syndrome: a randomized controlled trial. Clin J Am Soc Nephrol, 6, 1308–15. Razzak, M. (2014). Nephrotic syndrome: rituximab is safe and effective in FRNS and SDNS-but where to go from here? Nat Rev Nephrol, 10, 421. Ruth, E. M., Kemper, M. J., Leumann, E. P., et al. (2005). Children with steroid-sensitive nephrotic syndrome come of age: long-term outcome. J Pediatr, 147, 202–7. Rydel, J. J., Korbet, S. M., Borok, R. Z., et al. (1995). Focal segmental glomerular sclerosis in adults: presentation, course, and response to treatment. Am J Kidney Dis, 25, 534–42. Sato, M., Ito, S., Ogura, M., et al. (2013). Atypical Pneumocystis jiroveci pneumonia with multiple nodular granulomas after rituximab for refractory nephrotic syndrome. Pediatr Nephrol, 28, 145–9. Sellier-Leclerc, A. L., Baudouin, V., Kwon, T., et al. (2011). Rituximab in steroid-dependent idiopathic nephrotic syndrome in childhood-follow-up after CD19 recovery. Nephrol Dial Transplant. Sellier-Leclerc, A. L., Belli, E., Guerin, V., et al. (2013). Fulminant viral myocarditis after rituximab therapy in pediatric nephrotic syndrome. Pediatr Nephrol, 28(9), 1875–9. Siegel, N. J., Kashgarian, M., Spargo, B. H., et al. (1974). Minimal change and focal sclerotic lesions in lipoid nephrosis. Nephron, 13, 125–37. Sinha, A., and Bagga, A. (2013). Rituximab therapy in nephrotic syndrome: implications for patients' management. Nat Rev Nephrol, 9, 154–69. Sinha, A., Saha, A., Kumar, M., et al. (2015). Extending initial prednisolone treatment in a randomized control trial from 3 to 6 months

did not significantly influence the course of illness in children with steroid-sensitive nephrotic syndrome. Kidney Int, 87, 217–24. Sinha, M. D., Macleod, R., Rigby, E., et al. (2006). Treatment of severe steroid-dependent nephrotic syndrome (SDNS) in children with tacrolimus. Nephrol Dial Transplant, 21, 1848–54. Siu, Y. P., Tong, M. K., Leung, K., et al. (2008). The use of enteric-coated mycophenolate sodium in the treatment of relapsing and steroid-dependent minimal change disease. J Nephrol, 21, 127–31. Sureshkumar, P., Hodson, E. M., Willis, N. S., et al. (2014). Predictors of remission and relapse in idiopathic nephrotic syndrome: a prospective cohort study. Pediatr Nephrol, 29, 1039–46. Tanaka, H., Tsugawa, K., Tsuruga, K., et al. (2006). Single-dose daily treatment with cyclosporin A for relapsing nephrotic syndrome: report of a case showing poor response. Clin Nephrol, 66, 219–20. Tanphaichitr, P., Tanphaichitr, D., Sureeratanan, J., et al. (1980). Treatment of nephrotic syndrome with levamisole. J Pediatr, 96, 490–3. Trainin, E. B., Boichis, H., Spitzer, A., et al. (1975). Late nonresponsiveness to steroids in children with the nephrotic syndrome. J Pediatr, 87, 519–23. Trompeter, R. S., Evans, P. R., and Barratt, T. M. (1981). Gonadal function in boys with steroid-responsive nephrotic syndrome treated with cyclophosphamide for short periods. Lancet, 1, 1177–9. Trompeter, R. S., Lloyd, B. W., Hicks, J., et al. (1985). Long-term outcome for children with minimal-change nephrotic syndrome. Lancet, 1, 368–70. Ueda, N., Kuno, K., and Ito, S. (1990). Eight and 12 week courses of cyclophosphamide in nephrotic syndrome. Arch Dis Child, 65, 1147–50. Vester, U., Kranz, B., Zimmermann, S., et al. (2005). The response to cyclophosphamide in steroid-sensitive nephrotic syndrome is influenced by polymorphic expression of glutathion-S-transferases-M1 and -P1. Pediatr Nephrol, 20, 478–81. Waldman, M., Crew, R. J., Valeri, A., et al. (2007). Adult minimal-change disease: clinical characteristics, treatment, and outcomes. Clin J Am Soc Nephrol, 2, 445–53. Wang, F., Looi, L. M., and Chua, C. T. (1982). Minimal change glomerular disease in Malaysian adults and use of alternate day steroid therapy. Q J Med, 51, 312–28. Watson, A. R., Taylor, J., Rance, C. P., et al. (1986). Gonadal function in women treated with cyclophosphamide for childhood nephrotic syndrome: a long-term follow-up study. Fertil Steril, 46, 331–3. White, R. H., Glasgow, E. F., and Mills, R. J. (1970). Clinicopathological study of nephrotic syndrome in childhood. Lancet, 1, 1353–9. Williams, S. A., Makker, S. P., Ingelfinger, J. R., et al. (1980). Long-term evaluation of chlorambucil plus prednisone in the idiopathic nephrotic syndrome of childhood. N Engl J Med, 302, 929–33. Wynn, S. R., Stickler, G. B., and Burke, E. C. (1988). Long-term prognosis for children with nephrotic syndrome. Clin Pediatr (Phila), 27, 63–8. Yoshikawa, N., Nakanishi, K., Sako, M., et al. (2015). A multicenter randomized trial indicates initial prednisolone treatment for childhood nephrotic syndrome for two months is not inferior to six-month treatment. Kidney Int, 87, 225–32. Zagury, A., De Oliveira, A. L., De Moraes, C. A., et al. (2011). Long-term follow-up after cyclophosphamide therapy in steroid-dependent nephrotic syndrome. Pediatr Nephrol, 26, 915–20.

CHAPTER 57

Primary focal segmental glomerulosclerosis: clinical features and diagnosis Alain Meyrier and Patrick Niaudet Introduction Focal segmental glomerulosclerosis (FSGS) is the other histopathologic subset of idiopathic nephrotic syndrome. It accounts for 40% of cases in adults but < 20% of cases in children (Korbet, 2012). FSGS is often ‘idiopathic’ but cannot be considered as ‘primary’ without ruling out identified aetiologies of the podocytopathy that characterize it and leads to proteinuria, glomerular sclerosis, and end-stage renal failure. Therefore FSGS is not a disease but a clinicopathologic nephrotic condition, the description of which is primarily based on the lesions found on a kidney biopsy. Of note, the umbrella term FSGS is a misnomer as the lesions are not always focal, segmental, and/or sclerotic.

Demography FSGS is considered ‘primary’ when all causes of ‘secondary’ FSGS have been excluded (Table 57.1). This subset of idiopathic nephrotic syndrome accounts for 20% of children and 40% of adults with an estimated incidence of 7 per million (D’Agati et al., 2011; Korbet, 2012). For unknown reasons the incidence of idiopathic FSGS has steadily increased over recent decades, in children (Filler et  al., 2003) and in adults (Braden et al., 2000). This subset of idiopathic nephrotic syndrome represents a major cause of end-stage renal failure leading to renal replacement therapy with a prevalence of 4% in the United States. There is a strong predominance in black patients of African ancestry that is explained by inherited risk factors. This population develop more severe forms of FSGS than their white and Asian fellow sufferers, are less prone to enjoy remission with treatment (Crook et al., 2005) and follow a more rapid course to end-stage renal disease (ESRD) when their kidney histology reveals cellular forms of FSGS (Schwartz et al., 1999). This risk has been strongly associated with sequence variants (G1 and G2) in the gene encoding apolipoprotein L1 (APOL1) (Genovese et  al., 2010; see Chapter 341). This association of both G1 and G2 seems to account for the excess risk of FSGS in African Americans, both in primary FSGS and in HIV-associated nephropathy (HIVAN). The same variants were also associated with hypertension-attributed end-stage kidney disease, raising the question of whether the hypertension in these patients was in fact

a sign of kidney disease. Alternatively the genotype confers greater susceptibility to progression of different types of injury. The APOL1 gene is very close to the MYH9 gene and represents a 100-fold greater risk factor than had previously been associated with the latter (Kopp et al., 2008). APOL1 is a lysis factor for Trypanosoma brucei brucei, the parasite transmitted by the Tsetse fly and causing sleeping sickness. However, T. brucei gambiense resists APOL1 lysis and this could have induced a selection of African subjects genetically protected from T. brucei brucei sleeping illness but with a trade-off propensity to APOL1 G1 and G2 induced podocyte injury, the mechanism of which is still unknown.

Histopathology Distribution of lesions FSGS is characterized by focal lesions affecting a variable proportion of glomeruli at any time during the course of the disease (Churg et al., 1970; Habib and Gubler, 1973). They predominate in the deeper cortex and juxtamedullary glomeruli are mainly affected (Rich, 1957). Focal changes are frequently limited to part of the tuft. Glomerular hypertrophy is common in FSGS, and glomerulomegaly found in minimal change disease (MCD) is somewhat predictive of further development to FSGS (Fogo et al., 1990; Muda et al., 1994). By serial three-dimensional (3D) ultrathin sections the distribution of focal sclerosis is more widespread than what is seen by conventional microscopy. Fogo et al. studied kidney biopsies from 15 adults and six children (Fogo et al., 1995). Sclerosis assessed on a single section involved 31.5 ± 6.8% of glomeruli in adults, contrasting with only 11.7 ± 5.7% in children. After 3D screening, the percentage of glomeruli involved by sclerosis increased to 48.0 ± 6.6% in adults and 23.2 ± 7.4% in children. The greater increase in sclerosis after 3D analysis in children versus adults reflected the predominance of small peripheral, that is, more segmental, lesions in children than adults. Similar findings were described by Fuiano et al. (1996). They concur to indicate that FSGS is less segmental and probably less focal than previously thought. This may explain the fact that FSGS can be overlooked on an early initial kidney biopsy or in cases of sampling error with < 10 glomeruli studied by light microscopy.

516

Section 3  

the patient with glomerular disease

Table 57.1  Differential expression of cyclin-dependent kinase inhibitors in human glomerular disease Response (N)

All (187)

Cell (18) CG (53) GTL (33) NOS (83)

CR-PR-F (N)

51-21-115

7-1-10

4-3-46

20-5-8

20-12-51

CR+PR (%)

38.5

44.4

13.2

75.8

38.6

ESRD (%)

30.7

27.8

65.3

5.7

34.5

Kaplan–Meier months survival to ESRD (median)

72 ± 10

35 ± 6

23 ± 11

85 ± 8

90 ± 10

Cell = cellular FSGS; CG = collapsing FSGS; CR = complete remission; ESRD = end-stage renal disease; F = no remission; GTL = glomerular tip lesion variant; NOS = FSGS not otherwise specified; PR = partial remission. Part of Table 57.1 from Stokes et al. (2006).

The initial podocyte lesion FSGS is a podocyte disease that starts with changes that are initially only detectable by electron microscopy, as shown by kidney biopsy carried out when proteinuria relapses shortly, sometimes within hours, after kidney transplantation (Schwartz and Korbet, 1993; Kriz et al., 1994; Rydel et al., 1995; Schwartz et al., 1995; Bariéty et al., 1998a; Elger and Kriz, 1998). Relapse of nephrotic primary FSGS on a renal transplant offers a privileged model to study the incipient lesion in man and to follow its progression (Verani and Hawkins, 1986; Korbet et al., 1988; Bariéty et al., 2001). In such a case the initial appearance is often limited to foot process fusion that can be restricted to some segments of the glomerular basement membrane. As noted in Chapter 56, this fusion, or ‘flattening’ of the podocyte foot processes is due to a rearrangement of the actin cytoskeleton fibres. However, contrary to MCD, this rearrangement is usually not reversible. In fact the podocyte cytoskeleton derangement is an essential factor inherent in the pathogenesis of FSGS (Mathieson, 2010). Within weeks following recurrence of proteinuria, podocytes observed by electron microscopy appear swollen and vacuolated. Some vacuoles are round and by immunofluorescence appear to contain immunoglobulin (Ig)-A (Fig. 57.1). The podocytes exhibit strong mitotic activity, with multinucleation and expression of the (A)

PCNA and Ki-67 proliferation markers. The number of visceral epithelial cells seems to be increased, suggesting a mitotic activity and replication. Podocytes detached from the glomerular basement membranes assume a round shape and form grape-like clusters of cells on the outer aspect of the tuft. Their number may be such that they assume the appearance of a pseudocrescent. Some appear to drift free in the urinary chamber, and migrate into the tubular lumen (Oda et al., 1998) (Fig. 57.2). Other dysmorphic podocytes assume a ‘cobblestone’ pattern covering the outer aspect of the tuft. The parietal epithelial cells of Bowman’s capsule facing this line of cuboid cells usually proliferate and form a pluricellular stratum (Fig. 57.3). Underlying endocapillary lesions comprise foam cells, macrophages, and mesangial cell proliferation along with capillary loop collapse. This subset of FSGS is the ‘cellular lesion’ (Schwartz and Lewis, 1985). Further stages are characterized by extracellular matrix build-up, comprising ubiquitous collagen, leading to the scar lesion, variably hilar, peripheral, or central (Schwartz and Korbet, 1993; Schwartz et al., 1995; D’Agati et al., 2011).

The progression to segmental glomerular lesions Following a progression that may take a few weeks, the lesion of FSGS affects a few capillary loops, which stick together either at the hilum or at the periphery of the tuft, often at both. Hyaline material is often present within sclerotic areas, appearing as a peripheral rim or as round deposits obstructing lumens. Foamy endothelial cells and lipid inclusions may be found. At the periphery of sclerotic segments there is, in most cases, a clear ‘halo’ (Fig. 57.4). The segmental lesion has a different appearance depending on whether it affects a group of capillary loops free in Bowman’s space or is adherent to Bowman’s capsule. The ‘free’ sclerotic segments are surrounded by a layer of cuboid podocytes (‘cobblestones’) in close apposition to the clear ‘halo’. When the sclerotic lesion adheres to Bowman’s capsule, podocytes are no longer identifiable and a synechia forms an adherence between the collapsed capillary loops and Bowman’s capsule. The rest of the tuft and the non-sclerotic glomeruli show either ‘minimal changes’ or ‘mesangial proliferation’ with foot-process effacement (Figs 57.5, 57.6, and 57.7). In some areas the tuft is separated from Bowman’s capsule by a continuous layer of parietal epithelial cells progressing along the outer aspect of the tuft, thereby circumscribing an empty slit and (B) Droplets Vacuoles

Fig. 57.1  Dysmorphic dedifferentiated podocytes. (A) The upper part of this glomerulus (upper rectangle) shows a cluster of hypertrophied podocytes, each one forming a giant clear vacuole limited by a thin border of remaining cytoplasm. In the lower rectangle a grossly hypertrophied podocyte seems about to detach from the glomerular basement membrane. (Masson trichrome, ×350.) (B) Electron microscopy discloses clear vacuoles (lower arrow) and proteinaceous droplets (upper arrow). (Uranyl-acetate lead citrate ×4500.)

Chapter 57 

(A)

(B)

fsgs: clinical features and diagnosis (C)

Fig. 57.2  Microscopic preparation using a monoclonal anti-CD68 antibody that labels macrophagic epitopes in red. Detachment and migration of transdifferentiated podocytes. (A) A cluster of detached podocytes (thin arrow) migrates into Bowman’s chamber towards the glomerular outlet to the proximal tubule. An isolated macrophage-like cell (arrowhead) is bi-nucleated indicating mitosis. (B) Proximal tubule. The macrophagic-like cells drift along the tubules. The tubular epithelium is thinned and atrophic, limited to a single layer of tubular cells. The surrounding interstitium is fibrous and inflammatory. (C) Confocal laser microscopy using an anticytokeratin (CK) monoclonal antibody (AE1/AE3 anti-CK mAb.) that labels the tubular basement basement membrane in bright red (arrowhead) and a monoclonal antibody (PGM1) that identifies macrophagic epitopes (CD68). The macrophage-like cells appear in a yellowish-green colour (dotted arrow). One of the cells (broken arrow) assumes a brownish colour indicating the presence of CD68 and cytokeratin. This indicates that a process of epithelial-mesenchymatous transition (EMT) is at work. This cell is bi-nucleated indicating mitosis.

assuming the appearance of a pseudo-tubule. This pseudo-tubule persists in obsolescent glomeruli, which allows retrospective diagnosis of terminal forms of FSGS (Fig. 57.8). Glomerular obsolescence follows. FSGS is an irreversible scarring process, as demonstrated by repeat biopsies (Velosa et al., 1975; Nash et al., 1976). The whole tuft is sclerotic, often in association with conspicuous interstitial and tubular damage. In children, focal global sclerosis should be differentiated from ‘congenital glomerulosclerosis’, which is a developmental anomaly frequently found in kidneys of infants and young children (Kohaut et al., 1976) and is not associated with tubulointerstitial changes. In normal adults, the number of sclerotic glomeruli increases with age (Kaplan et al., 1975) and their interpretation on biopsies from patients > 40 years is difficult (Cameron et al., 1974). Electron microscopy shows that capillary obliteration is mainly due to paramesangial and subendothelial, finely granular deposits (Velosa et al., 1975) with endothelial cell disappearance or swelling, and increased mesangial matrix. Fatty vacuoles may be seen, among

Fig. 57.3  ‘Cobblestones’. Masson trichrome. A row of dedifferentiated podocytes lines a large scar lesion. Parietal epithelial cells (not shown here) may participate to ‘cobblestone’ formation and form a pseudotubule as shown in Fig. 57.8.

the abnormal deposits or in the cytoplasm of endothelial and mesangial cells. Synechiae are formed by apposition of a cloudy, acellular material containing thin and irregular layers of newly formed basement membrane (Fig. 57.9). Sclerotic lesions result from a marked increase in glomerular basement menbrane-like material with capillary wall wrinkling and capillary collapse (Rumpelt and Thoenes, 1974). Collagen fibres are seen within some segmental areas. Studies in animals (Couser and Stilmant, 1975) and in nephrotic patients have shown that proteinuria precedes the development to focal sclerotic lesions. The same sequence was reported in patients with recurrence after transplantation.

Cell dedifferentiation and transdifferentiation in cellular forms of FSGS Since 1998, a number of studies, especially but not exclusively regarding collapsing forms of FSGS, have shown that the process inducing the podocyte disease which characterizes this condition is accompanied by striking changes occurring in these visceral

Fig. 57.4  Focal segmental sclerosis/hyalinosis: segmental lesion of the tuft characterized by the deposition of hyaline material in the inner side of glomerular basement membrane, and a ring of detached podocytes separated from the glomerular basement membrane by a clear ‘halo’. (Masson’s trichrome, ×520.)

517

518

Section 3  

the patient with glomerular disease

Fig. 57.5  Focal segmental sclerosis/hyalinosis: obliteration of capillary lumens at the vascular pole of the glomerulus by a combination of sclerosis and hyalinosis. (Masson’s trichrome, ×320.)

epithelial cells. These changes have been attributed to a profound cell cycle derangement. The normal mature podocyte is unable to replicate and does not express the proliferation markers PCNA and Ki-67. It is characterized by the expression of numerous epitopes comprising Wilms tumour protein-1 (WT-1), common acute lymphoblastic leukaemia antigen (CALLA), C3b receptor (CR1), glomerular epithelial protein-1 (GLEPP-1), podocalyxin, synaptopodin, and vimentin. The first stages of FSGS, as observed when the lesions relapse on a renal transplant (Bariéty et al., 2001) and in ‘collapsing glomerulopathy’ (Bariéty et al., 1998b; Barisoni et al., 1999, 2000), are characterized by loss of normal podocyte epitopes and acquisition of macrophagic and cytokeratin markers. Nuclear proliferation markers PCNA and Ki-67 are expressed indicating strong mitotic activity. Such ‘podocyte dysregulation’ is accompanied by podocyte detachment from the basement membranes, and migration into the urinary chamber and the tubular lumens where they assume a round shape and occasionally co-express macrophagic and cytokeratin epitopes. By laser confocal microscopy some of these round cells

Fig. 57.6  Collapsing glomerulopathy. Masson trichrome, ×200. The glomerular tuft is shrunken and only a withered stalk remains (arrowhead). A great number of de-differentiated podocytes are completely vacuolated and occupy the urinary chamber (thin arrow). When more numerous they may assume the appearance of a pseudocrescent. Courtesy of Laure-Hélène Noël MD, Necker Hospital, Paris, France.

Fig. 57.7  Cell-FSGS. Methenamine silver–periodic acid–Schiff (Jones stain), ×185. This variant of FSGS was first described by Schwartz and Lewis in 1985 (Schwartz and Lewis, 1985).There is a collapse of all of the glomerular capillaries and diffuse proliferation and hypertrophy of the visceral epithelial cells. It is difficult to find a difference, if any, between this picture and severe forms of the ‘Collapsing variant’ of the Columbia classification. From Schwartz et al. (1999).

co-express CD68 or cytokeratin markers and original podocyte markers such as podocalyxin (Fig. 57.10). Shankland et al. showed that the expression of cyclins and cyclin inhibitors of the Cip/Kip family differ among patients with MCD or idiopathic membranous glomerulopathy and those with cellular focal segmental glomerulonephritis (Shankland et al., 2000). These remarkable findings are summarized in Table 57.2.

Reversibility of the cellular FSGS lesion That a circulating factor induces proteinuria followed by podocyte lesions, capillary collapse, and perpetuating to fibrosis is a well-established notion. The question at stake is to know if the initial podocytopathy is at some point reversible before sclerosis sets in. A  few privileged clinical observations point to a

Fig. 57.8  Appearance of terminal FSGS. Toluidine blue stain. ×250.This glomerulus is totally sclerotic. However a pseudotubule remains at the lower pole, presumably formed by a continuous joining of visceral and parietal epithelial cells, a feature indicating retrospectively that the initial glomerulopathy was FSGS.

Chapter 57 

fsgs: clinical features and diagnosis

demonstrates the reversibility of proteinuria and of podocyte changes. A patient with nephrotic FSGS was transplanted with a living donor kidney. Proteinuria recurred within 2  days. The recipient was fully nephrotic with a decline in renal function. Postoperative biopsies showed recurrence of the podocytopathy. The kidney was removed and re-transplanted in a patient whose primary renal disease was diabetic glomerulopathy. The evolution was to disappearance of proteinuria, recovery of renal function, and kidney biopsies showed that podocytes had recovered a normal appearance within 2 weeks (Gallon et al., 2012). Such observations show that removing or counteracting the elusive glomerular permeability factor (GPF) should be the first goal of treatment, a goal poorly achieved with plasmapheresis (see Chapter 289). In fact, the nature of the GPF and its source are still virtually unknown. Neutralizing the GPF and/or acting on its source would represent the real treatment of FSGS and, as analysed below, this goal is not necessarily based on our current corticosteroid and immunosuppressive therapy.

Fig. 57.9  Focal segmental sclerosis/hyalinosis (electron microscopy): note the presence around the collapsed capillary loops of multilayered basement-membrane material in a subepithelial location (arrows). (Uranyl acetate–lead citrate, ×2380.)

reversibility of both proteinuria and podocytopathy when a normal kidney is withdrawn from the ‘nephrotic environment’. This was shown in cases of pregnancy when after delivery of a woman with nephrotic FSGS the children were proteinuric at birth and proteinuria progressively waned in a matter of weeks (Lagrue et al., 1989; Kemper et al., 2001). A paradigmatic case

Associated tubulointerstitial lesions Tubular atrophy and interstitial fibrosis are common and proportional to the glomerular damage (Hyman and Burkholder, 1973; Newman et al., 1976). Overlooked FSGS should therefore be suspected when tubular and interstitial changes are associated with minimal glomerular changes. In rare instances tubular lesions

HLA-DR

HAM56

(B)

(A)

25F9

25F9

(C)

(D)

CD16

(E)

Fig. 57.10  Podocyte transdifferentiation in FSGS. (A) HAM56, a marker of the human macrophage, labels large round cells aligned at the periphery of the glomerular tuft (arrow). Note the numerous small spindle-shaped cells in the interstitium (arrowhead). (B) CRD/43, an mAb specific for HLA-DR, labels a row of large round cells at the periphery of the tuft (arrow), indicating that these cells can be considered activated. (C) 25F9 mAb, a marker of macrophage maturation, is detected on all cells of some tubules (arrowheads), whereas other tubules do not exhibit such transdifferentiation. Interstitial inflammatory cells are not tagged by 25F9. (D) 25F9 mAb labels two large round cells in a tubular lumen. One (arrow) is apparently free in the lumen. The other (arrowhead) seems to be inserted between tubular cells or, alternatively, could be a transdifferentiated tubular cell. (E) Anti-CD16 mAb, a marker of activated macrophages, tags large round cells within the tubular lumens. The interstitial inflammatory cells are CD16 negative. (Magnifications: ×120 in A and B; ×200 in C and D.) From Bariéty et al. (2001).

519

520

Section 3  

the patient with glomerular disease

Table 57.2  Differential expression of cyclin-dependent kinase inhibitors in human glomerular disease p57

p27

p21

Ki-67

Controls

+

+

–

–

Minimal change disease

+

+

–

–

Membranous glomerulopathy

+

+

–

–

FSGS, cellular variant

–

–

+

+

FSGS, collapsing variant

–

–

+

+

HIV-associated FSGS

–

–

+

+

The expression by podocytes of the cyclin-dependent kinase inhibitors p57, p27, and p21 and the proliferation marker Ki-67 differs among control renal tissue and glomerular disease tissue. Notably, the cellular and collapsing variants of idiopathic FSGS and HIV-associated FSGS share a similar podocyte phenotype. Data summarizing Shankland et al. (2000).

predominate (Hayslett et al., 1969). Foam cells may be seen in the interstitium and occasionally in the glomeruli. Detached podocytes that drift along the tubules and can be recovered in the urine might explain the tubulointerstitial lesions. These dysregulated podocytes express macrophagic markers and it is conceivable that along with heavy proteinuria they contribute to injuring the tubular cells.

Histological subtypes of FSGS A consensus workshop elaborated a classification of FSGS that is widely used and has the merit of making pathologists speak the same language (D’Agati et al., 2004). This ‘Columbia classification’ applies both to idiopathic and secondary forms of FSGS. It sorts out five histologic subtypes shown in Figs 57.6, 57.7, 57.11, 57.12, and 57.13. 1. FSGS not otherwise specified (NOS). This common variant is a diagnosis of exclusion based on absent features of the following forms. These other forms may evolve into FSGS NOS.

Fig. 57.11  FSGS ‘NOS’ (‘not otherwise specified’). Counterstained periodic acid–Schiff staining. A large area of the glomerulus is fibrous (upper half, bright red) and still contains sparse podocyte remnants. Another lesion is found in the lower pole of this glomerulus and involves several lobules. Courtesy of Ian Roberts MD, Pathology Department, John Radcliffe Hospital, Oxford, UK.

5. Glomerular tip lesion (GTL). The lesion involves the tubular pole. It might be due to an adaptive mechanism to protein-rich ultrafiltrate causing podocyte shear stress and prolapse of part of the tuft protruding into the initial segment of the tubule.

Clinicopathologic correlations Three variants of the above classification deserve a particular attention with regard to their clinical implications.

Collapsing glomerulopathy Collapsing glomerulopathy (CG) is a severe form of cellular FSGS. In 1986, Weiss et al. published on ‘A new clinicopathologic entity’ observed in a small group of patients with glomerular ‘collapse’ and nephrotic syndrome rapidly progressing to renal failure

2. Perihilar FSGS. This subtype consists of lesions located at the vascular pole of the glomerular tuft. It is considered as an ‘adaptive’ form of FSGS and is accompanied with glomerulomegaly. It is the common appearance in secondary forms with hyperfiltration. 3. Cellular FSGS. It is characterized by podocyte hyperplasia and endocapillary hypercellularity, foam cells and leucocyte infiltration. 4. Collapsing glomerulopathy. This ‘implosive’ type of FSGS is characterized by segmental or global wrinkling and collapse of the glomerular capillary walls with prominent hypertrophy and hyperplasia of the overlying podocytes that can form ‘pseudocrescents’. Biopsies containing any glomeruli with global collapse and/or > 20% of glomeruli with segmental collapse are considered to be ‘collapsing glomerulopathy’.

Fig. 57.12  Focal-segmental sclerosis/hyalinosis: obliteration of capillary lumens at the vascular pole of the glomerulus by a combination of sclerosis and hyalinosis. (Masson’s trichrome, ×320.)

Chapter 57 

Fig. 57.13  Typical tip lesion. The photomicrograph is focused on the ‘glomerular outlet’. Note the ‘cell-type’ of this form of focal–segmental glomerulonephritis, with abundant, swollen macrophages in the capillary lumens. (Masson’s trichrome, ×600.) Courtesy of Professor Jean Bariéty, Paris VI University, Broussais-Hôtel Dieu Medical School and INSERM U 430, Paris.

(Weiss et al., 1986). These features closely resembled those of the newly described HIVAN and in fact some of their patients were HIV-1 infected. Eight years later, Detwiler et al. reported on 16 HIV-negative patients with idiopathic FSGS, a rapid course, and collapsing glomerular features characterized by segmental or global wrinkling and collapse of the glomerular capillary walls with prominent hypertrophy and hyperplasia of the overlying podocytes (Detwiler et al., 1994). Biopsies containing any glomeruli with global collapse and/or > 20% of glomeruli with segmental collapse were considered to be ‘collapsing glomerulopathy’. In fact, earlier descriptions of FSGS comprised a collapsing component of some capillary loops (Velosa et  al., 1975). Korbet et  al. studied two cases of recurrent FSGS on renal allografts. Focal glomerular lesions consisted of segmental epithelial cell proliferation with mitotic figures and collapse of glomerular capillaries (Korbet et al., 1988). This cellular and collapsing lesion was later followed by a scar lesion. The same was found by Bariéty et al. after relapse of FSGS on renal transplants (Bariéty et al., 2001). Since then, numerous articles have appeared which all consider CG as a distinct form of FSGS with increasing incidence, distinct clinicopathologic features, black racial predominance, a relative steroid resistance, and a rapidly progressive course. Collapsing glomerulopathy may be ‘idiopathic’ (Nagata et al., 1998; Barisoni et  al., 1999, 2000), recur on a renal transplant (Clarkson et  al., 1998; Toth et al., 1998), represent a variety of de novo glomerular disease after renal transplantation (Meehan et al., 1998; Stokes et al., 1999), or be associated with Loa Loa (Pakasa et al., 1997), malaria (Niang et al., 2008), or viral infection other than HIVAN (see Chapter 187). Viral aetiologies are discussed below.

Cellular FSGS ‘Cell-FSGS’ described by Schwartz and Lewis is a common form of FSGS (Schwartz and Lewis, 1985). Schwartz et al. do not sort out ‘cell-FSGS’ from collapsing glomerulopathy and their histopathologic pictures of cell-FSGS are indistinguishable from CG (Schwartz et  al., 1999)  (Fig. 57.7). In the latter article, the prognosis of the cellular lesion was retrospectively studied in 100 patients with FSGS (43 had cell-FSGS and 57 had FSGS with the

fsgs: clinical features and diagnosis

classic segmental scar (CS)). Patients with the cell-FSGS lesion were more often black and severely proteinuric and developed more ESRD. They were more proteinuric at presentation than patients with FSGS-CS. ESRD developed more frequently in patients with cell-FSGS and patients with extensive cell-FSGS (≥ 20% glomeruli) were mainly black (94%), severely nephrotic (67%, >10 g/day), and had a poor response to treatment (23% remission). The rates of remission in treated nephrotic patients with cell-FSGS and FSGS-CS were the same (53% vs 52%), and patients in both groups who achieved a remission had a 5-year survival of 100%. Steroid responsiveness was the only variable that predicted remission. These data are not worse than those of Stokes et al. (Table 57.1) who in 187 cases of FSGS (all types) found 72 cases (38.5%) of complete plus partial remission (Stokes et al., 2006).

Glomerular tip lesion The GTL is the most benign form of FSGS. Stokes et al followed up 29 cases. The clinical onset of the nephrotic syndrome was rapid, as it is in MCD, there was no tubular disease and the response to treatment was favourable with 21 complete plus partial remissions and eight failures (Stokes et al., 2004). The authors concluded that ‘Within the MCD/FSGS spectrum, GTL is a distinctive and prognostically favourable clinico-pathologic entity whose presenting features and outcome more closely approximate those of MCD’.

Differential diagnosis Clinical Heavy proteinuria with or without the full-blown picture of nephrotic syndrome may be the consequence of virtually all glomerulopathies. This implies that the diagnosis of FSGS rests on a kidney biopsy comprising a sufficient number of glomeruli adequately processed for light, electron, and immunofluorescence microscopy. However, it is worth recalling that primary FSGS does not comprise extrarenal abnormalities. This not the case for some forms of secondary FSGS that can be syndromic as briefly analysed below.

Histopathologic FSGS is well defined histologically but some forms may still pose a diagnostic problem. This is true of elderly patients (although the definition of ‘elderly’ is subject to some interpretation), patients in whom other diseases may induce glomerular lesions (especially the hilar variant) that cannot be considered as true primary FSGS. A kidney biopsy disclosing only obsolescent glomeruli may also leave the diagnosis of the primary glomerulopathy pending, except when remnant pseudotubules (Fig. 57.8) suggest a diagnosis of terminal FSGS. This has practical implications in case of post-transplant relapse in a patient whose primary glomerulopathy had not been previously identified (see Chapter 283).

Aetiologic The renal biopsy appearance of FSGS can be secondary to diverse mechanisms (see Box 57.1). This is an important distinction as these are unlikely to respond to the same therapies.

521

522

Section 3  

the patient with glomerular disease

Box 57.1  Main conditions associated with ‘secondary’ focal segmental glomerulosclerosisa ◆ Hyperfiltration inducing shear stretch on podocytesb (Chapter 139) ◆ Unilateral renal agenesis or hypoplasia (Chapter 138) ◆ Renal ablation—remnant kidney ◆ Reflux interstitial nephropathy (chapter 355) ◆ Oligomeganephronia ◆ Morbid obesity ◆ Sickle cell disease (Chapter 167) ◆ Diabetic glomerulopathy (Chapter 149) ◆ Anabolic androgens ◆ Congenital cyanotic heart disease ◆ Viral infections: • HIV-associated nephropathy (HIVAN) • Parvovirus B 19c • Simian virus SV 40c • Cytomegalovirus (CMV) • Epstein–Barr virus (EBV) • Hepatitis C virus (HCV)

A common hypothesis is that FSGS is a secondary consequence of haemodynamic stretch on podocytes (Kriz et al., 2013) (see Chapter 136). It is argued that this could explain FSGS seen in obesity and nephron number reduction, such as, amongst others, remnant kidney, oligomeganephronia, reflux nephropathy, and various nephropathies with widespread glomerular obsolescence. A particular issue is that of viral aetiologies. Following the identification of HIV-associated nephropathy, researchers claimed that such viruses as parvovirus B 19 (Moudgil et al., 2001), simian virus SV 40 (Li et al., 2002), hepatitis B virus (Sakai et al., 2011), and hepatitis C virus (Sperati, 2012) may cause the collapsing variant of FSGS. In fact, a critical review of the literature based on case reports and small case series does link infection with some common viruses and glomerular injury, including anecdotal cases of collapsing glomerulopathy associated with cytomegalovirus (CMV) and Epstein–Barr virus (EBV; Chandra and Kopp, 2013). However evidence for a pathogenic role is generally stronger for glomerulonephritis than specifically for collapsing glomerulopathy. The evidence linking collapsing glomerulopathy with CMV is relatively strong but not yet conclusive, whilst the evidence for a pathogenic role for EBV and parvovirus B19 is weaker. The identification of genetic forms of FSGS has elicited an immense interest and a flurry of publications over the last two decades. Some forms are syndromic, that is, comprising extrarenal abnormalities and others non-syndromic with only kidney involvement. This major breakthrough is covered in Chapter 327. Interestingly, treatment responses are described for some genetic causes.

◆ Toxic agents (Chapter 82): • Heroin

References

• Pamidronate-alendronate

Bariéty, J., Bruneval, P., Hill, G., et al. (2001). Posttransplantation relapse of FSGS is characterized by glomerular epithelial cell transdifferentiation. J Am Soc Nephrol, 12, 261–74. Bariéty, J., Nochy, D., Jacquot, C., et al. (1998a). Diversity and unity of focal and segmental glomerular sclerosis. Adv Nephrol Necker Hosp, 28, 1–42. Bariéty, J., Nochy, D., Mandet, C., et al. (1998b). Podocytes undergo phenotypic changes and express macrophagic-associated markers in idiopathic collapsing glomerulopathy. Kidney Int, 53, 918–25. Barisoni, L., Kriz, W., Mundel, P., et al. (1999). The dysregulated podocyte phenotype: a novel concept in the pathogenesis of collapsing idiopathic focal segmental glomerulosclerosis and HIV-associated nephropathy. J Am Soc Nephrol, 10, 51–61. Barisoni, L., Mokrzycki, M., Sablay, L., et al. (2000). Podocyte cell cycle regulation and proliferation in collapsing glomerulopathies. Kidney Int, 58, 137–43. Braden, G. L., Mulhern, J. G., O’Shea, M. H., et al. (2000). Changing incidence of glomerular diseases in adults. Am J Kidney Dis, 35, 878–83. Cameron, J. S., Turner, D. R., Ogg, C. S., et al. (1974). The nephrotic syndrome in adults with 'minimal change' glomerular lesions. QJM, 43, 461–88. Chandra, P. and Kopp, J. B. (2013). Viruses and collapsing glomerulopathy: a brief critical review. Clin Kidney J, 6, 1–5. Churg, J., Habib, R., and White, R. H. (1970). Pathology of the nephrotic syndrome in children: a report for the International Study of Kidney Disease in Children. Lancet, 760, 1299–302. Clarkson, M. R., O’Meara, Y. M., Murphy, B., et al. (1998). Collapsing glomerulopathy—recurrence in a renal allograft. Nephrol Dial Transplant, 13, 503–6.

• Lithium • Interferon • Bevacizumab (anti VEGF monoclonal antibody) • Ciclosporin ◆ Ageing (Chapter 300) ◆ “Hypertensive nephrosclerosis” (see Chapters 100, 208) ◆ Hereditary conditions (see Chapter 45): • Gene defects in the slit diaphragm (Nephrin, podocin, CD2AP, actinin 4, TRCP6) • Mitochondrial cytopathies • WT1 mutation (Denys–Drash and Frasier syndromes) • Thin membrane disease with collagen mutations. a

These secondary forms require a supportive regimen. In some, such as HIVAN and HCV, antiviral treatment is required. When a toxic agent has been identified, its removal is mandatory. b The size (diameter and volume) of glomeruli is often distinctly greater than normal. This ‘glomerulomegaly’ may induce a shear stretch on podocytes, see Chapters 136 and 139. c Not definitely proven.

Chapter 57 

Couser, W. G. and Stilmant, M. M. (1975). Mesangial lesions and focal glomerular sclerosis in the aging rat. Lab Invest, 33, 491–501. Crook, E. D., Habeeb, D., Gowdy, O., et al. (2005). Effects of steroids in focal segmental glomerulosclerosis in a predominantly African-American population. Am J Med Sci, 330, 19–24. D’Agati, V. D., Fogo, A. B., Bruijn, J. A., et al. (2004). Pathologic classification of focal segmental glomerulosclerosis: a working proposal. Am J Kidney Dis, 43, 368–82. D’Agati, V. D., Kaskel, F. J., and Falk, R. J. (2011). Focal segmental glomerulosclerosis. N Engl J Med, 365, 2398–411. Detwiler, R. K., Falk, R. J., Hogan, S. L., et al. (1994). Collapsing glomerulopathy: a clinically and pathologically distinct variant of focal segmental glomerulosclerosis. Kidney Int, 45, 1416–24. Elger, M. and Kriz, W. (1998). Podocytes and the development of segmental glomerulosclerosis. Nephrol Dial Transplant, 13, 1368–73. Filler, G., Young, E., Geier, P., et al. (2003). Is there really an increase in non-minimal change nephrotic syndrome in children? Am J Kidney Dis, 42, 1107–13. Fogo, A., Glick, A. D., Horn, S. L., et al. (1995). Is focal segmental glomerulosclerosis really focal? Distribution of lesions in adults and children. Kidney Int, 47, 1690–6. Fogo, A., Hawkins, E. P., Berry, P. L., et al. (1990). Glomerular hypertrophy in minimal change disease predicts subsequent progression to focal glomerular sclerosis. Kidney Int, 38, 115–23. Fuiano, G., Comi, N., Magri, P., et al. (1996). Serial morphometric analysis of sclerotic lesions in primary “focal” segmental glomerulosclerosis. J Am Soc Nephrol, 7, 49–55. Gallon, L., Leventhal, J., Skaro, A., et al. (2012). Resolution of recurrent focal segmental glomerulosclerosis after retransplantation. N Engl J Med, 366, 1648–9. Genovese, G., Friedman, D. J., Ross, M. D., et al. (2010). Association of trypanolytic ApoL1 variants with kidney disease in African Americans. Science, 329, 841–5. Habib, R. and Gubler, M. C. (1973). Focal sclerosing glomerulonephritis. Perspect Nephrol Hypertens, 1 Pt 1, 263–78. Hayslett, J. P., Krassner, L. S., Bensch, K. G., et al. (1969). Progression of “lipoid nephrosis” to renal insufficiency. N Engl J Med, 281, 181–7. Hyman, L. R. and Burkholder, P. M. (1973). Focal sclerosing glomerulonephropathy with segmental hyalinosis. A clinicopathologic analysis. Lab Invest, 28, 533–44. Kaplan, C., Pasternack, B., Shah, H., et al. (1975). Age-related incidence of sclerotic glomeruli in human kidneys. Am J Pathol, 80, 227–34. Kemper, M. J., Wolf, G., and Muller-Wiefel, D. E. (2001). Transmission of glomerular permeability factor from a mother to her child. N Engl J Med, 344, 386–7. Kohaut, E. C., Singer, D. B., and Hill, L. L. (1976). The significance of focal glomerular sclerosis in children who have nephrotic syndrome. Am J Clin Pathol, 66, 545–50. Kopp, J. B., Smith, M. W., Nelson, G. W., et al. (2008). MYH9 is a major-effect risk gene for focal segmental glomerulosclerosis. Nat Genet, 40, 1175–84. Korbet, S. M. (2012). Treatment of primary FSGS in adults. J Am Soc Nephrol, 23, 1769–76. Korbet, S. M., Schwartz, M. M., and Lewis, E. J. (1988). Recurrent nephrotic syndrome in renal allografts. Am J Kidney Dis, 11, 270–6. Kriz, W., Elger, M., Nagata, M., et al. (1994). The role of podocytes in the development of glomerular sclerosis. Kidney Int Suppl, 45, S64–72. Kriz, W., Shirato, I., Nagata, M., et al. (2013). The podocyte's response to stress: the enigma of foot process effacement. Am J Physiol Renal Physiol, 304, F333–47. Lagrue, G., Niaudet, P., Guillot, F., et al. (1989). Pregnancy and glomerulonephritis. Lancet, 2, 1037.

fsgs: clinical features and diagnosis

Li, R. M., Branton, M. H., Tanawattanacharoen, S., et al. (2002). Molecular identification of SV40 infection in human subjects and possible association with kidney disease. J Am Soc Nephrol, 13, 2320–30. Mathieson, P. W. (2010). Podocyte actin in health, disease and treatment. Nephrol Dial Transplant, 25, 1772–3. Meehan, S. M., Pascual, M., Williams, W. W., et al. (1998). De novo collapsing glomerulopathy in renal allografts. Transplantation, 65, 1192–7. Moudgil, A., Nast, C. C., Bagga, A., et al. (2001). Association of parvovirus B19 infection with idiopathic collapsing glomerulopathy. Kidney Int, 59, 2126–33. Muda, A. O., Feriozzi, S., Cinotti, G. A., et al. (1994). Glomerular hypertrophy and chronic renal failure in focal segmental glomerulosclerosis. Am J Kidney Dis, 23, 237–41. Nagata, M., Hattori, M., Hamano, Y., et al. (1998). Origin and phenotypic features of hyperplastic epithelial cells in collapsing glomerulopathy. Am J Kidney Dis, 32, 962–9. Nash, M. A., Greifer, I., Olbing, H., et al. (1976). The significance of focal sclerotic lesions of glomeruli in children. J Pediatr, 88, 806–13. Newman, W. J., Tisher, C. C., McCoy, R. C., et al. (1976). Focal glomerular sclerosis: contrasting clinical patterns in children and adults. Medicine (Baltimore), 55, 67–87. Niang, A., Niang, S. E., Ka El, H. F., et al. (2008). Collapsing glomerulopathy and haemophagocytic syndrome related to malaria: a case report. Nephrol Dial Transplant, 23, 3359–61. Oda, T., Hotta, O., Taguma, Y., et al. (1998). Clinicopathological significance of intratubular giant macrophages in progressive glomerulonephritis. Kidney Int, 53, 1190–200. Pakasa, N. M., Nseka, N. M., and Nyimi, L. M. (1997). Secondary collapsing glomerulopathy associated with Loa loa filariasis. Am J Kidney Dis, 30, 836–9. Rumpelt, H. J. and Thoenes, W. (1974). Focal and segmental sclerosing glomerulopathy (-nephritis). Virchows Arch A Pathol Anat Histol, 362, 265–82. Rydel, J. J., Korbet, S. M., Borok, R. Z., et al. (1995). Focal segmental glomerular sclerosis in adults: presentation, course, and response to treatment. Am J Kidney Dis, 25, 534–42. Sakai, K., Morito, N., Usui, J., et al. (2011). Focal segmental glomerulosclerosis as a complication of hepatitis B virus infection. Nephrol Dial Transplant, 26, 371–3. Schwartz, M. M., Evans, J., Bain, R., et al. (1999). Focal segmental glomerulosclerosis: prognostic implications of the cellular lesion. J Am Soc Nephrol, 10, 1900–7. Schwartz, M. M., and Korbet, S. M. (1993). Primary focal segmental glomerulosclerosis: pathology, histological variants, and pathogenesis. Am J Kidney Dis, 22, 874–83. Schwartz, M. M., Korbet, S. M., Rydell, J., et al. (1995). Primary focal segmental glomerular sclerosis in adults: prognostic value of histologic variants. Am J Kidney Dis, 25, 845–52. Schwartz, M. M., and Lewis, E. J. (1985). Focal segmental glomerular sclerosis: the cellular lesion. Kidney Int, 28, 968–74. Shankland, S. J., Eitner, F., Hudkins, K. L., et al. (2000). Differential expression of cyclin-dependent kinase inhibitors in human glomerular disease: role in podocyte proliferation and maturation. Kidney Int, 58, 674–83. Sperati, J. (2012). Stabilization of hepatitis C associated collapsing focal segmental glomerulosclerosis with interferon a-2a and ribavirin. Clin Nephrol. Stokes, M. B., Davis, C. L., and Alpers, C. E. (1999). Collapsing glomerulopathy in renal allografts: a morphological pattern with diverse clinicopathologic associations. Am J Kidney Dis, 33, 658–66.

523

524

Section 3  

the patient with glomerular disease

Stokes, M. B., Markowitz, G. S., Lin, J., et al. (2004). Glomerular tip lesion: a distinct entity within the minimal change disease/focal segmental glomerulosclerosis spectrum. Kidney Int, 65, 1690–702. Stokes, M. B., Valeri, A. M., Markowitz, G. S., et al. (2006). Cellular focal segmental glomerulosclerosis: Clinical and pathologic features. Kidney Int, 70, 1783–92. Toth, C. M., Pascual, M., Williams, W. W., Jr., et al. (1998). Recurrent collapsing glomerulopathy. Transplantation, 65, 1009–10. Velosa, J. A., Donadio, J. V., Jr., and Holley, K. E. (1975). Focal sclerosing glomerulonephropathy: a clinicopathologic study. Mayo Clin Proc, 50, 121–33.

Verani, R. R. and Hawkins, E. P. 1986. Recurrent focal segmental glomerulosclerosis. A pathological study of the early lesion. Am J Nephrol, 6, 263–70. Weiss, M. A., Daquioag, E., Margolin, E. G., et al. (1986). Nephrotic syndrome, progressive irreversible renal failure, and glomerular “collapse”: a new clinicopathologic entity? Am J Kidney Dis, 7, 20–8.

CHAPTER 58

Primary focal segmental glomerulosclerosis: treatment and outcome Alain Meyrier and Patrick Niaudet Introduction The proportion of cases of primary focal segmental glomerulosclerosis (FSGS) responsive to treatment with corticosteroids is variable and depends on histological type, and duration and dose of steroid treatment, but overall complete remission rate is estimated at 20–25% in white and Asian patients, and lower in black patients. Partial response dependent on a high dose of steroids is possible. Despite anxieties about nephrotoxicity, there may be justification for adding calcineurin inhibitors to control nephrotic syndrome if it is severe. Data for additional agents are not very encouraging. Plasma exchange appears to remove a circulating factor that causes proteinuria in FSGS, as illustrated by responses to this treatment when proteinuria recurs after kidney transplantation.

Treatment The principles of management of FSGS are similar in children and adults. A genetic cause is more likely in young children, but may also present in adults, especially young adults.

General management The general management of nephrotic syndrome is described in Chapter 52.

Aetiologic treatment It can be considered as a paradox to envisage an aetiologic treatment of ‘primary’ nephrotic FSGS, as the aetiology of this condition is still unknown. This would be based on the assumption that FSGS is an immunological, autoimmune disease. We give further our reasons for believing that the favourable effect of some and not all immunosuppressive regimens are not a sufficient argument to endorse the concept of primary FSGS being an immunologic disease. This applies to corticosteroids, calcineurin inhibitors, antimetabolites, and rituximab (RTX) (see Table 45.1).

Corticosteroids The mode of action of corticosteroids in idiopathic nephrotic syndrome was long considered as being that of immunosuppression and of an anti-inflammatory direct effect (Buttgereit et al., 2005).

Recent research points to another interpretation based on a pharmacologic, specifically antiproteinuric effect of steroids on the podocyte. Xing et al. studied the direct effects of the glucocorticoid dexamethasone at concentrations designed to mimic in vivo therapeutic corticosteroid levels (Xing et al., 2006). A conditionally immortalized human podocyte cell line was transfected with a temperature-sensitive simian virus 40 (SV40) transgene. When the SV40 transgene was inactivated in vitro, these cells adopted the phenotype of differentiated podocytes. The study confirmed that expression of glucocorticoid receptors by podocytes was replicated by their podocyte cell line in vitro. Glucocorticoid receptors were present in both nuclear and cytoplasmic extracts. There was a suggestion that overall level of expression and nuclear localization of glucocorticoid receptors was upregulated by dexamethasone in a dose-dependent manner. Dexamethasone upregulated expression of nephrin and tubulin-a and downregulated vascular endothelial growth factor. Effects on the cell cycle comprised downregulation of cyclin kinase inhibitor p21 (that promotes podocyte proliferation) and augmentation of podocyte survival, without any effect on apoptosis. Cytokine production by podocytes, especially interleukin (IL)-6 and IL-8. IL-6 expression was suppressed by dexamethasone. Notwithstanding these recent notions and from a practical standpoint, nephrotic, primary FSGS represents a constant indication of first-line corticosteroid treatment (Korbet, 2002; Chun et  al., 2004; Braun et al., 2008; Meyrier, 2009b). However, the response of nephrotic FSGS to steroids depends on several factors. The degree of fibrotic glomerular and tubulointerstitial lesions when treatment is undertaken matters, with usually the poorest results when serum creatinine concentrations are > 150–200 µmoles per litre. This indicates that treatment should be undertaken early in the course of FSGS. There is some wishful thinking in this assertion as the clinical onset of nephrotic syndrome, which marks the time when a kidney biopsy is performed is not as explosive in FSGS as it is in MCD, except in the glomerular tip lesion (GTL) variant (Stokes et al., 2004). It is conceivable that the kidney biopsy be carried out following a protracted period of clinically silent glomerular injury by the offending factor that causes the disease. The histologic subtype according to the Columbia classification also matters as the rate of complete, partial remission, and failure is respectively approximately 59%, 14%, and 27% in the GTL

526

Section 3  

the patient with glomerular disease

(Stokes et al., 2004), whereas 72–93% of collapsing glomerulopathies are treatment failures (Albaqumi et al., 2006; Stokes et al., 2006). A series of 187 patients with FSGS was analysed by Stokes et al. according to the histopathologic type (Stokes et al., 2006). Seventy two patients were treated with steroids (the figures for the cell, the collapsing (CG), the GTL, and the NOS (not otherwise specified) variants were respectively 77.8%, 66.0%, 97.2%, and 63.0%). Other treatments comprised in a non-systematic fashion calcineurin inhibitors and/or alkylating agents in some but not all patients. The results show that on a whole the NOS variant reflects the overall rate of complete and partial remission and that the cell and the collapsing variant do not follow the same course, the latter enjoying remission in only 13% of cases whereas the former achieves complete or partial remission in nearly half of cases. The GTL in terms of response to treatment is quite comparable to that of MCD (Stokes et al., 2004). One-third of all patients and one-third of those with the NOS variant progressed to end-stage renal disease as compared with one-quarter for the cell subtype, two-thirds for CG, and only one-twentieth for the GTL. Two drugs are commonly used:  prednisone or prednisolone orally. The initial dosage in adults is in the order of 1 mg/kg/day without exceeding 80 mg/day in obese patients. The time to remission is slow and in adults the cumulative success rate of steroids requires about 4  months before pronouncing steroid resistance (Nakayama et al., 2002). Prednisone is typically given for 12 weeks, followed in cases of remission, even partial, by slow tapering over months to avoid a rebound effect. On a whole, patients with nephrotic FSGS do not enjoy complete (and more often partial) remission in more than approximately 20–25% of cases but with a wide spectrum of responses according to the mode of treatment and as illustrated above to the histologic subtype. In white and Asian patients the overall rate of complete remission with corticosteroids alone is in the order of 20–25% (Braun et al., 2008). Conversely the response to therapy is extremely low in black patients (Schwartz et al., 1999; Crook et al., 2005). Complete remission portends a favourable outcome (Korbet, 2002; Chun et al., 2004) and a partial remission is better in terms of renal function than no remission (Troyanov et al., 2005). However, a minority of patients with steroid-sensitive nephrotic FSGS achieve complete stable remission after tailing off steroids to a stop. Most are steroid dependent. When the threshold dose is high, > 10–15 mg/day, the patient is exposed to the well-known, long-term complications of steroids, including hip osteonecrosis. The high incidence of patients with FSGS and steroid dependency on a high threshold dose or steroid resistance leads to discussing other treatment options that imply adding or substituting their own toxic effects to the complications of steroid therapy (Philibert and Cattran, 2008).

and multirelapsing forms, the remission they yield is long-lasting. However, this is mostly true in MCD whereas multiple relapses following periods of complete remission are a very rare scenario in FSGS, although it has been published that, surprisingly, some cases of the worst form of FSGS, that is, collapsing glomerulopathy, achieved a spontaneous remission (Valeri et al., 1996). Steroid resistance, which is the usual case in FSGS is highly predictive of resistance to alkylating agents with the corresponding figures of 17%, 15%, and 69%. The results of cytotoxic drugs used in about 20% of all patients in the series of Stokes et al. were not different amongst the various histologic subtypes (Stokes et al., 2006). Heering et  al. conducted a prospective randomized study in two groups of patients with nephrotic FSGS (Heering et al., 2004). Thirty-four were treated with steroids and ciclosporin, whilst 23 received steroids and chlorambucil for 6 months. When FSGS was refractory to chlorambucil the patients were switched to ciclosporin. The results in terms of renal function and proteinuria were the same in the two groups. The authors concluded that adding treatment with chlorambucil to steroids was ineffective in FSGS. In fact, cytotoxic agents are of no avail in a majority of steroid-resistant patients. Their short- and long-term toxicity has been established. They entail a risk of definitive sterility in the young male and hypofertility in females (Chapman, 1983). Cyclophosphamide is carcinogenic (Faurschou et  al., 2008). Evidence-based analysis leads to considering that alkylating agents are dangerous and mostly inefficient in FSGS (Braun et al., 2008).

Alkylating agents

Ciclosporin

Two cytotoxic agents have been used in the treatment of steroid-resistant or -dependent idiopathic nephrotic syndrome: chlorambucil and cyclophosphamide. The results of cytotoxic agents in idiopathic nephrotic syndrome largely depend on the previous response to steroids. Their best indication is corticosteroid dependency, where the rates of complete remission, partial remission, and failure are, respectively, 51%, 23%, and 26% of cases (Korbet, 1999; 2002; Chun et al., 2004). The main advantage of alkylating agents is that, in steroid-dependent

The mode of action of ciclosporin in reducing proteinuria is not necessarily immunological (Meyrier, 1992, 1999, 2009a, 2009b). Ciclosporin was shown to reduce proteinuria in glomerulopathies with no immunologic background, such as diabetic glomerulopathy and Alport syndrome. This effect was initially attributed to renal vasoconstriction and considered as more noxious than beneficial. In fact ciclosporin is endowed with pharmacological properties that interfere with the glomerular permeability to albumin. Chen et al. in a model of Alport syndrome in the dog showed that

Immunophilin modulators Calcineurin inhibitors Calcineurin inhibitors operate on intracellular signal transduction pathways (Kaminuma, 2008; Rao, 2009). A T-cell-driven immune response develops in three phases. First, transcriptional activation of early genes such as the IL-2 receptor that elicits progression of T cells from the G0 to the G1 state. Second, T cells transduce the signal triggered by specific cytokines that permit entry into the cell cycle, resulting in clonal expansion and effector functions in the third phase of the immune response. Both ciclosporin and tacrolimus bind to the same family of immunophilins, cytosolic FK binding proteins (FKBP). Ciclosporin and FK-506 inhibit the first phase. Both inhibit the nuclear factor of activated T cells (NFAT) signalling in T cells. Two calcineurin inhibitors have been tried in the treatment of idiopathic nephrotic syndrome:  ciclosporin and tacrolimus (FK-506). The rationale for using them is the postulate that all subsets of idiopathic nephrotic syndrome are a T-cell driven immunologic disease. This hypothesis has some consistence in MCD but is not substantiated in FSGS where calcineurin inhibition cannot be considered as an aetiologic treatment (Meyrier, 2009b).

Chapter 58 

ciclosporin reduced albuminuria and preserved renal function, despite the fact that repeat electron microscopy showed that the glomerular basement menbrane lesions were progressing (Chen et al., 2003). Ambalavanan et al. treated nephrotic patients suffering from idiopathic membranous glomerulonephritis with ciclosporin (Ambalavanan et al., 1996); proteinuria diminished significantly and the glomerular filtration rate was unchanged. Repeat electron microscopy showed that the lesions had progressed. Other studies in membranous glomerulonephritis have confirmed a ‘proteinuria-only’ effect of calcineurin inhibitors (see Chapter 62). Recent research has shed a new light on the non-specific effect of ciclosporin on proteinuria. It appears that the drug does not only exert an immunosuppressive action but also an antiproteinuric effect of its own. Faul et al. published an elegant study on the mode of action of ciclosporin on the podocyte (Faul et al., 2008). They examined the effect of ciclosporin from various angles. By confocal laser microscopy they showed that synaptopodin specifically interacts with 14-3-3, an intermediate filament protein. 14-3-3β, E64, and ciclosporin block the cathepsin L-mediated degradation of synaptopodin. Ciclosporin and E64 also ameliorate lipopolysaccharide-induced proteinuria following the cathepsin L mediated degradation of synaptopodin. These experiments collectively demonstrated that the antiproteinuric effect of ciclosporin does not result from the inhibition of NFAT signalling but from blocking the calcineurin-mediated dephosphorylation of the actin-organizing protein synaptopodin, which confers a stabilization of the actin cytoskeleton in podocytes. This antiproteinuric effect of ciclosporin might explain partial, but clinically beneficial remissions in FSGS. It is conceivable that tacrolimus diminishes proteinuria through a similar mechanism, but so far no study has been undertaken on this matter. Since 1985, ciclosporin has been considered amongst the most useful immunosuppressive agents in the treatment of idiopathic nephrotic syndrome, including FSGS (Meyrier, 2009a). The efficacy of ciclosporin depends essentially on the previous response to steroids. In steroid-responsive cases, the percentages of complete remission, partial remission, and failure are, respectively, 73%, 7%, and 20%. In steroid-resistant cases of FSGS, that are by far the most frequent, the respective figures are 29%, 22%, and 49%. Evidence-based analysis of the results of three studies showed that the rate of complete plus partial remissions with 3.5–5 mg/kg of ciclosporin (Sandimmune®) in combination with low-dose (0.15 mg/kg/day) prednisone or prednisolone was significantly increased versus the glucocorticoid alone (Meyrier et al., 1994; Braun et al., 2008). Thus, ciclosporin is a steroid-sparing agent and steroids enhance the efficacy of ciclosporin. This synergic efficacy can be explained by the mode of action of both drugs described above (Xing et al., 2006; Faul et al., 2008). Treatment of FSGS with ciclosporin has been the subject of numerous publications since 1986. In the first article on the effect of ciclosporin in adults with idiopathic nephrotic syndrome (three with MCD and three with FSGS), the authors observed that ‘The results . . . suggest that minimal change lipoid nephrosis and focal segmental glomerulosclerosis are separate entities’ (Meyrier et al., 1986). This is still true. The last review found in 2013 in a Medline search on ciclosporin treatment of FSGS in adults dates back to 2007 (Cattran et al., 2007). Since, all publications deal with idiopathic nephrotic syndrome in children and do not adduce substantial progress on the matter.

fsgs: treatment and outcome

The acquired experience teaches that the drug, when used at low dosages in combination with steroids, has increased efficacy and lower toxicity (Braun et al., 2008; Meyrier, 2009a). A randomized study showed that despite its nephrotoxic potential, long-term ciclosporin treatment slowed the pace to end-stage renal disease in FSGS (Cattran et al., 1999). This unexpected finding might be interpreted as indicating a favourable effect of reduced proteinuria on the tubulointerstitium, more than reflecting an improvement of the glomerular lesions of FSGS (Meyrier et al., 1994). The only controlled study to compare the efficacy (induction of remission) and safety of ciclosporin alone with those of supportive therapy in patients with steroid-resistant idiopathic nephrotic syndrome was carried out by Ponticelli et al. in 14 patients with FSGS (Ponticelli et al., 1993). There were three complete remissions, five partial remissions, and six failures. Yet steroid resistance in adults was pronounced after only 2  months of this treatment, which makes the interpretation of results difficult. However, the same group 6 years later published results indicating that FSGS requires a long course of ciclosporin treatment for obtaining and maintaining a remission (Ponticelli et al., 1999). Guidelines regarding Sandimmune® dosage formulated nearly two decades ago (Meyrier et al., 1994; Niaudet, 1994) still apply and have been the subject of the consensus workshop cited earlier (Cattran et al., 2007). This update comprises caveats regarding the use of ciclosporin generics of the first galenic form of the drug in oily solution. The better bioavailability of the microemulsion (Neoral®) leads to recommending dosages distinctly < 5 mg/kg/day. Ciclosporin dependency was observed from the very first trials of treatment. The probability of remaining in remission after abruptly stopping ciclosporin was 50% at 2 months, 30% at 4 months, 20% at 6 months, and nil at 9 months. This implied the worrisome prospect of indefinite treatment with a nephrotoxic drug. However, the notion of ciclosporin dependency was reconsidered when a series of 36 adults having undergone a repeat kidney biopsy was analysed after 1–5 years of ciclosporin treatment (Meyrier et al., 1994). Fourteen had been treated for 26 ± 14.5 months including four with FSGS. After > 1 year of remission, ciclosporin was progressively tapered to a stop. Remission was durable and maintained without steroids in 11 and with low-dose steroids in three. In five cases patients remained in remission with very low dosages, in the order of 3 mg/kg/day of the oily solution (Sandimmune®), and even in one with 1 mg/kg/day. Ciclosporin dependency to a low dosage most probably entails little renal toxicity over the years. This justifies long-term ciclosporin treatment to maintain remission (Ponticelli et al., 1999).

Tacrolimus (FK506) Tacrolimus is a calcineurin inhibitor whose mode of action is similar to that of ciclosporin. However, FK506 also exerts a pharmacological action on the slit diaphragm molecules to diminish proteinuria in a hereditary autosomal dominant form of FSGS, that is, a mutation of the transient receptor potential cation channel 6 (TRPC6) (Mukerji et  al., 2007). Preliminary experiments reveal that FK-506 can inhibit TRPC6 in vitro through blocking the TRPC6 channels (Winn, 2003, 2008; Winn et  al., 2005; Mukerji et al., 2007). Trials of treatment of FSGS with FK-506 in adults are few and comprise short series of patients (Westhoff and van der Giet, 2007).

527

528

Section 3  

the patient with glomerular disease

Segarra et  al. presented the results of combined therapy with tacrolimus and steroids in 25 patients with nephrotic FSGS in whom ciclosporin had not obtained remission. At 6  months, ten patients enjoyed complete remission, two partial remission, and five a significant reduction of proteinuria (Segarra et  al., 2002). Time to remission was long (112 ± 24 days). Reversible nephrotoxicity was observed in 40%. A majority of patients were FK-506-dependent. Duncan et al. studied prospectively six patients with nephrotic FSGS treated with tacrolimus alone, and a further five in remission on ciclosporin, converted to FK-506 in an attempt to arrest a decline in renal function (Duncan et  al., 2004). They achieved partial remission after 6.5 ± 5.9 months. Proteinuria diminished by 75.2 ± 16.8%. Renal function declined modestly at 3 months and subsequently remained stable. Another group of five patients were converted from ciclosporin to FK-506. In two, FK506 maintained complete remission and the extant three patients had a further reduction in proteinuria. Overall, conversion from ciclosporin to FK-506 was followed by an improvement in renal function, but the follow up was short.

Sirolimus Sirolimus is another immunosuppressive agent used in transplantation, but it may be acutely nephrotoxic in proteinuric glomerulopathies, including FSGS (Fervenza et al., 2004) and may induce de novo FSGS in transplanted patients (Letavernier et  al., 2007). There is one report to the contrary (a paper based on 21 cases that presented rather favourable results of the drug (Tumlin et al., 2006)) but it cannot be recommended.

Antimetabolites Azathioprine Linshaw and Gruskin analysed the results of treatment with azathioprine and concluded that there was a lack of efficacy of this antimetabolite in children (Linshaw and Gruskin, 1974). Conversely, Cade et  al. treated with azathioprine 13 adults in whom idiopathic nephrotic syndrome had appeared in childhood in two and after age 14 in 11 (Cade et al., 1986). Five had FSGS. Six had been steroid resistant from the outset. Seven others were multirelapsers, of whom four evolved to steroid resistance. At 3  months, all patients showed clinical improvement. At 18 months the six patients with selective proteinuria were in remission. At 24  months, 12 out of 13 of the patients still followed up were in complete remission. Since 1986, azathioprine has occasionally been cited in papers dealing with treatment of idiopathic nephrotic syndrome, but there are no hard data on its place in the treatment of FSGS.

Mycophenolate mofetil Since preliminary publications dating back to 1998 (Briggs et al., 1998), mycophenolate mofetil (MMF) had been tried in a few cases of FSGS, amongst other causes of nephrotic syndrome (Choi et al., 2002; Bagga et al., 2003; Barletta et al., 2003; Cattran et al., 2004; Gellermann and Querfeld, 2004; Segarra et al., 2007). Of 11 adults with FSGS, two went into remission with stable renal function. In the others, proteinuria diminished whilst renal function declined. In the three paediatric series, a total of four children were treated with MMF. The drug allowed corticosteroid sparing and seemed to be beneficial in terms of renal function.

Cattran et al. performed an open 6-month trial of MMF in 18 patients resistant to a course of corticosteroid therapy (Cattran et  al., 2004). Seventy-five per cent had also failed to respond to a cytotoxic agent and/or a calcineurin inhibitor. A  substantial improvement in proteinuria was seen in 8/18 of the patients by 6 months. This was sustained for up to 1 year post treatment in 4/8 of this group. No patient had a complete remission. No deterioration in renal function was observed over the treatment period, but three patients progressed to chronic kidney failure during follow-up. Adverse effects were mild. Relapses were common, suggesting that more prolonged or combination therapy may be required. Segarra et  al. published on a series of 22 patients (Segarra et al., 2007). All had received a previous 6-month course of prednisone and a 6-month course of ciclosporin. Five had also been treated with cyclophosphamide and four with FK-506. Over a 12-month follow-up period of MMF, two went into complete remission, 10 into a partial remission, and 10 were failures. When obtained the time to remission was very long, in the order of 150 days. The authors concluded that ‘MMF causes a moderate decrease in proteinuria in 50% of the patients who do not have other treatment options. The response to therapy is largely influenced by a preserved renal function and requires sustained MMF treatment’.

Rituximab The majority of publications on RTX in steroid-dependent and steroid-resistant idiopathic nephrotic syndrome deal with the disease in children and are analysed in Chapter 56. RTX has been tried in very few adult patients with FSGS. From this sparse experience it appears that RTX in this setting is poorly efficient in FSGS that resisted other treatments, including in cases that relapse following transplantation (Yabu et al., 2008; Fernandez-Fresnedo et al., 2009; Kisner et al., 2012).

Other treatment options The hypothesis that FSGS is the consequence of a circulating factor that induces proteinuria, followed by podocyte lesions has led to attempts to remove it by plasmapheresis or by adsorption on staphylococcal protein A-coated columns. Plasmapheresis was considered of doubtful avail in primary FSGS (Feld et  al., 1998)  although a few studies observed a favourable result of plasma and LDL apheresis (Yokoyama et al., 2007). In fact plasmapheresis is widely used in the special case of nephrotic syndrome recurring after transplantation. Relapse of nephrotic syndrome and glomerular lesions are observed in approximately 30% of patients undergoing renal transplantation for end-stage FSGS (Ponticelli, 2010). Recurrence often leads to loss of the transplant, with an increased incidence when the primary disease was a collapsing glomerulopathy (Swaminathan et al., 2006). Repeated sessions of plasmapheresis (Artero et al., 1994) are carried out alone or with increased immunosuppression (Canaud et al., 2009). The efficacy of pre-emptive or curative plasmapheresis per se in these recurrent forms is not clearly established as large, randomized studies are lacking (Gohh et al., 2005). In fact pre-emptive plasmapheresis can be repeated over days and weeks in case of transplantation with a living donor and thus achieve substantial removal of the glomerular permeability factor, whereas in case of deceased donor transplantation

Chapter 58 

there is no sufficient time for performing more than three plasma exchanges before grafting. Plasma protein adsorption on columns coated with staphylococcal protein A have led to a dead end. Dantal et al. had published enticing results of this costly technique (Dantal et  al., 1994), but the excitement was damped when another group showed that immunoadsorption diminished proteinuria in various glomerulopathies with no specificity regarding FSGS (Esnault et al., 1999). Antifibrotic agents have been tried in FSGS as in other glomerulopathies. Endothelin has a fibrosing effect on the injured glomerulus (Barton, 2008)  and animal experiments indicate that endothelin receptor antagonists (ERAs) exert a preventive effect on the progression of glomerulosclerosis in the rat. So far these drugs have only been the subject of anecdotal reports in human FSGS. In fact clinical development of ERAs has been hampered by problems with dosing, with the make-up of study cohorts, and adverse events (Barton and Kohan, 2011). Pirfenidone, an orally available antifibrotic drug (Cho et al., 2007) and the monoclonal antibody adalimumab, an antitumour necrosis factor inhibitor associated with rosiglitazone—an antidiabetic agent—have been the subject of preliminary trials (Peyser et al., 2010).

Treatment of steroid-resistant FSGS in children The prognosis of steroid-resistant FSGS is poor, with a high proportion of children progressing to end-stage renal failure. This explains that intensive treatment regimens have been tried. The results of immunosuppressive treatments should take into account the fact that children with genetic forms of idiopathic nephrotic syndrome most often fail to respond to any therapy. However, many published trials include patients who had not been tested for mutations in the genes involved in steroid-resistant idiopathic nephrotic syndrome. Moreover, most studies are non-randomized and include a small number of patients.

Calcineurin inhibitors A combination of calcineurin inhibitor with low-dose steroid therapy for at least 6 months is presently the best known option as a first-line therapy.

Ciclosporin The rate of complete remission is significantly higher when ciclosporin is given in combination with steroids (Niaudet and Habib, 1994). Three randomized trials involving 49 children showed that complete remissions and partial remissions were observed in 31% and 38% of patients which was significantly higher than in the control arms (Garin et al., 1988; Ponticelli et al., 1993; Lieberman and Tejani, 1996). Ingulli et al. reported that prolonged ciclosporin treatment in children with steroid-resistant FSGS reduces proteinuria and blunts the progression to end-stage renal failure (Ingulli et al., 1995). The dose of ciclosporin (4–20 mg/kg/day) was titrated to the serum cholesterol level to achieve a remission. In this study, only 5 of the 21 treated patients (24%) progressed to end-stage renal failure compared to 42 of 54 patients from an historical group who had not received this treatment. Ehrich et al. reported a retrospective study including 25 children with steroid-resistant FSGS who received prolonged and

fsgs: treatment and outcome

intensified treatment with combined ciclosporin and steroids including methylprednisolone pulses. This treatment resulted in sustained remission in 84% of children with non-genetic forms of steroid resistant idiopathic nephrotic syndrome (Ehrich et al., 2007).

Tacrolimus There is evidence from case series that tacrolimus is effective in children with steroid-resistant FSGS (Loeffler et al., 2004; Bhimma et  al., 2006; Gulati et  al., 2008). Tacrolimus was found to be as effective as ciclosporin in a randomized trial involving 41 children (Choudhry et al., 2009). Both groups received alternate-day steroids and enalapril. The rate of remission at 6 months was 85.7% with tacrolimus and 80% with ciclosporin. Interestingly, the proportion of patients with relapses was significantly higher in the ciclosporin group. Conversely, Wang et al. found that tacrolimus was associated with higher efficacy and lower renal toxicity in comparison to ciclosporin, with no difference in the rate of relapse (Wang et al., 2012).

Pulse methylprednisolone Methylprednisolone pulse therapy has been proposed by Mendoza et al. It consists of methylprednisolone (30 mg/kg intravenously), administered every other day for 2 weeks, weekly for 8 weeks, every other week for 8 weeks, monthly for 9  months, and then every other month for 6 months in association with oral prednisone and, if necessary, cyclophosphamide or chlorambucil (Mendoza et al., 1990). At an average of > 6 years of follow-up, 21 of 32 children were in complete remission and the 5-year incidence of end-stage renal disease was approximately 5% versus 40% in historical controls (Tune et al., 1995). Two publications reported similar results (Yorgin et al., 2001; Pena et al., 2007). Although these results are better than those seen in any other study, other reports described less favourable results (Waldo et al., 1992; Hari et al., 2001).

Alkylating agents Although alkylating agents have little therapeutic effect in steroid-resistant FSGS, for unknown reasons they are still widely used either alone or in combination with corticosteroids. The International Study of Kidney Disease in Children reported on 60 children with steroid-resistant FSGS who were randomly allocated to receive either prednisone 40 mg/m2 on alternate days for 12 months (control group) or cyclophosphamide, 2.5 mg/kg body weight for 3 months plus prednisone 40 mg/m2 on alternate days for 12  months (Tarshish et  al., 1996). Complete remissions were observed in 28% of children in the control group and in 25% of children who received cyclophosphamide. The authors concluded that there was no beneficial effect of cyclophosphamide in these patients. Rennert et al. treated 10 children with steroid-resistant FSGS with cyclophosphamide pulses. Only two of the five initial non-responders went into remission whereas all five late non-responders achieved complete remission (Rennert et al., 1999). In a prospective study of 24 patients, Bajpai et al. also found that therapy with intravenous cyclophosphamide had limited efficacy in patients with initial corticosteroid resistance whilst sustained remission was likely to occur in patients with late resistance and those with absence of significant tubulointerstitial changes on renal histology (Bajpai et al., 2003).

Mycophenolate mofetil There is no convincing data for the beneficial effect of MMF in children with steroid-resistant FSGS. Menzibal et  al. treated five

529

530

Section 3  

the patient with glomerular disease

patients and only one achieved complete remission (Mendizabal et al., 2005). MMF in association with methylprednisolone pulses and angiotensin converting-enzyme inhibitors was reported to significantly reduce proteinuria (Montane et al., 2003). Another study involving 52 patients found a rate of complete remission of 23% and partial remission of 35.5% following therapy with MMF (de Mello et al., 2010). A prospective trial of the NIH compared ciclosporin to a combination of oral pulse dexamethasone and MMF (Gipson et al., 2011). Partial or complete remission was achieved in 22 of the 66 patients in the mycophenolate/dexamethasone group and 33 of the 72 ciclosporin-treated patients at 12 months. The authors concluded that the small sample size might have prevented detection of a moderate treatment effect.

Rituximab At present, there is no evidence that RTX is effective in patients with steroid-resistant nephrotic syndrome, although retrospective case series report that treatment with RTX was effective in some patients (Peters et al., 2008; Gulati et al., 2010; Prytula et al., 2010; Sinha and Bagga, 2013). Gulati et al. reported that RTX had induced a complete remission in 27% of 33 steroid-resistant patients and a partial remission in 21% of them. The rate of remission was higher in patients with minimal changes on renal biopsy and in patients who were late non-responders (Gulati et al., 2010). Magnasco et al. reported the results of an open-label randomized trial including 31 children aged 2 to 16 years. All received calcineurin inhibitors and prednisolone and 16 of them received in addition two RTX infusions. Proteinuria remained unchanged in these patients and none entered partial or complete remission (Magnasco et al., 2012).

Post-transplantation recurrence of FSGS (See also Chapter 289.) Post-transplantation recurrence of nephrotic FSGS is frequent, and affects about 30–50% of patients (Ponticelli, 2010). Recurrence, which often occurs within the first hours after transplantation, is characterized by profuse proteinuria with foot process fusion followed by histologic lesions of FSGS. Recurrence leads to loss of the graft in at least half of cases. The risk is particularly high in patients whose primary glomerulopathy was of the cellular and the collapsing variants. Other risk factors are a rapid progression to renal failure ( 60 or < 20  years), presence of weight loss, rash, arthritis, or risk factors for hepatitis, raise the likelihood that the MGN is secondary in nature. Due to the wide range of causes of secondary MGN, the clinical features present at diagnosis are more variable. History should carefully probe exposure to drugs and toxins. Skin-lightening creams have been repeatedly associated with MGN (see Chapter 82) and should be directly asked about. Laboratory testing may reveal positive autoimmune serology (i.e. antinuclear and anti-DNA antibodies, and/or low complement level), positive hepatitis serology, or cryoglobulin tests, in contrast with primary MGN, in which these investigations are negative or normal.

Pathology The hallmark finding in MGN is the presence of subepithelial immune complex deposits, best seen on electron microscopy (EM) (see Fig. 61.1). These are found in cases of both primary and secondary forms of the disease. The principal findings evident in MGN

Fig. 61.1  Silver stain revealing ‘spikes’, representing growth of the glomerular basement membrane matrix between and around subepithelial immune deposits. Image courtesy of Dr Andrew Herzenberg and Dr Rohan John.

Chapter 61 

membranous: clinical features and diagnosis

used system to describe the histopathologic variations seen on biopsy by EM: ◆ Stage

I—subepithelial deposits: at this stage of disease, LM findings are frequently normal. The GBM is usually normal in thickness, with minimal evidence of spike formation. A few small, flat, electron-dense deposits may be seen on the epithelial surface of the GBM.

◆ Stage

II—spike formation: spikes protruding from epithelial surface of GBM become clearly visible. The spikes extend between the electron-dense deposits, and are present in virtually every capillary loop. With progression, the spike tips may have a widened or clubbed appearance. The number and size of deposits are increased compared with stage I.

◆ Stage

Fig. 61.2  IgG immunofluorescence in membranous glomerulonephritis. Image courtesy of Dr Andrew Herzenberg and Dr Rohan John.

of the earliest changes seen by light microscopy (LM) is a ‘moth eaten’ appearance of the basement membrane when observed ‘en face’ using silver stains (Kern et al., 1999). Silver stain is also useful to observe the linear ‘spike’ projections protruding from the outer (epithelial) surface of the GBM (Fig. 61.2). With disease progression, and larger numbers of immune deposits, capillary walls and the GBM may appear globally thickened. The mesangium usually exhibits normal cellularity in idiopathic MGN. With advanced disease, segmental and global glomerulosclerosis may be observed. The tubulointerstitial compartment, and vessels are usually unremarkable in early disease. With disease chronicity, however, tubulointerstitial atrophy and fibrosis are frequently observed. Vascular injury with arterial and arteriolar sclerosis may be evident.

Immunofluorescence The characteristic finding on IF microscopy is granular capillary wall staining for immunoglobulin and complement. IgG is most commonly present. The predominant IgG subclass represented in idiopathic MGN biopsies is IgG4 (Doi et al., 1984). Complement C3 is commonly found by IF within deposits in MGN, despite the lack of predilection for IgG4 to fix complement. In patients where a broader range of immunoglobulins, or of IgG subtypes are found, secondary MGN is more likely (see ‘Pathologic findings in secondary MGN’). A more diverse spectrum of IgG subtypes may be observed in MGN secondary to lupus (Haas, 1994)  and IgG1 and IgG2 may be found more commonly in MGN diagnosed in the context of a malignancy (Ohtani et al., 2004). The finding of intense C1q staining, and the presence of other immunoglobulins such as IgM and IgA are suggestive of lupus-associated membranous nephropathy (Jennette et al., 1983).

Electron microscopy EM is essential for diagnosing MGN, and is especially useful in detecting changes that may suggest SLE as a primary cause. Several ‘staging’ systems have been used to describe the pathologic changes observed (Ehrenreich and Churg, 1968; Bariety et al., 1970; Gartner et al., 1977). A modified version of the Ehrenreich and Churg scale, which includes the addition of a stage V lesion, is the most widely

III  –incorporation of deposits:  electron-dense deposits become surrounded by and incorporated into the GBM. This results in an irregular thickening of the GBM, and capillary wall. The capillary wall may appear to have a split appearance due to interruption of the GBM by this immune-complex material.

◆ Stage IV—disappearing deposits: the deposits incorporated within

the GBM lose their electron density. As a result, the GBM has a very irregular, thickened appearance. Areas of the GBM occupied by deposit material which has lost electron density will have a vacuolated or lucent appearance.

◆ Stage

V—repair stage:  during this ‘healing’ phase, the deposits have become completely rarified, and the GBM appearance is returning to normal. The GBM appears partially thickened, and may still contain areas of lucency. Some consider stage V to include an ‘end-stage’ appearance, with glomerular obsolescence and sclerosis (Donadio et al., 1988).

In addition to the pathologic changes described by the staging system, diffuse effacement of visceral epithelial cell foot processes is also frequently observed on EM in MGN. This finding may be associated with heavy proteinuria. A review of 350 biopsies of patients with primary MGN at the University of North Carolina, United States, reveals that most patients present with stage I or II biopsy findings (total 70%) (Falk et al., 2000). Some disagreement exists regarding whether a biopsy should be staged according to the predominant lesion observed in the majority of glomeruli, versus the most advance lesion evident in the specimen. Tornroth et al. (1987) first demonstrated that in biopsies of patients with a protracted clinical course, lesions representing all stages of disease are often present in one specimen. They suggested that based upon serial biopsies of patients with MGN, the presence of subepithelial deposits correlate with an ‘active’ clinical stage, whereas replacement by lucent intramembranous lesions occurred during clinical remission. Cessation of formation of immune complexes may result in the presence of intramembranous healing changes, such as in a patient who has one ‘flare’ of proteinuria, then achieves a remission. However, ongoing formation of immune complexes, as may occur in the patient with a more protracted and progressive course, may result in the continued formation of new subepithelial deposits that will be seen in addition to more ‘chronic’ intramembranous lesions. In addition, deposits which are large and extend from the subepithelial space deep into the membrane are also suggestive of continuous deposition of immune complex material.

541

542

Section 3  

the patient with glomerular disease

Pathologic findings in secondary MGN Emerging data suggest that the presence or absence of anti-PLA2R antibodies may be informative to distinguish idiopathic versus secondary MGN; this is discussed further in Chapter 64. However some pathological features are suggestive of a secondary aetiology. A more diverse spectrum of IgG subtypes is characteristic of MGN secondary to lupus (Haas, 1994), and IgG1 and IgG2 may be found more commonly in MGN diagnosed in the context of a malignancy (Ohtani et al., 2004). The finding of intense C1q staining, and the presence of other immunoglobulins such as IgM and IgA are suggestive of lupus-associated membranous nephropathy (Jennette et al., 1983). The location of immune deposits differed in 28 patients with lupus from Jennette et al.’s series of 170 patients with a pathologic diagnosis of MGN (Jennette et  al., 1983). The final diagnosis of systemic lupus erythematosus (SLE) was based on serologic and clinical criteria. Mesangial, subendothelial, and tubular basement membrane electron dense deposits in addition to the subepithelial deposits were highly suggestive of SLE-related MGN. In primary MGN, immune complex formation occurs in situ, along the basal surface of the podocyte. Other mechanisms such as immune-complex formation in the circulation, or deposition of foreign antigen, may operate in secondary causes of MGN, so that deposits may be found throughout the glomerulus, as opposed to only in the subepithelial zone. Although not routinely sought, hepatitis antigens may be found in in the glomeruli of MGN due to hepatitis B; in one study, 100% of biopsies were positive for glomerular core antigen, and 88% for the e-antigen, with only a minority of biopsies demonstrating the presence of surface antigen deposition (Lin, 1990). Mesangial hypercellularity is uncommon in idiopathic MGN, and may be suggestive of a systemic disease. Other suggestive findings include tubuloreticular inclusions (associated with viral or SLE-associated MGN), and intense C1q staining (suggest SLE). The presence of crescents in MGN biopsies is a rare finding, and even in context of minimal subepithelial deposits and proliferation this lesion should elicit consideration of underlying severe lupus nephritis (Basford et al., 2011). Multiple reports have described cases of anti-GBM antibody glomerulonephritis, occurring in the presence of or diagnosed shortly after the development of biopsy-proven membranous nephropathy (see Chapter  72) (Klassen et  al., 1974, reviewed in Basford et  al., 2011). In rarer instances, a necrotizing crescentic glomerulonephritis may also be observed and this may be accompanied by the presence of antineutrophil cytoplasmic antibodies (Nasr et al., 2009).

References Abe, S., Amagasaki, Y., Konishi, K., et al. (1986). Idiopathic membranous glomerulonephritis: aspects of geographical differences. J Clin Pathol, 39, 1193–8. Abrass, C. K. (1985). Glomerulonephritis in the elderly. Am J Nephrol, 5, 409–18. Adu, D. and Cameron, J. S. (1989). Aetiology of membranous nephropathy. Nephrol Dial Transplant, 4, 757–8. Bariety, J., Druet, P., Lagrue, G., et al. (1970). ‘Extra-membranous’ glomerulopathies (E.M.G.). Morphological study with optic microscopy, electron microscopy and immunofluorescence. Pathol Biol (Paris), 18, 5–32. Basford, A. W., Lewis, J., Dwyer, J. P., et al. (2011). Membranous nephropathy with crescents. J Am Soc Nephrol, 22(10), 1804–8.

Braden, G. L., Mulhern, J. G., O’Shea, M. H., et al. (2000). Changing incidence of glomerular diseases in adults. Am J Kidney Dis, 35, 878–83. Cahen, R., Francois, B., Trolliet, P., et al. (1989). Aetiology of membranous glomerulonephritis: a prospective study of 82 adult patients. Nephrol Dial Transplant, 4, 172–80. D’Agati, V. (1994). The many masks of focal segmental glomerulosclerosis. Kidney Int, 46, 1223–41. Davison, A. M., Cameron, J. S., Kerr, D. N., et al. (1984). The natural history of renal function in untreated idiopathic membranous glomerulonephritis in adults. Clin Nephrol, 22, 61–7. Doi, T., Mayumi, M., Kanatsu, K., et al. (1984). Distribution of IgG subclasses in membranous nephropathy. Clin Exp Immunol, 58, 57–62. Donadio, J. V., Jr. (1990). Treatment of glomerulonephritis in the elderly. Am J Kidney Dis, 16, 307–11. Donadio, J. V., Jr., Torres, V. E., Velosa, J. A., et al. (1988). Idiopathic membranous nephropathy: the natural history of untreated patients. Kidney Int, 33, 708–15. Ehrenreich, T. and Churg, J. (1968). Pathology of membranous nephropathy. In S.C. Scommers (ed.) Pathology Annual, pp.145–86. New York : Appleton-Century-Crofts. Ehrenreich, T., Porush, J. G., Churg, J., et al. (1976). Treatment of idiopathic membranous nephropathy. N Engl J Med, 295, 741–6. Falk, R., Jennette, J. C., and Nachman, P. H. (2000). Primary glomerular disease. In B. M. Brenner (ed.) Brenner and Rector’s The Kidney, pp. 1263–92. Philadelphia, PA: W. B. Saunders. Gartner, H. V., Watanabe, T., Ott, V., et al. (1977). Correlations between morphologic and clinical features in idiopathic perimembranous glomerulonephritis. A study on 403 renal biopsies of 367 patients. Curr Top Pathol, 65, 1–29. Glassock, R. J. (1992). Secondary membranous glomerulonephritis. Nephrol Dial Transplant, 7 Suppl 1, 64–71. Gluck, M. C., Gallo, G., Lowenstein, J., et al. (1973). Membranous glomerulonephritis. Evolution of clinical and pathologic features. Ann Intern Med, 78, 1–12. Guella, A., Akhtar, M., Ronco, P. (1997). Idiopathic membranous nephropathy in identical twins. Am J Kidney Dis, 29, 115–8. Haas, M. (1994). IgG subclass deposits in glomeruli of lupus and nonlupus membranous nephropathies. Am J Kidney Dis, 23, 358–64. Hay, N. M., Bailey, R. R., Lynn, K. L., et al. (1992). Membranous nephropathy: a 19 year prospective study in 51 patients. N Z Med J, 105, 489–91. Hladunewich, M. A., Troyanov, S., Calafati, J., et al. (2009). The natural history of the non-nephrotic membranous nephropathy patient. Clin J Am Soc Nephrol, 4, 1417–22. Honkanen, E. (1986). Survival in idiopathic membranous glomerulonephritis. Clin Nephrol, 25, 122–8. Honkanen, E., Tornroth, T., and Grönhagen-Riska, C. (1992). Natural history, clinical course and morphological evolution of membranous nephropathy. Nephrol Dial Transplant, 7 (Suppl 1), 35–41. Jennette, J. C., Iskandar, S. S., and Dalldorf, F. G. (1983). Pathologic differentiation between lupus and nonlupus membranous glomerulopathy. Kidney Int, 24, 377–85. Kern, W. F., Silva, F. G., Laszik, Z. G., et al. (1999). Atlas of Renal Pathology. Philadelphia, PA: W. B. Saunders. Kida, H., Asamoto, T., Yokoyama, H., et al. (1986). Long-term prognosis of membranous nephropathy. Clin Nephrol, 25, 64–9. Klassen, J., Elwood, C., Grossberg, A. L., et al. (1974). Evolution of membranous nephropathy into anti-glomerular-basement-membrane glomerulonephritis. N Engl J Med, 290(24), 1340–4. Korbet, S. M., Genchi, R. M., Borok, R. Z., et al. (1996). The racial prevalence of glomerular lesions in nephrotic adults. Am J Kidney Dis, 27, 647–51. Lin, C. Y. (1990). Hepatitis B virus-associated membraneous nephropathy: clinical features, immunological profiles and outcome. Nephron, 55, 37–44.

Chapter 61 

MacTier, R., Boulton Jones, J. M., Payton, C. D., et al. (1986). The natural history of membranous nephropathy in the West of Scotland. QJM, 60, 793–802. Maisonneuve, P., Agodoa, L., Gellert, R., et al. (2000). Distribution of primary renal diseases leading to end-stage renal failure in the United States, Europe, and Australia/New Zealand: results from an international comparative study. Am J Kidney Dis, 35, 157–65. Mallick, N. P., Short, C.D., Manos, J. (1983). Clinical membranous nephropathy. Nephron, 34, 209–19. Murphy, B. F., Fairley, K.F., and Kincaid-Smith, P.S. (1988). Idiopathic membranous glomerulonephritis: long-term follow-up in 139 cases. Clin Nephrol, 30, 175–81. Nasr, S. H., Said, S. M., Valeri, A. M., et al. (2009). Membranous glomerulonephritis with ANCA-associated necrotizing and crescentic glomerulonephritis. Clin J Am Soc Nephrol, 4, 299–308. Noel, L. H., Zanetti, M., Droz, D., et al. (1979). Long-term prognosis of idiopathic membranous glomerulonephritis. Study of 116 untreated patients. Am J Med, 66, 82–90. Ohtani, H., Wakui, H., Komatsuda, A., et al. (2004). Distribution of glomerular IgG subclass deposits in malignancy-associated membranous nephropathy. Nephrol Dial Transplant, 19, 574–9. Passerini, P., Como, G., Viganò, E., et al. (1993). Idiopathic membranous nephropathy in the elderly. Nephrol Dial Transplant, 8, 1321–5.

membranous: clinical features and diagnosis

Ramzy, M. H., Cameron, J. S., Turner, D. R., et al. (1981). The long-term outcome of idiopathic membranous nephropathy. Clin Nephrol, 16, 13–9. Rosen, S. (1971). Membranous glomerulonephritis: current status. Human Pathology, 2, 209–31. Row, P. G., Cameron, J. S., Turner, D. R., et al. (1975). Membranous nephropathy. Long-term follow-up and association with neoplasia. QJM, 44, 207–39. Schena, F. P. (1997). Survey of the Italian Registry of Renal Biopsies. Frequency of the renal diseases for 7 consecutive years. The Italian Group of Renal Immunopathology. Nephrol Dial Transplant, 12, 418–26. Schieppati, A., Mosconi, L., Perna, A., et al. (1993). Prognosis of untreated patients with idiopathic membranous nephropathy. N Engl J Med, 329, 85–9. Tornroth, T., Honkanen, E., Pettersson, E. (1987). The evolution of membranous glomerulonephritis reconsidered: new insights from a study on relapsing disease. Clin Nephrol, 28, 107–17. United States Renal Data System (2009). USRDS 2009 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Zucchelli, P., and Pasquali, S. (1998). Membranous nephropathy. In A. M. Davison, J. S. Cameron, J. -P. Grunfeld, et al. (ed.) Oxford Textbook of Clinical Nephrology (2nd ed.), pp. 571–90. Oxford: Oxford University Press.

543

CHAPTER 62

Membranous glomerulonephritis: treatment and outcome Daniel C. Cattran and Heather N. Reich Natural history There have been few studies of untreated subjects with primary MGN, and multiple inconsistencies in the literature contribute to the difficulty establishing the natural history reliably. The outcomes of secondary MGN are considered in Chapter 63. Marx and Marx’s (Marx and Marx, 1997) review emphasizes the lack of uniformity with respect to diagnostic criteria, selection of endpoints, baseline severity, statistical techniques, and assessment of treatment effects. Noting these shortcomings, Table 62.1 shows studies examining the natural history of membranous glomerulonephritis (MGN) selected on the basis of relatively large sample size, and including a significant percentage of patients who received only supportive therapy. In addition, these studies provided data regarding the clinical status at diagnosis, and clear statistics regarding patient outcomes. The disease course of MGN is considered to be indolent when left untreated, as is evident in most of the studies quoted in this table. The 10-year survival rates are favourable, averaging 82%. The study by Murphy et al. (1988) also provides information regarding survival if the patients had nephrotic syndrome at presentation; the 5-year and 8-year survival were 85% and 82% respectively, which did not differ from the overall patient survival. This relatively good prognosis is likely an underestimate of today’s outcome, given the introduction within the last decade of more potent antihypertensive medications, angiotensin-converting enzyme inhibitors (ACEIs), and lowered ideal levels for target blood pressure. However, this assumption is speculative, as there are still no data indicating that these advances have yielded an improvement in the natural history of this disease. Despite intentional selection of studies in Table 62.1 based upon some degree of uniformity in data and sample size, the variability in outcome evident in the table illustrate some of the limitations of the available literature. For example, the year of diagnosis and entry into study has important implications; the accuracy of the diagnosis may even be questioned, given the lack of electron microscopy prior to around 1975. In addition, some discrepancies in outcome may reflect differences in the quality of ‘conservative therapy’ that have evolved over time. The study by Noel et al. (1979), for instance, included patients diagnosed between 1958 and 1975, and is potentially subject to inconsistencies in diagnosis, which may explain differences in outcome compared with the other studies.

Another element to be considered when interpreting the data is the presence or absence of nephrotic syndrome at diagnosis. In the Donadio et al. (1988) study, 83% had nephrotic syndrome at diagnosis, compared to only 60% in the study by Kida et al. (1986). This may explain part of the differences in survival observed, given that the presence of the nephrotic syndrome impacts negatively on prognosis. One large study by Zucchelli et al. (1987) including 82 subjects with at least 10 years of follow-up reported far more ominous prognosis, with 20% reaching end-stage renal disease (ESRD), 17% chronic renal insufficiency, and 20% of patients suffering from non-renal mortality. These surprising results may be due to the methodological differences described, and the study was therefore not included in the table. For instance, all patients were diagnosed prior to 1976, and 100% had the nephrotic syndrome. A final element to be considered is the control group used for comparison (if available). The study by Donadio et al. (1989) compares survival to actuarial life tables in healthy individuals, as opposed to comparing survival of patients who did or did not receive treatment (Zucchelli et al., 1987). Most recently, a large retrospective review of 328 subjects with the nephrotic syndrome, confirmed spontaneous remission in approximately one-third of subjects (Polanco et al., 2010). Although spontaneous remission rates were lower in patients with higher grades of proteinuria, approximately 20–25% of subjects with proteinuria of 8–12 g/day or more achieved spontaneous remission. The time to complete spontaneous remission, however, ranged from 25 to 41 months. Patients who did not achieve a spontaneous remission had a high rate of non-renal morbidity and mortality in addition to poor renal outcomes. Multiple reviews combining studies of pooled treated and untreated patients have examined overall survival rates. A review of 11 reports of the natural history of idiopathic MGN revealed an overall renal survival of 65–90% by 10 years (Cattran et al., 1992). Similarly, a large pooled analysis of 32 studies examined the course of 1189 patients, estimated renal survival at 86% at 5 years, 65% at 10 years, and 60% at 15 years (Hogan et al., 1995). Non-renal-related morbidity and mortality in patients with idiopathic MGN is difficult to determine from the literature. There are, however, quite alarming mortality rates in several of the earlier studies of the natural history of the disease. Reviews of these earlier papers have indicated that a high proportion of deaths in

Chapter 62 

Table 62.1  Selected studies of the natural history of idiopathic MGN. Survival rates reflect overall patient survival, that is, patient alive with adequate independent renal function, with number of patients at risk in parentheses when available

membranous: treatment and outcome

cardiovascular disease or malignancy (Honkanen, 1986). The reasons for this high rate of premature non-renal deaths may be comorbid conditions, the effects of the disease itself (i.e. complications of the nephrotic syndrome) or of the treatment. The precise cause of death in many of the aforementioned studies is unclear. An as yet unidentified element of glomerulonephritis may increase the risk of mortality. A compelling review of 2380 subjects who underwent biopsy for glomerulonephritis revealed that by 10 years following diagnosis, 32% of the subjects had died (Heaf et al., 1999). More recently, it was demonstrated that the achievement of a remission is important with respect to minimizing morbidity and mortality; in patients who do not achieve a spontaneous remission of nephrotic syndrome have a substantially higher rate of mortality compared to those who do achieve this endpoint (10.7% vs 1.9%, P = 0.002) (Polanco et al., 2010). This is an area in the natural history of MGN about which little information exists, and where further investigation is certainly warranted.

Study (year, country)

Number of subjects

Mean follow-up in months

Outcomes

Noel et al. (1979, France)

116

54

CR 23.5%, PR 14.5%, NR 43%, deterioration 19% (9.5% ESRD) 88% 5 YS (N = 60), 76% 10 YS (N = 12)

Davison et al. (1984, Great Britain)

64

N/A—range 2–15 years

47% no change, 42% doubling creatinine, or level > 400 μmol/L, 8% slow deterioration, 3% excluded

Kida et al. (1986, Japan)

104 (59 received no treatment)

138

CR 40%, PR 30%, 10–15% persistent disease 94.% 5 YS (N = 72), 90% 10 YS (N = 56), 80% 15 YS (N = 34)a

Prognostic factors

MacTier et al. (1986, Scotland)

37

64

30% CR, 16% persistent proteinuria, 19% proteinuria and CRF, 22% ESRD, 13% non-renal mortality

Sex

Donadio et al. (1988, USA)

140 (89 received no treatment)

127.4

64% stable, 20% ESRD, 16% CRF 85% 5 YS, 71% 10 YS.a

Murphy et al. (1988, Australia)

139 (79 received no treatment)

52

50% alive, normal function, 13% CRF, 6% ESRD, 11% non-renal death, 20% lost to follow-up 88% 5 YS (N = 68), 81% 10 YS (N = 38)a

Schieppati et al. (1993, Italy)

100

52

68% normal function, 18% CRF, 14% ESRD (including 6% mortality) 5YS (renal) 88% (N = 37), 8 YS 73%

a Survival rates calculated using both treated and untreated patients. Statistical analysis

indicated no difference in survival between treated and non-treated groups. CR = complete remission; CRF = chronic renal failure; ESRD = end-stage renal disease; NR = no response; PR = partial remission; YS = year survival.

this population—up to 60%—are non-renal related (Donadio et al., 1988; Laluck and Cattran, 1999). One early study of 32 subjects, for instance, indicated 41% mortality in the follow-up period (Franklin et al., 1973). A high mortality rate was confirmed in larger more recent series as well, where 6–20% of patients died by the end of the follow-up period (Kida et  al., 1986; Zucchelli et  al., 1987; Murphy et al., 1988; Wehrmann et al., 1989). These deaths tend to occur at a young age (mean 51 years), and are most often due to

In univariate analysis, gender, advanced age, hypertension, renal insufficiency at presentation, and urinary protein excretion all appear to be predictive. These can be combined into a predictive equation. Males have a worse prognosis. This was first observed in a study of prednisone for treatment of MGN (Collaborative Study of the Adult Idiopathic Nephrotic Syndrome, 1979), and later confirmed in a case series (Hopper et  al., 1981). Two studies in 1984 confirmed by univariate analysis that male gender is associated with an unfavourable prognosis (Davison et al., 1984; Tu et al., 1984). The gender ratio at presentation versus at end stage of disease also illustrates this effect. In several populations, it has been observed that the gender ratio is almost equal at presentation of MGN (Abe et  al., 1986; Harrison 1986; Simon et  al., 1994), whereas at end-stage renal disease it is consistently 2–3:1, male:female (United States Renal Data System, 1999; Maisonneuve et al., 2000). The role of gender in the progression of MGN has been evaluated in a large meta-analysis, which included 21 studies containing 1894 patients followed for an average of 84  months (Neugarten et al., 2000). The pooled analysis confirmed that male gender is highly significantly associated with a more rapid rate of progression in MGN (standardized effect size estimate 0.30, 95% confidence interval (CI) 0.16–0.36, P < 0.00001). Our own data suggest that females demonstrate a lower rate of renal function decline and lower risk of kidney failure compared to males (0.63, 95% CI 0.40–1.00, P = 0.05), and that this may be related, in part, to lower degrees of proteinuria and blood pressure both at presentation and follow-up (Cattran et al., 2008).

Age Older age is generally associated with a higher risk of development of chronic renal insufficiency. We performed a retrospective study of 74 patients > 60 years of age at time of diagnosis of idiopathic MGN and compared them to the younger subjects in the same registry, who presented over a 19-year period (Zent et al., 1997). Older subjects had significantly higher median creatinine levels (1.3 mg/dL vs 1.0 mg/dL, P < 0.001), and lower calculated creatinine

545

546

Section 3  

the patient with glomerular disease

clearance (55 mL/min vs 95 mL/min, P < 0.0001) than those in the younger-onset group. The incidence of chronic renal insufficiency defined as a clearance of < 50 mL/min was significantly higher in the elderly after a mean observation time of nearly 4 years (59% vs 25%, P < 0.0001), although the incidence of ESRD was not different. The rate of deterioration in renal function, however, was not different between younger and older patients. The observed difference in rates of chronic renal failure may therefore be due in large part to loss of reserve of functional nephron mass with age, such that there is less remaining to compensate for the additive detrimental effects of a glomerular disease. A  more recent large retrospective study including 44 patients presenting over 60  years of age confirmed that older patients more often have worse renal impairment than a younger cohort, and also found a worse prognosis for renal survival (O’Callaghan et al., 2002). Smaller individual studies have found a somewhat more variable association with age and renal prognosis (Row et  al., 1975; Davison et  al., 1984; Tu et  al., 1984; Zucchelli et al., 1987; Schieppati et al., 1993; Honkanen et al., 1994).

Proteinuria Nephrotic syndrome and nephrotic-range proteinuria were shown early on to be associated with a worse prognosis, when analysed in univariate analysis (Row et al., 1975; Noel et al., 1979; Mallick et  al., 1983; Davison et  al., 1984; Tu et  al., 1984; Gerstoft et  al., 1986; Donadio et al., 1988; Murphy et al., 1988; Wehrmann et al., 1989). Multivariable analysis, however, does not consistently reveal that the degree of proteinuria at the time of diagnosis predicts outcome (Tu et  al., 1984; Schieppati et  al., 1993; Marx and Marx, 1999), and this may suggest that if reduction of proteinuria over time is achieved prognosis will be favourable (Ponticelli et al., 1992a; Laluck and Cattran, 1999). In one of the largest series to date, it is not surprising that complete remission of proteinuria from the nephrotic range is associated with an excellent prognosis (Troyanov et al., 2004). However patients achieving a partial remission of proteinuria, defined as reduction of proteinuria of 50% from peak value to sub-nephrotic levels (< 3.5 g/day) also benefit from a marked reduction in risk of kidney failure (hazard ratio 0.08, 95% CI 0.03–0.19, P < 0.001) compared to those with no remission (Troyanov et al., 2004). Given the variable course of the disease, a dynamic view of proteinuria over time is likely more indicative of prognosis (see below). Hypertension and renal insufficiency at the time of biopsy have shown similar results to initial proteinuria with respect to prediction of adverse outcome (Troyanov et al., 2004, as well as the rate of renal function decline 2006). There is general agreement that impaired function at the time of biopsy and hypertension do not portend a favourable prognosis; however, studies indicate conflicting results when incorporating these factors in multivariable models. The prognosis of patients who have non-nephrotic range proteinuria has recently been described in a review of 395 patients with idiopathic biopsy-proven MGN. One hundred and eight (27% of the total) patients presented with sub-nephrotic proteinuria and almost 40% (42 of 108) remained sub-nephrotic throughout the average followed up of 68  months. Their long-term rate of renal function declined as measured by slope of creatinine clearance (slope) was −0.93 mL/min/year. In contrast, those who subsequently developed nephrotic range proteinuria had a progression rate almost four times faster (−3.52 mL/min/year). The majority who developed nephrotic syndrome did so within the first year of follow-up. The

only distinguishing baseline feature between the two groups was a higher level of urine protein in the group that subsequently developed nephrotic syndrome (1.98 (0.3–3.4) versus 2.43 (0.5–3.4) g/ day). Baseline creatinine clearance as well as baseline and follow-up mean arterial pressure were not significantly different nor were the mean number of blood pressure medications between those that remained some nephrotic versus those that converted to nephrotic range proteinuria (Hladunewich et al., 2009).

Pathological findings The relationship between pathologic findings and both cross-sectional and long-term clinical parameters has been an area of conflicting results. This likely relates in part to relatively small study sample sizes, a mixture of idiopathic and secondary disease in studies, as well as a lack of consensus regarding standardized pathologic staging for the disease. With respect to the stage of intramembranous deposits, more ‘advanced’ stages of pathology have been found to correlate with an adverse prognosis in some studies according to univariate analysis. For instance, several analyses suggest an association between stage III and IV deposits and an adverse long term prognosis, compared with stages I and II deposits (Gluck et al., 1973; Franklin et al., 1973; Noel et al., 1979; Zucchelli et al., 1986, 1987; Tornroth et al., 1987; Hay et al., 1992), but not in all studies (Abe et al., 1986). Furthermore, biopsies with heterogenous deposits at various phases of evolution have been associated with a worse outcome (Yoshimoto et al., 2004). However, when a large sample of 389 cases was reviewed, neither stage nor synchronicity of EM deposits was found to be associated with clinical variables evaluated at the time of biopsy or with progressive renal insufficiency (Troyanov et al., 2006). The extent of C3 deposition (measured on a semi-quantitative scale) did correlate with the degree of proteinuria at the time of biopsy, as well as the rate of renal function decline. However, there was no difference in renal survival or likelihood of remission according to the extent of deposition (Troyanov et  al., 2006). Mesangial immune complexes have been associated with a favourable prognosis in idiopathic MGN, although this finding may more commonly suggest a secondary aetiology (Davenport et al., 1994). Focal and segmental glomerulosclerosis lesions are more consistently associated with a trend towards worse renal survival (Wakai and Magil, 1992; Toth and Takebayashi, 1994; Dumoulin et al., 2003, Troyanov et al., 2006). The findings of interstitial fibrosis and tubular atrophy are more consistently associated with a poor prognosis (Noel et  al., 1979; Ramzy et  al., 1981)  even in multivariable analysis (Wehrmann et al., 1989; Ponticelli et al., 1989; Marx and Marx, 1999). While lesions of tubulointerstitial fibrosis correlate with a worse renal survival, these lesions do not affect the rate of renal function decline; this may reflect the fact that patients with interstitial fibrosis at the time of biopsy tend to have more impaired clearance at the time of kidney biopsy (Zent et al., 1997; Troyanov et al., 2006).

Other markers Several other factors have been correlated with outcome, but are not yet part of the usual diagnostic assessment. Urinary excretion of immunoglobulin (Ig)-G and alpha-1-microglobulin has been correlated with outcome more closely than total proteinuria (Bazzi, 2001). Further, urinary excretion of beta-2-microglobulin, potentially reflecting proximal tubular cell injury, is independently predictive of progression and adds to data obtained from routine

Chapter 62 

clinical evaluation (Branten et  al., 2005; Hofstra et  al., 2008). In addition, there has been significant investigation into the role of genetic factors that are associated with the course of MGN.

Predicting prognosis Given the potential toxicities of the medications involved in treating MGN, as well as the significant potential for a patient to undergo a spontaneous remission, it is desirable to predict the clinical course of an individual patient before making therapeutic decisions. Ideally, the data required to offer prognostic information should be obtainable as soon as possible period after diagnosis. The difficulty associating the above factors with long-term outcome is their poor specificity, qualitative nature, and the fact that they reflect only cross sectional data at diagnosis, which has varied in each study. Another approach is to use an observation period to gather further information on progression (and to allow time for spontaneous remission). It is highly likely that in the near future autoantibody titres (e.g. to PLA2R, see Chapter 64) will be able to usefully supplement this information.

Box 62.1  Prediction model for risk of progression in idiopathic MGN (Pei et al., 1992; Cattran, 2001) Logistic regression model: X = 1.26 + (0.3 × PP) – (0.3 × slopeCcr) – (0.05 × Ccri) PP:  the level of persistent proteinuria in g/24 hours. This is measured as the lowest level observed over a period of 6 months. SlopeCcr: the slope of the creatinine clearance over the period used to observe persistent proteinuria (e.g. 6 months). Measured in mL/min/month. Ccri: the initial creatinine clearance documented at the beginning of the observation period, in mL/min. Then, use the calculated X to obtain a probability of progression (R) by substituting as follows: R = ex/(1+ eX) Sample calculation of risk of progression Patient A: Would like to know patient A’s risk of progression. One requires 6 months of follow-up for the calculation. PP = 5 g/24 hours Month

Proteinuria (g/24 hours)

Creatinine clearance (mL/min)

0

10

95

3

5

75

6

7

75

SlopeCcr  =  Ccr final  – Ccr initial  =  75 mL/min  – 95 mL/ min = −3.33 Time 6 months X = 1.26 + (0.3 × PP) – (0.3 × slope Ccr) – (0.05 × Ccri) = 1.26 + (0.3 × 5) – (0.3 × −3.33) – (0.05 × 95) = −0.991 eX = 0.37 R = 0.37 / (1 + 0.37) = 0.27 or a 27% chance of progressing.

membranous: treatment and outcome

Box 62.1 presents one prediction model. A less mathematical variant is presented in Box 62.2 and Fig. 62.1. The prediction model in Box 62.1 incorporates the clinical parameters of proteinuria and creatinine clearance estimates over fixed periods of time. The approach demonstrates an impressive overall accuracy of predicting progression to chronic renal failure. When the sensitivity, specificity, and positive and negative predictive values are considered, the accuracy of this formula, meaning its ability to predict whether or not a patient will progress, can be determined. This accuracy changes according to proteinuria values over a 6-month observation period; when values were persistently > 4 g/day, for instance, the algorithm can accurately predict whether a patient will or will not progress with a rate of 71%. At ≥ 6 g/day accuracy is 79% and at ≥ 8 g/day accuracy is 84% (Cattran et al., 1997). If the patient’s renal function was impaired or deteriorated over the 6 months, sensitivity and specificity were even higher. In addition, the regression formula has been validated in two populations—101 patients from Italy, and 78 patients from Finland, and found to have consistent accuracy in identifying the risk of progression compared to the original Canadian (Cattran et al., 1997). There are several advantages to this algorithm. Firstly, all of the factors are easily obtainable standard laboratory measurements. In addition, the dynamic nature of the algorithm allows it to be recalculated and reapplied over the course of a patient’s disease. It is important to note that the individual risk factors of age, gender, biopsy findings, and presence of hypertension were not found to be independent of the factors in the model, and although relevant to each individual case, they do not add to the predictive value of the algorithm.

Treatment of primary membranous glomerulonephritis The general management of nephrotic syndrome is described in Chapter 52. Here we examine specific issues and specific therapies for MGN. Specific measures for secondary MGN are mentioned in Chapter 63.

Non-immunologic therapy Dietary manipulation alone has not been shown to induce a complete remission of the nephrotic syndrome in idiopathic MGN. However, some data suggests that protein restriction results in a reduction in proteinuria and possibly progression of disease, demonstrated specifically in patients with MGN, and as well in patients with other causes of the nephrotic syndrome (Cupisti et al., 1990; D’Amico 1992; Pedrini et  al., 1996; Kopple et  al., 1997). Dietary management is discussed in Chapters 47 and 101. The additional benefit of ACE inhibition has been demonstrated specifically in idiopathic MGN (Thomas et  al., 1991; Rostoker et al., 1995). It is assumed that there is benefit from HMG-CoA reductase inhibitors in all patients with long-term heavy proteinuria, including MGN. Thrombotic complications pose a significant risk in patients with nephrotic syndrome, and it may be greater when the syndrome is due to MGN (See Chapter 52) (Trew et al., 1978; Wagoner et al., 1983; Llach, 1985; Bellomo and Atkins, 1993; Bellomo et al., 1993; Rabelink et al., 1994). Even when adjusted for the degree of proteinuria, the disease-specific risk of clinically evident thrombotic events is highest in MGN compared to other forms of idiopathic

547

548

Section 3  

the patient with glomerular disease “Low Risk” Asymptomatic Proteinuria < 4g/day Normal renal function

“Medium Risk” Proteinuria 4–8g/day Normal renal function

“High Risk” Proteinuria ≥8g/day +/– deteriorating renal function

Spontaneous remission Maintain BP ≤ 130/80 using ACEI, monitor proteinuria and renal function

ACEI maintain BP≤ 130/80, observe for 6 months

ACEI, maintain BP ≤130/80, observe for 3–6 months

Persistent nephrotic-range Proteinuria

Persistent heavy proteinuria +/– decreasing renal function

Consider longer observation versus specific therapy (see Box 62.2 and text)

Consider specific therapy (see Box 62.2 and text)

Fig. 62.1  Treatment algorithm. Patients may change from one category to another during the course of follow-up. BP = blood pressure; ACEI = angiotensin-converting-enzyme inhibiting drug.

glomerulonephritis (Barbour et al., 2011). Some have observed that the risk is heavily weighted towards the first few months around the time of diagnosis (Kumar et al., 2012). Given the risks associated with anticoagulation in this population, prophylactic anticoagulation even in the high-risk group has not been adopted by all clinicians (see Chapter 52).

the presence of life-threatening complications of the nephrotic syndrome (i.e. refractory oedema) and declining renal function not explained by other complications (e.g. renal vein thrombosis or acute tubular necrosis) may prompt consideration for earlier initiation of immunotherapy. A summary of the discussed treatment algorithm is presented in Box 62.2 and Fig. 62.1.

Specific immunotherapy

Possible regimens

The majority of trials of immunosuppressive agents in idiopathic MGN do not meet rigorous standards; they are often non-controlled, have small patient numbers, and short observation periods (Geddes and Cattran, 2000). The regimens examined in controlled trials include corticosteroids alone or in combination with either chlorambucil, cyclophosphamide, or ciclosporin, and the latter three drugs used as single agents. These medications all have significant adverse effects, and therefore the decision to subject a patient to these risks, must be weighed against the potential benefits. They are reviewed with our recommendations. Evidence for other potential therapies is also discussed.

Corticosteroid monotherapy: ineffective

Recommendations for specific treatment Given the variable clinical outcomes of patients, it is useful to first establish categories of progression, as determined by factors in the previously discussed algorithm. We recommend an initial observation period of 6 months, during which time non-immunologic interventions should be maximized. In order to determine the categories of progression, specific therapeutic trials were retrospectively divided according to their subjects’ initial laboratory characteristics. This allows the separation of the effectiveness of any one therapy by the category of risk of progression of the patients in that trial. After making a therapeutic decision regarding a patient, one can then compare a patient with the most similar risk profile, and treatment strategy can be determined for each individual patient, and a better estimate of risk:benefit ratio can be determined. While in general a 6-month wait period is advocated to fully evaluate risk,

Corticosteroids alone have been shown to be ineffective in inducing remission. Some studies have shown some improvement in proteinuria, but corticosteroids have not been found to prevent progression when used as monotherapy in controlled studies (Collaborative Study of the Adult Idiopathic Nephrotic Syndrome, 1979; Kobayashi et  al., 1982; Cattran et  al., 1989). Entry criteria in these trials would have placed the participants into the ‘medium-risk’ category. Although the total follow-up periods were < 4years, and the protocol for administration differed, it is generally held that corticosteroids alone do not have a role in treatment of idiopathic MGN (Lewis, 1993; Muirhead, 1999). A possible exception is patients of Asian ancestry (specifically Japanese) who in several retrospective studies have shown better response to steroid monotherapy. This race-specific difference in response to steroid monotherapy has not been validated in randomized controlled prospective studies and there are insufficient data to warrant use as a single agent in Asian populations (Shiiki et al., 2004).

Corticosteroids (high risk): ineffective This subgroup of patients is relatively small, and very few trials have exclusively studied subjects in this risk category. Corticosteroid treatment alone was examined in the subgroup of 55 patients with renal insufficiency from the relatively large randomized trial of Canadian subjects (Cattran et al., 1989). All in this subgroup had an initial creatinine clearance of < 72 mL/min. Those treated with prednisone (dose of 45 mg/m2 on alternate days for 6 months) did

Chapter 62 

Box 62.2  Stratification of risk of progression in MGN

Low risk of progression Asymptomatic proteinuria, peak < 4g/day, with normal serum creatinine at presentation and stable creatinine clearance over 6 months of observation. Blood pressure management and antiproteinuric strategies with agents such as ACEIs are very important in this group. However, given the generally favourable outcome, immunosuppressive therapy is not recommended. The prognosis of these patients is generally good. Our study of three cohorts of patients from Canada (N = 184), Finland (N = 78), and Italy (N = 101) showed that 17–28% of patients present in this category. Of these, only 6%, 0%, and 24% developed sustained renal insufficiency (clearance < 60 mL/min/1.73 m2) after a mean follow-up of 70, 104, and 59 months respectively. The average overall risk was only 5%. However, the numbers of patients in this group included in trials are relatively small and observation time limited. A small percentage will progress, so monitoring of renal function, proteinuria, and blood pressure is necessary to assess if their category has changed.

Medium risk of progression Proteinuria consistently 4–8 g/day over 6 months, but normal creatinine and creatinine clearance at presentation and during observation. In addition to general therapy, only the combination of alkylating agents with high-dose corticosteroids for a period of 6 months is of unequivocal benefit. However, toxicity is significant, and the natural history of this group is varied. Longer monitoring and further consideration may be justified. Steroids alone, and probably MMF, have no lasting effect on the disease, though steroids may have a transient effect on proteinuria. Calcineurin inhibitors have not been proven to alter long-term outcome, but do reduce proteinuria, which is useful in managing intractable nephrotic syndrome and may translate into long-term improvement in renal survival.

High risk of progression Persistent proteinuria ≥ 8g/day over the 6 months of observation, and/or deteriorating renal function. In addition to general therapy, the best evidence is for alkylating agents delivered with high-dose corticosteroids for a period of 6  months. IV cyclophosphamide is probably safer and less toxic than chlorambucil, but toxicity of these regimens is significant. The rate of reaching renal endpoints remains high in treated patients. The role of steroids in these regimens is unproven, but it is known that they are ineffective when administered alone. Ciclosporin was no better than supportive management in a randomized study in this group with high-grade proteinuria in association with deteriorating renal function (Howman et  al., 2013)  although this was in contrast to earlier smaller studies (Cattran at al., 1995). We regard the role of anti-B-cell antibodies such as rituximab as still unproven but deserving of further investigation.

membranous: treatment and outcome

not demonstrate an improvement in rate of deterioration of renal function. In the one randomized study of corticosteroids alone in high risk subjects (mean proteinuria was 10.6 g/day) (Cameron et al., 1990), prednisolone at 100–150 mg on alternate days given for 8 weeks prior to taper did not confer benefit with respect to rate of deterioration of function or proteinuria. Somewhat in contrast, a small study of 15 patients and declining function suggested that 5  days of 1 g intravenous (IV) methylprednisolone followed by a tapering course of prednisolone was associated with an initial stabilization in renal function in nine of the subjects (Short et al., 1987). At last follow-up, however, two patients had died, and five had reached ESRD, suggesting that any positive effects of the treatment were transient.

Alkylating agents with corticosteroids: beneficial but toxic There is evidence of benefit, however, when corticosteroids are in combination with a cytotoxic agent used in this risk group. A significant increase in both remission of proteinuria and renal survival was demonstrated, with follow-up out to 10 years in a trial comparing a regimen of prednisone and chlorambucil to symptomatic treatment (Ponticelli et al., 1984, 1992b, 1995). The regimen consisted of 1 g of IV methylprednisolone daily for the first 3 days of months 1, 3, and 5, followed by 27 days of oral methylprednisolone 0.5 mg/kg/day for the remainder of the month. In alternating months (months 2, 4, and 6), chlorambucil 0.2 mg/kg/day was used instead of the corticosteroid. At 10 years, the probability of survival without dialysis was 92% in the treatment group, and 60% in the group receiving symptomatic therapy (P = 0.004). The probability of achieving a complete or partial remission was 83% in treated group, and only 38% in controls (P  25–50%. All patients were felt to be at high risk of progression based on urinary IgG and urine beta-2-microglobulin levels that were previously correlated with a high risk of progressive renal insufficiency. They found a more rapid remission in proteinuria in early-start patients, but no differences between the two groups in overall remission rates, serum creatinine levels, average proteinuria, relapse rates, or adverse events after 6 years (Hofstra et al., 2010). Five studies have examined high-risk patients treated with alkylating agents and corticosteroids. A substantial improvement in renal function in more than half of patients, and a decline in proteinuria was noted in one study of eight patients (Mathieson et al., 1988). Similarly, half of the 21 subjects in a subsequent study were noted to have a stabilization or improvement in renal function (Warwick et al., 1994), but subjects with an initial creatinine between 180 and 480  μmol/L continued to show deterioration in function. When the outcome of these subjects was compared to historical controls, however, there did appear to be a trend to improved renal survival (Stirling et al., 1998). The success noted by these small trials must, however, be balanced by the high incidence of serious complications; in the aforementioned study by Stirling et al., for instance, half of patients had significant side effects related to therapy (infectious and myelosuppressive), necessitating discontinuation of medications in 6 of 19 patients. This particular study population, particularly those with significantly impaired renal function, may be the group most vulnerable to drug toxicity. Most recently, one study of 39 subjects compared conservative therapy in patients treated between 1975 and 1989, to a group treated between 1990 and 2000 with a regimen of oral chlorambucil (0.15 mg/kg/ day for 14 weeks) with oral prednisone for 6 months (Torres et al., 2002). Those receiving the chlorambucil had a 90% probability of renal survival at 4  years of follow-up, compared with only 55% probability in subjects receiving only conservative therapy (P < 0.001). This benefit persisted at 7 years. A UK multicentre study randomized 108 patients with declining renal function (≥ 20% decline within 2 years of trial entry and average creatinine clearance 50 mL/min) to alternating months of chlorambucil/ high-dose steroids for 6 months, ciclosporin for 12 months, or supportive management. Risk of a 20% decline in GFR over 3 years was significantly reduced in the alkylating agent group but still high (58 vs 84%). The outcome for the ciclosporin group was not similar to that of supportive care alone. Adverse events were common in all groups but significantly more common in the chlorambucil/ steroids group. Less than 50% of subjects were evaluable at 12 months (Howman et al., 2013). It is possible that IV cyclophosphamide rather than oral chlorambucil would have caused fewer side effects. It is also worth questioning the need for corticosteroids in these regimens. Monthly pulse cyclophosphamide for 6  months plus prednisone compared to prednisone alone produced no additional benefit in a randomized trial including 36 who would be considered ‘high risk’ by virtue of renal insufficiency (Falk et al., 1992). Two non-randomized case–control studies in similar populations involving long-term oral cyclophosphamide with or without prednisone did indicate a benefit to the therapy (Bruns et al., 1991; Jindal et  al., 1992). However, aside from the limitations of the non-randomized design, long-term cyclophosphamide exposure carries the significant risks of infertility, infection, and

malignancy, limiting the applicability of these trials. Direct comparison of cyclophosphamide and chlorambucil was undertaken in two trials that included patients with progressive deterioration in renal function. The first compared a traditional Italian regimen with chlorambucil to a modified routine with IV cyclophosphamide pulses at months 2, 4, 6, and three daily 1 g IV methylprednisolone pulses at months 1, 3, and 5 (Reichert et al., 1994). The authors concluded that cyclophosphamide administered in this manner was not beneficial after 6–36  months of follow-up. The same group then examined 27 patients receiving one of two treatment strategies. The first (N = 15) received a regimen of 3 days of IV methylprednisolone followed by 0.5 mg/kg/day of oral prednisone given in months 1, 3, and 5 alternating with oral chlorambucil 0.15 mg/kg/day in months 2, 4, and 6. The comparison group (N = 17) received cyclophosphamide 1.5–2.0 mg/kg/day for 1 year, and steroids in comparable doses to the first group (Branten et al., 1998). Those treated with cyclophosphamide showed a greater benefit with a greater fall in serum creatinine (61 vs 121 μmol/L fall, P < 0.01), lower incidence of ESRD (4 vs 1 patients, P < 0.05), and more frequent remission of proteinuria, and fewer short-term side effects, in a total of 27 subjects.

Calcineurin inhibitors: weak evidence of benefit Given the high doses of steroid required in these studies, as well as the potential for both gonadal toxicity and malignancy with use of cyclophosphamide, efforts have focused on identification of alternate approaches to treatment of MGN that may spare some of the treatment-associated toxicity. In retrospect, many of these studies have reached misleading conclusions because of the propensity of calcineurin inhibitors to reduce proteinuria in a dose-related manner without necessarily impacting on the progression of the underlying disease (see Chapters 45, 58). One study of subjects in the medium risk category examined the effectiveness of ciclosporin in combination with low-dose prednisone (Cattran et  al., 2001). Fifty-one subjects were enrolled in this multicentre, placebo-controlled, single-blind randomized trial. All had failed to achieve remission after at least 8 weeks of therapy with prednisone 1 mg/kg/day. Study subjects receiving active treatment (N = 28) were given ciclosporin in a liquid formulation starting at 3.5 mg/kg/day divided in two doses, and adjusted to achieve whole blood 12-hour trough levels of 125–225 micrograms/L (monoclonal assay). Control subjects (N = 23) received a placebo liquid, and all subjects were given prednisone at a dose of 0.15 mg/ kg/day to a maximum of 15 mg/day. Subjects received 26 weeks of therapy, after which the ciclosporin/placebo was stopped, and steroid dose was tapered. By 26 weeks, 75% of treated subjects had reached a partial or complete remission, compared with only 22% of controls (P = 0.001). The relapse rate was significant in both the treatment and control groups. The fraction of patients remaining in remission, however, remained significantly different at the 1year mark —39% of ciclosporin-treated subjects remained in remission, versus 13% in the placebo group (P = 0.007). This improvement in remission rate was not at the expense of a change in renal function, since there was no significant change noted in creatinine clearance in either group. A significant number of subjects did not respond to therapy. Further investigations are necessary to determine if a longer course of treatment, higher dose, or re-treatment of relapses may increase the rate and perhaps the duration of response. The high relapse rate may suggest that ciclosporin controls rather than

Chapter 62 

cures this disease especially in those patients that only achieve partial remission status. Only one RCT in MGN patients has compared tacrolimus (N = 39, given for 6–9 months) to oral cyclophosphamide (N = 34 given for 4 months) (Chen et al., 2010). Both groups in this study of patients of Asian descent also received prednisone tapered off over 8 months. The results indicated no difference between treatment groups in terms of partial or complete remission of proteinuria (79% vs 69%), or adverse events at 12 months of follow-up. Relapses occurred in approximately 15% of each group. These data would suggest that the use of tacrolimus is an effective alternative to an oral alkylating-agent regimen with similar short-term outcomes. However, the long-term efficacy of a tacrolimus-based regimen for MGN remains to be determined. One RCT explored the use of tacrolimus monotherapy in MGN, in patients with normal kidney function and mean proteinuria of approximately 8 g per 24 hours (moderate to high risk, N = 25). The patients received tacrolimus 0.05 mg/kg/day for 12 months followed by a 6-month taper, and were compared to conservatively treated controls (N = 23) (Praga et al., 2007). After 18 months, the probability of remission was 94% in the tacrolimus group but only 35%, in the control group. Six patients in the control group and only one in the tacrolimus group reached the secondary endpoint of a 50% increase in serum creatinine. Almost half of the patients relapsed after tacrolimus was withdrawn, similar to patients treated with ciclosporin. Tacrolimus monotherapy is therefore appealing for patients who are intolerant of corticosteroids; however, as in the case of ciclosporin, concerns regarding relapse rate after a 12-month course also merit further investigation into appropriate duration of therapy. In our RCT of high-risk patients with renal insufficiency (mean creatinine was 195 μmol/L), 1 year of ciclosporin was administered at a dose of 3.8 mg/kg, and compared with placebo (Cattran et al., 1995). Ciclosporin-treated patients demonstrated significantly reduced proteinuria, and a slowed rate of progression of renal failure (P = 0.02, and P < 0.02 respectively). These positive results were sustained in more than half of the patients as late as 2 years after treatment. The number of patients in the study, however, was small, and there was a trend towards transient increases in creatinine noted in the treatment group. A similar benefit was noted in an uncontrolled study of 15 individuals with steroid-resistant progressive disease, however the relapse rate was high (Rostoker et al., 1993). A retrospective review from a large collaborative group included 41 patients considered high risk due to the severity of proteinuria (> 10 g/day), and resistance to other immunosuppressive drugs (Fritsche et  al., 1999). Thirty-four per cent achieved a complete remission after a mean treatment time of 225 days, at a mean dose of 3.3 mg/ kg, thus confirming the drug’s efficacy, but also the need for prolonged therapy before assuming resistance to the medication.

Mycophenolate mofetil: no evidence of benefit One uncontrolled study including 17 patients with MGN amongst a group of 46 subjects with primary glomerulonephritis, included a treatment protocol with a minimum of 3 months with mycophenolate mofetil (MMF) (Choi 2002). Patients with MGN were somewhat heterogeneous with regard to risk profile, and received variable doses of prednisone in addition to the MMF. There was a significant decrease in proteinuria, and trend towards improved renal function.

membranous: treatment and outcome

The findings of this preliminary study were supported by a small pilot RCT in 21 drug-naïve adults at medium risk of progression. Patients received either MMF 2 g/day with prednisone 0.5 mg/kg/ day for 2 of the 6 months of immunotherapy, or alternating cycles of steroids and cyclophosphamide. There was no significant difference in the proportion of patients achieving remission: 64% with MMF, 80% with Ponticelli routine (Senthil Nayagam et al., 2008). The frequency of relapses and incidence of infections were similar in both groups. Similar results were observed in a small study comparing the same two regimens in patients of Chinese ancestry (Chan et al., 2007). In the latter the relapse rate was 23% at 2 years. Uncontrolled studies of MMF in patients with MGN at high risk of progression have demonstrated conflicting results. MMF was evaluated in 16 patients with nephrotic syndrome due to MGN who would be categorized as either medium to high risk by virtue of their renal parameters alone (Miller et al., 2000). Nearly all had steroid-resistant disease, and half had failed cytotoxic and ciclosporin therapy. Moderate success was noted after a mean of 8 months of treatment, with six patients achieving a halving of their proteinuria. No difference was noted with respect to renal function; side effects were infrequent. In contrast, more recent studies using MMF as initial therapy in MGN have not consistently demonstrated efficacy in inducing remissions or delaying the onset of progressive CKD (Branten et al., 2007). Thirty-two patients with MGN and impairment of kidney function (creatinine > 132  μmol/L) were treated with oral MMF 1 g twice daily for 12 months, in combination with corticosteroids, and compared to 32 patients—historical controls—treated for the same duration with oral cyclophosphamide in combination with corticosteroids (cyclophosphamide; 1.5 mg/kg/day). Cumulative incidences of complete and partial remission of proteinuria at 12 months were 66% with MMF versus 72% with cyclophosphamide (P = 0.3). Adverse effects occurred at a similar rate in the two groups, but relapses were much more common with MMF beginning even before completion of the 12 months of treatment and approaching an 80% relapse rate within 10 months of stopping therapy. Data from another randomized controlled study in this population do not support an advantage of MMF over standard care; 36 patients with MGN and nephrotic syndrome received either conservative therapy (renin–angiotensin system blockade, statins, low-salt and low-protein diet, and diuretics) plus MMF (2 g/day, without concomitant steroids) (N  =  19) or conservative therapy alone (N = 17) for 12 months (Dussol et al., 2008). The probability of a complete or partial remission did not differ between the two groups after 12 months. In summary, in patients at high risk of progressive renal disease, the studies evaluating a role for MMF are of limited number, include small sample sizes, and demonstrate inconsistent results with limited data beyond short-term endpoints. A regimen of MMF plus steroids may have comparable efficacy to the standard regimen of cyclical alkylating agents and steroids but is associated with more relapses, substantially reducing enthusiasm to adopt this approach to therapy of MGN. In this population, monotherapy with MMF appears to be ineffective.

Anti-B-cell antibodies: possibly effective, RCTs needed A promising agent described is rituximab, a monoclonal antibody directed against the surface antigen CD20 of B cells. In addition to the lack of possible direct nephrotoxicity, this medication has the potential advantage of ease of administration with a limited

551

552

Section 3  

the patient with glomerular disease

number of infusions and no need for frequent drug monitoring. While this drug likely has broad effects on immune system regulation and inflammation, its primary mechanism of action is causing apoptosis of CD20+ B cells. Interference with B-cell function should inhibit production of antibodies to the putative human auto-antigen implicated in the pathogenesis of MGN. The lack of relationship between levels of CD20+ cells and response to therapy supports the concept that the immunomodulatory effects of this drug extend beyond its ability to deplete this B-cell line. A role for rituximab was first established in small observational studies. Seven patients received the IV medication in an uncontrolled trial (Remuzzi et al., 2002). The mean creatinine clearance was 68.7 mL/min/1.73 m2, and mean proteinuria was 8.6 g/day (all had > 3.5g/day). All were on full doses of ACEIs, and had not reached remission after an observation of 1 year following biopsy. By 20 weeks following drug administration (the last follow-up), urine protein had decreased to a mean of 3.7 g/day, two patients achieved a full remission, and three a partial remission. Adverse effects were mild, and infusion-related. A  second observational study from the same investigators suggested that rituximab is likely to be most effective in patients with minimal degrees of tubulointerstitial injury (Ruggenenti et al., 2006). To reduce the cumulative dose of rituximab, the investigators subsequently performed a matched-cohort controlled study using circulating B-cell counts to guide dosing. At 1 year, the proportion of patients who achieved disease remission with lymphocyte-guided dosing was identical to that of 24 historical patients who were given a standard rituximab protocol (four weekly doses of 375 mg/m2). Lymphocyte-guided therapy resulted in less cumulative exposure to rituximab with substantial cost-saving benefits (Cravedi et al., 2007). Two pilot studies have been completed in North American patients. The first (Fervenza et al., 2008) was a prospective observational study in 15 patients with MGN and proteinuria > 4 g per 24 hours—despite ACEI/ARB use for > 3 months and systolic blood pressure < 130 mmHg— treated with 1 g rituximab at days 0 and 15. At 6 months, patients who remained with proteinuria >3 g per 24 hours, and with some recovery of their CD19/20+ B-cell count received a second identical course of rituximab. The mean baseline proteinuria of participants was 13.0 ± 5.7 g per 24 hours. The mean decline in proteinuria from baseline to 12 months was 6.2 ± 5.1 g/ day and was statistically significant (P = 0.002). Rituximab was well tolerated, and was effective in reducing proteinuria in most of the patients. The complete and partial remission rate was almost 60%, higher than would have been expected based on known spontaneous remission rates. The second study carried out by the same group was a prospective observational study in 20 patients with MGN and baseline persistent proteinuria > 5.0 g/day (Fervenza et  al., 2010). It was designed to test whether the standard four-dose regimen would be more efficacious than the 1 g, two-dose regimen given in the first study. All patients received rituximab (375 mg/m2 weekly for four doses), with retreatment at 6 months regardless of proteinuria response. Baseline proteinuria was 11.9 g/day and decreased to 4.2 g/day and 2.0 g/day at 12 and 24 months, respectively, while creatinine clearance increased from 72.4 to 88.4 mL/min per 1.73 m2 at 24 months. Among 18 patients who completed 24 months of follow-up, four achieved complete remission, 12 achieved partial remission (total complete plus partial remission of 80%). One

patient relapsed during follow-up. More than half of the patients in this pilot trial had not responded to prior therapy. No short-term toxicity of rituximab was observed. This study also reinforced the observation, originally made in patients receiving alkylating agents, that proteinuria declines gradually in patients with MGN and many months may be required for proteinuria to reach its nadir. Comparison with alkylating agents in long term randomized studies is needed.

Adrenocorticotrophic hormone: experimental Adrenocorticotrophic hormone (ACTH) was originally administered for the treatment of dyslipidaemia. The observation that patients receiving this drug for dyslipidaemia associated with the nephrotic syndrome also resulted in reduction in proteinuria first prompted consideration of potential anti-proteinuric drug activity particularly in patients with MGN. One observational study and one small randomized controlled trial provide preliminary support for the use of long-acting ACTH as initial therapy in MGN. Depot synthetic ACTH (Synacthen®) administered for 1 year in an observational study decreased proteinuria in patients with MGN (Berg and Nilsson-Ehle, 1994; Berg et al., 1999; Berg and Arnadottir, 2004). More recently, a small, open-label, pilot RCT compared IV methylprednisolone and oral corticosteroids plus a cytotoxic agent (N = 16) versus synthetic ACTH (N = 16) as initial therapy in MGN, and found them to be of similar efficacy, at least over a short-term follow-up (Ponticelli et al., 2006). Side effects associated with the use of synthetic ACTH included dizziness, glucose intolerance, diarrhoea, and the development of bronze-coloured skin, all of which resolved after the end of therapy. Larger RCTs are required before synthetic ACTH can be considered as recommended therapy for MGN. Preliminary reports of uncontrolled studies showing a similar effect of native, intact (porcine) ACTH in a gel formulation have been published but no RCTs have been conducted with this formulation of ACTH (it should be noted that in the United States intact native ACTH in gel formulation is approved by the Food and Drug Administration for treatment of the nephrotic syndrome including MGN in the absence of renal failure) (Bomback et al., 2011; Hladunewich et al., 2014).

Strategies to reduce the side effects of immunosuppressive therapy Bone loss due to corticosteroid treatment, is related to both dose and duration of therapy, and is greatest in the first 3–6 months of treatment. Prolonged low-dose exposure, however, has also been associated with significant loss of bone density (Canalis and Giustina, 2001). The effects of steroids on the intestinal absorption of calcium and promotion of calciuria suggest that calcium and vitamin D should be incorporated into the treatment regimen. There are, however, significant data indicating the protective effects of antiresorptive medications on bone loss and fractures induced by glucocorticoids. Several agents, including etidronate and alendronate, have been shown in multicentred, well-designed trials to improve these outcomes (Adachi et  al., 1997; Saag et al., 1998). Avascular necrosis of the femoral head is another potentially serious skeletal complication of prednisone. Patients must be informed of this potential side effect, as it is not preventable, and is not necessarily dose related.

Chapter 62 

Ongoing monitoring for excessive myelosuppression should be employed in order to reduce the potential for infections, but individuals receiving alkylating agents and high-dose corticosteroids are at substantial risk of infectious complications. The use of trimethoprim-sulfamethoxazole significantly reduces the incidence of Pneumocystis pneumonia in this population (Ognibene et al., 1995). Gonadal toxicity due to alkylating agents is of significant concern but is linked to cumulative dose. IV cyclophosphamide given for 3 months is unlikely to reach levels of high concern, but may be enough to bring on early menopause in some patients. As re-treatment may be necessary, cryopreservation of sperm or oocytes should be considered prior to initiation of therapy, where it is available. Acrolein is the toxic metabolite produced from cyclophosphamide which induces urothelial damage. Hydration and MESNA can be used if administering the intravenous formulation of cyclophosphamide. Patients should be warned of the long-term risk of bladder and also of other malignancies, including skin cancers and haematological malignancy (Radis et al., 1995).

References Abe, S., Amagasaki, Y., Konishi, K., et al. (1986). Idiopathic membranous glomerulonephritis: aspects of geographical differences. J Clin Pathol, 39, 1193–8. Adachi, J. D., Bensen, W. G., Brown, J., et al. (1997). Intermittent etidronate therapy to prevent corticosteroid-induced osteoporosis. N Engl J Med, 337, 382–7. Barbour, S. J., Greenwald, A., Djurdjev, O., et al. (2011). Disease-specific risk of venous thromboembolic events is increased in idiopathic glomerulonephritis. Kidney Int, 81, 190–5. Bazzi, C. (2001). Urinary excretion of IgG and alpha(1)-microglobulin predicts clinical course better than extent of proteinuria in membranous nephropathy. Am J Kidney Dis, 38, 240–8. Bellomo, R. and Atkins, R.C. (1993). Membranous nephropathy and thromboembolism: is prophylactic anticoagulation warranted? Nephron, 63, 249–54. Bellomo, R., Wood, C., Wagner, I., et al. (1993). Idiopathic membranous nephropathy in an Australian population: the incidence of thromboembolism and its impact on the natural history. Nephron, 63, 240–1. Berg, A. L. and Arnadottir, M. (2004). ACTH-induced improvement in the nephrotic syndrome in patients with a variety of diagnoses. Nephrol Dial.Transplant, 19, 1305–7. Berg, A. L. and Nilsson-Ehle, P. (1994). Direct effects of corticotropin on plasma lipoprotein metabolism in man—studies in vivo and in vitro. Metabolism, 43, 90–7. Berg, A. L., Nilsson-Ehle, P., and Arnadottir, M. (1999). Beneficial effects of ACTH on the serum lipoprotein profile and glomerular function in patients with membranous nephropathy. Kidney Int, 56, 1534–43. Bomback, A. S., Tumlin, J. A., Baranski, J., et al. (2011). Treatment of nephrotic syndrome with adrenocorticotropic hormone (ACTH) gel. Drug Des Devel Ther, 5, 147–53. Branten, A. J., Reichert, L. J., Koene, R. A., et al. (1998). Oral cyclophosphamide versus chlorambucil in the treatment of patients with membranous nephropathy and renal insufficiency. QJM, 91, 359–66. Branten, A. J., du Buf-Vereijken, P. W., Klasen, I. S., et al. (2005). Urinary excretion of beta2-microglobulin and IgG predict prognosis in idiopathic membranous nephropathy: a validation study. Am Soc Nephrol, 16, 169–74. Branten, A. J., du Buf-Vereijken, P. W., Vervloet, M., et al. (2007). Mycophenolate mofetil in idiopathic membranous nephropathy: a clinical trial with comparison to a historic control group treated with cyclophosphamide. Am J Kidney Dis., 50, 248–56.

membranous: treatment and outcome

Bruns, F. J., Adler, S., Fraley, D. S., et al. (1991). Sustained remission of membranous glomerulonephritis after cyclophosphamide and prednisone. Ann Intern Med, 114, 725–30. Cameron, J. S., Healy, M. J., and Adu, D. (1990). The Medical Research Council trial of short-term high-dose alternate day prednisolone in idiopathic membranous nephropathy with nephrotic syndrome in adults. The MRC Glomerulonephritis Working Party. QJM, 74, 133–56. Canalis, E. and Giustina, A. (2001). Glucocorticoid-induced osteoporosis: summary of a workshop. J Clin Endocrinol Metab, 86, 5681–5. Cattran, D. C. (2001). Idiopathic membranous glomerulonephritis. Kidney Int, 59, 1983–94. Cattran, D. C., Appel, G. B., Hebert, L. A., et al. (2001). Cyclosporine in patients with steroid-resistant membranous nephropathy: a randomized trial. Kidney Int, 59, 1484–90. Cattran, D. C., Delmore, T., Roscoe, J., et al. (1989). A randomized controlled trial of prednisone in patients with idiopathic membranous nephropathy. N Engl J Med, 320, 210–15. Cattran, D. C., Greenwood, C., Ritchie, S., et al. (1995). A controlled trial of cyclosporine in patients with progressive membranous nephropathy. Canadian Glomerulonephritis Study Group. Kidney Int, 47, 1130–5. Cattran, D. C., Pei, Y., and Greenwood, C. (1992). Predicting progression in membranous glomerulonephritis. Nephrol Dial Transplant, 7 Suppl 1, 48–52. Cattran, D. C., Pei, Y., Greenwood, C. M., et al. (1997). Validation of a predictive model of idiopathic membranous nephropathy: its clinical and research implications. Kidney Int, 51, 901–7. Cattran, D. C., Reich, H. N., Beanlands, H. J., et al. (2008). The impact of sex in primary glomerulonephritis. Nephrol Dial Transplant, 23, 2247–53. Chan, T. M., Lin, A. W., Tang, S. C., et al. (2007). Prospective controlled study on mycophenolate mofetil and prednisolone in the treatment of membranous nephropathy with nephrotic syndrome. Nephrology, 12, 576–81. Chen, M., Li, H., Li, X. Y., et al. (2010). Tacrolimus combined with corticosteroids in treatment of nephrotic idiopathic membranous nephropathy: a multicenter randomized controlled trial. Am J Med Sci, 339, 233–8. Choi, M. J., Eustace, J. A., Gimenez, L. F., et al. (2002). Mycophenolate mofetil treatment for primary glomerular diseases. Kidney Int, 61, 1098–114. Collaborative Study of the Adult Idiopathic Nephrotic Syndrome (1979). A controlled study of short-term prednisone treatment in adults with membranous nephropathy. N Engl J Med, 301, 1301–6. Cravedi, P., Ruggenenti, P., Sghirlanzoni, M. C., et al. (2007). Titrating rituximab to circulating B cells to optimize lymphocytolytic therapy in idiopathic membranous nephropathy. Clin J Am Soc Nephrol, 2, 932–7. Cupisti, A., Morelli, E., Ciardella, F., et al. (1990). Dietary proteins affect proteinuria in primary membranous glomerulonephritis with nephrotic syndrome and normal renal function. Contrib Nephrol, 83, 166–9. D’Amico, G. and Gentile, M. G. (1992). Effect of dietary manipulation on the lipid abnormalities and urinary protein loss in nephrotic patients. Miner Electrolyte Metab, 18, 203–6. D’Amico, G., Gentile, M. G., Manna, G., et al. (1992). Effect of vegetarian soy diet on hyperlipidaemia in nephrotic syndrome. Lancet, 339, 1131–4. Davenport, A., Maciver, A. G., Hall, C. L., et al. (1994). Do mesangial immune complex deposits affect the renal prognosis in membranous glomerulonephritis? Clin Nephrol, 41, 271–6. Davison, A. M., Cameron, J. S., Kerr, D. N., et al. (1984). The natural history of renal function in untreated idiopathic membranous glomerulonephritis in adults. Clin Nephrol, 22, 61–7. Donadio, J. V., Jr., Torres, V. E., Velosa, J. A., et al. (1988). Idiopathic membranous nephropathy: the natural history of untreated patients. Kidney Int, 33, 708–15. Dumoulin, A., Hill, G. S., Montseny, J. J., et al. (2003). Clinical and morphological prognostic factors in membranous nephropathy: significance of focal segmental glomerulosclerosis. Am J Kidney Dis, 41, 38–48.

553

554

Section 3  

the patient with glomerular disease

Dussol, B., Morange, S., Burtey, S., et al. (2008). Mycophenolate mofetil monotherapy in membranous nephropathy: a 1-year randomized controlled trial. Am J Kidney Dis, 52, 699–705. Falk, R. J., Hogan, S. L., Muller, K. E., et al. (1992). Treatment of progressive membranous glomerulopathy. A randomized trial comparing cyclophosphamide and corticosteroids with corticosteroids alone. The Glomerular Disease Collaborative Network. Ann Intern Med, 116, 438–45. Fervenza, F. C., Abraham, R. S., Erickson, S. B., et al. (2010). Rituximab therapy in idiopathic membranous nephropathy: a 2-year study. Clin J Am Soc Nephrol, 5, 2188–98. Fervenza, F. C., Cosio, F. G., Erickson, S. B., et al. (2008). Rituximab treatment of idiopathic membranous nephropathy. Kidney Int, 73, 117–25. Franklin, W. A., Jennings, R.B., Earle, D.P. (1973). Membranous glomerulonephritis: long-term serial observations on clinical course and morphology. Kidney Int, 4, 36–56. Fritsche, L., Budde, K., Färber, L., et al. (1999). Treatment of membranous glomerulopathy with cyclosporin A: how much patience is required? Nephrol Dial Transplant, 14, 1036–8. Geddes, C. C. and Cattran, D.C. (2000). The treatment of idiopathic membranous nephropathy. Semin Nephrol, 20, 299–308. Gerstoft, J., Balsløv, J. T., Brahm, M., et al. (1986). Prognosis in glomerulonephritis. II. Regression analyses of prognostic factors affecting the course of renal function and the mortality in 395 patients. Calculation of a prognostic model. Report from a Copenhagen study group of renal diseases. Acta Med Scand, 219, 179–87. Gluck, M. C., Gallo, G., Lowenstein, J., et al. (1973). Membranous glomerulonephritis. Evolution of clinical and pathologic features. Ann Intern Med, 78, 1–12. Harrison, D. J., Thomson, D., and MacDonald, M.K. (1986). Membranous glomerulonephritis. J Clin Pathol, 39, 167–7. Hay, N. M., Bailey, R. R., Lynn, K. L., et al. (1992). Membranous nephropathy: a 19 year prospective study in 51 patients. N Z Med J, 105, 489–91. Heaf, J., Lokkegaard, H., and Larsen, S. (1999). The epidemiology and prognosis of glomerulonephritis in Denmark 1985–1997. Nephrol Dial Transplant, 14, 1889–97. Hladunewich, M. A., Cattran, D., Beck, L. H., et al. (2014). A pilot study to determine the dose and effectiveness of adrenocorticotrophic hormone (H.P. Acthar® Gel) in nephrotic syndrome due to idiopathic membranous nephropathy. Nephrol Dial Transplant, 29(8), 1570–7. Hladunewich, M. A., Troyanov, S., Calafati, J., et al. (2009). The natural history of the non-nephrotic membranous nephropathy patient. Clin J Am Soc Nephrol, 4, 1417–22. Hofstra, J. M., Branten, A. J., Wirtz, J. J., et al. (2010). Early versus late start of immunosuppressive therapy in idiopathic membranous nephropathy: a randomized controlled trial. Nephrol Dial Transplant, 25, 129–36. Hofstra, J. M., Deegens, J. K., Willems, H. L., et al. (2008). Beta-2-microglobulin is superior to N-acetyl-beta-glucosaminidase in predicting prognosis in idiopathic membranous nephropathy. Nephrol Dial Transplant, 23, 2546–51. Hogan, S. L., Muller, K. E., Jennette, J. C., et al. (1995). A review of therapeutic studies of idiopathic membranous glomerulopathy. Am J Kidney Dis, 25, 862–75. Honkanen, E. (1986). Survival in idiopathic membranous glomerulonephritis. Clin Nephrol, 25, 122–8. Honkanen, E., Törnroth, T., Grönhagen-Riska, C., et al. (1994). Long-term survival in idiopathic membranous glomerulonephritis: can the course be clinically predicted? Clin Nephrol, 41, 127–34. Hopper, J., Jr., Trew, P.A., and Biava, C.G. (1981). Membranous nephropathy: its relative benignity in women. Nephron, 29, 18–24. Howman, A., Chapman, T. L., Langdon, M. M., et al. (2013). Immunosuppression for progressive membranous nephropathy: a UK randomised controlled trial Lancet, 381, 744–51. Jha, V., Ganguli, A., Saha, T. K., et al. (2007). A randomized, controlled trial of steroids and cyclophosphamide in adults with nephrotic syndrome caused by idiopathic membranous nephropathy. Am Soc Nephrol, 18, 1899–904.

Jindal, K., West, M., Bear, R., et al. (1992). Long-term benefits of therapy with cyclophosphamide and prednisone in patients with membranous glomerulonephritis and impaired renal function. Am J Kidney Dis, 19, 61–7. Kida, H., Asamoto, T., Yokoyama, H., et al. (1986). Long-term prognosis of membranous nephropathy. Clin Nephrol, 25, 64–9. Kobayashi, Y., Tateno, S., Shigematsu, H., et al. (1982). Prednisone treatment of non-nephrotic patients with idiopathic membranous nephropathy. A prospective study. Nephron, 30, 210–19. Kopple, J. D., Levey, A. S., Greene, T., et al. (1997). Effect of dietary protein restriction on nutritional status in the Modification of Diet in Renal Disease Study. Kidney Int, 52, 778–91. Kumar, S., Chapagain, A., Nitsch, D., et al. (2012). Proteinuria and hypoalbuminemia are risk factors for thromboembolic events in patients with idiopathic membranous nephropathy: an observational study. BMC Nephrol, 13, 107. Laluck, B. J. Jr. and Cattran, D.C. (1999). Prognosis after a complete remission in adult patients with idiopathic membranous nephropathy. Am J Kidney Dis, 33, 1026–32. Lewis, E. J. (1993). Idiopathic membranous nephropathy—to treat or not to treat? N Engl J Med, 329, 127–9. Llach, F. (1985). Hypercoagulability, renal vein thrombosis, and other thrombotic complications of nephrotic syndrome. Kidney Int, 28, 429–39. MacTier, R., Boulton Jones, J. M., Payton, C. D., et al. (1986). The natural history of membranous nephropathy in the West of Scotland. QJM, 60, 793–802. Maisonneuve, P., Agodoa, L., Gellert, R., et al. (2000). Distribution of primary renal diseases leading to end-stage renal failure in the United States, Europe, and Australia/New Zealand: results from an international comparative study. Am J Kidney Dis, 35, 157–65. Mallick, N. P., Short, C.D., Manos, J. (1983). Clinical membranous nephropathy. Nephron, 34, 209–19. Marx, B. E. and Marx, M. (1997). Prognosis of idiopathic membranous nephropathy: a methodologic meta- analysis. Kidney Int, 51, 873–9. Marx, B. E. and Marx, M. (1999). Prediction in idiopathic membranous nephropathy. Kidney Int, 56, 666–73. Mathieson, P. W., Turner, A. N., Maidment, C. G., et al. (1988). Prednisolone and chlorambucil treatment in idiopathic membranous nephropathy with deteriorating renal function. Lancet, 2, 869–72. Miller, G., Zimmerman, R., 3rd, Radhakrishnan, J., et al. (2000). Use of mycophenolate mofetil in resistant membranous nephropathy. Am J Kidney Dis, 36, 250–6. Muirhead, N. (1999). Management of idiopathic membranous nephropathy: evidence-based recommendations. Kidney Int Suppl, 70, S47–55. Murphy, B. F., Fairley, K.F., Kincaid-Smith, P.S. (1988). Idiopathic membranous glomerulonephritis: long-term follow-up in 139 cases. Clin Nephrol, 30, 175–81. Neugarten, J., Acharya, A., Silbiger, S.R. (2000). Effect of gender on the progression of nondiabetic renal disease: a meta-analysis. Am Soc Nephrol, 11, 319–29. Noel, L. H., Zanetti, M., Droz, D., et al. (1979). Long-term prognosis of idiopathic membranous glomerulonephritis. Study of 116 untreated patients. Am J Med, 66, 82–90. O’Callaghan, C. A., Hicks, J., Doll, H., et al. (2002). Characteristics and outcome of membranous nephropathy in older patients. Int Urol Nephrol, 33, 157–65. Ognibene, F. P., Shelhamer, J. H., Hoffman, G. S., et al. (1995). Pneumocystis carinii pneumonia: a major complication of immunosuppressive therapy in patients with Wegener’s granulomatosis. Am J Respir Crit Care Med, 151, 795–9. Pedrini, M. T., Levey, A. S., Lau, J., et al. (1996). The effect of dietary protein restriction on the progression of diabetic and nondiabetic renal diseases: a meta-analysis. Ann Intern Med, 124, 627–32. Pei, Y., Cattran, D., Greenwood, C. (1992). Predicting chronic renal insufficiency in idiopathic membranous glomerulonephritis. Kidney Int, 42, 960–6.

Chapter 62 

Polanco, N., Gutiérrez, E., Covarsí, A., et al. (2010). Spontaneous remission of nephrotic syndrome in idiopathic membranous nephropathy. Am Soc Nephrol, 21, 697–704. Ponticelli, C., Altieri, P., Scolari, F., et al. (1998). A randomized study comparing methylprednisolone plus chlorambucil versus methylprednisolone plus cyclophosphamide in idiopathic membranous nephropathy. Am Soc Nephrol, 9, 444–50. Ponticelli, C. and Moroni, G. (1998). Renal biopsy in lupus nephritis—what for, when and how often? Nephrol Dial Transplant, 13, 2452–4. Ponticelli, C. and Passerini, P. (1990). The natural history and therapy of idiopathic membranous nephropathy. Nephrol Dial Transplant, 5 Suppl 1, 37–41. Ponticelli, C., Passerini, P., Altieri, P., et al. (1992a). Remissions and relapses in idiopathic membranous nephropathy. Nephrol Dial Transplant, 7 (Suppl 1), 85–90. Ponticelli, C., Passerini, P., Salvadori, M., et al. (2006). A randomized pilot trial comparing methylprednisolone plus a cytotoxic agent versus synthetic adrenocorticotropic hormone in idiopathic membranous nephropathy. Am J Kidney Dis, 47, 233–40. Ponticelli, C., Zucchelli, P., Imbasciati, E., et al. (1984). Controlled trial of methylprednisolone and chlorambucil in idiopathic membranous nephropathy. N Engl J Med, 310, 946–50. Ponticelli, C., Zucchelli, P., Passerini, P., et al. (1989). A randomized trial of methylprednisolone and chlorambucil in idiopathic membranous nephropathy. N Engl J Med, 320, 8–13. Ponticelli, C., Zucchelli, P., Passerini, P., et al. (1992b). Methylprednisolone plus chlorambucil as compared with methylprednisolone alone for the treatment of idiopathic membranous nephropathy. The Italian Idiopathic Membranous Nephropathy Treatment Study Group. N Engl J Med, 327, 599–603. Ponticelli, C., Zucchelli, P., Passerini, P., et al. (1995). A 10-year follow-up of a randomized study with methylprednisolone and chlorambucil in membranous nephropathy. Kidney Int, 48, 1600–4. Praga, M., Barrio, V., Juárez, G. F., et al. (2007). Tacrolimus monotherapy in membranous nephropathy: a randomized controlled trial. Kidney Int, 71, 924–30. Rabelink, T. J., Zwaginga, J. J., Koomans, H. A., et al. (1994). Thrombosis and hemostasis in renal disease. Kidney Int, 46, 287–96. Radis, C.D., Kahl, L. E., Baker, G. L., et al. (1995). Effects of cyclophosphamide on the development of malignancy and on long-term survival of patients with rheumatoid arthritis. A 20-year followup study. Arthritis Rheum, 38, 1120–7. Ramzy, M. H., Cameron, J. S., Turner, D. R., et al. (1981). The long-term outcome of idiopathic membranous nephropathy. Clin Nephrol, 16, 13–19. Reichert, L. J., Huysmans, F. T., Assmann, K., et al. (1994). Preserving renal function in patients with membranous nephropathy: daily oral chlorambucil compared with intermittent monthly pulses of cyclophosphamide. Ann Intern Med, 121, 328–33. Remuzzi, G., Chiurchiu, C., Abbate, M., et al. (2002). Rituximab for idiopathic membranous nephropathy (letter). Lancet, 360, 923–4. Research Group on Progressive Chronic Renal Disease (1999). Nationwide and long-term survey of primary glomerulonephritis in Japan as observed in 1,850 biopsied cases. Nephron, 82, 205–13. Rostoker, G., Belghiti, D., Ben Maadi, A., et al. (1993). Long-term cyclosporin A therapy for severe idiopathic membranous nephropathy. Nephron, 63, 335–41. Rostoker, G., Ben Maadi, A., Remy, P., et al. (1995). Low-dose angiotensin-converting-enzyme inhibitor captopril to reduce proteinuria in adult idiopathic membranous nephropathy: a prospective study of long-term treatment. Nephrol Dial Transplant, 10, 25–29. Row, P. G., Cameron, J. S., Turner, D. R., et al. (1975). Membranous nephropathy. Long-term follow-up and association with neoplasia. QJM, 44, 207–39. Ruggenenti, P., Chiurchiu, C., Abbate, M., et al. (2006). Rituximab for idiopathic membranous nephropathy: who can benefit? Clin J Am Soc Nephrol, 1, 738–48.

membranous: treatment and outcome

Saag, K. G., Emkey, R., Schnitzer, T. J., et al. (1998). Alendronate for the prevention and treatment of glucocorticoid-induced osteoporosis. Glucocorticoid-Induced Osteoporosis Intervention Study Group. N Engl J Med, 339, 292–9. Schieppati, A., Mosconi, L., Perna, A., et al. (1993). Prognosis of untreated patients with idiopathic membranous nephropathy. N Engl J Med, 329, 85–9. Senthil Nayagam, L., Ganguli, A., Rathi, M., et al. (2008). Mycophenolate mofetil or standard therapy for membranous nephropathy and focal segmental glomerulosclerosis: a pilot study. Nephrol Dial Transplant, 23, 1926–30. Shiiki, H., Saito, T., Nishitani, Y., et al. (2004). Prognosis and risk factors for idiopathic membranous nephropathy with nephrotic syndrome in Japan. Kidney Int, 65, 1400–7. Short, C. D., Solomon, L. R., Gokal, R., et al. (1987). Methylprednisolone in patients with membranous nephropathy and declining renal function. QJM, 65, 929–40. Simon, P., Ramée, M. P., Autuly, V., et al. (1994). Epidemiology of primary glomerular diseases in a French region. Variations according to period and age. Kidney Int, 46, 1192–8. Stirling, C. M., Simpson, K., Boulton-Jones, J.M. (1998). Immunosuppression and outcome in idiopathic membranous nephropathy. QJM, 91, 159–64. Thomas, D. M., Hillis, A. N., Coles, G. A., et al. (1991). Enalapril can treat the proteinuria of membranous glomerulonephritis without detriment to systemic or renal hemodynamics. Am J Kidney Dis, 18, 38–43. Tornroth, T., Honkanen, E., Pettersson, E. (1987). The evolution of membranous glomerulonephritis reconsidered: new insights from a study on relapsing disease. Clin Nephrol, 28, 107–17. Torres, A., Domínguez-Gil, B., Carreño, A., et al. (2002). Conservative versus immunosuppressive treatment of patients with idiopathic membranous nephropathy. Kidney Int, 61, 219–27. Toth, T. and Takebayashi, S. (1994). Factors contributing to the outcome in 100 adult patients with idiopathic membranous glomerulonephritis. Int Urol Nephrol, 26, 93–106. Trew, P. A., Biava, C. G., Jacobs, R. P., et al. (1978). Renal vein thrombosis in membranous glomerulonephropathy: incidence and association. Medicine (Baltimore), 57, 69–82. Troyanov, S., Roasio, L., Pandes, M., et al. (2006). Renal pathology in idiopathic membranous nephropathy: a new perspective. Kidney Int, 69, 1641–8. Troyanov, S., Wall, C. A., Miller, J. A., et al. (2004). Idiopathic membranous nephropathy: definition and relevance of a partial remission. Kidney Int, 66(3), 1199–205. Tu, W. H., Petitti, D. B., Biava, C. G., et al. (1984). Membranous nephropathy: predictors of terminal renal failure. Nephron, 36, 118–24. United States Renal Data System (2009). USRDS 2009 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Wagoner, R. D., Stanson, A. W., Holley, K. E., et al. (1983). Renal vein thrombosis in idiopathic membranous glomerulopathy and nephrotic syndrome: incidence and significance. Kidney Int, 23, 368–74. Wakai, S. and Magil, A. B. (1992). Focal glomerulosclerosis in idiopathic membranous glomerulonephritis. Kidney Int, 41(2), 428–34. Warwick, G. L., Geddes, C.G., and Boulton-Jones, J.M. (1994). Prednisolone and chlorambucil therapy for idiopathic membranous nephropathy with progressive renal failure. QJM, 87, 223–9. Wehrmann, M., Bohle, A., Bogenschütz, O., et al. (1989). Long-term prognosis of chronic idiopathic membranous glomerulonephritis. An analysis of 334 cases with particular regard to tubulo-interstitial changes. Clin Nephrol, 31, 67–76. Yoshimoto, K., Yokoyama, H., Wada, T., et al. (2004). Pathologic findings of initial biopsies reflect the outcomes of membranous nephropathy. Kidney Int, 65, 148–53.

555

556

Section 3  

the patient with glomerular disease

Zent, R., Nagai, R., and Cattran, D. C. (1997). Idiopathic membranous nephropathy in the elderly: a comparative study. Am J Kidney Dis, 29, 200–6. Zucchelli, P., Cagnoli, L., Pasquali, S., et al. (1986). Clinical and morphologic evolution of idiopathic membranous nephropathy. Clin Nephrol, 25, 282–8.

Zucchelli, P., Ponticelli, C., Cagnoli, L., et al. (1987). Long-term outcome of idiopathic membranous nephropathy with nephrotic syndrome. Nephrol Dial Transplant, 2, 73–8. Zucchelli, P. and Pasquali, S. (1998). Membranous nephropathy. In A. M. Davison, J. S. Cameron, J. -P. Grunfeld, et al. (ed.) Oxford Textbook of Clinical Nephrology (2nd ed.), pp. 571–90. Oxford: Oxford University Press.

CHAPTER 63

Secondary membranous glomerulonephritis Daniel C. Cattran and Heather N. Reich The approach to distinguishing primary from secondary causes of membranous glomerulonephritis (MGN) is described in Chapter 61. The major causes are considered in more detail here. The general conservative measures described for the treatment of primary MGN (ACE inhibition, blood pressure control, lipid-lowering therapy, etc.) are, in our opinion, also recommended for the treatment of almost all causes of secondary forms of MGN, despite a lack of clear evidence from controlled trials.

day steroids alone (Austin et al., 2009). At 1 year, the cumulative probability of remission was 27% with prednisone, 60% with cyclophosphamide and 83% with ciclosporin; not surprisingly relapse rates were lowest in the group treated with cyclophosphamide. Tacrolimus monotherapy and combination regimens are currently being explored as potential options to minimize steroid exposure, and in resistant disease (Tse et al., 2007; Uchino et al., 2010).

Lupus membranous

The natural history of SLE-related MGN is summarized in a review of the available data, which pooled together studies including 157 patients with SLE and predominant membranous lesion on biopsy (World Health Organization (WHO) class V), with follow-up ranging from 25 to 279 months (Kolasinski et al., 2002). Only 23% of patients with ‘pure’ MGN had complete resolution of proteinuria at follow-up. Most patients (84.5%) continue to have normal renal function, but the numbers with abnormal renal function (14.6%) and with ESRD (5.1%) are hardly negligible. When data were available, subjects were also examined according to WHO classification subgroups based upon the presence (previously WHO class Vc and Vd) or absence (previously WHO class Va and Vb) of focal segmental changes or diffuse proliferative changes. Those with WHO class Vc and Vd had a higher rate of ESRD (35.7% vs 24.6%), but WHO subclass generally does not appear to be an independent predictor of renal failure when adjusted for renal function at presentation, proteinuria, anaemia, or age (Donadio et al., 1995). Much of the prognosis of SLE-MGN seems related to the subsequent development of proliferative changes on biopsy (Pasquali et al., 1993; Sloan et al., 1996). It is interesting to note that although there are limited data on the subject, there was a rather striking rate of non-renal-related deaths of 8.2% due to a variety of causes, emphasizing the serious prognostic implications of renal involvement in SLE. Although overall survival in SLE patients has improved over time, the 5-year survival of patients with SLE regardless of histologic type is worse if there is renal involvement (not specifically MGN) (Cameron, 1999).

A majority of patients with MGN secondary to systemic lupus erythematosus (SLE; see Chapter 161) present with nephrotic-range proteinuria (64%), and preserved renal function (90%) at the time of initial presentation (Kolasinski et al., 2002). Haematuria is variable; a large number of red cell casts should lead the clinician to suspect coexistent proliferative activity, and the possible presence of a mixed lesion on biopsy. Overall, a pure membranous lesion on biopsy is uncommon in SLE, diagnosed in only 14% of biopsies in SLE patients (Gruppo Italiano per lo Studio della Nefrite Lupica, 1992). Furthermore, up to 50% of patients with renal involvement will change histologic classification on later biopsies (Ponticelli and Moroni, 1998). In the case of MGN secondary to SLE, the prognosis seems very dependent upon the coexistence of proliferative lesions on biopsy. The therapy is therefore usually guided by the degree of proliferation or necrosis seen, or by the presence of systemic features that require therapy. An isolated pure membranous nephropathy in the absence of any systemic features or proliferation on biopsy is relatively rare. Although patients with pure membranous lesion are generally regarded as having a lower risk of end-stage renal disease (ESRD), morbidity associated with the persistence of nephrotic syndrome has prompted interest in the role of immunosuppression for this scenario (Donadio, 1992). There is evidence of favourable response rates with modified versions of the Italian regimen of combination steroids and cytotoxic agents (Pasquali et al., 1993; Moroni et al., 1998; Chan et al., 1999). When data from two multicentre randomized clinical trials were pooled (Ginzler et al., 2005; Appel et al., 2009), there were no differences in response rate between those treated with induction therapy with mycophenolate versus cyclophosphamide (Radhakrishnan et al., 2010). More recently, evidence regarding the utility of ciclosporin has emerged (Radhakrishnan et  al., 1994; Austin et  al., 2000). One randomized study of 42 patients with membranous nephropathy compared treatment with ciclosporin for 11  months (with steroids), alternate month IV cyclophosphamide for six doses (with steroids) and alternate

Natural history

Membranous glomerulonephritis in hepatitis B infection Hepatitis B (HBV)-associated MGN is usually associated with nephrotic-range proteinuria, and normal renal function (Lin, 1990; Lai et al., 1991; Peña et al., 2001). Clinical evidence of liver disease is not required for the development of MGN, and although the worldwide prevalence of HBV is quite high, the development of MGN in the context of this infection is relatively infrequent. The features of

558

Section 3  

the patient with glomerular disease

this condition are described more extensively in paediatric populations, who may have a different natural history compared with adults. Hepatitis and the kidney are further discussed in Chapter 185. The majority although not all patients with HBV who develop MGN, are also seropositive for hepatitis B e antigen (HBeAg) (Lin, 1990; Lai et al., 1991). The correlation of the presence of HBeAg and clinical outcome is debatable, although remission of MGN with the development of antibodies to this antigen has been reported (Hsu et al., 1989).

immune deposits (Hsu et al., 1989). The membranous lesion has been documented to actually resolve with the gradual spontaneous remission of the nephrotic syndrome (Gonzalo et al., 1999). The natural history in adults may not be as benign. One study of adults infected in an endemic area (therefore vertical transmission possible) indicated that spontaneous remission was uncommon in this population with progressive renal failure seen in 29%, and 10% reaching ESRD by the end of the average 60-month follow-up period (Lai et al., 1991).

Treatment

Membranous glomerulonephritis associated with malignancy

Data regarding therapy for HBV-related MGN are largely from paediatric populations, in whom spontaneous remissions are relatively common in children. The use of immunosuppressants in this condition carries the risk of allowing uncontrolled viral replication, and possible future exacerbation of hepatitis (Lai et al., 1990), but broadly, treatments that clear the infection improve the nephropathy. The treatment of Hepatitis B in patients with renal disease is reviewed in detail in Chapter 185. Here the results specifically in MGN are discussed. In children, the utility of alpha-interferon (IFN) following trials of prednisone was been studied, in a relatively large group of approximately 40 patients (Lin, 1995). Twenty received IFN therapy for 12 months, and 20 received supportive treatment only. Both groups had failed steroid therapy. At the end of the study period, all patients in the IFN group were completely free of proteinuria, compared to none in the control group. Disappearance of HBeAg occurred more frequently in the treatment group. In adults, one study of the natural history of HBV-related MGN documented a poor response to IFN in the 5 patients who received the therapy (Lai et al., 1991). A study assessing the long-term outcome of 15 patients indicated that half of those treated with interferon had a sustained remission in their liver disease with loss of HBV DNA and HbeAg, as well as favourable aminotransferase levels (Conjeevaram et al., 1995). The ‘responders’, all of whom had MGN as the associated renal lesion, also demonstrated marked improvement in proteinuria. From the limited data available, there appears to be a role for IFN, particularly in the presence of serologic markers of active liver disease. Patients treated with lamivudine (who also had elevated alanine aminotransferase and HBV DNA titres) had a high rate of remission compared to historical controls (60% vs 25%) and superior renal survival (Tang et  al., 2005). Long-term therapy with lamivudine may carry a risk of development of mutations conferring viral resistance to the drug and therefore judicious use of antiviral therapy is warranted. In endemic regions, vaccination programmes have substantially reduced the incidence of hepatitis B virus-associated membranous nephropathy (Bhimma et  al., 2003; Xu et  al., 2003); these programmes are an important strategy for minimizing the complications of this infection on a global front.

Natural history HBV-associated MGN has a variable natural history, depending upon the population studied. The majority of information is available from paediatric populations. In children, the disease generally has a benign course. The cumulative probability of remission at 4 years is 64% (Gilbert and Wiggelinkhuizen, 1994), and is usually associated with younger age, and smaller burden of subepithelial

The association of MGN with malignancy approaches 22% in patients > 60 years of age (Keur et al., 1989; Burstein et al., 1993). A  series of 155 patients indicated that 10% of patients > 60 had MGN in association with a malignancy, versus only 1% of patients < 60 years old (O’Callaghan et al., 2002). However it is not clear whether this reflects a causal relationship versus simply an age-related association. Some studies have included tumours diagnosed long after presentation; many of these are likely to be coincidental in this age group. Patients with MGN associated with malignancy tend to be older—the mean age was 63 years with all patients over 52, in one study of nine patients (Burstein et al., 1993). The average rate of malignancy in a population or subjects with the nephrotic syndrome is 6–11% (Lee et al., 1966; Row et al., 1975; Cahen et al., 1989; Yamauchi et al., 1985). In the series described by Burstein et al., all patients had nephrotic-range proteinuria, and six out of nine had evidence of renal insufficiency at presentation (one was related to obstruction). Most patients manifested proteinuria prior to or at the time of diagnosis of the malignancy. Reappearance of proteinuria may herald relapse of the malignancy. Although there is inadequate evidence at this point, it seems possible that the target podocyte antigen in patients with malignancy-associated MGN may be different from that seen in primary MGN (see Chapter 64). Apart from usual general management, and treatment appropriate for the tumour and the individual patient, there is no evidence on specific approaches to MGN in this setting. There are several reported cases of resolution of the MGN with treatment of the primary malignancy by resection or medical therapy (Robinson et al., 1984; Coltharp et al., 1991; Burstein et al., 1993).

