Oxford Textbook of Neurological Surgery [1st ed.] 9780192519535

Table of contents :
Cover......Page 1
00-med-9780198746706-FM.pdf......Page 2
Series......Page 3
Series......Page 4
Copyright......Page 5
Contents......Page 10
Symbols and abbreviations......Page 16
List of contributors......Page 22
Section 1 Principles of neurological surgery......Page 28
1. The history of neurosurgery......Page 30
2. Clinical assessment......Page 38
3. Overview of neuroimaging......Page 58
4. The operating theatre environment......Page 72
5. Perioperative care of the neurosurgical patient......Page 84
Section 2 Tumours and skull base-​intrinsic tumour......Page 100
6. Low-​grade glioma......Page 102
7. High-​grade gliomas and molecular biology of neurosurgical oncology......Page 116
8. Intracranial metastasis......Page 134
9. Primary central nervous system lymphoma......Page 144
10. Glioneuronal and other epilepsy-​associated tumours......Page 156
11. Radiotherapy and radiosurgery for brain tumours......Page 168
12. Chemotherapy for brain tumours......Page 176
13. Surgical techniques in the management of intrinsic tumours......Page 190
Section 3 Tumours and skull base—​extra-​axial and
skull lesions......Page 200
14. Meningiomas and haemangiopericytoma
(HPC): ​solitary fibrous tumour (SFT)......Page 202
15. Chordomas and chondrosarcomas of the skull base......Page 216
16. Dermoid and epidermoid cysts......Page 224
17. Esthesioneuroblastoma......Page 232
18. Malignant skull base tumours......Page 240
19. Surgical management of tumours of the orbit......Page 248
20. Skull lesions......Page 256
21. Surgical management of anterolateral skull base lesions......Page 268
Section 4 Tumours and skull base—​CP angle lesions......Page 280
22. Schwannomas......Page 282
23. Glomus tumours......Page 298
24. Surgical management of cerebellopontine angle and petrous lesions......Page 308
Section 5 Tumours and skull base—​sellar and suprasellar tumours......Page 324
25. Pituitary tumours......Page 326
26. Craniopharyngioma and Rathke’s cleft cysts......Page 348
27. Surgical management of sellar and suprasellar tumours......Page 366
Section 6 Tumours and skull base—​posterior fossa......Page 378
28. Medulloblastoma......Page 380
29. Ependymoma......Page 388
30. Haemangioblastoma......Page 394
31. Surgical approaches to posterior fossa tumours......Page 402
Section 7 Tumours and skull base—​intraventricular......Page 410
32. Intraventricular tumours......Page 412
33. Colloid cyst......Page 420
34. Choroid plexus tumours......Page 430
35. Surgical management of intraventricular lesions......Page 438
Section 8 Tumours and skull base—​pineal......Page 452
36. Pineal tumours......Page 454
37. Surgical management of pineal region lesions......Page 466
Section 9 Tumours and skull base—​tumour
syndromes......Page 474
38. Neurophakomatoses......Page 476
39. Uncommon brain lesions......Page 488
Section 10 Neurotrauma and intensive care......Page 500
40. Epidemiology of head injury and outcome after head injury......Page 502
41. Pathophysiology of traumatic brain injury......Page 510
42. Intensive care management of head injury......Page 524
43. Surgical management of head injury......Page 536
44. Complications of head injury......Page 548
45. Concussion and sports-​related head injury......Page 558
Section 11 Vascular neurosurgery......Page 564
46. Normal cerebrovascular physiology and vascular anatomy......Page 566
47. The pathophysiology of aneurysms......Page 586
48. The pathophysiology of subarachnoid haemorrhage......Page 594
49. Management of subarachnoid haemorrhage......Page 602
50. Cerebral arteriovenous malformation and dural arteriovenous fistulae......Page 618
51. Carotid artery disease and cerebral ischaemia......Page 642
52. Extracranial-​intracranial bypass for cerebral ischaemia......Page 654
53. Giant aneurysms and bypass surgery......Page 660
54. Spontaneous intracranial haematoma......Page 670
55. Cavernomata and angiographically occult lesions......Page 678
Section 12 Spinal surgery—​principles......Page 686
56. Surgical principles in spinal surgery......Page 688
57. Spinal stability......Page 696
58. Spinal physiology......Page 706
59. Medical pathologies of the spinal cord......Page 714
Section 13 Spinal surgery......Page 722
60. Cervical spinal disease......Page 724
61. Thoracic spinal disease......Page 738
62. Lumbar spinal disease......Page 746
63. Spinal tumours......Page 760
64. Vascular lesions of the spinal cord......Page 778
65. Spinal cerebrospinal fluid dynamics......Page 788
66. Scoliosis and spinal deformity......Page 796
Section 14 Spinal trauma......Page 806
67. Managing spinal cord trauma......Page 808
68. Cervical spine injuries......Page 816
69. Thoracic and lumbar spine injuries......Page 828
70. Spinal cord injury rehabilitation......Page 840
Section 15 Peripheral nerve surgery......Page 852
71. Electrodiagnostics......Page 854
72. Entrapment syndromes......Page 866
73. Supraclavicular brachial plexus and peripheral nerve injuries......Page 874
74. Peripheral nerve tumours......Page 882
Section 16 Functional neurosurgery......Page 890
75. Principles of deep brain stimulation......Page 892
76. Movement disorders......Page 900
77. Spasticity......Page 912
78. Pain pathophysiology and surgical management......Page 922
79. Cranial nerve vascular compression syndromes......Page 936
80. Neurosurgical interventions for psychiatric disorders......Page 956
Section 17 Epilepsy......Page 960
81. Classification of seizures and epilepsy......Page 962
82. Epilepsy—​diagnosis and assessment......Page 972
83. Surgical management of epilepsy......Page 984
Section 18 Paediatrics......Page 996
84. Developmental disorders of the brain......Page 998
85. Spinal development and spinal dysraphism......Page 1010
86. Craniofaciosynostosis: Syndromic and
non-​syndromic craniofacial anomalies......Page 1020
87. Special considerations in paediatric head and spinal trauma......Page 1026
88. Paediatric brain tumours......Page 1036
89. Paediatric hydrocephalus......Page 1050
90. Paediatric neurovascular disorders......Page 1064
91. Paediatric epilepsy......Page 1074
Section 19 CSF disorders......Page 1082
92. Hydrocephalus and normal CSF dynamics......Page 1084
93. Shunt technology and endoscopic ventricular surgery......Page 1098
94. Normal pressure hydrocephalus......Page 1110
95. Pseudotumour cerebri syndrome......Page 1120
96. Arachnoid cysts......Page 1128
Section 20 Infection......Page 1134
97. Microbiology......Page 1136
98. Cranial infections......Page 1144
99. Spinal infection......Page 1156
Index......Page 1164