Drug- and toxin-associated membranous nephropathy Glomerulonephritis associated with drugs and toxins is considered in Chapter 82. It is generally held that discontinuation of the causative agent results in a resolution of the nephrotic syndrome in patients with medication-related MGN. This has been documented in a relatively large series of patients with MGN related to gold therapy for rheumatoid arthritis (Hall et al., 1987). All patients had a remission of their proteinuria, although a period of up to 18  months was required for complete resolution. Renal function remained stable during the course of the illness and subsequent resolution of proteinuria.

Chapter 63 

References Appel, G. B., Contreras, G., Dooley, M. A., et al. (2009). Mycophenolate mofetil versus cyclophosphamide for induction treatment of lupus nephritis. J Am Soc Nephrol, 20, 1103–12. Austin, H. A., Illei, G. G., Braun, M. J., et al. (2000). Lupus membranous nephropathy: controlled trial of prednisone, pulse cyclophosphamide, and cyclosporine A (abstract). J Am Soc Nephrol, 11, A439. Austin, H. A., III, Illei, G. G., Braun, M. J., et al. (2009). Randomized, controlled trial of prednisone, cyclophosphamide, and cyclosporine in lupus membranous nephropathy. J Am Soc Nephrol, 20, 901–11. Bhimma, R., Coovadia, H. M., Adhikari, M., et al. (2003). The impact of the hepatitis B virus vaccine on the incidence of hepatitis B virus-associated membranous nephropathy. Arch Pediatr Adolesc Med, 157, 1025–30. Burstein, D. M., Korbet, S.M., and Schwartz, M.M. (1993). Membranous glomerulonephritis and malignancy. Am J Kidney Dis, 22, 5–10. Cahen, R., Francois, B., Trolliet, P., et al. (1989). Aetiology of membranous glomerulonephritis: a prospective study of 82 adult patients. Nephrol Dial Transplant, 4, 172–80. Cameron, J. S. (1999). Lupus nephritis. Am Soc Nephrol, 10, 413–24. Chan, T., Li, F. K., Hao, W. K., et al. (1999). Treatment of membranous lupus nephritis with nephrotic syndrome by sequential immunosuppression. Lupus, 8, 545–51. Coltharp, W. H., Lee, S. M., Miller, R. F., et al. (1991). Nephrotic syndrome complicating adenocarcinoma of the lung with resolution after resection. Ann Thorac Surg, 51, 308–9. Conjeevaram, H.S., Hoofnagle, J. H., Austin, H. A., et al. (1995). Long-term outcome of hepatitis B virus-related glomerulonephritis after therapy with interferon alfa. Gastroenterology, 109, 540–6. Donadio, J. V., Jr. (1992). Treatment of membranous nephropathy in systemic lupus erythematosus. Nephrol Dial Transplant, 7 Suppl 1, 97–104. Donadio, J. V., Jr. Hart, G. M., Bergstralh, E. J., et al. (1995). Prognostic determinants in lupus nephritis: a long-term clinicopathologic study. Lupus, 4, 109–15. Gilbert, R. D. and Wiggelinkhuizen, J. (1994). The clinical course of hepatitis B virus-associated nephropathy. Pediatr Nephrol, 8, 11–14. Ginzler, E. M., Dooley, M. A., Aranow, C., et al. (2005). Mycophenolate mofetil or intravenous cyclophosphamide for lupus nephritis. N Engl J Med, 353, 2219–28. Gonzalo, A., Mampaso, F., Bárcena, R., et al. (1999). Membranous nephropathy associated with hepatitis B virus infection: long-term clinical and histological outcome. Nephrol Dial Transplant, 14, 416–18. Gruppo Italiano per lo Studio della Nefrite Lupica (GISNEL) (1992). Lupus nephritis: prognostic factors and probability of maintaining life- supporting renal function 10 years after the diagnosis. Am J Kidney Dis, 19, 473–9. Hall, C. L., Fothergill, N. J., Blackwell, M. M., et al. (1987). The natural course of gold nephropathy: long term study of 21 patients. Br Med J (Clin Res Ed), 295, 745–8. Hsu, H. C., Wu, C. Y., Lin, C. Y., et al. (1989). Membranous nephropathy in 52 hepatitis B surface antigen (HBsAg) carrier children in Taiwan. Kidney Int, 36, 1103–7. Keur, I., Krediet, R.T., and Arisz, L. (1989). Glomerulopathy as a paraneoplastic phenomenon. Neth J Med, 34, 270–84. Kolasinski, S. L., Chung, J. B., and Albert, D. A. (2002). What do we know about lupus membranous nephropathy? An analytic review. Arthritis and Rheumatism, 47, 450–5.

secondary membranous glomerulonephritis

Lai, K. N., Li, P. K., Lui, S. F. et al. (1991). Membranous nephropathy related to hepatitis B virus in adults. N Engl J Med, 324, 1457–63. Lai, K. N., Tam, J. S., Lin, H. J., et al. (1990). The therapeutic dilemma of the usage of corticosteroid in patients with membranous nephropathy and persistent hepatitis B virus surface antigenaemia. Nephron, 54, 12–7. Lee, J. C., Yamauchi, H., Hopper, J., Jr. (1966). The association of cancer and the nephrotic syndrome. Ann Intern Med, 64, 41–51. Lin, C. Y. (1990). Hepatitis B virus-associated membraneous nephropathy: clinical features, immunological profiles and outcome. Nephron, 55, 37–44. Lin, C. Y. (1995). Treatment of hepatitis B virus-associated membranous nephropathy with recombinant alpha-interferon. Kidney Int, 47, 225–30. Moroni, G., Maccario, M., Banfi, G., et al. (1998). Treatment of membranous lupus nephritis. Am J Kidney Dis, 31, 681–6. O’Callaghan, C. A., Hicks, J., Doll, H., et al. (2002). Characteristics and outcome of membranous nephropathy in older patients. Int Urol Nephrol, 33, 157–65. Pasquali, S., Banfi, G., Zucchelli, A., et al. (1993). Lupus membranous nephropathy: long-term outcome. Clin Nephrol, 39, 175–82. Peña, A. Débora, M. J., Melgosa, M., et al. (2001). Membranous nephropathy associated with hepatitis B in Spanish children. Clin Nephrol, 55, 25–30. Ponticelli, C. and Moroni, G. (1998). Renal biopsy in lupus nephritis—what for, when and how often? Nephrol Dial Transplant, 13, 2452–4. Radhakrishnan, J., Kunis, C. L., D’Agati, V., et al. (1994). Cyclosporine treatment of lupus membranous nephropathy. Clin Nephrol, 42, 147–54. Radhakrishnan, J., Moutzouris, D. A., Ginzler, E. M., et al. (2010). Mycophenolate mofetil and intravenous cyclophosphamide are similar as induction therapy for class V lupus nephritis. Kidney Int, 77, 152–60. Robinson, W. L., Mitas, J. A. 2nd, Haerr, R. W., et al. (1984). Remission and exacerbation of tumor-related nephrotic syndrome with treatment of the neoplasm. Cancer, 54, 1082–4. Row, P. G., Cameron, J. S., Turner, D. R., et al. (1975). Membranous nephropathy. Long-term follow-up and association with neoplasia. QJM, 44, 207–39. Sloan, R. P., Schwartz, M. M., Korbet, S. M., et al. (1996). Long-term outcome in systemic lupus erythematosus membranous glomerulonephritis. Lupus Nephritis Collaborative Study Group. Am Soc Nephrol, 7, 299–305. Tang, S., Lai, F. M., Lui, Y. H., et al. (2005). Lamivudine in hepatitis B-associated membranous nephropathy. Kidney Int, 68(4), 1750–8. Tse, K. C., Lam, M. F., Tang, S. C., et al. (2007). A pilot study on tacrolimus treatment in membranous or quiescent lupus nephritis with proteinuria resistant to angiotensin inhibition or blockade. Lupus, 16, 46–51. Uchino, A., Tsukamoto, H., Nakashima, H., et al. (2010). Tacrolimus is effective for lupus nephritis patients with persistent proteinuria. Clin Exp Rheumatol, 28, 6–12. Xu, H., Sun, L., Zhou, L. J., et al. (2003). The effect of hepatitis B vaccination on the incidence of childhood HBV-associated nephritis. Pediatr Nephrol, 18, 1216–9. Yamauchi, H., Linsey, M. S., Biava, C. G., et al. (1985). Cure of membranous nephropathy after resection of carcinoma. Arch Intern Med, 145, 2061–3.

559

CHAPTER 64

Membranous glomerulonephritis: pathogenesis Daniel C. Cattran and Heather N. Reich Heymann nephritis shows that membranous glomerulonephritis is an autoantibody disease The Heymann model of experimental membranous nephropathy in rats bears many similarities to both the clinical and pathologic findings observed in human membranous glomerulonephritis (MGN). First described over 40  years ago, the immunization of susceptible strains of rats with Freund’s adjuvant and ‘rat kidney suspension’ consisting of material extracted from the brush border of proximal convoluted tubule cells, produces proteinuria and subepithelial deposits identical to that seen in human MGN (Heymann et  al., 1959). Much of the focus of research in the Heymann nephritis model has been related to the identification of the responsible antigen(s) and the subsequent immune response (Shankland, 2000). The deposition of immune complexes results in the activation of many mediators of injury, including leucocytes, complement, products of arachidonic acid metabolism, a variety of cytokines, adhesion molecules, and growth factors. The initial subject of intense investigation involving the Heymann nephritis model was the antigenic target(s) of the immune response stimulated by injection of the nephritogenic kidney membrane extract known as Fx1A. The target has been identified as a large membrane glycoprotein, gp330, which is also known as megalin due to its large size of 515 kD (Kerjaschki and Farquhar, 1982; Saito et al., 1994). The protein is a polyspecific receptor, and is a member of the low-density lipoprotein (LDL) receptor family (Herz et  al., 1988; Raychowdhury et  al., 1989, Kerjaschki and Neale, 1996). It complexes with a specific epitope of the 44 kD protein known as receptor associated protein, and can be found expressed in the clathrin-coated pits on the bases of podocyte foot processes (Kerjaschki and Farquhar, 1982; Pietromonaco et al., 1990; Kerjaschki et al., 1992; Orlando et al., 1992; Farquhar, 1996). The anatomic location of megalin supports its role in receptor-mediated endocytosis. The receptor-associated protein likely functions as a chaperone assisting in the folding of megalin in the endoplasmic reticulum of the cell, and facilitating its transport to the cell surface. Circulating antibodies directed against megalin are hypothesized to penetrate the glomerular basement membrane and bind to megalin/RAP forming immune complexes in situ. Megalin was confirmed to be the putative antigen in the Heymann model based on the following criteria:  (a)  active immunization of rats with megalin produces subepithelial immune deposits (active

Heymann nephritis), (b) injection of autologous anti-megalin antibodies produces similar deposits (passive Heymann nephritis), and (c)  antibodies eluted from affected rat glomeruli recognize only megalin (Kerjaschki and Farquhar, 1982; Farquhar et  al., 1995). Furthermore, a binding site for immunoglobulin G has been localized to the LDL-receptor-like domain of megalin, and Heymann nephritis has been reproduced using recombinant segments of this domain (Saito et al., 1996; Raychowdhury et al., 1996). As megalin is not expressed in human podocytes, and could not be found in human immune deposits, the hunt was on for a comparable target in human disease.

Human antigens Neutral endopeptidase: rare One of the first potential human antigens was identified in a case of antenatal membranous glomerulonephritis, which developed following pregnancy-induced immunization of a neutral endopeptidase-deficient mother and subsequent transplacental passage of antibodies directed against fetal neutral endopeptidase, a protein expressed on podocytes (Debiec et al., 2002). While further cases of the antenatal form of the disease were described in families with maternal mutations in the neutral endopeptidase gene, this antigen was not found to be responsible for the development of sporadic idiopathic MGN (Debiec et al., 2004).

M-type PLA2 receptor: common A breakthrough in the identification of a putative antigen came only recently, as a result of a painstaking graded sieving of protein extracts from pooled human glomeruli from healthy deceased subjects, and use of immunoglobulin (Ig)G-depleted protein fractions in a modified Western blotting protocol (Beck et al., 2009). The serum of a majority (70%) of patients with idiopathic MGN was found to display a 185 kD band detected by Western blotting with a fraction of the glomerular extract, under non-reducing conditions. The glomerular antigen in this band was identified as the M-type phospholipase A2 receptor protein. Antibodies directed against the phospholipase A2 receptor (PLA2R) (predominantly of the IgG4 subtype), were confirmed to be present in the serum of these affected individuals, and while not present in all subjects with idiopathic MGN, were notably absent in subjects with MGN associated with lupus or hepatitis. The PLA2R was also eluted from kidney biopsy immune deposits of patients with idiopathic MGN (Debiec and Ronco, 2011).

Chapter 64 

Mechanisms of injury

(A)

% Maximum

100

Proteinuria

Anti-PLA2R

0

(B)

Time

Treatment or spontaneous remission

Relapse

100 Anti-PLA2R

Proteinuria 0

(C)

Cytotoxicity

Time

Treatmentresistance

Treatment

100 Anti-PLA2R

% Maximum

Data derived from the Heymann model suggest that complement is required for the development of tissue injury and proteinuria, as illustrated by the fact that anti-Fx1A antibodies and immune complex deposition in C6- or C8-deficient rats do not result in proteinuria (Salant et al., 1980; Cybulsky et al., 1986; Baker et al., 1989). After the deposition of the antigen–antibody complex in the subepithelial space, the complement system is activated, leading to insertion of the C5b-9 membrane attack complex (MAC) into the podocyte plasma membrane (Couser et al., 1992). The MAC is formed after proteolytic cleavage of C5 and combination with components C6-C9. Its insertion into the cell membranes of nucleated cells results in sublytic activity (Koski et al., 1983), and in combination with other stimuli, is capable of causing cell death. The MAC has been localized to the subepithelial deposits and along the surface of podocytes, particularly in the clathrin-coated pits of glomerular epithelial cells in passive Heymann nephritis. It is likely endocytosed and transported intracellularly in multivescicular bodies, and then exocytosed into the urinary space (Kerjaschki et  al., 1987). Podocyte injury and resulting proteinuria has been shown to be dependent upon the presence of the MAC, as depletion of complement with cobra venom factor is associated with lack of proteinuria, despite the formation of immune deposits. The MAC has been documented in the urine of Heymann’s nephritis

Treatment or spontaneous remission

% Maximum

A genome-wide association study linked not only a class II antigen (human leucocyte antigen (HLA)-DQ1), a type of association expected in almost any autoimmune disease, but also the PLA2R1 gene itself with susceptibility to idiopathic MGN (Stanescu et al., 2011). The mechanism of this association is not yet known. The PLA2R protein is a type 1 transmembrane receptor expressed by human podocytes. Analysis of immune deposits suggests that complement activation is induced despite the fact that the predominant immunoglobulin present in these complexes, IgG4, is not classically thought to activate complement. Further work is required to delineate the mechanism by which anti-PLA2R antibodies may be pathogenic. Cumulative data suggest that anti-PLA2R antibodies are detectable in the serum of up to 80% of subjects with idiopathic disease when the Western blot technique is applied (Beck et  al., 2009; Hofstra et al., 2011; Qin et al., 2011). While assay specificity may account, in part, for variable detection of antibodies, the absence of circulating antibodies does not preclude the presence of anti-PLA2R antigen-antibody complexes in the kidney (Debiec and Ronco, 2011). The level of circulating antibody may be undetectable by the time a patient is referred for kidney biopsy, potentially reflecting clearance of the antibody following resolution of the precipitating immunologic event. A rise in antibody levels may precede the development of proteinuria and conversely, proteinuria may persist following clearing of serum antibodies, given the time required for immune complex reabsorption and basement membrane turn-over (Beck and Salant, 2010) (Fig. 64.1). While further work is required to clarify the relationship between serum antibody detection, proteinuria, and disease activity, multiple studies have suggested correlations (e.g. Beck et al., 2011) and it seems certain that tests for PLA2R antibodies will enter clinical practice soon.

membranous glomerulonephritis: pathogenesis

0

Proteinuria

Time

Fig. 64.1  Hypothetical situations relating serum PLA2R antibodies to disease status in idiopathic MGN. (A) In panel A, representing remitting disease, a reduction in PLA2R antibody titre precedes the reduction in proteinuria; this may represent a resolution in immunologic disease activity preceding basement membrane regeneration and healing. (B) In panel B, representing relapsing disease, the rise in serum PLA2R antibody titre precedes and heralds clinical relapse of proteinuria. (C) In panel C, representing resistant disease, neither PLA2R antibody titre nor proteinuria demonstrate any change over time or in response to therapy.

561

562

Section 3  

the patient with glomerular disease

animals, and may be a marker of disease activity (Schulze et al., 1989; Pruchno et al., 1989). Although it appears that epithelial cell injury is dependent upon the presence of the MAC, the mechanisms by which the MAC causes injury and the podocyte’s maladaptive response to this injury remain areas of ongoing active investigation. Some of the proposed pathways of injury induced by the MAC include those mediated by the generation of reactive oxygen species (ROS), podocyte cytoskeleton modifications, and podocyte apoptosis. Injury to both the podocyte and the basement membrane is hypothesized to be mediated in part by local generation of reactive oxygen species (ROS) (Adler et al., 1986; Peng et al., 2002). The ROS may in fact be produced locally, as is suggested by in vitro experiments in which the application of C5b-9 to cultured epithelial cells results in the production of ROS (Adler et al., 1986). Subsequent injury to the filtration barrier may then be induced by peroxidation of membrane proteins and collagen (Kerjaschki and Neale, 1996). The MAC has also been documented to activate epithelial cell cytosolic phospholipase A2 in vivo and in vitro, resulting in the production of free arachidonic acid by post-translational regulation of the enzyme (Cybulsky et al., 1995, 2000). Arachidonic acids are known to be precursors of eicosanoids (e.g. prostaglandins, leukotrienes, and thromboxanes), that have potential haemodynamic effects within the glomerulus, and are present in increased levels in the Heymann model. The presence of ROS in tissue has been correlated with the development of proteinuria, and the administration of ROS scavengers (antioxidants) has been documented to cause a decrease in proteinuria (Shah, 1988, 1989). The ROS generated by the MAC are also postulated to mediate activation of a variety of stress-associated pathways, including protein kinase C and the extracellular-signal-related kinase ERK2 (Cybulsky et al., 1998).

Structural effects Observations from the Heymann model suggest a link between structural changes in the podocyte and interconnecting slit diaphragms with the immune complex formation, permitting the passage of protein through the filtration barrier. Microscopically, foot process effacement is readily appreciable in MGN. This may be related to disruption of the actin cytoskeleton that has been documented to be mediated by MAC exposure (Topham et al., 1999). It has been demonstrated that nephrin dissociates from the actin cytoskeleton in experimental membranous nephropathy, resulting in prominent changes in the morphology of the slit diaphragm (Yuan et  al., 2002). Slit diaphragms have also been observed to undergo morphologic and phenotypic changes, confirmed by documentation in human samples of decreased mRNA levels of nephrin, the primary component of the diaphragm (Furness et al., 1999). The C5b-9-mediated glomerular epithelial cell structural and slit diaphragm changes observed in the Heymann model may be modulated in part by cytochrome P450 2B1(CYP2B1) (Liu et al., 2010); cytoskeletal integrity appears to be maintained in cells deficient in this cytochrome protein, and inhibition of CYP2B1 moderates injury in the passive Heymann model.

Podocyte injury and death Podocyte apoptosis is thought to be mediated both directly and indirectly by C5b-9. The induction of MGN is associated with an overall decrease in podocyte number, and this has been attributed to MAC-induced podocyte apoptosis (Shankland et al., 1997) and

podocyte detachment (Petermann et al., 2003). Indirectly the MAC may also potentiate angiotensin II and transforming growth factor beta (TGF-β)-mediated podocyte apoptosis (Ding et al., 2002; Zoja et al., 2003). The podocyte reaction to injury induced by the MAC is one of hypertrophy, rather than proliferation. It is postulated that interruption of cell cycle regulation may cause the lack of proliferation of podocytes (Shankland et al., 1997, 1999). In addition, an increase in extracellular matrix is observed, and can be appreciated at a microscopic level as thickening of the basement membrane. The accumulation of extracellular matrix may be mediated to a large extent by TGF-β, as is suggested by experiments that indicate that only certain TGF-β/receptor isoforms lead to matrix expansion in the Heymann model (Shankland et al., 1996). The sublytic C5b-9 exposure may also contribute to podocyte elaboration of type I and type IV collagen (Kim et al., 1991; Minto et al., 1993).

Human studies In human MGN studies, the presence of C5b-9 has been documented within immune complex deposits (Hinglais et al., 1986). In one biopsy study, MAC was found in 50% of biopsies of idiopathic MGN, 75% of MGN secondary to lupus, and a minority of MGN due to hepatitis B (Lai et al., 1989). The presence of MAC on biopsy in one recent study was not, however, found to be correlated with the severity of proteinuria or the presence of the nephrotic syndrome; it was suggested that rather than MAC deposition, glomerular expression of various adhesion molecules may be more reflective of advanced disease (Honkanen et al., 1998; Papagianni et  al., 2002). The MAC has also been identified in the urine of human subjects with MGN. The clinical course of the disease has been shown to have some association with the presence of urinary C5b-9 in human studies; two studies have demonstrated that the presence of MAC is correlated with an ‘active’ clinical course with ongoing proteinuria, and declining function (Coupes et al., 1992, 1993; Kon et al., 1995). One study, however, correlated its presence with a shorter duration of disease and lower creatinine, but more proteinuria (Schulze et  al., 1991). The differences may relate to small sample sizes, cross-sectional design, and differing statistical methods used to prove the association.

Other mechanisms of injury Although much of the focus thus far has been on complement-mediated injury, cellular and humoral-mediated pathways are also thought to be important factors in the ultimate development of proteinuria, at least in the experimental model. A  role for cell-mediated injury is supported by the observations that depletion of the cytotoxic CD8+ T-cell reduces the injury and that monoclonal anti-CD4 and anti-CD8 treatment modifies the disease (Penny et al., 1998). However, increased protein permeability was demonstrated in isolated perfused rat glomeruli exposed to antibodies derived from rats with Heymann nephritis, in the complete absence of infiltrating leucocytes (Neale et al., 1982), supporting the notion that the importance of leucocytes in mediating glomerular permeability may differ according to the stage of disease pathogenesis (reviewed by Glassock, 2010). In addition to injury induced by MAC, pathways of tubulointerstitial injury and subsequent loss of kidney function are likely shared with many forms of progressive proteinuric kidney disease. These may include the loss of microvascular endothelium, and

Chapter 64 

proteinuria-induced upregulation of renal cytokines and growth factors that promote tubulointerstitial inflammation and fibrosis (Remuzzi and Bertani, 1998; Remuzzi et al., 1994).

Genetics Both susceptibility to and progression of MGN may have a genetic basis (Vaughan et  al., 1995; Kleta, 2008). There is evidence that HLA type may underlie susceptibility to this condition, and potentially account for rare familial variants of the disease (Muller et al., 2008; Stanescu et al., 2011). The HLA type may also be associated with tendency towards progressive disease epidemiology (Abe et al., 1986; Kida et al., 1986; Papiha et al., 1987; Freedman et al., 1994). Advances in laboratory technology that allow a broader look at whole genomes and changes in gene and protein expression may help to elucidate additional pathways responsible for the injury that produces MGN. In addition, further investigation into the factors responsible for an individual’s susceptibility to either progression or spontaneous remission in this type of immunologic injury is required.

References Abe, S., Amagasaki, Y., Konishi, K., et al. (1986). Idiopathic membranous glomerulonephritis: aspects of geographical differences. J Clin Pathol, 39, 1193–8. Adler, S., Baker, P. J., Johnson, R. J., et al. (1986). Complement membrane attack complex stimulates production of reactive oxygen metabolites by cultured rat mesangial cells. J Clin Invest, 77, 762–7. Baker, P. J., Ochi, R. F., Schulze, M., et al. (1989). Depletion of C6 prevents development of proteinuria in experimental membranous nephropathy in rats. Am J Pathol, 135, 185–94. Beck, L. H., Jr., Bonegio, R. G., Lambeau, G., et al. (2009). M-type phospholipase A2 receptor as target antigen in idiopathic membranous nephropathy. N Engl J Med, 361, 11–21. Beck, L. H., Jr. and Salant, D. J. (2010). Membranous nephropathy: recent travels and new roads ahead. Kidney Int, 77, 765–70. Coupes, B., Brenchley, P. E., Short, C. D., et al. (1992). Clinical aspects of C3dg and C5b-9 in human membranous nephropathy. Nephrol Dial Transplant, 7 (Suppl 1), 32–4. Coupes, B. M., Kon, S. P., Brenchley, P. E., et al. (1993). The temporal relationship between urinary C5b-9 and C3dg and clinical parameters in human membranous nephropathy. Nephrol Dial Transplant, 8, 397–401. Couser, W. G., Schulze, M., and Pruchno, C. J. (1992). Role of C5b-9 in experimental membranous nephropathy. Nephrol Dial Transplant, 7 Suppl 1, 25–31. Cybulsky, A. V., Monge, J. C., Papillon, J., et al. (1995). Complement C5b-9 activates cytosolic phospholipase A2 in glomerular epithelial cells. Am J Physiol, 269, F739–49. Cybulsky, A. V., Papillon, J., and McTavish, A.J. (1998). Complement activates phospholipases and protein kinases in glomerular epithelial cells. Kidney Int, 54, 360–72. Cybulsky, A. V., Rennke, H. G., Feintzeig, I. D., et al. (1986). Complement-induced glomerular epithelial cell injury. Role of the membrane attack complex in rat membranous nephropathy. J Clin Invest, 77, 1096–107. Cybulsky, A. V., Takano, T., Papillon, J., et al. (2000). Complement-induced phospholipase A2 activation in experimental membranous nephropathy. Kidney Int, 57, 1052–62. Debiec, H., Guigonis, V., Mougenot, B., et al. (2002). Antenatal membranous glomerulonephritis due to anti-neutral endopeptidase antibodies. N Engl J Med, 346, 2053–60. Debiec, H., Nauta, J., Coulet, F., et al. (2004). Role of truncating mutations in MME gene in fetomaternal alloimmunisation and antenatal glomerulopathies. Lancet, 364, 1252–9.

membranous glomerulonephritis: pathogenesis

Debiec, H. and Ronco, P. (2011). PLA2R autoantibodies and PLA2R glomerular deposits in membranous nephropathy. N Engl J Med, 364, 689–90. Ding, G., Reddy, K., Kapasi, A. A., et al. (2002). Angiotensin II induces apoptosis in rat glomerular epithelial cells. Am J Physiol Renal Physiol, 283, F173–80. Farquhar, M. G., Saito, A., Kerjaschki, D., et al. (1995). The Heymann nephritis antigenic complex: megalin (gp330) and RAP. Am Soc Nephrol, 6, 35–47. Freedman, B. I., Spray, B. J., Dunston, G. M., et al. (1994). HLA associations in end-stage renal disease due to membranous glomerulonephritis: HLA-DR3 associations with progressive renal injury. Southeastern Organ Procurement Foundation. Am J Kidney Dis, 23, 797–802. Furness, P. N., Hall, L. L., Shaw, J. A., et al. (1999). Glomerular expression of nephrin is decreased in acquired human nephrotic syndrome. Nephrol Dial Transplant, 14, 1234–7. Glassock, R. J. (2010). The pathogenesis of idiopathic membranous nephropathy: a 50-year odyssey. Am J Kidney Dis, 56, 157–67. Herz, J., Hamann, U., Rogne, S., et al. (1988). Surface location and high affinity for calcium of a 500-kd liver membrane protein closely related to the LDL-receptor suggest a physiological role as lipoprotein receptor. EMBO J, 7, 4119–27. Heymann, W., Hackel, D. B., Harwood, S., et al. (1959). Production of nephrotic syndrome in rats by Freund’s adjuvants and rat kidney suspensions. 1951. Proc Soc Exp Biol Med, 100, 600–7. Hinglais, N., Kazatchkine, M. D., Bhakdi, S., et al. (1986). Immunohistochemical study of the C5b-9 complex of complement in human kidneys. Kidney Int, 30, 399–410. Hofstra, J. M., Beck, L. H. Jr, Beck, D. M., et al. (2011). Anti-phospholipase A(2) receptor antibodies correlate with clinical status in idiopathic membranous nephropathy. Clin J Am Soc Nephrol, 6, 1286–91. Honkanen, E., von Willebrand, E., Teppo, A. M., et al. (1998). Adhesion molecules and urinary tumor necrosis factor-alpha in idiopathic membranous glomerulonephritis. Kidney Int, 53, 909–17. Kerjaschki, D. and Farquhar, M. G. (1982). The pathogenic antigen of Heymann nephritis is a membrane glycoprotein of the renal proximal tubule brush border. Proc Natl Acad Sci U S A, 79, 5557–81. Kerjaschki, D. and Neale, T.J. (1996). Molecular mechanisms of glomerular injury in rat experimental membranous nephropathy (Heymann nephritis). Am Soc Nephrol, 7, 2518–26. Kerjaschki, D., Miettinen, A., and Farquhar, M.G. (1987). Initial events in the formation of immune deposits in passive Heymann nephritis. gp330-anti-gp330 immune complexes form in epithelial coated pits and rapidly become attached to the glomerular basement membrane. J Exp Med, 166, 109–28. Kerjaschki, D., Ullrich, R., Diem, K., et al. (1992). Identification of a pathogenic epitope involved in initiation of Heymann nephritis. Proc Natl Acad Sci U S A, 89, 11179–83. Kida, H., Asamoto, T., Yokoyama, H., et al. (1986). Long-term prognosis of membranous nephropathy. Clin Nephrol, 25, 64–9. Kim, Y., Butkowski, R., Burke, B., et al. (1991). Differential expression of basement membrane collagen in membranous nephropathy. Am J Pathol, 139, 1381–8. Kleta, R. (2008). Fanconi or not Fanconi? Lowe syndrome revisited. Clin J Am Soc Nephrol, 3, 1244–5. Kon, S. P., Coupes, B., Short, C. D., et al. (1995). Urinary C5b-9 excretion and clinical course in idiopathic human membranous nephropathy. Kidney Int, 48, 1953–8. Koski, C. L., Ramm, L. E., Hammer, C. H., et al. (1983). Cytolysis of nucleated cells by complement: cell death displays multi- hit characteristics. Proc Natl Acad Sci U S A, 80, 3816–20. Lai, K. N., Lo, S.T., Lai, F.M. (1989). Immunohistochemical study of the membrane attack complex of complement and S-protein in idiopathic and secondary membranous nephropathy. Am J Pathol, 135, 469–76. Liu, H., Tian, N., Arany, I., et al. (2010). Cytochrome P450 2B1 mediates complement-dependent sublytic injury in a model of membranous nephropathy. J Biol Chem, 285, 40901–10.

563

564

Section 3  

the patient with glomerular disease

Minto, A. W., Fogel, M. A., Natori, Y., et al. (1993). Expression of type I collagen mRNA in glomeruli of rats with passive Heymann nephritis. Kidney Int, 43, 121–7. Muller, C., Alenabi, F., Chantrel, F., et al. (2008). Familial membranous glomerulopathy, toxic exposure and/or genetic sensibility? Clin Nephrol, 70, 422–3. Neale, T. J., Couser, W. G., Salant, D. J., et al. (1982). Specific uptake of Heymann's nephritic kidney eluate by rat kidney: studies in vivo and in isolated perfused kidneys. Lab Invest, 46, 450–3. Orlando, R. A., Kerjaschki, D., Kurihara, H., et al. (1992). gp330 associates with a 44-kDa protein in the rat kidney to form the Heymann nephritis antigenic complex. Proc Natl Acad Sci U S A, 89, 6698–702. Papagianni, A. A., Alexopoulos, E., Leontsini, M., et al. (2002). C5b-9 and adhesion molecules in human idiopathic membranous nephropathy. Nephrol Dial Transplant, 17, 57–63. Papiha, S. S., Pareek, S. K., Rodger, R. S., et al. (1987). HLA-A, B, DR and Bf allotypes in patients with idiopathic membranous nephropathy (IMN). Kidney Int, 31, 130–4. Peng, H., Takano, T., Papillon, J., et al. (2002). Complement activates the c-Jun N-terminal kinase/stress-activated protein kinase in glomerular epithelial cells. J Immunol, 169, 2594–601. Penny, M. J., Boyd, R.A., and Hall, B.M. (1998). Mycophenolate mofetil prevents the induction of active Heymann nephritis: association with Th2 cytokine inhibition. Am Soc Nephrol, 9, 2272–82. Petermann, A. T., Krofft, R., Blonski, M., et al. (2003). Podocytes that detach in experimental membranous nephropathy are viable. Kidney Int, 64, 1222–31. Pietromonaco, S., Kerjaschki, D., Binder, S., et al. (1990). Molecular cloning of a cDNA encoding a major pathogenic domain of the Heymann nephritis antigen gp330. Proc Natl Acad Sci U S A, 87, 1811–15. Pruchno, C. J., Burns, M. W., Schulze, M., et al. (1989). Urinary excretion of C5b-9 reflects disease activity in passive Heymann nephritis. Kidney Int, 36, 65–71. Qin, W., Beck, L. H. Jr., Zeng, C., et al. (2011). Anti-phospholipase A2 receptor antibody in membranous nephropathy. Am Soc Nephrol, 22, 1137–43. Raychowdhury, R., Niles, J. L., McCluskey, R. T., et al. (1989). Autoimmune target in Heymann nephritis is a glycoprotein with homology to the LDL receptor. Science, 244, 1163–5. Raychowdhury, R., Zheng, G., Brown, D., et al. (1996). Induction of Heymann nephritis with a gp330/megalin fusion protein. Am J Pathol, 148, 1613–23. Remuzzi, G., Schieppati, A., Garattini, S. (1994). Treatment of idiopathic membranous glomerulopathy. Curr Opin Nephrol Hypertens, 3, 155–63. Remuzzi, G. and Bertani, T. (1998). Pathophysiology of progressive nephropathies. N Engl J Med 339, 1448–56.

Saito, A., Pietromonaco, S., Loo, A. K., et al. (1994). Complete cloning and sequencing of rat gp330/ ‘megalin,’ a distinctive member of the low density lipoprotein receptor gene family. Proc Natl Acad Sci U S A, 91, 9725–9. Saito, A., Yamazaki, H., Rader, K., et al. (1996). Mapping rat megalin: the second cluster of ligand binding repeats contains a 46-amino acid pathogenic epitope involved in the formation of immune deposits in Heymann nephritis. Proc Natl Acad Sci U S A, 93, 8601–5. Salant, D. J., Belok, S., Madaio, M. P., et al. (1980). A new role for complement in experimental membranous nephropathy in rats. J Clin Invest, 66, 1339–50. Schulze, M., Baker, P. J., Perkinson, D. T., et al. (1989). Increased urinary excretion of C5b-9 distinguishes passive Heymann nephritis in the rat. Kidney Int, 35, 60–8. Schulze, M., Donadio, J. V. Jr., Pruchno, C. J., et al. (1991). Elevated urinary excretion of the C5b-9 complex in membranous nephropathy. Kidney Int, 40, 533–8. Shah, S. V. (1988). Evidence suggesting a role for hydroxyl radical in passive Heymann nephritis in rats. Am J Physiol, 254, F337–44. Shah, S. V. (1989). Role of reactive oxygen metabolites in experimental glomerular disease. Kidney Int, 35, 1093–106. Shankland, S. J. (2000). New insights into the pathogenesis of membranous nephropathy. Kidney Int, 57, 1204–5. Shankland, S. J., Floege, J., Thomas, S. E., et al. (1997). Cyclin kinase inhibitors are increased during experimental membranous nephropathy: potential role in limiting glomerular epithelial cell proliferation in vivo. Kidney Int, 52, 404–13. Shankland, S. J., Pippin, J.W., Couser, W.G. (1999). Complement (C5b-9) induces glomerular epithelial cell DNA synthesis but not proliferation in vitro. Kidney Int, 56, 538–48. Shankland, S. J., Pippin, J., Pichler, R. H., et al. (1996). Differential expression of transforming growth factor-beta isoforms and receptors in experimental membranous nephropathy. Kidney Int, 50, 116–24. Stanescu, H. C., Arcos-Burgos, M., Medlar, A., et al. (2011). Risk HLA-DQA1 and PLA(2)R1 alleles in idiopathic membranous nephropathy. N Engl J Med, 364, 616–26. Topham, P. S., Haydar, S. A., Kuphal, R., et al. (1999). Complement-mediated injury reversibly disrupts glomerular epithelial cell actin microfilaments and focal adhesions. Kidney Int, 55, 1763–75. Vaughan, R. W., Tighe, M. R., Boki, K., et al. (1995). An analysis of HLA class II gene polymorphism in British and Greek idiopathic membranous nephropathy patients. Eur J Immunogenet, 22, 179–86. Yuan, H., Takeuchi, E., Taylor, G. A., et al. (2002). Nephrin dissociates from actin, and its expression is reduced in early experimental membranous nephropathy. Am Soc Nephrol, 13, 946–56. Zoja, C., Morigi, M., and Remuzzi, G. (2003). Proteinuria and phenotypic change of proximal tubular cells. Am Soc Nephrol, 14 Suppl 1, S36–41.

CHAPTER 65

Immunoglobulin A nephropathy: overview Kar Neng Lai and Sydney C. W. Tang Immunoglobulin A (IgA) nephropathy is the most common primary glomerulonephritis. It runs a slow and sometimes relentless clinical course (see Chapter 66) with consequent end-stage renal failure in 35–40% of patients 25–30 years after first clinical presentation (see Chapter 68). The pathology is characterized by deposition of macromolecular (polymeric) IgA1 in the glomerular mesangium, proliferation of mesangial cells, increased synthesis of extracellular matrix, and infiltration of macrophages, monocytes, and T cells (Fig. 65.1) (see Chapter 67). The severity of glomerular and tubulointerstitial damage in IgA nephropathy correlates with the rate of renal function decline and long-term renal outcome. However, IgA deposition is a common incidental finding at autopsy and in some patients is associated with minimal or no overt renal disease. The kidney is believed to be an innocent bystander in IgA nephropathy (see Chapter 69). The primary defect appears to be aberrant glycosylation of O-linked glycans in the hinge region of a fraction of IgA1 molecules. Rather than terminating with galactose, the aberrant galactose-deficient O-glycans end with N-acetylgalactosamine or sialylated acetylgalactosamine. The absence of galactose in O-glycans reduces their uptake by the liver

and reticuloendothelial system by asialoglycoprotein receptor. The terminal N-acetylgalactosamine moiety on the aberrantly glycosylated IgA1 may in turn be recognized by antiglycan antibodies. The aberrant IgA1 or nephritogenic immune complexes may induce kidney injury. The aberrant underglycosylation of macromolecular IgA1 explains the recurrence of IgA nephropathy in transplanted kidney. Serum galactose-deficient IgA1 levels appear to be heritable in a dominant pattern in IgA nephropathy, although most relatives with high levels have no clinical manifestation of renal injury. Familial IgA nephropathy is uncommon but may be under-recognized. It may have a poorer prognosis. This supports that genetic factors are involved in the pathogenesis of IgA nephropathy and specific candidate genes have been detected in selected ethnic groups. Blockade of the renin–angiotensin system and blood pressure control remain the mainstay of treatment (see Chapter 68). Courses (months) of high-dose corticosteroids have antiproteinuric effects and seem to preserve glomerular filtration rate in selected patients. The efficacy of other immunosuppressive agents remains debatable.

Fig. 65.1  (A) Immunofluorescent staining for IgA deposits in the mesangium (×500). (B) Moderate mesangial matrix expansion with increased cell number (H&E ×400). (C) Electron micrograph showing mesangial expansion with electron-dense deposits (arrows) (×9600).

CHAPTER 66

Immunoglobulin A nephropathy: clinical features Kar Neng Lai and Sydney C. W. Tang Introduction Primary IgA nephropathy is the commonest form of idiopathic (primary) glomerulonephritis in the developed world and it is an important cause of end-stage kidney failure. In 1967, Drs Jean Berger and Nicole Hinglais at the Paris Necker Hospital first described a new glomerulopathic entity that they subsequently called mesangial IgA/IgG deposition with IgA predominance following the application of new technique of immunofluorescence staining (Berger and Hinglais, 1968). Levy et al. (1972) first used in print the term ‘Berger’s disease’. By 1975, Berger disease became an established glomerular entity: a condition with moderate proliferative glomerular changes, usually mesangial but often focal or segmental in distribution, associated with microscopic haematuria and about 15–20% with macroscopic haematuria. IgA nephropathies are characterized by the presence of diffuse mesangial deposition of IgA in the glomeruli in selected pathological entities such as Berger disease, Henoch–Schönlein purpura, and systemic lupus erythematosus. Secondary IgA nephropathy may occur in several systemic diseases when associated with an abnormal response of the IgA immune system. Cirrhosis and heavy alcohol intake may induce nephritis with IgA deposits. The association between staphylococcal infection and IgA-predominant or co-dominant glomerulonephritis (see Chapter  78) was first reported in Japan, and subsequently in other regions. Distinction of this entity from primary IgA nephropathy is important to avoid immunosuppressive treatment.

Epidemiology Geographical distribution Mesangial IgA deposits are also present in 4–16% of normal, healthy adults, living and cadaveric donors. Thus, the biopsy-proven IgA nephropathy cases represent a very small fraction of the total individuals with disease in the population as a whole (Levy and Berger, 1988). The systematic screening of urinary abnormalities could have influenced the higher prevalence reported both in Japan and in Singapore and different clinical policies for diagnostic renal biopsy may account for lower detection of IgA nephropathy (Schena et al., 1990). However, the finding of familial aggregation of IgA nephropathy has hinted that genetic factors are important in the aetiology of IgA nephropathy. The clinical onset of IgA nephropathy is usually in the second and third decade of life but may occur at any age. Males are

affected from two- to sixfold more frequently than females in North America and Europe but from 1.5- to twofold in Asia. It is more frequent in white people and Asians than in African Americans, and rarely reported in black people of direct African descent. The difference is still unexplained. Most of the worldwide studies report prevalence rates as a percentage of cases of primary glomerulonephritides or as a percentage of a total series of renal biopsies, while few epidemiologic studies focused on the real incidence of primary IgA nephropathy in various populations (Table 66.1).

Familial studies The best evidence for a genetic effect comes from reports of familial aggregation of the disease, sometimes recognized when screening for prospective kidney donors (Lavigne et al., 2010). As a result of the requirement for renal biopsy for diagnosis, and the intermittence of urinary abnormalities, no systematic study has reported the prevalence of familial IgA nephropathy or sibling-recurrence risk. A good family history should document occurrence of kidney disease or unexplained haematuria in first-degree and second-degree relatives, any childhood deaths, age of onset of disease in all cases, gender distribution, ethnic origin, presence of consanguinity, and potential environmental exposures. Familial aggregation of biopsy-confirmed IgA nephropathy was independently reported in two families in the late 1970s. Since then, this has been described with increasing frequency. Reports from the United Kingdom and Germany indicated that 4–10% of patients with IgA nephropathy had a family history of renal disease (Rambausek et al., 1987; Johnston et al., 1992). Schena et al. (1993) detected urinary abnormalities in 61 of 269 asymptomatic first-degree relatives from 48 families of IgA nephropathy patients. In another Italian survey lasting 25  years, IgA nephropathy was diagnosed in 185 patients; 26 of these (14%) were related to at least one other IgA nephropathy patient, forming ten multiplex pedigrees (Scolari et al., 1999). Levy (1989) discovered 40 families with two or three members who had biopsy-documented IgA nephropathy. The majority of families were of European origin, but there were also families from Asia and North America (Tam et al., 2009; Lavigne et al., 2010). These data indicate that familial IgA nephropathy is quite common, but probably underdiagnosed as urinalysis is not routinely performed and microhaematuria is intermittent. In addition to multiplex kindreds, epidemiologic investigations have identified an increased prevalence of IgA nephropathy in some isolated populations in geographically isolated areas. In such populations,

Chapter 66 

Table 66.1  Frequency of IgA nephropathy amongst renal biopsies of primary glomerulonephritis Frequency (%)

Number of renal biopsy

Brazil (Bahiense-Oliveira et al., 2004)

21.4 (A)

943

Peru (Hurtado et al., 2000)

0.9 (A)

1263

USA (Swaminathan et al., 2006)

21.4 (A)

5586

America

Asia China (Lai et al., 2004)

45.3 (A)

13519

Hong Kong (Lai et al., 1994)

35.0 (A)

961

India (Chandrika, 2007)

14.3 (A)

1544

Japan (Research Group on Progressive

47.4 (A)

1045

Chronic Renal Disease, 1999)

22.1 (A)

4514

Korea (Choi et al., 2001)

10.3 (P)

3555

Singapore (Sinniah, 1983)

34.0 (A)

Thailand (Parichatikanond et al., 2006)

17.9 (A)

34.1 (A)

2030

Croatia (Batinic et al., 2007)

18.1 (A)

565

France (Simon et al., 2004)

29.7 (A)

600

Germany (Werner et al., 2009)

26 (A)

359

Italy (Schena et al., 1997)

21.5 (A)

32862

Europe 20.0 (P)

18.8 (P) Netherlands (Tiebosch et al., 1987)

22.0 (A)

Portugal (Carvalho et al., 2006)

31.2 (A)

Romania (Covic et al., 2006)

28.9 (A)

635

Spain (Rivera et al., 2002)

17.2 (A)

7016

19.5 (P) United Kingdom (Ballardie et al., 1987)

after apparent onset of the disease, resulting in 41% renal survival rate compared with 94% in patients with sporadic IgA nephropathy (Schena et al., 2002; Tam et al., 2010). Abnormalities of IgA glycosylation have been reported in relatives of patients (see ‘Aberrant structure of the IgA molecule’ in Chapter 69). Families in which IgA nephropathy aggregate can be associated with other forms of glomerular disease including IgM nephropathy, thin basement membrane disease, and most importantly, Henoch–Schönlein purpura (Levy, 1989; Frasca et al., 2004).

Immunogenetic association Most IgA nephropathies are ‘sporadic’ rather than familial. IgA nephropathy is generally considered to be a complex disorder, that is, it is a multifactorial disease with both genetic and environmental factors likely contributing in the majority. However, in some families the disease segregates in an obviously autosomal dominant fashion. It therefore seems likely that the genetic contribution to the disease is heterogeneous, and can lie anywhere in the spectrum from monogenic, through oligogenic to polygenic, differing in individual cases and families. Recent studies suggest genes responsible for sporadic and familial IgA nephropathy could well be different.

Genes responsible for, or predisposing to, IgA nephropathy

Australia (Briganti et al., 2001)

immunoglobulin a nephropathy: clinical features

31.0 (A)

A = adult, P = paediatric.

individuals with sporadic IgA nephropathy have been shown to share common ancestors as many as seven to eight generations (Julian et al., 1985; Izzi et al., 2006). No familial clustering in nearby villages with similar population histories and lifestyles strongly indicates an inherited rather than environmental mechanism. There is little difference between familial and sporadic forms of IgA nephropathy with respect to clinical presentations. Additionally, histologic findings, frequency of the immunoglobulin isotype and presence of complement C3 in renal tissue do not differ between the two forms. Two studies reported a worse renal prognosis and a more severe histopathology for familial IgA nephropathy 20 years

Familial IgA nephropathy and linked loci 2q36 locus The most clear-cut report of familial IgA nephropathy and linkage is a study of a four- generation Canadian family of German-Austrian origin with 14 affected and 11 unaffected members (Paterson et al., 2007). The pedigree is consistent with autosomal dominant inheritance. Parametric and non-parametric linkage analysis produced significant logarithm of the ratio of odds (LOD) scores according to standard criteria for Mendelian disease. 6q22–q23 (IGAN1) and 3p24–p23 loci These were the first loci identified in linkage analysis of IgA nephropathy with a LOD score of 5.6 to 6q22–q23, which was named IGAN1 (Gharavi et al., 2000). The study analysed 24 Italian and six American families suggesting locus at 3p24–p23 with maximum LOD score of 2.8. 4q26–q31 (IGAN2) and 17q12–q22 (IGAN3) loci The European IgAN consortium identified suggestive loci at 4q26–q31 (LOD score 1.8) and 17q12–q22 (LOD score 2.6), in 22 Italian families with 59 affected and 127 unaffected members. The loci were named IGAN2 and IGAN3 (Bisceglia et al., 2006). These four loci have not been revealed in familial IgA nephropathy of other ethnicity (Karnib et al., 2007).

Genome-wide association study in sporadic IgA nephropathy By genome-wide association study (GWAS), a strong signal of association on chromosome 6p in the region of the major histocompatibility complex (P = 1 × 10−9) was identified in a cohort of 533 European patients with IgA nephropathy recruited from the United Kingdom Glomerulonephritis DNA Bank (Feehally et  al., 2010). Human leucocyte antigen (HLA) imputation analysis showed the strongest association signal from a combination of DQ loci with some support for an independent HLA-B signal. These results suggest that the HLA region contains the strongest common susceptibility alleles that predispose to IgA nephropathy in the European

567

568

Section 3  

the patient with glomerular disease

population. Using similar approach in 1194 patients of Chinese Han ethnicity, Gharavi et al. (2011) identified three independent loci in the major histocompatibility complex, as well as a common deletion of CFHR1 and CFHR3 at chromosome 1q32 and a locus at chromosome 22q12 that each surpassed genome-wide significance (P values for association between 1.59 × 10−26 and 4.84 × 10−9 and minor allele odds ratios of 0.63–0.80). However, these five loci explain 4–7% of the disease variance and up to a tenfold variation in interindividual risk. IgA nephropathy risk allele frequencies closely parallel the variation in disease prevalence among Asian, European, and African populations, suggesting complex selective pressures. In a GWAS replication study and geospatial risk analysis (Kiryluk et al., 2012), a seven-single nucleotide polymorphism genetic risk score, which explained 4.7% of overall IgA nephropathy

Table 66.2  Genes in which variants have been reportedly associated with susceptibility to IgA nephropathy Immune system genes PIGR (polymeric immunoglobulin receptor) IGHMBP2 (immunoglobulin mu binding protein 2) TRAC (T cell receptor alpha constant; T-cell receptor constant alpha chain) FCAR (Fc fragment of IgA, receptor for; CD89; FcalphaR, Fcalpha receptor) HLA-DRA (major histocompatibility complex, class II, DR α) FCGR3B (Fc fragment of IgG, low affinity IIIb, receptor; CD16b; FcgRIIIb) FCGR2B (Fc region of IgG) FCRLB (Fc receptor-like protein expressed on B linkage cells) Cytokine coding genes TNF-α (tumour necrosis factor alpha) IFNG (interferon gamma) TGF-β1 (transforming growth factor, beta1; TGF-beta1) IL10 (interleukin 10) IL5RA (interleukin 5 receptor, alpha; IL5RA) TNFRSF6B (tumour necrosis factor receptor superfamily, member 6b, decoy) Adhesion molecule genes SELE (selectin E; endothelial adhesion molecule 1) SELL (selectin L; lymphocyte adhesion molecule 1) Renin–angiotensin system genes ACE (angiotensin I converting enzyme 1) Glycosylation-related gene ST6GALNAC2 (ST6 alpha-N-acetyl-neuraminyl-2,3-beta-galactosyl-1,3-N-acety lgalactosaminide alpha-2, 6-sialyltransferase 2) C1GALT1 (core 1 synthase, glycoprotein-N-acetylgalactosamine 3-beta-galactosyltransferase, 1) Others EDN1 (endothelin 1; ET-1) NPHS2 (nephrosis 2, idiopathic, steroid-resistant; podocin) GNB3 (G protein beta polypeptide 3; G protein beta 3 subunit) SERPINB7 (serpin peptidase inhibitor, clade B, member 7; MEGSIN) SCGB1A1 (secretoglobin, family 1A, member 1; uteroglobin; UG) PAI-1 (plasminogen activator inhibitor-1) BMP2 (bone morphogenetic protein 2) Adapted and modified from Maxwell, P. H. and Wang, Y. (2009). Genetic contribution to IgA nephropathy. In K.N. Lai (ed.) Recent Advances in IgA nephropathy, pp 21–36. Copyright © World Scientific Press.

risk, increased sharply with Eastward and Northward distance from Africa (r = 0.30, P = 3 × 10−128). This model paralleled the known East–West gradient in disease burden. In another GWAS of 1434 Han Chinese, the genes encoding tumour necrosis factor (TNFSF13) and α-defensin (DEFA) were identified as susceptibility genes, and the association at 22q12 was confirmed (Yu et al., 2011).

Susceptibility or disease progression genes in non-familial IgA nephropathy Numerous association studies have reported different frequencies of genetic variants but none of these candidate genes is unique to IgA nephropathy (Tables 66.2 and 66.3). In an analysis of 123 candidate-gene association studies, Kiryluk et  al. (2010) found 31%, 32%, and 35% were associated with disease susceptibility, disease progression, or both respectively. One-third of the studies involved polymorphism of the renin–angiotensin system. Cohorts

Table 66.3  Genes in which variants have been reportedly associated with progression of IgA nephropathy Immune system genes FCGR3B (Fc fragment of IgG, low affinity IIIb, receptor; CD16b; FcgRIIIb) FCGR2A (Fc fragment of IgG, low affinity IIa, receptor; CD32; FcgammaRIIa) CD14 (CD14 molecule) Cytokine coding genes TNF-α (tumour necrosis factor alpha) IL10 (interleukin-10) IL4 (interleukin-4) TGF-β1 (transforming growth factor beta 1; TGF-beta1) CCL2 (chemokine C-C motif ligand 2; Monocyte chemoattractant protein -1; MCP-1) CCR5 (chemokine C-C motif receptor 5; CC-chemokine receptor five; chemokine receptor 5) Adhesion molecules SELE (selectin E; endothelial adhesion molecule 1, E-selectin) SELL (selectin L, lymphocyte adhesion molecule 1, L-selectin) Renin–angiotensin system genes ACE (angiotensin I converting enzyme 1) AGT (angiotensinogen; serpin peptidase inhibitor, clade A, member 8) Others SCGB1A1 (secretoglobin, family 1A, member 1; uteroglobin; Clara cell secretory protein; CC16) PON1 (paraoxonase 1) NPHS1 (nephrosis 1, congenital, Finnish type; Nephrin) NPHS2 (nephrosis 2, idiopathic, steroid-resistant; Podocin) VEGFA (Vascular endothelial growth factor A; VEGF) SERPINE1 (Serpin pepdase inhibitor, clade E, member 1; Plasminogen activator inhibitor-1; PAI-1) SERPINB7 (Serpin peptidase inhibitor, clade B, member 7; MEGSIN) PPARG (Peroxisome proliferator-activated receptor gamma) ACSM3 (Acyl-CoA synthetase medium-chain family member 3; SA) MUC20 (Mucin 20, cell surface associated) KLOTHO Adapted and modified from Genetic contribution to IgA nephropathy. Maxwell, P. H. and Wang, Y. (2009). In K.N. Lai (ed.) Recent Advances in IgA nephropathy, pp 21–36. Copyright © World Scientific Press.

Chapter 66 

Table 66.4  Age and presenting clinical features on the onset of 356 subjects with primary IgA nephropathy (IGAN-STET-CO: prospective cohort diagnosed from 1990 to 1999) Male (N = 255)

Female (N = 101)

No. (%)

No. (%)

Mean (SD) years

35.2 (15.4)

36.5 (15.2)

Median (range)

34.4 (5.1–76.6)

36.2 (2.7–71.6)

Isolated microhaematuria

54 (21.2 %)

40 (39.6 %)

Macrohaematuria ± microhaematuria

24 (9.4 %)

11 (10.9 %)

Isolated proteinuria

34 (13.3 %)

9 (8.9 %)

Proteinuria ± (microhaematuria/ macrohaematuria)

73 (28.6 %)

20 (19.8 %)

Hypertension ± (proteinuria/ microhaematuria/ macrohaematuria)

40 (15.7 %)

12 (11.9 %)

Renal failure (acute/chronic) ± (proteinuria/ microhaematuria/ macrohaematuria)

30 (11.8 %)

9 (8.9 %)

Age at onset

Modalities at the onset

Adapted from Clinical course of IgA nephropathy. Berthoux, F. C. and Mohey, H. (2009). In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 107–19. Copyright © 2009 World Scientific Press.

immunoglobulin a nephropathy: clinical features pharyngitic episode in contrast to 1–3 weeks in postinfectious glomerulonephritis. The macrohaematuria is sometimes accompanied by flank and loin pain. The urine colour is red or brown, but never contains clots. Asymptomatic microscopic haematuria is a more common presentation than macrohaematuria especially in Asian population often detected with health screening. Urine microscopy reveals dysmorphic red blood cells and red cell casts. Despite the early controversy between the prognostic value of microhaematuria versus macrohaematuria, multivariate Cox analysis fails to demonstrate any difference (Coppo and D’Amico, 2005).

Proteinuria Asymptomatic proteinuria may be found in 9–13% of cases (Table 66.5). The degree of proteinuria tends to fluctuate within a narrow range for most patients. Proteinuria is usually not heavy and < 30% have proteinuria exceeding 1 g/day. A transient increase of

Table 66.5  Characteristics at diagnosis (first renal biopsy) in 356 subjects with primary IgA nephropathy (IGAN-STET-CO: prospective cohort diagnosed from 1990 to 1999) Male (N = 255)

Female (N = 101)

No. (%)

No. (%)

98 (38.4 %)

28 (27.7 %)

Macrohaematuria

57 (22.4%)

21 (20.8 %)

Microhaematuria

175 (68.6 %)

79 (78.2 %)

Mean (SD)

1.08 (1.50)

0.57 (1.04)

Median (range)

0.50 (0–9.04)

0.08 (0–7.26)

0–0.30 g/day

92 (36.1 %)

60 (59.4 %)

0.31–0.99 g/day

77 (30.2 %)

24 (23.8 %)

1.00–2.99 g/day

58 (22.7 %)

13 (12.9 %)

> 3 g/day

28 (11.0 %)

4 (4.0 %)

Mean (SD)

76.3 (28.9)

74.6 (25.2)

Median (range)

81.6 (5.8–157)

76.7 (5.6–16.8)

CKD stage 1

89 (34.9 %)

27 (26.7 %)

CKD stage 2

103 (40.4 %)

52 (51.5 %)

CKD stage 3

40 (15.7 %)

15 (14.9 %)

CKD stage 4

12 (4.7 %)

6 (5.9 %)

CKD stage 5

11 (4.3 %)

1 (1.0 %)

End-stage renal failure/dialysis

4 (1.6 %)

0 (0 %)

Hypertension Yes (independent of treatment) Haematuria

in these studies were underpowered with inadequate cover of single nucleotide polymorphisms. It is therefore unclear which of the reported genes truly confers susceptibility.

Clinical manifestations Clarkson et  al. (1977) remarked that ‘Berger disease’ was a syndrome of uniform morphology, diverse clinical features, and uncertain prognosis. It is now recognized that primary IgA nephropathy has diverse clinical presentations (Table 66.4) but it still has uncertain prognosis. Among Caucasians, there is a male predominance. In a French analysis of 356 patients, synpharyngitic macrohaematuria, isolated microhaematuria, and arterial hypertension occur in 20%, 26%, and 22% of patients, respectively (Berthoux and Mohey, 2009). In East Asian subjects, the male-to-female ratio is close to unity (Lai et al., 1985; Yoshikawa et al., 1994). The clinical presentation, however, is not different between the two genders.

Haematuria The first episode of macroscopic haematuria generally occurs between 15 and 30  years of age—often a decade earlier than a biopsy diagnosis is made. The true onset of the glomerulopathy is likely in the teens or even earlier as inflammatory processes will take time to develop following mesangial IgA deposition. Not infrequently, patients may first present with macroscopic haematuria complicating mucosal infection (respiratory or gastrointestinal), and the former is often described as ‘synpharyngitic macrohaematuria’. Macrohaematuria occurs shortly (12–72 hours) following the

Proteinuria (g/day)

GFR (MDRD): mL/min/1.73 m2

Adapted from Clinical course of IgA nephropathy. Berthoux, F. C. and Mohey, H. (2009). In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 107–19. Copyright © 2009 World Scientific Press. CKD = chronic kidney disease; GFR = glomerular filtration rate; MDRD = Modification of Diet in Renal Disease Study Group.

569

570

Section 3  

the patient with glomerular disease

proteinuria occurs with gross haematuria complicating mucosal infection or urinary tract infection. In a small group of nephrotic patients, the renal biopsy shows histological and ultrastructural features of minimal change nephropathy but with mesangial IgA deposits. These are now referred to as ‘an overlapping syndrome of IgA nephropathy and lipoid nephrosis’ (Lai et  al., 1986). This entity occurs more frequently in children. Clinically, this entity behaves like minimal change nephropathy with good response to corticosteroids. In contrast, for nephrotic patients with advanced glomerular pathology, response to corticosteroid is poor with high probability of renal deterioration.

Other presentations Acute kidney injury is an uncommon presentation and the pathology is frequently associated with extensive crescentic formation (Lai et al., 1987). Occasionally, acute kidney injury may be a complication from bouts of macroscopic haematuria in IgA nephropathy. Biopsy reveals prominent tubular red blood cell casts, acute tubular necrosis, and interstitial extravasation of red cells (Gutiérrez et al., 2007). The abundance of interstitial macrophages expressing the haemoglobin scavenger receptor CD163 and oxidative stress markers suggests a pathogenetic role for free haemoglobin-induced tubulointerstitial renal injury (Martín-Cleary et al., 2010). Malignant hypertension in the setting of IgA nephropathy may lead to rapidly progressive renal failure. In advanced disease, it may be associated with thrombotic microangiopathy as a renovascular complication of IgA nephropathy. Renal prognosis is as poor as in patients with primary malignant hypertension (Chen et al., 2005). Fifteen per cent of patients have significant renal impairment (chronic kidney disease stage 3 or higher) at first presentation (Table 66.5). Occasionally, IgA nephropathy may present as end-stage renal failure requiring dialytic treatment at presentation, more commonly seen in population with under-privileged healthcare. Fifteen per cent of IgA nephropathy is diagnosed in patients undergoing investigation for hypertension. A renal cause is suggested by abnormal urinalysis, elevated serum creatinine, or raised serum IgA level.

References Bahiense-Oliveira, M., Saldanha, L. B., Mota, E. L., et al. (2004). Primary glomerular diseases in Brazil (1979-1999): is the frequency of focal and segmental glomerulosclerosis increasing? Clin Nephrol, 61, 90–7. Ballardie, F. W., O’Donoghue, D. J., and Feehally, J. (1987). Increasing frequency of adult IgA nephropathy in the UK? Lancet, 2, 1205. Batinic, D., Scukanec-Spoljar, M., Milosevic, D., et al. (2007). [Clinical and histopathological characteristics of biopsy-proven renal diseases in Croatia]. Acta Med Croatica, 61, 361–4. Berger, J. and Hinglais, N. (1968). [Intercapillary deposits of IgA-IgG]. J Urol Nephrol, 74, 694–5. Berthoux, F. C. and Mohey, H. (2009). Clinical course of IgA nephropathy. In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 107–19. Singapore: World Scientific Press. Bisceglia, L., Cerullo, G., Forabosco, P., et al. (2006). Genetic heterogeneity in Italian families with IgA nephropathy: suggestive linkage for two novel IgA nephropathy loci. Am J Hum Genet, 79, 1130–4. Briganti, E. M., Dowling, J., Finlay, M., et al. (2001). The incidence of biopsy-proven glomerulonephritis in Australia. Nephrol Dial Transplant, 16, 1364–7. Carvalho, E., do Sameiro, F. M., Nunes, J.P., et al. (2006). Renal diseases: a 27-year renal biopsy study. J Nephrol, 19, 500–7. Chandrika, B. K. (2007). Non-neoplastic renal diseases in Kerala, India—analysis of 1592 cases, a two year retrospective study. Indian J Pathol Microbiol, 50, 300–2.

Chen, Y., Tang, Z., Yang, G., et al. (2005). Malignant hypertension in patients with idiopathic IgA nephropathy. Kidney Blood Press Res, 28, 251–8. Choi, I. J., Jeong, H. J., Han, D. S., et al. (2001). An analysis of 4,514 cases of renal biopsy in Korea. Yonsei Med J, 42, 247–54. Clarkson, A. R., Seymour, A. E., Thompson, A. J., et al. (1977). IgA nephropathy: a syndrome of uniform morphology, diverse clinical features and uncertain prognosis. Clin Nephrol, 8, 459–71. Coppo, R. and D’Amico, G. (2005). Factors predicting progression of IgA nephropathies. J Nephrol, 18, 503–12. Covic, A., Schiller, A., Volovat, C., et al. (2006). Epidemiology of renal disease in Romania: a 10 year review of two regional renal biopsy databases. Nephrol Dial Transplant, 21, 419–24. Feehally, J., Farrall, M., Boland, A., et al. (2010). HLA has strongest association with IgA nephropathy in genome-wide analysis. J Amer Soc Nephrol, 21, 1791–7. Frasca, G. M., Soverini, L., Gharavi, A. G., et al. (2004). Thin basement membrane disease in patients with familial IgA nephropathy. J Nephrol, 17, 778–85. Gharavi, A. G., Kiryluk, K., Choi, M., et al. (2011). Genome-wide association study identifies susceptibility loci for IgA nephropathy. Nat Genet, 43, 321–7. Gharavi, A. G., Yan, Y., Scolari, F., et al. (2000). IgA nephropathy, the most common cause of glomerulonephritis, is linked to 6q22–23. Nat Genet, 26, 354–7. Gutiérrez, E., Gonzalez, E., Hernandez, E., et al. (2007). Factors that determine an incomplete recovery of renal function in macrohematuria-induced acute renal failure of IgA nephropathy. Clin J Am Soc Nephrol, 2, 51–7. Hurtado, A., Escudero, E., Stromquist, C. S., et al. (2000). Distinct patterns of glomerular disease in Lima, Peru. Clin Nephrol, 53, 325–32. Izzi, C., Sanna-Cherchi, S., Prati, E., et al. (2006). Familial aggregation of primary glomerulonephritis in an Italian population isolate: Valtrompia study. Kidney Int, 69, 1033–40. Johnston, P. A., Brown, J. S., Braumholtz, D. A., et al. (1992). Clinico-pathological correlations and long-term follow-up of 253 United Kingdom patients with IgA nephropathy. A report from the MRC Glomerulonephritis Registry. QJM, 84, 619–27. Julian, B. A., Quiggins, P. A., Thompson, J. S., et al. (1985). Familial IgA nephropathy. Evidence of an inherited mechanism of disease. N Engl J Med, 312, 202–8. Karnib, H. H., Sanna-Cherchi, S., Zalloua, P. A., et al. (2007). Characterization of a large Lebanese family segregating IgA nephropathy. Nephrol Dial Transplant, 22, 772–7. Kiryluk, K., Julian, B. A., Wyatt, R. J. et al. (2010). Genetic studies of IgA nephropathy: past, present, and future. Pediatr Nephrol, 25, 2257–68. Kiryluk, K., Li, Y., Sanna-Cherchi, S., et al. (2012). Geographic differences in genetic susceptibility to IgA nephropathy: GWAS replication study and geospatial risk analysis. PLoS Genet, 8, e1002765. Lai, K. N., Chan, L. Y., Tang, S. C., et al. (2004). Mesangial expression of angiotensin II receptor in IgA nephropathy and its regulation by polymeric IgA1. Kidney Int, 66, 1403–16. Lai, K. N., Ho, C. P., Chan, K. W., et al. (1985). Nephrotic range proteinuria—a good predictive index of disease in IgA nephropathy? QJM, 57, 677–88. Lai, K. N., Ho, R. T., Lai, C. K., et al. (1994). Increase of both circulating Th1 and Th2 T lymphocyte subsets in IgA nephropathy. Clin Exp Immunol, 96, 116–21. Lai, K. N., Lai, F. M., Chan, K. W., et al. (1986). An overlapping syndrome of IgA nephropathy and lipoid nephrosis. Am J Clin Pathol, 86, 716–23. Lai, K. N., Lai, F. M., Leung, A. C., et al. (1987). Plasma exchange in patients with rapidly progressive idiopathic IgA nephropathy: a report of two cases and review of literature. Am J Kidney Dis, 10, 66–70. Lavigne, K. A., Woodford, S. Y., Barker, C. V., et al. (2010). Familial IgA nephropathy in southeastern Kentucky. Clin Nephrol, 73, 115–21. Levy, M. (1989). Familial cases of Berger’s disease and anaphylactoid purpura: more frequent than previously thought. Am J Med, 87, 246–8.

Chapter 66 

Levy, M., Beaufils, H., Gubler, M.C., et al. (1972). Idiopathic recurrent macroscopic hematuria and mesangial IgA-IgG deposits in children (Berger’s disease). Clin Nephrol, 1, 63–9. Levy, M. and Berger, J. (1988). Worldwide perspective of IgA nephropathy. Am J Kidney Dis, 12, 340–7. Martín-Cleary, C., Moreno, J. A., Fernandez, B., et al. (2010). Glomerular haematuria, renal interstitial haemorrhage and acute kidney injury. Nephrol Dial Transplant, 25, 4103–6. Parichatikanond, P., Chawanasuntorapoj, R., Shayakul, C., et al. (2006). An analysis of 3,555 cases of renal biopsy in Thailand. J Med Assoc Thai, 89 Suppl 2, S106–11. Paterson, A. D., Liu, X. Q., Wang, K., et al. (2007). Genome-wide linkage scan of a large family with IgA nephropathy localizes a novel susceptibility locus to chromosome 2q36. J Am Soc Nephrol, 18, 2408–15. Rambausek, M., Hartz, G., Waldherr, R., et al. (1987). Familial glomerulonephritis. Pediatr Nephrol, 1, 416–18. Research Group on Progressive Chronic Renal Disease (1999). Nationwide and long-term survey of primary glomerulonephritis in Japan as observed in 1,850 biopsied cases. Nephron, 82, 205–13. Rivera, F., Lopez-Gomez, J. M., and Perez-Garcia, R. (2002). Frequency of renal pathology in Spain 1994–1999. Nephrol Dial Transplant, 17, 1594–602. Schena, F. P. (1997). Survey of the Italian Registry of Renal Biopsies. Frequency of the renal diseases for 7 consecutive years. The Italian Group of Renal Immunopathology. Nephrol Dial Transplant, 12, 418–26. Schena, F. P., Cerullo, G., Rossini, M., et al. (2002). Increased risk of end-stage renal disease in familial IgA nephropathy. J Am Soc Nephrol, 13, 453–60. Schena, F. P., Scivittaro, V., Di Cillo, M., et al. (1990). Is Berger’s disease a hereditary nephritis? Contrib Nephrol, 80, 118–25.

immunoglobulin a nephropathy: clinical features Schena, F. P., Scivittaro, V., and Ranieri, E. (1993). IgA nephropathy: pros and cons for a familial disease. Contrib Nephrol, 104, 36–45. Scolari, F., Amoroso, A., Savoldi, S., et al. (1999). Familial clustering of IgA nephropathy: further evidence in an Italian population. Am J Kidney Dis, 33, 857–65. Simon, P., Ramee, M. P., Boulahrouz, R. et al. (2004). Epidemiologic data of primary glomerular diseases in western France. Kidney Int, 66, 905–8. Sinniah, R. (1983). Occurrence of mesangial IgA and IgM deposits in a control necropsy population. J Clin Pathol, 36, 276–9. Swaminathan, S., Leung, N., Lager, D. J., et al. (2006). Changing incidence of glomerular disease in Olmsted County, Minnesota: a 30-year renal biopsy study. Clin J Am Soc Nephrol, 1, 483–7. Tam, K. Y., Leung, J. C., Chan, L. Y., et al. (2009). Macromolecular IgA1 taken from patients with familial IgA nephropathy or their asymptomatic relatives have higher reactivity to mesangial cells in vitro. Kidney Int, 75, 1330–9. Tam, K. Y., Leung, J. C., Chan, L. Y., et al. (2010). In vitro enhanced chemotaxis of CD25+ mononuclear cells in patients with familial IgAN through glomerulotubular interactions. Am J Physiol Renal Physiol, 299, F359–68. Tiebosch, A. T., Wolters, J., Frederik, P. F., et al. (1987). Epidemiology of idiopathic glomerular disease: a prospective study. Kidney Int, 32, 112–6. Werner, T., Brodersen, H. P., and Janssen, U. (2009). [Analysis of the spectrum of nephropathies over 24 years in a West German center based on native kidney biopsies]. Med Klin (Munich), 104, 753–9. Yoshikawa, N., Nakamura, H., and Ito, H. (1994). IgA nephropathy in children and adults. Springer Semin Immunopathol, 16, 105–20. Yu, X. Q., Li, M., Zhang, H., et al. (2011).A genome-wide association study in Han Chinese identifies multiple susceptibility loci for IgA nephropathy. Nat Genet, 44,178–82.

571

CHAPTER 67

Immunoglobulin A nephropathy: diagnosis Kar Neng Lai and Sydney C. W. Tang Urinalysis The presence of red cell casts and dysmorphic red cells (see Chapter 6), though less often undertaken by many nephrologists now, indicates glomerular bleeding. This should spare the patient from unnecessary urologic procedures such as cystoscopy or retrograde pyelography (see Chapter 46). Other slowly evolving glomerular disorders that are particularly likely to present in a similar way include hereditary nephritis (see Chapter 321) and thin basement membrane disease (see Chapter 325). Varying degrees of proteinuria may also be present.

Renal biopsy The diagnosis cannot be made with certainty without a renal biopsy, and this can offer additional prognostic information.

Light microscopy A typical biopsy of immunoglobulin A (IgA) nephropathy is characterized by an increase in mesangial cells with matrix expansion and normal glomerular basement membranes (GBMs) (Fig. 67.1B). However, light microscopic changes are highly variable, ranging from normal or minimal lesions in the glomerular architecture, diffuse mesangial proliferative changes to focal segmental glomerulosclerosis (FSGS), and rarely focal segmental necrotizing lesions with crescent formation. It is likely that the presence of focal segmental or global sclerosis represents the disease that has already been ongoing for some

time. Indeed, capsular adhesions without underlying abnormalities in the tuft, often the first sign of FSGS, are also frequent in IgA nephropathy. A  recent French study showed that some form of lesion resembling FSGS, including classical, prehilar, cellular, tip and collapsing variants, was present in 101 out of 128 biopsies IgA nephropathy (El Karoui et al., 2011). It was proposed that most cases of IgA nephropathy can therefore be interpreted as representing a variant of FSGS, highlighting the notion that podocytopathy plays a role in the pathogenesis and progression of IgA nephropathy (see Chapter 139). The presence of aggressive crescentic lesions may represent a different category of disease and may associate with the presence of antineutrophil cytoplasmic autoantibodies (ANCAs). In the overlapping syndrome of IgA nephropathy and lipoid nephrosis, the light microscopy is normal with mesangial IgA deposits. Apart from changes in the glomerulus, the tubulointerstitium and periglomerular arterioles may also display pathological changes, though these are often non-specific for IgA nephropathy and signify a final common pathway of renal damage in a variety of glomerulopathies in the advanced stage. Variable degrees of tubular atrophy/interstitial fibrosis can occur during the different stages of IgA nephropathy, and its severity carries prognostic implications (Chan et al., 2004; Coppo and D’Amico, 2005). In addition, excessive inflammatory cells may populate the cortical interstitium leading to interstitial inflammation, another feature of the chronic kidney irrespective of the original disease. The presence of granule membrane protein of 17 kDa (GMP-17)-positive cytotoxic T-lymphocytes in intact renal tubules serves as a marker of disease progression in early

Fig. 67.1  (A) Immunofluorescent staining for IgA deposits in the mesangium (×500). (B) Moderate mesangial matrix expansion with increased cell number (H&E ×400). (C) Electron micrograph showing mesangial expansion with electron-dense deposits (arrows) (×9600).

Chapter 67 

stage (Van Es et al., 2008). Arteriolar lesions are characterized by wall hyalinosis, intimal thickening, and subintimal fibrosis. These changes often accompany the presence of hypertension.

immunoglobulin a nephropathy: diagnosis

Table 67.1  Systemic diseases associated with predominant mesangial IgA deposition with probable pathogenetic association with secondary IgA nephropathy

Immune deposits

Diseases

Common

Rare

The defining histological hallmark is the presence of dominant or co-dominant deposition of IgA in the glomerular mesangium (Fig. 67.1A). Mesangial IgA deposits are predominantly polymeric in nature with λ-light chain (Lai et al., 1988). These consist mostly of underglycosylated polymeric IgA1 with the absence of the secretory component and the presence of the J chain. Apart from primary IgA nephropathy, a variety of systemic conditions may also be associated with secondary IgA nephropathy (Table 67.1). IgG, IgM, and C3 may co-distribute with IgA. C3 is detectable in up to 70–90% of cases often with same distribution as IgA. IgG is present in about 50–70%, and often assumes co-dominance with the IgA staining. IgM deposits are less common and are found in 31–66%. The early components of complement, such as C1q and C4, are infrequently present. Occasionally, C5b-9 with properdin is found, indicating activation of the alternative pathway. Fibrin/fibrinogen deposits have been reported in 30–40% of biopsies, mainly locate in the mesangium and the glomerular endothelium. Von Willebrand factor may be present in the endothelium, and together with platelet aggregates activates intraglomerular coagulation and contributes to glomerular sclerosis. The extent of fibrin deposition correlates with mesangial proliferation, indicating that the coagulation/fibrinolysis system participates in renal damage of IgA nephropathy.

Gastrointestinal

Coeliac disease

Crohn disease Ulcerative colitis

Hepatic

Alcoholic liver disease Non-alcoholic liver cirrhosis

Infections

HIV Hepatitis B Schistosomiasis Staphylococcal PIGN

Brucellosis Leprosy

Malignancies

Renal cell carcinoma

Monoclonal lymphoproliferative diseases Mixed cryoglobulinaemia Carcinoma (lung, larynx, pancreas) Mycosis fungoides

Rheumatic diseases

Ankylosing spondylitis Rheumatoid arthritis

Autoimmune diseases

Systemic lupus erythematosus Wegener’s granulomatosis Sjögren disease Hashimoto’s thyroiditis

Electron microscopy Typically, there are electron-dense deposits that are confined to mesangial or paramesangial areas (Fig.  67.1C). Occasionally, they may be found in the subepithelial or subendothelial space in 10–20% of cases, but are generally small and segmental in nature (Woodrow et al., 1989). The GBM is usually of normal thickness, but may display focal thinning in up to a third of patients. In the overlapping IgA nephropathy/lipoid nephrosis variant, there is extensive effacement of the foot processes.

Pathology-based classification and grading for IgA nephropathy There have been numerous attempts by pathologists to develop and refine different grading systems based on the extent of pathological lesions to predict clinical outcome (Table 67.2). A common feature of all these systems is that they are all based solely upon light microscopic findings of the renal biopsy which are the most highly variable pathological footprint in IgA nephropathy. These classifications, mostly favouring injury in glomeruli over that in the interstitium and vasculature, have individual strengths and limitations in predicting prognosis. Conflicting results may arise from differences in patient selection, cohort size, treatment, and clinical outcome measures. In general, studies that employ end-stage renal disease (ESRD) as an endpoint have shown chronic pathologic lesions, such as tubular atrophy, and glomerulosclerosis, to predict outcome. In contrast, studies that use rate of GFR decline or responsiveness to immunosuppressive treatment have shown active glomerular lesions, such as mesangial, endocapillary, or extracapillary hypercellularity, to have predictive power.

Dermatological disorders

Dermatitis herpetiformis

Psoriasis

Others

Diabetes mellitus/ metabolic syndrome

Idiopathic pulmonary haemosiderosis, sarcoidosis

PIGN: post-infectious glomerulonephritis

None of these pathological grading systems has gained widespread acceptance and the debates continue as to whether pathological findings may provide additional prognostic indicators beyond those from clinical features (Bartosik et al., 2001). In 2004, the International IgA Nephropathy Network () and the International Society of Nephrology/ Renal Pathology Society formed a working group to devise an international consensus on a new classification for IgA nephropathy. Renal biopsy materials from 206 adult and 59 paediatric patients were assessed objectively. The report is duly called ‘The Oxford Classification of IgA Nephropathy’ (Cattran et  al., 2009). After extensive iterative work by five pathologists confined to histological features exclusively on periodic acid–Schiff (PAS)-stained sections on light microscopy, the working group concluded that four glomerular and parenchymal parameters possess reproducible and independent predictive value on renal outcome: mesangial hypercellularity, segmental glomerulosclerosis, endocapillary proliferation, and tubular atrophy/interstitial fibrosis. A unique feature of this system is that it recommends the individual reporting on these four features without artificially clustering into different ‘grading’ in pathology reports. Thus, the nephrologist would interpret

573

574

Section 3  

the patient with glomerular disease

Table 67.2  Grading and classification systems for IgA nephrology from different eras Classification (author/year)

Grade 1

Grade 2

Grade 3

Grade 4

Grade 5

(Southwest Pediatric Nephrology Study Group, 1982)

Minimal glomerular changes

Mesangial proliferation only

Any focal or sclerotic lesion

(Haas, 1997)

Minimal or no mesangial Focal segmental hypercellularity; without glomerulosclerosis; sclerosis or crescents minimal increase in mesangial hypercellularity; no crescents

Focal proliferative: changes, < 50% of glomeruli are hypercellular

Diffuse proliferative: > 50% of glomeruli are hypercellular

Advanced sclerotic changes: ≥ 40% of glomeruli are globally sclerotic and/or ≥ 40% tubular atrophy or loss of cortex

(To et al., 2000)

Mean glomerular sclerosis  75% of glomeruli with proliferation, or < 25% of crescents, segmental or with crescents, segmental crescents, segmental or glomeruli with crescents, global sclerosis or global sclerosis global sclerosis segmental /global sclerosis

(Wakai et al., 2006)

Slight mesangial cell proliferation and increased matrix

Slight mesangial cell proliferation and increased matrix. Glomerulosclerosis, crescent formation or adhesion to Bowman’s capsule in < 10% of glomeruli

Moderate diffuse mesangial cell proliferation and increased matrix. Glomerulosclerosis, crescent formation or adhesion to Bowman’s capsule in 10–30% of glomeruli

(Manno et al., 2007)

Normal glomeruli or slight increase in mesangial matrix and/or cellularity

Moderate or diffuse mesangial proliferation and/or focal segmental sclerosis and/or endocapillary proliferative and/or cellular crescents ≤ 0% of glomeruli

Cellular crescents in > 50% of glomeruli and/ or global sclerosis and fibrous crescents involving > 1/3 of glomeruli and/or diffuse segmental sclerosis

Oxford (Roberts et al., 2009)

Forgoes arbitrary classification into different grades Emphasizes on the following four reproducible histologic elements, each independently predictive of renal outcome: (1) mesangial hypercellularity (M0 or M1)a (2) segmental glomerulosclerosis (S0 or S1)b (3) endocapillary proliferation (E0 or E1)c (4) tubular atrophy/interstitial fibrosis (T0, T1 or T2)d

Severe diffuse mesangial cell proliferation and increased matrix. Glomerulosclerosis, crescent formation or adhesion to Bowman’s capsule in > 30% of glomeruli

a Mesangial score should be assessed on PAS-stained sections; M0 ≤ 50% of the glomeruli showing hypercellularity (> 3 cells in a mesangial area), M1 > 50% of the glomeruli having hypercellularity. b Any amount of sclerosis in the glomerular tuft; S0 = absent, S1 = present. c Hypercellularity within glomerular capillary lumina causing its narrowing; E0 = absent, E1 = present. d Percentage cortical area involved; T0 = 0–25%, T1 = 26–50%, T2 > 50%.

Adapted and modified with permission from Clinicopathologic findings. Julian, B. A. and Wyatt, R. J. (2009). In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 83–106. Copyright © 2009 World Scientific Press.

these four features separately on top of the salient clinical features known to impact upon prognosis. It must be borne in mind that this classification suffers the same drawback as being a retrospective observational review, and the histopathological specimens and biochemical data were not uniformly collected in a standardized fashion at source. Nevertheless, a number of validation studies have

supported this classification. A  French study (Alamartine et  al., 2011) supported the validity of the Oxford classification in predicting outcome and also showed that categorizing the type of FSGS lesion carries prognostic value. Similar data have also been generated from Chinese (Shi et al., 2011) and Japanese (Shima et al., 2012) patients.

Chapter 67 

Other investigations Renal function Patients with estimated GFR (eGFR) < 60 mL/min/1.73 m2 at the time of renal biopsy have worse outcomes than those with normal eGFR. However, the overall rate of decline of eGFR is difficult to predict as some patients with mild-to-moderate renal impairment may remain stable for years, particularly when treated with renin–angiotensin blockade. In the Oxford classification of IgA nephropathy study cohort (Cattran et al., 2009), the rate of GFR decline correlated with segmental glomerulosclerosis and tubular atrophy/interstitial fibrosis, whereas mesangial hypercellularity and tubular atrophy/interstitial fibrosis are predictive of the composite endpoints of 50% reduction in eGFR or ESRD.

Serum IgA Serum IgA levels have been much studied but are not diagnostically very helpful. Levels are elevated in 33–50% of patients, but this is neither sensitive nor specific for IgA nephropathy. Serum IgA level also bears no prognostic value and the evolution of serum IgA during the course of the disease remains undefined. IgA is unique in its ability to form multimers. Elevated circulating levels of polymeric IgA occur in 25% of patients. The predominant subclass is IgA1, although IgA2 is also elevated in African American patients with IgA nephropathy. This finding is consistent with the mesangial deposition of IgA2 in 56% of these patients. Galactose-deficient IgA1 is implicated in pathogenesis (see Chapter 69). Serum levels determined by a lectin-based enzyme-linked immunosorbent assay were reported to have 90% specificity and 76% sensitivity for diagnosing sporadic IgA nephropathy in a large cohort of Caucasians from the United States (Moldoveanu et  al., 2007). Interestingly, a high serum galactose-deficient IgA1 level was present in the majority of index cases in this cohort, as well as among their parents (39%), siblings (28%), and children (30%). Levels in spouses were indistinguishable from controls, ruling out an environmental effect (Gharavi et al., 2008). Segregation analysis of galactose-deficient IgA1 suggested inheritance of a major dominant gene with an additional polygenic component. Inheritance of galactose-deficient IgA1 has been confirmed in Chinese patients with familial and sporadic adult IgAN (Lin et al., 2009; Tam et al., 2009). These data demonstrate that an elevated serum galactose-deficient IgA1 level is antecedent to disease but, because most family members with elevated levels are asymptomatic, IgA1 glycosylation abnormalities are not sufficient to produce IgA nephropathy, and additional cofactors must trigger formation of immune complexes.

Complement Serum complement levels are generally normal and their main value is in excluding other conditions. IgA deposits are commonly associated with the deposition of complement components, most notably C3, the membrane attack complex (C5b-9), and properdin (Couser et al., 1985; Rauterberg et al., 1987; Schena, 1992), suggesting the participation of complement activation in disease pathogenesis. Two Japanese groups (Komatsu et  al., 2004; Nakayama et  al., 2008)  reported that the serum IgA/C3 ratio may reflect the histological severity of IgA nephropathy and could serve as a marker for disease progression, and may be useful for prediction of diagnosis of IgA nephropathy and distinguishing it from other

immunoglobulin a nephropathy: diagnosis

glomerulonephritides. Another German study (Zwirner et  al., 1997)  indicated that only alternative pathway complement was activated in IgA nephropathy and its activation was associated with more severe renal disease. Serum complement levels of C3, C4, and factor B are usually normal, although small amounts of neoantigens of C3, such as iC3b and C3b, can be detected using sensitive methods. Increased activated plasma C3 levels are present in 30% of patients, particularly in those with proteinuria and haematuria, and these correlated with renal deterioration on follow-up. Deficiencies in certain complement components, such as factor H, and C9, have been described in some subjects.

Immune complexes Circulating IgA immune complex (IgA-IC) formation is thought to be universal in IgA nephropathy (see Chapter 69), but again this is not a diagnostically useful test.

Antineutrophil cytoplasmic autoantibodies Circulating ANCAs are not usually found but are described in some case reports (Bantis et al., 2010). ANCAs against myeloperoxidase are more commonly encountered than those against proteinase 3. When ANCAs are present, the disease often takes a more aggressive form with active urine sediments, rapidly progressive glomerulonephritis, and renal biopsies often reveal necrotizing and crescentic lesions. Therapeutic response to timely immunosuppressive treatment is usually good.

Advanced urinalysis A number of techniques show promise as predictors of disease or disease activity, but are mainly research tools at present. Cytokines and chemokines in urine are elevated in IgA nephropathy, particularly in patients with moderately advanced renal injury. Urinary monocyte chemoattractant protein-1 (MCP-1) levels correlate with proteinuria (Wasilewska et  al., 2011). Together with urinary interleukin-6 and epidermal growth factor, these cytokines may act as predictors of renal outcome (Stangou et  al., 2009). Urinary interleukin-8 levels have also been correlated with disease activity. Urinary angiotensinogen is a powerful tool for determining intrarenal renin–angiotensin system activity and is associated renal derangement (Nishiyama et al., 2011). Urinary complement factor H is shown to correlate closely with disease activity (Zhang et al., 2009). In crescentic IgA nephropathy, fractional excretion of IgG in relation to the degree of nephron loss predicts disease progression (Bazzi et al., 2009). Urinary secretory IgA (sIgA) has been associated with high-grade histological changes and proteinuria, and might be used as a non-invasive biomarker to evaluate kidney injury in IgA nephropathy (Tan et al., 2009). Urinary podocytes or their fragments can be identified by immunohistochemical approaches, or by mRNA analysis, or by protein analysis. Podocyte loss may be a marker for progressive renal disease (see Chapter 139). Urine uromodulin fragment has been suggested as a biomarker for the non-invasive diagnosis of IgA nephropathy (Wu et  al., 2010). Several urinary microRNAs (miR-200a, miR-200b, and miR-429) are downregulated in patients with IgA nephropathy, and the degree of reduction suggested to correlate with disease severity and rate of progression (Wang et al., 2010).

575

576

Section 3  

the patient with glomerular disease

References Alamartine, E., Sauron, C., Laurent, B., et al. (2011). The use of the Oxford classification of IgA nephropathy to predict renal survival. Clin J Am Soc Nephrol, 6, 2384–88. Bantis, C., Stangou, M., Schlaugat, C., et al. (2010). Is presence of ANCA in crescentic IgA nephropathy a coincidence or novel clinical entity? A case series. Am J Kidney Dis, 55, 259–68. Bartosik, L. P., Lajoie, G., Sugar, L., et al. (2001). Predicting progression in IgA nephropathy. Am J Kidney Dis, 38, 728–35. Bazzi, C., Rizza, V., Raimondi, S., et al. (2009). In crescentic IgA nephropathy, fractional excretion of IgG in combination with nephron loss is the best predictor of progression and responsiveness to immunosuppression. Clin J Am Soc Nephrol, 4, 929–35. Cattran, D. C., Coppo, R., Cook, H. T., et al. (2009). The Oxford classification of IgA nephropathy: rationale, clinicopathological correlations, and classification. Kidney Int, 76, 534–45. Chan, L. Y., Leung, J. C., and Lai, K. N. (2004). Novel mechanisms of tubulointerstitial injury in IgA nephropathy: a new therapeutic paradigm in the prevention of progressive renal failure. Clin Exp Nephrol, 8, 297–303. Coppo, R., Amore, A., Chiesa, M., et al. (2007). Serological and genetic factors in early recurrence of IgA nephropathy after renal transplantation. Clin Transplant, 21, 728–37. Couser, W. G., Baker, P. J., and Adler, S. (1985). Complement and the direct mediation of immune glomerular injury: a new perspective. Kidney Int, 28, 879–90. El Karoui, K., Hill, G. S., Karras, A., et al. (2011). Focal segmental glomerulosclerosis plays a major role in the progression of IgA nephropathy. II. Light microscopic and clinical studies. Kidney Int, 79, 643–54. Gharavi, A. G., Moldoveanu, Z., Wyatt, R. J., et al. (2008). Aberrant IgA1 glycosylation is inherited in familial and sporadic IgA nephropathy. J Amer Soc Nephrol, 19, 1008–14. Haas, M. (1997). Histologic subclassification of IgA nephropathy: a clinicopathologic study of 244 cases. Am J Kidney Dis, 29, 829–42. Julian, B. A. and Wyatt, R. J. (2009). Clinicopathologic findings. In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 83–106. Singapore: World Scientific Press. Komatsu, H., Fujimoto, S., Hara, S., et al. (2004). Relationship between serum IgA/C3 ratio and progression of IgA nephropathy. Intern Med, 43, 1023–8. Lai, K. N., Chui, S. H., Lai, F. M., et al. (1988). Predominant synthesis of IgA with lambda light chain in IgA nephropathy. Kidney Int, 33, 584–9. Lee, H. S., Lee, M. S., Lee, S. M., et al. (2005). Histological grading of IgA nephropathy predicting renal outcome: revisiting H. S. Lee’s glomerular grading system. Nephrol Dial Transplant, 20, 342–8. Lin, X., Ding, J., Zhu, L., et al. (2009). Aberrant galactosylation of IgA1 is involved in the genetic susceptibility of Chinese patients with IgA nephropathy. Nephrol Dial Transplant, 24, 3372–5. Manno, C., Strippoli, G. F., D’Altri, C., et al. (2007). A novel simpler histological classification for renal survival in IgA nephropathy: a retrospective study. Am J Kidney Dis, 49, 763–75. Moldoveanu, Z., Wyatt, R. J., Lee, J. Y., et al. (2007). Patients with IgA nephropathy have increased serum galactose-deficient IgA1 levels. Kidney Int, 71, 1148–54. Nakayama, K., Ohsawa, I., Maeda-Ohtani, A., et al. (2008). Prediction of diagnosis of immunoglobulin A nephropathy prior to renal biopsy and correlation with urinary sediment findings and prognostic grading. J Clin Lab Anal, 22, 114–18. Nishiyama, A., Konishi, Y., Ohashi, N., et al. (2011). Urinary angiotensinogen reflects the activity of intrarenal renin-angiotensin system in patients with IgA nephropathy. Nephrol Dial Transplant, 26, 170–7.

Rauterberg, E. W., Lieberknecht, H. M., Wingen, A. M., et al. (1987). Complement membrane attack (MAC) in idiopathic IgA-glomerulonephritis. Kidney Int, 31, 820–9. Roberts, I. S., Cook, H. T., Troyanov, S., et al. (2009). The Oxford classification of IgA nephropathy: pathology definitions, correlations, and reproducibility. Kidney Int, 76, 546–56. Schena, F. P. (1992). IgA nephropathies. In S. Cameron, A. M. Davison, J. P. Grunfield, et al. (eds.) Oxford Textbook of Clinical Nephrology, pp. 339–69. Oxford: Oxford University Press. Schena, F. P. (1997). Survey of the Italian Registry of Renal Biopsies. Frequency of the renal diseases for 7 consecutive years. The Italian Group of Renal Immunopathology. Nephrol Dial Transplant, 12, 418–26. Shi, S. F., Wang, S. X., Jiang, L., et al. (2011). Pathologic predictors of renal outcome and therapeutic efficacy in IgA nephropathy: validation of the oxford classification. Clin J Am Soc Nephrol, 6, 2175–84. Shima, Y., Nakanishi, K., Hama, T., et al. (2012). Validity of the Oxford classification of IgA nephropathy in children. Pediatr Nephrol, 27,783–92. Southwest Pediatric Nephrology Study Group (1982). A multicenter study of IgA nephropathy in children. A report of the Southwest Pediatric Nephrology Study Group. Kidney Int, 22, 643–52. Stangou, M., Alexopoulos, E., Papagianni, A., et al. (2009). Urinary levels of epidermal growth factor, interleukin-6 and monocyte chemoattractant protein-1 may act as predictor markers of renal function outcome in immunoglobulin A nephropathy. Nephrology, 14, 613–20. Tam, K. Y., Leung, J. C., Chan, L. Y., et al. (2009). Macromolecular IgA1 taken from patients with familial IgA nephropathy or their asymptomatic relatives have higher reactivity to mesangial cells in vitro. Kidney Int, 75, 1330–9. Tan, Y., Zhang, J. J., Liu, G., et al. (2009). The level of urinary secretory immunoglobulin A (sIgA) of patients with IgA nephropathy is elevated and associated with pathological phenotypes. Clin Exp Immunol, 156, 111–16. Tiebosch, A. T., Wolters, J., Frederik, P. F., et al. (1987). Epidemiology of idiopathic glomerular disease: a prospective study. Kidney Int, 32, 112–16. To, K. F., Choi, P. C., Szeto, C. C., et al. (2000). Outcome of IgA nephropathy in adults graded by chronic histological lesions. Am J Kidney Dis, 35, 392–400. Van Es, L. A., de, Heer, E., Vleming, L. J., et al. (2008). GMP-17-positive T-lymphocytes in renal tubules predict progression in early stages of IgA nephropathy. Kidney Int, 73, 1426–33. Wakai, K., Kawamura, T., Endoh, M., et al. (2006). A scoring system to predict renal outcome in IgA nephropathy: from a nationwide prospective study. Nephrol Dial Transplant, 21, 2800–8. Wang, G., Kwan, B. C., Lai, F. M., et al. (2010). Expression of microRNAs in the urinary sediment of patients with IgA nephropathy. Dis Markers, 28, 79–86. Wasilewska, A., Zoch-Zwierz, W., Taranta-Janusz, K., et al. (2011). Urinary monocyte chemoattractant protein-1 excretion in children with glomerular proteinuria. Scand J Urol Nephrol, 45, 52–9. Woodrow, D. F., Shore, I., Moss, J., et al. (1989). Immunoelectron microscopic studies of immune complex deposits and basement membrane components in IgA nephropathy. J Pathol, 157, 47–57. Wu, J., Wang, N., Wang, J., et al. (2010). Identification of a uromodulin fragment for diagnosis of IgA nephropathy. Rapid Commun Mass Spectrom, 24, 1971–8. Zhang, J. J., Jiang, L., Liu, G., et al. (2009). Levels of urinary complement factor H in patients with IgA nephropathy are closely associated with disease activity. Scand J Immunol, 69, 457–64. Zwirner, J., Burg, M., Schulze, M., et al. (1997). Activated complement C3: a potentially novel predictor of progressive IgA nephropathy. Kidney Int, 51, 1257–64.

CHAPTER 68

Immunoglobulin A nephropathy: treatment and outcome Kar Neng Lai and Sydney C. W. Tang Clinical course The clinical course of IgA nephropathy is highly variable. In general, the glomerulopathy usually runs an indolent but slowly progressive course leading to ESRD in 20–50% of patients over 30 years. The symptoms and prevalence vary between regions due to ethnic difference and biopsy criteria. Some asymptomatic patients are diagnosed after incidental finding of microscopic haematuria, low-grade proteinuria, or hypertension. Patients with these isolated features may not be biopsied in many centres. Some patients present with episodic synpharyngitic macrohaematuria, with or without significant proteinuria or hypertension. Some patients present with advanced renal failure, and a subset presents with features of rapidly progressive glomerulonephritis. The natural history of primary IgA nephropathy is progression to CKD and eventually ESRD at variable rates, which means some patients may not develop CKD in their lifetime, while some patients may develop ESRD shortly after diagnosis. Clearly, these patients differ in terms of clinical as well as histologic features in the kidney biopsy. Even patients with the most benign clinical features must be monitored at least yearly for life. Among 72 patients in Hong Kong who presented with isolated microhaematuria and were found to have minimal proteinuria of < 0.4 g/day, normal renal function and blood pressure, 33% developed proteinuria of > 1 g/day, 26% became hypertensive, and 7% developed impaired renal function after a median observation period of 7 years (Szeto et al., 2001). Similar rates of progression among clinically early disease patients are reproduced in a recent analysis of 177 patients from Shanghai (Shen et al., 2008).

Outcome prognostic markers In assessing prognosis, most studies have examined features at the time of renal biopsy. Due to the indolent clinical course in most instances, using the definitive end-point of ESRD is often beyond the scope of prospective studies. This has led to the use of alternative markers including time to doubling of serum creatinine, and slope of 1/creatinine, creatinine clearance, or eGFR against time. Some studies employ the absolute or percentage change in proteinuria as a surrogate marker as proteinuria per se indicates the degree of glomerular injury and correlates strongly with outcome (Chapter 50).

Predicting clinical outcome Predicting clinical outcome for IgA nephropathy remains an imprecise process. There are clinicopathological features that are generally, but not universally, accepted as indicating a less favourable prognosis in patients with preserved clearance function at diagnosis (Table 68.1). Clinical predictors of progression include raised serum creatinine at the time of diagnosis, arterial hypertension, significant proteinuria (> 1 g/day), male gender, and persistent microhaematuria. Clearly patients with raised serum creatinine at diagnosis are likely to have progressive loss of renal function. It is likely that impaired renal function at diagnosis simply reflects belated discovery of an indolent disease process that has progressed over a substantial period of time before the diagnosis was made. Up to 50% of patients have a chronic course, characterized by a slowly declining renal function over 10–20 years, eventually developing renal failure. The percentage of patients who will be in ESRD is roughly the same as the duration of the disease in years from the time of diagnosis (Johnston et al., 1992). Life-table analysis of several series from Asia, Australia, Europe, and North America shows a 10-year renal survival of 80–90% (D’Amico et al., 1993).

Proteinuria Traditionally, the severity of proteinuria upon presentation carries prognostic implication. More recently, a strong correlation is found between proteinuria at the time of biopsy and the degree of histologic lesions (Cattran et al., 2009). While early studies suggested that the cut-off value for poor prognosis was 2 g/day, subsequent reports showed a continuous effect, with an adverse impact on outcome starting at 500 mg/day. More importantly, rather than a single measurement upon presentation, the change in proteinuria over time is being regarded as a better yardstick for prognosis. A Canadian study has showed that proteinuria during follow-up was the most important predictor of the rate of GFR decline (Reich et al., 2007). Patients with sustained proteinuria > 3 g/day lost GFR 25 times faster than those with proteinuria maintained < 1 g/day. Furthermore, patients with proteinuria > 3 g/day upon presentation who achieved a partial remission (< 1 g/day) during follow-up had a similar clinical course to patients with proteinuria ≤ 1 g/day throughout, and fared better than patients who never achieved remission. Another study in Hong Kong (Tang et  al.,

578

Section 3  

the patient with glomerular disease

Table 68.1  Commonly accepted markers of a worse prognosis for patients with IgA nephropathy Demographic ◆ Male sex ◆

Older age at diagnosis

◆ Obesity.

Clinical No history of macroscopic haematuria ◆ Persistent microscopic haematuria ◆ Hypertension, persistent. ◆

Laboratory Proteinuria persistently > 1000 mg/day ◆ Hyperuricaemia ◆ Hypertriglyceridemia. ◆

Histological (see Chapter 67) Light microscopy: • Mesangial hypercellularity • Focal segmental glomerular sclerosis • Endocapillary cellular proliferation • Capillaritis • Interstitial fibrosis/tubular atrophy • Thrombotic microangiopathy • Loss of podocytes ◆ Immunofluorescence microscopy: • IgG in mesangial deposits ◆ Electron microscopy: • Electron-dense deposits in capillary loops. ◆

2010)  demonstrated change in the urine albumin-to-creatinine ratio at 1  year to be an independent predictor of progression to ESRD upon a 6-year follow-up period. Using multivariate Cox analysis, age and mean proteinuria at follow-up are powerful independent prognostic predictors in a cohort of Italian patients (Coppo and D’Amico, 2005).

Other factors The mechanisms leading to progression for some of the markers are not well understood (e.g. the apparent benefit of a history of macroscopic haematuria). Other markers, such as obesity and hyperuricaemia, may exert some of their adverse effects remotely. Light microscopic findings of the renal biopsy are the basis for several classifications that have correlated specific features with a progressively worse prognosis (Table 68.2; see Chapter 67). The traditional histologic predictors of poor renal outcome include glomerulosclerosis, interstitial fibrosis, tubular atrophy, and crescent formation. The Oxford classification (see Chapter 67) has redefined some of these histologic predictors in which mesangial hypercellularity and segmental glomerulosclerosis/adhesions predicted a 50% reduction in eGFR or ESRD and rate of eGFR decline, respectively, whereas tubular atrophy/interstitial fibrosis predicted both. Finally, endocapillary proliferation, though not predictive of any these outcome parameters, was associated with treatment responsiveness to immunosuppressive agents (Cattran et al., 2009; Roberts et al.,

2009). These findings were recently validated in a retrospective cohort of 128 adult patients in France in which mesangial hypercellularity, endocapillary proliferation, segmental glomerulosclerosis, and tubular atrophy/interstitial fibrosis each predicted bad outcome, defined as doubling of serum creatinine or need of dialysis after a follow up of 80 months. More importantly, the presence of FSGS lesions, particularly the collapsing and cellular forms, significantly worsened renal survival (El Karoui et al., 2011).

Predictive equations Several individual centres have derived formulae to predict the risk of ESRD or rate of GFR loss for an individual patient, using commonly measured laboratory and/or histological features (Table 68.3). To date, there has been no consensus as to which of the components of the formula or even the end-points bear the most significant prognostic implication. Some components are unique to a given population in a region. Furthermore, the relative weights of shared components (e.g. serum creatinine) differ between formulae. It is disappointing that these calculations often yield disparate estimates under commonly encountered clinical scenarios. Only one formula from Finland (Rauta et al., 2002) has been validated in a second cohort of patients. Another formula from Japan (Wakai et al., 2006) was validated in the same cohort of subjects with additional patients after an extension of follow-up (Goto et al., 2009).

Genetic factors Finally, some studies suggest that gene polymorphisms also make a prognostic difference. None are yet ready for use in the clinic. The angiotensin-converting enzyme (ACE) DD genotype is associated with an increased rate of progressive renal disease in Caucasians (Harden et  al., 1995)  and Japanese (Hunley et  al., 1996). More recently, the PREDICT-IgAN study group from Japan identified from a hundred atherosclerotic disease-related gene polymorphisms that glycoprotein Ia and intercellular adhesion molecule-1 polymorphisms are significantly associated with progression (Yamamoto et al., 2009). Among the Chinese population, the 2093C-2180T haplotype of the MEGSIN gene, which is predominantly expressed in the mesangial cell regulating its matrix metabolism, proliferation, and apoptosis and is upregulated in IgA nephropathy, is shown to be associated with more severe forms of IgA nephropathy and more rapid progression (Xia et  al., 2006). This finding, however, cannot be reproduced in a smaller cohort of Czech patients (Maixnerova et al., 2008).

Treatment There is a lack of large randomized controlled trials (RCTs) that provide a definitive immunosuppressive protocol for IgA nephropathy

Non-immunosuppressive treatment Blockade of the renin–angiotensin system remains the mainstay of treatment for IgA nephropathy. Cheng et al. (2009a) analysed 11 RCTs involving 585 patients with seven trials using placebo/no treatment as controls and four trials used other antihypertensive agents as controls. They reported angiotensin-converting enzyme inhibitors (ACEIs)/angiotensin receptor blockers (ARBs) had statistically significant effects on protecting renal function and reduction of proteinuria when compared with control group. A recent

Chapter 68 

iga nephropathy: treatment and outcome

Table 68.2  Formulae to predict clinical outcome in an individual patient with IgA nephropathy Authors, year (number of subjects, country of origin)

Component of score clinical factors

Histology

Endpoint measurement

Comment

Validated

Beukhof et al., 1986 75 patients, the Netherlands

History of macroscopic haematuria, microscopic haematuria, creatinine clearance, and 24-hr proteinuria.

None

ESRD at 5 years

Requires control of hypertension.

No

Alamartine et al., 1991 (282 patients, France)

Proteinuria, hypertension, and Global optical score HLA B35 (mesangial, tubular, interstitial, and vascular components)

Renal insufficiency (serum creatinine ≥ 1.5 mg/dL (135μmol/L)) at 10 and 20 years

Proteinuria was strongest factor

No

Radford et al., 1997 (206 patients, Midwest USA)

Serum creatinine and age

Total glomerular score

ESRD at 5 and 10 years

Serum creatinine dominates. Older age confers better score. No accounting for treatment.

No

Bartosik et al., 2001 (298 patients, Canada)

Time-averaged MAP and proteinuria

None

Slope of GFR (C-G) after 2–3 years

No demographic, clinical, laboratory or histological parameter at biopsy significantly predicted progressive loss of GFR. MAP and proteinuria accounted for only one third of the variability in loss of renal function.

No

Rauta et al., 2002 (161 patients, Finland)

Microscopic haematuria and hypertension

Arteriosclerosis and glomerular score >2

ESRD at 10 years

Starting GFR (C-G) > 85 mL/ min. Some patients had been treated.

No

Magistroni et al., 2006 (310 patients, Italy)

Serum creatinine >1.4 mg/dL, proteinuria >1g/d, hypertension and age >30 yr.

None

ESRD at 10 years

Serum creatinine was strongest risk factor.

Yes

Wakai et al., 2006 (1754 patients, Japan)

Sex, age, systolic BP, proteinuria (dipstick), haematuria (microscopic), serum total protein, and serum creatinine.

Total histological grade (glomerular + interstitial/vascular)

ESRD at 7 years

Serum creatinine dominates, better score with older age; no accounting for treatment.

No

Goto et al., 2009 (2283 patients, Japan)

Sex, age, systolic BP, proteinuria (dipstick), haematuria (microscopic), serum albumin, and serum creatinine.

Total histological grade (glomerular + interstitial/vascular)

ESRD at 10 years

An extended observation of the Yes study reported by Wakai et al. (2006)

BP = blood pressure; C-G = Cockcroft-Gault formula; ESRD = end-stage renal disease; GFR = glomerular filtration rate; MAP = mean arterial pressure. Adapted and modified with permission from Clinicopathologic findings. Julian, B. A. and Wyatt, R. J. (2009). In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 83–106. Copyright © 2009 World Scientific Press.

meta-analysis of 27 RCTs (1577 participants) using ACEIs, ARBs, or a combination of both, versus other antihypertensives, other agents or placebo reveals renin–angiotensin blockade appears to potentially outweigh the harms in patients with IgA nephropathy (Reid et al., 2011). The benefits are largely manifested as a reduction in proteinuria, a surrogate outcome. There is no evidence that treatment with any of the antihypertensive agents evaluated affects major renal and/or cardiovascular endpoints or long-term mortality risk beyond the benefit that arises from controlling hypertension. The RCT evidence is insufficiently robust to demonstrate efficacy for any of the other non-immunosuppressive therapies including fish oils, anticoagulants, and tonsillectomy.

Immunomodulatory treatment Research is hampered by the slowly progressive nature of the disease, with 10-year renal survival rates exceeding 85%, marked patient heterogeneity, and the lack of a good animal model that closely resembles human IgA nephropathy.

Corticosteroids: effective in those at high risk A relatively large quantity of data on corticosteroid was contributed by Japanese researchers in early years. Kobayashi et al. (1986) reported that steroid treatment in 14 patients with proteinuria 1–2 g/day versus 29 control subjects was effective in lowering proteinuria, particularly in patients with baseline GFR > 70 mL/min. After

579

580

Section 3  

the patient with glomerular disease Table 68.3  Recurrence rate of IgA nephropathy Follow-up duration (mean) (months)

No. of allograft

Recurrence ratea No. (%)

Graft loss due to recurrence No. (%)

Moroni et al., 2013

*113.1 (60.5–165)

190

42 (22.1%)

12 (6.3%)

Coppo et al., 2007

*62.4 (45.6–114)

116

36 (31%)

NA

Moriyama et al., 2005

67.8 ± 19.9

49

13(26.5%)

5 (10%)

Choy et al., 2003

100.0 ± 5.8

75

14(18.7%)

3 (4.0 %)

Briganti et al., 2002

12–120

532

NA

15 (2.8%)

Andresdottir et al., 2001

67.2 ± 54

79

17(21.5%)

1 (1.3 %)

Authors, year

Ponticelli et al., 2001 Wang et al., 2001 Kim et al., 2001 Freese et al., 1999 Bumgardner et al., 1998

70.4 ± 50.5

106

37(35%)

4 (3.8 %)

*52 (18–155)

48

14(29.2%)

4 (8.3 %)

2–164

90

19(21.1%)

2 (2.2 %)

*67 (11–159)

104

13(12.5%)

6 (5.8 %)

61 ± 37

61

18(29.5%)

7 (11.5 %)

Ohmacht et al., 1997c

54 (7–127)

61

20(29.9%)

10 (16.4 %)

Frohnert et al., 1997

*78 (3–156)

53

10(19%)

3 (5.7 %)

Kessler et al., 1996b

68.1 ± 37.2

84

13(15.5%)

4 (4.8 %)

Hartung et al., 1995

45.9 ± 10

128

47(36.7%)

9 (7.0 %)

Odum et al., 1994

3–183

51

17(33.3%)

5 (9.8 %)

Bachman et al., 1986

20 ± 13

13

6 (46.2%)

1 (7.6 %)

* = median; % = percentage was calculated from number of graft loss due to recurrent IgA nephropathy / total number of patients with primary IgA nephropathy. a Recurrence rate in patients with clinical symptoms of proteinuria/haematuria/renal impairment. b Included 13 patients suffered from underlying Henoch–Schönlein purpura. c Included four patients suffered from underlying Henoch–Schönlein purpura.

Updated and adopted from Recurrent IgA nephropathy in transplant. Choy, B. Y. and Lai, K. N. (2009). In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 149–59. Copyright © 2009 World Scientific Press.

10 years’ follow-up, renal survival was 80% versus 34% (Kobayashi et al., 1996). These results were reproduced in 86 Italian patients when Pozzi et al. (1999) reported renoprotective effects of a 6-month course of steroid treatment. At 10 years, renal survival was better in the steroid group (Pozzi et  al., 2004). On the other hand, Lai et al. (1986) found no therapeutic value of short-term (6-month) corticosteroid treatment in Chinese patients. A  meta-analysis of seven RCTs involving 386 subjects suggests that corticosteroids have statistically significant effects on protecting renal function and reduction of proteinuria, but gastrointestinal tract reaction is a concern (Cheng et al., 2009b). Another meta-analysis (including 15 controlled, quasi-randomized controlled and non-controlled trials with 1542 participants) suggests corticosteroid therapy is associated with a decrease of proteinuria and with a statistically significant reduction of the risk in ESRD. Moreover, subgroup analysis also suggested that long-term steroid therapy had a higher efficiency than standard and short-term therapy (Zhou et al., 2011). A more recent meta-analysis (including 536 patients who had urinary protein excretion >1 g/day and normal renal function from nine relevant trials) suggested that high-dose and short-term therapy produced significant renal protection, whereas low-dose, long-term steroid use did not (Lv et al., 2012). The KDIGO Clinical Practice Guideline for Glomerulonephritis (Kidney Disease:  Improving

Global Outcomes (KDIGO) Glomerulonephritis Work Group, 2012) suggested that there is low-quality evidence for corticosteroids to provide additional benefits on top of optimized supportive care, and suggested that patients with persistent proteinuria > 1 g/ day despite adequate ACEI or ARB and blood pressure control and GFR > 50 mL/min/1.73 m2 receive a 6-month course of steroid therapy. Trials studying the efficacy and safety of steroid treatment in IgA nephropathy are currently underway.

Cyclophosphamide plus corticosteroids: uncertain Evidence that pulse corticosteroid plus intravenous or oral cyclophosphamide slow the progression of advanced IgA nephropathy was provided by several groups worldwide. Ballardie and Roberts (2002) showed in 38 patients with progressive renal deterioration that renal survival in cyclophosphamide-treated patients was considerably better at 5  years (72% compared with 6% in controls). Proteinuria and erythrocyturia reduced from 12 and 6 months of treatment, respectively. This study may be faulted, however, for suboptimal blood pressure control and insufficient use of medications that block the angiotensin system, the unusually poor survival rate of the placebo group, and the small number of patients. In another prospective, uncontrolled, open-label trial, 12 patients with crescentic IgA

Chapter 68 

nephropathy received three doses of methylprednisolone at 15 mg/ kg/day, followed by intravenous cyclophosphamide at 0.5 g/m2/ month for 6 months (Tumlin et al., 2003). Serum creatinine fell from 2.7 to 1.5 mg/dL, and proteinuria decreased from 4 to 1.3 g/ day after treatment, suggesting a beneficial role of cyclophosphamide in crescentic glomerulonephritis. In another observational, uncontrolled study, similar results were reported in 21 patients with advanced IgA nephropathy treated with intravenous pulse cyclophosphamide (0.75 g/m2/month) for 6 months together with low-dose oral prednisolone (Rasche et al., 2003). More recently, Mitsuiki et al. (2007) retrospectively examined the outcome of 35 patients with histologically advanced IgA nephropathy, in whom 27 received prednisolone for 2 years and oral cyclophosphamide (50 mg/day) for 6 months, and the remaining eight received supportive treatment. Renal prognosis was significantly better in the treatment group. Overall, these studies suggest that combined cyclophosphamide/steroid therapy may benefit patients at very high risk of renal failure, namely those with a progressive decline in GFR and/or crescentic lesions before randomization. Due to the side effects, it is reasonable to use short-term cyclophosphamide with corticosteroids for IgA nephropathy patients with true crescentic or rapidly progressive glomerulonephritis. The KDIGO Clinical Practice Guideline for Glomerulonephritis (Kidney Disease:  Improving Global Outcomes (KDIGO) Glomerulonephritis Work Group, 2012)  suggested a similar approach (low-quality evidence).

Tonsillectomy: doubtful For a long time, tonsillectomy was considered a treatment option for IgA nephropathy, aimed at removing a relevant source of pathogens, which can multiply in tonsil crypts, and also in macrophages and B cells in lymphoid tonsil follicles. This specific antigen challenge was thought to induce a supernormal IgA synthesis, as tonsil lymphocytes from IgA nephropathy patients showed a higher production of dimeric and undergalactosylated IgA1 than control subjects. In Japan, tonsillectomy-steroid pulse therapy has frequently been used for treatment of early IgA nephropathy, and showed favourable outcomes (Moriyama and Nitta, 2011). A recent meta-analysis of seven studies (six from Japan and one from China) comprising 858 patients (534 underwent tonsillectomy and 324 did not) showed that tonsillectomy combined with either normal steroid or steroid pulse treatment, but not tonsillectomy or steroid treatment alone, resulted in higher remission rates with favourable long-term efficacy at both 5- and 10-year follow-up (Wang et al., 2011). Elsewhere, the benefits of tonsillectomy have been less impressive. A  retrospective review of 61 Caucasian patients showed that tonsillectomy was not associated with a different rate of disease progression after 20 years of follow-up (Piccoli et al., 2010). From the available evidence, it seems unlikely that a dysregulated mucosal immune system in IgA nephropathy could be substantially controlled by tonsillectomy alone. The role of tonsillectomy remains controversial. It is presently encouraged and practised only in Japan and certain parts of Asia but opposed in the rest of the world except where there are clear ENT indications. Randomized, controlled trials are needed to resolve this conflict. At present, the KDIGO Clinical Practice Guideline for Glomerulonephritis (Kidney Disease:  Improving Global Outcomes (KDIGO)

iga nephropathy: treatment and outcome

Glomerulonephritis Work Group, 2012) suggested that tonsillectomy should not be performed for IgA nephropathy (low-quality evidence).

Calcineurin inhibitors: no evidence of long term benefit There is minimal evidence for calcineurin inhibitors, though they may reduce proteinuria in the short term. Lai et  al. (1987) conducted a randomized prospective single blind study of ciclosporin in 19 patients with proteinuria > 1.5 g/day. Patients who received the drug had significant reduction of proteinuria, serum IgA, and increase of plasma albumin concentration compared with placebo. However, there was transient deterioration of renal function during treatment, despite within-range trough drug levels. The authors discourage indiscriminate use of ciclosporin in IgA nephropathy due to lack of efficacy and nephrotoxicity. A more recent study suggested that tacrolimus could induce remission of proteinuria in 14 patients with refractory IgA nephropathy, possibly by stabilizing podocyte cytoskeleton (Zhang, et al., 2012) as in other conditions (see Chapters 45, 58).

Azathioprine: ineffective A retrospective analysis of 74 IgA nephropathy patients followed for 10  years shows that long-term azathioprine combined with low-dose prednisone did not alter the clinical course compared to untreated controls (Goumenos et al., 2003). However, in a subgroup of patients with heavy proteinuria > 3 g/day and baseline serum creatinine between 1.4 and 2.5 mg/dL, this immunosuppressive regimen reduced the risk of doubling serum creatinine compared to controls (27% versus 78%) and delayed progression to end-stage renal failure (17% versus 55%). The Japanese Paediatric IgA Nephropathy Treatment Study Group randomized 78 children with newly diagnosed early IgA nephropathy to receive either prednisolone, azathioprine, heparin-warfarin, and dipyridamole or the combination of heparin-warfarin, and dipyridamole only (Yoshikawa and Ito, 1999). The study was flawed by a lack of data on baseline proteinuria and creatinine clearance as well as blood pressure control in both groups. A recent prospective randomized study of 207 subjects showed that the addition of azathioprine to corticosteroids did not provide additional benefits in terms of renal survival versus corticosteroids alone in patients with proteinuria ≥ 1 g/day and plasma creatinine ≤ 2.0 mg/dL (Pozzi et  al., 2010). Current data therefore suggest that the addition of azathioprine was ineffective and may even be potentially toxic.

Mycophenolate mofetil: probably ineffective To date, four randomized clinical trials have been published on the use of mycophenolate mofetil (MMF) in IgA nephropathy, which add more controversy than consensus. Although these trials have produced conflicting results, they differ significantly in patient selection and treatment duration and deserve attention. The first randomized study was conducted 62 Chinese patients with severe IgA nephropathy and urinary protein > 2.0 g/ day receiving MMF or oral prednisone for at least 12  months (Chen et al., 2002). After 18 months’ follow-up, the MMF group showed significant improvement in proteinuria and serum lipids than the prednisone group. In a study of 34 Belgian patients with impaired renal function, histologic unfavourable criteria and arterial hypertension, after instituting salt restriction and ACEI therapy in all, MMF failed

581

582

Section 3  

the patient with glomerular disease

to demonstrate a better beneficial effect after 3 years of evaluation (Maes et al., 2004). In a similar study that recruited patients with even more advanced renal insufficiency using MMF as a ‘salvage’ therapy, a worse outcome occurred in the MMF group (Frisch et al., 2005). Tang et al. (2005) treated 40 Chinese patients with mild tubulointerstitial lesions with persistent proteinuria > 1 g/day despite full angiotensin blockade for 6 months. Twelve months after stopping MMF, the overall remission rate was significantly higher in MMF-treated patients whose proteinuria dropped to 62% of baseline, whereas urine protein in control patients increased to 120% of baseline. Serum interleukin-6 levels and, more intriguingly, in vitro binding of IgA to mesangial cells were elevated at baseline in both groups compared with normal healthy subjects; however, after MMF treatment these parameters were comparable to those of the healthy subjects. Serum interleukin-6 concentration and mesangial binding of IgA in patients who did not receive MMF showed no change. After 6 years’ follow-up, the difference in proteinuria between the two groups was lost, but renal survival was significantly better in the MMF group (Tang et  al., 2010). In another Italian study, a subset of IgA nephropathy patients with florid glomerular changes treated with MMF and steroid showed remission of proteinuria and reversal of progressive renal failure (Roccatello et al., 2012). Overall, MMF appears to be effective in reducing proteinuria in Chinese but not Caucasoid subjects. Therefore, ethnic differences may be one possible reason to account for the differences observed in these studies. Another possibility is the mild histologic grade in Tang’s study versus the moderate-to-severe grades in the studies by Maes and Frisch. Further observation and studies are needed to provide more definitive answers on the efficacy of MMF in IgA nephropathy. The KDIGO Clinical Practice Guideline for Glomerulonephritis (Kidney Disease: Improving Global Outcomes (KDIGO) Glomerulonephritis Work Group, 2012) suggested not using MMF in IgA nephropathy (low-quality evidence).

New therapies Chapter 136 gives an overview of hypotheses of long term progression of renal disease, and chapter 99 lists recommendations for management. In vitro studies shows that peroxisome proliferator-activated receptor-γ (PPAR-γ) agonist attenuates inflammatory response in activated tubular epithelial cells in IgA nephropathy by downregulating the expression of ATR1 (Xiao et al., 2009). Dual treatment of PPAR-γ and ARB provide synergistic effect in reducing inflammation and angiotensin II signalling in renal tubular epithelial cells and the therapeutic benefit is confirmed in an animal model of IgA nephropathy (Lai et al., 2011). It is possible that this drug acts at the level of the podocyte (Chapter 45). Aliskiren, a direct renin inhibitor, confers an antiproteinuric effect in 25 IgA nephropathy patients with significant residual proteinuria despite receiving the adequate ARB treatment (Tang et al., 2012). Treatment for 12 months lowered the mean urinary protein:creatinine ratio by 26.3% with significant reductions in plasma renin activity, and serum interleukin-6 and transforming growth factor-β levels. The antiproteinuric efficacy of aliskiren on top of an ACEI/ARB has been confirmed in a randomized cross-over study of 22 patients (Szeto et  al., 2013). However, combining aliskiren with an ARB has not been found to confer

renoprotection and has been hailed to be potentially harmful due to the high incidence of hyperkalaemia and hypotension among type 2 diabetic subjects (Parving et al., 2012). Its application in the treatment of IgA nephropathy, therefore, remains to be investigated. Enteric budesonide targeted to the Peyer’s patches in the ileocaecal region has been shown to reduce proteinuria by 23% and modestly augment eGFR by 8% in 16 patients who received this new formulation for 6 months followed by 3 months’ further observation (Smerud et al., 2011). Based on these encouraging results, a multicentre phase IIb trial is currently being planned in Europe.

Immunoglobulin A nephropathy after renal transplantation Recurrence of mesangial IgA deposits in the renal allografts was first described by Berger et al. (1975). Subsequent studies reported a recurrence rate ranging from 13% to 60% of patients (Table 68.1) (Bachman et  al., 1986; Odum et  al., 1994; Hartung et  al., 1995; Kessler et  al., 1996; Frohnert et  al., 1997; Ohmacht et  al., 1997; Bumgardner et  al., 1998; Freese et  al., 1999; Andresdottir et  al., 2001; Kim et  al., 2001; Ponticelli et  al., 2001; Wang et  al., 2001; Briganti et al., 2002; Moriyama et al., 2005; Choy et al., 2003, 2006, 2007; Moroni et al., 2013). Great variation can partly be explained by the difference in biopsy policy of different transplant centres and the duration of follow-up. Most centres performed renal biopsy only when patients presented with clinical symptoms. This would potentially underestimate the rate of recurrence as patients who were clinically asymptomatic but with immunohistological changes in the graft kidneys would remain undiagnosed. In a Canadian epidemiologic study comprising 2026 sequential renal transplant recipients without loss to follow-up, IgA nephropathy was found to recur in 25.3% of patients after 15 years (Chailimpamontree et al., 2009). The cumulative risk of graft loss following the diagnosis of post-transplant glomerulonephritis (recurrent and de novo forms summed) was over sevenfold. Recurrent disease after transplantation is also discussed in Chapter 289.

Clinical course Recurrent disease exhibits considerable clinical similarities with primary IgA nephropathy. Microscopic haematuria and proteinuria are common presenting symptoms followed by slow decline in renal function. With increasing long-term data, it is apparent that recurrent disease is not as benign as had been reported previously (Berger et al., 1975). Graft loss from recurrence with histological features of diffuse mesangial proliferative expansion and glomerular sclerosis were reported between 2% and 16% depending on duration of follow up. Briganti et al. (2002) reported an estimated 10-year incidence of graft loss due to recurrent IgA nephropathy of 9.7% basing on data from the Australia and New Zealand Dialysis and Transplant Registry (ANZDATA) which contains 532 allograft recipients with primary IgA nephropathy. The renal allograft survival of recurrent IgA nephropathy for the first 5 years post-transplant is better compared to patients with other primary diseases. Lim et al. (1993) reported a superior 5-year graft survival rate in patients with IgA nephropathy as compared to patients with other primary diseases. The superior graft survival of IgA nephropathy patients for the early post-transplant period

Chapter 68 

is no longer observed on longer follow-up. Ponticelli et al. (2001) reported a comparable 10-year graft survival for patient with IgA nephropathy. Choy et  al. (2006) reported an inferior graft survival for primary IgA nephropathy patients with follow up beyond 12 years. At 15 years, IgA nephropathy had a higher cumulative incidence of graft failure with non-IgA nephropathy controls (Moroni et al., 2013). These observations suggest that impact of other factors including recurrent disease on graft survival becomes more apparent on long-term follow up and recurrent IgA nephropathy runs an indolent course similar to primary disease with favourable outcome in the initial 10 years post transplant and thereafter its contribution to graft loss becomes more significant (Kim et al., 2001; Ponticelli et al., 2001; Choy et al., 2003). Patients with prior graft loss due to recurrent IgA nephropathy have higher risk of recurrence in the second transplant (20–100%).

Potential risk factors for recurrence Donor type Pooling all available data from literature that contained information on graft recurrence and graft loss in relation to donor type showed a higher risk of disease recurrence amongst transplant recipients with related donors (common odds ratio 2.29, P < 0.001) but the risk of graft loss was not increased (common odds ratio 1.95, P = 0.24) (Choy and Lai, 2009). Given the fact that the graft survival of patients with primary IgA nephropathy is excellent for the first decade post transplant, it is inappropriate to refrain from living related donor transplantation even though there may be a slight risk of recurrence. In contrast, familial IgA nephropathy should be rigorously excluded in potential living related donors since this may be associated with high risk of development of the nephropathy in affected members with more severe pathology (Tam et al., 2010).

Human leucocyte antigens and degree of mismatch No specific type of HLA has been identified to be predictive of recurrence. An analysis by ANZDATA studying 1306 patients with primary IgA nephropathy reported a higher risk of recurrent disease only in patients who received zero HLA mismatch living donor grafts (McDonald and Russ, 2006).

Latent IgA deposition from donor kidney Incidental finding of glomerular mesangial IgA deposits in donor kidneys has been reported in 4–24 % of patients (Suzuki et al., 2003; Ji et al., 2004). These deposits usually disappear within 6 months post transplant if recipients do not have primary IgA nephropathy.

Serological and genetic factors High level of aberrantly glycosylated IgA1 in recipients did not predict recurrence and no association of ACE gene (insertion/deletion) polymorphism was detected with recurrence (Coppo et al., 2007).

Prevention and management Apparently, no effective therapy for prevention or treatment of recurrent IgA nephropathy is available at the moment. There is no evidence that any particular immunosuppressive regime alters the incidence or clinical course of recurrent disease. Systemic hypertension, glomerular hyperfiltration, and heavy proteinuria secondary to recurrent IgA nephropathy are detrimental to the graft function. Adequate blockade of the renin–angiotensin system provides a better graft survival for recurrent IgA nephropathy (Oka

iga nephropathy: treatment and outcome

et al., 2000; Courtney et al., 2006). However, a recent retrospective study from Italy revealed a significant reduction of recurrence of IgA nephropathy from 1981 to 2010 correlating with mycophenolate treatment and triple immunosuppressive therapy (Moroni et al., 2013).

References Alamartine, E., Sabatier, J. C., Guerin, C., et al. (1991). Prognostic factors in mesangial IgA glomerulonephritis: an extensive study with univariate and multivariate analyses. Am J Kidney Dis, 18, 12–19. Andresdottir, M. B., Hoitsma, A. J., Assmann, K. J., et al. (2001). Favorable outcome of renal transplantation in patients with IgA nephropathy. Clin Nephrol, 56, 279–88. Bachman, U., Biava, C., Amend, W., et al. (1986). The clinical course of IgA-nephropathy and Henoch-Schonlein purpura following renal transplantation. Transplantation, 42, 511–15. Ballardie, F. W. and Roberts, I. S. (2002). Controlled prospective trial of prednisolone and cytotoxics in progressive IgA nephropathy. J Amer Soc Nephrol, 13, 142–8. Bartosik, L. P., Lajoie, G., Sugar, L., et al. (2001). Predicting progression in IgA nephropathy. Am J Kidney Dis, 38, 728–35. Berger, J., Yaneva, H., Nabarra, B., et al. (1975). Recurrence of mesangial deposition of IgA after renal transplantation. Kidney Int, 7, 232–41. Beukhof, J. R., Kardaun, O., Schaafsma, W., et al. (1986). Toward individual prognosis of IgA nephropathy. Kidney Int, 29, 549–56. Briganti, E. M., Russ, G. R., McNeil, J. J., et al. (2002). Risk of renal allograft loss from recurrent glomerulonephritis. N Engl J Med, 347, 103–9. Bumgardner, G. L., Amend, W. C., Ascher, N. L., et al. (1998). Single-center long-term results of renal transplantation for IgA nephropathy. Transplantation, 65, 1053–60. Cattran, D. C., Coppo, R., Cook, H. T., et al. (2009). The Oxford classification of IgA nephropathy: rationale, clinicopathological correlations, and classification. Kidney Int, 76, 534–45. Cheng, J., Zhang, W., Zhang, X. H., et al. (2009a). ACEI/ARB therapy for IgA nephropathy: a meta analysis of randomised controlled trials. Int J Clin Pract, 63, 880–8. Cheng, J., Zhang, X., Zhang, W., et al. (2009b). Efficacy and safety of glucocorticoids therapy for IgA nephropathy: a meta-analysis of randomized controlled trials. Am J Nephrol, 30, 315–22. Choy, B. Y., Chan, T. M., and Lai, K. N. (2006). Recurrent glomerulonephritis after kidney transplantation. Am J Transplant, 6, 2535–42. Choy, B. Y., Chan, T. M., Lo, S. K., et al. (2003). Renal transplantation in patients with primary immunoglobulin A nephropathy. Nephrol Dial Transplant, 18, 2399–404. Choy, B. Y. and Lai, K. N. (2009). Recurrent IgA nephropathy in transplant. In K. N. Lai (ed.) Recent Advances in IgA Nephropathy, pp. 149–59. Singapore: World Scientific Press. Coppo, R., Amore, A., Chiesa, M., et al. (2007). Serological and genetic factors in early recurrence of IgA nephropathy after renal transplantation. Clin Transplant, 21, 728–37. Coppo, R. and D’Amico, G. (2005). Factors predicting progression of IgA nephropathies. J Nephrol, 18, 503–12. Courtney, A. E., McNamee, P. T., Nelson, W. E., et al. (2006). Does angiotensin blockade influence graft outcome in renal transplant recipients with IgA nephropathy? Nephrol Dial Transplant, 21, 3550–4. D’Amico, G., Ragni, A., Gandini, E., et al. (1993). Typical and atypical natural history of IgA nephropathy in adult patients. Contrib Nephrol, 104, 6–13. El Karoui, K., Hill, G. S., Karras, A., et al. (2011). Focal segmental glomerulosclerosis plays a major role in the progression of IgA nephropathy. II. Light microscopic and clinical studies. Kidney Int, 79, 643–54.

583

584

Section 3  

the patient with glomerular disease

Freese, P., Svalander, C., Norden, G., et al. (1999). Clinical risk factors for recurrence of IgA nephropathy. Clin Transplant, 13, 313–17. Frohnert, P. P., Donadio, J. V., Jr., Velosa, J. A., et al. (1997). The fate of renal transplants in patients with IgA nephropathy. Clin Transplant, 11, 127–33. Goto, M., Wakai, K., Kawamura, T., et al. (2009). A scoring system to predict renal outcome in IgA nephropathy: a nationwide 10-year prospective cohort study. Nephrol Dial Transplant, 24, 3068–74. Harden, P. N., Geddes, C., Rowe, P. A., et al. (1995). Polymorphisms in angiotensin-converting-enzyme gene and progression of IgA nephropathy. Lancet, 345, 1540–2. Hartung, R., Livingston, B., Excell, L., et al. (1995). Recurrence of IgA deposits/disease in grafts. An Australian Registry Survey 1980-1990. Contrib Nephrol, 111, 13–16. Hunley, T. E., Julian, B. A., Phillips, J. A., et al. (1996). Angiotensin converting enzyme gene polymorphism: potential silencer motif and impact on progression in IgA nephropathy. Kidney Int, 49, 571–7. Ji, S., Liu, M., Chen, J., et al. (2004). The fate of glomerular mesangial IgA deposition in the donated kidney after allograft transplantation. Clin Transplant, 18, 536–40. Johnston, P. A., Brown, J. S., Braumholtz, D. A., et al. (1992). Clinico-pathological correlations and long-term follow-up of 253 United Kingdom patients with IgA nephropathy. A report from the MRC Glomerulonephritis Registry. QJM, 84, 619–27. Kessler, M., Hiesse, C., Hestin, D., et al. (1996). Recurrence of immunoglobulin A nephropathy after renal transplantation in the cyclosporine era. Am J Kidney Dis, 28, 99–104. Kidney Disease: Improving Global Outcomes (KDIGO) Glomerulonephritis Work Group (2012). KDIGO clinical practice guideline for glomerulonephritis. Kidney Int Suppl, 2(2), 139–274. Kim, Y. S., Moon, J. I., Jeong, H. J., et al. (2001). Live donor renal allograft in end-stage renal failure patients from immunoglobulin A nephropathy. Transplantation, 71, 233–8. Kobayashi, Y., Fujii, K., Hiki, Y., et al. (1986). Steroid therapy in IgA nephropathy: a prospective pilot study in moderate proteinuric cases. QJM, 61, 935–43. Kobayashi, Y., Hiki, Y., Kokubo, T., et al. (1996). Steroid therapy during the early stage of progressive IgA nephropathy. A 10-year follow-up study. Nephron, 72, 237–42. Lai, K. N., Chan, L. Y., Guo, H., et al. (2011). Additive effect of PPAR-gamma agonist and ARB in treatment of experimental IgA nephropathy. Pediatr Nephrol, 26, 257–66. Lai, K. N., Lai, F. M., Ho, C. P., et al. (1986). Corticosteroid therapy in IgA nephropathy with nephrotic syndrome: a long-term controlled trial. Clin Nephrol, 26, 174–80. Lai, K. N., Lai, F. M., Li, P. K., et al. (1987). Cyclosporin treatment of IgA nephropathy: a short term controlled trial. Br Med J (Clin Res Ed), 295, 1165–8. Lai, K. N., To, W. Y., Li, P. K., et al. (1996). Increased binding of polymeric lambda-IgA to cultured human mesangial cells in IgA nephropathy. Kidney Int, 49, 839–45. Lv, J., Xu, D., Perkovic, V., et al. (2012). TESTING Study Group. Corticosteroid therapy in IgA nephropathy. J Am Soc Nephrol, 23, 1108–16. Magistroni, R., Furci, L., Leonelli, M., et al. (2006). A validated model of disease progression in IgA nephropathy. J Nephrol, 19, 32–40. Maixnerova, D., Merta, M., Reiterova, J. et al. (2008). The influence of two megsin polymorphisms on the progression of IgA nephropathy. Folia Biol (Praha), 54, 40–5. McDonald, S. P. and Russ, G. R. (2006). Recurrence of IgA nephropathy among renal allograft recipients from living donors is greater among those with zero HLA mismatches. Transplantation, 82, 759–62. Mitsuiki, K., Harada, A., Okura, T., et al. (2007). Histologically advanced IgA nephropathy treated successfully with prednisolone and cyclophosphamide. Clin Exp Nephrol, 11, 297–303. Moriyama, T., Nitta, K., Suzuki, K., et al. (2005). Latent IgA deposition from donor kidney is the major risk factor for recurrent IgA

nephropathy in renal transplantation. Clin Transplant, 19 Suppl 14, 41–8. Moriyama, T. and Nitta, K. (2011). Tonsillectomy and steroid pulse therapy for IgA nephropathy. Tohoku J Exp Med, 224, 243–50. Moroni, G., Longhi, S., Quaglini, S., et al. (2013). The long-term outcome of renal transplantation of IgA nephropathy and the impact of recurrence on graft survival. Nephrol Dial Transplant , 28, 1305–14. Odum, J., Peh, C. A., Clarkson, A. R., et al. (1994). Recurrent mesangial IgA nephritis following renal transplantation. Nephrol Dial Transplant, 9, 309–12. Ohmacht, C., Kliem, V., Burg, M., et al. (1997). Recurrent immunoglobulin A nephropathy after renal transplantation: a significant contributor to graft loss. Transplantation, 64, 1493–6. Oka, K., Imai, E., Moriyama, T., et al. (2000). A clinicopathological study of IgA nephropathy in renal transplant recipients: beneficial effect of angiotensin-converting enzyme inhibitor. Nephrol Dial Transplant, 15, 689–95. Parving, H. H., Brenner, B. M., McMurray, J. J., et al. (2012). ALTITUDE Investigators. Cardiorenal end points in a trial of aliskiren for type 2 diabetes. N Engl J Med, 367, 2204–13. Piccoli, A., Codognotto, M., Tabbi, M. G. et al. (2010). Influence of tonsillectomy on the progression of mesangioproliferative glomerulonephritis. Nephrol Dial Transplant, 25, 2583–9. Ponticelli, C., Traversi, L., Feliciani, A., et al. (2001). Kidney transplantation in patients with IgA mesangial glomerulonephritis. Kidney Int, 60, 1948–54. Pozzi, C., Andrulli, S., Pani, A., et al. (2010). Addition of Azathioprine to Corticosteroids Does Not Benefit Patients with IgA Nephropathy. J Am Soc Nephrol, 21, 1783–90. Pozzi, C., Bolasco, P. G., Fogazzi, G. B. et al. (1999). Corticosteroids in IgA nephropathy: a randomised controlled trial. Lancet, 353, 883–7. Radford, M. G., Jr., Donadio, J. V., Jr., Bergstralh, E. J., et al. (1997). Predicting renal outcome in IgA nephropathy. J Am Soc Nephrol, 8, 199–207. Rasche, F. M., Klotz, C. H., Czock, D., et al. (2003). Cyclophosphamide pulse therapy in advanced progressive IgA nephropathy. Nephron Clin Pract, 93, c131–c136. Rauta, V., Finne, P., Fagerudd, J., et al. (2002). Factors associated with progression of IgA nephropathy are related to renal function—a model for estimating risk of progression in mild disease. Clin Nephrol, 58, 85–94. Reich, H. N., Troyanov, S., Scholey, J. W., et al. (2007). Remission of proteinuria improves prognosis in IgA nephropathy. J Amer Soc Nephrol, 18, 3177–83. Reid, S., Cawthon, P. M., Craig, J. C., et al. (2011). Non-immunosuppressive treatment for IgA nephropathy. Cochrane Database Syst Rev, 3, CD003962. Roberts, I. S., Cook, H. T., Troyanov, S., et al. (2009). The Oxford classification of IgA nephropathy: pathology definitions, correlations, and reproducibility. Kidney Int, 76, 546–56. Samuels, J.A., Strippoli, G.F., Craig, J.C., Schena, F.P., and Molony, D.A. (2004). Immunosuppressive treatments for immunoglobulin A nephropathy: a meta-analysis of randomized controlled trials. Nephrology, 9, 177–85. Shen, P., He, L., and Huang, D. (2008). Clinical course and prognostic factors of clinical early IgA nephropathy. Neth J Med, 66, 242–7. Smerud, H. K., Barany, P., Lindstrom, K., et al. (2011). New treatment of IgA nephropathy: enteric budesonide targeted to the ileocecal region ameliorates proteinuria. Nephrol Dial Transplant, 26, 3237–42. Suzuki, K., Honda, K., Tanabe, K., et al. (2003). Incidence of latent mesangial IgA deposition in renal allograft donors in Japan. Kidney Int, 63, 2286–94. Szeto, C. C., Lai, F. M., To, K. F., et al. (2001). The natural history of immunoglobulin a nephropathy among patients with hematuria and minimal proteinuria. Am J Med, 110, 434–7. Szeto, C. C., Kwan, B. C., Chow, K. M., et al. (2013). The safety and short-term efficacy of aliskiren in the treatment of immunoglobulin a nephropathy—a randomized cross-over study. PLoS One, 8, e62736. Tam, K. Y., Leung, J. C., Chan, L. Y., et al. (2010). In vitro enhanced chemotaxis of CD25+ mononuclear cells in patients with familial IgAN

Chapter 68 

through glomerulotubular interactions. Am J Physiol Renal Physiol, 299, F359–68. Tang, S., Leung, J. C., Chan, L. Y. et al. (2005). Mycophenolate mofetil alleviates persistent proteinuria in IgA nephropathy. Kidney Int, 68, 802–12. Tang, S. C. and Lai, K. N. (2009). The ubiquitin-proteasome pathway and IgA nephropathy: a novel link? Kidney Int, 75, 457–9. Tang, S. C., Lin, M., Tam, S., et al. (2012). Aliskiren combined with losartan in immunoglobulin A nephropathy: an open-labeled pilot study. Nephrol Dial Transplant, 27, 613–18. Tang, S. C., Tang, A. W., Wong, S. S., et al. (2010). Long-term study of mycophenolate mofetil treatment in IgA nephropathy. Kidney Int, 77, 543–9. Tumlin, J. A., Lohavichan, V., and Hennigar, R. (2003). Crescentic, proliferative IgA nephropathy: clinical and histological response to methylprednisolone and intravenous cyclophosphamide. Nephrol Dial Transplant, 18, 1321–9. Wakai, K., Kawamura, T., Endoh, M., et al. (2006). A scoring system to predict renal outcome in IgA nephropathy: from a nationwide prospective study. Nephrol Dial Transplant, 21, 2800–8. Wang, A. Y., Lai, F. M., Yu, A. W., et al. (2001). Recurrent IgA nephropathy in renal transplant allografts. Am J Kidney Dis, 38, 588–96.

iga nephropathy: treatment and outcome

Wang, Y., Chen, J., Wang, Y., et al. (2011). A meta-analysis of the clinical remission rate and long-term efficacy of tonsillectomy in patients with IgA nephropathy. Nephrol Dial Transplant, 26, 1923–31. Xia, Y., Li, Y., Du, Y., et al. (2006). Association of MEGSIN 2093C-2180T haplotype at the 3’ untranslated region with disease severity and progression of IgA nephropathy. Nephrol Dial Transplant, 21, 1570–4. Xiao, J., Leung, J. C., Chan, L. Y., et al. (2009). Crosstalk between peroxisome proliferator-activated receptor-gamma and angiotensin II in renal tubular epithelial cells in IgA nephropathy. Clin Immunol, 132, 266–76. Yamamoto, R., Nagasawa, Y., Shoji, T., et al. (2009). A candidate gene approach to genetic prognostic factors of IgA nephropathy—a result of Polymorphism REsearch to DIstinguish genetic factors Contributing To progression of IgA Nephropathy (PREDICT-IgAN). Nephrol Dial Transplant, 24, 3686–94. Zhang, Q., Shi, S.F., Zhu, L., et al. (2012). Tacrolimus improves the proteinuria remission in patients with refractory IgA nephropathy. Am J Nephrol, 35, 312–20. Zhou, Y. H., Tang, L. G., Guo, S. L., et al. (2011). Steroids in the treatment of IgA nephropathy to the improvement of renal survival: a systematic review and meta-analysis. PLoS One, 6, e18788.

585

CHAPTER 69

Immunoglobulin A nephropathy: pathogenesis Kar Neng Lai and Sydney C. W. Tang to five O-linked glycan chains. O-glycans on circulatory IgA1 consist of N-acetylgalactosamine (GalNAc) with a β1,3-linked galactose; both residues may be sialylated. Carbohydrate composition of O-linked glycans on normal serum IgA1 is variable (Figs. 69.1, 69.2). Prevailing forms include the galactose-GalNAc disaccharide and its mono- and disialylated forms. Galactose-deficient variants with terminal GalNAc or sialylated GalNAc are more common in IgA nephropathy patients. These aberrantly glycosylated galactose-deficient forms predominate in glomerular IgA deposits

Aberrant structure of the immunoglobulin A molecule The primary defect of IgA nephropathy seems to lie in the structure of IgA molecule. In humans, IgA1 represents one of the two structurally and functionally distinct subclasses of IgA. Unlike IgA2, IgM, and IgG, IgA1 has heavy chains that contain a unique hinge-region segment between the first and second constant-region domains, which is the site of attachment of three CH1 FAB CH1

CL Hinge region

CH2

FC CH3

α1,3

GalNAc A Ser/Thr Val C1GALT1 Pro Ser C1GALTIC1 Thr 255 α1,3 β1,3 Pro Pro B Ser/Thr GalNAc Gal Thr 228 Pro Ser 230 Pro Ser 232 α1,3 Thr C Ser/Thr GalNAc Pro α2,6 Pro NeuNAc Thr 236 Pro Ser α1,3 Pro β1,3 Ser GalNAc Gal D Ser/Thr Cys NeuNAc α2,3 CH2

α1,3 E

Ser/Thr

β1,3

GalNAc NeuNAc

F

α2,6

Gal

Tn antigen

C1GALT1 + Cosmc GalNac-T2 β1,3

α1,3 Ser/Thr

T antigen Core 1 structure

Sialyl-T

MonosialyI-T

MonosialyI-T

α1,3 β1,3 NeuNAc Disialyl-T α2,3 Ser/Thr GalNAc Gal α 2,6

NeuNAc

Fig. 69.1  The microheterogeneity of the O-glycans at the hinge region of IgA1 molecule. O-Glycosylation of protein is initiated by the addition of GalNAc to serine or threonine residues through the activity of UDP-N-acetyl-α-D-galactosamine:polypeptide N-acetylgalactosaminyltransferases (pp-GalNAc-Ts), which is followed by β-galactosylation through core 1 synthase 1, also called glycoprotein-N-acetylgalactosamine 3-β-galactosyltransferase 1, core 1 β3-Gal-T, or T-synthase (C1GALT1). The moiety is galactosylated in the β1,3 configuration by core-1 β1-3 galactosyltransferase-1 (C1Gal-T1), which requires the presence of the core-1-β3-Gal-T-specific molecular chaperone (Cosmc). Cosmc is encoded by the CIGALT1C1 gene.

Chapter 69 

IgA1

immunoglobulin a nephropathy: pathogenesis AGTR1

Glomerulo-podocytic crosstalk via TNF and TGF-β

Glomerulo-tubular crosstalk via TNF, IL-6 and Ang II AGTR2

AGTR1 Tubular epithelial cell

MR AGTR2

Glomerulo-tublar crosstalk favoring glomerulosclerosis

Podocyte

Mesangial cell

TNFR1

Aldosterone

TNF

TNFR2

Ang II

Apoptosis

Proteinuria Tubular atrophy

TNF IL-6

Apoptosis Bcl-2 BAX

Proinflammatory cellular response

Fig. 69.2  Proposed pathways leading to glomerular damage, podocyte dysfunction, and tubulointerstitial injury in IgAN. Mesangial deposition of IgA-IC leads to activation of mesangial cells (HMC), triggering mesangial cell proliferation and release of proinflammatory and profibrotic mediators including tumour necrosis factor alpha (TNF-α), transforming growth factor beta (TGF-β), interleukin 6 (IL-6), and angiotensin II (AngII). There is insignificant binding of IgA-IC to podocytes or tubular epithelial cells (TEC). Tumour necrosis factor-α released from mesangium after IgA deposition induces TNF-α synthesis by podocytes. Podocyte-derived TNF-α further upregulates the TNF-α production in an autocrine manner. TNF-α upregulates the expression of TNF-α receptors. The binding to TNF-α receptor 1 (TNFR1) leads to IL-6 synthesis and apoptosis, while binding to TNF-α receptor 2 (TNFR2) maintains proinflammatory cellular responses. Podocytes enhance interstitial damage in IgAN by amplifying the TEC activation with enhanced TNF-α synthesis. In the renal tubulointerstitium the interaction of AngII and angiotensin receptor subtype-1 (AGTR1) will lead to inflammatory responses through the upregulation of PKC and MAPK pathways. The activation of angiotensin receptor subtype 2 (AGTR2) leads to apoptosis through downregulation of the MAPK pathway. The aldosterone released from HMC following IgA-IC deposition acts synergistically with AngII to induce apoptosis in renal tubular epithelial cells. The mesangial-derived AngII maintains the tubulointerstitial injury. MR represents mineralocorticoid receptor. From Lai (2012) with permission with permission from Nature Publishing Group.

and circulating complexes in IgA nephropathy (Tomana et  al., 1997, 1999; Hiki et al., 2001). The synthesis of O-linked glycans of circulatory IgA1 follows a step-wise manner. O-Glycosylation of protein is initiated by the addition of GalNAc to serine or threonine residue through the activity of UDP-N-acetyl-α-D-galactosamine:  polypeptide N-acetylgalactosaminyltransferases (pp-GalNAc-Ts), which is followed by β-galactosylation through core 1 synthase 1, also called glycoprotein-N-acetyl-galactosamine 3-β-galactosyltransferase 1, core 1  β3-Gal-T, or T-synthase (C1GALT1). The moiety is galactosylated in the β1,3 configuration by core-1  β1-3 galactosyltransferase-1 (C1Gal-T1), which requires the presence of the core-1-β3-Gal-T-specific molecular chaperone (Cosmc). The formation of the glycan structure is accomplished by the α-2,6-sialyltransferase 2 (ST6GalNAc2) and α-2,3-sialyltransferase (ST3Gal) that attach sialic acid to the GalNAc and galactose residues, respectively. Sialic acid may also be added to terminal GalNAc by ST6GalNAc2. This represent a terminal step of O-glycan formation as the addition prevents any further modification of the molecule (Raska et al., 2007; Suzuki et al., 2008). Genes encoding C1GALT1, Cosmc (C1GALT1C1), and ST6GalNAc2 are located in chromosome 7p13–14, chromosome Xq24, and chromosome 17q25.1, respectively. Earlier Chinese and Italian studies reported risk haplotypes in ST6GALNAc2 and C1GALT1 for IgAN (Li et al., 2007; Pirulli et al., 2009; Zhu et  al., 2009). However, later studies revealed that raised serum galactose-deficient IgA1 was derived from IgA1-producing cells and the serum level correlated with that in supernatant of cultured IgA1-producing cells isolated from peripheral blood of the same IgAN patients (Suzuki et  al., 2008). These data suggest that the aberrant glycosylation in IgAN is an intrinsic and specific defect originated from IgA1-producing cells instead of attribution to

modification of IgA1 during the immune complexes formation. The expression and enzymatic activity of C1GALT1 appear to be dissociated from O-glycosylation in IgAN (Buck et al., 2008). Such aberrant glycosylation does not exist in other glycoproteins with O-linked glycans, such as IgD (Smith et al., 2006). These observations indicate the IgA1 glycosylation defect lies in the upstream regulatory pathway(s) rather than in glycosylation enzymes located in the downstream of IgA synthesis. However, a study by Suzuki et al. (2008) in EBV-immortalized cells from patients with IgAN demonstrating a decrease in C1Gal-T1 activity and an increase in ST6GalNAc2 favours premature sialylation. Undergalactosylated IgA1 molecules are prone to self-aggregate and to form complexes with IgG antibodies (Tomana et al., 1997, 1999). Epitopes at the hinge region now exposed in the absence of galactose are recognized by IgG and IgA1 with antiglycan specificities (Suzuki et al., 2009). IgA-IC are formed following the binding of glycan-specific IgG from IgAN patients with undergalactosylated IgA1. A possibility of mesangial IgA1-IC formation in situ following the initial deposition of IgA1 alone has also been raised. The binding of plasma polymeric IgA (pIgA) to human mesangial cells is charge dependent. pIgA from IgA nephropathy patients with the highest net anionic charge binds more to human mesangial cells. Pre-incubation with polyanion decreases the binding of pIgA1 to mesangial cells indicating the anionic charge of IgA1 plays an important role in mesangial deposition (Leung et al., 2001). The over-sialylation of the IgA increases the negative charge and also enhances steric hindrance to binding (Leung et al., 1999).

Immunoglobulin A receptors There are five known IgA receptors: FcαR1 (CD89), asialoglycoprotein receptor (ASGPR), polymeric Ig receptors (pIgR), transferrin

587

588

Section 3  

the patient with glomerular disease

receptor (TfR), and Fcα/μ receptor. The FcαR1 binds both the monomeric and dimeric forms of IgA1 and IgA2 (Monteiro et al., 1992; Morton et al., 1996). Transfection studies in leucocytes showed that the FcαR1 does not bind IgG (Reterink et  al., 1997). FcαR1 was originally found to be expressed by neutrophils, monocytes, macrophages, and eosinophils (Monteiro et al., 1990). It was proposed that FcαR1 plays a role in the removal of IgA-antigen complexes from the circulation (Grossetete et al., 1998). Previous works have suggested that mesangial cells possess Fc receptors for IgA (Mostov et al., 1984; Gomez-Guerrero et al., 1993). The pathogenic role of this receptor was suggested by the development of mesangial IgA deposits, glomerular and interstitial macrophage infiltration, haematuria and mild proteinuria in transgenic mice expressing human CD89 (Launay et al., 2000). Moura et al. (2008) proposed that, as a second event, activation of the classic, FcRγ-associated transmembrane FcαRI expressed on circulating myeloid leucocytes takes place. FcαRI/γ2 cross-linking in human FcαRI transgenic animals promotes disease progression by enhancing leucocyte chemotaxis and cytokine production. However, other investigators failed to demonstrate the expression of FcαR1 by human mesangial cells (Diven et al., 1998; Leung et al., 2000), despite the fact that mesangial cells showed Fc-dependent IgA binding that was saturable and dose dependent. The asialoglycoprotein receptor is a C-type lectin that recognizes galactose and N-acetylgalactosamine residues of desialylated glycoproteins and mediates endocytosis of serum glycoproteins (Stockert, 1995). The human ASGPR is an integral transmembrane glycoprotein composed of two units, H1 and H2. Although ASGPR is thought to be exclusively present in liver cells, several investigators have shown that mRNA for rat RHL-1 and RHL2/3 are widely expressed in different tissues and cell lines (Pacifico et al., 1995; Park et al., 1998). The absence of galactose in galactose-deficient IgA1 with its anionic charge reduces the hepatic clearance of IgA1 leading to increased serum IgA1 in IgA nephropathy (Leung et  al., 1999). Gomez-Guerrero et  al. (1998) first demonstrated that human and rat mesangial cells were able to specifically bind, internalize, and degrade iodine-125-labelled asialo-orosomucoid that was rich in terminal galactose. They also detected RHL-1 and RHL-2 transcripts in RNA extracted from rat mesangial cells. With IgA1 partially inhibiting the binding of asialo-orosomucoid to mesangial cells, they concluded that human mesangial cells expressed ASGPR. pIgR is an integral membrane secretory component localized on the basolateral surface of secretory epithelial cells. It mediates the transepithelial transport of pIg, particularly, pIgA (Piskurich et al., 1997). pIgR is detected in most human secretory epithelia (Krajci et al., 1989). The pIgR neutralizes extracellular and intracellular pathogens in mucous membranes by epithelial transport of pIgA-pathogen complexes and then excretes them via epithelial transcytosis (Mostov et al., 1984). Transferrin receptor (TfR) has been suggested as an IgA1 receptor as TfR binds IgA1 but not IgA2, co-localizes with mesangial IgA1 deposits, and is overexpressed in patients with IgA nephropathy (Moura et al., 2001, 2004). However, TfR is expressed ubiquitously in different kidney cells. Fcα/μ receptor was also examined as novel IgA1 receptor on mesangial cells based on in vitro culture studies (McDonald et al., 2002). Subsequent in-depth studies failed to detect these five IgA receptors in mesangial cells, podocytes or renal tubular epithelial cells (Leung et al., 2000; Chan et al., 2005;

Lai et al., 2008). Hence, the predominant binding of human IgA to human mesangial cells is mediated by other mechanisms yet to be revealed.

Immunoglobulin A immune complexes Early studies found high titres of IgA-ICs and a low frequency of IgG-ICs in 30–70% of adult and paediatric patients (Coppo et al., 1995). The former is present during acute mucosal infection and in remission, while the latter appears only in relapses. Circulating IgA-fibronectin aggregates are useful as a serologic marker for IgA nephropathy in Caucasian patients, but their values could not be reproduced in Chinese subjects (Lai et  al., 1996). More recently, soluble CD89–IgA complexes are identified as a potential new biomarker of risk of progression. In this Swedish study (Vuong et al., 2010), there was no difference in levels of sCD89–pIgA complexes between IgA nephropathy patients, healthy controls, and matched subjects with biopsy-proven glomerulonephritis from non-IgA nephropathy causes. However, within the IgA nephropathy group, there was a significant association between levels of sCD89–pIgA complexes and the likelihood of developing progressive renal disease. Such findings have implications for risk stratification and suggest a role for CD89 in the formation of pathogenic IgA-ICs in IgA nephropathy. Animal data suggest IgA1 interacts with CD89 on mononuclear cells that induces the release of sCD89 and the formation of IgA1–CD89 complexes. These complexes then interact with the transferrin receptor (CD71/TfR1) on mesangial cells and further enhance the expression of TfR1 via transglutaminase-2 (TGase2). TfR1 and TGase2 were both shown to bind sCD89, but also to directly interact with each other, providing an amplification step for IgA1 accumulation and inflammation in the kidney (Berthelot et al., 2012).

T and B cells in circulation Patients with IgA nephropathy have increased number of IgA-bearing B cells and activated Tα helper cells in circulation. They demonstrate increase of both circulating Th1 and Th2 T lymphocyte subsets (Lai et al., 1994a). The cytokine expression in these cells is characterized by a predominance of interleukin-4, interleukin-5, interleukin-10, and interleukin-13 belonging to Th2 cells (Lai et  al., 1994a; Ebihara et  al., 2001). The increased interleukin-4 production may explain hyperproduction of IgA while the enhanced interleukin-5 production favours the IgA isotype switch and differentiation (Lai et al., 1994b). The proportion of γδT cells in peripheral blood mononuclear cells is high in IgA nephropathy and correlates well with surface IgA-positive B cells. Both γδ− and CD4-positive T cells produce a large amount of transforming growth factor-β1, which induces IgA class switching on B cells (Lai et al., 1994c; Toyabe et al., 2001). Plasma IgA levels are determined by the rate of IgA production, uptake by leucocytes, and removal by hepatocytes. In IgA nephropathy, there is increased binding of endogenous IgA to circulating granulocytes and monocytes (Lai et al., 2002). FcαR1 expression on leucocytes is increased, independently of plasma IgA levels, as FcαR1 is not saturated in leucocytes, because of internalization of IgA after uptake. There is binding mechanism other than FcαR1 for pIgA uptake by leucocytes. Migration and/or sequestration of ‘activated’ leucocytes with predominant λ-IgA in the mononuclear

Chapter 69 

immunoglobulin a nephropathy: pathogenesis

Second or multiple hit So matic mutation, vial infection, genetic factor, autoimmune dysregulation

Production of anti-glycan antibodies

Mucosa-bone marrow axis

Inherited defect in B cell producing IgA1

Production of galactose-deficient IgA1

Production of IgA1 in bone marrow

Immune complex formation Decreased clearance of IgA1 by hepatocytes or reticuloendothelial system

Tonsils

Abnormality in mucosal immune system and/or abnormal systemic response to mucosal antigens

Glomerular IgA1 deposition

Inflammation with mesangial cells proliferation and complement activation

Glomerular sclerosis

Haematuria, proteinuria, renal failure

Fig. 69.3  Proposed pathways involved in mesangial IgA deposition in IgAN: multi-hit mechanism. Fundamental to immune complex formation is the enhanced synthesis of aberrant IgA1 with undergalactosylation (hit-1). Genetic factors heavily influence the production of undergalactosylated IgA1 and familial clustering has been well recognized. However, the presence these IgA1 O-glycoforms alone is insufficient to cause IgAN. The second hit is the formation of glycan-specific IgG and IgA antibodies that recognize the undergalactosylation IgA molecule (hit-2). These antibodies often with reactivity against antigens from extrinsic microorganisms may arise from recurrent mucosal infection (subsequent hits). Emerging evidence indicates that B cells in the mucosal infections, particularly in tonsillitis, may produce the nephritogenic IgA1. With increased immune complex formation and its decreased clearance, IgA1 (mainly polymeric in nature) binds to glomerular mesangium via yet unidentified receptor. Glomerular IgA1 deposits trigger the local production of cytokines and growth factors, leading to mesangial cell activation and complement activation.

phagocytic system or inflammatory tissues, after the initial binding of λ-pIgA occurs in IgA nephropathy.

Bone marrow–mucosa axis Most patients with IgA nephropathy have a higher memory repertoire of IgA-bearing B cells in the bone marrow. The displacement of mucosal B cells to systemic lymphoid organs and bone marrow may arise from abnormal trafficking of lymphocytes along the mucosa-bone marrow axis involving changes of chemokines and adhesion molecules (Yu et al., 2010). The connection between the bone marrow compartment and the mucosal immune system acts through the trafficking of antigen presenting cells and/or antigen-specific lymphocytes. An increased synthesis of both monomeric and pIgA1 occurs in IgA nephropathy with increased number of IgA1-producing plasma cells (van den Wall Bake et  al., 1988, 1989). A  shift towards IgA1 subclass is present in circulating IgA and mesangial deposits may originate from the bone marrow (Harper et al.,

1994a). High serum levels of IgA, IgA-immune complexes, and hyper-responsiveness of lymphocytes to antigens, in vitro and in vivo, are present in these patients. An abnormal systemic response to tetanus toxoid immunization has been demonstrated in IgA nephropathy (Layward et al., 1992). Animal studies suggest that bone marrow-derived Th1 cells initiate the disease activity and mucosal IgA responses to antigens are altered by Th2-biased background or dysregulation of innate immunity in this disease (Suzuki and Tomino, 2007; Suzuki et al., 2007). These data suggest that the excess marrow-derived IgA may be of ‘mucosal’ type pIgA, which has abnormal access to the circulation and hence the mesangium. Increased percentage and number of IgA-positive cells are found in tonsillar tissues of IgA nephropathy patients (Bene et al., 1991). The germinal centres of tonsils in these patients are constituted by follicular dendritic cells with preferential IgA1 localization (Kusakari et al., 1994). A reduced expression of J chain mRNA in duodenal IgA plasma cells is found in IgA nephropathy (Harper

589

590

Section 3  

the patient with glomerular disease

et al., 1994b). Clinical exacerbation of the disease with macrohaematuria is frequently associated with mucosal infection.

Glomerulo-podocyte-tubular crosstalk The mechanism by which mesangial IgA-triggered inflammation leads to the varied types and rates of glomerular lesion is still not well understood, but some hypotheses are summarized in Fig. 69.2 and reviewed by Lai (2012). Fig. 69.3 summarizes our view on the pathogenesis of IgA nephropathy.

References Bene, M. C., Hurault De, L. B., Kessler, M., et al. (1991). Confirmation of tonsillar anomalies in IgA nephropathy: a multicenter study. Nephron, 58, 425–8. Berthelot, L., Papista, C., Maciel, T.T., et al. (2012). Transglutaminase is essential for IgA nephropathy development acting through IgA receptors. J Exp Med, 209, 793–806. Buck, K. S., Smith, A. C., Molyneux, K., et al. (2008). B-cell O-galactosyltransferase activity, and expression of O-glycosylation genes in bone marrow in IgA nephropathy. Kidney Int, 73, 1128–36. Chan, L. Y., Leung, J. C., Tsang, A. W., et al. (2005). Activation of tubular epithelial cells by mesangial-derived TNF-alpha: glomerulotubular communication in IgA nephropathy. Kidney Int, 67, 602–12. Coppo, R., Amore, A., Gianoglio, B., et al. (1995). Macromolecular IgA and abnormal IgA reactivity in sera from children with IgA nephropathy. Italian Collaborative Paediatric IgA Nephropathy Study. Clin Nephrol, 43, 1–13. Diven, S. C., Caflisch, C. R., Hammond, D. K., et al. (1998). IgA induced activation of human mesangial cells: independent of FcalphaR1 (CD 89). Kidney Int, 54, 837–47. Ebihara, I., Hirayama, K., Yamamoto, S., et al. (2001). Th2 predominance at the single-cell level in patients with IgA nephropathy. Nephrol Dial Transplant, 16, 1783–9. Gomez-Guerrero, C., Duque, N., and Egido, J. (1998). Mesangial cells possess an asialoglycoprotein receptor with affinity for human immunoglobulin A. J Am Soc Nephrol, 9, 568–76. Gomez-Guerrero, C., Gonzalez, E., and Egido, J. (1993). Evidence for a specific IgA receptor in rat and human mesangial cells. J Immunol, 151, 7172–81. Grossetete, B., Launay, P., Lehuen, A., et al. (1998). Down-regulation of Fc alpha receptors on blood cells of IgA nephropathy patients: evidence for a negative regulatory role of serum IgA. Kidney Int, 53, 1321–35. Harper, S.J., Allen, A.C., Layward, L., et al. (1994a). Increased immunoglobulin A and immunoglobulin A1 cells in bone marrow trephine biopsy specimens in immunoglobulin A nephropathy. Am J Kidney Dis, 24, 888–92. Harper, S. J., Pringle, J. H., Wicks, A. C. et al. (1994b). Expression of J chain mRNA in duodenal IgA plasma cells in IgA nephropathy. Kidney Int, 45, 836–44. Hiki, Y., Odani, H., Takahashi, M., et al. (2001). Mass spectrometry proves under-O-glycosylation of glomerular IgA1 in IgA nephropathy. Kidney Int, 59, 1077–85. Krajci, P., Solberg, R., Sandberg, M., et al. (1989). Molecular cloning of the human transmembrane secretory component (poly-Ig receptor) and its mRNA expression in human tissues. Biochem Biophys Res Commun, 158, 783–9. Kusakari, C., Nose, M., Takasaka, T., et al. (1994). Immunopathological features of palatine tonsil characteristic of IgA nephropathy: IgA1 localization in follicular dendritic cells. Clin Exp Immunol, 95, 42–8. Lai, K. N., Chan, L. Y., Tang, S. C., et al. (2002). Characteristics of polymeric lambda-IgA binding to leukocytes in IgA nephropathy. J Am Soc Nephrol, 13, 2309–19.

Lai, K. N., Ho, R. T., Lai, C. K., et al. (1994a). Increase of both circulating Th1 and Th2 T lymphocyte subsets in IgA nephropathy. Clin Exp Immunol, 96, 116–21. Lai, K. N., Ho, R. T., Leung, J. C., et al. (1994b). CD4-positive cells from patients with IgA nephropathy demonstrate increased mRNA of cytokines that induce the IgA switch and differentiation. J Pathol, 174, 13–22. Lai, K. N., Ho, R. T., Leung, J. C., et al. (1994c). Increased mRNA encoding for transforming factor-beta in CD4+ cells from patients with IgA nephropathy. Kidney Int, 46, 862–8. Lai, K. N., Leung, J. C., Chan, L. Y., et al. (2008). Activation of podocytes by mesangial-derived TNF-alpha: glomerulo-podocytic communication in IgA nephropathy. Am J Physiol Renal Physiol, 294, F945–F955. Lai, K. N., To, W. Y., Leung, J. C., et al. (1996). Serologic study of immunoglobulin A-fibronectin aggregates in immunoglobulin A nephropathy. Am J Kidney Dis, 27, 622–30. Launay, P., Grossetete, B., Arcos-Fajardo, M., et al. (2000). Fcalpha receptor (CD89) mediates the development of immunoglobulin A (IgA) nephropathy (Berger’s disease). Evidence for pathogenic soluble receptor-Iga complexes in patients and CD89 transgenic mice. J Exp Med, 191, 1999–2009. Layward, L., Allen, A. C., Harper, S. J., et al. (1992). Increased and prolonged production of specific polymeric IgA after systemic immunization with tetanus toxoid in IgA nephropathy. Clin Exp Immunol, 88, 394–8. Leung, J. C., Poon, P. Y., and Lai, K. N. (1999). Increased sialylation of polymeric immunoglobulin A1: mechanism of selective glomerular deposition in immunoglobulin A nephropathy? J Lab Clin Med, 133, 152–60. Leung, J. C., Tang, S. C., Lam, M. F., et al. (2001). Charge-dependent binding of polymeric IgA1 to human mesangial cells in IgA nephropathy. Kidney Int, 59, 277–85. Leung, J. C., Tsang, A. W., Chan, D. T., et al. (2000). Absence of CD89, polymeric immunoglobulin receptor, and asialoglycoprotein receptor on human mesangial cells. J Am Soc Nephrol, 11, 241–9. Li, G. S., Zhang, H., Lv, J. C., et al. (2007). Variants of C1GALT1 gene are associated with the genetic susceptibility to IgA nephropathy. Kidney Int, 71, 448–53. McDonald, K. J., Cameron, A. J., Allen, J. M., et al. (2002). Expression of Fc alpha/mu receptor by human mesangial cells: a candidate receptor for immune complex deposition in IgA nephropathy. Biochem Biophys Res Commun, 290, 438–42. Monteiro, R. C., Cooper, M. D., and Kubagawa, H. (1992). Molecular heterogeneity of Fc alpha receptors detected by receptor-specific monoclonal antibodies. J Immunol, 148, 1764–70. Monteiro, R. C., Kubagawa, H., and Cooper, M. D. (1990). Cellular distribution, regulation, and biochemical nature of an Fc alpha receptor in humans. J Exp Med, 171, 597–613. Morton, H. C., van, E. M., and van de Winkel, J. G. (1996). Structure and function of human IgA Fc receptors (Fc alpha R). Crit Rev Immunol, 16, 423–40. Mostov, K. E., Friedlander, M., and Blobel, G. (1984). The receptor for transepithelial transport of IgA and IgM contains multiple immunoglobulin-like domains. Nature, 308, 37–43. Moura, I. C., Benhamou, M., Launay, P., et al. (2008). The glomerular response to IgA deposition in IgA nephropathy. Semin Nephrol, 28, 88–95. Moura, I. C., Centelles, M. N., Arcos-Fajardo, M., et al. (2001). Identification of the transferrin receptor as a novel immunoglobulin (Ig)A1 receptor and its enhanced expression on mesangial cells in IgA nephropathy. J Exp Med, 194, 417–25. Moura, I. C., Arcos-Fajardo, M., Sadaka, C., et al. (2004). Glycosylation and size of IgA1 are essential for interaction with mesangial transferrin receptor in IgA nephropathy. J Am Soc Nephrol, 15, 622–34. Pacifico, F., Laviola, L., Ulianich, L., et al. (1995). Differential expression of the asialoglycoprotein receptor in discrete brain areas, in kidney and thyroid. Biochem Biophys Res Commun, 210, 138–44.

Chapter 69 

Park, J. H., Cho, E. W., Shin, S. Y., et al. (1998). Detection of the asialoglycoprotein receptor on cell lines of extrahepatic origin. Biochem Biophys Res Commun, 244, 304–11. Pirulli, D., Crovella, S., Ulivi, S., et al. (2009). Genetic variant of C1GalT1 contributes to the susceptibility to IgA nephropathy. J Nephrol, 22, 152–9. Piskurich, J. F., Youngman, K. R., Phillips, K. M., et al. (1997). Transcriptional regulation of the human polymeric immunoglobulin receptor gene by interferon-gamma. Mol Immunol, 34, 75–91. Raska, M., Moldoveanu, Z., Suzuki, H., et al. (2007). Identification and characterization of CMP-NeuAc:GalNAc-IgA1 alpha2,6-sialyltransferase in IgA1-producing cells. J Mol Biol, 369, 69–78. Reterink, T. J., van Zandbergen, G., van Egmond, M., et al. (1997). Size-dependent effect of IgA on the IgA Fc receptor (CD89). Eur J Immunol, 27, 2219–24. Smith, A. C., de Wolff, J. F., Molyneux, K., et al. (2006). O-glycosylation of serum IgD in IgA nephropathy. J Am Soc Nephrol, 17, 1192–9. Stockert, R. J. (1995). The asialoglycoprotein receptor: relationships between structure, function, and expression. Physiol Rev, 75, 591–609. Suzuki, H., Fan, R., Zhang, Z., et al. (2009). Aberrantly glycosylated IgA1 in IgA nephropathy patients is recognized by IgG antibodies with restricted heterogeneity. J Clin Invest, 119, 1668–77. Suzuki, H., Moldoveanu, Z., Hall, S., et al. (2008). IgA1-secreting cell lines from patients with IgA nephropathy produce aberrantly glycosylated IgA1. J Clin Invest, 118, 629–39. Suzuki, H., Suzuki, Y., Aizawa, M., et al. (2007). Th1 polarization in murine IgA nephropathy directed by bone marrow-derived cells. Kidney Int, 72, 319–27.

immunoglobulin a nephropathy: pathogenesis

Suzuki, Y. and Tomino, Y. (2007). The mucosa-bone-marrow axis in IgA nephropathy. Contrib Nephrol, 157, 70–9. Tomana, M., Matousovic, K., Julian, B.A., et al. (1997). Galactose-deficient IgA1 in sera of IgA nephropathy patients is present in complexes with IgG. Kidney Int, 52, 509–16. Tomana, M., Novak, J., Julian, B. A., et al. (1999). Circulating immune complexes in IgA nephropathy consist of IgA1 with galactose-deficient hinge region and antiglycan antibodies. J Clin Invest, 104, 73–81. Toyabe, S., Harada, W., and Uchiyama, M. (2001). Oligoclonally expanding gammadelta T lymphocytes induce IgA switching in IgA nephropathy. Clin Exp Immunol, 124, 110–7. Van den Wall Bake, A. W., Daha, M. R., Haaijman, J. J., et al. (1989). Elevated production of polymeric and monomeric IgA1 by the bone marrow in IgA nephropathy. Kidney Int, 35, 1400–4. Van den Wall Bake, A. W., Daha, M. R., Radl, J., et al. (1988). The bone marrow as production site of the IgA deposited in the kidneys of patients with IgA nephropathy. Clin Exp Immunol, 72, 321–5. Vuong, M. T., Hahn-Zoric, M., Lundberg, S., et al. (2010). Association of soluble CD89 levels with disease progression but not susceptibility in IgA nephropathy. Kidney Int, 78, 1281–7. Yu, H. H., Chu, K. H., Yang, Y. H., et al. (2010). Genetics and immunopathogenesis of IgA nephropathy. Clin Rev Allergy Immunol, 41, 198–213. Zhu, L., Tang, W., Li, G., et al. (2009). Interaction between variants of two glycosyltransferase genes in IgA nephropathy. Kidney Int, 76, 190–8.

591

CHAPTER 70

Crescentic (rapidly progressive) glomerulonephritis Neil Turner Introduction Crescentic nephritis was recognized in the nineteenth century (Fig. 70.1). The term rapidly progressive glomerulonephritis (RPGN) was used by Ellis (1942) to describe patients who developed renal failure over weeks or months, most of whom were found to have crescentic nephritis.

Pathological characterization When immunofluorescence techniques became available, crescentic nephritis was divided into different types depending on fluorescence pattern of staining for immunoglobulins: linear, none, or granular. However, crescent formation can be a consequence of any aggressive nephritis, so it is more important to define the underlying disease than to subclassify crescentic nephritis. Discontinuation of this terminology was aided by the recognition that most patients with the condition previously described as pauci-immune, idiopathic crescentic nephritis have small vessel vasculitis (see Chapter 159). Pauci-immune is the most common histological finding in patients who have the clinical syndrome of RPGN (Table 70.1). Crescents originate in proliferation of epithelial cells, mostly derived from parietal epithelial cells lining Bowman’s capsule (see ‘Formation

of crescents’) (Fig. 70.2). This is termed extracapillary proliferation. In more severe disease, it extends circumferentially around the glomerulus so that in cross-section it has the classic crescentic form (Fig. 70.3). A key additional abnormality is the glomerular basement membrane (GBM) breaks that lead to crescent initiation, seen on silver stains or others that show the structure of the GBM clearly, and on electron microscopy. Other features depend on the initiating disease. Some diseases may have a stuttering focal nature so that crescents of different ages may be seen in the same biopsy. Most patients with small vessel vasculitis show this to some extent. There may be evidence of inactive focal scarring in affected glomeruli, and fibrotic partial or circumferential crescents. Crescents may also resolve without scarring in some circumstances (see below).

Clinical features of rapidly progressive glomerulonephritis There are two key characteristics to RPGN: ◆ Its

nature: it is from the ‘nephritic’ end of the spectrum of glomerulonephritis (see Chapter 45) so is characterized by marked haematuria and a variable amount of proteinuria

◆

Its tempo: kidney function declines over days or weeks (Fig. 70.4).

In acute and swiftly moving RPGN there may be visibly red urine, or it may be described as smoky or Coca-Cola urine. Blood clots are not a feature of glomerular bleeding. Microscopy shows glomerular-type dysmorphic red cells and usually also red cell casts (see Chapter 6). There may, however, be very little proteinuria if the disease is very acute. Nephrotic-range proteinuria usually implies that there has been an exacerbation of a pre-existing disease. The patient may complain of loin pain and kidneys may be moderately enlarged on imaging. Hypertension is not a prominent or early feature in most types of RPGN. Lupus nephritis and post-infectious glomerulonephritis are prominent exceptions. Hypertension is also likely when RPGN is caused by exacerbation of a prior nephritis, most commonly seen in immunoglobulin A (IgA) nephropathy.

Lung haemorrhage with RPGN

Fig. 70.1  Crescentic nephritis as illustrated by Volhard and Fahr (1914). The glomerular tuft shows proliferative changes, and a cellular crescent occupies the lower segment of Bowman’s capsule.

This combination encompasses the classic causes of pulmonary renal syndrome, although there is an important differential diagnosis with other causes of simultaneous pulmonary and renal failure (see Table 72.1 and Box 72.1 in Chapter 72). Alveolar and glomerular capillaries appear to share enough antigens and/or properties

Chapter 70 

crescentic (rapidly progressive) glomerulonephritis

for immune attack on both to occur in more than one disease. In anti-GBM disease the target is the basement membrane (see Chapter 72) and in small vessel vasculitis the target appears to be endothelium (see Chapter 158). Life-threatening disease with aggressive pulmonary and renal disease is most commonly seen in anti-GBM disease and small vessel vasculitis.

Investigations Although it may be possible to establish the aetiology from associated features, and from serological and other tests, in most cases a renal biopsy is urgently required. This is often the most rapid way to prove the diagnosis with certainty, and may also establish a likely prognosis which may affect decisions about treatment. Useful investigations are shown in Table 70.1.

Causes of crescentic nephritis Aggressive nephritis The major causes are shown in Table 70.1. They are inflammatory systemic diseases and types of glomerulonephritis from the ‘nephritic’ end of the spectrum (see Fig. 45.2, Chapter 45). Children, particularly in series from developing regions, are more likely to have post-infectious disease, but the range of causes is similar to that in adults (Southwest Pediatric Study Group, 1985; Jardim et al., 1992; Srivastava et al., 1992; Dewan et al., 2008; Sinha et al., 2013). Table 70.1  Distinguishing characteristics of key causes of RPGN. Frequency in different series taken from Heilman (1987), Keller et al. (1989), Andrassy et al. (1991), Angangco et al. (1994), and Levy and Pusey (2005) Disease

Frequency

Tests

Comments

Small vessel vasculitis

43–68%

MPO and Pr3 More specific antibodies than ANCA by immunofluorescence. Usually positive in the most frequent primary vasculitides. Note that false positive results occur. Renal biopsy confirms (Chapter 159)

Anti-GBM

6–20%

Anti-GBM antibodies

Lupus nephritis

3–8%

Complement Anti-ds DNA almost (C3, C4) always positive Anti-ds DNA, (Chapter 162) ANF

Post-infectious glomerulonephritis

0–13%

Complement Complement is the most valuable clinical Throat or test. Anti-streptococcal other swab titres can be historic (Chapter 77)

Other glomerulonephritis

2–23%

Fig. 70.2  A small segmental crescent in a patient with IgA nephropathy (haematoxylin and eosin).

There are some informative exceptions where crescent formation is encountered in the context of diseases not usually characterized by aggressive inflammation.

Nephrosis transforming into aggressive nephritis Membranous nephropathy rarely transforms into crescentic nephritis. About half of reported cases have anti-GBM disease (discussed

Always positive in florid disease (Chapter 72). Renal biopsy is the most sensitive and specific test and often also the most rapid.

Features of that nephritis

Fig. 70.3  A large cellular crescent surrounds and partially compresses the glomerular tuft in a patient with ANCA-associated vasculitis (methenamine silver–haematoxylin and eosin). GBM breaks can be seen.

593

Section 3  

the patient with glomerular disease Renal biopsy showing 12% crescents Treatment

1200

Dialysis

1000 Creatinine (µmoL/L)

594

800 600 400 200 0

0

10

20 Days

30

40

Fig. 70.4  Deterioration of kidney function in crescentic nephritis caused by anti-GBM disease. This is typically the most rapidly progressive RPGN and exacerbation of established disease by infection is well recognized (Rees et al., 1977; Guillen et al., 1995; Erwig et al., 2001).

in Chapters  61 and 74). Diabetes mellitus may also occasionally transform into anti-GBM disease. In both nephropathies a thickened GBM contains increased amounts of the collagen 345 network that includes the Goodpasture antigen. This plus some ‘second hit’ may enable breaking tolerance to the antigen (see Chapter 74).

Disorders associated with GBM fragility Several disorders leading to GBM fragility other than through nephritis have been occasionally reported to develop crescentic changes, for example, in amyloidosis (Nagata et  al., 2001; Schafernak et  al., 2005), in experimental models of Alport syndrome and rarely in humans (Ryu et  al., 2012), and in fibrillary nephritis (Sethi et al., 2001).

Formation of crescents Anti-GBM antibodies define the best-characterised, archetypal crescentic nephritis, but animal experiments clearly define that the characteristic histopathological changes are dependent on cell mediated immunity including macrophages and lymphocytes. Fig. 70.5 illustrates the steps in crescent formation in cartoon form.

Breaks in the GBM Breaks in the glomerular basement membrane allow plasma proteins to enter Bowman’s space. As described above this is usually caused by aggressive immune attack but occasionally crescents are seen when GBM breaks spontaneously because of abnormal composition. Fig. 70.6 shows a scanning electron micrograph of rat capillary basement membranes after induction of nephrotoxic nephritis, a model of anti-GBM disease (Bonsib, 1985, 1988).

Fibrin provokes proliferation of parietal epithelial cells The formation of fibrin in Bowman’s space seems to be critical; crescent formation can be reduced or blocked by defibrination with the snake venom ancrod or by fibrinolysis with urokinase (Naish et al., 1972; Ryu et al., 2012), and is diminished in mice lacking fibrinogen (Drew et al., 2001). Tissue factor is presumed

to trigger activation of fibrin formation and has been considered as a therapeutic target, but anticoagulation after a renal biopsy in patients who may have lung haemorrhage would be very hazardous (McCluskey, 1997). Furthermore it is not clear that preventing crescent formation would be enough, if the process causing the underlying glomerular damage continues (Browne et  al., 2004). Fig. 70.7 shows an early, fibrin-containing lesion in human disease. The large cells that participate in the first phase of crescent formation are mostly derived from parietal epithelial cells lining Bowman’s capsule, with only occasional cells that seem to have a podocyte origin (Smeets et al., 2009). Indeed this was the origin first suggested by Langerhans (1885). A similar proliferative response of parietal epithelial cells (‘the forgotten fourth cell of the glomerulus’) is seen in collapsing glomerulopathy and can be induced by injury to parietal epithelial cells themselves (Sicking et al., 2012; Shankland et al., 2014).

Macrophage infiltration and cell mediated immunity There are many macrophages in more mature crescents (Tipping et al., 1986; Lan et al., 1997) and these are important for mediating injury in many experimental models. It is also clear that crescentic nephritis is dependent on lymphocytes and other effectors of cell-mediated immunity, and that factors that modulate cellular infiltration or inflammatory state (interleukin 1, tumour necrosis factor alpha, and others) will modulate the injury caused by a particular immune insult (Holdsworth et al., 1999; Kitching et al., 2000; Kitching and Holdsworth, 2011; and see Fig. 70.4). Most of these studies have been carried out in animal models of anti-GBM disease.

Breaks in Bowman’s capsule In the final stage of the most destructive disease the damage extends to Bowman’s capsule, where breaks seem associated with influx of further cells including fibrocytes. It has been suggested that this determines whether a glomerulus affected by crescentic change recovers or fibroses (Boucher et al., 1987; Lan et al., 1997).

Resolution of crescents It is clear from reports of outcomes in post-streptococcal glomerulonephritis, from some patients with vasculitis, and rare patients with anti-GBM disease, that recovery can occur despite high proportions of crescents in renal biopsies. Post-streptococcal disease in children probably has the greatest propensity to do this. It has been termed ‘pneumonia of the glomerulus’, and like pneumococcal pneumonia, can be associated with good recovery and repair despite severe functional impairment and extreme histological hypercellularity and infiltration (Roy et al., 1981). In anti-GBM disease, crescent score usually predicts the outcome quite accurately (Fig. 70.8). This may be because this is such a destructive diseases that most crescentic glomeruli are irrecoverable, and sclerose. However, one individual in Fig. 70.8 recovered to a serum creatinine of 200  μmol/L at 1  year despite 86% crescents.

Management of crescentic nephritis The management depends on the cause, but there is some urgency, so it is important to reach a rapid diagnosis. As suggested in

Chapter 70 

(A)

crescentic (rapidly progressive) glomerulonephritis (B)

GBM breaks; fibrin leaks (C)

(D)

Parietal cell proliferation

(E)

Inflammatory cell migration (F)

Bowman’s capsule perforated

Macrophage infiltration (and other cells) (G)

Influx of fibrocytes

Fig. 70.5  Mechanism of crescent formation. A. Three schematic capillary loops. B. Damage to GBM mediated by inflammatory cells, antibodies with complement, etc. C. Fibrin formation from local tissue factor acting on fibrinogen promotes proliferation of parietal epithelial cells. D. Macrophages and later lymphocytes, fibrocytes migrate in, leading to further chemokine release. E. Macrophages found in crescents. F. Bowman’s capsule perforated. G. Further migration of cells from periglomerular space, including fibrocytes.

Table 70.1, renal biopsy is often the most informative test. It not only determines the cause, but it also gives prognostic information (number of affected glomeruli, age of lesions, degree of architectural damage, and presence or absence of interstitial fibrosis). As many of the therapies for crescentic nephritis involve significant risk, this is important information. If you can be certain that the patient does not have an infection underlying their disease, and that RPGN is the diagnosis, therapy with cyclophosphamide and prednisolone may be commenced in advance of a firm diagnosis.

elevated by the practice. There is some experimental support for use of high-dose corticosteroids alone, but studies have tended to be short and look more at histological appearances than outcome (Yamamoto-Shuda et al., 1999; Ou et al., 2001). It undoubtedly does ‘reduce the appearance of crescents’ and reduce leucocyte infiltration. It may improve outcome when used in high dose as sole therapy, but it is associated with risk. Plasma exchange is proven to add benefit in small vessel vasculitis and is accepted therapy in anti-GBM disease (see Chapters 73 and 160).

Specific treatments

Specific diseases

Pulses of methylprednisolone have been used in these circumstances following reports by Bolton and Sturgill (1989). There is little good evidence that this is any more effective than more moderate doses, and as patients are often treated with other immunosuppressive agents the risk of infection is probably significantly

In IgA nephropathy there is little evidence that immunosuppressive agents add additional benefits above those of high-dose steroids (see Chapter 68). In anti-GBM disease, a significant proportion of patients have no realistic chance of salvaging renal function (Fig. 70.8) (see

595

Section 3  

the patient with glomerular disease ≥1000

Creatine (µmoL/L)

596

800 600 400 200 0

0

20

40

60

80

100

Crescents (%) Independent renal function at 1 year On dialysis at 1 year Dead at 1 year

Fig. 70.6  Scanning electron micrograph of glomerular capillary loops of an animal with experimental anti-GBM disease, from Bonsib (1985). The glomerulus has been denuded of cells to show only the fixed GBM. Following anti-GBM antibody binding, fixation of complement and recruitment of cell-mediated effectors has blown holes in the GBM allowing serum proteins and red blood cells to leak into the urinary space. In most human nephritis GBM breaks are more subtle than this.

Chapter 73). If they do not have lung haemorrhage, a decision not to treat with immunosuppressive agents may be considered. In systemic vasculitis, the pace of the disease is highly variable, and many patients are elderly and have comorbid conditions which increases the risk of many therapies. Therapeutic strategies need to take this into account (see Chapter 160).

Fig. 70.7  A cellular segmental crescent is seen at 6 o’clock in this patient with Henoch–Schönlein purpura. In the centre of the picture fibrin can be seen leaking into the urinary space from a break in the GBM, stimulating the proliferation of parietal epithelial cells.

Fig. 70.8  Plasma creatinine concentration at presentation, and the proportion of glomeruli with crescents, in 38 patients treated at Hammersmith Hospital, London. Those who did not receive the combination of plasma exchange, cyclophosphamide, and prednisolone, or who had < 10 glomeruli in their renal biopsies, have been excluded. The following are illustrated: (1) the correlation between the creatinine at presentation and histological evidence of glomerular damage, except in one patient with acute tubular necrosis; (2) the close relationship between the severity of renal damage at presentation and outcome; (3) death from pulmonary haemorrhage occurs predominantly in those with severe renal disease.

References Andrassy, K., Kuster, S., Waldherr, R., et al. (1991). Rapidly progressive glomerulonephritis: analysis of prevalence and clinical course. Nephron, 59 (2), 206–12. Angangco, R., Thiru, S., Esnault, V. L., et al. (1994). Does truly ‘idiopathic’ crescentic glomerulonephritis exist? Nephrol Dial Transplant, 9, 630–6. Bolton, W. K. and Sturgill, B. C. (1989). Methylprednisolone therapy for acute crescentic rapidly progressive glomerulonephritis. Am J Nephrol, 9, 368–75. Bonsib, S. M. (1985). GBM discontinuities. Scanning electron microscopic study of a cellular glomeruli. Am J Pathol, 119, 357–60. Bonsib, S. M. (1988). GBM necrosis and crescent organization. Kidney Int, 33, 966–74. Boucher, A., Droz, D., Adafer, E., et al. (1987). Relationship between the integrity of Bowman’s capsule and the composition of cellular crescents in human crescentic glomerulonephritis. Lab Invest, 56, 526–33. Browne, G., Brown, P. A., Tomson, C. R., et al. (2004). Retransplantation in Alport post-transplant anti-GBM disease. Kidney Int, 65, 675–81. Dewan, D., Gulati, S., Sharma, R. K., et al. (2008). Clinical spectrum and outcome of crescentic glomerulonephritis in children in developing countries. Pediatr Nephrol, 23, 389–94. Drew, A. F., Tucker, H. L., Liu, H., et al. (2001). Crescentic glomerulonephritis is diminished in fibrinogen-deficient mice. Am J Physiol Renal Physiol, 281, F1157–63. Ellis, A. (1942). Natural history of Bright’s disease. Clinical, histological and experimental observations. Lancet, i, 34–6. Erwig, L. P., Kluth, D. C., and Rees, A. J. (2001). Macrophages in renal inflammation. Curr Opin Nephrol Hypertens, 10, 341–7. Guillen, E. L., Ruiz, A. M., Fernandez, M. A., et al. (1995). Goodpasture syndrome: re-exacerbations associated with intercurrent infections. Revista Clinica Española, 195, 761–4. Holdsworth, S. R., Kitching, A. R., and Tipping, P. G. (1999). Th1 and Th2 helper cell subsets affect patterns of injury and outcomes in glomerulo-nephritis. Kidney Int, 55, 1198–216. Jardim, H. M., Leake, J., Risdon, R. A., et al. (1992). Crescentic glomerulonephritis in children. Pediatr Nephrol, 6, 231–5.

Chapter 70 

crescentic (rapidly progressive) glomerulonephritis

Keller, F., Oehlenberg, B., Kunzendorf, U., et al. (1989). Long-term treatment and prognosis of rapidly progressive glomerulonephritis. Clin Nephrol, 31, 190–7. Kitching, A. R. and Holdsworth, S. R. (2011). The emergence of TH17 cells as effectors of renal injury. J Am Soc Nephrol, 22(2), 235–8. Kitching, A. R., Holdsworth, S. R., and Tipping, P. G. (2000). Crescentic glomerulonephritis—a manifestation of a nephritogenic Th1 response? Histol Histopathol, 15, 993–1003. Lan, H. Y., Nikolic-Paterson, D. J., Mu, W., et al. (1997). Local macrophage proliferation in the pathogenesis of glomerular crescent formation in rat anti-glomerular basement membrane (GBM) glomerulonephritis. Clin Exp Immunol, 110, 233–40. Langerhans, T. (1885). Über die entzundlichen Veranderungen der Glomeruli und die acute Nephritis. Virchows Archiv, 99, 193–204. Levy, J. and Pusey, C. D. (2005). Crescentic glomerulonephritis. In J. S. Cameron, A. M. Davison AM, J. -P. Grunfeld, et al. (eds.) Oxford Textbook of Clinical Nephrology (3rd ed), pp. 559–78. Oxford: Oxford University Press. McCluskey, R. T. (1997). Tissue factor in crescentic glomerulonephritis. Am J Pathol, 150, 787–92. Nagata, M., Shimokama, T., Harada, A., et al. (2001). Glomerular crescents in renal amyloidosis: an epiphenomenona or distinct pathology? Pathol Int, 51, 179–86. Naish, P., Penn, G. B., Evans, D. J., et al. (1972). The effect of defibrination on nephrotoxic serum nephritis in rabbits. Clin Sci, 42, 643. Ou, Z. L., Nakayama, K., Natori, Y., et al. (2001). Effective methylprednisolone dose in experimental crescentic glomerulonephritis. Am J Kidney Dis, 37, 411–17. Rees, A. J., Lockwood, C. M., and Peters, D. K. (1977). Enhanced allergic tissue injury in Goodpasture’s syndrome by intercurrent bacterial infection. Br Med J, 2, 723–6. Roy, S., Murphy, W. M., and Arant, B. S. (1981). Post-streptococcal glomerulonephritis in children: comparison of quintuple therapy versus supportive care. Pediatrics, 98, 403–10.

Ryu, M., Migliorini, A., Miosge, N., et al. (2012). Plasma leakage through glomerular basement membrane ruptures triggers the proliferation of parietal epithelial cells and crescent formation in non-inflammatory glomerular injury. J Pathol, 228, 482–94. Schafernak, K. T., Chugh, S. S., and Kanwar, Y. S. (2005). Co-existent crescentic glomerulonephritis and renal amyloidosis: a case report and literature review. J Nephrol, 18, 616–22. Sethi, S., Adeyi, O. A., and Rennke, H. G. (2001). A case of fibrillary glomerulonephritis with linear immunoglobulin G staining of the glomerular capillary walls. Arch Pathol Lab Med, 125(4), 534–6. Shankland, S. J., Smeets, B., Pippin, J. W., et al. (2014). The emergence of the glomerular parietal epithelial cell. Nat Rev Nephrol, 10, 158–73. Sicking, E. M., Fuss, A., Uhlig, S., et al. (2012). Subtotal ablation of parietal epithelial cells induces crescent formation. J Am Soc Nephrol, 23, 629–40. Sinha, A., Puri, K., Hari, P., et al. (2013). Etiology and outcome of crescentic glomerulonephritis. Indian Pediatr, 50(3), 283–8. Smeets, B., Uhlig, S., Fuss, A., et al. (2009). Tracing the origin of glomerular extracapillary lesions from parietal epithelial cells. J Am Soc Nephrol, 20, 2604–15. Southwest Pediatric Nephrology Study Group (1985). A clinico- pathologic study of crescentic glomerulonephritis in children. A report of the Southwest Pediatric Nephrology Study Group. Kidney Int, 27, 450–8 Srivastava, R. N., Moudgil, A., Bagga, A., et al. (1992). Crescentic glomerulonephritis in children: a review of 43 cases. Am J Nephrol, 12, 155–61. Tipping, P. G. and Holdsworth, S. R. (1986). The participation of macrophages, glomerular procoagulant activity, and factor VIII in glomerular fibrin deposition. Studies on anti-GBM antibody-induced glomerulonephritis in rabbits. Am J Pathol, 124, 10–17. Volhard, F. and Fahr, T. (1914). Die Brightsche Nierenkrankheit. Berlin: Springer. Yamamoto-Shuda, Y., Nakayama, K., Saito, T., et al. (1999). Therapeutic effect of glucocorticoid on experimental crescen- tic glomerulonephritis. J Lab Clin Med, 134, 410–18.

597

CHAPTER 71

Antiglomerular basement membrane disease: overview Zhao Cui, Neil Turner, and Ming-hui Zhao Antiglomerular basement membrane (anti-GBM) disease is characteristically the most rapidly progressive (crescentic) nephritis. It is often accompanied by lung haemorrhage, and occasionally causes lung disease alone (see Chapter 72). Its hallmark is linear deposition of immunoglobulin G (IgG) along the GBM. There are usually few systemic symptoms apart from any related to the lung disease. Urine shows haematuria, often macroscopic in very acute disease. Diagnosis depends on identification of anti-GBM antibodies in association with tissue damage (Fig. 71.1) (see Chapter 72). Renal biopsy is important for confirming the diagnosis and gives important prognostic information. Early treatment (see Chapter 73) with cyclophosphamide, plasma exchange, and prednisolone arrests lung haemorrhage and can salvage renal function, but the disease often progresses very rapidly so that renal destruction is advanced by the time the diagnosis is made, and renal recovery partial or absent. The value of alternative therapies is unproven. Anti-GBM disease tends to be an acute disease not requiring long-term immunosuppression. Anti-GBM antibody formation and/or disease may be provoked by small vessel vasculitis affecting the kidney, by lithotripsy, and occasionally in other circumstances. There is a strong association with class II human leucocyte antigen (HLA)-DR15 and -DR4 (see Chapter 74). The disease is caused by autoimmunity to the carboxy-terminal (NC1) domain of one of the tissue-specific basement membrane (type IV) collagen chains, α3(IV)NC1 (see Chapter 74). The antigen is also found in the alveolus, and causes lung haemorrhage in about half of patients with the disease (see Chapter 73). This can be life-threatening and associated with severe renal disease, but it can also occur with minimal renal disease. The antigen in lung is cryptic:  additional insults are required to expose it to the immune system leading to lung haemorrhage. Most patients who develop lung haemorrhage in anti-GBM disease are cigarette smokers. Alport post-transplant anti-GBM disease (see Chapter 75) is a rare post-transplant complication that occurs in a small minority of patients with Alport syndrome. It appears clinically very like spontaneous anti-GBM disease, but causes isolated kidney disease. The target of antibodies is usually the type IV collagen chain that carries the Alport mutation. It is difficult to treat.

(A)

C-terminals of adjacent collagen (IV) molecules (B)

N

NC1 Hexamer

NC1 Dimers, monomers

Collagenase

SDS-PAGE

A

G

G 60 kDa

Dimers 40 kDa 30 kDa Monomers 25 kDa

Fig. 71.1  (A) Western blotting of tissue-specific type IV collagen NC1 domains. Collagenase-solubilised basement membrane from bovine testis has been separated by SDS–PAGE in non-reducing conditions and transferred to nitrocellulose. (See also Chapter 74, Fig. 74.2.) (B) Goodpasture sera (lanes marked ‘G’) recognize monomers of α3(IV) NC1 domains of approximately 28 kDa and a cluster of dimers. Some Alport post-transplant anti-GBM sera (e.g. the lane marked ‘A’) contain antibodies to α5(IV)NC1, which can be seen as a monomer of approximately 26 kDa. Normal sera (‘N’) show no monomer recognition but weak staining of dimer regions after prolonged incubation.

CHAPTER 72

Antiglomerular basement membrane disease: clinical features and diagnosis Zhao Cui, Neil Turner, and Ming-hui Zhao Introduction The name Goodpasture was first applied in 1958 by Stanton and Tange (1958), who described a group of nine patients with similar clinical features. They recognized these as similar to an 18-year-old man presenting with lung haemorrhage and crescentic nephritis at autopsy during an epidemic of influenza, reported by Ernest Goodpasture in 1919 (Goodpasture, 1919). The term ‘Goodpasture syndrome’ is sometimes used to describe the combination of severe glomerulonephritis and lung haemorrhage, irrespective of cause. Actually, the pulmonary-renal syndrome is more commonly caused by systemic vasculitis, usually associated with antineutrophil cytoplasmic antibodies (ANCAs), but with a wider range of possible causes described below. Patients with pulmonary-renal syndrome and pathogenic antibodies against GBM are best described as anti-GBM disease, but the description ‘Goodpasture disease’ is still in use. In 1967, the pathogenicity of autoantibodies against GBM was demonstrated by the antibody transfer experiment of Lerner, Glassock, and Dixon (Lerner et al., 1967). Now, Goodpasture disease is known to be caused by autoimmunity to a target on the non-collagenous domain 1 of the α3 chain of type IV collagen (α3(IV)NC1), present in alveolar and glomerular basement membranes, and now commonly described as the Goodpasture antigen.

Epidemiology Anti-GBM disease is rare, with an estimated and possibly rising incidence of up to 1 case per million of the population per annum (Savage et al., 1986). It occurs across all racial groups, but not equally affected. Although there is no large study of incidence for Asian populations, cases have been reported from Japan (47 cases from 715 patients with rapidly progressive glomerulonephritis (RPGN)) (Hirayama et al., 2008) and China (221 cases reported, accounting for 16% of crescentic nephritis) (Li et al., 2004; Lin et al., 2010; Cui et al., 2011a). It seems to be particularly rare in black races, although cases in black Americans have been described (Kelly and Haponik, 1994). The disease has been estimated to cause up to 5% of glomerulonephritis (Wilson and Dixon, 1973), 10–20% of crescentic glomerulonephritis (Couser, 1988; Andrassy et al., 1991; Lin et al., 2010), and 2% of end-stage renal disease (Disne, 1986).

All age groups can be affected by the disease. The youngest reported case appears to be 11 months old (Bigler et al., 1997); several patients in their 90s have been described (Cui et al., 2011b). Generally, there are two periods of peak incidence. The first peak is in the second and third decades of life, the second peak is in the sixth and seventh decades (Kluth and Rees, 1999; Pusey, 2003). Early series showed a striking preponderance of young male patients with high frequency of pulmonary haemorrhage. The wide application of immunoassays and immunohistology, and increased awareness of the disease, has led to later series showing a greater proportion of older patients (Savage et al., 1986; Herody et al., 1993; Merkel et al., 1994; Daly et al., 1996; Levy et al., 2001; Fischer and Lager, 2006; Hirayama et al., 2008; Cui et al., 2011b). Glomerulonephritis alone is more common in older patients. At the time of diagnosis, their renal dysfunction is mild or moderate, but the outcome is similar to younger patients. This age and gender distribution is notably different to that of other organ-specific autoimmune disorders. Lung haemorrhage is more common in men (Fig. 72.1). Glomeru­ lonephritis alone is more common in women. The male:female ratio in 71 cases from the United Kingdom was 1.4, but for those with both pulmonary haemorrhage and glomerulonephritis it was 3.0, and for those with glomerulonephritis alone it was 0.9 (Levy et al., 2001). These differences were probably related to smoking history. Anti-GBM disease has been associated with a number of other conditions, the most common shown in Table 72.1. Some of these may be causally related, inducing the anti-GBM response. Probably examples of this include small vessel vasculitis (common), and rarely lithotripsy, urinary tract obstruction, and perhaps other examples. An association with lymphocyte depletion (e.g. in alemtuzumab therapy and HIV) may also give clues to the natural control of immunity to this autoantigen (see Chapter 74).

Clinical features Patients may present with renal disease alone, with renal disease plus lung haemorrhage, or occasionally with lung haemorrhage alone. Although there are hints that the disease may have a long prodrome, presentation is usually acute following an accelerated phase. The prodromal phase can be seen in some patients with low titres of antibodies associated with pulmonary haemorrhage alone,

Section 3  

the patient with glomerular disease

Number of patients

(A) 15

(B)

Male Female

10

5

0

Number of patients

600

0–10 11–20 21–30 31–40 41–50 51–60 61–70 >70 Age (years)

20

+ Pulmonary haemorrhage – Pulmonary haemorrhage

10

0

0–10 11–20 21–30 31–40 41–50 51–60 61–70 >70 Age (years)

Fig. 72.1  (A) Age and gender of 68 patients with Goodpasture disease treated at Hammersmith Hospital, London. (B) Pulmonary haemorrhage in 68 patients with Goodpasture disease treated at Hammersmith Hospital, London.

that sometimes progresses, and in historic studies of antibody titres (see Chapter 74).

General manifestations The clinical features of anti-GBM disease have been well established from the relatively large series of immunologically defined cases published over the last 40 years (Wilson and Dixon, 1973; Beirne et al., 1977; Teague et al., 1978; Briggs et al., 1979; Peters et al., 1982; Johnson et al., 1985; Walker et al., 1985; Savage et al., 1986; Herody et al., 1993; Merkel et al., 1994; Daly et al., 1996; Levy et al., 2001; Shah and Hugghins, 2002; Segelmark et al., 2003; Li et al., 2004; Cui Table 72.1  Diseases recurrently associated with Goodpasture disease. Alport syndrome following transplantation is bracketed as it has some notable differences (see Chapter 75) Disease

Number of reports (approx.)

ANCA-associated vasculitis (mostly with anti-myeloperoxidase ANCA)

Hundreds

Membranous nephropathy

< 20

Diabetes mellitus

< 20

Malignancy—lymphoma, bronchial carcinoma

10

Lithotripsy to intrarenal stones

< 10

Lymphocyte depletion (alemtuzumab, HIV)

< 10

(Alport syndrome following renal transplantation)

(Tens)

et al., 2005, 2011a; Fischer and Lager, 2006; Hirayama et al., 2008). Systemic symptoms, such as malaise, fever, or weight loss, are less frequently seen and generally mild. Anaemia is common and frequently symptomatic, even in patients who have had little or no haemoptysis. Anti-GBM disease should be considered in patients who present with moderate to severe anaemia together with acute kidney injury. The anaemia is usually microcytic and hypochromic, but sometimes shows evidence of microangiopathy as can be the case with other types of RPGN. The iron deficiency probably reflects subclinical pulmonary haemorrhage, but can on occasion be confused with gastrointestinal disease, especially if uraemia is causing nausea and vomiting.

Renal manifestations Abnormalities of the urine sediment, usually microscopic haematuria, are the earliest sign of renal damage. Later, the urine contains numerous dysmorphic red blood cells and red cell casts. In severe disease, macroscopic haematuria can occur. Proteinuria is generally mild or moderate (< 3.5 g/24 hours), but some patients may have severe proteinuria and present with nephrotic syndrome. Oliguria is a late feature and a bad prognostic sign. However, the chance of superimposed acute tubular necrosis in hypoxic and severely ill patients is always high. There may be loin pain when inflammation is severe. Hypertension is generally a late feature that accompanies advanced renal failure and fluid retention. Kidney size is usually normal or enlarged due to inflammation. Disease onset is typically abrupt with oliguria or anuria, haematuria and proteinuria, and end-stage renal disease. Renal function is usually already reduced at presentation and may deteriorate from normal to dialysis requiring levels in a matter of days to weeks. Several studies found that a subgroup of patients, 3–36%, has normal renal function or only minimal renal dysfunction and mild glomerular lesions (Mathew et al., 1975; Zimmerman et al., 1979; Bailey et al., 1981; Bell et al., 1990; Knoll et al., 1993; Min et al., 1996; Ang et al., 1998; Cui et al., 2007). These patients present mainly with lung haemorrhage, with varying degree of haematuria and proteinuria, but macroscopic haematuria and nephrotic range proteinuria are rare. During follow-up, renal function is preserved in most of this subgroup, although slow progression to renal failure has been seen in some cases, and a typical catastrophic deterioration has occurred after an interval in others.

Renal pathology An adequate renal biopsy is an essential part of the assessment of patients with the disease, and has prognostic as well as diagnostic importance. The pathologic finding of linear staining of immunoglobulins along glomerular capillary wall by direct immunofluorescence is indicative of anti-GBM glomerulonephritis (Fig. 72.2). This is predominantly IgG, however, rare patients with IgA-dominant anti-GBM glomerulonephritis have also been reported (Border et al., 1979; Fivush et al., 1986). Most specimens (60–70%) have discontinuous linear or granular capillary wall staining for C3, but a minority has little or no C3 staining. Occasional reports mention IgM alone, IgA alone, or C3 alone (Savage et al., 1986). Linear staining for IgG may also occur along tubular basement membranes in some but not all cases. Fibrin-related antigens are commonly present within the crescents and segmental necrotizing lesions.

Chapter 72 

anti-gbm disease: clinical features and diagnosis

and chronic lesions; however, the glomerular lesions of anti-GBM glomerulonephritis tend to be more in synchrony than those of ANCA-glomerulonephritis, which more often show admixtures of acute and chronic injury. Tubulointerstitial changes are commensurate with the degree of glomerular injury. Glomeruli with extensive necrosis and disruption of Bowman capsule typically have intense periglomerular inflammation, including occasional multinucleated giant cells (Fig. 72.3). There also is focal tubular epithelial acute simplification or atrophy, focal interstitial oedema and fibrosis, and focal interstitial infiltration of predominantly mononuclear leucocytes.

(A)

Fig. 72.2  Direct immunofluorescence for IgG in a typical glomerulus from a patient with anti-GBM disease, showing linear fixation of antibody to the GBM (FITC anti-IgG).

Additional granular deposits, associated with subepithelial or sometimes intramembranous or subendothelial deposits ultrastructurally, have been noted in some patients (Rajaraman et al., 1984) and may be associated with resolution. However, there are clear-cut examples of patients with membranous nephropathy developing anti-GBM disease, so this circumstance should be considered in patients with a history of nephrotic syndrome or heavy proteinuria. The glomeruli may be abnormal even in patients with normal renal function. The earliest and mildest changes consist of segmental mesangial matrix expansion and hypercellularity, progressing to a more generalized, but still focal and segmental, proliferative glomerulonephritis with increased numbers of neutrophils in the glomeruli. Later, glomeruli show a diffuse glomerulonephritis with segmental or total necrosis and extensive crescent formation. At the time of biopsy, 95% of patients have some degree of crescent formation and 81% have crescents in 50% or more of glomeruli. On average, 77% of glomeruli have crescents. Early crescents are formed by proliferating epithelial cells and infiltrating T lymphocytes, monocytes, and polymorphonuclear leucocytes, whereas older ones are composed predominantly of spindled fibroblast-like cells, with few if any infiltrating leucocytes (see Chapter 70). Glomeruli with crescents typically have fibrinoid necrosis in adjacent glomerular segments. Non-necrotic segments may look entirely normal by light microscopy, or may have slight infiltration by neutrophils or mononuclear leucocytes. This differs from crescentic immune complex glomerulonephritis, which typically has capillary wall thickening and endocapillary hypercellularity in the intact glomeruli. Special stains that outline basement membranes, such as Jones silver methenamine or periodic acid–Schiff stains, often demonstrate focal breaks in glomerular basement membranes in areas of necrosis, and also show focal breaks in Bowman capsule. The most severely injured glomeruli have global glomerular necrosis, circumferential cellular crescents, and extensive disruption of Bowman capsule. There may be a mixture of acute

(B)

Fig. 72.3  Light microscopy on renal biopsy from patients with anti-GBM disease. (A) A glomerulus showing mild proliferation of mesangial cells, from renal biopsy of a patient with microscopic haematuria and proteinuria, and normal serum creatinine. (B) A typical glomerulus from the renal biopsy of a patient with RPGN and requiring dialysis at presentation, showing global glomerular necrosis, circumferential cellular crescents, and extensive disruption of Bowman capsule (periodic acid–Schiff and silver methenamine stains).

601

602

Section 3  

the patient with glomerular disease

There are no specific changes in arteries or arterioles. If necrotizing inflammation is observed in arteries or arterioles, the possibility of concurrent ANCA should be considered. The findings by electron microscopy reflect those seen by light microscopy. Ultrastructurally, the GBM usually shows a widespread irregular broadening, often with mottled thickening of the lamina rara interna. Breaks in the GBM are common. Endothelial and epithelial cells are swollen, and epithelial foot processes may be effaced. An important negative observation is the absence of immune complex type electron-dense deposits. These occur only in anti-GBM disease patients who have concurrent immune complex disease.

Pulmonary manifestations The most frequent extrarenal presentation is pulmonary involvement. Patients present with cough, dyspnoea, shortness of breath, and haemoptysis, although the severity can vary widely—ranging from mild to life-threatening and requiring mechanical ventilation. The prevalence of pulmonary haemorrhage is reported from 50% to 90%, varying on different criteria for diagnosis of pulmonary haemorrhage. Pulmonary haemorrhage presents typically as haemoptysis that may be episodic, varies from the trivial to torrential and is a poor reflection of the actual quantity of pulmonary bleeding. It can also occur without haemoptysis. In contrast with renal injury, lung disease shows a very poor correlation with antibody titre, even though the autoantigen is present in alveolar as well as glomerular basement membrane. This may reflect the lack of direct contact between circulating antibodies and alveolar basement membrane (Jennings et al., 1981; Downie et al., 1982; Yamamoto and Wilson, 1987). It is consistent with the clear association between pulmonary haemorrhage and cigarette smoking or exposure to other inhaled toxins, notably gasoline or other hydrocarbons (discussed further in Chapter  74). Fluid overload and pulmonary infections have also been shown to provoke lung haemorrhage in anti-GBM disease. Isolated lung disease is reported regularly, though haematuria is probably always present. It may occur weeks to months before presentation with fulminant renal/ pulmonary disease, and is often associated with false-negative serological tests for anti-GBM antibodies.

Signs Physical examination can be normal in patients with mild to moderate pulmonary haemorrhage, but the more severely affected are usually tachypnoeic and may be cyanosed. They may expectorate fresh blood and have rather dry-sounding inspiratory crackles on auscultation that are most prominent over the lower lung fields and may be accompanied by areas of bronchial breathing.

Radiology Most episodes of pulmonary haemorrhage are associated with changes in the chest radiograph (Fig. 72.4). Usually the shadows involve the central lung fields, with peripheral and upper-lobe sparing. The abnormalities are generally symmetrical, but can be markedly asymmetrical. Changes range from ill-defined nodules of size 1–4 mm to confluent consolidation with an air bronchogram. Shadowing is rarely limited or entirely confined by a fissure, and radiographs that show this or those that demonstrate shadowing at

Fig. 72.4  Chest radiograph of a patient with anti-GBM disease who presented with pulmonary haemorrhage and RPGN. Symmetrical parenchymal patchy shadows are shown with some confluent consolidation.

the apex strongly suggest infection, either alone or superimposed on pulmonary haemorrhage. Shadows caused by bleeding usually start to clear within 48 hours. Residual minor changes are usually gone within 2 weeks. The diagnosis of pulmonary haemorrhage presents few problems in the majority of patients, and difficulties only arise in the minority whose haemoptysis is absent. Other indicators include a sudden otherwise unexplained drop in haemoglobin and new shadows on the chest radiograph.

Pulmonary function Bleeding into the lung also causes an acute increase in the transfer factor corrected for lung volumes and the patient’s haemoglobin, KCO, the diffusion coefficient for carbon monoxide (CO). This is the most sensitive and specific test for fresh pulmonary haemorrhage (Ewan et al., 1976). The cause is the additional free haemoglobin within the alveoli that is able to bind inspired carbon monoxide and so increase values for KCO. These results contrast with the usual situation in renal failure in which the KCO is about 30% lower than predicted, and for most patients with pulmonary oedema who also show reduced values. The lung lesions resolve almost completely and chronic lung disease is not a described outcome even in those who have survived life-threatening pulmonary haemorrhage (Conlon et al., 1994).

Pulmonary pathology At autopsy of patients with pulmonary haemorrhage, the lungs are characteristically heavy, showing patchy congestion or haemorrhage. Histologically, intra-alveolar haemorrhage is accompanied by haemosiderin-containing macrophages, deposits of fibrin, and alveolar cell hyperplasia. Electron microscopy shows thickening of the alveolar basement membrane, often with defects. Thickened alveolar walls may show oedema, fibrosis, and modest inflammatory cell infiltration, mainly with polymorphs and lymphocytes. Immunofluorescence investigations are more difficult in the lung, but linear fixation of immunoglobulin can be detected at autopsy in patients with lung haemorrhage. However, the binding is patchy, so

Chapter 72 

that, although a high success rate in obtaining diagnostic materials by transbronchial biopsy has been reported, the technique is unreliable for diagnostic use (Johnson et al., 1985; Nakajima et al., 1999).

Diagnosis The specific diagnosis of anti-GBM disease rests on the demonstration of anti-GBM antibodies in the circulation or fixed to the kidney. Circulating anti-GBM antibodies are predominantly IgG, although there are rare cases in which only anti-GBM IgA could be detected (Border et al., 1979; Savage et al., 1986; Maes et al., 1999; Ho et  al., 2008). Indirect immunofluorescence assay using sections of normal kidney is specific but not sensitive and has been replaced by solid phase immunoassays. Radioimmunoassay and enzyme-linked immunosorbent assay (ELISA) for the antibodies are relatively specific and sensitive (Kluth and Rees, 1999; Pusey, 2003). Purified bovine or sheep GBM soluble proteins enriched for α3(IV)NC1 are widely used in commercial available assays with high specificity (> 95%) and sensitivity (> 90%) (Sinico et al., 2006). In some diagnostic centres, recombinant human α1 to α5 (IV)NC1 are used to confirm and validate the presence of anti-GBM antibodies using solid phase immunoassays and Western blotting analysis (Salama et al., 2002; Jia et al., 2012; Mahler et al., 2012). Circulating anti-GBM antibodies can be detected in > 90% of patients using these solid phase immunoassays, but some assays may still have varying degree of false-negativity or false-positivity (Litwin et al., 1996; Jaskowski et al., 2002; Salama et al., 2002; Jia et al., 2012). False-negative immunoassay results are most likely in patients with isolated pulmonary or low-grade/early renal disease. These patients will have linear antibody binding to the GBM, and the presence of typical but low-titre circulating antibodies can often be detected by more sensitive assays or by Western blotting. Rare non-IgG antibodies are identified by direct immunofluorescence studies of the renal biopsy. False-positive assays may occasionally be seen in states of polyclonal activation including other autoimmune conditions. Renal biopsy will not confirm linear binding of antibody to the GBM. It is the rapidity of deterioration, and the rapidly changing prognosis for renal recovery, that makes prompt diagnosis and therapy of anti-GBM disease imperative. In the correct clinical context (i.e. raised serum creatinine concentration and active urinary sediment or proteinuria, with or without haemoptysis), positive circulating anti-GBM antibodies can be used as an indication to start treatment in cases where renal biopsy cannot be performed immediately. Renal biopsy should, however, be performed as soon as possible, since in rare cases, circulating anti-GBM antibodies are not detectable despite the presence of antibodies deposited along the GBM in the kidney (Salama et al., 2002; Jia et al., 2012). From 10% to 38% of patients have both positive anti-GBM antibodies and ANCA usually directed against myeloperoxidase (Rutgers et al., 2005; Cui et al., 2011a). Other serologic tests such as anti-streptolysin O, antinuclear antibodies, serum immunoglobulin and complement levels, rheumatoid factor, cryoglobulins, and circulating immune complex are either negative or normal. Although a strong HLA association is recognised (see Chapter 74), this is not useful diagnostically. Direct immunofluorescence on the renal biopsy is the most sensitive technique of all, if adequate renal tissue is obtained and

anti-gbm disease: clinical features and diagnosis Box 72.1  Causes of linear staining on direct immunofluorescence of renal tissuea ◆ Anti-GBM disease ◆ Diabetes mellitus ◆ Older patients with hypertensive vascular disease ◆ Alport syndrome after renal transplantation ◆ Systemic lupus erythematosus ◆ Normal autopsy kidneys ◆ Cadaver kidneys after perfusion ◆ Transplant biopsies ◆ Fibrillary nephritis. (Wilson and Dixon, 1974; Quérin et  al., 1986; Alpers et  al., 1987; Peten et al., 1991) a

In most cases, the appearances are of weak binding than in anti-GBM disease, but confirmation of specificity can only be achieved by elution studies or by testing serum for antibodies to the Goodpasture antigen.

glomerular destruction is not severe; this is the main method of diagnosis in many centres. However, the typical linear staining of IgG involving GBM cannot be revealed in all patients, in part because of severe destruction of glomerular capillary walls. In such condition, the immunofluorescence may display segmental linear deposits or absolutely negative. There may also be occasional Table 72.2  Differential diagnosis of pulmonary-renal syndrome (Leatherman et al., 1984; Holdsworth et al., 1985; Clutterbuck and Pusey, 1987; Leatherman, 1987; Vats et al., 1999; Masson et al., 1992; Espinosa et al., 2002). Small vessel vasculitis and anti-GBM disease are the key disorders to differentiate when there is true pulmonary haemorrhage with glomerulonephritis Causing lung haemorrhage and RPGN

Causing simultaneous renal and pulmonary failure by various means

ANCA-associated small vessel vasculitis

Pulmonary oedema secondary to hypervolemia in acute renal failure of any aetiology

Anti-GBM disease

Pulmonary embolism secondary to nephrotic syndrome (most commonly seen in membranous nephropathy)

Systemic lupus erythematosus

Severe pneumonia (including Legionella pneumonia) with acute tubular necrosis

Antiphospholipid syndrome

Pulmonary tuberculosis, especially during steroid treatments for glomerulonephritis

Henoch–Schönlein purpura

Hantavirus infections

Behçet disease

Bacterial endocarditis with pulmonary oedema

Mixed essential cryoglobulinaemia Paraquat poisoning Haemolytic uraemic syndrome Rheumatoid vasculitis

Leptospirosis

603

604

Section 3  

the patient with glomerular disease

false-positive results, as linear fluorescence not attributable to anti-GBM disease has been noted in a number of circumstances, mostly seen in diabetic nephropathy and transplant biopsies (Box 72.1). The clinical data and light microscopic findings should help make this distinction. Serologic confirmation should always be obtained to substantiate the diagnosis of anti-GBM disease. Direct immunofluorescence is not useful for following disease activity as linear antibody fixation may be demonstrable for a year or more after diagnosis, after circulating antibodies have become undetectable, and in the absence of clinical disease (Teague et al., 1978). It is important to distinguish the syndrome of lung haemorrhage and RPGN from other causes of renal and pulmonary failure (Table 72.2).

References Alpers, C. E., Rennke, H. G., Hopper, J., Jr., et al. (1987). Fibrillary glomerulonephritis: an entity with unusual immunofluorescence features. Kidney Int, 31, 781–9. Andrassy, K., Kuster, S., Waldherr, R., et al. (1991). Rapidly progressive glomerulonephritis: analysis of prevalence and clinical course. Nephron, 59, 206–12. Ang, C., Savige, J., Dawborn, J., et al. (1998). Anti-glomerular basement membrane (GBM)-antibody-mediated disease with normal renal function. Nephrol Dial Transplant, 13, 935–9. Bailey, R. R., Simpson, I. J., Lynn, K. L., et al. (1981). Goodpasture’s syndrome with normal renal function. Clin Nephrol, 15, 211–15. Beirne, G. J., Wagnild, J. P., Zimmerman, S. W., et al. (1977). Idiopathic crescentic glomerulonephritis. Medicine (Baltimore), 56, 349–81. Bell, D. D., Moffatt, S. L., Singer, M., et al. (1990). Antibasement membrane antibody disease without clinical evidence of renal disease. Am Rev Respir Dis, 142, 234–7. Bigler, S. A., Parry, W. M., Fitzwater, D. S., et al. (1997). An 11-month-old with anti-glomerular basement membrane disease. Am J Kidney Dis, 30, 710–12. Border, W. A., Baehler, R. W., Bhathena, D., et al. (1979). IgA antibasement membrane nephritis with pulmonary hemorrhage. Ann Intern Med, 91, 21–5. Briggs, W. A., Johnson, J. P., Teichman, S., et al. (1979). Antiglomerular basement membrane antibody-mediated glomerulonephritis and Goodpasture’s syndrome. Medicine (Baltimore), 58, 348–61. Clutterbuck, E. J. and Pusey, C. D. (1987). Severe alveolar haemorrhage in Churg-Strauss syndrome. Eur J Respir Dis, 71, 158–63. Conlon, P. J., Jr., Walshe, J. J., Daly, C., et al. (1994). Antiglomerular basement membrane disease: the long-term pulmonary outcome. Am J Kidney Dis, 23, 794–6. Couser, W. G. (1974). Goodpasture’s syndrome: a response to nitrogen mustard. Am J Med Sci, 268, 175–9. Cui, Z., Zhao, J., Jia, X. Y., et al. (2011a). Anti-glomerular basement membrane disease: outcomes of different therapeutic regimens in a large single-center chinese cohort study. Medicine (Baltimore), 90, 303–11. Cui, Z., Zhao, J., Jia, X. Y., et al. (2011b). Clinical features and outcomes of anti-glomerular basement membrane disease in older patients. Am J Kidney Dis, 57, 575–82. Cui, Z., Zhao, M. H., Singh, A. K., et al. (2007). Antiglomerular basement membrane disease with normal renal function. Kidney Int, 72, 1403–8. Cui, Z., Zhao, M. H., Xin, G., et al. (2005). Characteristics and prognosis of Chinese patients with anti-glomerular basement membrane disease. Nephron Clin Pract, 99, c49–55. Daly, C., Conlon, P. J., Medwar, W., et al. (1996). Characteristics and outcome of anti-glomerular basement membrane disease: a single-center experience. Ren Fail, 18, 105–12. Disney, A. P. S. (1986). Tenth Report of the Australian and New Zealand Combined Dialysis and Transplant Registry (ANZ-DATA). Adelaide: Queen Elizabeth Hospital.

Downie, G. H., Roholt, O. A., Jennings, L., et al. (1982). Experimental anti-alveolar basement membrane antibody-mediated pneumonitis. II. Role of endothelial damage and repair, induction of autologous phase, and kinetics of antibody deposition in Lewis rats. J Immunol, 129, 2647–52. Espinosa, G., Cervera, R., Font, J., et al. (2002). The lung in the antiphospholipid syndrome. Ann Rheum Dis, 61, 195–8. Ewan, P. W., Jones, H. A., Rhodes, C. G., et al. (1976). Detection of intrapulmonary hemorrhage with carbon monoxide uptake. Application in goodpasture’s syndrome. N Engl J Med, 295, 1391–6. Fischer, E. G. and Lager, D. J. (2006). Anti-glomerular basement membrane glomerulonephritis: a morphologic study of 80 cases. Am J Clin Pathol, 125, 445–50. Fivush, B., Melvin, T., Solez, K., et al. (1986). Idiopathic linear glomerular IgA deposition. Arch Pathol Lab Med, 110, 1189–91. Goodpasture, E. W. (1919). The significance of certain pulmonary lesions in relation to the etiology of influenza. Am J Med Sci, 158, 863–70. Herody, M., Bobrie, G., Gouarin, C., et al. (1993). Anti-GBM disease: predictive value of clinical, histological and serological data. Clin Nephrol, 40, 249–55. Hirayama, K., Yamagata, K., Kobayashi, M., et al. (2008). Anti-glomerular basement membrane antibody disease in Japan: part of the nationwide rapidly progressive glomerulonephritis survey in Japan. Clin Exp Nephrol, 12, 339–47. Ho, J., Gibson, I. W., Zacharias, J., et al. (2008). Antigenic heterogeneity of IgA anti-GBM disease: new renal targets of IgA autoantibodies. Am J Kidney Dis, 52, 761–5. Holdsworth, S., Boyce, N., Thomson, N. M., et al. (1985). The clinical spectrum of acute glomerulonephritis and lung haemorrhage (Goodpasture’s syndrome). QJM, 55, 75–86. Jaskowski, T. D., Martins, T. B., Litwin, C. M., et al. (2002). Comparison of four enzyme immunoassays for the detection of immunoglobulin G antibody against glomerular basement membrane. J Clin Lab Anal, 16, 143–5. Jennings, L., Roholt, O. A., Pressman, D., et al. (1981). Experimental anti-alveolar basement membrane antibody-mediated pneumonitis. I. The role of increased permeability of the alveolar capillary wall induced by oxygen. J Immunol, 127, 129–34. Jia, X. Y., Qu, Z., Cui, Z., et al. (2012). Circulating anti-GBM autoantibodies against alpha3(IV)NC1 undetectable by commercial available enzyme-linked immunosorbent assays. Nephrology (Carlton), 17(2), 160–6. Johnson, J. P., Moore, J., Jr., Austin, H. A., 3rd., et al. (1985). Therapy of anti-glomerular basement membrane antibody disease: analysis of prognostic significance of clinical, pathologic and treatment factors. Medicine (Baltimore), 64, 219–27. Kelly, P. T. and Haponik, E. F. (1994). Goodpasture syndrome: molecular and clinical advances. Medicine (Baltimore), 73, 171–85. Kluth, D. C. and Rees, A. J. (1999). Anti-glomerular basement membrane disease. J Am Soc Nephrol, 10, 2446–53. Knoll, G., Rabin, E., and Burns, B. F. (1993). Antiglomerular basement membrane antibody-mediated nephritis with normal pulmonary and renal function. A case report and review of the literature. Am J Nephrol, 13, 494–6. Leatherman, J. W. (1987). Immune alveolar hemorrhage. Chest, 91, 891–7. Leatherman, J. W., Davies, S. F., and Hoidal, J. R. (1984). Alveolar hemorrhage syndromes: diffuse microvascular lung hemorrhage in immune and idiopathic disorders. Medicine (Baltimore), 63, 343–61. Lerner, R. A., Glassock, R. J., and Dixon, F. J. (1967). The role of anti-glomerular basement membrane antibody in the pathogenesis of human glomerulonephritis. J Exp Med, 126, 989–1004. Levy, J. B., Turner, A. N., and Rees, A. J. (2001). Long-term outcome of anti-glomerular basement membrane antibody disease treated with plasma exchange and immunosuppression. Ann Intern Med, 134, 1033–42.

Chapter 72 

Li, F. K., Tse, K. C., Lam, M. F., et al. (2004). Incidence and outcome of antiglomerular basement membrane disease in Chinese. Nephrology (Carlton), 9, 100–4. Lin, W., Chen, M., Cui, Z., et al. (2010). The immunopathological spectrum of crescentic glomerulonephritis: a survey of 106 patients in a single Chinese center. Nephron Clin Pract, 116, c65–74. Litwin, C. M., Mouritsen, C. L., Wilfahrt, P. A., et al. (1996). Anti-glomerular basement membrane disease: role of enzyme-linked immunosorbent assays in diagnosis. Biochem Mol Med, 59, 52–6. Maes, B., Vanwalleghem, J., Kuypers, D., et al. (1999). IgA antiglomerular basement membrane disease associated with bronchial carcinoma and monoclonal gammopathy. Am J Kidney Dis, 33, E3. Mahler, M., Radice, A., Sinico, R. A., et al. (2012). Performance evaluation of a novel chemiluminescence assay for detection of anti-GBM antibodies: an international multicenter study. Nephrol Dial Transplant, 27(1), 243–52. Masson, R. G., Rennke, H. G., and Gottlieb, M. N. (1992). Pulmonary hemorrhage in a patient with fibrillary glomerulonephritis. N Engl J Med, 326, 36–9. Mathew, T. H., Hobbs, J. B., Kalowski, S., et al. (1975). Goodpasture’s syndrome: normal renal diagnostic findings. Ann Intern Med, 82, 215–18. Merkel, F., Pullig, O., Marx, M., et al. (1994). Course and prognosis of anti-basement membrane antibody (anti-BM-Ab)-mediated disease: report of 35 cases. Nephrol Dial Transplant, 9, 372–6. Min, S. A., Rutherford, P., Ward, M. K., et al. (1996). Goodpasture’s syndrome with normal renal function. Nephrol Dial Transplant, 11, 2302–5. Nakajima, I., Sasaki, M., Ito, T., et al. (1999). [Goodpasture’s syndrome initially presenting with alveolar hemorrhage]. Nihon Kokyuki Gakkai Zasshi, 37, 652–7. Peten, E., Pirson, Y., Cosyns, J.P., et al. (1991). Outcome of 30 patients with Alport’s syndrome after renal transplantation. Transplantation, 52, 823–6. Peters, D. K., Rees, A. J., Lockwood, C. M., et al. (1982). Treatment and prognosis in antibasement membrane antibody-mediated nephritis. Transplant Proc, 14, 513–21. Pusey, C. D. (2003). Anti-glomerular basement membrane disease. Kidney Int, 64, 1535–50. Quérin, S., Noël, L. H., Grünfeld, J. P., et al. (1986). Linear glomerular IgG fixation in renal allografts: incidence and significance in Alport’s syndrome. Clin Nephrol, 25, 134–40.

anti-gbm disease: clinical features and diagnosis

Rajaraman, S., Pinto, J. A., and Cavallo, T. (1984). Glomerulonephritis with coexistent immune deposits and antibasement membrane activity. J Clin Pathol, 37, 176–81. Rutgers, A., Slot, M., van Paassen, P., et al. (2005). Coexistence of anti-glomerular basement membrane antibodies and myeloperoxidase-ANCAs in crescentic glomerulonephritis. Am J Kidney Dis, 46, 253–62. Salama, A. D., Dougan, T., Levy, J. B., et al. (2002). Goodpasture’s disease in the absence of circulating anti-glomerular basement membrane antibodies as detected by standard techniques. Am J Kidney Dis, 39, 1162–7. Savage, C. O., Pusey, C. D., Bowman, C., et al. (1986). Antiglomerular basement membrane antibody mediated disease in the British Isles 1980-4. Br Med J (Clin Res Ed), 292, 301–4. Segelmark, M., Hellmark, T., and Wieslander, J. (2003). The prognostic significance in Goodpasture’s disease of specificity, titre and affinity of anti-glomerular-basement-membrane antibodies. Nephron Clin Pract, 94, c59–68. Shah, M. K. and Hugghins, S. Y. (2002). Characteristics and outcomes of patients with Goodpasture’s syndrome. South Med J, 95, 1411–18. Sinico, R. A., Radice, A., Corace, C., et al. (2006). Anti-glomerular basement membrane antibodies in the diagnosis of Goodpasture syndrome: a comparison of different assays. Nephrol Dial Transplant, 21, 397–401. Stanton, M. C. and Tange, J. D. (1958). Goodpasture’s syndrome (pulmonary haemorrhage associated with glomerulonephritis). Australas Ann Med, 7, 132–44. Teague, C. A., Doak, P. B., Simpson, I. J., et al. (1978). Goodpasture’s syndrome: an analysis of 29 cases. Kidney Int, 13, 492–504. Vats, K. R., Vats, A., Kim, Y., et al. (1999). Henoch-Schonlein purpura and pulmonary hemorrhage: a report and literature review. Pediatr Nephrol, 13, 530–4. Walker, R. G., Scheinkestel, C., Becker, G. J., et al. (1985). Clinical and morphological aspects of the management of crescentic anti-glomerular basement membrane antibody (anti-GBM) nephritis/Goodpasture’s syndrome. QJM, 54, 75–89. Wilson, C. B. and Dixon, F. J. (1973). Anti-glomerular basement membrane antibody-induced glomerulonephritis. Kidney Int, 3, 74–89. Yamamoto, T. and Wilson, C. B. (1987). Binding of anti-basement membrane antibody to alveolar basement membrane after intratracheal gasoline instillation in rabbits. Am J Pathol, 126, 497–505. Zimmerman, S. W., Varanasi, U. R., and Hoff, B. (1979). Goodpasture’s syndrome with normal renal function. Am J Med, 66, 163–71.

605

CHAPTER 73

Antiglomerular basement membrane disease: treatment and outcome Zhao Cui, Neil Turner, and Ming-hui Zhao Treatment The rarity and fulminant course of antiglomerular basement membrane (GBM) disease prevent the initiation of any large randomized studies to investigate therapeutic benefits. The only reported randomized controlled trial (RCT) tested the need for plasma exchange, but it was small, groups unevenly matched for severity, and the degree of plasmapheresis and doses of cyclophosphamide were lower than those generally used (Johnson et al., 1985). However, it appeared that less severely affected individuals did relatively well without plasma exchange. Retrospective analyses based on a large number of patients provide useful information (Couser, 1988; Cui et al., 2011a). The best summary of this approach is a large retrospective study of anti-GBM disease from Hammersmith Hospital (Levy et al., 2001), including 85 patients seen over 25 years. Seventy-one patients were treated with high-dose prednisone (1 mg/kg/day) tapered over 6–9 months, oral cyclophosphamide for 2–3 months, and daily plasmapheresis for 14 days, or until anti-GBM antibody was no longer detectable. The kidney outcome for this cohort was strongly influenced by kidney function at presentation. Patients who had an initial serum creatinine (SCr) < 5.7 mg/dL (500 μmol/L) had 1-year overall survival of 100% and kidney survival of 95%. If the initial SCr was > 5.7 mg/dL (500 μmol/L) but dialysis was not required immediately, the patient and kidney survivals were 83% and 82% at 1 year, respectively. However, among patients who needed dialysis at presentation, patient and kidney survival were reduced to 65% and 8% at 1 year, respectively. Compared to nearly 100% mortality from pulmonary haemorrhage and kidney failure in historical series, this treatment strategy represent a significant improvement and is widely adopted in multiple centres. See Table 73.1.

Antibody removal Plasma exchange is ineffective if used alone (Guillen et al., 1995) but hastens the disappearance of circulating antibody when used in combination with immunosuppressive agents (Johnson et al., 1985; Savage et al., 1986). There is a strong clinical impression of its efficacy, and its role as essential component in the treatment of severe anti-GBM disease is accepted. Couser (1988) reviewed 22 published uncontrolled studies involving 186 patients, and reported a favourable effect of plasmapheresis on pulmonary disease in around 90% of patients, and a favourable effect of plasmapheresis

on renal disease in about 40% of patients. In a retrospective analysis involving 221 patients Beijing, it was demonstrated that, in comparison with corticosteroids alone and corticosteroids plus cyclophosphamide, the combination of plasma exchange, corticosteroids and cyclophosphamide has an overall beneficial effect on patient and kidney survival. Particular benefit was seen for patients with pulmonary haemorrhage and, in relation to kidney survival, for those with anti-GBM nephritis and initial SCr higher than 6.8 mg/ dL (600 µmol/L) (Cui et al., 2011a). Using immunoadsorption against staphylococcal protein A  to adsorb circulating IgG antibodies (Bygren et  al., 1985; Laczika et  al., 2000)  lowers antibody titres more rapidly than plasma exchange and has a number of theoretical advantages. However, it is a more time-consuming and complicated technique, and is not widely available.

Corticosteroids Prednisolone has been a universal part of treatment regimens. We do not recommend the use of bolus doses of methylprednisolone. This (as methylprednisolone 10 mg/kg intravenously once daily for 1–3 days) has been advocated when there is severe pulmonary haemorrhage or very rapidly declining renal function (Johnson et al., 1985), but may increase the risk of later infection, a major concern because as well as threatening survival directly, it may exacerbate renal and pulmonary injury.

Cytotoxic and immunosuppressive agents Cyclophosphamide was first used successfully by Couser in 1974 (Couser, 1974), and is a critical element of therapy. Alternative immunosuppressive agents have been tried in only a few cases, including azathioprine, ciclosporin, mycophenolate mofetil, and the anti-B-cell antibody rituximab (Pepys et al., 1982; Quérin et al., 1992; Garcia-Canton et al., 2000). Most of these seem likely to be less effective or only slowly effective.

Outcomes and prognostic factors Outcomes of the Hammersmith cohort (Levy et al., 2001) are representative of what can be expected with a uniform, aggressive approach to therapy as outlined. In other series, not all necessarily using the same treatment regimens, and encompassing patients

Chapter 73 

Table 73.1  Therapy of anti-GBM disease Corticosteroids

1 mg/kg/day ideal body weight (maximum 80 mg), tailed off over 3–6 months (typically 60 mg, 45 mg, 30 mg, 25 mg, 20 mg at 1–2 weekly intervals then more slowly. There is disagreement over possible value for higher dose steroids (e.g. pulsed methylprednisolone) at initiation of treatment; it may increase risk of infection and disease amplification

Cyclophosphamide

2 mg/kg/day orally for 3 months, rounded down to nearest 50 mg, and at lower dose if > 60 years

Plasmapheresis

One 4-L (or plasma-volume) exchange per day with 5% albumin. Add 150–300 mL fresh frozen plasma at the end of each pheresis session if patients have pulmonary haemorrhage, or have had recent surgery, including kidney biopsy. Plasmapheresis should be continued for 14 days or until anti-GBM antibodies are no longer detectable

from the United States, Europe, China, and Japan, patient survival at 6–12  months is approximately 67–94%, and kidney survival is about 15–58% (Johnson et al., 1985; Kelly and Haponik, 1994; Merkel et al., 1994; Daly et al., 1996; Shah and Hugghins, 2002; Li et al., 2004; Hirayama et al., 2008; Cui et al., 2011a). Haemoptysis is no longer the major cause of death. Titres of antibodies against GBM and coexistence of ANCA are independent predictors for death (Cui et al., 2011a). An association between high anti-GBM antibody titres and poor renal outcomes is also reported. There are conflicting findings on the implications of ANCA for renal outcome. Predictors of kidney survival are serum creatinine at presentation, the need for dialysis at presentation, and the percentage of glomerular crescents (Johnson et al., 1985; Levy et al., 2001; Cui et al., 2011a). Serum creatinine at presentation is an independent predictor for renal failure (Cui et al., 2011a). Patients who required dialysis at presentation may not be able to come off dialysis, despite aggressive treatment (Levy et al., 2004; Li et al., 2004). The most optimistic study observed that no patients with a combination of dialysis at presentation plus 100% crescents on kidney biopsy recovered kidney function sufficiently to come off dialysis (Levy et al., 2001). A survey of several studies shows dialysis dependence at diagnosis in a median of 55% (range 12–83%) of patients, 100% crescents on kidney biopsy in 20.5% (range 7–50%) of patients, and a median initial SCr of 6.8 mg/dL (600 μmol/L) (range 4.9–7.2 mg/ dL (430–630 μmol/L)), underscoring the importance of early diagnosis and urgent intervention. The prognosis of patients with mild renal dysfunction or normal kidney function is good when they are treated aggressively with plasmapheresis and immunosuppressive regimens used for classic anti-GBM disease (Levy et al., 2001; Cui et al., 2007). Indeed, this treatment is indicated in patients who have haemoptysis (Zimmerman et  al., 1979; Bell et  al., 1990)  and in those in whom renal histological abnormalities are seen (Ang et al., 1998). The use of less aggressive strategies or merely supportive treatment

anti-gbm disease: treatment and outcome

has occasionally been associated with spontaneous recovery and persistently normal renal function. However, given the fulminant course and possible severe outcomes of anti-GBM disease, a nonaggressive approach is still not recommended in these patients. Johnson’s randomized study (Johnson et al., 1985) suggested that an approach that uses cyclophosphamide without plasma exchange might be reasonable however.

Deciding not to treat The kidney prognosis of dialysis-dependent patients with 100% crescents in adequate biopsy sample is extremely poor. It has been suggested that, in the absence of pulmonary haemorrhage, aggressive immunosuppression should be withheld, for the risks of therapy outweigh the potential benefits (Flores et  al., 1986). However, there are occasional reports of patients who have recovered despite acute kidney injury requiring dialysis (Cohen et al., 1976; Johnson et al., 1978; Schindler et al., 1998; Laczika et al., 2000; Cui et al., 2011a). Their hallmark is that they have either mild glomerular change with superimposed acute tubular necrosis or very new cellular crescents without fibrosis on renal biopsy. These patients emphasize the value of an urgent renal biopsy even when the diagnosis has been established serologically. In those patients, aggressive treatment should continue for at least 4 weeks. If there is no restoration of renal function by 4–8 weeks, and in the absence of pulmonary bleeding, immunosuppression should be discontinued.

Relapse and recurrence In the early phase, immunological relapse should be distinguished from haemodynamic changes or drug effects, and from exacerbation of injury in association with infection. In contrast to most other autoimmune diseases, anti-GBM disease is not usually characterized by a recurrent course. Autoantibodies seem to disappear spontaneously after 12–18 months (Wilson and Dixon, 1981; Levy et al., 1996). Recurrences of pulmonary haemorrhage over many years are described in the literature (Dahlberg et al., 1978; Mehler et al., 1987; Klasa et al., 1988), and in these circumstances haematuria may also. Recurrent renal disease is exceptional (Levy et al., 1996). Autoantibody reproduction may occur without signs of tissue damage (Hind et al., 1984). Treatment of recurrences is identical to that of initial disease and probably has a better chance of success.

Kidney transplantation There is little evidence as to the timing of transplant after anti-GBM disease has caused end-stage renal disease. Most transplant centres require at least 6 months of undetectable anti-GBM antibody levels before kidney transplantation. Recurrent anti-GBM disease in a kidney allograft has been described but is now very unusual (Choy et al., 2006; Joshi et al., 2007). Alport anti-GBM disease (see Chapter  75) is an alloimmune, rather than autoimmune disease, that occurs in < 5% of patients with Alport syndrome after transplantation. It is caused by the development of antibodies to antigens that are missing or altered in the recipient. Most commonly the missing antigen is α5(IV), rather than α3(IV). Clinically these anti-GBM variants are nearly identical, but Alport anti-GBM disease attacks antigens present in the kidney, so lung haemorrhage is not to be expected (it has been reported in one case). Alloimmune responses like this are hard to

607

608

Section 3  

the patient with glomerular disease

suppress and outcomes of treatment are poor. Alport anti-GBM disease is described further in Chapter 75.

References Ang, C., Savige, J., Dawborn, J., et al. (1998). Anti-glomerular basement membrane (GBM)-antibody-mediated disease with normal renal function. Nephrol Dial Transplant, 13, 935–9. Bell, D. D., Moffatt, S. L., Singer, M., et al. (1990). Antibasement membrane antibody disease without clinical evidence of renal disease. Am Rev Respir Dis, 142, 234–7. Bygren, P., Freiburghaus, C., Lindholm, T., et al. (1985). Goodpasture’s syndrome treated with staphylococcal protein A immunoadsorption. Lancet, 2, 1295–6. Choy, B. Y., Chan, T. M., and Lai, K.,N. (2006). Recurrent glomerulonephritis after kidney transplantation. Am J Transplant, 6, 2535–42. Cohen, L. H., Wilson, C. B., and Freeman, R. M. (1976). Goodpasture syndrome: recovery after severe renal insufficiency. Arch Intern Med, 136, 835–7. Couser, W. G. (1974). Goodpasture’s syndrome: a response to nitrogen mustard. Am J Med Sci, 268, 175–9. Couser, W. G. (1988). Rapidly progressive glomerulonephritis: classification, pathogenetic mechanisms, and therapy. Am J Kidney Dis, 11, 449–64. Cui, Z., Zhao, J., Jia, X. Y., et al. (2011a). Anti-glomerular basement membrane disease: outcomes of different therapeutic regimens in a large single-center chinese cohort study. Medicine (Baltimore), 90, 303–11. Cui, Z., Zhao, M. H., Singh, A. K., et al. (2007). Antiglomerular basement membrane disease with normal renal function. Kidney Int, 72, 1403–8. Dahlberg, P. J., Kurtz, S. B., Donadio, J. V., et al. (1978). Recurrent Goodpasture’s syndrome. Mayo Clin Proc, 53, 533–7. Daly, C., Conlon, P. J., Medwar, W., et al. (1996). Characteristics and outcome of anti-glomerular basement membrane disease: a single-center experience. Ren Fail, 18, 105–12. Flores, J. C., Taube, D., Savage, C. O., et al. (1986). Clinical and immunological evolution of oligoanuric anti-GBM nephritis treated by haemodialysis. Lancet, 1, 5–8. Garcia-Canton, C., Toledo, A., Palomar, R., et al. (2000). Goodpasture’s syndrome treated with mycophenolate mofetil. Nephrol Dial Transplant, 15, 920–2. Guillen, E. L., Ruiz, A. M., Fernandez, M. A., et al. (1995). Goodpasture syndrome: re-exacerbations associated with intercurrent infections, Revista Clinica Española, 195, 761–4. Hind, C. R., Bowman, C., Winearls, C. G., et al. (1984). Recurrence of circulating anti-glomerular basement membrane antibody three years after immunosuppressive treatment and plasma exchange. Clin Nephrol, 21, 244–6. Hirayama, K., Yamagata, K., Kobayashi, M., et al. (2008). Anti-glomerular basement membrane antibody disease in Japan: part of the nationwide rapidly progressive glomerulonephritis survey in Japan. Clin Exp Nephrol, 12, 339–47.

Johnson, J. P., Moore, J., Jr., Austin, H. A., 3rd., et al. (1985). Therapy of anti-glomerular basement membrane antibody disease: analysis of prognostic significance of clinical, pathologic and treatment factors. Medicine (Baltimore), 64, 219–27. Johnson, J. P., Whitman, W., Briggs, W. A., et al. (1978). Plasmapheresis and immunosuppressive agents in antibasement membrane antibody-induced Goodpasture’s syndrome. Am J Med, 64, 354–9. Joshi, K., Nada, R., Minz, M., et al. (2007). Recurrent glomerulopathy in the renal allograft. Transplant Proc, 39, 734–6. Kelly, P. T. and Haponik, E. F. (1994). Goodpasture syndrome: molecular and clinical advances. Medicine (Baltimore), 73, 171–85. Klasa, R. J., Abboud, R. T., Ballon, H. S., et al. (1988). Goodpasture’s syndrome: recurrence after a five-year remission. Case report and review of the literature. Am J Med, 84, 751–5. Laczika, K., Knapp, S., Derfler, K., et al. (2000). Immunoadsorption in Goodpasture’s syndrome. Am J Kidney Dis, 36, 392–5. Levy, J. B., Hammad, T., Coulthart, A., et al. (2004). Clinical features and outcome of patients with both ANCA and anti-GBM antibodies. Kidney Int, 66, 1535–40. Levy, J. B., Lachmann, R. H., and Pusey, C. D. (1996). Recurrent Goodpasture’s disease. Am J Kidney Dis, 27, 573–8. Levy, J. B., Turner, A. N., and Rees, A. J. (2001). Long-term outcome of anti-glomerular basement membrane antibody disease treated with plasma exchange and immunosuppression. Ann Intern Med, 134, 1033–42. Li, F. K., Tse, K. C., Lam, M. F., et al. (2004). Incidence and outcome of antiglomerular basement membrane disease in Chinese. Nephrology (Carlton), 9, 100–4. Mehler, P. S., Brunvand, M. W., Hutt, M. P., et al. (1987). Chronic recurrent Goodpasture’s syndrome. Am J Med, 82, 833–5. Merkel, F., Pullig, O., Marx, M., et al. (1994). Course and prognosis of anti-basement membrane antibody (anti-BM-Ab)-mediated disease: report of 35 cases. Nephrol Dial Transplant, 9, 372–6. Pepys, E. O., Rees, A. J., and Pepys, M. B. (1982). Enumeration of lymphocyte populations in whole peripheral blood of patients with antibody-mediated nephritis during treatment with cyclosporin A. Immunol Lett, 4, 211–14. Quérin, S., Schurch, W., and Beaulieu, R. (1992). Ciclosporin in Goodpasture’s syndrome. Nephron, 60, 355–9. Savage, C. O., Pusey, C. D., Bowman, C., et al. (1986). Antiglomerular basement membrane antibody mediated disease in the British Isles 1980-4. Br Med J (Clin Res Ed), 292, 301–4. Schindler, R., Kahl, A., Lobeck, H., et al. (1998). Complete recovery of renal function in a dialysis-dependent patient with Goodpasture syndrome. Nephrol Dial Transplant, 13, 462–6. Shah, M. K. and Hugghins, S. Y. (2002). Characteristics and outcomes of patients with Goodpasture’s syndrome. South Med J, 95, 1411–18. Wilson, C. B. and Dixon, F. J. (1981). The renal response to immunological injury. In The kidney (ed. B.M. Brenner and F.C. Rector), pp. 1237–350. Philadelphia, P: W.B. Saunders. Zimmerman, S. W., Varanasi, U. R., and Hoff, B. (1979). Goodpasture’s syndrome with normal renal function. Am J Med, 66, 163–71.

CHAPTER 74

Antiglomerular basement membrane disease: pathogenesis Zhao Cui, Neil Turner, and Ming-hui Zhao Introduction Elucidating the aetiology of autoimmune disease in man is a formidable task, but anti-GBM disease provides a model in which pathogenesis has been quite well defined. Loss of tolerance with the consequent development of autoimmune disease requires both an underlying genetic susceptibility and exposure to an environmental trigger. There are also many reports of anti-GBM disease occurring in association with other disorders, especially those with damage on the kidneys. One hypothesis suggests that in individuals with susceptible HLA alleles, things that alter antigen presentation in quantity or quality, or that expose epitopes sequestered within the basement membranes, trigger an autoimmune response to α3(IV)NC1 (Nachman et al., 2007). There is no direct evidence to prove this mechanism, only suggestive or circumstantial evidence from animal experiments and case observations.

Predisposing factors Inherited susceptibility Genetic predisposition to mount an anti-GBM response is an important requirement for the disease. There is clear evidence of inherited susceptibility from reports of the disease occurring in four sibling pairs (Stanton and Tange, 1958; Gossain et al., 1972) and two sets of identical twins (D’Apice et al., 1978; Simonsen et al., 1982). Further support comes from mouse models of anti-GBM disease in which crescentic glomerulonephritis and lung haemorrhage are restricted to only certain major histocompatibility complex (MHC) haplotypes, despite the ability of mice of all haplotypes to produce anti-α3(IV)NC1 antibodies (Kalluri et al., 1997). In common with many other human autoimmune disorders, anti-GBM disease has been associated with inheritance of specific class  II HLA alleles. A  strong association with HLA-DR2 specificity (Rees et al., 1978) has been extended. Meta-analysis of published series confirms that the primary association is with the DRB1 locus (Phelps et al., 2000). It also demonstrates a hierarchy of associations, ranging from a strong positive association with DRB1*1501 (odds ratios (OR) = 8.5), through weaker positive associations with DRB1*0401 and DRB1*0301, to neutral effects and then increasingly strong negative associations with DRB1*0101 and DRB1*0701 (OR  =  0.6 and 0.3 respectively). Most patients (63.9–92%) with anti-GBM disease inherit the DRB1*1501 allele (Dunckley et al., 1991; Huey et al., 1993; Fisher et al., 1997; Phelps and Rees, 1999; Kitagawa et al., 2008; Yang et al., 2009). Gene dosage does not affect susceptibility.

The molecular mechanisms that may underlie the strong HLA class II associations with anti-GBM disease has been explored with the precisely defined autoantigen α3(IV)NC1. DRB1*1501 in general bound the α3(IV)NC1-derived peptides with low affinity compared to DRB1*0101 and DRB1*0701. All the major α3(IV)NC1 peptides would bind preferentially to DRB1*01/07 in DRB1*15, 01/07 heterozygote antigen-presenting cells (APC), which present the peptides to T-helper cells (Phelps et al., 2000). Thus, DRB1*0101 and DRB1*0701 could protect by capturing α3(IV)NC1 peptides and preventing their display bound to DRB1*1501. The immunoglobulin heavy chain Gm locus which encodes the IgG heavy chain constant region was identified as a second genetic influence on anti-GBM disease (Rees et al., 1984). Anti-GBM antibodies derived from Wistar Kyoto (WKY) rats are only able to transfer crescentic glomerulonephritis to WKY rats but not Lewis rats, suggesting that factors related to the inflammatory response to deposited antibody contribute to disease susceptibility (Reynolds et al., 2006). The genetic basis of this is shown to be owing to a copy number polymorphism of Fcgr3 (Aitman et  al., 2006). In a Chinese cohort susceptibility of the disease was linked to gene polymorphism of FCGR2B (232T/I) and the copy number variation of FCGR3A (Zhou et al., 2010a, 2010b). These findings need to be tested in different populations.

Lymphocyte depletion and breaking tolerance Lymphocyte depletion has been associated with autoimmunity in a number of clinical settings and in animal models. Autoimmune phenomena are seen as lymphocytes begin to recover, so presumably with a limited range of affinities and perhaps lacking the usual balance of regulatory and potentially inflammatory cells. In patients with multiple sclerosis the anti-CD52 lymphocyte depleting antibody alemtuzumab (CamPath®) has been associated with a high incidence of autoimmmune thyroiditis and less frequently with other autoimmune conditions, including several instances of anti-GBM disease (Coles et al., 2012; Jones et al., 2013). Multiple sclerosis also has an association with HLA-DR15, so perhaps this is understandable. Several case reports have also reported an association of anti-GBM disease with HIV infection (Monteiro et  al., 2006; Wechsler et al., 2008; reviewed by Hartle et al., 2013). In a cohort of 105 HIV-infected individuals, 18 (17%) sera were identified as containing anti-GBM antibodies, but they were associated with polyclonal activation and had no clinical features (Savige et al., 1994; Szczech et al., 2006).

610

Section 3  

the patient with glomerular disease

Environmental triggers Infectious agents have long been suspected as causative, since an upper respiratory infection precedes disease onset in 20–60% of cases (Appel et al., 2007). Influenza virus has been specifically mentioned in a number of reports, including Goodpasture’s original description (1919), but infection with the virus is common and no studies have consistently shown an association with any particular infection (Wilson and Dixon, 1973; Wilson and Smith, 1972). The association with HIV infection may be different (see above paragraph). For bacterial infections, it was suggest that mimicry of T-cell epitopes by microbial antigens derived from Clostridium botulinum is sufficient to induce anti-GBM disease in WKY rats (Arends et al., 2006). However, no relationship has been found between C. botulinum (or any other infection) and human anti-GBM disease.

Pulmonary haemorrhage The link between pulmonary involvement and smoking was established by Donaghy and Rees (1983). They found that pulmonary haemorrhage occurred in nearly all patients with anti-GBM disease who smoke, whereas this complication was rare in non-smokers. Exposure to organic solvents and hydrocarbons has been associated with disease onset (Bombassei and Kaplan, 1992; Stevenson et  al., 1995). Around 6% of all patients with anti-GBM disease in the literature have hydrocarbon exposure (Shah, 2002) in retrospective analyses, and a causal relationship has been suggested, but the evidence is weak. It is more likely that it exposes disease in those with a pre-existing immune response. The direct role of smoke inhalation and hydrocarbon exposure is further demonstrated by recurrence of anti-GBM disease. The majority of them are young male patients with recurrent pulmonary haemorrhage, in association with cigarette smoking or hydrocarbon re-exposure after apparent remission (Keller and Nekarda, 1985; Levy et al., 1996). Similarly, anti-GBM disease has been associated with intranasal or smoked cocaine (crack) (Garcia-Rostan y Perez et al., 1997; Peces et al., 1999), but the same caveats about causation apply. In experimental models, cigarette smoke and other non-specific irritants (gasoline, oxygen toxicity) act in a similar way of a non-specific irritant or toxic effect on the lungs and precipitates pulmonary haemorrhage in the presence of circulating anti-GBM antibodies (Jennings et al., 1981; Downie et al., 1982; Yamamoto et al., 1987). Similarly, intercurrent infection amplifies the intensity of inflammatory responses and can aggravate disease and so make it clinically apparent (Rees et al., 1978; Daly et al., 1996). It may be difficult to distinguish agents that exacerbate previously covert disease from true aetiological agents that initiate disease in individual cases, but it is worth noting that in animal models of active immunity the disease takes weeks to months to develop.

Disease associations Anti-GBM disease has been reported in association with several forms of other diseases, including autoimmune disorders, glomerulonephritis and others.

Dual positivity for anti-GBM and ANCA Ten to 38% of patients with anti-GBM disease also have ANCA detectable in their sera with specificity mainly for myeloperoxidase (MPO) (Short et al., 1995; Hellmark et al., 1997; Levy et al., 2004;

Rutgers et al., 2005; Yang et al., 2007; Zhao et al., 2007; Cui et al., 2011a). These patients are termed as double positive. Anti-GBM response could occur in genetically predisposed patients following damage to the glomerular or alveolar basement membrane by small vessel vasculitis. However, there are occasional instances where the clinical history suggests that anti-GBM disease antedated the development of vasculitis (O’Donoghue et al., 1989; Peces et al., 2000). Double-positive patients are mostly older (average age of 55–66  years) and present with higher prevalence of systemic involvement, including muscle pain, arthralgia, skin rash, nasal, ear, eye, throat, pulmonary, gastrointestinal, and nervous system involvement. Pulmonary haemorrhage occurs in about half of the patients, presenting no difference from the patients with anti-GBM antibodies alone or those with ANCA only. Renal involvement behaves more like anti-GBM disease than vasculitis (Table 74.1). Serum creatinine levels and percentages of patients presenting with oliguria or anuria are higher in double-positive and anti-GBM positive patients, compared with MPO-ANCA single-positive patients (Rutgers et al., 2005). Renal biopsy shows extensive glomerular cellular crescents in most patients and some of them show linear binding of antibodies to GBM by direct immunofluorescence. Nevertheless, granulomatous periglomerular inflammation is found more common in double-positive patients, but not in anti-GBM-positive patients (Levy et al., 2004; Rutgers et al., 2005). Renal prognosis is worse than that of patients with ANCA-associated vasculitis, but similar to anti-GBM disease. Recovery from severe renal failure is rare. Patient survival is worse than that of anti-GBM disease, but similar to vasculitis (Levy et al., 2004; Rutgers et al., 2005; Cui et al., 2011a). Thus, intensive plasmapheresis and immunosuppressive therapy are crucial in the early stage of treatment, and maintenance therapy may be necessary for patients in remission. In view of the high frequency of positive ANCA in patients with anti-GBM disease, the coexistence of these two types of autoantibodies cannot be explained by chance. There is no significant correlation observed between the titres of anti-GBM antibodies and ANCA. Irrespective of whether the antibodies are non-pathogenic in healthy individuals or pathogenic in patients with anti-GBM disease, no cross reaction is seen among antibodies against GBM, MPO, or PR3 (Hellmark et al., 1997; Cui et al., 2010). In double-positive patients, the prevalence of antibodies against GBM targeting α1(IV)NC1, α4(IV)NC1 and α5(IV)NC1 is significantly higher, while the prevalence of anti-GBM antibodies targeting α3(IV)NC1, specific epitope EA and EB on α3(IV)NC1 are lower, than that in patients with anti-GBM antibodies alone. Thus, the double-positive patients have a broader spectrum of antibodies to type IV collagen than patients with anti-GBM antibodies alone (Yang et al., 2007). This may support the hypothesis that the anti-GBM response is usually secondary to the small vessel vasculitis. Olson et al. (2011) used historical samples to identify elevated anti-GBM titres in advance of diagnosis of 30 patients with anti-GBM disease who did not in general have ‘double positivity’ at diagnosis. Four had elevated titres months in advance of the disease, but elevated ANCA, particularly anti-Pr3, were found further in advance in a larger proportion, raising the possibility that subclinical vasculitis is the trigger more often than currently thought.

Chapter 74 

anti-gbm disease: pathogenesis

Table 74.1  Characteristics of patients with MPO-ANCA, anti-GBM antibodies and double positive (Rutgers et al., 2005)

Age (years) Male sex (%)

MPO-ANCA (N = 46)

Double-positive (N = 10)

Anti-GBM (N = 13)

63 ± 12.7

64 ± 8.7

52 ± 20.6

67

80

39

Blood pressure (mmHg)

150/83

158/84

145/85

Proteinuria (g/24 hours)

2.0 (0.0–4.1)

1.3 (0.0–2.6)

2.3 (0.6–5.4)

12

63

50

Serum creatinine (mg/dL)

5.0 ± 2.9

10.3 ± 5.6

9.6 ± 8.1

Cellular crescents (%)

18 (0–75)

27 (0–75)

29 (0–90)

Fibrous crescents (%)

Anuira/oliguria (%)

27 (0–100)

31 (0–100)

43 (0–100)

Periglomerular granuloma (%)

11

40

0

Linear deposits of IgG on GBM (%)

0

40

77

No deposit (%)

59

0

0

Scanty, not linear deposits (%)

33

50

15

Renal survival at 1 year (%)

64

10

15

Patient survival at 1 year (%)

75

79

100

Note: to convert serum creatinine in mg/dL to μmol/L, multiply by 88.4.

Membranous nephropathy An association with idiopathic membranous nephropathy (Fig. 74.1) was first described in 1974 (Klassen et al., 1974), and the literature

now includes nearly 30 cases. In some of the reported cases, a sudden deterioration of previously diagnosed membranous nephropathy has been associated with anti-GBM antibodies (Klassen et al., 1974; Kurki et  al., 1984). In other examples, both membranous nephropathy and anti-GBM disease were present on initial assessment (Pasternack et al., 1978; Savige et al., 1989; Cui et al., 2006; Basford et al., 2011). Occasionally, there are cases having anti-GBM disease followed by membranous nephropathy (Agodoa et al., 1976; Kielstein et al., 2001; Hecht et al., 2008). It is hypothesized that the GBM may be damaged in membranous nephropathy, or its turnover increased, altering antigen processing so that new (cryptic) epitopes to which there is little immune tolerance are released.

Other conditions There are numerous but mostly isolated case reports of antiGBM disease in association with other possible relevant diseases (Table 74.2). Some are likely to be coincidental.

Animal models

Fig. 74.1  Direct immunofluorescence for IgG on the glomerulus of a 65-year-old man with nephrotic syndrome that deteriorated rapidly to severe renal failure. Linear binding to the endothelial surface of the GBM is accompanied by granular deposits of IgG subepithelially, typical of membranous nephropathy. This was confirmed by electron microscopy. Typical anti-GBM antibodies were detected in serum.

Studies in animal models have shown that both antibodies against GBM and antigen-specific T cells are pathogenic. Therefore, humoral and cellular immunity both contribute to glomerular damage in anti-GBM disease. Two experimental models of anti-GBM disease, mostly in mice or rats, are widely used. In experimental autoimmune glomerulonephritis (EAG), animals from susceptible strains are immunized with homologous or heterologous GBM or α3(IV)NC1 and after some delay develop an autoimmune response that targets their own kidneys. In nephrotoxic nephritis (NTN), animals are injected with heterologous antibodies to GBM, which deposit in kidney and cause transient injury (the heterologous phase). The animal then mounts its own immune response to the foreign immunoglobulin, which acts as a planted antigen on GBM (the autologous phase). This

611

612

Section 3  

the patient with glomerular disease

Table 74.2  Disorders in association with anti-GBM disease (Kalderon et al., 1973; Curtis et al., 1976; Ma et al., 1978; Blake et al., 1980; Kleinknecht et al., 1980; Blake et al., 1980; Wilson and Dixon 1981; Guerin et al., 1990; Tan and Cumming, 1993; Umekawa et al., 1993; Kalluri et al., 1994; Kelly and Haponik, 1994; Wuthrich et al., 1994; Savage et al., 1996; Bindi et al., 1997; Drube et al., 1997; Ahuja et al., 1998; Henderson et al., 1998; Komatsu et al., 1998; Xenocostas et al., 1999; Morello et al., 2001; Curioni et al., 2002; Shaer et al., 2003; Li et al., 2006; Lv et al., 2009; Torok et al., 2010) Immune (presumed)

Glomerulonephritis

Others Lithotripsy

Systemic lupus erythematosus Myasthenia gravis Thrombotic thrombocytopenic purpura Dermatomyositis Thyroiditis Primary biliary cirrhosis Multiple sclerosis Cryoglobulinaemia

Lymphoma Castleman disease Partial lipodystrophy Coeliac disease Inflammatory bowel disease Vitamin B12 deficiency Pneumococcal vaccination Nail-patella syndrome

Probably coincidental

IgA nephropathy Membranoproliferative glomerulonephritis Focal segmental glomerulosclerosis Henoch–Schönlein purpura

Likely real

ANCA associated Membranous vasculitis nephropathy Lymphocyte Diabetic nephropathy depletion (alemtuzumab, HIV) Alport post-transplant Penicillamine therapy

exclusively to the C-terminal globular NC1 domain, a3(IV)NC1 (Leinonen et al., 1999). Epitope mapping has defined major and minor conformational epitopes of α3(IV)NC1, designated as EA, which encompasses residues 17–31, and EB, which encompasses residues 127–141 (Netzer et  al., 1999; David et  al., 2001). The epitopes appear to be structurally sequestered (hidden) to some extent by adjacent α5(IV)NC1 and α4(IV)NC1 in α3.α4.α5(IV) hexamer (Fig. 74.2) (Borza et al., 2002). Perturbation of the quaternary structure of α3.α4.α5(IV) hexamer by denaturation or possibly in conditions of inflammation increases the accessibility of epitopes (Kalluri et  al., 2000). This is the reason that Goodpasture epitopes are described as ‘cryptic’. In addition to reactivity to α3(IV)NC1, which is detectable in all patients, other antibodies recognizing other α chains of collagen IV may be present, such as α1, α2, α4 and α5(IV)NC1 (Johansson et al., 1993; Hellmark et al., 1994; Kalluri et al., 1995, 1996; Dehan et  al., 1996; Ghohestani et  al., 2000, 2003; Yang et  al., 2007; Zhao et  al., 2009b; Pedchenko et  al., 2010). A  few reports suggest that relatively minor disease can occasionally be associated with such antibodies in the absence of reactivity for α3(IV)NC1 (Ghohestani et al., 2000; Chen et al., 2003; Zhao et al., 2009b; Pedchenko et al., 2010). In the presence of linear binding on immunofluorescence and active disease, low α3 titres are a likely alternative explanation for most negative immunoassay results. Western blotting or individual immunoassays at a specialist centre would be required for certainty. Alport anti-GBM disease is a special example which is considered in Chapter 75.

Antibodies against GBM There is good evidence that antibodies against GBM have a directly causal role. In the classic adoptive transfer experiments by Lerner et al. (1967), antibodies eluted from the kidneys of patients with anti-GBM disease and injected into squirrel monkeys could fix to (A)

model can be made more severe by pre-immunizing the animal with Ig from the species in which anti-GBM antibodies are raised (the accelerated model of NTN). These animal models permit studies of the role of various inflammatory mediators in the development of disease (Sado and Naito, 1987; Bolton et  al., 1993; Sado et  al., 1998; Chen et al., 2003, 2006). Conclusions about the pathogenesis of human anti-GBM disease drawn from animal models, however, must be tempered by the realization that findings in animal models might not exactly reflect outcomes in human disease.

Antibody responses Localization of the Goodpasture antigen GBM composition is described in Chapter 320. The Goodpasture antigen-bearing type IV collagen chain, α3, is a significant component of the glomerular and alveolar basement membranes (Saus et  al., 1988; Kalluri et  al., 1995; Pedchenko et  al., 2010), and has been found also in testis, choroid plexus, Bruch’s membrane in the eye, cochlea and neuromuscular junction. However, clinical manifestations outside the kidney and lung are uncommon. Autoantibodies from patients with anti-GBM disease bind

(C)

Collagen network (B)

Crystal structure NC1 domains

Fig. 74.2  (A) Cartoon of the network of the type IV collagen network within GBM. (B) An NC1 hexamer, as produced by digestion of GBM with collagenase. (C) The crystal structure of a type IV collagen NC1 domain hexamer. A model of the α1/α2 hexamer crystal structure. The NC1 domain of a single type IV collagen chain is shown in blue, with its N terminus (where it joins the collagenous part of the molecule) in light blue. The equivalent location of the major B cell epitope on α3(IV)NC1 is shown in red. Fig. 71.1 in Chapter 71 shows the Western blotting pattern of these domains. Modified from Than et al. (2002) by Dr R. Phelps.

Chapter 74 

the GBM of squirrel monkeys in vivo and cause pathological glomerular changes and pulmonary haemorrhage. In patients with anti-GBM disease, antibody concentrations are correlated with the severity of kidney damage at presentation and the raised titres are independently prognostic for kidney outcome and patient survival (Hellmark et al., 1994, 1999; Ang et al., 1998; Cui et al., 2005, 2011a; Yang et al., 2009). The removal of circulating antibodies with plasmapheresis is associated with clinical recovery after lung haemorrhage and renal dysfunction (Lockwood et al., 1975; Jindal, 1999; Levy et al., 2001). Furthermore, anti-GBM disease recurs immediately in renal allografts if circulating anti-GBM antibodies remain positive in recipients (Wilson and Dixon, 1973). The α3(IV)NC1-specific human anti-GBM antibodies can be of any IgG subclass but are typically IgG1 (high affinity for Fcγ receptors, probably Th1 associated) or IgG4 (low affinity for Fcγ receptors, probably Th2 associated), accounting for less than 1% of circulating IgG (Bowman et  al., 1987; Segelmark et  al., 1990; Zhao et al., 2009a). Clinical reports, together with passive transfers of monoclonal anti-GBM antibodies of different IgG subclasses in WKY rats suggests that high affinity for Fcγ receptors is important as an IgG1 (in rodents as IgG2a), but not an IgG4 (rodent IgG1) skewed humoral immune response is pathogenic (Kohda et  al., 2004). Anti-GBM antibodies have high affinity, with relatively high on (binding) rates and slow off (dissociation) rates (Rutgers et al., 2000). The specificities of circulating and tissue bound antibodies are identical (Pedchenko et  al., 2010), with predominant target always α3(IV)NC1, and conformational epitopes, EA and EB, as described above. In healthy individuals, natural antibodies against GBM can be purified from IgG fractions in sera or plasma. They recognize human α3(IV)NC1 and are specific for EA and EB, as are antibodies against GBM in patients, but affinity is lower and antibodies are restricted to IgG2 and IgG4 subclasses (Cui et al., 2010; 181, 182). Their significance is not entirely clear. Studies investigating disease development focus on the differences in immunologic characteristics between natural antibodies

anti-gbm disease: pathogenesis

and disease-associated ones (Table 74.3) (Cui and Zhao, 2005; Cui et al., 2006, 2007; Zhao et al., 2009a, 2009b). In patients with anti-GBM disease and normal kidney function, anti-GBM antibodies in circulation generally present lower titre, lower affinity, less of IgG1 subclass, and limited reaction to α3(IV)NC1 and α5(IV)NC1. In patients with mild and moderate renal dysfunction, the circulating antibodies reveal higher titre, higher affinity, more of IgG1 subclass and broader spectrum of target antigens for the five α chains of type IV collagen. In patients with dialysis-dependent acute renal failure, the antibodies have the highest titre, highest affinity, of IgG1 predominance (Sado et al., 1998; Rutgers et al., 2000; Radeke et al., 2002; Kohda et  al., 2004; Bolton et  al., 2005; Chen et  al., 2006). Table 74.3 shows some of the characteristics of antibodies in different circumstances.

Cell-mediated immunity Although anti-GBM disease is seen as a prototypic autoantibody-mediated disease, there is strong experimental evidence that the full expression of disease is dependent on cell-mediated autoimmunity, in particular autoreactive T cells. T-cell involvement can be implied from the strong HLA associations in human anti-GBM disease. Help from T cells is required for affinity maturation, antigen specificity, subclass switching and epitope spreading of antibody response. Effector T cells also contribute directly to tissue injury of anti-GBM disease with the finding of CD4+ and CD8+ T cells in affected glomeruli (Nolasco et al., 1987). In rodent models, one T-cell epitope, pCol(28–40), derived from α3(IV)NC1, induces severe glomerulonephritis in WKY rats and triggers a diversified anti-GBM antibody response through B-cell spreading of epitopes (Wu et al., 2001). In the absence of antibodies against GBM, α3(IV)NC1-specific CD4+ T cells alone is sufficient to initiate glomerular injury and reveals a direct pathogenic role (Wu et al., 2001, 2003, 2004; Robertson et al., 2005; Arends et  al., 2006). The disease can be inhibited by anti-T-cell therapies, including CD28–B7 or CD154–CD40 co-stimulatory blockade (Reynolds et al., 2000, 2004), and anti-CD4 and anti-CD8

Table 74.3  Different characteristics of anti-GBM antibodies in healthy versus diseased individuals Patient group

Percentage of IgG

Titre

Affinity (aK)

IgG subtype

Target antigens

IgG2 (100%) IgG4 (100%)

α3(IV)NC1 (100%) α4(IV)NC1 (100%)

Healthy individuals with natural anti-GBM antibodies

0.5%

1:60

9.09 × 106 M–1

Patients with pathogenic anti-GBM antibodies and normal renal function

Unknown

1:200

4.0 × 107 M–1

IgG1 (8%) IgG2 (62%) IgG4 (39%)

α3(IV)NC1 (100%) α4(IV)NC1 (14%) α5(IV)NC1 (86%)

Patients with pathogenic anti-GBM antibodies and moderate renal dysfunction

Unknown

1:400

2.4 × 108 M–1

IgG1 (69%) IgG2 (59%) IgG3 (35%) IgG4 (66%)

α3(IV)NC1 (100%) α1(IV)NC1 (40%) α2(IV)NC1 (35%) α4(IV) NC1 (45%) α5(IV)NC1 (70%)

Patients with pathogenic anti-GBM antibodies and severe renal injury and dependent on dialysis

1%

1:800

3.26 × 108 M–1

IgG1 (94%) IgG2 (87%) IgG3 (32%) IgG4 (61%)

α3(IV)NC1 (100%) α1(IV)NC1 (50%) α2(IV)NC1 (43%) α4(IV) NC1 (60%) α5(IV)NC1 (93%)

aK = affinity constant of antibodies against GBM, measured as the reciprocal value of molar concentration of α(IV)NC1 needed for 50% inhibition of the binding capacity; GBM = glomerular basement membrane.

613

614

Section 3  

the patient with glomerular disease

monoclonal antibodies (Reynolds et  al., 2002). Furthermore, T-cell tolerance can be induced via oral or nasal administration of Goodpasture antigen prior to induction of disease (Reynolds et al., 2005). In EAG models, Th1 responses are found pathogenetic, while Th2 responses are protective (Holdsworth et al., 1999; Phoon et al., 2008). A  polarity shift from Th2 to Th1 response seems feasible during disease process. Besides that, Th17 cell subset, which is maintained by IL-23, plays a dominant part in the development of disease (Holdsworth et al., 1999; Hopfer et al., 2003; Phoon et al., 2008; Ooi et al., 2009; Summers et al., 2009). Mice deficient in IL-23 immunized with α3(IV)NC1 develop anti-GBM disease, although autoantibody titres are lower, cellular reactivity is reduced, and renal injury is less severe than in mice with IL-23 (Ooi et al., 2009). A move from an antibody-associated phase to a more aggressive phase associated with Th1 and Th17 CD4+ T cells was described in DBA/1 mice (Hopfer et al. 2012). α3 and α4 and α5 chains of type IV collagen are all expressed in normal human thymus (Wong et  al., 2001). It is not known what proportion of T cells that recognize the respective α chains are deleted in the thymus, but some α3(IV)NC1-specific CD4+ T cells escape thymic deletion and exist in the periphery circulation in healthy individuals (Zou et al., 2008). Peripheral blood T cells from healthy individuals have similar specificities to the α3(IV)NC1-reactive T cells found in patients with anti-GBM disease. However, T cells that proliferate in response to α3(IV)NC1 are much less abundant in healthy individuals, in the absence of stimulation. Therefore, the naïve autoimmunity to α3(IV)NC1 can be classified as quiescent. In patients with active anti-GBM disease, concentrations of autoreactive T cells at onset are higher than that in healthy controls, decline in remission and reach to similar levels of healthy controls after several years (Salama et al., 2001). Unlike most autoimmune diseases, anti-GBM disease does not follow a relapsing and remitting course. Treatment reduces the time to disappearance of antibody, but even without treatment most patients no longer have detectable circulating antibody after 12–18 months (Levy et al., 1996). Re-establishment of autoimmune tolerance coincides with the alteration and persistence of α3(IV) NC1 reactive T cells from T-helper cells to T-regulatory cells during the evolution of disease (Cairns et al., 2003; Salama et al., 2003). These could be responsible for terminating the anti-GBM response. Regulatory CD4+CD25+ T cell depletion in peripheral blood mononuclear cells from convalescent patients markedly increases the number of Goodpasture antigen-specific IFN-γ producing cells (by ELISPOT) (Salama et al., 2003). In EAG models, transfer of regulatory CD4+CD25+ T cells prior to disease induction significantly attenuates glomerular injury (Wolf et al., 2005).

T-cell epitope T cells recognize antigen only when presented in the form of peptides bound to MHC class  II molecules on the surface of APCs. Thus, T-cell epitopes are critically dependent on the peptides generated by digestion within APCs (antigen processing) and the ability of these peptides to bind MHC class II molecules (antigen presentation). α3(IV)NC1-derived peptides have been characterized that are naturally processed and presented by cells bearing HLA-DRB1*1501, which consists of nested sets centred on core MHC-binding motifs and overlaps the major autoantibody epitope. The linear epitopes of α3(IV)NC1 recognized by autoreactive T

cells have proven hard to define (Hellmark et al., 1996). A mapping study of human T-cell epitopes has defined the specificity of α3(IV)NC1 reactive T cells as being highly focused on two peptides: α371–90 and α3131–150 (Cairns et al., 2003). The epitopes are highly susceptible to early endopeptidases, such as cathepsins D and E, and might be destroyed by APCs before presentation in the thymus. The α3131–150 epitope overlaps with EB region and binds with high affinity to the disease-associated allele HLA-DRB1*1501. The key residues have been mapped to residues α3134–148, which stimulate T cells from patients with anti-GBM disease to proliferate and secrete IFN-γ (Zou et al., 2007). The same peptide was identified by a different method of mapping and modelling in murine models by Ooi et al. (2013).

References Abrahamson, D. R., Hudson, B. G., Stroganova, L., et al. (2009). Cellular origins of type IV collagen networks in developing glomeruli. J Am Soc Nephrol, 20, 1471–9. Agodoa, L. C., Striker, G. E., George, C. R., et al. (1976). The appearance of nonlinear deposits of immunoglobulins in Goodpasture’s syndrome. Am J Med, 61, 407–13. Ahuja, T. S., Velasco, A., Deiss, W., Jr., et al. (1998). Diabetic nephropathy with anti-GBM nephritis. Am J Kidney Dis, 31, 127–30. Aitman, T. J., Dong, R., Vyse, T. J., et al. (2006). Copy number polymorphism in Fcgr3 predisposes to glomerulonephritis in rats and humans. Nature, 439, 851–5. Ang, C., Savige, J., Dawborn, J., et al. (1998). Anti-glomerular basement membrane (GBM)-antibody-mediated disease with normal renal function. Nephrol Dial Transplant, 13, 935–9. Appel, G. B., Radhakrishnan, J., and D’Agati, V. (2007). Secondary glomerular disease. In B. M. Brenner (ed.) Brenner and Rector’s The Kidney, pp. 1067–126. Philadelphia, PA: Saunders. Arends, J., Wu, J., Borillo, J., et al. (2006). T cell epitope mimicry in antiglomerular basement membrane disease. J Immunol, 176, 1252–8. Basford, A. W., Lewis, J., Dwyer, J. P., et al. (2011). Membranous nephropathy with crescents. J Am Soc Nephrol, 22, 1804–8. Bindi, P., Gilson, B., Aymard, B., et al. (1997). Antiglomerular basement membrane glomerulonephritis following D-penicillamine-associated nephrotic syndrome. Nephrol Dial Transplant, 12, 325–7. Blake, D. R., Rashid, H., McHugh, M., et al. (1980). A possible association of partial lipodystrophy with anti-GBM nephritis (Goodpasture’s syndrome). Postgrad Med J, 56, 137–9. Bolton, W. K., Chen, L., Hellmark, T., et al. (2005). Epitope spreading and autoimmune glomerulonephritis in rats induced by a T cell epitope of Goodpasture’s antigen. J Am Soc Nephrol, 16, 2657–66. Bolton, W. K., May, W. J., and Sturgill, B. C. (1993). Proliferative autoimmune glomerulonephritis in rats: a model for autoimmune glomerulonephritis in humans. Kidney Int, 44, 294–306. Bombassei, G. J. and Kaplan, A. A. (1992). The association between hydrocarbon exposure and anti-glomerular basement membrane antibody-mediated disease (Goodpasture’s syndrome). Am J Ind Med, 21, 141–53. Borza, D. B., Bondar, O., Colon, S., et al. (2005). Goodpasture autoantibodies unmask cryptic epitopes by selectively dissociating autoantigen complexes lacking structural reinforcement: novel mechanisms for immune privilege and autoimmune pathogenesis. J Biol Chem, 280, 27147–54. Borza, D. B., Bondar, O., Todd, P., et al. (2002). Quaternary organization of the goodpasture autoantigen, the alpha 3(IV) collagen chain. Sequestration of two cryptic autoepitopes by intrapromoter interactions with the alpha4 and alpha5 NC1 domains. J Biol Chem, 277, 40075–83. Bowman, C., Ambrus, K., and Lockwood, C. M. (1987). Restriction of human IgG subclass expression in the population of auto-antibodies to glomerular basement membrane. Clin Exp Immunol, 69, 341–9.

Chapter 74 

Cairns, L. S., Phelps, R. G., Bowie, L., et al. (2003). The fine specificity and cytokine profile of T-helper cells responsive to the alpha3 chain of type IV collagen in Goodpasture’s disease. J Am Soc Nephrol, 14, 2801–12. Chen, L., Hellmark, T., Pedchenko, V., et al. (2006). A nephritogenic peptide induces intermolecular epitope spreading on collagen IV in experimental autoimmune glomerulonephritis. J Am Soc Nephrol, 17, 3076–81. Chen, L., Hellmark, T., Wieslander, J. et al. (2003). Immunodominant epitopes of alpha3(IV)NC1 induce autoimmune glomerulonephritis in rats. Kidney Int, 64, 2108–20. Coles, A. J., Fox, E., Vladic, A., et al. (2012). Alemtuzumab more effective than interferon β-1a at 5-year follow-up of CAMMS223 clinical trial. Neurology, 78(14), 1069–78. Cui, Z. and Zhao, M. H. (2005). Avidity of anti-glomerular basement membrane autoantibodies was associated with disease severity. Clin Immunol, 116, 77–82. Cui, Z., Wang, H. Y., and Zhao, M. H. (2006). Natural autoantibodies against glomerular basement membrane exist in normal human sera. Kidney Int, 69, 894–9. Cui, Z., Zhao, J., Jia, X. Y., et al. (2011). Anti-glomerular basement membrane disease: outcomes of different therapeutic regimens in a large single-center chinese cohort study. Medicine (Baltimore), 90, 303–11. Cui, Z., Zhao, J., Jia, X. Y., et al. (2011b). Clinical features and outcomes of anti-glomerular basement membrane disease in older patients. Am J Kidney Dis, 57, 575–82. Cui, Z., Zhao, M. H., Segelmark, M., et al. (2010). Natural autoantibodies to myeloperoxidase, proteinase 3, and the glomerular basement membrane are present in normal individuals. Kidney Int, 78, 590–7. Cui, Z., Zhao, M. H., Singh, A. K., et al. (2007). Antiglomerular basement membrane disease with normal renal function. Kidney Int, 72, 1403–8. Cui, Z., Zhao, M. H., Wang, S. X., et al. (2006). Concurrent antiglomerular basement membrane disease and immune complex glomerulonephritis. Ren Fail, 28, 7–14. Cui, Z., Zhao, M. H., Xin, G., et al. (2005). Characteristics and prognosis of Chinese patients with anti-glomerular basement membrane disease. Nephron Clin Pract, 99, c49–55. Curioni, S., Ferrario, F., Rastaldi, M. P., et al. (2002). Anti-GBM nephritis complicating diabetic nephropathy. J Nephrol, 15, 83–7. Curtis, J. J., Bhathena, D., Leach, R. P., et al. (1976). Goodpasture’s syndrome in a patient with the Nail-Patella syndrome. Am J Med, 61, 401–6. D’Apice, A. J., Kincaid-Smith, P., Becker, G. H., et al. (1978). Goodpasture’s syndrome in identical twins. Ann Intern Med, 88, 61–2. Daly, C., Conlon, P. J., Medwar, W., et al. (1996). Characteristics and outcome of anti-glomerular basement membrane disease: a single-center experience. Ren Fail, 18, 105–12. David, M., Borza, D. B., Leinonen, A., et al. (2001). Hydrophobic amino acid residues are critical for the immunodominant epitope of the Goodpasture autoantigen. A molecular basis for the cryptic nature of the epitope. J Biol Chem, 276, 6370–7. Dehan, P., Weber, M., Zhang, X., et al. (1996). Sera from patients with anti-GBM nephritis including goodpasture syndrome show heterogenous reactivity to recombinant NC1 domain of type IV collagen alpha chains. Nephrol Dial Transplant, 11, 2215–22. Donaghy, M. and Rees, A. J. (1983). Cigarette smoking and lung haemorrhage in glomerulonephritis caused by autoantibodies to glomerular basement membrane. Lancet, 2, 1390–3. Downie, G. H., Roholt, O. A., Jennings, L., et al. (1982). Experimental anti-alveolar basement membrane antibody-mediated pneumonitis. II. Role of endothelial damage and repair, induction of autologous phase, and kinetics of antibody deposition in Lewis rats. J Immunol, 129, 2647–52. Drube, S., Maurin, N., and Sieberth, H. G. (1997). Coincidence of myasthenia gravis and antiglomerular basement membrane glomerulonephritis: a combination of two antibody-mediated autoimmune diseases on day 15. Nephrol Dial Transplant, 12, 1478–80. Duncan, D. A., Drummond, K. N., Michael, A. F. et al. (1965). Pulmonary hemorrhage and glomerulonephritis. report on six cases and study of

anti-gbm disease: pathogenesis

the renal lesion by the fluorescent antibody technique and electron microscopy. Ann Intern Med, 62, 920–38. Dunckley, H., Chapman, J. R., Burke, J., et al. HLA-DR and -DQ genotyping in anti-GBM disease. Dis Markers, 9, 249–56. Fisher, M., Pusey, C. D., Vaughan, R. W., et al. (1997). Susceptibility to anti-glomerular basement membrane disease is strongly associated with HLA-DRB1 genes. Kidney Int, 51, 222–9. Garcia-Rostan y Perez, G. M., Garcia Bragado, F., and Puras Gil, A.,M. (1997). Pulmonary hemorrhage and antiglomerular basement membrane antibody-mediated glomerulonephritis after exposure to smoked cocaine (crack): a case report and review of the literature. Pathol Int, 47, 692–7. Ghohestani, R. F., Hudson, B. G., Claudy, A., et al. (2000). The alpha 5 chain of type IV collagen is the target of IgG autoantibodies in a novel autoimmune disease with subepidermal blisters and renal insufficiency. J Biol Chem, 275, 16002–6. Ghohestani, R. F., Rotunda, S. L., Hudson, B., et al. (2003). Crescentic glomerulonephritis and subepidermal blisters with autoantibodies to alpha5 and alpha6 chains of type IV collagen. Lab Invest 83, 605–11. Goodpasture, E. W. (1919). The significance of certain pulmonary lesions in relation to the etiology of influenza. Am J Med Sci, 158, 863–70. Gossain, V. V., Gerstein, A. R., and Janes, A. W. (1972). Goodpasture’s syndrome: a familial occurrence. Am Rev Respir Dis, 105, 621–4. Guerin, V., Rabian, C., Noel, L. H., et al. (1990). Anti-glomerular-basement-membrane disease after lithotripsy. Lancet, 335, 856–7. Hartle, P. M., Carlo, M. E., Dwyer, J. P., et al. (2013) AKI in an HIV patient. J Am Soc Nephrol, 24, 1204–8. Hecht, N., Omoloja, A., Witte, D., et al. (2008). Evolution of antiglomerular basement membrane glomerulonephritis into membranous glomerulonephritis. Pediatr Nephrol, 23, 477–80. Hellmark, T., Brunmark, C., Trojnar, J., et al. (1996). Epitope mapping of anti-glomerular basement membrane (GBM) antibodies with synthetic peptides. Clin Exp Immunol, 105, 504–10. Hellmark, T., Johansson, C. and Wieslander, J. (1994). Characterization of anti-GBM antibodies involved in Goodpasture’s syndrome. Kidney Int, 46, 823–9. Hellmark, T., Niles, J. L., Collins, A. B., et al. (1997). Comparison of anti-GBM antibodies in sera with or without ANCA. J Am Soc Nephrol, 8, 376–85. Hellmark, T., Segelmark, M., Unger, C., et al. (1999). Identification of a clinically relevant immunodominant region of collagen IV in Goodpasture disease. Kidney Int, 55, 936–44. Henderson, R. D., Saltissi, D., and Pender, M. P. (1998). Goodpasture’s syndrome associated with multiple sclerosis. Acta Neurol Scand, 98, 134–5. Holdsworth, S. R., Kitching, A. R., and Tipping, P. G. (1999). Th1 and Th2 T helper cell subsets affect patterns of injury and outcomes in glomerulonephritis. Kidney Int, 55, 1198–216. Hopfer, H., Holzer, J., Hunemorder, S., et al. (2012). Characterization of the renal CD4+ T-cell response in experimental autoimmune glomerulonephritis. Kidney Int, 82, 60–71. Hopfer, H., Maron, R., Butzmann, U., et al. (2003). The importance of cell-mediated immunity in the course and severity of autoimmune anti-glomerular basement membrane disease in mice. FASEB J, 17, 860–8. Hudson, B. G., Tryggvason, K., Sundaramoorthy, M., et al. (2003). Alport’s syndrome, Goodpasture’s syndrome, and type IV collagen. N Engl J Med, 348, 2543–56. Huey, B., McCormick, K., Capper, J., et al. Associations of HLA-DR and HLA-DQ types with anti-GBM nephritis by sequence-specific oligonucleotide probe hybridization. Kidney Int, 44, 307–12. Jennings, L., Roholt, O. A., Pressman, D., et al. (1981). Experimental anti-alveolar basement membrane antibody-mediated pneumonitis. I. The role of increased permeability of the alveolar capillary wall induced by oxygen. J Immunol, 127, 129–34. Jindal, K. K. (1999). Management of idiopathic crescentic and diffuse proliferative glomerulonephritis: evidence-based recommendations. Kidney Int, Suppl 70, S33–40.

615

616

Section 3  

the patient with glomerular disease

Johansson, C., Butkowski, R., Swedenborg, P., et al. (1993). Characterization of a non-Goodpasture autoantibody to type IV collagen. Nephrol Dial Transplant, 8, 1205–10. Jones, J. L., Thompson, S. A., Loh, P., et al. (2013). Human autoimmunity after lymphocyte depletion is caused by homeostatic T-cell proliferation. Proc Natl Acad Sci U S A, 110(50), 20200–5. Kalderon, A. E., Bogaars, H. A., and Diamond, I. (1973). Ultrastructural alterations of the follicular basement membrane in Hashimoto’s thyroiditis. Report of eight cases with basement deposits. Am J Med, 55, 485–91. Kalluri, R., Cantley, L. G., Kerjaschki, D., et al. (2000). Reactive oxygen species expose cryptic epitopes associated with autoimmune goodpasture syndrome. J Biol Chem, 275, 20027–32. Kalluri, R., Danoff, T.M., Okada, H., et al. (1997). Susceptibility to anti-glomerular basement membrane disease and Goodpasture syndrome is linked to MHC class II genes and the emergence of T cell-mediated immunity in mice. J Clin Invest, 100, 2263–75. Kalluri, R., Petrides, S., Wilson, C. B., et al. (1996). Anti-alpha1(IV) collagen autoantibodies associated with lung adenocarcinoma presenting as the Goodpasture syndrome. Ann Intern Med, 124, 651–3. Kalluri, R., Weber, M., Netzer, K. O., et al. (1994). COL4A5 gene deletion and production of post-transplant anti-alpha 3(IV) collagen alloantibodies in Alport syndrome. Kidney Int, 45, 721–6. Kalluri, R., Wilson, C. B., Weber, M., et al. (1995). Identification of the alpha 3 chain of type IV collagen as the common autoantigen in antibasement membrane disease and Goodpasture syndrome. J Am Soc Nephrol, 6, 1178–85. Keller, F. and Nekarda, H. (1985). Fatal relapse in Goodpasture’s syndrome 3 years after plasma exchange. Respiration, 48, 62–6. Kelly, P. T. and Haponik, E. F. (1994). Goodpasture syndrome: molecular and clinical advances. Medicine (Baltimore), 73, 171–85. Kielstein, J. T., Helmchen, U., Netzer, K. O., et al. (2001). Conversion of Goodpasture’s syndrome into membranous glomerulonephritis. Nephrol Dial Transplant, 16, 2082–5. Kitagawa, W., Imai, H., Komatsuda, A., et al. (2008). The HLA-DRB1*1501 allele is prevalent among Japanese patients with anti-glomerular basement membrane antibody-mediated disease. Nephrol Dial Transplant, 23, 3126–9. Klassen, J., Elwood, C., Grossberg, A. L., et al. (1974). Evolution of membranous nephropathy into anti-glomerular-basement-membrane glomerulonephritis. N Engl J Med, 290, 1340–4. Kleinknecht, D., Morel-Maroger, L., Callard, P., et al. (1980). Antiglomerular basement membrane nephritis after solvent exposure. Arch Intern Med, 140, 230–2. Kohda, T., Okada, S., Hayashi, A., et al. (2004). High nephritogenicity of monoclonal antibodies belonging to IgG2a and IgG2b subclasses in rat anti-GBM nephritis. Kidney Int, 66, 177–86. Komatsu, T., Utsunomiya, K., and Oyaizu, T. (1998). Goodpasture’s syndrome associated with primary biliary cirrhosis. Intern Med, 37, 611–13. Kurki, P., Helve, T., von Bonsdorff, M., et al. (1984). Transformation of membranous glomerulonephritis into crescentic glomerulonephritis with glomerular basement membrane antibodies. Serial determinations of anti-GBM before the transformation. Nephron, 38, 134–7. Leinonen, A., Netzer, K. O., Boutaud, A., et al. (1999). Goodpasture antigen: expression of the full-length alpha3(IV) chain of collagen IV and localization of epitopes exclusively to the noncollagenous domain. Kidney Int, 55, 926–35. Lerner, R. A., Glassock, R. J., and Dixon, F. J. (1967). The role of anti-glomerular basement membrane antibody in the pathogenesis of human glomerulonephritis. J Exp Med, 126, 989–1004. Levy, J. B., Hammad, T., Coulthart, A., et al. (2004). Clinical features and outcome of patients with both ANCA and anti-GBM antibodies. Kidney Int, 66, 1535–40. Levy, J. B., Lachmann, R. H., and Pusey, C. D. (1996). Recurrent Goodpasture’s disease. Am J Kidney Dis, 27, 573–8. Levy, J. B., Turner, A. N., and Rees, A. J. (2001). Long-term outcome of anti-glomerular basement membrane antibody disease treated with

plasma exchange and immunosuppression. Ann Intern Med, 134, 1033–42. Li, C. H., Li, Y. C., Xu, P. S., et al. (2006). Clinical significance of anti-glomerular basement membrane antibodies in a cohort of Chinese patients with lupus nephritis. Scand J Rheumatol, 35, 201–8. Lionaki, S., Jennette, J. C. and Falk, R. J. (2007). Anti-neutrophil cytoplasmic (ANCA) and anti-glomerular basement membrane (GBM) autoantibodies in necrotizing and crescentic glomerulonephritis. Semin Immunopathol, 29, 459–74. Lockwood, C. M., Boulton-Jones, J. M., Lowenthal, R. M., et al. (1975). Recovery from Goodpasture’s syndrome after immunosuppressive treatment and plasmapheresis. Br Med J, 2, 252–54. Lockwood, C. M., Rees, A. J., Pearson, T. A., et al. (1976). Immunosuppression and plasma-exchange in the treatment of Goodpasture’s syndrome. Lancet, 1, 711–15. Lv, J., Zhang, H., Zhou, F., et al. (2009). Antiglomerular basement membrane disease associated with Castleman disease. Am J Med Sci, 337, 206–9. Ma, K. W., Golbus, S. M., Kaufman, R., et al. (1978). Glomerulonephritis with Hodgkin’s disease and herpes zoster. Arch Pathol Lab Med, 102, 527–9. McCoy, R. C., Johnson, H. K., Stone, W. J., et al. (1982). Absence of nephritogenic GBM antigen(s) in some patients with hereditary nephritis. Kidney Int, 21(4), 642–52. Monteiro, E. J., Caron, D., Balda, C. A., et al. (2006). Anti-glomerular basement membrane glomerulonephritis in an HIV positive patient: case report. Braz J Infect Dis, 10, 55–8. Morello, R., Zhou, G., Dreyer, S. D., et al. (2001). Regulation of glomerular basement membrane collagen expression by LMX1B contributes to renal disease in nail patella syndrome. Nat Genet, 27, 205–8. Nachman, P. H., Jennette, J. C., and Falk, R. J. (2007). Primary glomerular disease. in B.M. Brenner (ed.) Brenner and Rector’s The Kidney, pp. 987–1279. Philadelphia, PA: Saunders. Nagashima T, Ubara Y, Tagami T, et al. (2002). Anti-glomerular basement membrane antibody disease: a case report and a review of Japanese patients with and without alveolar hemorrhage. Clin Exp Nephrol, 6, 49–57. Netzer, K. O., Leinonen, A., Boutaud, A., et al. (1999). The goodpasture autoantigen. Mapping the major conformational epitope(s) of alpha3(IV) collagen to residues 17-31 and 127-141 of the NC1 domain. J Biol Chem, 274, 11267–74. Nolasco, F. E., Cameron, J. S., Hartley, B., et al. (1987). Intraglomerular T cells and monocytes in nephritis: study with monoclonal antibodies. Kidney Int, 31, 1160–66. O’Donoghue, D. J., Short, C. D., Brenchley, P. E., et al. (1989). Sequential development of systemic vasculitis with anti-neutrophil cytoplasmic antibodies complicating anti-glomerular basement membrane disease. Clin Nephrol, 32, 251–5. Olson, S. W., Arbogast, C. V., Baker, T. P., et al. (2011). Asymptomatic autoantibodies associate with future anti-glomerular basement membrane disease. J Am Soc Nephrol, 22, 1946–52. Ooi, J. D., Chang, J., O’Sullivan, K. M., et al. (2013). The HLA-DRB1*15:01-restricted Goodpasture’s T cell epitope induces GN. J Am Soc Nephrol, 24, 419–31. Ooi, J. D., Phoon, R. K., Holdsworth, S. R., et al. (2009). IL-23, not IL-12, directs autoimmunity to the Goodpasture antigen. J Am Soc Nephrol, 20, 980–9. Pasternack, A., Tornroth, T., and Linder, E. (1978). Evidence of both anti-GBM and immune complex mediated pathogenesis in the initial phase of Goodpasture’s syndrome. Clin Nephrol, 9, 77–85. Peces, R., Navascues, R. A., Baltar, J., et al. (1999). Antiglomerular basement membrane antibody-mediated glomerulonephritis after intranasal cocaine use. Nephron, 81, 434–8. Peces, R., Rodriguez, M., Pobes, A., et al. (2000). Sequential development of pulmonary hemorrhage with MPO-ANCA complicating anti-glomerular basement membrane antibody-mediated glomerulonephritis. Am J Kidney Dis, 35, 954–7.

Chapter 74 

Pedchenko, V., Bondar, O., Fogo, A. B., et al. (2010). Molecular architecture of the Goodpasture autoantigen in anti-GBM nephritis. N Engl J Med, 363, 343–54. Phelps, R. G. and Rees, A. J. (1999). The HLA complex in Goodpasture’s disease: a model for analyzing susceptibility to autoimmunity. Kidney Int, 56, 1638–53. Phelps, R. G., Jones, V., Turner, A. N., et al. (2000). Properties of HLA class II molecules divergently associated with Goodpasture’s disease. Int Immunol, 12, 1135–43. Phoon, R. K., Kitching, A. R., Odobasic, D., et al. (2008). T-bet deficiency attenuates renal injury in experimental crescentic glomerulonephritis. J Am Soc Nephrol, 19, 477–85. Radeke, H. H., Janssen-Graalfs, I., Sowa, E. N., et al. (2002). Opposite regulation of type II and III receptors for immunoglobulin G in mouse glomerular mesangial cells and in the induction of anti-glomerular basement membrane (GBM) nephritis. J Biol Chem, 277, 27535–44. Rees, A. J., Demaine, A. G., and Welsh, K. I. (1984). Association of immunoglobulin Gm allotypes with antiglomerular basement membrane antibodies and their titer. Hum Immunol, 10, 213–20. Rees, A. J., Peters, D. K., Compston, D. A., et al. (1978). Strong association between HLA-DRW2 and antibody-mediated Goodpasture’s syndrome. Lancet, 1, 966–8. Reynolds, J., Albouainain, A., Duda, M. A., et al. (2006). Strain susceptibility to active induction and passive transfer of experimental autoimmune glomerulonephritis in the rat. Nephrol Dial Transplant, 21, 3398–408. Reynolds, J., Khan, S. B., Allen, A. R., et al. (2004). Blockade of the CD154-CD40 costimulatory pathway prevents the development of experimental autoimmune glomerulonephritis. Kidney Int, 66, 1444–52. Reynolds, J., Norgan, V. A., Bhambra, U., et al. (2002). Anti-CD8 monoclonal antibody therapy is effective in the prevention and treatment of experimental autoimmune glomerulonephritis. J Am Soc Nephrol, 13, 359–69. Reynolds, J., Prodromidi, E. I., Juggapah, J. K., et al. (2005). Nasal administration of recombinant rat alpha3(IV)NC1 prevents the development of experimental autoimmune glomerulonephritis in the WKY rat. J Am Soc Nephrol, 16, 1350–9. Reynolds, J., Tam, F. W., Chandraker, A., et al. (2000). CD28-B7 blockade prevents the development of experimental autoimmune glomerulonephritis. J Clin Invest, 105, 643–51. Robertson, J., Wu, J., Arends, J., et al. (2005). Activation of glomerular basement membrane-specific B cells in the renal draining lymph node after T cell-mediated glomerular injury. J Am Soc Nephrol, 16, 3256–63. Rutgers, A., Meyers, K. E., Canziani, G., et al. (2000). High affinity of anti-GBM antibodies from Goodpasture and transplanted Alport patients to alpha3(IV)NC1 collagen. Kidney Int, 58, 115–22. Rutgers, A., Slot, M., van Paassen, P., et al. (2005). Coexistence of anti-glomerular basement membrane antibodies and myeloperoxidase-ANCAs in crescentic glomerulonephritis. Am J Kidney Dis, 46, 253–62. Sado, Y. and Naito, I. (1987). Experimental autoimmune glomerulonephritis in rats by soluble isologous or homologous antigens from glomerular and tubular basement membranes. Br J Exp Pathol, 68, 695–704. Sado, Y., Boutaud, A., Kagawa, M., et al. (1998). Induction of anti-GBM nephritis in rats by recombinant alpha 3(IV)NC1 and alpha 4(IV)NC1 of type IV collagen. Kidney Int, 53, 664–71. Salama, A. D., Chaudhry, A. N., Holthaus, K. A., et al. (2003). Regulation by CD25+ lymphocytes of autoantigen-specific T-cell responses in Goodpasture’s (anti-GBM) disease. Kidney Int, 64, 1685–94. Salama, A. D., Chaudhry, A. N., Ryan, J. J., et al. (2001). In Goodpasture’s disease, CD4(+) T cells escape thymic deletion and are reactive with the autoantigen alpha3(IV)NC1. J Am Soc Nephrol, 12, 1908–15. Saus, J., Wieslander, J., Langeveld, J. P., et al. (1988). Identification of the Goodpasture antigen as the alpha 3(IV) chain of collagen IV. J Biol Chem, 263, 13374–80.

anti-gbm disease: pathogenesis

Savage, C. O., Pusey, C. D., Bowman, C., et al. (1986). Antiglomerular basement membrane antibody mediated disease in the British Isles 1980-4. Br Med J (Clin Res Ed), 292, 301–4. Savige, J. A., Chang, L., Horn, S., et al. (1994). Anti-nuclear, anti-neutrophil cytoplasmic and anti-glomerular basement membrane antibodies in HIV-infected individuals. Autoimmunity, 18, 205–11. Savige, J. A., Dowling, J., and Kincaid-Smith, P. (1989). Superimposed glomerular immune complexes in anti-glomerular basement membrane disease. Am J Kidney Dis, 14, 145–53. Segelmark, M., Butkowski, R., and Wieslander, J. (1990). Antigen restriction and IgG subclasses among anti-GBM autoantibodies. Nephrol Dial Transplant, 5, 991–6. Shaer, A. J., Stewart, L. R., Cheek, D. E., et al. (2003). IgA antiglomerular basement membrane nephritis associated with Crohn’s disease: a case report and review of glomerulonephritis in inflammatory bowel disease. Am J Kidney Dis, 41, 1097–109. Shah, M. K. (2002). Outcomes in patients with Goodpasture’s syndrome and hydrocarbon exposure. Ren Fail, 24, 545–55. Short, A. K., Esnault, V. L., and Lockwood, C. M. (1995). Anti-neutrophil cytoplasm antibodies and anti-glomerular basement membrane antibodies: two coexisting distinct autoreactivities detectable in patients with rapidly progressive glomerulonephritis. Am J Kidney Dis, 26, 439–45. Simonsen, H., Brun, C., Thomsen, O. F., et al. (1982). Goodpasture’s syndrome in twins. Acta Med Scand, 212, 425–8. Stanton, M. C. and Tange, J. D. (1958). Goodpasture’s syndrome (pulmonary haemorrhage associated with glomerulonephritis). Australas Ann Med, 7, 132–44. Stevenson, A., Yaqoob, M., Mason, H., et al. (1995). Biochemical markers of basement membrane disturbances and occupational exposure to hydrocarbons and mixed solvents. QJM, 88, 23–8. Summers, S. A., Steinmetz, O. M., Li, M., et al. (2009). Th1 and Th17 cells induce proliferative glomerulonephritis. J Am Soc Nephrol, 20, 2518–24. Szczech, L. A., Anderson, A., Ramers, C., et al. (2006). The uncertain significance of anti-glomerular basement membrane antibody among HIV-infected persons with kidney disease. Am J Kidney Dis, 48, e55–9. Tan, S. Y. and Cumming, A. D. (1993). Vaccine related glomerulonephritis. BMJ, 306, 248. Than, M. E., Henrich, S., Huber, R., et al. (2002). The 1.9-A crystal structure of the noncollagenous (NC1) domain of human placenta collagen IV shows stabilization via a novel type of covalent Met-Lys cross-link. Proc Natl Acad Sc U S A, 99, 6607–12. Torok, N., Niazi, M., Al Ahwel, Y., et al. (2010).Thrombotic thrombocytopenic purpura associated with anti-glomerular basement membrane disease. Nephrol Dial Transplant, 25, 3446–9. Umekawa, T., Kohri, K., Iguchi, M., et al. (1993). Glomerular-basement-membrane antibody and extracorporeal shock wave lithotripsy. Lancet, 341, 556. Vanacore, R. M., Ham, A. J., Cartailler, J. P., et al. (2008). A role for collagen IV cross-links in conferring immune privilege to the Goodpasture autoantigen: structural basis for the crypticity of B cell epitopes. J Biol Chem, 283, 22737–48. Vanacore, R., Ham, A. J., Voehler, M., et al. (2009). A sulfilimine bond identified in collagen IV. Science, 325, 1230–4. Wang, X. P., Fogo, A. B., Colon, S., et al. (2005). Distinct epitopes for anti-glomerular basement membrane alport alloantibodies and goodpasture autoantibodies within the noncollagenous domain of alpha3(IV) collagen: a janus-faced antigen. J Am Soc Nephrol, 16, 3563–71. Wechsler, E., Yang, T., Jordan, S. C., et al. (2008). Anti-glomerular basement membrane disease in an HIV-infected patient. Nat Clin Pract Nephrol, 4, 167–71, Wilson, C. B. and Dixon, F. J. (1973). Anti-glomerular basement membrane antibody-induced glomerulonephritis. Kidney Int, 3, 74–89. Wilson, C. B. and Smith, R. C. (1972). Goodpasture’s syndrome associated with influenza A2 virus infection. Ann Intern Med, 76, 91–4.

617

618

Section 3  

the patient with glomerular disease

Wolf, D., Hochegger, K., Wolf, A. M., et al. (2005). CD4+CD25+ regulatory T cells inhibit experimental anti-glomerular basement membrane glomerulonephritis in mice. J Am Soc Nephrol, 16, 1360–70. Wong, D., Phelps, R. G. and Turner, A. N. (2001). The Goodpasture antigen is expressed in the human thymus. Kidney Int, 60, 1777–83. Wu, J., Arends, J., Borillo, J., et al. (2004). A self T cell epitope induces autoantibody response: mechanism for production of antibodies to diverse glomerular basement membrane antigens. J Immunol, 172, 4567–74. Wu, J., Borillo, J., Glass, W. F., et al. (2003). T-cell epitope of alpha3 chain of type IV collagen induces severe glomerulonephritis. Kidney Int, 64, 1292–301. Wu, J., Hicks, J., Ou, C., et al. (2001). Glomerulonephritis induced by recombinant collagen IV alpha 3 chain noncollagen domain 1 is not associated with glomerular basement membrane antibody: a potential T cell-mediated mechanism. J Immunol, 167, 2388–95. Wuthrich, R. P. (1994). Pernicious anemia, autoimmune hypothyroidism and rapidly progressive anti-GBM glomerulonephritis. Clin Nephrol, 42, 404. Xenocostas, A., Jothy, S., Collins, B., et al. (1999). Anti-glomerular basement membrane glomerulonephritis after extracorporeal shock wave lithotripsy. Am J Kidney Dis, 33, 128–32. Yamamoto, T. and Wilson, C. B. (1987). Binding of anti-basement membrane antibody to alveolar basement membrane after intratracheal gasoline instillation in rabbits. Am J Pathol, 126, 497–505. Yang, R., Cui, Z., Hellmark, T., et al. (2007). Natural anti-GBM antibodies from normal human sera recognize alpha3(IV)NC1 restrictively and recognize the same epitopes as anti-GBM antibodies from patients with anti-GBM disease. Clin Immunol, 124, 207–12. Yang, R., Cui, Z., Zhao, J., et al. (2009). The role of HLA-DRB1 alleles on susceptibility of Chinese patients with anti-GBM disease. Clin Immunol, 133, 245–50.

Yang, R., Hellmark, T., Zhao, J., et al. (2007). Antigen and epitope specificity of anti-glomerular basement membrane antibodies in patients with goodpasture disease with or without anti-neutrophil cytoplasmic antibodies. J Am Soc Nephrol, 18, 1338–43. Yang, R., Hellmark, T., Zhao, J., et al. (2009). Levels of epitope-specific autoantibodies correlate with renal damage in anti-GBM disease. Nephrol Dial Transplant, 24, 1838–44. Zhao, J., Cui, Z., Yang, R., et al. (2009b). Anti-glomerular basement membrane autoantibodies against different target antigens are associated with disease severity. Kidney Int, 76, 1108–15. Zhao, J., Yan, Y., Cui, Z., et al. (2009a). The immunoglobulin G subclass distribution of anti-GBM autoantibodies against rHalpha3(IV)NC1 is associated with disease severity. Hum Immunol, 70, 425–9. Zhao, J., Yang, R., Cui, Z., et al. (2007). Characteristics and outcome of Chinese patients with both antineutrophil cytoplasmic antibody and antiglomerular basement membrane antibodies. Nephron Clin Pract, 107, c56–62. Zhou, X. J., Lv, J. C., Bu, D. F., et al. (2010a). Copy number variation of FCGR3A rather than FCGR3B and FCGR2B is associated with susceptibility to anti-GBM disease. Int Immunol, 22, 45–51. Zhou, X. J., Lv, J. C., Yu, L., et al. (2010a). FCGR2B gene polymorphism rather than FCGR2A, FCGR3A and FCGR3B is associated with anti-GBM disease in Chinese. Nephrol Dial Transplant, 25, 97–101. Zou, J., Hannier, S., Cairns, L. S., et al. (2008). Healthy individuals have Goodpasture autoantigen-reactive T cells. J Am Soc Nephrol, 19, 396–404. Zou, J., Henderson, L., Thomas, V., et al. (2007). Presentation of the Goodpasture autoantigen requires proteolytic unlocking steps that destroy prominent T cell epitopes. J Am Soc Nephrol, 18, 771–9.

CHAPTER 75

Alport post-transplant antiglomerular basement membrane disease Zhao Cui, Neil Turner, and Ming-hui Zhao Introduction When studied by indirect immunofluorescence, the Goodpasture antigen is absent or greatly diminished in most patients with Alport syndrome (see Chapter 321). Transplant of a normal kidney may allow the development of anti-GBM antibodies to foreign antigen(s) (alloantigens) in the donor kidney. Severe disease is an unusual occurrence though. Overall the outcome of transplantation in Alport syndrome is better than average (Byrne et al., 2002; Temme et al., 2012).

Clinical features Linear IgG fixation to the GBM occurs in about 15% of Alport recipients (Quérin et al., 1986; Byrne et al., 2002) but only a minority of these (possibly 10–20%, 2–3% in total) progress to crescentic nephritis or other glomerular abnormalities characteristic of anti-GBM disease. As the immune response is to allo-antigens in the allograft, lung haemorrhage does not usually occur. The histological appearances in the kidney and clinical progression are indistinguishable from those of spontaneous Goodpasture disease (see Chapter 72). Many examples of the phenomenon have now been described, and the graft has been lost in the majority (summarized in Browne et al. 2004). The typical sequence of events is that a first allograft is biopsied months to years after the transplant with a suspicion of chronic rejection, and unexpectedly crescentic changes are seen in glomeruli. A second allograft is lost more rapidly, in weeks to months. The response to a third allograft is brisker still and disease may be apparent in days. However, there are examples of slower tempo disease in the literature, and also of patients in whom a second graft has not triggered an aggressive immune response.

Diagnosis Immunoassays for anti-GBM antibodies may be falsely negative as Alport anti-GBM antibodies are usually directed towards the molecule in which the genetic defect lies. The NC1 domain is targeted as in spontaneous anti-GBM disease, but in the most common X-linked variety of the disease the target is usually the NC1 domain of the α5 chain of type IV collagen, encoded by COL4A5 (Brainwood et al., 1998). Therefore it is important to maintain a

high index of suspicion, undertake early renal biopsy, and if there is evidence of glomerular disease to look for linear binding of IgG to the GBM. Linear binding to GBM without glomerular damage is common and most do not progress to overt disease (Quérin et al., 1986; Peten et al., 1991; Byrne et al., 2002). Reliable techniques for distinguishing anti-α3 from anti-α5 antibodies are not routine, though anti-α5 antibodies are more likely to bind strongly to Bowman’s capsule. Other techniques for distinguishing the molecular target do not affect management and are not routinely available, but Figs 75.1 and 71.1 in Chapter  71 illustrate some characteristics.

Differential diagnosis It is not uncommon for this diagnosis to be made in patients who had not previously been diagnosed with Alport syndrome—often labelled as aetiology unknown, or some type of glomerulonephritis. Important features include other signs of Alport syndrome, notably sensorineural deafness, and a family history of renal disease. The major differential diagnosis is with spontaneous anti-GBM disease. However, recurrence of anti-GBM disease while taking immunosuppressive drugs is very unusual (see Chapter 74).

Treatment and outcome Treatment is as for spontaneous anti-GBM disease (see Chapter 73), and although it has not been successful in most reported cases, success or partial success has been achieved in some. Diagnosis and treatment have often been late, and it is possible that the most severe examples are over-represented in published cases. In addition to treatment of the type used in anti-GBM disease, patients with this complication have been treated with anti-rejection therapies of almost every type, sometimes before the true diagnosis has been reached. Overall though, none of the therapies used in published reports have been very successful, even in retransplantation when the diagnosis must have been expected. Browne et al. (2004) reviewed 16 cases of retransplantation. Anti-GBM nephritis was seen in 15 out of the 16 cases, with 12 of those grafts damaged irrevocably. It is not yet possible to suggest a better treatment strategy. Foreign antigen immune responses like this are notoriously hard to suppress.

620

Section 3  

the patient with glomerular disease degenerative disorder of certain specialized basement membranes that we recognize as Alport syndrome. Patients with the (most common) X-linked form of Alport syndrome who develop this complication are more likely to have a substantial α5 chain gene deletion, perhaps because this prevents any α5 chain production, so that the antigen in the allograft is truly foreign to the recipient’s immune system (Ding et al., 1994). In a European study (Jais et al., 2000), only 3 of the 118 transplanted male patients with identified COL4A5 mutation developed post-transplant anti-GBM disease. All three had a large deletion. The risk for these patients of developing this complication was 15%, which represents a sixfold increase compared to the total Alport syndrome population. However, 16 other patients with large COL4A5 rearrangements and 32 with mutations expected to produce a truncated protein lacking the NC1 domain, did not develop anti-GBM glomerulonephritis in the graft, showing that other factors must contribute to alloimmunization. Other genetic factors may also be important, as may the immunosuppression given to prevent rejection of the allograft. Detailed studies show that the target of anti-GBM antibodies in these circumstances is most likely to be the NC1 domain of the α5 chain if the genetic abnormality is in the gene (COL4A5) encoding it (Brainwood et  al., 1998), that is, different from the classic Goodpasture antigen. In patients with autosomal disease the target may be α3(IV)NC1, the Goodpasture antigen. It seems clear that the disease is, like spontaneous anti-GBM disease (see Chapter 73), mediated by both antibody and cellular (T lymphocyte, macrophage) components. Possibly the disease may be triggered by increased presentation of the inciting antigen in an inflammatory context, but at present the reasons for the unpredictability of development of this complication are unknown.

References Fig. 75.1  Moderately severe anti-a5(IV) collagen-mediated Alport post-transplant anti-GBM disease in the third renal transplant of a patient with a large COL4A5 deletion. Biopsy at day 44 showed cellular crescents (A) with focal necrotizing lesions in three out of five glomeruli. By direct immunofluorescence, linear staining for IgG and C3 was seen on the GBM, and strongly on Bowman’s capsule and in distal tubules (B). From Browne et al. (2004).

Diagnosis in a second transplant is likely to be more prompt as suspicions will be high. Our current recommendation is that patients who lost a second allograft from this complication should not usually be put forward for further attempts at transplantation unless there are strong reasons to believe that a new approach is likely to be more successful.

Pathogenesis The molecular defect in Alport syndrome involves a gene encoding one of the tissue-specific type IV collagen chains. The disease is usually X-linked, involving COL4A5, but mutations in one chain may destabilize the 345 supramolecular network involving α3, α4, and α5 chains (further described in Chapter 320), producing the

Brainwood, D., Kashtan, C., Gubler, M. C., et al. (1998). Target of alloantibodies in Alport anti-glomerular basement membrane disease after renal transplantation. Kidney Int, 53, 762–6. Browne, G., Brown, P. A., Tomson, C. R., et al. (2004). Retransplantation in Alport post-transplant anti-GBM disease. Kidney Int, 65, 675–81. Bygren, P., Freiburghaus, C., Lindholm, T., et al. (1985). Goodpasture’s syndrome treated with staphylococcal protein A immunoadsorption. Lancet, 2, 1295–6. Byrne, M. C., Budisavljevic, M. N., Fan, Z., et al. (2002). Renal transplant in patients with Alport’s syndrome. Am J Kidney Dis, 39, 769–75. Ding, J., Zhou, J., Tryggvasson, K., et al. (1994). COL4A5 deletions in three patients with Alport syndrome and posttransplant antiglomerular basement membrane nephritis. J Am Soc Nephrol, 5, 161–8. Jais, J. P., Knebelmann, B., Giatras, I., et al. (2000). X-linked Alport syndrome: natural history in 195 families and genotype- phenotype correlations in males. J Am Soc Nephrol, 11, 649–57. Peten, E., Pirson, Y., Cosyns, J. P., et al. (1991). Outcome of thirty patients with Alport’s syndrome after renal transplantation. Transplantation, 52, 823–6. Quérin, S., Noël, L. H., Grünfeld, J. P., et al. (1986). Linear glomerular IgG fixation in renal allografts: incidence and significance in Alport’s syndrome. Clin Nephrol, 25, 134–40. Temme, J., Gross, O., Jager, K. J., et al. (2012). Outcomes of male patients with Alport syndrome undergoing renal replacement therapy. Clin J Am Soc Nephrol, 7, 1969–76.

CHAPTER 76

Post-infectious glomerulonephritis: overview Bernardo Rodriguez-Iturbe and Mark Haas Post-infectious glomerulonephritis (GN) defines an inflammatory lesion involving exclusively or predominantly the glomeruli that is a consequence of an infectious disease. There are numerous bacterial, viral, and fungal infections associated with GN (Table 76.1). In the following chapters, we will discuss only post-streptococcal GN (Chapter  77), immunoglobulin A  (IgA)-dominant GN usually associated with staphylococcal infections (Chapter  78), and GN associated with bacterial endocarditis, with infected ventriculo-atrial shunts (‘shunt nephritis’), and with deep-seated infections (osteomyelitis, visceral abscesses, pleural suppuration, pneumonia) (Chapter 79). Almost all of these lesions result from antigen–antibody reactivity that causes local activation of the complement system and of the coagulation cascade. Specific types of post-infectious GN, such as those caused by staphylococcal infections associated with IgA deposition may involve superantigens causing intense T-cell activation that engage cytokine-mediated polyclonal B-cell responses. In disadvantaged populations with poor hygienic conditions and limited access to early medical care, post-infectious GNs, particularly post-streptococcal GN, are frequent. In more affluent societies, post-infectious glomerulonephritis is less common, and tends to occur in older patients or those with significant comorbidity already receiving medical care and/or with implanted catheters or devices. Recent studies indicate that bacteria causing GN are more frequently Staphylococcus (46%), Streptococcus (16%), and Gram-negative organisms (Nasr et al., 2011) and the most common sites of infection are the upper respiratory tract (23%), skin (17%), lung (17%), and heart valves (11.6%). Chronic GN develops in about 25% of these patients (Nasr et al., 2008). The clinical presentation and the pathological characteristics of post-infectious GN are not uniform. Even in a specific disease these characteristics may have considerable variability. For instance, post-streptococcal GN may be clinically asymptomatic, may present with acute nephritic or nephrotic syndrome, or have a rapidly progressive course, and the typical endocapillary GN may be associated with crescent formation followed by sclerosis. This lack of uniformity results from the variability of size, load, and charge of the antigen, the intensity of the antibody response, the site of the immune reactivity, and the efficiency of immune-complex clearance. In post-infectious GN, the immune complexes may be predominantly located in the mesangium, in and on the glomerular basement membrane (GBM), or both. Prominent deposition of immune complexes in and on the GBM is usually associated with heavy proteinuria while mesangial deposits alone are usually

Table 76.1  Infectious agents associated with endocapillary glomerulonephritis Associated with infectious syndromes Skin and throat (Streptococcus group A) IgA-dominant glomerulonephritis (methicillin-resistant Staphylococcus) Bacterial endocarditis (Staphylococcus aureus, Streptococcus viridans) Pneumonia (Diplococcus pneumoniae, Mycoplasma) Meningitis (Meningococcus pneumoniae, Mycoplasma) Visceral abscesses and osteomyelitis (Staphylococcus aureus, Escherichia coli, Pseudomonas aeruginosa, Proteus mirabilis, Klebsiella, Clostridium perfringens) ‘Shunt’ nephritis (Staphylococcus aureus, Staph. albus, Streptococcus viridans) Infected vascular prosthesis (Staphylococcus aureus) Guillain–Barré syndrome (G-B infectious agent?). Associated with specific bacterial diseases Typhoid fever (Salmonella typhi) Leprosy Yersiniosis Brucellosis Leptospirosis. Associated with viral infections Hepatitis A, B, and C Epstein–Barr virus Parvovirus B19 Cytomegalovirus Measles Mumps Varicella Coxsackie virus. Associated with parasitic infestation Malaria (Plasmodium falciparum, P. malariae) Schistosomiasis (Schistosoma haematobium, S. mansoni) Toxoplasmosis Trichinosis (Trichinella spiralis) Filiarasis (Onchocerca volvulus, Loa Loa). Associated with other infectious organisms Rickettsiae (Coxiella) Fungi (Candida albicans, Coccidioides immitis, Histoplasma capsulatum).

622

Section 3  

the patient with glomerular disease

associated with only mild-to-moderate proteinuria and microscopic haematuria. As demonstrated in serum sickness experimental models, circulating immune complexes are usually deposited in subendothelial and mesangial regions and, if the condition is self-limited, induce a transient glomerulonephritis. In situ immune complex formation, a consequence of the penetration of free antigen through the GBM, tends to form subepithelial immune complexes. This mechanism is favoured when there is antigen excess that facilitates dissociation of circulating immune complexes and

when the antigen is cationic and therefore attracted by the negatively charged GBM.

References Nasr, S. H., Fidler, M. E., Valeri, A. M., et al. (2011). Postinfectious glomerulonephritis in the elderly. J Am Soc Nephrol, 22, 187–95. Nasr, S. H., Markowitz, G. S., Stokes, M. B., et al. (2008). Acute postinfectious glomerulonephritis in the modern era: experience with 86 adults and review of the literatura. Medicine, 87, 21–32.

CHAPTER 77

Post-streptococcal glomerulonephritis Bernardo Rodríguez-Iturbe and Mark Haas Introduction In autopsy studies during the nineteenth century, the renal lesion most frequently found in patients who developed oliguria during scarlet fever was, surprisingly, interstitial nephritis (Councilman, 1898) but the observation that ‘dark and scanty urine’ was a serious complication of the convalescent dates back to descriptions of epidemics in the fourteenth century (Becker and Murphy, 1968). The disease was very common in central Europe in the pre-antibiotic era and acute post-streptococcal glomerulonephritis (PSGN) (Fig. 77.1) was probably the cause of the untimely death of Wolfgang Amadeus Mozart in 1791 (Zegers et al., 2009). In 1910, Clemens von Pirquet, in a landmark paper that opened the field of immune diseases (von Pirquet 1910), postulated that the post-scarlatinal nephritis was caused by the development of harmful antibodies (as opposed to the beneficial antibodies in vaccination) and coined the term ‘allergy’ (altered reactivity) to define this pathogenic modality. The streptococcal aetiology of scarlet fever (Dochez and Sherman, 1924) prompted the recognition of PSGN as the first and most extensively studied glomerulonephritis (GN) associated with bacterial infections. Acute rheumatic fever and GN are both non-infectious complications of streptococcal infection, but have epidemiological and biological differences and only rarely if ever occur in the same

Fig. 77.1  Acute post-streptococcal glomerulonephritis: proliferative and exudative glomerulonephritis. The glomerulus shows marked global hypercellularity with many neutrophils, and is representative of all of the glomeruli on this biopsy. (Haematoxylin and eosin stain, original magnification 400×.)

patient. Therefore, Seegal and Earle (1941) postulated the existence of distinct rheumatogenic and nephritogenic strains of the bacterium. Since recurrence of PSGN is a very rare event it is likely that putative antigen(s) shared among nephritogenic strains confer long-lasting immunity.

Epidemiology The incidence of PSGN has decreased in recent decades in industrialized countries (Ahn and Ingulli, 2008). At present it is usually associated with diabetes, alcoholism, intravenous drug addiction, and debilitating conditions. The reduction of the incidence in the United States, United Kingdom, and Western continental Europe is probably the result of easier and earlier access to appropriate medical care for streptococcal infections and perhaps widespread fluorination of the water since virulence factors in Streptococcus pyogenes are reduced with fluoride exposure (Thongboonkerd et al., 2002). The rarity of PSGN in affluent societies has been cited as a factor for delayed diagnosis in patients who do not have gross haematuria (Pais et al., 2008). The incidence of PSGN is also decreasing in developing countries (Rodríguez-Iturbe and Mezzano, 2005) but it is not uncommon. Two independent studies (Carapetis et al., 2005, Rodriguez-Iturbe and Musser, 2008) have estimated that the incidence of PSGN in developing countries is at least 9.3–9.8 cases per 100,000 population per year and may be in fact as much as three times these values (Carapetis et al., 2005). PSGN is particularly frequent in populations with deficient hygienic conditions and substandard medical services (Sarkissian et al., 1997; Orta and Moriyón, 2001) and particularly in aboriginal communities (Currie and Brewster, 2001; Marshall et al., 2011). The importance of educational programmes in the control of the disease is underlined by the finding that programmes designed to reduce the incidence of rheumatic fever have led also to the decline in the incidence of GN in French Caribbean islands (Bach et al., 1966). Pharyngitis and tonsillitis are the usual sites of antecedent infection in the winter and the spring in temperate climates and impetigo is more frequent in the tropics in the summer months. PSGN may occur in sporadic cases, in clusters of cases and in epidemic outbreaks. Clusters of cases are more frequently reported in poor communities in industrialized countries with adequate health systems that allow detection and documentation of the disease, while epidemics of > 100 cases are usually reported from countries in the middle range of Human Development Index and a mean annual

624

Section 3  

the patient with glomerular disease

health expenditure per capita of about US$550 (Rodríguez-Iturbe and Musser, 2008). Epidemics have been reported in periodic outbreaks in specific regions in Minnesota (Anthony et al., 1969), Trinidad (Poon-King et  al., 1967), Venezuela (Rodríguez-Iturbe, 1984), and in the Northern Territory of Australia (Marshall et al., 2011). The risk of nephritis in epidemics of streptococcal infections varies significantly, ranging from 5% in throat infections, to as high as 25% in pyoderma caused by M type 49 streptococci. Clinical observations suggest a genetic predisposition to PSGN and prospective studies have found that that 38% of the siblings of index cases have evidence of symptomatic or subclinical nephritis, an incidence that is higher than the attack rate in the general population in epidemic conditions that ranges between 5% and 28% (Rodríguez-Iturbe, 1984). Associations of PSGN with human leucocyte antigen (HLA)-DR4 and DR-1 have been reported but definite genetic associations have not been detected.

Clinical features The infection preceding nephritis The usual sites of infection with a nephritogenic Streptococcus are the skin or the throat, but other foci are possible. Unusual sites of infection include spider bites (Lung and Mallory, 2000)  and infected circumcision wounds (Tasic and Polenakovic, 2000). Streptococcal pharyngitis may cause only sore throat or be accompanied by tonsillar exudate, fever, and cervical lymphadenopathy. Scarlatinal rash is due to an erythrogenic toxin that is produced by the bacteria. Vomiting may be a prominent initial symptom in scarlet fever. Streptococcal impetigo is characterized by clusters of small vesicles that appear in exposed skin areas. They break rapidly and leave lesions covered with a thick yellow crust. Regional lymphadenopathy is usually present in patients with active skin infection. Impetigo is frequently associated with scabies and a history of intense itching, particularly if it is also present in other family members, is a diagnostic clue. Outbreaks of ulcerated ecthyma skin lesions in soldiers have also been associated with acute PSGN (Wasserzug et al., 2009). The latent period after infection is shorter after throat infections (about 2 weeks) than after pyodermitis (several weeks). When throat and skin infections are present at the same time, the throat infection is usually due to a contamination from the skin (Anthony et al., 1969).

Clinical features of acute nephritis The clinical features of PSGN are different in adults and in children (Table 77.1). In children the typical presentation of acute PSGN is the acute nephritic syndrome; however, the disease may be subclinical, or it may be manifested by nephrotic syndrome or rapidly progressive GN. Asymptomatic disease is recognized by microscopic haematuria and reduced complement levels with or without hypertension. Patients with subclinical PSGN are 1.5 times more frequent than patients with clinical disease in epidemics. In non-epidemic situations, prospective family studies of index cases indicate that the ratio of subclinical/clinical disease is 4.0 (Rodríguez-Iturbe et al., 1981b) to 5.3 (Dodge et al., 1967). The acute nephritic syndrome is the classic clinical presentation of acute PSGN. The typical patient is a boy (male:female ratio 2:1) 4–14 years of age who

Table 77.1  Clinical manifestations of acute post-streptococcal glomerulonephritis in children and elderly adultsa Children (%)

Elderly patients (%)

Haematuria

100

100

Proteinuria

80

92

Oedema

90

75

Hypertensionb

60–80

80–86

Oliguria

10–50

58

Dyspnoea/heart failure

80% of patients, but only 50% require drug treatment. Haemodynamic measurements indicate that increased plasma volume, increased cardiac output, and elevated peripheral vascular resistance contribute to the hypertension in the acute nephritic syndrome. Plasma renin and aldosterone levels are suppressed, consistent with the volume-dependent nature of arterial hypertension in the acute nephritic syndrome. Oliguria is noted as a symptom on admission by less than half of the children or their parents/guardians. Non-specific symptoms, such as dull

Chapter 77 

lumbar pain, malaise, weakness, and nausea frequently accompany the cardinal manifestations of the acute nephritic syndrome. In a typical case of post-streptococcal nephritis, improvement is observed after 2–7 days when the urine volume increases, followed rapidly by resolution of oedema and return of the blood pressure to normal levels.

Diagnosis of post-streptococcal glomerulonephritis The differential diagnosis of a patient presenting with acute nephritic syndrome includes distinguishing a primary renal disease from the initial manifestation of an otherwise silent systemic disease. Systemic lupus erythematosus, essential cryoglobulinaemia, subacute and acute bacterial endocarditis, ‘shunt’ nephritis, visceral abscess, antineutrophil cytoplasmic antibody (ANCA)-associated vasculitis and GN, and anti-glomerular basement membrane (anti-GBM) antibody disease may all potentially present as an isolated nephritis. Measurement of serum complement activity has been suggested as a first-line test in evaluation of acute GN because low serum complement is a feature of 90% or more of the cases of acute PSGN and the finding of a normal serum complement would make this diagnosis unlikely. Low levels of C3 and normal or mildly depressed values of C4 are typical of diseases in which there is preferential activation of the alternative pathway, such as dense deposit disease and PSGN. Serum complement returns to normal usually within 1  month in patients with uncomplicated PSGN. Serum immunoglobulin G (IgG) and IgM are elevated in 80% of the cases, and in contrast with another post-streptococcal disease, rheumatic fever, IgA serum level is normal. Cryoglobulins, elevated rheumatoid factor, and anti-C1q antibodies are present in up to one-third of patients, and rarely patients may have low titters of anti-DNA and ANCA (Rodríguez-Iturbe, 1984). While the diagnosis of post-streptococcal aetiology may be suspected on clinical grounds its confirmation requires positive bacterial culture or serological evidence of recent streptococcal infection. Positive cultures maybe obtained in as many as 70% of the cases in epidemics but at best in 25% of sporadic cases. Increasing or high titres of serum anti-streptolysin O are found in 60–80% of the patients with streptococcal throat infection and high anti-DNAse B titres in about 70% of the cases with pyodermitis. The streptozyme test, which includes four antigens (DNAse B, Streptolysin O, hyaluronidase and streptokinase) may be found elevated in nearly 80% of the cases. Antibodies against a nephritis-associated plasmin receptor (Yamakami et al., 2000) and antibodies against the zymogen precursor of streptococcal pyrogenic exotoxin B (Parra et al., 1998), both of which are assumed to be nephritogenic antigens (see below), are more sensitive and specific for nephritogenic strains than the rest of anti-streptococcal antibodies but they are not clinically available. In children with uncomplicated PSGN, renal biopsy is only indicated when there are unusual features that make the diagnosis doubtful. Among these features are proteinuria in the nephrotic range, progressive azotaemia suggesting crescentic GN, and serum complement levels that are not depressed in the acute phase or remain reduced for > 1 month. In adult patients, kidney biopsy is the norm because in adults PSGN carries a much poorer immediate and long-term prognosis.

post-streptococcal glomerulonephritis

Aetiology and pathogenesis Nephritogenic antigens Nephritogenic strains of group A streptococcus pyogenes include impetigo-associated M types 47, 49 and throat infection-associated M types 1, 2, 4 and 12. One recent large epidemic (Balter et  al., 2000) and several clusters of cases (Francis et al 1993; Nicholson el al., 2000)  have been related to the ingestion of unpasteurized milk contaminated with group C streptococcus (Streptococcus zooepidemicus). Throughout the years, many putative streptococcal nephritogenic antigens have been suggested without confirmation of a causal relationship with GN (Rodríguez-Iturbe and Batsford, 2007). At the present time two antigens are associated with PSGN:  the so-called nephritis-associated plasmin receptor (NAPLr) identified as glyceraldehyde-3-phosphate dehydrogenase (GAPDH) (Yoshizawa et al., 2004) and the cationic (pK 8.0) streptococcal proteinase exotoxin B (SPEB) and its more immunogenic precursor zymogen (Vogt et al., 1983; Poon-King et al., 1993). Both antigens injected intravenously have affinity for the glomeruli. Antibody titres to these two antigens are present in the serum at the time of convalescence. GAPDH has been localized in areas of the glomeruli with plasmin-like activity, suggesting a direct mechanism for glomerular inflammatory damage but is not co-localized with complement or immunoglobulin (Oda et al., 2005, 2010). Patients with acute PSGN have increased urinary plasmin activity resistant to alpha(2)-AP, likely due to urinary excretion of NAPlr (Oda et al., 2008). SPEB is co-localized with both complement and IgG in the glomeruli suggesting a participation in immune-mediated renal damage and is the only putative antigen demonstrated in the subepithelial electron dense deposits (‘humps’) that are the typical ultrastructural lesion of PSGN (Batsford et al., 2005). Vogt et al. (1990) have emphasized that cationic antigens, such as SPEB and its zymogen precursor are attracted by the negatively charged GBM, and penetrate towards subepithelial locations where they react in situ with the antibody. SPEB/zymogen induce leucocyte infiltration in the glomerulus, possibly as a result of the induction of chemotactic and migration inhibition factor activities (Romero et  al., 1999) and increased angiotensin II production in mesangial cells (Viera et al., 2009). In situ immune complex formation would easily explain the formation of subepithelial electron dense deposits (humps) which are extremely difficult to produce by the injection of preformed immune complexes and, in contrast, are the rule with the injection of cationic antigens. Heavy, confluent immunoglobulin and complement deposition in and on the GBM (garland-type deposits) are associated with the existence of a large number of humps in biopsies, heavy proteinuria and worse prognosis (see later), therefore the in situ mechanism may have significant clinical relevance. Nevertheless, SPEB cannot be the single nephritogenic antigen of Streptococcus, since the SPEB gene was not present in the Streptococcus zooepidemicus responsible for a recent major epidemic (Beres et al., 2008). Other cationic components that could have pathogenetic relevance are the streptococcal histones that may readily accumulate in the glomeruli (Choi et al., 1995). After streptococcal lysis, histones can enter the blood and bind to the anionic proteoglycans to trigger an in situ immune complex formation or directly induce the production of proinflammatory cytokines (Zhang et al., 1999).

625

626

Section 3  

the patient with glomerular disease

Complement activation in PSGN In PSGN, serum C3 levels are more frequently and more intensively depressed than C4. Activation of the alternate complement pathway may be mediated by transient expression of C3 NeF autoantibody (Frémeaux-Bacchi et al., 1994). Berge et al. (1997) have shown that protein H, a surface protein of Streptococcus pyogenes, in association with IgG is capable of activating the classic complement pathway. These findings may explain why 15–30% of the patients with PSGN show evidence of reduction of the C1 and C4 complement components. The lectin pathway of complement may also be activated in PSGN, likely as a result of the recognition of the starter molecule mannose-binding lectin by the N-acetyl glucosamine residues of the bacterial wall (Oshawa et al., 1999) but individuals genetically unable to activate this pathway may still develop PSGN (Skattum et al., 2006). The complement system recruits inflammatory cells but, in addition, individual components have a direct nephritogenicity. C3a and C5a cause histamine release and increased capillary permeability, and the terminal C5b-C9 membrane attack complex has a direct effect on the glomerular capillary membrane. Plasma levels of C5b-C9 complement complexes are consistently elevated at the onset of acute PSGN (Matsell et  al., 1994). Non-lytic effects of the membrane attack complex may result from platelet activation; reduction of platelet counts and survival time and increased platelet-activating factor are present in acute PSGN (Mezzano et al., 1993).

Cell-mediated immunity Early studies demonstrated the infiltration of CD4 T lymphocytes (Parra et  al., 1984)  and the overexpression of intercellular adhesion molecules in renal biopsies of patients with acute PSGN (Parra et al., 1994). Subsequent studies showed associations between interleukin (IL)-6, IL-8, transforming growth factor beta (TGFβ), and tumour necrosis factor alpha (TNFα) with nephritogenic antigens and proliferative changes in PSGN (Mezzano et al., 1997; Pedreanez et al., 2006).

Autoimmunity Autoimmune reactivity is present in many cases of acute PSGN. Elevated titres of IgG rheumatoid factor were found in over one-third of the patients in the first week of the disease and anti-IgG deposition has been demonstrated in the kidney (Rodríguez-Iturbe, 1984). Anti-IgG reactivity may result from the loss of sialic acid from autologous IgG due to streptococcal neuraminidase (sialidase). Neuraminidase activity and free sialic acid have been found in the plasma of patients with acute PSGN (Rodríguez-Iturbe et al., 1981a; Asami et al., 1985). Neuraminidase may have the additional effect of facilitating the infiltration of desialized leucocytes in the glomeruli (Marín et  al., 1997). Duvic et  al. (2000) reported the simultaneous presentation of acute PSGN and thrombotic microangiopathy and suggested a role for neuraminidase in the combined clinical picture. Another possible mechanism for the production of anti-Ig is the binding of the Fc fragment of IgG to type II receptors on the surface of group A  streptococcus. This binding induces intense anti-IgG reactivity and GN with anti-IgG deposits. Burova et  al. (1998, 2012) have postulated that PSGN might be triggered by this mechanism. Additional autoimmune phenomena have been found

in acute PSGN patients. Anti-DNA antibodies, anti-C1q antibodies (Kozyro et al., 2006), and ANCAs are present in some patients. Interestingly, the latter have been found in two-thirds of the patients with azotaemia and 70% of the patients with crescentic acute PSGN and a rapidly progressive course (Ardiles et al., 1997). Despite these findings the clinical relevance of autoimmune reactivity in PSGN remains undefined.

Pathology Acute post-streptococcal glomerulonephritis Light microscopic findings The majority of cases, including 72% in a recently published series of adult patients (Nasr et al., 2008), show diffuse proliferative and exudative GN (Fig. 77.1), with the most of the remainder showing focal proliferative and exudative GN or predominantly mesangial proliferative GN (Fig. 77.2) (Rosenberg et al., 1985; Montseny et al., 1995; D’Agati et al., 2005; Nasr et al., 2008). A membranoproliferative GN (MPGN) pattern is rarely seen (Montseny et al., 1995; Nasr et al., 2008). With the Masson’s trichrome stain, fuschinophilic, red-orange subepithelial and mesangial deposits may be evident (Fig. 77.3). Crescents, primarily segmental cellular crescents, are present in up to half of cases, and may be accompanied by segmental fibrinoid necrosis with disruption of the GBM evident on the silver methenamine stain (Montseny et al., 1995; Nasr et al., 2008). Interstitial inflammation, typically comprised of a mixture of lymphocytes, monocytes, plasma cells, and neutrophils, is present in most cases. Focal intratubular neutrophils are not infrequent, with these cells coming from the inflamed glomeruli. In the recent study of adult patients by Nasr et al. (2008), histologic evidence of acute tubular injury, characterized by flattening of proximal tubular epithelium with loss of brush borders and nuclear enlargement, was seen in 60% of cases. Mild to moderate arteriosclerosis was also seen in the majority of these cases; cases with underlying diabetic nephropathy tended to have more frequent and more severe arteriosclerosis, as well as arteriolar hyalinization and thickening (Nasr et al., 2008).

Fig. 77.2  Post-streptococcal glomerulonephritis: mesangial proliferative glomerulonephritis. This pattern is most often seen in lesions undergoing resolution. (Haematoxylin and eosin stain, original magnification 400×.)