Citation preview

Oxford Textbook of

Neurological Surgery

Oxford Textbooks in Surgery SERIES EDITOR Professor Sir Peter J. Morris Nuffield Professor of Surgery Emeritus, and former Chairman of the Department of Surgery and Director of the Oxford Transplant Centre, University of Oxford and Oxford Radcliffe Hospitals, UK

PUBLISHED Oxford Textbook of Trauma and Orthopaedics 2e edited by Christopher Bulstrode, James Wilson-​ MacDonald, Deborah M. Eastwood, John McMaster, Jeremy Fairbank, Parminder J. Singh, Sandeep Bawa, Panagoitis D. Gikas, Tim Bunker, Grey Giddins, Mark Blyth, and David Stanley Oxford Textbook of Fundamentals of Surgery edited by William E.  G. Thomas, Malcolm W. R. Reed, and Michael G. Wyatt Oxford Textbook of Vascular Surgery edited by Matthew M. Thompson, Robert Fitridge, Jon Boyle, Matt Thompson, Karim Brohi, Robert J. Hinchliffe, Nick Cheshire, A. Ross Naylor, Ian Loftus, and Alun H. Davies Oxford Textbook of Urological Surgery edited by Freddie C. Hamdy and Ian Eardley Oxford Textbook of Neurological Surgery edited by Ramez W. Kirollos, Adel Helmy, Simon Thomson, and Peter J.A. Hutchinson