Chapter 77 

Fig. 77.3  Acute post-streptococcal glomerulonephritis: Masson’s trichrome stain. Distinct immune complex deposits are noted that stain a red-orange colour; note especially the subepithelial, ‘hump-like’ deposit near the centre of the field. (Masson’s trichrome stain, original magnification 1000×, scale bar 50 microns.)

post-streptococcal glomerulonephritis

Fig. 77.5  Resolving post-streptococcal glomerulonephritis: immunofluorescence microscopy for C3. Light microscopy from this same biopsy shown in Fig. 77.2. Deposits of C3 appear to be limited to mesangial areas, although electron microscopy also showed subepithelial deposits localized to mesangial ‘waist’ regions; such deposits cannot be distinguished from mesangial deposits by immunofluorescence and contribute to the mesangial pattern. (FITC conjugated anti-human C3, original magnification 400×.)

Immunofluorescence microscopy Immunofluorescence findings in evolving stages of PSGN were elegantly defined by Sorger et  al. (1982). These investigators defined three patterns of glomerular staining, and correlated these with time after disease onset, relative intensity of IgG and C3 staining, and location and frequency of immune complex deposits by electron microscopy (see below). In the early phase of the disease (initial two or three weeks), the glomeruli show finely granular deposits of C3 and usually IgG in the capillary walls and mesangial areas, in what has been termed a ‘starry sky’ pattern (Fig. 77.4). Later in the disease, with resorption of

Fig. 77.4  Acute post-streptococcal glomerulonephritis: immunofluorescence microscopy for C3. Numerous granular deposits typical of immune complexes are present in the glomerular capillary walls and mesangial areas, giving a ‘starry sky’ pattern. (Fluorescein isothiocynate (FITC) conjugated anti-human C3, original magnification 400×.)

many of the subepithelial and subendothelial deposits, there is a predominantly mesangial pattern of staining (Fig. 77.5) with a predominance of C3. In truth, not all of the deposits contributing to the mesangial pattern are actually within the mesangium; as discussed below many are subepithelial deposits within the mesangial ‘waist’ that are resorbed more slowly than deposits in peripheral portions of the glomerular capillaries. A third pattern of staining, characterized by coarse granular to confluent staining along the glomerular capillary walls (Fig. 77.6), termed the ‘garland’ pattern, is most often seen early in the disease may but may be seen later as well. It is this pattern that best shows the individual subepithelial ‘humps’ that are the characteristic ultrastructural feature of this disease.

Fig. 77.6  Acute post-streptococcal glomerulonephritis: immunofluorescence microscopy for IgG. There is global, coarse to confluent granular staining mainly along the glomerular capillary walls, corresponding to subepithelial deposits (including large subepithelial ‘humps’) and likely small subendothelial deposits as well. This ‘garland’ pattern is most often seen early in the disease, particularly when there is IgG as well as C3 deposition, although C3 deposits with this pattern may also be seen later in the disease course. (FITC conjugated anti-human IgG, original magnification 400×.)

627

628

Section 3  

the patient with glomerular disease

In acute (and subacute) PSGN, deposits of C3 are invariably present, accompanied by IgG in most cases, and IgM in approximately 50%, although the latter staining tends to be of low intensity (D’Agati et al., 2005; Nasr et al., 2008). IgA staining is uncommon and of low intensity when present, although IgA is often the dominant immunoglobulin present in post-staphylococcal lesions, as discussed below. Nasr et al. (2008) found C1q staining, typically of low intensity, in approximately one-third of their cases. Staining for kappa and lambda light chains mirrors that for IgG with respect to pattern with similar staining intensity for both light chains. Focal and segmental, blotchy to amorphous staining for fibrinogen, most typically at the periphery of glomerular tufts, is frequently noted within cellular crescents when these are present.

Resolving post-streptococcal glomerulonephritis

The characteristic ultrastructural finding of acute PSGN is the presence of large subepithelial electron-dense deposits with a ‘hump-like’ appearance (Fig. 77.7). The number of these deposits varies considerably between different cases; they can be quite segmental or rather numerous, although not so much so as to suggest a membranous nephropathy. The size of the subepithelial ‘humps’ may also vary considerably within any given glomerulus. In early post-infectious lesions, these deposits are distributed at various points along glomerular capillaries, although even at this stage there is some tendency for the greatest number of deposits to be concentrated at or near the glomerular basement membrane reflection over mesangial areas (the mesangial ‘waist’ or ‘notch’; Fig. 77.7), a finding first noted by Heptinstall (1974). In subacute and resolving cases, the fraction of subepithelial deposits localized to mesangial ‘waist’ areas increases as more peripheral deposits are resorbed (Sorger et al., 1982; Haas, 2003; Nasr et al., 2008). Mesangial deposits are present in the great majority of cases of acute PSGN and may be abundant, and most cases also show subendothelial deposits, although these tend to be small and segmental (D’Agati et  al., 2005; Nasr et al., 2008) (Fig. 77.7). Extraglomerular deposits are not a feature of this disease.

The resolution of PSGN proceeds through various stages that are best identified by correlating light, immunofluorescence, and electron microscopic (EM) findings. After the first 1–2 weeks of the disease, in which the glomeruli appear enlarged and markedly hypercellular with prominent numbers of neutrophils (Fig. 77.1), there is a progressive decline in cellularity, initially with the loss of the neutrophils resulting in a combined mesangial and endocapillary proliferative GN (Fig. 77.8). At this stage, the histologic appearance of the glomeruli may resemble that of an early MPGN (without the numerous double contours of the GBM) or even a highly active IgA nephropathy, although subepithelial ‘humps’ are still quite evident by EM, often in early stages of resorption (Fig. 77.9). Over the ensuing weeks, endocapillary hypercellularity is lost, together with ‘humps’ in peripheral portions of the GBM. The result is a predominantly mesangial proliferative GN by light microscopy (Fig. 77.2), with an accompanying mesangial pattern of C3 deposits by immunofluorescence (Fig. 77.5) and subepithelial deposits limited largely to mesangial ‘waist’ regions by EM (Fig. 77.10), the latter often showing evidence of partial resorption. Notably, weak, granular staining for C3 (most typically in a mesangial pattern) and partially or largely resorbed subepithelial deposits (often partially or even completely incorporated into the GBM) remain evident months to years after the resolution of haematuria, renal insufficiency, and light microscopic findings of proliferative GN (Tornroth, 1976; Sorger et al., 1982; Rosenberg et al., 1985; Haas, 2003). These largely resorbed deposits are variably electron-lucent, containing granular, vesicular, or membrane-like structures (Tornroth, 1976; Haas, 2003) (Fig. 77.10). These findings are not infrequently associated with sub-nephrotic proteinuria, and may occasionally represent the only pathologic findings on a renal biopsy done for such proteinuria (Baldwin et al., 1974; Haas, 2003). It has been our experience that incidental, ultrastructural findings of old, largely healed post-infectious GN, namely partially or largely resorbed subepithelial deposits with at least some localized to mesangial ‘waist’ areas, are evident in approximately 10%

Fig. 77.7  Acute post-streptococcal glomerulonephritis: electron microscopy. There are three large subepithelial ‘humps’; the one at left is in a mesangial ‘waist’ region. There are also some small mesangial deposits underlying this latter deposit, and a small subendothelial deposit (arrow). (Uranyl acetate and lead citrate stain, original magnification 6300×.)

Fig. 77.8  Subacute post-streptococcal glomerulonephritis. The glomerulus, representative all of glomeruli on this biopsy, still shows mesangial and endocapillary hypercellularity. However, the degree of cellularity less than that in Fig. 77.1, and neutrophils are no longer prominently seen. (Haematoxylin and eosin stain, original magnification 400×.)

Electron microscopy

Chapter 77 

post-streptococcal glomerulonephritis

Subepithelial humps, while characteristic of post-infectious GN, are by no means specific for this entity and may be present in other forms of GN, most notably C3 glomerulonephritis (C3GN) and dense deposit disease, both of the latter characterized by abnormal regulation of the alternative pathway of complement (Pickering et  al., 2013). Interestingly, Sethi et  al. (2012) likewise found evidence of abnormalities in the alternative pathway in 10 of 11 cases of atypical (persistent) post-infectious GN, including mutations in complement factor H (CFH) and CFH-related proteins, and C3 Nef autoantibody.

Treatment Antibiotic treatment Fig. 77.9  Subacute post-streptococcal glomerulonephritis: electron microscopy. The subepithelial deposits, including the large ‘hump’, have a somewhat variegated appearance consistent with early resorption, and there is relative electron-lucency at the periphery of the large deposit. (Uranyl acetate and lead citrate stain, original magnification 8000×.)

of native renal biopsy specimens (Haas, 2003). In the vast majority of these cases there is no documented clinical evidence of an acute episode of post-infectious GN. Findings of incidental, healed post-infectious GN are particularly common in biopsies from patients with diabetic nephropathy, being evident in 30–40% of such biopsies (Haas, 2003), although whether these lesions have any clinical significance remains unknown.

The diagnosis of impetigo is usually straightforward. In contrast, the clinical judgement may incorrectly diagnose a sore throat as being streptococcal in 20–40% of the cases (Cebul and Poses, 1986). Clinical scores have been proposed to increase the accuracy of this diagnosis (McIsaac et al., 1998). Newly developed rapid high sensitivity tests require culture confirmation if the results are negative (American Academy of Pediatrics, 2000) but the decision to withhold treatment based on this rapid diagnostic test of streptococcal sore throat does not carry increased risk of post-streptococcal complications (Webb et al., 2000). We recommend antibiotic treatment in all patients with acute PSGN at the time of diagnosis whether the streptococcal infection is obvious or not, to ensure the elimination of the responsible antigen. Penicillin or cephalosporins (Casey and Pichichero, 2004) are adequate treatment. Erythromycin is the treatment of choice in patients allergic to penicillin. Prophylactic antibiotic treatment should be prescribed in household members of index cases, particularly in high-risk communities since most of them have evidence of recent streptococcal infection and about one-third develop GN (Rodríguez-Iturbe et al., 1981b). This strategy has resulted in a reduction of cases in aboriginal communities (Johnston et al., 1999). As discussed earlier, educational programmes have resulted in a reduction of the incidence of PSGN (Bach et al., 1966).

Treatment of acute PSGN

Fig. 77.10  Resolving post-streptococcal glomerulonephritis: electron microscopy. There are two large, partially resorbed subepithelial deposits in mesangial ‘waist’ regions (arrows). The lower deposit is more electron-dense at its centre and less so at its periphery; the upper deposit contains a membrane-like structure. (Uranyl acetate and lead citrate stain, original magnification 7200×.)

Children with subclinical disease may be followed in the clinic but patients with the acute nephritic syndrome usually require hospitalization. Strict bed rest is of doubtful value. Restriction of water and particularly sodium intake should be prescribed to all patients with the acute nephritic syndrome. Early studies (Powell et  al., 1980)  showed that loop diuretics increase urine output severalfold in most patients and entail a 50% reduction in the time required for normalization of blood pressure and disappearance of oedema, including pulmonary oedema. Antihypertensive medication may be needed in approximately half the children with PSGN and nearly 80% of adults. A  rare patient with convulsions will require sedation and intubation. Pulmonary oedema may complicate the clinical course. Dialysis may be required if there is severe azotaemia or hyperkalaemia. Posterior reversible leucoencephalopathy has been reported in acute PSGN (Ahn and Ingulli, 2008), probably as a manifestation of severe hypertension, and is manifested by drowsiness progressing to stupor, visual hallucinations, headache, and convulsions. Firm diagnosis requires magnetic resonance studies. Immune-mediated

629

630

Section 3  

the patient with glomerular disease

pneumonitis has been also reported in an adult patient with PSGN (Wiles et al., 2011). Patients with crescentic PSGN are sometimes given intravenous pulses of methylprednisolone, but it has not been demonstrated that intravenous steroids, immunosuppression, or anticoagulation improve the outcome of crescentic PSGN (Zaffanello et al., 2010).

Prognosis The early mortality of acute endocapillary GN is very low in children but significant in adults (Melby et  al., 1987)  (Table 77.1). Cardiovascular complications are the main cause of death in acute PSGN. Irreversible renal failure may follow acute GN if widespread extracapillary (crescentic) proliferation develops, but crescentic PSGN has a more favourable prognosis than other types of rapidly progressive GN. The medium-term prognosis of PSGN has been studied extensively. Initial reports in 1930 and 1940 indicated an excellent prognosis but follow-up periods were relatively short. Subsequent studies have produced variable results. The incidence of abnormal laboratory findings during the follow-up varies from 3.5% (Potter et al., 1982) to 60% (Baldwin et al., 1974). Discrepancies may result, in part, from the different prognosis of PSGN in adults and in children, which is not always taken into account in the reported series. The worse prognosis in adults may result in part from age-related tendency to fibrosis (see Chapter 140), or other changes such as impairment of the Fc-receptor function of the mononuclear phagocyte system (Mezzano et al., 1991). Several risk factors have a definite influence on the long-term prognosis of acute PSGN. A subgroup of adult patients that had massive proteinuria as the initial manifestation had an incidence of chronic renal failure as high as 77% (Vogl et  al., 1986). In an outbreak of PSGN in adults resulting from consumption of cheese contaminated with Streptococcus zooepidemicus, there was an alarming incidence of chronic renal disease: impaired renal function was found in 30% of the patients after 2  years of follow-up (10% of them in chronic dialysis therapy) (Balter et al., 2000; Pinto et al., 2001). Recent studies suggest that deficiency of complement factor H-related protein 5 may predispose to the development of chronic renal disease (Vernon et al., 2012). Studies after 1970 reporting the findings in children, 10–20 years after the acute episode, found that approximately 20% of the patients have an abnormal urinalysis or creatinine clearance, but < 1% develop end-stage kidney disease. However even 20 years leaves many more years at risk for individuals not yet in their middle years. Proteinuria and hypertension occur in 8–13% of the patients in most studies (range 1.4–46%). The incidence of glomerular sclerosis and fibrosis is nearly 50% (Gallo et al., 1980), but the clinical relevance of these histological characteristics is uncertain. Our own data, that include 110 children with epidemic and sporadic PSGN followed prospectively over 15–18 years after the acute episode, indicate an incidence of 7.2% of proteinuria, 5.4% of microhaematuria, 3.0% of arterial hypertension, and 0.9% of azotaemia. These values are essentially similar to those found in the general population. We have also followed 10 cases of subclinical PSGN for 10–11 years and the prognosis is excellent. Studies in Australian aboriginal communities (Hoy et al., 1998) indicate that patients who had acute PSGN have an increased risk for albuminuria (adjusted odds ratio (OR) 6.1, 95% confidence interval (CI)

2.2–16.9) and haematuria (OR 3.7, 95% CI 1.8–8.0) in relation to controls who did not have acute PSGN. Finally, the long-term prognosis of acute PSGN may be influenced by the coexistence of other risk factors of chronic renal failure. The association of PSGN, diabetes, and metabolic syndrome are likely responsible for the high incidence of end-stage renal disease in aboriginal communities in Northern Australia (White et al., 2001; Hoy et al., 2012). It is interesting that persisting arterial stiffness, as determined by brachial–ankle pulse wave velocity, is found in patients with acute PSGN who develop chronic renal disease (Yu et al., 2011).

References Ahn, S. Y. and Ingulli, E. (2008). Acute poststreptococcal glomerulonephritis: an update. Curr Opin Pediatr, 20, 157–62. American Academy of Pediatrics, Committee on Infectious Diseases. (2000). Group A streptococcal infections. In The Red Book, pp. 526–92. Elk Groove Village, II: American Academy of Pediatrics. Anthony, B. F., Kaplan, E. L., Wannamaker, L. W., et al. (1969). Attack rates of acute nephritis after type 49 streptococcal infections of the skin and of the respiratory tract. J Clin Invest, 48, 1697–704. Ardiles, L. G., Valderrama, G., Moya, P., et al. (1997). Incidence and studies on antigenic specificities of antineutrophil-cytoplasmic autoantibodies (ANCA) in poststreptococcal glomerulonephritis. Clin Nephrol, 47, 1–5. Asami, T., Tanaka, A., Gunji, T., et al. (1985). Elevated serum and urine sialic acid levels in renal diseases of childhood. Clin Nephrol, 23, 112–19. Bach, J. F., Chalons, S., Forier, E., et al. (1966). 10-year educational programme aimed at rheumatic fever in two French Caribbean Islands. Lancet, 347, 644–8. Baldwin, D. S., Gluck, M. C., Schacht, R. G., et al. (1974). The long-term course of poststreptococcal glomerulonephritis. Ann Intern Med, 80, 342–58. Balter, S., Benin, A., Pinto, S. W., et al. (2000). Epidemic nephritis in Nova Serrana, Brazil. Lancet, 355, 1776–80. Batsford, S. R., Mezzano, S., Mihatsch, M., et al. (2005). Is the nephritogenic antigen in poststreptococcal glomerulonephritis pyrogenic exotoxin B (SPE B) or GAPDH? Kidney Int, 68, 1120–29. Becker, C. G. and Murphy, G. E. (1968). The experimental induction of glomerulonephritis like that in man by infection with Group A streptococci. J Exp Med, 127, 1–23. Beres, S. B., Sesso, R., Pinto, S. W., et al. (2008). Genome sequence of a Lancefield group C Streptococcus zooepidemicus strain causing epidemic nephritis: New information about an old disease. PLoS ONE, 3, e3026. Berge, A., Kihlberg, B. M., Sjöholm, A. G., et al. (1997). Streptococcal protein H forms soluble complement-activating complexes with IgG, but inhibits complement activation by IgG-coated targets. J Biol Chem, 272, 20774–81. Burova, L. A., Nagornev, V. A., Pigarevsky, P. V., et al. (1998). Triggering of renal tissue damage in the rabbit by IgG Fc receptor-positive group A streptococci. APMIS, 106, 277–87. Burova, L., Pigarevsky, P., Seliverstova, V., et al. (2012). Experimental poststreptococcal glomerulonephritis elicited by IgG Fc-binding M family proteins and blocked by IgG Fc fragment. APMIS, 120, 221–30. Carapetis, J. R., Steer, A. C., Mullholand, E. K., et al. (2005). The global burden of group A streptococcal diseases. Lancet Infect Dis, 5, 685–94. Casey, J. R. and Pichichero, M. E. (2004). Meta-analysis of cephalosporin versus penicillin treatment of group A streptococcal tonsillopharyngitis in children. Pediatrics, 113, 866–82. Cebul, D. R. and Poses, R. M. (1986). The comparative cost-effectiveness of statistical decision rules and experienced physicians in pharyngitis management. JAMA, 256, 3353–7. Choi, S. H., Zhang, X., and Stinton, M. W. (1995). Dynamics of streptococcal histone retention by mouse kidneys. Clin Immunol Immunopathol, 6, 68–74.

Chapter 77 

Councilman, W. T. (1898). Acute interstitial nephritis. J Exp Med, 3, 393–420. Currie, B. J. and Brewster, D. R. (2001). Childhood infections in the tropical north of Australia. J Paediatr Child Health, 37, 326–30. D’Agati, V. D., Jennette, J. C., and Silva, F. G. (2005). Non-Neoplastic Kidney Diseases. Atlas of Nontumor Pathology, First Series, Fascicle 4, pp. 269–96. Washington, DC: American Registry of Pathology. Dochez, A. R. and Sherman, L. (1924). The significance of Streptococcus hemolyticus in scarlet fever and the preparation of a specific antiscarlatinal serum by immunization of the horse to Streptococcus hemolyticus scarlatinae. JAMA, 2, 542–4. Dodge, W. F., Spargo, B. F., and Travis, L. B. (1967). Occurrence of acute glomerulonephritis in sibling contacts of children with sporadic acute glomerulonephritis. Pediatrics, 40, 1028–30. Duvic, C., Desramé, J., Hérody, M., et al. (2000). Acute poststreptococcal glomerulonephritis associated with thrombotic microangiopathy in an adult. Clin Nephrol, 54, 169–73. Francis, A. J., Nimmo, G. R., Efstratiou, A., et al. (1993). Investigation of milk-borne Streptococcus zooepidemicus infection associated with glomerulonephritis in Australia. J Infect, 27, 317–23. Frémeaux-Bacchi, V., Weiss, L., Demouchy, C., et al. (1994). Hypocomplementaemia of poststreptococcal acute glomerulonephritis is associated with a C3 nephritic factor (C3NeF) IgG autoantibody activity. Nephrol Dial Transplant, 9, 1747–50. Gallo, G. R., Feiner, H. D., Steele, J. M. Jr., et al. (1980). Role of intrarenal vascular sclerosis in progression of poststreptococcal glomerulonephritis. Clin Nephrol, 13, 449–57. Haas, M. (2003). Incidental healed postinfectious glomerulonephritis: a study of 1012 renal biopsy specimens examined by electron microscopy. Hum Pathol, 34, 3–10. Heptinstall, R. H. (1974). Pathology of the Kidney (2nd ed.). Boston, MA: Little, Brown. Hoy, W. E., Mathews, J. D., McCredie, D. A., et al. (1998). The multidimensional nature of renal disease: rates and association of albuminuria in an Australian Aboriginal community. Kidney Int, 54, 1296–304. Hoy, W. E., White, A. V., Dowling, A., et al. (2012) Post-streptococcal glomerulonephritis is a strong risk factor for chronic kidney disease in later life. Kidney Int, 81, 1026–32. Johnston, F., Carapetis, J., Patel, M. S., et al. (1999). Evaluating the use of penicillin to control outbreaks of acute poststreptococcal glomerulonephritis. Pediatr Infect Dis J, 18, 327–32. Kasahara, T., Hayakawa, H., Okubo, S., et al. (2001). Prognosis of acute poststreptococcal glomerulonephritis (APSGN) is excellent in children, when adequately diagnosed. Pediatr Int, 43, 364–7. Kozyro, I., Perahud, I., Sadallah, S., et al. (2006). Clinical value of autoantibodies against C1q in children with glomerulonephritis. Pediatrics, 117, 1663–8. Lung, J. M., and Mallory, S. B. (2000). A child with spider bite and glomerulonephritis: a diagnostic challenge. Int J Dermatol, 39, 287–9. Marín, C., Mosquera, J., and Rodríguez-Iturbe, B. (1995). Neuraminidase promotes neutrophil, lymphocyte and macrophage infiltration in the normal rat kidney. Kidney Int, 47, 88–95. Marshall, C. S., Cheng, A. C., Markey, P. G., et al. (2011). Acute post-streptococcal glomerulonephritis in the Northern Territory of Australia: A review of 16 years data and comparison with the literature. Am J Trop Med Hyg, 85, 703–10. Matsell, D. G, Wyatt, R. J., and Glaber, L. W. (1994). Terminal complement complexes in acute poststreptococcal glomerulonephritis. Pediatr Nephrol, 8, 671–6. Melby, P. C., Musick, W. D., Luger, A. M., et al. (1987). Poststreptococcal glomerulonephritis in the elderly. Report of a case and review of the literature. Am J Nephrol, 7, 235–40. McIsaac, W. J., White, D., Tannenbaum, D., et al. (1998). A clinical score to reduce unnecessary antibiotic use in patients with sore throat. Can Med Assoc J, 158, 75–83. Mezzano, S., Kunick, M., Olavarria, F., et al. (1993). Detection of platelet-activating factor activity in plasma of patients with streptococcal nephritis. J Am Soc Nephrol, 4, 235–42.

post-streptococcal glomerulonephritis

Mezzano, S., Burgos, M. E., Olavarría, F., et al. (1997). Immunohistochemical localization of IL-8 and TGF-beta in streptococcal glomerulonephritis. J Am Soc Nephrol, 8, 234–41. Mezzano, S., Lopez, M. I., Olavarria, F., et al. (1991). Age influence on mononuclear phagocyte system Fc-receptor function in poststreptococcal nephritis. Nephron, 57, 16–22. Montseny, J. -J., Meyrier, A., Kleinknecht, D., et al. (1995). The current spectrum of infectious glomerulonephritis: experience with 76 patients and review of the literature. Medicine, 74, 63–73. Nasr, S. H., Markowitz, G. S., Stokes, M. B., et al. (2008). Acute postinfectious glomerulonephritis in the modern era: experience with 86 adults and review of the literatura. Medicine, 87, 21–32. Nicholson, M. L., Ferdinand, L., Sampson, J. S., et al. (2000). Analysis of immunoreactivity to a Streptococcus equi subsp. zooepidemicus M-like protein to confirm an outbreak of poststreptococcal glomerulonephritis, and sequences of M-liked proteins from isolates obtained from different host species. J Clin Microbiol, 38, 4126–30. Oda, T., Tamura, K., Yoshizawa, N., et al. (2008). Elevated urinary plasmin activity resistant to alpha2-antiplasmin in acute poststreptococcal glomerulonephritis. Nephrol Dial Transplant, 23, 2254–9. Oda, T., Yamakami, K., Omasu, F., et al. (2005). Glomerular plasmin-like activity in relation to nephritis-associated plasmin receptor in acute poststreptococcal glomerulonephritis. J Am Soc Nephrol, 16, 247–54. Oda, T., Yoshizawa, N., Yamakami, K., et al. (2010). Localization of nephritis-associated plasmin receptor in acute poststreptococcal glomerulonephritis. Hum Pathol, 41, 1276–85. Orta, N. and Moriyón, J. C. (2001). Epidemiología de las enfermedades renales en niños en Venezuela. Arch Venz Pueric Pediatr, 64, 76–83. Oshawa, I., Ohi, H., Endo, M., et al. (1999). Evidence of lectin complement pathway activation in postsreptococcal glomerulonephritis. Kidney Int, 56, 1158–9. Pais, P. J., Kump, T., and Greenbaum, L. A. (2008). Delay in diagnosis in poststreptococcal glomerulonephritis. J Pediatr, 153, 560–4. Parra G., Rodríguez-Iturbe, B., Batsford, S., et al. (1998). Antibody to streptococcal zymogen in the serum of patients with acute glomerulonephritis: amulticentric study. Kidney Int, 54, 509–17. Parra, G., Platt, J. L., Falk, R. J., et al. (1984). Cell populations and membrane attack complex in glomeruli and patients with poststreptococcal glomerulonephritis: identification using monoclonal antibodies by indirect immunofluorescence. Clin Exp Immunol Immunopathol, 33, 324–32. Parra, G., Romero, M., Henríquez-LaRoche, C., et al. (1994). Expression of adhesion molecules in poststreptococcal glomerulonephritis. Nephrol Dial Transplant, 9, 1412–4. Pedreanez, A., Viera, N., Rincon, J., et al. (2006). Increased IL-6 in supernatant of rat mesangial cell cultures treated with erythrogenic toxin type B and its precursor isolated from nephritogenic streptococci. Am J Nephrol, 26, 75–81. Pickering, M.C., D’Agati, V.D., Nester, C.M., et al. (2013). C3 glomerulopathy: consensus report. Kidney Int, 84, 1079–89. Pinto, S. L. W., Sesso, R., Vasconcelos, E., et al. (2001). Follow-up of patients with epidemic postreptococcal glomerulonephritis. Am J Kidney Dis, 38, 249–55. Poon-King, T., Bannan, J., Viteri, A., et al. (1993). Identification of an extracellular plasmin binding protein from nephritogenic streptococci. J Exp Med, 178, 759–63. Poon-King, T., Mohammed, I., Cox, R., et al. (1967). Recurrent epidemic nephritis in South Trinidad. N Engl J Med, 277, 728–33. Potter, E. V., Lipschultz, S. A., Abidh, S., et al. (1982). Twelve to seventeen-year follow-up of patients with poststreptococcal acute glomerulonephritis in Trinidad. N Engl J Med, 307, 725–9. Powell, H. T., McCredie, D., and Rotenberg, F. (1980). Response to furosemide in acute renal failure. Dissociation of renin and diuretic responses. Clin Nephrol, 14, 55–9. Rodríguez-Iturbe, B. (1984). Epidemic poststreptococcal glomerulonephritis. (Nephrology Forum). Kidney Int, 25, 129–36. Rodríguez-Iturbe, B. and Batsford, S. (2007). Pathogenesis of poststreptococcal glomerulonephritis a century after Clemens von Pirquet. Kidney Int, 71, 1094–104.

631

632

Section 3  

the patient with glomerular disease

Rodríguez-Iturbe, B. and Mezzano, S. (2005). Infections and kidney diseases: a continuing global challenge. In M. El Nahas (ed.) Kidney Diseases in the Developing World and Ethnic Minorities, pp. 59–82. London: Taylor & Francis, London. Rodríguez-Iturbe, B., Katiyar, V.N., and Coello, J. (1981a). Neuraminidase activity and free sialic acid levels in the serum of patients with acute poststreptococcal glomerulonephritis. N Engl J Med, 304, 1506–10. Rodríguez-Iturbe, B. and Musser, J.M. (2008). The current state of poststreptococcal nephritis. J Am Soc Nephrol, 19, 1855–64. Rodríguez-Iturbe, B., Rubio, L., Garcia, R. (1981b). Attack rate of poststreptococcal glomerulonephritis in families. A prospective study. Lancet, 1, 401–3 Romero, M., Mosquera, J., Novo, E., et al. (1999). Erythrogenic toxin type B and its precursor isolated from streptococci induce leukocyte infiltration in normal rat kidneys. Nephrol Dial Transplant, 14, 1867–74. Rosenberg, H. G., Vial, S. U., Pomeroy, J., et al. (1985) Acute glomerulonephritis in children. An evolutive morphologic and immunologic study of the glomerular inflammation. Pathol Res Pract, 180, 633–43. Sarkissian, A., Papazian, M., Azatian, G., et al. (1997). An epidemic of acute postinfectious glomerulonephritis in Armenia. Arch Dis Child, 77, 342–4. Seegal, D. and Earle, D. P. (1941). A consideration of certain biological differences between glomerulonephritis and rheumatic fever. Am J Med Sci, 201, 528–39. Sethi, S., Fervenza, F.C., Zhang, Y., et al. (2012). Atypical postinfectious glomerulonephritis is associated with abnormalities in the alternative pathway of complement. Kidney Int, 83, 293–9. Skattum, L., Akesson, P., Truedsson, L., et al. (2006). Antibodies against four proteins from a Streptococcus pyogenes serotype M1 strain and levels of circulating mannan-binding lectin in acute poststreptococcal glomerulonephritis. Int Arch Allergy Immunol, 140, 9–19. Sorger, K., Gessler, U., Hubner, F.K., et al. (1982) Subtypes of acute postinfectious glomerulonephritis. Synopsis of clinical and pathological features. Clin Nephrol, 17, 114–28. Tasic. V. and Polenakovic, M. (2000). Acute poststreptococcal glomerulonephritis following circumcision. Pediatr Nephrol, 15, 274–5. Thongboonkerd, V., Luengpailin, J., Cao, J., et al. (2002). Fluoride exposure attenuates expression of Streptococcus pyogenes virulence factors. J Biol Chem, 277, 16599–605. Tornroth, T. (1976) The fate of subepithelial deposits in acute poststreptococcal glomerulonephritis. Lab Invest, 35, 461–74 Vernon, K. A., Goicoechea de Jorge, E., Hall, A. E., et al. (2012). Acute presentation and persistent glomerulonephritis following streptococcal infection in a patient with heterozygous complement factor h-related protein 5 deficiency. Am J Kidney Dis, 60, 121–5. Viera, N., Pedreanez, A., Rincon, J., et al. (2009). Streptococcal zymogen type B induces angiotensin II in mesangial cells and leukocytes. Pediatr Nephrol, 24, 1005–11.

Vogl, W., Renke, M., Mayer-Eichberger, D., et al. (1986). Long-term prognosis for endocapillary glomerulonephritis of poststreptococcal type in children and adults. Nephron, 44, 58–65. Vogt, A., Batsford, S., Rodríguez-Iturbe, B., et al. (1983). Cationic antigens in poststreptococcal glomerulonephritis. Clin Nephrol, 20, 271–9. Vogt, A., Schmiedeke, T., Stöckl, F., et al. (1990). The role of cationic proteins in the pathogenesis of immune complex glomerulonephritis. Nephrol Dial Transplant, 5 Suppl 1, 6–9. Von Pirquet, C. (1910). Ergebn Inn Med Kinderheilk, 5, 459–539. [Translated into English in: von Pirquet, C. (1911). Allergy. Arch Intern Med, 7, 259–88, 382–436.] Wasserzug, O., Valinsky, L., Klement, E., et al. (2009). A cluster of ecthyma outbreaks caused by a single clone of invasive and highly infective Streptococcus pyogenes. Clin Infect Dis, 48, 1213–9. Webb, K. H., Needham, C. A., and Kurtz, S. R. (2000). Use of a high-sensitivity rapid strep test without culture confirmation of negative results: 2 years’ experience. J Fam Pract, 49, 34–6. White, A. V., Hoy, W. E., and McCredie, D. A. (2001). Childhood post-streptococcal glomerulonephritis as a risk factor for chronic renal disease in later life. Med J Aust, 174, 492–4. Wiles, K. S., Lee, M., Brindle, R., et al. (2011). Rare mmune-mediated pneumonitis in association with post-streptococcal glomerulonephritis. Nephrol Dial Transplant, 26, 4140–2. Yamakami, K., Yoshizawa, N., Wakabayashi, K., et al. (2000). The potential role for nephritis-associated plasmin receptor in acute poststreptococcal glomerulonephritis. Methods, 21, 185–97. Yoshizawa, N., Yamakami, K., Fujino, M., et al. (2004). Nephritis-associated plasmin receptor and acute poststreptococcal glomerulonephritis: characterization of the antigen and associated immune response. J Am Soc Nephrol, 15, 1785–93. Yu, M. C., Yu, M. S., Yu, M. K., et al. (2011). Acute reversible changes of brachial-ankle pulse wave velocity in children with acute poststreptococcal glomerulonephritis. Pediatr Nephrol, 26, 233–9. Zaffanello, M., Cataldi, L., Franchini, M., et al. (2010). Evidence-based treatment limitations prevent any therapeutic recommendation for acute poststreptococcal glomerulonephritis in children. Med Sci Monitor, 16, RA79–84. Zegers, R. H., Weigl, A., and Steptoe, A. (2009). The death of Wolfgang Amadeus Mozart: an epidemiologic perspective. Ann Intern Med, 151, 274–8. Zhang, L., Ignatowski, T. A., Spengler, R. N., et al. (1999). Streptococcal histone induces murine macrophages to produce interleukin-1 and tumor necrosis factor alpha. Infect Immunol, 67, 6473–7.

CHAPTER 78

Immunoglobulin A-dominant post-infectious glomerulonephritis Bernardo Rodriguez-Iturbe and Mark Haas Introduction The uncommon IgA-dominant variant of post-infectious glomerulonephritis has been particularly associated with infections with Staphylococcus, which may induce immune complex-mediated glomerulonephritis (GN) (see Chapter 77) as well as this variant. It was first described by Koyama et al. (1995) in association with methicillin-resistant Staphylococcus aureus (MRSA) infection that caused a severe form of GN, but the case for there being a specific type of nephritis associated with methicillin resistance is not strong (Usui et al., 2011). The frequency of this condition is unclear but reports have appeared from several centres around the world and it represents 1.6% of the biopsies in adult patients in a single institution (Satoskar et al., 2006)

Clinical characteristics The typical patient presents with acute renal injury and massive proteinuria and the renal biopsy not infrequently shows focal glomerular crescent formation (Haas et al., 2008). The defining characteristic is the immunoglobulin A (IgA) dominant or co-dominant immune deposition in the biopsy, frequently increased serum IgA levels (Wen and Chen, 2011), and specific T-cell receptor Vβ+ subsets in the serum. Staphylococcal superantigens have been implicated in the pathogenesis (Koyama et al., 1995). The clinical picture has been reviewed by Nasr and D’Agati (2011). IgA-dominant post-infectious GN is most frequent in older patients (mean age 60 years) and is more than three time more frequent in males than in females. The most common site of infection is skin, reported in 51% of patients in whom the site of infection was identified. Other frequent sites include surgical wounds and intravenous lines. Staphylococcus aureus was responsible in 94% of the reported cases, including MRSA. Gross haematuria was present in 20% of the cases, proteinuria in the nephrotic range was found in 51%, and rapidly progressive renal failure was frequent. Diabetes is a major risk factor and two-thirds of the patients who developed end-stage renal disease had diabetic nephropathy. Complement levels are frequently low but may be normal. Since non-infectious IgA nephropathy only rarely presents with massive proteinuria and acute kidney injury, and usually shows normal

complement levels, these findings may be useful in the differential diagnosis (Satoskar et al., 2006).

Pathogenesis The pathogenesis of this GN is not entirely clear. There may be activation of a selective IgA response or, in some cases, an intense T-cell activation and T cells activate B cells to produce polyclonal IgA and IgG. The T-cell activation is the result of direct binding of staphylococcal superantigens to the major histocompatibility complex class II molecules in antigen presenting cells that then engage the Vβ+ T-cell receptor region. Koyama et al. (2004) have identified an antigen in Staphylococcus aureus that is co-localized with glomerular IgA deposits and developed an experimental model of the disease immunizing Balb/c mice with this antigen (Sharmin et al., 2004).

Pathology There have been a number of reports of GN with IgA-containing and often IgA-dominant or co-dominant immune complex deposits in association with infections with Staphylococcus species, including methicillin-sensitive S. aureus (MSSA), MRSA, and S. epidermidis (Koyama et al., 1995; Nasr et al., 2003; Satoskar et al., 2006; Haas et al., 2008). The histologic features of these cases have been quite variable, including predominantly mesangial proliferative GN, membranoproliferative-type GN with or without crescents, in some instances resembling cryoglobulinaemic GN, and diffuse proliferative and exudative GN resembling acute post-streptococcal GN (Koyama et al., 1995; Nasr et al., 2003; Satoskar et al., 2006). Nasr et al. (2003) reported five of the latter cases, three following MSSA infections and two following S. epidermidis infections, each superimposed on underlying diabetic nephropathy; a similar case is illustrated in Figs 78.1 and 78.2. Notably, many of these reported cases, including those of Nasr et al. (2003), did not show prominent subepithelial ‘humps’ by electron microscopy. A recent series of 13 cases of IgA-dominant post-infectious GN (Haas et al. 2008), identified based on morphologic features including subepithelial ‘humps’ with or without evidence of resorption, demonstrated the full gamut of pathologic features described for post-streptococcal GN (see Chapter 77). Two of the biopsies

634

Section 3  

the patient with glomerular disease

Fig. 78.1  Acute post-infectious (post-staphylococcal) glomerulonephritis superimposed on diabetic glomerulosclerosis. Note the prominent neutrophils, as well as the nodular expansion of the mesangial matrix. (Periodic acid–Schiff stain, original magnification 400×, scale bar 50 microns.)

showed diffuse proliferative and exudative GN with prominent subepithelial ‘humps’ (acute post-infectious GN), three showed diffuse proliferative GN with only occasional neutrophils and subepithelial ‘humps’ showing evidence of resorption (as shown in Chapter  77, Figs 77.8 and 77.9; subacute), and the remaining eight showed mainly mesangial proliferative changes with largely resorbed subepithelial deposits, often localized to the mesangial ‘waist’ region of glomeruli (as shown in Chapter 77, Figs 77.2 and 77.10; resolving or persistent). Five of the 13 patients were diabetic and three had diabetic nephropathy, and seven had documented infections prior to the onset of GN (three MRSA, three MSSA, one hepatitis C; two additional patients were HIV positive). IgA was the dominant immunoglobulin present in all cases, with five biopsies showing staining for IgG and 10 for IgM; C3 was present in all cases with a mean staining intensity approximately equivalent to that for IgA. Biopsies with histologic changes of acute or subacute post-infectious GN showed a ‘starry sky’ pattern of IgA and C3 staining by immunofluorescence (Fig. 78.2), while those with resolving/persistent histology had mainly a mesangial pattern (Haas et al., 2008).

Distinction from IgA nephropathy While acute, IgA-dominant post-infectious GN is rather easily diagnosed by light microscopy and immunofluorescence alone, subacute and resolving/persistent cases often cannot be distinguished from IgA nephropathy without electron microscopy. This is particularly true with cases of resolving/persistent IgA-dominant post-infectious GN, which show predominantly mesangial proliferative changes histologically, a mesangial pattern of IgA and C3 deposits by immunofluorescence, and in our experience often lack a clear infection history (Haas et al. 2008). As with IgA nephropathy, C1q staining by immunofluorescence is usually absent in IgA-dominant post-infectious GN. The latter typically shows equivalent staining intensity for kappa and lambda light chains, while some cases of IgA nephropathy exhibit a clear lambda predominance (Lai et al., 1986; Jennette, 1988) which when present can be helpful in resolving this differential diagnosis. However, the most definitive method for distinguishing resolving/persistent IgA-dominant post-infectious GN

Fig. 78.2  Subacute post-staphylococcal glomerulonephritis superimposed on diabetic glomerulosclerosis: immunofluorescence microscopy for IgA. There is granular staining in the peripheral capillary walls and mesangial areas; the overall appearance is that of a ‘starry sky’ pattern transitioning into a mesangial pattern (see also Figs 77.4 and 77.5 in Chapter 77). This biopsy showed strong staining for IgA and C3, with weak and segmental staining for IgM and minimal staining for IgG. In the background at right, the underlying nodular expansion of the mesangial matrix can be appreciated. (FITC conjugated anti-human IgA, original magnification 400×.)

from IgA nephropathy remains the identification of prominent although partially or largely resorbed subepithelial deposits by electron microscopy, with at least some of these localized to the mesangial ‘waist’ region of glomeruli.

Treatment Antibiotic treatment of staphylococcal infection may be associated with recovery of renal function. Steroid or immunosuppressive treatment is contraindicated if infection is present but it has been used in selected cases with protracted azotaemia when active infection is no longer present (Okuyama et  al., 2008; Chen and Wen, 2011).

References Chen, Y. R. and Wen, Y. K. (2011). Favorable outcome of crescentic IgA nephropathy associated with methicillin-resistant Staphylococcus aureus infection. Renal Fail, 33, 96–100. Haas, M., Racusen, L. C., and Bagnasco, S. M. (2008). IgA-dominant postinfectious glomerulonephritis: a report of 13 cases with common ultrastructural features. Hum Pathol, 39, 1309–16. Jennette, J. C. (1988). The immunohistology of IgA nephropathy. Am J Kidney Dis, 12, 348–52. Koyama, A., Kobayashi, M., Yamaguchi, N., et al. (1995). Glomerulonephritis associated with MRSA infection: a possible role of bacterial superantigen. Kidney Int, 47, 207–16. Koyama, A., Sharmin, S., Sakurai, H., et al. (2004). Staphylococcus aureus cell envelope antigen is a new candidate for the induction of IgA nephropathy. Kidney Int, 66, 121–32.

Chapter 78 

Lai, K. -N., Chan, K. W., Lai, F. M. -M., et al. (1986). The immunochemical characterization of the light chains in the mesangial IgA deposits in IgA nephropathy. Am J Clin Pathol, 85, 548–51. Nasr, S. H. and D’Agati, V. D. (2011). IgA-dominant postinfectious glomerulonephritis: a new twist in an old disease. Nephron Clin Pract, 119, c18–25. Nasr, S. H., Markowitz, G. S., Whelan, J. D., et al. (2003). IgA-dominant acute poststaphylococcal glomerulonephritis complicating diabetic nephropathy. Hum Pathol, 34, 1235–41 Okuyama, S., Wakui, H., Maki, N., et al. (2008). Successful treatment of post-MRSA infection glomerulonephritis with steroid therapy. Clin Nephrol, 70, 344–7.

iga-dominant post-infectious glomerulonephritis Satoskar, A. A., Nadasdy, G., Plaza, J. A., et al. (2006). Staphylococcus infection-associated glomerulonephritis mimicking IgA nephropathy. Clin J Am Soc Nephrol, 1, 1179–86. Sharmin, S., Shimizu, Y., Hagiwara, M., et al. (2004). Staphylococcus aureus antigens induce IgA-type glomerulonephritis in Balb/c mice. J Nephrol, 17, 504–11. Usui, J., Kobayashi, M., Ebihara, I., et al. (2011). Methicillin-resistant Staphylococcus-aureus-associated glomerulonephritis on the decline: decreased incidence since the 1990s. Clin Exp Nephrol, 15, 184–6. Wen, Y. K. and Chen, M. L. (2011). IgA-dominant postinfectious glomerulonephritis: not peculiar to staphylococcal infection and diabetic patients. Renal Fail, 33, 480–5.

635

CHAPTER 79

Glomerulonephritis associated with endocarditis, deep-seated infections, and shunt nephritis Bernardo Rodriguez-Iturbe and Mark Haas Glomerulonephritis associated with endocarditis Introduction and epidemiology Bacterial endocarditis may have an acute or subacute clinical presentation and is characterized by the formation of vegetations, composed of platelets, fibrin, inflammatory cells, and bacteria, on heart endothelial surfaces. The incidence of community-acquired native valve endocarditis in population studies in the United States and Western Europe has remained essentially unchanged in the last three decades. Indications for prophylactic antibiotic therapy have been reduced without increment in the incidence of infective endocarditis (Duval et al., 2012). Population studies indicate 1.7–6.2 cases per 100,000 person years (Tleyjeh et al., 2007). Not surprisingly, there is a decline in the number of cases associated with rheumatic heart disease and an increase in cases associated with valve surgery and intravenous drug abuse (Tleyjeh et al., 2005). The heart lesion found to predispose most frequently to endocarditis is mitral valve prolapse and the microorganism most frequently involved is Staphylococcus aureus (Mylonakis and Calderwood, 2001). The most important change in the epidemiology of bacterial endocarditis is the increased incidence related to healthcare interventions. Instead of being a disease that affects predominantly young adults with rheumatic heart disease, it presents now in older patients and in patients at risk, which include drug users, patients with prosthetic valves and implantable devices, and patients with HIV infection. Staphylococcus aureus and Staphylococcus epidermidis are the most common pathogens in hospital-acquired infective endocarditis while Streptococcus infections are more frequent in community acquired and native valve endocarditis (Furuno et al., 2011; Que and Moreillon, 2011); however, the incidence of infections by methicillin-resistant Staphylococcus aureus (MRSA) is increasing in community-acquired endocarditis (David and Daum, 2010). In 1910, Löhlein described ‘embolic’ non-suppurative glomerulonephritis caused by bacterial endocarditis (Löhlein, 1910). Subsequent descriptions emphasized the immune complex pathogenesis of this disease, particularly in patients with Streptococcus viridians infections of rheumatic or congenital valve disease (Neugarten and Baldwin, 1984; Eknoyan, 1985). In recent

reports, glomerulonephritis (GN) is present in 26% of patients with heart valve infection frequently associated with renal vasculitis. Localized infarcts (31%), many of which were septic, were also described. Interstitial nephritis, mostly attributable to antibiotic therapy (10%) and cortical necrosis (10%) were also significant autopsy findings (Majumdar et al., 2000). Immune deposits are sometimes present (Kirkpantur et  al., 2007; Lee et  al., 2007; Nasr et al., 2008) but may be absent (Majumdar et al., 2000). The increased incidence of bacterial endocarditis in association with healthcare interventions is particularly evident in patients treated with maintenance haemodialysis. Bacterial endocarditis is 20–60 times more frequent in these patients than in the general population and carries a near 50% mortality risk (Hoen, 2004; Rekik et  al., 2009); infections with MRSA and vancomycin-resistant enterococci are associated with a poorer outcome (Leone and Suter, 2010). Emphasis has recently been placed on the need of strict infection control policies, prompt conversion to arteriovenous access from catheters and appropriate antibiotic prescriptions (Fitzgibbons et al., 2011).

Clinical features The clinical picture of subacute bacterial endocarditis includes splinter haemorrhages, Osler nodules, and Janeway lesions but these manifestations are seen only occasionally. Heart murmurs, fever, anaemia, leucocytosis, elevated sedimentation rate, and purpura are frequent. In particular, purpura on the neck is highly suggestive of endocarditis. Patients at risk with fever should be evaluated for endocarditis since 38% of cases diagnosed at autopsy were not diagnosed clinically (Fernández Guerrero, et al., 2012). In subacute forms of endocarditis, anorexia and weight loss are commonly observed. Infective endocarditis in haemodialysis patients usually is associated with infections originating from the vascular access, particularly in arteriovenous grafts or dialysis catheters in use for > 1  year (McCarthy and Steckelberg, 2000; Nori et al., 2006; and see Chapter 269). The most frequently infected heart valves in haemodialysis patients are the mitral and aortic valves and the vegetations are best demonstrated with trans-oesophageal echocardiography (Tao et  al., 2010). Systemic embolization with large emboli is usually associated with endocarditis caused by fungi or Haemophilus.

Chapter 79 

One-third of patients with bacterial endocarditis develop azotaemia and the risk increases with age, a history of hypertension, thrombocytopenia, and prosthetic valve infection (Conlon et al., 1998). Renal manifestations of GN in patients with bacterial endocarditis include microscopic haematuria and proteinuria. A rapidly progressive azotaemia is a feature of crescentic GN. A clinical presentation with nephrotic syndrome is distinctly unusual. GN in a patient with bacterial endocarditis may be not easy to differentiate from renal involvement resulting from embolization or with interstitial nephritis due to antibiotic treatment. Large emboli may produce flank pain and haematuria while microscopic emboli produce microabscesses. Eosinophilia suggests antibiotic-induced interstitial nephritis but eosinophiluria is not a specific finding (see Chapter 6). The serological findings frequently include decreased C3 and C4 levels (except in superantigen-mediated GN), high titres of rheumatoid factor, serum cryoglobulins, and occasionally positive anti-PR3 antineutrophil cytoplasmic antibodies (ANCAs) (Fukuda et al., 2006; Satake et al., 2011).

Aetiology and pathogenesis Staphylococcus aureus is the most common aetiologic agent, followed by S. epidermidis, Streptococcus viridians, S. pyogenes, and Enterococcus fecalis. Less frequent are Gram-negative infections (Escherichia coli, Proteus). Experimental studies have confirmed that high-level bacteraemia is not required for endocarditis and a cumulative exposure to low-level bacteraemia is a genuine risk of endocarditis (Veloso et  al., 2011). As mentioned above, there are several kidney lesions associated with bacterial endocarditis but the pathogenesis of GN in this condition involves the deposition of immune complexes formed with bacterial antigens (Yum et al., 1978) and, in some cases, the participation of superantigens directly activating T cells and inducing a polyclonal hypergammaglobulinaemia (Koyama et al., 1995; Yoh et al., 2000).

Pathology Recent North American and European studies of acute post-infectious GN in adults have identified bacterial endocarditis as the underlying infectious process in approximately 10% of cases (Nasr et al., 2008; Montseny et al., 1995). In the older literature, including autopsy studies from the pre-antibiotic era, focal GN was the most common pattern of glomerular disease associated with bacterial endocarditis, although diffuse GN (in some instances exudative) was not uncommonly seen (Baehr, 1912; Bell, 1932). However, in more recent studies two histologic patterns of GN predominate: diffuse endocapillary proliferative and exudative GN, similar to acute post-infectious GN resulting from infections at other sites such as the upper respiratory tract and the skin (see Fig. 77.1 in Chapter 77) and necrotizing and crescentic GN (Fig. 79.1), often resembling that seen in ANCA-associated vasculitis (ANCA-GN) (Majumdar et al., 2000; Nasr et  al., 2008) (See Chapter 157). Notably, the latter form of GN, which was pauci-immune by immunofluorescence similar to ANCA-GN, was seen in 11/16 patients with GN and confirmed bacterial endocarditis reported in a series from the United Kingdom by Majumdar et al. (2000), and this form was predominant in patients diagnosed both by renal biopsy and autopsy, although only one of five patients tested had a positive ANCA

nephritis in endocarditis and similar infections

Fig. 79.1  Pauci-immune necrotizing and crescentic glomerulonephritis. This biopsy is from a patient with ANCA-associated vasculitis. The silver stain, which stains the basement membrane matrix black, shows segmental necrosis of the glomerular tuft with pink-staining fibrin, and disruption of the glomerular basement membrane. Adjacent to the area of necrosis is an early cellular crescent. Note the lack of hypercellularity in portions of the glomerulus not involved by necrosis or crescent formation; this latter feature is typical of pauci-immune necrotizing and crescentic glomerulonephritis and contrasts with the hypercellularity of crescentic forms of immune complex-mediated glomerulonephritides, including post-streptococcal glomerulonephritis. (Jones silver methenamine stain, original magnification 400×, scale bar 50 microns.)

serology. Three patients had diffuse endocapillary proliferative and exudative GN, and two had MPGN-like lesions (Majumdar et al., 2000). GN with crescents was also seen in 7 out of 10 patients with bacterial endocarditis from the French series (all biopsies) of Montseny et al. (1995), although the GN in these cases tended not to resemble ANCA-GN as in most instances endocapillary as well as extracapillary hypercellularity was present, and immunofluorescence studies showed deposits of complement in most cases, with or without accompanying immunoglobulin. The three remaining cases showed diffuse endocapillary proliferative and exudative GN (Montseny et al., 1995). In the biopsy series of Nasr et al. (2008) from the United States, 8 out of 10 patients with infective endocarditis had diffuse endocapillary proliferative and exudative GN, while 2 out of 10 (one with a positive C-ANCA) has necrotizing/ crescentic GN, like ANCA-GN without significant endocapillary proliferation. In the latter cases, immune complex deposits were limited to the mesangium (Nasr et al., 2008). The reason for the differences between studies is not entirely clear, although focal GN seen in older autopsy series could represent partially resolved lesions detected later in the disease course than lesions diagnosed by renal biopsy. The high fraction of patients with pauci-immune necrotizing/crescentic GN in the series of Majumdar et al. (2000) was seen in both patients diagnosed by renal biopsy and at autopsy, and does not appear to be related to concurrent ANCA vasculitis. Still, while the fraction of patients with this form of glomerular pathology in this study was higher than in other studies, it should be emphasized that other studies have reported lesions resembling ANCA-GN in patients with bacterial endocarditis (Morel-Maroger et  al., 1972; Boulton-Jones et  al., 1974; Neugarten and Baldwin, 1984; Nasr et al., 2008), and immune complexes do not appear to account for all glomerular lesions in these patients.

637

638

Section 3  

the patient with glomerular disease

Treatment and outcome Complete eradication of the infection usually requires 4–6 weeks of antibiotic treatment. Serological abnormalities are usually corrected in this time but proteinuria, haematuria, and mild elevation of serum creatinine may persist for several months. As in other varieties of crescentic GN, patients with rapidly progressive GN have been treated with ‘pulse’ steroid therapy and plasmapheresis (Couzi et  al., 2004; Koya et  al., 2004)  but the value of these treatments remains unproven. Mortality of patients with bacterial endocarditis is as high as 36% in patients who develop kidney failure (Baddour et  al., 2005). Plasmapheresis has been anecdotally reported to improve crescentic GN resulting from bacterial endocarditis (Daimon et al., 1998), but this and any other potential immunosuppressive manoeuvres carry significant additional risks in this patient group.

Glomerulonephritis associated with ventriculoatrial shunt infections Shunt nephritis is now exceptionally rare, but a similar syndrome may occur with infections of other long term intravascular devices such as central vein catheters or pacing wires. Atrioventricular and peritoneoventricular shunts are used for alleviating the intracranial pressure in congenital and acquired hydrocephalus. Previous studies indicate that 30% of the atrioventricular shunts become infected from 2 months to many years after insertion (Haffner et al., 1997; Kubota et al., 2001). However, in recent studies, the incidence of infection in cerebrospinal fluid shunts in adults appears to be considerably less. A recent prospective 8-year surveillance study found that 6.1% of adult patients developed shunt infection (Korinek et al., 2011). The initial description of a patient with infected atrioventricular shunt who developed nephrotic syndrome was made by Black et al. (1965). It is now recognized that GN may occur in 1–2% of infected atrioventricular shunts; in contrast, this complication is seldom seen with infected ventriculoperitoneal shunts. The infecting organisms are usually Staphylococcus epidermidis (75% of the cases) and S.  aureus and less frequently Propionibacterium acne, diphtheroids, Pseudomonas, and Serratia species. The clinical picture of infected ventriculoatrial or ventriculoperitoneal shunts includes low-grade fever, hepatosplenomegaly, arthralgias, weight loss, anaemia, and skin rash, with or without increased intracranial pressure. The diagnosis may sometimes be difficult by standard methods and when cultures are negative. Eosinophilia in the cerebrospinal fluid, which is a sign of malfunctioning shunt (Heidemann et al., 2010; Rehman et al., 2011) and positron emission tomography may be of help in the diagnosis (Rehman et al., 2011). GN, when present, is manifested by microscopic haematuria and proteinuria, frequently in the nephrotic range. The full picture of nephrotic syndrome occurs in 25% of the cases. Serological findings include high rheumatoid factor titres, cryoglobulinaemia, depressed serum complement levels, and, in some patients, positive proteinase 3 ANCA titres (Dobrin et  al., 1975; ter Borg et al., 1991; Iwata et al., 2004).

Pathology The initial description of shunt nephritis noted a ‘lobular proliferative’ GN (Black et al., 1965). From a number of published series since that time (Arze et al., 1983; Vella et al., 1995; Haffner et al., 1997;

Ozcan et al., 2001; D’Agati et al., 2005), it is now well recognized that the most common histologic pattern of GN associated with infected ventriculoatrial shunts is an MPGN (type I) pattern, with diffuse, global endocapillary hypercellularity and glomerular capillary wall thickening with double contours of the glomerular basement membrane (GBM) appreciated on periodic acid–Schiff (PAS) and silver methenamine stains (Fig. 79.2). Less commonly, other patterns of GN may be seen, including diffuse endocapillary proliferative and exudative GN, and predominantly mesangial proliferative GN that may be diffuse or focal. While crescents may be seen, the presence of these in > 50% of glomeruli is rare (D’Agati et al., 2005). Immunofluorescence microscopy shows immune complex deposits along glomerular capillary walls that typically contain immunoglobulin (Ig)-M and C3, frequently C1q, and more variably IgG and IgA (D’Agati et  al., 2005). Interestingly, while the most common underlying infection is with staphylococcal species, mainly S. epidermidis (Vella et al., 1995; Haffner et al., 1997), staining for IgA, when present, is typically of lower intensity than that for IgM and IgG (D’Agati et  al., 2005). This contrasts with IgA-dominant post-infectious GN associated with staphylococcal infections (see Chapter 78) and together with the different histologic patterns would indicate different pathogenic mechanisms for these lesions. Electron microscopy in the majority of cases of shunt nephritis shows immune complex deposits with a distribution consistent with an MPGN-like lesion, mainly in mesangial and subendothelial locations, the latter often associated with GBM duplication (Vella et al., 1995; D’Agati et al., 2005). Subepithelial ‘humps’ are seen in a minority of cases, including those with endocapillary proliferative and exudative GN, and in biopsies with mesangial proliferative GN deposits may be restricted to mesangial areas (D’Agati et al., 2005). Bacterial antigens have been demonstrated in the glomeruli (Dobrin et al., 1975).

Treatment and outcomes Treatment consists of antibiotic therapy and removal of the infected atrioventricular shunt that is substituted at a later date by a ventriculoperitoneal shunt. If dialysis is necessary, haemodialysis is the modality indicated to avoid potential peritoneal infection that

Fig. 79.2  Shunt nephritis. There is a membranoproliferative-type glomerulonephritis with endocapillary hypercellularity, double contours of the glomerular capillary basement membrane, and lobular accentuation. (PAS stain, original magnification 400×, scale bar 50 microns.)

Chapter 79 

carries the risk of meningitis in patients with ventriculoperitoneal shunts. Complete recovery is the outcome in more than half of the patients but persistent urinary abnormalities and end-stage renal failure are reported in 22% and 6% of the patients, respectively (Haffner et al., 1997).

Glomerulonephritis associated with deep-seated infections (abscesses, osteomyelitis) Osteomyelitis, intra-abdominal abscesses, pneumonia, and dental abscesses are sometimes associated with glomerulonephritis. Osteomyelitis and intra-abdominal abscesses are generally present for several months prior to renal disease that can manifest with mild urinary abnormalities, rapidly progressive azotaemia or, more frequently, by nephrotic syndrome (Beaufils et al., 1976; Ho et al., 2008). In contrast with other infection-related GN, serum complement levels are often normal. Hypergammaglobulinaemia is sometimes present. Treatment of the infection, if started early, improves the renal disease. Lobar pneumonia caused by Streptococcus pneumoniae has been reported to cause GN usually manifested by proteinuria and haematuria (Phillips et al., 2005; Hoshino et al., 2007). The pathogenesis is mediated by immune complexes and pneumococcal antigens (type 14) that have been localized in the glomeruli. The bacterial capsular antigen is capable of activating the alternative complement pathway. In some cases of pneumococcal pneumonia the patient may develop haemolytic-uraemic syndrome due to unmasking of the Thomsen–Friedenreich antigen by the bacterial neuraminidase (Krysan et al., 2001) (see Chapter 174).

Pathology A number of different histologic patterns of GN have been reported with these infections, although the most common appears to be diffuse proliferative GN, often with crescents (Beaufils et  al., 1976; Whitworth et  al., 1976). Prominent monocytic infiltration of glomeruli has been reported, albeit not to the extent seen with cryoglobulaemic GN (Magil, 1984). Membranoproliferative-type GN has been infrequently reported, while a predominantly mesangial proliferative pattern may indicate partial resolution. Immunofluorescence studies, even in diffuse proliferative lesions with crescents, tend to show glomerular capillary wall and mesangial deposits containing mainly C3, with little or no specific staining for immunoglobulins (Beaufils et al., 1976; D’Agati et al., 2005).

References Arze, R. S., Rashid, H., Morley, R., et al. (1983). Shunt nephritis. Report of two cases and review of the literature. Clin Nephrol, 19, 48–53. Baddour, L. M., Wilson, W. R., Bayer, A. S. et al. (2005). Infective endocarditis: diagnosis, antimicrobial therapy, and management of complications. A statement for healthcare professionals from the Committee on Rheumatic Fever, Endocarditis, and Kawasaki Disease, Council on Cardiovascular Disease in the Young and the Councils on Clinical Cardiology, Stroke, and Cardiovascular Surgery and Anesthesia, American Heart Association: endorsed by the Infectious Disease Society of America. Circulation, 111, e394–434. Baehr, G. (1912). Glomerular lesions of subacute bacterial endocarditis. J Exp Med, 15, 330–47. Beaufils, M., Morel-Maroger, L., Sraer, J. D., et al. (1976). Acute renal failure of glomerular origin during visceral abscesses. N Engl J Med, 295, 185–9.

nephritis in endocarditis and similar infections

Bell, E. T. (1932). Glomerular lesions associated with endocarditis. Am J Pathol, 8, 639–63. Black, J. A., Chaacombe, D. N., and Ockenden, B. G. (1965). Nephrotic syndrome associated with bacteraemia after shunt operations for hydrocephalus. Lancet, 2, 921–4. Boulton-Jones, J. M., Sissons, J. G. P., Evans, D. J., et al. (1974). Renal lesions of subacute infective endocarditis. Br Med J, 2, 11–14. Conlon, P. J., Jefferies, F., Krigman, H. R., et al. (1998). Predictors of prognosis and risk of acute renal failure in bacterial endocarditis. Clin Nephrol, 49, 96–101. Couzi, L., Morel, D., Deminiére, C., et al. (2004). An unusual endocarditis-induced crescentic glomerulonephritis treated by plasmapheresis. Clin Nephrol, 62, 461–4. D’Agati, V. D., Jennette, J. C., and Silva, F. G. (2005). Non-neoplastic kidney diseases. In Atlas of Nontumor Pathology (First Series, Fascicle 4), pp. 269–96. Washington, DC: American Registry of Pathology. Daimon, S., Mizuno, Y., Fujii, S., et al. (1998). Infective endocarditis-induced crescentic glomerulonephritis dramatically improved by plasmapheresis. Am J Kidney Dis, 32, 309–13. David, M. Z. and Daum, R. S. (2010). Community-associated methicillin-resistant Staphylococcus aureus: epidemiology and clinical consequences of an emerging epidemic. Clin Microbiol Rev, 23, 616–87. Dobrin, R. S., Day, N. K., Quie, P. G., et al. (1975). The role of complement, immunoglobulin and bacterial antigens in coagulase-negative Staphylococcus shunt nephritis. Am J Med, 59, 660–73. Duval, X., Delahaye, F., Alla, F., et al. (2012). Temporal trends in infective endocarditis in the context of prophylaxis guideline modifications: three successive population-based surveys. J Am Coll Cardiol, 29, 1968–76. Eknoyan, G. (1984). Renal complications of bacterial endocarditis. Am J Nephrol, 5, 457–69. Fitzgibbons, L. N., Puls, D. L., Mackay, K., et al. (2011). Management of gram-positive coccal bacteremia and hemodialysis. Am J Kidney Dis, 57, 624–40. Fernández Guerrero, M. L., Álvarez, B., Manzarbeitia, F., et al. (2012). Infective endocarditis at autopsy: a review of pathologic manifestations and clinical correlates. Medicine, 91, 152–64. Fukuda, M., Motokawa, M., Usami, T., et al. (2006). PR3-ANCA-positive crescentic necrotizing glomerulonephritis accompanied by isolated pulmonic valve infective endocarditis, with reference to previous reports of renal pathology. Clin Nephrol, 66, 202–9. Furuno, J. P., Johnson, J. K., Schweizer, M. L., et al. (2011). Community-associated methicillin-resistant Staphylococcus aureus bacteremia and endocarditis among HIV patients: a cohort study. BMC Infect Dis, 11, 298. Haffner, D., Schinderas, F., and Aschoff, A. (1997). The clinical spectrum of shunt nephritis. Nephrol Dial Transplant, 12, 1143–8. Heidemann, S. M., Fiore, M., Sood, S., et al. (2010). Eosinophil activation in the cerebrospinal fluid of children with shunt obstruction. Pediatr Neurosur, 46, 255–8. Ho, C. I., Wen, Y. K., and Chen, M. L. (2008). Glomerulonephritis with acute renal failure related to osteomyelitis. Journal of Chinese Medical Association, 71, 315–17. Hoen, B. (2004). Infective endocarditis: a frequent disease in dialysis patients. Nephrol Dial Transplant, 19, 1360–2. Hoshino, C., Satoh, N., Sugawara, S., et al. (2007). Community-acquired Staphylococcus aureus pneumonia accompanied by rapidly progressive glomerulonephritis and hemophagocytic syndrome. Intern Med, 46, 1047–53. Iwata, Y., Ohta, S., Kawai, K., et al. (2004) Shunt nephritis with positive titers for ANCA specific for proteinase 3. Am J Kidney Dis, 43, e11–16. Kirkpantur, A., Altinbas, A., Arici, M., et al. (2007). Enterococcal endocarditis associated with crescentic glomerulonephritis. Clin Exp Nephrol, 11, 321–5. Korinek, A. M., Fulla-Oller, L., Boch, A. L., et al. (2011). Morbidity of ventricular cerebrospinal fluid shunt surgery in adults an 8-year study. J Neurosurg Sci, 55, 161–3.

639

640

Section 3  

the patient with glomerular disease

Koya, D., Shibuya, K., Kikkawa, R., et al. (2004). Successful recovery of infective endocarditis-induced rapidly progressive glomerulonephritis by steroid therapy combined with antibiotics: a case report. BMC Nephrol, 5, 18. Koyama, A., Kobayashi, M., Yamaguchi, N., et al. (1995). Glomerulonephritis associated with MRSA infection: a possible role of bacterial superantigen. Kidney Int, 47, 207–16. Krysan, D. J. and Flynn, J. T. (2001). Renal transplantation after Streptococcus pneumoniae-associated hemolytic uremic syndrome. Am J Kidney Dis, 37, E15. Kubota, M., Sakata, Y., Saeki, N., et al. (2001). A case of shunt nephritis diagnosed 17 years after ventriculoatrial shunt implantation. Clin Neurol Neurosurg, 103, 245–6. Landry, D. L., Braden, G. L., Gobeille, S. L., et al. (2010). Emergence of gentamicin-resistant bacteremia in hemodialysis patients receiving gentamicin lock catheter prophylaxis. Clin J Am Soc Nephrol, 5, 1799–804. Lee, L. C., Lam, K. K., Lee, C. T., et al. (2007). ‘Full house’ proliferative glomerulonephritis: an unreported presentation of subacute infective endocarditis. J Nephrol, 20, 745–9. Leone, S. and Suter, F. (2010). Severe bacterial infections in haemodialysis patients. Infezione Medicale, 18, 79–85. Löhlein, M. (1910). Über hämorrhagische Nierenalffek bei chronischer ulzeröser Endokarditis (embolische nichteiterige Herdnephritis). Medizinische Klinik, 6, 375–9. Magil, A. B. (1984). Monocytes and glomerulonephritis associated with remote visceral infection. Clin Nephrol, 22, 169–75. Majumdar. A., Chowdhary, S., Ferreira, M. A. A. S., et al. (2000). Renal pathological findings in infective endocarditis. Nephrol Dial Transplant, 15, 1782–7. McCarthy, J. T. and Steckelberg, J. M. (2000). Infective endocarditis in patients receiving long-term hemodialysis. Mayo Clin Proc, 75, 1008–14. Montseny, J. -J., Meyrier, A., Kleinknecht, D., et al. (1995). The current spectrum of infectious glomerulonephritis: experience with 76 patients and review of the literature. Medicine, 74, 63–73. Morel-Maroger, L., Sraer, J. D., Herreman, G., et al. (1972). Kidney in subacute endocarditis. Pathological and immunofluorescent findings. Arch Pathol, 94, 205–13. Mylonakis, E. and Calderwood, S.B. (2001). Infective endocarditis in adults. N Engl J Med, 345, 1318–30. Nasr, S. H., Markowitz, G. S., Stokes, M. B., et al. (2008). Acute postinfectious glomerulonephritis in the modern era: experience with 86 adults and review of the literatura. Medicine, 87, 21–32. Neugarten, J. and Baldwin, D. S. (1984). Glomerulonephritis in bacterial endocarditis. Am J Med, 77, 297–304.

Nori, U. S., Manoharan, A., Thornby, J. I., et al. (2006). Mortality risk factors in chronic haemodialysis patients with infective endocarditis. Nephrol Dial Transplant, 21, 2184–90. Ozcan, F., Sabel, M., Heering, P., et al. (2001). Glomerulonephritis secondary to chronic infection of a ventriculoatrial shunt. Deutsche Medizinische Wochenschrift, 126, 1229–32. Phillips, J., Palmer, A. and Baliga, R. (2005). Glomerulonephritis associated with acute pneumococcal pneumonia: a case report. Pediatr Nephrol, 20, 1494–5. Que, Y. A. and Moreillon P. (2011). Infective endocarditis. Nat Rev Cardiol, 8, 322–36. Rehman, T., Chohan, M. and Yonas, H. (2011). Diagnosis of ventriculoperitoneal shunt infection using [F-18]-FDG PET: a case report. J Neurosurg Sci, 55, 161–3. Rekik, S., Trabelsi, I., Hentati M., et al. (2009). Infective endocarditis in hemodialysis patients: clinical features, echocardiographic data and outcome: a 10-year descriptive analysis. Clin Exp Nephrol, 13, 350–4. Satake, K., Ohsawa, I., Kobayashi, N., et al. (2011). Three cases of PR3-ANCA positive subacute endocarditis caused by attenuated bacteria (Propionibacterium, Gemella, and Bartonella) complicated with kidney injury. Modern Rheumatol, 21, 536–41. Tao, J. L., Ma, J., Ge, G. L., et al. (2010). Diagnosis and treatment of infective endocarditis in chronic hemodialysis patients. Chin Med Sci J, 25,135–9. Ter Borg, E. J., Van Rijswijk, M. H. and Kallenberg, C. G. (1991). Transient arthritis with positive tests for rheumatoid factor as presenting sign of shunt nephritis. Ann Rheum Dis, 50, 182–3. Tleyjeh, I. M., Abdel-Latif, A., Rahbi, H., et al. (2007). A systematic review of population-based studies of infective endocarditis. Chest, 132, 1025–35. Tleyjeh, I. M., Steckelberg, J. M., Murad, H. S., et al. (2005). Temporal trends in infective endocarditis: a population-based study in Olmsted County, Minnesota. JAMA, 293, 3022–8. Vella, J., Carmody, M., Campbell, E., et al. (1995). Glomerulonephritis after ventriculo-atrial shunt. QJM, 88, 911–18. Veloso, T. R., Amiguet, M., Rousson, V., et al. (2011). Induction of experimental endocarditis by continous low-grade bacteremia mimicking spontaneous bacteremia in humans. Infect Immun, 79, 2006–11. Whitworth, J. A., Morel-Maroger, L., Mignon, F., et al. (1976). The significance of extracapillary proliferation. Clinicopathological review of 60 patients. Nephron, 16, 1–19. Yum, M., Wheat, L. J., Maxwell, D., et al. (1978). Immunofluorescent localization of Staphylococcus aureus antigen in acute bacterial endocarditis nephritis. Am J Clin Pathol, 70, 832–5.

CHAPTER 80

Membranoproliferative glomerulonephritis and C3 glomerulopathy Daniel P. Gale and Terry Cook Historical perspective and nomenclature The term membranoproliferative glomerulonephritis (MPGN) was first introduced in 1961 (Habib et al., 1961) and originally cases were subdivided purely on the basis of light microscopic appearances (Fig. 80.1) into the categories classical, lobular, crescentic, and focal. With the introduction of electron microscopy and immunostaining during the following decade and the observation that electron-dense deposits were an almost invariable ultrastructural feature, MPGN was subclassified into three types based on the location of deposits. In type 1 there are deposits in the mesangium and between the endothelium and glomerular basement membrane (GBM)—subendothelial deposits (Fig. 80.2). Type 2 is defined by dense transformation of the lamina densa of the GBM and is alternatively known as dense deposit disease (DDD) (Fig. 80.3). Type 3 MPGN is a less well-defined condition with subendothelial, subepithelial, and variable intramembranous deposits. It was also recognized that, while types 1 and 3 MPGN were associated with the presence of glomerular immunoglobulin, complement C1q and complement C3 deposition, immunostaining in type 2 MPGN (DDD) was usually positive for C3 only, implicating a primary defect of complement alternative pathway regulation (see Fig. 80.4). (A)

Current nomenclature Although the subdivision of MPGN described above was clinically useful in identifying patients in whom a defect of complement regulation was likely, several factors have led to the current classification of MPGN into immune-complex MPGN and those cases that are examples of C3 glomerulopathy, according to the likely underlying pathophysiology (Fig. 80.5 and see algorithm in Fig 18.11). ◆ Recognition

of the importance of diseases, such as hepatitis C, other viral infections, paraproteinaemia, and cryoglobulinaemia in causing glomerulonephritis. This reduced the number of patients for whom ‘type 1 MPGN’ was a diagnosis (rather than simply the kidney biopsy appearance).

◆ Appreciation

that DDD was more commonly associated with light microscopic appearances that do not fulfil the criteria for MPGN (Walker et al., 2007).

◆ Realization

that type 1 or type 3 MPGN can also occur without significant deposition of immunoglobulins, and that in these cases an underlying systemic defect of complement regulation can frequently be identified.

(B)

Fig. 80.1  Light microscopic appearances of MPGN. (A) H&E stained section. The glomeruli show enhanced lobulation, mesangial expansion and mesangial hypercellularity, and capillary wall thickening (B) Methenamine sliver-stained section. In the silver-stained section the double-contour appearance of the capillary walls is clearly seen. Courtesy of Dr P. Walker.

642

Section 3  

the patient with glomerular disease

These and other considerations led to the introduction of the term ‘C3 glomerulopathy’ which encompasses the disorders in which complement C3 accumulates in the glomerulus in the absence of significant immunoglobulin deposition. This is the hallmark of complement alternative pathway dysregulation and represents underlying aetiology, pathophysiology, and prognosis distinct from those associated with cases of proliferative glomerulonephritis (GN) in which immunoglobulins are present, whatever the morphology may be by light or electron microscopy. With the recent development of therapeutic agents that directly modulate complement activity it is more clinically relevant to define these disorders by the underlying pathophysiology rather than the morphological appearances of the kidney biopsy. Light microscopic appearances resembling MPGN are sometimes seen in patients with chronic thrombotic microangiopathies. In this context, rather than immunoglobulin and complement deposition with electron-dense deposits visible on electron microscopy, immunostains are typically negative and there is accumulation of electron-lucent, flocculent material (thought to be fibrin and its metabolites) beneath the endothelial cells. Thrombotic microangiopathies are discussed in Chapter  174. This differential diagnosis is also considered in Chapter 18, see Fig 18.11.

Epidemiology The incidence of MPGN varies across different regions, with higher rates documented in developing countries—an observation usually attributed to the relative prevalence of infectious disease (see below). In the United Kingdom, MPGN is more common in children and young adults and is thought to account for approximately 2% of kidney biopsies, although it is reported in up to 10% of biopsies performed to investigate nephrotic syndrome. Longitudinal evidence suggests that the rates of MPGN are declining (Covic et al., 2006; Hanko et al., 2009; Woo et al., 2010), possibly as a result of reductions in chronic infectious disease and perhaps also due to better treatment and diagnosis of autoimmune conditions.

Fig. 80.2  Electron micrograph of a case of immune complex-associated MPGN type 1. There are many subendothelial and mesangial deposits and rare small subepithelial deposits. There is reduplication of the glomerular basement membrane and mesangial cell interposition.

Immune complex membranoproliferative glomerulonephritis MPGN associated with immunoglobulin deposition can occur in a very wide range of diseases—usually those in which there is increased or abnormal immunoglobulin production. In up to one-third of cases immune complexes are detectable in the circulation, but many people with such immune complexes do not develop glomerulonephritis. The aetiologies of immune complex MPGN can be divided into infectious, autoimmune, or neoplastic. In these conditions, aberrant production of immunoglobulins can lead to accumulation in the glomerulus with consequent complement activation, inflammation and damage to the kidney. The fact that most people with immunological activation do not develop renal disease suggests that other factors (perhaps relating to characteristics of the antigen or immunoglobulins, or to variation in complement regulators or local glomerular architecture) determine which patients will develop renal disease.

Infection-related membranoproliferative glomerulonephritis Chronic infection was probably the most common cause of MPGN historically, and still accounts for a significant proportion of cases—especially in regions where infectious disease is common. Bacterial infection-related GN is reviewed in Chapter  76. Generation of antibody–antigen complexes leads to deposition in the glomerulus, local complement activation, and MPGN. Although overall the risk of MPGN with infection is low, some infections, such as endocarditis and viral hepatitis (with or without associated cryoglobulinaemia) seem more likely to result in this renal lesion. Schistosoma mansoni infection (see Chapter 182) is a common cause of MPGN in endemic countries, and it has been

Fig. 80.3  Electron micrograph of the glomerulus from a case of dense deposit disease showing prominent osmiophilic transformation of the glomerular basement membrane. Courtesy of Dr C. Nast.

Chapter 80 

mpgn and c3 glomerulopathy

Host cells CFHRs Circulating regulators

Cell surface regulators

Complement factor H (CFH) Complement factor I

Membrane cofactor protein (CD46) Decay accelerating factor

+ + Class ic path al way

Antibody activation

Factor Bb C3bBb

C1 C4, C2 C3b

MBLs

Alternative pathway

MBL ay w path

Membrane attack complex

C5–C9 Terminal pathway C5–C9 deposited on attacked surfaces

Fig. 80.4  The complement system. Complement is a cascade of circulating proteins (designated C1 to C9) that forms an important component of the innate immune system. It can be activated via the ‘classical pathway’ in which antibodies bind to antigen, and (via the recruitment of the C1 complex and other proteins) lead to the formation of a C3 convertase that catalyses the cleavage of C3 to generate C3b. This C3 convertase can also be generated by mannose binding lectins (MBLs) that recognize mannose residues present on the surface of bacteria (the ‘MBL pathway’) and there is also spontaneous ‘tickover’ activation of C3 in the circulation by hydrolysis. C3b binds to activated factor B to form C3bBb, a non-covalently bound alternative C3 convertase enzyme that catalyses the cleavage of C3 to generate more C3b in a positive feedback amplification loop known as the ‘alternative pathway’. The C3bBb convertase can freely dissociate but is stabilized by a number of factors, including biological surfaces. C3b also forms part of a complex that catalyses the cleavage of C5 to activate the ‘terminal pathway’ that leads to lysis of a targeted cell. In order to prevent runaway activation and damage to host tissues, the alternative pathway is regulated by a number of cell-surface and circulating regulators, including complement factor I (CFI), membrane cofactor protein (CD46), and complement factor H (CFH). Regulation by CFH is modulated by some or all of the CFH-related proteins CFHRs 1–5. Genetic defects of these regulators can lead to diseases to which the kidney is particularly susceptible.

postulated that hepatosplenic involvement in the disease results in reduced clearance of immune complexes by Kupffer cells, leading to increased systemic (and therefore renal) exposure (Barsoum et al., 1988). Presentation of infection-associated MPGN is usually with haematuria, proteinuria, and renal dysfunction which may be progressive if the underlying cause does not resolve. Classic post-streptococcal GN (see Chapter  77) is sometimes seen several days after resolution of an acute bacterial infection. Presentation is typically with haematuria, proteinuria (sometimes with nephrotic syndrome), and renal impairment. It is usually self-limiting, although supportive renal replacement therapy is occasionally needed. There are frequently detectable circulating antibodies against streptolysin O and DNase B, and hypocomplementaemia is common (Blyth et al., 2007). Because some antibodies and bacterial antigens are able to stabilize the alternative pathway C3 convertase, low C3 is not always accompanied by low C4. Kidney biopsy in post-infectious GN typically shows diffuse mesangial proliferation, neutrophil infiltration and large, hump-like subepithelial deposits on electron microscopy. Immunostains of the glomerulus are usually positive for IgG, IgM, C1q, and C3, although C3-predominant staining in this context is well recognized (Sethi et al., 2013). Distinguishing post-infectious GN from C3 glomerulopathy (see below) on a single biopsy is therefore not always possible, and lack of the typical clinical course and resolution should raise the suspicion of an underlying disorder of complement regulation.

Cryoglobulinaemic membranoproliferative glomerulonephritis Cryoglobulins are immunoglobulins that reversibly precipitate at 4°C. Cryoglobulinaemia can arise as a result of an aberrant plasma

cell clone (types 1 and 2 cryoglobulinaemia) or due to polyclonal production (type 3)  and these are discussed in more detail in Chapter  151. Tests for rheumatoid factor are often positive and hypocomplementaemia is common in this condition. The association of cryoglobulins with infection is well recognized and they are detectable in 15–20% of people infected with HIV and up to 50% of those infected with hepatitis C.  Cryoglobulinaemia may also occur in autoimmune diseases, most commonly Sjögren syndrome. Clinically significant disease resulting from cryoglobulinaemia is only seen in a small proportion of patients with detectable circulating cryoglobulins. Deposition of cryoglobulins may occur anywhere in the body, leading to local activation of complement and thrombosis, and leading to Meltzer’s clinical triad of palpable purpura, joint pains, and muscle weakness. Renal involvement usually manifests as proteinuria (with nephrotic syndrome in approximately one-fifth of patients), haematuria, and renal impairment. Around a third of patients have concurrent extrarenal disease at presentation, although renal involvement can be the presenting feature of hepatitis C infection, with over half of patients with this underlying cause having normal liver function tests at presentation. Kidney biopsy in cryoglobulinaemic GN frequently shows MPGN, sometimes with very prominent hypercellularity due to macrophage infiltration into the glomeruli. There may also be florid accumulation of eosinophilic material in intraluminal thrombi, likely representing cryoprecipitate in the glomerular capillaries (Fig. 80.6). Small or medium vessel vasculitis is seen in approximately one-third of patients. Progression of cryoglobulinaemic GN to end-stage renal disease is seen in approximately 10% patients, usually after at 10 years or

643

644

Section 3  

the patient with glomerular disease

Autoimmune disease

MPGN Immune complex GN

immunogloblin and C3 deposition

C3 glomerulopathies C3 but minimal or no immunogloblin deposited

DDD

(formerly MPGN type 2)

C3GN

(Includes cases of MPGN types 1 and 3)

Fig. 80.5  Classification of proliferative glomerulonephritis. C3GN = C3 glomerulonephritis; DDD = dense deposit disease. (Compare with Fig. 18.11D)

more, and treatment is usually aimed at the underlying cause (e.g. clearing hepatitis C infection (see Chapter 186) or suppressing a plasma cell clone). Plasma exchange is sometimes used to reduce the cryoglobulin load (or ‘cryocrit’), especially where there is hyperviscosity syndrome, but although good outcomes have been reported in case series, clinical trial data is lacking (Dammacco and Sansonno, 2013).

Monoclonal immunoglobulin G deposition and MPGN Proliferative GN has also been described in association with monoclonal immunoglobulin G (IgG) deposition in the absence of detectable cryoglobulinaemia. This is sometimes associated with hypocomplementaemia and/or a monoclonal band detectable in the serum or urine. Published series indicate there is a low risk of development of subsequent haematological malignancy, at least within 2–3 years of diagnosis. However, it is also recognized that monoclonal gammopathy associated with established lymphocytic leukaemia can also result in MPGN with monoclonal immunoglobulin deposition, implying that thorough haematological assessment of patients presenting with monoclonal immunoglobulin deposition in the glomerulus is warranted. Follow-up data in proliferative GN with monoclonal IgG deposition is limited, but in the largest published series > 20% of patients progressed to end-stage renal disease over 30  months of follow-up (Nasr et  al., 2009). Unsurprisingly, disease recurrence following transplantation is also described in this condition.

The presence of circulating antibodies against autoantigens can result in immune complexes that can lead to glomerular disease. Most commonly, this results in proliferation confined to either mesangial regions or the glomerular capillaries. In this context, positive immunostaining for IgA (but not IgG or IgM) is diagnostic of underlying IgA nephropathy, and immunostaining for all immunoglobulins is suggestive of lupus nephritis (although a diagnosis of systemic lupus erythematosus properly relies on additional clinical and/or serological features). These diseases are discussed in Chapter  162. Occasionally, proliferative GN with staining for IgM but not IgG or IgA is seen, and this is sometimes termed ‘IgM nephropathy’. Whether the IgM is relevant to disease pathogenesis in this context is not known. Where an underlying autoimmune, infectious, or lymphoproliferative disorder is not identified these cases are now regarded as ‘immune complex MPGN of unknown cause’ rather than the former diagnostic categories of ‘primary’ or ‘idiopathic’ MPGN type 1 or type 3. Interestingly, an underlying disorder of complement regulation can frequently be identified in such cases (Servais et al., 2012) (see Fig. 80.2). This suggests that regarding the histomorphological and immunohistochemical findings as the ‘gold standard’ in the diagnosis in MPGN may not always be warranted and that a clinically useful diagnosis is most likely to come from a complete clinical, genetic, serological and pathological assessment of each patient.

Treatment of immune-complex associated MPGN Where an underlying systemic disease (such as bacterial infection, hepatitis, or autoimmune condition) is identified its successful treatment is likely to ameliorate the renal disease. In addition, aggressive treatment of high blood pressure and proteinuria with angiotensin blockade is widely accepted to be beneficial. Where a cause is not identified there is very little trial data to guide treatment for MPGN. Outcomes in immune-complex MPGN seem to relate most closely to the degree of inflammation seen on the initial biopsy (Little et al., 2006)  and therapy aimed at suppressing the immune system is sometimes used. Alternate day corticosteroids has been used with some evidence of efficacy in children with MPGN, and antiproliferative agents have shown efficacy in small observational studies, but controlled trials with long-term follow-up have not been published. Kidney Disease: Improving Global Outcomes (KDIGO) guidelines suggest, in the presence of frank nephrotic syndrome and declining function, that a trial of 6 months of corticosteroids and either cyclophosphamide or mycophenolate mofetil is reasonable to treat MPGN of unknown cause.

C3 glomerulopathy

Fig. 80.6  Glomerulus in a case of cryoglobulinaemia showing prominent intraluminal thrombi (periodic acid–Schiff stain).

Proliferative GN sometimes occurs with deposition of C3 but without significant quantities of immunoglobulins or C1q in the glomerulus. This pattern of immunostaining is characteristic of DDD (formerly MPGN type 2)  (Fig. 80.7) where there is dense transformation of the basement membrane, but it is also seen in some cases where there are discrete electron-dense deposits in distributions that would formerly have been characterized as MPGN type 1 (subendothelial and mesangial) or type 3 (subepithelial as well as subendothelial and mesangial). These appearances are now termed

Chapter 80 

mpgn and c3 glomerulopathy

pathway. Examples include homozygous mutations of the complement regulator CFH gene (Licht et al., 2006) and heterozygosity for a particular activating mutation of C3 (Martinez-Barricarte et al., 2010). Of note, fluid phase dysregulation of complement associated with DDD contrasts to mutations of CFH and other complement regulators that lead to complement dysregulation particularly at endothelial surfaces—these are most often described in association with atypical haemolytic uraemic syndrome (see Chapter 174). In addition to these individual monogenic disorders, certain common allelic variants of CFH and its homologue complement factor H-related protein 5 (CFHR5) are more common in patients with DDD than in controls (Abrera-Abeleda et al., 2006). While these findings do not have direct implications for diagnostic or therapeutic decisions concerning individual patients, they do highlight the importance of the fine control of complement regulation in the disease. Fig. 80.7  Immunofluorescence for C3 in a case of dense deposit disease. There is staining along glomerular capillary walls and also granular mesangial staining.

C3 glomerulonephritis (C3GN). Together, DDD and C3GN are referred to as the C3 glomerulopathies and in the majority of cases a genetic or acquired defect of complement alternative pathway can be identified (Servais et al., 2012). Although an MPGN pattern is often seen, in some cases there may be a mesangial proliferative pattern or even normal glomeruli. Whatever the basic pattern of glomerular injury, there may also be variable segmental or global endocapillary hypercellularity and/or crescent formation.

Dense deposit disease DDD is rare, occurring at a frequency of approximately 2–4 per million of the population. It is usually diagnosed in children (where it accounts for around 15–20% of MPGN) or, less commonly, young adults. Clinical presentation is usually with proteinuria which may be accompanied by the nephrotic syndrome and/or haematuria. The diagnosis is made by observing the characteristic highly osmiophilic dense transformation of the lamina densa of the GBM. Presentation is often in the aftermath of an infectious disease and there is usually progressive renal impairment, with 50% of patients requiring renal replacement therapy within 10 years of diagnosis. Prognosis is worse with greater age or greater degree of renal impairment at diagnosis. Recurrence following renal transplantation is common, with 50% grafts lost to recurrent disease at 5 years (Braun et al., 2005).

Aetiology of DDD DDD is usually associated with systemic evidence of activation of the complement alternative pathway. In around 70% of patients, an autoantibody that recognizes the alternative pathway C3 convertase (termed a C3 nephritic factor, or C3NeF) is detectable. C3NeFs stabilize the convertase complex (C3bBb), leading to runaway activation of the alternative pathway in the circulation, usually resulting in very low C3 levels. C4 levels are typically unaffected as the classical pathway is not activated. In individuals in whom a C3NeF is not detected, antibodies that bind to both C3b and factor B, or antibodies that recognize complement factor H (CFH) have been reported (Chen et al., 2011). Familial occurrence of DDD is exceedingly rare, but has been described in association with defects that result in fluid phase (as opposed to cell-surface) dysregulation of the complement alternative

Systemic features associated with DDD Retinal drusen, which result from the accumulation of complement components in Bruch’s membrane of the eye are also visible by fundoscopy in DDD. Visual loss may occur, as a consequence of retinal atrophy or occasional subretinal neovascular membrane formation, usually over several decades. Ophthalmological assessment is therefore recommended in patients with DDD. Acquired partial lipodystrophy (loss of fat in the upper half of the body) is also associated with the presence of a C3NeF and is thought to result from complement-mediated damage to adipocytes caused by serum containing a C3NeF (Mathieson et al., 1993). Partial lipodystrophy can precede the renal disease and can sometimes be triggered following a minor infection.

C3 glomerulonephritis Glomerular inflammation with C3 deposition, without significant immunoglobulin deposition, and in the absence of dense transformation of the basement membrane is termed C3 glomerulonephritis (C3GN). C3GN is rare, probably accounting for < 10% of proliferative GN and most commonly occurs in young adulthood, although can occur at any age. The term C3GN has only recently gained widespread acceptance, and it is believed that most cases would previously have been regarded as variants of MPGN types 1 or 3 (Fig. 80.8). Presentation is with one or more of haematuria, proteinuria, or nephrotic syndrome and varies according to underlying aetiology. Although traditionally regarded as a benign disease, recent evidence suggests that outcomes are similar to those in DDD, with progressive renal impairment, usually over several years. Recurrence following transplantation is common in the disease, presumably because of the systemic nature of the underlying defect(s) of complement regulation.

Aetiology of C3GN C3GN is usually sporadic but in published series potential defects of complement regulation are often present. A C3NeF is seen in approximately 40% of cases and mutations of the complement regulators CFH, complement factor I, and membrane cofactor protein (MCP) have also been described in sporadic cases (Servais et al., 2012). Interestingly, these mutations overlap with mutations also observed in patients with atypical haemolytic uraemic syndrome (see Chapter 174), implicating complement dysregulation particularly at surfaces in C3GN. These defects of complement regulation

645

Section 3  

the patient with glomerular disease 100 Percent renal survival

646

Female Male

80

N = 91

60 40 20 0

0

20

40

60

80

100

Age at ESRD

Fig. 80.10  Outcomes in CFHR5 nephropathy. From Athanasiou et al. (2011).

Fig. 80.8  Electron micrograph from a case of C3 glomerulonephritis. There is electron-dense material in the mesangium and the glomerular basement membrane is thickened with intramembranous deposits of variable density. This appearance has previously been referred to as MPGN type 3.

are summarized and compared with those identified in immune complex MPGN and DDD in Fig. 80.9. In addition, an association exists between C3GN and paraproteinaemia (Zand et al., 2013), and it is important that this possibility is considered, especially in older patients. Familial forms of C3GN have been described in association with mutations and rearrangements of the CFHR genes, the most common of which, CFHR5 nephropathy, is endemic in Cypriots (Gale et al., 2010).

CFHR5 nephropathy CFHR5 nephropathy is a highly penetrant monogenic disease with autosomal dominant inheritance. It is caused by heterozygous mutation of CFHR5 leading to production of an elongated version of the protein due to duplication of the N-terminal 2 domains, denoted CFHR512123–9. The disease is endemic in Cyprus with an allele frequency of 1:6000 in this population. Kidney biopsy in the disease invariably shows C3GN and CFHR5 nephropathy is by far the commonest single known cause of this renal lesion. 70%

MPGN with immune complexes DDD C3GN

60% 50% 40% 30% 20% 10% 0%

The clinical features of CFHR5 nephropathy are distinct from those of C3GN due to other causes and are highly consistent: there is microscopic haematuria (present in > 90% of mutation carriers) and episodes of macroscopic haematuria occur at times of upper respiratory tract or other infections. There may also be acute kidney injury during these episodes, and the clinical resemblance to IgA nephropathy is striking. Proteinuria is usually only seen after the development of established chronic renal impairment and is usually low grade (i.e. < 1 g/day). Nephrotic syndrome has not been reported in this disease and extrarenal manifestations are not described. The association between disease flare and intercurrent infection, also seen in DDD, attests to the importance of environmental immunological exposures in modulating complement alternative pathway activity. While microscopic haematuria and C3GN appear to be present in both sexes, episodes of macroscopic haematuria and renal dysfunction are markedly less common in females than males. Over 80% men with the disease have established renal dysfunction by the age of 50, whereas this is observed in < 20% of women (Athanasiou et al., 2011). A corollary of this is that end-stage renal disease is very much more likely in men with the disease, and this was borne out in a large cohort study where all the males but fewer than half the females had either died or required renal replacement therapy by the age of 80 (see Fig. 80.10). The disease recurs following transplantation, but overall graft survival is not demonstrably affected, presumably because of the slow pace of progression. Other mutations resulting in rearrangements of the CFHR genes and also leading to production of elongated versions of the proteins (e.g. a CFHR3-1 hybrid protein) have been described in association with autosomal dominant inheritance of C3GN in single families (Malik et al., 2012; Tortajada et al., 2013), strongly implicating the CFHR proteins in the modulation of complement activity in humans.

Investigation of C3 glomerulopathies C3NEF

CFH deficiency mutation

CFH missense mutation

CFI mutation

MCP mutation

Fig. 80.9  Complement abnormalities in sporadic cases of MPGN and C3 glomerulopathy. Data from Servais et al. (2012).

A kidney biopsy showing a C3 glomerulopathy (i.e. DDD or C3GN) should prompt investigation of alternative pathway regulation. Tests for paraproteinaemia, complement C3, and C4 are widely available and C3NeF measurement is available in regional or national reference centres. These tests should therefore be performed in all patients. In patients with C3 glomerulopathy who may have Cypriot

Chapter 80 

ancestry, a genetic test for the CFHR512123–9 mutation should be performed (this test is available at the Institute of Child Health in London, ). Additional tests such as serum CFH and CFI levels and tests for autoantibodies against factor B and CFH should be considered if a C3NeF is not identified. Where these serological tests do not identify a likely cause, additional genetic testing by mutation and copy number variation screening of the genes for complement regulators, including CFH, MCP, CFHR1-5, CFI, CFB, and C3 should be considered, especially if there is a family history of kidney disease. Interpretation of specialized serological and genetic tests is not always straightforward so referral to a centre with experience in this area may be advisable.

Treatment of C3 glomerulopathies It is generally assumed that blood pressure control and angiotensin system blockade should be introduced to delay progression of renal damage in the presence of hypertension or proteinuria. Probably because of the rarity, slowly progressive course, and heterogeneity of aetiology of C3 glomerulopathies, there are currently no controlled trial data to guide the treatment of these conditions. As with immune complex MPGN, a kidney biopsy showing highly cellular or inflammatory appearances predicts a worse prognosis and may prompt the use of immunosuppressive therapy. Where a pathogenic antibody (e.g. a C3NeF) is detected or suspected, therapeutic strategies aimed at depleting antibody production (e.g. with an anti-CD20 monoclonal antibody, corticosteroids, or mycophenolate mofetil) have been employed, with some anecdotal evidence of success. Plasma exchange has been used to treat patients with a C3NeF, CFH deficiency, and CFHR5 nephropathy, but there are no controlled outcome data from these manoeuvres. Eculizumab, a humanized monoclonal antibody directed against complement C5, is effective at blocking the terminal complement

Light microscopic morphology

Immunostaining and electron microscopy

mpgn and c3 glomerulopathy

pathway in humans. Prospective trials showing efficacy in the complement-associated diseases paroxysmal nocturnal haemoglobinuria and atypical haemolytic uraemic syndrome have been published, but reports of its use in DDD and C3GN (in very small numbers of patients) have shown mixed and somewhat disappointing responses (Bomback et al., 2012).

Conclusions Identification of proliferative GN by kidney biopsy should prompt a search as to the underlying cause. The presence of immunoglobulins suggests underlying increased or aberrant antibody production, which can have a variety of causes, including infection, autoimmune disease, and lymphoproliferative disease. The presence of complement C3 without significant immunoglobulin is diagnostic of a C3 glomerulopathy and suggests an underlying disorder of complement regulation. An approach to differential diagnosis based on histomorphology, immunohistochemistry, and electron microscopy is summarized in Fig. 80.11. It has long been evident that immunological activation leading to excessive or aberrant antibody production can lead to MPGN. In this context there is evidence of the immunoglobulins and complement C3 in the kidney. Appreciation that in C3 glomerulopathy activation of the alternative complement pathway alone is also sufficient to cause MPGN in the absence of antibody production or deposition suggests that renal complement activation per se is central to the pathogenesis of these diseases, and this has been supported by experiments in animal models. In this paradigm, MPGN is a consequence of complement activation, resulting from either immunological diseases that lead to the generation of antibody–antigen complexes, or from defects in the regulation of the complement system itself. In clinical practice, determining which process is driving the renal disease is instructive in determining the appropriate therapy: treatments aimed at combating infections or suppressing the adaptive immune system

Proliferative glomerulonephritis (including MPGN)

Glomerular C3 with immunoglobulin deposition

IgG ± IgM ± IgA deposition

IgA only deposition

Glomerular C3 with minimal or no immunoglobulin deposition

Monoclonal light chain restriction

Dense transformation of GBM

Discrete electron dense deposits

Differential diagnosis

Infection

Autoimmune disease

Cryoglobulinaemia

IgA nephropathy

Immune-complex proliferative glomerulonephritis

B cell dyscrasia

Cryoglobulinaemia

DDD

C3GN

CFHR5 nephropathy

C3 glomerulopathy

Fig. 80.11  Approach to the diagnosis of proliferative glomerulonephritis. DDD = dense deposit disease; C3GN, C3 glomerulonephritis. Compare the algorithm in Fig. 18.11 which is based on the appearance of basement membrane splitting (double contour).

647

648

Section 3  

the patient with glomerular disease

(e.g. using cytotoxic or antiproliferative agents) are not necessarily effective in restoring normal regulation of complement alternative pathway activity. Currently available treatments are unsatisfactory but it is hoped that therapies, currently under development, that are able to block or modulate C3 activation may be efficacious in both C3 glomerulopathies and immune complex-driven GN.

Further reading Bomback, A. S. and Appel, G. B. (2012). Pathogenesis of the C3 glomerulopathies and reclassification of MPGN. Nat Rev Nephrol, 8(11), 634–42. Gale, D. P. and Maxwell, P. H. (2013). C3 glomerulonephritis and CFHR5 nephropathy. Nephrol Dial Transplant, 28(2), 282–8. Pickering, M. C., D’Agati, V. D., Nester, C. M., et al. (2013). C3 glomerulopathy: consensus report. Kidney Int, 84(6), 1079–89. Sethi, S., Nester, C. M., and Smith, R. J. (2012). Membranoproliferative glomerulonephritis and C3 glomerulopathy: resolving the confusion. Kidney Int, 81(5), 434–41. Smith, R. J., Harris, C. L., and Pickering, M. C. (2011). Dense deposit disease. Mol Immunol, 48(14), 1604–10.

References Abrera-Abeleda, M. A., Nishimura, C., Smith, J. L., et al. (2006). Variations in the complement regulatory genes factor H (CFH) and factor H related 5 (CFHR5) are associated with membranoproliferative glomerulonephritis type II (dense deposit disease). J Med Genet, 43(7), 582–9. Athanasiou, Y., Voskarides, K., Gale, D. P., et al. (2011). Familial C3 glomerulopathy associated with CFHR5 mutations: clinical characteristics of 91 patients in 16 pedigrees. Clin J Am Soc Nephrol, 6(6), 1436–46. Barsoum, R. S., Sersawy, G., Haddad, S., et al. (1988). Hepatic macrophage function in schistosomal glomerulopathy. Nephrol Dial Transplant, 3(5), 612–16. Blyth, C. C., Robertson, P. W., and Rosenberg, A. R. (2007). Post-streptococcal glomerulonephritis in Sydney: a 16-year retrospective review. J Paediatr Child Health, 43(6), 446–50. Bomback, A. S., Smith, R. J., Barile, G. R., et al. (2012). Eculizumab for dense deposit disease and C3 glomerulonephritis. Clin J Am Soc Nephrol, 7(5), 748–56. Braun, M. C., Stablein, D. M., Hamiwka, L. A., et al. (2005). Recurrence of membranoproliferative glomerulonephritis type II in renal allografts: the North American Pediatric Renal Transplant Cooperative Study experience. J Am Soc Nephrol, 16(7), 2225–33. Chen, Q., Muller, D., Rudolph, B., et al. (2011). Combined C3b and factor B autoantibodies and MPGN type II. N Engl J Med, 365(24), 2340–2. Covic, A., Schiller, A., Volovat, C., et al. (2006). Epidemiology of renal disease in Romania: a 10 year review of two regional renal biopsy databases. Nephrol Dial Transplant, 21(2), 419–24. Dammacco, F. and Sansonno, D. (2013). Therapy for hepatitis C virus-related cryoglobulinemic vasculitis. N Engl J Med, 369(11), 1035–45.

Gale, D. P., de Jorge, E. G., Cook, H. T., et al. (2010). Identification of a mutation in complement factor H-related protein 5 in patients of Cypriot origin with glomerulonephritis. Lancet, 376(9743), 794–801. Habib, R., Michielsen, P., de Montera, E., et al. (1961). Clinical, microscopic and electron microscopic data in the nephrotic syndrome of unknown origin. In G. E. W. Wolstenholme and M. P. Cameron (eds.) Ciba Foundation Symposium—Renal Biopsy: Clinical and Pathological Significance, pp. 70–102. Chichester: John Wiley & Sons Ltd. Hanko, J. B., Mullan, R. N., O’Rourke, D. M., et al. (2009). The changing pattern of adult primary glomerular disease. Nephrol Dial Transplant, 24(10), 3050–4. Licht, C., Heinen, S., Józsi, M., et al. (2006). Deletion of Lys224 in regulatory domain 4 of Factor H reveals a novel pathomechanism for dense deposit disease (MPGN II). Kidney Int, 70(1), 42–50. Little, M. A., Dupont, P., Campbell, E., et al. (2006). Severity of primary MPGN, rather than MPGN type, determines renal survival and post-transplantation recurrence risk. Kidney Int, 69(3), 504–11. Malik, T. H., Lavin, P. J., Goicoechea de Jorge, E., et al. (2012). A hybrid CFHR3-1 gene causes familial C3 glomerulopathy. J Am Soc Nephrol, 23(7), 1155–60. Martinez-Barricarte, R., M. Heurich, Vazquez-Martul, E., et al. (2010). Human C3 mutation reveals a mechanism of dense deposit disease pathogenesis and provides insights into complement activation and regulation. J Clin Invest, 120(10), 3702–12. Mathieson, P. W., Wurzner, R., Oliveria, D. B., et al. (1993). Complement-mediated adipocyte lysis by nephritic factor sera. J Exp Med, 177(6), 1827–31. Nasr, S. H., Satoskar, A., Markowitz, G. S., et al. (2009). Proliferative glomerulonephritis with monoclonal IgG deposits. J Am Soc Nephrol, 20(9), 2055–64. Servais, A., L. Noel, H., Roumenina, L. T., et al. (2012). Acquired and genetic complement abnormalities play a critical role in dense deposit disease and other C3 glomerulopathies. Kidney Int, 82(4), 454–64. Sethi, S., Fervenza, F. C., Zhang, Y., et al. (2013). Atypical postinfectious glomerulonephritis is associated with abnormalities in the alternative pathway of complement’ Kidney Int, 83(2), 293–9. Tortajada, A., Yebenes, H., Abarrategui-Garrido, C., et al. (2013). C3 glomerulopathy-associated CFHR1 mutation alters FHR oligomerization and complement regulation. J Clin Invest, 123(6), 2434–46. Walker, P. D., Ferrario, F., Joh, K., et al. (2007). Dense deposit disease is not a membranoproliferative glomerulonephritis. Mod Pathol, 20(6), 605–16. Woo, K. T., Chan, C. M., Mooi, C. Y., et al. (2010). The changing pattern of primary glomerulonephritis in Singapore and other countries over the past 3 decades. Clin Nephrol, 74(5), 372–83. Zand, L., Kattah, A., Fervenza, F. C., et al. (2013). C3 glomerulonephritis associated with monoclonal gammopathy: a case series. Am J Kidney Dis, 62(3), 506–14.

CHAPTER 81

Fibrillary and immunotactoid glomerulopathy Stephen M. Korbet, Melvin M. Schwartz, and Edmund J. Lewis Introduction In 1977, Rosenmann and Eliakim reported an unusual glomerular lesion in a 45-year-old woman presenting with the nephrotic syndrome and renal insufficiency (Rosenmann and Eliakim, 1977). Electron microscopy demonstrated electron-dense deposits with a high degree of organization in the form of fibrils which measured 10 nM in diameter. The deposits were associated with mesangial expansion and immune deposits of immunoglobulin (Ig)-G, IgM, and C3 in a mesangial pattern. Congo-red stain of the deposits was negative and there was no clinical or serologic evidence of a systemic disease. The deposits were interpreted to be ‘amyloid-like’ and it was speculated that they might represent a ‘pre-amyloid’ state. Shortly thereafter, Schwartz and Lewis (1980) reported a case of a 49-year-old man presenting with the nephrotic syndrome, with no evidence of systemic disease, who had a similar renal lesion: immune aggregates were associated with highly organized electron-dense deposits composed of microtubules. During 7 years of follow-up the patient progressed to renal failure but never demonstrated any clinical or serologic evidence of a systemic disease. In order to distinguish this lesion from other disorders with renal lesions having glomerular immune deposits associated with highly organized microtubular or fibrillary structures such as amyloidosis, cryoglobulinaemia, paraproteinaemias, and systemic lupus erythematosus, the term ‘immunotactoid glomerulopathy’ (ITG) was introduced, reflecting the immunoglobulin composition (immuno-) and polymeric morphology (tactoid) of the glomerular deposits. Since these initial reports, > 300 cases of ITG have been reported. Various synonyms have been used to refer to the lesion described in these reports, including fibrillary glomerulonephritis (FGN), non-amyloidotic fibrillary glomerulopathy, amyloid-like glomerulopathy, and amyloid-stain-negative microfibrillary glomerulopathy, but we believe they all represent the same or a similar disease process. The unifying feature in all of the cases is the finding of highly organized ultrastructural deposits that appear to be composed of immunoglobulin and complement and are negative for amyloid by Congo-red stain. Despite the increasing recognition of this lesion, ITG is an uncommon glomerulopathy found in < 1% of renal biopsies (Korbet et  al., 1991; Iskander et  al., 1992; Fogo et  al., 1993; Pronovost et al., 1996; Brady, 1998; Rosenstock et al., 2003; Nasr

et al., 2011). The clinical diagnosis of ITG is applied only after the exclusion of diseases known to be associated with organized glomerular immune deposits including amyloidosis, cryoglobulinaemia, paraproteinaemias, and systemic lupus erythematosus (SLE). Along with ITG, these disorders comprise the family of histopathologic lesions referred to as the ‘fibrillary glomerulopathies’ (Table 81.1 and Fig. 81.1). The disorders included in this classification are defined histochemically. In this schema, ITG represents one of the non-amyloid, immunoglobulin-mediated fibrillary glomerulopathies of which there is a differential diagnosis with diseases which must be excluded before the diagnosis of ITG is made. Many of the diseases associated with the fibrillary glomerulopathies have specific therapies and prognoses which differ significantly from that of ITG. As a result, it is critical that the clinician use a combined histologic, clinical and serologic approach in reaching the correct diagnosis (Fig. 81.1 and Table 81.2).

Congo red Positive

Negative

Amyloid

Non-amyloid Immunofluorescence Negative

Positive Immunoglobulin derived

Non-immunoglobulin derived —Diabetic fibrillosis —Fibronectin nephropathy —Other

Cryoglobulinaemia

Monoclonal gammopathy

—Mixed essential —Multiple myeloma —CLL

—‘Benign’ —Multiple myeloma —MIDD —CLL

Systemic lupus

Immunotactoid glomerulopathy

Fig. 81.1  Algorithm for the evaluation of a patient with a fibrillary glomerulopathy. CLL = chronic lymphocytic leukaemia; MIDD = monoclonal deposition disease. Reproduced with permission from Korbet et al. (1994).

650

Section 3  

the patient with glomerular disease

Table 81.1  Classification of the fibrillary glomerulopathies Amyloid (Congo-red positive) AL amyloid: Primary Multiple myeloma AA amyloid: Rheumatic diseases Chronic suppurative and granulomatous inflammation Tumours Familial Mediterranean fever Non-amyloid (Congo-red negative) Immunoglobulin-derived fibrils: Cryoglobulinaemias: Mixed essential Multiple myeloma Chronic lymphocytic leukaemia Monoclonal gammopathies: ’Benign’ Multiple myeloma Monoclonal immunoglobulin deposition disease Chronic lymphocytic leukaemia Systemic lupus erythematosus Immunotactoid (fibrillary) glomerulopathy Non-immunoglobulin-derived fibrils: Diabetes mellitus (diabetic fibrillosis) Fibronectin nephropathy Others

Clinical and laboratory features Proteinuria is the presenting feature in patients with ITG and on clinical grounds there is nothing unique about the presentation or course of ITG that would allow one to distinguish this disorder from other primary glomerulopathies (Table 81.3). Patients with ITG range in age from 10 to 80  years but on average have been 44–57  years old (Korbet et  al., 1991, 1994; Iskander et  al., 1992; Pronovost et al., 1996; Brady, 1998; Rosenstock et al., 2003; Nasr et  al., 2011). In excess of 90% of the patients are white, and the distribution between men and women is approximately equal. The level of proteinuria at presentation ranges from 0.4–25 g/day, with > 60% having the nephrotic syndrome. Hypertension and microscopic haematuria are common and present in > 65% and > 50% of patients, respectively, and > 45% of patients have some degree of renal insufficiency at the time of diagnosis, indicating the chronic and progressive nature of ITG. In two large reviews of ITG (Pronovost et al., 1996; Rosenstock et al., 2003) there was no difference in the prevalence of hypertension, haematuria, nephrotic syndrome, or renal insufficiency at presentation when the diagnosis

of ITG was subdivided based on differences in fibril size (> 30 nM vs ≤ 30 nM) or arrangement (random vs parallel bundles). The serologic evaluation for cryoglobulins and paraproteins (by immunoelectrophoresis or immunofixation of serum and urine) is, by our definition, negative in ITG and the serum complement levels are generally normal. However, up to 19% of patients have a low-titre of antinuclear antibodies, often in a speckled pattern (Korbet et al., 1985; Iskander et al., 1992; Pronovost et al., 1996; Rosenstock et  al., 2003; Nasr et  al., 2011). Nonetheless, patients with ITG do not have clinical SLE and in general, have no evidence of a systemic disease process. ITG appears to represent a primary glomerulopathy as extrarenal manifestations associated with organized immunoglobulin deposits has been described in only four cases (Ozawa et al., 1991; Masson et  al., 1992; Wallner et  al., 1996; Sabatine et  al., 2002). Liver involvement was described in two cases (Ozawa et al., 1991; Wallner et al., 1996) with lung (Masson et al., 1992) and cardiac involvement (Sabatine et  al., 2002)  found in the remaining two cases. Unlike amyloidosis and other forms of monoclonal immunoglobulin deposition diseases, deposits have not been demonstrated in clinically uninvolved organs studied at autopsy (Korbet et al., 1985; Satoskar et al., 2008). While the overall prevalence of a lymphoproliferative malignancy in patients with ITG is low (≤ 3%), it has been suggested that lymphoproliferative disorders and dysproteinaemias are more frequently seen in patients whose deposits are comprised of larger (> 30 nM) microtubules (Alpers, 1992, 1993; Fogo et al., 1993; Rosenstock et al., 2003). A study by Pronovost et al. (1996) found that this observation is usually a result of the inclusion criteria used and when patients with paraproteinaemia are excluded, there is no difference in the prevalence of lymphoproliferative disease based on differences in fibril diameter (≤ 3% of patients). Furthermore, patients with ITG rarely go on to develop clinical or serologic evidence of a systemic disease or a dysproteinaemia (Korbet et al., 1991). Thus, the pathogenetic process in ITG primarily involves the glomeruli, which distinguishes the lesion from the other immunoglobulin-derived fibrillary glomerulopathies (Korbet et al., 1994).

Pathology The primary pathology of ITG is almost exclusively confined to the glomeruli, reflecting the location of the microfibrils in the mesangium and the glomerular capillary walls (Korbet et al., 1991; Schwartz et al., 2002). By light microscopy, mesangial expansion by periodic acid–Schiff (PAS)-positive material with only a mild mesangial hypercellularity is almost always observed (Fig. 81.2). Glomerular capillary wall pathology is also common and may be focal or diffuse, and consists of thickening and complex staining patterns seen with methenamine silver-PAS (Jones) stain, including reticular patterns, spikes, and double contours (Fig. 81.3). In a study of 66 patients, Nasr et al. (2011) found that 71% had a mesangial proliferative, 15% a membranoproliferative, and 2% a membranous pattern of glomerulonephritis by light microscopy. Proliferative glomerulonephritis with cellular and fibrocellular crescents and segmental necrotizing lesions have been described in a few patients (Duffy et al., 1983; Alpers et al., 1987; Iskander et al., 1992; Brady 1998; Rosenstock et al., 2003; Nasr et al., 2011). However, we have not seen these lesions in ITG when systemic diseases such as cryoglobulinaemia, paraproteinaemia, and SLE have been excluded.

Chapter 81 

fibrillary and immunotactoid glomerulopathy

Table 81.2  Diagnostic features of fibrillary glomerulopathies Primary amyloid

Mixed essential cryoglobulinaemia

Light chain deposition disease

Systemic lupus

Immunotactoid glomerulopathy

Systemic symptoms

75%

75%

75%

100%

0%

Cryoglobulins

0%

100%

0%

50%

0%

Paraproteins

100%

75%

100%

0%

0%

ANA

0%

0%

0%

100%

20%

C3/C4

Normal

Low

Normal

Low

Normal

8–10

6–62

10–15

8–25

10–49

Microfibril diameter (nM)

Most importantly, the glomeruli, tubulointerstitium, and vessels are negative for amyloid by Congo red and thioflavin-T stains. Evaluation of extraglomerular structures demonstrates no specific vascular or tubulointerstitial lesions in ITG. The principal findings by fluorescence microscopy are the presence of immunoglobulins and complement in a pattern that precisely reflects the glomerular mesangial and capillary wall pathology seen by light microscopy (Fig. 81.4) and the distribution of the fibrils by electron microscopy (Korbet et al., 1991; Schwartz et al., 2002; Rosenstock et al., 2003; Nasr et al., 2011). The capillary wall deposits are either diffuse and coarsely granular or discontinuous and pseudo-linear. Tubular basement membrane deposits have only rarely been described, but interstitial and vascular deposits, as determined by fluorescence microscopy, have not been observed. The immunoglobulin class is IgG in > 90% of cases, and the deposits usually contains both κ and λ light chains (Table 81.4). Despite the absence of a paraproteinaemia, monoclonal immunoglobulin deposits have been seen in approximately 20% of ITG cases studied with light chain antisera, and κ light chain restriction was present in the majority cases, usually in combination with an IgG heavy chain (Korbet et al., 1991; Nasr et al., 2011; Schwartz et al., 2002). Evaluations of IgG subgroups have demonstrated deposits comprised of both IgG1 and IgG4 but IgG2 and IgG3 were absent (Iskander et al., 1992; Rosenstock et al., 2003). Monoclonal IgG3κ was reported in one case, which was Table 81.3  Presenting clinical features Korbet et al. (1991, 1994) N

Pronovost et al. (1996)

Rosenstock et al. (2003)

not associated with a paraproteinaemia, with 35 nM microtubular deposits (Schwartz and Lewis, 1980). The ultrastructural appearance of ITG is characterized by the glomerular deposition of extracellular elongated, non-branching microfibrils/microtubules which have neither periodicity nor substructure. The microfibrils are seen in the same locations as the immune deposits seen by immunofluorescence microscopy suggesting that they are comprised of immunoglobulin and complement. Thus, the microfibrils are seen in the mesangium, the primary site of deposition and often also seen in the glomerular capillary wall. The amount of tactoidal material present in the glomerular capillary wall seems to correlate with the extent of glomerular damage. Most commonly, they are present within a thickened basal lamina, but they also are present beneath the epithelial cell where they form large deposits that alternate with projections of basement membrane (spikes). Occasionally, the deposits are seen in the subendothelial space and within the capillary lumen. When fibrils are subepithelial or subendothelial, new layers of basement membrane form over them and incorporate the fibrils into a thickened, irregular capillary wall. Extraglomerular fibrillar deposits have not been described in the interstitium or vasculature but tubular basement membrane involvement has been demonstrated in a few cases of ITG (Duffy et al., 1983; Korbet et al., 1985; Alpers et al., 1987; Korbet et al., 1991; Rosenstock et al., 2003; Nasr et al., 2011). The size of the fibrils varies, but they are distinguished from amyloid by a larger diameter (Table 81.2). In most series the diameters have a mean value of 18–22 nM (Table 81.5 and Fig. 81.5).

Nasr et al. (2011)

62

186

61

66

Age range

10–80

10–81

28–81

19–81

Mean

44 ± 15

57 ± 2

53 ± 12

Male

61%

47%

39%

45%

White

92%

90%

92%

95%

Hypertension

66%

70%

77%

71%

Proteinuria

100%

100%

100%

100%

Nephrotic

61%

72%

52%

55%

Haematuria

78%

71%

60%

52%

Renal insufficiency

47%

53%

69%

66%

Fig. 81.2  Glomerulus with diffuse increase in mesangial matrix and normal capillary walls (PAS, ×100).

651

652

Section 3  

the patient with glomerular disease Table 81.4  Immunofluorescence features

Fig. 81.3  The mesangium is diffusely expanded. The glomerular basement membrane has a complex appearance with diffuse thickening, focal spikes and a reticular pattern (Jones, ×100.)

However, the reported diameters have varied from slightly larger than amyloid (10–12 nM) to as large as 49 nM (Fig. 81.6). Even though there is variability of fibril size among cases, the fibrils in a given case are remarkably consistent in appearance wherever they appear in the glomerulus. The cross-sectional appearance varies from a solid dot to microtubules with either a thin or a thick wall. Examination of the fibrils at high magnification reveals a central core and a wall of varying thickness. Fibrils have a variable length and can appear long and straight or short and curved. They are usually present within a granular, electron-dense matrix suggesting that only part of the deposit is aggregated into fibrils. In most cases the microtubules are randomly arranged on cross-section with various elongated profiles seen in adjacent areas (Fig. 81.5). In other cases, the microtubules appear to be in tightly packed parallel bundles on cross-section, especially with larger fibrils, that have a paracrystalline appearance (Fig. 81.6). It has been suggested that ITG should be separated into two categories based upon arbitrary ultrastructural criteria regarding fibril size and/or organization. The proponents of subdividing ITG suggest the diagnosis of ITG be reserved for cases with larger (> 30 nM), parallel microtubules, and that FGN be applied to cases with smaller (≤ 30 nM), randomly arranged fibrils (Alpers, 1992; Iskander et al., 1992; Fogo et al., 1993). The rationale used for this subdivision is that the different morphological categories have significant clinical

Schwartz et al. (2002)

Rosenstock et al. (2003)

Nasr et al. (2011)

IgG

94%

96%

100%

IgA

29%

30%

28%

IgM

60%

52%

47%

C3

96%

83%

92%

Kappa or lambda only

19%

3%

11%

Kappa and lambda

72%

96%

84%

implications (D’Agati et  al., 1991; Iskander et  al., 1992; Alpers, 1993; Fogo et al., 1993). Presently there is no compelling reason to separately diagnose ITG and FGN on the basis of morphology alone as it has not been demonstrated that the ultrastructural features have significant pathogenetic or clinical implications (Brady, 1998; Pronovost et al., 1996). Thus, we use the diagnosis of ITG to describe patients with both types of deposits, and reserve the term fibrillary glomerulopathy to denote the broader category of diseases (Fig. 81.1) that are characterized morphologically by fibrils seen by electron microscopy without regard to their biochemical composition (Churg and Venkataseshan, 1993; Korbet et al., 1994; Brady, 1998). While there continues to be debate on the issue of classification, what is agreed upon is the importance of distinguishing these patients from amyloidosis and being sure to assess patients for cryoglobulinaemia, a paraproteinaemia, and systemic lupus are these diagnoses carry important therapeutic and prognostic implications (Korbet et al., 2006; Alpers and Kowalewska, 2008).

Pathogenesis and pathophysiology The term immunotactoid was chosen to stress the organized orientation of the deposits and their immunoglobulin composition (Korbet et al., 1985). Using immuno-electron microscopy, it has been shown that the fibrils in patients with ITG contain immunoglobulins (both heavy and light chains) and complement as well as amyloid P component but do not contain other amyloid-associated, basement membrane-associated (type IV collagen and heparan-sulphate proteoglycans) or microfibril-associated (fibronectin and fibrillin) proteins (Casanova et al., 1992; Yang et al., 1992). The presence of

Table 81.5  The range of microfibril diameter in ITG Microfibril diameter (nM)

Fig. 81.4  Mesangial and peripheral capillary wall deposits of IgG. (Fluorescein isothiocynate conjugated rabbit anti-human immunoglobulin G (IgG), ×100.)

% of patients

< 12

13%

13–17

11%

18–22

44%

23–27

12%

28–32

14%

> 32 Data from Korbet et al. (1994).

6%

Chapter 81 

Fig. 81.5  Electron micrograph of glomerular fibrillar deposits showing a random arrangement and measuring 20 nM in diameter. (Uranyl acetate and lead citrate, ×32,000.)

amyloid P component raises the possibility that fibrillogenesis in ITG may be analogous to amyloidosis but without resulting in the critical β-pleated sheet formation. In the usual physiologic environment, intact normal immunoglobulins do not crystallize readily. In ITG, the propensity to form microtubular structures or tactoids suggests that the deposits are composed of a uniform substructure with strong inter-molecular attraction. Therefore, one can speculate that the formation of immunotactoid deposits is the result of immune complexes having a uniform structure or an abnormal production of monoclonal immunoglobulins which perhaps have an unusual or abnormal structure. These may be produced in such small quantities that they escape detection with standard serologic evaluation as patients with ITG, by definition, do not have evidence of a circulating cryoglobulin or a paraprotein. The deposition of the immunoglobulins within the glomerulus along the filtration surface of the glomerular capillary wall may be a consequence of the unique environment created by the ultrafiltration of plasma (Korbet et al., 1985). The increased concentration of protein occurring along the glomerular capillary as a consequence of ultrafiltration may account for the tendency of the deposits to form exclusively within the kidney. Structural alterations along the filtration surface of the glomerulus may also be important and predispose to fibril formation. In mice, absence of CD2 associated protein (CD2ap), a protein which binds to nephrin and is important in the function of the podocyte slit diaphragm, results in congenital nephrotic syndrome and glomerular ultrastructural pathology similar to that seen in ITG in

fibrillary and immunotactoid glomerulopathy

humans (Shih et al., 1999; Li et al., 2000; Shaw, 2000; Kim et al., 2003). Thus, glomerular deposits in ITG may result from acquired defects in critical podocyte cellular functions involved in the clearance of filtered and retained immunoglobulin. Although the cause of ITG is unknown, the heterogeneity of the immunopathology suggests that more than one aetiology is responsible for the production of fibrils with a common morphologic appearance. In this respect it may be similar to amyloid where it is well known that various disease states are capable of producing different proteins which have in common the capacity to form the highly organized beta pleated sheet structure. Since immunotactoids may be composed of either immune complexes or monoclonal proteins which are capable of forming tactoids or microtubules, the variability in the size and orientation of the tactoids from one patient to another, may be a result of concentration or biochemical composition of the protein similar to that described in cryoglobulinaemia. Alternatively, the variability in ultrastructural morphology among patients with ITG may be analogous to the morphologic heterogeneity in haemoglobin S described in the haemoglobinopathy of sickle cell anaemia (Eaton and Hofrichter, 1987). The morphology of deoxygenated haemoglobin S in sickle cell disease is dependent upon the concentration of Hb S and the rate of tactoid formation. Under circumstances where they form slowly, the tactoids are aligned in parallel forming a paracrystalline structure. In contrast, the more rapidly the tactoids are formed the more random the orientation to one another (Eaton and Hofrichter, 1987). Similarly, a patient with ITG has been described with biochemically identical fibrils in the glomeruli and in a serum precipitate that formed after 4 months in cold storage, however, the ultrastructural morphology differed significantly between the fibrils (Rostagno et al., 1996). The fibrils in the glomeruli were 15–20 nM in diameter while those in the serum precipitate were 90 nM. Thus, as in haemoglobin S, the variability in morphology observed in ITG may result from physio-chemical factors involved in fibrillogenesis. The pathogenesis of ITG may be immunochemically diverse with the unifying feature being the ultrastructural organization of the deposits. In the appropriate setting, immune complexes or immunoglobulins are capable of forming fibrils or microtubules (tactoids) in the glomerular capillary wall or mesangium. Unfortunately, the disease(s) responsible for the production of the immune material in the tactoids of ITG has not been determined in the patients that have been described to date.

Prognosis and treatment

Fig. 81.6  Electron micrograph of glomerular microtubular deposits showing a parallel, packed arrangement and measuring 35 nM in diameter. (Uranyl acetate and lead citrate, ×32,000.)

The course of patients with ITG is one of progressive renal failure to the requirement of dialysis over 2–5 years in 50% of patients (Korbet et al., 1991; Iskander et al., 1992; Pronovost et al., 1996; Rosenstock et al., 2003; Nasr et al., 2011). The progression to end-stage renal disease (ESRD) is slower in those patients with predominantly mesangial proliferative or membranous lesions and more rapid in those with diffuse proliferative or membranoproliferative glomerular lesions (Rosenstock et al., 2003; Nasr et al., 2011). This is similar to other primary glomerulopathies but is distinct from that of the other fibrillary glomerulopathies (i.e. amyloid and monoclonal immunoglobulin deposition diseases) which experience a more rapid decline to ESRD (Korbet et al., 1994; Korbet and Schwartz, 2006). Clinical features at presentation which portend a poor renal prognosis include older age, hypertension, level of proteinuria,

653

654

Section 3  

the patient with glomerular disease Table 81.6  Post-transplant course in ITG Reference

Patients (N)

Follow-up (years)

Recurrence

End-stage renal disease

Alpers et al. (1987)

1

5

1 (5 years)

0

Sturgil et al. (1989)

1

2

0

0

Korbet et al. (1990)

2

5, 6

1 (4.5 years)

1

Fogo et al. (1993)

2

??

0

0

Pronovost et al. (1996)

4

4–11

2 (? years)

1

Carles et al. (2000)

1

4

1 (1.5 years)

0

Samaniego et al. (2000)

4

3–13

2 (2 and 9 years)

1

Rosenstock et al. (2003)

2

4, 8

0

0

Czarnecki et al. (2009)/Nasr et al. (2011)

14

0.5–13

5 (0.25–7 years)

2

Total

31

2–13

12 (39%)

5 (16%)

nephrotic syndrome, and the level of renal insufficiency (Korbet et al., 1985, 1991; Pronovost et al., 1996; Rosenstock et al., 2003; Nasr et al., 2011). Pathologic features which portend a poor renal prognosis include the percentage of globally sclerotic glomeruli, the severity of tubular atrophy and interstitial fibrosis and the extent of glomerular deposits (Korbet et al., 1985; Rosenstock et al., 2003; Nasr et al., 2011). The response to immunosuppressive treatment in nephrotic ITG patients has generally been poor. The use of prednisone (Dickenmann et al., 2002) and rituximab (Sathyan et al., 2009) has resulted in a remission of the nephrotic syndrome in a small number of patients with predominantly mesangial disease and well preserved renal function. However, the overall experience with treatment of ITG with steroids alone, steroids with immunosuppressive agents, and steroids with plasmapheresis have rarely (< 10% of patients) resulted in clinical remission of proteinuria or altered the progression to ESRD (Schwartz and Lewis, 1980; Alpers et  al., 1987; Schifferli et  al., 1987; D’Agati et  al., 1991; Minami et al., 1997; Kurihara et al., 1998; Rosenstock et al., 2003; Nasr et al., 2011). The overall patient survival of the patient with ITG is as one might expect with a primary glomerulopathy and no systemic disease. The survival at 1 year is 100% with > 80% of patients alive at 5 years (Korbet et al., 1985, 1991; Nasr et al., 2011)}. As a result, renal transplantation is a treatment consideration for ITG patients with ESRD.

Renal transplantation The outcome of renal transplantation (Table 81.6) has been reported in 31 ITG patients with 2–13  years of post-transplant follow-up (Alpers et al., 1987; Sturgil et al., 1989; Korbet et al., 1990; Fogo et al., 1993; Carles et al., 2000; Samaniego et al., 2001; Rosenstock et al., 2003; Czarnecki et al., 2009; Nasr et al., 2011). Recurrence of ITG has been demonstrated in 39% of these patients from 0.25 to 9 years after transplantation, and in five cases this resulted in the loss of the graft. In the remaining patients with recurrent disease, renal function continued to be adequate after 5–13 years of follow-up. The rate of deterioration in renal function in patients with recurrent disease has been shown to be slower than with their

original disease. One possible explanation could be the effect of immunosuppression (Pronovost et al., 1996). In those patients with recurrent disease, the ultrastructural morphology in the transplants was similar to that originally seen in the native kidneys (Alpers et al., 1987; Korbet et al., 1990; Carles et al., 2000; Samaniego et al., 2001). Thus, while recurrent disease does occur in ITG, it does not inevitably result in graft loss.

References Alpers, C. E. (1992). Immunotactoid (microtubular) glomerulopathy: an entity distinct from fibrillary glomerulonephritis? Am J Kidnet Dis, 19, 185–91. Alpers, C. E. (1993). Fibrillary glomerulonephritis and immunotactoid glomerulopathy: two entities, not one. Am J Kidney Dis, 22, 448–51. Alpers, C. E. and Kowalewska, J. (2008). Fibrillary glomerulonephritis and immunotactoid glomerulopathy. J Am Soc Nephrol, 19(1), 34–7. Alpers, C. E., Rennke, H. G., Hopper, J., et al. (1987). Fibrillary glomerulonephritis: An entity with unusual immunofluorescence features. Kidney Int, 31, 781–9. Brady, H. R. (1998). Fibrillary glomerulopathy [clinical conference]. Kidney Int, 53(5), 1421–9. Carles, X., Rostaing, L., Modesto, A., et al. (2000). Successful treatment of recurrence of immunotactoid glomerulopathy in a kidney allograft recipient. Nephrol Dial Transplant, 15(6), 897–900. Casanova, S., Donini, U., Zucchelli, P., et al. (1992). Immunohistochemical distinction between amyloidosis and fibrillary glomerulopathy. Am J Clin Pathol, 97, 787–95. Churg, J. and Venkataseshan, S. (1993). Fibrillary glomerulonephritis without immunoglobulin deposits in the kidney. Kidney Int, 44, 837–42. Czarnecki, P. G., Lager, D. J., Leung, N., et al. (2009). Long-term outcome of kidney transplantation in patients with fibrillary glomerulonephritis or monoclonal gammopathy with fibrillary deposits. Kidney Int, 75(4), 420–7. D’Agati, V., Sacchi, G., Truong, L., et al. (1991). Fibrillary glomerulopathy: defining the disease spectrum. J Am Soc Nephrol, 2, 591. Dickenmann, M., Schaub, S., Nickeleit, V., et al. (2002). Fibrillary glomerulonephritis: early diagnosis associated with steroid responsiveness. Am J Kidney Dis, 40(3), E9. Duffy, J. L., Khurana, E., Susin, M., et al. (1983). Fibrillary renal deposits and nephritis. Am J Pathol, 113, 279–90. Eaton, W. and Hofrichter, J. (1987). Hemoglobin S gelation and sickle cell disease. Blood, 70, 1245–66. Fogo, A., Qureshi, N., and Horn, R. G. (1993). Morphologic and clinical features of fibrillary glomerulonephritis versus immunotactoid glomerulopathy. Am J Kidney Dis, 22(3), 367–77.

Chapter 81 

Iskander, S. S., Falk, R. J., and Jennette, C. (1992). Clinical and pathological features of fibrillary glomerulonephritis. Kidney Int, 42, 1401–7. Kim, J. M., Wu, H., Green, G., et al. (2003). CD2-associated protein haploinsufficiency is linked to glomerular disease susceptibility. Science, 300(5623), 1298–300. Korbet, S. M., Rosenberg, B. F., Schwartz, M. M., et al. (1990). Course of renal transplantation in immunotactoid glomerulopathy. Am J Med, 89, 91–5. Korbet, S. M. and Schwartz, M. M. (2006a). Multiple myeloma. J Am Soc Nephrol, 17(9), 2533–45. Korbet, S. M., Schwartz, M. M., and Lewis, E. J. (1991). Immunotactoid glomerulopathy. Am J Kidney Dis, 17(3), 247–57. Korbet, S. M., Schwartz, M. M., and Lewis, E. J. (1994). The fibrillary glomerulopathies. Am J Kidney Dis, 23(5), 751–65. Korbet, S. M., Schwartz, M. M., and Lewis, E. J. (2006). Immuotactoid glomerulopathy (fibrillary glomerulonephritis). Clin J Am Soc Nephrol, 1(6), 1351–6. Korbet, S. M., Schwartz, M. M., Rosenberg, B. F., et al. (1985). Immunotactoid glomerulopathy. Medicine, 64(4), 228–43. Kurihara, I., Saito, T., Sato, H., et al. (1998). Successful treatment with steroid pulse therapy in a case of immunotactoid glomerulopathy with hypocomplementemia. Am J Kidney Dis, 32(1), E4. Li, C., Ruotsalainen, V., Tryggvason, K., et al. (2000). CD2AP is expressed with nephrin in developing podocytes and is found widely in mature kidney and elsewhere. Am J Physiol Renal Physiol, 279, F785–92. Masson, R. G., Rennke, H. G., and Gottlieb, M. N. (1992). Pulmonary hemorrhage in a patient with fibrillary glomerulonephritis. N Engl J Med, 326, 36–9. Minami, J., Ishimitsu, T., Inenaga, T., et al. (1997). Immunotactoid glomerulopathy: report of a case. Am J Kidney Dis, 30(1), 160–3. Nasr, S. H., Valeri, A. M., Cornell, L. D., et al. (2011). Fibrillary glomerulonephritis: a report of 66 cases from a single institution. Clin J Am Soc Nephrol, 6(4), 775–84. Ozawa, K., Yamabe, H., Fukushi, K., et al. (1991). Case report of amyloid-like glomerulopathy with hepatic involvement. Nephron, 58, 347–50. Pronovost, P. H., Brady, H. R., Gunning, M. E., et al. (1996). Clinical features, predictors of disease progression and results of renal transplantation in fibrillary/immunotactoid glomerulopathy. Nephrol Dial Transplant, 11, 837–42.

fibrillary and immunotactoid glomerulopathy

Rosenmann, E. and Eliakim, M. (1977). Nephrotic syndrome associated with amyloid-like glomerular deposits. Nephron, 18, 301–8. Rosenstock, J. L., Markowitz, G. S., Valeri, A. M., et al. (2003). Fibrillary and immunotactoid glomerulonephritis: Distinct entities with different clinical and pathologic features. Kidney Int, 63(4), 1450–61. Rostagno, A., Vidal, R., Kumar, A., et al. (1996). Fibrillary glomerulonephritis related to serum fibrillar immunoglobulin-fibrinectin complexes. Am J Kidney Dis, 28, 676–84. Sabatine, M. S., Aretz, H. T., Fang, L. S., et al. (2002). Images in cardiovascular medicine. Fibrillary/immunotactoid glomerulopathy with cardiac involvement. Circulation, 105(15), e120–e121. Samaniego, M., Nadasdy, G. M., Laszik, Z., et al. (2001). Outcome of renal transplantation in fibrillary glomerulonephritis. Clin Nephrol, 55(2), 159–66. Sathyan, S., Khan, F. N., and Ranga, K. V. (2009). A case of recurrent immunotactoid glomerulopathy in an allograft treated with rituximab. Transplant Proc, 41(9), 3953–5. Satoskar, A. A., Calomeni, E., Nadasdy, G., et al. (2008). Fibrillary glomerulonephritis with splenic involvement: a detailed autopsy study. Ultrastruct Pathol, 32(3), 113–21. Schifferli, J. A., Merot, Y., and Chatelanat, F. (1987). Immunotactoid glomerulopathy with leucocytoclastic skin vasculitis and hypocomplementemia: a case report. Clin Nephrol, 27, 151–5. Schwartz, M. M., Korbet, S. M., and Lewis, E. J. (2002). Immunotactoid glomerulopathy. J Am Soc Nephrol, 13(5), 1390–7. Schwartz, M. M. and Lewis, E. J. (1980). The quarterly case: nephrotic syndrome in a middle-aged man. Ultrastruct Pathol, 1, 575–82. Shaw, A. S. (2000). Congenital nephrotic syndrome in mice lacking CD2-associated protein. J Am Soc Nephrol, 11, 19. Shih, N. Y., Li, J., Karpitskii, V., et al. (1999). Congenital nephrotic syndrome in mice lacking CD2-associated protein. Science, 286, 312–5. Sturgil, B. C. and Bolton, W. K. (1989). Non-amyloidotic fibrillary glomerulopathy. Kidney Int, 35, 233. Wallner, M., Prischl, F. C., Hobling, W., et al. (1996). Immunotactoid glomerulopathy with extrarenal deposits in the bone, and chronic cholestatic liver disease. Nephrol Dial Transplant, 11, (8) 1619–24. Yang, G. C. H., Nieto, R., Stachura, I., et al. (1992). Ultrastructural immunohistochemical localization of polyclonal IgG, C3, and amyloid P component on the Congo red-negative amyloid-like fibrils of fibrillary glomerulopathy. Am J Pathol, 141, 409–19.

655

CHAPTER 82

Drug-induced and toxic glomerulopathies Alexander Woywodt and Diana Chiu Introduction The kidney is frequently the site of injury from drugs and its metabolites, as highly protein-bound substances are unbound and tubular concentrations are often much higher than those seen in plasma or tissues (see Chapter 362). Many offending substances therefore cause damage, primarily to tubular cells, as some concentrated solutes in the tubular fluid reach urinary:plasma concentration ratios in excess of 1000:1. Furthermore, tubular cells host a multitude of transport mechanisms and receptors, which render them even more vulnerable to injury as is illustrated in the role of megalin in aminoglycoside-induced nephrotoxicity. Therefore, drug-induced nephrotoxicity is in its overwhelming majority due to lesions of the tubules and interstitium, followed by drug-induced endothelial and vascular syndromes, such as drug-induced thrombotic microangiopathy. In comparison, drug-induced glomerulopathy is relatively rare and also much less appreciated by many clinicians. However, the idea that many drugs and substances are capable of inducing glomerular lesions is very old and substances such as the aminonucleoside puromycin or phorbol myristate acetate (PMA) and Adriamycin® (doxorubicin ) (Lee and Harris, 2011) have been used for decades to induce glomerular damage in animal models. A  broad variety of histological lesions have been described in drug-induced glomerulopathy (Izzedine et al., 2006). Large-scale studies are lacking but the commonest disease entity seen in our clinical practice is probably minimal change disease (MCD) associated with non-steroidal anti-inflammatory drug (NSAID) use. Drug-induced focal and segmental glomerulosclerosis (FSGS) and membranous GN (MGN) are rare, at least in our practice, although historically MGN associated with gold treatment used to be common while this substance was still in widespread use for rheumatoid arthritis (Hill, 1986). Other forms of drug-induced glomerulopathy are rarer still. Some drugs are almost exclusively linked to one particular histological lesion, for example, pamidronate, which is associated with the collapsing variant of FSGS. Other substances, such as the NSAIDs, penicillamine, or heroin, are capable of causing a variety of glomerular lesions. Table 82.1 provides an overview of lesions and commonly implicated drugs and toxins. Drug-induced thrombotic microangiopathy (see Chapter 174), glomerular disease due to calcineurin inhibitors (see Chapter 362), and the putative role of hydrocarbons in renal disease associated with antibodies to the glomerular basement membrane (see Chapter 74) are all discussed

elsewhere. In the following, we will first discuss common forms of drug-induced glomerulopathy by histological phenotype. We will then review other forms of glomerulopathy, which are associated with individual substances or substance classes that do not fit into any of the histological types of glomerulonephritis discussed already. This will include glomerulopathy due to inhibitors of the mammalian target of rapamycin (mTOR) and due to therapeutic inhibition of vascular endothelial growth factor (VEGF).

Drug-induced minimal change glomerulopathy MCD is one of the more common forms of drug-induced glomerulopathy and NSAIDs are often implicated, among a variety of other drugs (Table 82.1). The spectrum of NSAID-induced nephrotoxicity is well characterized. It also includes peripheral oedema, acute interstitial nephritis, papillary necrosis, tubular damage, and acute kidney injury (see Chapter 362). NSAID-induced MCD is well described and some 100 cases have been reported in the literature. In a large series of adult patients with MCD as many as 10% were associated with NSAID use (Warren et al., 1989). Various mechanisms have been implicated (Izzedine et al., 2006). Interestingly, NSAID can also reduce proteinuria (Alavi et  al., 1986) as is illustrated by their historic or last-ditch use in the treatment of nephrotic syndrome (Arisz et al., 1976; Alavi et al., 1986 and see Chapter 52). All NSAIDS and atypical variant substances, such as piroxicam, or cyclooxygenase (COX)-2 inhibitors seem to be capable of causing MCD (Izzedine et al., 2006). The duration of NSAID use prior to the onset of proteinuria is variable, ranging from 2 weeks to 18 months. In our experience, many patients with NSAID-induced MCD will have used the drug for a considerable period of time. MCD associated with NSAIDs may or may not remit after stopping the offending drug. Data as to who may benefit from steroid treatment are currently lacking; the optimum dose and duration of treatment are also unclear. Some authors have suggested a 2-month regimen, that is, a shorter treatment than is usually advocated for MCD not associated with NSAID (Izzedine et al., 2006). The different renal lesions associated with NSAID use are not mutually exclusive. Indeed MCD and acute interstitial nephritis often coexist (Clive and Stoff, 1984; Warren et al., 1989) (Fig. 82.1). The same observation has been described for celecoxib (Alper

Chapter 82 

drug-induced and toxic glomerulopathies

Table 82.1  Drug-induced glomerulopathy: histological lesions, and commonly implicated drugs Lesion

Commonly implicated drug

Minimal change glomerulonephritis (Izzedine et al., 2006)

D-penicillamine (Herve et al., 1980), Gold (Francis et al., 1984; Wolters et al., 1987), interferon alpha (Dizer et al., 2003), interferon beta (Nakao et al., 2002), lithium, mercury (Tang et al., 2006), NSAIDS (Warren et al., 1989; Izzedine et al., 2006), including atypical NSAIDs, such as piroxicam (Fellner, 1985), pamidronate (Barri et al., 2004), and sulfasalazine (Molnar et al., 2010)

Focal and segmental glomerulosclerosis (Izzedine et al., 2006) Collapsing variant NOS Perihilar variant Tip lesion

Pamidronate (Markowitz et al., 2001; Kunin et al., 2004), probably also alendronate (Pascual et al., 2007), anabolic steroids (Herlitz et al., 2010), interferon alpha, beta, and gamma (Markowitz et al., 2010), sirolimus (Dogan et al., 2011), valproic acid (Ackoundou-N’guessan et al., 2007) Lithium (Markowitz et al., 2000), heroin (or an adulterant), particularly in black patients (Rao et al., 1974), interferon alpha (Traynor et al., 1994), mTOR inhibitors (Letavernier et al., 2007; Izzedine et al., 2009), norfloxacin (Traynor et al., 1996), toluene (Bosch et al., 1988) Anabolic steroids (Herlitz et al., 2010) NSAID (Sekhon et al., 2005)

Membranous glomerulonephritis (Izzedine et al., 2006)

Adalimumab (den Broeder et al., 2003), aprotinine (Boag et al., 1985), captopril (Hoorntje et al., 1979), celecoxib (Markowitz et al., 2003b), diuretics, gold (Francis et al., 1984; Hall et al., 1987), lithium (Phan et al., 1991), mercury (Li et al., 2010; George, 2011), NSAIDs and COX-2 inhibitors (Markowitz et al., 2003b), penicillamine (Neild et al., 1979), probenecid (Izzedine et al., 2007), methiamazole (Reynolds and Bhathena, 1979), TNF inhibitors (Stokes et al., 2005)

Membranoproliferative glomerulonephritis

Heroin, particularly in Caucasian patients (Jaffe and Kimmel, 2006)

Crescentic glomerulonephritis and vasculitis (Izzedine et al., 2006) ANCA positive (mostly pANCA with MPO specificity) ANCA negative

Allopurinol (Choi et al., 1998), D-penicillamine (Nanke et al., 2000; Bienaime et al., 2007), hydralazine (Dobre et al., 2009; Kalra et al., 2012), minocycline (pANCA with specificities for cathepsin G (Elkayam et al., 1998), elastase (Elkayam et al., 1998), and bactericidal permeability increasing protein (Elkayam et al., 1998) rather than against MPO and anecdotal cases of cANCA with PR3 specificity (Sethi et al., 2003), propylthiouracil (Yu et al., 2007), TNF inhibitors (Stokes et al., 2005), levamisole (Simms et al., 2008; Zwang et al., 2011) Adalimumab (Fournier et al., 2009), foscarnet (Trolliet et al., 1995), isoniazid (Brik et al., 1998), D-penicillamine (Ntoso et al., 1986), penicillin, phenylbutazone (Leung et al., 1985), rifampicin (Ogata et al., 1998), TNF alpha antagonists (Simms et al., 2008), thiazides

Lupus nephritis (all subtypes) (Izzedine et al., 2006)

Alpha-methyldopa, procainamide (Sheikh et al., 1981; McLaughlin et al., 1998), hydralazine (Shapiro et al., 1984), quinidine (Alloway and Salata, 1995), TNF inhibitors (Neradova et al., 2009) and others (Stokes et al., 2007; Chang and Gershwin, 2011).

Others/unclassified forms of glomerulopathy

Chloroquine and hydroxychloroquine (Bracamonte et al., 2001; Woywodt et al., 2007), cocaine (Jaffe et al., 2006), foscarnet (Goldfarb and Coe, 1998; Maurice-Estepa et al., 1998), various herbs and natural products (Blowey, 2005), heroin (Jaffe et al., 2006), mercury (George, 2011), mTOR inhibitors (Bertoni et al., 2009), uranium (Vicente-Vicente et al., 2010), VEGF inhibitors (George et al., 2007; Stokes et al., 2008; Izzedine et al., 2010; Manjunath et al., 2011)

et al., 2002). This is a common clinical pitfall in that the presence of active urinary sediment together with the nephrotic syndrome may tempt the clinician to exclude MCD from the differential diagnosis. Fever, rash, and eosinophilia are usually absent (Izzedine et al., 2006) although in our experience serum immunoglobulin (Ig)-E is often elevated. Many other drugs are also associated with MCD (Table 82.1). The second most commonly implicated drug is probably lithium, although large studies into the incidence of drug-associated MCD are lacking. Lithium nephropathy usually spares the glomeruli and instead features chronic tubulointerstitial nephritis with tubular cysts, polydipsia/polyuria, and impaired renal concentrating ability (see Chapter 362). In comparison little is known about glomerular damage due to lithium (Markowitz et al., 2000). However, a number of cases of lithium-induced MCD have been reported, often featuring rapid remission of nephrotic syndrome after drug withdrawal (Tam et  al., 1996). There are also reports of relapse after reintroduction (Aliabadi et al., 2008). Wood and colleagues in 1989 reviewed nine cases and also reported a favourable outcome after cessation of lithium although two patients required steroid treatment to attain remission (Wood et al., 1989). Of note, FSGS

in association with lithium has been reported as well (Markowitz et al., 2000), suggesting that lithium exerts direct effects on podocytes (Markowitz et al., 2000) The mechanism of this effect, however, is as yet controversial and poorly defined. Dai and colleagues demonstrated that lithium causes β-catenin activation, causing podocyte injury (Dai et  al., 2009)  whereas Tam and co-workers postulated an involvement of the phosphoinositol pathway (Tam et al., 1996).

Drug-induced focal and segmental glomerulosclerosis FSGS has become a common histological diagnosis in adults with nephrotic syndrome and now represents the most common primary glomerular disease underlying end-stage renal disease (ESRD) in the United States (Kitiyakara et al., 2004). Recent years have seen not only a marked increase in the incidence of FSGS but also a revised classification of the disease (see Chapter 57). Among a variety of secondary forms, the drug-induced variant of FSGS is rare. Pamidronate (Fig. 82.2), lithium, anabolic steroids, and heroin are most commonly implicated (Table 82.1).

657

658

Section 3  

the patient with glomerular disease

(A)

(B)

Fig. 82.1  Renal biopsy obtained from a 73-year-old patient who presented with nephrotic syndrome (proteinuria 30 g/L) and renal impairment (serum creatinine 375 μmol/L) after a long history of using various NSAIDs, most recently naproxen. (A) Light microscopy shows normal glomeruli with an infiltrate composed of lymphocytes and occasional eosinophils. (B) Electron microscopy shows widespread fusion of podocyte foot processes. The diagnosis is NSAID-associated minimal change disease with concurrent acute interstitial nephritis. The patient made a full recovery with steroid treatment and renal function returned to normal. Courtesy of Dr Beena Nair, Department of Pathology, Royal Preston Hospital, and Dr Ajay Dhaygude, Department of Nephrology, Royal Preston Hospital.

In 2001, Markowitz and others first reported collapsing-type FSGS in seven patients who had received pamidronate (Markowitz et al., 2001). Of note, MCD has also been described in association with pamidronate (Markowitz et al., 2004). Recovery of nephrotic syndrome is possible after cessation of pamidronate treatment (Desikan et al., 2002) but reversal may be incomplete or even absent in some cases, particularly those with markedly impaired renal function (Izzedine et al., 2006). Progression to ESRD is also seen (Izzedine et al., 2006). It is not clear whether other bisphosphonates are also capable of causing FSGS. Deterioration of pre-existing FSGS in association with alendronate use has been reported (Miura et al., 2009), as well as worsening proteinuria and renal impairment associated with alendronate use in a liver transplant recipient with pre-existing chronic kidney disease (Pascual et  al., 2007). Most cases reported so far have occurred in the context of high-dose pamidronate treatment for multiple myeloma (Perazella and Markowitz, 2008). Newer bisphosphonates such as pamidronate differ in mode of action in that they inhibit farnesyl synthase and

Fig. 82.2  Renal biopsy obtained from a 52-year-old patient with multiple myeloma and pamidronate treatment. The biopsy shows a glomerulus with global collapse of the glomerular tuft with surrounding tubules exhibiting atrophy. Interstitial fibrosis is also present. (Silver methenamine stain ×200.) From Kunin et al. (2004), with permission.

thereby interfere with guanosine-5'-triphosphate (GTP)-binding proteins. It has been speculated that this difference accounts for the differential effects of pamidronate on the podocyte when compared to older substances from the same class (Markowitz et al., 2001). Of note, bisphosphonates are also capable of causing acute kidney injury through tubular damage (Markowitz et al., 2003). The spectrum of bisphosphonate nephrotoxicity is reviewed in detail elsewhere (Perazella and Markowitz, 2008). Finally, it has to be noted that the true incidence of podocyte injury, proteinuria, and FSGS in association with bisphosphonates remains unclear as long as good data from large trials are lacking (Perazella and Markowitz, 2008). Lithium is another well-described cause of FSGS and the substance can also cause MCD as discussed above. Previously viewed almost exclusively as a tubulointerstitial disease, lithium nephropathy must now be regarded as a disease that often affects the glomerulus as well (Markowitz et al., 2000). FSGS due to lithium was considered truly rare until Markowitz and colleagues reported FSGS in 50% of 24 renal biopsies taken from patients on lithium treatment (Markowitz et al., 2000). It is difficult to exclude with certainty hyperfiltration due to nephron loss as the causative mechanisms (as opposed to a true effect of the podocyte per se). However, this study reported a high incidence of foot process fusion on electron microscopy, suggesting a direct podocytopathic effect of lithium treatment (Markowitz et al., 2000). Finally, it is worthwhile to note that FSGS and tubulointerstitial changes due to lithium often coexist (Markowitz et al., 2000).

Chapter 82 

In 2010, Herlitz and co-workers described FSGS in patients who had previously used anabolic steroids as body builders (Herlitz et al., 2010). In this cohort, collapsing variant FSGS was seen in three cases while the perihilar variant was seen in four patients. One patient progressed to ESRD. The authors speculated that secondary FSGS resulted from a combination of post-adaptive glomerular changes driven by increased lean body mass and a potential direct nephrotoxic effect of anabolic steroids (Herlitz et al., 2010). The causality must remain unclear, since FSGS in conjunction with obesity (Kambham et al., 2001) but also in patients with increased muscle mass is well described (Schwimmer et al., 2003).

Drug-induced membranous glomerulonephritis Secondary membranous glomerulonephritis (MGN) is described in Chapter 63. Overall, drug-induced MGN is rare:  Rihova and colleagues reported 18 such cases in a series of 129 cases of MGN (Rihova et al., 2005). Furthermore, it must be emphasized that causality can be difficult to establish. A  broad variety of drugs has been implicated (Table 82.1), particularly captopril (Hoorntje et  al., 1979), D-penicillamine (Neild et  al., 1979; Hall et  al., 1988), gold (Hall et al., 1987), mercury, NSAIDs, and COX-2 inhibitors (Sennesael et al., 1986; Markowitz et al., 2003). Much of the literature relates to gold or penicillamine, neither of which is commonly used today. Mechanisms remain ill-defined and derive from our current understanding of MGN in general. Here, the antigen may be the drug itself or its components or metabolites or the drug may act as a hapten (Nadasdy et al., 1998). A good example is the case of captopril where it has been suggested that the sulfhydryl moiety of the captopril molecule is involved (Izzedine et al., 2006), either directly or through an immune-mediated mechanism (Kallenberg et al., 1981, 1982). Accordingly, there is no convincing reports in conjunction with the use of other angiotensin-converting enzyme inhibitors. It has been proposed that the drug-induced variant tends to be mild (Izzedine et al., 2006). Epi-membranous deposits can be small and widely spaced on electron microscopy (Izzedine et al., 2006). MGN associated with gold (Hall et al., 1987) has now become exceedingly rare as the use of gold in the treatment of rheumatoid arthritis has been largely superseded. But the complication was uncommon. Katz and co-workers detected proteinuria in 3% of 1283 auranofin-treated patients, with four cases of GN (Katz et al., 1984). Nephropathy due to mercury was described in 1811 by Scottish-American physician William Charles Wells (George, 2011). Exposure in contemporary cases is usually from mercury-containing skin cream (Soo et  al., 2003), hair-dying agents, or inhalation of mercury vapour at the workplace. Li and colleagues reported on a series of 11 patients with proteinuria and normal renal function and described MN with a particular pattern of IgG subclass deposition (Li et al., 2010). MCD due to mercury has also been reported (Tang et al., 2006). Withdrawal of the offending substance led to complete recovery in the majority of cases (Li et al., 2010). Generally, it has been suggested that the course of drug-induced MGN is often favourable once the offending agent has been stopped (Izzedine et al., 2006).

drug-induced and toxic glomerulopathies

Drug-induced crescentic glomerulonephritis and lupus/lupus-like nephritis Drug-induced crescentic glomerulonephritis (see Chapter 70) is rare. The causality may be difficult to establish, but in general the mechanism is usually via causing a small vessel vasculitis (see Chapter 157). ANCA may be either positive or negative (Table 82.1). Some cases of drug-induced crescentic GN occur in the context of drug-associated systemic vasculitis, which is often associated with ANCA, overwhelmingly with a perinuclear pattern and specificity for myeloperoxidase (MPO) (Table 82.1). The only exception to this rule seems to be minocycline, which is capable of inducing a whole variety of ANCA specificities, such as cathepsin G (Elkayam et al., 1998), elastase (Elkayam et al., 1998), and bactericidal permeability increasing protein (Elkayam et al., 1998). Other cases occur as crescentic glomerulonephritis without systemic involvement. The topic of drug-induced ANCA-associated vasculitis is discussed in detail by Choi and colleagues (Choi et al., 2000). A vigorous accompanying interstitial nephritis is sometimes seen (Abt and Gordon, 1985). The underlying mechanisms remain essentially unknown. Penicillamine also deserves special attention in that the substance is capable of causing a surprising variety of different renal lesions, including membranous glomerulonephritis, ANCA-negative crescentic GN and, finally, an ANCA-positive systemic vasculitis (Table 82.1). Remission after cessation of the offending drug has been described, as has been successful immunosuppressive treatment in more severe cases (Ntoso et al., 1986). At present, it is therefore difficult to provide any valid guidance for treatment, chiefly because the number of cases overall is so small. Drug-induced lupus or lupus-like nephritis is an equally uncommon condition and differs in its manifestation from drug-induced vasculitis. The topic of drug-induced lupus is reviewed in detail elsewhere (Chang and Gershwin, 2011). Even among cases with drug-induced lupus, renal involvement is believed to be rare, although good data are lacking and subtle renal involvement may be under-diagnosed. Drug-induced lupus nephritis is typically associated with only a handful of drugs (Table 82.1). Sheikh and co-workers documented a case of crescentic lupus nephritis in drug-induced lupus due to procainamide (Sheikh et al., 1981). McLaughlin and others reported a similar case, again in association with procainamide. Nephrotic syndrome has been reported as well (Zech et al., 1979). Several cases of lupus nephritis have been described in conjunction with the use of tumour necrosis factor inhibitors in rheumatoid arthritis (Stokes et al., 2005). These agents are capable of inducing vasculitis as well. The clinical spectrum of autoimmune phenomena cause by these agents is reviewed in detail elsewhere (Ramos-Casals et al., 2007).

Other forms of drug-induced glomerular damage associated with individual substances or classes of substances Chloroquine/hydroxychloroquine-induced storage disorder Both chloroquine and hydroxychloroquine can cause a rare pseudo-Fabry (see Chapter 335) storage disorder with glomerular involvement. Findings include ‘zebra bodies’ on electron

659

660

Section 3  

the patient with glomerular disease

microscopy, which can be indistinguishable from those seen in Fabry disease (Fig. 82.3). The lack of extrarenal manifestations of Fabry’s or family history should lead to questioning the diagnosis of Fabry Syndrome and scrutiny of the medication (Woywodt et al., 2007). Albay and co-workers suggested criteria to distinguish chloroquine-induced storage disorder from true Fabry disease (Albay et al., 2005). Bracamonte and co-workers review this topic in great detail (Bracamonte et al., 2006).

Glomerulopathy due to foscarnet Foscarnet is a second-line antiviral drug used for ganciclovir-resistant cytomegalovirus (CMV) infection. Nephrotoxicity through widespread intrarenal deposition of tricalcium or mixed sodium/calcium foscarnet crystals is well described (Goldfarb and Coe, 1998; Deray et al., 1989; Zanetta et al., 1999) and occurs in as many as 20–30% of patients (Wagstaff and Bryson, 1994). Good supportive care, hydration, and dose adjustment are effective for prevention and nephrotoxicity is often reversible (Wagstaff and Bryson, 1994). Tubular toxicity appears to be the predominant mechanisms causing but there is unequivocal evidence of glomerular involvement (Goldfarb and Coe, 1998; Maurice-Estepa et al., 1998). Proteinuria around 1 g/day is therefore common in foscarnet-induced nephrotoxicity. The urinary sediment can be unremarkable but microscopic haematuria has been described as well (Maurice-Estepa et al., 1998). Glomerular capillaries are also involved in the disease process (Beaufils et  al., 1990)  and crescentic glomerulonephritis has been reported (Trolliet et al., 1995).

Glomerulopathy caused by heavy metals Several heavy metals, such as cadmium, lead, mercury, and uranium, are well-described nephrotoxins. However, their nephrotoxicity is determined chiefly by their ability to cause tubulointerstitial damage with varying degrees of chronicity. There are two exceptions, membranous glomerulonephritis due to mercury (discussed above) and the glomerulopathy caused by uranium. The nephrotoxicity of uranium is described in detail elsewhere (Arzuaga et al., 2010; Vicente-Vicente et al., 2010; see Chapter 362) but very few

cases have been reported. Uranium causes both tubulointerstitial and glomerular damage (Bentley et al., 1985) and a variety of reversible changes of the podocyte foot processes is seen in rats (Kobayashi et  al., 1984). Others have suggested that glomerular endothelial cells, rather than podocytes, are the prime target of uranium-induced nephrotoxicity (Avasthi et al., 1980).

Glomerulopathy due to opiates and cocaine Evidence that use of street heroin may be associated with proteinuria and progressive renal failure first emerged in small case series in the 1970s (Rao et al., 1974). Renal biopsy showed FSGS in the majority of cases (Rao et  al., 1974). Subsequent reports have emphasized rapid progression to ESRD (Hill, 1986). Other histological lesions have been described as well, particularly immune-complex glomerulonephritis and membranoproliferative glomerulonephritis (MPGN). Renal amyloidosis has been reported in association with subcutaneous injection of heroin (‘skin popping’) (Hill, 1986). Some authors have questioned the existence of heroin nephropathy as an entity in its own right (Jaffe and Kimmel, 2006) while others have speculated that additives, rather than heroin itself, may be responsible (Friedman and Tao, 1995). This is supported by the observation that despite continuing use of the drug the incidence of heroin-associated renal failure has decreased markedly since the end of the 1980s (Friedman and Tao, 1995). Opiate use is still associated with ESRD (Perneger et al., 2001), but immune complex glomerulonephritis due to endocarditis and hepatitis C as well as HIV nephropathy and renal amyloidosis are also prevalent in this population. The use of cocaine, crack and other drugs also shows a weak association with ESRD although robust data are lacking (Jaffe and Kimmel, 2006). A variety of histological lesions has been described in cocaine users, including accelerated vascular nephropathy, glomerulosclerosis and tubulointerstitial damage (Jaffe and Kimmel, 2006). Cocaine is also capable of causing renal infarction through vasospasm (Madhrira et al., 2009) as well as interstitial nephritis (Wojciechowski et al., 2008).

Glomerulopathy due to mTOR inhibitors

Fig. 82.3  Laminated intra-cytoplasmic inclusions (‘zebra bodies’) and myelin figures within podocytes in chloroquine-induced storage disorder. (Renal biopsy, transmission electron microscopy, 8000× magnification.) From Woywodt et al. (2007), with permission.

mTOR is a highly conserved serine/threonine kinase, which controls cell growth and metabolism in response to nutrients, growth factors, cellular energy, and stress. mTOR inhibitors are sometimes used in renal transplantation, either in transplant recipients with malignancy or in those with chronic allograft nephropathy (see Chapter 286). The fact that these substances cause proteinuria is now well appreciated (Rangan, 2006) although both the clinical course of this disorder and its mechanisms remain ill-defined. Studies by Letavernier and co-workers (Letavernier et al., 2005) and by Ruiz and colleagues (Ruiz et al., 2006) provided first robust evidence of proteinuria in renal transplant recipients converted to sirolimus. Proteinuria can be substantial and reach the subnephrotic and even nephrotic range in some cases (Perlman et al., 2007). Some authors have proposed predictive factors, such as pre-conversion proteinuria and blood pressure (Padiyar et al., 2010). The effect is usually reversible after withdrawal of the offending drug (Perlman et al., 2007). Histological correlates of this entity remain particularly ill-defined although FSGS (Straathof-Galema et al., 2006; Franco et  al., 2007; Letavernier et  al., 2007)  and its collapsing variant (Dogan et al., 2011) have been described. Others have proposed a

Chapter 82 

drug-induced and toxic glomerulopathies Growth factors amino acids

?

Proteinuria

TSC2

TSC1

?

TSC Rheb

mTOR Rictor

mTORC1 Podocyte

Raptor mTOR

Slit diaphragm Mislocalization of slit diaphragm

PKC

AKT

Actin cytoskeleton

FoxO

GBM

Protein translation

ER stress

Cell Mesenchymal change growth

Detachment from GBM

Proteinuria sclerosis

Fig. 82.4  Putative mechanisms of glomerulopathy due to mTOR inhibitors. Podocyte integrity depends on careful balance of the two mTOR complexes, mTORC1 and 2. Raptor and rictor are the co-factors complexed in mTORC1 and 2, respectively. Growth factors and amino acids induce mTORC1 activity, their counterparts for mTORC2 activation are unknown. Activation of mTORC1, as seen in diabetic nephropathy, can lead to a variety of changes, including a more mesenchymal phenotype, detachment from the basement membrane, proteinuria, and sclerosis. Conversely, inhibition of mTORC1, can also cause podocyte injury. Added mTORC2 inhibition causes more severe podocyte injury. Sirolimus is capable of inhibiting both mTORC1 and 2, and of disturbing the mTORC1/2 balance. From Fogo (2011), with permission.

tubular mechanism (Lieberthal et al., 2001; Straathof-Galema et al., 2006) although it is difficult to understand how a tubular mechanism should cause nephrotic-range proteinuria (Letavernier et al., 2005). Sirolimus is capable of inducing proteinuria not only in renal transplant recipients but also in native kidneys as described in the context of bone marrow (Jhaveri et al., 2008) and cardiac transplantation (Aliabadi et al., 2008). Proteinuria appears to be a class effect of all mTOR inhibitors and is also seen with everolimus, both in animal models (Vogelbacher et al., 2007), and in humans (Bertoni et al., 2009) and with temsirolimus (Izzedine et al., 2009). Some authors report a decrease in proteinuria after conversion from sirolimus to everolimus (Neau-Cransac et al., 2009), but this is difficult to assess. Possible mechanisms include interference with VEGF synthesis in podocytes. But they also decrease Akt phosphorylation in vivo, in parallel with changes of cytoskeleton and cell phenotype (Letavernier et al., 2009). More recently, Stallone and others confirmed a dose-dependent effect of mTOR inhibitors on key podocyte structures (Stallone et al., 2011). Recent studies by Inoki et al. (2011) and Goedel et al. (2011) have also underscored the crucial role of the mTOR pathway in podocytes. It is now believed that podocyte maintenance is dependent on a fine-tuned balance of two mTOR complexes, mTORC1 and mTORC2 (Fogo, 2011) and that this balance can be disturbed by mTOR inhibitors (Fig. 82.4). There

may also be a genetic basis for the susceptibility to the effects of mTOR inhibitors on podocytes (Fogo, 2011).

Glomerulopathy due to VEGF inhibition Glomerulopathy has been repeatedly described in association with inhibitors of vascular endothelial growth factor (VEGF) (Izzedine et  al., 2010). George and colleagues, in 2007, first reported nephrotic syndrome in a patient treated with Bevacizumab for pancreatic adenocarcinoma (George et  al., 2007). Renal biopsy showed focal glomerulonephritis with immune complex deposition (George et al., 2007). In a meta-analysis of published cancer trials with Bevacizumab, Wu and co-workers reported that treatment increases the risk of severe proteinuria (Wu et al., 2010). The renal histology in published cases appears to be heterogenous and includes cryoglobulinaemic glomerulonephritis (Johnson et  al., 2004) collapsing glomerulopathy (Miller et al., 2005), and immune complex-associated focal proliferative glomerulonephritis (George et al., 2007). VEGF inhibition has also been associated with thrombotic microangiopathy (Eremina et al., 2008). Electron microscopy showed endothelial injury in these cases (Eremina et  al., 2008). Other VEGF-signalling blockers (axitinib, sorafenib, sunitinib, VEGF-Trap) have also been associated with severe proteinuria (Wu et  al., 2010). An algorithm for management of proteinuria due to VEGF inhibition has been suggested elsewhere (Izzedine

661

662

Section 3  

the patient with glomerular disease

et al., 2010). Manjunath and colleagues very recently reported proteinuria in a patient treated with Vargatef®, an investigational agent that blocks not only VEGF but also platelet-derived growth factor receptor (PDGF) and fibroblast growth factor (FGF) receptor (Manjunath et al., 2011). Renal biopsy showed cytoplasmic vacuoles with osmiophilic material in the podocytes, mesangial and endothelial cells (Manjunath et al., 2011). It is possible that concomitant interference with the PDGF and FGF pathway accounted for the distinct histological phenotype in this case. The putative mechanisms of this peculiar disorder have been the focus of considerable scientific interest. It is well known that VEGF is produced by podocytes and that glomerular endothelial cells possess VEGF receptors (Eremina et  al., 2003). More recent evidence has further emphasized the crucial role of VEGF in crosstalk between glomerular endothelial cells and podocytes as reviewed in great detail elsewhere (Fogo and Kon, 2010). In animal models, VEGF inhibition causes glomerular endothelial cell detachment and hypertrophy, in association with downregulation of nephrin (vet al., 2003). Some have compared renal injury due to VEGF inhibition with pre-eclampsia (see Chapter 298), with which it has multiple similarities (Muller-Deile and Schiffer, 2011) where increased sFLT-1 receptor is believed to lead to low levels of free VEGF, with consecutive glomerular damage (Maynard et al., 2003).

References Abt, A. B. and Gordon, J. A. (1985). Drug-induced interstitial nephritis. Coexistence with glomerular disease. Arch Intern Med, 145, 1063–7. Ackoundou-N’guessan, C., Canaud, B., Leray-Moragues, H., et al. (2007). Collapsing focal segmental glomerulosclerosis as a possible complication of valproic acid. S Afr Med J, 97, 388, 90. Alavi, N., Lianos, E. A., Venuto, R. C., et al. (1986). Reduction of proteinuria by indomethacin in patients with nephrotic syndrome. Am J Kidney Dis, 8, 397–403. Albay, D., Adler, S. G., Philipose, J., et al. (2005). Chloroquine-induced lipidosis mimicking Fabry disease. Mod Pathol, 18, 733–8. Aliabadi, A. Z., Pohanka, E., Seebacher, G., et al. (2008). Development of proteinuria after switch to sirolimus-based immunosuppression in long-term cardiac transplant patients. Am J Transplant, 8, 854–61. Alloway, J. A. and Salata, M. P. (1995). Quinidine-induced rheumatic syndromes. Semin Arthritis Rheum, 24, 315–22. Alper, A. B., Jr., Meleg-Smith, S., and Krane, N. K. (2002). Nephrotic syndrome and interstitial nephritis associated with celecoxib. Am J Kidney Dis, 40, 1086–90. Arisz, L., Donker, A. J., Brentjens, J. R., et al. (1976). The effect of indomethacin on proteinuria and kidney function in the nephrotic syndrome. Acta Med Scand, 199, 121–5. Arzuaga, X., Rieth, S. H., Bathija, A., et al. (2010). Renal effects of exposure to natural and depleted uranium: a review of the epidemiologic and experimental data. J Toxicol Environ Health B Crit Rev, 13, 527–45. Avasthi, P. S., Evan, A. P., and Hay, D. (1980). Glomerular endothelial cells in uranyl nitrate-induced acute renal failure in rats. J Clin Invest, 65, 121–7. Barri, Y. M., Munshi, N. C., Sukumalchantra, S., et al. (2004). Podocyte injury associated glomerulopathies induced by pamidronate. Kidney Int, 65, 634–41. Beaufils, H., Deray, G., Katlama, C., et al. (1990). Foscarnet and crystals in glomerular capillary lumens. Lancet, 336, 755. Bentley, K. W., Stockwell, D. R., Britt, K. A., et al. (1985). Transient proteinuria and aminoaciduria in rodents following uranium intoxication. Bull Environ Contam Toxicol, 34, 407–16. Bertoni, E., Bruschi, M., Candiano, G., et al. (2009). Posttransplant proteinuria associated with everolimus. Transplant Proc, 41, 1216–17.

Bienaime, F., Clerbaux, G., Plaisier, E., et al. (2007). D-Penicillamine-induced ANCA-associated crescentic glomerulonephritis in Wilson disease. Am J Kidney Dis, 50, 821–5. Blowey, D. L. (2005). Nephrotoxicity of over-the-counter analgesics, natural medicines, and illicit drugs. Adolesc Med Clin, 16, 31–43, x. Boag, F., Chappell, M., Beckett, A. G., et al. (1985). Lipohypertrophy and glomerulonephritis after the use of aprotinin in an insulin-dependent diabetic. N Engl J Med, 312, 245–6. Bosch, X., Campistol, J. M., Montoliu, J., et al. (1988). Myelofibrosis and focal segmental glomerulosclerosis associated with toluene poisoning. Hum Toxicol, 7, 357–61. Bracamonte, E. R., Kowalewska, J., Starr, J., et al. (2006). Iatrogenic phospholipidosis mimicking Fabry disease. Am J Kidney Dis, 48, 844–50. Brik, R., Magen, D., Ben-Yzhak, O., et al. (1998). Isoniazid-induced crescentic glomerulonephritis in a child with a positive tuberculin skin test. Am J Nephrol, 18, 430–2. Chang, C. and Gershwin, M. E. (2011). Drug-induced lupus erythematosus: incidence, management and prevention. Drug Saf, 34, 357–74. Choi, H. K., Merkel, P. A., and Niles, J. L. (1998). ANCA-positive vasculitis associated with allopurinol therapy. Clin Exp Rheumatol, 16, 743–4. Choi, H. K., Merkel, P. A., Walker, A. M., et al. (2000). Drug-associated antineutrophil cytoplasmic antibody-positive vasculitis: prevalence among patients with high titers of antimyeloperoxidase antibodies. Arthritis Rheum, 43, 405–13. Clive, D. M. and Stoff, J. S. (1984). Renal syndromes associated with nonsteroidal antiinflammatory drugs. N Engl J Med, 310, 563–72. Dai, C., Stolz, D. B., Kiss, L. P., et al. (2009). Wnt/beta-catenin signaling promotes podocyte dysfunction and albuminuria. J Am Soc Nephrol, 20, 1997–2008. Den Broeder, A. A., Assmann, K. J., van Riel, P. L., et al. (2003). Nephrotic syndrome as a complication of anti-TNFalpha in a patient with rheumatoid arthritis. Neth J Med, 61, 137–41. Deray, G., Martinez, F., Katlama, C., et al. (1989). Foscarnet nephrotoxicity: mechanism, incidence and prevention. Am J Nephrol, 9, 316–21. Desikan, R., Veksler, Y., Raza, S., et al. (2002). Nephrotic proteinuria associated with high-dose pamidronate in multiple myeloma. Br J Haematol, 119, 496–9. Dizer, U., Beker, C. M., Yavuz, I., et al. (2003). Minimal change disease in a patient receiving IFN-alpha therapy for chronic hepatitis C virus infection. J Interferon Cytokine Res, 23, 51–4. Dobre, M., Wish, J., and Negrea, L. (2009). Hydralazine-induced ANCA-positive pauci-immune glomerulonephritis: a case report and literature review. Ren Fail, 31, 745–8. Dogan, E., Ghanta, M., and Tanriover, B. (2011). Collapsing glomerulopathy in a renal transplant recipient: potential molecular mechanisms. Ann Transplant, 16, 113–16. Elkayam, O., Levartovsky, D., Brautbar, C., et al. (1998). Clinical and immunological study of 7 patients with minocycline-induced autoimmune phenomena. Am J Med, 105, 484–7. Eremina, V., Jefferson, J. A., Kowalewska, J., et al. (2008). VEGF inhibition and renal thrombotic microangiopathy. N Engl J Med, 358, 1129–36. Eremina, V., Sood, M., Haigh, J., et al. (2003). Glomerular-specific alterations of VEGF-A expression lead to distinct congenital and acquired renal diseases. J Clin Invest, 111, 707–16. Fellner, S. K. (1985). Piroxicam-induced acute interstitial nephritis and minimal-change nephrotic syndrome. Am J Nephrol, 5, 142–3. Fogo, A. B. and Kon, V. (2010). The glomerulus—a view from the inside—the endothelial cell. Int J Biochem Cell Biol, 42, 1388–97. Fogo, A. B. (2011). The targeted podocyte. J Clin Invest, 121, 2142–5. Fournier, A., Nony, A., and Rifard, K. (2009). [Antineutrophil cytoplasmic antibody associated vasculitis in a patient treated with adalimumab for a rheumatoid arthritis]. Nephrol Ther, 5, 652–7. Francis, K. L., Jenis, E. H., Jensen, G. E., et al. (1984). Gold-associated nephropathy. Arch Pathol Lab Med, 108, 234–8.

Chapter 82 

Franco, A. F., Martini, D., Abensur, H., et al. (2007). Proteinuria in transplant patients associated with sirolimus. Transplant Proc, 39, 449–52. Friedman, E. A. and Tao, T. K. (1995). Disappearance of uremia due to heroin-associated nephropathy. Am J Kidney Dis, 25, 689–93. George, B. A., Zhou, X. J., and Toto, R. (2007). Nephrotic syndrome after bevacizumab: case report and literature review. Am J Kidney Dis, 49, e23–9. George, C. R. (2011). Mercury and the kidney. J Nephrol, 24 Suppl 17, S126–32. Gödel, M., Hartleben, B., Herbach, N., et al. (2011). Role of mTOR in podocyte function and diabetic nephropathy in humans and mice. J Clin Invest, 121, 2197–209. Goldfarb, D. S. and Coe, F. L. (1998). Foscarnet crystal deposition and renal failure. Am J Kidney Dis, 32, 519–20. Hall, C. L., Fothergill, N. J., Blackwell, M. M., et al. (1987). The natural course of gold nephropathy: long term study of 21 patients. Br Med J (Clin Res Ed), 295, 745–8. Hall, C. L., Jawad, S., Harrison, P. R., et al. (1988). Natural course of penicillamine nephropathy: a long term study of 33 patients. Br Med J (Clin Res Ed), 296, 1083–6. Hanson, B., D’Hondt, A., Depierreux, M., et al. (1996). Nephrotic Syndrome after Norfloxacin. Nephron, 74, 446. Herlitz, L. C., Markowitz, G. S., Farris, A. B., et al. (2010). Development of focal segmental glomerulosclerosis after anabolic steroid abuse. J Am Soc Nephrol, 21, 163–72. Herve, J. P., Leguy, P., Cledes, J., et al. (1980). [Nephrotic syndrome with minimal glomerular lesions during treatment with D-penicillamine]. Nouv Presse Med, 9, 2847. Hill, G. S. (1986). Drug-associated glomerulopathies. Toxicol Pathol, 14, 37–44. Hoorntje, S. J., Weening, J. J., Kallenberg, C. G., et al. (1979). Serum-sickness-like syndrome with membranous glomerulopathy in patient on captopril. Lancet, 2, 1297. Inoki, K., Mori, H., Wang, J., et al. (2011). mTORC1 activation in podocytes is a critical step in the development of diabetic nephropathy in mice. J Clin Invest, 121, 2181–96. Izzedine, H., Boostandoot, E., Spano, J. P., et al. (2009). Temsirolimus-induced glomerulopathy. Oncology, 76, 170–2. Izzedine, H., Brocheriou, I., Becart, J., et al. (2007). Probenecidinduced membranous nephropathy. Nephrol Dial Transplant, 22, 2405–6. Izzedine, H., Launay-Vacher, V., Bourry, E., et al. (2006). Drug-induced glomerulopathies. Exp Opin Drug Saf, 5, 95–106. Izzedine, H., Massard, C., Spano, J. P., et al. (2010). VEGF signalling inhibition-induced proteinuria: Mechanisms, significance and management. Eur J Cancer, 46, 439–48. Jaffe, J. A. and Kimmel, P. L. (2006). Chronic nephropathies of cocaine and heroin abuse: a critical review. Clin J Am Soc Nephrol, 1, 655–67. Jhaveri, K. D., Schatz, J. H., Young, J. W., et al. (2008). Sirolimus (rapamycin) induced proteinuria in a patient undergoing allogeneic hematopoietic stem cell transplant. Transplantation, 86, 180–1. Johnson, D. H., Fehrenbacher, L., Novotny, W. F., et al. (2004). Randomized phase II trial comparing bevacizumab plus carboplatin and paclitaxel with carboplatin and paclitaxel alone in previously untreated locally advanced or metastatic non-small-cell lung cancer. J Clin Oncol, 22, 2184–91. Kallenberg, C. G., Hoorntje, S. J., Smit, A. J., et al. (1982). Antinuclear and antinative DNA antibodies during captopril treatment. Acta Med Scand, 211, 297–300. Kallenberg, C. G., van der Laan, S., and de Zeeuw, D. (1981). Captopril and the immune system. Lancet, 2, 92. Kalra, A., Yokogawa, N., Raja, H., et al. (2010). Hydralazine-induced pulmonary-renal syndrome: a case report. Am J Ther, 19(4), e136–8. Kambham, N., Markowitz, G. S., Valeri, A. M., et al. (2001). Obesity-related glomerulopathy: an emerging epidemic. Kidney Int, 59, 1498–509.

drug-induced and toxic glomerulopathies

Katz, W. A., Blodgett, R. C., Jr., and Pietrusko, R. G. (1984). Proteinuria in gold-treated rheumatoid arthritis. Ann Intern Med, 101, 176–9. Kitiyakara, C., Eggers, P., and Kopp, J. B. (2004). Twenty-one-year trend in ESRD due to focal segmental glomerulosclerosis in the United States. Am J Kidney Dis, 44, 815–25. Kobayashi, S., Nagase, M., Honda, N., et al. (1984). Glomerular alterations in uranyl acetate-induced acute renal failure in rabbits. Kidney Int, 26, 808–15. Kunin, M., Kopolovic, J., Avigdor, A., et al. (2004). Collapsing glomerulopathy induced by long-term treatment with standard-dose pamidronate in a myeloma patient. Nephrol Dial Transplant, 19, 723–6. Kunin, M., Kopolovic, J., Avigdor, A., et al. (2004). Collapsing glomerulopathy induced by long-term treatment with standard-dose pamidronate in a myeloma patient. Nephrol Dial Transplant, 19, 723–6. Lee, V. W. and Harris, D. C. (2011). Adriamycin nephropathy: a model of focal segmental glomerulosclerosis. Nephrology (Carlton), 16, 30–8. Letavernier, E., Bruneval, P., Mandet, C., et al. (2007). High sirolimus levels may induce focal segmental glomerulosclerosis de novo. Clin J Am Soc Nephrol, 2, 326–33. Letavernier, E., Bruneval, P., Vandermeersch, S., et al. (2009). Sirolimus interacts with pathways essential for podocyte integrity. Nephrol Dial Transplant, 24, 630–8. Letavernier, E., Pe’raldi, M. N., Pariente, A., et al. (2005). Proteinuria following a switch from calcineurin inhibitors to sirolimus. Transplantation, 80, 1198–203. Leung, A. C., McLay, A., Dobbie, J. W., et al. (1985). Phenylbutazone-induced systemic vasculitis with crescentic glomerulonephritis. Arch Intern Med, 1985; 145, 685–7. Li, S. J., Zhang, S. H., Chen, H. P., et al. (2010). Mercury-induced membranous nephropathy: clinical and pathological features. Clin J Am Soc Nephrol, 5, 439–44. Lieberthal, W., Fuhro, R., Andry, C. C., et al. (2001). Rapamycin impairs recovery from acute renal failure: role of cell-cycle arrest and apoptosis of tubular cells. Am J Physiol Renal Physiol, 281, F693–706. Madhrira, M. M., Mohan, S., Markowitz, G. S., et al. (2009). Acute bilateral renal infarction secondary to cocaine-induced vasospasm. Kidney Int, 76, 576–80. Manjunath, V., Moeckel, G. W., Mahnensmith, R., et al. (2011). Proteinuria and glomerular injury associated with the anti-angiogenesis drug Vargatef™. NDT Plus, 4(6), 430–3. Markowitz, G. S., Appel, G. B., Fine, P. L., et al. (2001). Collapsing focal segmental glomerulosclerosis following treatment with high-dose pamidronate. J Am Soc Nephrol, 12, 1164–72. Markowitz, G. S., Falkowitz, D. C., Isom, R., et al. (2003b). Membranous glomerulopathy and acute interstitial nephritis following treatment with celecoxib. Clin Nephrol, 59, 137–42. Markowitz, G. S., Fine, P. L., Stack, J. I., et al. (2003a). Toxic acute tubular necrosis following treatment with zoledronate (Zometa). Kidney Int, 64, 281–9. Markowitz, G. S., Nasr, S. H., Stokes, M. B., et al. (2010). Treatment with IFN-{alpha}, -{beta}, or -{gamma} is associated with collapsing focal segmental glomerulosclerosis. Clin J Am Soc Nephrol, 5, 607–15. Markowitz, G. S., Radhakrishnan, J., Kambham, N., et al. (2000). Lithium nephrotoxicity: a progressive combined glomerular and tubulointerstitial nephropathy. J Am Soc Nephrol, 11, 1439–48. Maurice-Estepa, L., Daudon, M., Katlama, C., et al. (1998). Identification of crystals in kidneys of AIDS patients treated with foscarnet. Am J Kidney Dis, 32, 392–400. Maynard, S. E., Min, J. Y., Merchan, J., et al. (2003). Excess placental soluble fms-like tyrosine kinase 1 (sFlt1) may contribute to endothelial dysfunction, hypertension, and proteinuria in preeclampsia. J Clin Invest, 111, 649–58. McLaughlin, K., Gholoum, B., Guiraudon, C., et al. (1998). Rapid development of drug-induced lupus nephritis in the absence of extrarenal disease in a patient receiving procainamide. Am J Kidney Dis, 32, 698–702. Miller, K. D., Chap, L. I., Holmes, F. A., et al. (2005). Randomized phase III trial of capecitabine compared with bevacizumab plus capecitabine in

663

664

Section 3  

the patient with glomerular disease

patients with previously treated metastatic breast cancer. J Clin Oncol, 23, 792–9. Miura, N., Mizuno, N., Aoyama, R., et al. (2009). Massive proteinuria and acute renal failure after oral bisphosphonate (alendronate) administration in a patient with focal segmental glomerulosclerosis. Clin Exp Nephrol, 13, 85–8. Molnar, T., Farkas, K., Nagy, F., et al. (2010). Sulfasalazine-induced nephrotic syndrome in a patient with ulcerative colitis. Inflamm Bowel Dis, 16, 552–3. Muller-Deile, J. and Schiffer, M. (2011). Renal involvement in preeclampsia: similarities to VEGF ablation therapy. J Pregnancy, 2011, 176973. Nadasdy, T. and Racusen, L. (1998). Renal injury caused by therapuetic and diagnostic agents and abuse of analgesics and narcotics. In J. C. Jennette, J. L. Olson, M. M. Schwartz, et al. (eds.) Heptinstall’s Pathology of the Kidney, pp. 811–61. Philadelphia: Lippincott-Raven. Nakao, K., Sugiyama, H., Makino, E., et al. (2002). Minimal change nephrotic syndrome developing during postoperative interferon-beta therapy for malignant melanoma. Nephron, 90, 498–500. Nanke, Y., Akama, H., Terai, C., et al. (2000). Rapidly progressive glomerulonephritis with D-penicillamine. Am J Med Sci, 320, 398–402. Neau-Cransac, M., Moreau, K., Deminiere, C., et al. (2009). Decrease in sirolimus-induced proteinuria after switch to everolimus in a liver transplant recipient with diabetic nephropathy. Transpl Int, 22, 586–7. Neild, G. H., Gartner, H. V., and Bohle, A. (1979). Penicillamine induced membranous glomerulonephritis. Scand J Rheumatol Suppl, 1979, 79–90. Neradova, A., Stam, F., van den Berg, J. G., et al. (2009). Etanercept-associated SLE with lupus nephritis. Lupus, 18, 667–8. Ntoso, K. A., Tomaszewski, J. E., Jimenez, S. A., et al. (1986). Penicillamine-induced rapidly progressive glomerulonephritis in patients with progressive systemic sclerosis: successful treatment of two patients and a review of the literature. Am J Kidney Dis, 8, 159–63. Ogata, H., Kubo, M., Tamaki, K., et al. (1998). Crescentic glomerulonephritis due to rifampin treatment in a patient with pulmonary atypical mycobacteriosis. Nephron, 78, 319–22. Padiyar, A., Bodziak, K. A., Hricik, D. E., et al. (2010). Clinical predictors of proteinuria after conversion to sirolimus in kidney transplant recipients. Am J Transplant, 10, 310–4. Pascual, J., Torrealba, J., Myers, J., et al. (2007). Collapsing focal segmental glomerulosclerosis in a liver transplant recipient on alendronate. Osteoporos Int, 18, 1435–8. Perazella, M. A. and Markowitz, G. S. (2008). Bisphosphonate nephrotoxicity. Kidney Int, 74, 1385–93. Perlman, A. S., Kim, E. H., Kallakury, B., et al. (2007). Clinically significant proteinuria following the administration of sirolimus to renal transplant recipients. Drug Metab Lett, 1, 267–71. Perneger, T. V., Klag, M. J., and Whelton, P. K. (2001). Recreational drug use: a neglected risk factor for end-stage renal disease. Am J Kidney Dis, 38, 49–56. Phan, L., Coulomb, F., Boudon, M., et al. (1991). [Extramembranous glomerulonephritis induced by lithium]. Nephrologie, 12, 185–7. Ramos-Casals, M., Brito-Zeron, P., Munoz, S., et al. (2007). Autoimmune diseases induced by TNF-targeted therapies: analysis of 233 cases. Medicine (Baltimore), 86, 242–51. Rangan, G. K. (2006). Sirolimus-associated proteinuria and renal dysfunction. Drug Saf, 29, 1153–61. Rao, T. K., Nicastri, A. D., and Friedman, E. A. (1974). Natural history of heroin-associated nephropathy. N Engl J Med, 290, 19–23. Reynolds, L. R. and Bhathena, D. (1979). Nephrotic syndrome associated with methimazole therapy. Arch Intern Med, 139, 236–7. Rihova, Z., Honsova, E., Merta, M., et al. (2005). Secondary membranous nephropathy—one center experience. Ren Fail, 27, 397–402. Ruiz, J. C., Campistol, J. M., Sanchez-Fructuoso, A., et al. (2006). Increase of proteinuria after conversion from calcineurin inhibitor to sirolimus-based treatment in kidney transplant patients with chronic allograft dysfunction. Nephrol Dial Transplant, 21, 3252–7.

Schwimmer, J. A., Markowitz, G. S., Valeri, A. M., et al. (2003). Secondary focal segmental glomerulosclerosis in non-obese patients with increased muscle mass. Clin Nephrol, 60, 233–41. Sekhon, I., Munjal, S., Croker, B., et al. (2005). Glomerular tip lesion associated with nonsteroidal anti-inflammatory drug-induced nephrotic syndrome. Am J Kidney Dis, 46, e55–8. Sennesael, J., Van den Houte, K., and Verbeelen, D. (1986). Reversible membranous glomerulonephritis associated with ketoprofen. Clin Nephrol, 26, 213–5. Sethi, S., Sahani, M., and Oei, L. S. (2003). ANCA-positive crescentic glomerulonephritis associated with minocycline therapy. Am J Kidney Dis, 42, E27–31. Shapiro, K. S., Pinn, V. W., Harrington, J. T., et al. (1984). Immune complex glomerulonephritis in hydralazine-induced SLE. Am J Kidney Dis, 3, 270–4. Sheikh, T. K., Charron, R. C., and Katz, A. (1981). Renal manifestations of drug-induced systemic lupus erythematosus. Am J Clin Pathol, 75, 755–62. Simms, R., Kipgen, D., Dahill, S., et al. (2008). ANCA-associated renal vasculitis following anti-tumor necrosis factor alpha therapy. Am J Kidney Dis, 51, e11–4. Soo, Y. O., Chow, K. M., Lam, C. W., et al. (2003). A whitened face woman with nephrotic syndrome. Am J Kidney Dis, 41, 250–3. Stallone, G., Infante, B., Pontrelli, P., et al. (2011). Sirolimus and proteinuria in renal transplant patients: evidence for a dose-dependent effect on slit diaphragm-associated proteins. Transplantation, 91, 997–1004. Stokes, M. B., Erazo, M. C., and D’Agati, V. D. (2008). Glomerular disease related to anti-VEGF therapy. Kidney Int, 74, 1487–91. Stokes, M. B., Foster, K., Markowitz, G. S., et al. (2005). Development of glomerulonephritis during anti-TNF-alpha therapy for rheumatoid arthritis. Nephrol Dial Transplant, 20, 1400–6. Straathof-Galema, L., Wetzels, J. F., Dijkman, H. B., et al. (2006). Sirolimus-associated heavy proteinuria in a renal transplant recipient: evidence for a tubular mechanism. Am J Transplant, 6, 429–33. Sugimoto, H., Hamano, Y., Charytan, D., et al. (2003). Neutralization of circulating vascular endothelial growth factor (VEGF) by anti-VEGF antibodies and soluble VEGF receptor 1 (sFlt-1) induces proteinuria. J Biol Chem, 278, 12605–8. Tam, V. K., Green, J., Schwieger, J., et al. (1996). Nephrotic syndrome and renal insufficiency associated with lithium therapy. Am J Kidney Dis, 27, 715–20. Tang, H. L., Chu, K. H., Mak, Y. F., et al. (2006). Minimal change disease following exposure to mercury-containing skin lightening cream. Hong Kong Med J, 12, 316–8. Traynor, A., Kuzel, T., Samuelson, E., et al. (1994). Minimal-change glomerulopathy and glomerular visceral epithelial hyperplasia associated with alpha-interferon therapy for cutaneous T-cell lymphoma. Nephron, 67, 94–100. Trolliet, P., Dijoud, F., Cotte, L., et al. (1995). Crescentic glomerulonephritis and crystals within glomerular capillaries in an AIDS patient treated with foscarnet. Am J Nephrol, 15, 256–9. Vicente-Vicente, L., Quiros, Y., Perez-Barriocanal, F., et al. (2010). Nephrotoxicity of uranium: pathophysiological, diagnostic and therapeutic perspectives. Toxicol Sci, 118, 324–47. Vogelbacher, R., Wittmann, S., Braun, A., et al. (2007). The mTOR inhibitor everolimus induces proteinuria and renal deterioration in the remnant kidney model in the rat. Transplantation, 84, 1492–9. Wagstaff, A. J. and Bryson, H. M. (1994). Foscarnet. A reappraisal of its antiviral activity, pharmacokinetic properties and therapeutic use in immunocompromised patients with viral infections. Drugs, 48, 199–226. Warren, G. V., Korbet, S. M., Schwartz, M. M., et al. (1989). Minimal change glomerulopathy associated with nonsteroidal antiinflammatory drugs. Am J Kidney Dis, 13, 127–30. Wojciechowski, D., Kallakury, B., and Nouri, P. (2008). A case of cocaine-induced acute interstitial nephritis. Am J Kidney Dis, 52, 792–5.

Chapter 82 

Wolters, J., Frederik, P., van Rie, H., et al. (1987). Minimal change nephropathy during gold treatment. A case with unusual histopathological and immunopathological features. Neth J Med, 31, 234–40. Wood, I. K., Parmelee, D. X., and Foreman, J. W. (1989). Lithium-induced nephrotic syndrome. Am J Psychiatry, 146, 84–7. Woywodt, A., Hellweg, S., Schwarz, A., et al. (2007). A wild zebra chase. Nephrol Dial Transplant, 22, 3074–7. Wu, S., Kim, C., Baer, L., et al. (2010). Bevacizumab increases risk for severe proteinuria in cancer patients. J Am Soc Nephrol, 21, 1381–9.

drug-induced and toxic glomerulopathies

Yu, F., Chen, M., Gao, Y., et al. (2007). Clinical and pathological features of renal involvement in propylthiouracil-associated ANCA-positive vasculitis. Am J Kidney Dis, 49, 607–14. Zanetta, G., Maurice-Estepa, L., Mousson, C., et al. (1999). Foscarnet-induced crystalline glomerulonephritis with nephrotic syndrome and acute renal failure after kidney transplantation. Transplantation, 67, 1376–8. Zech, P., Colon, S., Labeeuw, M., et al. (1979). Nephrotic syndrome in procainamide induced lupus nephritis. Clin Nephrol, 11, 218–21.

665

SECTION 4

The patient with interstitial disease

83 Acute tubulointerstitial nephritis: overview  669 Richard Baker

84 Drug-induced acute tubulointerstitial nephritis  678 Hassan Izzedine and Victor Gueutin

85 Other toxic acute tubulointerstitial nephritis  687 Benjamin J. Freda and Gregory L. Braden

86 Chronic tubulointerstitial nephritis: overview  690 Adalbert Schiller, Adrian Covic, and Liviu Segall

87 Drug-induced chronic tubulointerstitial nephritis  695 Hassan Izzedine and Victor Gueutin

88 Heavy metal-induced tubulointerstitial nephritis  702 Patrick C. D’Haese, Benjamin A. Vervaet, and Anja Verhulst

89 Aristolochic acid nephropathy caused by ingestion of herbal medicinal products  709 M. Refik Gökmen and Graham M. Lord

90 Balkan endemic nephropathy  714 Milan Radović and Adalbert Schiller

91 Radiation nephropathy  719 Lisa M. Phipps and David C. H. Harris

92 Urate nephropathy  724 Duk-Hee Kang and Mehmet Kanbay

93 Immune-mediated tubulointerstitial nephritis  726 Liviu Segall and Adrian Covic

CHAPTER 83

Acute tubulointerstitial nephritis: overview Richard Baker Introduction Acute tubulointerstitial nephritis (ATIN) is a clinical syndrome, usually associated with the development of acute kidney injury (AKI), which is characterized by the presence of inflammatory cells (and often oedema) within the renal interstitium. Some of these cells may cross the tubular basement membrane (TBM) to invade the tubules, resulting in tubulitis. The glomeruli and blood vessels are usually unaffected or only minimally abnormal. Notably, this definition of ATIN excludes both pyelonephritis, due to direct bacterial invasion (see Chapter 177), and AKI secondary to glomerular or vascular diseases, which may both display prominent interstitial infiltrates (see Section 11). Since ATIN is a morphologically defined process, the diagnosis can only be confirmed by renal biopsy. The characteristic interstitial and tubular inflammation do not always leads to AKI, but sometimes to milder forms of renal disease such as asymptomatic urinary abnormalities. The exact incidence of ATIN is difficult to define, since indications for performing a renal biopsy vary according to local practice and, furthermore, patients are often treated presumptively, without a biopsy-based diagnosis. In a study of Finnish army recruits who underwent biopsy for the evaluation of urinary abnormalities, the incidence of ATIN was approximately 1% (Pettersson et al., 1984). However, in another series of 109 renal biopsies performed in a large centre because of unexplained AKI, this incidence reached 27% (Farrington et al., 1989). A similar study from the United States described 259 older patients (age ≥ 60 years) who were biopsied for AKI and ATIN was found to be responsible for 18.6% of cases (Haas et al., 2000). Data extracted from two large European registries suggests that the diagnosis of ATIN constitutes approximately 11% of renal biopsies performed for the evaluation of AKI (Schena, 1997; Lopez-Gomez et al., 2008). It is in this latter context that ATIN is particularly important, since it represents a relatively common and potentially reversible cause of AKI, which requires prompt and specific treatment. There is some evidence that early therapy leads to a quicker and more complete recovery of renal function. ATIN is sometimes called ‘acute interstitial nephritis’; however, since the tubulitis is often a prominent feature, the term ‘ATIN’ is preferred. This entity was first described by Councilman in 1898 at Harvard Medical School, in a series of 42 autopsies from patients dead of diphtheria and scarlet fever, all of whom had typical renal ‘cellular and fluid exudation in the interstitial tissue’ (Councilman, 1898). The inflammation within the kidneys was characterized by an exudate that was not purulent and the tissue itself was sterile. He speculated that the cells might accumulate because ‘soluble

substances may exert a positive chemotaxis’. Crucially, he made the observation that the tissue damage was not due to direct microbial invasion, but secondary to an allergic-type phenomenon. In 1946 a series of patients with similar histological findings was described, all of whom had been treated with sulphonamides, but it was not clear at the time whether the inciting agent was the drug itself or the underlying infection (More et al., 1946). During the 1960s the first reports emerged of ATIN associated with penicillins, particularly methicillin (Hewitt et al., 1961; Baldwin et al., 1968), and over the ensuing decades links with other drugs (e.g. phenindione, rifampicin, and azathioprine) were established, confirming drug allergy as a common cause of ATIN (Hewitt et al., 1961; McMenamin et al., 1976; Ditlove et al., 1977; Nolan and Abernathy, 1977; Galpin et al., 1978; Linton et al., 1980). With the burgeoning use of percutaneous renal biopsy throughout the 1950s and 1960s, ATIN became increasingly recognized as a cause of AKI. The exact pathophysiology of ATIN is unclear, but there are a number of observations, predominantly derived from patients with drug-related ATIN, which suggest that the renal injury is immune-mediated in an allergic-type reaction.

Aetiology Drugs

Approximately 80% of ATIN cases are related to the administration of drugs (Chapter 84), with antibiotics and non-steroidal anti-inflammatory drugs (NSAIDs) being the two main culprits (see Tables 83.1 and 83.2) (Rossert, 2001; Izzedine et  al., 2007; Perazella and Markowitz, 2010; Praga and Gonzalez, 2010). However, the list of agents that may cause ATIN continues to grow, and ongoing vigilance is required, as new drugs are constantly being involved, including levetiracetam, etanercept, sorafenib, and highly active antiretroviral therapy (HAART) among the most recent ones (Izzedine et al., 2007; Said et al., 2007; Schmid et al., 2007; Sugimoto et al., 2008). Although reported series are currently dominated by drug-related ATIN (see Chapter 84), the causative agent is often difficult to pinpoint precisely, for a number of different reasons. Firstly, antibiotics are often being used to treat infections and, in such cases, it is not always clear, in retrospect, whether it was the infection itself or the antibiotic that caused the ATIN. Secondly, although the latent period between exposure to a neoantigen and the development of a primary immune response is usually approximately 10 days, there are numerous exceptions in clinical practice. For example ATIN may develop after short exposure to drugs such as rifampicin, especially on re-exposure (Covic et al., 1998, 2004; Muthukumar et al., 2002; Schubert et al.,

670

Section 4  

the patient with interstitial disease

Table 83.1  Agents causing ATIN from seven published series including 237 patients Causative agent

%

Common examples

Antibiotics

39.7

β-lactams, sulphonamides, rifampicin, macrolides, aminoglycosides, chloramphenicol, quinolones, tetracyclines

NSAIDs

22.4

Indomethacin, ketoprofen, fenoprofen, naproxen, ibuprofen, diclofenac

Other drugs

16.8

Diuretics, H2-antagonists, proton pump inhibitors, anticonvulsants, allopurinol, analgesics, warfarin

Infections

5.9

Pneumococcus, Streptococcus, Mycobacteria, Leptospira, Legionella, syphilis

Other

15.2

Idiopathic, TINU, sarcoidosis

2010), and after much longer periods, between 6 and 18 months after NSAIDs (Pirani et al., 1987). Finally, patients suffering from ATIN are often on multiple drugs and, in this case, clinicians usually attribute causality to the most commonly associated agent, thus generating a self-fulfilling prophecy, whereby NSAIDs and antibiotics will continue to be the commonest drugs associated with ATIN. Drug-related ATIN is covered in more detail in Chapter 84.

Infection Infectious disease constitutes the second major group of causes for ATIN. As mentioned above, ATIN was originally described in association with diphtheria and scarlet fever. Once again, the list of causal agents is a long one, with new associations continually evolving. For example, there have been recent descriptions of ATIN associated with hydatid disease and histoplasmosis in native kidneys, and BK virus and adenovirus in renal allografts (Randhawa et al., 1999; Dall et al., 2008; Nickavar et al., 2011; Qian et al., 2011; Storsley et al., 2011). It should be emphasized that these infectious diseases cause interstitial and tubular tissue injury by an indirect mechanism, not by direct microbial invasion, as in ascending infection of the urinary tract (see Chapter 177). In ATIN the interstitium is sterile, as originally described by Councilman, the damage resulting from immunological and inflammatory mechanisms. The link between tuberculosis and ATIN is an intriguing one (see Chapter 196). Classically, renal tuberculosis has been associated with lower urinary tract calcification and scarring. Renal parenchymal involvement consists of caseating granulomas, and mycobacteria can often be cultured from early morning urine samples. More recently, histological changes of ATIN have been reported in patients with extrarenal tuberculosis (Chapagain et  al., 2011; Eastwood et al., 2011). In these cases, urine cultures did not grow mycobacteria and, where available, molecular testing of renal tissue (e.g. polymerase chain reaction) was negative. Since end-stage renal disease (ESRD), particularly due to interstitial nephritis, is overrepresented in immigrant Asian populations in the United Kingdom, it has been speculated that ATIN secondary to tuberculosis could evolve into a chronic form of nephropathy, contributing to the increased incidence of ESRD amongst these populations (Clark et al., 1993; Lightstone et al., 1995; Ball et al., 1997, 2001).

Table 83.2  Causes of ATIN Drug-related Antibiotics: β-lactams—methicillin, ampicillin, benzylpenicillin, flucloxacillin, cephalosporins Sulphonamides, co-trimoxazole Gentamicin Tetracycline Vancomycin, teicoplanin Quinolones—ciprofloxacin, levofloxacin Macrolides—erythromycin, clarithromycin, azithromycin Chloramphenicol Rifampicin, ethambutol, isoniazid Antivirals: Aciclovir, HAART (indinavir), NSAIDs: Fenoprofen, indomethacin, ketoprofen, naproxen, ibuprofen, diclofenac, phenylbutazone, tolmetin, aspirin, celecoxib, rofecoxib, and most others Antiulcer medications: Omeprazole, lansoprazole, famotidine, ranitidine, cimetidine Diuretics: Furosemide, thiazides, triamterene Anticonvulsants: Carbamazepine, phenytoin, levetiracetam, valproate Anticoagulants: Warfarin, phenindione Analgesics Others: Allopurinol, mesalazine, propranolol, amlodipine, azathioprine, etanercept, sorafenib, captopril, clofibrate, cocaine, creatine, diltiazem, pranlukast, propylthiouracil, quinine Infectious Bacterial: Streptococcus, Staphylococcus, Pneumococcus, Legionella, Corynebacterium diphtheriae, Yersinia, Brucella, Campylobacter, Escherichia coli, Salmonella, Mycobacterium tuberculosis Viral: Cytomegalovirus, Epstein–Barr virus, herpes simplex virus, Hantavirus, hepatitis A, hepatitis B, hepatitis C, HIV, measles, mumps, Polyoma (BK) virus, adenovirus Other: Leptospira, Treponema, Mycoplasma, Rickettsia, Toxoplasma, Chlamydia, Leishmania Systemic autoimmune disease Connective tissue disease Systemic lupus erythematosus, Sjögren syndrome, cryoglobulinaemia, ANCA-associated small vessel vasculitis Tubulointerstitial nephritis and uveitis (TINU) IgG4-related disease Sarcoidosis Idiopathic: With anti-TBM antibodies Without anti-TBM antibodies Secondary: Associated with glomerular disease Associated with light chain nephropathy Associated with vascular disease

Chapter 83 

Systemic diseases Numerous systemic diseases are also associated with ATIN, but this is usually secondary to extensive glomerular involvement. For example, ATIN is often present on biopsies of patients with systemic lupus erythematosus (SLE) (see Chapter 161) or in myeloma (see Chapter 153). Occasionally, isolated primary ATIN may occur in SLE with absent or only minor glomerular changes (Cunningham et al., 1978; Gur et al., 1987; Mori et al., 2005).

Sarcoidosis Renal involvement in sarcoidosis (see Chapter 156) has been described in a minority of patients, manifested by nephrocalcinosis and tubulointerstitial nephritis, with or without granulomas on renal biopsy (Bergner et al., 2003). Conversely, approximately 90% of patients with renal sarcoidosis will have extrarenal symptoms (Hannedouche et al., 1990; Robson et al., 2003; Rajakariar et al., 2006; Joss et al., 2007; Mahevas et al., 2009). ATIN has also been described in patients with Sjögren syndrome (see Chapter 93), although this disorder more commonly presents with chronic tubular dysfunction (Goules et al., 2000). ATIN associated with both sarcoidosis and Sjögren syndrome has been reported to show a good response to corticosteroids.

IgG4-related TINU IgG4-related TIN occurs in association with autoimmune pancreatitis. The association between autoimmune pancreatitis and hypergammaglobulinaemia was first described in 1961 and later on these antibodies have been characterized as of IgG4 isotype (Sarles et al., 1961). Furthermore, tissue deposition of IgG4-positive plasma cells has been demonstrated in a number of different organs (Cornell, 2010; Saeki et al., 2010). The condition is described further in Chapter 93. Studies suggest that most patients are men over the age of 50 and they present with either acute or chronically progressive renal failure, often with hypocomplementaemia. Approximately 80% of patients will have multiorgan involvement (especially lymphadenopathy, adenitis, and pancreatitis), 80% have radiological renal abnormalities (enlarged kidneys or patchy hypoattenuated lesions) and 80% have either total IgG or IgG4 raised serum levels (Raissian et al., 2011). Histological examination reveals a plasma cell-rich infiltrate, sometimes accompanied by prominent eosinophils. Diffuse interstitial fibrosis and deposition of immune complexes along the TBM is common. Immunostaining for IgG4 appears to be highly suggestive of the diagnosis and a good response to steroids has been described.

Tubulointerstitial nephritis with uveitis (TINU syndrome) TINU was first reported in 1975 as an association between ATIN and anterior uveitis, sometimes with additional bone marrow granulomas (Dobrin et al., 1975). A number of such patients have now been described (Mandeville et al., 2001). Although associations with both Chlamydia and Mycoplasma infections have been suggested, the aetiology remains obscure (Stupp et al., 1990). The uveitis may occur several weeks before or up to 3 months after the ATIN. The TINU syndrome commonly occurs in adolescence and early adulthood, with a preponderance of females (3:1). Patients generally suffer from weight loss, myalgia, fever, and anaemia. Elevated serum inflammatory markers are common and prolonged steroid therapy usually leads to improvement in both renal function and uveitis, though the latter may relapse (Rodriguez-Perez et al., 1995; Mackensen and Billing et al., 2009). Curiously, a relapse of TINU with both renal an ocular involvement has been described

acute tubulointerstitial nephritis: overview

after renal transplantation, despite ongoing immunosuppression (Onyekpe et al., 2011).

Idiopathic Occasionally ATIN will occur without any precipitant or associated factors and these cases are termed idiopathic. Only rarely will such cases have evidence of anti-TBM antibodies (Bergstein and Litman, 1975; Rakotoarivony et al., 1981).

Pathogenesis The pathogenesis of ATIN is poorly understood, but there are a number of features that suggest that the disease is triggered by an immune hypersensitivity reaction to either drugs or infectious agents: 1. Drug reactions are idiosyncratic and occur only in a small percentage of patients. 2. Older series that were dominated by antibiotic-associated ATIN commonly described clinical features associated with allergic-type phenomena, for example, arthralgia, rash, fever, and eosinophilia (Ditlove et al., 1977; Galpin et al., 1978; Linton et al., 1980; Rossert, 2000). These associated manifestations are much less frequently seen now. 3. ATIN often occurs within 10–14 days of antigen exposure, the archetypal timing of a primary immune response. 4. Inadvertent rechallenge with a drug that previously caused ATIN leads to a rapid recrudescence of disease, suggesting a classical memory response (Sloth and Thomsen, 1971; Saltissi et al., 1979; Pusey et al., 1983; Covic et al., 2004; Schubert et al., 2010). 5. The histology of ATIN is dominated by lymphocyte infiltration, with the formation of granulomas in some cases, the hallmark of a delayed-type hypersensitivity (DTH) reaction. 6. In infection-related ATIN, the renal parenchyma is sterile, with relatively rare neutrophils, as originally described by Councilman (1898). 7. Some patients show a DTH-like reaction on intradermal injection of the offending drug (Border et al., 1974; Baldwin et al., 1977). 8. A proliferation of autologous lymphocytes in vitro after exposure to the culprit drug has been demonstrated (Joh et al., 1990; Shibasaki et al., 1991; Spanou et al., 2006). 9. Animal models of ATIN, like the kd/kd mouse, develop spontaneous interstitial nephritis, the expression of which is dependent on a finely tuned balance between nephritogenic and regulatory T cells (Neilson et al., 1984; Kelly and Neilson, 1987; Kelly, 1990). 10. Immunization of animals with renal proteins, such as the Tamm–Horsfall protein or components of the normal human TBM, like glycoprotein H3M-1 and TIN-antigen, can induce ATIN (Clayman et al., 1986; Neilson, 1989; Butkowski et al., 1990; Wilson, 1991). The nature of the actual antigens involved in clinical ATIN is unknown. Drugs themselves are relatively small molecules, which are capable of generating only weak immune responses. For this reason, it has been proposed that they may act as haptens to alter the immunogenicity of associated proteins. According to this model, drugs

671

672

Section 4  

the patient with interstitial disease

are filtered in the kidney and become associated either with residual renal proteins or with other antigens that have become entrapped in the kidney. This mechanism is supported by the occasional demonstration of anti-TBM antibodies in human drug-related disease (Border et al., 1974; Bergstein and Litman, 1975). Interestingly, in patients with rapidly progressive glomerulonephritis associated with anti-GBM antibodies, the subgroup who also develop anti-TBM antibodies have the most florid interstitial inflammation, suggesting a possible pathogenic role (Andres et al., 1978). The isolation of anti-TBM antibodies from patients with ATIN has led to the identification of a 48kD antigen, R3M-1, which is the target of these antibodies (Clayman et al., 1986). However, the evidence for a major humoral component in the pathogenesis of ATIN is lacking, since the vast majority of the patients have no anti-TBM antibodies, normal complement levels, and no evidence of immune deposits, either by immunofluorescence or electron microscopy. An alternative explanation could involve molecular mimicry, whereby an immune response to a pathogen or drug results in cross-reactivity against renal antigens. One experimental model of ATIN has shown nephritogenic T cells cross-reactive to heat shock proteins (Weiss et  al., 1994)  and another demonstrated the development of ATIN after immunization with Escherichia coli in adjuvant (Sherlock, 1977). The characteristic pathological features of ATIN are found in a number of different circumstances (e.g. allograft rejection and BK virus nephropathy), suggesting that the histological picture probably represents the outcome of a common downstream inflammatory pathway. It is certainly possible that mechanisms other than priming of the adaptive immune system exist to induce this pattern of inflammation. For example, it has been shown that highly conserved lipoproteins (LipL32) in the outer membrane proteins of pathogenic strains of Leptospira can trigger interstitial inflammation via Toll-like receptor-2 (TLR-2) on renal tubular cells (Yang et  al., 2002; Yang, 2007). Indeed, it has been demonstrated that renal inflammation can be triggered by both damage-associated molecular patterns (DAMPs) and pathogen-associated molecular patterns (PAMPs) via Toll-like and other receptors, thus triggering renal inflammation independent of upstream adaptive immune responses (Anders et al., 2004; Anders, 2010). Such a mechanism has been demonstrated in ATIN caused by BK virus in renal allografts (Anders et al., 2004; Ribeiro et al., 2012). It is likely that these mechanisms (and probably others) act in concert to bring about the histological picture that characterizes ATIN. Under the influence of chemotactic stimuli, inflammatory leucocytes, which are predominantly mononuclear in nature, leave the circulation from the peritubular capillaries. They cross the vascular endothelium and capillary basement membrane to infiltrate the interstitial space, pushing apart the tubules which are normally juxtaposed. From here, some cells continue to cross the TBM and inflict injury to the tubular epithelial cells (tubulitis). Following the acute phase, the inflammation may resolve, with return to baseline renal function; however, in some cases the inflammatory infiltrates persist and evolve into a more fibrotic phenotype, associated with deposition of extracellular matrix, leading to progressive renal disease (Neilson, 2006; Zeisberg and Neilson, 2010).

Fig. 83.1  Tubulitis in ATIN.

back, but in fact there are two types of cells normally present between the tubules and the peritubular capillaries. These are renal fibroblasts, which provide the extracellular skeleton upon which the tubules and capillaries are suspended, and dendritic cells, which are positive for class II major histocompatibility complex molecules and are thought to play a role in antigen presentation (Kaissling and Le Hir, 2008). Macroscopically, in ATIN the kidneys are often oedematous and enlarged. Under the microscope, the normal renal architecture is disturbed by an infiltration of inflammatory cells in the renal interstitium, which cross the TBM to invade the tubules. This process, termed ‘tubulitis’, may cause breaks in the TBM, with necrosis and atrophy of tubular epithelial cells (Fig. 83.1) Thus, the tubular epithelial cells may appear flattened. This inflammatory process is often focal and may be accompanied by interstitial oedema. Mononuclear cells, specifically lymphocytes and macrophages, usually dominate the picture and create new space between the tubules. Immunophenotyping studies have suggested that in most cases the predominant mononuclear cell is the CD4+ T cell, but sometimes CD8+ cells may be preeminent (Bender et al., 1984). Neutrophil infiltrates may be present, but if prominent they arouse suspicion of pyelonephritis. When drug allergy is implicated, there may be large numbers of eosinophils (Fig. 83.2). If the inflammation persists, then macrophages and

Pathology The normal renal tubulointerstitium consists of tubular epithelial cells attached to the TBM and surrounded by small peritubular capillaries. The tubules appear on light microscopy to be virtually back to

Fig. 83.2  Interstitial infiltrate in ATIN.

Chapter 83 

Fig. 83.3  Granuloma in ATIN.

monocytes often prevail and eventually plasma cells and histiocytes may become evident. The inflammatory process is finely balanced between repair and fibrosis (Tanaka and Nangaku, 2011). In chronic forms of the disease and with certain agents interstitial fibrosis may ensue, with proliferation of extracellular matrix and ascendancy of fibroblasts. Uncommonly, granulomas may develop, a process that is associated with certain drug reactions, tuberculosis, and sarcoidosis (Fig. 83.3) (Magil, 1983; Mignon et al., 1984; Bijol et al., 2006; Joss et al., 2007). Electron microscopy, immunohistochemical, and immunofluorescence studies are rarely helpful, except in rare cases of linear staining secondary to anti-TBM disease. The glomeruli are typically normal or have only minor abnormalities. Examination of the blood vessels may reveal evidence of ageing or hypertension, but nothing specific to ATIN. Overall, the morphology is non-specific with regards to the underlying cause and, while certain findings (e.g. eosinophils) may suggest certain aetiological agents, it is not possible to be specific.

Clinical manifestations The renal disease associated with ATIN is variable and may range from completely asymptomatic to fulminant and irreversible AKI. However, it usually presents with slowly progressive AKI (over a few weeks). Epidemiological evidence suggests that, when there is no obvious precipitant for AKI and an ultrasound reveals normal sized kidneys, then there is a high likelihood that ATIN is the underlying cause (10–25% of cases). The first published reports of ATIN were dominated by cases related to the use of penicillin and many of these patients had allergic-type clinical features, such as rash, arthralgia, fever, and eosinophilia (Baldwin et al., 1968; Ditlove et al., 1977; Nolan et al., 1977; Galpin et al., 1978). In recent years these features have become less common (as shown in Table 83.3), particularly in cases associated with NSAID usage (Clive and Stoff, 1984; Baker and Pusey, 2004). The classical allergic triad of fever, arthralgia, and rash is now only seen in 10% of patients. Other rare symptoms consistent with a hypersensitivity reaction that have been described include haemolysis, hepatitis and elevated serum IgE levels. Clinical findings are inconsistent, but the history and physical examination can be guided by certain fundamental principles.

acute tubulointerstitial nephritis: overview

A thorough history is mandatory, including enquiry about recent episodes of infection or drug exposure. Details should be sought regarding over-the-counter medications such as NSAID creams and herbal remedies. When discussing drug exposure it should be borne in mind that drug-related ATIN is an idiosyncratic reaction and not dose-related. It can occur at any time after starting drug treatment and prior tolerance of a medication does not necessarily preclude its involvement; this is particularly true for ATIN associated with NSAIDs (Kleinknecht, 1995). Symptoms such as malaise, anorexia, and fatigue are common, particularly in severe AKI. Occasionally, flank pain may be described, presumably due to renal capsular stretching. There may symptoms related to extrarenal disease, especially if the ATIN is part of a multisystem disease (e.g. TINU, IgG4-related ATIN, or sarcoidosis). The clinical examination will usually be normal, although rarely there may be evidence of a rash, swollen joints or uveitis. Most patients with ATIN are normotensive and have no peripheral oedema. Investigations will usually yield a raised serum creatinine and biochemical evidence of tubular dysfunction may be present (e.g. renal tubular acidosis or Fanconi syndrome). The renal function is impaired severely enough to require dialysis in approximately 40% of patients. There may be a low fractional excretion of sodium. Eosinophilia is present in approximately one-third of cases and serum inflammatory markers may be raised. Urinalysis usually reveals low-grade proteinuria (< 2 g/24 hours), but occasionally nephrotic-range proteinuria may be present, indicating the coexistence of significant glomerular disease. This may occur in the elderly, with the occasional association of anti-TBM disease with membranous glomerulopathy or in some NSAID- and methicillin-related cases (Nolan and Abernathy, 1977; Clive and Stoff, 1984; Porile et al., 1990; Katz et al., 1992; Haas et al., 2000). Microscopic haematuria is common, but macroscopic haematuria is unusual and suggests an alternative diagnosis. Sediment-free urine does not exclude the diagnosis of ATIN. White blood cells Table 83.3  Clinical features of ATINa AKI

100%

AKI requiring renal replacement therapy

40%

Arthralgia

45%

Rash

18%

Fever

32%

‘Allergic triad’ (fever, rash and arthralgia)

10%

Eosinophilia

31%

Microscopic haematuria

67%

Macroscopic haematuria

5%

Leucocyturia

82%

Non-nephrotic-range proteinuria

93%

Nephrotic-range proteinuria

2.5%

Nephrotic syndrome

0.8%

a Data pooled from several studies (Buysen et al., 1990; Schwarz et al., 2000; Baker and

Pusey, 2004; Clarkson et al., 2004; Gonzalez et al., 2008).

673

674

Section 4  

the patient with interstitial disease

and white cell casts may also be present in the urine and examination for eosinophiluria (eosinophils > 1% of total urine white cells) with Hansel’s stain is often positive. This test is not routinely performed in many centres and detailed analysis demonstrates a low sensitivity and low positive predictive value for the diagnosis of ATIN (Nolan et al., 1986; Landais et al., 1987; Ruffing et al., 1994; Fletcher, 2008; Kaye and Gagnon, 2008). It is possible that urinary biomarkers may prove useful in the future but they are not currently validated for the diagnosis of ATIN (Waanders et al., 2010; Chen et al., 2011).

Diagnosis Enlarged renal bipolar length is a non-specific finding in ATIN, but it is not sensitive enough to aid diagnosis. Gallium-67 scintigraphy was originally described as a useful diagnostic test in ATIN, the tracer being taken up by the mononuclear phagocytic cells within the renal interstitium (Wood et al., 1978; Linton et al., 1985). However, subsequent reports have not confirmed these observations and this investigation is no longer considered useful (Graham et al., 1983; Kodner and Kudrimoti, 2003). The lack of any non-invasive test with a satisfactory sensitivity and specificity to make the diagnosis of ATIN means that a percutaneous renal biopsy remains the gold standard for diagnosis. If the clinical situation allows, an early biopsy is recommended, since there is some evidence that early treatment ensures both a more rapid and more complete renal recovery (see below).

Prognosis Full recovery from AKI caused by ATIN is the usual, but by no means universal, outcome. Historical data demonstrates that the majority of patients with methicillin-induced ATIN recovered renal function after either discontinuation of the drug or steroid therapy, but other reports suggest a lower proportion of patients recovering renal function after ATIN secondary to other drugs (Galpin et al., 1978; Schwarz et al., 2000). Acute dialysis is sometimes required, but only a few patients become dialysis-dependent (Laberke and Bohle, 1980; Kida et al., 1984). Renal function does not return to baseline in up to 40% of patients with ATIN, but the final creatinine level does not correlate with its peak value (Rossert, 2001). Attempts have been made to gain prognostic information from the renal biopsy, with inconsistent results. The degree of tubular atrophy predicted renal outcome in one series (Baker and Pusey, 2004), whereas some authors have reported that patchy cellular infiltration is associated with a better prognosis than diffuse disease (Laberke and Bohle, 1980; Kida et al., 1984; Chen et al., 2011). Other studies showed no correlation between the degree of cellular infiltration or tubulitis and outcome (Buysen et al., 1990; Ivanyi et al., 1996). The degree of interstitial fibrosis has been correlated to outcome in some studies (Bhaumik et al., 1996; Ivanyi et al., 1996; Gonzalez et al., 2008), but no such relationship has been found in others (Cheng et  al., 1989). These conflicting observations may be due to the patchy nature of the disease and the random sampling on renal biopsy. This was recognized over 100 years ago by Councilman, who commented on the local nature of the histological changes (Councilman, 1898). The infiltrate is generally most prominent at the corticomedullary junction, with the medulla being relatively spared. Some authors have linked the presence of

granulomas with a good response to steroids and a favourable outcome (Joss et al., 2007). It has previously been suggested that the long-term outcome is worse if renal failure lasts for > 3 weeks (Ditlove et al., 1977; Laberke and Bohle, 1980). However, this is clearly not useful prospectively. Two series have demonstrated worse outcomes with increasing age (Ditlove et al., 1977; Kida et al., 1984).

Management Most authors favour early treatment with steroids in most circumstances. Unfortunately, recommendations for the management are blighted by the absence of any prospective randomized controlled studies in the treatment of ATIN. The published literature consists entirely of relatively small retrospective and uncontrolled studies, usually extracted from a single centre. Interpretation is further complicated by the transformation in aetiology that has taken place over the last five decades. Patients with antibiotic-related ATIN, in particular methicillin, dominated early series, whereas contemporary series are more heterogeneous, although still dominated by drug allergy. The following discussion is an attempt to assimilate the evidence from these studies. Since ATIN is thought to be driven by a pathological response to an allergenic antigen, it follows that removal of the offending agent remains a cornerstone of treatment. Therefore, in drug-related ATIN it is incumbent upon the clinician to identify the most likely agent and stop it. In clinical practice the patient is often taking multiple drugs and deciding which of them to stop is based on probability derived from epidemiological data. It is perhaps prudent to stop all non-essential medications, at least in the short term. In the case of infection-related ATIN it is important to treat the primary infection, especially if it represents an indolent process such as tuberculosis. The nature of the inflammatory infiltrate and the putative allergic-type mechanism has led clinicians to use corticosteroids to treat ATIN for many years. However, the exact role of steroids in treatment remains to be defined. Several small uncontrolled series have suggested that the administration of steroids is superior to conservative therapy alone (Galpin et al., 1978; Laberke and Bohle, 1980; Pusey et al., 1983; Buysen et al., 1990). A retrospective study of 27 patients with ATIN, by Laberke et al., demonstrated that the seven patients who were treated with steroids had a significantly better renal outcome than those who were not (Laberke, 1980). Buysen et al. (1990) described 27 patients with ATIN, 17 of whom improved spontaneously with conservative measures and drug discontinuation. The remaining 10 showed further deterioration of renal function in the 2 weeks following admission, and were then treated with steroids. In all of these patients, renal function subsequently improved, returning to baseline in 6 weeks. More recently Clarkson et  al. described a retrospective series of 60 patients from a single centre, presenting between 1988 and 2001, in over 90% of whom drugs were invoked as the aetiological agents (44% NSAIDs) (Clarkson et al., 2004). Full follow-up data was available in 42 patients (60%), of whom 26 were treated with corticosteroids and 16 were not. Treatment was not uniform, but involved daily 500 mg methylprednisone intravenously for 2–4 days, followed by tapering oral steroids for a further 3–6 weeks. Two patients died, but the remaining patients had good short- and medium-term outcomes of renal function, although

Chapter 83 

a noteworthy proportion was left with chronic renal impairment. There was no significant difference between the patients who were treated with steroids and those who were not. It is important to note that patients in the steroid group were treated at a median of 4 weeks after the onset of symptoms and had a mean creatinine of 545 μmol/L (339–1110 μmol/L) at presentation. It has been argued that this cohort of patients may have been treated too late for steroids to make any significant difference (Praga and Gonzalez, 2010). In contrast, Gonzalez et al. have reported a retrospective series of 61 patients with drug-induced ATIN from multiple Spanish centres, presenting between 1975 and 2006 (Gonzalez et al., 2008). Fifty-six per cent of cases were ascribed to antibiotics and 37% to NSAIDs. Fifty-two patients (85%) were treated with cortico­ steroids. Treatment was not uniform, but most patients received methylprednisolone (250–500 mg daily for 3–4 consecutive days), followed by oral prednisone (1 mg/kg/day) tapered off over 8–12 weeks. The outcome at a median of 19 months was considerably better in the patients receiving steroid treatment, both in terms of requirement for chronic renal replacement therapy (3.8 vs 44%) and final serum creatinine (185 ± 185 vs 326 ± 255 μmol/L). There were no other significant differences at baseline between the two groups and the duration and doses of steroids were similar. However, those patients who experienced incomplete renal function recovery had a significantly longer interval between withdrawal of the offending drug and the start of steroid treatment (34 ± 17 vs 13 ± 10 days). This study suggests that steroids are indeed effective, but to gain the maximum benefit they need to be administered early. Given that up to half of the patients with ATIN will not regain their baseline renal function, most authors favour the early use of steroids (Baker and Pusey, 2004; Appel, 2008; Perazella and Markowitz, 2010; Praga and Gonzalez, 2010). Experimental data showing that fibrosis can develop within 7 days of an inflammatory stimulus also supports this approach (Neilson, 2006; Zeisberg and Neilson, 2010). Clearly, a multicentre prospective randomized trial of early versus late steroid treatment is required. Until such a study is performed, it is recommended to perform an early renal biopsy, if clinically safe, and start steroid therapy immediately following diagnosis, with a course that tapers over 8 to 12 weeks (Appel, 2008). The duration and strength of steroid treatment is also uncertain. Most use prednisolone at 1 mg/kg/day to a maximum of 60 mg/ day, tapering after 2 weeks, when the creatinine has started to fall and continuing for 2–3 months. Methylprednisolone has been used to initiate therapy in patients with severe renal impairment. Using pulsed methylprednisolone for a few days in the hospital setting may be more attractive in certain clinical situations, for example if there are questions about potential adherence or there are anxieties over the morbidities associated with steroids, such as psychosis or diabetes. Ciclosporin has been shown to ameliorate animal models of autoimmune ATIN and might be used as a second-line agent in humans (Shih et  al., 1988), although there is limited evidence that it may be effective (Zuliani et al., 2005). Both mycophenolate mofetil and cyclophosphamide have also been used as alternative agents with some success, predominantly in steroid-resistant cases (Preddie et al., 2006).

References Anders, H. J. (2010). Toll-like receptors and danger signaling in kidney injury. J Am Soc Nephrol, 21(8), 1270–4.

acute tubulointerstitial nephritis: overview

Anders, H. J., Banas, B., and Schlondorff, D. (2004). Signaling danger: toll-like receptors and their potential roles in kidney disease. J Am Soc Nephrol, 15(4), 854–67. Andres, G., Brentjens, J., Kohli, R., et al. (1978). Histology of human tubulo-interstitial nephritis associated with antibodies to renal basement membranes. Kidney Int, 13(6), 480–91. Appel, G. B. (2008). The treatment of acute interstitial nephritis: more data at last. Kidney Int, 73(8), 905–7. Baker, R. J. and Pusey, C. D. (2004). The changing profile of acute tubulointerstitial nephritis. Nephrol Dial Transplant, 19(1), 8–11. Baldwin, D. S., Levine, B. B., McCluskey, R. T., et al. (1968). Renal failure and interstitial nephritis due to penicillin and methicillin. N Engl J Med, 279(23), 1245–52. Ball, S., Cook, T., Hulme, B., et al. (1997). The diagnosis and racial origin of 394 patients undergoing renal biopsy: an association between Indian race and interstitial nephritis. Nephrol Dial Transplant, 12(1), 71–7. Ball, S., Lloyd, J., Cairns, T., et al. (2001). Why is there so much end-stage renal failure of undetermined cause in UK Indo-Asians? QJM, 94(4), 187–93. Bender, W. L., Whelton, A., Beschorner, W. E., et al. (1984). Interstitial nephritis, proteinuria, and renal failure caused by nonsteroidal anti-inflammatory drugs. Immunologic characterization of the inflammatory infiltrate. Am J Med, 76(6), 1006–12. Bergner, R., Hoffmann, M., Waldherr, R., et al. (2003). Frequency of kidney disease in chronic sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis, 20(2), 126–32. Bergstein, J. and Litman, N. (1975). Interstitial nephritis with anti-tubular-basement-membrane antibody. N Engl J Med, 292(17), 875–8. Bhaumik, S. K., Kher, V., Arora, P., et al. (1996). Evaluation of clinical and histological prognostic markers in drug-induced acute interstitial nephritis. Ren Fail, 18(1), 97–104. Bijol, V., Mendez, G. P., Nose, V., et al. (2006). Granulomatous interstitial nephritis: a clinicopathologic study of 46 cases from a single institution. Int J Surg Pathol, 14(1), 57–63. Border, W. A., Lehman, D. H., Egan, J. D., et al. (1974). Antitubular basement-membrane antibodies in methicillin-associated interstitial nephritis. N Engl J Med, 291(8), 381–4. Butkowski, R. J., Langeveld, J. P., Wieslander, J., et al. (1990). Characterization of a tubular basement membrane component reactive with autoantibodies associated with tubulointerstitial nephritis. J Biol Chem, 265(34), 21091–8. Buysen, J. G., Houthoff, H. J., Krediet, R. T., et al. (1990). Acute interstitial nephritis: a clinical and morphological study in 27 patients. Nephrol Dial Transplant, 5(2), 94–9. Chapagain, A., Dobbie, H., Sheaff, M., et al. (2011). Presentation, diagnosis, and treatment outcome of tuberculous-mediated tubulointerstitial nephritis. Kidney Int, 79(6), 671–7. Chen, D., Luo, C., Tang, Z., et al. (2011). Delayed renal function recovery from drug-induced acute interstitial nephritis. Am J Med Sci, 2, 2. Cheng, H. F., Nolasco, F., Cameron, J. S., et al. (1989). HLA-DR display by renal tubular epithelium and phenotype of infiltrate in interstitial nephritis. Nephrol Dial Transplant, 4(3), 205–15. Clark, T. J., Richards, N. T., Adu, D., et al. (1993). Increased prevalence of dialysis-dependent renal failure in ethnic minorities in the west Midlands. Nephrol Dial Transplant, 8(2), 146–8. Clarkson, M. R., Giblin, L., O’Connell, F. P., et al. (2004). Acute interstitial nephritis: clinical features and response to corticosteroid therapy. Nephrol Dial Transplant, 19(11), 2778–83. Clayman, M. D., Michaud, L., Brentjens, J., et al. (1996). Isolation of the target antigen of human anti-tubular basement membrane antibody-associated interstitial nephritis. J Clin Invest, 77(4), 1143–7. Clive, D. M. and Stoff, J. S. (1984). Renal syndromes associated with nonsteroidal antiinflammatory drugs. N Engl J Med, 310(9), 563–72. Cornell, L. D. (2010). IgG4-related tubulointerstitial nephritis. Kidney Int, 78(10), 951–3.

675

676

Section 4  

the patient with interstitial disease

Councilman, W. (1898). Acute interstitial nephritis. J Exp Med, 3, 393. Covic, A., Goldsmith, D. J., Segall, L., et al. (1998). Rifampicin-induced acute renal failure: a series of 60 patients. Nephrol Dial Transplant, 13(4), 924–9. Covic, A., Golea, O., Segall, L., et al. (2004). A clinical description of rifampicin-induced acute renal failure in 170 consecutive cases. J Indian Med Assoc, 102(1), 20, 22–5. Cunningham, E., Provost, T., Brentjens, J., et al. (1978). Acute renal failure secondary to interstitial lupus nephritis. Arch Intern Med, 138(10), 1560–1. Dall, A. and Hariharan, S. (2008). BK virus nephritis after renal transplantation. Clin J Am Soc Nephrol, 3 Suppl 2(2), S68–75. Ditlove, J., Weidmann, P., Bernstein, M., et al. (1977). Methicillin nephritis. Medicine (Baltimore), 56(6), 483–91. Dobrin, R. S., Vernier, R. L., and Fish, A. L. (1975). Acute eosinophilic interstitial nephritis and renal failure with bone marrow-lymph node granulomas and anterior uveitis. A new syndrome. Am J Med, 59(3), 325–33. Eastwood, J. B., Corbishley, C. M., and Grange, J. M. (2011). Tuberculosis and tubulointerstitial nephritis: an intriguing puzzle. Kidney Int, 79(6), 579–81. Farrington, K., Levison, D. A., Greenwood, R. N., et al. (1989). Renal biopsy in patients with unexplained renal impairment and normal kidney size. QJM, 70(263), 221–33. Fletcher, A. (2008). Eosinophiluria and acute interstitial nephritis. N Engl J Med, 358(16), 1760–1. Galpin, J. E., Shinaberger, J. H., Stanley, T. M., et al. (1978). Acute interstitial nephritis due to methicillin. Am J Med, 65(5), 756–65. Gonzalez, E., Gutierrez, E., Galeano, C., et al. (2008). Early steroid treatment improves the recovery of renal function in patients with drug-induced acute interstitial nephritis. Kidney Int, 73(8), 940–6. Goules, A., Masouridi, S., Tzioufas, A. G., et al. (2000). Clinically significant and biopsy-documented renal involvement in primary Sjogren syndrome. Medicine (Baltimore), 79(4), 241–9. Graham, G. D., Lundy, M. M., and Moreno, A. J. (1983). Failure of gallium-67 scintigraphy to identify reliably noninfectious interstitial nephritis: concise communication. J Nucl Med, 24(7), 568–70. Gur, H., Kopolovic, Y., and Gross, D. J. (1987). Chronic predominant interstitial nephritis in a patient with systemic lupus erythematosus: a follow up of three years and review of the literature. Ann Rheum Dis, 46(8), 617–23. Haas, M., Spargo, B. H., Wit, E. J., et al. (2000). Etiologies and outcome of acute renal insufficiency in older adults: a renal biopsy study of 259 cases. Am J Kidney Dis, 35(3), 433–47. Hannedouche, T., Grateau, G., Noel, L. H., et al. (1990). Renal granulomatous sarcoidosis: report of six cases. Nephrol Dial Transplant, 5(1), 18–24. Hewitt, W., Finegold, S., and Monzon, O. (1961). Untoward side effects associated with methicillin therapy. Antimicrob Agents Chemother, 1, 765–9. Ivanyi, B., Hamilton-Dutoit, S. J., Hansen, H. E, et al. (1996). Acute tubulointerstitial nephritis: phenotype of infiltrating cells and prognostic impact of tubulitis. Virchows Arch, 428(1), 5–12. Izzedine, H., Brocheriou, I., Rixe, O., et al. (2007). Interstitial nephritis in a patient taking sorafenib. Nephrol Dial Transplant, 22(8), 2411. Joh, K., Aizawa, S., Yamaguchi, Y., et al. (1990). Drug-induced hypersensitivity nephritis: lymphocyte stimulation testing and renal biopsy in 10 cases. Am J Nephrol, 10(3), 222–30. Joss, N., Morris, S., Young, B., et al. (2007). Granulomatous interstitial nephritis. Clin J Am Soc Nephrol, 2(2), 222–30. Kaissling, B. and Le Hir, M. (2008). The renal cortical interstitium: morphological and functional aspects. Histochem Cell Biol, 130(2), 247–62. Katz, A., Fish, A. J., Santamaria, P., et al. (1992). Role of antibodies to tubulointerstitial nephritis antigen in human anti-tubular basement membrane nephritis associated with membranous nephropathy. Am J Med, 93(6), 691–8.

Kaye, M. and Gagnon, R. F. (2008). Acute allergic interstitial nephritis and eosinophiluria. Kidney Int, 73(8), 980. Kelly, C. J. (1990). T cell regulation of autoimmune interstitial nephritis. J Am Soc Nephrol, 1(2), 140–9. Kelly, C. J. and Neilson, E. G. (1987). Contrasuppression in autoimmunity. Abnormal contrasuppression facilitates expression of nephritogenic effector T cells and interstitial nephritis in kdkd mice. J Exp Med, 165(1), 107–23. Kida, H., Abe, T., Tomosugi, N., et al. (1984). Prediction of the long-term outcome in acute interstitial nephritis. Clin Nephrol, 22(2), 55–60. Kleinknecht, D. (1995). Interstitial nephritis, the nephrotic syndrome, and chronic renal failure secondary to nonsteroidal anti-inflammatory drugs. Semin Nephrol, 15(3), 228–35. Kodner, C. M. and Kudrimoti, A. (2003). Diagnosis and management of acute interstitial nephritis. Am Fam Physician, 67(12), 2527–34. Laberke, H. G. (1980). Treatment of acute interstitial nephritis. Klin Wochenschr, 58(10), 531–2. Laberke, H. G. and Bohle, A. (1980). Acute interstitial nephritis: correlations between clinical and morphological findings. Clin Nephrol, 14(6), 263–73. Landais, P., Goldfarb, B., and Kleinknecht, D. (1987). Eosinophiluria and drug-induced acute interstitial nephritis. N Engl J Med, 316(26), 1664. Lightstone, L., Rees, A. J., Tomson, C., et al. (1995). High incidence of end-stage renal disease in Indo-Asians in the UK. QJM , 88(3), 191–5. Linton, A. L., Clark, W. F., Driedger, A. A., et al. (1980). Acute interstitial nephritis due to drugs: Review of the literature with a report of nine cases. Ann Intern Med, 93(5), 735–41. Linton, A. L., Richmond, J. M., Clark, W. F., et al. (1985). Gallium67 scintigraphy in the diagnosis of acute renal disease. Clin Nephrol, 24(2), 84–7. Lopez-Gomez, J. M. and Rivera, F. (2008). Renal biopsy findings in acute renal failure in the cohort of patients in the Spanish Registry of Glomerulonephritis. Clin J Am Soc Nephrol, 3(3), 674–81. Mackensen, F. and Billing, H. (2009). Tubulointerstitial nephritis and uveitis syndrome. Curr Opin Ophthalmol, 20(6), 525–31. Magil, A. B. (1983). Drug-induced acute interstitial nephritis with granulomas. Hum Pathol, 14(1), 36–41. Mahevas, M., Lescure, F. X., Boffa, J. J., et al. (2009). Renal sarcoidosis: clinical, laboratory, and histologic presentation and outcome in 47 patients. Medicine (Baltimore), 88(2), 98–106. Mahta, A., Kim, R. Y., and Kesari, S. (2011). Levetiracetam-induced interstitial nephritis in a patient with glioma. J Clin Neurosci, 7, 7. Mandeville, J. T., Levinson, R. D., and Holland, G. N. (2001). The tubulointerstitial nephritis and uveitis syndrome. Surv Ophthalmol, 46(3), 195–208. McMenamin, R. A., Davies, L. M., and Craswell, P. W. (1976). Drug induced interstitial nephritis, hepatitis and exfoliative dermatitis. Aust N Z J Med, 6(6), 583–7. Mignon, F., Mery, J. P., Mougenot, B., et al. (1984). Granulomatous interstitial nephritis. Adv Nephrol Necker Hosp, 13, 219–45. More, R., McMillan, G., and Duff, G. (1946). The pathology of sulphonamide allergy in man. Am J Pathol, 22, 703–25. Mori, Y., Kishimoto, N., Yamahara, H., et al. (2005). Predominant tubulointerstitial nephritis in a patient with systemic lupus nephritis. Clin Exp Nephrol, 9(1), 79–84. Muthukumar, T., Jayakumar, M., Fernando, E. M., et al. (2002). Acute renal failure due to rifampicin: a study of 25 patients. Am J Kidney Dis, 40(4), 690–6. Neilson, E. G. (1989). Pathogenesis and therapy of interstitial nephritis. Kidney Int, 35(5), 1257–70. Neilson, E. G. (2006). Mechanisms of disease: fibroblasts—a new look at an old problem. Nat Clin Pract Nephrol, 2(2), 101–8. Neilson, E. G., McCafferty, E., Feldman, A., et al. (1984). Spontaneous interstitial nephritis in kdkd mice. I. An experimental model of autoimmune renal disease. J Immunol, 133(5), 2560–5. Nickavar, A. (2011). Hydatid disease and tubulointerstitial nephritis. Pediatr Nephrol, 26(7), 1171–2.

Chapter 83 

Nolan, C. M. and Abernathy, R. S. (1977). Nephropathy associated with methicillin therapy. Prevalence and determinants in patients with staphylococcal bacteremia. Arch Intern Med, 137(8), 997–1000. Nolan, C. R., 3rd, Anger, M. S., and Kelleher, S. P. (1986). Eosinophiluria—a new method of detection and definition of the clinical spectrum. N Engl J Med, 315(24), 1516–19. Onyekpe, I., Shenoy, M., Denley, H., et al. (2001). Recurrent tubulointerstitial nephritis and uveitis syndrome in a renal transplant recipient. Nephrol Dial Transplant, 26(9), 3060–2. Perazella, M. A. and Markowitz, G. S. (2010). Drug-induced acute interstitial nephritis. Nat Rev Nephrol, 6(8), 461–70. Pettersson, E., von Bonsdorff, M., Tornroth, T., et al. (1984). Nephritis among young Finnish men. Clin Nephrol, 22(5), 217–22. Pirani, C. L., Valeri, A., D’Agati, V., et al. (1987). Renal toxicity of nonsteroidal anti-inflammatory drugs. Contrib Nephrol, 55, 159–75. Porile, J. L., Bakris, G. L., and Garella, S. (1990). Acute interstitial nephritis with glomerulopathy due to nonsteroidal anti-inflammatory agents: a review of its clinical spectrum and effects of steroid therapy. J Clin Pharmacol, 30(5), 468–75. Praga, M. and Gonzalez, E. (2010). Acute interstitial nephritis. Kidney Int, 77(11), 956–61. Preddie, D. C., Markowitz, G. S., Radhakrishnan, J., et al. (2006). Mycophenolate mofetil for the treatment of interstitial nephritis. Clin J Am Soc Nephrol, 1(4), 718–22. Pusey, C. D., Saltissi, D., Bloodworth, L., et al. (1983). Drug associated acute interstitial nephritis: clinical and pathological features and the response to high dose steroid therapy. QJM, 52(206), 194–211. Qian, Q., Humayun, H., Humayun ,Y., et al. (2011). Granulomatous interstitial nephritis associated with disseminated histoplasmosis in an immunocompetent patient. Am J Kidney Dis, 58(6), 1018–21. Raissian, Y., Nasr, S. H., Larsen, C. P., et al. (2011). Diagnosis of IgG4-related tubulointerstitial nephritis. J Am Soc Nephrol, 22(7), 1343–52. Rajakariar, R., Sharples, E. J., Raftery, M. J., et al. (2006). Sarcoid tubulo-interstitial nephritis: long-term outcome and response to corticosteroid therapy. Kidney Int, 70(1), 165–9. Rakotoarivony, J., Orfila, C., Segonds, A., et al. (1981). Human and experimental nephropathies associated with antibodies to tubular basement membrane. Adv Nephrol Necker Hosp, 10, 187–212. Randhawa, P. S., Finkelstein, S., Scantlebury, V., et al. (1999). Human polyoma virus-associated interstitial nephritis in the allograft kidney. Transplantation, 67(1), 103–9. Ribeiro, A., Wornle, M., Motamedi, N., et al. (2012). Activation of innate immune defense mechanisms contributes to polyomavirus BK-associated nephropathy. Kidney Int, 81(1), 100–11. Robson, M. G., Banerjee, D., Hopster, D., et al. (2003). Seven cases of granulomatous interstitial nephritis in the absence of extrarenal sarcoid. Nephrol Dial Transplant, 18(2), 280–4. Rodriguez-Perez, J. C., Cruz-Alamo, M., Perez-Aciego, P., et al. (1995). Clinical and immune aspects of idiopathic acute tubulointerstitial nephritis and uveitis syndrome. Am J Nephrol, 15(5), 386–91. Rossert, J. (2001). Drug-induced acute interstitial nephritis. Kidney Int, 60(2), 804–17. Ruffing, K. A., Hoppes, P., Blend, D., et al. (1994). Eosinophils in urine revisited. Clin Nephrol, 41(3), 163–66. Saeki, T., Nishi, S., Imai, N., et al. (2010). Clinicopathological characteristics of patients with IgG4-related tubulointerstitial nephritis. Kidney Int, 78(10), 1016–23. Said, S. M., Nasr, S. H., Samsa, R., et al. (2007). Nephrotoxicity of antiretroviral therapy in an HIV-infected patient. Kidney Int, 71(10), 1071–5. Saltissi, D., Pusey, C. D., and Rainford, D. J. (1979). Recurrent acute renal failure due to antibiotic-induced interstitial nephritis. Br Med J, 1(6172), 1182–3.

acute tubulointerstitial nephritis: overview

Sarles, H., Sarles, J. C., Muratore, R., et al. (1961). Chronic inflammatory sclerosis of the pancreas—an autonomous pancreatic disease? Am J Dig Dis, 6, 688–98. Schena, F. P. (1997). Survey of the Italian Registry of Renal Biopsies. Frequency of the renal diseases for 7 consecutive years. The Italian Group of Renal Immunopathology. Nephrol Dial Transplant, 12(3), 418–26. Schmid, S., Opravil, M., Moddel, M., et al. (2007). Acute interstitial nephritis of HIV-positive patients under atazanavir and tenofovir therapy in a retrospective analysis of kidney biopsies. Virchows Arch, 450(6), 665–70. Schubert, C., Bates, W. D., and Moosa, M. R. (2010). Acute tubulointerstitial nephritis related to antituberculous drug therapy. Clin Nephrol, 73(6), 413–19. Schwarz, A., Krause, P. H., Kunzendorf, U., et al. (2000). The outcome of acute interstitial nephritis: risk factors for the transition from acute to chronic interstitial nephritis. Clin Nephrol, 54(3), 179–90. Sherlock, J. E. (1977). Interstitial nephritis in rats produced by E. coli in adjuvant: immunological findings. Clin Exp Immunol, 30(1), 154–9. Shibasaki, T., Ishimoto, F., Sakai, O., et al. (1991). Clinical characterization of drug-induced allergic nephritis. Am J Nephrol, 11(3), 174–80. Shih, W., Hines, W. H., and Neilson, E. G. (1988). Effects of cyclosporin A on the development of immune-mediated interstitial nephritis. Kidney Int, 33(6), 1113–18. Sloth, K. and Thomsen, A. C. (1971). Acute renal insufficiency during treatment with azathioprine. Acta Med Scand, 189(3), 145–8. Spanou, Z., Keller, M., Britschgi, M., et al. (2006). Involvement of drug-specific T cells in acute drug-induced interstitial nephritis. J Am Soc Nephrol, 17(10), 2919–27. Storsley, L. and Gibson, I. W. (2011). Adenovirus interstitial nephritis and rejection in an allograft. J Am Soc Nephrol, 22(8), 1423–7. Stupp, R., Mihatsch, M. J., Matter, L., et al. (1990). Acute tubulo-interstitial nephritis with uveitis (TINU syndrome) in a patient with serologic evidence for Chlamydia infection. Klin Wochenschr, 68(19), 971–5. Sugimoto, T., Yasuda, M., Sakaguchi, M., et al. (2008). Acute interstitial nephritis associated with etanercept. Rheumatol Int, 28(12), 1283–4. Tanaka, T. and Nangaku, M. (2011). Pathogenesis of tubular interstitial nephritis. Contrib Nephrol, 169, 297–310. Waanders, F., Navis, G., and van Goor, H. (2010). Urinary tubular biomarkers of kidney damage: potential value in clinical practice. Am J Kidney Dis, 55(5), 813–16. Weiss, R. A., Madaio, M. P., Tomaszewski, J. E., et al. (1994). T cells reactive to an inducible heat shock protein induce disease in toxin-induced interstitial nephritis. J Exp Med, 180(6), 2239–50. Wilson, C. B. (1991). Nephritogenic tubulointerstitial antigens. Kidney Int, 39(3), 501–17. Wood, B. C., Sharma, J. N., Germann, D. R., et al. (1978). Gallium citrate Ga 67 imaging in noninfectious interstitial nephritis. Arch Intern Med, 138(11), 1665–6. Yang, C. W. (2007). Leptospirosis renal disease: understanding the initiation by Toll-like receptors. Kidney Int, 72(8), 918–25. Yang, C. W., Wu, M. S., Pan, M. J., et al. (2002). The Leptospira outer membrane protein LipL32 induces tubulointerstitial nephritis-mediated gene expression in mouse proximal tubule cells. J Am Soc Nephrol, 13(8), 2037–45. Zeisberg, M. and Neilson, E. G. (2010). Mechanisms of tubulointerstitial fibrosis. J Am Soc Nephrol, 21(11), 1819–34. Zuliani, E., Zwahlen, H., Gilliet, F., et al. (2005). Vancomycin-induced hypersensitivity reaction with acute renal failure: resolution following cyclosporine treatment. Clin Nephrol, 64(2), 155–8.

677

CHAPTER 84

Drug-induced acute tubulointerstitial nephritis Hassan Izzedine and Victor Gueutin Introduction Acute tubulointerstitial nephritis (ATIN) defines a pattern of renal disease usually associated with acute kidney injury (AKI), inflammation, and oedema of the renal interstitium. The incidence of ATIN remains unknown. Available estimates show ATIN in 1% of autopsies, 1% of renal biopsies for the evaluation of haematuria and/ or proteinuria (Michel and Kelly, 1998), 1–3% of all cases of AKI, and 15% of renal biopsies for unexplained AKI (Farrington et al., 1989; Neilson, 1989; Kodner and Kudrimoti, 2003). Drug-induced ATIN is responsible for approximately 70% of all ATIN cases (Michel and Kelly, 1998; Kabakus et al., 1999; Nishitarumizu et al., 2000; Baker and Pusey, 2004). Medication-induced adverse events may be classified using a qualitative study assessment, based upon the classification of causal criteria for adverse effects of medications from the World Health Organization, which rates causality as certain, probable, possible, improbable, and conditional or insufficient (Table 84.1) (Edwards and Aronson, 2000). Drug-induced ATIN was first described in 1968, when it was associated with the use of methicillin (Baldwin et  al., 1968), occurring in up to 17% of patients who have been treated for > 10 days (Nolan and Abernathy, 1977; Galpin et  al., 1978; Rossert, 2001), and the clinical manifestations of methicillin-induced ATIN were considered the prototypical presentation of ATIN. There are several case reports of cross-sensitivity to beta-lactam antibiotics, eliciting acute allergic TIN. Since then, many other drugs have been implicated, of which antimicrobial agents, non-steroidal anti-inflammatory drugs (NSAIDs), and, more recently, proton pump inhibitors (PPIs) have been most commonly involved. Currently, the list of drugs that can cause ATIN continues to expand. Table 84.2 summarizes the most common of these (Nessi et al., 1976; Galpin et al., 1978; Ten et al., 1988; Neilson, 1989; Allon et  al., 1990; Gaughan et  al., 1993; Lo et al., 1993; Neelakantappa et al., 1993; World et al., 1996; Abadín et  al., 1998; Cruz and Perazella, 1998; Fang and Huang, 1998; Markowitz et al., 1998; Michel and Kelly, 1998; Schurman et al., 1998; Wai et al., 1998; Andrews and Robinson, 1999; Koshy et al., 1999; Corrigan and Stevens, 2000; Ejaz et  al., 2000; Jaradat et al., 2000; Post et al., 2000; Torpey et al., 2004; Esteve et al., 2005; Audimoolam and Bhandari, 2006; Tomlinson et al., 2006; Hoppes et al., 2007; Brosnahan et al., 2008; Hunter et al., 2009; Wang et  al., 2009; Chatzikyrkou et  al., 2010; Korsten et  al., 2010). Recently, antivascular endothelial growth factor agents

(bevacizumab, sorafenib, sunitinib) used in clinical trials to treat cancer, have also been reported to cause ATIN (Barakat et al., 2007; Izzedine et al., 2007; Winn et al., 2009).

Pathogenesis Drug-induced ATIN occurs in an idiosyncratic, dose-independent manner. The pathogenesis of ATIN involves an allergic response that is prompted by exposure to a drug. A  type-IV (delayed) hypersensitivity response is often implicated in the pathogenesis of drug-induced ATIN. The presence of helper-inducer and suppressor-cytotoxic T lymphocytes in the renal interstitial inflammatory infiltrate suggests that T-cell-mediated hypersensitivity reactions and cytotoxic T-cell-induced injury are involved in pathogenesis of ATIN (Toto, 1990). While antibiotics often produce a systemic allergic reaction (including fever, skin rash, and eosinophilia), NSAIDs trigger a cell-mediated or delayed-type hypersensitivity response. A humoral response underlies rare cases of ATIN, in which a portion of a drug molecule (i.e. methicillin) may act as a hapten, bind to the tubular basement membrane (TBM), and elicit anti-TBM antibodies (Border et al., 1974; Perazella and Markowitz, 2010). The most widely accepted theory is that drugs behave as haptens after binding either to extrarenal proteins that later will be planted in the kidney, or to renal proteins (Rossert, 2001). The reaction to the agent is presumably caused by previous sensitization, and, indeed, patients may have a history of exposure to the ingested drug or to a similar drug. The inflammation in the kidney is often part of a systemic hypersensitivity reaction, which may include fever, arthralgias, and skin rash. Eosinophils are commonly a significant component of the renal inflammatory infiltrate, and, as noted earlier, peripheral blood eosinophilia is often seen, as well. Reactions involving immune complex deposition are of two types: those with formation of immune complexes that are deposited around tubules and those owing to formation of antibodies directed against antigens of the TBM. Rarely, drug antigens may be planted in the TBM. The inciting drug may serve as a hapten, leading to antibody formation (Nadasdy and Sedmak, 2007). Cell-mediated immunity has also been involved in the genesis of drug-induced ATIN. Reactions to antibiotics are often associated with infiltrates consisting mainly of mononuclear cells and eosinophils. Most of the mononuclear cells are CD4+ T lymphocytes (Bender et al., 1984; Pamukcu et al., 1984; Kobayashi et al., 1998).

Chapter 84 

drug-induced acute tubulointerstitial nephritis

Table 84.1  Classification of causal criteria for adverse effects of medications Classification

Criteria

Certain

Clinical event including abnormalities in the laboratory, which occurs in a coherent temporal relation to the administration of the drug, and which cannot be explained by concurrent illness, drugs or other chemicals The response to withdrawal of the medication can be clinically demonstrated The phenomenon can be pharmacologically demonstrable, using the re-administration of the medication if necessary

Probable

Clinical event including abnormalities in the laboratory, which occurs within a ‘reasonable’ time period after the administration of the drug, and for which it would be improbable to explain by concurrent illness, drugs or other chemicals In clinical follow-ups there is a reasonable response upon withdrawal of the medication It is not necessary to re-administer the drug

Possible

Clinical event including abnormalities in the laboratory, occurring within a reasonable time period upon administering the medication, but which can be explained by illness, other medications, or concomitantly used chemicals Information about suspending the drug is hidden or is unclear

Improbable

Clinical event including abnormalities in the laboratory, within a reasonable time period upon administering the drug, but for which a causal relation is improbable due to the fact that other drugs, chemicals, or illness can provide a causal explanation

Conditional

Clinical event including abnormalities in the laboratory, reported as an adverse reaction to the drug but for which more essential data are necessary to make an appropriate evaluation or for which additional data are beginning to be evaluated

Insufficient

The report suggests an adverse reaction that cannot be judged because the information is insufficient or cannot be verified or corroborated

However, patients with drug-induced ATIN and nephrotic-range proteinuria were found to have a predominance of CD8+ T cells in the interstitial infiltrate (Bender et al., 1984; Pamukcu et al., 1984; D’Agati et al., 1986; Kobayashi et al., 1998). Eosinophils are commonly seen, but their absence does not exclude drug-induced ATIN (Hawkins et al., 1989). After a few days or weeks from the onset of the disease, a variable accumulation of plasma cells and histiocytes may occur. Although rare in other types of ATIN, granulomas may sometimes develop in drug-induced cases (Nadasdy, 2007); the presence of granulomas is consistent with delayed-type hypersensitivity. T-cell reactivity has been documented in some patients with drug-induced hypersensitivity reactions (Joh et al., 1990; Shibasaki et  al., 1991). Cytotoxic lymphocytes reactive against autologous renal cell lines have been isolated from one patient treated with recombinant interleukin-2 (IL-2) (Vlasveld et al., 1993).

Clinical presentation General features The clinical presentation of ATIN is highly variable. A  lag of 7–10 days typically exists between drug exposure and the development of AKI (Ten et al., 1988; Neilson, 1989), but this lag can be considerably shorter following repeated exposure (Ten et al., 1988; Neilson, 1989; Schubert et al., 2010) or markedly longer with certain drugs (e.g. up to 18 months with NSAIDs) (Clive and Stoff, 1984). The clinical presentation may vary from isolated abnormal urinary sediment or asymptomatic elevation in serum creatinine to generalized hypersensitivity syndrome, with fever, rash, eosinophilia, and oliguric AKI. Skin rash, fever, eosinophilia, and the classic triad (including all three of the above) were observed in 21%, 30%, 36%, and < 10% of cases, respectively (Clarkson et al., 2004).

Table 84.2  Drugs associated with ATIN Drug class

Examples

Antibiotics

Almost all agents, in particular ampicillin, cephalosporins, ciprofloxacin, co-trimoxazole, ethambutol (myambutol), isoniazid, macrolides, penicillin G, rifampicin, sulphonamides, tetracycline, vancomycin

NSAIDs (including salicylates and selective COX2 inhibitors)

Almost all agents, in particular aspirin, fenoprofen, ibuprofen, indomethacin, naproxen, phenylbutazone, indomethacin, naproxen, phenylbutazone, piroxicam, tolemetin, zomepirac

Gastric acid suppressants

Proton pump inhibitors (omeprazole, lansoprazole, rabeprazole), cimetidine, ranitidine (very rarely)

Diuretics

Chlorthalidone, ethacrynic acid, furosemide, thiazides, indapamide, tienilic acid, triamterene

Miscellaneous

Abacavir, aciclovir, adalimumab, allopurinol, amlodipine, anti-CD4 antibody, atazanavir, azathioprine, bethanidine, captopril, carbimazole, chlorpropamide, cimetidine, carbamazepine, clofibrate, clometacin, cocaine, creatine, deferasirox, diltiazem, famotidine, fenofibrate, floctafenin, foscarnet, glafenin, indinavir, interferon, interleukin-2, lenalidomide, mesalazine, phentermine, phenindione, phenytoin, pranlukast, propylthiouracil, quinine, sunitinib, sorafenib, TNF-alpha inhibitors etc.

Bold: frequent or clinically important. Italic: with interstitial granuloma formation.

679

680

Section 4  

the patient with interstitial disease

AKI can often be irreversible, resulting in chronic kidney disease (CKD) (Michel and Kelly, 1998; Kodner and Kudrimoti, 2003). AKI is typically non-oliguric. It develops 7–10 days after drug exposure in 80% of patients (Rossert, 2001), with a slow increase in serum creatinine; however, patients with severe AKI can present with oliguria and a rapidly progressive course (Alexopoulos, 1998). Other associated symptoms may include flank pain (distension of the renal capsule by inflammation and parenchymal swelling may occur, particularly with rifampicin), gross haematuria, and other general manifestations, such as myalgias, arthralgias, and myositis. Occasionally, the kidneys are palpable, when markedly enlarged (Baker and Williams 1963; Simenhoff et al., 1968; Toto, 1990). Drug-induced hypersensitivity syndrome (DiHS), also called ‘drug rash with eosinophilia and systemic symptoms’ (DRESS), is a severe drug hypersensitivity reaction involving rash, fever (38–40°C), and multiorgan failure, affecting liver, kidneys, heart, and/or lungs (Mauri-Hellweg et al., 1995; Peyrière et al., 2006; Ben M’rad et al., 2009). Debate is ongoing about the most accurate name for this syndrome, as fewer than one-half of cases show eosinophilia (whereas, for example, those caused by drugs like abacavir or lamotrigine typically do not). Granulomatous interstitial nephritis (GIN) is present in 0.5–0.9% of renal biopsies and has been associated with anticonvulsants, antibiotics, NSAIDs, allopurinol, and diuretics (Table 84.2). Histologic features do not seem to distinguish the underlying cause of GIN. Treatment with a moderate dosage of prednisolone is associated with a good prognosis, irrespective of the underlying cause and the degree of interstitial fibrosis (Joss et al., 2007).

Drug classes Different drug classes are associated with some particular clinical features of ATIN, as summarized in Table 84.3.

Non-steroidal anti-inflammatory drugs NSAIDs, including cyclooxygenase 2 (COX-2) inhibitors, may precipitate AKI, particularly in vulnerable patients. NSAID use accounts for an estimated 15% of all cases of drug-induced AKI. A case–control study estimated a 3.2 relative risk (95% confidence

interval (CI) 1.8–5.8) of AKI in otherwise healthy current users of NSAIDs. The NSAID-associated ATIN typically occurs after treatments > 1 year (in contrast to 12 days, on average, with beta-lactam antibiotics) and has a much lower incidence of fever, rash, and eosinophilia than other drug-induced ATIN (Pirani et al., 1987). NSAID-associated ATIN exhibits a low grade of interstitial inflammation and it may often be accompanied by a nephrotic syndrome. Proteinuria, usually in the nephritic range, occurs in 70% of cases (Rossert, 2001). The onset of NSAID-induced nephritic syndrome is usually delayed, with a mean time to onset of 5.4 months after the start of NSAID therapy and ranging from 2 weeks to 18  months (Abraham and Keane, 1984; Clive and Stoff, 1984). NSAID-induced nephritic syndrome is usually reversible between 1 month and 1 year after discontinuation of NSAID therapy. During the recovery period, some 20 % of patients require dialysis. Steroids should be employed as in patients with idiopathic minimal change disease (Murray and Brater, 1993). Changes in the glomeruli in these patients are minimal and resemble those of idiopathic minimal change disease, with marked epithelial-foot process fusion (Murray and Brater, 1993; Rotellar et  al., 1989; Morgenstern et  al., 1989). The mechanism of NSAID-induced nephrotic syndrome has not been fully characterized. It is thought to be the result of leukotriene release, from arachidonic acid via the lipooxygenase pathway, when the cyclooxy­genase pathway is blocked. Leukotrienes increase glomerular and peritubular permeability, which may lead to the induction of TIN and proteinuria (Abraham and Keane, 1984; Clive and Stoff, 1984; Warren et al., 1989). Such patients should avoid subsequent administration of NSAIDs, as nephrotic syndrome relapse may occur with re-challenge (Mohammed and Stevens, 2000). Radford et  al. (1996), using the Mayo Clinic biopsy registry, reported that > 10% of biopsy-proven membranous nephropathy was attributable to NSAIDs, with a median duration of 43 weeks of drug ingestion. Nephrotic-range proteinuria was present for < 8 weeks, but reversed after discontinuation of the drug. Other rarer kidney injury mechanisms have also been reported with NSAIDs, including acute papillary necrosis (Atta and Whelton, 1997) and renal vasculitis (Leung et al., 1985). Adverse renal effects

Table 84.3  Clinical features of ATIN associated with specific drug classes Drug class

Clinical features

Antibiotics

Common (e.g. 17% of methicillin-treated patients) Fever, rash, arthralgias, eosinophilia, eosinophiluria, and pyuria Cross-sensitivity between penicillin and cephalosporins Rifampicin-induced ATIN occurs with intermittent or discontinuous drug administration Associated allergic and hypersensitivity reactions and vasculitis on histology

NSAIDs

Common: 15% of all causes of AKI Oedema, congestive heart failure, hyponatraemia, hyperkalaemia, nephrotic syndrome may also occur Possibly associated renal lesions: papillary necrosis, minimal change disease, membranous nephropathy

Acid suppressants

Rash (rarely), inconsistent pyuria

Diuretics

Unknown incidence Rare rash, inconsistent pyuria Cross-reactivity between furosemide and sulphonamide antibiotics Vasculitis on histology

Chapter 84 

are generally reversible after discontinuation of NSAID treatment (Murray and Brater, 1993). Cases of ATIN, membranous nephropathy, and minimal change disease following treatment with COX-2 inhibitors have also been reported (Rocha and Fernandez-Alonso, 2001; Alper et al., 2002; Henao et al., 2002; Markowitz et al., 2003; Brewster and Perazella, 2004). The rapid and complete resolution of these conditions following discontinuation of COX-2 inhibitors strongly supports their implication in the disease pathogenesis.

Proton pump inhibitors It appears that PPIs have now become a leading cause of drug-induced ATIN, with an incidence of 8 cases per 10,000 patient-years (Simpson et al., 2006). ATIN develops, on average, 11 weeks after starting the PPI treatment (Geevasinga et al., 2006). Since 1992, when the first case of ATIN induced by omeprazole was reported (Ruffenach et al., 1992), all other agents in this class have been associated with ATIN (Torpey et al., 2004; Geevasinga et al., 2005; Simpson et al., 2006; Ricketson et al., 2009). In a study by Geevasinga et al. the outcome of patients with PPI-induced ATIN was good when the disease was early recognized; these patients rarely required renal replacement therapy and end-stage renal disease was not seen in any of them, although many developed mild-to-moderate CKD (Geevasinga et  al., 2006). Accidental or unrecognized drug re-challenge after an initial episode of suspected PPI-induced ATIN is associated with a rapid onset of AKI, within a few days from exposure (Ruffenach et al., 1992; Christensen et al., 1993; Assouad et al., 1994; Gronich et al., 1994).

Antimicrobials

β-lactam antibiotics β-lactam antibiotics (penicillins and cephalosporins) are frequently involved in the development of hypersensitivity syndrome (Baldwin et al., 1968; Border et al., 1974). Most patients who develop ATIN after treatment with a cephalosporin have a history of penicillin allergy (Toto, 1990; Alexopoulos, 1998; Rossert, 2001). The duration of exposure to the causative drug is relatively short, ranging from a few days to a few weeks. Fever, rash and/or eosinophilia are seen in > 75% of patients. Urinary abnormalities, like proteinuria, leucocyturia and haematuria, also occur in approximately 75% of cases. Although most patients with β-lactam-induced ATIN recover their renal function, irreversible CKD may sometimes develop (Baldwin et  al., 1968; Border et  al., 1974; Schellie and Groshong, 1999; Papachristou et al., 2006). Other antibiotics ATIN can also occur with other antibiotics, such as rifampicin, sulphonamides, and quinolones. Rifampicin-induced ATIN is associated with the production of anti-rifampicin antibodies and commonly manifests with oliguric AKI, proximal tubular injury, haemolytic anaemia, thrombocytopenia, and hepatitis. Approximately two-thirds of patients require renal replacement therapy (Campese et  al., 1973; Toto, 1990; Alexopoulos, 1998; Rossert, 2001). ATIN induced by sulphonamide antibiotics may be associated with typical hypersensitivity reactions, such as fever, rash, and eosinophilia (Kleinknecht et  al., 1983). HIV patients, transplant recipients, and those with pre-existing CKD are prone to develop sulphonamide-induced ATIN more than other individuals (Perazella, 2000, 2003). However, this may be due to the frequent use of sulphonamide antibiotics in these patients. Fluoroquinolones

drug-induced acute tubulointerstitial nephritis

may also cause ATIN. In contrast to rifampicin and sulphonamides, hypersensitivity syndrome associated with fluoroquinolones is rare. Ciprofloxacin is the most common causative agent in this class, but cases of ATIN from norfloxacin, ofloxacin, and levofloxacin use have also been described (Toto, 1990; Alexopoulos, 1998; Rossert, 2001) .

Protease inhibitors Protease inhibitors have become the mainstay of therapy for patients with AIDS. Renal complications, particularly crystalluria, were early recognized as adverse effects of these drugs. However, more recent reports indicate that ATIN (with foreign body-type giant cells) may also occur, particularly with indinavir (Sarcletti et al., 1998; Olyaei et al., 2000).

Diagnosis A diagnosis of ATIN should be considered in any patient with clinical manifestations of a hypersensitivity reaction, unexplained AKI, and a history of recent exposure to a possibly offending drug (Rossert, 2001; Toto, 2001). Some clinical features (Table 84.3) and ancillary tests may be helpful in diagnosing drug-induced ATIN. Unfortunately, none of these tests has sufficient accuracy for a certain diagnosis, except renal biopsy. However, biopsy is indicated only when diagnosis is unclear or when the patient does not improve after discontinuation of the suspected medication.

Laboratory tests Characteristic laboratory findings include an acute rise in plasma creatinine concentration, eosinophilia, leucocyturia with white cell casts and eosinophiluria. Eosinophiluria helps confirm the diagnosis of ATIN with an estimated sensitivity and specificity of 67% and 83%, respectively (Rossert, 2001); however, it can also be found in patients with rapidly progressive glomerulonephritis and with renal atheroembolism. Gross or microscopic haematuria and urinary red blood cell casts have also been described. However, occasional patients show bland sediment, with few cells or casts (Lo et al., 1993). Thus, a relatively normal urinalysis should not exclude the diagnosis of ATIN. Proteinuria is usually mild (< 1 g/day), but nephrotic-range proteinuria may occur in cases with associated minimal change disease or membranous nephropathy induced by NSAIDs or, sometimes, by ampicillin, rifampicin, interferon or ranitidine (Neugarten et al., 1983; Averbuch et al., 1984; Clive and Stoff, 1984; Gaughan et  al., 1993; Praga and González, 2010). Tubular function disorders have also been described, including salt-wasting syndrome, proximal tubular damage with type II renal tubular acidosis and Fanconi syndrome, and distal tubular damage, with type I renal tubular acidosis and sodium or potassium abnormalities. Involvement of collecting ducts in the medulla and papillae may be associated with polyuria (Choudhury and Ahmed, 2006). Increased serum immunoglobulin E levels, suggesting an allergic response, have been reported in some patients with drug-induced ATIN (Linton et al., 1980), but this is an inconsistent finding.

Imaging studies There are no imaging studies to accurately confirm or exclude ATIN. Renal ultrasonography and CT scanning may demonstrate

681

682

Section 4  

the patient with interstitial disease

normal or enlarged kidneys, with increased cortical echogenicity or density. The role of gallium-67 scanning for the diagnosis of ATIN remains unclear; however, in patients who are poor candidates for renal biopsy, it may be useful in distinguishing ATIN from acute tubular necrosis (Graham et al., 1983; Linton et al., 1985).

(A)

Kidney biopsy Kidney biopsy remains the gold standard for the diagnosis of drug-induced ATIN. The main pathological findings are interstitial inflammation and tubulitis (Fig. 84.1A). The hallmark of ATIN is the inflammatory infiltrate, consisting of T lymphocytes, monocytes, and a variable number of plasma cells and eosinophils (Fig. 84.1B), associated with oedema, within the renal interstitium, sparing the glomeruli and the blood vessels (Michel and Kelly, 1998; Rastegar and Kashgarian, 1998). Occasionally, granuloma formation with epithelioid giant cells may be seen (Fig. 84.1C). In an immunohistochemical study of interstitial infiltrates in cases of ATIN secondary to NSAIDs and β-lactam antibiotics, the mononuclear cell component was found to consist of 71.7% T cells (with equal numbers of CD4+ and CD8+ cells), 15.2% monocytes, and 7.4% B cells (D’Agati et al., 1989). Peritubular infiltration and occasional lymphocytic invasion beyond the TBM may occur with mild-to-severe tubular damage. Many drugs can induce a granulomatous reaction. In a series of 46 patients with GIN, the most frequent aetiologies were drug-induced disease and sarcoidosis, representing 44.7% and 28.9% of cases, respectively (Bijol et  al., 2006). In this study, drug-induced GIN was mainly caused by antibiotics and NSAIDs. The pathological findings in drug-induced GIN are indistinguishable from those seen in the setting of sarcoidosis or other forms of GIN (Bijol et al., 2006). Interestingly, a higher incidence of persistent renal impairment is found in cases with interstitial granulomas than in those without granulomas (Grunfeld et al., 1993). Immunofluorescence microscopy is usually negative, although rare cases with linear deposits of immunoglobulins and complement along the TBM, suggesting an antibody-mediated inflammatory response, have been described in patients with methicillin-induced ATIN (Border et al., 1974).

(B)

(C)

Outcomes Recovery of renal function was observed in the great majority of cases, after discontinuation of the offending drug, with or without associated glucocorticoid therapy (Galpin et  al., 1978; Schwarz et  al., 2000; Rossert, 2001). The probability of recovery depends on the duration of AKI prior to diagnosis, and ideally this should not exceed two weeks. The recovery of the kidney function is often incomplete, with persistent elevation of serum creatinine in up to 40% of cases (Rossert, 2001; Baker and Pusey, 2004). Acute dialysis may be required (Handa, 1986; Koselj et al., 1993; Bhaumik et al., 1996), but only about 10% of patients remain dialysis-dependent on long-term (Baker and Pusey, 2004; Clarkson et al., 2004; Laberke and Bohle, 1980). Poor prognostic factors include prolonged AKI (> 3 weeks), ATIN associated with NSAID use, and certain histologic findings, such as interstitial granulomas (Grunfeld et  al., 1993), diffuse versus patchy inflammation, interstitial fibrosis, and tubular atrophy (Laberke and Bohle, 1980; Bhaumik et al., 1996; Schwarz et al., 2000).

Fig. 84.1  Drug-induced ATIN with: (A) Prominent interstitial inflammation with lymphocytes, eosinophils, and focal plasma cells, associated with tubulitis (trichrome Masson; magnification ×100). (B) Eosinophils (magnification ×400). (C) Interstitial granuloma, composed of a well-circumscribed aggregate of epithelioid histiocytes and multinucleated giant cells (magnification ×200).

Treatment The optimal therapy of drug-induced ATIN remains to be defined, since there are no randomized controlled trials. A proposed management algorithm is presented in Fig. 84.2 (adapted from Kodner and Kudrimoti, 2003).

Chapter 84 

drug-induced acute tubulointerstitial nephritis

Drug-induced ATIN suspected Withdraw offending medications Clinical improvement within 1 week

No clinical improvement within 1 week

Supportive management observation

Contraindication to renal biopsy or patient refuses biopsy? No

Yes

Perform renal biopsy Acute TIN

No

Consider other diagnostic studies (gallium-67 scan or renal ultrasound)

Yes

Continue evaluation for other causes and treat appropriately

Results consistent with ATIN

Contraindication to steroid therapy? Yes

Results not consistent with ATIN

Continue evaluation for other causes of AKI

No Steroid therapy (prednisone 0.5–1 mg/kg/day) Improvement in renal function? No

Yes

Discuss alternative immunosuppressive therapy

Continue steroid therapy

Fig. 84.2  Algorithm for the management of drug-induced ATIN. Source data from Kodner and Kudrimoti (2003).

Supportive care Discontinuation of the potentially causative drug is the mainstay of therapy and the first necessary step in the early management of suspected or biopsy-proven drug-induced ATIN (Michel and Kelly, 1998; Rossert, 2001). No other therapy is required if a patient’s renal function shows improvement within 1 week following drug cessation (Buysen et al., 1990). However, a considerable proportion of patients (36%) may develop CKD (Baker and Pusey, 2004).

Glucocorticoid therapy There are no randomized controlled trials of glucocorticoid therapy in patients with drug-induced ATIN and the available data is conflicting. Thus, the decision to initiate steroids should be guided by the patient’s clinical course following withdrawal of the offending drug. Several retrospective series demonstrated no benefit from glucocorticoid therapy (Koselj et al., 1993; Bhaumik et al., 1996; Schwarz et al., 2000; Clarkson et al., 2004). In one series, including 42 cases

of biopsy-proven drug-induced ATIN (NSAIDs 44%, antibiotics 33%, and PPIs 7%), 26 patients were treated with steroids (intravenous methylprednisolone for 3 days followed by oral prednisone tapered over 3–6 weeks), whereas the remaining 16 were not; there was no difference in serum creatinine between the two groups at 1, 6, and 12 months (Clarkson et al., 2004). Improvement in kidney function following glucocorticoid therapy has been suggested by several uncontrolled studies (Galpin et al., 1978; Handa, 1986; Buysen et al., 1990). A multicentre retrospective study involving 61 patients (Gonzalez et al., 2008) suggested a beneficial influence of corticosteroids on the outcome of drug-induced ATIN. An earlier start of therapy (day 13 vs day 34) was associated with a better recovery of renal function (Appel, 2008; González et al., 2008), a lower need for dialysis after 18 months (4% vs 44% of cases), and lower serum creatinine levels (2.1 mg/dL vs 3.7 mg/dL). Among treated patients, those who started steroids within 7 days from withdrawal of the offending drug were significantly more likely to recover renal

683

684

Section 4  

the patient with interstitial disease

function than those who received steroids after this period (odds ratio 6.6, 95% CI 1.3–33.6). Given the potential benefit and the relative safety of short-term steroid therapy, it seems reasonable to treat patients with corticosteroids if they do not show significant improvement in serum creatinine within 3–7 days after discontinuation of the offending agent. The optimal dose and duration of therapy are unclear; one suggested regimen consists of prednisone at a dose of 0.5–1 mg/ kg per day (without exceeding 60 mg/day) for 1 month, beginning a gradual taper after the serum creatinine has returned to or near baseline, for a total therapy duration of 3 months (Clarkson et al., 2004). Most patients are likely to improve within the first 1 or 2 weeks (Clarkson et al., 2004). In patients with more severe AKI, therapy may be initiated with intravenous methylprednisolone (0.5–1 g/day for 3 consecutive days) (Clarkson et al., 2004).

Other therapies There is limited experience with alternative agents in patients who are steroid-dependent, steroid-resistant (as with NSAID-induced ATIN), or cannot tolerate glucocorticoids. There are a few case reports and small series using mycophenolate mofetil (MMF) (Preddie et  al 2006)  and ciclosporin (Zuliani et  al., 2005). The largest study (Preddie et  al., 2006)  included eight patients with biopsy-proven ATIN who had received glucocorticoids for at least 6  months and became steroid-dependent. MMF was then given for 13–34 months. All patients were subsequently able to discontinue corticosteroids and all but two patients showed significant improvement in serum creatinine.

Summary and recommendations 1. Antibiotics, NSAIDs, and PPIs are the most common drugs inducing ATIN. 2. Only 10% of patients present with the classical triad of rash, fever, and eosinophilia. Urinalysis usually reveals leucocyturia and haematuria; white cell casts may also be present. Eosinophiluria occurs in 80% of cases. Proteinuria is usually mild to moderate. Tubular function abnormalities may be seen. 3. The potentially inducing agent must be immediately discontinued. Most cases improve spontaneously thereafter, within 3–7  days. Indications for kidney biopsy include uncertainty regarding the diagnosis and lack of spontaneous recovery following cessation of the offending drug. 4. In cases with no improvement of renal function within a week after drug discontinuation, early initiation of corticosteroids (0.5–1 mg/kg per day for 1 month, followed by gradual taper for a total therapy duration of 3 months) will typically improve the course of the disease within 1–2 weeks. 5. In patients who are steroid-dependent, steroid-resistant, or cannot tolerate glucocorticoids, the optimal therapy approach is not known, but MMF, ciclosporin, or cyclophosphamide may be considered.

References Abadín, J. A., Durán, J. A., and Pérez de León, J. A. (1998). Probable diltiazem-induced acute interstitial nephritis. Ann Pharmacother, 32(6), 656–8.

Abraham, P. A. and Keane, W. F. (1984). Glomerular and interstitial disease induced by nonsteroidal anti-inflammatory drugs. Am J Nephrol, 4, 1–6. Alexopoulos, E. (1998). Drug-induced acute interstitial nephritis. Ren Fail, 20, 809–19. Allon, M., Lopez, E. J., and Min, K. W. (1990). Acute renal failure due to ciprofloxacin. Arch Intern Med, 150, 2187–9. Alper, A. B., Meleg-Smith, S., and Krane, N. K. (2002). Nephrotic syndrome and interstitial nephritis associated with celecoxib. Am J Kidney Dis, 40, 1086–90. Andrews, P. A. and Robinson, G. T. (1999). Intravascular haemolysis and interstitial nephritis in association with ciprofloxacin. Nephron, 83(4), 359–60. Appel, G. B. (2008). The treatment of acute interstitial nephritis: More data at last. Kidney Int, 73(8), 905–7. Assouad, M., Vicks, S. L., Pokroy, M. V., et al. (1994). Recurrent acute interstitial nephritis on rechallenge with omeprazole. Lancet, 344(8921), 549. Atta, M. G. and Whelton, A. (1997). Acute renal papillary necrosis induced by ibuprofen. Am J Ther, 4(1), 55–60. Audimoolam, V. K. and Bhandari, S. (2006). Acute interstitial nephritis induced by glucosamine. Nephrol Dial Transplant, 21(7), 2031. Audimoolam, V. K. and Bhandari, S. (2006). Clarithromycin-induced granulomatous tubulointerstitial nephritis. Nephrol Dial Transplant, 21(9), 2654–5. Averbuch, S. D., Austin, H. A., 3rd., Sherwin, S. A., et al. (1984). Acute interstitial nephritis with the nephrotic syndrome following recombinant leukocyte a interferon therapy for mycosis fungoides. N Engl J Med, 310(1), 32–5. Baker, R. J. and Pusey, C. D. (2004). The changing profile of acute tubulointerstitial nephritis. Nephrol Dial Transplant, 19(1), 8–11. Baker, S. B. and Williams, R. T. (1963). Acute interstitial nephritis due to drug sensitivity. Br Med J, 1, 1655–8. Baldwin, D. S., Levine, B. B., McCluskey, R. T., et al. (1968). Renal failure and interstitial nephritis due to penicillin and methicillin. N Engl J Med, 279(23), 1245–52. Barakat, R. K., Singh, N., and Lal, R., et al. (2007). (2007). Interstitial nephritis secondary to bevacizumab treatment in metastatic leiomyosarcoma. Ann Pharmacother, 41(4), 707–10. Bender, W. L., Whelton, A., Beschorner, W. E., et al. (1984). Interstitial nephritis, proteinuria, and renal failure caused by nonsteroidal anti-inflammatory drugs. immunologic characterization of the inflammatory infiltrate. Am J Med, 76, 1006–12. Ben M’rad, M., Leclerc-Mercier, S., Blanche, P., et al. (2009). Drug-induced hypersensitivity syndrome: clinical and biologic disease patterns in 24 patients. Medicine, 88(3), 131–40. Bhaumik, S. K., Kher, V., Arora, P., et al. (1996). Evaluation of clinical and histological prognostic markers in drug-induced acute interstitial nephritis. Ren Fail, 18(1), 97–104. Bijol, V., Mendez, G. P., Nose, V., et al. (2006). Granulomatous interstitial nephritis: a clinicopathologic study of 46 cases from a single institution. Int J Surg Pathol, 14, 57–63. Border, W. A., Lehman, D. H., Egan, J. D., et al. (1974). Antitubular basement-membrane antibodies in methicillin-associated interstitial nephritis. N Engl J Med, 291(8), 381–4. Brewster, U. C. and Perazella, M. A. (2004). Acute tubulolinterstitial nephritis associated with celecoxib. Nephrol Dial Transplant, 19, 1017–18. Brosnahan, G., Gokden, N., and Swaminathan, S. (2008). Acute interstitial nephritis due to deferasirox: a case report. Nephrol Dial Transplant, 23(10), 3356–8. Buysen, J. G., Houthoff, H. J., Krediet, R. T., et al. (1990). Acute interstitial nephritis: a clinical and morphological study in 27 patients. Nephrol Dial Transplant, 5(2), 94–9. Campese, V. M., Marzullo, F., Schena, F. P., et al. (1973). Acute renal failure during intermittent rifampicin therapy. Nephron, 10, 256–61. Chatzikyrkou, C., Hamwi, I., Clajus, C., et al. (2010). Biopsy proven acute interstitial nephritis after treatment with moxifloxacin. BMC Nephrol, 11, 19.

Chapter 84 

Choudhury, D. and Ahmed, Z. (2006). Drug-associated renal dysfunction and injury. Nat Clin Pract Nephrol, 2(2), 80–91 Clarkson, M. R., Giblin, L., O’Connell, F. P., et al. (2004). Acute interstitial nephritis: clinical features and response to corticosteroid therapy. Nephrol Dial Transplant, 19(11), 2778–83. Clive, D. M. and Stoff, J. S. (1984). Renal syndromes associated with nonsteroidal antiinflammatory drugs. N Engl J Med, 310(9), 563–72. Corrigan, G. and Stevens, P. E. (2000). Interstitial nephritis associated with the use of mesalazine in inflammatory bowel disease. Aliment Pharmacol Ther, 14, 1–6. Cruz, D. N. and Perazella, M. A. (1998). Drug-induced acute tubulointerstitial nephritis: the clinical spectrum. Hosp Pract (Minneap), 33(2), 151–2, 157–8, 161–4. D’Agati, V. D., Appel, G. B., Estes, D., et al. (1986). Monoclonal antibody identification of infiltrating mononuclear leukocytes in lupus nephritis. Kidney Int, 30, 573 D’Agati, V. D., Theise, N. D., Pirani, C. L., et al. (1989). Interstitial nephritis related to nonsteroidal anti- inflammatory agents and â-lactam antibiotics: a comparative study of the interstitial infiltrates using monoclonal antibodies. Modern Pathol, 2, 390–6. Edwards, I. R. and Aronson, J. K. (2000). Adverse drug reactions: definitions, diagnosis, and management. Lancet, 356(9237), 1255–9. Ejaz, A. A., Fitzpatrick, P. M., Haley, W. E., et al. (2000). Amlodipine besylate induced acute interstitial nephritis. Nephron, 85(4), 354–6. Esteve, J. B., Launay-Vacher, V., Brocheriou, I., et al. (2005). COX-2 inhibitors and acute interstitial nephritis: case report and review of the literature. Clin Nephrol, 63, 385–9. Fang, J. T. and Huang, C. C. (1998). Propylthiouracil-induced acute interstitial nephritis with acute renal failure requiring haemodialysis: successful therapy with steroids. Nephrol Dial Transplant, 13(3), 757–8. Farrington, K., Levison, D. A., Greenwood, R. N., et al. (1989). Renal biopsy in patients with unexplained renal impairment and normal kidney size. QJM, 70, 221–33 Galpin, J. E., Shinaberger, J. H., Stanley, T.M., et al. (1978). Acute interstitial nephritis due to methicillin. Am J Med, 65, 756–65. Gaughan, W. J., Sheth, V. R., Francos, G. C., et al. (1993). Ranitidine-induced acute interstitial nephritis with epithelial cell foot process fusion. Am J Kidney Dis, 22(2), 337–40. Geevasinga, N., Coleman, P. L., Webster, A. C., et al. (2006). Proton pump inhibitors and acute interstitial nephritis. Clin Gastroenterol Hepatol, 4, 597–604. Geevasinga, N., Kairaitis, L., Rangan, G. K., et al. (2005). Acute interstitial nephritis secondary to esomeprazole. Med J Aust, 182, 235–6. González, E., Gutiérrez, E., Galeano, C., et al. (2008). Early steroid treatment improves the recovery of renal function in patients with drug-induced acute interstitial nephritis. Kidney Int, 73(8), 940–6. Graham, G. D., Lundy, M. M., and Moreno, J. J. (1983). Failure of 67Gallium scintigraphy to identify reliably non-infectious interstitial nephritis. J Nucl Med, 24, 568–70. Grunfeld, J. P., Kleinknecht, D., and Droz, D. (1993). Acute interstitial nephritis. In R. W. Schrier and C. W. Gottschalk (eds.) Diseases of the Kidney, pp.1331–53. Boston, MA: Little Brown. Handa, S. P. (1986). Drug-induced acute interstitial nephritis: report of 10 cases. CMAJ, 135(11), 1278–81. Hawkins, E. P., Berry, P. L., and Silva, F. G. (1989). Acute tubulointerstitial nephritis in children: clinical, morphologic, and lectin studies. A report of the Southwest Pediatric Nephrology Study Group. Am J Kidney Dis, 14, 466–71. Henao, J., Hisamuddin, I., Nzerue, C. M., et al. (2002). Celecoxib-induced acute interstitial nephritis. Am J Kidney Dis, 39, 1313–17. Hoppes, T., Prikis, M., and Segal, A. (2007). Four cases of nafcillin-associated acute interstitial nephritis in one institution. Nat Clin Pract Nephrol, 3, 456–61. Hunter, R., Gaughan, T., Queirazza, F., et al. (2009). Clozapine-induced interstitial nephritis—a rare but important complication: a case report. J Med Case Reports, 3, 8574. Izzedine, H., Brocheriou, I., Rixe, O., et al. (2007). Interstitial nephritis in a patient taking sorafenib. Nephrol Dial Transplant, 22(8), 2411.

drug-induced acute tubulointerstitial nephritis

Jaradat, M., Phillips, C., Yum, M. N., et al. (2000). Acute tubulointerstitial nephritis attributable to indinavir therapy. Am J Kidney Dis, 35(4), E16. Joh, K., Aizawa, S., Yamaguchi, Y., et al. (1990). Drug-induced hypersensitivity nephritis: Lymphocyte stimulation testing and renal biopsy in 10 cases. Am J Nephrol, 10, 222–30. Joss, N., Morris, S., Young, B., et al. (2007). Granulomatous interstitial nephritis. Clin J Am Soc Nephrol, 2(2), 222–30. Kabakuş, N., Aydinoğlu, H., Bakkaloğlu, S. A., et al. (1999). Mumps interstitial nephritis: a case report. Pediatr Nephrol, 13(9), 930–1. Kleinknecht, D., Vanhille, P., Morel-Maroger, L., et al. (1983). Acute interstitial nephritis due to drug hypersensitivity. An up-to-date review with a report of 19 cases. Adv Nephrol Necker Hosp, 12, 277–308. Kobayashi, Y., Honda, M., Yoshikawa, N., et al. (1998). Immunohistological study in sixteen children with acute tubulointerstitial nephritis. Clin Nephrol, 50, 14–20. Kodner, C. M. and Kudrimoti, A. (2003). Diagnosis and management of acute interstitial nephritis. Am Fam Physician, 67(12), 2527–34. Korsten, P., Sweiss, N. J., Nagorsnik, U., et al. (2010). Drug-induced granulomatous interstitial nephritis in a patient with ankylosing spondylitis during therapy with adalimumab. Am J Kidney Dis, 56(6), e17–21. Koselj, M., Kveder, R., Bren, A. F., et al. (1993). Acute renal failure in patients with drug-induced acute interstitial nephritis. Ren Fail, 15(1), 69–72. Koshy, K. M., Griswold, E., and Schneeberger, E. E. (1999). Interstitial nephritis in a patient taking creatine. N Engl J Med, 340(10), 814–15. Laberke, H.G. and Bohle, A. (1980). Acute interstitial nephritis: correlations between clinical and morphological findings. Clin Nephrol, 14(6), 263–73. Leung, A. C., McLay, A., Dobbie, J. W., et al. (1985). Phenylbutazone-induced systemic vasculitis with crescentic glomerulonephritis. Arch Intern Med, 145(4), 685–7. Linton, A. L., Clark, W. F., Driedger, A. A., et al. (1980). Acute interstitial nephritis due to drugs. Ann Intern Med, 93, 735–41. Linton, A. L., Richmond, J. M., Clark, W. F., et al. (1985). Gallium67 scintigraphy in the diagnosis of acute renal disease. Clin Nephrol, 24, 84–7. Lo, W. K., Rolston, K. V., Rubenstein, E. B., et al. (1993). Ciprofloxacin-induced nephrotoxicity in patients with cancer. Arch Intern Med, 153(10), 1258–62 Markowitz, G. S., Falkowitz, D. C., Isom, R., et al. (2003). Membranous glomerulopathy and acute interstitial nephritis following treatment with celecoxib. Clin Nephrol, 59, 137–42. Markowitz, G. S., Tartini, A., and D’Agati, V. D. (1998). Acute interstitial nephritis following treatment with anorectic agents phentermine and phendimetrazine. Clin Nephrol, 50(4), 252–254. Mauri-Hellweg, D., Bettens, F., Mauri, D., et al. (1995). Activation of drug-specific CD4+ and CD8+ T cells in individuals allergic to sulfonamides, phenytoin, and carbamazepine. J Immunol, 155(1), 462–472. Michel, D. M. and Kelly, C. J. (1998). Acute interstitial nephritis. J Am Soc Nephrol, 9(3), 506–15. Mohammed, E. P. and Stevens, J. M. (2000). Recurrence of Arthrotec-associated nephrotic syndrome with re-challenge. Clin Nephrol, 53, 483–5. Morgenstern, S. J., Bruns, F. J., Fraley, D. S., et al. (1989). Ibuprofen-associated lipoid nephrosis without interstitial nephritis. Am J Kidney Dis, 14(1), 50–2 Murray, M. D. and Brater, D. C. (1993). Renal toxicity of the nonsteroidal anti-inflammatory drugs. Annu Rev Pharmacol Toxicol, 33, 435–65. Nadasdy, T. and Sedmak, D. (2007). Acute and chronic tubulointerstitial nephritis. In J. C. Jennette and R. H. Heptinstall (eds.) Hepinstall’s Pathology of the Kidney (6th ed.), pp. 1083–137. Philadelphia, PA: Lippincott Williams & Wilkins. Neelakantappa, K., Gallo, G. R., Lowenstein, J. (1993). Ranitidine-associated interstitial nephritis and Fanconi syndrome. Am J Kidney Dis, 22, 333–6. Neilson, E. G. (1989). Pathogenesis and therapy of interstitial nephritis. Kidney Int, 35(5), 1257–70. Nessi, R., Bonoldi, G. L., Redaelli, B., et al. (1976). Acute renal failure after rifampicin: a case report and survey of the literature. Nephron, 16, 148–59.

685

686

Section 4  

the patient with interstitial disease

Neugarten, J., Gallo, G. R., and Baldwin, D. S. (1983). Rifampin-induced nephrotic syndrome and acute interstitial nephritis. Am J Nephrol, 3(1), 38–42. Nishitarumizu, K., Tokuda, Y., Uehara, H., et al. (2000). Tubulointerstitial nephritis associated with Legionnaires’ disease. Intern Med, 39(2), 150–3. Nolan, C. M. and Abernathy, R. S. (1977). Nephropathy associated with methicillin therapy. Prevalence and determinants in patients with staphylococcal bacteremia. Arch Intern Med, 137(8), 997–1000. Olyaei, A. J., deMattos, A. M., and Bennett, W. M. (2000). Renal toxicity of protease inhibitors. Curr Opin Nephrol Hypertens, 9, 473–6. Pamukcu, R., Moorthy, V., Singer, J. R., et al. (1984). Idiopathic acute interstitial nephritis: Characterization of the infiltrating cells in the renal interstitium as T helper lymphocytes. Am J Kidney Dis, 4, 24–9. Papachristou, F., Printza, N., Farmaki, E., et al. (2006). Antibiotics-induced acute interstitial nephritis in 6 children. Urol Int, 76(4), 348–52 Perazella, M. A. (2000). Acute renal failure in HIV-infected patients: a brief review of common causes. Am J Med Sci, 319, 385–91. Perazella, M. A. (2003). Drug-induced renal failure: update on new medications and unique mechanisms of nephrotoxicity. Am J Med Sci, 325, 349–62. Perazella, M. A. and Markowitz, G. S. (2010). Drug-induced acute interstitial nephritis. Nat Rev Nephrol, 6(8), 461–70. Peyrière, H., Dereure, O., Breton, H., et al. (2006). Variability in the clinical pattern of cutaneous side-effects of drugs with systemic symptoms: does a DRESS syndrome really exist? Br J Dermatol, 155(2), 422–8. Pirani, C. L., Valeri, A., D’Agati, V., et al. (1987). Renal toxicity of nonsteroidal anti-inflammatory drugs. Contrib Nephrol, 55, 159–75. Post, A. T., Voorhorst, G., and Zanen, A. L. (2000). Reversible renal failure after treatment with omeprazole. Neth J Med, 57(2), 58–61. Praga, M. and González, E. (2010). Acute interstitial nephritis. Kidney Int, 77(11), 956–61. Preddie, D. C., Markowitz, G. S., Radhakrishnan, J., et al. (2006). Mycophenolate mofetil for the treatment of interstitial nephritis. Clin J Am Soc Nephrol, 1(4), 718–22. Radford, M. G. Jr., Holley, K. E., Grande, J. P., et al. (1996). Reversible membranous nephropathy associated with the use of nonsteroidal anti-inflammatory drugs. JAMA, 276(6), 466–9. Rastegar, A. and Kashgarian, M. (1998). The clinical spectrum of tubulointerstitial nephritis. Kidney Int, 54(2), 313–27. Ricketson, J., Kimel, G., Spence, J., et al. (2009). Acute allergic interstitial nephritis after use of pantoprazole. CMAJ, 180, 535–8. Rocha, J. L. and Fernandez-Alonso, J. (2001). Acute tubulointerstitial nephritis associated with the selective COX-2 enzyme inhibitor, rofecoxib. Lancet, 357, 1946–7. Rossert, J. (2001). Drug-induced acute interstitial nephritis. Kidney Int, 60, 804–17. Rotellar, J. A., Garcia Ruiz, C., Martinez Vea, A., et al. (1989). Minimal-change glomerulopathy associated with diclofenac: response to prednisone. Am J Kidney Dis, 14(6), 530–1 Ruffenach, S. J., Siskind, M. S., and Lien, Y. H. (1992). Acute interstitial nephritis due to omeprazole. Am J Med, 93, 472–3.

Sarcletti, M., Petter, A., and Zangerle, R. (1998). Indinavir and interstitial nephritis. Ann Intern Med, 128, 320–1. Schellie, S. F. and Groshong, T. (1999). Acute interstitial nephritis following amoxicillin overdose. Mo Med, 96(6), 209–11. Schubert, C., Bates, W. D., and Moosa, M. R. (2010). Acute tubulointerstitial nephritis related to antituberculous drug therapy. Clin Nephrol, 73(6), 413–19. Schurman, S. J., Alderman, J. M., Massanari, M., et al. (1998). Tubulointerstitial nephritis induced by the leukotriene receptor antagonist pranlukast. Chest, 114(4), 1220–3. Schwarz, A., Krause, P. H., Kunzendorf, U., et al. (20000). The outcome of acute interstitial nephritis: risk factors for the transition from acute to chronic interstitial nephritis. Clin Nephrol, 54(3), 179–90. Shibasaki, T., Ishimoto, F., Sakai, O., et al. (1991). Clinical characterization of drug-induced allergic nephritis. Am J Nephrol, 11, 174–80. Simenhoff, M. L., Guild, W. R., and Dammin, G. J. (1968). Acute diffuse interstitial nephritis: review of the literature and case report. Am J Med, 44, 618–25. Simpson, I. J., Marshal, M. R., Pilmore, H., et al. (2006). Proton pump inhibitors and acute interstitial nephritis: report and analysis of 15 cases. Nephrology, 11, 381–5. Ten, R. M., Torres, V. E., Milliner, D. S., et al. (1988). Acute interstitial nephritis: immunologic and clinical aspects. Mayo Clin Proc, 63, 921–30. Tomlinson, L., Davies, K., Wright, D. A., et al. (2006). Granulomatous interstitial nephritis treated with a tumour necrosis factor-alpha inhibitor. Nephrol Dial Transplant, 21(8), 2311–14. Torpey, N., Barker, T., and Ross, C. (2004). Drug-induced tubulo-interstitial nephritis secondary to proton pump inhibitors: experience from a single UK renal unit. Nephrol Dial Transplant, 19, 1441–6. Toto, R. D. (1990). Acute tubulointerstitial nephritis. Am J Med Sci, 299, 392–410. Vlasveld, L. T., van de Wiel-van Kemenade, E., de Boer, A. J., et al. (1993). Possible role for cytotoxic lymphocytes in the pathogenesis of acute interstitial nephritis after recombinant interleukin-2 treatment for renal cell cancer. Cancer Immunol Immunother, 36, 210–13. Wai, A. O., Lo, A. M., Abdo, A., et al. (1998). Vancomycin-induced acute interstitial nephritis. Ann Pharmacother, 32(11), 1160–4. Wang, Y. C., Lin, Y. F., Chao, T. K., et al. (2009). Acute interstitial nephritis with prominent eosinophil infiltration. Clin Nephrol, 71, 187–91. Warren, G. V., Korbet, S. M., Schwartz, M. M., et al. (1989). Minimal change glomerulopathy associated with nonsteroidal antiinflammatory drugs. Am J Kidney Dis, 13, 127–30. Winn, S. K., Ellis, S., Savage, P., et al. (2009). Biopsy-proven acute interstitial nephritis associated with the tyrosine kinase inhibitor sunitinib: a class effect? Nephrol Dial Transplant, 24(2), 673–5. World, M. J., Stevens, P. E., Ashton, M. A., et al. (1996). Mesalazine-associated interstitial nephritis. Nephrol Dial Transplant, 11, 614–21. Zuliani, E., Zwahlen, H., Gilliet, F., et al. (2005). Vancomycin-induced hypersensitivity reaction with acute renal failure: resolution following cyclosporine treatment. Clin Nephrol, 64(2), 155–8.

CHAPTER 85

Other toxic acute tubulointerstitial nephritis Benjamin J. Freda and Gregory L. Braden Ethylene glycol Ethylene glycol intoxication produces severe neurologic changes, increased osmolal and anion gap acidosis, from glycolic, glyoxylic, and oxalic acids, and may cause acute kidney injury (AKI) within 72 hours of ingestion. However, AKI from ethylene glycol often leads to chronic kidney disease (CKD) and sometimes to end-stage renal disease (ESRD) requiring long-term dialysis. Autopsy studies of patients who died 22–44 hours after ethylene glycol ingestion, before acute peritoneal dialysis or haemodialysis were available, demonstrated significant calcium oxalate monohydrate (COM) deposition, occluding the lumens of proximal tubules, along with cytoplasmic deposition in the tubular cytoplasm, easily detectable by birefringence on polarized microscopy, but there was no tubular necrosis (Pons and Custer, 1946). In patients with delayed therapy or delayed presentation, proximal tubular degeneration and necrosis accompanied proximal tubule COM deposition. Serial renal biopsies in these patients revealed the disappearance of tubular COM crystals, but persistent COM crystals in the interstitium and gradual interstitial fibrosis. Some patients remained on dialysis permanently and others had variable stages of CKD after dialysis was discontinued. Patients with repeated exposure to low-dose ethylene glycol all had severe CKD, leading to long-term dialysis (DeSilva and Mueller, 2009). There are several mechanisms whereby ethylene glycol induces renal tubular toxicity. The ethylene glycol metabolites, glycolaldehyde and glyoxylate, incubated in vitro with human proximal tubule HK-2 cells caused ATP depletion and phospholipid and enzyme destruction (Poldelski et al, 2001). Moreover, studies of cultured human proximal tubule cells have shown that COM crystals may kill these cells at concentrations in the range measured in human intoxications (McMartin, 2009). In addition, studies of ethylene glycol toxicity in animals correlate the extent of renal damage to the level of tubular COM crystal deposition. COM crystals can alter phospholipid membrane structure and function and induce reactive oxygen species and mitochondrial dysfunction (McMartin, 2009). Taken together, these studies support multiple pathways whereby ethylene glycol induces severe renal tubular damage leading to CKD and the need for chronic dialysis, particularly in patients with delayed presentation or delayed therapy. Ethylene glycol intoxication usually presents with abdominal pain, severe nausea and vomiting, mental status changes and severe anion gap metabolic acidosis, usually associated with an osmolal gap > 10 mOsm/kg H2O. More severe central nervous system

complications include seizures, delirium, and coma. The urinalysis usually shows an excess of crystals, including calcium oxalate dihydrate (envelope-shaped) and/or monohydrate (resembling dog biscuits or spindles). Urgent haemodialysis is indicated to remove both ethylene glycol and its toxic metabolites. In addition, fomepizole inhibits hepatic alcohol dehydrogenase and prevents formation of toxic metabolites from ethylene glycol. This agent should be administered intravenously (IV) as early as possible, at a dose of 15 mg/kg, with subsequent dosing every 12 hours, until dialysis has removed all excess ethylene glycol and the serum bicarbonate, anion gap, and the osmolal gap have returned to normal. During haemodialysis, a continuous infusion of fomepizole 1–2 mg/kg per hour should be administered, since it is removed by dialysis. If fomepizole is not available, ethanol can be administered IV to inhibit alcohol dehydrogenase, but this leads to temporary alcohol intoxication which could lead to further complications in these acutely ill patients. Where available, ethylene glycol blood levels can be used to guide therapy. Haemodialysis is usually continued for 2 hours after the serum bicarbonate and anion gap have returned to normal, to prevent rebound intoxication after dialysis is completed.

Chlorinated hydrocarbons Carbon tetrachloride (CCl4) has been used as an antihelminthic agent, as a solvent in dry cleaning or as an insecticide. However, due to severe hepatic and renal toxicity, its use has recently been limited to the synthesis of chlorofluorocarbon refrigerants. The two carbon chlorinated solvents, trichloroethylene and tetrachloroethylene are still used in dry cleaning, synthesis of fumigants, and as solvents in paints and varnishes. When the latter agents are abused by sniffing paint thinners or cleaning solvents, both agents can cause acute tubular necrosis and hepatic necrosis, but neither cause CKD. In contrast, irreversible acute tubular necrosis from CCl4 has been reported with numerous crystals in proximal tubular lumens, which stain as calcium by Von Kossa stain, associated with interstitial fibrosis (Morrin, et al., 1961). However, two series of 19 patients intoxicated with CCl4, primarily by inhalation, described AKI in 50–100% of patients, often supported by peritoneal dialysis or haemodialysis, but none developed CKD (Ruprah et al., 1985). Although these and other solvents were thought to cause glomerulonephritis, better epidemiologic studies have found no association between solvents and glomerulonephritis (Harrington et  al., 1989). The mechanism of

688

Section 4  

the patient with interstitial disease

CCl4 toxicity is likely due to its liver metabolism via cytochrome P450 2E1 to highly reactive free trichloromethyl radicals causing lipid peroxidation. Toxicity from these agents is dominated by hepatic necrosis due to direct hepatocyte toxicity, leading to irreversible liver failure, hepatic encephalopathy, ascites, circulatory failure, and death. These patients most often die before liver transplantation can be considered. Management of chlorinated hydrocarbon intoxication is largely supportive, as in any haemodynamically unstable patient. If the skin is contaminated, all clothing should be rapidly removed and all exposed skin should be washed with large volumes of water; however, soap should not be used, as it may enhance the skin absorption of chlorinated hydrocarbons. Gastric aspiration can be attempted to lessen the gastrointestinal absorption of the agent. Although there is no specific antidote, N-acetylcysteine may minimize hepatotoxicity and hyperbaric oxygen may be useful. Dialysis is ineffective in removing chlorinated hydrocarbons.

Paraquat Paraquat is a commonly used herbicide, which causes human toxicity by the generation of reactive oxygen species. Even small amounts of ingested paraquat can be lethal and the case fatality rate is > 50% (Pond, 1990; Gawarammana and Buckley, 2011). Paraquat is rapidly absorbed and eliminated unchanged in the urine over 24 hours. Poisoning can be confirmed by measuring paraquat in plasma or urine. Clinical presentation depends on the amount ingested, but starts with nausea, vomiting, and a burning sensation in the mouth, throat, chest, and abdomen. Paraquat is concentrated in the lungs and oxidant damage ensues, with severe inflammation and fibrosis. Pulmonary oedema can develop within 24–48 hours and many patients develop acute respiratory distress syndrome. Severe impairment in gas exchange can occur in the absence of significant radiographic changes (Kim et al., 2009). Renal dysfunction is common during the first week after ingestion and can occur as early as 24–48 hours in more severely poisoned patients. Proximal tubular dysfunction may occur, including glucosuria, tubular proteinuria, aminoaciduria, and impaired sodium, urate, and phosphorus handling. In one study, > 50% of 278 patients poisoned with paraquat developed AKI, with approximately 35% having RIFLE class ‘Failure’ (Kim et al., 2009). Most patients were non-oliguric. When the initial serum creatinine was ≥ 1.2 mg/dL, the survival rate was only 14%. AKI usually developed within the first 5 days after ingestion. The amount of paraquat ingested correlated with the severity of AKI; importantly, however, renal recovery was the rule in survivors. In addition, baseline serum uric acid is an independent predictor of mortality and AKI, possibly as a marker of the severity of oxidative damage (Kim et al., 2011). Initial treatments include standard resuscitation measures, gastrointestinal decontamination and enhancing the renal elimination of paraquat. IV saline should be used to optimize intravascular volume and renal excretion of paraquat. In patients with airway stability, bentonite, Fuller’s earth, and activated charcoal should be orally administered to enhance gastrointestinal adsorption and elimination (Gawarammana and Buckley, 2011). Gastric lavage should be avoided in most cases, because paraquat is caustic. Various antioxidant and anti-inflammatory agents have also been used (Gawarammana and Buckley, 2011). Although haemodialysis,

haemofiltration, and haemoperfusion can remove paraquat, the clinical impact of these procedures is small, given the rapid accumulation of paraquat in lung tissue. In patients with severe poisoning, it is unlikely that extracorporeal removal will be beneficial, and palliative care is the most suitable approach. Haemodialysis, haemofiltration, or haemoperfusion can be considered in patients thought to have a chance to survive; however, it is unclear if one modality is preferred over the others or whether extracorporeal clearance improves clinical outcomes (Fienfeld, 2006).

Toxic mushrooms Of the thousands of different species of mushrooms, < 100 are known to be toxic to humans. The clinical presentation of toxic mushroom exposure varies from mild gastrointestinal discomfort to organ failure and rarely, death (Berger and Guss, 2005). Depending on the type of ingested mushroom and predominant toxin, 14 clinical syndromes have been described. Early onset toxicity (< 6 hours after exposure) mainly involves the gastrointestinal tract and the central nervous system. Importantly, delayed toxicity can appear > 6–24 hours after exposure to certain mushrooms and include encephalopathy, liver and renal failure, as well as rhabdomyolysis (Diaz, 2005). Multiple mushroom-derived toxins have been associated with the development of nephrotoxicity. The most common expression of nephrotoxicity is AKI. Depending on the timing of presentation and type of mushroom, AKI has been identified as early as 6 hours after exposure and as late as 1–2 weeks (West et al., 2009; Talmud et al., 2011).

Cortinarius The genus Cortinarius contains around 2000 different species of mushrooms, several of which are poisonous to humans. Most poisonings have been reported with Cortinarius orellanus (‘Fool’s webcap’) and the toxicity is attributed to the bipyridine compound, orellanine. This mushroom is found mostly in wooded areas throughout Europe. Orellanine is not destroyed by cooking and the toxicity arises mainly from the generation of toxic free radicals, resulting in damage to renal tubular cells (Mount et al., 2002). Gastrointestinal symptoms are followed by acute tubulopathy and variable progression to CKD. In 90 cases of Cortinarius poisoning, the development of nephrotoxicity was delayed, with a median time to onset of 8.5 days (Danel et al., 2001). Thirty-five patients had renal biopsies 1–9 weeks after onset of AKI. Most cases showed TIN or tubular necrosis, with variable degrees of interstitial oedema and fibrosis (Danel et  al., 2001). Many patients required acute dialysis (74%) and about 50% developed CKD, with 68% of these requiring either chronic dialysis or kidney transplantation. Treatment of Cortinarius toxicity is mainly supportive. Various techniques aimed at toxin removal, including haemoperfusion and plasma exchange, have been tried without success. This may be a result of the short-lived presence (likely < 2–3 days) of the toxin in plasma. Thus, at the time of patient’s presentation, it is possible that most of the toxin is already concentrated in renal tissue (Danel et al., 2001).

Amanita phalloides Several species from the genus Amanita are associated with significant nephrotoxicity, especially Amanita phalloides, otherwise

Chapter 85 

known as ‘Deathcap’. This mushroom is associated with the majority of fatal mushroom ingestions. Amatoxins are not denatured by heating and are readily absorbed from the gastrointestinal tract. They act as potent inhibitors of DNA transcription in tissues with a high rate of protein synthesis (Karlson-Stiber and Persson, 2003). These compounds are extremely hepatotoxic and ingestion of one mushroom may be fatal. Patients present with acute diarrhoea and abdominal colic after 6–24 hours from ingestion. This can be followed by elevations in serum transaminase levels and in some patients, fulminant renal, liver and cardiac failure, within 2–4 days. Confirmation of amatoxin poisoning is difficult, because clinical assays are not readily available and bedside tests are cumbersome and non-specific (Beuhler et al., 2004). Efforts should be made to obtain a sample of the ingested mushroom for analysis and identification by an experienced mycologist. Post mortem renal biopsies showed severe acute tubular necrosis in proximal convoluted tubules with some interstitial oedema and mononuclear cell infiltration (Fineschi et al., 1996). Beyond supportive care and gastrointestinal decontamination with repeated doses of activated charcoal, many unproven strategies have been used to enhance toxin elimination and removal, including plasmapheresis, charcoal haemoperfusion, thioctic acid, silibinin, intravenous penicillin, and N-acetylcysteine (Diaz, 2005). For patients with progressive liver failure, the Molecular Absorbent Regenerating System (MARS ®) has been used as a bridge to liver transplantation. In survivors, hepatic and renal recovery may occur, with some patients developing immune-mediated chronic active hepatitis. One case has been reported with hepatic recovery, but with dialysis-dependent renal failure (Garrouste et al, 2009).

Amanita smithiana Amanita smithiana can be found mainly in the western part of North America. Most patients present with nausea, vomiting and abdominal pain, within 5–6 hours of ingestion. This contrasts to most Amanita phalloides poisonings, where diarrhoea is the predominant gastrointestinal symptom, occurring commonly > 6 hours after ingestion (West et  al., 2009). The renal failure in Amanita smithiana poisoning is also delayed, seen usually 3–5  days after ingestion. The main renal lesion is acute tubular necrosis. Treatment is supportive and most patients recover renal function, even when dialysis is required temporarily (West et al., 2009).

Other toxic mushrooms Gyromitra spp. mushrooms can cause gastrointestinal symptoms with toxicity progressing to hepatic dysfunction, methaemoglobinaemia and haemolysis. Haem-induced AKI, rhabdomyolysis, and hepatorenal syndrome may also develop. Tricholomas equestre mushrooms can be associated with AKI from rhabdomyolysis (Bedry et al., 2001). Paxillus involutus ingestion can lead to an allergic immune-mediated haemolysis and resultant haem-induced AKI (Schmidt et al., 1971).

other toxic acute tubulointerstitial nephritis

References Bedry, R., Baudrimont, I., Deffieux, G., et al. (2001). Wild-mushroom intoxication as a cause of rhabdomyolysis. N Engl J Med, 345(11), 798–802. Berger, K. J. and Guss, D. A. (2005). Mycotoxins revisited: Part I. J Emerg Med, 28(1), 53–62. Beuhler, M., Lee, D. C., and Gerkin, R. (2004). The Meixner test in the detection of alpha-amanitin and false-positive reactions caused by psilocin and 5-substituted tryptamines. Ann Emerg Med, 44(2), 114–20. Danel, V. C., Saviuc, P. F., and Garon, D. (2001). Main features of Cortinarius spp. poisoning: a literature review. Toxicon, 39(7), 1053–60. DeSilva, M. B. and Mueller, P. S. (2009). Renal consequences of long-term, low-dose intentional ingestion of ethylene glycol. Ren Fail, 31, 586–8. Diaz, J. H. (2005). Syndromic diagnosis and management of confirmed mushroom poisonings. Crit Care Med, 33(2), 427–36. Fineschi, V., Di Paolo, M., Centini, F. (1996). Histological criteria for diagnosis of amanita phalloides poisoning. J Forensic Sci, 41, 429–32. Garrouste, C., Hémery, M., Boudat, A. M., et al. (2009). Amanita phalloides poisoning-induced end-stage renal failure. Clin Nephrol, 71(5), 571–4. Gawarammana, I. B. and Buckley, N. A. (2011). Medical management of paraquat ingestion. Br J Clin Pharmacol, 72(5), 745–57. Harrington, J. M., Witby, H., Gray, C. N., et al. (1989). Renal disease and occupational exposure to organic solvents: a case referent approach. Br J Ind Med, 46, 643–50. Karlson-Stiber, C. and Persson, H. (2003). Cytotoxic fungi—an overview. Toxicon, 42 (4), 339–49. Kim, J. H., Gil, H. W., Yang, J. O., et al. (2011). Serum uric acid level as a marker for mortality and acute kidney injury in patients with acute paraquat intoxication. Nephrol Dial Transplant, 26(6), 1846–52. Kim, S. J., Gil, H. W., Yang, J. O., et al. (2009). The clinical features of acute kidney injury in patients with acute paraquat intoxication. Nephrol Dial Transplant, 24(4), 1226–32. McMartin, K. (2009). Are calcium oxalate crystals involved in the mechanism of acute renal failure in ethylene glycol poisoning? Clin Toxicol, 47, 859–69. Morrin, P. A., Gedney, W. B., Barth, W., et al. (1961). Acute tubular necrosis: report of a case with failure to recover after sixty-seven days of oliguria. Ann Intern Med, 925–30. Mount, P., Harris, G., Sinclair, R., et al. (2002). Acute renal failure following ingestion of wild mushrooms. Intern Med J, 32(4), 187–90. Pond, S. M. (1990). Manifestations and management of paraquat poisoning. Med J Aust, 152(5), 256–9. Pons, C. A. and Custer, R. P. (1946). Acute ethylene glycol poisoning: a clinico-pathologic report of eighteen fatal cases. Am J Med Sci, 211(5), 544–52. Ruprah, M., Mant, T.G.K., and Flanagan, R.J. (1985). Acute carbon tetrachloride poisoning in 19 patients: implications for diagnosis and treatment. Lancet, 1(8436), 1027–9. Schmidt, J., Hartmann, W., Würstlin, A., et al. (1971). ‘[Acute kidney failure due to immunohemolytic anemia following consumption of the mushroom Paxillus involutus]’ (in German). Deutsche Medizinische Wochenschrift, 96 (28), 1188–91. Talmud, D., Wynckel, A., Grossenbacher, F., et al. (2011). Four family cases of acute renal failure. Diagnosis: Orellanus syndrome. Pediatr Nephrol, 26(3), 385–8. West, P. L., Lindgren, J., and Horowitz, B. Z. (2009). Amanita smithiana mushroom ingestion: a case of delayed renal failure and literature review. J Med Toxicol, 5(1), 32–8.

689

CHAPTER 86

Chronic tubulointerstitial nephritis: overview Adalbert Schiller, Adrian Covic, and Liviu Segall Introduction

Aetiology

The term ‘interstitial nephritis’ was first used by Councilman in 1898 (Baker and Pusey, 2004) to describe renal lesions he observed in several patients who had died from scarlet fever and diphtheria. Those lesions included interstitial oedema and sterile cell infiltrate. Sixteen years later, Volhard and Fahr included interstitial nephritis in their classification of kidney diseases (Weening and Jennette, 2012). The interstitium and the tubules—representing around 80% of the renal mass—are distinct but interrelated structural components of the kidneys, and damage to one of them is typically associated with damage to the other. Therefore, the term ‘tubulointerstitial nephropathy’ (TIN) is now preferred. TIN describes a group of renal diseases characterized by interstitial infiltration with inflammatory cells, interstitial oedema and/or fibrosis, as well as tubular atrophy. According to the clinical presentation, TIN is classified into acute TIN (ATIN)—with sudden onset and acute kidney injury (AKI)—and chronic TIN (CTIN)—with insidious onset and slow progression towards end-stage renal disease (ESRD). However, ATIN may sometimes become CTIN, when both acute and chronic lesions coexist at some point. TIN may further be classified as primary (with inflammation limited to the tubules and interstitium, without significant changes in the glomeruli and the vessels) and secondary (when tubulointerstitial lesions are associated with dominant glomerular or vascular disease). In this chapter, we will focus on the general features of primary CTIN.

The aetiologic spectrum of CTIN is remarkably wide. The most common causes are shown in Table 86.1 (Eknoyan and Truong, 1999; Braden et al., 2005; Remuzzi et al., 2007). Among these, drugs are responsible for > 70% of cases, followed by infections—15.8% (Baker and Pusey, 2004).

Epidemiology The prevalence of CTIN varies with geographical area, diagnostic criteria employed, and indications for renal biopsy. In an autopsy series reported in 1978, CTIN was found in 0.2% of cases (Zollinger and Mihatsch, 1978, pp. 407–10). In a study from 1975, among patients who underwent renal biopsy for chronic kidney disease (CKD) of unknown aetiology, the prevalence of CTIN was found to be 22% (Murray and Goldberg, 1975). In more recent renal biopsy registries, CTIN comprise between 1.5% of cases in Romania (Covic et al., 2006) and 4.4% in the Czech Republic (Rychlík et  al., 2004). The prevalence of CTIN among ESRD patients on renal replacement therapy was reported to be 9.6% in the ERA-EDTA registry in 2003 (including nine European countries) (Stengel et  al., 2003)  and 7.2% in the US in 2010 (United States Renal Data System, 2010).

Clinical manifestations The onset of CTIN is insidious and the course is slowly progressive towards ESRD, typically over several years. Patients are often asymptomatic and the disease is diagnosed during routine laboratory check-ups, screening for CKD, or assessment of arterial hypertension. However, the blood pressure is usually normal until advanced stages of CTIN (López-Novoa et al., 2011). The clinical presentation of CTIN is dependent of the aetiology and the severity of renal lesions. Thorough history taking may reveal drug intake (such as analgesics or NSAIDs), herbal therapy, or exposure to industrial chemicals. A history of haematological malignancies (such as multiple myeloma, leukaemia, or lymphoma) or solid tumours treated with chemotherapy is also significant. Some patients may have urological diseases like kidney stones and recurrent urinary tract infections, or a history of urologic surgery. Clinical signs suggestive of systemic diseases such as systemic lupus erythematosus, Sjögren syndrome, cryoglobulinaemia, sarcoidosis, or amyloidosis may be found on physical examination. Tubular dysfunctions are common and their pattern and severity depend on the location and extension of tubular lesions, as well as on the aetiology. Agents that damage the proximal tubule—like heavy metals and immunoglobulin light chains deposition—may induce type 2 (proximal) renal tubular acidosis (RTA) or Fanconi syndrome, consisting of bicarbonaturia, hyperphosphaturia, glucosuria, aminoaciduria, uricosuria, and tubular proteinuria (Rastegar and Kashgarian, 1998; Eknoyan and Truong, 1999). Agents affecting the loop of Henle and the collecting duct—like analgesics, hypercalciuria/hypercalcaemia, urate nephropathy, and sickle cell disease—may cause a decrease of sodium or water reabsorption ability, the latter resulting in nephrogenic diabetes insipidus. Distal tubular damage can be induced by light chain deposition disease, chronic pyelonephritis or vesicoureteric reflux, and typically presents with type 1 (distal) RTA, associated with hypokalaemia, renal stone formation, and nephrocalcinosis (Eknoyan and Truong,

Chapter 86 

Table 86.1  Aetiology of CTIN Drugs (see Chapter 87)

Analgesics: see Chapter 87 NSAIDs: aspirin, COX-2 inhibitors (see Chapter 87) Antiviral agents (nucleoside inhibitors): cidofovir, tenofovir, adefovir Calcineurin inhibitors: ciclosporin, tacrolimus Antineoplastic agents: cisplatin, ifosfamide, nitrosourea, methotrexate Lithium

Infectious diseases

Acute bacterial pyelonephritis (Chapter 177), leptospirosis (Chapter 191), haemorrhagic fever with renal syndrome (Chapter 188), HIV/ AIDS (Chapter 187), tuberculosis (Chapter 196)

Immune-mediated diseases (Chapter 93)

Sarcoidosis, Sjögren’s syndrome, systemic lupus erythematosus, cryoglobulinaemia

Heavy metals (Chapter 88)

Lead, cadmium, mercury

Metabolic disorders

Hyperuricaemia/hyperuricosuria, hypercalcaemia/ hypercalciuria, hyperoxaluria, hypokalaemia, methylmalonic acidaemia

Haematologic disorders

Multiple myeloma (Chapter 153), light-chain deposition disease (Chapter 155), amyloidosis (Chapter 152), sickle-cell disease (Chapter 167)

Genetic disorders

Autosomal dominant interstitial kidney disease including medullary cystic kidney disease (Chapter 318), cystinosis (Chapter 339), Dent disease, adenine-phosphoribosyl-transferase deficiency, autosomal dominant hypoparathyroidism, karyomegalic interstitial nephropathy, mitochondrial mutations (Chapter 340), autosomal dominant polycystic kidney disease (Chapter 306)

Urinary tract obstruction (Chapter 356)

Tumours, stones, bladder outlet obstruction, vesicoureteric reflux (Chapter 355)

Miscellaneous

Balkan endemic nephropathy (Chapter 90), Chinese herbs nephropathy (aristolochic acid; Chapter 89), radiation nephropathy (Chapter 91)

Idiopathic AIDS = acquired immunodeficiency syndrome; COX = cyclooxygenase; HIV = human immunodeficiency virus; NSAIDs = non-steroidal anti-inflammatory drugs; SLE = systemic lupus erythematosus.

1999; Braden, 2005). In most cases of CTIN, variable combinations of the three tubular syndromes are seen. Low-molecular-weight proteinuria (usually < 1.5 g/day and not > 2.5 g/day) is characteristic, as well as leucocyturia, white blood cell casts, and haematuria (Rastegar and Kashgarian, 1998). During the course of the disease, the glomerular filtration rate (GFR) progressively decreases. Renal imaging (ultrasound, intravenous pyelogram, computed tomography, or magnetic resonance) may reveal urinary tract obstruction, vesicoureteric reflux, kidney stones, or parenchymal calcifications, as causes of CTIN. In advanced stages, the kidneys appear shrunken and may have irregular outlines. Additionally, a kidney biopsy is often needed for the diagnosis of CTIN and its aetiology.

chronic tubulointerstitial nephritis: overview

Pathology In patients with CTIN, the size of the kidneys usually decreases with the progression of the disease. The renal surface may be scarred, as in analgesic nephropathy and chronic pyelonephritis, or finely granular, as in Balkan endemic nephropathy. Papillary necrosis, sclerosis or calcification may be associated with analgesic nephropathy or chronic pyelonephritis. On light microscopy, tubular atrophy, interstitial fibrosis, and a variable degree of interstitial cell infiltrate are the hallmarks of CTIN. Tubular changes are usually patchy, with areas of atrophic tubules alternating with dilated tubules with compensatory hypertrophy. The atrophic tubules have variably thickened and lamellated tubular basement membrane (TBM), simplified and flattened epithelia, and sometimes the lumen is filled with an eosinophilic periodic acid–Schiff (PAS) stain-positive material (Nadasdy and Sedmak, 2007). The interstitial fibrosis may be focal or diffuse and the extracellular matrix may contain various types of collagen, derived from interstitial fibroblasts and from tubular epithelial cells (Tang et  al., 1994). A  variable interstitial infiltrate may be seen, consisting of activated T lymphocytes, macrophages, and, more rarely, B lymphocytes, plasma cells, and eosinophils (Eknoyan and Truong, 1999). In sarcoidosis, certain forms of vasculitis, and infections with mycobacteria, fungi, and parasites, interstitial granulomas may develop (Joss et al., 2007). Glomerular changes are often associated. Periglomerular fibrosis, thickening of the Bowman’s capsule, focal segmental glomerulosclerosis, or ischaemic glomerular lesions can be found. Vascular changes are also present in patients with hypertension. Immunofluorescence or immunohistochemistry can add useful information. Immunoglobulin and complement granular deposits may be seen along the TBM in some immune complex-mediated CTIN. Electron microscopy has limited diagnostic value; it shows the lamellated structure of the TBM and, possibly, the presence of granular aggregates at this level (Nadasdy and Sedmak, 2007).

Pathogenesis In CTIN, various triggering factors activate tubulointerstitial inflammation and repair mechanisms, followed by fibrosis and progressive renal parenchyma destruction (López-Novoa et al., 2011). Major causes fall into genetic, immune/autoimmune, infective, and toxic categories (see Table 86.1).

Genetics including Autosomal Dominant Interstitial Kidney Disease Some genetic causes are mentioned in Table 86.1. Genetic discoveries have emphasised how different genes may lead to similar or overlapping phenotypes and given rise to the concept of Autosomal Dominant Interstitial Kidney Disease (ADIKD) (see Chapter 318), which may be caused by mutations in a number of genes including UMOD, REN1, HNF1B, and MUC1. Some of these mutations add particular clinical features (e.g. typically gout, diabetes with UMOD, HNF1B), but others are characterized only by a clinically rather featureless CTIN.

Antigens No nephritogenic antigens are known for human interstitial nephritis, not even the target of the rare condition of anti-tubular

691

692

Section 4  

the patient with interstitial disease

basement membrane (anti-TBM) disease, in which an interstitial nephritis is accompanied by antibody fixation to the TBM, but not the GBM (Yoshioka et al., 2002; Remuzzi et al., 2008). Tubular injury may lead to expression of nephritogenic antigens, either derived from tubular cells and TBM or from exogenous antigens processed by tubular cells (Remuzzi et al., 2008). Exogenous and endogenous antigens processed by tubular epithelial cells may also trigger inflammatory reactions. Tubular cells may act as antigen-presenting cells in response to proinflammatory cytokines—including interferon (IFN)-gamma, interleukin (IL)-1, and tumour necrosis factor alpha (TNF-α)—may enhance the antigen-presenting capacity of tubular cells by increasing the expression of MHC class II molecules and of intercellular adhesion molecule (ICAM)-1 (Rubin-Kelley and Jevnikar, 1991). Furthermore, tubular cells may have co-stimulatory effects on the interstitial T lymphocytes, by expressing CD40, ICAM-1, and vascular cell adhesion molecule (VCAM)-1 (Remuzzi et al., 2008). Exposure of tubular epithelial cells to immune, chemical or biomechanical triggering factors leads to activation of NF-κB and downstream release of proinflammatory chemokines, cytokines and growth factors, such as plasminogen activator inhibitor (PAI)-1, IL-1, IL-6, monocyte chemoattractant protein-1 (MCP-1)/ chemokine ligand 2 (CCL2), CCL5 (RANTES), and TNF-α (Tashiro et al., 2003; Gong et al., 2004).

Urinary tract obstruction In obstructive uropathies, increased intratubular pressure and stretch activate tubular epithelial cells. As a consequence, transforming growth factor beta 1 (TGF-β1) is upregulated and induces epithelial-to-mesenchymal transition and tubulointerstitial fibrosis via tubular SMAD3 signalling. Other reactions to stretch have also been demonstrated in vitro, including activation of mitogen-activated protein kinases (MAPKs), with subsequent generation of arachidonic acid metabolites, caspase activation, and apoptosis. In addition, activation of epidermal growth factor receptor (EGFR) stimulates the production of inducible nitric oxide synthase (iNOS) via nuclear factor kappa B (NF-κB) and signal transducer and activator of transcription (STAT)-3 (Broadbelt et al., 2009; Rohatgi and Flores, 2010).

Cell-mediated immunity In experimental models, the initiation of the interstitial inflammatory process may vary, depending on the triggering agent and on the model design. Rag-2 null mice lack mature B and T lymphocytes and are protected from fibrosis induced by ureteric obstruction. Transfer of CD4+ T cells in these animals stimulates fibrogenesis (Tapmeier et al., 2010). Renal fibrosis after ischaemia-reperfusion injury also depends on persistent infiltration of activated and effector-memory T lymphocytes (Ascon et  al., 2009). Most frequently, the CD4+, CD8+, and CD3+ cells are the effector lymphocytes infiltrating the tubulointerstitial compartment involved in renal fibrogenesis. For example, CD8+ cells are considered the predominant effector cells in anti-TBM disease, in Heymann nephritis, and in the murine doxorubicin nephrosis (Remuzzi et al., 2008; Zeisberg and Neilson, 2010). In infection-induced CTIN, experimental unilateral renal artery stenosis, protein overload models, or reduction of renal mass, macrophages are initially the dominant infiltrating cells (Remuzzi et al.,

2008). For example, in the protein overload model of kidney disease, macrophage infiltration is an early process, whereas helper and cytotoxic T cells become involved only about 2 weeks later. The helper T cells tend to decrease in number after 3 weeks, and cytotoxic T cells after 7 weeks. It seems that T-cell depletion using anti-T-cell monoclonal antibodies does not affect the macrophage infiltration (Eddy, 1989). In this model, it seems that macrophage infiltration is dependent on signalling molecules expressed by tubular cells, like MCP-1, VCAM-1, and ICAM-1. A strong correlation between the degree of macrophage infiltration and the extension of fibrosis has been demonstrated in CTIN. Furthermore, macrophage depletion was shown to prevent fibrogenesis in mice (Duffield et al., 2005; Nishida and Hamaoka, 2008). Macrophages are versatile cells, which are able to modulate the expression of their surface receptor proteins and secreted cytokines in dependence of local stimuli and can fulfil various functions. Macrophages have been classified into M1 and M2. The M1 macrophages are attracted and activated by IFN-γ (synthesized mainly by T-helper (Th)-1 cells) and by lipopolysaccharides via toll-like receptor (TLR)-4 and are involved in fibrogenesis. Following their activation, these macrophages will stimulate iNOS activity and enhance their phagocytic capacity. Interleukin IL-l-α/β induces the release of IL-8 and CCL2, which further stimulate neutrophil and macrophage recruitment. The M2 macrophages are activated by interleukins IL-4, IL-13 and IL-10. The activation by IL-4 results in increased expression of scavenger receptor proteins (such as the mannose receptor, CD36) on the M2 cell. These cells have less phagocytic capacities, but higher macropinocytosis abilities, and their function seems to be the clearance of debris when the inflammatory process extinguishes (Ricardo et al., 2008). Mast cells and dendritic cells have also been found in interstitial infiltrates of patients with CTIN. Dendritic cells are antigen-presenting cells, playing an essential part in antigen processing after tubulointerstitial injury. During proteasomal processing of proteins, they may create new antigenic targets. Mast cells are involved in lung and liver fibrosis and in the synthesis of proinflammatory cytokines and chemokines; however, mast cell-deficient mice were shown to develop severe renal interstitial fibrosis, associated with high levels of TGF-β (Zeisberg and Neilson, 2010). The roles of mast and dendritic cells in CTIN are still unclear.

Fibrosis Interstitial fibrosis encompasses excess deposition of extracellular matrix components (collagen types I, III, V, VII, and XV, and fibronectin), components of the TBM (collagen IV and laminin), and de novo synthesized proteins (such as tenascin, fibronectin isoforms, and laminin chains), in parallel with a reduction of the renal parenchyma (López-Novoa et al., 2011). Upregulation of several cytokines, chemokines, and growth factors during the inflammatory response is the driving force for the recruitment and activation of fibroblasts. Several Th-2 cytokines (IL-4, IL-5, IL-13, and IL-2) play a very important role in this process, with IL-13 being considered the dominant effector (Blease et al., 2001). These Th-2 cytokines cooperate with TGF-β to induce fibrosis. On the other hand, Th-1-associated cytokines, IFN-γ and IL-12, as well as IL-10 (also known as human cytokine synthesis inhibitory factor) are fibrosis inhibitors (Wynn, 2008). TGF-β inhibits the expression of metalloproteases (MMPs) and

Chapter 86 

stimulates PAI-1, an MMP inhibitor (Roberts et al., 1992). A number of other molecules in the renal interstitium, including TGF-α, EGF, platelet-derived growth factor (PDGF), TNF-α, IL-1, IL-6, oncostatin M, endotoxin, and thrombin may enhance the expression of tissue inhibitor of metalloprotease (TIMP)-1 (Eddy, 2000). Overexpression of MMP-2 induces fibrosis (Cheng et  al., 2006), by degradation of TBM and promotion of epithelial-mesenchymal transition. Aberrant matrix synthesis by collagen remodelling cannot explain alone the reduction of kidney size in patients with CTIN. Most probably, the collapse of the kidney parenchyma also contributes to this process (Hewitson, 2009). Myofibroblasts (activated fibroblasts) synthesize extracellular matrix components and also remodel the matrix to increase its density via β-1-integrins (Kelynack et al., 1999). Interstitial myofibroblasts may have the origin in the perivascular area or may result from local cell proliferation, circulating mesenchymal cells, tubular epithelial cells (by epithelial-to-mesenchymal transition) or endothelial cells (by endothelial-to-mesenchymal transition). In a unilateral ureteral obstruction model of CTIN, around 38% of interstitial fibroblasts originated from epithelial-to-mesenchymal transition, 9% from circulating precursors, and 53% from local proliferation (Hewitson, 2009). The differentiation and proliferation of myofibroblasts, as well as collagen synthesis, is stimulated by TGF-β-1 and PDGF, which are released by tubular and interstitial cells and by infiltrating inflammatory cells (Boor et al., 2010).

Ischaemia An important role in the development of interstitial fibrosis is played by chronic ischaemia. Elevated synthesis of angiotensin II and decreased production of nitric oxide may contribute to ischaemia, by inducing vasoconstriction (Nangaku, 2006). In the early stages of CTIN, the renal vasculature also seems to be damaged by apoptosis. The developing interstitial fibrosis, the endothelial-tomesenchymal transition (with loss of endothelial cells), as well as the downregulation of vascular endothelial growth factor (VEGF), all contribute to the rarefaction of peritubular microvessels and ensuing interstitial hypoxia. With the expansion of the interstitial matrix, the diffusion distance of oxygen from the peritubular vessels to the tubular epithelia increases, thus further impairing the oxygen supply (Remuzzi et al., 2008). In its turn, hypoxia contributes to the progression of tubulointerstitial fibrosis by further stimulating the epithelial-to-mesenchymal transition and promoting matrix components synthesis by fibroblasts (Higgins et al., 2007). On the other hand, the endogenous hepatocyte growth factor (HGF) can decrease TGF-β1 expression and prevent epithelial-tomesenchymal transition (Mizuno et al., 2001).

Therapy Therapeutic intervention in CTIN has the following objectives and methods: 1. Stopping the action of triggering agents—for example, early discontinuation of nephrotoxic drugs and exposure to nephrotoxins, treating pyelonephritis with antibiotics, or surgical removal of urinary tract obstructions 2. Specific treatment of underlying systemic autoimmune, haematological, and metabolic diseases 3. Reducing the progression of interstitial inflammation and fibrosis.

chronic tubulointerstitial nephritis: overview

Angiotensin II exerts profibrotic effects in the heart, liver, and kidney, by directly inducing NADPH oxidase activity and synthesis of reactive oxygen species, by increasing the expression of TGF-β1, and by triggering fibroblast proliferation and activation. Angiotensin-converting enzyme inhibitors and angiotensin II receptor blockers have been shown to reduce cardiac, hepatic, and renal fibrogenesis and to slow down the progression of CKD, though not specifically in ATIN (Wynn, 2008). Other interventions, such as blood pressure control, reduction of proteinuria, and control of blood glucose and lipids are recommended (Hewitson, 2009). Regression of established fibrosis by stimulating collagen degradation is a possible therapeutic approach for CTIN. Pirfenidone, a pyridone compound, may reduce fibrosis by inhibiting fibroblast growth factor (FGF), EGF, PDGF, and TGF-β1. In 2011, the European Commission approved the use of pirfenidone in Europe for idiopathic pulmonary fibrosis. In some experimental studies of kidney disease (anti-Thy-1 glomerulonephritis and the ureteral obstruction model), pirfenidone has produced interesting results (Shimizu et al., 1997, 1998). More recent clinical studies tend to confirm its efficacy in patients with focal segmental glomerulosclerosis (Cho et al., 2007) and diabetic nephropathy (Sharma et al., 2011). Encouraging results have also been obtained in experimental models with relaxin, a naturally occurring hormone (Hewitson et al., 2010). It seems that relaxin signalling, by inhibiting SMAD2 phosphorylation, can interfere with TGF-β1-mediated renal myofibroblast differentiation and collagen production (Mookerjee et al., 2009).

References Ascon, M., Ascon, D. B., Liu, M., et al. (2009). Renal ischemia-reperfusion leads to long term infiltration of activated and effector-memory T lymphocytes. Kidney Int, 75, 526–35. Baker, R. J. and Pusey, C. D. (2004). The changing profile of acute tubulointerstitial nephritis. Nephrol Dial Transplant, 19, 8–11. Boor, P., Ostendorf, T., and Floege, J. (2010). Renal fibrosis: novel insights into mechanisms and therapeutic targets. Nat Rev Nephrol, 6, 643–56. Broadbelt, N. V., Chen, J., Silver, R. B., et al. (2009). Pressure activates epidermal growth factor receptor leading to the induction of iNOS via NFkappaB and STAT3 in human proximal tubule cells. Am J Physiol Renal Physiol, 297, F114–24. Cheng, S., Pollock, A. S., Mahimkar, R., et al. (2006). Matrix metalloproteinase 2 and basement membrane integrity: a unifying mechanism for progressive renal injury. FASEB J, 20, 1898–1900. Cho, M. E., Smith, D. C., Branton, M. H., et al. (2007). Pirfenidone slows renal function decline in patients with focal segmental glomerulosclerosis. Clin J Am Soc Nephrol, 2, 906–13. Covic, A., Schiller, A., Volovat, C., et al. (2006). Epidemiology of renal disease in Romania: a 10 year review of two regional renal biopsy databases. Nephrol Dial Transplant, 21(2), 419–24. Duffield, J. S., Forbes, S. J., Constandinou, C. M., et al. (2005). Selective depletion of macrophages reveals distinct, opposing roles during liver injury and repair. J Clin Invest, 115, 56–65. Eddy, A. A. (1989). Interstitial nephritis induced by protein-overload proteinuria. Am J Pathol, 135, 719–33. Eddy, A. A. Molecular basis of renal fibrosis. Pediatr Nephrol, 15, 290–301. Eknoyan, G. and Truong, L. D. (1999). Renal interstitium and major features of chronic tubulointerstitial nephritis. In R. J. Glassock, A. H. Cohen, and J. P. Grünfeld (eds.) Atlas of Diseases of the Kidney (Vol. 2), pp. 6.1–6.25. New York: Current Medicine Inc.

693

694

Section 4  

the patient with interstitial disease

Gong, R., Rifai, A., Tolbert, E. M., et al. (2004). Hepatocyte growth factor ameliorates renal interstitial inflammation in rat remnant kidney by modulating tubular expression of macrophage chemoattractant protein-1 and RANTES. J Am Soc Nephrol, 15, 2868–81. Hewitson, T. D. (2009). Renal tubulointerstitial fibrosis: common but never simple. Am J Physiol Renal Physiol, 296, F1239–44. Higgins, D. F., Kimura, K., Bernhardt, W. M., et al. (2007). Hypoxia promotes fibrogenesis in vivo via HIF-1 stimulation of epithelial-tomesenchymal transition. J Clin Invest, 117, 3810–20. Joss, N., Morris, S., Young, B., et al. (2007). Granulomatous interstitial nephritis. Clin J Am Soc Nephrol, 2, 222–30. Kelynack, K. J., Hewitson, T. D., Pedagogos, E., et al. (1999). Renal myofibroblasts contract collagen I lattices in vitro. Am J Nephrol, 19, 694–701. López-Novoa, J. M., Rodríguez-Peña, A. B., Ortiz, A., et al. (2011). Etiopathology of chronic tubular, glomerular and renovascular nephropathies: clinical implications. J Transl Med, 9, 13. Markowitz, G. S. and Perazella, M. A. (2005). Drug-induced renal failure: a focus on tubulointerstitial disease. Clin Chim Acta, 351, 31–47. Mizuno, S., Matsumoto, K., and Nakamura, T. (2001). Hepatocyte growth factor suppresses interstitial fibrosis in a mouse model of obstructive nephropathy. Kidney Int, 59, 1304–14. Mookerjee, I., Hewitson, T. D., Halls, M. L., et al. (2009). Relaxin inhibits renal myofibroblast differentiation via RXFP1, the nitric oxide pathway, and Smad2. FASEB J, 23, 1219–29. Murry, T. and Goldberg, M. (1975). Chronic interstitial nephritis: etiologic factors. Ann Intern Med, 82, 453–459. Nadasdy, T. and Sedmak, D. (2007). Acute and chronic tubulointerstitial nephritis. In J. C. Jennette, J. L. Olson, M. M. Schwartz, et al. (eds.) Hepinstall’s Pathology of the Kidney (6th ed.), pp. 1084–137. Philadelphia, PA: Lippincott, Williams & Wilkins. Nagai, J., Christensen, E. I., Morris, S. M., et al. (2005). Mutually dependent localization of megalin and Dab2 in the renal proximal tubule. Am J Physiol Renal Physiol, 289, F569–76. Nangaku, M. (2006). Chronic hypoxia and tubulointerstitial injury: a final common pathway to end-stage renal failure. J Am Soc Nephrol, 17, 17–25. Nishida, M. and Hamaoka, K. (2008). Macrophage phenotype and renal fibrosis in obstructive nephropathy. Nephron Exp Nephrol, 110, e31–6. Rastegar, A. and Kashgarian, M. (1998). The clinical spectrum of tubulointerstitial nephritis. Kidney Int, 54, 313–27. Remuzzi, G., Perico, N., and De Broe, M. E. (2008). Tubulointerstitial diseases. In B. M. Brenner (ed.) Brenner & Rector’s The Kidney (8th ed.), pp. 1174–202. Philadelphia, PA: Saunders. Ricardo, S. D., van Goor, H., and Eddy, A. A. (2008). Macrophage diversity in renal injury and repair. J Clin Invest, 118, 3522–30.

Roberts, A. B., McCune, B. K., and Sporn, M. B. (1992). TGF-beta: regulation of extracellular matrix. Kidney Int, 41, 557–9. Rohatgi, R. and Flores, D. (2010). Intratubular hydrodynamic forces influence tubulointerstitial fibrosis in the kidney. Curr Opin Nephrol Hypertens, 19, 65–71. Rubin-Kelley, V. E. and Jevnikar, A. M. (1991). Antigen presentation by renal tubular epithelial cells. J Am Soc Nephrol, 2, 13–26. Rychlík, I., Jancová, E., Tesar, V., et al. (2004). The Czech registry of renal biopsies. Occurrence of renal diseases in the years 1994-2000. Nephrol Dial Transplant, 19, 3040–9. Sharma, K., Ix, G. H., Mathew, A. V, et al. (2011). Pirfenidone for diabetic nephropathy. J Am Soc Nephrol, 22, 1144–51. Shimizu, F., Fukagawa, M., Yamauchi, S., et al. (1997). Pirfenidone prevents the progression of irreversible glomerular sclerotic lesions in rats. Nephrology, 3, 315–22. Shimizu, T., Kuroda, T., Hata, S., et al. (1998). Pirfenidone improves renal function and fibrosis in the post-obstructed kidney. Kidney Int, 54, 99–109. Stengel, B., Billon, S., van Dijk, P. C. W., et al. (2003). Trends in the incidence of renal replacement therapy for end-stage renal disease in Europe, 1990–1999. Nephrol Dial Transplant, 18, 1824–33. Tang, W. W., Feng, L., Xia, Y., et al. (1994). Extracellular matrix accumulation in immune-mediated tubulointerstitial injury. Kidney Int, 45, 1077–84. Tapmeier, T. T., Fearn, A., Brown, K., et al. (2003). Pivotal role of CD4(+) T cells in renal fibrosis following ureteric obstruction. Kidney Int, 78, 351–62. Tashiro, K., Tamada, S., Kuwabara, N., et al. (2003). Attenuation of renal fibrosis by proteasome inhibition in rat obstructive nephropathy: possible role of nuclear factor kappa B. Int J Mol Med, 12, 587–92. United States Renal Data System (2010). USRDS 2010 Annual Data Report: Atlas of End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Diabetes and Digestive and Kidney Diseases. Weening, J. J. and Jennette, J. C. (2012). Historical milestones in renal pathology. Virchows Arch, 461, 3–11. Wynn, T. A. (2008). Cellular and molecular mechanisms of fibrosis. J Pathol, 214, 199–210. Yoshioka, K., Takemura, T., and Hattori, S. (2002). Tubulointerstitial nephritis antigen: primary structure, expression and role in health and disease. Nephron, 90, 1–7. Zeisberg, M. and Neilson, E. G. (2010). Mechanisms of tubulointerstitial fibrosis. J Am Soc Nephrol, 21, 1819–34. Zollinger, H. U. and Mihatsch, M. J. (1978). Renal Pathology in Biopsy. Berlin: Springer-Verlag.

CHAPTER 87

Drug-induced chronic tubulointerstitial nephritis Hassan Izzedine and Victor Gueutin Introduction Several drugs can cause chronic tubulointerstitial nephritis (CTIN), including analgesics, lithium, antineoplastic chemotherapeutic agents (cisplatin, nitrosoureas), and immunosuppressive drugs (ciclosporin, tacrolimus). Many of the drugs that cause acute tubulointerstitial nephritis (ATIN) may also induce CTIN. Patients who develop the acute form of drug-induced TIN generally recover fully (Fig. 87.1). A few, however, do not recover and progress to CTIN; such is the case of cisplatin toxicity. On the other hand, drugs like analgesics, lithium, and nitrosourea, typically induce CTIN only after prolonged exposure (several months or years); in such cases, the renal disease has an insidious onset and a slow course. There are several ways in which injury to the tubulointerstitium can occur, and these can involve either immune-mediated or non-immune-mediated (direct toxicity) mechanisms. The diagnosis largely depends on the history of exposure to such nephrotoxic drugs. The recognition of a potential association between a patient’s renal disease and the previous administration of certain drugs is crucial, because, unlike in other forms of renal disease, the progression of these nephropathies can be prevented and even reversed, by simply avoiding additional exposure. The functional abnormalities depend on the site of the nephron that is mainly involved, which in its turn depends on the offending drug; for example, proximal tubular dysfunction is associated with cidofovir toxicity, distal tubular dysfunction (with salt wasting, acidosis, and hyperkalaemia) is commonly seen in lithium nephropathy, whereas medullary injury with impaired urine concentrating ability is characteristic of analgesic nephropathy. Renal biopsy can confirm the diagnosis. Although renal biopsy is indispensable for assessing the severity of pathologic lesions in drug-induced CTIN, it is not acceptable in some cases and cannot be performed serially because of its invasive nature. In those cases, urinary monocyte chemotactic peptide-1 (MCP-1) levels correlated with and were predictive of the severity of acute lesions in drug-induced TIN, whereas neutrophil gelatinase-associated lipocalin (NGAL) and α1-microglobulin levels showed the highest correlation coefficient with tubular atrophy (Wu et al., 2010). The lack of effective therapies for advanced CTIN, in general, highlights the importance of making an early diagnosis, when the progression of the disease can effectively be stopped or even

reversed, essentially by avoiding any further exposure of the patient to the offending drug.

Analgesic nephropathy In the 1970s and 1980s, analgesic nephropathy was the cause of end-stage renal disease (ESRD) in up to 20% of patients on dialysis in some countries (including Australia and Belgium), but it has now become a relatively rare condition, following market withdrawal of phenacetin in most countries (Table 87.1). Although initially thought to be exclusively associated with phenacetin-containing combinations, analgesic nephropathy can also be caused by other drugs, including acetaminophen, aspirin, and non-steroidal anti-inflammatory drugs (NSAIDs) (De Broe and Elseviers, 2009). For half a century, a large number of epidemiologic studies have linked prolonged and excessive consumption of analgesic mixtures to a renal disease characterized by papillary necrosis and CTIN (Sandler et al., 1989; Perneger et al., 1994; Elseviers and De Broe, 1995, Elseviers et al., 1995). The incidence varies greatly from study to study, depending primarily on the region or country where the investigation was performed. In Europe, the percentage of analgesic nephropathy among patients with ESRD undergoing long-term dialysis varies widely, from only 0.1% in Ireland, Norway, Poland, and Hungary to 18.1% in Switzerland (Elseviers and De Broe, 1993). According to the Analgesic Nephropathy Network of Europe study, the average European incidence of analgesic nephropathy among patients who were started on renal replacement therapy in 1991 to 1992 was 6.4% (Elseviers et al., 1995). In Australia and Canada, 11% and 2.5% incidence rates have been reported, respectively (Gault and Wilson, 1978; Kincaid-Smith, 1990). In the United States, 1.7– 10% of the ESRD cases are thought to be the result of analgesic nephropathy in various regions (Gonwa et al., 1981; Perneger et al., 1994).These large geographic differences may be explained by differences in local habits, psychosocial factors, availability of these drugs, and probably also the frequency of correct diagnosis and reporting. Analgesic nephropathy occurs in about 4 out of 100,000 people with long-term consumption of large amounts of analgesics, most often combinations of acetaminophen (paracetamol), phenacetin, aspirin, and NSAIDs. The nephrotoxicity of phenacetin is likely to be due to its major metabolite, paracetamol. Aspirin achieves higher medullary and cortical concentrations than paracetamol (Elseviers et al., 1995). The risk of developing renal disease is dependent on

696

Section 4  

the patient with interstitial disease

Fig. 87.1  Drug-induced TIN with marked chronicity. The tubules appear shrunken and atrophic, and are separated by extensive interstitial fibrosis. (TM; magnification ×400.)

the frequency and duration of analgesic abuse. In epidemiologic studies, patients with analgesic nephropathy used analgesics virtually every day for several years, so that the cumulative amount of individual analgesic intake was > 1–3  kg. A  total dose of at least 1  kg of phenacetin (i.e. 1 g/day for 3  years) is necessary to cause analgesic nephropathy (Elseviers et al., 1995). This usually occurs as a result of self-medication, for some type of chronic pain. The prospective controlled longitudinal epidemiological studies reported by Dubach et al. (1975, 1978) clearly demonstrated a significant association between analgesic consumption and renal disease. Nevertheless, in only a small number of analgesic abusers does nephropathy develop, and progression of chronic kidney disease (CKD) may be stopped with discontinuation of analgesics. Nowadays, the decrease in availability of phenacetin-containing and other analgesic mixtures has led to a marked reduction in the number of new cases of analgesic nephropathy. The clinical manifestations associated with analgesic abuse are well recognized and have been reviewed extensively (Nanra, 1993). Analgesic nephropathy is most common in women (85% of cases) in the fourth and fifth decades of life, who have a history of low back pain, migraine headaches, or other chronic musculoskeletal pain. Analgesic nephropathy may show a familial clustering and the increased prevalence of human leucocyte antigen (HLA)-B12 in patients with this disease suggests a possible role for some genetic factor(s) (Nanra, 1993). Table 87.1  Clinical findings in analgesic nephropathy Finding

Frequency

Headache

35–100%

Pyuria

50–100%

Anaemia

60–90%

Hypertension

15–70%

Gastrointestinal symptoms

40–60%

Urinary tract infection

30–60%

From Murray and Goldberg (1978).

Analgesic nephropathy is often asymptomatic. The clinical manifestations are non-specific and may include slowly progressive CKD, sterile pyuria, mild proteinuria, and haematuria (Table 87.1). Virtually all patients with analgesic nephropathy have a urine concentrating defect, and 25% also have an acidifying defect; however, frank renal tubular acidosis is seen in only 10% of cases, when the glomerular filtration rate (GFR) is significantly reduced. These tubular function abnormalities may be associated with clinical manifestations and complications like nocturia, muscle cramps, renal stones, medullary nephrocalcinosis, and renal osteodystrophy. Clinical gout occurs in 4.5% of patients with normal renal function and in 26.5% of those with low GFR. Urinary tract infections are a late complication, which occurs in 30–60%, and may be recurrent (Murray and Goldberg, 1978). Haematuria is found in up to 35% of patients (Nanra, 1980) and may be related to urinary tract infections, stones, malignant hypertension, associated glomerular disease, or uroepithelial tumours. In some patients, episodes of papillary necrosis may occur, manifested with gross haematuria and flank pain, occasionally accompanied by obstruction and infection (Griffin et al., 1995). Glomerular disease may be associated with CTIN in approximately 60% of patients with proteinuria > 3.5–5.0 g per day. Fifteen to 70% of patients develop hypertension, sometimes secondary to renal artery stenosis. Malignant hypertension has been observed in 6.9% of cases. Urinary tract obstruction is a serious complication and may be due to a fragment of necrotic or calcified papilla, a stone, a transitional cell pelvic or ureteral tumour, or a postinflammatory ureteral stricture. Concomitant urinary tract infection may lead to septicaemia or pyonephrosis, potentially fatal complications (Nanra, 1993). The diagnosis of analgesic nephropathy should not be solely based on renal biopsy. Renal imaging techniques, such as sonography and particularly computed tomography, are the best methods for diagnosis in the appropriate clinical context (Elseviers et al., 1995). The Analgesic Nephropathy Network of Europe study showed that shrinkage of renal mass, bumpy renal contours and the presence of papillary calcifications are the most useful criteria in diagnosing analgesic nephropathy, with a sensitivity of 96%, 57%, and 85%, respectively, and a specificity of 37%, 92%, and 93%, respectively. The combination of these three criteria resulted in a sensitivity of 85% and a specificity of 93% (Elseviers et al., 1995; De Broe and Elseviers, 1998). Radiocontrast examinations may be helpful in the diagnosis of papillary necrosis. The pathological renal changes in analgesic nephropathy have been well documented. The primary lesion is renal papillary necrosis; the CTIN lesions are secondary, resulting from obstruction. Renal papillary necrosis extending into the medulla may involve several papillae (Nanra, 1993; Griffin et al., 1995). Gross appearance and early light microscopic findings are most distinctive of this entity. The kidneys are small and shrunken, with irregular contours and papillary calcifications. On histology, the capillaries beneath the urothelium in the renal pelvis exhibit basement membrane thickening and calcification. This vascular injury leads to papillary ischaemia and eventual necrosis. Necrotic tissue may slough off or become calcified. There is compensatory hypertrophy of the columns of Bertin, while the suprapapillary cortex undergoes atrophy. Non-renal manifestations of the analgesic syndrome include gastrointestinal manifestations (peptic ulcer in 40%), haematological

Chapter 87 

abnormalities (anaemia in 60–90%, which may be haemolytic, splenomegaly in 10%), headache (80%), psychiatric disorders (90%), cardiovascular complications, premature ageing, skin hyperpigmentation, and gonadal and pregnancy-related manifestations. Malignancy may occur after 20 years of analgesic abuse, on average, in about 10% of patients. However, it can sometimes occur even long after cessation of analgesic consumption (Bengtsson et al., 1978). The major analgesic-associated tumour is transitional cell carcinoma of the uroepithelium; however, hypernephroma, sarcoma, and chorioepithelioma have also been reported (Bengtsson et  al., 1978). The tumour tends to be multifocal and, in 5% of cases, bilateral simultaneous renal pelvic carcinomas have been described (Bengtsson et  al., 1978). In one study in patients with analgesic-associated tumours, the mean intake of phenacetin was 9.1 kg, the mean drug exposure time was 17 years, and the mean induction time was 21 years (Bengtsson et al., 1978). The diagnosis of analgesic nephropathy obviously relies on the history of heavy analgesic abuse. Computed tomography may reveal reduced renal size, ‘bumpy’ renal contours, as well as bilateral microcalcifications at the papillary tips and ring shadows, typical of papillary necrosis. These findings are usually lacking or less prominent in other forms of CTIN (Mackinnon et al., 2003; Pintér et al., 2004). The exact pathogenesis of the toxicity of analgesic compounds and the primary target of the toxic reactions are unknown. Inhibition of prostaglandin synthesis and immunologic reactions are unlikely causes (Mihatsch and Zollinger, 1993). It is possible that metabolites of phenacetin, aspirin, or paracetamol, under the influence of cytochrome P450 mono-oxygenase, bind covalently to cellular proteins and cause toxic damage (Nanra, 1993). Long-term follow-up studies have shown that the main complications of long-term analgesic abuse are progression to ESRD, accelerated atherosclerosis, and increased incidence of uroepithelial carcinomas. In a study of 323 consecutive patients with analgesic nephropathy followed for up to 66 months, the renal function improved in 17%, remained stable in 50%, and worsened in 23% of cases; 12% presented with ESRD and either died or required initiation of dialysis within 6 months (Nanra, 1980). Nanra (1980) found that patients with analgesic nephropathy and ESRD, treated either with dialysis or with renal transplantation had worse outcomes in comparison with patients with glomerulonephritis, experiencing a significantly higher mortality rate over a 6-year period (50.8% vs 15.8%). There is no specific therapy for analgesic nephropathy. Treatment is supportive and includes discontinuation of analgesic use and abundant fluid intake. The decline in GFR can be expected to progress if drug consumption is continued. On the other hand, the renal function can stabilize or even improve in many patients if analgesic abuse is stopped in time (De Broe and Elseviers, 1998).

Lithium nephropathy Lithium is commonly used in the treatment of bipolar disorder. It may induce acute kidney injury, as well as CKD. Possible renal complications of lithium treatment are shown in Box 87.1. The major risk factors for lithium nephrotoxicity appear to be the duration of drug exposure and the cumulative dose (Presne et al., 2003); other risk factors include episodes of acute intoxication, older age, comorbidity (such as hypertension, diabetes mellitus,

drug-induced chronic tubulointerstitial nephritis Box 87.1  Lithium nephrotoxicity ◆ Nephrogenic diabetes insipidus and impairment of urinary concentration ◆ Incomplete distal renal tubular acidosis ◆ Chronic tubulointerstitial nephropathy ◆ Hypercalcaemia ◆ Distal tubular microcysts ◆ Acute kidney injury ◆ Glomerulonephritis. hyperparathyroidism, and hyperuricaemia), and concomitant use of other antipsychotic medications. The first suggestion that progressive impairment of GFR may occur in lithium-treated patients came from Hestbech et al. (1977). Long-term lithium use is associated with CKD in 15–20% of patients, who develop a slow decline in GFR (Boton et al., 1987; Bendz et al., 1994; Presne et al., 2003) of about 3 mL/min per year and an average course to ESRD of 20 years (Presne et al., 2003). The degree of interstitial fibrosis on renal biopsy may be directly related to the duration of therapy and the cumulative dose of lithium (Presne et al., 2003). Among patients with affective disorders, Walker et al. (1982) found that those treated with lithium had higher serum creatinine and β2-microglobulin and lower 51Cr-EDTA clearance, compared to those not treated with lithium. However, they failed to show any difference in renal histology between the two groups, except for microcyst formation in the lithium-treated patients. In rats, prolonged lithium administration at high doses was associated with an increase in the size of the tubules and of the kidneys (Kling et al., 1984), but no evidence of nephron loss or progressive interstitial lesions. Rabbits treated with lithium chloride (50–250 mmol/kg of food, over 12 months) developed significant interstitial fibrosis, tubular atrophy, glomerular sclerosis, and cystic tubular lesions. Microcysts have been demonstrated on magnetic resonance imaging and ultrasonographic studies (Farres et al., 2003) and located histologically in the distal and collecting tubules. Nephrogenic diabetes insipidus (with polyuria, polydipsia, and impaired renal concentrating capacity) is the most usual renal complication of maintenance lithium therapy (Walker, 1993). This disorder results from downregulation of aquaporin-2 water channels in the collecting duct (Christensen et al., 2004) and occurs in up to 50% of patients receiving lithium therapy (Boton et al., 1987). Impaired concentrating ability is usually reversible after lithium discontinuation, although it may persist for as long as 12–18 months (Rabin et al., 1979). CTIN may develop in a small subset of patients who have had frequent episodes of acute lithium toxicity, with high serum drug levels (Hestbech et al., 1977; Bucht et al., 1980), and it is probably the result of repeated tubulointerstitial injury and repair. Cases of lithium-induced nephrotic syndrome have rarely been reported (Tam et  al., 1996; Markowitz et  al., 2000; Presne et  al., 2003). A study from the Columbia University showed that 25% of patients who underwent kidney biopsy and were diagnosed with lithium nephropathy also had nephrotic syndrome (Markowitz et al., 2000); these patients had a histologic pattern of focal segmental glomerulosclerosis. The nephrotic syndrome could be the

697

698

Section 4  

the patient with interstitial disease

result of lithium interaction with anionic sites on the glomerular basement membrane (Tam et al., 1996). Lithium nephropathy appears to be a slowly progressive disease, unless lithium administration is stopped early enough. Progression to ESRD may occur even after lithium discontinuation in patients with an initial serum creatinine > 2.5 mg/dL (221  μmol/L). Additionally, systemic and intraglomerular hypertension may induce secondary glomerulosclerosis (Hansen et  al., 1979; Markowitz et  al., 2000), thus contributing to the progression of CKD. One study showed that the prevalence of lithium nephropathy is 0.2% in ESRD patients on maintenance dialysis (Presne et al., 2003). The authors calculated that the duration of lithium intake until ESRD was 19.8 years and the estimated cumulative lithium dose was 5231 g per patient. Annual monitoring of serum sodium, creatinine, and GFR is recommended in patients receiving lithium therapy. Polyuria and polydipsia usually resolve rapidly following drug withdrawal. However, since lithium is so clearly beneficial in most treated psychiatric patients, polyuria is often considered an acceptable side effect and does not prompt the discontinuation of therapy. On the other hand, it is more difficult to decide appropriate management in a patient who has been on lithium for many years and in whom there is evidence of progressive glomerular and tubular dysfunction. Close monitoring of serum lithium is essential, because nephrotoxicity is usually dose-dependent; maintaining levels between 0.4 and 0.8 mmol/L is recommended. Since renal handling of lithium resembles that of sodium, the elevation of its serum levels usually occurs in states of volume depletion, renal insufficiency, and concomitant therapy with diuretics and/or NSAIDs. Amiloride may be used in the treatment of lithium-associated polyuria, since it prevents lithium entry into the distal tubule. Caffeic acid phenethyl ester (CAPE), a known component of honeybee propolis, can be protective against oxidative stress in ischaemia-reperfusion and toxic renal injuries; used in experimental rat models, CAPE was shown to prevent lithium-induced tubular damage (Oktem et  al., 2005). N-acetylcysteine, a drug that is effective in preventing radiocontrast-induced nephropathy, was also capable to reduce lithium-induced tubular injury in Sprague–Dawley rats (Efrati et al., 2005).

Calcineurin inhibitor-induced nephropathy Although indispensable in the management of solid organ transplantation, calcineurin inhibitors ciclosporin and tacrolimus can cause acute and chronic nephrotoxicity. The mechanism appears to be largely dependent on the potent vasoconstrictive effects of these drugs. CTIN induced by ciclosporin or tacrolimus is common among patients receiving kidney, heart, liver, and pancreas transplants. In renal transplant recipients, ciclosporin- and tacrolimus-induced CTIN is similar to chronic allograft nephropathy. Most of these patients have a slow course, with mild impairment of renal function remaining stable for a long time. On the other hand, up to 10% of heart transplant recipients develop rapidly progressive renal insufficiency and eventually require dialysis. This condition is rare in bone marrow transplant recipients, because such patients receive these immunosuppressive drugs for a short time and generally at lower doses. Patients treated with calcineurin inhibitors are at high risk of developing renal injury (Burdmann et al., 2003), manifested either

as acute azotaemia, which is largely reversible after dose reduction, or as chronic progressive renal disease, usually irreversible (Kopp and Klotman, 1990; de Mattos et al., 2000, Naesens et al., 2009). Chronic nephrotoxicity typically occurs after 6–12 months of therapy. A similar pattern of renal injury is seen with the use of both ciclosporin and tacrolimus, suggesting a drug class effect. However, tacrolimus has less renal toxicity at lower doses, without compromising overall outcomes (Ekberg et al., 2007; Shihab et al., 2008). The pathologic features of calcineurin inhibitors-induced chronic nephropathy include vascular changes (arteriolopathy), associated with patchy (striped) interstitial fibrosis, tubular atrophy, and glomerular sclerosis (Burdmann et al., 2003). Ciclosporin arteriolopathy is characterized by thickening of the arteriolar wall, infarction of myocytes, protein deposits in the vessel wall, and hyalinosis. Such vascular changes compromise the blood supply to the tubulointerstitium, possibly setting the stage for the development of CTIN. Ciclosporin is directly toxic to the tubular epithelium, where it induces epithelial vacuolization and swelling of mitochondria, and it also promotes the expression of mRNA for type I collagen, which may contribute to renal fibrosis. Although the pathogenesis of such lesions is clearly multifactorial, a dominant factor is ciclosporin-induced vasoconstriction, an effect that is directed largely to the afferent arteriole and one that reduces the GFR and renal blood flow. Tubular blood supply is compromised as postglomerular blood flow is reduced, thereby incurring tubular ischaemia. The factors responsible for chronic calcineurin inhibitor nephrotoxicity are not well understood. The development of interstitial fibrosis is associated with increased expression of osteopontin, a potent macrophage chemoattractant secreted by the tubular epithelial cells (Pichler et al., 1995), chemokines, a class of cytokines that are strong chemoattractants for a variety of haematopoietic cells (Benigni et  al., 1999), and transforming growth factor-beta (TGF-β), a powerful stimulator of extracellular matrix production (Shihab et al., 1997; Islam et al., 2001). TGF-β appears to be induced in part by decreased secretion of nitric oxide (Shihab et al., 2000), as well as by increased local concentrations of angiotensin II, possibly explaining some of the beneficial effects observed with angiotensin-converting enzyme inhibitors (ACEIs) and angiotensin II receptor antagonists (Burdmann et al., 1995; Shihab et al., 1997). Ciclosporin is a substrate for the transmembrane pump P-glycoprotein. There is some experimental evidence that decreased expression of this pump may contribute to increased ciclosporin levels, leading to nephrotoxicity (Del Moral et al., 1998; Koziolek et al., 2001), as in several polymorphisms of its gene (Hauser et al., 2005). This suggests that underlying genetic factors that increase ciclosporin concentrations in the kidney may play a role in chronic nephrotoxicity. Short-term studies suggest that low doses of ciclosporin may not lead to renal dysfunction (Deray et al., 1992; Ekberg et al., 2007), as shown in the CAESAR study (Ekberg et al., 2007). The replacement of ciclosporin with non-nephrotoxic immunosuppressive agents may improve renal dysfunction in patients with ciclosporin-induced nephrotoxicity. Several metabolic abnormalities are recognized as complications of calcineurin inhibitors nephrotoxicity, including hyperkalaemic (type IV) renal tubular acidosis, renal magnesium wasting, and hyperuricaemia by decreased urate clearance, which can be severe enough to provoke gout attacks and tophi. Both ciclosporin and tacrolimus frequently cause hypertension. Sometimes, thrombotic

Chapter 87 

microangiopathy may also occur, as a result of the vasoconstrictive, salt-retaining, and nephrotoxic effects of these drugs, and can further contribute to the renal disease. Many agents have been tried aiming to reduce the nephrotoxic effects of calcineurin inhibitors, including fish oil, calcium channel blockers, thromboxane synthesis inhibitors, and pentoxifylline; however, none of these has proved to be clearly effective. Animal and human studies suggest that concurrent administration of calcium channel blockers may be protective against ciclosporin nephrotoxicity, probably by counteracting the renal vasoconstriction (Palmer et al., 1991); on the other hand, some calcium-channel blockers (such as verapamil, diltiazem, and nicardipine) may cause elevations in plasma ciclosporin concentrations. ACEIs seem to be less effective and carry the risk of hyperkalaemia in patients who may have ciclosporin-induced type IV renal tubular acidosis. The best way of minimizing calcineurin inhibitors nephrotoxicity is to reduce the doses and target trough levels of these drugs. Completely stopping their administration or switching to other immunosuppressive agents (like rapamycin), especially in patients with more advanced renal disease, should also be considered.

Aminosalicylates The association between the use of 5-aminosalicylic acid (5-ASA) and the development of CTIN in patients with inflammatory bowel disease (IBD) gained recognition in the 1990s, after the publication of several case reports. The adverse effects of 5-aminosalicylates are similar and include the common occurrence of fever and rash in > 10% of patients. Hypersensitivity responses have been described in multiple organ systems, most commonly the kidney (Moss and Peppercorn, 2007). Aminosalicylate-associated nephrotoxicity most frequently takes the form of an indolent, slowly progressive CTIN (Corrigan et al., 2000; Arend and Springate 2004). The disease is more prevalent in males, with a male/female ratio of 5.3/1. The age of reported cases ranged from 14 to 45  years. There is no relationship between the duration of 5-ASA treatment and the risk of renal disease (Riley et  al., 1992; Ransford and Langman, 2002). In contrast with analgesic nephropathy, where renal lesions were only observed after several years of drug abuse, CTIN associated with 5-ASA may occur during the first year of treatment in 50% of cases (World et al., 1996; Corrigan and Stevens, 2000; Cunliffe, 2002). The incidence of TIN among patients taking 5-aminosalicylates is between 1 in 200 to 1 in 500 patients (World et  al. 1996; Arend and Springate, 2004; Gisbert et al., 2007), although some studies suggest a much lower rate of occurrence. Elseviers et al. prospectively evaluated 1529 patients with IBD, followed up to 1 year, in 27 European centres. Although renal impairment occurred in 2.2% of patients, there was no relation with 5-ASA use. In fact, a possible association with 5-ASA had an estimated prevalence of 1.3–3.3 cases per 1000 patients (Elseviers et  al., 2004). The incidence of nephrotoxicity in IBD patients taking 5-ASA therapy seems to be < 0.5%, as shown by pooled data from 2671 patients receiving this treatment for a total of 3070 years of follow-up, and in whom serum creatinine or creatinine clearance were measured regularly (Gisbert et al., 2007). Based on data from the UK General Practice Research Database (Van Staa et al., 2004), mesalazine and sulfasalazine had comparable risks of nephrotoxicity in adult patients with IBD (0.17 vs 0.29 cases per 100 person-years, respectively).

drug-induced chronic tubulointerstitial nephritis It has been demonstrated that high pharmacological doses of 5-ASA induced necroses of proximal convoluted tubules and papillary necroses, similarly to salicylates—not surprisingly, as the molecular structure of 5-ASA is very close to that of salicylic acid, phenacetin, and acetaminophen (Bilyard et  al., 1990; Schreiber et al., 1997). Drug withdrawal leads to restoration of renal function in 60% of cases, if the diagnosis is made within 10 months from initiation of treatment (Gisbert et al., 2007), and renal outcome depends on the degree of renal damage at diagnosis (Gisbert et al., 2007). Serial monitoring of serum creatinine and urinalysis is recommended for all patients on 5-aminosalicylate therapy: before initiation of treatment, each month for the first 3 months of treatment, quarterly for the remainder of the first year, and annually (World et al. 1996; Corrigan et al., 2000) or bi-annually thereafter (Arend and Springate, 2004).

Antineoplastic agents Cisplatin is an agent used in the treatment of various solid tumours. The kidney is its major route of excretion. The drug and its metabolites are highly concentrated in the renal cortex, thereby predisposing to nephrotoxicity. Acute toxicity is usually reversible and may in part be derived from the vasoconstrictive effects of cisplatin. Hypomagnesaemia is one of the most serious side effects of the drug and could be life-threatening; it develops in over 70% of patients and persists for months in 50% of these, even after the drug is stopped. Toxicity is mitigated or prevented by adequate hydration, diuresis, and slow intravenous infusion of the drug. Nitrosoureas, carmustine (BCNU), lomustine (CCNU), and semustine (methyl-CCNU) are used in the treatment of malignant melanoma, brain tumours, and lymphomas. Dose-dependent nephrotoxicity may be insidious and occurs months after cessation of therapy. ESRD can occur in up to 50% patients after receiving 1.2–1.5 g/m² of semustine.

Others Cidofovir, a monophosphate nucleotide analogue of deoxycytidine, has been employed almost exclusively to treat cytomegalovirus retinitis in patients with the acquired immunodeficiency syndrome (AIDS) and is also being increasingly used as a therapeutic option against other viral infections. The most important adverse effect of cidofovir is dose-dependent nephrotoxicity. Approximately 50% of patients receiving cidofovir in clinical trials developed proteinuria, an increase in serum creatinine by at least 0.4 mg/dL or a decrease in GFR below 55 mL/min (Gilead Sciences, Inc., 1996). Renal dysfunction is usually reversible after discontinuation of the drug. However, a few cases of ESRD associated with the use of cidofovir in HIV-positive individuals have been reported (Vandercam et al., 1999; Meier et al., 2002). Topical or intralesional use of cidofovir may also be rarely associated with renal dysfunction (Bienvenu et al., 2002; Naiman et al., 2004). Propriothiouracil has been widely used to treat hyperthyroidism since the 1940s. Major side effects are agranulocytosis, hepatitis, vasculitis, and lupus-like syndrome, whereas kidney impairment is uncommon. Only a few cases of CTIN have been reported, with diffuse interstitial collagen accumulation (Nakahama et al., 1999).

699

700

Section 4  

the patient with interstitial disease

Conclusion Drug-induced CTIN is probably an underestimated cause of CKD. At the same time, it is often preventable and easily treatable, if diagnosed early. Physicians should be familiar with the wide range of medications potentially harmful to the kidneys, and be aware of the damage they may induce. The incidence of drug nephrotoxicity is only expected to rise, in parallel with the worldwide ageing population. This is due to frequent comorbidities, polypharmacy, and age-related structural renal changes. The diagnosis of drug-induced CTIN requires vigilance and knowledge of drug pharmacokinetics and pharmacodynamics; it is a multidisciplinary task, involving clinicians, pharmacists, and clinical chemists.

References Bendz, H., Aurell, M., Balldin, J., et al. (1994). Kidney damage in long-term lithium patients: a cross-sectional study of patients with 15 years or more on lithium. Nephrol Dial Transplant, 9, 1250–4. Bengtsson, U., Johansson, S., and Angervall, L. (1978). Malignancies of the urinary tract and their relation to analgesic abuse. Kidney Int, 13(1), 107–13. Benigni, A., Bruzzi, I., Mister, M., et al. (1999). Nature and mediators of renal lesions in kidney transplant patients given cyclosporine for more than one year. Kidney Int, 55, 674–85. Bienvenu, B., Martinez, F., Devergie, A., et al. (2002). Topical use of cidofovir induced acute renal failure. Transplantation, 73, 661–2. Bilyard, K. G., Joseph, E. C., and Metcalf, R. (1990). Mesalazine: an overview of key preclinical studies. Scand J Gastroenterol Suppl, 172, 52–5 Bolomey, A. A., Michie, A. J., Michie, C., et al. (1949). Simultaneous measurement of effective renal blood flow and cardiac output in resting normal subjects and patients with essential hypertension. J Clin Invest, 28(1), 10–17. Boton, R., Gaviria, M., and Batlle, D. C. (1987). Prevalence, pathogenesis, and treatment of renal dysfunction associated with chronic lithium therapy. Am J Kidney Dis, 10, 329–45. Bucht, G., Wahlin, A., Wentzel, T., et al. (1980). Renal function and morphology in long-term lithium and combined lithium-neuroleptic treatment. Acta Med Scand, 208, 381–5. Burdmann, E. A., Andoh, T. F., Nast, C. C., et al. (1995). Prevention of experimental cyclosporin-induced interstitial fibrosis by losartan and enalapril. Am J Physiol, 269, F491–9. Burdmann, E. A., Andoh, T. F., Yu, L., et al. (2003). Cyclosporine nephrotoxicity. Semin Nephrol, 23, 465–76. Christensen, B. M., Marples, D., Kim, Y. H., et al. (2004). Changes in cellular composition of kidney collecting duct cells in rats with lithium-induced NDI. Am J Physiol Cell Physiol, 286, C952. De Broe, M. E. and Elseviers, M. M. (1998). Analgesic nephropathy. N Engl J Med, 338, 446–52. De Broe, M. E. and Elseviers, M. M. (2009). Over-the-counter analgesic use. J Am Soc Nephrol, 20, 2098–103. De Mattos, A. M., Olyaei, A. J., and Bennett, W. M. (2000). Nephrotoxicity of immunosuppressive drugs: long-term consequences and challenges for the future. Am J Kidney Dis, 35, 333–46. Del Moral, R. G., Olmo, A., Osuna, A., et al. (1998). Role of P-glycoprotein in chronic cyclosporine nephrotoxicity and its relationship to intrarenal angiotensin II deposits. Transplant Proc, 30, 2014–16. Deray, G., Benhmida, M., Le Hoang, P., et al. (1992). Renal function and blood pressure in patients receiving long-term, low-dose cyclosporine therapy for idiopathic autoimmune uveitis. Ann Intern Med, 117, 578–83. Dubach, U. C., Rosner, B., Levy, P. S., et al. (1978). Epidemiological study in Switzerland. Kidney Int, 13(1), 41–9. Dubach, U. C., Rosner, B., Müller, A., et al. (1975). Relation between regular intake of phenacetin-containing analgesics and laboratory evidence

for urorenal disorders in a working female population of Switzerland. Lancet, 1(7906), 539–43. Efrati, S., Averbukh, M., Berman, S., et al. (2005). N-Acetylcysteine ameliorates lithium induced renal failure in rats. Nephrol Dial Transplant, 20, 65–70. Ekberg, H., Grinyó, J., Nashan, B., et al. (2007). Cyclosporine sparing with mycophenolate mofetil, daclizumab and corticosteroids in renal allograft recipients: the CAESAR Study. Am J Transplant, 7, 560–70. Ekberg, H., Tedesco-Silva, H., Demirbas, A., et al. (2007). Reduced exposure to calcineurin inhibitors in renal transplantation. N Engl J Med, 357, 2562–75. Elseviers, M. and De Broe, M. (1993). The implication of analgesic in human kidney disease. In J. Stewart (ed.) Analgesic and NSAID-Induced Kidney Disease, pp. 32–47. Oxford: Oxford University Press. Elseviers, M. M. and De Broe, M. E. (1995). A long-term prospective controlled study of analgesic abuse in Belgium. Kidney Int, 48, 1912–19. Elseviers, M. M., Waller, I., Nenoy, D., et al. (1995). Evaluation of diagnostic criteria for analgesic nephropathy in patients with end-stage renal failure: Results of the ANNE study. Analgesic Nephropathy Network of Europe. Nephrol Dial Transplant, 10, 808–14. Elseviers, M. M., D’Haens, G., Lerebours, E., et al. (2004). 5-ASA Study Group. Renal impairment in patients with inflammatory bowel disease: association with aminosalicylate therapy? Clin Nephrol, 61(2), 83–9. Farres, M. T., Ronco, P., Saadoun, D., et al. (2003). Chronic lithium nephropathy: MR imaging for diagnosis. Radiology, 229, 570–4. Gault, M. H. and Wilson, D. R. (1978). Analgesic nephropathy in Canada: clinical syndrome, management, and outcome. Kidney Int, 13, 58–63. Gilead Sciences, Inc. (1996). VISTIDE Prescribing Information. Foster City, CA: Gilead Sciences, Inc. Gonwa, T. A., Hamilton, R. W., Buckalew, V. M. Jr. (1981). Chronic renal failure and end-stage renal disease in northwest North Carolina. Importance of analgesic-associated nephropathy. Arch Intern Med, 141, 462–5. Griffin, M. D., Bergstralhn, E. J., and Larson, T. S. (1995). Renal papillary necrosis: a sixteen-year clinical experience. J Am Soc Nephrol, 6, 248–56. Hansen, H. E., Hestbech, J., Sorensen, J. L., et al. (1979). Chronic interstitial nephropathy in patients on long-term lithium treatment. QJM, 48, 577–91. Hauser, I. A., Schaeffeler, E., Gauer, S., et al. (2005). ABCB1 genotype of the donor but not of the recipient is a major risk factor for cyclosporine-related nephrotoxicity after renal transplantation. J Am Soc Nephrol, 16, 1501–11. Hestbech, J., Hansen, H. E., Amdisen, A., et al. (1977). Chronic renal lesions following long-term treatment with lithium. Kidney Int, 12(3), 205–13. Islam, M., Burke, J. F. Jr., McGowan, T. A., et al. (2001). Effect of anti-transforming growth factor beta antibodies in cyclosporine-induced renal dysfunction. Kidney Int, 59, 498–506. Kincaid-Smith, P. (1990). Analgesic nephropathy and the effect of non-steroidal anti-inflammatory drugs on the kidney. In G. Catto (ed.) Drugs and the Kidney, Chapter 1. Dordrecht: Kluwer Academic. Kling, M. A., Fox, J. G., Johnston, S. M., et al. (1984). Effects of long-term lithium administration on renal structure and function in rats. A distinctive tubular lesion. Lab Invest, 50(5), 526–35. Kopp, J. B. and Klotman, P. E. (1990). Cellular and molecular mechanisms of cyclosporin nephrotoxicity. J Am Soc Nephrol, 1, 162–79. Koziolek, M. J., Riess, R., Geiger, H., et al. (2001). Expression of multidrug resistance Pglycoprotein in kidney allografts from cyclosporine A-treated patients. Kidney Int, 60, 156–66. Mackinnon, B., Boulton-Jones, M., and McLaughlin, K. (2003). Analgesic-associated nephropathy in the West of Scotland: a 12-year observational study. Nephrol Dial Transplant, 18, 1800–5. Markowitz, G. S., Radhakrishnan, J., Kambham, N., et al. (2000). Lithium nephrotoxicity: a progressive combined glomerular and tubulointerstitial nephropathy. J Am Soc Nephrol, 11, 1439–48.

Chapter 87 

Meier, P., Dautheville-Guibal, S., Ronco, P. M., et al. (2002). Cidofovir-induced end-stage renal failure. Nephrol Dial Transplant, 17, 148–9. Mihatsch, M. and Zollinger, H. (1993). The pathology of analgesic nephropathy. In J Stewart (ed.) Analgesic and NSAID-Induced Kidney Disease, pp. 67–85. Oxford: Oxford University Press. Murray, T. G. and Goldberg, M. (1978). Analgesic-associated nephropathy in the U.S.A.: epidemiologic, clinical and pathogenetic features. Kidney Int, 13(1), 64–71. Naesens, M., Kuypers, D. R., and Sarwal, M. (2009). Calcineurin inhibitor nephrotoxicity. Clin J Am Soc Nephrol, 4, 481–508. Naiman, A. N., Roger, G., Gagnieu, M. C., et al. (2004). Cidofovir plasma assays after local injection in respiratory papillomatosis. Laryngoscope, 114, 1151–6. Nakahama, H., Nakamura, H., Kitada, O., et al. (1999). Chronic drug-induced tubulointerstitial nephritis with renal failure associated with propylthiouracil therapy. Nephrol Dial Transplant, 14(5), 1263–5. Nanra, R. S. (1980). Clinical and pathological aspects of analgesic nephropathy. Br J Clin Pharmacol, 10 Suppl 2, 359S–68S. Nanra, R. (1993). Functional defects in analgesic nephropathy. In J. Stewart (ed.) Analgesic and NSAID-Induced Kidney Disease, p. 102. Oxford: Oxford University Press. Oktem, F., Ozguner, F., Sulak, O., et al. (2005). Lithium-induced renal toxicity in rats: protection by a novel antioxidant caffeic acid phenethyl ester. Mol Cell Biochem, 277, 109–15. Palmer, B. F., Dawidson, I., Sagalowsky, A., et al. (1991). Improved outcome of cadaveric renal transplantation due to calcium channel blockers. Transplantation, 52, 640–5. Perneger, T. V., Whelton, P. K., and Klag, M. J. (1994). Risk of kidney failure associated with the use of acetaminophen, aspirin, and nonsteroidal antiinflammatory drugs. N Engl J Med, 331, 1675–9. Pichler, R. H., Franceschini, N., Young, B. A., et al. (1995). Pathogenesis of cyclosporine nephropathy: roles of angiotensin II and osteopontin. J Am Soc Nephrol, 6, 1186–96. Pintér, I., Mátyus, J., Czégány, Z., et al. (2004). Analgesic nephropathy in Hungary: the HANS study. Nephrol Dial Transplant, 19, 840–3.

drug-induced chronic tubulointerstitial nephritis Presne, C., Fakhouri, F., Noel, L. H., et al. (2003). Lithium-induced nephropathy: Rate of progression and prognostic factors. Kidney Int, 64, 585–92. Rabin, E. Z., Garston, R. G., Weir, R. V., et al. (1979). Persistent nephrogenic diabetes insipidus associated with long-term lithium carbonate treatment. Can Med Assoc J, 121(2), 194–8. Sandler, D. P., Smith, J. C., Weinberg, C. R., et al. (1989). Analgesic use and chronic renal disease. N Engl J Med, 320, 1238–43. Schreiber, S., Hämling, J., Zehnter, E., et al. (1997). Renal tubular dysfunction in patients with inflammatory bowel disease treated with aminosalicylate. Gut, 40(6), 761–6 Shihab, F. S., Bennett, W. M., Tanner, A. M., et al. (1997). Angiotensin II blockade decreases TGF-beta1 and matrix proteins in cyclosporine nephropathy. Kidney Int, 52, 660–73. Shihab, F. S., Waid, T. H., Conti, D. J., et al. Conversion from cyclosporine to tacrolimus in patients at risk for chronic renal allograft failure: 60-month results of the CRAF Study. Transplantation, 85, 1261–9. Shihab, F. S., Yi, H., Bennett, W. M., et al. (2000). Effect of nitric oxide modulation on TGFbeta1 and matrix proteins in chronic cyclosporine nephrotoxicity. Kidney Int, 58, 1174–85. Tam, V. K., Green, J., Schwieger, J., et al. (1996). Nephrotic syndrome and renal insufficiency associated with lithium therapy. Am J Kidney Dis, 27, 715–20. Vandercam, B., Moreau, M., Goffin, E., et al. (1999). Cidofovir-induced end-stage renal failure. Clin Infect Dis, 29, 948–9. Walker, R. G., Bennett, W. M., Davies, B. M., et al. (1982). Structural and functional effects of long-term lithium therapy. Kidney Int Suppl, 11, S13–19. Walker, R. G. (1993). Lithium nephrotoxicity. Kidney Int, 42(Suppl), S93. World, M. J., Stevens, P. E., Ashton, M. A., et al. (1996). Mesalazineassociated interstitial nephritis. Nephrol Dial Transplant, 11(4), 614–21. Wu, Y., Yang, L., Su, T., et al. (2010). Pathological significance of a panel of urinary biomarkers in patients with drug-induced tubulointerstitial nephritis. Clin J Am Soc Nephrol, 5(11), 1954–9.

701

CHAPTER 88

Heavy metal-induced tubulointerstitial nephritis Patrick C. D’Haese, Benjamin A. Vervaet, and Anja Verhulst Introduction Humans are exposed to various potentially toxic agents in their living and occupational environments. Some of these agents are metals, which may enter the human body through oral, inhalation, or transdermal routes, and may exert effects on all organ systems (Soderland et al., 2010). Several well-known, as well as lesser-known associations exist between exposure to heavy metals and chronic kidney disease (CKD). Extremely high environmental and occupational exposure to metals is rarely seen nowadays. Nevertheless, exposure to substantially lower levels is still regarded as an important cause of acute kidney injury (AKI) and CKD, particularly in the developing world. During the last years, there has been an increasing interest in the potential synergistic toxic effect of low exposure to multiple metals (Hambach et al., 2013). Due to its important blood flow, large endothelial surface, high metabolic activity through multiple enzyme systems, the high concentration of filtered chemicals in tubular fluid and in tubular cells, and the biotransformation of chemicals and protein unbinding, the kidney is highly vulnerable to the effects of toxic agents. Exposure to heavy metals may occur at the workplace, but industrial contamination of ground water, as well as inhalation of polluted air, and consumption of drugs or contaminated food are also recognized as important sources of heavy metals intoxication, which may induce CKD also in individuals without occupational exposure (Soderland et al., 2010). In this chapter, the renal effects of heavy metals, including lead, cadmium, arsenic, chromium, mercury, and uranium, are discussed in general, with a special focus on tubulointerstitial nephritis (TIN).

Lead The toxic effects of lead have been known for more than 2000 years, and the first reported case of nephrotoxicity associated with lead dates back to the nineteenth century. Environmental exposure may occur via lead paint and lead pipes, which may still be present in older houses, contamination of food during processing, indoor firing ranges, cigarette smoke, and contaminated air and soil near lead-processing industries. Occupational exposure occurs during battery manufacturing, welding, and use of lead solder. At present, exposure to extremely high concentrations of lead is less common than it once was, due to improved industrial management and the fact that this metal is no longer added to fuel and paint. Nevertheless, increased exposure to lead

is still a public health problem in some developing countries in Africa, Asia, and Latin America, due to contaminated water and soil resulting from poor industrial preventive measures (Sabath and Robles-Osorio, 2012). Lead nephropathy typically occurs when the blood lead concentration exceeds 400 micrograms/L. The disease presents with minimal proteinuria, a bland urinary sediment, hyperuricaemia, and often hypertension. The kidneys have a granular surface and reduced size. Renal biopsies show tubular atrophy and interstitial fibrosis, without cellular infiltration. In the proximal tubules, acid-fast nuclear inclusion bodies, consisting of a lead-binding protein complex, can be detected (Moore et al., 1973; Goyer, 1989; Evans and Elinder, 2011). Renal effects, however, may also be seen at much lower levels. A  study, in which 4813 individuals with or without high blood pressure were included, having mean blood lead levels of 42 micrograms/L and 33 micrograms/L, respectively, revealed that the prevalence of elevated serum creatinine was 11.5% and 1.8%, whilst it was seen in 10.0% and 1.1% of the subjects with CKD, respectively. These data made the authors conclude that low-level exposure to lead is associated with CKD, particularly in patients with hypertension (Muntner et  al., 2003). Results from follow-up studies carried out in Taiwan by Lin et al. indicated that individuals with chronic nephritis and a glomerular filtration rate (GFR) < 60 mL/minute and low-to-moderate exposure to lead (urinary lead excretion 80–600 micrograms/24 hours post ethylenediaminetetraacetic acid (EDTA) administration) experienced faster deterioration of renal function (Lin et  al., 2003). In an experimental study, Roncal et  al. showed that low-lead exposure accelerates CKD in 5/6th nephrectomized rats, primarily by raising blood pressure and accelerating microvascular and tubulointerstitial injury (Roncal et  al., 2007). The pathophysiological mechanism(s) underlying lead-induced CKD is not yet fully understood and various hypotheses have been put forward, including the induction of oxidative stress, generation of free radicals, and interference with calcium-dependent enzymatic reactions, which in turn may result in high blood pressure, inflammation, apoptosis, and, ultimately, development of chronic renal lesions (Sabath and Robles-Osorio, 2012) (Fig. 88.1). Diagnosing chronic lead nephropathy is difficult, given the unreliability of non-invasive tests. It relies mainly on a history of exposure to lead, in patients with significant and otherwise unexplained renal abnormalities. The EDTA mobilization test,

Chapter 88 

heavy metal-induced tubulointerstitial nephritis

EXPOSURE TO LEAD

Oxidative stress Guanylate cyclase

Production of nitric oxide

Interference with Ca-dependent enzymatic reactions

Free radicals Activation of NFkβ

Inflammation cGMP

Calcium

Apoptosis

Systemic vascular resistance High blood pressure

KIDNEY INJURY

Fig. 88.1  Pathophysiological mechanisms of lead-induced kidney injury. Adapted from Sabath and Robles-Osorio (2012).

which consists of measuring whole-blood or 24-hour urine lead levels over 1–4  days after parenteral administration of 1–3 g of calcium disodium ethylenediaminetetraacetic acid (CaNa2EDTA), may be helpful to assess the body lead burden (Wedeen, 2008). Alternatively, X-ray fluorescence may be used to detect increased bone lead concentrations, reflecting cumulative lead exposure. Although lead-induced AKI can sometimes be reversed by increasing the rate of lead excretion through chelation therapy, there is no evidence that such therapy reverses established TIN (Wedeen, 2008).

Cadmium Environmental sources of cadmium include combustion of fuels, industrial and household waste, tobacco smoke, sewage, contaminated sea food, vegetables and cereals, and (Indian) medicinal herbs (Hellström et al., 2001; Dey et al., 2009; Soderland et al., 2010). Cadmium is a by-product of mining and is used industrially in steel plating and manufacturing of plastics and nickel-cadmium batteries. Low serum metallothionein levels, iron deficiency, older age, female gender, and residence in the proximity of industrial cadmium sources have all been reported to hold an increased risk for cadmium toxicity (Berglund et al., 1994; Staessen et al., 1994). During the 1950s, Japanese doctors began to recognize an association between environmental exposure to cadmium and an increased incidence of renal tubular dysfunction, CKD, and a type of osteomalacia known as ‘itai-itai’ (Emmerson, 1970). Studies carried out later on in occupationally exposed workers also revealed an association between cadmium and an increased risk of developing kidney disease and osteomalacia (Adams et al., 1969). After the publication of several studies by Bernard and colleagues (Lauwerys et al., 1993), the scientific community became aware that even low-level cadmium exposure was associated with nephrotoxic effects and that up to 7% of the exposed population developed CKD. In many studies, the nephrotoxic effects of cadmium have been assessed by measurements of kidney biomarkers, such as β2-microglobulin (β2M), N-acetyl-β-D-glucosaminidase (NAG),

kidney injury molecule-1 (KIM-1), intestinal alkaline phosphatase (IAP), and retinol binding protein (RBP) (Nishijo et al., 2006; Prozialeck et al., 2007; Hambach et al., 2013). The results of these studies again suggested that cadmium toxicity can occur at much lower levels of exposure than those recognized by the World Health Organization. Moreover, it has recently been shown that nephrotoxicity from low-level cadmium exposure is aggravated by co-exposure to lead (Hambach et al., 2013). However, it is still a matter of debate whether increased urinary excretion of such biomarkers can predict later development of CKD; prospective studies over several years are required to clarify this issue. In this respect, it is worth mentioning that Nishijo et  al., in a 15-year follow-up study in inhabitants of the cadmium-polluted Kakehashi river basin area in Japan, showed an increased mortality risk in those with urinary β2M levels as high as 10,000 micrograms/g creatinine (Nishijo et  al., 2006). The histopathological renal examination revealed that the glomeruli were relatively well preserved in number and size, but the tubules were markedly damaged, with luminal obstruction. The mechanism responsible for the decrease in GFR due to cadmium nephrotoxicity is still uncertain; some authors suggested that cadmium exerts a direct effect on the glomeruli, whilst others postulated that cadmium-induced tubular damage can lead to TIN, which in turn results in an alteration of GFR (Takebayashi, 1980; Nordberg et al., 2008). The mechanisms of cadmium nephrotoxicity have been summarized in an elegant review by Sabath and Robles-Osorio (2012). Cadmium circulates in blood as cadmium-metallothionein-1 (MT-1) complex, which is filtered by the glomeruli and entirely reabsorbed in the S1 segment of the proximal tubular cells, by megalin- and cubulin-mediated endocytosis (Klassen et al., 2004). Within these cells, the cadmium-MT-1 complex is stored and broken down by lysosomes, after which free cadmium is released and transported to the cytoplasm by divalent metal transporter-1 (DMT-1) (Nordberg et al., 2008; Cucu et al., 2011) (Fig. 88.2). The activation of protein kinase C increases the expression of DMT-1, thereby increasing cadmium-induced tubular toxicity (Olivi et al., 2001). Free cadmium accumulates in mitochondria, resulting in mitochondrial dysfunction and formation of free radicals, which in turn activate caspase enzymes and apoptosis. Free cadmium

703

704

Section 4  

the patient with interstitial disease

may also bind to protein sulfhydryl groups and thus affect the structure and function of these proteins. The accumulation of free cadmium in tubular cells induces the expression of DMT-1 mRNA, haem oxygenase-1 (HO-1) mRNA, and pro-apoptotic genes—all involved in renal toxicity; however, no such effects are seen when cells are exposed to the cadmium-MT-1 complex (Cucu et al., 2011). Determination of cadmium concentration in blood reflects acute exposure, whilst urinary excretion of cadmium is used to assess the body cadmium burden and to evaluate chronic exposure. An isolated increase in urinary β2M is an important evidence of early proximal tubular dysfunction. Several studies have reported that the presence of diabetes, increased levels of tissue anti-MT-1 antibodies, and/or concomitant exposure to organic arsenic hold an increased risk for cadmium-induced renal dysfunction, even with cadmium exposure levels as low as those occurring in the general population in many countries. Although further studies are needed to confirm these findings, it has been postulated that for individuals with such associated risk factors there may not exist a threshold for cadmium renal toxicity (Hambach et al., 2013). There is no specific treatment for cadmium-induced renal disease, other than supportive care and change of residence area, to avoid further cadmium exposure.

The kidney is the main excretion route and target organ for mercury accumulation. With regard to its renal handling, mercury shares common pathways, to a certain extent, with lead and cadmium (Barbier et  al., 2005)  (Fig. 88.3). It is filtered by the glomeruli and reabsorbed by the proximal convoluted tubules. Acute mercury poisoning may result in acute tubular necrosis, particularly involving the proximal tubules. Increased exposure to mercury can cause tubular damage, characterized by low-molecular-weight proteinuria and possible further development of chronic TIN (Li et al., 2010; Soderland et al., 2010). In addition to tubulointerstitial lesions, a typical nephrotic syndrome may also occur. Some patients developed severe tubular damage, with excessive urinary losses of sodium, and a nephrotic syndrome, with only trivial morphological glomerular damage (i.e. minimal change disease), following prolonged treatment of psoriasis with mercurial diuretics or mercury-containing ointments. Nephrotic syndrome has also been reported after long-term exposure to mercury-containing paint additives, antirheumatoid medication, skin-lightening cream, hair-dyeing agents, and mercury vapour. Examination of renal biopsies in such cases revealed a typical membranous nephropathy, with minimal or no tubular injury (see Chapter 82) (Becker et al., 1962; Fowler et al., 2008; Li et al., 2010). Renal lesions may depend on the type of mercury compounds involved (organic versus inorganic), as well as on the valence of the metal. Biological monitoring of mercury concentrations is useful for assessing both the exposure level and the health risks (Elinder et al., 1994), but it may be complicated by the fact that both organic and inorganic mercury compounds are produced in the body and can be found in blood and urine. Urinary mercury is thought to correlate best with the amount of mercury within the kidneys (Zalups, 2000) and it is mainly associated with exposure to metallic mercury vapour or inorganic mercury compounds. In individuals who are not occupationally exposed, the urinary mercury concentrations seldom exceed 10 micrograms/L. There is now general consensus that if mercury-to-creatinine ratio in 24-hour urine is > 50 micrograms/g, nephrotoxicity is highly probable and comprises cytotoxic lesions of the proximal tubule (e.g. enzymuria and increase in tubular antigens) and functional changes (e.g. proteinuria, increase in serum β2M) (Roels et al., 1999). It is still a matter of debate whether combined measurements of mercury levels and biomarkers like leucine aminopetidase (LAP) or NAG may yield a better prediction of renal disease (Mason et al., 2001).

Mercury Mercury intoxication can result from consumption of water, fish, or cereals contaminated by ethyl mercury, used as a pesticide. Dental amalgam fillings have also been reported as sources of mercury. Additionally, accidental exposure to mercury has been described from breakage of mercury-containing thermometers and use of metallic mercury or mercury-containing ointments, creams, and drugs. Occupational exposure may occur in dental, chloralkali, and recycling industries, as well as in battery manufacturing (Fowler et al., 2008; Soderland et al., 2010). The first well-documented outbreak of acute methyl mercury (MeHg) poisoning occurred in Minamata, Japan, in 1953 and was due to the consumption of fish contaminated by waste drain from a chemical factory. The clinical picture of this poisoning, described in 1956 under the name of ‘Minamata disease’, was dominated by neurological symptoms, but low-molecular-weight proteinuria was also reported (Ekinoa et al., 2007).

liver cell

glomerular membrane

to bile

plasma

GSH Cd -GSH

?

Cd -Alb

Cd Cd -MT

?

so lys o me Cd -MT

Cd -Alb Cd

Cd -MT

Cd -MT MT

MT to urine

Fig. 88.2  Pathways of cadmium uptake and interaction with target sites in the kidney. Adapted from Nordberg et al. (2008).

renal tubular cell

tubular fluid

Cd -MT

damage to sensitive membrane site

Chapter 88 

heavy metal-induced tubulointerstitial nephritis

Cd Pb H

Cd Pb Hg

Cd Pb Hg

GSH-Cd/Pb/Hg Cys-Cd/Pb/Hg

Transporters: ZnT1 ZIP ABC type Proximal tubule

Paracellular

ɣ-GT

Cd

DMTI transporter SACs transporter

Distal tubule and terminal segments

Na -amino aci d transporter Na

MT-Cd/Pb/Hg GSH-Cd/Pb/Hg

Endocytosi s

Apical

Basolateral

Cd Pb H

DMTI transporter

Loop of Henle

Fig. 88.3  Mechanisms involved in the uptake of cadmium (Cd), lead (Pb), and mercury (Hg) along the nephron. Adapted from Barbier et al. (2005).

Treatments currently available for mercury poisoning involve the use of thiol-based chelating agents, such as British anti-Lewisite (BAL or dimercaprol), penicillamine, 2,3-dimercaptopropane-1-su lphonate (DMPS), and 2,3-dimercaptosuccinic acid (DMSA) (Sällsten et  al., 1994; Fowler et  al., 2008). Clinical studies have shown that chelation therapy successfully lowers the mercury body burden and increases urinary mercury excretion. Complete reversal of mercury-induced nephrotic syndrome has been reported in adults, as well as in infants, after dimercaprol treatment and withdrawal of the source of exposure (Wilson et al., 1952; Williams and Bridge, 1958).

Uranium Uranium is the heaviest of all naturally occurring elements. Its biological effects were described in literature as early as the 1820s (Soderland et al., 2010). Human beings are constantly exposed to a certain amount of uranium, because it is widely present in its natural form in food, air, soil, and water. The repercussions of this exposure on human physiology and pathophysiology are not yet fully understood (Vicente-Vicente et al., 2010). Natural exposure, overexposure, and intoxication can occur by ingestion, inhalation, or skin contact. Uranium accumulates mainly in the bones (66%), kidneys (8%), and liver (16%). The metal is excreted in the

urine. It is rapidly eliminated from the blood, whilst removal from organ depots occurs slowly (La Touche et al., 1987; International Commission for Radiation Protection, 1996). Animal studies, as well as studies in occupationally exposed workers, have shown that the major health hazard of uranium is chemical kidney toxicity, rather than radiation (Zamora et  al., 2009). In animal studies, renal effects have been reported after acute uranium intoxication, but it is not clear if these effects are able to trigger chronic renal lesions and if such lesions progress irreversibly and independently of the presence of the metal. The findings of Bijlsma et al., who studied urinary uranium excretion and kidney function in professional assistance workers several years after their acute exposure to uranium following the air disaster in Amsterdam, suggest that this would not be the case (Bijlsma et al., 2008). Chronic ingestion of uranium from drinking water has been associated with glucosuria, microalbuminuria, β2-microglobinuria, phosphaturia and hypercalciuria (Zamora et al., 1998). Studies in occupationally exposed populations have also reported aminoaciduria and low-molecular-weight proteinuria. A  non-significant trend towards higher serum creatinine levels has been observed in an epidemiological study in individuals residing in close proximity to a uranium processing plant in Ohio (Pinney et al., 2003). In a 20-year follow-up study of a Gulf War depleted uranium (DU) cohort, renal biomarkers showed minimal DU-related effects on proximal tubular

705

706

Section 4  

the patient with interstitial disease

function and cytotoxicity, but a significant increase in some urinary biomarkers was observed when urine concentrations of multiple metals, including uranium, were examined together (McDiarmid et al., 2013). Histopathological data in humans are scarce; however, studies in rats showed that chronic DU contamination, in addition to functional disturbances, can also induce structural renal damage, including interstitial fibrosis (Zhu et al., 2008). Possible mechanisms of uranium toxicity have been suggested by Vicente-Vicente et al. in an excellent review (2010) (Fig. 88.4). To date, the best method of diagnosing uranium exposure is the detection of the metal in urine. There are no specific biomarkers for uranium-related nephrotoxicity, but general markers of kidney injury may be useful in cases suspected to have been exposed to this metal (Vicente-Vicente et al., 2010). Cessation of exposure is the first line of therapy. Once uranium has been taken up by target organs, therapy with chelating agents, like EDTA or diethylenetriaminepentaacetic acid (DTPA), should be initiated. Although these agents can remove the metal from critical tissues and prevent it from binding to target cells, there is no evidence that this can help the recovery of renal damage.

Arsenic Arsenic is one of the most widespread environmental pollutants and millions of people (mostly in Asia and Latin America) suffer from exposure to the element, since it is a common contaminant of drinking water. Other sources of arsenic are seafood, pesticides (causing food contamination), and products for wood preservation. A  less common source is medication, such as arsenic trioxide, used in the treatment of acute promyelocytic leukaemia, and certain drugs used for sleeping sickness and leishmaniasis (Mahmudur-Rahman et al., 2009; Sabath and Robles-Osorio, 2012). Although renal involvement in arsenic poisoning has been reported, it remains an underdiagnosed cause of kidney disease. High concentrations of arsenic in the drinking water have been associated with an increased mortality from CKD (Meliker et al.,

2007; Smith et al., 2012). Few reports in the literature deal with the effects of arsenic on renal function in the general population. Hsueh et al. studied 125 individuals with a GFR < 60 mL/ minute and 229 subjects with normal renal function and found a weak but significant association between urinary arsenic levels and decreased renal function (r2  =  0.04, P < 0.001) (Hsueh et  al., 2009). Arsenic levels in serum and blood cells correlate with the worsening of kidney disease and with the development and progression of CKD. These effects have been attributed to arsenic-induced oxidative stress (Zhang et al., 1995; Sasaki et al., 2007). Huang et  al. evaluated urinary biomarkers, the possible role of oxidative stress, and the effect of co-exposure to environmental low-levels of arsenic and cadmium in 290 adults from the general population. They found NAG, malondialdehyde (MDA), and 8-hydroxydeoxyguanosine (8-OHdG) to positively correlate with both arsenic and cadmium exposure. Interestingly, the effects of concomitant exposure to both metals on these biomarkers were more pronounced than those of exposure to only one of them. Based on these findings, it was suggested that chronic exposure to low levels of arsenic and/or cadmium may produce tubular damage in the human kidney through oxidative stress (Huang et al., 2009). Little is known about the histopathology of arsenic-induced renal injury in humans. A  case of TIN associated with elevated urinary arsenic concentration was reported by Prasad and Rossi (1995). In a patient with no history of diabetes, hypertension, heart disease or hepatitis and no family history of CKD, increased urinary arsenic levels (91.0 micrograms/L) were found, together with normal cadmium and mercury concentrations. Light microscopy of the renal biopsy showed normal glomeruli, extensive interstitial fibrosis, with tubular atrophy and a focus of cellular infiltrate, mainly consisting of lymphocytes. These findings were interpreted as chronic TIN. Following changes in the patient’s diet, which was based on ‘organically grown health food’, and cessation of intake of over-the-counter vitamins, the arsenic content in the urine dropped to 6.5 micrograms/L, concomitantly with a decrease in serum creatinine from 2.0 mg/dL to 1.7 mg/dL.

EXPOSURE TO URANIUM Alterations in transport mechanisms Na/K ATPase

Na ATPase

Na Pi II

SGLT

Pi

Glucose

Na Tubular reuptake dysfunction

Oxidative stress Inhibition of mitochondrial oxidative phosphorylation

Lipid peroxidation

Caspases

Suppression of cell respiration CELL DEATH

Fig. 88.4  Possible mechanisms involved in uranium nephrotoxicity. Adapted from Vicente-Vicente et al. (2010).

ROS

Chapter 88 

Data on the pathophysiological mechanisms of arsenic nephrotoxicity are scarce. The multidrug resistance-associated protein 2 (MRP-2) transporter, together with aquaporins, favours the entry of arsenic in proximal tubular cells. Once taken up by these cells, arsenic seems to induce its toxic effects through glutathione depletion, which in turn increases oxidative stress by induction of free radicals. This hypothesis is supported by the reduction of arsenic toxicity by administration of selenium, a well-known antioxidant agent (Messarah et al., 2012).

Chromium Whilst the role of chromium has been intensively studied in oncology (Seidler et al., 2013), its potential renal toxicity has been largely overlooked. Exposure to chromium can occur mainly by intake of contaminated food, inhalation of polluted air, or skin contact during chromium handling at the workplace. Other sources of chromium exposure may consist in drinking contaminated well water and residing in the proximity of uncontrolled hazardous waste sites or industrial plants that use or process chromium. Nephrotoxic effects of chromium have been demonstrated in animal studies (Hojo and Satomi, 1991). Chromates and chromic acid used in the treatment of certain skin diseases have been reported to cause fatal cases of acute nephritis. Necropsies of such cases revealed acute tubular necrosis, without glomerular lesions (Petersen et  al., 1994). Other studies reported renal function impairment in subjects with a high urinary chromium concentration (Hsueh et al., 2009). Petersen et al. (1994) described the case of a 48-year-old man, who developed chronic TIN after long-term occupational exposure to chromium. His renal biopsy showed totally or partially sclerotic glomeruli, focal interstitial fibrosis, with scattered lymphocytes, and tubular atrophy, but no tubular necrosis (Petersen et al., 1994).

References Adams, R. G., Harrison, J. F., and Scott, P. (1969). The development of cadmium-induced proteinuria, impaired renal function, and osteomalacia in alkaline battery workers. QJM, 38, 425–43. Barbier, O., Jacquillet, G., Tauc, M., et al. (2005). Effect of heavy metals on, and handling by, the kidney. Nephron Physiol, 99, 105–10. Becker, C. G., Becker, E. L., Maher, J. F., et al. (1962). Nephrotic syndrome after contact with mercury. Arch Intern Med, 83, 178–86. Berglund, M., Akesson, A., Nermell, B., et al. (1994). Intestinal absorption of dietary cadmium in women depends on body iron stores and fiber intake. Environ Health Perspect, 102(12), 1058–66. Bijlsma, J. A., Slottje, P., Huizink, A. C., et al. (2008). Urinary uranium and kidney function parameters in professional assistance workers in the Epidemiological Study Air Disaster in Amsterdam (ESADA). Nephrol Dial Transplant, 23, 249–55. Cucu, D., D’Haese, P. C., De Beuf, A., et al. (2011). Low doses of cadmium chloride and methallothionein-1-bound cadmium display different accumulation kinetics and induce different genes in cells of the human nephron. Nephron Extra, 1, 24–37. Dey, S., Saxena, A., Dan, A., et al. (2009). Indian medicinal herb: a source of lead and cadmium for humans and animals. Arch Environ Occup Health, 64(3), 164–7. Ekinoa, S., Susab, M., Ninomiyaa, T., et al. (2007). Minamata disease revisited: an update on the acute and chronic manifestations of methyl mercury poisoning. J Neurol Sci, 262, 131–144. Elinder, C. -G., Friberg, L., Nordberg, G. F., et al. (1994). Biological monitoring of metals. Chemical safety monographs. International Programme on Chemical Safety. WHO/EHG, 94(2), 1–86.

heavy metal-induced tubulointerstitial nephritis Emmerson, B. T. (1970). ‘Ouch-Ouch’ disease: the osteomalacia of cadmium nephropathy. Ann Intern Med, 73, 854–5. Evans, M. and Elinder, G.-F. (2011). Chronic renal failure from lead: myth or evidence-based fact? Kidney Int, 79, 272–9. Fowler, B. A., Whittaker, M. H., and Elinder, C. -G. (2008). Mercury-induced renal effects. In M. E. De Broe, G. A. Porter, W. M. Bennett, et al. (eds.) Clinical Nephrotoxins—Renal Injury from Drugs and Chemicals, pp. 811–26. New York: Springer Science + Business Media, LLC. Goyer, R. (1989). Mechanisms of cadmium and lead nephropathy. Toxicol Lett, 46, 153–62. Hambach, R., Lison, D., D’Haese, P. C., et al. (2013). Co-exposure to lead increases the renal response to low levels of cadmium in metallurgy workers. Toxicol Lett, 222(2), 233–8. Hambach, R., Lison, D., D’Haese, P., et al. (2013). Adverse effects of low occupational cadmium exposure on renal and oxidative stress biomarkers in solderers. Occup Environ Med, 70(2), 108–13. Hellström, L., Elinder, C. G., Dahlberg, B., et al. (2001). Cadmium exposure and end-stage renal disease. Am J Kidney Dis, 38(5), 1001–8. Hojo, Y. and Satomi, Y. (1991). In vivo nephrotoxicity induced in mice by chromium(VI). Involvement of glutathione and chromium(V). Biol Trace Elem Res, 31(1), 21–3. Hsueh, Y. M., Chung, C. J., Shiue, H. S., et al. (2009). Urinary arsenic species and CKD in a Taiwanese population: a case-control study. Am J Kidney Dis, 54, 859–70. Huang, M., Choia, S. -J., Kim, D.W., et al. (2009). Risk assessment of low-level cadmium and arsenic on the kidney. J Toxicol Environ Health A, 72, 1493–98. International Commission for Radiation Protection (ICRP) (1996). Age-Dependent Doses to Members of the Public from Intake of Radionuclides: Part 4, Inhalation Dose Coefficients. Oxford: Pergamon Press. Klassen, R. B., Crenshaw, K., Kozyraki, R., et al. (2004). Megalin mediates renal uptake of heavy metal methallothionein complexes. Am J Physiol Renal Physiol, 287, F393–403. La Touche, Y. D., Willis, D. L., and Dawydiak, O. I. (1987). Absorption and biokinetics of U in rats following an oral administration of uranyl nitrate solution. Health Phys, 53, 147–62. Lauwerys, R. R., Bernard, A. M., Buchet, J. P., et al. (1993). Assessment of the health impact of environmental exposure to cadmium: contribution of the epidemiologic studies carried out in Belgium. Environ Res, 62(2), 200–6. Li, S. -J., Zhang, S. -H., Chen, H. -P., et al. (2010). Mercury-induced membranous nephropathy: clinical and pathological feature. Clin J Am Soc Nephrol, 5(3), 439–44. Lin, J. L., Lin-Tan, D. T., Hsu, K. H., et al. (2003). Environmental lead exposure and progression of chronic renal diseases in patients without diabetes. N Engl J Med, 348(4), 277–86. Mahmudur-Rahman, R., Ng, J. C., and Naidu, R. (2009). Chronic exposure of arsenic via drinking water and its adverse health impacts on humans. Environ Geochem Health, 31, 189–200. Mason, H. J., Hindell, P., and Williams, N. R. (2001). Biological monitoring and exposure to mercury. Occup Med, 51(1), 2–11. McDiarmid, M. A., Gaitens, J. M., Hines, S., et al. (2013). The Gulf War depleted uranium cohort at 20 years: bioassay results and novel approaches to fragment surveillance. Health Phys, 104(4), 347–61. Meliker, J. R., Wahl, R. L., Cameron, L. L., et al. (2007). Arsenic in drinking water and cerebrovascular disease, diabetes mellitus, and kidney disease in Michigan: a standardized mortality ratio analysis. Environ Health, 6, 4. Messarah, M., Klibet, F., Boumendjel, A., et al. (2012). Hepatoprotective role and antioxidant capacity of selenium on arsenic-induced liver injury in rats. Exp Toxicol Pathol, 64(3), 167–74. Moore, J. Goyer, R., and Wilson, M. (1973). Lead induced inclusion bodies. Solubility, amino acid content, and relationship to residual acidic nuclear proteins. Lab Invest, 29, 488–94.

707

708

Section 4  

the patient with interstitial disease

Muntner, P., He, J., Vupputuri, S., et al. (2003). Blood lead and chronic kidney disease in the general United States population: Results from NHANES III. Kidney Int, 63, 1044–50. Nishijo, M., Morikawa, Y., Nakagawa, H., et al. (2006). Causes of death and renal tubular dysfunction in residents exposed to cadmium in the environment. Occup Environ Med, 63(8), 545–50. Nordberg, G. F., Kido, T., Roels, H. A., et al. (2008). Cadmium-induced renal effects. In M. E. De Broe, G. A. Porter, W. M. Bennett, et al. (eds.) Clinical Nephrotoxins—Renal Injury from Drugs and Chemicals, pp. 785–810. New York: Springer Science + Business Media, LLC. Olivi, L., Sisk, J., and Bressler, J. (2001). Involvement of DMT1 in uptake of Cd in MDCK cells: role of protein kinase C. Am J Physiol Cell Physiol, 281, C793–800. Petersen, R., Mikkelsen, S., and Thomsen, O. F. (1994). Chronic interstitial nephropathy after plasma cutting in stainless steel. Occup Environ Med, 51, 259–61. Pinney, S. M., Freyberg, R. W., Levine, G. E., et al. (2003). Health effects in community residents near a uranium plant at Fernald, Ohio, USA. Int J Occup Med Environ Health, 16(2), 139–53. Prasad, G. V. R. and Rossi, N. F. (1995). Arsenic intoxication associated with tubulointerstitial nephritis. Am J Kidney Dis, 26(2), 373–6. Prozialeck, W. C., Vaidya, V. S., Liu, J., et al. (2007). Kidney injury molecule-1 is an early biomarker of cadmium nephrotoxicity. Kidney Int, 72, 985–93. Roels, H. A., Hoet, P., and Lison, D. (1999). Usefulness of biomarkers of exposure to inorganic mercury, lead, or cadmium in controlling occupational and environmental risks of nephrotoxicity. Ren Fail, 21(3–4), 251–62. Roncal, C., Mu, W., Reungjui, S., et al. (2007). Lead, at low levels, accelerates arteriolopathy and tubulointerstitial injury in chronic kidney disease. Am J Physiol Renal Physiol, 293(4), F1391–6. Sabath, E. and Robles-Osorio, M. L. (2012). Renal health and the environment: heavy metal nephrotoxicity. Nefrologia, 32(3), 279–86. Sällsten, G., Barregård, L., and Schütz, A. (1994). Clearance half life of mercury in urine after the cessation of long term occupational exposure: influence of a chelating agent (DMPS) on excretion of mercury in urine. Occup Environ Med, 51(5), 337–42. Sasaki, A., Oshima, Y., and Fujimura, A. (2007). An approach to elucidate potential mechanism of renal toxicity of arsenic trioxide. Exp Hematol, 35(2), 252–62. Seidler, A., Jähnichen, S., Hegewald, J., et al. (2013). Systematic review and quantification of respiratory cancer risk for occupational

exposure to hexavalent chromium. Int Arch Occup Environ Health, 86(8), 943– 55. Smith, A. H., Marshall, G., Liaw, J., et al. (2012). Mortality in young adults following in utero and childhood exposure to arsenic in drinking water. Environ Health Perspect, 120(11), 1527–31. Soderland, P., Lovekar, S., Weiner, D. E., et al. (2010). Chronic kidney disease associated with environmental toxins and exposures. Adv Chronic Kidney Dis, 17(3), 254–64. Staessen, J. A., Lauwerys, R. R., Ide, G., et al. (1994). Renal function and historical environmental cadmium pollution from zinc smelters. Lancet, 343, 1523–7. Takebayashi, S. (1980) First autopsy case, suspicious of cadmium intoxication, from the cadmium-polluted area in Tsushima, Nagasaki Prefecture. In I. Shigematsu and K. Nomiyama (eds.) Cadmium-Induced Osteopathy, pp. 124–38. Tokyo: Japan Public Health Association. Vicente-Vicente, L., Quiros, Y., Pérez-Barriocanal, F., et al. (2010). Nephrotoxicity of uranium: pathophysiological, diagnostic and therapeutic perspectives. Toxicol Sci, 118(2), 324–47. Wedeen, R. P. (2008). Lead nephropathy. In M. E. De Broe, G. A. Porter, W. M. Bennett, et al. (eds.) Clinical Nephrotoxins—Renal Injury from Drugs and Chemicals, pp. 773–83. New York: Springer Science + Business Media, LLC. Williams, N. E. and Bridge, H. G. (1958). Nephrotic syndrome after the application of mercury ointment. Lancet, 20(2), 602. Wilson, V. K., Thomson, M. L., and Holzel, A. (1952). Mercury nephrosis in young children, with special reference to teething powders containing mercury. Br Med J, 16(1), 358–60. Zalups, R. K. (2000). Molecular interactions with mercury in the kidney. Pharmacol Rev, 52(1), 113–43. Zamora, M. L., Tracy, B. L., Zielinski, J. M., et al. (1998). Chronic ingestion of uranium in drinking water: a study of kidney bioeffects in humans. Toxicol Sci, 43(1), 68–77. Zamora, M. L., Zielinski, J. M., Moodie, G. B., et al. (2009). Uranium in drinking water: renal effects of long-term ingestion by an aboriginal community. Arch Environ Occup Health, 64(4), 228–41. Zhang, X., Cornelis, R., De Kimpe, J., et al. (1995). Determination of total arsenic in serum and packed cellsof patients with renal insufficiency. Anal Bioanal Chem, 353(2), 143–7. Zhu, G., Xiang, X., Chen, X., et al. (2008). Renal dysfunction induced by long-term exposure to depleted uranium in rats. Arch Toxicol, 83(1), 37–46.

CHAPTER 89

Aristolochic acid nephropathy caused by ingestion of herbal medicinal products M. Refik Gökmen and Graham M. Lord Introduction and epidemiology The association between aristolochic acid (AA) ingestion and progressive renal interstitial fibrosis first came to light in the context of an outbreak of severe renal failure in more than 100 patients in Belgium starting in the early 1990s. The initial report described nine women who presented either requiring dialysis or with rapidly progressive renal impairment, all of whom had taken a slimming regimen prescribed by the same clinic (Vanherweghem et  al., 1993). This regimen had been modified in June 1990 to include extracts of two Chinese herbs, labelled as Stephania tetrandra and Magnolia officinalis. Although these products were banned by the Belgian authorities in 1992, it soon emerged that these plant species were not in fact the true culprits. Subsequent investigations showed that S.  tetrandra in this regimen had in fact been replaced with Aristolochia fangchi, and phytochemical analysis further showed the presence of AA, rather than the expected tetrandrine (Vanhaelen et  al., 1994). More recently, the detection of AA-DNA adducts in renal tissue from patients with ‘Chinese herbal nephropathy’ has provided more definitive confirmation of AA involvement in affected patients (Schmeiser et al., 1996; Nortier et al., 2000). Together with data demonstrating the characteristic pattern of nephrotoxicity caused by AA administration in experimental animals, these observations have led to the replacement of the term ‘Chinese herbal nephropathy’ with the more accurate ‘aristolochic acid nephropathy’ (Debelle et al., 2008). Since the first description of the Belgian cohort, other cases and case series have been reported in a number of European countries, as well as the United States, Japan, Korea, China, Taiwan, and Hong Kong (Debelle et  al., 2008; Gökmen et  al., 2013). These reports attest that AA-containing remedies have been (and continue to be) used for a variety of indications, including eczema, acne, liver disease, arthritis, and chronic pain. However, the number of people affected by aristolochic acid nephropathy (AAN) worldwide remains unclear. Investigators from China have reported that thousands of cases have been identified among patients previously labelled as having chronic tubulointerstitial nephritis (CTIN) of unknown origin, and describe 300 cases identified between 1997 and 2006 in one centre in Beijing (Yang et al., 2011). In addition, Aristolochia species are known to be used for a wide variety of

indications in many regions of the world where AAN has not yet been described, including Africa, South America, and the Indian subcontinent (Vanherweghem, 1997; Heinrich et al., 2009).

Risk factors It is estimated that 1500–2000 people were exposed to AA as part of the slimming regimen in the Belgian outbreak; of these approximately 100 are known to have developed renal disease (Vanherweghem, 1998). The only risk factor for the development and progression of renal disease which has so far been defined with any certainty is the cumulative dose of AA. In the Belgian cohort, where patients received a number of pharmaceutical products including fenfluramine and acetazolamide alongside herbal products, the ingested dose of A.  fangchi emerged in multiple regression analysis as the only significant drug predicting the rate of progression of CKD (Martinez et al., 2002). Although theoretical reasons why co-administration of fenfluramine or acetazolamide might potentiate AA nephrotoxicity have been suggested (Martinez et  al., 2002), the worldwide prevalence of AAN outside the Belgian outbreak suggests that AA alone is sufficient to cause the disease. In a Chinese cohort, among 280 patients with chronic AAN, the median cumulative intake of aristolochic acid I (AAI), the main component of the plant extract AA was found to be 1.01 g (Yang et al., 2011), although reported exposures as small as 0.025 g were seen in those affected. A retrospective review of 199,843 persons in the Taiwanese National Health Insurance reimbursement database observed that reported ingestion of > 60 g of Fangchi or of > 30 g of Mu-Tong was associated with an increased risk of CKD (Lai et al., 2009). However, in light of the variations in the concentration of AA in different herbal preparations, there is unlikely to be a ‘safe dose’ of these products. All the published AAN case series, including those from China and Taiwan, note a marked female preponderance among affected individuals. The available epidemiological data do not allow assessment of whether female gender represents a genuine risk factor, or whether this reflects the known increased use of Chinese herbal products by women. Finally, genetic studies in Balkan endemic nephropathy (BEN) may provide new insights into the genetic factors conferring risk to those exposed to AA (Stefanovic et al., 2006).

710

Section 4  

the patient with interstitial disease

Association with Balkan endemic nephropathy BEN is a CTIN found in farming villages close to tributaries of the Danube river in Bosnia, Bulgaria, Croatia, Romania, and Serbia (see Chapter 90). First described in the 1950s (Tanchev et al., 1956; Danilovic, 1958), BEN is notable for showing a familial, but not inherited, pattern of distribution, and a strong association with urothelial malignancy (Stefanovic et al., 2006). In the early stages of the disease, BEN is characterized by tubular dysfunction, whereas marked tubular atrophy and hypocellular interstitial fibrosis is seen in more advanced disease. The role of environmental exposure to AA in the aetiology of BEN is now well documented. The striking pathological similarity between the first described cases of AAN and BEN was noted in the original report of the Belgian outbreak. Indeed, environmental exposure to AA had first been suggested as a cause of BEN in 1969, when Ivić found contamination of wheat flour by the seeds of Aristolochia clematitis, a weed that is common in wheat fields in endemic areas (Ivić, 1969; Hranjec et  al., 2005). More recently, definitive proof of AA exposure in patients with BEN has come from the use of the 32P-postlabelling technique to detect AA-specific DNA adducts in renal cortical and urothelial malignant tissue (Arlt et al., 2002; Grollman et al., 2007; Arlt et al., 2007, Jelaković et al., 2012). Furthermore, 78% of the examined tumours were found to have A:T→T:A transversion mutations in the TP53 gene, which is a hallmark mutation in AA-induced carcinogenesis and rarely seen in urothelial malignancies unrelated to AA (Arlt et al., 2001). With BEN pathologically indistinguishable from AAN caused by ingestion of Chinese herbal products, definitive evidence of AA exposure in affected individuals, and evidence of AA as an environmental toxin in endemic areas, it is highly likely that BEN represents a form of AAN.

Clinical features and investigations Clinical features and basic investigations Most patients with AAN present with renal insufficiency, anaemia, a urine sediment with a few red and white blood cells, and mild proteinuria (typically, < 1.5 g/24 hours). In keeping with the known toxic effects of AA on proximal tubular cells, proteinuria tends to consist of low-molecular-weight proteins that are usually filtered at the glomerulus, but reabsorbed by proximal tubular cells (Kabanda et al., 1995). Small numbers of patients presenting with acute kidney injury or with a Fanconi syndrome of tubular dysfunction have also been reported (Tanaka et al., 2000; Yang et al., 2000; Krumme et al., 2001). Both in Chinese herb-associated AAN and BEN, the degree of anaemia seen has been noted to be more severe than expected with the decrease in glomerular filtration rate (GFR), perhaps as a result of early destruction of erythropoietin-producing peritubular cells (Reginster et  al., 1997). Renal ultrasonography reveals shrunken kidneys, which can be asymmetrical and irregular in cortical outline. No serum or urinary biomarkers have so far been demonstrated to have clinical utility in the diagnosis of AAN. Recent studies have reported that a number of urinary proteins, including beta-2-microglobulin and alpha-1-microglobulin, are elevated in patients with AAN, compared with either those with glomerulonephritis and hypertensive nephrosclerosis (Stefanovic et al., 2011), or those with diabetic nephropathy, acute kidney injury and healthy controls (Pešić et al., 2011). Although this is promising, it is not yet

clear whether these findings simply represent non-specific tubular damage. Further studies are required to determine and evaluate novel non-invasive biomarkers for AAN.

Renal histology Given the non-specific clinical features, pathological examination of renal biopsy tissue is usually essential for a diagnosis of AAN. The characteristic findings are extensive interstitial fibrosis and tubular atrophy, typically with more marked fibrosis of the outer renal cortex (Fig. 89.1). Infiltration of the interstitium by inflammatory cells is seen in some patients (Pozdzik et al., 2010), although the degree of inflammation is typically less than in other fibrotic interstitial diseases. While the glomeruli are relatively preserved, global glomerular obsolescence and ischaemic changes are common in more advanced disease. Vascular involvement typically consists of fibrous hyperplasia of arteriolar walls. Urothelial atypia is observed almost universally (Fig. 89.2); in 40–46% of patients there is also multifocal transitional cell carcinoma in situ (Nortier et al., 2003).

DNA adduct analysis Exposure to AA leads to the formation of covalent AA-DNA adducts, which persist as a specific long-term biomarker of AA exposure. Where possible, identification of AA-DNA adducts using the 32P-postlabelling technique can form an important part of establishing the diagnosis of AAN (Lord et al., 2001; Arlt et al., 2004). In most cases, this requires the extraction of DNA from a fresh (unfixed) biopsy core of renal tissue; alternatively a presumptive diagnosis of AAN can be confirmed following the identification of adducts in nephrectomy specimens.

Natural history The majority of patients diagnosed with AAN show a relentless course towards end-stage renal disease (ESRD). In a follow-up study of the original Belgian cohort, the 2-year actuarial survival rate without ESRD was only 17%, compared with 74% in a control group with other CTIN (Reginster et al., 1997). The experience in other centres has been similar, with the median rate of change in eGFR being −3.5 mL/min/year in the largest Chinese case series (Yang et al., 2000). Those with a relatively preserved GFR at presentation (< 2 mg/dL or 176 μmol/L) appear to have a reduced risk of progression to stage 5 CKD (Reginster et al., 1997), although, as

Fig. 89.1  Low-power view of renal cortex from a patient with AAN, showing extensive hypocellular interstitial fibrosis associated with marked tubular atrophy.

Chapter 89 

Fig. 89.2  High-power view of ureteric epithelium showing cellular atypia and dysplasia. Pathological images provided by Professor T. H. Cook, Imperial College London.

noted above, the single best predictor of the degree of renal insufficiency and the rate of decline of residual renal function has been found to be the cumulative dose of AA ingested.

Association with urothelial malignancy A key feature of the natural history of AAN, irrespective of the mode of exposure, is the risk of urothelial malignancy. Although almost all documented cases of urothelial malignancy have been in AAN patients requiring renal replacement therapy, extensive urothelial cell atypia is a common finding in biopsy specimens from patients with less severe renal impairment (Nortier et  al., 2000). A number of observational studies have helped in quantifying the risk of malignancy in individuals exposed to AA, with a reported prevalence of high-grade carcinoma in situ or invasive lesions in 40–46% of AAN patients who have kidneys and ureters removed prophylactically. The majority of tumours arise in the upper urinary tract, although an increased incidence of late-onset bladder tumours has also been reported (Lemy et al., 2008).

Aetiology and pathogenesis Experimental studies have shown that the main targets of AA nephrotoxicity are the tubular and interstitial compartments, with an early phase of acute tubular necrosis preceding the development of tubular atrophy and interstitial fibrosis (Okada et al., 2003; Sato et al., 2004; Lebeau et al., 2005). Tubular cells undergo apoptosis after exposure to AA (Gao et al., 2000; Hsin et al., 2006), while both epithelial-to-mesenchymal transition of tubular cells and activation of resident tissue fibroblasts are likely to contribute to the extensive interstitial fibrosis seen in AAN (Pozdzik et al., 2008). In common with other fibrotic processes, transforming growth factor (TGF)-β and its downstream signalling pathways have been implicated in AA-induced fibrosis (Pozdzik et al., 2008; Zhou et al., 2010). The mechanisms of AA-induced carcinogenesis are well characterized, although it is unclear whether the nephrotoxic and carcinogenic effects of AA are related; there has been at least one report of AA-associated urothelial malignancy in the absence of severe renal impairment (Nortier et al., 2003). AA is genotoxic and mutagenic after metabolic activation (Schmeiser et al., 2009). A number of enzymes are involved in the activation of AA, with NAD(P) H:quinone oxidoreductase (NQO1) being the most important (Stiborova et al., 2008). The key mutagenic moiety resulting from

aristolochic acid nephropathy

this process is an electrophilic cyclic N-acylnitrenium ion that reacts preferentially with purine bases in DNA, forming covalent DNA adducts. These adducts can be detected by 32P-postlabelling analysis and serve as a long-lasting biomarker of AA exposure (Fernando et al., 1993; Nortier et al., 2000; Jelaković et al., 2012). The majority of oncogenic mutations associated with AA exposure are A:T→T:A transversion mutations arising from the incorporation of an adenine residue opposite the adduct during DNA replication (Broschard et  al., 1994; Attaluri et  al., 2010). A:T→T:A transversions mutations are commonly identified in the oncogene TP53 in areas of urothelial atypia and in urothelial malignancies (Lord et  al., 2004)  both in patients with herbal product-related AAN and in BEN (Grollman et al., 2007). The total cumulative dose of AA has emerged as the key risk factor for the development of renal disease and malignancy following AA exposure. Few other risk factors have been identified; the female preponderance of reported cases worldwide is deemed likely to be the result of increased uptake of AA-containing medicinal products among females. A number of studies have also reported polymorphisms in genes encoding AA-activating enzymes such as NQO1 and cytochrome P450 enzymes as being associated with increased cancer risk in AAN and BEN patients, although some of these associations have not been confirmed in subsequent reports(Stefanovic et al., 2006; Atanasova et al., 2005; Toncheva et al., 2004).

Treatment and outcome Prevention Given the severe consequences of AA ingestion, prevention of exposure to AA is a key public health priority. In the European Union, the 2004 Traditional Herbal Products Directive has, since 1 May 2011, required that all traditional herbal medicines must be registered and approved; no products containing AA have been approved (European Union, 2004). In the United States, the Food and Drug Administration issued an alert in 2001 warning consumers and the herbal medicine industry of the dangers of AA (Food and Drug Administration, 2001), and import alerts are in force allowing the seizure of any product containing AA. In the Far East, the medicinal use of most AA-containing plant species has been banned in Hong Kong, Taiwan, and mainland China, although certain AA-containing products are still permitted in China under the supervision of Chinese medicine practitioners (World Health Organization, 2004; Poon, 2007). Despite these regulatory measures, there is still cause for concern. Products containing AA are still available over the Internet (Gold and Slone, 2003), and some practitioners continue to dispute the evidence that AA causes serious harm. Tighter local regulation of practitioners and outlets of alternative and herbal medicine is required, as well as a robust international system of surveillance to identify products containing AA (Gökmen and Lord, 2012).

Disease-specific management AAN is notable for the rapid progression to ESRD, despite cessation of AA-containing products. Although there are no randomized trials in humans, there is currently no evidence that renin–angiotensin system blockade with ACE-inhibitors or angiotensin receptor blockers can improve renal function or delay progression. A study in a rat model of AAN showed no difference in outcome with sodium restriction or with treatment with enalapril,

711

712

Section 4  

the patient with interstitial disease

or enalapril and candesartan combined (Debelle et al., 2004). There is, however, some evidence that steroid treatment can modify the course of the disease. A non-randomized study of steroid therapy in the original Belgian cohort of AAN patients showed a significant slowing in the progression of CKD in 12 treated patients compared with well-matched historical controls from the Belgian registry (Vanherweghem et  al., 1996). Based on the available evidence, a therapeutic trial of steroids may be warranted in patients with a diagnosis of AAN and an estimated GFR of > 20 mL/min. Aside from steroid therapy for appropriately selected patients, the priorities in the management of AAN are similar to those in other causes of CKD: careful blood pressure control, cardiovascular risk reduction, management of metabolic complications, and timely preparation for renal replacement therapy.

Dialysis and transplantation As with most patients with end-stage kidney disease, transplantation is the treatment of choice for AAN patients requiring renal replacement therapy. Patients should be evaluated for pre-emptive living donor transplantation with combined bilateral nephro-ureterectomy. The disease has not been found to recur following transplantation (Reginster et al., 1997), although, if bilateral nephro-ureterectomy is not performed at or before transplantation, the risk of malignancy in the native urinary tract after transplantation has been reported to be as high as 52.9% over a median follow-up period of 57 months (Yuan et al., 2009). If transplantation is delayed, or is not possible on medical grounds, haemodialysis is preferred to peritoneal dialysis, given the extensive surgical intervention that patients with AAN are likely to require.

Urological management Given the substantial risk of malignancy, many nephrologists would advise that all patients with a confirmed diagnosis of AAN be offered bilateral nephro-ureterectomy at the point of needing renal replacement therapy. As noted above, this would best be performed in the context of planned living donor renal transplantation. In patients with AAN who do not yet require renal replacement therapy, ongoing surveillance with computed tomography imaging and ureteroscopy is warranted. The role of cystectomy is unclear, as the incidence of bladder tumours has been found to be lower than that of upper tract malignancy. Patients and clinicians will need to consider the relative merits of regular urine cytology, cystoscopy, non-invasive imaging and prophylactic radical cystectomy on an individual basis. However, many individuals may choose to undergo cystectomy if AA-DNA adducts are detected in bladder specimens. It is hoped that non-invasive biomarkers may be identified that will better guide these difficult management decisions.

References Arlt, V. M., Alunni-Perret, V., Quatrehomme, G., et al. (2004). Aristolochic acid (AA)-DNA adduct as marker of AA exposure and risk factor for AA nephropathy-associated cancer. Int J Cancer, 111(6), 977–80. Arlt, V. M., Schmeiser, H. H., and Pfeifer, G. P. (2001). Sequence-specific detection of aristolochic acid-DNA adducts in the human p53 gene by terminal transferase-dependent PCR. Carcinogenesis, 22(1), 133–40. Arlt, V. M., Stiborová, M., vom Brocke, J., et al. (2007). Aristolochic acid mutagenesis: molecular clues to the aetiology of Balkan endemic nephropathy-associated urothelial cancer. Carcinogenesis, 28(11), 2253–61.

Arlt, V. M., Stiborová, M., vom Brocke, J., et al. (2002). Is aristolochic acid a risk factor for Balkan endemic nephropathy-associated urothelial cancer? Int J Cancer, 101(5), 500–2. Atanasova, S. Y., von Ahsen, N., Toncheva, D. I., et al. (2005). Genetic polymorphisms of cytochrome P450 among patients with Balkan endemic nephropathy (BEN). Clin Biochem, 38(3), 223–8. Attaluri, S., Bonala, R. R., Yang, I. Y., et al. (2010). DNA adducts of aristolochic acid II: total synthesis and site-specific mutagenesis studies in mammalian cells. Nucleic Acids Res, 38(1), 339–52. Broschard, T. H., Wiessler, M., von der Lieth, C. W., et al. (1994). Translesional synthesis on DNA templates containing site-specifically placed deoxyadenosine and deoxyguanosine adducts formed by the plant carcinogen aristolochic acid. Carcinogenesis, 15(10), 2331–2340. Danilovic, V. (1958). Chronic nephritis due to ingestion of lead-contaminated flour. Br Med J, 1(5061), 27–8. Debelle, F. D., Nortier, J. L., Husson, C. P., et al. (2004). The renin-angiotensin system blockade does not prevent renal interstitial fibrosis induced by aristolochic acids. Kidney Int, 66(5), 1815–25. Debelle, F. D., Vanherweghem, J. -L., and Nortier, Joëlle, L. (2008). Aristolochic acid nephropathy: a worldwide problem. Kidney Int, 74(2), 158–69. European Parliament and Council of the European Union (2004). Directive 2004/24/EC on Traditional Herbal Products. OJ, L136, 85–90. Food and Drug Administration (2001). Alerts—Dietary Supplements: Aristolochic Acid. [Online] Fernando, R. C., Schmeiser, H. H., Scherf, H. R., et al. (1993). Formation and persistence of specific purine DNA adducts by 32P-postlabelling in target and non-target organs of rats treated with aristolochic acid I. IARC Sci Publ, 124, 167–71. Gao, R., Zheng, F., Liu, Y., et al. (2000). Aristolochic acid I-induced apoptosis in LLC-PK1 cells and amelioration of the apoptotic damage by calcium antagonist. Chin Med J, 113(5), 418–24. Gold, L. S. and Slone, T. H. (2003). Aristolochic acid, an herbal carcinogen, sold on the Web after FDA alert. N Engl J Med, 349(16), 1576–7. Gökmen, M. R., Cosyns, J. P., Arlt, V. A., et al. (2013). The epidemiology, diagnosis and management of aristolochic acid nephropathy: a narrative review. Ann Intern Med, 158, 469–77 Gökmen, M. R. and Lord, G.M. (2012). Aristolochic acid nephropathy. BMJ, 344, e4000 Grollman, A. P., Shibutani, S., Moriya, M., et al. (2007). Aristolochic acid and the etiology of endemic (Balkan) nephropathy. Proc Natl Acad Sci U S A, 104(29), 12129–34. Heinrich, M., Chan, J., Wanke, S., et al. (2009). Local uses of Aristolochia species and content of nephrotoxic aristolochic acid 1 and 2—a global assessment based on bibliographic sources. J Ethnopharmacol, 125(1), 108–44. Hranjec, T., Kovac, A., Kos, J., et al. (2005). Endemic nephropathy: the case for chronic poisoning by aristolochia. Croat Med J, 46(1), 116–25. Hsin, Y. -H., Cheng, C. H., Tzen, J. T., et al. (2006). Effect of aristolochic acid on intracellular calcium concentration and its links with apoptosis in renal tubular cells. Apoptosis, 11(12), 2167–77. Ivić, M. (1969). [Etiology of endemic nephropathy]. Liječnički vjesnik, 91(12), 1273–81. Jelaković, B., Karanović, S., Vuković-Lela, I., et al. (2012). Aristolactam-DNA adducts are a biomarker of environmental exposure to aristolochic acid. Kidney Int, 81(6), 559–67. Kabanda, A., Jadoul, M., Lauwerys, R., et al. (1995). Low molecular weight proteinuria in Chinese herbs nephropathy. Kidney Int, 48(5), 1571–6. Krumme, B., Endmeir, R., Vanhaelen, M., et al. (2001). Reversible Fanconi syndrome after ingestion of a Chinese herbal ‘remedy’ containing aristolochic acid. Nephrology Dial Transplant, 16(2), 400–2. Lai, M. -N., Lai, J. N., Chen, P. C., et al. (2009). Increased risks of chronic kidney disease associated with prescribed Chinese herbal products suspected to contain aristolochic acid. Nephrology, 14(2), 227–34.

Chapter 89 

Lebeau, C., Debelle, F. D., Arlt, V. M., et al. (2005). Early proximal tubule injury in experimental aristolochic acid nephropathy: functional and histological studies. Nephrology Dial Transplant, 20(11), 2321–32. Lemy, A., Wissing, K. M., Rorive, S., et al. (2008). Late onset of bladder urothelial carcinoma after kidney transplantation for end-stage aristolochic acid nephropathy: a case series with 15-year follow-up. Am J Kidney Dis, 51(3), 471–7. Lord, G. M., Cook, T., Arlt, V. M., et al. (2001). Urothelial malignant disease and Chinese herbal nephropathy. Lancet, 358(9292), 1515–16. Lord, G. M., Hollstein, M., Arlt, V. M., et al. (2004). DNA adducts and p53 mutations in a patient with aristolochic acid-associated nephropathy. Am J Kidney Dis, 43(4), e11–7. Martinez, M. -C. M., Nortier, J., Vereerstraeten, P., et al. (2002). Progression rate of Chinese herb nephropathy: impact of Aristolochia fangchi ingested dose. Nephrology Dial Transplant, 17(3), 408–12. Nortier, J. L., Martinez, M. C., Schmeiser, H. H., et al. (2000). Urothelial carcinoma associated with the use of a Chinese herb (Aristolochia fangchi). N Engl J Med, 342(23), 1686–92. Nortier, J. L., Schmeiser, H. H., Muniz Martinez, M. C., et al. (2003). Invasive urothelial carcinoma after exposure to Chinese herbal medicine containing aristolochic acid may occur without severe renal failure. Nephrology Dial Transplant, 18(2), 426–8. Okada, H., Watanabe, Y., Inoue, T., et al. (2003). Transgene-derived hepatocyte growth factor attenuates reactive renal fibrosis in aristolochic acid nephrotoxicity. Nephrology Dial Transplant, 18(12), 2515–23. Pešić, I., Stefanović, V., Müller, G. A., et al. (2011). Identification and validation of six proteins as marker for endemic nephropathy. J Proteomics, 74(10), 1994–2007. Poon, W. (2007). Aristolochic acid nephropathy: the Hong Kong perspective. Hong Kong J Nephrol, 9(1), 7–14. Pozdzik, A. A., Berton, A., Schmeiser, H. H., et al. (2010). Aristolochic acid nephropathy revisited: a place for innate and adaptive immunity? Histopathology, 56(4), 449–63. Pozdzik, A. A., Salmon, I. J., Debelle, F. D., et al. (2008). Aristolochic acid induces proximal tubule apoptosis and epithelial to mesenchymal transformation. Kidney Int, 73(5), 595–607. Reginster, F., Jadoul, M., and van Ypersele de Strihou, C. (1997). Chinese herbs nephropathy presentation, natural history and fate after transplantation. Nephrology Dial Transplant, 12(1), 81–6. Sato, N., Takahashi, D., Chen, S. M., et al. (2004). Acute nephrotoxicity of aristolochic acids in mice. J Pharm Pharmacol, 56(2), 221–9. Schmeiser, H., Stiborova, M., and Arlt, V. M. (2009). Chemical and molecular basis of the carcinogenicity of Aristolochia plants. Curr Opin Drug Discov Devel, 12(1), 141–8. Schmeiser, H. H., Bieler, C. A., Wiessler, M., et al. (1996). Detection of DNA adducts formed by aristolochic acid in renal tissue from patients with Chinese herbs nephropathy. Cancer Res, 56(9), 2025–8.

aristolochic acid nephropathy

Stefanovic, V., Djukanović, L., Cukuranović, R., et al. (2011). Beta2-microglobulin and alpha1-microglobulin as markers of Balkan endemic nephropathy, a worldwide disease. Ren Fail, 33(2), 176–83. Stefanovic, V., Toncheva, D., Atanasova, S., et al. (2006). Etiology of Balkan endemic nephropathy and associated urothelial cancer. Am J Nephrol, 26(1), 1–11. Stiborova, M., Frei, E., and Schmeiser, H.H. (2008). Biotransformation enzymes in development of renal injury and urothelial cancer caused by aristolochic acid. Kidney Int, 73(11), 1209–11. Tanaka, A., Nishida, R., Maeda, K., et al. (2000). Chinese herb nephropathy in Japan presents adult-onset Fanconi syndrome: could different components of aristolochic acids cause a different type of Chinese herb nephropathy? Clin Nephrol, 53(4), 301–6. Tanchev, I., Evstatiev, T. S., Dorosiev, D., et al. (1956). [Study of nephritis in Vrattsa district]. Suvr Med (Sofiia), 7(9), 14–29. Toncheva, D.I., Von Ahsen, N., Atanasova, S. Y., et al. (2004). Identification of NQO1 and GSTs genotype frequencies in Bulgarian patients with Balkan endemic nephropathy. J Nephrol, 17(3), 384–9. Vanhaelen, M., Vanhaelen-Fastre, R., But, P., et al. (1994). Identification of aristolochic acid in Chinese herbs. Lancet, 343(8890), 174. Vanherweghem, J. L. (1997). Aristolochia sp and chronic interstitial nephropathies in Indians. Lancet, 349(9062), 1399. Vanherweghem, L. J. (1998). Misuse of herbal remedies: the case of an outbreak of terminal renal failure in Belgium (Chinese herbs nephropathy). J Altern Complement Med, 4(1), 9–13. Vanherweghem, J. L., Abramowicz, D., Tielemans, C., et al. (1996). Effects of steroids on the progression of renal failure in chronic interstitial renal fibrosis: a pilot study in Chinese herbs nephropathy. Am J Kidney Dis, 27(2), 209–15. Vanherweghem, J. L., Depierreux, M., Tielemans, C., et al. (1993). Rapidly progressive interstitial renal fibrosis in young women: association with slimming regimen including Chinese herbs. Lancet, 341(8842), 387–91. World Health Organization (2004). Aristolochic acid—to be replaced by Stephania tetrandra and Inula helenium. Pharmaceut Newslett, 5. [Online] Yang, C. S., Lin, C. H., Chang, S. H., et al. (2000). Rapidly progressive fibrosing interstitial nephritis associated with Chinese herbal drugs. Am J Kidney Dis, 35(2), 313–18. Yang, L., Su, T., Li, X. M., et al. (2011). Aristolochic acid nephropathy: variation in presentation and prognosis. Nephrology Dial Transplant, 27(1), 292–8. Yuan, M., Shi, Y. B., Li, Z. H., et al. (2009). De novo urothelial carcinoma in kidney transplant patients with end-stage aristolochic acid nephropathy in China. Transplantat Proc, 41(5), 1619–23. Zhou, L., Fu, P., Huang, X. R., et al. (2010). Mechanism of chronic aristolochic acid nephropathy: role of Smad3. Am J Physiol Renal Physiol, 298(4), F1006–17.

713

CHAPTER 90

Balkan endemic nephropathy Milan Radović and Adalbert Schiller Epidemiology During World War II, clinical observations of a slowly progressive form of chronic kidney disease (CKD) with familial clustering led to the initial descriptions of Balkan endemic nephropathy (BEN). After 1950, independently from each other, authors from three Balkan countries (Danilovic in Yugoslavia, Foarta, Tonea, Bruckner, Lazarescu, and Zosin in Romania, and Dimitrov, Tanchev, and Ivanov in Bulgaria) reported high death rates from uraemia in certain small geographical areas. The first cases were detected in the Kolubara region in Serbia (Danilović et al., 1957), Mehedinti and Caras-Severin counties in Romania (Bruckner et al., 1979), and the Vratsa region in Bulgaria (Puchlev et  al., 1979). Soon after, BEN was also reported in Croatia (Pichler et al., 1959; Čeović et al., 1979), Bijeljina-Posavina, in Bosnia and Herzegovina, and Kosovska Vitina, Pomoravlje, and West Morava, in Serbia (Danilović, 1979). BEN affects both genders evenly. Most patients are farmers. The peak incidence of BEN onset was initially described to be in the third and fourth decades of life; however, this has recently shifted towards older ages (Bukvić et al., 2009). In endemic areas, the incidence of BEN ranges from 55 to 85 new cases per 100,000 inhabitants and per year (Janković et al., 2011), whereas the prevalence is 0.5–4% (Stefanović and Cosyns, 2005). In the same areas, the incidence of end-stage renal disease (ESRD) due to BEN varies between 5.4 and 12.8 patients per million-population; however, a decreasing trend has been noticed during the past few decades (Čukuranović et al., 2007). Approximately 11% of all patients treated with haemodialysis in countries with endemic regions had BEN as a cause of ESRD (Djukanović et al., 2002); this prevalence has remained relatively stable during the last 10 years (Čukuranović et al., 2007; Janković et al., 2011).

Clinical manifestations BEN is a disease with an insidious onset and slow, progressive course. When associated with urothelial cancers, haematuria, pain, and urinary tract obstruction may occur. The diagnosis of BEN is based on a patient’s medical history, physical examination, laboratory tests, kidney imaging, and additional investigations. The diagnostic criteria were established by Danilović in 1979 and updated by Djukanovic et  al. in 2007, as follows: 1. Farmers residing in endemic areas 2. Family history of BEN 3. Mild proteinuria (< 500 mg/day)

4. Low urinary specific gravity (< 1.010), or low urinary osmolality (< 850 mOsm/kg in patients < 20 years old to < 600 mOsm/kg in those > 60 years old) 5. Anaemia (haemoglobin < 13.0 g/dL in men and post-menopausal women and < 12.0 g/dL in pre-menopausal women) 6. Azotaemia (defined as serum creatinine > 125 μmol/L or serum urea > 8.0 mmol/L) 7. Symmetrically small kidneys, with smooth outlines The presence of the first three criteria and at least one of the remaining four raises the suspicion of BEN, whereas definite BEN requires five criteria (of which the first two are mandatory). However, these criteria enable only the detection of advanced BEN cases and offer little support for the differential diagnosis with other CKDs. More recently, other investigations, such as urinary alfa-1-microglobulin and microalbuminuria, have been proposed as additional diagnostic criteria for BEN, but they still need validation (Radović and Djukanović, 2004; Djukanović et al., 2007, 2008; Imamović et al., 2008; Gluhovschi et al., 2011; Pešić et al., 2011). Typically, BEN patients are farmers in their fifth or sixth decade of life, with a positive family history of CKD, residing in villages where other inhabitants also have CKD. Physical examination is often unremarkable. Patients may be pale, or with a grey or yellowish skin colour (xanthodermia). Mild hypertension can be found. Laboratory examination of urine sediment is typically negative. If haematuria is found, the patient should be promptly investigated for urothelial cancer or glomerular disease. Urine specific gravity is low. Urinalysis reveals mild (usually < 500 mg/day) tubular proteinuria. Decreased glomerular filtration rate (GFR) is a common finding. Additionally, normochromic normocytic anaemia is often present. Renal imaging exams (ultrasound, intravenous pyelogram, computed tomography, and magnetic resonance) reveal markedly reduced kidney size, with a smooth outline. Ultrasound is the most cost-effective and reliable method for diagnostic screening (Ležaić et al., 2008; Hanjangsit et al., 2010). Radiology studies are also performed in search for upper urothelial cancers. Kidney biopsy is usually not indicated in BEN patients, for the following reasons: (a) diagnostic criteria are not based on histology, (b) the kidneys have reduced size and cortical thickness, and (c) biopsy findings do not influence the therapy options. However, the lack of pathological studies is one of the most important reasons why the pathophysiology of BEN is still unclear. The differential diagnosis of BEN should particularly consider Chinese herbs nephropathy, as well as other toxic chronic tubulointerstitial nephritis (CTIN), such as analgesic nephropathy,

Chapter 90 

where kidneys have bumpy contours and papillary calcifications and medical history reveals analgesic abuse (Elseviers et al., 1995).

Aetiology and pathogenesis The aetiology of BEN has remained unclear for the last 50 years. However, some of its features (its limited geographical distribution and its association with urothelial cancers) suggest that both genetic and environmental factors are probably involved.

Genetic BEN does not show a typical Mendelian inheritance pattern. Nevertheless, offspring of BEN patients are at high risk of developing CKD and they were sometimes found to have reduced kidney size and cortex thickness, elevated blood pressure (Dimitrov et al., 2006, 2007), as well as high CRP levels, suggesting an early inflammatory state, which is inversely related to the renal cortex width (Karmaus et al., 2009). Reduced GFR and tubular proteinuria have also been detected in family members of BEN (especially female) patients (Stefanović et al., 2002; Arsenović et al., 2005; Dimitrov et al., 2006). Several genetic factors have been implicated in the aetiopathogenesis of BEN, including abnormalities in chromosome 3q25–3q26, transforming growth factor beta (TGF-β) gene, genetic heterogeneity of xenobiotic-metabolizing enzymes or immune system genetic defects. The predisposition for urothelial cancers of BEN patients could be due to germline mutations in tumour suppressor genes or acquired somatic mutations in oncogenes (Toncheva et al., 1998, 2002). Since the angiotensin receptor gene is located on the 3q chromosome (Goodfriend et al., 1996; Heiber et al., 1996), the potential role of angiotensin-converting enzyme (ACE) gene polymorphism in the development of interstitial fibrosis in BEN has been suggested (Huskić et al., 1996), but not yet confirmed (Krcunović et al., 2010). Genetic deficiencies of enzymes such as lecithin-cholesterol acyltransferase (LCAT), erythrocyte delta-aminolevulinate dehydratase (ALA-D), and cytochrome P450 2D6 have also been suggested to play a role in BEN (Djordjevic et al., 1991; Nikolov et al., 1991; Pavlovic et al., 1991). More recently, cytochrome P450 3A5*1 (CYP3A5*1) (Atanasova et al., 2005) and glutathione-S-transferase M1 (GSTM1) wild-type (Andonova et al., 2004) allele carriers were found to be associated with a higher risk of BEN.

Aristolochic acid The occurrence of BEN in immigrants to (Čeović et al., 1985) and in emigrants from endemic regions (Nikolić et al., 2006) strongly suggests the role of environmental factors in the aetiology of this disease. Many such factors have been suspected so far, but aristolochic acid currently appears to be the most convincing one. The relation between aristolochic acid and BEN was first described by Ivić in 1969 and was later supported by the detection of Aristolochia clematitis in wheat, wheat flour, and bread consumed by inhabitants of endemic regions. The incorporation of aristolochic acid into the genome of patients with BEN and urothelial cancers was confirmed a few years ago (Grollman and Jelaković, 2007; Grollman et al., 2007). Aristolochic acid nephropathy is discussed in detail in Chapter 89. It is a genotoxic mutagen, forming DNA adducts after metabolic activation. Aristolochic acid DNA and aristolactam-DNA adducts in renal cortex and AT-TA p53 mutations were found in patients with both BEN and urothelial cancers (Arlt et al., 2007; Slade et al., 2009; Jelaković et al., 2012).

balkan endemic nephropathy

Aristolochic acid is able to promote both renal interstitial fibrosis and urothelial carcinogenesis. The role of aristolochic acid in BEN is also suggested by similarities in the clinical course and histological features between BEN and aristolochic acid (Chinese herbs) nephropathy. However, there may be a genetic predisposition for a particular aristolochic acid metabolic processing and an individual susceptibility for BEN (Vanherweghem et al., 1993; Cosyns et al., 1994; Depierreux et al., 1994; Cosyns et al., 2001; Arlt et al., 2002; de Jonge and Vanrenterghem, 2008), to explain the high incidence in the region.

Other toxins Poisoning of the drinking water with water-soluble carcinogenic hydrocarbons from Pliocene-epoch lignites (Orem et al., 2004) was another explored hypothesis. Almost all endemic areas in the Balkans are lying over Pliocene lignites. These lignites contain water-soluble organic compounds, including polar polycyclic aromatic hydrocarbons and aromatic amines, which are known to be carcinogenic. This hypothesis has not yet been confirmed, despite some preliminary findings (Schiller et al., 2008). The role of environmental trace elements, like cadmium, lead, silica, manganese, copper, and selenium, has also been suggested (Diven et al., 1979; Stefanović et al., 2006), but still remains unclear. Ochratoxin A  (OTA), a product of Penicillium or Aspergillus strains, which may contaminate foods, is a nephrotoxic, teratogenic, immunotoxic, and carcinogenic mycotoxin. OTA may cause tissue injury via oxidative DNA damage and lethal or sublethal cellular cytotoxicity (Kamp et al., 2005). It has also been shown to have a dose-dependent oncogenetic effect in vitro (Arlt et al., 2001; Cosyns et  al., 2001; Mally et  al., 2007; Slade et  al., 2009). OTA-induced DNA adducts have been shown in animal models. It has been suggested that CYP3A5*1 allele carriers can convert more efficiently OTA into genotoxic metabolites (Pfohl-Leszkowicz, 2009). This mycotoxin has been considered as a possible aetiologic factor for BEN; however, a direct relation between OTA contamination and BEN has not been demonstrated. Furthermore, the clinical course of BEN is typically much slower than the one seen in OTA-induced nephropathy (Cosyns et al., 2001; Hassen et al., 2004).

Infection A viral aetiology for BEN was a topic of research during the 1970s and 1980s. Coronavirus-like particles were found in the blood and kidney biopsies of BEN patients (Apostolov and Spasić, 1975; Georgescu et al., 1978; Uzelac-Keserović et al., 2000). Other studies suggested a role for adenoviruses (Georgescu et al., 1978) and papovaviruses (Stoian et al., 1983). In Bulgarian BEN patients, elevated urinary levels of neopterin—a marker of activated T-helper-1 immune response, viral infections, and malignancies—have been identified (Toncheva et al., 2003). The viral hypothesis of BEN aetiology still awaits confirmation.

Metabolomics Recently, proton nuclear magnetic resonance (1H NMR) spectroscopic analysis has been used for metabolomic studies of the urine of BEN patients from Romania and Bulgaria. Principal component analysis clustered healthy controls from both countries together. Bulgarian BEN patients clustered separately from Bulgarian controls, whereas Romanian patients not only clustered away from controls, but also clustered separately from the Bulgarian patients.

715

716

Section 4  

the patient with interstitial disease

Moreover, the urinary metabolomics of two individuals among Romanian controls clustered within the Romanian BEN group. One of these had been suspected of incipient BEN at the time of selection as a ‘healthy’ control. At least two conclusions emerged from this pilot study:  (1)  metabolomic urinalysis could predict BEN before the development of clinical signs, and (2)  considering the separate clustering of BEN patients from Romania and Bulgaria, a different aetiology in different geographic areas is likely (Mantle et al., 2011).

Models There is no specific or validated experimental model for BEN. However, aristolochic acid-, OTA-, and ciclosporin-induced experimental nephropathies can mimic some of the BEN features (Ferluga et al., 1991; Pichler et al., 1995).

Pathology The kidneys of BEN patients are smoothly outlined and symmetrically shrunken. In advanced stages, a kidney’s weight may be reduced to as little as 30 g. In up to 35% of cases, urothelial carcinoma of the renal pelvis or ureter can be found, and in 13% of cases, it is bilateral (Djokić et al., 1999). The histopathology of BEN is characterized by diffuse fibrosis of the cortical interstitium, tubular atrophy, and absence of cellular interstitial infiltration, which classifies BEN among non-inflammatory CTIN. The study by Ferluga and colleagues, including the largest number of kidney biopsies in BEN patients (50 cases), found interstitial fibrosis in 98% and tubular atrophy in 96% (Ferluga et al., 1991; Trnačević et al., 1991). BEN kidneys show striped cortical, hypocellular interstitial fibrosis, interstitial oedema, and tubular atrophy. These changes, scattered in the early stages, become diffuse in the later stages, when hypocellular interstitial fibrosis is prevailing. Fibrosis surrounds the proximal tubules, whose epithelia become flattened and atrophic. The structure of the glomeruli is normal in the early phase of the disease, but later may have an ischaemic appearance (Hall et al., 1978; Dojčinov et al., 1979; Sindjić et al., 1979; Ferluga et al., 1991). Blood vessels may show afferent artery hyalinosis and, occasionally, peritubular capillary wall thickening (Sindjić et  al., 1979)  (Fig. 90.1). In some cases, rare foci of segmental sclerosis or intracapillary lesions mimicking thrombotic microangiopathy were seen. The medulla is usually preserved. There is a striking histologic resemblance of BEN with chronic ciclosporin nephropathy (Ferluga et al., 1991). Immunofluorescent microscopy findings are non-specific. Scanty, granular segmental immunoglobulin M (IgM) or C3 deposits may occasionally be present in the glomeruli. Rarely, the tubular basement membrane may show granular C3 deposits. Arterial walls may also contain C3 and, rarely, IgM deposits. On immunohistochemistry staining, overexpression of laminin in the interstitial capillaries and in the tubules, as well as a co-expression of vimentin and cytokeratin in the tubular epithelial cells, has been described (Savić et al., 2002). Electron microscopy shows increased interstitial bundles of collagen fibres and elongated, stellate fibroblast-like cells. The intercellular junctions of tubular cells are widened. The tubular basement membrane is thickened and it splits up the peritubular capillary basement membrane. Peritubular endothelial cell

Fig. 90.1  Balkan endemic nephropathy (optical microscopy, haematoxylin-eosin): striped cortical, diffuse hypocellular interstitial fibrosis, tubular atrophy with interstitial oedema. Fibrosis surrounds the proximal tubules, whose epithelia are flattened and atrophic. The structure of the glomeruli shows an ischaemic appearance. Afferent artery hyalinosis and peritubular capillary wall thickening is seen. Courtesy of Prof. Dr Jasmina Marković—Lipkovski, Institute of Pathology, University of Belgrade, School of Medicine, Belgrade, Serbia.

swelling is present. The glomeruli are preserved, with occasional mild thickening of the glomerular basement membrane (Hvala et al., 2005).

Treatment and outcome Since the aetiology is unknown, there is no specific treatment and no prevention strategies for BEN. Avoidance of exposure to aristolochic acid is difficult (Schiller et al., 2008). Even in such a scenario, an embedded genetic imprint may reside within the kidneys and contribute to the occurrence of tubulointerstitial nephritis in emigrants from endemic regions. The treatment of BEN is non-specific and similar to other CKD cases of unknown aetiology, including blood pressure control, preferably by using ACE inhibitors or angiotensin receptor blockers, treatment of dyslipidaemia, anaemia, and CKD-related mineral and bone disorder. Patients with ESRD caused by BEN can be treated by either dialysis or kidney transplantation. Careful evaluation for urothelial cancers is mandatory. Since BEN is characterized by acellular urine sediment, the finding of haematuria in a BEN patient is an indication for thorough diagnostic imaging workup (Djokić et al., 2001, 2006). Upper urothelial cancers in renal transplant recipients occur 50 times more frequently in patients with BEN than in those without BEN (Bašić-Jukić et al., 2007; Zivcić-Kosić et al., 2007). The life expectancy of BEN patients is similar to that of the general population. BEN specific mortality is 0.65 per 100,000 inhabitants; however, when urothelial cancers also develop, the mortality raises significantly to 4.7–9.0 per 100,000 inhabitants (Bukvić et al. 2000; Miletić-Medved et al., 2005).

References Apostolov, K. and Spasić, P. (1975). Evidence of a viral aetiology in endemic (Balkan) nephropathy. Lancet, 2(7948), 1271–3.

Chapter 90 

Andonova, R., Sarueva, A., Horvath, V., et al. (2004). Balkan endemic nephropathy and genetic variants of glutathione S-transferases. J Nephrol, 17, 390–8. Arlt, V. M., Pfohl-Leszkowicz, A., Cosyns, J., et al. (2001). Analyses of DNA adducts formed by ochratoxin A and aristolochic acid in patients with Chinese herbs nephropathy. Mutat Res, 494, 143–50. Arlt, V. M., Striborova, M., and Schmeiser, H. H. (2002). Aristolochic acid as a probable human cancer hazard in herbal remedies: a review: Mutagenesis, 174, 265–77. Arlt, V. M., Stiborova, M., vom Brocke, J., et al. (2007). Aristolochic acid mutagenesis: molecular clues to the aetiology of Balkan endemic nephropathy-associated urothelial cancer. Carcinogenesis, 28, 253–61. Arsenović, A., Bukvić, D., Trbojević, S., et al. (2005). Detection of renal dysfunctions in family members of patients with Balkan endemic nephropathy. Am J Nephrol, 25, 50–4. Atanasova, S. Y., Armstrong, V. W., von Ahsen, N., et al. (2005). Genetic polymorphisms of cytochrome P450 among patients with Balkan endemic nephropathy (BEN). Clin Biochem, 38, 223–8. Bašić-Jukić, N., Hršak-Puljić, I., Kes, P., et al. (2007). Renal transplantation in patients with Balkan endemic nephropathy. Transplant Proc, 39, 1432–5. Bruckner, I., Nicifor, F., and Rusu G. (1979). Endemic nephropathy in Romania, In S. Strahinjić and V. Stefanović (eds.) Endemic (Balkan) Nephropathy. Proceeding of the 4th Symposium on Endemic (Balkan) Nephropathy, pp. 11–14. Niš: University Press. Bukvić, D., Janković, S., Dukanović, L., et al. (2000). Survival of Balkan endemic nephropathy patients. Nephron, 86, 63–6. Bukvić, D., Janković, S., Marić, I., et al. (2009). Today Balkan endemic nephropathy is a disease of the elderly with a good prognosis. Clin Nephrol, 72, 105–13. Cosyns, J. P., Dehoux, J. P., Guiot, Y., et al. (2001). Chronic aristolochic acid toxicity in rabbits: a model of Chinese herbs nephropathy. Kidney Int, 59, 2164–73. Cosyns, J. P., Jadoul, M., Squifflet, J. P., et al. (1994). Chinese herbs nephropathy: a clue to Balkan endemic nephropathy. Kidney Int, 45, 1680–8. Čeović, S., Hrabar, A., and Radonić, M. (1985). An etiological approach to Balkan endemic nephropathy based on investigation of two genetically different populations. Nephron, 40, 175–9. Čeović, S., Radonić, M., Hrabar, A., et al. (1979). Endemic nephropathy in Brodska Posavina—in a twenty-year period. In S. Strahinjić and V. Stefanović (eds.) Endemic (Balkan) Nephropathy. Proceeding of the 4th Symposium on Endemic (Balkan) Nephropathy, pp. 223–7. Niš: University Press. Čukuranović, R., Jovanović, I., Miljković, S., et al. (2007). Hemodialysis treatment in patients with Balkan endemic nephropathy: an epidemiological study. Ren Fail, 29, 805–10. Danilović, V. (1979). Endemic nephropathy in Yugoslavia. In S. Strahinjić and V. Stefanović (eds.) Endemic (Balkan) Nephropathy. Proceeding of the 4th Symposium on Endemic (Balkan) Nephropathy, pp. 1–5. Niš: University Press. Danilović, V., Djurišić, M., Mokranjac, M., et al. (1957). Néphrites chroniques provoquées par l’intoxication au plomb par voie digestive (farine). Presse Méd, 65, 2039–40. De Jonge, H. and Vanrenterghem. Y. (2008). Aristolochic acid: the common culprit of Chinese herbs nephropathy and Balkan endemic nephropathy. Nephrol Dial Transplant, 23, 39–41. Depierreux, M., Van Damme, B., Vanden Houte, K., et al. (1994). Pathologic aspects of a newly described nephropathy related to the prolonged use of Chinese herbs. Am J Kidney Dis, 24, 172–80. Dimitrov, P., Tsolova, S., Georgieva, R., et al. (2006). Clinical markers in adult offspring of families with and without Balkan endemic nephropathy. Kidney Int, 69, 723–9. Dimitrov, P., Simeonov, V. A., Tsolova, S. D., et al. (2007). Increased blood pressure in adult offspring of families with Balkan endemic nephropathy: a prospective study. BMC Nephrol, 7, 12. Diven, I. H., Makarov, V. J., Topkbashjan, S. A., et al. (1979). Level of trace elements in biological materials of patients with Balkan endemic

balkan endemic nephropathy

nephropathy (BEN). In S. Strahinjić and V. Stefanović (eds.) Endemic (Balkan) Nephropathy. Proceeding of the 4th Symposium on Endemic (Balkan) Nephropathy, pp. 53–5. Niš: University Press. Djokić, M., Dragičević, D., Nikolić, J., et al. (2006). Bilateral tumors of the upper urothelium. Srp Arh Celok Lek, 134, 509–15. Djokić, M., Hadži-Djokić, J., Nikolić, J., et al. (1999). Comparison of upper urinary tract tumors in the region of Balkan nephropathy with those of other regions of Yugoslavia. Prog Urol, 9, 61–8. Djokić, M., Hadži-Djokić, J., Nikolić, J., et al. (2001). Tumors of the upper urinary tract: results of conservative surgery. Prog Urol, 11, 1231–8. Djordjevic, V. B., Strahinjic, S., Koracevic, D., et al. (1991). d-Aminolevulinate dehydratase measurements in Balkan endemic nephropathy. Kidney Int, 40, S93–6. Djukanović, L., Marić, I., Marinković, J., et al. (2007). Evaluation of criteria for the diagnosis of Balkan endemic nephropathy. Ren Fail, 29, 607–14. Djukanović, L., Marinković, J., Marić, I., et al. (2008). Contribution to the definition of diagnostic criteria for Balkan endemic nephropathy. Nephrol Dial Transplant, 23, 3932–8. Djukanović, L., Radović, M., Baković, J., et al. (2002). Epidemiology of end-stage renal disease and current status of hemodialysis in Yugoslavia. Int J Artif Organs, 25, 852–9. Dojčinov, D., Strahinjić, S., and Stefanović, V. (1979). Ultrastructure of the kidney in the early phases of endemic (Balkan) nephropaty. In S. Strahinjić and V. Stefanović (eds.) Endemic (Balkan) Nephropathy. Proceeding of the 4th Symposium on Endemic (Balkan) Nephropathy, pp. 105–12. Niš: University Press. Elseviers, M. M., De Schepper, A., Corthouts, R., et al. (1995). High diagnostic performance of CT scan for analgesic nephropathy in patients with incipient to severe renal failure. Kidney Int, 48, 1316–23. Ferluga, D., Hvala, A., Vizjak, A., et al. (1991). Renal function, protein excretion, and pathology of Balkan endemic nephropathy. III. Light and electron microscopic studies. Kidney Int Suppl, 34, S57–67. Georgescu, L., Diosi, P., Butiu, I., et al. (1978). Porcine coronavirus antibodies in endemic (Balkan) nephropathy. Lancet, 1(8056), 163. Georgescu, L., Diosi, P., Plavoşin, L., et al. (1978). Adenovirus in Balkan nephropathy. Lancet, 2(8087), 482. Gluhovschi, G., Margineanu, F., Velciov, S., et al. (2011). Fifty years of Balkan endemic nephropathy in Romania: some aspects of the endemic focus in the Mehedinti county. Clin Nephrol, 75, 34–48. Goodfriend, T. L., Elliot, M. E., and Catt, K. J. (1996). Angiotensin receptors and their antagonists. N Engl J Med, 334, 1649–55. Grollman, A. P. and Jelaković, B. (2007). Role of environmental toxins in endemic (Balkan) nephropathy. J Am Soc Nephrol, 18, 2817–23. Grollman, A., Shibutani, S., Moriya, M., et al. (2007). Aristolochic Acid and the etiology of endemic (Balkan) nephropathy. Proc Natl Acad Sci U S A, 104, 12129–34. Hall, P. W. and Dammin, G. J. (1978). Balkan nephropathy. Nephron, 22, 281–300. Hanjangsit, K., Dimitrov, P., Karmaus, W., et al. (2010). Reduced kidney size in adult offspring of Balkan endemic nephropathy patients and controls: a prospective study. Am J Med Sci, 340, 94–102. Hassen, W., Abid, S., Achour, A. et al. (2004). Ochratoxin A and beta -2–microglobulinuria in healthy individuals and in chronic interstitial nephropathy patients in the Centre of Tunisia: a hot spot of Ochratoxin A exposure. Toxicology, 199, 185–93. Heiber, M., Marchese, A., Nguyen, T., et al. (1996). A novel human gene encoding a G-protein—coupled receptor (GPR 15) is located on chromosome 3. Genomics, 32, 462–5. Huskić, J., Kulenović, H. and Culo, F. (1996). Serum angiotensin-converting enzyme activity in patients with endemic nephropathy. Nephron, 74, 120–4. Hvala, A., Ferluga, D., Rott, T., et al. (2005). Peritubular capillary changes in Alport syndrome, diabetic glomerulopathy, Balkan endemic nephropathy and hemorrhagic fever with renal syndrome. Ultrastruct Pathol, 29, 451–9. Imamović, G., Batuman, V., Sinanović, O., et al. (2008). Microalbuminuria as a possible marker of risk of Balkan endemic nephropathy. Nephrology, 13, 616–21.

717

718

Section 4  

the patient with interstitial disease

Ivić, M. (1969). Etiology of endemic nephropathy. Lijec Vjesn, 91, 1273–81. Janković, S., Bukvić, D., Djukanović, L., et al. (2011). Time trends in Balkan endemic nephropathy incidence in the most affected region in Serbia, 1977–2009: the disease has not yet disappeared. Nephrol Dial Transplant, 26 (10), 3171–6. Jelaković, B., Karanović, S., Vuković-Lela, I., et al. (2012). Aristolactam-DNA adducts in the renal cortex: biomarker of environmental exposure to aristolochic acid. Kidney Int, 81 (6), 559–67. Kamp, H. G., Eisenbrand, G., Schlatter, B., et al. (2005). Ochratoxin A: induction of (oxidative) DNA damage, cytotoxicity and apoptosis in mammalian cell lines and primary cells. Toxicology, 206, 413–25. Karmaus, W., Dimitrov, P., Simeonov, V., et al. (2009). Offspring of parents with Balkan endemic nephropathy have higher C-reactive protein levels suggestive of inflammatory processes: a longitudinal study BMC Nephrol, 10, 10. Krcunović, Z., Novaković, I., Maksimović, N., et al. (2010). Genetic clues to the etiology of Balkan endemic nephropathy: investigating the role of ACE and AT1R polymorphisms. Arch Biol Sci, 62(4), 957–65. Ležaić, V., Marić, I., Jovanović, D., et al. (2008). Comparison of kidney size between patients with Balkan endemic nephropathy and other kidney diseases. Kidney Blood Press Res, 31, 307–12. Mally, A., Hard, G. C., and Dekant, W. (2007). Ochratoxin A as a potential etiologic factor in endemic nephropathy: lessons from toxicity studies in rats. Food Chem Toxicol, 45, 2254–60. Mantle, P., Modalca, M., Nicholls, A., et al. (2011). Comparative 1H NMR metabolomic urinalysis of people diagnosed with Balkan endemic nephropathy, and healthy subjects, in Romania and Bulgaria: a pilot study. Toxins, 3, 815–33. Miletić-Medved, M., Domijan, A. M., andPeraica, M. (2005). Recent data on endemic nephropathy and related urothelial tumors in Croatia. Wien Klin Wochenschr, 117, 604–9. Nikolov, I., Chernozemsky, I., and Idle, J. (1991). Genetic predisposition to Balkan endemic nephropathy: ability hydroxylate debrisoquine as a host factor. In M. Castegnaro, R. Plestina, G. Dirheimer, et al. (eds.) Mycotoxins, Endemic Nephropathy and Urinary Tract Tumors, pp. 289–96. Lyon: IARC. Nikolić, J., Djokic, M., Ignjatović, I., et al. (2006). Upper urothelial tumors in emigrants from Balkan endemic nephropathy areas in Serbia. Urol Int, 77, 240–4. Orem, W. H., Tatu, C. A., Lerch, H. E., et al. (2004). Identification and environmental significance of the organic compounds in water supplies associated with Balkan endemic nephropathy region in Romania. J Environ Health, 3, 49–57. Pavlovic, N. M., Varghese, Z., Persaud, J. W., et al. (1991). Partial lecithin:cholesterol acyltransferase deficiency in Balkan endemic nephropathy. Kidney Int, 40, S102–5. Pešić, I., Stefanović, V., Mueller, G., et al. (2011). Identification and validation of six proteins as marker for endemic nephropathy. J Prot, 74(10), 1994–2007. Pichler, O., Bobinac, E., Minjuš, B., and Sindik, A. (1959). O učestaloj pojavi bubrežnih oboljenja u okolici Slavonskog Broda. Lij Vjes, 81, 295. Pfohl-Leszkowicz, A. (2009). Ochratoxin A and aristolochic acid involvement in nephropathies and associated urothelial tract tumors. Arh Hig Rada Toksikol, 60, 465–83. Pichler, R. H., Francheschini, N., Young, B. A., et al. (1995). Pathogenesis of cyclosporine nephropathy: roles of angiotensin II and osteopontin. J Am Soc Nephrol, 6, 1186–96.

Puchlev, A., Dimitrov, T. S., Dinev, I., and Doichinov, D. (1979). Clinical investigations in patients with endemic nephropathy. In S. Strahinjić and V. Stefanović (eds.) Endemic (Balkan) Nephropathy. Proceeding of the 4th Symposium on Endemic (Balkan) Nephropathy, pp. 7–10. Niš: University Press. Radović, M. and Djukanović, L. (2004). Has the time come for new diagnostic criteria of Balkan endemic nephropathy. Kidney Int, 65, 1970–1. Savić, V., Čukuranović, R., Stefanović, N., et al. (2002). Damage to the kidney in Balkan endemic nephropathy: initial lesions, target structures and pathomorphogenesis. Facta Universitatis, 9, 92–4. Schiller, A., Gusbeth-Tatomir, P., Pavlović, N., et al. (2008). Balkan endemic nephropathy: a still unsolved puzzle. J Nephrol, 21, 673–80. Shihab, F. S., Yi, H., Bennett, W. H., et al. (2000). Effect of nitric oxide modulation on TGF-beta1 and matrix proteins in chronic cyclosporine nephropathy. Kidney Int, 58(3), 1174–85. Sindjić, M., Čalić-Perišić, N., Velimirović, D., et al. (1979). Renal vascular changes and their possible role in pathogenesis and morphogenesis of endemic Balkan nephropathy. In S. Strahinjić and V. Stefanović (eds.) Endemic (Balkan) Nephropathy. Proceeding of the 4th Symposium on Endemic (Balkan) Nephropathy, pp. 113–22. Niš: University Press. Slade, N., Moll, U., Brdar, B., et al. (2009). p53 mutations as fingerprints for aristolochic acid—an environmental carcinogen in endemic (Balkan) nephropathy. Mutation Res, 663, 1–6. Stefanović, V., Atanasova, S., Polenaković, M., et al. (2006). Etiology of Balkan endemic nephropathy and associated urothelial cancer. Am J Nephrol, 26, 1–11. Stefanović, V. and Cosyns J.P. (2005). Balkan nephropathy. In A. M. Davison, J. Stewart Cameron, J. -P. Gruenfeld, et al. (eds.) Oxford Textbook of Clinical Nephrology (3rd ed.), pp. 1095–101. Oxford: Oxford University Press. Stefanović, V., Čukuranović, R., Mitić-Zlatković, M., et al. (2002). Increased urinary albumin excretion in children from families with Balkan nephropathy. Pediatr Nephrol, 17, 913–16. Stoian, M., Hozoc, M., Iosipenco, M., et al. (1983). Serum antibodies to papova viruses (BK and SV 40) in subjects from the area with Balkan endemic nephropathy. Virologie, 34, 113–17. Toncheva, D. I., Atanassova, S. Y., Gergov, T. D., et al. (2002). Genetic changes in uroepithelial tumors of patients with Balkan endemic nephropathy. J Nephrol, 15, 387–93. Toncheva, D., Dimitrov, T., and Stojanova, S. (1998). Etiology of Balkan endemic nephropathy: a multifactorial disease? Eur J Epidemiol, 14, 389–94. Toncheva, D., Galabov, A. S., Laich, A., et al. (2003). Urinary neopterin concentrations in patients with Balkan endemic nephropathy. Kidney Int, 64, 1817–21. Trnačević, S., Halilbašić, A., Ferluga, D., et al. (1991). Renal function, protein excretion and pathology of Balkan endemic nephropathy. I. Renal function. Kidney Int Suppl, 34, S49–51. Uzelac-Keserović, B., Vasić, D., Ikonomovski J., et al. (2000). Isolation of a coronavirus from urinary tract tumours of endemic Balkan nephropathy patients. Nephron, 86, 93–4. Vanherweghem, J. L., Depierreux, M., Tielmans, C., et al. (1993). Rapidly progressive interstitial renal fibrosis in young women: association with slimming regimen including Chinese herbs. Lancet, 341, 387–91. Zivcić-Cosić, S., Grzetić, M., Valencić, M., et al. (2007). Urothelial cancer in patients with Endemic Balkan Nephropathy (EN) after renal transplantation. Ren Fail, 29, 861–5.

CHAPTER 91

Radiation nephropathy Lisa M. Phipps and David C. H. Harris Introduction Radiation nephropathy is defined as renal injury caused by ionizing radiation. This insult can occur when a patient receives total body irradiation (TBI) as part of the conditioning prior to bone marrow transplantation or local field irradiation for malignancy; damage results from inclusion of the kidney in the field of radiotherapy. It has also been described in patients receiving targeted radionuclide therapy with small molecules radiolabelled with high doses of beta-emitting radionuclides (Stoffel et al., 2001; Breitz, 2004). Radiation nephropathy is difficult to define and diagnose, as many patients undergoing radiation therapy are also receiving potentially nephrotoxic antineoplastic, antibacterial, antifungal, and antiviral agents, and may also be affected by tumour lysis syndrome. Radiation nephropathy may also result from radiation accidents.

Epidemiology Patients may develop symptoms and evidence of abnormal renal function from as early as 6 months to as long as 19 years after irradiation treatment (Thompson et al., 1971). The extent of damage is both volume- and dose-related (Cheng et al., 2008). Radiation nephropathy results in a progressive loss of renal function. Previous or concurrent antineoplastic chemotherapy may potentiate the effect of radiation on the kidney (Cohen et al., 1995). The kidneys are inherently radiosensitive organs and are thus major dose-limiting structures in abdominal radiotherapy fields (Yang et al., 2010). There remains debate about the primary site of injury. The endothelium of both glomerular and peritubular capillaries appears to have the highest proliferative activity and thus is postulated to be the most susceptible to radiation damage (Nadasdy et al., 1994). However, renal tubular epithelial cells seem to be more radiosensitive than epithelial cells from other tissues (Krochak and Baker, 1986). The exact threshold of radiation beyond which radiation nephropathy occurs is yet to be determined. Fractionation decreases the risk. In rodent models, a single dose of irradiation that resulted in radiation nephropathy did not cause the disease when fractionated into multiple, separated doses (Stewart et  al., 1994). However, once damage has occurred, the pathological findings are the same, regardless of how the radiation was delivered (Lawton et al., 1991). The growing kidney appears to be more sensitive to irradiation than the adult kidney (Mitus et al., 1969). Doses of 10 Gy or more involving the abdomen are associated with a > 5% decrease in the size of the primarily irradiated kidney

within 1 year of exposure (Yang et al., 2010). The tolerance dose associated with a 5% risk of renal dysfunction at 5 years after single, whole-kidney irradiation was reported to be 23 Gy, and that associated with a 50% risk at 5 years was found to be 28 Gy (Keane et al., 1976). The tolerance dose after TBI is 14 Gy in adults (Tarbell et al., 1988). It is recommended that renal shielding be implemented for doses above this range (Henk et al., 1967). It is generally accepted that the kidney has some capacity to repair sublethal radiation injury during fractionated irradiation (Dewit et al., 1990). This has been demonstrated experimentally in Wistar rats (Yildiz et al., 2000). Previous renal irradiation results in a loss of re-irradiation tolerance. This was demonstrated in renal-irradiated mice administered a range of second (single) doses after either a 2- or 26-week period. Doses required to give a 50% incidence of damage (RD50) were significantly lower than those causing RD50 following initial treatment alone (1.4 Gy compared to 3.3 Gy) (Stewart et al., 1994). With regards to environmental radiation, such as atomic bomb or nuclear power plant explosions, the doses that people close to the event are exposed to will likely result in death from bone marrow or gastrointestinal failure before renal impairment would have time to develop (Cohen, 2000). Long-term, low-dose environmental exposure (levels of radiation above background levels, yet below that which could induce acute effects, usually associated with cell death) may result in chronic radiation nephropathy. Evidence of histopathological changes consistent with radiation nephropathy (glomerular and tubular degeneration, desquamation, regeneration, and nuclear pyknosis) has been detected in Ukrainian patients living in areas radiocontaminated by the Chernobyl nuclear reactor accident (Romanenko et al., 2001). An increased incidence of pre-eclampsia and renal impairment in pregnant women has also been associated with environmental radiation exposure. A  fivefold increase in pre-eclampsia was demonstrated in women from Belarus in the 4 years after the Chernobyl reactor accident (Petrova et al., 1997) although other factors could explain this.

Clinical features The clinical presentation of radiation nephropathy falls into four broad categories, based on clinical features and timing of onset (Krochak and Baker, 1986) (Table 91.1).

Acute radiation nephropathy This presents with an abrupt onset of renal dysfunction, 6–12  months after exposure. The reason that nephropathy does not manifest until at least 6  months after radiotherapy is due in part to mitotic rate of renal tubule cells, as radiation-induced cell

720

Section 4  

the patient with interstitial disease

Table 91.1  Clinical syndromes associated with radiation nephropathy Syndrome

Onset

Clinical features

Acute radiation nephritis

6–12 months

Abrupt onset, raised serum creatinine, anaemia, hypertension, proteinuria, microscopic haematuria

Chronic radiation nephritis

1–19 years

Insidious onset, features of chronic renal failure

Hypertension

From 18 months

May be benign or malignant, with retinopathy, congestive heart failure, and encephalopathy

From 18 months

Intermittent proteinuria, normal serum creatinine

Asymptomatic proteinuria

death is expressed at the time of the next mitosis (Soranson and Denekamp, 1986). Patients may present with symptoms of advanced renal failure, with lethargy, oedema, headaches, and shortness of breath. Hypertension is an almost universal finding. This is thought to be due in part to the systemic effect of radiation, resulting in increased peripheral resistance, which induces an increase in systemic blood pressure. Volume expansion and renovascular disease may also play a role. Renal artery stenosis as a complication of abdominal irradiation has been well described (Dean and Abels, 1945; Staab et al., 1976; Gerlock et al., 1977; McGill et al., 1979). Patients have increased serum creatinine, proteinuria, and microscopic haematuria. There may be an associated significant anaemia, contributing to the patient’s symptomatology (Luxton and Kunkler, 1964). The anaemia is usually hypochromic microcytic and can be due in part to haemolysis, although a positive response to erythropoietin therapy in patients with radiation nephropathy suggests that erythropoietin deficiency plays a substantial role (Cohen, 2000). Haemolytic anaemia is due to thrombotic microangiopathy (TMA) (Breitz, 2004). The presentation of TMA may range from mild and subclinical, with fragmented cells on blood smear and low platelet count, to a fulminant presentation with thrombotic thrombocytopenic purpura or haemolytic uraemic syndrome. How much TMA is attributable specifically to radiation exposure is difficult to ascertain, as affected patients are often exposed to other agents known to cause this syndrome (e.g. chemotherapy, immunosuppressants). The prognosis of acute radiation nephropathy has been linked to the severity of the hypertension (Krochak and Baker, 1986). Patients who survive the acute phase are left with varying degrees of chronic renal impairment. The renal dysfunction is not always progressive.

Chronic radiation nephropathy This may occur as a sequela of acute radiation nephropathy, or present more indolently. Cases have been reported to present as late as 19 years after exposure to radiotherapy. Patients present with features of chronic renal impairment. Imaging reveals small atrophic kidneys. Chronic radiation nephropathy has also been described in patients exposed to long-term, low-dose environmental radiation (Romanenko et al., 2001).

Benign or malignant hypertension Hypertension is a prominent feature of both acute and chronic radiation nephropathy, but may occur in isolation (Cohen, 2000). Benign hypertension can occur from 18 months to years after exposure. Benign hypertension may develop into malignant hypertension (see Chapter 216) over many years. Malignant hypertension occurs within the same time frame as benign hypertension. Patients may present with symptoms of retinopathy, congestive heart failure, pleural and pericardial effusions, encephalopathy, and seizures (Luxton, 1961; Tarbell et al., 1988). Renal impairment is not usually a prominent feature.

Asymptomatic proteinuria Proteinuria may occur in the absence of abnormal renal function. It may be intermittent. It has not been demonstrated to progress to renal impairment (Krochak and Baker, 1986; Breitz, 2004).

Investigations Serum biochemistry Serum urea and creatinine are raised consistent with the fall in GFR. There is a tendency towards hyperkalaemia, due to suppression of the renin–aldosterone axis (Cohen, 2000). Lactic dehydrogenase may be raised if haemolysis is present.

Haematology The degree of anaemia is disproportionately worse than would usually be expected for the degree of renal impairment. Blood film may reveal evidence of haemolysis. Thrombocytopenia may be noted, in the presence of underlying TMA.

Urine Urinalysis may reveal proteinuria, microscopic haematuria, and occasional pyuria. Proteinuria is generally non-nephrotic, up to 2 g/day (Cohen, 2000). Urinary excretion of β2 microglobulin may be increased, reflecting underlying tubular cell damage (Dewit et al., 1990).

Nuclear medicine scans Technetium-labelled diethylene triamine pentaacetic acid (Tc-DTPA) can detect a fall in GFR as early as 6 months post irradiation in affected individuals, even in the presence of a normal serum creatinine (Dewit et  al., 1990). Dimercaptosuccinic acid (DMSA) scintigraphy can demonstrate a decline in tubular function from as early as 6 months after exposure (Dewit et al., 1990).

Renal biopsy There are limited biopsy studies of humans early in the disease process. As a result, the majority of specimens show changes of end-stage kidney damage, in which the initial injury can no longer be recognized (Fajardo et al., 1976). There are morphological similarities between radiation nephropathy and haemolytic-uraemic syndrome (see Chapter 174). There are no pathognomic changes, and diagnosis is suspected based on clinical features, consistent history, and the following pathological findings:

Glomeruli There may be aneurysmal dilatation of capillary loops, obliteration of tufts, and segmental or total glomerulosclerosis (Guinan et al.,

Chapter 91 

1988; Borg et al., 2002). There is arteriolar intimal thickening and endothelial cell dropout. There may be evidence of obstruction of the glomerular capillaries by platelets and fibrin (Fajardo et  al., 1976). Mesangial hyperplasia is a prominent feature (Tarbell et al., 1988). There is often amorphous material separating glomerular endothelium from the glomerular basement membrane (Figs 91.1 and 91.2). On electron microscopy this material appears to be basement membrane-like material deposited on the endothelial aspect of the basement membrane. Tubules and interstitium: atrophic tubules with hyaline casts are prominent, as is interstitial fibrosis (Kapur et  al., 1977; Lawton et al., 1991). Vessels: segmental arteries may display endothelial proliferation, with evidence of complete occlusion. Smaller arteries and arterioles demonstrate evidence of fibrinoid necrosis. The electron microscopy changes consist of folded thickened glomerular basement membranes, with areas of attenuation. Endothelial and mesangial cells reveal hypertrophic changes, with an increase both in cell size and cytoplasmic organelles. The nuclei are lobulated and show margination of chromatin. These changes are thought to be a direct effect of radiation damage, as they are not seen in biopsies of patients with malignant hypertension (Kapur et al., 1977). (See Fig. 91.1.)

radiation nephropathy

Pathophysiology and pathology Glomerular damage Glomerular endothelial injury was visible at 3 weeks post irradiation in a porcine model (Robbins et al., 1993). Endothelial disruption and leucocyte adherence is followed by subendothelial expansion (Jaenke et al., 1977). There is a concurrent increase in platelet adhesion (Verheij et al., 1994) and subsequent vascular and glomerular microthrombi. Glomerular scarring ensues. Similar changes occur in the peritubular capillaries (Fig. 91.1). Experimental models have demonstrated significant increases in glomerular von Willebrand factor along with decreased levels of ADPase (a potent inhibitor of platelet adhesion) after irradiation, with an associated increase in deposition of fibrinogen, thought to contribute to microthrombi formation (Stewart et al., 2001). The initial endothelial disruption results in filtrate extruding from the capillaries, with protein and other high-molecular-weight blood components escaping into the extravascular space. This protein gradually becomes insoluble, resulting in impaired diffusion of oxygen and other essential metabolites (Breitz, 2004).

Tubular damage Tubular epithelial cells appear to be more radiosensitive than epithelial cells from other tissues (Emery et al., 1970). Radiobiological data reveal that tubular epithelial cells have limited capacity to repair lethal and sublethal damage (Deschavanne et al., 1980), particularly if exposed to unfractionated radiation. The tubular compartment may be further damaged indirectly by ischaemia secondary to radiation injury of the renal microvasculature (Krochak and Baker, 1986). Constriction of the tubular lumen at the origin of the proximal tubule (the glomerulotubular neck) has been demonstrated in porcine and rat radiation nephropathy (Cohen et al., 2000).

Hypertension

Fig. 91.1  Glomerulus from a patient with radiation nephropathy, showing mesangiolysis, mesangial expansion, thickening, and splitting (arrows) of glomerular capillary wall.

Fig. 91.2  Low power view from a patient with radiation nephropathy demonstrating glomerular thrombosis (arrow) and sclerosis (arrowhead) with tubular degeneration and fibrosis (asterisk).

In large field or total body irradiation, increases in peripheral resistance may also induce a compensatory increase in systemic blood pressure (Krochak and Baker, 1986). In experimental rats hypertension occurred after TBI, irrespective of renal shielding, however the degree of hypertension was less in the renal-shielded group (Wachholz and Casarett, 1970).

Treatment and outcome Renal protection in the form of blocking or shielding devices to decrease the dose of irradiation to the kidneys has been shown to decrease the incidence of radiation nephropathy. Partial renal shielding during TBI has been demonstrated to reduce the incidence of radiation nephropathy from 29% with no shielding to 14% with 15% renal shielding, and to 0% with 30% renal shielding (Lawton et  al., 1992). The Wilms Tumour Study group from the United Kingdom recommends renal shielding during radiotherapy after unilateral nephrectomy for Wilms tumour (Taylor, 1997). As with all renal diseases, hypertension needs to be controlled. In experimental models, the use of angiotensin-converting enzyme inhibitors (ACEIs) or angiotensin II receptor antagonists (ARBs) appears to have additional benefits above those of other classes of antihypertensive agents (Cohen, 2000). The mechanistic basis for the increased efficacy of these agents is uncertain but may include their anti-inflammatory, antifibrogenic, and antimitogenic activity

721

722

Section 4  

the patient with interstitial disease

(Molteni et  al., 2001; Cohen et  al., 2002; Suzuki et  al., 2003). It has been hypothesized that they act to limit the consequences of endothelial cell damage (Moulder et al., 1998b; Molteni et al., 2001) by prevention of radiation-induced proliferation (Moulder et  al., 2002a). The initiation of ACEI or ARB therapy before the clinical onset of experimental radiation nephropathy has been demonstrated to provide additional benefits in slowing the progression of the renal disease (Cohen et al., 1996). Whether a similar benefit of early therapy occurs in human subjects is yet to be determined. Aspirin may have a preventative effect by inhibiting the increased platelet adhesion (Sinzinger and Firbas, 1985). There is experimental evidence for a beneficial effect of dexamethasone in rats (Geraci et al., 1992); however, such a benefit has not been demonstrated in humans (Baldwin and Hagstrom, 1962). Despite evidence of protection in other radiation-induced injuries (such as that of leucocytes) there has been no demonstrated benefit from the use of antioxidants such as vitamin A (Balabanli et al., 2006) and retinoic acid (Moulder et al., 1998a, 2002b) in the treatment of radiation nephropathy (Cohen et al., 2009).

References Balabanli, B., Turkozkan, N., Akmansu, M., et al. (2006). Role of free radicals on mechanism of radiation nephropathy. Mol Cell Biochem, 293, 183–6. Baldwin, J. N. and Hagstrom, J. W. (1962). Acute radiation nephritis. Calif Med, 97, 359–62. Borg, M., Hughes, T., Horvath, N., et al. (2002). Renal toxicity after total body irradiation. Int J Radiat Oncol Biol Phys, 54, 1165–73. Breitz, H. (2004). Clinical aspects of radiation nephropathy. Cancer Biother Radiopharmaceut, 19, 359–62. Cheng, J. C., Schultheiss, T. E., and Wong, J. Y. (2008). Impact of drug therapy, radiation dose, and dose rate on renal toxicity following bone marrow transplantation. Int J Radiat Oncol Biol Phys, 71, 1436–43. Cohen, E. P. (2000). Radiation nephropathy after bone marrow transplantation. Kidney Int, 58, 903–18. Cohen, E. P., Fish, B. L., Irving, A. A., et al. (2009). Radiation nephropathy is not mitigated by antagonists of oxidative stress. Radiat Res, 172, 260–4. Cohen, E. P., Fish, B. L., and Moulder, J. E. (2002). The renin-angiotensin system in experimental radiation nephropathy. J Lab Clin Med, 139, 251–7. Cohen, E. P., Lawton, C. A., and Moulder, J. E. (1995). Bone marrow transplant nephropathy: radiation nephritis revisited. Nephron, 70, 217–22. Cohen, E. P., Molteni, A., Hill, P., et al. (1996). Captopril preserves function and ultrastructure in experimental radiation nephropathy. Lab Invest, 75, 349–60. Cohen, E. P., Regner, K., Fish, B. L., et al. (2000). Stenotic glomerulotubular necks in radiation nephropathy. J Pathol, 190, 484–8. Dean, A. L. and Abels, J. C. (1945). Study by the newer renal function tests of an unusual case of hypertension following irradiation of one kidney and the relief of the patient by nephrectomy. Trans Am Assoc Genitourin Surg, 37, 239–44. Deschavanne, P. J., Guichard, M., and Malaise, E. P. (1980). Radiosensitivity of mouse kidney cells determined with an in vitro colony method. Int J Radiat Oncol Biol Phys, 6, 1551–7. Dewit, L., Anninga, J. K., Hoefnagel, C. A., et al. (1990). Radiation injury in the human kidney: a prospective analysis using specific scintigraphic and biochemical endpoints. Int J Radiat Oncol Biol Phys, 19, 977–83. Emery, E. W., Denekamp, J., Ball, M. M., et al. (1970). Survival of mouse skin epithelial cells following single and divided doses of x-rays. Radiat Res, 41, 450–66. Fajardo, L. F., Brown, J. M., and Glatstein, E. (1976). Glomerular and juxta-glomerular lesions in radiation nephropathy. Radiat Res, 68, 177–83.

Geraci, J. P., Taylor, D. A., Mariano, M. S., et al. (1992). Effects of dexamethasone on the development of radiation nephropathy in the rat. Radiat Res, 131, 186–91. Gerlock, A. J., Jr., Goncharenko, V. A., and Ekelund, L. (1977). Radiation-induced stenosis of the renal artery causing hypertension: case report. J Urol, 118, 1064–5. Guinan, E. C., Tarbell, N. J., Niemeyer, C. M., et al. (1988). Intravascular hemolysis and renal insufficiency after bone marrow transplantation. Blood, 72, 451–5. Henk, J. M., Cottrall, M. F., and Taylor, D. M. (1967). Radiation dosimetry of the 131-I-hippuran renogram. Br J Radiol, 40, 327–34. Jaenke, R. S., Phemister, R. D., Angleton, G. M., et al. (1977). Characterization of renal damage following perinatal gamma radiation in the beagle. Radiat Res, 72, 277–90. Kapur, S., Chandra, R., and Antonovych, T. (1977). Acute radiation nephritis. Light and electron microscopic observations. Arch Pathol Lab Med, 101, 469–73. Keane, W. F., Crosson, J. T., Staley, N. A., et al. (1976). Radiation-induced renal disease. A clinicopathologic study. Am J Med, 60, 127–37. Krochak, R. J. and Baker, D. G. (1986). Radiation nephritis. Clinical manifestations and pathophysiologic mechanisms. Urology, 27, 389–93. Lawton, C. A., Barber-Derus, S. W., Murray, K. J., et al. (1992). Influence of renal shielding on the incidence of late renal dysfunction associated with T-lymphocyte deplete bone marrow transplantation in adult patients. Int J Radiat Oncol Biol Phys, 23, 681–6. Lawton, C. A., Cohen, E. P., Barber-Derus, S. W., et al. (1991). Late renal dysfunction in adult survivors of bone marrow transplantation. Cancer, 67, 2795–800. Luxton, R. W. (1961). Radiation nephritis. A long-term study of 54 patients. Lancet, 2, 1221–4. Luxton, R. W. and Kunkler, P. B. (1964). Radiation nephritis. Acta Radiol Ther Phys Biol, 2, 169–78. McGill, C. W., Holder, T. M., Smith, T. H., et al. (1979). Postradiation renovascular hypertension. J Pediatr Surg, 14, 831–3. Mitus, A., Tefft, M., and Fellers, F. X. (1969). Long-term follow-up of renal functions of 108 children who underwent nephrectomy for malignant disease. Pediatrics, 44, 912–21. Molteni, A., Moulder, J. E., Cohen, E. P., et al. (2001). Prevention of radiation-induced nephropathy and fibrosis in a model of bone marrow transplant by an angiotensin II receptor blocker. Exp Biol Med, 226, 1016–23. Moulder, J. E., Fish, B. L., and Cohen, E. P. (1998a). Brief pharmacological intervention in experimental radiation nephropathy. Radiat Res, 150, 535–41. Moulder, J. E., Fish, B. L., and Cohen, E. P. (1998b). Radiation nephropathy is treatable with an angiotensin converting enzyme inhibitor or an angiotensin II type-1 (AT1) receptor antagonist. Radiother Oncol, 46, 307–15. Moulder, J. E., Fish, B. L., Regner, K. R., et al. (2002a). Angiotensin II blockade reduces radiation-induced proliferation in experimental radiation nephropathy. Radiat Res, 157, 393–401. Moulder, J. E., Fish, B. L., Regner, K. R., et al. (2002b). Retinoic acid exacerbates experimental radiation nephropathy. Radiat Res, 157, 199–203. Nadasdy, T., Laszik, Z., Blick, K. E., et al. (1994). Proliferative activity of intrinsic cell populations in the normal human kidney. J Am Soc Nephrol, 4, 2032–9. Petrova, A., Gnedko, T., Maistrova, I., et al. (1997). Morbidity in a large cohort study of children born to mothers exposed to radiation from Chernobyl. Stem Cells, 15 Suppl 2, 141–50. Robbins, M. E., Jaenke, R. S., Bywaters, T., et al. (1993). Sequential evaluation of radiation-induced glomerular ultrastructural changes in the pig kidney. Radiat Res, 135, 351–64. Romanenko, A., Morell-Quadreny, L., Nepomnyaschy, V., et al. (2001). Radiation sclerosing proliferative atypical nephropathy of peritumoral tissue of renal-cell carcinomas after the Chernobyl accident in Ukraine. Virchows Arch, 438, 146–53.

Chapter 91 

Sinzinger, H. and Firbas, W. (1985). Irradiation depresses prostacyclin generation upon stimulation with the platelet-derived growth factor. Br J Radiol, 58, 1023–6. Soranson, J. and Denekamp, J. (1986). Precipitation of latent renal radiation injury by unilateral nephrectomy. Br J Cancer Suppl, 7, 268–72. Staab, G. E., Tegtmeyer, C. J., and Constable, W. C. (1976). Radiation-induced renovascular hypertension. AJR. Am J Roentgenol, 126, 634–7. Stewart, F. A., Oussoren, Y., Van Tinteren, H., et al. (1994). Loss of reirradiation tolerance in the kidney with increasing time after single or fractionated partial tolerance doses. Int J Radiat Biol, 66, 169–79. Stewart, F. A., Te Poele, J. A., Van der Wal, A. F., et al. (2001). Radiation nephropathy—the link between functional damage and vascular mediated inflammatory and thrombotic changes. Acta Oncol, 40, 952–7. Stoffel, M. P., Pollok, M., Fries, J., et al. (2001). Radiation nephropathy after radiotherapy in metastatic medullary thyroid carcinoma. Nephrol Dial Transplant, 16, 1082–3. Suzuki, Y., Ruiz-Ortega, M., Lorenzo, O., et al. (2003). Inflammation and angiotensin II. Int J Biochem Cell Biol, 35, 881–900.

radiation nephropathy

Tarbell, N. J., Guinan, E. C., Niemeyer, C., et al. (1988). Late onset of renal dysfunction in survivors of bone marrow transplantation. Int J Radiat Oncol Biol Phys, 15, 99–104. Taylor, R. E. (1997). Morbidity from abdominal radiotherapy in the First United Kingdom Children’s Cancer Study Group Wilms’ Tumour Study. United Kingdom Children’s Cancer Study Group. Clin Oncol, 9, 381–4. Thompson, P. L., MacKay, I. R., Robson, G. S., et al. (1971). Late radiation nephritis after gastric x-irradiation for peptic ulcer. QJM, 40, 145–57. Verheij, M., Dewit, L. G., Boomgaard, M. N., et al. (1994). Ionizing radiation enhances platelet adhesion to the extracellular matrix of human endothelial cells by an increase in the release of von Willebrand factor. Radiat Res, 137, 202–7. Wachholz, B. W., and Casarett, G. W. (1970). Radiation hypertension and nephrosclerosis. Radiat Res, 41, 39–56. Yang, G. Y., May, K. S., Iyer, R. V., et al. (2010). Renal atrophy secondary to chemoradiotherapy of abdominal malignancies. Int J Radiat Oncol Biol Phys, 78, 539–46. Yildiz, F., Atahan, I. I., and Yildiz, O. (2000). Radiation nephropathy in rats and its modification by the angiotensin converting enzyme inhibitor enalapril. Radiat Med, 18, 153–9.

723

CHAPTER 92

Urate nephropathy Duk-Hee Kang and Mehmet Kanbay Introduction

Urate (gout) nephropathy

Gout is recorded from ancient times, but hyperuricaemia seems to be increasing in incidence now. Its association with nephropathy dates back to lead poisoning (Chapter 88), but it is now clear that there are genetic (and very likely other) explanations for this coincidence (Chapter 316), it is not simply that urate levels or crystals are necessarily nephrotoxic.

Gout is a disorder of purine metabolism, characterized by hyperuricaemia and urate crystal deposition within and around the joints (Richette and Bardin, 2010). The most important single risk factor for developing gout is the raised serum uric acid level. The recognition of increased comorbidity burden in patients with gout rendered it as a systemic disorder rather than simply a musculoskeletal disease. Older studies reported that 25% of gout patients had proteinuria, 50% had renal insufficiency, and 10% to 25% developed end-stage renal disease (Brochner-Mortensen, 1958; Talbott and Terplan, 1960). More recently, in a large retrospective cohort study, Primatesta et al. (2011) found that, out of 177,637 gout patients, more than half (58.1%) had one or more comorbidities, including hypertension (36.1%), dyslipidaemia (27.0%), diabetes (15.1%), and ischaemic heart disease (10.2%), as the most common ones. The prevalence of CKD in this population was 3.2%. However, in other studies, the prevalence of CKD among patients with gout was much higher: 39% (Fuldeore et al., 2011) and 30% (Wu et al., 2012). These discrepant findings may be due to differences in patient selection criteria and in the definitions of CKD. In autopsy studies, renal histologic abnormalities have been described in as many as 75–99% of patients with gout. Gout nephropathy (also known as chronic uric acid nephropathy or urate nephropathy) is a form of CKD induced by the deposition of monosodium urate crystals in the distal collecting ducts and the medullary interstitium, associated with a secondary inflammatory reaction. Other histologic findings include arteriolosclerosis, glomerulosclerosis, and tubulointerstitial fibrosis. Urate crystal deposition has previously been considered as the mediator of renal injury (Greenbaum et al., 1961); however, these deposits were found to be focal rather than diffuse (Linnane et al., 1981). The diagnosis of gout nephropathy may often be problematic, as histologic findings may be indistinguishable from benign nephrosclerosis or from age-associated renal changes. In patients with gout, the renal blood flow was found to be disproportionately low in comparison with the glomerular filtration rate (Berger et al., 1975; Yu et al., 1979; Yu and Berger, 1982). The fractional excretion of uric acid is usually < 10%. Proteinuria occurs in a minority of cases, and it is usually mild to moderate. The urinary sediment also shows minor or no changes. On the other hand, hypertension is very common, occurring in 50–60% of patients, and its prevalence increases as the renal function gets worse. Serum creatinine is usually normal or only mildly increased (Berger and Yu, 1975; Yu and Berger, 1982; Yu et al., 1979). An excessively high serum uric acid in relation to serum creatinine

Renal damage induced by hyperuricaemia In urate oxidase knockout mice, uric acid rapidly increases in serum, and precipitates in the renal tubules, resulting in acute kidney injury (AKI) (Wu et al., 1994). This experimental model may share some features with the human tumour lysis syndrome; however, it may not be appropriate for understanding the renal effects of protracted milder hyperuricaemia. Recent data have shown that moderate and persistent uric acid elevations may also be detrimental to the kidney. A  rat model of hyperuricaemia, using uricase inhibitor oxonic acid, enabled studies of renal damage associated with this condition. These hyperuricaemic rats showed preglomerular arterial disease, renal inflammation, and hypertension, via activation of the renin–angiotensin system (Kang et al., 2002; Mazzali et al., 2002; Nakagawa et al., 2003). It has been speculated that the resulting thickening and macrophage infiltration of the afferent arteriole walls may induce postglomerular ischaemia. The reduction in vascular lumen could also provide a stimulus for the increase in renin expression, which contributes to the development of severe arterial hypertension (Mazzali et  al., 2001, 2003; Kang et  al., 2002). Furthermore, there is evidence that the arteriolopathy also leads to ineffective autoregulation and increased transmission of systemic pressures to the glomerulus (Sanchez-Lozada et al., 2002, 2005), which can worsen the renal damage. In addition, uric acid turned out to be pro-oxidative under certain circumstances (Bagnati et al., 1999). Several epidemiological studies have found that serum uric acid is an independent risk factor for the development and progression of chronic kidney disease (CKD). In one study, hyperuricaemia was associated with a 10.8-fold higher risk in women and a 3.8-fold higher risk in men for the development of CKD, compared to subjects with normal uric acid levels (Iseki et al., 2004). This increased risk was independent of age, body mass index, systolic blood pressure, smoking, and proteinuria. An elevated serum uric acid was also associated with a significantly increased risk of CKD in a study on > 49,000 male individuals (Tomita et al., 2000).

Chapter 92 

(e.g. > 9 mg/dL vs < 1.5 mg/dL, > 10 mg/dL vs 1.5–2.0 mg/dL, or > 12 mg/dL vs > 2.0 mg/dL) may evoke the diagnosis of gout nephropathy. Urate (gout) nephropathy seems to be related to other factors than hyperuricaemia alone. This is suggested by the focal nature of renal uric acid crystal deposition, the inconsistent response to uric acid-lowering therapy, and the common association of gout with other risk factors for CKD.

Management of urate nephropathy Some studies suggested that reduction of serum uric acid could improve gout nephropathy (Briney et al., 1975; Patial and Sehgal, 1992), whereas others did not confirm such benefits (Rosenfeld, 1974). However, it seems reasonable to try to lower serum uric acid in subjects with hyperuricaemia, especially when it is markedly elevated (> 10 mg/dL). The consumption of uric acid-raising foods should be reduced, including those with high purine content, fructose, and alcohol drinks. In patients with CKD, uricosuric agents are known to be ineffective (Perez-Ruiz et al., 1998). On the other hand, xanthine oxidase inhibitor allopurinol is a potent uric acid-lowering agent. However, severe hypersensitivity reactions to allopurinol have been reported, including Stevens–Johnson syndrome, with fever, liver dysfunction, and AKI (Anderson and Adams, 2002). Most individuals with allopurinol hypersensitivity were found to be human leucocyte antigen (HLA)-B58 positive; therefore, screening for HLA-B58 may be considered before allopurinol prescription, in order to avoid such potentially life-threatening events (Jung et al., 2011). Furthermore, high doses of allopurinol may cause xanthine or allopurinol crystal intratubular deposition, leading to a worsening of the renal disease. An alternative to allopurinol is febuxostat, a non-purine-analogue inhibitor of xanthine oxidase. Dose adjustment of this drug is not required in patients with impaired renal function and no cases of hypersensitivity syndrome have been reported with its use (Becker et al., 2005).

References Anderson, B. E. and Adams, D. R. (2002). Allopurinol hypersensitivity syndrome. J Drugs Dermatol, 1(1), 60–2. Bagnati, M., Perugini, C., Cau, C., et al. (1999). When and why a water-soluble antioxidant becomes pro-oxidant during copper-induced low-density lipoprotein oxidation: a study using uric acid. Biochem J, 340(Pt 1), 143–52. Becker, M. A., Schumacher, H. R., Jr., Wortmann, R. L., et al. (2005). Febuxostat, a novel nonpurine selective inhibitor of xanthine oxidase: a twenty-eight-day, multicenter, phase II, randomized, double-blind, placebo-controlled, dose-response clinical trial examining safety and efficacy in patients with gout. Arthritis Rheum, 52(3), 916–23. Berger, L. and Yu, T. F. (1975). Renal function in gout. IV. An analysis of 524 gouty subjects including long-term follow-up studies. Am J Med, 59(5), 605–13. Briney, W. G., Ogden, D., Bartholomew, B., et al. (1975). The influence of allopurinol on renal function in gout. Arthritis Rheum, 18(6 Suppl), 877–81. Brochner-Mortensen, K. (1958). Gout. Ann Rheum Dis, 17(1), 1–8. Fuldeore, M. J., Riedel, A. A., Zarotsky, V., et al. (2011). Chronic kidney disease in gout in a managed care setting. BMC Nephrol, 12(1), 36.

urate nephropathy

Greenbaum, D., Ross, J. H., and Steinberg, V. L. (1961). Renal biopsy in gout. Br Med J, 1(5238), 1502–4. Iseki, K., Ikemiya, Y., Inoue, T., et al. (2004). Significance of hyperuricemia as a risk factor for developing ESRD in a screened cohort. Am J Kidney Dis, 44(4), 642–50. Jung, J. W., Song, W. J., Kim, Y. S., et al. (2011). HLA-B58 can help the clinical decision on starting allopurinol in patients with chronic renal insufficiency. Nephrol Dial Transplant, 26(11), 3567–72. Kang, D. H., Nakagawa, T., Feng, L., et al. (2002). A role for uric acid in the progression of renal disease. J Am Soc Nephrol, 13(12), 2888–97. Linnane, J. W., Burry, A. F., and Emmerson, B. T. (1981). Urate deposits in the renal medulla. Prevalence and associations. Nephron, 29(5–6), 216–22. Mazzali, M., Hughes, J., Kim, Y. G., et al. (2001). Elevated uric acid increases blood pressure in the rat by a novel crystal-independent mechanism. Hypertension, 38(5), 1101–6. Mazzali, M., Jefferson, J. A., Ni, Z., et al. (2003). Microvascular and tubulointerstitial injury associated with chronic hypoxia-induced hypertension. Kidney Int, 63(6), 2088–93. Mazzali, M., Kanellis, J., Han, L., et al. (2002). Hyperuricemia induces a primary renal arteriolopathy in rats by a blood pressure-independent mechanism. Am J Physiol Renal Physiol, 282(6), F991–7. Nakagawa, T., Mazzali, M., Kang, D. H., et al. (2003). Hyperuricemia causes glomerular hypertrophy in the rat. Am J Nephrol, 23(1), 2–7. Patial, R. K. and Sehgal, V. K. (1992). Non-oliguric acute renal failure in gout. Indian J Med Sci, 46(7), 201–4. Perez-Ruiz, F., Alonso-Ruiz, A., Calabozo, M., et al. (1998). Efficacy of allopurinol and benzbromarone for the control of hyperuricemia. A pathogenic approach to the treatment of primary chronic gout. Ann Rheum Dis, 57(9), 545–9. Primatesta, P., Plana, E., and Rothenbacher, D. (2011). Gout treatment and comorbidities: a retrospective cohort study in a large US managed care population. BMC Musculoskelet Disord, 12, 103. Richette, P. and Bardin, T. (2010). Gout. Lancet, 375(9711), 318–28. Rosenfeld, J. B. (1974). Effect of long-term allopurinol administration on serial GFR in normotensive and hypertensive hyperuricemic subjects. Adv Exp Med Biol, 41, 581–96. Sanchez-Lozada, L. G., Tapia, E., Avila-Casado, C., et al. (2002). Mild hyperuricemia induces glomerular hypertension in normal rats. Am J Physiol Renal Physiol, 283(5), F1105–10. Sanchez-Lozada, L. G., Tapia, E., Santamaría, J., et al. (2005). Mild hyperuricemia induces vasoconstriction and maintains glomerular hypertension in normal and remnant kidney rats. Kidney Int, 67(1), 237–47. Talbott, J. H. and Terplan, K. L. (1960). The kidney in gout. Medicine (Baltimore), 39, 405–67. Tomita, M., Mizuno, S., Yamanaka, H., et al. (2000). Does hyperuricemia affect mortality? A prospective cohort study of Japanese male workers. J Epidemiol, 10(6), 403–9. Wu, E. Q., Forsythe, A., Guérin, A., et al. (2012). Comorbidity burden, healthcare resource utilization, and costs in chronic gout patients refractory to conventional urate-lowering therapy. Am J Ther, 19(6), e157–66. Wu, X., Wakamiya, M., Vaishnav, S., et al. (1994). Hyperuricemia and urate nephropathy in urate oxidase-deficient mice. Proc Natl Acad Sci U S A, 91(2), 742–6. Yu, T. F. and Berger, L. (1982). Impaired renal function gout: its association with hypertensive vascular disease and intrinsic renal disease. Am J Med, 72(1), 95–100. Yu, T. F., Berger, L., Dorph, D. J., et al. (1979). Renal function in gout. V. Factors influencing the renal hemodynamics. Am J Med, 67(5), 766–71.

725

CHAPTER 93

Immune-mediated tubulointerstitial nephritis Liviu Segall and Adrian Covic Introduction Immune-mediated tubulointerstitial nephropathies (TINs) are generally encountered in the context of systemic or extrarenal autoimmune diseases, such as sarcoidosis, Sjögren’s syndrome, systemic lupus erythematosus, inflammatory bowel disease, TINU syndrome, and IgG4-related disease. The pathogenesis of these TINs is complex and more or less unclear; it usually involves leukocyte activation, autoantibodies, immune complex deposition, complement activation, and release of inflammatory cytokines and growth factors. Immune-mediated glomerulonephritis or other complications, such as renal stone disease, may sometimes be associated with TIN and contribute to renal damage. Tubulointerstitial inflammation most commonly has a chronic pattern, although acute forms of TIN may also occur. Furthermore, inflammation may be granulomatous (as in sarcoidosis or Crohn’s disease) or non-granulomatous. Immunofluorescence staining can sometimes reveal immune complex deposits or anti-tubular basement membrane (TBM) autoantibodies. Systemic immunosuppressive therapies are almost always required to prevent progression to irreversible interstitial fibrosis, tubular atrophy, and end-stage renal disease.

Sarcoidosis Sarcoidosis is a systemic disorder of unknown aetiology that usually occurs in young people, 20–40 years old, and is characterized by the presence of non-caseating granulomas in various organs, particularly the respiratory tract (Nunes et al., 2007). It has a slight female predominance (1.5:1) (James et al., 1984; Baughman et al., 2001) and it affects African Americans 3–20 times more often than white people (Iannuzzi et al., 2007). It is additionally discussed in Chapter 156. The pathogenesis of sarcoidosis is unclear, but it is postulated that exposure to unknown environmental factors triggers an inflammatory response involving macrophages and CD4+ helper T-cell activation, which participate in the formation of granulomas (Newman et al., 1997). A critical role in the development of granulomas is played by cytokines like interferon gamma (IFN-γ), tumour necrosis factor alpha (TNF-α), interleukin 12 (IL-12), and interleukin 18 (IL-18) (Roach et al., 2002; Baughman et al., 2005; Semenzato et al., 2005). Infectious agents such as Propionibacterium acnae (Ishige et  al., 1999) and a genetically determined abnormal response to inhaled antigens (Valentonyte et al., 2005) have been suggested, among other factors, as possible contributors to the origin of sarcoidosis.

Many patients with sarcoidosis are initially asymptomatic and diagnosis is evoked by incidental findings of bilateral hilar lymphadenopathy on chest radiographs—a typical feature of this disease. Other patients present with complaints such as persistent dry cough, fatigue, weight loss, fever, night sweats, eye redness, enlarged peripheral lymph nodes, or erythema nodosum (ATS/ ERS/WASOG Committee, 1999; Nunes et al., 2007). At onset, 84% of patients have intrathoracic disease and up to 30% show extrapulmonary involvement (Rizzato, 2001); the latter may include skin, liver, spleen, eyes, lymph nodes, central nervous system, salivary glands, mucosae, joints, heart, bone marrow, muscles, and kidneys (Baughman et al., 2003; Nunes et al., 2007). Moreover, sarcoidosis often induces disturbances of calcium metabolism: hyperproduction of vitamin D by activated macrophages in granulomatous lesions causes increased absorption of dietary calcium, which may subsequently lead to hypercalcaemia, hypercalciuria, and renal stone formation (Muther et al., 1981; Singer and Adams, 1986). The evolution and severity of the disease are highly variable. Spontaneous remission occurs in most patients within a few years; however, in some cases there may be a more chronic and unfavourable course. Pulmonary fibrosis is the most common severe complication (ATS/ERS/WASOG Committee, 1999). Death may result from lung, heart, and central nervous system involvement (Reich, 2002). The diagnosis of sarcoidosis is based on suggestive clinical and radiographic findings, presence of non-caseating granulomas on biopsies, and exclusion of all other granulomatous disorders, such as mycobacterial, spirochaetal, fungal, and parasitic infections (Nunes et al., 2007; Rao and Sabanegh, 2009). Hypercalciuria (more commonly seen than hypercalcaemia) and elevated serum levels of angiotensin-converting enzyme (ACE) further support the diagnosis; the latter abnormality results from production of ACE by active epithelioid cells within granulomas, but it is not specific for sarcoidosis (Johns and Michele, 1999). The therapy of sarcoidosis mainly depends on its clinical severity: 30–70% of patients with mild disease never need to be treated. On the other hand, those with cardiac, neurological, renal, or ocular involvement not responding to topical agents or with malignant hypercalcaemia always require systemic anti-inflammatory and/ or immunosuppressive drugs. Corticosteroids are the mainstay of therapy, as they impede granuloma formation and are generally efficient against most active clinical manifestations; however, frequent relapses may occur, especially if steroids are stopped too early (Nunes et al., 2007). Antimalarial drugs are indicated in

Chapter 93 

mild isolated skin involvement (Baughman, 2002). Azathioprine is a useful steroid-sparing agent, as well as methotrexate, which is beneficial in both pulmonary and extrapulmonary disease, particularly cutaneous, ophthalmic, neurologic, and musculoskeletal (Baughman, 2002). Cyclophosphamide has sometimes been used, with good results, in severe steroid-resistant sarcoidosis. More recently, the TNF-α inhibitor infliximab was also shown to be effective in a double-blind study in patients with severe lung disease (Baughman et al., 2006), as well as in some small series and case reports with refractory lupus pernio, uveitis, or central nervous system involvement (Yee and Pochapin, 2001; Baughman and Lower, 2001; Mallbris et al., 2003; Pettersen et al., 2002). Renal involvement in sarcoidosis (Table 93.1) is generally considered to be rare, but it is probably underestimated (Mahevas et al., 2009). In a series of > 800 patients with sarcoidosis (James 1984), the incidence of clinical kidney disease was reported to be 1%. On the other hand, in small series of biopsy reports, some degree of morphological renal involvement has been described in as many as 50% of cases (Bergner et al., 2003). In practice, renal manifestations can be seen in association with other localizations of sarcoidosis or as the initial and/or sole presentation of the disease (Berliner et al., 2006). Table 93.1  Renal involvement in sarcoidosis (Berliner et al. 2006) Hypercalcaemia

Affects 10–20% of patients, can cause acute kidney injury or chronic kidney disease secondary to nephrocalcinosis

Hypercalciuria

Most common renal manifestation, affecting up to 50% of patients, caused by glomerular filtration of excess blood calcium and suppression by calcitriol of parathyroid hormone activity in the nephron; risk factor for nephrolithiasis

Granulomatous interstitial nephritis

The classic renal lesion of sarcoidosis with non-caseating granulomatous inflammation; although found in a substantial number of kidneys at autopsy in patients with sarcoidosis, only represents a very small proportion of clinically relevant renal failure

Glomerular disease

Rare; many different associated lesions, with membranous nephropathy perhaps the most common

Renal tubular dysfunction

Common; may include proximal or distal renal tubular acidosis, Fanconi syndrome, mild urinary concentrating defects or frank diabetes insipidus, and metabolic alkalosis

Renovascular disease

Rare; associated with severe hypertension; may be caused by renal artery stenosis from granulomatous angiitis or renal artery encasement by an external inflammatory mass

Obstructive uropathy

Rare; genitourinary tract structures may be obstructed by external lymphoid masses or direct sarcoid involvement in genitourinary tissues, causing obstructive uropathy

immune-mediated tubulointerstitial nephritis

Most often, the renal involvement is secondary to the vitamin D and calcium metabolism abnormalities associated with sarcoidosis. Nephrolithiasis has been reported in up to 14% of patients (Muther et al., 1981; Singer and Adams, 1986), whereas nephrocalcinosis is thought to be the main cause of end-stage renal disease (Muther et al., 1981; Casella and Allon, 1993). Obstructive uropathy may also occur, as a result of retroperitoneal granulomas, retroperitoneal fibrosis, renal stones, or ureteral involvement (Gil et al., 2010). Sarcoidosis-related glomerular diseases are rare and their mechanisms are unclear (Göbel et al., 2001); cases of membranous nephropathy (Toda et al., 1999), diffuse proliferative or crescentic glomerulonephritis (Van Uum et al., 1997), and focal segmental glomerulosclerosis (Casella and Allon, 1993) have been described so far. Sarcoid tubulointerstitial nephritis (TIN) is a less common cause of renal impairment. Autopsies of patients with sarcoidosis found TIN in 7–23% of cases, but many of these were asymptomatic (Longcope and Freiman, 1952; Berliner et al., 2006). The two largest series of patients with sarcoid TIN published so far included 94 (Berliner et al., 2006) and 47 cases (Mahevas et al., 2009), respectively. These studies have shown that, in contrast with the female predominance of pulmonary sarcoidosis, the sex ratio in patients with TIN is approximately 1.76:1 in favour of males. The most common urinary abnormality is moderate proteinuria, usually < 1.0 g/day. Microscopic haematuria and sterile pyuria are found each in 20–30% of cases, while glycosuria and hypercalciuria are less frequent. At presentation, most patients have severe renal dysfunction, with a mean serum creatinine of 4.8 mg/dL (Berliner et al., 2006) and an estimated glomerular filtration rate (GFR) of 20 mL/min per 1.73 m2 (Mahevas et al., 2009). The large majority also have extrarenal localizations of sarcoidosis, although chest radiographs are often normal. Serum ACE is typically increased; however, due to its low specificity and lack of correlation with kidney disease severity, it is not useful as a diagnostic tool, but mainly as a marker of disease activity and response to therapy (Berliner et al., 2006). On renal biopsy, the interstitial inflammatory infiltrate is confined mainly to the renal cortex (Longcope and Freiman, 1952) and has a granulomatous pattern in about 80% of cases; yet, no granulomas are found in the remainder 20% (Mahevas et al., 2009). Sarcoid granulomas are non-caseating and consist of macrophages, CD4+ T lymphocytes, and multinucleated giant cells (Utas et al., 1999; Göbel et  al., 2001). Gallium-67 radiotracer scanning has been used to help diagnosis or to monitor disease activity, especially in patients with pulmonary sarcoidosis; however, the sensitivity and specificity of this test for the renal disease are unknown and its value as a diagnostic tool is questionable (Pagniez and Delvallez, 1989; Berliner et al., 2006). Corticosteroids are very efficient for the treatment of renal sarcoidosis (Hannedouche et  al., 1990; Brause et  al., 2002; Robson et al., 2003; Rajakariar et al., 2006) and are capable to induce remission even in patients with advanced kidney disease (Simonsen and Thysell, 1985). However, renal function recovery is often incomplete, particularly in cases with chronic and irreversible lesions (Hannedouche et  al., 1990; O’Riordan et  al., 2001; Brause et  al., 2002). Serial renal biopsies in treated patients may show disappearance of granulomas, but no changes or even worsening of interstitial fibrosis (Farge et al., 1986; Hannedouche et al., 1990). The appropriate intensity and duration of therapy is unclear; however, it appears that prolonged administration is required to

727

728

Section 4  

the patient with interstitial disease

prevent progression to end-stage renal disease. Most protocols suggest an induction dose for prednisolone of 0.5–1 mg/kg per day, given for at least 6–12 months, with subsequent tapering to the minimal effective maintenance dose (Newman et al., 1997; O’Riordan et al., 2001; Brause et al., 2002; Berliner et al., 2006). If steroids are tapered or withdrawn too rapidly, relapse and progression of sarcoidosis can occur (Singer and Evans 1986; Robson et  al., 2003; Joss et al., 2007; Mayer et al., 2008). Fortunately, relapses usually respond to increased steroid doses (Hannedouche et al., 1990; Mills et al., 1994). In the aforementioned study by Mahevas et al., all 47 patients were initially treated with prednisone, for a median duration of 18 months, and 10 of them also received intravenous pulse methylprednisolone. The mean estimated GFR (eGFR) gradually increased from 20 to 44 (after one month), to 47 (after 1 year), and to 49 mL/min per 1.73 m2 (after a median follow-up of 24 months). Renal function long-term improvement was directly correlated with the response obtained at 1 month and inversely related to the initial histologic fibrosis score. Relapses occurred in 17 patients and these were purely renal (N = 3), purely extrarenal (N = 10), or both (N = 4) (Mahevas et al., 2009). Various steroid-sparing drugs have been tried in steroid-dependant, intolerant, or refractory cases, but experience with such agents is very limited (Thumfart et al., 2005). In a 13-yearold boy with sarcoid TIN (initially treated with steroids, but having developed steroid toxicity), therapy with infliximab led to partial recovery of renal function and disappearance of renal granulomas on biopsy (Thumfart et al., 2005). Mycophenolate mofetil (Moudgil et al., 2006) and mizoribine (Ito et al., 2009) were also shown to successfully maintain remission in two isolated cases of paediatric patients with TIN. Patients with sarcoidosis have received transplanted hearts, lungs, livers, and kidneys without an apparent increase in morbidity compared with other transplant recipients (Padilla et al., 1997). However, little is known about the incidence rate and outcomes of sarcoidosis recurrence in renal allografts. In a series of 18 patients (Aouizerate et al., 2010), after a median follow-up of 42 months, the patient and graft survival rates were 94% and the mean GFR was 60 mL/min per 1.73 m2; recurrence of sarcoidosis was seen in five patients (27%), with extrarenal involvement in two and renal involvement in three cases.

Systemic lupus erythematosus Lupus nephritis is discussed more broadly in Chapter 161 and following chapters. Systemic lupus erythematosus (SLE) is characterized by abnormalities of immune regulation and loss of self-tolerance, triggering systemic autoimmune reactions. These reactions include the activation of autoreactive B cells that produce antibodies against nuclear and other antigens. Circulating immune complexes are deposited at multiple sites, including the kidneys, where they induce complement activation and a massive cascade of inflammatory events (Benigni et al., 2007). Immune complex deposition and inflammation occur in both the glomeruli and the tubulointerstitium and, if left untreated, they may result in scarring and irreversible chronic kidney disease (Lahita 2004). Glomerulonephritis (GN) is the most prominent and best studied feature of lupus nephritis (LN). Moreover, the current approach to the treatment of LN is largely guided by the International Society

of Nephrology/Renal Pathology Society classification of the GN (Weening et al., 2004). Lupus nephritis, including its pathogenesis, clinical manifestations, diagnosis, and therapy, is described in detail elsewhere in this book; therefore, we shall focus in the following only on some particularities of the tubulointerstitial component. Tubulointerstitial involvement is very common in patients with LN (Hsieh et  al., 2011), although its clinical manifestations are hardly remarkable. The urinary signs of tubulointerstitial damage, like haematuria and pyuria, are non-specific and difficult to distinguish from those of GN. Isolated tubular proteinuria is exceptional and renal tubular dysfunctions, such as impaired urine concentrating ability and renal tubular acidosis (RTA), are rarely clinically significant (Balow, 2005). However, the presence and severity of tubulointerstitial damage on renal biopsy is recognized as a risk factor for progression to end-stage renal disease (Schwartz et al., 1982; Esdaile et al., 1989; Nath, 1992; Hsieh et al., 2011), while, in contrast, the NIH activity index, which primarily assesses glomerular inflammation, does not have a similar prognostic value (Austin et al., 1983; Esdaile et al., 1989; Neumann et al., 1995). Furthermore, the presence of tubulointerstitial scarring is more predictive of subsequent renal failure than glomerular scarring (Schwartz et al., 1982; Esdaile et al., 1989; Hsieh et al., 2011). Unlike GN, the severity of TIN does not correlate with titres of serum anti-double-stranded DNA (anti-dsDNA) antibodies (Hsieh et  al., 2011)  and, occasionally, TIN can occur independently of GN (Singh et al., 1996; Mori et al., 2005; Moyano et al., 2009). These data indicate that, in LN, the pathogenic mechanisms of TIN may be somewhat different from those involved in the GN (Chang et al., 2011). Immune complexes are found in the tubular basement membrane (TBM) in about 50% of patients with LN, more frequently in those with endothelial rather than mesangial GN pattern (Stewart Cameron, 1999). Tubulitis (active infiltration and invasion of tubules by mainly lymphocytes and monocytes) is often seen in active disease, whereas in more chronic disease, the interstitium is invaded by a variable amount of collagen (Molino et al., 2009). The tubulointerstitial infiltrate is often organized into well-circumscribed T:B cell aggregates or into germinal centres containing follicular dendritic cells, both of these formations being strongly associated with the TBM immune complexes (Chang et al., 2011). Proximal tubular epithelial cells seem to play an active role in tubulointerstitial damage, by releasing proinflammatory cytokines when exposed to anti-dsDNA antibodies (Yung et al., 2005). Interestingly, titres of autoantibodies against monomeric C-reactive protein correlate positively with interstitial inflammation, tubular atrophy, and interstitial fibrosis, suggesting a possible role for these antibodies in the pathogenesis of TIN (Tan et al., 2008). Cases have been reported in which tubulointerstitial inflammation was linked to immune complex deposition in the capillary walls of the interstitium (Hayakawa et al., 2006) or to CD8+ cytotoxic T cells (Omokawa et al., 2008).

Immunoglobulin G4-related disease Immunoglobulin G4-related disease (IgG4-RD), also known as IgG4-related multiorgan lymphoproliferative syndrome (IgG4-MOLPS), is a systemic disorder characterized by increased levels of serum IgG4 and infiltrates of IgG4-producing plasma cells, together with fibrosis, involving multiple organs (Zhang and Smyrk, 2010; Kim et al., 2011; Masaki et al., 2011).

Chapter 93 

The disease was initially described in the pancreas, as an unusual form of chronic pancreatitis. It was originally called ‘primary inflammatory sclerosis of the pancreas’ (Sarles et al., 1961), but various other names were later coined, such as ‘lymphoplasmacytic sclerosing pancreatitis’, ‘chronic sclerosing pancreatitis’, ‘non-alcoholic duct-destructive chronic pancreatitis’, and ‘inflammatory pseudotumour’ (Kawaguchi et al., 1991; Sood et al., 1995; Ectors et al., 1997; Wreesmann et al., 2001). The concept of ‘autoimmune pancreatitis’ (AIP) was suggested in 1995 by Yoshida et al., who reported a patient with chronic pancreatitis, hyperglobulinaemia, circulating autoantibodies, and corticosteroid sensitivity (Yoshida et al., 1995). Subsequent studies confirmed the autoimmune nature of the disease and, in 2001, Hamano et al., discovered its association with elevated serum levels of IgG4 (Hamano et al., 2001). The idea of a systemic IgG4-related disease emerged in 2003, when Kamisawa et al., described widespread IgG4+ plasma cell infiltrates in patients with AIP (Kamisawa et al., 2003). Indeed, in the following years, it was found that almost every organ can be involved in this disease, including kidneys, liver, gallbladder, gastrointestinal tract, salivary and lacrimal glands, lungs, orbits, breasts, retroperitoneum, aorta, lymph nodes, skin, pituitary gland, and prostate (Kitagawa et al., 2005; Uehara et al., 2005; Zen et al., 2005; Deshpande et al., 2006; Shrestha et al., 2009; Cheuk and Chan, 2010). Although reported worldwide, IgG4-RD has been studied mostly in Japan, where it appears to have a relatively high incidence, possibly due to better diagnosis (Masaki et al., 2011). The diagnostic criteria for IgG4-RD have been recently defined by the group of Umehara et al. (Masaki et al., 2011), as shown in Table 93.2. Renal involvement in IgG4-RD mainly consists of a TIN with typical plasma cell infiltration and TBM immune complex deposition (Watson et al., 2006; Cornell et al., 2007; Saeki et al., 2010). Glomerular disease may also be present, most commonly as membranous nephropathy (Watson et al., 2006; Saeki et al., 2009; Hill et  al., 2009; Raissian et  al., 2011)  or, sometimes, as membranoproliferative (Morimoto et  al., 2009)  or crescentic GN (Katano et al., 2007). In the largest published series of patients with IgG4-related kidney disease (n = 35) (Raissian et al., 2011), the average age at the time of diagnosis was 65 years (range 20–81) and 86% were men. Most patients (83%) also had extrarenal involvement. High serum IgG, IgG4 or total gamma globulin levels were found in 88% of cases, while 56% had hypocomplementaemia, with low C3 and/or C4 concentrations, and 33% had peripheral blood eosinophilia. Urinalysis revealed proteinuria > 1.0 g/day in eight and haematuria in six of 27 patients. The mean serum creatinine at presentation was 3.6 mg/ dL (range, 0.9–9). Renal imaging showed abnormalities in 78% of cases, consisting of small low-attenuation lesions (usually bilateral and multiple), tumour masses or markedly enlarged kidneys. All 35 renal biopsies showed diffuse or multifocal interstitial infiltrates, consisting of plasma cells, mononuclear cells, and eosinophils. Focal mild mononuclear cell tubulitis was seen in most cases. In 30 cases, moderate-to-severe interstitial fibrosis and tubular atrophy was found, whereas five cases showed an acute TIN pattern, with minimal fibrosis. The glomeruli appeared normal or only with a mild mesangial matrix expansion. Immunohistochemistry showed moderate or marked increase in IgG4+ plasma cells in all specimens, and TBM IgG4-containing immune complex deposits in 83%. The diagnosis of IgG4-related TIN should be considered especially in elderly male patients, presenting with unexplained urinary

immune-mediated tubulointerstitial nephritis

Table 93.2  Diagnostic criteria for IgG4-RD (Masaki et al. 2011) 1. Elevated serum IgG4 (> 135 mg/dL) AND 2. Histopathological features including lymphocyte and IgG4+ plasma cell infiltration (IgG4+ plasma cells/IgG+ plasma cells > 40%) with typical tissue fibrosis or sclerosis. Note: ◆ It is necessary to distinguish IgG4-RD from other disorders, including sarcoidosis, Castleman disease, Wegener granulomatosis, lymphoma, and cancer ◆ Patients fulfilling only one of the above criteria are classified as ‘suspected IgG4-RD’ ◆ Patient fulfilling both (1) and (2) and having other distinct disorders (designated as ‘XX’), are classified as having ‘XX disease with suspected association with IgG4-RD’ ◆ Patients diagnosed with IgG4-RD, but refractory to glucocorticoid treatment, should be re-diagnosed Suspicious IgG4-RD: 1. Presence of only one can be enough for the suspicious IgG4-RD lesion: a. Symmetrical swelling of one of the lacrimal, parotid, or submandibular glands b. Autoimmune pancreatitis c. Inflammatory pseudotumour d. Retroperitoneal fibrosis e. Histopathological findings are similar to lymphoplasmacytosis or suspected Castleman disease 2. Presence of at least two would be sufficient for suspected IgG4-RD: (a) unilateral swelling of one of the lacrimal, parotid, or submandibular glands, (b) orbital tumorous lesion, (c) autoimmune hepatitis, (d) sclerosing cholangitis, (e) prostatitis, (f) patchy meningitis, (g) interstitial pneumonitis, (h) interstitial nephritis, (i) mediastinal fibrosis, (j) thyroiditis or hypothyroidism, (k) hypophysitis, (l) inflammatory aneurysm 3. Common findings in patients with IgG4-RD: (a) polyclonal hyper-IgG-gammopathy, (b) elevation of serum IgE or eosinophilia, (c) hypocomplementaemia or presence of immune complex in serum, (d) tumorous lesion or lymphadenopathy with strong accumulation in 67Ga-scan or 18FDG-PET-scan

abnormalities and/or decreased eGFR, hypergammaglobulinaemia, and a history of other inflammatory organ involvement (Kim et  al., 2011; Raissian et  al., 2011). Radiologic methods such as galium-67 scintigraphy (Saeki et  al., 2007), contrast-enhanced computed tomographic (CT) imaging (Takahashi et  al., 2007; Kim et al., 2011), and 18FDG-positron emission tomography scan (Nakajo et al., 2007; Lee et al., 2009) are helpful to detect renal and other organ lesions; however, the use of contrast media should be avoided in cases with impaired renal function. Raissian et  al. recently proposed a set of criteria for the diagnosis of IgG4-related TIN (Raissian et al., 2011) (Table 93.3) The differential diagnosis of IgG4-related TIN should take into consideration the so-called idiopathic hypocomplementaemic TIN with extensive tubulointerstitial deposits, an extremely rare disorder, with only 12 cases described so far (Kambham et al., 2001; Vaseemuddin et al., 2007; Gupta et al., 2010); however, it is currently believed that this entity might in fact be nothing else but unrecognized IgG4-TIN (Raissian et al., 2011). Other diseases that need to be excluded are Sjögren syndrome and lupus nephritis, which may have clinical similarities but usually no IgG4+ plasma

729

730

Section 4  

the patient with interstitial disease

Table 93.3  Diagnostic criteria for IgG4-related TIN (Raissian et al. 2011) Histology

Plasma cell–rich tubulointerstitial nephritis with > 10 IgG4 + plasma cells/high power field in the most concentrated field (mandatory criterion) Tubular basement membrane immune complex deposits by immunofluorescence, immunohistochemistry, and/or electron microscopy (supportive criterion, present in > 80% of cases)

Imaging

Small peripheral low-attenuation cortical nodules, round or wedge-shaped lesions, or diffuse patchy involvement Diffuse marked enlargement of kidneys

Serology

Elevated serum IgG4 or total IgG level

Other organ involvement

Includes autoimmune pancreatitis, sclerosing cholangitis, inflammatory masses in any organ, sialadenitis, inflammatory aortic aneurysm, lung involvement, retroperitoneal fibrosis

The diagnosis of IgG4-TIN requires the histologic feature of plasma cell–rich TIN with increased IgG4+ plasma cells and at least one other feature from the categories of ‘imaging’, ‘serology’, or ‘other organ involvement’

cell interstitial infiltrates; furthermore, in SLE, GN is almost always the dominant feature (Raissian et al., 2011). IgG4-RD is typically very sensitive to corticosteroids, which is the standard treatment for this disease. The initial response is spectacular, and even fibrotic lesions may show some improvement with this therapy (Masaki et al., 2011). On the other hand, spontaneous remissions, without any treatment, have also been described. As there are no randomized controlled studies comparing steroids to placebo, it is difficult to decide when or which patients should be treated. However, functional impairment of the pancreas, kidneys, lungs or liver can significantly reduce a patient’s quality of life and survival duration; therefore, it seems reasonable to initiate steroids in those cases where any of these organs is involved, in order to prevent irreversible damage (Masaki et al., 2011). Unfortunately, there is yet no consensus regarding the starting dose, the duration of the initial therapy, the tapering schedule, or the maintenance doses of steroids. A clinical prospective study has recently been started by a Japanese group, aiming to establish the optimal treatment strategy for IgG4-RD (Masaki et al., 2011). In this study, glucocorticoid treatment consists in oral prednisolone at an initial dose of 0.6 mg/kg per day, divided into three doses, with tapering by 10% every 2 weeks. A maintenance dose of 10 mg per day is continued for at least 3 months, and a further daily dose of prednisolone is left at the discretion of the attending physician; the final maintenance dose is to be decided in dependence of symptoms and clinical data in each case. The authors have noticed that most patients require 5–10 mg per day of prednisolone as a maintenance dose to prevent relapses, which may occur at a rate of 30–40% after steroid discontinuation (Masaki et al., 2011). Typically, response to treatment can be confirmed within several days. Superficial organs, such as the lacrimal, parotid, and submandibular glands, and lymph nodes, can be monitored by physical examination, but deep organs, like the pancreas or the kidneys, require imaging examination (CT) after 2 weeks of treatment; if the response is insufficient, a differential diagnosis work-up should be redone, to exclude

other diseases, such as cancer, lymphoma, Castleman disease, or sarcoidosis (Masaki et al., 2011). The response to treatment of the IgG4-related TIN has not been well studied but available data suggest favourable outcomes. In the series of Raissian et al., 21 patients were treated with prednisone and two patients additionally received mycophenolate mofetil. After a mean follow-up of 14.5 months, 17 of 19 patients with renal insufficiency showed a decrease in serum creatinine (from 3.5 mg/dL to 1.7 mg/dL), whereas only two patients failed to improve and developed end-stage renal disease. Interestingly, there was no correlation between the renal biopsy findings and the response to therapy, and even patients with extensive interstitial fibrosis responded to steroid treatment. In two cases there was a relapse upon steroid taper. Two patients treated with prednisone had normal renal function at baseline and remained stable on follow-up. In contrast, the five patients who received no treatment showed increasing or persistently elevated serum creatinine (Raissian et al., 2011). As well as for other organ involvement in IgG4-RD, low-dose steroid maintenance therapy is most likely necessary to prevent relapses of TIN, although there is no consensus upon the recommended duration and dose. Long-term careful observation is required in all patients, regardless of maintenance therapy (Saeki et al., 2007; Aoki et al., 2009). There is little evidence concerning the treatment for relapsed and refractory cases; another course of steroids is usually effective, but other options may also be considered, including azathioprine (Chari, 2007), cyclophosphamide, methotrexate, mizoribine (Nanke et al., 2010), rituximab (Topazian et al., 2008; Khosroshahi et al., 2010), and bortezomib (Khan et al., 2010).

Sjögren syndrome Primary Sjögren syndrome (pSS) is an autoimmune disorder involving the lacrimal and salivary glands. Its clinical presentation consists of keratoconjunctivitis and xerostomia (sicca syndrome), and its characteristic pathological feature is a lymphocytic infiltrate around the epithelial ducts of these glands. Extraglandular manifestations are seen in about 25% of patients and may include interstitial lung disease, cutaneous vasculitis, peripheral neuropathy, and lymphoma. Secondary Sjögren syndrome occurs in relation with other systemic autoimmune diseases, such as rheumatoid arthritis, systemic sclerosis, and SLE (Aasarød et al., 2000; Maripuri et al., 2009). Sjogren syndrome and related conditions are further considered in Chapter 166. Kidney involvement has been reported in 4–67% of patients with pSS—this wide variation being probably explained by the different classification criteria used in the studies and the selection of patients (Pokorny et al., 1989; Eriksson et al., 1995; Goules et al., 2000; Bossini et al., 2001; Lin et al., 2010). TIN is the most common underlying renal pathological lesion (Enestrom et al., 1995), typically consisting of a plasmacytoid lymphocytic infiltrate (Maripuri et al., 2009). GN is generally less frequent, tends to develop late in the course of the disease (Skopouli et al., 2000), and may consist in focal segmental glomerulosclerosis, membranous nephropathy, or mesangial proliferative nephritis (Ren et  al., 2008; Lin et  al., 2010). In some of the largest published series of patients with pSS and biopsy-proven renal involvement, Maripuri et al. found TIN in 71% (chronic TIN in 46% and acute TIN in 25%) and GN in 29% of cases (Maripuri et al., 2009), Ren et al. described TIN in 85% and GN in 15% of cases (Ren et al., 2008), whereas Lin et al.

Chapter 93 

reported pure TIN in approximately 34%, pure GN in 37%, and combined TIN and GN in 29% of cases (Lin et al., 2010). TIN can manifest with distal (type I) RTA and less often with Fanconi syndrome (Kassan and Talal, 1987; Siamopoulos et al., 1992; Kobayashi et al., 2006). Hyposthenuria (Kassan et al., 1987) and hypokalaemia (Aasarød et al., 2000; Toy and Jasin 2008) may also occur. Distal tubular acidosis is generally asymptomatic but it increases the risk of stone disease and nephrocalcinosis (Moutsopoulos et al., 1991). Among 573 patients with pSS, Lin et al. found 192 patients (33.5%) with renal involvement, of which 126 (65.6%) had proteinuria (0.5 g/day, on average), 96 (50%) had RTA (type I in 91.7% of these), 45 (23.4%) had kidney stones and/or nephrocalcinosis, and 41 (21.3%) had renal insufficiency (Lin et al., 2010). Patients with pSS and extraglandular manifestations are usually treated with systemic corticosteroids and, sometimes, other immunosuppressive drugs. However, there is limited evidence on the outcomes of organ involvement with such therapies, since available controlled and prospective studies were small and specifically designed to evaluate the sicca syndrome (Ramos-Casals et al., 2010). In the case series of Maripuri et  al., 20 patients with pSS and biopsy-proven renal disease were treated with steroids (with a median initial dose of 40 mg/day and a median duration of 30 weeks); eight patients remained on maintenance corticotherapy for > 1 year, with a median dose of 5 mg/day. Two patients with severe TIN on biopsy received cyclophosphamide and one patient received rituximab, in addition to steroids. In 16 patients that were followed for more than 12 months, there was a significant improvement in eGFR and proteinuria. None of the patients progressed to end-stage renal disease, after a median follow-up of 38 months (range 3–192). The treatment was well tolerated, with no severe adverse effects attributable to immunosuppressive therapy. In conclusion, the authors suggested that all patients with pSS and renal involvement should receive a course of corticosteroids as first-line treatment (Maripuri et al., 2009). Rituximab might be an option for steroid-refractory cases, as it has shown good results in a controlled trial in patients with pSS and vasculitis (Ramos-Casals et al., 2010); however, the precise role of this drug remains to be defined by future research.

Inflammatory bowel disease Inflammatory bowel disease (IBD) is a chronic relapsing inflammatory disease characterized by mucosal ulcerations of the digestive tract. It includes two main disorders:  Crohn disease and ulcerative colitis. IBD results from a genetic predisposition to abnormal interactions between intestinal epithelial cells and luminal bacteria (Riis et al., 2007). Crohn disease generally involves the ileum and the colon, but it can affect any part of the intestine, often in a discrete manner, whereas ulcerative colitis involves the rectum and, sometimes, other parts of the colon, in a continuous pattern. Crohn disease can be associated with intestinal granulomas, strictures, and fistulas, but these are not characteristic of ulcerative colitis (Abraham and Cho 2009). Extraintestinal manifestations are common in IBD and may occur in up to 47% of patients (Danese et al., 2005), involving the skin, the eyes, the joints, the biliary tract, and the kidneys (Rothfuss et al., 2006). Renal and urinary tract manifestations are seen in 4–23% of patients with IBD (Pardi et al., 1998) and often have a significant impact on the quality of life, morbidity, and mortality of these

immune-mediated tubulointerstitial nephritis

patients (Rothfuss et  al., 2006). Malabsorption, bacterial overgrowth, and short bowel syndrome after resection may complicate with enteric hyperoxaluria; this, in turn, may lead to calcium oxalate stone formation, obstructive uropathy, chronic pyelonephritis, and fistulization of the urinary tract, which are probably the most common reno-urinary manifestations of IBD (Banner, 1987; Oikonomou et al., 2011). Proximal tubular dysfunction may often be detected in this setting (Kreisel et al., 1996; Fraser et al., 2001; Herrlinger et al., 2001; Mahmud et al., 2002), sometimes associated with proximal RTA and osteomalacia (Victorino et al., 1986; Pardi et al., 1998). IgA nephropathy and renal amyloidosis have also been described in IBD (Pardi et al., 1998; Filiopoulos et al., 2010). Tubulointerstitial nephritis may occur in patients with IBD mainly as a result of treatment with 5-aminosalicylic acid (Mahmud et al., 2002; Rothfuss et al., 2006; Jose et al., 2009); however, cases of TIN have also been described independently of drug intake and these are thought to have an autoimmune pathogenesis (Kreisel et al., 1996; Mahmud et al., 1996; Fraser et al., 2001; Herrlinger et  al., 2001; Izzedine et  al., 2002; Poulou et  al., 2006; Marcus et al., 2008; Oikonomou et al., 2011). Autoimmune TIN has been reported in only a few patients with Crohn disease (both children and adults); its occurrence was associated with exacerbation of the bowel disease. Renal biopsies showed a predominant lymphocytic infiltrate with characteristic non-necrotizing granulomas (Archimandritis and Weetch, 1993; Izzedine et  al., 2002; Tovbin et al., 2000; Marcus et al., 2008). Response to treatment seems to be rather poor, with both steroids and infliximab (Marcus et al., 2008); all published cases progressed to end-stage renal disease, 30% of them in < 3 years (Oikonomou et al., 2011). One reported patient who underwent renal transplantation experienced deterioration in graft function during a post-transplantation relapse of Crohn disease (Archimandritis and Weetch, 1993). In ulcerative colitis, autoimmune TIN is also very rare; renal biopsy shows intense interstitial mononuclear infiltration with occasional eosinophils, fibrosis and tubular atrophy (Khosroshahi and Shoja, 2006). Hypokalaemia (due to chronic diarrhoea) may also play a role in inducing or aggravating TIN (Oikonomou et al., 2011).

Primary biliary cirrhosis Primary biliary cirrhosis (PBC) is an autoimmune cholestatic chronic liver disease, characterized by non-suppurative destruction of interlobular bile ducts, associated with high levels of serum IgM and the presence of circulating anti-mitochondrial antibodies. About 70% of patients with PBC have extrahepatic organ involvement, including the kidneys (Talwalkar and Lindor 2003). The main feature of PBC-associated renal disease is distal RTA, which is found in one-third of patients, but usually with no clinical significance (Parés et al., 1981; Komatsuda et al., 2010). Only nine patients with TIN and PBC have been reported so far (Macdougall et al., 1987; Kamouchi et al., 1991; Kodama et al., 1996; Lino et al., 2005; Terrier et al., 2008), recently reviewed by Komatsuda et al. (2010). All of those patients were females, with a median age of 56 years (range 36–68). Eight of them had reduced eGFR, five had distal RTA, and five had Fanconi syndrome. The clinical manifestations of tubular involvement were hypokalaemia, bone pains and fractures, and urinary abnormalities, including mild proteinuria, glycosuria, and high urinary levels of β2-microglobulin and N-acetyl-β-D-glucosaminidase. Renal biopsies showed severe

731

732

Section 4  

the patient with interstitial disease

interstitial lymphocytic infiltration, tubulitis, and mild-to-moderate tubular atrophy and fibrosis. On the other hand, only mild or no glomerular or vascular abnormalities were found. In some cases, immunofluorescence revealed glomerular deposits of IgM in the mesangium and/or along the capillary walls. All nine patients were treated with steroids, with good results in seven patients and no effect in two patients (Komatsuda et al., 2010).

Tubulointerstitial nephritis and uveitis syndrome Tubulointerstitial nephritis and uveitis syndrome (TINU) is a rare disease, of unknown origin, associating ocular and renal inflammation. It usually occurs in young females (Mackensen and Billing, 2009). Since its first description in 1975 (Dobrin et al., 1975), some 200 cases have been published to date. The pathogenesis of TINU is unclear but it is thought to be immune mediated. Renal tubular and ciliary body epithelia possibly share similar antigens that may account for a cross-reactivity (Izzedine 2008). Abed et al. reported the presence of autoantibodies against a common tubular and uveal antigen in the serum of a TINU paediatric patient (Abed et al., 2008), while Shimazaki et al. published a TINU case in which serum antibodies against a 125-kDa renal and retinal protein were detected (Shimazaki et al., 2008). However, immunofluorescence studies of renal biopsies from patients with TINU could very rarely identify anti-TBM antibody deposition (Morino et al., 1983; Wakaki et al., 2001). On the other hand, TINU has occasionally been described in association with other autoimmune diseases, such as thyroiditis and rheumatoid arthritis, and in some patients with TINU serum antinuclear antibodies, rheumatoid factor, and anticardiolipin antibodies have been found (Mandeville et al., 2001). A recent report of a patient with recurrence of TINU after renal transplantation (Onyekpe et al., 2011) also suggests a role for circulating autoantibodies in the pathogenesis of this disease. Genetic factors are also thought to be involved, since strong associations with human leucocyte antigen (HLA) haplotypes HLA-DRB1*01 and HLA-DQA1*01 (Levinson et  al., 2003; Mackensen et al., 2008), as well as sporadic cases of monozygotic twins (Gianviti et al., 1994; Howarth et al., 2004) and familial clustering (Tanaka et al., 2001b; Dusek et al., 2008) have been reported. Furthermore, a role for recent infections with Epstein–Barr virus, herpes zoster virus, or Chlamydia trachomatis (Stupp et al., 1990; Cigni et al., 2003; Mandeville et al., 2001) and exposure to antibiotics or non-steroidal anti-inflammatory drugs (Mandeville et al., 2001)  has also been suggested, without conclusive evidence. An interesting finding by Kase et al. was a significant increase in serum levels of Krebs von den Lunge-6 (KL-6) protein (a human glycoprotein secreted by type II pulmonary alveolar cells) in 17 patients with TINU syndrome, as compared to controls. Renal distal tubules also stained strongly with anti-KL-6 antibody, suggesting that high KL-6 levels may reflect the renal lesion of this disease (Kase et al., 2006). The clinical onset of TINU is often with non-specific symptoms, such as malaise, fatigue, and fever, especially in children. The most common symptoms and laboratory findings are shown in Table 93.4. Ocular symptoms may precede (in 20% of cases) or follow the systemic ones (in 65%), by up to 14 months. (Mandeville et al., 2001; Izzedine 2008). Patients complain of eye redness, pain, photophobia, and blurred vision. The eye involvement consists of uveitis, which is typically non-granulomatous, bilateral, and limited to

Table 93.4  The most common symptoms and laboratory findings in TINU (Mackensen and Billing 2009) General symptoms

%

Fever

53

Weight loss

47

Fatigue

44

Anorexia

28

Weakness

28

Abdominal or flank pain

28

Arthralgias

17

Polyuria, nocturia

8

Ocular symptoms Eye pain or redness

77

Blurred vision

20

Photophobia

14

Laboratory abnormalities Blood: Anaemia

96

High creatinine

90

High ESR

89

High IgG levels

83

Urine: Proteinuria

86

High β2-microglobulin

92

Pyuria

55

Haematuria

42

the anterior segment in about 80% of cases (Mandeville et al., 2001; Mackensen et al., 2007). The renal signs may include pyuria (sometimes, with eosinophiluria), haematuria, and moderate proteinuria. Manifestations of proximal and/or distal tubular dysfunction, such as Fanconi syndrome and distal RTA, are also common (Igarashi et al., 1992; Yao et al., 2011). Renal ultrasound often shows enlarged kidneys with increased echogenicity, but these findings are non-specific (Michel and Kelly, 1998; Kodner and Kudrimoti 2003). Gallium scanning may be more sensitive, but also more invasive and equally non-specific (Mackensen and Billing 2009). The kidney biopsy remains the key diagnostic tool for TIN. Usually, light microscopy reveals a mixed inflammatory infiltrate with mononuclear cells including lymphocytes, plasma cells, and macrophages and, sometimes, also with eosinophils and neutrophils. Non-caseating granulomas have been found in 13% of cases (Dobrin et al., 1975; Mandeville et al., 2001; Herlitz et al., 2007). Immunofluorescence is generally negative, although tubular and glomerular immunoglobulin staining is occasionally observed (Mandeville et al., 2001). Bone marrow, lymph node, and hepatic granulomas have also been described in some patients (Dobrin et al., 1975; Mandeville et al., 2001; Herlitz et al., 2007).

Chapter 93 

The differential diagnosis of TINU implies the exclusion of other possible causes of associated eye and kidney disease, such as sarcoidosis, Sjögren syndrome, SLE, Wegener’s granulomatosis, Behçet disease, and infections (Mackensen et al., 2007). Other organ involvement (lungs, joints, skin), serology (ANA, ANCA), and biopsy findings are usually decisive. The prognosis of TINU is favourable in the majority of patients. In several studies, ocular complications have occurred in 21% to 45% of cases, depending on follow-up duration, and consisted mainly of posterior synechiae and, more rarely, of optic disc oedema, cystoid macular oedema, and cataract (Mandeville et al., 2001; Goda et al., 2005; Mackensen et al., 2007). Concerning the kidney disease, the necessity of dialysis, for either acute or chronic renal failure, has rarely been reported (Mandeville et  al., 2001); however, no long-term follow-up of nephritis patients has yet been published. Although uveitis has sometimes been treated with topical steroids, TINU syndrome should be regarded as a systemic disorder and, therefore, as an indication for systemic anti-inflammatory and/or immunosuppressive therapy. Patients with nephritis left untreated may develop end-stage renal disease (Suzuki et al., 2004). However, no clinical trials have been undertaken to test the necessity or the type of treatment for TINU, and the only existing evidence comes from a few reported case series (Mandeville et al., 2001). Systemic corticosteroids are used in most patients. The response is generally good for both the ocular and the renal disease (Takemura et al., 1999). However, the uveitis recurrence rate may be as high as 50% (Mandeville et al., 2001; Goda et al., 2005) to 100% (Takemura et al., 1999); therefore, its prevention is thought to require a relatively long-term treatment—possibly up to 12 months, (Mackensen et al., 2007). Doses of prednisone about 1 mg/kg per day for 2–3 weeks with subsequent taper are usually administered when renal impairment is severe or prolonged (Takemura et al., 1999). Usually, the renal function quickly returns to normal (Vohra et al., 1999), although increased urinary β2-microglobulin levels may last for several months (Kobayashi et al., 2000). Even patients who require renal replacement therapy can hope to improve and discontinue dialysis (van Leusen and Assmann, 1988; Mandeville et al., 2001). Nevertheless, a number of patients fail to respond to steroids and show persistent acute inflammatory changes on renal biopsies even after 6–9  months of treatment (Tanaka et  al., 2001a; Yanagihara et al., 2009). In cases with such prolonged or with relapsing course, mycophenolate mofetil or ciclosporin seem to be effective (Hinkle and Foster, 2008).

References Aasarød, K., Haga, H. J., Berg, K. J., et al. (2000). Renal involvement in primary Sjögren’s syndrome. QJM, 93(5), 297–304. Abed, L., Merouani, A., Haddad, E., et al. (2008). Presence of autoantibodies against tubular and uveal cells in a patient with tubulointerstitial nephritis and uveitis (TINU) syndrome. Nephrol Dial Transplant, 23(4), 1452–5. Abraham, C. and Cho, J. H. (2009). Inflammatory bowel disease. N Engl J Med, 361, 2066–78. Aoki, A., Sato, K., Itabashi, M., et al. (2009). A case of Mikulicz’s disease complicated with severe interstitial nephritis associated with IgG4. Clin Exp Nephrol, 13, 367–72. Aouizerate, J., Matignon, M., Kamar, N., et al. (2010). Renal transplantation in patients with sarcoidosis: a French multicenter study. Clin J Am Soc Nephrol, 5(11), 2101–18.

immune-mediated tubulointerstitial nephritis

Archimandritis, A. J. and Weetch, M. S. (1993) Kidney granuloma in Crohn’s disease. BMJ, 307, 540–1. ATS/ERS/WASOG Committee (1999). Statement on sarcoidosis. Am J Respir Crit Care Med, 160, 736–55. Austin, H. R., Muenz, L. R., Joyce, K. M., et al. (1983). Prognostic factors in lupus nephritis. Contribution of renal histologic data. Am J Med, 75, 382–91. Balow, J. E. (2005). Clinical presentation and monitoring of lupus nephritis. Lupus, 14(1), 25–30. Banner, M. P. (1987). Genitourinary complications of inflammatory bowel disease. Radiol Clin North Am, 25, 199–209. Baughman, R. P. (2002). Therapeutic options for sarcoidosis: new and old. Curr Opin Pulm Med, 8, 464–9. Baughman, R. P., Drent, M., Kavuru, M., et al. (2006). Infliximab therapy in patients with chronic sarcoidosis and pulmonary involvement. Am J Respir Crit Care Med, 174, 795–802. Baughman, R. and Lower, E. (2001). Infliximab for refractory sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis, 18, 70–4. Baughman, R. P., Lower, E. E., and du Bois, R. (2003). Sarcoidosis. Lancet, 361, 1111–18. Baughman, R., Teirstein, A., Judson, M., et al. (2001). Clinical characteristics of patients in a case control study of sarcoidosis. Am J Respir Crit Care Med, 164, 1885–9. Benigni, A., Caroli, C., Longaretti, L., et al. (2007). Involvement of renal tubular Toll-like receptor 9 in the development of tubulointerstitial injury in systemic lupus. Arthritis Rheum, 56(5), 1569–78. Bergner, R., Hoffmann, M., Waldherr, R., et al. (2003). Frequency of kidney disease in chronic sarcoidosis. Sarcoidosis Vasc Diffuse Lung Dis, 20, 126–32. Berliner, A. R., Haas, M., and Choi, M. J. (2006). Sarcoidosis: the nephrologist’s perspective. Am J Kidney Dis, 48(5), 856–70. Bossini, N., Savoldi, S., Franceschini, F., et al. (2001). Clinical and morphological features of kidney involvement in primary Sjogren’s syndrome. Nephrol Dial Transplant, 16, 2328–36. Brause, M., Magnusson, K., Degenhardt, S., et al. (2002). Renal involvement in sarcoidosis–a report of 6 cases. Clin Nephrol, 57, 142–8. Casella, F. and Allon, M. (1993). The kidney in sarcoidosis. J Am Soc Nephrol, 3, 1555–62. Chang, A., Henderson, S. G., Brandt, D., et al. (2011). In situ B cell-mediated immune responses and tubulointerstitial inflammation in human lupus nephritis. J Immunol, 186(3), 1849–60. Chari, S. T. (2007). Current concepts in the treatment of autoimmune pancreatitis. JOP, 8, 1–3. Cheuk, W. and Chan, J. K. (2010). IgG4-related sclerosing disease: a critical appraisal of an evolving clinicopathologic entity. Adv Anatom Pathol, 17, 303–32. Cigni, A., Soro, G., Faedda, R., et al. (2003). A case of adult-onset tubulointerstitial nephritis and uveitis (‘TINU syndrome’) associated with sacroileitis and Epstein-Barr virus infection with good spontaneous outcome. Am J Kidney Dis, 42, E4–10. Cornell, L. D., Chicano, S. L., Deshpande, V., et al. (2007). Pseudotumors due to IgG4 immune-complex tubulointerstitial nephritis associated with autoimmune pancreatocentric disease. Am J Surg Pathol, 31, 1586–97. Danese, S., Semeraro, S., Papa, A., et al. (2005). Extraintestinal manifestations in inflammatory bowel disease. World J Gastroenterol, 14, 227–36. Deshpande, V., Chicano, S., Finkelberg, D., et al. (2006). Autoimmune pancreatitis: a systemic immune complex mediated disease. AmJ Surg Pathol, 30, 1537–45. Dobrin, R. S., Vernier, R. L., and Fish, F. J. (1975). Acute eosinophilic interstitial nephritis and renal failure with bone marrow-lymph node granulomas and anterior uveitis. Am J Med, 596, 325–33. Dusek, J., Urbanova, I., Stejskal, J., et al. (2008). Tubulointerstitial nephritis and uveitis syndrome in a mother and her son. Pediatr Nephrol, 23, 2091–3. Ectors, N., Maillet, B., Aerts, R., et al. (1997). Non-alcoholic duct destructive chronic pancreatitis. Gut, 41, 263–8.

733

734

Section 4  

the patient with interstitial disease

Enestrom, S., Denneberg, T., and Eriksson, P. (1995). Histopathology of renal biopsies with correlation to clinical findings in primary Sjögren’s syndrome. Clin Exp Rheumatol, 13, 697–703. Eriksson, P., Denneberg, T., Larsson, L., et al. (1995). Biochemical markers of renal disease in primary Sjogren’s syndrome. Scand J Urol Nephrol, 29, 383–92. Esdaile, J. M., Levinton, C., Federgreen, W., et al. (1989). The clinical and renal biopsy predictors of long-term outcome in lupus nephritis: a study of 87 patients and review of the literature. QJM, 72, 779–833. Farge, D., Liote, F., Turner, M., et al. (1986). Granulomatous interstitial and chronic renal failure in sarcoidosis: long-term follow-up studies in two patients. Am J Nephrol, 6, 21–7. Filiopoulos, V., Trompouki, S., Hadjiyannakos, D., et al. (2010). IgA nephropathy in association with Crohn’s disease: a case report and brief review of the literature. Ren Fail, 32(4), 523–7. Fraser, J. S., Muller, A. F., Smith, D. J., et al. (2001). Renal tubular injury is present in acute inflammatory bowel disease prior to the introduction of drug therapy. Aliment Pharmacol Ther, 15, 1131–7. Gianviti, A., Greco, M., Barsotti, P., et al. (1994). Acute tubulointerstitial nephritis occurring with 1-year lapse in identical twins. Pediatr Nephrol, 8, 427–30. Gil, A., Ofer, Y., Yigla ,M., et al. (2010). Obstructive uropathy caused by retroperitoneal fibrosis presents as a late manifestation of sarcoidosis. Ren Fail, 32(9), 1118–20. Göbel, U., Kettritz, R., Schneider, W., et al. (2001). The protean face of renal sarcoidosis. J Am Soc Nephrol, 12, 616–23. Goda, C., Kotake, S., Ichiishi, A., et al. (2005). Clinical features in tubulointerstitial nephritis and uveitis (TINU) syndrome. Am J Ophthalmol, 140, 637–41. Goules, A., Masouridi, S., Tzioufas, A. G., et al. (2000). Clinically significant and biopsy-documented renal involvement in primary Sjogren syndrome. Medicine, 79, 241–9. Gupta, A., Jothy, S., Somerville, P., et al. (2010). Hypocomplementaemic immune complex tubulointerstitial nephritis. NDT Plus, 3, 78–80. Hamano, H., Kawa, S., Horiuchi, A., et al. (2001). High serum IgG4 concentrations in patients with sclerosing pancreatitis. N Engl J Med, 344, 732–8. Hannedouche, T., Grateau, G., Noel, L. H., et al. (1990). Renal granulomatous sarcoidosis: report of six cases. Nephrol Dial Transplant, 5, 18–24. Hayakawa, S., Nakabayashi, K., Karube, M., et al. (2006). Tubulointerstitial immune complex nephritis in a patient with systemic lupus erythematosus: role of peritubular capillaritis with immune complex deposits in the pathogenesis of the tubulointerstitial nephritis. Clin Exp Nephrol, 10(2), 146–51. Herlitz, L. C., Chun, M. J., Stokes, M. B., et al. (2007). Uveitis and acute renal failure. Kidney Int, 72, 1554–7. Herrlinger, K. R., Noftz, M. K., Fellermann, K., et al. (2001). Minimal renal dysfunction in inflammatory bowel disease is related to disease activity but not to 5-ASA use. Aliment Pharmacol Ther, 15, 363–9. Hill, P., Russell, P., Sammartino, C., et al. (2009). Acute kidney injury and proteinuria in a patient with diabetes and a submandibular mass. Am J Kidney Dis, 54, 375–80. Hinkle, D. M. and Foster, C. S. (2008). Tubulointerstitial nephritis and uveitis syndrome. Int Ophthalmol Clin, 48, 9–13. Howarth, L., Gilbert, R. D., Bass, P., et al. (2004). Tubulointerstitial nephritis and uveitis in monozygotic twin boys. Pediatr Nephrol, 19, 917–19. Hsieh, C., Chang, A., Brandt, D., et al. (2011). Predicting outcomes of lupus nephritis with tubulointerstitial inflammation and scarring. Arthritis Care Res (Hoboken), 63(6), 865–74. Iannuzzi, M. C., Rybicki, B. A., and Teirstein, A. S. (2007). Sarcoidosis. N Engl J Med, 357, 2153–65. Igarashi, T., Kawato, H., Kamoshita, S., et al. (1992). Acute tubulointerstitial nephritis with uveitis syndrome presenting as multiple tubular dysfunction including Fanconi’s syndrome. Pediatr Nephrol, 6, 547–9. Ishige, I., Usui, Y., and Takemura, T. (1999). Quantitative PCR of mycobacterial and propionibacterial DNA in lymph nodes of Japanese patients with sarcoidosis. Lancet, 354, 120–3.

Ito, S., Harada, T., Nakamura, T., et al. (2009). Mizoribine for renal sarcoidosis: effective steroid tapering and prevention of recurrence. Pediatr Nephrol, 24(2), 411–14. Izzedine, H., Simon, J., Piette, A. M., et al. (2002). Primary chronic interstitial nephritis in Crohn’s disease. Gastroenterology, 123, 1436–40. Izzedine, H. (2008). Tubulointerstitial nephritis and uveitis syndrome (TINU): a step forward to understanding an elusive oculorenal syndrome. Nephrol Dial Transplant, 23(4), 1095–7. James, D. (1984). Sarcoidosis. Postgrad Med J, 60, 234–41. Johns, C. and Michele, T. (1999). The clinical management of sarcoidosis: a 50 year experience at the Johns Hopkins Hospital. Medicine, 78, 65–111. Jose, F. A., Garnett, E. A., Vittinghoff, E., et al. (2009). Development of extraintestinal manifestations in pediatric patients with inflammatory bowel disease. Inflamm Bowel Dis, 15, 63–8. Joss, N., Morris, S., Young, B., et al. (2007). Granulomatous interstitial nephritis. Clin J Am Soc Nephrol, 2, 222–30. Kambham, N., Markowitz, G. S., Tanji, N., et al. (2001). Idiopathic hypocomplementemic interstitial nephritis with extensive tubulointerstitial deposits. Am J Kidney Dis, 37, 388–99. Kamisawa, T., Funata, N., Hayashi, Y., et al. (2003). A new clinicopathological entity of IgG4-related autoimmune disease. J Gastroenterol, 38, 982–4. Kamouchi, M., Tsuji, H., Hirakata, H., et al. (1991). Tubulointerstitial disorders in the kidney associated with primary biliary cirrhosis (PBC). Clin Nephrol, 35, 134–5. Kase, S., Kitaichi, N., Namba, K., et al. (2006). Elevation of serum Krebs von den Lunge-6 levels in patients with tubulointerstitial nephritis and uveitis syndrome. Am J Kidney Dis, 48, 935–41. Kassan, S. S. and Talal, N. (1987). Renal disease with Sjögren’s syndrome. In N. Talal, H. M. Moutsopoulos, and S. S. Kassan (eds.) Sjögren’s Syndrome: Clinical and Immunological Aspects, pp. 96–102. Berlin: Springer-Verlag. Katano, K., Hayatsu, Y., Matsuda, T., et al. (2007). Endocapillary proliferative glomerulonephritis with crescent formation and concurrent tubulointerstitial nephritis complicating retroperitoneal fibrosis with a high serum level of IgG4. Clin Nephrol, 68, 308–14. Kawaguchi, K., Koike, M., Tsuruta, K., et al. (1991). Lymphoplasmacytic sclerosing pancreatitis with cholangitis: a variant of primary sclerosing cholangitis extensively involving pancreas. Hum Pathol, 22, 387–95. Khan, M. L., Colby, T. V., Viggiano, R. W., et al. (2010). Treatment with bortezomib of a patient having hyper IgG4 disease. Clin Lymphoma Myeloma Leuk, 10, 217–19. Khosroshahi, A., Bloch, D. B., Deshpande, V., et al. (2010). Rituximab therapy leads to rapid decline of serumIgG4 level and prompt clinical improvement in IgG4-related systemic disease. Arthritis Rheum, 62, 1755–62. Khosroshahi, H. T. and Shoja, M. M. (2006). Tubulointerstitial disease and ulcerative colitis. Nephrol Dial Transplant, 21(8), 2340. Kim, F., Yamada, K., Inoue, D., et al. (2011). IgG4-related tubulointerstitial nephritis and hepatic inflammatory pseudotumor without hypocomplementemia. Intern Med, 50(11), 1239–44. Kitagawa, S., Zen, Y., Harada, K., et al. (2005). Abundant IgG4-positive plasma cell infiltration characterizes chronic sclerosing sialadenitis (Kuttner’s tumor). Am J Surg Pathol, 29, 783–91. Kobayashi, T., Muto, S., Nemoto, J., et al. (2006). Fanconi’s syndrome and distal (type 1) renal tubular acidosis in a patient with primary Sjogren’s syndrome with monoclonal gammopathy of undetermined significance. Clin Nephrol, 65, 427–32. Kobayashi, Y., Honda, M., Yoshikawa, N., et al. (2000). Acute tubulointerstitial nephritis in 21 Japanese children. Clin Nephrol, 54, 191–7. Kodama, T., Imai, H., Wakui, H., et al. (1996). Tubulointerstitial nephritis with renal tubular acidosis and asymptomatic primary biliary cirrhosis accompanied by antibody to a 52-kDa mitochondrial protein alone. Clin Nephrol, 45, 401–5. Kodner, C. M. and Kudrimoti, A. (2003). Diagnosis and management of acute interstitial nephritis. Am Fam Physician, 67, 2527–34. Komatsuda, A., Wakui, H., Ohtani, H., et al. (2010). Tubulointerstitial nephritis and renal tubular acidosis of different types are rare but

Chapter 93 

important complications of primary biliary cirrhosis. Nephrol Dial Transplant, 25(11), 3575–9. Kreisel, W., Wolf, L. M., Grotz, W., et al. (1996). Renal tubular damage: an extraintestinal manifestation of chronic inflammatory bowel disease. Eur J Gastroenterol Hepatol, 8, 461–8. Lahita, R. G. (2004). Systemic Lupus Erythematosus. San Diego, CA: Elsevier Academic Press. Lee, T. Y., Kim, M. H., Park do, H., et al. (2009). Utility of 18F-FDG PET/CT for differentiation of autoimmune pancreatitis with atypical pancreatic imaging findings from pancreatic cancer. Am J Roentgenol, 193, 343–8. Levinson, R. D., Park, M. S., Rikkers, S. M., et al. (2003). Strong associations between specific HLA-DQ and HLA-DR alleles and the tubulointerstitial nephritis and uveitis syndrome. Invest Ophthalmol Vis Sci, 44, 653–7. Lin, D. F., Yan, S. M., Zhao, Y., et al. (2010). Clinical and prognostic characteristics of 573 cases of primary Sjögren’s syndrome. Chin Med J (Engl), 123(22), 3252–7. Lino, M., Binaut, R., Noël, L. -H., et al. (2005). Tubulointerstitial nephritis and Fanconi syndrome in primary biliary cirrhosis. Am J Kidney Dis, 46, e41–6. Longcope, W. T. and Freiman, D. G. (1952). A study of sarcoidosis; based on a combined investigation of 160 cases including 30 autopsies from The Johns Hopkins Hospital and Massachusetts General Hospital. Medicine, 31, 1–132. Macdougall, I. C., Isles, C. G., Whiteworth, J. A., et al. (1987). Interstitial nephritis and primary biliary cirrhosis: a new association? Clin Nephrol, 27, 36–40. Mackensen, F. and Billing, H. (2009). Tubulointerstitial nephritis and uveitis syndrome. Curr Opin Ophthalmol, 20(6), 525–31. Mackensen, F., David, F., Schwenger, V., et al. (2008). HLA DRB1_0102, which is associated with tubulointerstitial nephritis and uveitis syndrome (TINU) is also frequent in patients with sudden onset anterior bilateral uveitis but not in patients with tubulointerstitial nephritis alone. Invest Ophthalmol Vis Sci, 49, E-Abstract 813. Mackensen, F., Smith, J. R., and Rosenbaum, J. T. (2007). Enhanced recognition, treatment, and prognosis of tubulointerstitial nephritis and uveitis syndrome. Ophthalmology, 114, 995–9. Mahevas, M., Lescure, F. X., Boffa, J. J., et al. (2009). Renal sarcoidosis: clinical, laboratory, and histologic presentation and outcome in 47 patients. Medicine, 88, 98–106. Mahmud, N., McDonald, G. S., Kelleher, D., et al. (1996). Microalbuminuria correlates with intestinal histopathological grading in patients with inflammatory bowel disease. Gut, 38, 99–103. Mahmud, N., O’Toole, D., O’Hare, N., et al. (2002). Evaluation of renal function following treatment with 5-aminosalicylic acid derivatives in patients with ulcerative colitis. Aliment Pharmacol Ther, 16, 207–15. Mallbris, L., Ljungberg, A., Hedblad, M., et al. (2003). Progressive cutaneous sarcoidosis respondingto anti-tumor necrosis factor-alpha therapy. J Am Acad Dermatol, 48, 290–3. Mandeville, J. T., Levinson, R. D., and Holland, G. N. (2001). The tubulointerstitial nephritis and uveitis syndrome. Surv Ophthalmol, 46, 195–208. Marcus, S. B., Brown, J. B., Melin-Aldana, H., et al. (2008). Tubulointerstitial nephritis: an extraintestinal manifestation of Crohn disease in children. J Pediatr Gastroenterol Nutr, 46, 338–41. Maripuri, S., Grande, J. P., Osborn, T. G., et al. (2009). Renal involvement in primary Sjögren’s syndrome: a clinicopathologic study. Clin J Am Soc Nephrol, 4(9), 1423–31. Masaki, Y., Kurose, N., and Umehara, H. (2011). IgG4-related disease: a novel lymphoproliferative disorder discovered and established in Japan in the 21st century. J Clin Exp Hematop, 51(1), 13–20. Mayer, C., Müller, A., Halbritter, J., et al. (2008). Isolated renal relapse of sarcoidosis under low-dose glucocorticoid therapy. J Gen Intern Med, 23(6), 879–82. Michel, D. M. and Kelly, C. J. (1998). Acute interstitial nephritis. J Am Soc Nephrol, 9, 506–15. Mills, P., Burns, A., Dorman, A., et al. (1994). Granulomatous sarcoid nephritis presenting as frank hematuria. Nephrol Dial Transplant, 9, 1649–51.

immune-mediated tubulointerstitial nephritis

Molino, C., Fabbian, F., and Longhini, C. (2009). Clinical approach to lupus nephritis: recent advances. Eur J Intern Med, 20(5), 447–53. Mori, Y., Kishimoto, N., Yamahara, H., et al. (2005). Predominant tubulointerstitial nephritis in a patient with systemic lupus nephritis. Clin Exp Nephrol, 9, 79–84. Morimoto, J., Hasegawa, Y., Fukushima, H., et al. (2009). Membranoproliferative glomerulonephritis-like glomerular disease and concurrent tubulointerstitial nephritis complicating IgG4-related autoimmune pancreatitis. Intern Med, 48, 157–62. Morino, M., Uryu, T., and Yoshikawa, T. (1983). A case of uveitis associated with tubulointerstitial nephritis. Nippon Jinzo Gakkai Shi, 25, 1027–34. Moudgil, A., Przygodzki, R., and Kher, K. (2006). Successful steroid-sparing treatment of renal limited sarcoidosis with mycophenolate mofetil. Pediatr Nephrol, 21, 281–5. Moutsopoulos, H. M., Cledes, J., Skopouli, F. N., et al. (1991). Nephrocalcinosis in Sjogren’s syndrome: a late sequela of renal tubular acidosis. J Intern Med, 230, 187–91. Moyano Franco, M. J., Amor Sánchez, J., Ortega Ruano, R., et al. (2009). Isolated tubular interstitial nephritis in a patient with systemic lupus erythematosus Nefrologia, 29(5), 501–2. Muther, R. S., McCarron, D. A., and Bennett, W. M. (1981). Renal manifestations of sarcoidosis. Arch Intern Med, 141, 643–5. Nakajo, M., Jinnouchi, S., Fukukura, Y., et al. (2007). The efficacy of whole-body FDG-PET or PET/CT for autoimmune pancreatitis and associated extrapancreatic autoimmune lesions. Eur J Nucl Med Mol Imaging, 34, 2088–95. Nanke, Y., Kobashigawa, T., Yago, T., et al. (2010). A case of Mikulicz’s disease, IgG4-related plasmacytic syndrome, successfully treated by corticosteroid and mizoribine, followed by mizoribine alone. Intern Med, 49, 1449–53. Nath, K. A. (1992). Tubulointerstitial changes as a major determinant in the progression of renal damage. Am J Kidney Dis, 20, 1–17. Neumann, K., Wallace, D. J., Azen, C., et al. (1995). Lupus in the 1980s: III. Influence of clinical variables, biopsy, and treatment on the outcome in 150 patients with lupus nephritis seen at a single center. Semin Arthritis Rheum, 25, 47–55. Newman, L. S, Rose, C. S., and Maier, L. A. (1997). Sarcoidosis. N Engl J Med, 336, 1224–34. Nunes, H., Bouvry, D., Soler, P., et al. (2007). Sarcoidosis. Orphanet J Rare Dis, 2, 46. O’Riordan, E., Willert, R. P., Reeve, R., et al. (2001). Isolated sarcoid granulomatous interstitial nephritis: review of five cases at one center. Clin Nephrol, 55, 297–302. Oikonomou, K., Kapsoritakis, A., Eleftheriadis, T., et al. (2011). Renal manifestations and complications of inflammatory bowel disease. Inflamm Bowel Dis, 17(4), 1034–45. Omokawa, A., Wakui, H., Okuyama, S., et al. (2008). Predominant tubulointerstitial nephritis in a patient with systemic lupus erythematosus: phenotype of infiltrating cells. Clin Nephrol, 69(6), 436–44. Onyekpe, I., Shenoy, M., Denley, H., et al. (2011). Recurrent tubulointerstitial nephritis and uveitis syndrome in a renal transplant recipient. Nephrol Dial Transplant, 26(9), 3060–2. Padilla, M., Schilero, G., and Teirstein, A. (1997). Sarcoidosis and transplantation. Sarcoidosis Vasc Diffuse Lung Dis, 14, 16–22. Pagniez, D. and Delvallez, L. (1989). Renal gallium scintigraphy and sarcoid nephritis. Nephrol Dial Transplant, 4, 76. Pardi, D. S., Tremaine, W. J., Sandborn, W. J., et al. (1998). Renal and urologic complications of inflammatory bowel disease. Am J Gastroenterol, 93, 504–14. Parés, A., Rimole, A., Bruguera, M., et al. (1981). Renal tubular acidosis in primary biliary cirrhosis. Gastroenterology, 80, 681–6. Pettersen, J., Zochodne, D., Bell, R., et al. (2002). Refractory neurosarcoidosis responding to infliximab. Neurology, 59, 1660–1. Pokorny, G., Sonkodi, S., Ivanyi, B., et al. (1989). Renal involvement in patients with primary Sjogren’s syndrome. Scand J Rheumatol, 18, 231–4. Poulou, A. C., Goumas, K. E., Dandakis, D. C., et al. (2006). Microproteinuria in patients with inflammatory bowel disease: is

735

736

Section 4  

the patient with interstitial disease

it associated with the disease activity or the treatment with 5-aminosalicylic acid? World J Gastroenterol, 12, 739–46. Raissian, Y., Nasr, S. H., Larsen, C. P., et al. (2011). Diagnosis of IgG4-related tubulointerstitial nephritis. J Am Soc Nephrol, 22(7), 1343–52. Rajakariar, R., Sharples, E. J., Raftery, M. J., et al. (2006). Sarcoid tubulo-interstitial nephritis: long-term outcome and response to corticosteroid therapy. Kidney Int, 70, 165–9. Ramos-Casals, M., Tzioufas, A. G., Stone, J. H., et al. (2010). Treatment of primary Sjögren syndrome: a systematic review. JAMA, 304(4), 452–60. Rao, P. K. and Sabanegh, E. S. (2009). Genitourinary sarcoidosis. Rev Urol, 11(2), 108–13. Reich, J. R. (2002). Mortality of intrathoracic sarcoidosis in referral vs population-based settings: influence of stage, ethnicity, and corticosteroid therapy. Chest, 121, 32–9. Ren, H., Wang, W. M., Chen, X. N., et al. (2008). Renal involvement and follow-up of 130 patients with primary Sjogren’s syndrome. J Rheumatol, 35, 278–84. Riis, L., Vind, I., Vermeire, S., et al. (2007). The prevalence of genetic and serologic markers in an unselected European population-based cohort of IBD patients. Inflamm Bowel Dis, 13, 24–32. Rizzato, G. (2001). Extrapulmonary presentation of sarcoidosis. Curr Opin Pulm Med, 7, 295–7. Roach, D. R., Bean, A. G., Demangel, C., et al. (2002). TNF regulates chemokine induction essential for cell recruitment, granuloma formation, and clearance of mycobacterial infection. J Immunol, 168, 4620–7. Robson, M. G., Banerjee, D., Hopster, D., et al. (2003). Seven cases of granulomatous interstitial nephritis in the absence of extrarenal sarcoid. Nephrol Dial Transplant, 18, 280–4. Rothfuss, K. S., Stange, E. F., and Herrlinger, K. R. (2006). Extraintestinal manifestations and complications in inflammatory bowel diseases. World J Gastroenterol, 14, 4819–31 Saeki, T., Imai, N., Ito, T., et al. (2009). Membranous nephropathy associated with IgG4-related systemic disease and without autoimmune pancreatitis. Clin Nephrol, 71, 173–8. Saeki, T., Nishi, S., Imai, N., et al. (2010). Clinicopathological characteristics of patients with IgG4-related tubulointerstitial nephritis. Kidney Int, 78, 1016–23. Saeki, T., Nishi, S., Ito, T., et al. (2007). Renal lesions in IgG4-related systemic disease. Intern Med, 46, 1365–71. Saeki, T., Saito, A., Yamazaki, H., et al. (2007). Tubulointerstitial nephritis associated with IgG4-related systemic disease. Clin Exp Nephrol, 11, 168–73. Sarles, H., Sarles, J. C., Muratore, R., et al. (1961). Chronic inflammatory sclerosis of the pancreas—an autonomous pancreatic disease? Am J Dig Dis, 6, 688–98. Schwartz, M., Fennell, J. S., and Lewis, E. J. (1982). Pathologic changes in the renal tubule in systemic lupus erythematosus. Hum Pathol, 13, 534–47. Semenzato, G., Bortoli, M., Brunetta, E., et al. (2005). Immunology and pathophysiology. Eur Respir Monograph, 10, 49–63. Shimazaki, K., Jirawuthiworavong, G. V., Nguyen, E. V., et al. (2008). Tubulointerstitial nephritis and uveitis syndrome: a case with an autoimmune reactivity against retinal and renal antigens. Ocul Immunol Inflamm, 16, 51–3. Shrestha, B., Sekiguchi, H., Colby, T. V., et al. (2009). Distinctive pulmonary histopathology with increased IgG4-positive plasma cells in patients with autoimmune pancreatitis: report of 6 and 12 cases with similar histopathology. Am J Surg Pathol, 33, 1450–62. Siamopoulos, K. C., Elisaf, M., Drosos, A. A., et al. (1992). Renal tubular acidosis in primary Sjogren’s syndrome. Clin Rheumatol, 11, 226–30. Simonsen, O. and Thysell, H. (1985). Sarcoidosis with normocalcemic granulomatous nephritis. Nephron, 40, 411–7. Singer, D. and Evans ,D. (1986). Renal impairment in sarcoidosis: granulomatous nephritis as an isolated cause (two case reports and review of the literature). Clin Nephrol, 26, 250–6.

Singer, F. R. and Adams, J. S. (1986). Abnormal calcium homeostasis in sarcoidosis. N Engl J Med, 315, 755–7. Singh, A. K., Ucci, A., and Madias, N. E. (1996). Predominant tubulointerstitial lupus nephritis. Am J Kidney Dis, 27, 273–8. Skopouli, F. N., Dafni, U., Ioannidis, J. P., et al. (2000). Clinical evolution, and morbidity and mortality of primary Sjogren’s syndrome. Semin Arthritis Rheum, 29, 296–304. Sood, S., Fossard, D. P., and Shorrock, K. (1995). Chronic sclerosing pancreatitis in Sjogren’s syndrome: a case report. Pancreas, 10, 419–21. Stewart Cameron, J. (1999). Lupus nephritis. J Am Soc Nephrol, 10, 413–24. Stupp, R., Mihatsch, M. J., Matter, L., et al. (1990). Acute tubulointerstitial nephritis with uveitis (TINU syndrome) in a patient with serologic evidence for Chlamydia infection. Klin Wochenschr, 68, 971–5. Suzuki, K., Tanaka, H., Ito, E., et al. (2004). Repeat renal biopsy in children with severe idiopathic tubulointerstitial nephritis. Pediatr Nephrol, 19, 240–3. Takahashi, N., Kawashima, A., Fletcher, J. G., et al. (2007). Renal involvement in patients with autoimmune pancreatitis: CT and MR imaging findings. Radiology, 242, 791–801. Takemura, T., Okada, M., Hino, S., et al. (1999). Course and outcome of tubulointerstitial nephritis and uveitis syndrome. Am J Kidney Dis, 34, 1016–21. Talwalkar, J. A. and Lindor, K. D. (2003). Primary biliary cirrhosis. Lancet, 362, 53–61. Tan, Y., Yu, F., Yang, H., et al. (2008). Autoantibodies against monomeric C-reactive protein in sera from patients with lupus nephritis are associated with disease activity and renal tubulointerstitial lesions. Hum Immunol, 69(12), 840–4. Tanaka, H., Suzuki, K., Nakahata, T., et al. (2001a). Repeat renal biopsy in a girl with tubulointerstitial nephritis and uveitis syndrome. Pediatr Nephrol, 16, 885–7. Tanaka, H., Waga, S., Nakahata, T., et al. (2001b). Tubulointerstitial nephritis and uveitis syndrome in two siblings. Tohoku J Exp Med, 194, 71–4. Terrier, B., Fakhouri, F., Berezne, A. et al. (2008). Osteomalacia revealing celiac disease and primary biliary cirrhosis-related Fanconi syndrome in a patient with systemic sclerosis. Clin Exp Rheumatol, 26, 467–70. Thumfart, J., Müller, D., Rudolph, B., et al. (2005). Isolated sarcoid granulomatous interstitial nephritis responding to infliximab therapy. Am J Kidney Dis, 45, 411–14. Toda, T., Kimoto, S., Nishio, Y., et al. (1999). Sarcoidosis with membranous nephropathy and granulomatous interstitial nephritis. Intern Med, 38, 882–6. Topazian, M., Witzig, T. E., Smyrk, T. C., et al. (2008). Rituximab therapy for refractory biliary strictures in immunoglobulin G4-associated cholangitis. Clin Gastroenterol Hepatol, 6, 364–6. Tovbin, D., Kachko, L., and Hilzenrat, N. (2000). Severe interstitial nephritis in a patient with renal amyloidosis and exacerbation of Crohn’s disease. Clin Nephrol, 53, 147–51. Toy, W. C. and Jasin, H. E. (2008). An unusual case of hypokalemic paralysis associated with primary Sjogren’s syndrome. J Ark Med Soc, 104(12), 286–7. Uehara, T., Hamano, H., Kawa, S., et al. (2005). Distinct clinicopathological entity ‘autoimmune pancreatitis-associated sclerosing cholangitis’. Pathol Int, 55, 405–11. Utas, C., Dogukan, A., Patiroglu, T., et al. (1999). Granulomatous interstitial nephritis in extrapulmonary sarcoidosis. Clin Nephrol, 51, 252–4. Valentonyte, R., Hampe, J., Huse, K., et al. (2005). Sarcoidosis is associated with a truncating splice site mutation in BTNL2. Nat Genet, 37, 357–64. Van Leusen, R. and Assmann, K. J. (1988). Acute tubulo-interstitial nephritis with uveitis and favourable outcome after five months of continuous ambulatory peritoneal dialysis (CAPD). Neth J Med, 33, 133–9. Van Uum, S. H., Cooreman, M. P., Assmann, K. J., et al. (1997). A 58-year-old man with sarcoidosis complicated by focal crescentic glomerulonephritis. Nephrol Dial Transplant, 12, 2703–7.

Chapter 93 

Vaseemuddin, M., Schwartz, M. M., Dunea, G., et al. (2007). Idiopathic hypocomplementemic immune-complex-mediated tubulointerstitial nephritis. Nat Clin Pract Nephrol, 3, 50–8. Victorino, R. M., Lucas, M. B., and de Moura, M. C. (1986). Severe osteomalacia associated with renal tubular acidosis in Crohn’s disease. Dig Dis Sci, 31, 322–6. Vohra, S., Eddy, A., Levin, A. V., et al. (1999). Tubulointerstitial nephritis and uveitis in children and adolescents. Four new cases and a review of the literature. Pediatr Nephrol, 13, 426–32. Wakaki, H., Sakamoto, H., and Awazu, M. (2001). Tubulointerstitial nephritis and uveitis syndrome with autoantibody directed to renal tubular cells. Pediatrics, 107, 1443–6. Watson, S. J., Jenkins, D. A., and Bellamy, C. O. (2006). Nephropathy in IgG4-related systemic disease. Am J Surg Pathol, 30, 1472–7. Weening, J. J., D’Agati, V. D., Schwartz, M. M., et al. (2004). The classification of glomerulonephritis in systemic lupus erythematosus revisited. J Am Soc Nephrol, 15, 241–50. Wreesmann, V., van Eijck, C. H., Naus, D. C., et al. (2001). Inflammatory pseudotumour (inflammatory myofibroblastic tumour) of the pancreas: a report of six cases associated with obliterative phlebitis. Histopathology, 38, 105–10.

immune-mediated tubulointerstitial nephritis

Yanagihara, T., Kitamura, H., Aki, K., et al. (2009). Serial renal biopsies in three girls with tubulointerstitial nephritis and uveitis syndrome. Pediatr Nephrol, 24, 1159–64. Yao, Y. H., Lin, C. C., Chung, Y. M., et al. (2011). Tubulointerstitial nephritis and uveitis syndrome (TINU) with Fanconi’s syndrome. Clin Nephrol, 75 Suppl 1, 75–8. Yee, A. and Pochapin, M. (2001). Treatment of complicated sarcoidosis with infliximab anti-tumor necrosis factor-alpha therapy. Ann Intern Med, 135, 27–31. Yoshida, K., Toki, F., Takeuchi, T., et al. (1995). Chronic pancreatitis caused by an autoimmune abnormality. Proposal of the concept of autoimmune pancreatitis. Dig Dis Sci, 40, 1561–8. Yung, S., Tsang, R. C., Sun, Y., et al. (2050). Effect of human anti-DNA antibodies on proximal renal tubular epithelial cell cytokine expression: implications on tubulointerstitial inflammation in lupus nephritis. J Am Soc Nephrol, 16, 3281–94. Zen, Y., Kasahara, Y., Horita, K., et al. (2005). Inflammatory pseudotumor of the breast in a patient with a high serum IgG4 level: histologic similarity to sclerosing pancreatitis. Am J Surg Pathol, 29, 275–8. Zhang, L. and Smyrk, T. C. (2010). Autoimmune pancreatitis and IgG4-related systemic diseases. Int J Clin Exp Pathol, 3(5), 491–504.

737

SECTION 5

The patient with reduced renal function

94 Chronic kidney disease: definition, classification, and approach to management  743 Ashish Upadhyay, Lesley A. Inker, and Andrew S. Levey

95 Chronic kidney disease in the developed world  755 Morgan E. Grams and Josef Coresh

96 Chronic kidney disease in developing countries  762 Luxia Zhang and Haiyan Wang

97 Chronic kidney disease long-term outcomes: progression, death, cardiovascular disease, infections, and hospitalizations  767 Monica Beaulieu, Catherine Weber, Nadia Zalunardo, and Adeera Levin

98 Cardiovascular disease and chronic kidney disease: overview  777 David J. Goldsmith

99 Recommendations for management of high renal risk chronic kidney disease  780 Charles J. Ferro and Khai Ping Ng

100 Hypertension as a cause of chronic kidney disease: what is the evidence?  787 Teena Tandon and Rajiv Agarwal

101 Diet and the progression of chronic kidney disease  793 Juan Jesús Carrero, Hong Xu, and Bengt Lindholm

102 Lipid disorders of patients with chronic kidney disease  800 Alan G. Jardine and Rajan K. Patel

103 Smoking in chronic kidney disease  807 Stephan R. Orth

104 Analytical aspects of measurements and laboratory values in kidney disease  813 Edmund J. Lamb and Finlay MacKenzie

105 Effect of lifestyle modifications on patients with chronic kidney disease  824 Andrew D. Williams, Robert G. Fassett, Erin J. Howden, and Jeff S. Coombes

106 Malnutrition, obesity, and undernutrition in chronic kidney disease  830 Navneet Kumar, Heather Henderson, Beverly D. Cameron, and Peter A. McCullough

107 Left ventricular hypertrophy in chronic kidney disease  837 Carmine Zoccali, Davide Bolignano, and Francesca Mallamaci

108 Sudden cardiac death in chronic kidney disease  853 Reza Hajhosseiny, Kaivan Khavandi, and David J. Goldsmith

109 Epidemiology of calcium, phosphate, and parathyroid hormone disturbances in chronic kidney disease  869 Patrick Biggar, Hansjörg Rothe, and Markus Ketteler

740

110 The role of inflammation in chronic kidney disease  877 Juan Jesús Carrero and Peter Stenvinkel

111 Vascular stiffness in chronic kidney disease: pathophysiology and implications  884 Jonathan N. Townend and Charles J. Ferro

112 Oxidative stress and its implications in chronic kidney disease  895 Nosratola D. Vaziri

113 Abnormal endothelial vasomotor and secretory function  903 Thimoteus Speer and Danilo Fliser

114 Endothelins and their antagonists in chronic kidney disease  910 Neeraj Dhaun and David J. Webb

115 Chronic kidney disease-mineral and bone disorder: overview  916 Stuart M. Sprague and Menaka Sarav

116 Imaging for detection of vascular disease in chronic kidney disease patients  923 Paolo Raggi and Luis D’Marco

117 Pathophysiology of chronic kidney disease-mineral and bone disorder  934 Alexandra Voinescu, Nadia Wasi Iqbal, and Kevin J. Martin

118 Management of chronic kidney disease-mineral and bone disorder  939 Alexandra Voinescu, Nadia Wasi Iqbal, and Kevin J. Martin

119 Fibroblast growth factor 23, Klotho, and phosphorus metabolism in chronic kidney disease  947 Orlando M. Gutiérrez

120 Vascular calcification  957 Adrian Covic, Mugurel Apetrii, Luminita Voroneanu, and David J. Goldsmith

121 Fractures in patients with chronic kidney disease  970 Alastair J. Hutchison and Michael L. Picton

122 Spectrum of bone pathologies in chronic kidney disease  976 Stuart M. Sprague and James M. Pullman

123 Clinical aspects and overview of renal anaemia  983 Iain C. Macdougall

124 Erythropoiesis-stimulating agents in chronic kidney disease  991 Iain C. Macdougall

125 Iron metabolism in chronic kidney disease  998 Jolanta Malyszko and Iain C. Macdougall

126 Iron management in renal anaemia  1010 Iain C. Macdougall

127 Pleiotropic effects of vitamin D  1016 Kaivan Khavandi, Halima Amer, Sarah Withers, and Behdad Afzali

128 Immunity 1038 Behdad Afzali and Claudia Kemper

129 The epidemiology of hepatitis viruses in chronic kidney disease  1049 Fabrizio Fabrizi and Michel Jadoul

130 Gastroenterology and renal medicine  1052 Rishi M. Goel, Kamal V. Patel, and Terry Wong

131 Cutaneous manifestations of end-stage renal disease  1064 Timur A. Galperin, Kieron S. Leslie, and Antonia J. Cronin

132 The patient with reduced renal function: endocrinology  1072 Tomas Thor Agustsson and Paul Carroll

133 Sexual dysfunction  1091 Tomas Thor Agustsson and Paul Carroll

134 Health-related quality of life and the patient with chronic kidney disease  1099 Fredric O. Finkelstein and Susan H. Finkelstein

135 Coagulopathies in chronic kidney disease  1102 Seema Shrivastava, Beverley J. Hunt, and Anthony Dorling

136 Mechanisms of progression of chronic kidney disease: overview  1105 Neil Turner

137 Proteinuria as a direct cause of progression  1107 Jeremy Hughes

138 Nephron numbers and hyperfiltration as drivers of progression  1112 Valerie A. Luyckx

139 Podocyte loss as a common pathway to chronic kidney disease  1118 Wilhelm Kriz

140 Disordered scarring and failure of repair  1126 Jeremy S. Duffield

741

141 Modality selection for renal replacement therapy  1136 Raj Thuraisingham and Cormac Breen

142 Patient education and involvement in pre-dialysis management  1142 Sue Cox and Nicola Thomas

143 Preparation for renal replacement therapy  1148 Muh Geot Wong, Bruce A. Cooper, and Carol A. Pollock

144 Choices and considerations for in-centre versus home-based renal replacement therapy  1157 Karthik K. Tennankore and Christopher T. Chan

145 Conservative care in advanced chronic kidney disease  1165 Aine Burns and Fliss E. M. Murtagh

146 Palliative care in end-stage renal disease  1180 Katie Vinen, Fliss E. M. Murtagh, and Irene J. Higginson

147 Patient selection when resources are limited  1187 Bernadette A. Thomas and Christopher R. Blagg

148 Acidosis in chronic kidney disease  1192 Muhammad M. Yaqoob

CHAPTER 94

Chronic kidney disease: definition, classification, and approach to management Ashish Upadhyay, Lesley A. Inker, and Andrew S. Levey Introduction Chronic kidney disease (CKD) is an important global public health problem (Levey et al., 2007). More than 2 million people worldwide are estimated to be receiving treatment with dialysis or transplantation for chronic kidney failure, and this population has been growing at an approximate rate of 7% per year (Lysaght, 2002). However, poor outcomes from CKD are not limited to kidney failure but also include a wide array of morbidity and mortality related to complications, particularly from decreased kidney function and cardiovascular diseases (CVD). The National Kidney Foundation (NKF) sponsored the Kidney Disease Outcomes Quality Initiative (KDOQI) clinical practice guidelines in 2002, which described the conceptual model, definition, and classification of CKD (Kidney Disease Outcomes Quality Initiative (K/DOQI), 2002). These guidelines were subsequently adopted with minor modifications by the international guideline group Kidney Disease Improving Global Outcomes (KDIGO) in 2004 (Levey et  al., 2005). The CKD guidelines represented a fundamental paradigm shift, from viewing kidney disease as a life-threatening condition affecting few people requiring care by nephrologists, to a common condition meriting attention by general internists, and requiring a concerted public health approach for prevention, early detection, and management (Rettig et al., 2008; Levey et al., 2009a). In less than a decade, the guidelines have had a major effect on clinical practice, research, and public health, but have also generated substantial controversy (Eckardt et  al., 2009; James et  al., 2010; Levey and Coresh, 2012). In 2009, KDIGO held a controversies conference to re-examine the CKD definition and classification. Participants at this conference reached a consensus to retain the 2002 KDOQI definition of CKD, but recommended including the cause of CKD and the level of albuminuria in the revised classification system (Levey et  al., 2011). Based on these recommendations, KDIGO recently updated the 2002 KDOQI guidelines in 2012 (Kidney Disease:  Improving Global Outcomes (KDIGO) CKD Work Group, 2013). The goal of this chapter is to review the current conceptual model and definition, current and proposed classification systems, the rationale for the definition and classification, as well as approach to the care of patients with CKD.

Definition and classification of chronic kidney disease Conceptual model of CKD Fig. 94.1 shows the KDOQI conceptual model for the development, progression, and complications of CKD (Levey et al., 2009a; Kidney Disease Outcomes Quality Initiative (K/DOQI), 2002). This model describes the natural history of CKD. Kidney failure is identified as the end stage of CKD, which is preceded by the stages of decreased glomerular filtration rate (GFR), kidney damage, and antecedent conditions associated with higher risk for developing CKD. The model suggests that kidney disease worsens over time by transitioning through a defined sequence of stages, regardless of the cause and rate of progression through each stage. Thus, it should be possible to detect CKD prior to kidney failure by testing for markers of kidney damage and estimating the level of GFR. The horizontal arrows pointing from left to right emphasize the progressive nature of CKD. However, the rate of progression is variable and not all CKD progresses; thus, a diagnosis of CKD does not equate with eventual development of kidney failure. Interventions in earlier stages may slow or prevent the progression to later stages. Early stages of kidney disease may be reversible, and individuals with kidney failure can revert to earlier stages through kidney transplantation, shown as dashed arrowheads pointing from right to left. Recent studies suggest that CKD is a risk factor for development of acute kidney injury (AKI), and that episodes of AKI may increase the risk for progression of CKD (Ishani et al., 2009; Lo et al., 2009; Pannu et al., 2011). The model also highlights that adverse outcomes of CKD are not limited to progressive decline in kidney function. Other complications include metabolic and endocrine complications of decreased GFR, such as anaemia, bone and mineral disorders, malnutrition and neuropathy, representing mild forms of uraemic manifestations, nephrotic syndrome in patients with marked albuminuria, and CVD, often leading to death. More recently recognized complications are threats to patient safety from systemic toxicity and nephrotoxicity due to drugs and procedures, infections, and impaired cognitive and physical function. These complications may arise at earlier stages and patients with CKD can suffer from complications of CKD without progression to kidney failure. Strategies for prevention, early detection, and treatment of CKD complications

744

Section 5  

the patient with reduced renal function

Complications

Normal Screening for CKD risk factors

Increased risk CKD risk reduction; screening for CKD

Damage

↓ GFR

Kidney failure

Death

Estimate Diagnosis Replacement and treatment; progression; by dialysis treat treat and transplant comorbid complications; prepare for conditions; replacement slow progression

Fig. 94.1  Conceptual model for chronic kidney disease. This diagram presents the continuum of development, progression, and complications of chronic kidney disease (CKD) and strategies to improve outcomes. Green circles represent stages of CKD; aqua circles represent potential antecedents of CKD; lavender circles represent consequences of CKD; and thick arrows between circles represent the development, progression, and remission of CKD. ‘Complications’ refers to all complications of CKD, including complications of decreased GFR and cardiovascular disease. Complications may also arise from adverse effects of interventions to prevent or treat the disease. The horizontal arrows pointing from left to right emphasize the progressive nature of CKD. Dashed arrowheads pointing from right to left signify that remission is less frequent than progression. Modified from American Journal of Kidney Diseases (Kidney Disease Outcomes Quality Initiative (K/DOQI), 2002; Levey et al., 2009a).

may prolong survival and improve quality of life even if there is no effect on kidney disease progression. Finally, the conceptual model also identifies a population at increased risk for developing CKD. Attributes which differentiate the population at higher risk from the population at lower risk are defined as ‘risk factors’ for development of CKD. Increased risk can arise from exposure to factors that cause kidney damage, such as hypertension or diabetes, or increased susceptibility to kidney damage, such as older age or reduced nephron mass. Some risk factors may be modifiable and, in principle, detection and modification of these risk factors could delay or prevent the development of CKD.

Definition of chronic kidney disease: diagnostic criteria and their rationale Definition CKD is defined as the presence of kidney damage or GFR < 60 mL/min/1.73 m2 (GFR in mL/min/1.73 m2 may be converted to mL/s/1.73 m2 by multiplying by 0.01667) for ≥ 3  months, irrespective of cause (Table 94.1) (Kidney Disease Outcomes Quality Initiative (K/DOQI), 2002; Levey et al., 2005). An important aspect of the definition is that the criteria are objective and can be ascertained by simple laboratory tests, and can be irrespective of cause. Kidney failure is defined as either (a) GFR < 15 mL/min/1.73 m2 (which in most cases will be accompanied by signs and symptoms of uraemia) or (b)  a need to start kidney replacement therapy (dialysis or transplantation). Kidney failure is not synonymous with end-stage renal disease (ESRD), the administrative term in the United States and elsewhere that indicates treatment by dialysis or transplantation. The term ESRD does not include patients with kidney failure who are not treated with dialysis and transplantation.

Rationale for kidney damage as a diagnostic criterion for the definition of CKD Presence of kidney damage qualifies for the definition of CKD even in the absence of low GFR as kidney damage portends a poor

prognosis for the major outcomes related to CKD. Ascertainment of damage is usually made without kidney biopsy. Because most kidney disease is due to diabetes or hypertension, persistent proteinuria or albuminuria is the principal marker. Epidemiologic studies in diverse populations have shown graded relations between higher albuminuria and mortality and kidney outcomes, in addition to, and independent of, low GFR and risk factors for CVD (Chronic Kidney Disease Prognosis Consortium et al., 2010; Hemmelgarn et al., 2010; Astor et al., 2011; Gansevoort et al., 2011). In addition, albuminuria has also been shown to be independently associated with other CKD complications like anaemia, acidosis, hypoalbuminaemia, hyperparathyroidism, and hypertension (Inker et  al., 2011). The generally accepted threshold for albuminuria as a marker of kidney damage is 30 mg/day, roughly equivalent to a urinary albumin to creatinine ratio of > 30 mg/g or > 3 mg/mmol. Apart from albuminuria and proteinuria, abnormalities in urine sediment (e.g. presence of red blood cells, white blood cells, tubular cells, or casts), imaging studies (e.g. hydronephrosis, asymmetry in kidney size, polycystic kidney disease, small echogenic kidneys), and blood and urine chemistry measurements (those related to altered tubular function, such as renal tubular acidosis) may reflect kidney damage and can fulfil the criterion for kidney damage.

Rationale for decreased GFR as a diagnostic criterion for the definition of CKD Reduced kidney function, specifically a GFR < 60 mL/min/1.73 m2, is also defined as CKD. The level of GFR is usually accepted as the best overall index of kidney function in health and disease (Smith, 1937). The normal level of GFR varies according to age, sex, and body size. GFR in healthy young adults is approximately 120–130 mL/min/1.73 m2 and declines by approximately 1 mL/min/1.73 m2 per year after the third decade (Davies and Shock, 1950; Lindeman et al., 1985). Thus, a GFR level of < 60 mL/min/1.73 m2 represents the loss of half or more of the adult level of normal kidney function. More than 25% of individuals aged 70 years and older have a GFR < 60 mL/min/1.73 m2 in the United States, which may reflect intrinsic

Table 94.1  Criteria for definition of chronic kidney disease Criteria

Comment

Duration ≥ 3 months, based on documentation or inference

Duration is necessary to distinguish chronic from acute kidney diseases: ◆ Clinical evaluation can often suggest duration ◆ Documentation of duration is usually not available in epidemiologic studies

Glomerular filtration rate (GFR) 300 mg/day correspond to microalbuminuria and macroalbuminuria, respectively. Normal urine contains small amounts of albumin, low molecular weight serum proteins, and proteins derived from renal tubules and the lower urinary tract. In most kidney diseases, albumin is the predominant urine protein, comprising approximately 60% of urine total protein when total protein is very high. Values corresponding to normal, high-normal, high, very high and nephrotic range total protein are approximately < 50, 20–50, 50–500, > 500, and > 3500 mg/day. Reproduced from The Lancet (Levey and Coresh, 2012).

degenerative processes or the high prevalence of systemic vascular diseases that affect the kidney (Coresh et al., 2007). The definition of decreased GFR as a criterion for CKD does not vary with age. Whatever its cause, a GFR < 60 mL/min/1.73 m2 in the elderly is an independent predictor of adverse outcomes such as death and CVD, and is associated with an increased prevalence of systemic complications (Chronic Kidney Disease Prognosis Consortium et al., 2010). Similar to younger patients, adjustment of drug doses is required in the elderly with lower GFR.

Classification of chronic kidney disease The NKF/KDOQI classification of CKD is based solely on the severity of the disease as indicated by the level of GFR, with higher stages representing lower GFR levels: GFR > 90 mL/min/1.73 m2 (stage 1), 60–89 mL/min/1.73 m2 (stage 2), 30–59 mL/min/1.73 m2 (stage 3), 15–29 mL/min/1.73 m2 (stage 4), and < 15 mL/ min/1.73 m2 (stage 5) (Kidney Disease Outcomes Quality Initiative (K/DOQI), 2002). Patients receiving treatment with dialysis are subclassified as GFR stage 5D to highlight the specialized care required for dialysis. As growing volume of evidence suggested an important role of albuminuria in the pathogenesis of disease progression and complications, the KDIGO sponsored an international conference in 2009 to examine the relationship of GFR and albuminuria to mortality and kidney outcomes (Levey et al., 2011). On the basis of data from 45 cohorts with > 1.5 million participants, the KDIGO conference recommended to modify the KDOQI classification by adding albuminuria stages, subdivision of stage 3, and the underlying cause of CKD (Levey et al., 2011). Based on these recommendations, KDIGO recently updated the 2002 KDOQI classification system (Tables 94.2, 94.3, and 94.4, ) (Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group, 2013). The combination of GFR, albuminuria and cause will help to classify patients according to their risk for CKD complications (Table 94.5).

Use of GFR and albuminuria categories to assess for future complications Both lower GFR and higher albuminuria are independently associated with a higher risk for all of the major CKD complications, including all-cause mortality, cardiovascular mortality, progressive

kidney disease, kidney failure, and AKI (Figs 94.2 and 94.3) (Levey et  al., 2011). The KDIGO conference put forth a ‘heat map’ that combines GFR and albuminuria categories into four groups based on a composite of risks for the five outcomes: low (green), moderate (yellow), high (orange), and very high (red) (Fig. 94.4) (Levey et al., 2011).

Use of GFR and albuminuria categories to assess for concurrent complications Concurrent complications of decreased GFR include anaemia, acidosis, bone and mineral disorders, malnutrition, hypertension, neuropathy, and decreased quality of life (Kidney Disease Outcomes Quality Initiative (K/DOQI), 2002). As shown in Fig. 94.5, the burden of complications is especially high in CKD GFR stages 4–5 (GFR < 30 mL/min/1.73 m2) (Inker et al., 2011). A recent analysis of data from the United States-based National Health and Nutrition Examination Surveys (NHANES) showed a minimal association between higher levels of albuminuria and a higher prevalence of anaemia, hypoalbuminaemia, acidosis, hypertension, and hyperparathyroidism (Inker et  al., 2011). Higher albuminuria was not associated with hyperphosphataemia. Lower estimated GFR, on the other hand, was strongly associated with all six of these complications. Susceptibility to side effects of medications or diagnostic and therapeutic procedures, such as imaging studies, are increased at lower GFR, and are not known to vary by the level of albuminuria.

Approach to the care of patients with chronic kidney disease CKD care is directed by the cause of CKD and the levels of GFR and albuminuria. Identification of the cause may allow for a ‘specific’ therapy directed at the underlying pathologic processes. Thereafter, staging based on GFR and albuminuria can be used to guide ‘non-specific’ therapies to slow progression and reduce complications (Tables 94.2 and 94.3).

Evaluation Fig. 94.6 provides a five-step guide for the detection and evaluation of CKD (Levey and Coresh, 2012). During routine health

Table 94.2  Categories of CKD by the level of GFR Category

GFR levels (mL/ min/1.73 m2)

Terms

Clinical action plan

G1a

> 90

Normal or high

Diagnose and treat the cause Treat comorbid conditions Evaluate for CKD risk factors Start measures to slow CKD progression Start measures to reduce CVD risk

G2a

60–89

Mildly decreasedb

Estimate progression

G3a

45–59

Mildly to moderately decreased

G3b

30–44

Moderately to severely decreased

Adjust medication dosages as indicated Evaluate and treat complications

G4

15–29

Severely decreased

Prepare for kidney replacement therapy (transplantation and/or dialysis)

G5

< 15

Kidney failure (add D if treated by dialysis)

Start kidney replacement therapy (if uraemia present)

a GFR stages G1 or G2 without markers of kidney damage do not fulfil the criteria for CKD. b Relative to young adult level.

GFR in mL/min/1.73 m2 may be converted to mL/s/1.73 m2 by multiplying by 0.01667. CVD = cardiovascular diseases; GFR = glomerular filtration rate.

Table 94.3  Categories of CKD by the level of albuminuria Category

AER (mg/day)

Approximately equivalent ACR (mg/mmol)

(mg/g)

Terms

Clinical action plan

A1

< 30

105 eGFR 90–105 eGFR 75–90 eGFR 60–75 eGFR 45–60 eGFR 30–45 eGFR 15–30

Summary of relative risks from categorical meta-analysis (dipstick included [–, ±, +, ≥++])

Kidney failure (ESRD) ACR ACR ACR 105 eGFR 90–105 eGFR 75–90 eGFR 60–75 eGFR 45–60 eGFR 30–45 eGFR 15–30

Ref

Ref

7.8

18

Ref

Ref

11

20

Ref

Ref

3.8

48

Ref

Ref

7.4

67

5.2

22

40

147

56

74

294

763

433

1044

1056

2286

Cardiovascular mortality

ACR 105 eGFR 90–105 eGFR 75–90 eGFR 60–75 eGFR 45–60 eGFR 30–45 eGFR 15–30

ACR ≥300

Ref

Ref

2.7

8.4

Ref

Ref

2.4

5.8

Ref

Ref

2.5

4.1

Ref

Ref

3.3

6.4

2.2

4.9

6.4

5.9

7.3

10

12

20

17

17

21

29

ACR 300

(mg/mmol)

< 15

15–50

> 50

(mg/g)

< 150

150–500

> 500

Negative to +

+ or greater

ACR

PCR

Protein reagent strip Negative to trace

ACR = albumin-to-creatinine ratio; AER = albumin excretion rate; PCR = protein-to-creatinine ratio; PER = protein excretion rate.

Fig. 94.6  Detection and evaluation of chronic kidney disease. CVD = cardiovascular disease; GFR = glomerular filtration rate; HBV = hepatitis B virus; HCV = hepatitis C virus; HIV = human immunodeficiency virus; NSAID = non-steroidal anti-inflammatory drugs.

Table 94.7  Criteria for nephrology referral Guideline group

Referral criteria

National Kidney Foundation/Kidney Disease Outcomes Quality Initiative (NKF/KDOQI) (2002, 2004)

Age  30 mg/mmol Haematuria not secondary to urological conditions Inability to identify a presumed cause of CKD Increased risk for progression of kidney disease GFR decline > 30% within 4 months without explanation Difficult to manage complications of CKD such as anaemia secondary to CKD requiring erythropoietin stimulating therapy, or abnormalities of bone and mineral metabolism requiring phosphorus binders or vitamin D preparations Hyperkalaemia (serum potassium concentration > 5.5 mEq/L) Resistant hypertension (BP > 130/80 mm Hg despite adherence to a three-drug antihypertensive regimen that includes a diuretic)a Difficult-to-manage complications of blood pressure lowering agentsa

National Institute for Health and Clinical Excellence (NICE) (2008)

eGFR < 30 mL/min/1.73 m2 Proteinuria (ACR ≥ 70 mg/mmol, PCR ≥ 100 mg/mmol, or 24-hour protein excretion ≥ 1g) unless due to diabetes and already treated Proteinuria (ACR ≥ 30 mg/mmol, PCR ≥ 50 mg/mmol, or 24-hour protein excretion ≥0.5g) with haematuria Rapidly declining eGFR (>5 mL/min/1.73m2 per year, or >10 mL/min/1.73m2 per 5 years) People with, or suspected of having, rare or genetic causes of CKD CKD with renal artery stenosis

Caring for Australasians with Renal Impairment (CARI) (Levin et al., 2008; Johnson, 2011)

eGFR 5 mL/min/1.73 m2 per 6 months) CKD with difficult to control blood pressure with at least 3 agents CKD with unexplained anaemia (Haemoglobin < 10 g/dL)

(Continued)

Chapter 94 

ckd: definitions and overview

Table 94.7 Continued Guideline group

Referral criteria

Canadian Society of Nephrology (Levin et al., 2008)

Persistent eGFR < 30 mL/min/1.73 m2 Progressive decline in kidney function Proteinuria (ACR ≥ 60 mg/mmol, PCR ≥ 100 mg/mmol, 24-hour protein excretion ≥ 900mg, 24-hour albumin excretion ≥ 500mg) Inability to achieve treatment targets Rapid changes in kidney function

Japanese Society of Nephrology (2009)

eGFR < 50 mL/min/1.73 m2 Proteinuria (PCR > 0.5 g/g or 2+ by dipstick) Dipstick positive for proteinuria and haematuria Rapidly progressive kidney disease CKD stages 1–3 with difficult to control blood pressure

a Kidney disease or hypertension specialist.

ACR = albumin-to-creatinine ratio; eGFR = estimated glomerular filtration rate; PCR = protein-to-creatinine ratio.

References American Diabetes Association (2011). Standards of medical care in diabetes—2011. Diabetes Care, 34 Suppl 1, S11–61. Astor, B. C., Matsushita, K., Gansevoort, R. T., et al. (2011). Lower estimated glomerular filtration rate and higher albuminuria are associated with mortality and end-stage renal disease. A collaborative meta-analysis of kidney disease population cohorts. Kidney Int, 79, 1331–40. Chobanian, A. V., Bakris, G. L., Black, H. R., et al. (2003). The Seventh Report of the Joint National Committee on Prevention, Detection, Evaluation, and Treatment of High Blood Pressure: the JNC 7 report. JAMA, 289, 2560–72. Chronic Kidney Disease Prognosis Consortium, Matsushita, K., Van Der Velde, M., et al. (2010). Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet, 375, 2073–81. Coresh, J., Selvin, E., Stevens, L. A., et al. (2007). Prevalence of chronic kidney disease in the United States. JAMA, 298, 2038–47. Davies, D. F. and Shock, N. W. (1950). Age changes in glomerular filtration rate, effective renal plasma flow, and tubular excretory capacity in adult males. J Clin Invest, 29, 496–507. Earley, A., Miskulin, D., Lamb, E. J., et al. (2012). Estimating equations for glomerular filtration rate in the era of creatinine standardization: a systematic review. Ann Intern Med, 156(11), 785–95. Eckardt, K. U., Berns, J. S., Rocco, M. V., et al. (2009). Definition and classification of CKD: the debate should be about patient prognosis—a position statement from KDOQI and KDIGO. Am J Kidney Dis, 53, 915–20. Gansevoort, R. T., Matsushita, K., Van der Velde, M., et al. (2011). Lower estimated GFR and higher albuminuria are associated with adverse kidney outcomes in both general and high-risk populations. A collaborative meta-analysis of general and high-risk population cohorts. Kidney Int, 80, 93–104. Ginsberg, J. M., Chang, B. S., Matarese, R. A., et al. (1983). Use of single voided urine samples to estimate quantitative proteinuria. N Engl J Med, 309, 1543–6. Gupta, S. K., Eustace, J. A., Winston, J. A., et al. (2005). Guidelines for the management of chronic kidney disease in HIV-infected patients: recommendations of the HIV Medicine Association of the Infectious Diseases Society of America. Clin Infect Dis, 40, 1559–85. Hemmelgarn, B. R., Manns, B. J., Lloyd, A., et al. (2010). Relation between kidney function, proteinuria, and adverse outcomes. JAMA, 303, 423–9. Inker, L. A., Coresh, J., Levey, A. S., et al. (2011). Estimated GFR, albuminuria, and complications of chronic kidney disease. J Am Soc Nephrol, 22(12), 2322–31.

Inker, L. A., Schmid, C. H., Tighiouart, H., et al. (2012). Estimating glomerular filtration rate from serum creatinine and cystatin C. N Engl J Med, 367, 20–9. Ishani, A., Xue, J. L., Himmelfarb, J., et al. (2009). Acute kidney injury increases risk of ESRD among elderly. J Am Soc Nephrol, 20, 223–8. James, M. T., Hemmelgarn, B. R., and Tonelli, M. (2010). Early recognition and prevention of chronic kidney disease. Lancet, 375, 1296–309. Japanese Society of Nephrology (2009). Chapter 14. Timing for referral of CKD patients to nephrologists. Clin Exp Nephrol, 13, 222–3. Johnson, D. (2011). When to refer for Specialist Renal Care—The CARI Guidelines Caring for Australasians with Renal Impairment [Online]

Kidney Disease: Improving Global Outcomes (KDIGO) (2008). KDIGO clinical practice guidelines for the prevention, diagnosis, evaluation, and treatment of hepatitis C in chronic kidney disease. Kidney Int Suppl, S1–99. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group (2013). KDIGO 2012 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease. Kidney Int Suppl, 3, 1–150. Kidney Disease Outcomes Quality Initiative (K/DOQI) (2002). K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis, 39, S1–266. Kidney Disease Outcomes Quality Initiative (K/DOQI) (2004). K/DOQI clinical practice guidelines on hypertension and antihypertensive agents in chronic kidney disease. Am J Kidney Dis, 43, S1–290. Levey, A. S., Atkins, R., Coresh, J., et al. (2007). Chronic kidney disease as a global public health problem: approaches and initiatives—a position statement from Kidney Disease Improving Global Outcomes. Kidney Int, 72, 247–59. Levey, A. S., Bosch, J. P., Lewis, J. B., et al. (1999). A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med, 130, 461–70. Levey, A. S. and Coresh, J. (2012). Chronic kidney disease. Lancet, 379, 165–80. Levey, A. S., De Jong, P. E., Coresh, J., et al. (2011). The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report. Kidney Int, 80, 17–28. Levey, A. S., Eckardt, K. U., Tsukamoto, Y., et al. (2005). Definition and classification of chronic kidney disease: a position statement from Kidney Disease: Improving Global Outcomes (KDIGO). Kidney Int, 67, 2089–100. Levey, A. S. and Stevens, L. A. (2010). Estimating GFR using the CKD Epidemiology Collaboration (CKD-EPI) creatinine equation: more

753

754

Section 5  

the patient with reduced renal function

accurate GFR estimates, lower CKD prevalence estimates, and better risk predictions. Am J Kidney Dis, 55, 622–7. Levey, A. S., Stevens, L. A. and Coresh, J. (2009a). Conceptual model of CKD: applications and implications. Am J Kidney Dis, 53, S4–16. Levey, A. S., Stevens, L. A., Schmid, C. H., et al. (2009b). A new equation to estimate glomerular filtration rate. Ann Intern Med, 150, 604–12. Levin, A., Hemmelgarn, B., Culleton, B., et al. (2008). Guidelines for the management of chronic kidney disease. CMAJ, 179, 1154–62. Lindeman, R. D., Tobin, J., and Shock, N. W. (1985). Longitudinal studies on the rate of decline in renal function with age. J Am Geriatr Soc, 33, 278–85. Lo, L. J., Go, A. S., Chertow, G. M., et al. (2009). Dialysis-requiring acute renal failure increases the risk of progressive chronic kidney disease. Kidney Int, 76, 893–9. Lysaght, M. J. (2002). Maintenance dialysis population dynamics: current trends and long-term implications. J Am Soc Nephrol, 13, S37–40. Matsushita, K., Mahmoodi, B. K., Woodward, M., et al. (2012). Comparison of risk prediction using the CKD-EPI equation and the MDRD study equation for estimated glomerular filtration rate. JAMA, 307, 1941–51. National Institute for Health and Clinical Excellence (2008). Chronic Kidney Disease: National Clinical Guideline for Early Identification and Management in Adults in Primary and Secondary Care. London: Royal College of Physicians. Pannu, N., James, M., Hemmelgarn, B. R., et al. (2011). Modification of outcomes after acute kidney injury by the presence of CKD. Am J Kidney Dis, 58, 206–13. Peralta, C. A., Katz, R., Sarnak, M. J., et al. (2011a). Cystatin C identifies chronic kidney disease patients at higher risk for complications. J Am Soc Nephrol, 22, 147–55. Peralta, C. A., Shlipak, M. G., Judd, S., et al. (2011b). Detection of chronic kidney disease with creatinine, cystatin C, and urine albumin-tocreatinine ratio and association with progression to end-stage renal disease and mortality. JAMA, 305, 1545–52. Rettig, R. A., Norris, K., and Nissenson, A. R. (2008). Chronic kidney disease in the United States: a public policy imperative. Clin J Am Soc Nephrol, 3, 1902–10.

Rule, A. D., and Teo, B. W. (2009). GFR estimation in Japan and China: what accounts for the difference? Am J Kidney Dis, 53, 932–5. Shlipak, M. G., Matsushita, K., Arnlov, J., et al. (2013). Cystatin C versus creatinine in determining risk based on kidney function. N Engl J Med, 369, 932–43. Shlipak, M. G., Sarnak, M. J., Katz, R., et al. (2005). Cystatin C and the risk of death and cardiovascular events among elderly persons. N Engl J Med, 352, 2049–60. Smith, H. W. (1937). The Physiology of the Kidney. New York: Oxford University Press. Stevens, L. A., Coresh, J., Feldman, H. I., et al. (2007a). Evaluation of the modification of diet in renal disease study equation in a large diverse population. J Am Soc Nephrol, 18, 2749–57. Stevens, L. A., Coresh, J., Schmid, C. H., et al. (2008). Estimating GFR using serum cystatin C alone and in combination with serum creatinine: a pooled analysis of 3,418 individuals with CKD. Am J Kidney Dis, 51, 395–406. Stevens, L. A. and Levey, A. S. (2009). Measured GFR as a confirmatory test for estimated GFR. J Am Soc Nephrol, 20, 2305–13. Stevens, L. A., Li, S., Kurella Tamura, M., et al. (2011). Comparison of the CKD Epidemiology Collaboration (CKD-EPI) and Modification of Diet in Renal Disease (MDRD) Study Equations: Risk Factors for and Complications of CKD and Mortality in the Kidney Early Evaluation Program (KEEP). Am J Kidney Dis, 57, S9–16. Stevens, L. A., Manzi, J., Levey, A. S., et al. (2007b). Impact of creatinine calibration on performance of GFR estimating equations in a pooled individual patient database. Am J Kidney Dis, 50, 21–35. Stevens, L. A., Schmid, C. H., Greene, T., et al. (2009). Factors other than glomerular filtration rate affect serum cystatin C levels. Kidney Int, 75, 652–60. Stevens, L. A., Schmid, C. H., Greene, T., et al. (2010). Comparative performance of the CKD Epidemiology Collaboration (CKD-EPI) and the Modification of Diet in Renal Disease (MDRD) Study equations for estimating GFR levels above 60 mL/min/1.73 m2. Am J Kidney Dis, 56, 486–95.

CHAPTER 95

Chronic kidney disease in the developed world Morgan E. Grams and Josef Coresh Overview Chronic kidney disease (CKD) is an important and increasingly common public health issue within the developed world. Both reduced glomerular filtration rate (GFR) and albuminuria are independently and continuously associated with acute kidney injury (AKI), cardiovascular disease (CVD), end-stage renal disease (ESRD), and death. A disease of multiple aetiologies but most commonly attributed to diabetes and hypertension, CKD rates and progression vary widely by population. Some explanations of the variation are intrinsic to the population studied—the age distribution, the frequency of high-risk genetic alleles (such as those in polycystic kidney disease (PKD) and apolipoprotein 1 (APOL1), for example), and the burden of chronic conditions such as diabetes and hypertension—but some are extrinsic, such as variation in diagnostic scrutiny and CKD staging, creatinine assay standardization, and GFR estimating equations, as well as the threshold for renal replacement therapy (RRT). For these reasons, comparing CKD rates across countries is a daunting task. Nonetheless, accurate identification and staging of CKD in the developed world is essential to optimal care, including adherence to current guidelines and the development of new and effective interventions.

Introduction Often asymptomatic until the advanced stages of disease, CKD has been historically under-recognized by both patient and provider. Increased appreciation of CKD-related sequelae, coupled with advances in CKD reporting and staging, may improve CKD detection and treatment. Recent progress includes the standardization of serum creatinine assays (Myers et  al., 2006; College of American Pathologists, 2011), the introduction and refinement of GFR estimating equations (Levey et al., 1999, 2009), the issuance of clinical practice guidelines (Kidney Disease Outcomes Quality Initiative (KDOQI)) for CKD staging (Table 95.1) (National Kidney Foundation, 2009), and the incorporation of routine estimated GFR (eGFR) reporting by most laboratory systems (College of American Pathologists, 2011). As a result, CKD identification—even at the population level (e.g. )—is increasingly feasible. A  suggested CKD classification system incorporating albuminuria may further improve CKD recognition (Levey et al., 2011); and, once standardized, cystatin C-based determination of CKD should help improve uniformity

in national and international evaluations, independent of diet and muscle mass.

Epidemiology of chronic kidney disease Prevalence of CKD While many countries maintain registries to track the prevalence and incidence of kidney disease requiring dialysis or transplantation, estimating the rates of earlier stages of kidney disease is more difficult. True and comparable estimates require a population-representative sample, standardized biomarker assays, and uniform estimating equations. For example, using the creatinine-based Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) equation tends to result in a lower prevalence of CKD than does the creatinine-based Modification of Diet in Renal Disease (MDRD) Study equation introduced in 1999 (Levey et al., 2009; Stevens et al., 2011). Prevalence determined by the newer biomarker cystatin C is similar to or even greater than that based on serum creatinine (Astor et  al., 2009; Peralta et  al., 2011). International comparisons are additionally complicated by the use of population-specific GFR estimating equations (e.g. the Japanese eGFR equations (Matsuo et  al., 2010; Kitiyakara et  al., 2012)). In the United States, the gold standard for population-based studies may be the National Health and Nutrition Examination Survey (NHANES) (Coresh et al., 2007). In this study, prevalence of CKD stages 3–5 was 7% in the 1999–2004 survey. Estimates of CKD prevalence in other general population cohorts range from 1% to 57% (Matsushita et al., 2012) with variations largely dependent on the age and ethnic distribution of the study participants, as well as the prevalence of chronic conditions—diabetes, hypertension, obesity, and cardiovascular disease—thought to be the primary causes of CKD in the developed world (Table 95.2) (de Boer et al., 2011).

Incidence of CKD Only limited information exists on the incidence of CKD in the developed world, primarily because the calculation of incidence requires longitudinal follow-up and multiple comparable measures of kidney function within an individual. Hence, the most accurate estimations of CKD incidence stem from long-running prospective cohorts with planned, interval measurement of kidney function. Retrospective analyses of administrative databases are less ideal, as patients at higher risk for CKD may have their

756

Section 5  

the patient with reduced renal function cohort of predominantly white individuals with a relatively low rate of diabetes (2.7%), found that 9.4% of an initially CKD-free population developed Stage 3 or worse CKD over an 18.5-year follow-up (Fox et al., 2004). A 10-year increase in age more than doubled the odds of developing CKD. The Atherosclerosis Risk in Communities (ARIC) study, a population-based cohort of 45–64-year-old black (25.6%) and white (74.4%) individuals with a much higher rate of diabetes (11.4%), reported 7.3% CKD incidence over 8.8 years, or 10.4 cases per 1000 person-years (Bash et al., 2009). In the Cardiovascular Health Study (CHS), among older adults without baseline CKD, incident CKD was detected at year 7 in 10% using creatinine and 19% using cystatin C (P < 0.001) (Shlipak et al., 2009). Older age was associated with an increased risk of CKD in all of these studies. Consistent results emerged in a retrospective review from the United Kingdom, which found an overall CKD incidence of 1.3 cases per 1000 person-years, ranging from 0.02 in the youngest (< 20 years) to 12 cases per 1000 person years in the oldest (80+) age group (Drey et al., 2003). Substantial variation in CKD rates by ethnicity has been demonstrated in three long-running US cohorts. The Coronary Artery Risk Development in Young Adults (CARDIA) cohort, a study of black and white individuals aged 18–30  years, reported a 1% incidence of CKD over 20 years of follow-up; however, the risk of CKD among black participants was 2.6 times that among white individuals (Muntner et al., 2012). This was also true in the older

Table 95.1  K/DOQI stages of CKD based on eGFRa Stage

Description

GFR (mL/min/ 1.73 m2)

1

Kidney damage with normal or increased GFR

≥ 90

2

Kidney damage with mild reduced GFR

60–89

3

Moderate reduced GFR

30–59

4

Severe reduced GFR

15–29

5

Kidney failure

< 15 or dialysis

Based on data, a consensus statement from the Kidney Disease: Improving Global Outcome (KDIGO) conference suggested adding albuminuria categories (albumin to creatinine ratio of < 30, 30–299, 300+ mg/g) and cause of CKD as additional components of a CKD staging system (Levey et al., 2011).

kidney function measured more often, creating an ascertainment bias, and trends in CKD reporting/assay standardization are difficult to control for. In addition, measures of CKD incidence have not been standardized, with some studies evaluating CKD onset and others requiring a minimal level of change to rule out random variation around the CKD cut-point (e.g. 25% decline in eGFR and eGFR < 60 mL/min/1.73 m2) (Bash et al., 2009). In the United States, a few large prospective cohorts have measured incident CKD. The Framingham study, a population-based

Table 95.2  Prevalence of CKD (eGFR < 60 by CKD-EPI) in the developed world % (Matsushita et al., 2012) Study name

Study region

Aichi

Japan

ARIC

Study size

Mean age (years)

% female

4731

48

20

0

26

7

1

USA

11,441

63

56

22

48

17

7

AusDiab

Australia

11,179

52

55

0

33

9

6

Beaver Dam

USA

4885

62

56

0

51

10

15

CHS

USA

2988

78

59

17

64

16

21

CIRCS

Japan

11,871

54

61

0

36

5

3

ESTHER

Germany

9641

62

55

0

60

19

14

Framingham

USA

2956

59

53

0

40

10

7

Gubbio

Italy

1682

55

56

0

39

5

1

HUNT

Norway

9659

62

55

0

82

18

11

IPHS

Japan

95,451

59

66

0

50

5

4

MESA

USA

6733

62

53

28

45

13

9

MRC

UK

12,371

81

61

0

34

8

57

NHANESIII

USA

15,563

47

53

28

29

12

7

Ohasama

Japan

1956

63

64

0

41

10

5

PREVEND

Netherlands

8385

49

50

1

33

4

4

Rancho Bernardo

USA

1477

71

60

0

55

12

22

REGARDS

USA

27,306

65

54

40

59

21

11

ULSAM

Sweden

1103

71

0

0

75

19

8

CKD = chronic kidney disease; DM = diabetes mellitus; HTN = hypertension.

% black

% HTN

% DM

% with CKD

Chapter 95 

Multi-Ethnic Study of Atherosclerosis (MESA) cohort. Here, black people and Hispanics had higher rates of incident CKD than white people and those of Chinese ethnicity, although the effect was not statistically significant after adjustment for baseline hypertension and diabetes (Peralta et al., 2011). In the ARIC study, African Americans had higher incidence rates than white people for CKD hospitalizations and a creatinine rise but not for incidence of eGFR < 60 by the MDRD Study equation (Bash et al., 2009).

Awareness of CKD Recognition of CKD is low, particularly in the earliest stages of disease. In a survey of US physicians, only 59% of family medicine physicians and 78% of general internal medicine physicians correctly identified stage 3–4 CKD in a hypothetical scenario (Boulware et  al., 2006). In clinical practice, physician awareness may be even lower. For example, in an Italian study, general practitioners recognized CKD in only 10.8% of their patients with stage 3 and in 72.6% of those with stage 4–5 (Ravera et al., 2011). Clinical studies of physician awareness should be interpreted cautiously, however: recognition is generally assessed from claims, which represent physician documentation translated to a billing code, a variably sensitive and specific process. Since CKD diagnosis hinges on laboratory tests, patient awareness of CKD understandably lags that of physicians. In the United States, many studies of patient recognition use the NHANES questionnaire, which asks, ‘Have you ever been told you have weak or failing kidneys (excluding kidney stones, bladder infections, or incontinence)?’ The proportion of CKD patients answering this question in the affirmative is very low: 5.2% in 1999–2000, compared with 6.7% in 2001–2002, and 6.0% in 2003–2004 (Plantinga et al., 2008). Estimates of awareness are minimally affected by the estimating equation used; in the Kidney Early Evaluation Program (KEEP), 9.5% of participants were aware of MDRD-based CKD, compared with 10.0% of participants aware of CKD-EPI based CKD (Kurella Tamura et al., 2011).

Genetics of CKD The large variation in CKD rates by ethnicity—more specifically, the strong predisposition to non-diabetic kidney disease among those of African descent—has been the subject of recent intense investigation (Friedman and Pollak, 2011). In 2008, two genome-wide admixture association studies demonstrated a strong link between a locus on chromosome 22 and non-diabetic kidney disease among African Americans (Kao et al., 2008; Kopp et al., 2008). Subsequent studies identified two genetic variants in APOL1 (a gene in close linkage disequilibrium with myosin heavy chain 9 gene (MYH9), the gene originally thought to contain the causal variants) that, in individuals with two risk alleles, confers a 10- to 7-fold increased risk of focal segmental glomerulosclerosis and hypertension-associated ESRD (Genovese et  al., 2010; Tzur et  al., 2010). Present in 50% of African Americans (and 10–15% have two risk alleles), the G1 and G2 APOL1 gene variants may be even more strongly linked to HIV-associated nephropathy (Kopp et al., 2011; Fine et al., 2012). In contrast to diabetic nephropathy, which is considered a complex disease with multiple genetic and environmental causes with surprisingly little effect of APOL1, some suggest that APOL1-mediated nephropathy in African-Americans is a predominantly Mendelian disease, with 3 million African Americans carrying the high-risk homozygous genotype (Friedman and Pollak, 2011).

chronic kidney disease in the developed world

Epidemiology of end-stage renal disease Prevalence of ESRD ESRD, or the most advanced stage of CKD, is defined both by the degree of kidney failure and by treatment: kidney disease requiring RRT. This functional definition merits consideration when performing international comparisons and evaluating ESRD trends over time. As countries develop economically, ESRD rates increase, and this chapter covers only the most highly developed countries in the world (20 out of 40 countries included in the United States Renal Data System (USRDS) Annual Data Report) (USRDS, 2010). The threshold for initiating dialysis may vary substantially by country:  in 2001, for example, the mean serum creatinine at dialysis initiation was 7.3 mg/dL in the United States compared with 8.5 mg/dL in Australia and New Zealand (Stewart et al., 2004; USRDS, 2010). In addition, there are strong trends towards earlier delivery (i.e. at higher levels of GFR) of RRT in recent years (Rosansky et al., 2009; Grams et al., 2011). The prevalence of ESRD has grown at a far greater rate (annualized mean, 1998 to 2009, 3.9%) in comparison with ESRD incidence (2.0% over the same period). This is attributable both to the increased numbers of patients initiating RRT and the better survival of patients already on RRT. In absolute terms, the United States has the largest number of ESRD patients (558,239 in 2009), followed by Japan (281,212). Scaled by population, Taiwan (2447 cases per million population), Japan (2205 cases per million population), and the United States (1811 cases per million population) have the highest prevalence rates, followed by Belgium, Canada, France, Greece, and Spain (1141, 1119, 1094, 1065, and 1034, respectively). Luxembourg and Iceland had the lowest prevalence rates. The rate of growth in ESRD prevalence was fairly uniform across developed countries with the exception of Taiwan, in which the prevalence rate averaged a 6.3% increase per year (Fig. 95.1) (USRDS, 2010).

Incidence of ESRD Data regarding ESRD incidence are more readily available than for CKD incidence for two reasons. First, in contrast to the earlier stages of CKD, the functional definition of ESRD renders it necessarily symptomatic, apparent to both patient and provider. Second, many countries track ESRD patients with national registries. In the United States, for example, the USRDS follows all patients initiated on maintenance haemodialysis and/or receiving kidney transplantation (USRDS, 2010). The United States and Taiwan have by far the highest rates of ESRD among the developed world, at 371 and 357 incident cases per million population in 2009, respectively (Fig. 95.2). While unadjusted rates have increased steadily over time, rates adjusted for age, sex, and ethnicity have remained fairly stable from 1996 to 2008, with a 1.1% uptick in 2009 (USRDS, 2010). Similar to the ethnic variation seen in CKD rates, ESRD incidence is 3.5 times higher among African Americans than white Americans (USRDS, 2010). Throughout the developed world, only Japan is close to the United States and Taiwanese incidence rates, at 287 new cases per million population. The lowest rates of ESRD were seen in Finland and Iceland (83 and 88 cases per million population, respectively). Greece, Belgium, and Luxembourg presented the highest rates in Europe (204, 201, and 227 cases per million population, respectively). Trends over time were fairly uniform:  the unadjusted average annualized increase in incidence rate within developed

757

3000

ESRD prevalence per million population

2500

Taiwan Japan

2000 United States 1500

1000

500

0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Australia Austria Belgium, Dutch speaking Belgium, French speaking Canada Denmark Finland France Greece Iceland Japan Luxembourg Netherlands New Zealand Norway Scotland Spain Sweden Taiwan U.K., England, Wales & N Ireland United States

Fig. 95.1  Prevalence rates of ESRD, 1998–2009.

Australia

500

Austria 450

Belgium, Dutch speaking

Taiwan

Belgium, French speaking

ESRD incidence per million population

400

Canada United States

350 300

Denmark Finland

Japan

France Greece Iceland

250

Japan Luxembourg

200

Netherlands New Zealand

150

Norway Scotland

100

Spain Sweden

50

Taiwan U.K., England, Wales & N Ireland

0 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

United States

Fig. 95.2  Unadjusted ESRD incidence rates, 1998–2009. Diabetic nephropathy—clearly a term fraught with uncertainty, as the majority of ESRD patients receiving this diagnosis are never biopsied—comprised a differing proportion of incident ESRD cases by country. In 2009, for example, the Netherlands reported that 14.9% of the incident ESRD cases were due to diabetes, whereas Taiwan, Japan, New Zealand, and the United States all attributed > 40% of incident ESRD cases to diabetic nephropathy. Rates of ESRD due to diabetic nephropathy were highest among the 65 to 74-year-old age group; particularly in Taiwan and the United States, where the rates were estimated at 713 and 628 cases per million population, respectively (USRDS, 2010).

Chapter 95 

chronic kidney disease in the developed world

Table 95.3  Worldwide distribution of haemodialysis in 2009 Top 5 haemodialysis Country

% using modality

Luxembourga 99.0

Top 5 home haemodialysis

Greece United States

Top 5 peritoneal dialysis

Country

% using modality

Country

% using modality

New Zealand

16.3

New Zealand

35.0

Luxembourg Belgium, French speaking Belgium, Dutch speaking New Zealand Iceland Denmark

Japan

96.7

Australia

9.3

Sweden

23.6

Greece

92.1

Denmark

4.9

Denmark

21.7

United States 90.9

Finland

3.8

Finland

21.5

Scotland

Austria

Canada

3.5

Australia

21.1

Australia

90.6

Canada France

Sweden

a2008 figures.

Austria

countries was 2.0% from 1998 to 2009, with a median of 1.9%. Spain did experience a declining rate of ESRD (175 cases per million population in 2004 compared with 129 cases per million population in 2009), while Sweden and Scotland remained stable over the 12-year period (USRDS, 2010).

U.K., England, Wales & N Ireland Finland Netherlands Spain Norway 0%

Dialysis modality in ESRD The distribution of dialysis modality differs by country (Table 95.3). In the United States, for example, there were 113,636 incident dialysis cases in 2009, with 6.9% and 1.1% started on peritoneal dialysis and home haemodialysis, respectively (USRDS, 2010). In contrast, the use of modalities other than in-centre haemodialysis was much more common in New Zealand, Australia, and Finland—perhaps due to their relatively low population densities.

Relationship between CKD prevalence and ESRD incidence The relationship between CKD prevalence and ESRD incidence is complex. The determinants of ESRD incidence include underlying CKD prevalence, rates of CKD progression, incidence of mortality prior to ESRD, and the rates of acceptance of patients into RRT programmes. National policy and practice patterns may contribute: a country with a lower threshold for RRT initiation (e.g. dialysis is initiated at higher levels of GFR) will have a higher rate of ESRD than a country that forestalls RRT initiation or offers conservative management alternatives to individuals in ill health. As an example, although CKD prevalence is roughly similar in Europe and the United States, the incidence of ESRD is dramatically different. A comparison of rates between Norway and the United States revealed a similar distribution of patients among CKD stages, but 2.5 times the relative risk of ESRD in US white people versus Norwegians (98% white) (Hallan et al., 2006). Disparities between the countries’ ESRD incidence were larger among patients older than 60 (relative risk, 3.0) and women (relative risk, 3.5). Whether this reflects differences in underlying comorbidities or differences in practice patterns is unclear. The proportion of incident ESRD due to diabetic nephropathy was much higher in the United States (41% vs 11% from 1995 to 1997), and the number of pre-dialysis visits with a nephrologist was much lower (41% had five visits or more, compared with 73% in Norway). Other detailed international comparisons have not been done; however, in the United States,

10% 20% 30% 40% 50% Transplant rate/incident ESRD rate

60%

Fig. 95.3  Rates of kidney transplantation per incident ESRD, 1998–2009.

there is a strong inverse association between per capita income and incidence rates of ESRD (Young et al., 1994).

Renal replacement therapy: transplantation Within the developed world, rates of kidney transplantation ranged from 14.9 (Greece) to 63.1 (Canada) per million population in 2009, or 7.3% (Greece) to 52.1% (Norway) when scaled by ESRD incidence (USRDS, 2010). These rates were similar when averaged over 1998–2009 (Fig. 95.3). The prevalence of kidney transplant recipients with a still-functioning allograft ranged from 215 (Greece) to 562 (United States) per million population in 2009; however, scaled by ESRD prevalence, this ranged from 31% (United States) to 70% (Canada). Transplantation rates in Japan, not available for recent years, have historically ranked among the lowest in the developed world (Satayathum et al., 2005). Internationally, rates of transplantation are consistently highest among young, white, better educated, and wealthier patients, with shorter dialysis vintage (Satayathum et al., 2005).

Prognosis CKD CKD is an independent risk factor for multiple adverse outcomes. The two commonly measured components of CKD, reduced GFR and albuminuria, independently contribute to the risk of AKI, ESRD, cardiovascular disease, and mortality (Matsushita et  al., 2010; Astor et al., 2011; Gansevoort et al., 2011). Adverse outcomes are interrelated—AKI can be a precursor to ESRD (Hsu et al., 2009), which may in turn increase the risk of cardiovascular disease and death—and the relative incidence of outcomes differs by patient

759

760

Section 5  

the patient with reduced renal function

population. In many studies of renal disease progression, such as the African American Study of Kidney Disease and Hypertension (AASK) (Alves et al., 2010) and the MDRD Study (Menon et al., 2009), the rates of ESRD are greater than that of pre-ESRD death. However, this is likely not true in the general population, where the incidence of CKD far exceeds that of ESRD. Indeed, in a US study using administrative data, the rates of death were consistently higher than those of ESRD in each stage of CKD (e.g. the rates of death and ESRD were 45.7% and 19.9%, respectively, for patients with stage 4 CKD) (Keith et al., 2004). In general, lower GFR, higher proteinuria, and younger age independently increase the relative risk of ESRD versus pre-ESRD death (O’Hare et  al., 2007). In contrast, diabetes, vascular disease, and older age confer an increased risk of pre-ESRD death. Among US veterans ages 85 and older, the rate of death exceeded ESRD regardless of level of kidney function (O’Hare et al., 2007).

Dialysis Mortality rates on dialysis remain exceedingly high. In a recent observational cohort of 32,065 nationally-representative haemodialysis patients in the United States, the mortality rate was 18.6 deaths per 100 person-years (Foley et al., 2011). Certain patient characteristics may predispose to a heightened mortality risk—older age and diabetes, for example, as among the non-ESRD population (Villar et al., 2007)—and ethnicity may modify these relationships (Kucirka et al., 2011). The interval between haemodialysis sessions may play a role: the majority of adverse events may occur on days after the 2-day gap in treatment (Foley et  al., 2011). The effect of dialysis modality remains a subject of debate—confounded by selection bias (Quinn et al., 2011) and frequent modality switches—and may vary by patient subgroup (Weinhandl et al., 2010). Because of the importance of age, ethnicity, and comorbidity distributions, as well as profound differences in local policy and practice, international comparisons of dialysis survival require cautious interpretation. That stated, a striking difference in mortality has been noted across different countries: the 1-year mortality rate reported in 2003 was 6.6% in Japan, 15.6% in Europe, and 21.7% in the United States (Goodkin et al., 2003). Within Europe, the United Kingdom had the highest mortality rate (18.6 deaths per 100 patient-years), followed by Germany (16.3 deaths per 100 patient-years), Spain (15.3 deaths per 100 patient-years), Italy, and France (13.8 and 13.3 deaths per 100 patient-years, respectively) (Rayner et al., 2004). Plausible explanations for these differences include variations in underlying comorbidities (e.g. diabetes and atherosclerotic cardiovascular disease), types of vascular access at initiation of dialysis, delivered dose of dialysis, level of care and education, and nutritional status and supplementation (Foley and Hakim, 2009).

Transplantation Kidney transplantation is the preferred mode of RRT in ESRD, imparting a significant survival advantage over remaining on dialysis (Wolfe et al., 1999; Oniscu et al., 2005). The survival benefit associated with transplantation varies by recipient age, comorbidities, and quality of the donor organ (Merion et al., 2005). In the United States, post-transplant mortality has improved over time, coinciding with the advent of modern immunosuppressive regimens. As of 2008, the 5-year crude survival rate was 91% for recipients of living donor kidneys and 84% for recipients of standard deceased

donor kidneys (Axelrod et  al., 2010). Comparisons with other countries are difficult because of differences in allograft quality and the make-up of transplant recipients; however, long-term mortality may be slightly higher in the United States compared with Canada (adjusted hazard ratio, 1.35) (Kim et al., 2006).

References Alves, T. P., Wang, X., Wright, J. T., Jr., et al. (2010). Rate of ESRD exceeds mortality among African Americans with hypertensive nephrosclerosis. J Am Soc Nephrol, 21(8), 1361–9. Astor, B. C., Levey, A. S., Stevens, L. A., et al. (2009). Method of glomerular filtration rate estimation affects prediction of mortality risk. J Am Soc Nephrol, 20(10), 2214–22. Astor, B. C., Matsushita, K., Gansevoort, R. T., et al. (2011). Lower estimated glomerular filtration rate and higher albuminuria are associated with mortality and end-stage renal disease. A collaborative meta-analysis of kidney disease population cohorts. Kidney Int, 79(12), 1331–40. Axelrod, D. A., McCullough, K. P., Brewer, E. D., et al. (2010). Kidney and pancreas transplantation in the United States, 1999–2008: the changing face of living donation. Am J Transplant, 10(4 Pt 2), 987–1002. Bash, L. D., Coresh, J., Kottgen, A., et al. (2009). Defining incident chronic kidney disease in the research setting: The ARIC Study. Am J Epidemiol, 170(4), 414–24. Boulware, L. E., Troll, M. U., Jaar, B. G., et al. (2006). Identification and referral of patients with progressive CKD: a national study. Am J Kidney Dis, 48(2), 192–204. College of American Pathologists (2011). Current Status of Reporting Estimated Glomerular Filtration Rate (eGFR). [Online] Coresh, J., Selvin, E., Stevens, L. A., et al. (2007). Prevalence of chronic kidney disease in the United States. JAMA, 298(17), 2038–47. De Boer, I. H., Rue, T. C., Hall, Y. N., et al. (2011). Temporal trends in the prevalence of diabetic kidney disease in the United States. JAMA, 305(24), 2532–9. Drey, N., Roderick, P., Mullee, M., et al. (2003). A population-based study of the incidence and outcomes of diagnosed chronic kidney disease. Am J Kidney Dis, 42(4), 677–84. Fine, D. M., Wasser, W. G., Estrella, M. M., et al. (2012). APOL1 risk variants predict histopathology and progression to ESRD in HIV-related kidney disease. J Am Soc Nephrol, 23(2), 343–50. Foley, R. N., Gilbertson, D. T., Murray, T., et al. (2011). Long interdialytic interval and mortality among patients receiving hemodialysis. N Engl J Med, 365(12), 1099–107. Foley, R. N. and Hakim, R. M. (2009). Why is the mortality of dialysis patients in the United States much higher than the rest of the world? J Am Soc Nephrol, 20(7), 1432–5. Fox, C. S., Larson, M. G., Leip, E. P., et al. (2004). Predictors of new-onset kidney disease in a community-based population. JAMA, 291(7), 844–50. Friedman, D. J. and Pollak, M. R. (2011). Genetics of kidney failure and the evolving story of APOL1. J Clin Invest, 121(9), 3367–74. Gansevoort, R. T., Matsushita, K., van der Velde, M., et al. (2011). Lower estimated GFR and higher albuminuria are associated with adverse kidney outcomes. A collaborative meta-analysis of general and high-risk population cohorts. Kidney Int, 80(1), 93–104. Genovese, G., Friedman, D. J., Ross, M. D., et al. (2010). Association of trypanolytic ApoL1 variants with kidney disease in African Americans. Science, 329(5993), 841–5. Goodkin, D. A., Bragg-Gresham, J. L., Koenig, K. G., et al. (2003). Association of comorbid conditions and mortality in hemodialysis patients in Europe, Japan, and the United States: the Dialysis Outcomes and Practice Patterns Study (DOPPS). J Am Soc Nephrol, 14(12), 3270–7. Grams, M. E., Massie, A. B., Coresh, J., et al. (2011). Trends in the timing of pre-emptive kidney transplantation. J Am Soc Nephrol, 22(9), 1615–20.

Chapter 95 

Hallan, S. I., Coresh, J., Astor, B. C., et al. (2006). International comparison of the relationship of chronic kidney disease prevalence and ESRD risk. J Am Soc Nephrol, 17(8), 2275–84. Hsu, C. Y., Chertow, G. M., McCulloch, C. E., et al. (2009). Nonrecovery of kidney function and death after acute on chronic renal failure. Clin J Am Soc Nephrol, 4(5), 891–8. Kao, W. H., Klag, M. J., Meoni, L. A., et al. (2008). MYH9 is associated with nondiabetic end-stage renal disease in African Americans. Nat Genet, 40(10), 1185–92. Keith, D. S., Nichols, G. A., Gullion, C. M., et al. (2004). Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med, 164(6), 659–63. Kim, S. J., Schaubel, D. E., Fenton, S. S., et al. (2006). Mortality after kidney transplantation: a comparison between the United States and Canada. Am J Transplant, 6(1), 109–14. Kitiyakara, C., Yamwong, S., Vathesatogkit, P., et al. (2012). The impact of different GFR estimating equations on the prevalence of CKD and risk groups in a Southeast Asian cohort using the new KDIGO guidelines. BMC Nephrol, 13, 1. Kopp, J. B., Nelson, G. W., Sampath, K., et al. (2011). APOL1 genetic variants in focal segmental glomerulosclerosis and HIV-associated nephropathy. J Am Soc Nephrol, 22(11), 2129–37. Kopp, J. B., Smith, M. W., Nelson, G. W., et al. (2008). MYH9 is a major-effect risk gene for focal segmental glomerulosclerosis. Nat Genet, 40(10), 1175–84. Kucirka, L. M., Grams, M. E., Lessler, J., et al. (2011). Association of race and age with survival among patients undergoing dialysis. JAMA, 306(6), 620–6. Kurella Tamura, M., Anand, S., Li, S., et al. (2011). Comparison of CKD awareness in a screening population using the Modification of Diet in Renal Disease (MDRD) study and CKD Epidemiology Collaboration (CKD-EPI) equations. Am J Kidney Dis, 57(3 Suppl 2), S17–23. Levey, A. S., Bosch, J. P., Lewis, J. B., et al. (1999). A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med, 130(6), 461–70. Levey, A. S., de Jong, P. E., Coresh, J., et al. (2011). The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report. Kidney Int, 80(1), 17–28. Levey, A. S., Stevens, L. A., Schmid, C. H., et al. (2009). A new equation to estimate glomerular filtration rate. Ann Intern Med, 150(9), 604–12. Matsuo, S., Imai, E., Horio, M., et al. (2009). Revised equations for estimated GFR from serum creatinine in Japan. Am J Kidney Dis, 53(6), 982–92. Matsushita, K., Mahmoodi, B. K., Woodward, M., et al. (2012). Comparison of risk prediction using the CKD-EPI equation and the MDRD study equation for estimated glomerular filtration rate. JAMA, 307(18), 1941–51. Matsushita, K., van der Velde, M., Astor, B. C., et al. (2010). Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet, 375(9731), 2073–81. Menon, V., Kopple, J. D., Wang, X., et al. (2009). Effect of a very low-protein diet on outcomes: long-term follow-up of the Modification of Diet in Renal Disease (MDRD) Study. Am J Kidney Dis, 53(2), 208–17. Merion, R. M., Ashby, V. B., Wolfe, R. A., et al. (2005). Deceased-donor characteristics and the survival benefit of kidney transplantation. JAMA, 294(21), 2726–33. Muntner, P., Newsome, B., Kramer, H., et al. (2012). Racial differences in the incidence of chronic kidney disease. Clin J Am Soc Nephrol, 7(1), 101–7. Myers, G. L., Miller, W. G., Coresh, J., et al. (2006). Recommendations for improving serum creatinine measurement: a report from the Laboratory Working Group of the National Kidney Disease Education Program. Clin Chem, 52(1), 5–18. National Kidney Foundation (2002). K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis, 39(2 Suppl 1), S1–266.

chronic kidney disease in the developed world

O’Hare, A. M., Choi, A. I., Bertenthal, D., et al. (2007). Age affects outcomes in chronic kidney disease. J Am Soc Nephrol, 18(10), 2758–65. Oniscu, G. C., Brown, H., and Forsythe, J. L. (2005). Impact of cadaveric renal transplantation on survival in patients listed for transplantation. J Am Soc Nephrol, 16(6), 1859–65. Peralta, C. A., Katz, R., DeBoer, I., et al. (2011). Racial and ethnic differences in kidney function decline among persons without chronic kidney disease. J Am Soc Nephrol, 22(7), 1327–34. Peralta, C. A., Shlipak, M. G., Judd, S., et al. (2011). Detection of chronic kidney disease with creatinine, cystatin C, and urine albumin-tocreatinine ratio and association with progression to end-stage renal disease and mortality. JAMA, 305(15), 1545–52. Plantinga, L. C., Boulware, L. E., Coresh, J., et al. (2008). Patient awareness of chronic kidney disease: trends and predictors. Arch Intern Med, 168(20), 2268–75. Quinn, R. R., Hux, J. E., Oliver, M. J., et al. (2011). Selection bias explains apparent differential mortality between dialysis modalities. J Am Soc Nephrol, 22(8), 1534–42. Ravera, M., Noberasco, G., Weiss, U., et al. (2011). CKD awareness and blood pressure control in the primary care hypertensive population. Am J Kidney Dis, 57(1), 71–7. Rayner, H. C., Pisoni, R. L., Bommer, J., et al. (2004). Mortality and hospitalization in haemodialysis patients in five European countries: results from the Dialysis Outcomes and Practice Patterns Study (DOPPS). Nephrol Dial Transplant, 19(1), 108–20. Rosansky, S. J., Clark, W. F., Eggers, P., et al. (2009). Initiation of dialysis at higher GFRs: is the apparent rising tide of early dialysis harmful or helpful? Kidney Int, 76(3), 257–61. Satayathum, S., Pisoni, R. L., McCullough, K. P., et al. (2005). Kidney transplantation and wait-listing rates from the international Dialysis Outcomes and Practice Patterns Study (DOPPS). Kidney Int, 68(1), 330–7. Shlipak, M. G., Katz, R., Kestenbaum, B., et al. (2009). Rate of kidney function decline in older adults: a comparison using creatinine and cystatin C. Am J Nephrol, 30(3), 171–8. Stevens, L. A., Li, S., Kurella Tamura, M., et al. (2011). Comparison of the CKD Epidemiology Collaboration (CKD-EPI) and Modification of Diet in Renal Disease (MDRD) study equations: risk factors for and complications of CKD and mortality in the Kidney Early Evaluation Program (KEEP). Am J Kidney Dis, 57(3 Suppl 2), S9–16. Stewart, J. H., McCredie, M. R., Williams, S. M., et al. (2004). Interpreting incidence trends for treated end-stage renal disease: implications for evaluating disease control in Australia. Nephrology (Carlton), 9(4), 238–46. Tzur, S., Rosset, S., Shemer, R., et al. (2010). Missense mutations in the APOL1 gene are highly associated with end stage kidney disease risk previously attributed to the MYH9 gene. Hum Genet, 128(3), 345–50. United States Renal Data System (2010). USRDS 2010 Annual Data Report: Atlas of Chronic Kidney Disease and End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Villar, E., Remontet, L., Labeeuw, M., et al. (2007). Effect of age, gender, and diabetes on excess death in end-stage renal failure. J Am Soc Nephrol, 18(7), 2125–34. Weinhandl, E. D., Foley, R. N., Gilbertson, D. T., et al. (2010). Propensity-matched mortality comparison of incident hemodialysis and peritoneal dialysis patients. J Am Soc Nephrol, 21(3), 499–506. Wolfe, R. A., Ashby, V. B., Milford, E. L., et al. (1999). Comparison of mortality in all patients on dialysis, patients on dialysis awaiting transplantation, and recipients of a first cadaveric transplant. N Engl J Med, 341(23), 1725–30. Young, E. W., Mauger, E. A., Jiang, K. H., et al. (1994). Socioeconomic status and end-stage renal disease in the United States. Kidney Int, 45(3), 907–11.

761

CHAPTER 96

Chronic kidney disease in developing countries Luxia Zhang and Haiyan Wang Introduction The spread of non-communicable diseases (NCDs) presents a global crisis, which is a barrier to the development of goals including reduction of poverty, health equity, economic stability, and human security (Beaglehole et al., 2011). NCDs accounted for 61% of the estimated 58 million deaths and 46% of the global burden of diseases worldwide in 2005 (Wagner and Brath, 2012). Among NCDs, chronic kidney disease (CKD) is of particular significance. It is recognized that the burden of CKD is not only limited to its impact on demands for renal replacement therapy (RRT) but has equally major impacts on the health of the overall population. For example, it is now well established that among the general population as well as in the diabetic or hypertensive population, the prognosis, especially the mortality and acceleration of cardiovascular events, depends on kidney involvement (Go et al., 2004; Bello et al., 2005; Matsushita et al., 2010). Also, CKD is associated with other major serious consequences including increased risk of acute kidney injury (AKI), increased risk of mineral and bone disease, adverse metabolic and nutritional consequences, infections, and reduced cognitive function. As the consequence of these amplifying effects, the financial expenditure and medical resources consumed for the management of CKD patients is much higher than expected. The burden of CKD is likely to have profound socioeconomic and public health consequences, especially in developing countries. This chapter focuses on the situation of CKD in developing countries.

Prevalence of chronic kidney disease In 2002, the Kidney Disease Outcomes Quality Initiative (KDOQI) of the National Kidney Foundation released a practice guideline for CKD (National Kidney Foundation, 2002). Since then, numerous cross-sectional surveys on the prevalence of CKD have been published. For example, results of the National Health and Nutrition Examination Surveys (NHANES) 1988–1994 in the United States revealed a prevalence of 10.0%; this number increased to 13.1% in 1999–2004 (Coresh et al., 2007). Studies from other developed countries reported similar prevalences (Chadban et  al., 2003; Hallan et al., 2006). Data on CKD burden in developing countries is relatively scanty. A  recent systematic review of CKD in population-based studies included 26 studies which were published before July 2006 (Zhang and Rothenbacher, 2008). Among them, only four (15.4%)

were from developing countries, including Mexico, China, and Thailand (Zhang and Rothenbacher, 2008). Furthermore, data from developing countries provided heterogeneous results, which makes comparisons difficult. There are three major methodological factors that may contribute to this heterogeneity. Firstly, most of the studies did not use a sampling scheme to obtain a representative sample of the general population. For example, in a study from Mexico by Amato et al. (2005), participants were randomly selected from lists of patients assigned to primary care facilities. Other studies involved participants of certain professions (such as government employees (Varma et  al., 2010)  or employees of the Electric Generation Authority (Domrongkitchaiporn et al., 2005)), or participants from certain areas (such as rural areas (Mani, 2006; O’Donnellet al., 2011) or certain big cities (Agarwal et al., 2005; Zhang et al., 2008; A. Chen et al., 2009; W. Chen et al., 2009b)). The limited representativeness of the study population jeopardizes the generalizability of the results, and also makes it hard to compare between studies. Secondly, the definition of ‘renal function decline’ is different among studies. Older studies provided estimates based on serum creatinine cut-offs or the Cockroft–Gault (CG) equation to define CKD (Barton et al., 2004; Agarwal et al., 2005; Singh et al., 2009), all of which are proved to be less accurate than the currently recommended creatinine-based estimations of glomerular filtration rate (GFR). Recent studies have employed an estimated glomerular filtration rate (eGFR) < 60 mL/min/1.73 m2 to define renal function decline; this is now recommended in recently released Kidney Disease: Improving Global Outcomes (KDIGO) guidelines for CKD classification and management (Wheeler et  al., 2013). In that guideline, it is also recommended that the 2009 Chronic Kidney Disease Epidemiology Collaboration (CKD-EPI) creatinine equation should be used for estimating GFR. An alternative creatinine-based GFR estimating equation is acceptable only if this equation has a comparable performance. Thirdly, the method of evaluating proteinuria is different among studies. Some earlier studies did not provide information on proteinuria (J. Chen et al., 2005). Hence, the available data are not sufficient to provide the prevalence of CKD in the general population for most developing countries, making between inter-country comparisons difficult. Among the developing countries, there are two recent national surveys of CKD employing standard protocols recommended by KDOQI guidelines. The first one is the Thai Screening and Early Evaluation of Kidney Disease (SEEK) study (Ingsathit et al., 2010). A stratified-cluster sampling method was used to obtain a sample

Chapter 96 

of 3459 participants (Ingsathit et al., 2010). eGFR was calculated by using the Modification of Diet in Renal Disease (MDRD) Study equation, and albuminuria was defined by the urinary albumin to creatinine ratio (uACR) of 30 mg/g creatinine or higher (Ingsathit et al., 2010). The prevalence of CKD was reported to be 17.5% (95% confidence interval (CI) 14.6–20.4%) (Ingsathit et al., 2010). The second study is The China National Survey of Chronic Kidney Disease. In this study, a multistage, stratified sampling method was used to obtain a representative sample of people aged 18 years or older in the general population of China (Zhang et al., 20120. eGFR was calculated by the Chinese equation (Ma et al., 2006), and the definition of albuminuria was the same as in the Thai SEEK study. The prevalence of CKD was reported to be 10.8% (95% CI 10.2–11.3%) (Zhang et  al., 2012). Both studies revealed a prevalence of CKD comparable to (if not higher than) that of previous reports from developed countries, indicating that CKD has become a leading public health problem in certain developing countries. As for the prevalence of end-stage renal disease (ESRD), lack of organized ESRD treatment programmes has precluded establishment of ESRD registries in most large developing countries. Some of reported data are rough estimates based on individual experience. According to the 2011 annual data report from the United States Renal Data System (USRDS), the prevalence of ESRD in developing countries varied substantially, from 110 per million population (Philippines) to 1314 per million population (Jalisco, Mexico) (USRDS, 2011). As for the incidence of ESRD in developing countries, it also varied from 13 per million population (Bangladesh) to 597 per million population (Morelos, Mexico). However, the reports from hospital-based data cannot accurately provide prevalence and incidence estimates in developing countries because of incomplete coverage, inability of many patients to reach a hospital, and lack of a proper referral system.

Causes of chronic kidney disease Diabetes contributes heavily to the burden of CKD in developing countries. One should, however, realize that the term ‘diabetic nephropathy’ is fraught with uncertainty as the majority of patients with ESRD who receive this diagnosis are not biopsied (see Chapter 149). A study of ESRD patients in a large urban population in India indicated that diabetic kidney disease comprised 40–47% of incident cases between 2002 and 2005 (Modi and Jha, 2006). Several hospital-based studies in India suggested that around 30% of referred CKD cases were diabetic kidney disease (Agarwal and Srivastava, 2009). Reports from Latin America also indicated that diabetes is the leading cause of ESRD (30.3% of incident population) (Cusumano and Gonzalez Bedat, 2008). Furthermore, it is predicted that the proportion of CKD attributable to diabetes will continue to rise in the near future, due to the following facts. Firstly, the rapid surge in diabetes has been observed in almost all developing countries. For example, the prevalence of diabetes in China increased from 1% in 1980 (Zhong, 1982) to 9.7% in 2008 (Yang et al., 2010). In urban Indian adults, diabetes prevalence increased from 3% in the early 1970s to 12% in 2000, with a narrowing rural–urban gradient (Ramachandran, 2005). A similar trend has been observed in other Asian countries such as Bangladesh, Nepal, and Indonesia (Chan et al., 2009), sub-Saharan African (Mbanya et al., 2010), and Latin America (Escobedo et al., 2009). It is speculated that factors including general and abdominal obesity, nutrition

chronic kidney disease in developing countries

transition and changes in diet and lifestyle, and smoking contribute to the increasing number of diabetes in developing countries (Chan et al., 2009). Secondly, the rate of underdiagnosed diabetes is high in developing countries, and a disparity in healthcare is observed between urban and rural areas. In a national survey of diabetes in China, 59.7% of patients were diagnosed through screening (Yang et al., 2010). In South Africa, > 50% of people were aware of their diabetic condition in an urban area (Levitt et al., 1993), compared with only 15% of people who have documented diabetes in rural area (Motala et al., 2008). Finally, the burden of diabetes in some developing countries is disproportionately high or is predicted to escalate in young to middle-aged adults (Chan et al., 2009; Mbanya et  al., 2010), which means long disease exposure and therefore more chronic complications, including CKD. Hypertension is also one of important causes of CKD in developing countries, which leads to 13–21% of ESRD cases (Barsoum, 2006) (see Section 10 of this textbook). The status of hypertension is similar to diabetes, including escalating prevalence, low awareness rate, and suboptimal treatment, especially in rural area. For example, a national survey of hypertension in 1991 suggested that the overall prevalence of hypertension among people aged > 15 years in China was 13.6% (Tao et al., 1995). Ten years later, the number was reported to be 23% in urban areas and 18% in rural areas (Wu et al., 2008). The awareness rate and control rate of hypertension was 24% and 19% (Wu et al., 2008), which is lower than reported from developed countries. Chronic glomerulonephritis is an import cause of CKD in developing countries, based on the data from dialysis registry and hospital-based data (see Section 3 of this textbook). According to a report from the Dialysis and Transplantation Registration Group of China in 1999 (Dialysis and Transplantation Registration Group, 2001), 49.9% of patients receiving chronic dialysis were diagnosed as chronic glomerulonephritis. A recent report from Beijing, China (Beijing Hemodialysis Quality Control and Improvement Center, 2012)  indicated that chronic glomerulonephritis remained the leading cause of haemodialysis in 2011, especially among young patients. However, an increasing tendency of diabetic nephropathy was noticed among incident patients, especially in patients aged > 50 years. Data from Indonesia and Malaysia also revealed similar results (Liu and Hooi, 2007; Prodjosudjadi and Suhardjono, 2009). The histopathological types of glomerulonephritis in the developing countries vary considerably. Immunoglobulin A (IgA) nephropathy predominates in China, Southeast Asia, and the Pacific region. For example, a study involving 5398 consecutive patients receiving a renal biopsy indicated that IgA nephropathy comprised 50.7% of primary glomerulonephritis (Zhou et al., 2009). By contrast, focal segmental glomerulosclerosis is the most common type among the black populations of Africa (15–25%), Saudi Arabia (40%), India (up to 46%), and South America (up to 43%) (Barsoum, 2006). Infectious diseases also contribute to the burden of CKD in developing countries. The number of people infected with human immunodeficiency virus (HIV) is estimated to be 33 million, and 67% of them are in sub-Saharan Africa (AIDS Foundation, n.d.). Ninety per cent of new HIV infections are in developing countries (Piot and Tezzo, 1990). It is known that infection with HIV and its treatment can produce a variety of kidney diseases, including glomerular (e.g. collapsing focal segmental glomerulosclerosis), vascular (e.g. thrombotic microangiopathy), and tubulointerstitial disorders (Wyatt et al., 2008, 2009) (see also Chapter 187).

763

764

Section 5  

the patient with reduced renal function

In cross-sectional surveys, the prevalence of CKD with HIV infection varied substantially (from 3.5% to 48.5%) (Naicker and Fabian, 2010), partly due to different study populations and different methodologies. Besides HIV infection, other infectious diseases including malaria, schistosomiasis, hepatitis B and hepatitis C also contribute to the burden of CKD in developing countries (Hossain et al., 2009; Nugent et al., 2011) (see also Section 8 in this book). A significant proportion of the population in developing countries, especially in Asia and Africa, depends on indigenous local medical systems. A longitudinal study from Taiwan indicated that regular users of Chinese herbal medicines have a 20% increased risk of developing CKD (Wen et al., 2008). A recent study (Zhang et al., 2013) using a national representative sample in China indicated that long-term intake of herbs containing aristolochic acid was independently associated with eGFR < 60 mL/min/1.73 m2 and albuminuria, with an odds ratio of 1.83 (95% CI, 1.22–2.74) and 1.39 (95% CI, 1.03–1.87), respectively. Longitudinal studies from Taiwan revealed that use of aristolochic acid-containing herbs, especially > 60 g of Mu Tong or Fangchi from herbal supplements, is associated with increased risk of developing kidney failure (Lai et al., 2010). Despite improving maternal and infant mortality rates, a high prevalence of maternal malnutrition and low-birth-weight deliveries is still seen in developing countries. For example, it is reported that 30% of infants are underweight in India (Yajnik et al., 2009). And the percentage of babies not weighed or with unknown birth weight is high in developing countries, due to the absence of scales and trained staff (Goto, 2011). Low birth weight has been associated with later hypertension, congenital low nephron number, and accelerated kidney senescence, which would have an impact on kidney damage in the later adult life (Luyckx and Brenner, 2005; Luyckx et al., 2009). A meta-analysis indicates that low birth weight is associated with subsequent risk of CKD, with an odds ratio of 1.73 (95% CI 1.44–2.08) (White et al., 2009).

The health and economic burden of chronic kidney disease A recent editorial (The Lancet, 2013) and subsequent series of articles in The Lancet between July and August 2013, discuss many aspects related to the global burden of AKI and CKD, in particular low-income countries. One of the ultimate outcomes of CKD is ESRD, which necessitates ever-growing dialysis and transplantation programmes, and therefore places an unaffordable financial burden on developing countries. A recent national survey in China estimates the number of patients with CKD in China to be 120 million (Zhang et al., 2012). If 1% of them progress to ESRD, the total costs of dialysis would be twice of the current healthcare budget in China. Actually the dialysis rate is quite low in many developing countries compared to that in developed countries, which is limited by the affordability and accessibility of the treatment. A session of haemodialysis costs US$100 in Nigeria, twice the minimum monthly wage paid to federal government workers (Katz et  al., 2011). In India, the average cost of haemodialysis per year is around US$9000–14,000, while the average annual income in India is US$8000. Hence, affordability is a major constraint to dialysis treatment. Another obstacle is accessibility of treatment. For example, in China almost all haemodialysis centres are located in cities (Zhang et al., 2009).

Nowadays the majority of haemodialysis centres in major cities are believed to be running at capacity. In the last 2 years, peritoneal dialysis (PD) has grown at a rate of 30% annually in China (Zhang et al., 2008), but most of the PD centres are still located in cities. A recent cohort study (Xu et al., 2012) revealed that low personal income was independently associated with all-cause and cardiovascular death, as well as initial peritonitis in patients receiving PD. The situation is similar in India (Agarwal and Srivastava, 2009) and in Africa (Katz et al., 20110, which places a further burden on patients who often have to travel (often with families) to a dialysis centre. Furthermore, health equity (regarding both affordability and accessibility) remains a major challenge to policymakers in developing countries despite the resurgence of interest to promote it. The sheer inadequacy of financial and human resources for health and the progressive undermining of state capacity in many under-resourced settings have made it extremely difficult to promote and achieve significant improvements in equity in health and access to healthcare. The economic cost associated with milder forms of CKD was even higher. For instance, according to data from USRDS, costs for Medicare patients with CKD reached US$34 billion, and accounted for nearly 16% of total Medicare dollars in 2009 (USRDS, 2012). The expenditures further increased in presence of diabetes and heart failure (USRDS, 2012). Part of the high cost of CKD is driven by its close association with other non-communicable chronic diseases. Cardiovascular disease (CVD) and CKD share common risk factors such as hypertension, diabetes, smoking, obesity, hyperlipidaemia, and ageing as well as some non-traditional risk factors such as vitamin D deficiency, hyperphosphataemia, anaemia, albuminuria, and HIV infection (Sarnak et al., 2003). In a meta-analysis conducted by pooling 45 general population cohorts involving 105,872 individuals, eGFR and ACR are multiplicatively associated with risk of overall and cardiovascular (CV) mortality (Matsushita et al., 2010). Even CKD stages 1 or 2 are associated with an increased risk of adverse overall, CV, and renal outcomes (Matsushita et al., 2010; Gansevoort et al., 2011). Even though the Chinese population is thought to have a relatively lower risk of CVD (Ma et al., 2006; Zhang et al., 2012), a recent study (USRDS, 2011) indicated that individuals with subtle decreased renal function seem much more likely to have multiple CV risk factors and have higher prevalence of CVD than those without CKD. Arterial changes such as elevated carotid artery intima-media thickness were observed in early stage CKD patients (Modi and Jha, 2006). Furthermore, Chinese studies revealed that CKD is associated with high burden of other chronic conditions, including metabolic syndrome (Zhang et  al., 2007), cognitive decline (Wang et al., 2010), and ocular fundus pathology (Gao et al., 2011), which all contribute to the adverse patients’ outcome and to elevated healthcare costs.

Prevention and early detection of chronic kidney disease In the developing world, CKD prevention programmes are relatively non-existent. Although CKD shares common risk factors and/or coexists with other NCDs, it has not received the same kind of attention. Lifestyle intervention for common risk factors for NCDs is important in the prevention of CKD. Several priority interventions were chosen to cope with the global NCD crisis, including

Chapter 96 

accelerated tobacco control, salt intake reduction, promotion of healthy diets and physical activity, and reduction of harmful alcohol consumption (Beaglehole et al., 2011). Those interventions are cost-effective in countries with a variety range of incomes (Gaziano et al., 2007). Cost-effective lifestyle and behavioural changes can be achieved on a population basis through legislation, government influence, manufacturing changes, mass education campaigns, and bans on negative advertisements (Gaziano et al., 2007). In addition, optimal control of diabetes and hypertension should be pursued (Barsoum et al., 2006). Based on data from developed countries, screening for CKD among ‘high-risk’ population (e.g. aged > 60 years or with hypertension or diabetes) is proved to be cost-effective (Boulware et al., 2003). It has been shown that community-based screening programmes in developing countries are feasible (Perico et al., 2009), provided the screening tools are simple and cheap. Even if a urinary dipstick test is a less precise measure of albuminuria, it is still useful for risk stratification and initial screening (Matsushita et al., 2010). Studies in developed countries on the cost-effectiveness of tertiary prevention of CKD by treatment of hypertension, albuminuria, and use of renin–angiotensin system inhibitors have shown that early intervention is more cost-effective than late intervention (Palmer et al., 2000, 2004; Ruggenenti et al., 2001). Therefore, screening for CKD among high-risk populations is warranted, especially considering the low awareness rate of CKD in developing countries (Zhang et al., 2012). Along these lines, the Research and Prevention Committee of the International Society of Nephrology  – Global Outreach (ISN-GO) has developed a global early detection and intervention programme for emerging countries that can be implemented according to the peculiar needs and organization facilities of the given country (Perico et al., 2009). In view of the close linkage and overlapping management strategies, programmes to combat CKD, diabetes, hypertension, and CVD need to be closely integrated in developing countries. As experience from Indian medical society has shown, these integrated early stages of prevention and management could be performed at low cost by medical assistants and nurses. However, in the face of an immature primary care system in developing countries, the involvement of a nephrologist is still needed to train the primary care medical care level and to establish a referral system. Studies are required to analyse successful experiences and to examine the cost-effectiveness of such approaches in different countries.

References Agarwal, S. K., Dash, S. C., Irshad, M., et al. (2005). Prevalence of chronic renal failure in adults in Delhi, India. Nephrol Dial Transplant, 20(8), 1638–42. Agarwal, S. K. and Srivastava, R. K. (2009). Chronic kidney disease in India: challenges and solutions. Nephron Clin Pract, 111(3), c197–203. AIDS Foundation (n.d.). HIV/AIDS in South Africa. [Online] Amato, D., Alvarez-Aguilar, C., Castaneda-Limones, R., et al. (2005). Prevalence of chronic kidney disease in an urban Mexican population. Kidney Int Suppl, 97, S11–17. Barsoum, R. S. (2006). Chronic kidney disease in the developing world. N Engl J Med, 354(10), 997–9. Barton, E. N., Sargeant, L. A., Samuels, D., et al. (2004). A survey of chronic renal failure in Jamaica. West Indian Med J, 53(2), 81–4. Beaglehole, R., Bonita, R., Horton, R., et al. (2011). Priority actions for the non-communicable disease crisis. Lancet, 377(9775), 1438–47.

chronic kidney disease in developing countries

Beijing Hemodialysis Quality Control and Improvement Center (2012). The Annual Data Report of Beijing Hemodialysis Quality Control and Improvement Center, 2012. Chinese Journal of Blood Purification, z1, 42. Bello, A. K., Nwankwo, E., and El Nahas, A. M. (2005). Prevention of chronic kidney disease: a global challenge. Kidney Int Suppl, 98, S11–17. Boulware, L. E., Jaar, B. G., Tarver-Carr, M. E., et al. (2003). Screening for proteinuria in US adults: a cost-effectiveness analysis. JAMA, 290(23), 3101–14. Chadban, S. J., Briganti, E. M., Kerr, P. G., et al. (2003). Prevalence of kidney damage in Australian adults: The AusDiab kidney study. J Am Soc Nephrol, 14(7 Suppl 2), S131–8. Chan, J. C., Malik, V., Jia, W., et al. (2009). Diabetes in Asia: epidemiology, risk factors, and pathophysiology. JAMA, 301(20), 2129–40. Chen, J., Wildman, R. P., Gu, D., et al. (2005). Prevalence of decreased kidney function in Chinese adults aged 35 to 74 years. Kidney Int, 68(6), 2837–45. Chen, N., Wang, W., Huang, Y., et al. (2009). Community-based study on CKD subjects and the associated risk factors. Nephrol Dial Transplant, 24(7), 2117–23. Chen, W., Wang, H., Dong, X., et al. (2009). Prevalence and risk factors associated with chronic kidney disease in an adult population from southern China. Nephrol Dial Transplant, 24(4), 1205–12. Coresh, J., Selvin, E., Stevens, L. A., et al. (2007). Prevalence of chronic kidney disease in the United States. JAMA, 298(17), 2038–47. Cusumano, A. M. and Gonzalez Bedat, M. C. (2008). Chronic kidney disease in Latin America: time to improve screening and detection. Clin J Am Soc Nephrol, 3(2), 594–600. Dialysis and Transplantation Registration Group (2001). The report about the registration of dialysis and transplantation in China 1999. Chin J Nephrol, 17, 77–9. Domrongkitchaiporn, S., Sritara, P., Kitiyakara, C., et al. (2005). Risk factors for development of decreased kidney function in a southeast Asian population: a 12-year cohort study. J Am Soc Nephrol, 16(3), 791–9. Escobedo, J., Buitron, L. V., Velasco, M. F., et al. (2009). High prevalence of diabetes and impaired fasting glucose in urban Latin America: the CARMELA Study. Diabet Med, 26(9), 864–71. Gansevoort, R. T., Matsushita, K., van der Velde, M., et al. (2011). Lower estimated GFR and higher albuminuria are associated with adverse kidney outcomes in both general and high-risk populations. A collaborative meta-analysis of general and high-risk population cohorts. Kidney Int, 80(1), 93–104. Gao, B., Zhu, L., Pan, Y., et al. (2011). Ocular fundus pathology and chronic kidney disease in a Chinese population. BMC Nephrol, 12, 62. Gaziano, T. A., Galea, G., and Reddy, K. S. (2007). Scaling up interventions for chronic disease prevention: the evidence. Lancet, 370(9603), 1939–46. Go, A. S., Chertow, G. M., Fan, D., et al. (2004). Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med, 351(13), 1296–305. Goto, E. (2011). Meta-analysis: identification of low birthweight by other anthropometric measurements at birth in developing countries. J Epidemiol, 21(5), 354–62. Hallan, S. I., Coresh, J., Astor, B. C., et al. (2006). International comparison of the relationship of chronic kidney disease prevalence and ESRD risk. J Am Soc Nephrol, 17(8), 2275–84. Hossain, M. P., Goyder, E. C., Rigby, J. E., et al. (2009). CKD and poverty: a growing global challenge. Am J Kidney Dis, 53(1), 166–74. Ingsathit, A., Thakkinstian, A., Chaiprasert, A., et al. (2010). Prevalence and risk factors of chronic kidney disease in the Thai adult population: Thai SEEK study. Nephrol Dial Transplant, 25(5), 1567–75. Katz, I. J., Gerntholtz, T., and Naicker, S. (2011). Africa and nephrology: the forgotten continent. Nephron Clin Pract, 117(4), c320–7. Lai, M. N., Lai, J. N., Chen, P. C., et al. (2010). Risks of kidney failure associated with consumption of herbal products containing Mu Tong or Fangchi: a population-based case-control study. Am J Kidney Dis, 55(3), 507–18.

765

766

Section 5  

the patient with reduced renal function

Levitt, N. S., Katzenellenbogen, J. M., Bradshaw, D., et al. (1993). The prevalence and identification of risk factors for NIDDM in urban Africans in Cape Town, South Africa. Diabetes Care, 16(4), 601–7. Liu, W. J. and Hooi, L. S. (2007). Patients with end stage renal disease: a registry at Sultanah Aminah Hospital, Johor Bahru, Malaysia. Med J Malaysia, 62(3), 197–200. Luyckx, V. A. and Brenner, B. M. (2005). Low birth weight, nephron number, and kidney disease. Kidney Int Suppl, 97, S68–77. Luyckx, V. A., Compston, C. A., Simmen, T., et al. (2009). Accelerated senescence in kidneys of low-birth-weight rats after catch-up growth. Am J Physiol Renal Physiol, 297(6), F1697–705. Ma, Y. C., Zuo, L., Chen, J. H., et al. (2006). Modified glomerular filtration rate estimating equation for chinese patients with chronic kidney disease. J Am Soc Nephrol, (10), 2937–44. Mani, M. K. (2006). Nephrologists sans frontieres: preventing chronic kidney disease on a shoestring. Kidney Int, 70(5), 821–3. Matsushita, K., van der Velde, M., Astor, B. C., et al. (2010). Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet, 375(9731), 2073–81. Mbanya, J. C., Motala, A. A., Sobngwi, E., et al. (2010). Diabetes in sub-Saharan Africa. Lancet, 375(9733), 2254–66. Modi, G. K. and Jha, V. (2006). The incidence of end-stage renal disease in India: a population-based study. Kidney Int, 70(12), 2131–3. Motala, A. A., Esterhuizen, T., Gouws, E., et al. (2008). Diabetes and other disorders of glycemia in a rural South African community: prevalence and associated risk factors. Diabetes Care, 31(9), 1783–8. Naicker, S. and Fabian, J. (2010). Risk factors for the development of chronic kidney disease with HIV/AIDS. Clin Nephrol, 74 Suppl 1, S51–6. National Kidney Foundation (2002). K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis, 39(2 Suppl 1), S1–266. Nugent, R. A., Fathima, S. F., Feigl, A. B., et al. (2011). The burden of chronic kidney disease on developing nations: a 21st century challenge in global health. Nephron Clin Pract, 118(3), c269–c277. O’Donnell, J. K., Tobey, M., Weiner, D. E., et al. (2011). Prevalence of and risk factors for chronic kidney disease in rural Nicaragua. Nephrol Dial Transplant, 26(9), 2798–805. Palmer, A. J., Annemans, L., Roze, S., et al. (2004). An economic evaluation of the Irbesartan in Diabetic Nephropathy Trial (IDNT) in a UK setting. J Hum Hypertens, 18(10), 733–8. Palmer, A. J., Weiss, C., Sendi, P. P., et al. (2000). The cost-effectiveness of different management strategies for type I diabetes: a Swiss perspective. Diabetologia, 43(1), 13–26. Perico, N., Bravo, R. F., De Leon, F. R., et al. (2009). Screening for chronic kidney disease in emerging countries: feasibility and hurdles. Nephrol Dial Transplant, 24(5), 1355–8. Piot, P. and Tezzo, R. (1990). The epidemiology of HIV and other sexually transmitted infections in the developing world. Scand J Infect Dis Suppl, 69, 89–97. Prodjosudjadi, W. and Suhardjono, A. (2009). End-stage renal disease in Indonesia: treatment development. Ethn Dis, 19(1 Suppl 1), S1-33–36. Ramachandran, A. (2005). Epidemiology of diabetes in India–three decades of research. J Assoc Physicians India, 53, 34–8. Ruggenenti, P., Pagano, E., Tammuzzo, L., et al. (2001). Ramipril prolongs life and is cost effective in chronic proteinuric nephropathies. Kidney Int, 59(1), 286–94. Sarnak, M. J., Levey, A. S., Schoolwerth, A. C., et al. (2003). Kidney disease as a risk factor for development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention. Circulation, 108(17), 2154–69. Singh, N. P., Ingle, G. K., Saini, V. K., et al. (2009). Prevalence of low glomerular filtration rate, proteinuria and associated risk factors

in North India using Cockcroft-Gault and Modification of Diet in Renal Disease equation: an observational, cross-sectional study. BMC Nephrol, 10, 4. Tao, S., Wu, X., Duan, X., et al. (1995). Hypertension prevalence and status of awareness, treatment and control in China. Chin Med J (Engl), 108(7), 483–9. The Lancet (2013). The global issue of kidney disease. Lancet, 382 (9887), 101. United States Renal Data System (2012). USRDS 2011 Annual Data Report. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Varma, P. P., Raman, D. K., Ramakrishnan, T. S., et al. (2010). Prevalence of early stages of chronic kidney disease in apparently healthy central government employees in India. Nephrol Dial Transplant, 25(9), 3011–17. Wagner, K. H. and Brath, H. (2012). A global view on the development of non-communicable diseases. Prev Med, 54 Suppl, S38–41. Wang, F., Zhang, L., Liu, L., et al. (2010). Level of kidney function correlates with cognitive decline. Am J Nephrol, 32(2), 117–21. Wen, C. P., Cheng, T. Y., Tsai, M. K., et al. (2008). All-cause mortality attributable to chronic kidney disease: a prospective cohort study based on 462 293 adults in Taiwan. Lancet, 371(9631), 2173–82. Wheeler, D. C. and Becker, G. J. (2013). Summary of KDIGO guideline. What do we really know about management of blood pressure in patients with chronic kidney disease? Kidney Int, 83(3), 377–83. White, S. L., Perkovic, V., Cass, A., et al. (2009). Is low birth weight an antecedent of CKD in later life? A systematic review of observational studies. Am J Kidney Dis, 54(2), 248–61. Wu, Y., Huxley, R., Li, L., et al. (2008). Prevalence, awareness, treatment, and control of hypertension in China: data from the China National Nutrition and Health Survey 2002. Circulation, 118(25), 2679–86. Wyatt, C. M., Morgello, S., Katz-Malamed, R., et al. (2009). The spectrum of kidney disease in patients with AIDS in the era of antiretroviral therapy. Kidney Int, 75(4), 428–34. Wyatt, C. M., Rosenstiel, P. E., and Klotman, P. E. (2008). HIV-associated nephropathy. Contrib Nephrol, 159, 151–61. Xu, R., Han, Q. F., Zhu, T. Y., et al. (2012). Impact of individual and environmental socioeconomic status on peritoneal dialysis outcomes: a retrospective multicenter cohort study. PLoS One, 7(11), e50766. Yajnik, C. S. (2009). Nutrient-mediated teratogenesis and fuel-mediated teratogenesis: two pathways of intrauterine programming of diabetes. Int J Gynaecol Obstet, 104 Suppl 1, S27–31. Yang, W., Lu, J., Weng, J., et al. (2010). Prevalence of diabetes among men and women in China. N Engl J Med, 362(12), 1090–101. Zhang, J., Zhang, L., Wang, W., et al. (2013). Association between aristolochic acid and CKD: a cross-sectional survey in China. Am J Kidney Dis, 61(6), 918–22. Zhang, L., Wang, F., Wang, L., et al. (2012). Prevalence of chronic kidney disease in China: a cross-sectional survey. Lancet, 379(9818), 815–22. Zhang, L. and Wang, H. (2009). Chronic kidney disease epidemic: cost and health care implications in China. Semin Nephrol, 29(5), 483–6. Zhang, L., Zhang, P., Wang, F., et al. (2008). Prevalence and factors associated with CKD: a population study from Beijing. Am J Kidney Dis, 51(3), 373–84. Zhang, L., Zuo, L., Wang, F., et al. (2007). Metabolic syndrome and chronic kidney disease in a Chinese population aged 40 years and older. Mayo Clin Proc, 82(7), 822–7. Zhang, Q. L. and Rothenbacher, D. (2008). Prevalence of chronic kidney disease in population-based studies: systematic review. BMC Public Health, 8, 117. Zhong, X. L. (1982). Diabetes mellitus survey in China. Chin Med J (Engl), 95(6), 423–30. Zhou, F. D., Zhao, M. H., Zou, W. Z., et al. (2009). The changing spectrum of primary glomerular diseases within 15 years: a survey of 3331 patients in a single Chinese centre. Nephrol Dial Transplant, 24(3), 870–6.

CHAPTER 97

Chronic kidney disease long-term outcomes: progression, death, cardiovascular disease, infections, and hospitalizations Monica Beaulieu, Catherine Weber, Nadia Zalunardo, and Adeera Levin Progression to end-stage renal disease (dialysis or transplantation) Data describing progression to renal replacement therapy (RRT) comes from large observational cohort studies in Australia, Canada, the United Kingdom, and the United States (Hsu et al., 2002; Iseki et al., 1996; Wen et al., 2008; Jafar et al., 2009; White et al., 2010; McDonald et al., 2010; Jolly et al., 2011; Hallan et al., 2012). Of note, in general or non-referred cohorts, the progression to end-stage renal disease (ESRD) is much lower than in those patients known to nephrologists. It is important to consider the patient population in describing the probability of progression. The Chronic Kidney Disease Prognosis Consortium (CKD-PC) has described the probability of ESRD in general populations, high-risk populations, and CKD populations (Matsushita et al., 2010; Astor et al., 2011; Gansevoort et al., 2011; Levey et al., 2011; Nitsch et al., 2013). On aggregate, there is the highest probability of progression in the CKD group, with the expected dose–response reduced risk in those in high-risk and general populations, respectively. The CKD-PC has used the parameters of proteinuria and estimated glomerular filtration rate (eGFR) to develop risk ratios for important outcomes, including progression to ESRD (Matsushita et  al., 2010; Van der Velde et  al., 2011)  in 45 different cohorts obtained from clinical trials, clinical databases, and administrative datasets. The findings therein have been corroborated in multiple other studies: those with the highest levels of proteinuria are at greatest risk for progression, irrespective of eGFR, those with lower eGFR values are also at risk, but those with intermediate values of eGFR and little or no proteinuria do not appear to be at risk for progression (Iseki et al., 2003; Halbesma et al., 2006; Hemmelgarn et al., 2006, 2010). In referred cohorts, those known to nephrologists, it appears that approximately 70% of patients demonstrate some rate of decline in eGFR, of which a much lesser number end up on dialysis in any given year (Keith et al., 2004; Levin et al., 2008; Astor et al., 2011). Over the long term, predictors of progression include male

gender, younger age, heavy proteinuria, lower eGFR, diabetes, and hypertension. Recent prediction models published have included additional laboratory variables such as phosphate, bicarbonate, and albumin to improve the precision of the model, but from a clinical standpoint, the first six clinical variables are easily incorporated into clinical decision-making as simple parameters (Hunsicker et al., 1997; Keane et al., 2006; Halbesma et al., 2011; Tangri et al., 2011). The search for additional and better prediction models and refinement of existing ones is ongoing. There are ethnic differences in progression rates, whereby African Americans, Asians, Aboriginal peoples, and Hispanics all appear to have faster rates of decline than Caucasians. While this may be confounded by socioeconomic factors and access to care, there is accruing data that true genetic factors play an important role (Hallan et al., 2006; Barbour et al., 2008, 2010; Pakov et al., 2008; Bui et al., 2009; Chen et al., 2009; Sood et al., 2010; Conley et al., 2012). The diversity of outcomes of people with CKD cannot be overstated. There is remarkable variability in outcomes of patients at each stage of CKD, and the trajectories of progression are not linear. This leads to confusion among patients and care providers, and difficulties in decision-making and planning. The trajectories can be impacted by hospitalizations, episodes of acute kidney injury (AKI) which often but not necessarily occur in the context of hospitalizations, infections, and cardiovascular (CV) events (either natural or investigation related). Thus, the complexity of the prediction of progression to ESRD is underscored by the events that occur to people with CKD, which impact both the trajectory as well as the probability of surviving to ‘achieve’ that outcome. Li et al. describe a number of different trajectories, as a function of time of follow-up, in that the longer people are followed, the more variable is the trajectory over time; O’Hare et al. describe four trajectories using dialysis within 2 years as the outcome; others have described non-progression over time (John et  al., 2004; Imai et  al., 2008; Li et al., 2012; O’Hare et al., 2012) (see Figs 97.1 and 97.2).

Section 5  

the patient with reduced renal function 63 50

53

20 10

40

43

eGFR

30

eGFR

eGFR

40

33

10 0

13 2

0

4 6 Year

8

30 20

23

0

2

0

4

6 Year

8

10

0

1

2

3 4 Year

5

6

7

Fig. 97.1  Variability in trajectories of change in renal function over time. From Li et al. (2012).

100 Quarterly median eGFR (mL/min/1.73 m2)

768

90

Group 4

80 70

Group 3

60 50 Group 2

40 30

Group 1

20

25 20 15 10 5 0

0 8

7

6

5 4 3 Quarters to first dialysis

2

1

0

Fig. 97.2  Different trajectories in those who achieved dialysis within 2 years. From O’Hare et al. (2012).

In the following sections, the probability of these outcomes, their predictors, and interactions with other factors are described in more detail.

Markers of progression While progression has long been thought to be associated with age, it is better understood as a consequence of the comorbidities that accrue as a function of ageing, and thus markers of ‘true progression’ have been sought (Lindeman, 1984; Fliser et al., 1997; Lindeman et  al., 1998). As has been discussed above, the rate at which this decline occurs varies according to the underlying population, presence of albuminuria/proteinuria, comorbidities, and exposures to nephrotoxins or AKI events (Coca et  al., 2009; La France et al., 2010; James et al., 2011). Nonetheless, there is accruing data which helps clinicians and researchers to identify true progression, within individuals, as opposed to biological variability or simple, slow ‘age’-related decline. In a general population study, Prevention of Renal and Vascular End-Stage Disease (PREVEND), decline in kidney function at the population level is described and reports eGFR decline which is variable as a function of proteinuria (Halbesma et al., 2006). The variability in rates of progression was closely linked to the amount of proteinuria:  those with no proteinuria experienced a loss in

eGFR of 0.2 mL/min/1.73 m2 over 4 years whereas those with an elevated urinary albumin to creatinine ratio (UACR) experienced a much faster rate of 2.3–7.2 mL/min/1.73m2 over 4  years. In a Japanese general population, followed for > 10 years, the decline was less at 0.36 mL/min/1.73m2 (Imai et al., 2008). Proteinuria is recognized as an important marker of progression, as has been demonstrated in numerous studies (Halbesma et  al., 2006; Imai et al., 2008; Hemmelgarn et al., 2010). In general, the studies suggest progression rates of approximately 0.3–1 mL/min/1.73m2 per year among participants without proteinuria or co-morbidity, with much higher rates (two to three times) in those with any degree of with proteinuria or co-morbidity. There are some population-based studies, identifying those with impaired kidney function, but in the absence of proteinuria measurements, which describe similarly low rates of renal decline (John et al., 2004; Halbesma et al., 2006). This apparent paradox is best understood in the context of the populations studied (i.e. general populations versus referred cohorts). Additional predictors of progression, above and beyond the aetiology of CKD, include the presence of hyperphosphataemia, hyperuricaemia, dyslipidaemia, and acidosis. Many of these would be considered to be markers of severity of CKD, and thus it is potentially difficult to sort what is ‘chicken and egg’ with respect to marker of severity versus marker of progression per se. Nonetheless, biological and experimental data do support these metabolic and laboratory abnormalities in the context of progression (Krowlewski et al., 1994; Ravid et al., 1998; Iseki et al., 2001; Obermayr et al., 2008; Kovedsky et al., 2009; Menon et al., 2010; Tangri et al., 2011). Cohort studies have described more rapid progression in those with hyperuricaemia and acidosis, and small studies of interventions for these two abnormalities, either alone or together, have demonstrated promising results to attenuate reduction in GFR (Siu et al., 2006; Goicoechea et al., 2010). Larger studies will improve our understanding of both predictors of CKD progression and effective interventions. Markers of oxidative stress, inflammation, and fibrosis have been examined in the context of predicting progression, but are not yet appropriate for use in clinical care. They do, however, serve to inform pathological mechanisms and thus avenues for clinical interventions. For a comprehensive overview of biomarkers important in progression of CKD, the reader is referred to an excellent review (Fassett et al., 2011). N-terminal pro-brain natriuretic peptide (NT-pro-BNP) and troponin I have also been identified as predictive of progression (Desai et al., 2011). The role of specific molecules like asymmetric dimethylarginine (ADMA),

Chapter 97 

transforming growth factor beta (TGF-β), and interleukin 6 (IL-6) remain areas of active study (Roberts et al., 2006; Young et al., 2009; Baretto et al., 2010). Progressive kidney disease remains problematic for clinicians and researchers alike. Improved understanding of the predictors of progression, best interventions to slow rates of progression, and the impact of progressive disease on resource utilization and patient outcomes remain active areas of investigation.

Chronic kidney disease and risk of death CKD can lead to a significant reduction in lifespan. Most of this reduction is due to the increased burden of cardiovascular disease (CVD), but it has been increasingly recognized that CKD increases risk of non-cardiac death as well. The mortality rates on maintenance dialysis have been the most extensively reported but recent research has helped to quantify the increased risk of mortality attributed to CKD in patients not on dialysis (CKD-ND). In addition, the presence of albuminuria, and episodes of AKI have also been shown to impact mortality. This section will highlight what is currently known with respect to the increased risk of mortality due to kidney disease.

Risk of death in CKD patients not on dialysis The risk of death attributed to a reduced eGFR has been evaluated in several population-based studies (Garg et  al., 2002; Muntner et al., 2002; Go et al., 2004; Nitsch et al., 2013). In a large American study, > 1.1 million community-based adults with CKD-ND (eGFR < 60 mL/min/1.73 m2) were followed for a median of 2.84 years (Go et al., 2004). An independent, graded association was noted between a reduced eGFR and the risk of death. As eGFR declined, the risk of death from any cause increased sharply. The adjusted hazard ratios (HRs) for death were 1.2, 1.8, 3.2, and 5.9 for eGFR 45–59, 30–44, 15–29, and < 15 mL/min/1.73 m2, respectively. This translates to a mortality risk increase from 17% at an eGFR of 45–59 mL/min to a staggering 343% at an eGFR of < 15 mL/min. In 2006, a systematic review of the association between CKD-ND and the risk for all-cause and CV mortality was conducted (Tonelli et al., 2006). Thirty-nine studies that followed > 1.3 million participants were included. The unadjusted relative risk for mortality in those with CKD versus those without, ranged from 0.94 to 5.0. The absolute risk for death increased exponentially with decreasing function. Adjusting for common associated patient factors such as diabetes and hypertension reduced, but did not negate the increased risk of death. In this meta-analysis, CV deaths (representing 58% of deaths) were the largest driver for the increased risk of death. In 2010, the results of a large global collaborative meta-analysis were reported. Results of the CKD-PC highlighted that both lower eGFR and the presence of albuminuria independently predicted mortality in the general population independent of cardiac risk factors (Matsushita et al., 2010). In a collaborative meta-analysis of > 1 million patients in the general population cohort, the consortium found that lower eGFR and higher albuminuria were risk factors for all-cause and CV mortality, independent of each other and of CV risk factors. The HRs for all-cause mortality at eGFRs of 60, 45, and 15 mL/min/1.73 m2 were 1.18, 1.57, and 3.14 respectively when compared to an eGFR of 95 mL/min/1.73 m2. The presence of albuminuria also carried an increased risk of mortality. Compared with

chronic kidney disease long-term outcomes

UACR of 0.6 mg/mmol, adjusted HRs for all-cause mortality were 1.2, 1.63, and 2.22 for UACR 1.1, 3.4, and 33.9 mg/mmol respectively. Subsequently, the analysis of cohorts at risk for CKD (Astor et al., 2011) or with CKD (van der Velde et al., 2011) has demonstrated similar associations. In addition, a recent meta-analysis of these cohorts assessed for the presence of a sex interaction in the associations of eGFR and UACR with all-cause mortality, CV mortality, and ESRD (Nitsch et al., 2013). This study demonstrated that both sexes face an increased risk of all-cause mortality, CV mortality, and ESRD with reduced eGFR and increasing UACR. Importantly, the association of kidney disease measures, eGFR, and albuminuria with mortality or ESRD has also been consistently found in those with or without hypertension (Mahmoodi et  al., 2012), diabetes (Fox et al., 2012), and also regardless of age (Hallan et al., 2012). There are multiple possible explanations for the increased risk of death for patients with CKD. Prior CVD and known risk factors for CVD represent some of the risk. However, when adjusted for these known risks, CKD still carries a strong, independent, and graded risk of death. Several possible explanations have been given including endothelial dysfunction, inflammation, pro-coagulability, anaemia, left ventricular hypertrophy, arterial stiffness, and calcification (Levin et al., 1999; Hsu et al., 2002; Raggi et al., 2002; Shiplak et al., 2003, 2005; Muntner et  al., 2004). The answer probably lies in a combination of some, if not all of the above factors. The mechanism by which this multitude of factors leads to increased mortality in CKD is still poorly elucidated.

Risk of death in CKD patients on dialysis Although dialysis is a life-saving treatment for patients with ESRD, the mortality rate for those on dialysis remains exceedingly high. In general, becoming dialysis dependent carries a prognosis worse than most cancers. Although many factors must be taken into consideration, the lifespan of a 60–64-year-old starting dialysis in the United States is approximately 4.5  years and 8  years for a 40–44-year-old patient (United States Renal Data System (USRDS), 2009). Factors that have been associated with decreased survival on dialysis include burden of co-morbidities, length of time on dialysis (Chertow et al., 2000), and either low or high pre-dialysis potassium levels (Kovesdy et al., 2007). In terms of specific modality of dialysis (haemodialysis (HD) or peritoneal dialysis (PD)), data is still conflicting regarding whether one modality has improved long-term outcomes over another. After several trials to address this question, we do know that there is no large survival advantage between modalities of dialysis as conventionally prescribed. However, there is currently a significant amount of research looking at outcomes with short daily and nocturnal HD (Nesrallah et al., 2004; Nesrallah et al., 2006). Preliminary results have been encouraging, with studies showing an improved survival among patients using nocturnal and short daily compared to conventional dialysis in one recent study (Johansen et al., 2009). Another study showed that patients on nocturnal HD had a similar survival to patients who received a deceased donor transplant (Pauly et al., 2009). However, recent randomized trials investigating whether more frequent HD sessions compared to the conventional three-times-weekly regimen showed some clinical and/or biochemical advantages, but none of these trials could give answers to the question of whether frequent dialysis ameliorates long-term dialysis patient survival; none of them was indeed

769

770

Section 5  

the patient with reduced renal function

designed for addressing this most important outcome parameter (Lameire et al., 2012). CVD remains the major cause of death in this population, accounting for up to half of all deaths. Of the cardiac causes, sudden cardiac death (SCD), most commonly due to ventricular arrhythmias, is the most common specific cause of death, representing close to 27% of all-cause mortality in dialysis patients (USRDS 2010). It is well known that obstructive coronary artery disease is not the only contributor to SCD in dialysis patients. Other myocardial abnormalities, including left ventricular hypertrophy, endothelial dysfunction, and myocardial perfusion abnormalities are also thought to be factors (Young, 2011). Studies have also shown an enhanced risk of SCD in the first HD session of the week after a 2-day dialysis-free interval (Bleyer et al., 1999). This study compared the risk of SCD on Monday (for patients dialysing Monday, Wednesday, and Friday) or on Tuesday (for patients dialysing Tuesday, Thursday, and Saturday) and found a 50% higher risk on the Monday or Tuesday. An increased risk of SCD has also been shown in patients using a low potassium dialysate ( 3 mL/min/1.73 m2 based on cystatin C measurements found after adjustment for demographics, CVD risk factors, and baseline kidney function, that rapid kidney function decline was significantly associated with heart failure (HR 1.32), myocardial infarction (HR 1.48), and peripheral arterial disease (PAD) (HR 1.67), and did not differ by the presence or absence of CKD at baseline (Shlipak et al., 2009). Both traditional and non-traditional risk factors likely contribute to the increased CV risk observed in CKD. Hypertension, smoking, diabetes, dyslipidaemia, and older age are highly prevalent in CKD populations (Abboud and Henrich, 2010). While they do not entirely explain the increased CV risk, they do appear to be the most important contributors in this population (Muntner et  al., 2005; Shlipak et al., 2005). In a large population-based cohort, traditional risk factors had an area under the curve (AUC) of 0.73 to explain CV mortality among older patients with eGFR < 60 mL/ min/1.73 m2; adding novel risk factors to the model only increased

Chapter 97 

the AUC to 0.74 (Shlipak et al., 2005). The relationship between certain risk markers and CV events in CKD stage 5D often differs from that in the general population, a phenomenon termed reverse epidemiology; whether the same is true in non-dialysis CKD remains unknown. Atherosclerotic disease may not follow the same pathways in people with CKD as compared to those with normal renal function, and there are several non-traditional risk factors unique to CKD that may contribute to or accelerate the CV risk. There is a multitude of observational data linking these non-traditional risk factors to poor CV outcomes throughout the spectrum of CKD; however, none have been proven a necessary cause in CVD. Purported non-traditional risk factors include uric acid, sympathetic over-activity, bone mineral metabolism (serum phosphate concentration, calcium loading, and more recently FGF23 (Isakova et  al., 2011)), anaemia, arterial stiffness, chronic overhydration, inflammatory markers (especially ADMA and IL-6), uraemic toxins, and oxidative stress. Outcomes after an acute coronary syndrome (unstable angina and myocardial infarction) in CKD patients are worse compared to the non-CKD population. Observational studies and a large meta-analysis in the general population demonstrate that in those with a cardiac event, short- and long-term CV morbidity and mortality are inversely and independently associated with kidney function, especially at eGFR < 15 mL/min/1.73 m2 (Shlipak et  al., 2002; Anavekar et  al., 2004; Cardinal et al., 2010; Campbell et al., 2012). Observational data also consistently shows increased risk of serious operative complications in CKD patients. The incidence of postoperative death after coronary artery bypass grafting is three- to sevenfold higher in CKD stages 4–5ND than in non-CKD patients (Cooper et al., 2006).

Non-atherosclerotic CVD considerations in CKD Congestive heart failure (CHF) is very common in CKD patients (USRDS, 2004). The mechanisms that lead to CHF are pressure overload, volume overload, and CKD-associated non-haemodynamic factors that affect the myocardium (Ronco et  al., 2009). CKD is associated with increased mortality in heart failure (Smith et al., 2006), with a higher mortality observed in diastolic compared to systolic heart failure in CKD patients (Ahmed 2007). An increased risk of stroke is observed in CKD patients (USRDS, 2009). A large meta-analysis found that an eGFR < 60mL/min/1.73 m2 was associated with an increased risk of incident stroke (pooled relative risk 1.43) compared to those with eGFR > 90 mL/min/1.73 m2 (Lee et al., 2010). Proteinuria or albuminuria also increases stroke risk. CKD also portends a worse prognosis after stroke; the risk of fatal stroke is higher than overall stroke risk in those with eGFR

25,000 (27% with CKD) from Canada. Over a median follow-up of 3.2  years, 3.1% were hospitalized with bloodstream infection. Compared to those with Modification of Diet in Renal Disease (MDRD; MDRD Study Group, 1991) formula-estimated GFR ≥ 60 mL/min/1.73 m2, the adjusted HRs for death were 1.34, 1.61, and 4.1 for GFR 45–59, 30–44, and < 30 mL/min/1.73 m2, respectively. Results for hospitalization were very similar (James et al., 2008). Risk of pneumonia-related hospitalization and death increased with declining MDRD formula-estimated GFR in a second retrospective cohort study from Canada of > 250,000 adults (aged 18 or greater with at least one measure of creatinine available in 2003 or 2004)  followed for a median of 2.5  years. Not unexpectedly, the elderly accounted for the greatest absolute number of hospital admissions and deaths; however, the relative risk of these outcomes was greatest in younger individuals (James et al., 2009). Rates of all infection-related hospitalizations have been reported in US cohorts where CKD was defined using diagnosis codes. For elderly US Medicare beneficiaries (aged 66 or greater) in 2009, the adjusted hospital admission rate due to any infection was 1.5 times greater in those with CKD (56 per 1000 patient-years in the non-CKD population versus 82 per 1000 patient-years in non-dialysis-dependent CKD). Hospitalization rates increased as CKD worsened:  from 69 per 1000 patient-years in stages 1 and 2 CKD, to 78 per 1000 patient-years in stage 3 CKD, to 107 per 1000 patient-years in stages 4 and 5 (non-dialysis) CKD. Similar trends were observed regardless of the site of primary infection (skin, circulatory, lung, genitourinary, musculoskeletal, and abdominal) (USRDS, 2011). Two recent reports have reinforced these findings. In an analysis of the Cardiovascular Health Study participants (5142 people aged 65  years or greater, excluding GFR < 15 mL/min/1.73 m2), 30% were hospitalized at least once for infection over a median follow-up period of 11.5 years. Reduced kidney function, in this study estimated using a cystatin C-based method, was again associated with a graded increase in adjusted risk of all infectious hospitalizations, including pulmonary and genitourinary infections (Dalrymple et al., 2012). Examination of infection-related mortality among NHANES III participants yielded very similar results. In addition, heavier proteinuria was a risk factor for hospitalization due to infection in this study (Wang et al. 2011).

Infections in CKD patients on dialysis Infection is often listed as the second leading cause of mortality in CKD-D (Foley, 2007). It accounted for about 20% of deaths in dialysis patients in the United Kingdom in 2009 (Caskey et al. 2010). The first few months following dialysis initiation appear to be a particularly high-risk period for infectious hospitalizations, especially for HD patients (Collins et al., 2009). The risk of septicaemia or bacteraemia varies by dialysis and vascular access type. In HD, they occur more frequently when temporary or cuffed catheters are used than when an arteriovenous graft of

fistula (AVG or AVF) is used. Rates of sepsis and bacteraemia in PD are similar to those reported in HD with an AVF (Ishani et al., 2005). Reported rates of catheter-related bacteraemia in HD vary between and within countries, but range from < 1 to 4 per 1000 catheter days in most reports. While catheter use for HD remains a concern in certain countries, it should be remembered that that the majority of septic events are accounted for by non-vascular access-related causes. For example, among the 1846 participants in the prospective HEMO Study, only one-fifth of all infection-related first hospitalizations were felt to be related to vascular access (Allon et al., 2003). Vascular access-related infections in HD are discussed in detail elsewhere (see Chapter 269). In comparison to the general population, mortality rates from sepsis in dialysis are alarmingly high. In the United States for the years 1994–1996, the adjusted annual mortality was between 100 and 300 times that observed from sepsis in the general population. The impact is especially significant in younger individuals (Sarnak and Jaber, 2000). Peritonitis is a major infectious complication of PD, and is discussed in detail elsewhere (see Chapter  266). Reported rates of PD-associated peritonitis vary by jurisdiction. In a large North American survey, peritonitis occurred at a rate of about one episode per 30 patient-months (Mujais, 2006). Mortality from an episode of peritonitis is < 4%, and peritonitis accounts for about 20% of infection-related deaths on PD (Li et al., 2010). The significance of this finding is highlighted in Australia and New Zealand, where the risk of infection-related death was found to be greater in PD than in HD, largely due to an excess of deaths from bacterial and fungal peritonitis episodes (Johnson et al., 2009). Pneumonia (inpatient and outpatient) is common in dialysis-dependent CKD, with a reported incidence of 21% (27.9 per 100 patient-years) within the first year of dialysis initiation in US Medicare beneficiaries starting dialysis between 1996 and 2001 (hospitalization rate 6.1 per 100 patient-years). The cumulative 1-year survival following pneumonia was only 51% at 1 year. Interestingly, the risk of pneumonia was higher in HD than PD patients (29.0 compared to 18.2 per 100 patient-years, respectively) in this report (Guo et al., 2008). In contrast, no difference in pneumonia rates between modalities was reported in Australia and New Zealand (Johnson et al., 2009). In a US study comparing pulmonary infectious mortality in dialysis to the general population, the annual mortality rate in all age groups combined was reported to be 14–16 times greater in those on dialysis (Sarnak and Jaber, 2001).

Long-term implications of serious infections in CKD In addition to the immediate risks posed by infections, these events also portend a poor longer-term prognosis in the subsequent months and years. In US dialysis patients, increased risks of CV morbidity (including heart failure, myocardial infarction, and stroke) and death have been reported up to 5 years following an infectious event (Ishani et al., 2005; Guo et al., 2008). Inflammatory pathways may be the pathophysiologic link between infections and adverse CV outcomes (Foley, 2007).

Opportunities for prevention of infections in CKD In view of the high risk of infectious events in CKD and the associated poor outcomes, efforts to prevent infections are extremely

Chapter 97 

important. This includes addressing modifiable individual risk factors (e.g. nutrition), reducing pathogen exposure (e.g. reducing central venous catheter use in HD, avoiding hospitalization and procedures where possible) and vaccination (e.g. influenza, pneumococcus, and hepatitis B) (Dalrymple and Go, 2008). The topic of vaccination in CKD is discussed in Chapter 128.

chronic kidney disease long-term outcomes

Hospitalizations

A recent analysis including > 10,000 community-dwelling elderly individuals (aged 75 years or greater) from 53 general practices in the United Kingdom supports the findings in the previous studies. Those with CKD-Epidemiology Collaboration formula-estimated GFR < 30 mL/min/1.73 m2 were at increased risk of hospitalization. Dipstick positivity for proteinuria was also independently associated with hospital admission over a 2-year follow-up period (Nitsch et al., 2011).

Risk of hospitalization

Hospitalization in CKD patients on dialysis

Given the multitude of complications associated with CKD, it is not surprising that hospitalization rates are high in this population, especially dialysis patients. The most common reason for hospital admission is CVD, followed by infection (USRDS, 2011). Hospitalization represents a major source of morbidity and healthcare expenditure throughout the entire CKD spectrum.

Among individuals with CKD, the hospitalization risk is highest for those on dialysis. The first 3 months after initiation of dialysis are a particularly high-risk period for hospitalization (Collins et al., 2009). This has been attributed in part to the high frequency of catheter use in incident HD patients. As in the non-dialysis-dependent CKD population, the most common reason for hospital admission is CVD, followed by infection (USRDS, 2011). In US prevalent dialysis patients, the adjusted all-cause hospitalization rate has remained fairly stable over the past decade, at about 1.9 per patient-year, although the infectious hospitalization rate has been steadily increasing. In 2009, the adjusted hospital admission rate for CVD was 0.55 per patient-year, and for infections was 0.47 per patient-year. Rehospitalization occurred in 36% within 30 days of discharge, indicating that this is a major cause of morbidity in dialysis patients. Overall, PD patients had fewer hospitalizations than their HD counterparts matched for various comorbidities and indicators of disease severity (USRDS, 2011). Hospital admission and readmission rates have recently been reported in the United Kingdom. In a retrospective, 5-year observational study in Northern Ireland, the median number of hospital admissions was three per HD patient. Similar to the observations reported by the USRDS, hospital admissions were especially common within the first 100 days of dialysis initiation, accounting for just over 50% of all hospital admissions. Importantly, duration of hospital stay was also noted to be far longer in HD patients by a factor of 3.75 (Quinn et al., 2011).

Hospitalization in CKD patients not on dialysis The increased risk of hospitalization with declining kidney function in CKD-ND has been described recently in several cohorts in the United States, Canada, and the United Kingdom. It has been consistently reported that the risk of hospital admission increases as GFR declines. Proteinuria is also an important independent risk factor for hospitalization. In an American study of > 1.1 million adults (mean age 52 years) in the Kaiser Permanente Renal Registry in California with at least one serum creatinine measured between 1996 and 2000 and followed for a median of 2.8 years, the adjusted HRs for any hospital admission were 1.1, 1.5, 2.1, and 3.1 for MDRD formula-estimated GFR 45–59, 30–44, 15–29, and < 15 (non-dialysis dependent) mL/min/1.73 m2 compared to ≥ 60 mL/min/1.73 m2, respectively (Go et al., 2004). Findings were similar in a report from the USRDS of elderly Medicare beneficiaries (age 66 years or greater). Overall hospital admission rates were three to five times higher in CKD (defined using diagnosis codes) than in those without a CKD diagnosis; however, with adjustment for disease severity (including comorbidities and prior hospitalization) the risk was 1.4 times greater. In 2009, the adjusted hospital admission rate for non-dialysis dependent CKD was 444 per 1000 patient-years compared to 318 per 1000 patient-years in those without CKD. Rates increased as CKD worsened (407 per 1000 patient-years in stages 1 and 2, 438 per 1000 patient-years in stage 3, 560 per 1000 patient-years in stage 4 and stage 5 (non-dialysis) CKD). The most common reasons for hospital admission were CVD (about 30%) and infections (about 20%) (USRDS, 2011). A series of studies of a community-based cohort from Alberta, Canada used a province-wide laboratory registry to determine the associations between MDRD formula-estimated GFR and proteinuria with the risk of hospitalization for various CV events and with AKI. About 1  million adults were included with a median follow-up of 35 months. Risk of hospitalization with AKI increased with declining kidney function, as did the risk of hospitalization for myocardial infarction, congestive heart failure, peripheral vascular disease, and transient ischaemic attack/cerebrovascular attack. Worsening proteinuria was also consistently and independently associated with increased risk of these outcomes, highlighting its importance as a risk factor for adverse events in CKD (Hemmelgarn et al., 2010; James et al., 2010; Bello et al., 2011).

Summary: chronic kidney disease long-term outcomes CKD, defined as reduction in kidney function or evidence of kidney damage present for a duration of 3 months or longer, has been definitively associated with poor long-term outcomes in numerous populations. While heterogeneity within populations is appreciated, the fact remains that those with CKD are at risk for progression of CKD, AKI, hospitalizations, infections, CVD (atherosclerotic and non-atherosclerotic), and death. Increasingly data have become available which identify both traditional and non-traditional risk factors for these adverse outcomes but few intervention trials have been conducted with help to guide therapy. Of note, an increasing commitment is observed within the international renal community to conduct large RCTs to answer important questions relevant for patients. Examples include the SHARP study (Study of Heart and Renal Protection), which determined that use of a combined lipid-lowering strategy of simvastatin and ezetimibe does reduce atherosclerotic events in those with CKD, with little adverse effects; the IDEAL study which determined that early start of dialysis was not associated with benefits and may have associated increased costs; and the TREAT study which determined that higher target

773

774

Section 5  

the patient with reduced renal function

haemoglobins using erythropoietic-stimulating agents in diabetic CKD-ND patients did not result in improved mortality and was associated with harm (Pfeffer et  al., 2009; Baigent et  al., 2011; Cooper et al., 2011). Thus, while we continue to better define high-risk populations, best interventions for delaying CKD progression, and complications such as infections, hospitalizations, and death, it would behove the community to review and apply results of well-conducted RCTs, and design future studies to address important clinical questions. These collaborative efforts will serve to improve understanding and outcomes in years to come.

References Abboud, H. and Henrich, W. L. (2010). Stage IV chronic kidney disease. N Engl J Med, 362, 56–65. Ahmed, A., Rich, M. W., Sanders, P. W., et al. (2007). Chronic kidney disease associated mortality in diastolic versus systolic heart failure: a propensity matched study. Am J Cardiol. 99, 393–8. Allon, M., Depner, T. A., Radeva, M., et al. (2003). Impact of dialysis dose and membrane on infection-related hospitalization and death: results of the HEMO Study. J Am Soc Nephrol, 14(7), 1863–70. Allon, M., Radeva, M., Bailey, J., et al. (2005). The spectrum of infection-related morbidity in hospitalized haemodialysis patients. Nephrol Dial Transplant, 20(6), 1180–6. Anavekar, N. S., McMurray, J. J., Velazquez, E. J., et al. (2004). Relation between renal dysfunction and cardiovascular outcomes after myocardial infarction. N Engl J Med, 351, 1285–95. Astor, B. C., Matsushita, K., Gansevoort, R. T., et al. (2011). Lower estimated glomerular filtration rate and higher albuminuria are associated with mortality and end-stage renal disease. A collaborative meta-analysis of kidney disease population cohorts. Kidney Int, 79(12), 1331–40. Baigent, C., Landray, M. J., Reith, C., et al. (2011). The effects of lowering LDL cholesterol with simvastatin plus ezetimibe in patients with chronic kidney disease (Study of Heart and Renal Protection): randomised placebo-controlled trial. Lancet, 377 (9784), 2181–92. Barbour, S. J., Er, L., Djurdjev, O., et al. (2008). The prevalence of hematologic and metabolic abnormalities during chronic kidney disease stages in different ethnic groups. Kidney Int, 74(1), 108–14. Barbour, S. J., Er, L., Djurdjev, O., et al. (2010). Differences in progression of CKD and mortality amongst Caucasian, Oriental Asian and South Asian CKD patients. Nephrol Dial Transplant, 25(11), 3663–72. Bello, A. K., Hemmelgarn, B., Lloyd, A., et al. (2011). Associations among estimated glomerular filtration rate, proteinuria, and adverse cardiovascular outcomes. Clin J Am Soc Nephrol, 6(6), 1418–26. Bleyer, A. J., Russell, G. P., and Satko, S. G. (1999). Sudden and cardiac death rates in hemodialysis patients. Kidney Int, 55(4), 1553–9. Bucaloiu, I. D., Kirchner, H. L., Norfolk, E. R., et al. (2011). Increased risk of death and de novo chronic kidney disease following reversible acute kidney injury. Kidney Int, 81(5), 477–85. Bui, A. L., Katz, R., Kestenbaum, B., et al. (2009). Cystatin C and carotid intima-media thickness in asymptomatic adults: the Multi-Ethnic Study of Atherosclerosis (MESA). Am J Kidney Dis, 53(3), 389–98. Campbell, N. G., Varagunam, M., Sawhney, V., et al. (2012). Mild chronic kidney disease is an independent predictor of long-term mortality after emergency angiography and primary percutaneous intervention in patients with ST-elevation myocardial infarction. Heart, 98, 42–7. Cardinal, H., Bogaty, P., Madore, F. et al. (2010). Therapeutic management in patients with renal failure who experience an acute coronary syndrome. Clin J Am Soc Nephrol, 5, 87–94. Caskey, F., Dawnay, A., Farrington, K., et al. (2011). UK Renal Registry, Bristol, UK. UK Renal Registry 2010. 13th Annual Report of the Renal Association. Nephron Clin Pract, 119, Suppl 2. Chen, N., Wang, W., Huang, Y., et al. (2009). Community based study on CKD subjects and associated risk factors. Nephrol Dial Transplant, 24, 2117–23.

Chen, Y. C., Su, C. T., Wang, S. T., et al. (2009). A preliminary investigation of the association between serum uric acid and impaired renal function. Chang Gung Med J, 32, 66–71. Chertow, G. M., Johansen, K. L., Lew, M., et al. (2000). Vintage, nutritional status and survival in hemodialysis patients. Kidney Int, 57(3), 1176–81. Coca, S. G., Yusuf, B., Shlipak, M. G., et al. (2009). Long-term risk of mortality and other adverse outcomes after acute kidney injury: a systematic review and meta-analysis. Am J Kidney Dis, 53(6), 961–73. Collins, A. J., Foley, R. N., Gilbertson, D. T., et al. (2009). The state of chronic kidney disease, ESRD, and morbidity and mortality in the first year of dialysis. Clin J Am Soc Nephrol, 4 Suppl 1, S5–11. Cooper, B. A., Branley, P., Bulfone, L., et al. (2010). A randomized, controlled trial of early versus late initiation of dialysis. N Engl J Med, 363, 609–19. Cooper, W. A., O’Brien, S. M., Thourani, V. H., et al. (2006). Impact of renal dysfunction on outcomes of coronary artery bypass surgery: results from the Society of Thoracic Surgeons National Adult Cardiac Database. Circulation, 113, 1063–70. Dalrymple, L. S. and Go, A. S. (2008). Epidemiology of acute infections among patients with chronic kidney disease. Clin J Am Soc Nephrol, 3(5), 1487–93. Dalrymple, L. S., Johansen, K. L., Chertow, G. M., et al. (2010). Infection-related hospitalizations in older patients with ESRD. Am J Kidney Dis, 56(3), 522–30. Dalrymple, L. S., Katz, R., Kestenbaum, B., et al. (2012). The risk of infection-related hospitalization with decreased kidney function. Am J Kidney Dis, 59(3), 356–63. Desai, A. S., Toto, R., Jarolim, P., et al. (2011). Association between cardiac biomarkers and the development of ESRD in patients with type 2 diabetes mellitus, anemia, and CKD. Am J Kidney Dis, 58(5), 717–28. Di Angelantonio, E., Chowdhury, R., Sarwar, N., et al. (2010). Chronic kidney disease and risk of major cardiovascular disease and non-vascular mortality: prospective population based cohort study. BMJ, 341, c4986. Fang, M. C., Go, A. S., Chang, Y., et al. (2011). A new risk scheme to predict warfarin-associated hemorrhage—The ATRIA (Anticoagulation and Risk Factors in Atrial Fibrillation) Study. J Am Coll Cardiol, 58, 395–401. Fassett, R. G., Venuthurupalli, S. K., Gobe, G. C., et al. (2011). Biomarkers in chronic kidney disease: a review. Kidney Int, 80, 806–21. Fliser, D., Franek, E., Joest, M., et al. (1997). Renal function in the elderly: impact of hypertension and cardiac function. Kidney Int, 51, 1196–204. Foley, R. N. (2007). Infections in patients with chronic kidney disease. Infect Dis Clin North Am, 21(3), 659–72, viii. Foster, M. C., Rawlings, A. M., Marrett, E., et al. (2013). Cardiovascular risk factor burden, treatment, and control among adults with chronic kidney disease in the United States. Am Heart J, 166, 150–6. Fox, C., Matsushita, K., Woodward, M., et al. (2012). Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without diabetes: a meta-analysis. Lancet, 380, 1662–73. Gansevoort, R. T., Matsushita, K., van der Velde, M., et al. (2011). Lower estimated GFR and higher albuminuria are associated with adverse kidney outcomes in both general population and high risk cohorts. A collaborative meta-analysis of general and high-risk population cohorts. Kidney Int, 80(1), 93–104. Garg, A. X., Clark, W. F., Haynes, R. B., et al. (2002). Moderate renal insufficiency and the risk of cardiovascular mortality: results from the NHANES I. Kidney Int, 61, 1486–94. Go, A. S., Chertow, G. M., Fan, D., et al. (2004). Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med, 351, 1296–305. Goicoechea, M., Garcia de Vinuesa, S., and Verdalles, U. (2010). Effect of allopurinol in CKD progression and cardiovascular risk. Clin J Am Soc Nephrol, 5, 1388–93.

Chapter 97 

Guo, H., Liu, J., Collins, A. J., et al. (2008). Pneumonia in incident dialysis patients—the United States Renal Data System. Nephrol Dial Transplant, 23(2), 680–6. Gupta, R., Plantinga, L. C., Fink, N. E., et al. (2007). Statin use and sepsis events [corrected] in patients with chronic kidney disease. JAMA, 297(13), 1455–64. Halbesma, N., Jansen, D. F., Heymans, M. W., et al. (2011). Development and validation of a general population risk score. Clin J Am Soc Nephrol, 6, 1731–8. Halbesma, N., Kuiken, D. S., Brantsma, A. H., et al. (2006). Macroalbuminuria is a better risk marker than low estimated GFR to identify individuals at risk for accelerated GFR loss in population screening. JAMA, 17, 2582–90. Hallan, S. I., Coresh, J., Astor, B. C., et al. (2006). International comparison of the relationship of chronic kidney disease prevalence and ESRD risk. J Am Soc Nephrol, 17, 2275–84. Hallan, S. I., Matsushita, K., Sang, Y., et al. (2012). Age and association of kidney measures with mortality and end-stage renal disease. JAMA, 308(22), 2349–60. Hemmelgarn, B. R., Manns, B. J., Lloyd, A., et al. (2010). Relation between kidney function, proteinuria, and adverse outcomes. JAMA, 303, 423–9. Hemmelgarn, B. R., Zhang, J., Manns, B. J., et al. (2006). Progression of kidney dysfunction in the community-dwelling elderly. Kidney Int, 69, 2155–61. Hsu, C. Y., Iribarren, C., McCulloch, C. E., et al. (2009). Risk factors for end stage renal disease: 25 year follow up. Arch Intern Med, 169, 342–50. Hsu, C. Y., McCulloch, C. E., and Curhan, G. C. (2002). Epidemiology of anemia associated with chronic renal insufficiency among adults in the US; results from the Third NHANES. J Am Soc Nephrol, 13, 504–10. Hunsicker, L. G., Adler, S., Caggiula, A., et al. (1997). Predictors of the progression of renal disease in the Modification of Diet in Renal Disease Study. Kidney Int, 51(6), 1908–19. Imai, E., Horio, M., Yamagata, K., et al. (2008). Slower decline of glomerular filtration rate in the Japanese general population: a longitudinal 10-year follow-up study. Hypertens Res, 31, 433–41. Isakova, T., Xie, H., Yang, W., et al. (2011). Fibroblast growth factor 23 and risks of mortality and end-stage renal disease in patients with chronic kidney disease. JAMA, 305, 2432–9. Iseki, K., Ikemiya, Y., Iseki, C., et al. (2003). Proteinuria and the risk of developing end stage renal disease. Kidney Int, 63, 1468–74. Iseki, K., Iseki, C., Ikemiya, Y., et al. (1996). Risk of developing end-stage renal disease in a cohort of mass screening. Kidney Int, 49, 800–5. Iseki, K., Oshiro, S., Tozawa, M., et al. (2001). Significance of hyperuricemia on the early detection of renal failure in a cohort of screened subjects. Hypertens Res, 24, 691–7. Ishani, A., Collins, A. J., Herzog, C. A., et al. (2005). Septicemia, access and cardiovascular disease in dialysis patients: the USRDS Wave 2 study. Kidney Int, 68(1), 311–18. Jafar, T. H., Qadri, Z., and Hashmi, S. (2009). Prevalence of microalbuminuria and associated electrocardiographia abnormalities in an Indo-Asian population. Nephrol Dial Transplant, 24(7), 2111–16. James, M. T., Hemmelgarn, B. R., Wiebe, N., et al. (2010). Glomerular filtration rate, proteinuria, and the incidence and consequences of acute kidney injury: a cohort study. Lancet, 376(9758), 2096–103. James, M. T., Laupland, K. B., Tonelli, M., et al. (2008). Risk of bloodstream infection in patients with chronic kidney disease not treated with dialysis. Arch Intern Med, 168(21), 2333–9. James, M. T., Quan, H., Tonelli, M., et al. (2009). CKD and risk of hospitalization and death with pneumonia. Am J Kidney Dis, 54(1), 24–32. John, R., Webb, M., Young, A., et al. (2004). Unreferred chronic kidney disease: a longitudinal study. Am J Kidney Dis, 43, 825–35. Johansen, K. L., Zhang, R., Huang, Y., et al. (2009). Survival and hospitalization among patients using nocturnal and short daily compared to conventional hemodialysis: a USRDS study. Kidney Int, 76(9), 984–90. Johnson, D. W., Dent, H., Hawley, C. M., et al. (2009). Associations of dialysis modality and infectious mortality in incident dialysis patients in Australia and New Zealand. Am J Kidney Dis, 53(2), 290–7.

chronic kidney disease long-term outcomes

Keane, W. F., Zhang, Z., Lyle, P. A., et al. (2006). Risk scores for predicting outcomes in patients with type 2 diabetes and nephropathy: the RENAAL study. Clin J Am Soc Nephrol, 1, 761–7. Keith, D. S., Nichols, G. A., Gullion, C. M., et al. (2004). Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization. Arch Intern Med, 164, 659–63. Kovesdy, C. P., Anderson, J. E., Kalantar-Zadeh, K. (2009). Association of serum bicarbonate levels with mortality in patients with non-dialysis dependent CKD. Nephrol Dial Transplant, 24, 1232–7. Kovesdy, C. P., Regidor, D. L., Mehrotra, R. et al. (2007). Serum and dialysate potassium concentrations and survival in hemodialysis patients. Clin J Am Soc Nephrol, 2(5), 999–1007. Kurth, T., de Jong, P. E., Cook, N. R., et al. (2009). Kidney function and risk of cardiovascular disease and mortality in women: a prospective cohort study. BMJ, 338, b2392. Lafrance, J. P., Djurdjev, O., and Levin, A. (2010). Incidence and outcomes of acute kidney injury in a referred chronic kidney disease cohort. Nephrol Dial Transplant, 25(7), 2203–9. Lameire, N., Eloot, S., Van Biesen, W., et al. (2012). Ongoing controversies on the role of frequent hemodialysis and its benefits in patients with ESRD. NephSAP, 11, 1–6. Lee, M., Saver, J. L., Chang, K. H., et al. (2010). Low glomerular filtration rate and risk of stroke: meta-analysis. BMJ, 341, c4249. Levey, A. S., de Jong, P. E., Coresh, J., et al. (2011). The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report. Kidney Int, 80, 17–28. Levin, A., Djurdjev, O., Beaulieu, M., et al. (2008). Variability and risk factors for kidney disease progression and death following attainment of stage 4 CKD in a referred cohort. Am J Kidney Dis, 52(4), 661–71. Levin, A., Thompson, C. R., Ethier, J., et al. (1999). Left ventricular mass index increase in early renal disease: impact of decline in hemoglobin. Am J Kidney Dis, 34, 125–34. Li, L., Astor, B., Lewis, J., et al. (2012). Longitudinal progression trajectory of GFR among patients with CKD. Am J Kidney Dis, 59(4), 504–12. Li, P. K., Szeto, C. C., Piraino, B., et al. (2010). Peritoneal dialysis-related infections recommendations: 2010 update. Perit Dial Int, 30(4), 393–423. Lindeman, R. D. (1998). Is the decline in renal function with normal aging inevitable? Geriatr Nephrol Urol, 8, 7–9. Lindeman, R. D., Tobin, J. D., and Shock, N. W. (1984). Association between blood pressure and the rate of decline in renal function with age. Kidney Int, 26, 861–8. Mahmoodi, B. K., Matsushita, K., Woodward, M., et al. (2012). Associations of kidney disease measures with mortality and end-stage renal disease in individuals with and without hypertension: a meta-analysis. Lancet, 380, 1649–61. Martínez-Castelao, A., Górriz, J., Portolés, J. M., et al. (2011). Baseline characteristics of patients with chronic kidney disease stage 3 and stage 4 in Spain: the MERENA observational cohort study. BMC Nephrol, 12, 53. Matsushita, K., van der Velde, M., Astor, B. C., et al. (2010). Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet, 375, 2073–81. McDonald, S., Excell, L., and Livingston, B. (eds.) (2010). ANZDATA Registry Report 2010. Adelaide: Australia and New Zealand Dialysis and Transplant Registry. MDRD Study Group (1991). Assessing the progression of renal disease in clinical studies: effects of duration of follow-up and regression to the mean. J Am Society Nephrol, 1, 1087–94. Menon, V., Tighiouart, H., Smith Vaughn, N., et al. (2010). Serum bicarbonate and long term outcomes in CKD. Am J Kidney Dis, 56, 907–14. Mujais, S. (2006). Microbiology and outcomes of peritonitis in North America. Kidney Int Suppl, 103, S55–62. Muntner, P., Hamm, L., Kusek, J. W., et al. (2004). The prevalence of nontraditional risk factors for coronary heart disease in patients with chronic kidney disease. Ann Intern Med, 140, 9–17.

775

776

Section 5  

the patient with reduced renal function

Muntner, P., He, J., Astor, B. C., et al. (2005). Traditional and non-traditional risk factors predict coronary heart disease in chronic kidney disease: results from the atherosclerosis risk in communities study. J Am Soc Nephrol, 16, 529–38. Muntner, P., He, J., Hamm, L., et al. (2002). Renal insufficiency and subsequent death resulting from cardiovascualr disease in the United States. J Am Soc Nephrol, 13, 45–53. Nesrallah, G. E., Moist, L. M., Awaraji, C. et al. (2004). An international registry to compare quotidian dialysis regimen with conventional thrice weekly hemodialysis: why, how, and potential pitfalls. Sem Dial, 17, 131–135. Nesrallah, G. E., Suri, R. S., Zoller, R. et al. (2006). The international quotidian dialysis registry. Annual Reporst 2006. Hemodial Int, 10, 219–24. Nitsch, D., Grams, M., Sang, Y., et al. (2013). Associations of estimated glomerular filtration rate and albuminuria with mortality and renal failure by sex: a meta-analysis. Chronic Kidney Disease Prognosis Consortium. BMJ, 346, f324. Nitsch, D., Nonyane, B. A., Smeeth, L., et al. (2011). CKD and hospitalization in the elderly: a community-based cohort study in the United Kingdom. Am J Kidney Dis, 57(5), 664–72. O’Hare, A. M., Batten, A., Ríos Burrows, N., et al. (2012). Trajectories of kidney function decline in the 2 years before initiation of long-term dialysis. Am J Kidney Dis, 59(4), 513–22. Obermayr, R. P., Temml, C., Gutjahr, G., et al. (2008). Elevated uric acid increases the risk for kidney disease. J Am Soc Nephrol, 19, 2407–13. Pauly, R. P., Gill, J. S., Rose, C. L., et al. (2009). Survival among nocturnal home haemodialysis patients compared to kidney transplant recipients. Nephrol Dial Transplant, 24(9), 2915–19. Pfeffer, M., Burdmann, E. A., Chen, C. Y., et al. (2009). The TREAT Investigators. A Trial of Darbepoetin Alfa in Type 2 Diabetes and Chronic Kidney Disease. N Engl J Med, 361, 2019–32. Powe, N. R., Jaar, B., Furth, S. L., et al. (1999). Septicemia in dialysis patients: incidence, risk factors, and prognosis. Kidney Int, 55(3), 1081–90. Pun, P. H., Lehrich, R. W., Honeycutt, E. F., et al. (2011). Modifiable risk factors associated with sudden cardiac arrest within hemodialysis clinics. Kidney Int, 79(2), 218–27. Quinn, M. P., Cardwell, C. R., Rainey, A., et al. (2011). Patterns of hospitalisation before and following initiation of haemodialysis: a 5 year single centre study. Postgrad Med J, 87(1028), 389–93. Raggi, P., Boulay, A., Chasan-Taber, S., et al. (2002). Cardiac calcification in adult hemodialysis patients: a link between end-stage renal disease and cardiovascular disease? J Am Coll Cardiol, 39, 695–701. Ravid, M., Broshd, D., Ravid-Safran, D., et al. (1998). Main risk factors for nephropathy in type 2 diabetes mellitus are plasma cholesterol levels, mean blood pressure and hyperglycemia. Arch Intern Med, 158, 998–1004. Roberts, M. A., Hare, D. L., Ratnaike, S., et al. (2006). Cardiovascular biomarkers in CKD: pathophysiology and implications for clinical management of cardiac disease. Am J Kidney Dis, 48(3), 341–60. Ronco, C., Chionh, C. Y., Haapio, M., et al. (2009). The cardiorenal syndrome. Blood Purif, 27, 114–26. Sarnak, M. J. and Jaber, B. L. (2000). Mortality caused by sepsis in patients with end-stage renal disease compared with the general population. Kidney Int, 58(4), 1758–64. Sarnak, M. J. and Jaber, B. L. (2001). Pulmonary infectious mortality among patients with end-stage renal disease. Chest, 120(6), 1883–7. Sarnak, M. J., Levey, A. S., Schoolwerth, A. C., et al. (2003). Kidney disease as a risk factor for development of cardiovascular disease: a statement from the American Heart Association Councils on Kidney in Cardiovascular Disease, High Blood Pressure Research, Clinical Cardiology, and Epidemiology and Prevention. Circulation, 108, 2154–69. Shlipak, M. G., Fired, L. F., Crump, C., et al. (2003). Elevations of inflammatory and procoagulant biomarkers in elderly persons with renal insufficiency. Circulation, 107, 87–92. Shiplak, M. G., Fired, L. F., Cushman, M., et al. (2005). Cardiovascular mortality risk in chronic kidney disease—comparison of traditional and novel risk factors. JAMA, 293, 1737–45. Shlipak, M. G., Heidenreich, P. A., Noguchi, H., et al. (2002). Association of renal insufficiency with treatment and outcomes after myocardial infarction in elderly patients. Ann Intern Med, 137, 555–62.

Shlipak, M. G., Katz, R., Kestenbaum, B., et al. (2009). Rapid decline of kidney function increases cardiovascular risk in the elderly. J Am Soc Nephrol, 20(12), 2625–30. Siu, Y. P., Leung, K. T., Tong, M. K., et al. (2006). Use of allopurinol in slowing the progression of renal disease through its ability to lower serum uric acid level. Am J Kidney Dis, 47, 51–9. Slinin, Y., Foley, R. N., Collins, A. J. (2006). Clinical epidemiology of pneumonia in hemodialysis patients: the USRDS waves 1, 3, and 4 study. Kidney Int, 70(6), 1135–41. Smith, G. L., Lichtman, J. H., Bracken, M. B., et al. (2006). Renal impairment and outcomes in heart failure: systematic review and meta-analysis. J Am Coll Cardiol, 47, 1987–96. Sood, M. M., Komenda, P., Sood, A. R., et al. (2010). Adverse outcomes among Aboriginal patients receiving peritoneal dialysis. CMAJ, 182(13), 1433–9. Tangri, N., Stevens, L. A., Griffith, J., et al. (2011). A predictive model for progression of chronic kidney disease to kidney failure. JAMA, 305(15), 1553–7. Tonelli, M., Wiebe, N., Culleton, B., et al. (2006). Chronic kidney disease and mortality risk: a systematic review. J Am Soc Nephrol, 17, 2034–47. Uchino, S., Kellum, J. A., Bellomo, R., et al. (2005). Acute renal failure in critically ill patients: a multinational, multicentre study. JAMA, 294(7), 813–18. United States Renal Data System (2004). USRDS 2004 Annual Data Report: Atlas of End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. United States Renal Data System (2009). USRDS 2009 Annual Data Report: Volume One: Atlas of Chronic Kidney Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. United States Renal Data System (2010). Excerpts from USRDS 2009 Annual Data Report. US. Department of Health and Human Services. Am J Kidney Dis, 55(Suppl 1), S1. United States Renal Data System (2011). USRDS 2011 Annual Data Report: Atlas of End-Stage Renal Disease in the United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Van der Velde, N., Matsushita, K., Coresh, J., et al. (2011). Lower estimated glomerular filtration rate and higher albuminuria are associated with all-cause and cardiovascular mortality. A collaborative meta-analysis of high risk population cohorts. Kidney Int, 79(12), 1341–52. Wang, A. Y., Lam, C. W., Chan, I. H., et al. (2010). Sudden cardiac death in end-stage renal disease patients: a 5-year prospective analysis. Hypertension, 56, 210–16. Wang, H. E., Gamboa, C., Warnock, D. G., et al. (2011). Chronic kidney disease and risk of death from infection. Am J Nephrol, 34(4), 330–6. Watanabe, H., Watanabe, T., Sasaki, S., et al. (2009). Bidirectional relationship between chronic kidney disease and atrial fibrillation: the Niigata preventive medicine study. Am Heart J, 158, 629–36. Wen, C. P., Cheng, T. Y., Tsai, M. K., et al. (2008). All-cause mortality attributable to chronic kidney disease: a prospective cohort study based on 462 293 adults in Taiwan. Lancet, 371(9631), 2173–82. Wen, C. P., David Cheng, T. Y., Chan, H. T., et al. (2010). Is high serum uric acid a risk marker or a target for treatment? Examination of its independent effect in a large cohort with low cardiovascular risk. Am J Kidney Dis, 56, 273–88. White, S. L., Polkinghorne, K. R., Atkins, R. C., et al. (2010). Comparison of the Prevalence and Mortality Risk of CKD in Australia Using the CKD Epidemiology Collaboration (CKD-EPI) and Modification of Diet in Renal Disease (MDRD) Study GFR Estimating Equations: The AusDiab (Australian Diabetes, Obesity and Lifestyle) Study. Am J Kidney Dis, 55(4), 660–70. Young, B. A. (2011). Prevention of sudden cardiac arrest in dialysis patients: can we do more to improve outcomes? Kidney Int, 79, 147–9. Young, J. M., Terrin, N., Wang, X., et al. (2009). Asymmetric dimethylarginine and mortality in stages 3 to 4 chronic kidney disease. Clin J Am Soc Nephrol, 4(6), 1115–22.

CHAPTER 98

Cardiovascular disease and chronic kidney disease: overview David J. Goldsmith

Pathophysiology Increased cardiovascular risk in individuals with CKD is due partly to the high prevalence of traditional risk factors, such as raised BP and diabetes. Smoking too has a most malign influence on both the heart and the kidney in CKD (see Chapter 103).

Raised BP is both a familiar and strong risk factor for development of CKD. Even in its earliest manifestations, CKD tends to raise BP, which is likely to increase cardiovascular risk in affected patients (see Chapter 100). Data from cohort studies suggest that antihypertensive therapy would more effectively reduce cardiovascular risk in patients with CKD than in those without this disorder. Few prospective studies, however, support this assumption. A  target BP of < 140/90 mmHg is considered appropriate to prevent CVD in patients with CKD, but a BP target of < 130/80 mmHg is now widely recommended for albuminuric patients (see Chapter 99).

(A) HR for CVD mortality (ACR studies)

Many studies in various populations have reported that reduced estimated glomerular filtration rate (eGFR) and raised albuminuria are associated with cardiovascular disease (CVD). Data from 31 published and unpublished cohort studies involving > 1.4 million individuals were assessed in meta-analyses. After adjustment for traditional cardiovascular risk factors and albuminuria, the risk gradient for cardiovascular mortality changed little when eGFR was > 75 mL/min/1.73 m2 but increased linearly once < 75 mL/min/1.73 m2. Cardiovascular mortality was about twice as high in patients with stage 3 chronic kidney disease (CKD) and three times higher at stage 4 than that in individuals with normal kidney function. In contrast to the non-linear risk relationship for eGFR, the association of albuminuria with cardiovascular risk has no threshold effect, even after adjustment for traditional cardiovascular risk factors and eGFR. The adjusted risk of cardiovascular mortality is more than doubled at the upper end of the microalbuminuria category (3–30 mg/mmol, 30–300 mg/g), compared with the risk in individuals with normal albuminuria. This lack of threshold effect indicates that albuminuria even at the upper end of the normal range (threshold 3 mg/mmol, 30 mg/g) confers cardiovascular risk. This is independent of, but additive to, diabetes and raised blood pressure (BP) (Fig. 98.1) (see Chapter 97). The proportion of deaths from CVD increases as eGFR declines. In a Canadian cohort, when adjusted for age and sex, CVD accounted for 27.5% of deaths in individuals with normal kidney function versus 58.0% in those with kidney failure. A similar increase in the proportion of deaths due to CVD is observed among people with raised albuminuria. When adjusted for gender, life expectancy due to CVD is shortened by 1.3, 7.0, 12.5, and 16.7 years, respectively, in patients aged 30 years with eGFR stages 3A, 3B, 4, and 5 (45–59, 30–44, 15–29, and < 15 mL/min/1.73 m2, respectively), compared with that in individuals with normal kidney function. Kidney transplantation does not completely correct this deficit in longevity.

Raised blood pressure

4

Significant values Non-significant values

2

1

0.5 15

30

45

60

75

90

105

120

eGFR (mL/min per 1.73 m2)

(B) HR for CVD mortality (ACR studies)

Introduction

4

2

1

0.5 2.5

5

10

30

300

1000

ACR (mg/g)

Fig. 98.1  The proportion of deaths from cardiovascular disease increases as eGFR declines. From Gansevoort et al. (2013).

778

Section 5  

the patient with reduced renal function

Left ventricular hypertrophy, sudden cardiac death, and cardiomyopathy In patients with early or advancing CKD, the prevalence of left ventricular hypertrophy (LVH) (see Chapter  107) is strikingly increased. When eGFR is < 30 mL/min/1.73 m2, around 50% of patients develop LVH, of which most is concentric hypertrophy. Apart from hypertension, renal anaemia (see Chapter  123) and increased vascular stiffness (see Chapter 111) seem to have pivotal roles in development of LVH and thus reduced coronary reserve. The latter could be aggravated by reduced cardiac capillary density in CKD and impaired coronary dilatory responses. Expression of endothelial nitric-oxide synthase is downregulated, which suggests a possible mechanism for coronary endothelial dysfunction in early stages of CKD. Histologically, LVH in CKD is characterized by myocardial fibrosis that is thought to impair contractility when severe. The high prevalence of LVH, with its associated risk of cardiac-rhythm disturbances, could at least partly explain why the prevalence of sudden cardiac death (see Chapter 108) is increased in people with CKD. In the general population, sudden cardiac death accounts for roughly one death per 1000 person-years and for around 5–10% of all deaths, whereas among individuals with kidney failure, the rates are around 50–60 deaths per 1000 person-years and 25% of total mortality (nearly half of all CV mortality). Besides the high prevalence of LVH, abnormal electrolyte concentrations and increased prevalence of coronary artery disease are predisposing factors for sudden cardiac death in patients with CKD. Atherosclerosis and valvular heart disease are frequently seen in patients with kidney failure, but also occur in those with early CKD. The key modulators in this field have not been elucidated in intervention trials, but seem likely to include calcification inhibitors (e.g. fetuin-A and matrix Gla protein), promoters (e.g. hyperphosphataemia), and calcium-phosphate product, parathyroid hormone, and leptin. (See Chapters 115 and 120). Distinguishing heart failure from volume overload in patients with advanced CKD can be difficult, but there seems little doubt that volume control is important in preventing heart failure as well as in treating it (see Chapters 107 and 268). Cardiorenal syndrome (CRS) is a group of diagnoses defined as disorders of the heart and kidney, whereby dysfunction in one organ may induce dysfunction in the other. The umbrella term and Table 98.1  Discrete subtypes of cardiorenal syndrome as defined by the ADQI Name

Description

Acute cardiorenal syndrome

Type 1 Acute worsening of cardiac function leading to renal dysfunction

Chronic cardiorenal syndrome

Type 2 Chronic abnormalities in cardiac function leading to renal dysfunction

Acute renocardiac syndrome

Type 3 Acute worsening of renal function causing cardiac dysfunction

Chronic renocardiac syndrome

Type 4 Chronic abnormalities in renal function leading to cardiac disease

Secondary cardiorenal Type 5 Systemic conditions causing simultaneous syndromes dysfunction of the heart and kidney ADQI: Acute Dialysis Quality Initiative.

discrete subtypes (Table 98.1) were developed by the Acute Dialysis Quality Initiative (ADQI) to emphasize the bidirectional pathways and to provide context for identification of the complex pathophysiological interactions occurring in different types of combined heart and kidney disorders (House et al., 2010). The value of a concept of a specific (or several) cardiorenal syndromes is not clear.

Lipids, oxidative stress, inflammation, endothelial function, and advanced glycation end products Dyslipidaemia and low-grade inflammation are also caused by CKD (see Chapters 102 and 110). In patients with impaired kidney function and high albuminuria, lipid profiles become atherogenic, owing partly to defective high-density lipoprotein cholesterol function and excessive oxidation of low-density lipoprotein cholesterol. Mechanisms of increased systemic inflammation in CKD are unclear, but increased production of inflammatory mediators has been attributed to raised oxidative stress (see Chapter 112) and accumulation of postsynthetically modified proteins and toxins that are cleared with normal renal function. Advanced glycation end products (AGEs) and receptors for AGEs (RAGEs) are much studied in CKD and may be both markers and drivers of CVD.

Renin–angiotensin system, renalase, and asymmetric dimethylarginine Other factors that raise cardiovascular risk in patients with CKD include increased activity of the renin–angiotensin system and sympathetic nerve activity in CKD. Angiotensin stimulates production of superoxide, interleukin 6, and other cytokines (see Chapter 110). Bioavailability of nitric oxide, which is involved with vascular smooth-muscle contraction and growth, platelet aggregation, and leucocyte adhesion to the endothelium, becomes decreased (see Chapter  113). Activity of renalase is lowered in individuals with CKD. This enzyme is produced by the kidney and inactivates catecholamines. All these vasoactive factors impinge on endothelial function (see Chapter 113). Another key factor for endothelial function seems to be asymmetric dimethylarginine (ADMA). Plasma ADMA concentrations increase with decreasing kidney function and predict mortality and cardiovascular complications in CKD patients (see Chapter  113). ADMA inhibits generation of nitric oxide, reduces cardiac output, and raises both systemic vascular resistance and BP. Increased concentrations of ADMA and sympathetic overactivity are strongly associated with concentric LVH.

Mineral and bone metabolites Patients with impaired kidney function frequently develop deficiency of active vitamin D because of a lack of its precursor, impaired activity of the kidney enzyme 1  α-hydroxylase, which converts this precursor to the active hormone, or both. Observational studies in patients with CKD have shown associations between vitamin D deficiency and increased risk of cardiovascular events, and experimental data suggest that the vitamin D pathway is involved in modifying cardiac structure and function. Elevated serum phosphate, fibroblast growth factor 23, and calcium concentrations have both been associated with the development of future CVD in CKD, and also in the general, populations—as yet no interventional study has tested whether these associations are causal. (See Chapter 115.)

Chapter 98 

Conclusions Too few randomized controlled trials have been conducted into prevention of CVD in this most susceptible CKD population; indeed, much effort has been expended to exclude CKD patients from the majority of CVD trials. Until this ends, while we can make long lists of associations and risk factors, we can make comparatively much shorter lists of useful, practical, proven interventions.

Further reading Blood Pressure Lowering Treatment Trialists’ Collaboration, Ninomiya, T., Perkovic, V., et al. (2013). Blood pressure lowering and major cardiovascular events in people with and without chronic kidney disease: meta-analysis of randomised controlled trials. BMJ, 347, f5680. Block, G. A., Ix, J. H., Ketteler, M., et al. (2013). Phosphate homeostasis in CKD: report of a scientific symposium sponsored by the National Kidney Foundation. Am J Kidney Dis, 62(3), 457–73. Clementi, A., Virzì, G. M., Goh, C. Y., et al. (2013). Cardiorenal syndrome type 4: a review. Cardiorenal Med, 3(1), 63–70. Gustafsson, D. and Unwin, R. (2013). The pathophysiology of hyperuricaemia and its possible relationship to cardiovascular disease, morbidity and mortality. BMC Nephrol, 14, 164. Huang, Y., Cai, X., Zhang, J., et al. (2014). Prehypertension and Incidence of ESRD: a Systematic Review and Meta-analysis. Am J Kidney Dis, 63(1), 76–83.

cardiovascular disease in ckd: overview

Jain, G. and Jaimes, E. A. (2013). Nicotine signaling and progression of chronic kidney disease in smokers. Biochem Pharmacol, 86(8), 1215–23. Khan, U. A., Garg, A. X., Parikh, C. R., et al. (2013). Prevention of chronic kidney disease and subsequent effect on mortality: a systematic review and meta-analysis. PLoS One, 8(8), e71784. Lambers Heerspink, H. J. and de Zeeuw, D. (2013). Novel drugs and intervention strategies for the treatment of chronic kidney disease. Br J Clin Pharmacol, 76(4), 536–50. Lv, J., Ehteshami, P., Sarnak, M. J., et al. (2013). Effects of intensive blood pressure lowering on the progression of chronic kidney disease: a systematic review and meta-analysis. CMAJ, 185(11), 949–57. Moradi, H., Sica, D. A., and Kalantar-Zadeh, K. (2013). Cardiovascular burden associated with uremic toxins in patients with chronic kidney disease. Am J Nephrol, 38(2), 136–48. Waheed, A. A., Pedraza, F., Lenz, O., et al. (2013). Phosphate control in end-stage renal disease: barriers and opportunities. Nephrol Dial Transplant, 28(12), 2961–8.

References Gansevoort, R. T., Correa-Rotter, R., Hemmelgarn, B. R., et al. (2013). Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Lancet, 382(9889), 339–52. House, A. A., Anand, I., Bellomo, R., et al. (2010). Definition and classification of Cardio-Renal Syndromes: workgroup statements from the 7th ADQI Consensus Conference. Nephrol Dial Transplant, 25(5), 1416–20.

779

CHAPTER 99

Recommendations for management of high renal risk chronic kidney disease Charles J. Ferro and Khai Ping Ng Introduction Worsening chronic kidney disease (CKD) is associated with increasing morbidity and mortality, mainly as a consequence of cardiovascular disease (Fig. 99.1) (Matsushita et al., 2010). Population studies have consistently shown that people with CKD identified by screening have a far greater likelihood of cardiovascular death than progression to end-stage renal disease (ESRD) (Keith et al., 2008). Preventing further deterioration of renal function and reducing the cardiovascular risk in this population is essential to improve the clinical outcomes of CKD patients. The concept that CKD is a single entity with generic treatment is clearly a huge oversimplification and requires that CKD be viewed as a single process despite a range of primary diseases. Nevertheless it is a useful concept and is supported by the effectiveness of some generic therapies and on data suggesting final common physiological pathways underlie the progression of CKD irrespective of the initiating causation (Turner et al., 2012) (see Chapter 136). Indeed, extensive studies have shown that the rate of loss of renal function may be largely due to common secondary factors often unrelated to the original renal disease. These factors are often also cardiovascular risk factors and so preventing the progression of CKD and reducing cardiovascular risk often are not mutually exclusive but indeed complementary. Even though the risk of progressive kidney disease and cardiovascular disease increases with even minimal renal impairment and microalbuminuria (Fig. 99.1), there is a much higher risk when the estimated glomerular filtration rate (eGFR) falls significantly below 60 mL/min/1.73 m2 especially in the presence of proteinuria (Levey et al., 2011; Turner et al., 2012). This chapter will focus on the management of these high-risk patients. Given that many local, national, and international societies produce their own, often frequently updated, guidelines on the management of CKD a selection of useful websites is given in Box 99.1. The increasing use of treatments to attenuate CKD progression has coincided with a plateau of ESRD incidence in the United States and United Kingdom in the last few years (Byrne et  al., 2010). However, incidence is still high and the field has not seen a truly new preventative therapy in over a decade. Nevertheless there is a growing literature that multidisciplinary programmes including patient education and using evidence-based guidelines can slow

down the progression of CKD, decrease hospitalizations, improve arteriovenous fistula creation before haemodialysis, and decrease mortality both before and after the initiation of dialysis (Lee et al., 2007; Snyder et al., 2009). These programmes have been found to be cost-effective.

Interventions influencing prognosis Hypertension Whereas the evidence that blood pressure lowering confers renal and cardiovascular protection is clear, the optimal level of blood pressure control is less well established. The MDRD Study showed that the level of proteinuria significantly modulates the effect of blood pressure lowering such that a lower blood pressure target (≤ 125/75 mmHg vs ≤ 140/90 mmHg) was associated with a slower rate of GFR decline among patients with > 1 g/day of proteinuria. Secondary analysis revealed significant correlations between rate of GFR decline and achieved blood pressure and long-term follow-up showed lower mortality and incidence of ESRD randomized to the low blood pressure target (Peterson et al., 1995). The African American Study of Kidney Disease and Hypertension (AASK) study did not show any additional benefit in targeting a blood pressure lower than 140/90 mmHg although there was a trend for this in patients with proteinuria consistent with the MDRD study (Wright et al., 2002). Several studies have suggested worse patient and renal outcomes with low achieved blood pressure (systolic 110–120 mmHg) (Cushman et al., 2010) There is therefore no evidence to support lowering blood pressure to below a systolic pressure of 120 mmHg and caution should be exercised particularly in patients who may suffer harm from excessive lowering of blood pressure such as patients with autonomic neuropathy or postural hypotension. The inherent difficulty associated with setting specific numeric targets for blood pressure control as well as the limited nature of the supporting evidence has recently been discussed in detail (Lewis, 2010). On balance, systolic blood pressure should be controlled to < 140 mmHg and the diastolic blood pressure to < 90 mmHg. For patients with significant proteinuria (urine albumin:creatinine ratio > 70 mg/mmol; urine protein:creatinine ratio > 100 mg/mmol; 24-hour urinary protein excretion > 1 g) or diabetes and CKD, a lower blood pressure (< 130/80 mmHg) is desirable. Lowering

Chapter 99 

(A)

8

management of high renal risk ckd

All-cause mortality; eGFR

(B)

All-cause mortality; ACR

Cardiovascular mortality; eGFR

(D)

Cardiovascular mortality; ACR

HR (95% Cl)

4 2 1 0.5 (C) 8

HR (95% Cl)

4 2 1 0.5

15

30

45

60

75

90

105

120

eGFR (mL/min per 1.73 m2)

2.5 5 10 (0.3) (0.6) (1.1)

30 (3.4)

300 (33.9)

1000 (113.0)

ACR (mg/g [mg/mmoL])

Fig. 99.1  Hazard ratios and 95% confidence intervals (CIs) for all-cause and cardiovascular mortality according to spline estimated glomerular filtration rate (eGFR) and albumin-to-creatinine ratio (ACR). Hazard ratios and 95% CIs (shaded areas) according to eGFR (A, C) and ACR (B, D) adjusted for each other, age, sex, ethnic origin, history of cardiovascular disease, systolic blood pressure, diabetes, smoking, and total cholesterol. The reference (diamond) was eGFR 95 mL/min/1.73 m2 and ACR 5 mg/g (0.6 mg/mmol), respectively. Circles represent statistically significant and triangles represent not significant. ACR plotted in mg/g. To convert ACR in mg/g to mg/ mmol multiply by 0.113. Approximate conversions to mg/mmol are shown in parentheses. From Gansevoort et al. (2013).

systolic blood pressure to < 120 mmHg should be avoided according to some guidance. However, this probably should not apply to teenagers or young adults and on occasion it may be useful to relate blood pressures to age norms. In the major paediatric study on control of blood pressure in CKD, a target of below the median for age was used rather than an absolute value (Escape Trial Group, 2009).

Use of angiotensin converting enzyme inhibitors and angiotensin receptor blockers Several large prospective randomized controlled trials among different groups of patients with CKD provide evidence that angiotensin converting enzyme inhibitors (ACEIs) or angiotensin receptor blockers (ARBs) afford significant renal protection in addition to that attributable to blood pressure lowering. Specifically, ACEI treatment reduces CKD progression in patients with diabetic nephropathy (Lewis et al., 1993) and in non-diabetic patients with proteinuria > 1 g/day (Wright et al., 2002). Treatment with ACEIs, compared with other antihypertensive drugs, significantly lowers the incidence of ESRD after adjustment for differences in level of blood pressure control. However, the benefit appears to be confined to patients with proteinuria > 0.5g/day (Casas et al., 2005). ARB treatment has also been shown to afford significant renal protection in patients with established diabetic nephropathy (Brenner et al., 2001; Lewis et al., 2001). Microalbuminuria is strongly positively associated with an increased risk of cardiovascular morbidity and mortality (Fig. 99.1). ACEIs and ARBs appear to be superior to more conventional antihypertensive therapies at reducing urinary albumin excretion

in diabetic, non-diabetic, and hypertensive populations (Brenner et al., 2001; Lewis et al., 2001). There is also a large body of evidence indicating that ACEI or ARB treatment reduces or delays progression from microalbuminuria to overt nephropathy and reduces cardiovascular risk in diabetic patients with microalbuminuria (Jafar et al., 2003). A number of small studies have suggested that using higher doses of an ACEI or ARB or indeed using a combination of these agents might result in superior outcomes, especially in reducing progression of kidney disease, compared to using conventional doses of single agents. This approach has, however, recently been called into question by both the withdrawal of the COOPERATE study (Kunz et al., 2008) and the publication of the ONgoing Telmisartan Alone and in combination with Ramipril Global Endpoint Trial (ONTARGET) (Mann et  al., 2008). This large-scale trial found that using a fixed dose combination of an ACEI and ARB resulted in worse renal outcomes (Mann et al., 2008). The injudicious use of ACEIs and ARBs in patients with CKD has further been called into question by investigators suggesting that they may increase the rate of progression in some patient groups, especially those with advanced CKD (Goncalves et  al., 2011). Current evidence does not support the use of ACEI or ARB treatment for all patients with CKD. Large prospective randomized controlled studies have reported significant reductions in cardiovascular morbidity and mortality associated with ACEI treatment (Yusuf et al., 2000; Fox, 2003). Interestingly, secondary analysis of the Prevention of Events with ACEI (PEACE) trial found a significant reduction in all-cause

781

782

Section 5  

the patient with reduced renal function

Box 99.1  Online sources of guidelines

Nephrology guidelines Kidney Disease: Improving Global Outcome (KDIGO)

European Renal Best Practice (ERBP)

Renal Association Guidelines (United Kingdom)

National Institute for Health and Care Excellence (NICE) guidelines on CKD

National Kidney Foundation Kidney Disease Outcome Quality Initiative (NKF KDOQI) (United States)

Caring for Australasians with Renal Impairment (CARI) guidelines (Australia)

Cardiovascular guidelines World Health Organization cardiovascular guidelines

National Institute for Health and Care Excellence (NICE) guidelines on cardiovascular disease

American College of Cardiology/American Heart Association (ACC/AHA) joint guidelines

National Heart Lung and Blood Institute guidelines (United States)

Hypertension guidelines World Health Organization/International Society of Hypertension guidelines

British Hypertension Society guidelines

America Society of Hypertension guidelines

National Heart Lung and Blood Institute guidelines (United States)

mortality associated with ACEI treatment in the CKD subgroup but not in patients with normal renal function (Solomon et  al., 2006). It would, therefore, seem reasonable to use ACEI or ARB treatment for reduction of cardiovascular risk as well as slowing of CKD progression. Direct renin inhibitors and mineralocorticoid receptor inhibitors have shown some promise when combined with ACEIs or ARBs in small studies looking at surrogate endpoints (Parving et al., 2008; Edwards et  al., 2009). However, larger hard outcome studies are needed before these agents are recommended for routine use in CKD patients. Indeed, the recent ALTITUDE trial examined the use of a direct renin inhibitor (DRI) combined with ARB therapy in patients with diabetic nephropathy was terminated early due to safety concerns in view of the elevated incidence adverse events. In summary, ACEI or ARB treatment should form part of the antihypertensive therapy of patients with CKD and urinary protein excretion of > 0.5 g/day unless there is a specific contraindication. Patients with CKD and proteinuria > 0.5 g/day should have their ACEI or ARB and other antihypertensive treatment escalated to achieve the lowest possible level of proteinuria. Similarly, patients

with diabetes mellitus and microalbuminuria should be treated with an ACEI or ARB, unless there is a specific contraindication. The use of ‘dual-blockade’ with various combinations of ACEI/ ARB/DRI has so far been proven disappointing.

Hyperlipidaemia Despite prior evidence to suggest that lipid-lowering with statins had a small but significant effect on CKD progression by reducing annual decline by 1.22 mL/min/year compared with placebo (Sandhu et al., 2006) as well as reducing proteinuria (Douglas et al., 2006), the largest study of lipid-lowering in CKD patients (SHARP; Study of Heart and Renal Protection) failed to demonstrate benefit in terms of slowing CKD progression (Baigent et  al., 2011). The study did however show cardiovascular benefits. Lipids in CKD are discussed in Chapter 102. Statins are a well-established treatment to reduce cardiovascular risk in a wide range of populations for both primary and secondary prevention. The benefits are effective regardless of baseline cholesterol levels, with an approximate 20% reduction in the 5-year risk of cardiovascular events per 1 mmol/L reduction in serum cholesterol.

Chapter 99 

Post hoc analysis and meta-analysis of studies, which included some patients with CKD, demonstrated that statins remained effective at reducing cardiovascular risk and mortality (Strippoli et al., 2008). The CKD patients included in these studies had relatively mild CKD and most were included in these trials because of other cardiovascular risk factors and not because of CKD per se. In general, CKD patients have a very similar reduction (~ 20%) in cardiovascular mortality to the general population with a 1 mmol/L reduction in low-density lipoprotein cholesterol (Strippoli et al., 2008). These findings have been confirmed by the SHARP study (Baigent et al., 2011). To justify primary prevention with lipid-lowering therapy, an estimate of cardiovascular risk is required. Available tools to assess cardiovascular risk do not adequately incorporate the impact of reduced GFR or increased proteinuria. Until such revised tools are available, and given the lack of specific evidence for intervention in the CKD population, the standard tools should probably continue to be used despite the expected higher risk. See also Chapter 102.

Dietary protein Dietary protein restriction was one of the earliest treatments used in CKD and it is reviewed in Chapter 101. In summary, while protein restriction has been shown to ameliorate symptoms associated with advanced CKD its effects on progression of earlier CKD are less clear-cut. It lowers phosphate and uric acid intake, as well as improving metabolic acidosis (Turner et  al., 2012). Protein restriction has been shown in many animal studies to reduce CKD progression possibly by lowering glomerular hyperfiltration and fibrosis (Turner et  al., 2012). However, clinical trials have been much less clear as to its efficacy, perhaps because blood pressure and other factors were better controlled. The largest and best study so far, the MDRD Study showed no effect (Klahr et  al., 1994). A meta-analysis did find a reduced requirement for dialysis associated with restricted protein diets in non-diabetic CKD (Fouque et al., 2009) but this is difficult to interpret when the best study produced a different result. In the meta-analysis, because of study heterogeneity an optimum level of protein intake could not be suggested. While most agree that high-protein diets are likely to be hazardous (Chapter  101) substantial protein restriction requires a significant amount of effort from physicians, dieticians, and patients if it is also to be safe. Careful monitoring of nutritional status is required to avoid malnutrition. Provided the burden is acceptable to the patient and malnutrition is avoided, a target of 0.8 g of protein per kg body weight per day seems reasonable (Turner et al., 2012). See also Chapter 101.

Dietary salt Western populations consume substantially higher amounts of sodium than is necessary, and this is associated with a higher prevalence of hypertension and cardiovascular disease (Bibbins-Domingo et al., 2010). There is relatively little evidence on whether a high sodium intake is specifically associated with poorer renal outcomes, or whether reducing sodium intake improves renal outcomes. However, given the impact of sodium reduction on blood pressure (Vollmer et al., 2001), and the known impact of high blood pressure on renal function and proteinuria, it seems sensible to adopt the general population recommendation of maintaining dietary sodium at < 2.4 g/day (100 mmol/day or < 6 g/day of salt). See also Chapter 101.

management of high renal risk ckd

Acidosis The concentration of hydrogen ions is normally managed by several buffering and elimination systems, including the kidney. Lower blood pH and reduced plasma bicarbonate levels thus accompany deterioration of renal function, especially when the GFR falls below 20mL/min/1.73 m2 (Hsu et al., 2002). Some experimental evidence as well as small studies in humans have suggested that lightening the acid load could stabilize or temporarily improve renal function. A randomized controlled trial in 134 patients with CKD demonstrated improvements in the rate of decline of renal function as well as a decrease in the need for dialysis in patients treated with sodium bicarbonate (de Brito-Ashurst et al., 2009). While further larger and confirmatory studies are awaited it would seem reasonable to treat patients with evidence of acidosis on blood tests with oral sodium bicarbonate given its simplicity, safety, and low cost. This is described further in Chapter 148.

Smoking Smoking has been identified as a risk factor for the development and progression of CKD (Yoshida et  al., 2008)  (Chapter  103). A small study has also intriguingly suggested that stopping smoking may slow down the rate of renal function deterioration (Schiffl et  al., 2002). Regardless, the clear evidence of smoking as a risk factor for cardiovascular and respiratory disease makes smoking cessation a critical intervention for improving survival in CKD patients. See also Chapter 103.

Interventions with uncertain or no effect on prognosis Chronic kidney disease–mineral and bone disorder Although there are strong associations between markers of mineral bone disease in CKD (CKD-MBD), including associations of mortality with serum phosphate levels in the general population as well as in patients with advanced CKD (Fig. 99.2) (Chue et al., 2010, 2011), it has been hard to prove that any interventions benefit progression of CKD or mortality, even in patients on dialysis. This is discussed further in Chapters 109, 117, and 118.

Anaemia Renal anaemia is considered in detail in Chapter 123. Anaemia was hypothesized to play a role in the early development of cardiovascular disease but the publication of several trials including CREATE (Drueke et al., 2006), CHOIR (Singh et al., 2006), and in particular the placebo-controlled TREAT (Pfeffer et al., 2009) have challenged the ‘cardiovascular’ rationale behind anaemia correction to higher targets. It is now recognized that complete anaemia correction is not advisable but partial correction may still be important, with different targets applying to different subgroups (Locatelli et al., 2009).

Uric acid Hyperuricaemia increases as GFR declines. Uric acid is clearly toxic to the kidney in very high concentrations such as in tumour lysis. However, whether more modest elevations are detrimental is much less clear (Filiopoulos et al., 2012), although there is evidence from small trials that lowering uric acid concentrations with allopurinol is associated with improvements in surrogate markers of cardiovascular disease (Filiopoulos et al., 2012). The little evidence available

783

Section 5  

the patient with reduced renal function 25 People suffering events (%)

784

Placebo Simvastatin plus ezetimibe

20

Rate reduction 17% (95% Cl 6–26%) Log-rank p = 0.0021

15 10 5 0 0

Number at risk Placebo 4620 Simvastatin 4650 plus ezetimibe

1

4204 4271

2

3

Years of follow-up 3849 3469 3939 3546

4

5

2566 2655

1269 1265

Fig. 99.2  Life-table plot of effects of allocation to simvastatin plus ezetimibe versus placebo on major atherosclerotic events. Numbers remaining at risk of a first major atherosclerotic event at the beginning of each year are shown for both treatment groups. From Baigent et al. (2011).

together with the occasional risk of severe allergic reactions to allopurinol make recommending routine lowering of uric acid difficult except in the context of recurrent gout.

Obesity Obesity is associated with a number of conditions known to increase the risk of CKD including hypertension, diabetes mellitus, and heart failure (Guh et al., 2009), and its impact on CKD is considered in detail in Chapter 106. However, the evidence that obesity is an independent risk factor for CKD or associated with increased rate of progression is limited. Intriguingly, small studies in patients after bariatric surgery show improvements in blood pressure control, proteinuria, and inflammatory markers as well as in GFR although this last parameter needs to be interpreted with caution and confirmed in larger studies with harder endpoints (Navaneethan et al., 2009). Nevertheless, given the negative impact of obesity on general and cardiovascular health it makes sense to recommend weight loss using lifestyle modification, in patients with CKD.

Exercise Patients with CKD have reduced exercise capacity and muscle strength which can improve with the introduction of an exercise programme (Kosmadakis et al., 2012). Regular, moderate exercise reduces cardiovascular risk in the general population (Paffenbarger, 2000). Although there is no solid evidence that regular exercise improves cardiovascular or renal outcomes in patients with CKD, the positive impact exercise has on general health and as part of a weight-loss programme makes it seem reasonable to recommend regular moderate exercise to CKD patients.

Glycaemic control in diabetes mellitus Observational studies suggest that HbA1C influences cardiovascular event rates and survival (Coresh et al., 2007). There is also strong evidence that improved glycaemic control prevents the development of microalbuminuria, progression of renal dysfunction, and microvascular complications in patients with both type 1 and 2 diabetes mellitus as well as reducing cardiovascular events and improving overall survival (de Boer et al., 2011). The precise level of glycaemic control to be achieved remains somewhat controversial. Recent randomized trials have indicated that aiming for more intensive glycaemic control may be associated

with adverse events, mainly hypoglycaemia, and increased overall mortality (Gerstein et al., 2011). Whereas the benefits of good glycaemic control are well established, the potential risks should also be considered and therapeutic targets should therefore be individualized and agreed with the patient.

Hyperhomocystinuria, vitamin B12, and folate deficiency

Higher homocysteine levels are associated with increased mortality and vascular disease in the general population and in patients with CKD (Zoccali et al., 2007). Homocysteine levels are higher in patients with all levels of renal impairment (Ganji et al., 2003), and are approximately three times higher in patients with ESRD (Heinz et al., 2009). Homocysteine levels can be reduced by supplementation with folic acid by approximately 25%, and vitamin B12 by approximately 7%. Trials of folate supplementation in both the general and CKD population, however, have failed to show any benefit (Lonn et al., 2006; Vianna et al., 2007). There are of course sound haematopoietic rationales for the use of folate supplements if there is anaemia or macrocytosis which can be attributed to folate deficiency. Indeed, correction of folate deficiency is generally good clinical practice irrespective of putative effects on homocysteine levels or vascular disease risk.

References Baigent, C., Landray, M. J., Reith, C., et al. (2011). The effects of lowering LDL cholesterol with simvastatin plus ezetimibe in patients with chronic kidney disease (Study of Heart and Renal Protection): a randomised placebo-controlled trial. Lancet, 377, 2181–92. Bibbins-Domingo, K., Chertow, G. M., Coxson, P. G., et al. (2010). Projected effect of dietary salt reductions on future cardiovascular disease. N Engl J Med, 362, 590–9. Brenner, B. M., Cooper, M. E., de Zeeuw, D., et al. (2001). Effects of losartan on renal and cardiovascular outcomes in patients with type 2 diabetes and nephropathy. N Engl J Med, 345, 861–9. Byrne, C., Ford, D., Gilg, J., et al. (2010). UK Renal Registry 12th Annual Report (December 2009): ­chapter 3: UK ESRD incident rates in 2008: national and centre-specific analyses. Nephron Clin Pract, 115 Suppl 1, c9–39. Casas, J. P., Chua, W., Loukogeorgakis, S., et al. (2005). Effect of inhibitors of the renin-angiotensin system and other antihypertensive drugs on renal outcomes: systematic review and meta-analysis. Lancet, 366, 2026–33. Chue, C.D., Edwards, N.C., Davis, L.J., et al. (2011). Serum phosphate but not pulse wave velocity predicts decline in renal function in

Chapter 99 

patients with early chronic kidney disease. Nephrol Dial Transplant, 26, 2576–82. Chue, C. D., Townend, J. N., Steeds, R. P., or delete from list. (2010). Arterial stiffness in chronic kidney disease: causes and consequences. Heart, 96, 817–23. Coresh, J., Selvin, E., Stevens, L. A., et al. (2007). Prevalence of chronic kidney disease in the United States. JAMA, 298, 2038–47. Cushman, W. C., Evans, G. W., Byington, R. P., et al. (2010). Effects of intensive blood-pressure control in type 2 diabetes mellitus. N Engl J Med, 362, 1575–85. De Boer, I. H., Rue, T. C., Cleary, P. A., et al. (2011). Long-term renal outcomes of patients with type 1 diabetes mellitus and microalbuminuria: an analysis of the Diabetes Control and Complications Trial/ Epidemiology of Diabetes Interventions and Complications cohort. Arch Intern Med, 171, 412–20. De Brito-Ashurst, I., Varagunam, M., Raftery, M. J., et al. (2009). Bicarbonate supplementation slows progression of CKD and improves nutritional status. J Am Soc Nephrol, 20, 2075–84. Douglas, K., O’Malley, P. G., and Jackson, J. L. (2006). Meta-analysis: the effect of statins on albuminuria. Ann Intern Med, 145, 117–24. Drueke, T. B., Locatelli, F., Clyne, N., et al. (2006). Normalization of hemoglobin level in patients with chronic kidney disease and anemia. N Engl J Med, 355, 2071–84. Edwards, N. C., Steeds, R. P., Stewart, P. M., et al. (2009). Effect of spironolactone on left ventricular mass and aortic stiffness in early-stage chronic kidney disease: a randomized controlled trial. J Am Coll Cardiol, 54, 505–12. Escape Trial Group (2009). Strict blood-pressure control and progression of renal failure in children. N Engl J Med, 361, 1639–50. Filiopoulos, V., Hadjiyannakos, D. and Vlassopoulos, D. (2012). New insights into uric acid effects on the progression and prognosis of chronic kidney disease. Ren Fail, 34(4), 510–20. Fouque, D. and Laville, M. (2009). Low protein diets for chronic kidney disease in non diabetic adults. Cochrane Database Syst Rev, 3, CD001892. Fox, K. M. (2003). Efficacy of perindopril in reduction of cardiovascular events among patients with stable coronary artery disease: randomised, double-blind, placebo-controlled, multicentre trial (the EUROPA study). Lancet, 362, 782–8. Ganji, V. and Kafai, M. R. (2003). Demographic, health, lifestyle, and blood vitamin determinants of serum total homocysteine concentrations in the third National Health and Nutrition Examination Survey, 1988-1994. Am J Clin Nutr, 77, 826–33. Gansevoort, R. T., Correa-Rotter, R., Hemmelgarn, B. R., et al. (2013). Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Lancet, 382(9889), 339–52. Gerstein, H. C., Miller, M. E., Genuth, S., et al. (2011). Long-term effects of intensive glucose lowering on cardiovascular outcomes. N Engl J Med, 364, 818–28. Goncalves, A. R., Khwaja, A., Ahmed, A. K., et al. (2011). Stopping renin-angiotensin system inhibitors in chronic kidney disease: predictors of response. Nephron Clin Pract, 119, c348–54. Guh, D. P., Zhang, W., Bansback, N., et al. (2009). The incidence of co-morbidities related to obesity and overweight: a systematic review and meta-analysis. BMC Public Health, 9, 88. Heinz, J., Domrose, U., Luley, C., et al. (2009). Influence of a supplementation with vitamins on cardiovascular morbidity and mortality in patients with end-stage renal disease: design and baseline data of a randomized clinical trial. Clin Nephrol, 71, 363–5. Hsu, C. Y. and Chertow, G. M. (2002). Elevations of serum phosphorus and potassium in mild to moderate chronic renal insufficiency. Nephrol Dial Transplant, 17, 1419–25. Jafar, T. H., Stark, P. C., Schmid, C. H., et al. (2003). Progression of chronic kidney disease: the role of blood pressure control, proteinuria, and angiotensin-converting enzyme inhibition: a patient-level meta-analysis. Ann Intern Med, 139, 244–52. Keith, D., Ashby, V. B., Port, F. K., et al. (2008). Insurance type and minority status associated with large disparities in prelisting dialysis among candidates for kidney transplantation. Clin J Am Soc Nephrol, 3, 463–70.

management of high renal risk ckd

Klahr, S., Levey, A. S., and Beck, G. J. (1994). The effects of dietary protein restriction and blood-pressure control on the progression of chronic renal disease. N Engl J Med, 330, 877–84. Kosmadakis, G. C., John, S. G., Clapp, E. L., et al. (2012). Benefits of regular walking exercise in advanced pre-dialysis chronic kidney disease. Nephrol Dial Transplant, 27, 997–1004. Kunz, R., Wolbers, M., Glass, T., et al. (2008). The COOPERATE trial: a letter of concern. Lancet, 371, 1575–6. Lee, W., Campoy, S., Smits, G., et al. (2007). Effectiveness of a chronic kidney disease clinic in achieving K/DOQI guideline targets at initiation of dialysis—a single-centre experience. Nephrol Dial Transplant, 22, 833–8. Levey, A. S., de Jong, P. E., Coresh, J., et al. (2011). The definition, classification, and prognosis of chronic kidney disease: a KDIGO Controversies Conference report. Kidney Int, 80, 17–28. Lewis, E. J., Hunsicker, L. G., Bain, R. P., et al. (1993). The effect of angiotensin-converting-enzyme inhibition on diabetic nephropathy. The Collaborative Study Group. N Engl J Med, 329, 1456–62. Lewis, E. J., Hunsicker, L. G., Clarke, W. R., et al. (2001). Renoprotective effect of the angiotensin-receptor antagonist irbesartan in patients with nephropathy due to type 2 diabetes. N Engl J Med, 345, 851–60. Lewis, J. B. (2010). Blood pressure control in chronic kidney disease: is less really more? J Am Soc Nephrol, 21, 1086–92. Locatelli, F., Covic, A., Eckardt, K. U., et al. (2009). Anaemia management in patients with chronic kidney disease: a position statement by the Anaemia Working Group of European Renal Best Practice (ERBP). Nephrol Dial Transplant, 24, 348–54. Lonn, E., Yusuf, S., Arnold, M. J., et al. (2006). Homocysteine lowering with folic acid and B vitamins in vascular disease. N Engl J Med, 354, 1567–77. Mann, J. F., Schmieder, R. E., McQueen, M., et al. (2008). Renal outcomes with telmisartan, ramipril, or both, in people at high vascular risk (the ONTARGET study): a multicentre, randomised, double-blind, controlled trial. Lancet, 372, 547–53. Matsushita, K., van der Velde, M., Astor, B. C., et al. (2010). Association of estimated glomerular filtration rate and albuminuria with all-cause and cardiovascular mortality in general population cohorts: a collaborative meta-analysis. Lancet, 375, 2073–81. Menon, V., Kopple, J. D., Wang, X., et al. (2009) Effect of a very low protein diet on outcomes: long-term follow-up of the MDRD study. Am J Kidney Dis, 53, 208–17. Navaneethan, S. D. and Yehnert, H. (2009). Bariatric surgery and progression of chronic kidney disease. Surg Obes Relat Dis, 5, 662–5. Paffenbarger, R. (2000). Physical exercise to reduce cardiovascular disease risk. Proc Nutr Soc, 59, 421–2. Parving, H. H., Persson, F., Lewis, J. B., et al. (2008). Aliskiren combined with losartan in type 2 diabetes and nephropathy. N Engl J Med, 358, 2433–46. Peterson, J. C., Adler, S., Burkart, J. M., et al. (1995). Blood pressure control, proteinuria, and the progression of renal disease. The Modification of Diet in Renal Disease Study. Ann Intern Med, 123, 754–62. Pfeffer, M. A., Burdmann, E. A., Chen, C. Y., et al. (2009). A trial of darbepoetin alfa in type 2 diabetes and chronic kidney disease. N Engl J Med, 361, 2019–32. Sandhu, S., Wiebe, N., Fried, L. F., et al. (2006). Statins for improving renal outcomes: a meta-analysis. J Am Soc Nephrol, 17, 2006–16. Schiffl, H., Lang, S. M. and Fischer, R. (2002). Stopping smoking slows accelerated progression of renal failure in primary renal disease. J Nephrol, 15, 270–4. Singh, A. K., Szczech, L., Tang, K. L., et al. (2006). Correction of anemia with epoetin alfa in chronic kidney disease. N Engl J Med, 355, 2085–98. Snyder, J. J. and Collins, A. J. (2009). Association of preventive health care with atherosclerotic heart disease and mortality in CKD. J Am Soc Nephrol, 20, 1614–22. Solomon, S. D., Rice, M. M., Jablonski, J. A., et al. (2006). Renal function and effectiveness of angiotensin-converting enzyme inhibitor therapy in patients with chronic stable coronary disease in the Prevention of Events with ACE inhibition (PEACE) trial. Circulation, 114, 26–31.

785

786

Section 5  

the patient with reduced renal function

Strippoli, G. F., Navaneethan, S. D., Johnson, D. W., et al. (2008). Effects of statins in patients with chronic kidney disease: meta-analysis and meta-regression of randomised controlled trials. BMJ, 336, 645–51. Turner, J. M., Bauer, C., Abramowitz, M. K., et al. (2012). Treatment of chronic kidney disease. Kidney Int, 81, 351–62. Vianna, A. C., Mocelin, A. J., Matsuo, T., et al. (2007). Uremic hyperhomocysteinemia: a randomized trial of folate treatment for the prevention of cardiovascular events. Hemodial Int, 11, 210–16. Vollmer, W. M., Sacks, F. M., Ard, J., et al. (2001). Effects of diet and sodium intake on blood pressure: subgroup analysis of the DASH-sodium trial. Ann Intern Med, 135, 1019–28.

Wright, J. T., Jr., Bakris, G., Greene, T., et al. (2002). Effect of blood pressure lowering and antihypertensive drug class on progression of hypertensive kidney disease: results from the AASK trial. JAMA, 288, 2421–31. Yoshida, T., Takei, T., Shirota, S., et al. (2008). Risk factors for progression in patients with early-stage chronic kidney disease in the Japanese population. Intern Med, 47, 1859–64. Yusuf, S., Sleight, P., Pogue, J., et al. (2000). Effects of an angiotensin-converting-enzyme inhibitor, ramipril, on cardiovascular events in high-risk patients. The Heart Outcomes Prevention Evaluation Study Investigators. N Engl J Med, 342, 145–53. Zoccali, C., Mallamaci, F. and Tripepi, G. (2007). It is important to lower homocysteine in dialysis patients. Semin Dial, 20, 530–3.

CHAPTER 100

Hypertension as a cause of chronic kidney disease: what is the evidence? Teena Tandon and Rajiv Agarwal Introduction Hypertension is a major public health problem worldwide. In the United States, it is estimated that one out of every three people has hypertension (Centers for Disease Prevention, 2013). After diabetes, hypertension is the second most common cause of end-stage renal disease (ESRD) (United States Renal Data System, 2013). But is it? Or is the hypertension in many of these patients caused by subclinical renal disease? To insinuate hypertension as a cause of chronic kidney disease (CKD), there are two levels of evidence—observational and interventional. First, the observational data are reviewed. Next, we review those randomized trials that have targeted blood pressure lowering to different levels in people with CKD to evaluate whether such interventions have any effect on CKD progression or cardiovascular disease. Finally, there is emerging evidence that specific genes may confer susceptibility to CKD rather than hypertension per se.

Observational data associate hypertension with end-stage renal disease Several large observational studies associate hypertension with CKD. Those that have used ESRD as an endpoint are discussed further. The Multiple Risk Factor Intervention Trial (MRFIT) (Klag et al., 1996) was a randomized primary prevention trial to test the effect of a multifactor intervention programme such as blood pressure control, diet modification for reducing blood cholesterol, and cessation of smoking on mortality from coronary heart disease in 12,866 high-risk men aged 35–57 years. In this trial, 332,544 men screened from 1973 to 1975 were followed prospectively over 16 years. Over this time, 814 cases of ESRD were identified. A continual and progressive increase in risk of ESRD was noted with increasing level of blood pressure. With optimal blood pressure (systolic blood pressure < 120  mmHg) as reference, stage 1 hypertension (systolic 140–159  mmHg or diastolic 90–99  mmHg) was associated with adjusted relative risk (RR) for ESRD of 3.1, stage 2 hypertension (systolic 160–179 mmHg or diastolic 100–109 mmHg) with adjusted RR of 6.0, stage 3 hypertension (systolic 180–209 mmHg or diastolic 110–119 mmHg) with adjusted RR 11.0, and stage 4

(systolic ≥ 210 mmHg or diastolic ≥ 120 mmHg) hypertension with even a higher adjusted RR for ESRD of 20 (Table 100.1). These data suggest that risk for CKD increases above systolic blood pressure >140 mmHg. Systolic blood pressure higher by 1 standard deviation (SD) was associated with a doubling of the risk of ESRD (P < 0.001) whereas diastolic blood pressure higher by 1 SD was associated with a 2.5-fold increase in the risk of ESRD (P < 0.001). Although MRFIT was limited to men, the Okinawa study (Tozawa et al., 2003) in Japan followed both men and women. Unlike the MRFIT study, the Okinawa study was without interventions. An observational cohort of 46,881 men and 51,878 women in Okinawa, Japan were evaluated at baseline and followed over 17  years. In both men and women, there was a significant positive association between systolic and diastolic blood pressure and the risk of development of ESRD. The RR for ESRD with each 10 mmHg increase in systolic blood pressure was 1.29 among men and 1.34 among women; for each 10 mmHg increase in diastolic blood pressure, the RR for ESRD was 1.56 in men and 1.69 in women. This study also concluded that in both men and women, high-normal blood pressure and hypertension are independent risk factors for ESRD. Jafar et al. (2001, 2003) performed a meta-analysis that included 11 randomized trials in 1860 patients with non-diabetic renal disease to evaluate the effect of achieved systolic blood pressure on renal outcome. Achieved systolic blood pressure of 110–129 mmHg and urine protein excretion < 2.0 g/day were associated with the lowest risk for CKD progression. Higher levels of systolic blood pressure were associated with a steep increase in the risk of CKD progression. Furthermore, achieved systolic blood pressure < 110 mmHg was also associated with increased risk of CKD progression. Although this meta-analysis included randomized trials, most of these trials did not randomize patients to two different blood pressure goals. In fact, all the trials in this meta-analysis tested the notion whether angiotensin converting enzyme inhibitor (ACEI) use can slow the progression of CKD. Thus, a cause and effect relationship is difficult to establish. Given this limitation, the authors appropriately concluded that ‘reverse causation cannot be excluded with certainty’. Moreover, although achieved systolic blood pressure of 110–129 mmHg was ideal, the adjusted RR for CKD progression was not statistically significant in when systolic blood pressure during follow-up was between 120–129 mmHg (adjusted RR 1.23; 95% confidence interval

788

Section 5  

the patient with reduced renal function

Table 100.1  MRFIT trial showed that risk of ESRD increases with stage of hypertension Stage of hypertension

Systolic blood pressure

Diastolic blood pressure

Adjusted RR of ESRD

Stage 1

140–159 mmHg

90–99 mmHg

3.1

Stage 2

160–179 mmHg

100–109 mmHg

6.0

Stage 3

180–209 mmHg

110–119 mmHg

11.0

Stage 4

>210 mmHg

>120 mmHg

20.0

(CI) 0.63–2.40) or between 130–139 mmHg (adjusted RR 1.83; 95% CI 0.97–3.44). Thus, achieved blood pressure of 140 mmHg or more was associated with increased CKD progression. Thus, high blood pressure is strongly associated with ESRD. This risk for ESRD is present even when hypertension is mild and occurs in the absence of history of diabetes or any other identifiable cause of ESRD. Nonetheless, association of blood pressure with ESRD does not establish causation. To insinuate that blood pressure is a cause of progressive CKD and therefore ESRD, randomized controlled data will be reviewed.

Randomized trials linking hypertension to chronic kidney disease Most of the evidence evaluating the level to which blood pressure should be lowered to has emerged from observational studies and post hoc analyses of randomized controlled trials that did not lower blood pressure to pre-specified goals. There are only three prospective randomized multicentre trials that have asked the question of how far blood pressure should be lowered to prevent the occurrence of ESRD and slow the progression of CKD. These trials were the Modification of Diet in Renal Disease (MDRD) Study, the African American Study of Kidney Disease (AASK), and the Ramipril Efficacy in Nephropathy (REIN-2) trial. The MDRD Study (Klahr et  al., 1994)  was a randomized controlled trial that tested the interventions of dietary modification and blood pressure lowering among 840 patients with CKD. Two studies were performed. In study 1, 585 patients with glomerular filtration rates (GFRs) of 25–55 mL/min/1.73 m2 body surface area were randomly assigned using a 2 × 2 factorial design to dietary or blood pressure modifications. The diets were either a usual-protein diet or a low-protein diet and blood pressure levels were usual- or a low-blood pressure group. Blood pressure assignments were by mean arterial pressure, 107 or 92 mmHg respectively in the two groups. In study 2, 255 patients with GFRs of 13–24 mL/min/1.73 m2 body surface area were randomly assigned to the low-protein diet or a very-low-protein diet with a keto acid-amino acid supplement, and a usual- or low-blood pressure groups. The mean follow-up was 2.2 years. In study 1, the mean decline in the GFR at 3  years did not differ significantly between the diet groups or between the blood pressure groups. There was no delay in the time to the occurrence of ESRD or death. In both studies, in post hoc analyses, patients who had more pronounced proteinuria at baseline had a significantly slower rate of decline in the GFR when they were randomized to the lower blood pressure group.

The AASK (Wright et  al., 2002)  used a 2  × 3 factorial design to compare the effects of two levels of blood pressure control and three antihypertensive drug classes on GFR decline in black patients with hypertension and CKD. African Americans (N  =  1094) aged 18–70  years with GFR between 20 and 65 mL/ min/1.73 m2 from 21 clinical centres throughout the United States were randomly assigned to one of two mean arterial pressure goals, 102–107 mmHg (usual goal; N = 554) or 92 mmHg or less (lower goal; N = 540), and to initial treatment with either a beta blocker (metoprolol 50–200 mg/day; N = 441), an ACEI (ramipril 2.5−10 mg/day; N = 436) or a dihydropyridine calcium channel blocker (amlodipine 5−10 mg/day; N = 217). Open-label agents were added to achieve the assigned blood pressure goals. Participants were followed up for 3–6.4  years. The main outcome measures were the following: rate of change in GFR (GFR slope); clinical composite outcome of reduction from baseline in GFR by 50% or more (or 25 mL/min/1.73 m2), ESRD, or death. Achieved blood pressure averaged (SD) 128/78 (12/8) mmHg in the lower-blood pressure group and 141/85 (12/7) mmHg in the usual-blood pressure group. The mean GFR slope from baseline through 4 years did not differ significantly between the lower-blood pressure group (−2.21 (0.17) mL/min per 1.73 m2 per year) and the usual-blood pressure group (−1.95 (0.17) mL/min per 1.73 m2 per year; P = 0.24), and the lower blood pressure goal did not significantly reduce the rate of the clinical composite outcome (risk reduction for lower blood pressure group = 2%; 95% CI −22% to 21%; P = 0.85). Thus, no additional benefit of slowing progression was observed with the lower blood pressure goal. Blood pressure control for renoprotection in patients with non-diabetic chronic renal disease (REIN-2) (Ruggenenti et al., 2005)  assessed the effect of intensified versus conventional blood pressure control on progression to ESRD. Participants with non-diabetic proteinuric nephropathies on ACEI ramipril (2.5–5 mg/day) were randomly assigned to either conventional (diastolic < 90 mmHg; N = 169) or intensified (systolic/diastolic < 130/80 mmHg; N = 169) blood pressure control. To achieve the intensified blood pressure level, patients received add-on therapy with the dihydropyridine calcium-channel blocker felodipine (5–10 mg/day). The primary outcome measure was time to ESRD over 36 month follow-up, and analysis was by intention to treat. Over a median follow-up of 19  months, 38/167 (23%) patients assigned to intensified blood pressure control and 34/168 (20%) allocated conventional control progressed to ESRD (hazard ratio (HR) 1.00; 95% CI 0.61–1.64; P = 0.99). The study was terminated early for futility. Among patients with non-diabetic proteinuric nephropathies receiving background ACEI therapy, no additional benefit from further blood pressure reduction by felodipine could be shown. Taken together, no study testing two different levels of blood pressure control was able to demonstrate that lower blood pressure delayed the rate of progression of renal failure. Due to lack to robust data for lower blood pressure, the 2012 KDIGO guidelines on blood pressure control in CKD patients recommend changing blood pressure target to 6.66 mg/g; women: > 15.24 mg/g) had an approximately twofold risk of developing hypertension (adjusted OR 1.93; P = 0.006) and 1.5-fold risk of hypertension progression (adjusted OR 1.45; P = 0.03). Among 1065 postmenopausal women from the first Nurses’ Health Study and 1114 premenopausal women from the second Nurses’ Health Study who had UACR < 25 mg/g and who did not have diabetes or hypertension, Forman et al. (2008) examined the association of UACR and incident hypertension. Among the older women, 271 incident cases of hypertension occurred during 4 years of follow-up, and among the younger women, 296 incident cases of hypertension occurred during 8 years of follow-up. Participants who had UACR in the highest quartile (4.34–24.17 mg/g for older women and 3.68–23.84 mg/g for younger women) were more likely to develop hypertension than those who had UACR in the lowest quartile (HR 1.76 (95% CI 1.21–2.56) and HR 1.35 (95% CI 0.97–1.91) for older and younger women, respectively). Higher UACRs, even within the normal range, are independently associated with increased risk for development of hypertension among women without diabetes. The effects of UACR on incident hypertension was independent of age, body mass index (BMI), estimated

Table 100.2  Details of studies relating albuminuria to hypertension Study

Method

Results

PREVEND trial (Brantsma et al., 2006)

Community-based prospective cohort study Examined association of urinary albumin excretion (UAE) (measured by two 24-hour urine collections) and hypertension in 4635 patients with normal blood pressure followed for 4.3 years

Baseline UAE associated with risk of developing hypertension (OR 2.29; 95% CI 1.77–2.95 per tenfold increase in UAE)

Framingham offspring study (Wang et al., 2005)

Community-based prospective cohort study Followed 1499 patients for 2.9 years

15% developed hypertension and 33% progressed to higher blood pressure. Thus, UAE predicts hypertension (adjusted OR 1.2; 95% CI 1.01–1.44)

Forman et al. (2008)

Nurses’ Health Study (NHS) cohort 1065 nurses from NHS I followed over 4 years 1114 nurses from NHS II followed over 8 years Participants did not have hypertension and diabetes and submitted a spot urine sample at enrolment

Compared to lowest quartile of UAE, highest quartile had HR 1.76 (P = 0.004) of incident hypertension in NHS I and HR 1.35 (P = 0.06 for trend) in NHS II

Jessani et al. (2012)

Cluster randomized controlled trial, nested cohort study Followed 1272 normotensive non-diabetic people in Pakistan for 2 years

Odds of developing hypertension higher in top quartile of UAE compared to lowest quartile

Palatini et al. (2005)

1033 patients screened for hypertension (ambulatory blood pressure UAE not helpful for predicting hypertension progression monitoring) and never treated for disease. Required hypertension to be present for admission into the cohort. UAE assessed by 24-hour urine

789

790

Section 5  

the patient with reduced renal function

GFR, baseline blood pressure, physical activity, smoking, and family history of hypertension. Similar findings have emerged from an Indo Asian population (Jessani et al., 2012). In a nested cohort study with a cluster randomized controlled trial of 1272 normotensive, non-diabetic Pakistani adults aged ≥ 40  years with UACR < 30 mg/g, subjects were followed for up to 2 years for incident hypertension. Incident hypertension was defined as new onset of systolic blood pressure ≥ 140 mmHg, or diastolic blood pressure ≥ 90 mmHg, or initiation of antihypertensive therapy. A total of 920 (72.3%) participants completed the 2-year final follow-up at which time 11.4% developed hypertension. The odds (95% CI) for incident hypertension were 2.45(1.21–4.98) for those in the fourth (top) quartile (≥ 6.1 mg/g) and 2.12 (1.04–4.35) in the third quartile (3.8–6.1 mg/g) compared to those in the lowest quartile (< 2.8 mg/g) of UACR. In addition, a significant interaction between UACR and baseline systolic blood pressure was observed suggesting that the odds (95% CI) of incident hypertension with UACR were greater at lower baseline systolic blood pressure (interaction P = 0.044). Some studies have questioned the association of albuminuria with hypertension. The Hypertension and Ambulatory Recording Venetia STudy (HARVEST) (Palatini et al., 2005) study in 17 outpatient clinics in Italy evaluated 1041 young stage 1 hypertensive subjects. Thus, the HARVEST participants were required to have a baseline systolic blood pressure of 140–159  mm Hg or a diastolic blood pressure of 90–99  mm Hg. Such individuals would be classified as having prevalent hypertension. Urinary albumin excretion rate was measured from 24-hour collections and 24-hour ambulatory monitoring was performed to assess blood pressure. The end point the development of sustained hypertension as defined by supine office systolic blood pressure ≥ 150 mmHg or supine office diastolic blood pressure ≥ 95 mmHg on two consecutive visits. At baseline, when compared to subjects with normoalbuminuria and normal filtration, those with microalbuminuria with hyperfiltration (P < 0.001) and or normal filtration (P = 0.04) had higher 24-hour systolic blood pressure. During the 12 years of observation, 461 of 1033 available subjects developed sustained hypertension. In a Cox analysis, neither microalbuminuria nor hyperfiltration were significant predictors of development of sustained hypertension. Given the greater correlation of urinary albumin excretion with 24-hour ambulatory blood pressure, it is possible that when 24-hour ambulatory blood pressure monitoring is available urinary albumin excretion offers little to improve the prediction of incident hypertension. However, the results of the HARVEST study do not apply to a population that is truly normotensive at baseline. Thus, hypertension is simply not a cause of kidney damage but taken together, the above studies suggest that early kidney damage may antedate hypertension in many people.

Nephron mass and hypertension Loss of nephrons due to peripheral cysts supports the hypothesis proposed by Brenner et al. (1988) and others that reduced number of nephrons contributes to hypertension in the general hypertension. This is reviewed in Chapter  138. This nephron number hypothesis was tested in autopsy study in Germany (Keller et al., 2003). A three-dimensional stereologic technique

was used to compare the number and volume of glomeruli in 10 middle-aged white patients (35–59 years of age) with a history of hypertension or left ventricular hypertrophy (or both) and renal arteriolar lesions with the number and volume in 10 normotensive subjects matched for sex, age, height, and weight. All 20 subjects had died in accidents. People with hypertension had significantly fewer glomeruli per kidney than matched normotensive controls (median 702,379 vs 1,429,200). Those with hypertension also had a significantly greater glomerular volume than the controls (median 6.50 × 10−3 mm3 vs 2.79 × 10−3 mm3; P < 0.001) but very few obsolescent glomeruli. These data provide evidence to support the concept that a low number of nephrons are associated with hypertension.

Genetic mutations as a cause of ‘hypertensive’ chronic kidney disease In 2008, landmark papers reviewed in Chapter  341 showed a very strong association between the increased incidence of focal segmental glomerulosclerosis (FSGS) in African Americans and markers associated with the MYH9 gene on chromosome 22. These analyses used the new technique of admixture analysis that depended on tracking segments on ‘African’ DNA in a mixed population. Further work showed that the same genetic variants could also account for much of the excess of renal disease in African Americans (Kao et al., 2008; Freedman et al., 2009). Renal disease in most of these individuals had been classified as hypertensive. African American patients have a high incidence rate of ESRD (Kidney Disease:  Improving Global Outcomes (KDIGO) CKD Work Group, 2013) attributed to hypertension at 33% compared to < 25% in all other racial groups. Hypertensive ESRD is found to aggregate in select African American families and relatively few patients with mild to moderate essential hypertension ultimately develop CKD. The MYH9 gene is associated with autosomal dominant renal disease, but Genovese et  al. (2010) showed that the strongest association signal came not from the MYH9 gene but from the adjacent APOL1 gene. APOL1 (Freedman et al., 2010; Friedman and Pollak, 2011; Freedman and Langefeld, 2012)  mutations are found exclusively among individuals of African descent and have traditionally been known to provide protection from sleeping sickness (trypanosomiasis). So far, little is known about how these variants might predispose to renal disease, but further evidence continues to emerge about their importance (see Chapter 341). These studies have shown that APOL1 mutations produce a distinct category of kidney disease that manifests in African ancestry population and support the notion that it is a genetic cause and not necessarily hypertension that predisposes to kidney disease. This disease association has altered our understanding of the factors that initiate hypertensive ESRD. Several studies have previously shown that high blood pressure does not commonly lead to progressive nephropathy in African American patients. It is now apparent that many patients with African American ancestry and ESRD attributed to hypertension may have been misdiagnosed and actually have APOL1-associated FSGS. This could explain the failure of intensive blood pressure reduction, including with the use of ACE inhibition, to slow nephropathy progression in hypertensive African Americans with CKD.

Chapter 100 

Once translated from research to bedside genetic screening, these risk markers will have potential clinical uses especially in transplantation. Genetic screening could be used to identify those donors with better graft survival, as well as for high-risk, black, live donors, reducing risk of subsequent nephropathy.

Unusual renal causes of hypertension Fatty kidney Kidneys have the potential to accumulate ectopic fat in the renal sinus. Excessive accumulation of fat within the renal sinus (fatty kidney) displaces and compresses the low-pressure renal lymphatics and veins, as well as the ureters. It is thought that compression of these structures increases renal hydrostatic pressure and activates the renin–angiotensin–aldosterone system (RAAS) which leads to hypertension. Chughtai et  al. (2010) tested this association among 205 patients aged 55–85 years who were recruited as part of the Pulmonary Edema and Stiffness of the Vascular System (PREDICT) trial, an ongoing prospective observational study the aim of which is to identify abnormalities of the cardiovascular system that predicts the first episode of congestive heart failure in middle-aged and elderly individuals. After accounting for age, sex, height, BMI, and intraperitoneal fat, multivariable linear regression revealed that renal sinus fat was associated with the number of prescribed antihypertensive medications (P = 0.010), stage II hypertension (P = 0.02), and renal size (P ≤ 0.001). Foster et al. (2011) showed similar findings in the Framingham Heart Study cohort. In this cross-sectional study, 2923 patients (mean age: 54 years; 51% women and mainly white) underwent quantification of renal sinus fat area using computed tomography. They found that prevalence of fatty kidney was 30.1% (N = 879). Individuals with fatty kidney had a higher risk of hypertension (odds ratio (OR) 2.12; P < 0.0001), which persisted after adjustment for BMI (OR 1.49; P < 0.0001). In this study, fatty kidney was also associated with an increased risk for CKD (OR 2.30; P = 0.005), even after additionally adjusting for BMI (OR 1.86; P = 0.04). However, further research is necessary to evaluate the longitudinal associations of renal sinus fat with markers of renal function.

Renal cysts Hypertension has classically been associated with autosomal dominant polycystic kidney disease (ADPKD) (Chapman et  al., 2010) and in fact predates the loss of renal function in these individuals (see Chapter 306). It has been postulated that in ADPKD, increased cyst growth and renal volume leads to compression of renal vasculature resulting in activation of the RAAS. RAAS activation leads to sodium retention, increased system vascular resistance, further cyst growth, and renal fibrosis, all of which lead to hypertension and renal disease. Not only ADPKD but simple renal cysts have also been associated with hypertension (Luscher et al., 1986). A simple renal cyst is one of the most common types of acquired renal cysts whose prevalence increases with age and varies according to gender. In 1942, Farrell and Young first reported the association of a simple renal cyst in an 18-year-old student with hypertension following trauma to the right kidney at age 6 (Farrell and Young, 1942). Following nephrectomy, hypertension resolved. Since then several authors have reported cure or improvement of hypertension after decompression of large cysts while others have refuted this finding.

hypertension as a cause of ckd

To examine this association, investigators in Korea (Chin et  al., 2006)  retrospectively identified 436 persons with a simple renal cyst(s) and 436 matched controls from among 6603 patients having routine health check-ups and abdominal ultrasounds. They then analysed the medical record of study subjects with careful attention to the confounding effects of age, gender, proteinuria, GFR, and hypertension on the existence of a simple renal cyst, hypertension, and renal dysfunction. The presence of a cyst was related to hypertension but not to renal dysfunction. The number and the size of cysts were independent risk factors to the prevalence of hypertension. In this study, subjects with peripheral cysts had higher prevalence of hypertension compared to perihilar cysts refuting the role of RAAS activation by cysts as a cause of hypertension. The authors postulated that the loss of nephrons along with ageing was involved in both development of hypertension and formation of a simple renal cyst. Ageing process along with decreased cortical mass causes aberrant tubular growth to generate cyst formation, especially in peripheral region as well as hypertension subsequent to reduced number of nephrons. Additional loss of renal cortex by peripheral cysts may increase the probability of hypertension.

References Brantsma, A. H., Bakker, S. J., de Zeeuw, D., et al. (2006). Urinary albumin excretion as a predictor of the development of hypertension in the general population. J Am Soc Nephrol, 17, 331–5. Brenner, B. M., Garcia, D. L., and Anderson, S. (1988). Glomeruli and blood pressure. Less of one, more the other? Am J Hypertens, 1, 335–47. Centers for Disease Prevention (2013). High Blood Pressure. [Online]

Chapman, A. B., Stepniakowski, K., and Rahbari-Oskoui, F. (2010). Hypertension in autosomal dominant polycystic kidney disease. Adv Chronic Kidney Dis, 17, 153–63. Chin, H. J., Ro, H., Lee, H. J., et al. (2006). The clinical significances of simple renal cyst: Is it related to hypertension or renal dysfunction? Kidney Int, 70, 1468–73. Chughtai, H. L., Morgan, T. M., Rocco, M., et al. (2010). Renal sinus fat and poor blood pressure control in middle-aged and elderly individuals at risk for cardiovascular events. Hypertension, 56, 901–6. Farrell, J. I. and Young, R. H. (1942). Hypertension caused by unilateral renal compression. JAMA, 118, 711–12. Forman, J. P., Fisher, N. D., Schopick, E. L., et al. (2008). Higher levels of albuminuria within the normal range predict incident hypertension. J Am Soc Nephrol, 19, 1983–8. Foster, M. C., Hwang, S. J., Porter, S. A., et al. (2011). Fatty kidney, hypertension, and chronic kidney disease: the Framingham Heart Study. Hypertension, 58, 784–90. Freedman, B. I., Hicks, P. J., Bostrom, M. A., et al. (2009). Polymorphisms in the non-muscle myosin heavy chain 9 gene (MYH9) are strongly associated with end-stage renal disease historically attributed to hypertension in African Americans. Kidney Int, 75, 736–45. Freedman, B. I., Kopp, J. B., Langefeld, C. D., et al. (2010). The apolipoprotein L1 (APOL1) gene and nondiabetic nephropathy in African Americans. J Am Soc Nephrol, 21, 1422–6. Freedman, B. I. and Langefeld, C. D. (2012). The new era of APOL1-associated glomerulosclerosis. Nephrol Dial Transplant, 27, 1288–91. Freedman, B. I., Soucie, J. M., McClellan, W. M. (1997). Family history of end-stage renal disease among incident dialysis patients. J Am Soc Nephrol, 8, 1942–5. Friedman, D. J. and Pollak, M. R. (2011). Genetics of kidney failure and the evolving story of APOL1. J Clin Invest, 121, 3367–74. Genovese, G., Tonna, S. J., Knob, A. U., et al. (2010). A risk allele for focal segmental glomerulosclerosis in African Americans is located within a region containing APOL1 and MYH9. Kidney Int, 78, 698–704.

791

792

Section 5  

the patient with reduced renal function

Jafar, T. H., Schmid, C. H., Landa, M., et al. (2001). Angiotensin-converting enzyme inhibitors and progression of nondiabetic renal disease. A meta-analysis of patient-level data. Ann Intern Med, 135, 73–87. Jafar, T. H., Stark, P. C., Schmid, C. H, et al. (2003). Progression of chronic kidney disease: the role of blood pressure control, proteinuria, and angiotensin-converting enzyme inhibition: a patient-level meta-analysis. Ann Intern Med, 139, 244–52. Jessani, S., Levey, A. S., Chaturvedi, N., et al. (2012). High normal levels of albuminuria and risk of hypertension in Indo-Asian population. Nephrol Dial Transplant, 27 Suppl 3, iii58–64. Kao, W. H., Klag, M. J., Meoni, L. A., et al. (2008). MYH9 is associated with nondiabetic end-stage renal disease in African Americans. Nat Genet, 40, 1185–92. Keller, G., Zimmer, G., Mall, G., et al. (2003). Nephron number in patients with primary hypertension. N Engl J Med, 348, 101–8. Kidney Disease: Improving Global Outcomes (KDIGO) CKD Work Group (2013). KDIGO 2012 Clinical Practice Guideline for the Evaluation and Management of Chronic Kidney Disease. Kidney Int Suppl, 3, 1–150. Klag, M. J., Whelton, P. K., Randall, B. L., et al. (1996). Blood pressure and end-stage renal disease in men. N Engl J Med, 334, 13–8. Klahr, S., Levey, A. S., Beck, G. J., et al. (1994). The effects of dietary protein restriction and blood-pressure control on the progression of chronic renal disease. Modification of Diet in Renal Disease Study Group. N Engl J Med, 330, 877–84.

Luscher, T. F., Wanner, C., Siegenthaler, W., et al. (1986). Simple renal cyst and hypertension: cause or coincidence? Clin Nephrol, 26, 91–5. Palatini, P., Mormino, P., Mos, L., et al. (2005). Microalbuminuria, renal function and development of sustained hypertension: a longitudinal study in the early stage of hypertension. J Hypertens, 23, 175–82. Ruggenenti, P., Perna, A., Loriga, G., et al. (2005). Blood-pressure control for renoprotection in patients with non-diabetic chronic renal disease (REIN-2): multicentre, randomised controlled trial. Lancet, 365, 939–46. Svenningsen, P., Bistrup, C., Friis, U. G., et al. (2009). Plasmin in nephrotic urine activates the epithelial sodium channel. J Am Soc Nephrol, 20, 299–10. Tozawa, M., Iseki, K., Iseki, C., et al. (2003). Blood pressure predicts risk of developing end-stage renal disease in men and women. Hypertension, 41, 1341–5. United States Renal Data System (2013). USRDS 2012 Annual Data Report: Atlas of CKD and ESRD in United States. Bethesda, MD: National Institutes of Health, National Institute of Diabetes and Digestive and Kidney Diseases. Wang, T. J., Evans, J. C., Meigs, J. B., et al. (2005). Low-grade albuminuria and the risks of hypertension and blood pressure progression. Circulation, 111, 1370–6. Wright, J. T., Jr., Bakris, G., Greene, T., et al. (2002). Effect of blood pressure lowering and antihypertensive drug class on progression of hypertensive kidney disease: results from the AASK trial. JAMA, 288, 2421–31.

CHAPTER 101

Diet and the progression of chronic kidney disease Juan Jesús Carrero, Hong Xu, and Bengt Lindholm Introduction The dietary management of non-dialysed CKD patients has focused on limiting the intake of substances which lead to accumulation of urea, potassium, phosphorus, and sodium. The hypothesis that a high dietary protein intake leads to progressive CKD through a mechanism of glomerular hyperfiltration has been taught for decades. However, the evidence that low-protein diets (LPDs) halt CKD progression is weak. Recent advances in nutritional epidemiology have given us the opportunity to examine the relationships between diet and CKD. This chapter is an overview of dietary factors and interventions that can affect progression. We restrict this review to patients in the earlier stages of CKD; conservative management of Stage 5 CKD is considered in Chapter 145. Nutritional epidemiology is complex. The estimation of individuals’ nutritional intake depends mostly on semi-quantitative food frequency questionnaires and food diaries, which are dependent on the accuracy of the reporters, and the precision by which these records are translated into estimation of actual intakes by use of food composition tables. The study of diet and disease progression or incidence also assumes that dietary habits remain more or less constant over time. Given the mixture of nutrients (we eat meals, not single nutrients), the effect attributed to a single item may be influenced by other components in the foods. Finally, diet is inevitably linked to lifestyle-related factors such as obesity, smoking, and physical activity which also influence patient outcomes.

Macronutrients and progression of chronic kidney disease Protein intake: not high, but weak evidence for restriction From early studies it was postulated that a high-protein diet results in glomerular hyperfiltration, leading to kidney injury. Thus, it is plausible that the high protein intake would affect the kidneys of healthy adults or those with CKD. In the Nurse’s Health Study, a high intake of non-dairy animal protein in women with mild kidney insufficiency was associated with albuminuria and a significantly greater loss in glomerular filtration rate (GFR) (Lin et al., 2011). This agreed with observational data linking higher intake

of protein-rich foods with albuminuria in the community and in individuals with hypertension and diabetes. High-protein diets are therefore not recommended in patients with clear-cut CKD. Apart from its effect on renal blood flow, dietary protein is a source of nitrogen, phosphorus, potassium, and metabolic acids that need to be excreted by the kidneys. Conversely, dietary protein restriction may protect against the progression of CKD by haemodynamic-mediated reductions in intraglomerular pressure as well as by changes in cytokine expression and matrix synthesis. Various randomized controlled trials (RCTs) have evaluated both the efficacy and safety of a LPD in patients with progressive CKD, but the evidence remains inconclusive. Protein restriction to approximately 0.6–0.8 g/kg/day improved a number of uraemic symptoms and was associated with a modest and not significant benefit on CKD progression (Aparicio et  al., 2000; Bernhard et al., 2001). Other RCTs including those in type 1 diabetic nephropathy patients showed that LPDs resulted in slower decline of GFR (Walker et al., 1989; Zeller et al., 1991). However, in the Modification of Diet in Renal Diseases (MDRD) Study, 600 individuals with moderate CKD, a low- versus usual-protein diet (0.58 vs 1.3 g/kg/day) did not result in a slower decline of kidney function (Klahr et  al., 1994). A  longer-term follow-up using national registries of these patients suggested that there might be an early benefit with low protein intake on the need for renal replacement therapy initiation or the composite of ESRD or all-cause mortality (Levey et al., 2006). Several meta-analyses have examined the relationship between a LPD and the initiation of renal replacement therapy by pooling data from previous interventions. The largest of these, based on 10 RCTs, concluded that CKD patients taking a LPD had a reduced risk for end-stage renal disease (ESRD) than those consuming higher amounts of protein (Fouque and Laville, 2009). To date, given the lack of conclusive data, dietary protein restriction cannot be recommended as a routine kidney protective strategy for patients with CKD. However, the safety of LPD and its beneficial effects on uraemic symptoms are obvious so its prescription may be justified in selected cases. Apart from the quantity of protein intake, the source of protein may also be important. Whereas animal protein and amino acid mixtures may increase GFR, vegetable protein and egg whites produce little or no effect. However, it should be noted that

794

Section 5  

the patient with reduced renal function

animal protein intake can increase serum creatinine directly as well as indirectly by increasing muscle mass, which could have influenced the interpretation of those trials. Interventional studies with non-serum creatinine-based measures of GFR as the outcome are needed before strong recommendations can be made about animal versus vegetable sources of protein in early CKD. In more advanced CKD, however, vegetable sources of protein might have additional benefits, such as lower absorption rate of bound-phosphate. Indeed, a cross-over trial in CKD stage 4 patients showed that 7  days of vegetarian diet with equivalent nutrients to a meat diet led to lower serum phosphate and decreased fibroblast growth factor 23 (FGF23) levels (Moe et al., 2011). Vegetarian diets are also associated with a decreased production of uraemic toxins such as p-cresyl sulphate and indoxyl sulphate, which have been implicated in CKD progression. Finally, a diet high in vegetable sources of protein might lead to lower endogenous production of acid. The 2012 KDIGO guidelines suggest the use of a lower, high-quality protein diet of 0.8 g/kg/day among select pre-dialysis patients who are highly motivated to follow such a diet (Stevens et al., 2013). Other recommendations, such as those from the US National Kidney Foundation or International Society of Renal Nutrition and Metabolism suggest a lower range of protein restriction in CKD stages 3–5 to 0.6–0.8 g/kg/day, provided there are no signs of malnutrition. A concern of LPDs in non-dialysed CKD is that they might eventually compromise nutritional status. RCTs have not shown that a LPD results in malnutrition (Klahr et  al., 1994; Aparicio et al., 2000; Bernhard et al., 2001), possibly because these studied patients are subjected to more stringent dietary counselling and control. This supports the advantages of appropriate dietary counselling and monitoring in CKD. Adequate caloric intake must be ensured and at least 60% of the ingested protein must be of high biologic value or contain a high percentage of essential amino acids to ensure net nitrogen balance. Because dietary protein is a source of metabolic acids that stimulate skeletal muscle protein breakdown, LPDs may be associated with less metabolic acidosis. If the patients appear to be at risk of malnutrition, the LPD can be supplemented with essential amino acids and/or keto acids. The use of hypercaloric renal-specific supplements in the context of a LPD was associated with protein intake closer to the target values, better nutritional measures, and better adherence to therapy than a LPD alone (Montes-Delgado et al., 1998). Likewise, keto-acids of essential amino acids can also be used because of their capacity to neutralize the excessive nitrogen residues through transamination and limit the production of urea, while, at the same time, allowing the preservation of nutritional status (Mircescu et al., 2007). Adherence to a LPD is often difficult for an individual. The question of the effectiveness of LPD in slowing the progression of CKD is irrelevant if patients are not able to follow it. Nutritional education programmes and dietitians’ advice are effective in increasing patient adherence (Paes-Barreto et al., 2013), but the support of family members and whoever prepares the food is also critical. Ideally, the LPD prescribed has to be pleasant, varied, and not too restrictive. A recent study proposed a simplified approach to a LPD prescription (Piccoli et al., 2013), using a vegetarian LPD supplemented with ketoanalogues. The simplified diet was based on the idea of forbidden and allowed foods (forbidden: fish, meat,

milk, eggs and derivatives (except in the context of the free-choice meals); everything else is allowed). The diet was essentially vegan, with an average of 0.6 g/kg/day protein intake and of 30–35 kcal/kg/ day, and was supplemented with ketoanalogues. To improve compliance, one to three free-choice meals per week were allowed and the foods were not weighed. Adherence and feasibility were good, suggesting this as an easier way to put this theory into practice. Compliance with dietary protein restriction can be estimated from measurement of urea in a 24-hour urine collection, assuming that daily intake is relatively constant and the patient is in a steady state (shown by a stable blood urea nitrogen and/or body weight). In this setting, nitrogen (N)  excretion roughly equals nitrogen intake. The former can be estimated from: N excretion  urinary urea N excretion  non-urea N excretion Non-urea nitrogen excretion (sum of non-urea urine nitrogen plus faecal nitrogen) is relatively constant, averaging 31 mg/kg per day. Moderate urinary protein loss can be ignored, but each gram excreted above 5 g/day should be added to the above formula. Each gram of nitrogen is derived from 6.25 g of protein. Thus: Estimated dietary protein intake  6.25 urine urea N  31 mg / kg If, for example, 24-hour urine urea nitrogen excretion is 8.2 g in a 60 kg woman excreting 3.5 g of protein per day, then: Estimated dietary protein intake  6.25 8.2  1.86  62.99 g Thus, protein intake in this example is approximately 1 g/kg per day.

Fat intake Although there is limited evidence that reduction in total dietary fat intake per se decreases CVD, replacement of dietary saturated fat and trans-fat with unsaturated fat is recommended for prevention of CVD in the general population. The potential of dietary fat interventions for CKD progression has so far been mainly limited to the study of n-3 polyunsaturated fatty acids (PUFAs) derived from fish oil. Less is known of the health risks of a diet with an unhealthy fat profile. There are three main types of fatty acids in humans: saturated fatty acids (SFAs), monounsaturated fatty acids (MUFAs), and PUFAs. The latter two are further classified into n-3, n-6, and n-9 (or omega-3, -6, and -9) subfamilies (IUPAC-IUB Commission on Biochemical Nomenclatur, 1978). Major dietary sources of fatty acids are summarized in Table 101.1. Fatty acids have many biological functions, for example, a source of energy, basic building blocks of lipids and membranes, second messengers regulating gene expression, precursors of eicosanoids, and as antioxidants, anti-inflammatory agents, and anticoagulants. Epidemiological studies investigating associations between SFAs and kidney function suggest a negative effect of SFAs on kidney function. In the Nurses’ Health Study, women within the highest quartile of SFA intake had increased odds of GFR decline ≥ 30% during 12 years (Lin et al., 2010). In contrast, most, but not all epidemiological studies investigating associations between n-3 PUFA and kidney function suggest a protective effect against the decline in GFR. Owing to the purported salutary effects of n-3 PUFAs,

Chapter 101 

Table 101.1  Major dietary sources of fatty acids Common name

Dietary sources

Saturated fatty acids Palmitic acid

Meats, cheeses, butter, palm oil

Stearic acid

Animal fat, cocoa butter and shea butter

Monounsaturated fatty acids Palmitoleic acid

Macadamia oil, sea buckthorn oil

Oleic acid

Sunflower oil, safflower oil

Polyunsaturated fatty acids n-6 subfamily: Linoleic acid

Sunflower seed, corn, soya, sesame, canola, safflower and their oils

Arachidonic acid

Meat, eggs

n-3 subfamily: Alpha-linolenic acid

Rapeseed, soybeans, walnuts, flaxseed, perilla, chia, hemp and their oils

Eicosapentaenoic acid

Oily fish, seafood, seaweed, krill oil, seal oil

Docosahexaenoic acid

Oily fish, seafood, seaweed, krill oil, seal oil

several RTCs have tested benefits of n-3 PUFA supplementation on proteinuria/albuminuria and renal function in patients mostly with diabetic nephropathy, immunoglobulin A  (IgA) nephropathy, or lupus nephritis. These studies are characterized by their small sample size and short duration. Only a few found significant effects of reducing proteinuria/albuminuria or retarding the reduction of GFR. Though not reaching statistical significance, the other RCTs generally showed benefits on markers of kidney damage. A meta-analysis pooling all these studies concluded that there was a greater reduction in urine protein excretion in the intervention group compared to the control group (Miller et  al., 2009). Moreover, this pooled analysis showed that the decline in GFR is slower in those with n-3 PUFA supplementation than in the controls, but this effect did not reach statistical significance (Miller et  al., 2009). Currently, there are no specific recommendations regarding dietary fat intake in patients with CKD. Along with the recommendations made at the general population level, a reduction in SFA intake and an increase in n-3 PUFA intake should be advocated for cardioprotection. When recommending fish intake in CKD patients, prescribers should advice on the phosphate to protein ratio of different fish species.

Carbohydrate intake Sugar and sweeteners There is no macronutrient whose consumption has increased as much in humans as sugars and sweeteners, especially after the introduction of soda drinks in the 1940s. Much research is devoted to the association between these nutrients and the development of obesity. What are the potential roles in the development of CKD? A number of epidemiological studies have suggested an association between consumption of sugary drinks/sodas and the presence of CKD in the community. The consumption of these items has also been associated with the development of albuminuria or CKD and

diet and the progression of chronic kidney disease with a faster GFR decline (Bomback et al., 2009; Lin et al., 2011). There is currently no evidence of an effect of interventional studies restricting sugar/sweetener intake in CKD patients. A RCT in patients with CKD stages 2–3 demonstrated, however, that a low fructose diet significantly reduced a number of inflammatory biomarkers and blood pressure (Brymora et al., 2012). There are some plausible explanations to explain these links. First, dietary sugar (and particularly fructose) may increase uric acid levels, which in turn may promote hypertension and kidney damage. Second, excessive sugar intake may promote obesity, which in turn may promote kidney damage. Third, excessive sugar intake increases the risk of diabetes mellitus, which is the most common cause of CKD (Karalius and Shoham, 2013).

Fibre intake Because potassium concentrations are high in fruits and vegetables, CKD and dialysis patients are counselled against taking these. This might result in a decreased intake of fibre. A recent study found that CKD patients with lower fibre intake have elevated serum C-reactive protein levels and increased mortality (Krishnamurthy et al., 2012). In a series of RCTs, a higher intake of fruits and vegetables in patients with CKD stages 1–4 was compared to the efficacy of oral bicarbonate. The rationale was that high-alkali fruit intake would reduce the dietary acidic load and be able to control acidic status and subsequent kidney injury; In CKD stage 2, both treatments attenuated kidney injury to a similar extent (Goraya et  al., 2012). In CKD stages 3–4, metabolic acidosis was significantly reduced by both treatments (although as expected more in the bicarbonate group) (Goraya et al., 2013). With increased but controlled fruit/vegetable intake, hyperkalaemia was not induced (Goraya et al., 2013). There is no evidence that fibre or fruit/vegetable intake can delay CKD progression.

Micronutrients and progression of chronic kidney disease Dietary salt: usually restrict intake Hypertension is a well-established cause and complication of CKD and a mediator of CKD progression. Epidemiological evidence suggests that salt consumption at a community level may impair kidney function. For instance, dietary sodium was associated with an increased odds ratio for albuminuria in obese, but not in normal-weight, adults (Aaron et al., 2011). In a 14-years follow-up study, higher dietary sodium intake was independently associated with an estimated GFR decline ≥ 30% from baseline (Lin et  al., 2010). Dietary programmes restricting salt intake in the community are effective in controlling hypertension, for example, the Dietary Approaches to Stop Hypertension (DASH)-Sodium study, in which lowering sodium intake to 1 or 0.6–1.0 g/day reduced systolic blood pressure by 7.1 and 11.5 mmHg, respectively (Sacks et  al., 2001). There is, however, no evidence from interventional trials to support the belief that a reduction of dietary sodium will prevent the development of CKD. The kidney is the major organ of salt excretion and so plays a key role in the maintenance of sodium balance. Most patients with moderate-advanced CKD have decreased ability to excrete a high sodium load and develop salt sensitive hypertension. Indeed, the prevalence of salt sensitive hypertension increases with age especially when kidney function is declining. In contrast, CKD subjects

795

Section 5  

the patient with reduced renal function

with tubulointerstitial disease may maintain normal blood pressure because they excrete salt more readily. An interventional study (Yu et al., 2012) assigned 176 patients with non-dialysed hypertensive CKD to a low-sodium diet (2300 mg/day) over 7 days. Compared with a control group, the sodium restriction group had greater reduction in systolic and diastolic blood as well as a reduction in urine protein excretion. The ideal sodium intake for non-dialysed CKD patients is a matter of debate. Most guidelines specific to the nutritional management of CKD patients recommend an upper limit of 2.3 g/ day (Fouque et  al., 2007; Levin et  al., 2008). The United States dietary guidelines recommend a stricter target of 1.5 g/day for people with CKD. The nature of the Western diet and the excess of salt in processed foods, mean that a careful dietary plan may not be sufficient to achieve these goals. In fact, mean 24-hour urinary sodium excretion in large studies suggests that CKD patients commonly have sodium intakes far above those recommendations. Identified barriers to adherence to a salt restrictive diet are (a) perceived taste/palatability of low-sodium foods, (b) convenience/difficulty (e.g. time, availability of low-sodium foods, interference with socialization, and cost) or, (c)  lack of knowledge or understanding (e.g. lack of perceived benefit and inability to identify low-sodium foods). Bread, baked products, pre-cooked foods, and sausages are the most common sources of sodium in a Western diet besides the salt added to meals. Sodium content is often reported on food labels, and as a rule of thumb, foods containing < 0.5 g of sodium/100 g can be considered low in sodium. Spices and aromatic herbs may replace sodium to enhance taste and appetite. A skilled nutritionist is needed to assist patients in restricting dietary salt. Various RCTs suggest that dietary sodium restriction enhances the renal and cardiovascular protective effects of first-choice antihypertensive therapies in patients with diabetic and non-diabetic kidney disease. For instance, treatment with angiotensin receptor blockers (ARBs) compared with non-renin–angiotensin–aldosterone system (RAAS) intervention produced the greatest long-term effects on renal and cardiovascular events in the lowest tertile of sodium intake (Fig. 101.1) (Lambers Heerspink et  al., 2012). Conversely, drug-resistant hypertension, including therapy with angiotensin-converting enzyme inhibitors (ACEIs) or ARBs, is often caused by a high level of salt intake.

4 3 2

* *

1 0

Placebo

losartan

losartan+HCT

Fig. 101.1  Effect of dietary salt reduction, hydrochlorothiazide, and their combination on proteinuria in proteinuric patients treated with losartan. Reproduced from Lambers Heerspink et al. (2012).

Dietary potassium: restrict only when indicated The kidney excretes 90% of dietary potassium and in the presence of kidney dysfunction, the intestinal excretion of potassium is increased as a compensatory mechanism. With advanced renal insufficiency, acidosis, or other chronic conditions, the patient’s ability to excrete potassium can be impaired. Impaired potassium excretion is the consequence not only of the loss of kidney function, but also an impairment in the action of protective hormones (e.g. aldosterone) or use of medication than can influence potassium levels, such ACEIs, ARBs, or non-steroidal anti-inflammatory drugs. Fortunately, there are adaptative mechanisms that increase the excretion of potassium through both the kidney and gut, making it possible to comply with potassium restriction approaches. There have been few investigations of the influence of dietary potassium modifications on CKD progression, but the clinical consequences of hyperkalaemia justify the need to reduce potassium intake. On the other hand, there is substantial evidence that diets rich in potassium (e.g. fruits and vegetables) reduce the likelihood of developing chronic diseases, such as coronary heart disease and diabetes. The National Kidney Foundation’s expert panel recommended potassium restriction for individuals with advanced CKD stages 4–5 (Kidney Disease Outcomes Quality Initiative (K/DOQI), 2004). However, in pre-dialysis patients, the need for potassium restriction is variable and limits the validity of general recommendations. For instance, potassium supplementation is seldom justified in those on diuretics. In contrast, potassium restriction is needed in those with hyperkalaemia receiving RAAS blockers. Steroids, ACEIs, and potassium-sparing diuretics may induce hyperkalaemia. Loss of intracellular potassium is promoted by acidosis and hyperglycaemia. Hence, serum potassium levels should guide the potassium intake in these patients, and before restricting dietary potassium, a careful exploration for other causes of hyperkalaemia is necessary. As a practical recommendation to patients, leaching or boiling vegetables in water reduces their mineral content and may be a useful adjunct to a diet low in potassium. However, such techniques may induce loss of nutrients other than potassium, so the patient’s nutritional status must be continuously monitored. An excessive consumption of fruits and vegetables may lead to hyperkalaemia, but fruits and vegetables also provide other nutrients of interest for CKD management. Finally, it should also be remembered that food additives may contain significant quantities of ‘hidden’ potassium.

Dietary phosphorus: restrict when indicated

High sodium Low sodium

* Proteinuria (g/d)

796

Organic phosphorus is bound to protein, so the amount of protein eaten will predict phosphorus intake. The intestinal absorption of dietary phosphorus is lower if it comes from animal sources (40–60% absorbed) than that from vegetable sources (10–30% absorbed). In addition, inorganic phosphorus, which is added to processed foods for conservation and enhancement of taste, is almost entirely absorbed in the intestine and represents an important ‘hidden’ source of phosphorus in the diet. As the amount of inorganic phosphorus is often not reported in food labels, accurate estimation of the real phosphorus intake of CKD patients is difficult. Besides the amount of protein in the diet and its absorption, serum phosphorus concentration depends on additional processes that regulate phosphorus metabolism, including renal excretion of phosphorus, and changes in bone turnover.

Chapter 101 

Because there are physiologic adaptations in the stages of earlier CKD that prevent excessive phosphorus retention, the inability to promote phosphorus excretion to avoid phosphorus accumulation and hyperphosphataemia occurs when GFR level decreases below 40 mL/min (Moranne et  al., 2009). The major adaptation is the stimulation of parathyroid hormone (PTH) release, which increases phosphorus excretion by the proximal tubule. In addition, FGF23 activation will reduce serum phosphorus concentration because phosphorus excretion is stimulated, and the production of calcitriol is reduced, suppressing intestinal phosphorus absorption. The metabolism of these molecules is described in more detail in Chapter 119. In patients with moderate/advanced CKD, higher serum phosphorus levels have been associated with increased mortality risk (Eddington et  al., 2010). There is no clinical evidence that high phosphorus intake accelerates CKD progression. In CKD individuals with moderate/advanced CKD in whom the phosphorus is high, it is reasonable to restrict phosphorus intake and to prescribe phosphorus binders. A low-phosphorus diet decreases serum phosphorus and FGF23 levels (Di Iorio et al., 2012), but the combined prescription of this diet and phosphate binders was more effective than either of those approaches alone (Isakova et  al., 2013). The rationale for limiting phosphorus intake in non-dialysed CKD should not be CKD slowing CKD progression but preventing secondary hyperparathyroidism and renal bone disease. It will also reduce the need for calcium-containing phosphate binders which carry a risk of promoting soft tissue calcification. Current recommendations for phosphorus intake in CKD stages 3 and 4 are similar to dialysis patients, that is, to reduce intake to 800–1000 mg/day, in conjunction with use of phosphate binders if appropriate (Ikizler et al., 2013).

Dietary calcium Calcium supplementation of 2–4 g/day reduces PTH levels in advanced CKD (Barsotti et al., 1998). Given that administration of calcium-containing phosphate binders in haemodialysis patients results in increased vascular calcification, concern has been raised regarding the safety of the excess calcium intake in non-dialysed individuals with CKD. There are no RCTs establishing the safe level of calcium intake in pre-dialysis CKD or its role in disease progression. In a calcium balance study, normal individuals and patients with CKD stages 3 and 4 were in slightly negative to neutral calcium balance on an 800-mg calcium diet (Spiegel and Brady, 2012). Normal individuals were in modest positive calcium balance on a 2000-mg diet, whereas CKD patients on the same diet were in marked positive calcium balance. Furthermore, increased calcium intake significantly decreased 1,25-dihydroxy-vitamin D and PTH levels, but it did not alter the serum calcium concentration. On the basis of this study, a diet of 2000 mg/day of calcium in CKD patients might result in a positive calcium balance with the extra calcium deposited in tissues leading to metastatic calcification. Therefore, it is reasonable to restrict total calcium intake to the currently recommended dose of 1500 mg/day (National Kidney Foundation, 2000).

Vitamins Vitamins are essential co-factors in normal metabolism. Besides an insufficient intake, vitamin deficiency can occur because of proteinuria with losses of protein-bound elements or decreased

diet and the progression of chronic kidney disease intestinal absorption of micronutrients, impaired cellular metabolism, circulating inhibitors, or increased losses during dialysis. There is very little information about the minimum requirements or recommended dietary vitamin allowances in patients with CKD. With the exception of vitamin D, no single study has provided evidence of a possible role of vitamin supplementation in retarding CKD progression. Because one of the premises of the science of nutrition is to restore levels of defined nutrients, vitamin supplementation probably does little harm and may provide benefits. Because peripheral neuropathy and hyperoxalaemia can occur with high doses of vitamin B6 (pyridoxine) and vitamin C (ascorbic acid), respectively, megavitamin therapy should be avoided. Plasma vitamin A  (retinol) levels are usually increased in CKD patients and, as excess can cause anaemia, dry skin, pruritus, and even hepatic dysfunction in uraemic patients, vitamin A  supplementation is not recommended. Vitamin E has the potential to suppress oxidative injury of cells and therefore protect against progressive renal insufficiency in CKD. In experimental models of CKD, vitamin E supplementation reduced renal injury in rats with IgA nephropathy or glomerulosclerosis following subtotal nephrectomy or diabetes. Plasma vitamin E concentrations are usually normal in non-dialysed CKD patients. Serum vitamin D concentrations in individuals with CKD not on dialysis decrease progressively with the reduction in GFR. Low vitamin D levels have been associated with inflammation and insulin resistance and predict mortality in non-dialysed CKD patients. A recent study demonstrated an association between vitamin D metabolites and ESRD initiation in patients with moderate/advanced CKD (Kendrick et  al., 2012). Recommendations on vitamin D at these disease stages are complex. Opinion-based guidelines () indicate vitamin D supplementation in CKD patients with serum vitamin D levels < 30 ng/mL. There are still no RCTs demonstrating a benefit of vitamin D supplementation in this patient population.

Diet as a whole Excessive calorie intake Caloric restriction is the only intervention that consistently reduces the primary ageing process across multiple species. In addition, and regardless of protein intake, caloric restriction seems to slow kidney injury in various animal models of glomerular adaptation. It has been speculated that in conjunction with the epidemic of obesity as a risk factor for CKD progression, excessive calorie intake rather than specific micro- or macronutrients may influence this risk (Kramer, 2013). Interestingly, a RCT allocated overweight and obese adults with proteinuria from diabetic and non-diabetic kidney diseases to either a usual diet or a diet that reduced their caloric intake by 500 kcal/day (Morales et al., 2003). The study maintained the same protein intake in both groups during 5 months of treatment. Results showed that the group with caloric restriction lost only approximately 4% of their initial body weight, yet proteinuria decreased by > 30%. In comparison, the group that continued to follow the usual diet showed increased proteinuria (Morales et al., 2003). When considering the risk of CKD progression, total caloric intake along with the caloric source should at least be considered as risk factors.

797

Section 5  

the patient with reduced renal function

Dietary patterns and CKD progression Because of the correlation of micro- and macronutrients within dietary patterns and the consistent dietary behaviour over time for most individuals, it is likely that the overall dietary pattern and cumulative exposure to a particular dietary pattern may be more influential on CKD than excess or deficiency of one specific macroor micronutrient (Kramer, 2013). Much has been said about the problems of the ‘Western diet’, which is characterized by high intake of red meat; animal fat; sweets and desserts; low intake of fresh fruits and vegetables and low-fat dairy products; and contains a high amount of highly processed foods containing refined sugars and saturated fats and trans-fats. In the Nurses’ Health Study, increased urine albumin excretion and higher odds of rapid decline in estimated GFR have been reported among individuals with consistent Western dietary patterns (Lin et al., 2011). In contrast, more healthy dietary patterns are associated with decreased risk for chronic diseases. This is true of the Mediterranean diet, characterized by a high intake of vegetables, legumes, fruits, nuts, cereals, and olive oil; a moderately high intake of fish; a low-to-moderate intake of dairy products; a low intake of saturated fats, meat, and poultry; and a regular but moderate intake of wine during meals. In a population-based study, a greater adherence to a Mediterranean diet associated with lower odds on presenting with CKD (Fig. 101.2) and was a predictor of mortality in those with manifest CKD (Huang et al., 2013). A recent RCT showed as an ancillary outcome, an improvement of renal function after 1-year following a Mediterranean-like dietary pattern in elderly individuals at high risk of coronary heart disease (Diaz-Lopez et al., 2012). Because blood pressure reduction may be one of the most important means of mitigating progression of established CKD, the DASH diet, especially when combined with salt restriction, could theoretically delay kidney disease progression as long as serum phosphorous levels could be controlled despite the high intake of dairy products in this diet. Adherence to a DASH-type dietary pattern was associated with incident microalbuminuria (Chang et al., 2013) and more rapid kidney function decline (Lin et al., 2011). P for trend 0.002

60

P for trend 0.01

High

Medium

Low

High

Low

20

Medium

40 CKD, %

798

0 All individuals

Adequate reporters

CKD, n

131

319

56

57

167

26

CKD, %

53

45

38

50

42

32

Fig. 101.2  Proportions of men with manifest CKD (GFR < 60 mL/min/1.73 m2) across different degrees of adherence to a Mediterranean diet in a population-based study of Swedish men. On the left all participants are shown. On the right only those identified as adequate dietary reporters. Reproduced from Huang et al. (2013).

A  secondary analysis of the original DASH study also showed that a diet high in fruits and vegetables decreases urinary albumin excretion in those with urinary albumin excretion > 7 mg/24 hours (Jacobs et al., 2009). It is possible that physical activity and smoking or occupational exposures that affect kidney disease risk correlate with dietary patterns. Thus, a recent study from the National Health and Nutrition Examination Survey (NHANES) III cohort of about 3000 participants with GFR < 60 mL/min/1.73 m2 showed that adherence to a healthy lifestyle (assessed on the basis of smoking habits, body mass index, physical activity, and dietary quality) was associated with lower all-cause mortality risk (Ricardo et al., 2013).

Summary There is anecdotal and experimental evidence that high protein intake may be harmful in moderate CKD, and cause symptoms in advanced CKD. Evidence for low-protein diets slowing the rate of progression of CKD comes from animal experiments, but in studies in CKD in humans in which blood pressure is well controlled, its effects are weak or absent. Furthermore the diets are different to comply with and carry some risk of malnutrition. Salt intake should be restricted in most patients with reduced GFR or proteinuria, except where there is a salt-wasting tubular problem. Diet otherwise should be generally healthy and mixed, without substantial restriction unless or until blood levels (in particular potassium, phosphate) dictate otherwise.

References Aaron, K. J., Campbell, R. C., Judd, S. E., et al. (2011). Association of dietary sodium and potassium intakes with albuminuria in normal-weight, overweight, and obese participants in the Reasons for Geographic and Racial Differences in Stroke (REGARDS) Study. Am J Clin Nutr, 94, 1071–8. Aparicio, M., Chauveau, P., De Precigout, V., et al. (2000). Nutrition and outcome on renal replacement therapy of patients with chronic renal failure treated by a supplemented very low protein diet. J Am Soc Nephrol, 11, 708–16. Barsotti, G., Cupisti, A., Morelli, E., et al. (1998). Secondary hyperparathyroidism in severe chronic renal failure is corrected by very-low dietary phosphate intake and calcium carbonate supplementation. Nephron, 79, 137–41. Bernhard, J., Beaufrere, B., Laville, M., et al. (2001). Adaptive response to a low-protein diet in predialysis chronic renal failure patients. J Am Soc Nephrol, 12, 1249–54. Bomback, A. S., Katz, R., He, K., et al. (2009). Sugar-sweetened beverage consumption and the progression of chronic kidney disease in the Multi-Ethnic Study of Atherosclerosis (MESA). Am J Clin Nutr, 90, 1172–8. Brymora, A., Flisinski, M., Johnson, R. J., et al. (2012). Low-fructose diet lowers blood pressure and inflammation in patients with chronic kidney disease. Nephrol Dial Transplant, 27, 608–12. Chang, A., Van Horn, L., Jacobs, D. R., Jr., et al. (2013). Lifestyle-related factors, obesity, and incident microalbuminuria: the CARDIA (Coronary Artery Risk Development in Young Adults) Study. Am J Kidney Dis, 62(2), 267–75. Di Iorio, B., Di Micco, L., Torraca, S., et al. (2012). Acute effects of very-low-protein diet on FGF23 levels: a randomized study. Clin J Am Soc Nephrol, 7, 581–7. Diaz-Lopez, A., Bullo, M., Martinez-Gonzalez, M. A., et al. (2012). Effects of Mediterranean diets on kidney function: a report from the PREDIMED Trial. Am J Kidney Dis, 60, 380–9.

Chapter 101 

Eddington, H., Hoefield, R., Sinha, S., et al. (2010). Serum phosphate and mortality in patients with chronic kidney disease. Clin J Am Soc Nephrol, 5, 2251–7. Fouque, D. and Laville, M. (2009). Low protein diets for chronic kidney disease in non diabetic adults. Cochrane Database Syst Rev, 3, CD001892. Fouque, D., Vennegoor, M., ter Wee, P., et al. (2007). EBPG guideline on nutrition. Nephrol Dial Transplant, 22 Suppl 2, ii45–87. Goraya, N., Simoni, J., Jo, C., et al. (2012). Dietary acid reduction with fruits and vegetables or bicarbonate attenuates kidney injury in patients with a moderately reduced glomerular filtration rate due to hypertensive nephropathy. Kidney Int, 81, 86–93. Goraya, N., Simoni, J., Jo, C. H., et al. (2013). A comparison of treating metabolic acidosis in CKD stage 4 hypertensive kidney disease with fruits and vegetables or sodium bicarbonate. Clin J Am Soc Nephrol, 8, 371–81. Huang X, Jimenez-Moleon JJ, Lindholm B, et al. (2013). Mediterranean diet, kidney function, and mortality in men with CKD. Clin J Am Soc Nephrol, 8(9), 1548–55. Ikizler, T. A., Cano, N. J., Franch, H., et al. (2013). Prevention and treatment of protein energy wasting in chronic kidney disease patients: a consensus statement by the International Society of Renal Nutrition and Metabolism. Kidney Int, 84(6), 1096–107. Isakova, T., Barchi-Chung, A., Enfield, G., et al. (2013). Effects of dietary phosphate restriction and phosphate binders on FGF23 levels in CKD. Clin J Am Soc Nephrol, 8, 1009–18. IUPAC-IUB Commission on Biochemical Nomenclature (1978). The nomenclature of lipids (recommendations 1976). IUPAC-IUB Commission on Biochemical Nomenclature. J Lipid Res, 19, 114–28. Jacobs, D. R., Jr., Gross, M. D., Steffen, L., et al. (2009). The effects of dietary patterns on urinary albumin excretion: results of the Dietary Approaches to Stop Hypertension (DASH) Trial. Am J Kidney Dis, 53, 638–46. Karalius, V. P. and Shoham, D. A. (2013). Dietary sugar and artificial sweetener intake and chronic kidney disease: a review. Adv Chronic Kidney Dis, 20, 157–64. Kendrick J, Cheung AK, Kaufman JS, et al. (2012). Associations of plasma 25-hydroxyvitamin D and 1,25-dihydroxyvitamin D concentrations with death and progression to maintenance dialysis in patients with advanced kidney disease. Am J Kidney Dis, 60, 567–75. Kidney Disease Outcomes Quality Initiative (K/DOQI) (2004). K/DOQI clinical practice guidelines on hypertension and antihypertensive agents in chronic kidney disease. Am J Kidney Dis, 43, S1–290. Klahr, S., Levey, A. S., Beck, G. J., et al. (1994). The effects of dietary protein restriction and blood-pressure control on the progression of chronic renal disease. Modification of Diet in Renal Disease Study Group. N Engl J Med, 330, 877–84. Kramer, H. (2013). Dietary patterns, calories, and kidney disease. Adv Chronic Kidney Dis, 20, 135–40. Krishnamurthy, V. M., Wei, G., Baird, B. C., et al. (2012). High dietary fibre intake is associated with decreased inflammation and all-cause mortality in patients with chronic kidney disease. Kidney Int, 81, 300–6. Lambers Heerspink, H. J., Holtkamp, F. A., Parving, H. H., et al. (2012). Moderation of dietary sodium potentiates the renal and cardiovascular protective effects of angiotensin receptor blockers. Kidney Int, 82, 330–7. Levey, A. S., Greene, T., Sarnak, M. J., et al. (2006). Effect of dietary protein restriction on the progression of kidney disease: long-term follow-up of the Modification of Diet in Renal Disease (MDRD) Study. Am J Kidney Dis, 48, 879–88. Levin, A., Hemmelgarn, B., Culleton, B., et al. (2008). Guidelines for the management of chronic kidney disease. CMAJ, 179, 1154–62.

diet and the progression of chronic kidney disease Lin, J. and Curhan, G. C. (2011). Associations of sugar and artificially sweetened soda with albuminuria and kidney function decline in women. Clin J Am Soc Nephrol, 6, 160–6. Lin, J., Fung, T. T., Hu, F. B., et al. (2011). Association of dietary patterns with albuminuria and kidney function decline in older white women: a subgroup analysis from the Nurses’ Health Study. Am J Kidney Dis, 57, 245–54. Lin, J., Hu, F. B., and Curhan, G. C. (2010). Associations of diet with albuminuria and kidney function decline. Clin J Am Soc Nephrol, 5, 836–43. Miller, E. R., 3rd, Juraschek, S. P., Appel, L. J., et al. (2009). The effect of n-3 long-chain polyunsaturated fatty acid supplementation on urine protein excretion and kidney function: meta-analysis of clinical trials. Am J Clin Nutr, 89, 1937–45. Mircescu, G., Garneata, L., Stancu, S. H., et al. ((2007). Effects of a supplemented hypoproteic diet in chronic kidney disease. J Ren Nutr, 17, 179–88. Moe, S. M., Zidehsarai, M. P., Chambers, M. A., et al. (2011). Vegetarian co