IN PRESS Oxford Textbook of Plastic and Reconstructive Surgery edited by Simon Kay, Daniel Wilks, and David McCombe Oxford Textbook of Paediatric Surgery edited by Paul Johnson, Eleri Cusick, and Spencer Beasley

Oxford Textbook of

Neurological Surgery Ramez W. Kirollos Consultant Neurosurgeon, Addenbrooke’s Hospital, Cambridge, UK

Adel Helmy University Lecturer Neurosurgery, University of Cambridge, UK/Honorary Consultant Neurosurgeon, Addenbrooke’s Hospital, Cambridge, UK

Simon Thomson Consultant Neurosurgeon, Leeds Teaching Hospitals NHS Trust, Leeds, UK

Peter J.A. Hutchinson Professor of Neurosurgery, University of Cambridge, UK

1

3 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 2019 The moral rights of the authors have been asserted First Edition Published in 2019 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: 2019946066 ISBN 978–​0–​19–​874670–​6 Printed in Great Britain by Bell & Bain Ltd., 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.

Series Preface This is a new development in surgical publishing; the first two editions of the Oxford Textbook of Surgery are to be replaced by a series of specialty-​specific textbooks in surgery. This change was precipitated by the ever-​increasing size of a single textbook of surgery which embraced all specialties (the second edition of the Oxford Textbook of Surgery was three volumes), and a decision to adapt the textbooks to meet the needs of the audience; firstly, to suit the requirements of Higher Surgical trainees and, secondly, to make it available online. Thus, we have produced a key book to deal with the fundamentals of surgery, such as Anatomy, Physiology, Biochemistry, Evaluation of Evidence, and so forth. Then there are to be separate volumes covering individual specialties, each appearing as an independent textbook and available on Oxford Medicine Online. It is planned that each textbook in each specialty will be independent although there obviously will be an overlap between different specialties and, of course, the core book on Fundamentals of Surgery will underpin the required scientific knowledge and practice in each of the other specialties.

This ambitious programme will be spread over several years, and the use of the online platform will allow for regular updates of the different textbooks. Each textbook will include the proposed requirements for training and learning as defined by the specialist committees (SACs) of surgery recognized by the four Colleges of Surgery in Great Britain and Ireland, and will continue to be applicable to a global audience. This ambitious programme will be spread over several years, and the use of the online platform will allow for regular updates of the different textbooks. When completed, the Oxford Textbooks in Surgery series will set standards for a long time to come. Professor Sir Peter J. Morris Nuffield Professor of Surgery Emeritus, and former Chairman of the Department of Surgery and Director of the Oxford Transplant Centre, University of Oxford and Oxford Radcliffe Hospitals, UK

Forewords The Oxford Textbook of Neurological Surgery is a new book complementing the Oxford Textbooks in Surgery series. It is arguably the first British-​led comprehensive textbook covering the breadth of neurological surgery since Northfield’s Surgery of the Central Nervous System. Although the book was conceived and led from the United Kingdom, it has an extensive international contribution and will be of value to specialists from all countries. The book bridges the gap between short handbook-​type texts and the large encyclopaedic multivolume tomes. It is suited to junior and senior neurosurgical trainees and consultants, and will also be useful to specialists from other disciplines. In addition, it will be of immense benefit to those studying for the UK Intercollegiate Neurosurgical Examination, the European examination, and equivalent examinations in other continents. It is written in such way as to emphasize the clinical implications of the science. The book is divided into 20 main sections with 99 chapters, 1000 pages in print, and over 1000 figures and tables. The chapters are carefully arranged such that it can be read from cover to cover or used as a reference for specific topics, given the comprehensive index. Each chapter follows a uniform format with abstract and key words, followed by comprehensive coverage of the topic illustrated by clear multicoloured figures. The hard-​copy print edition is complemented by an online version, which provides an alternative format to enabling instant access to specific topics. The ability to update at the proof stage means the book is up-​ to-​date in covering recent advances; for example, the results of clinical trials. I congratulate the authors on bringing together experts from all neurosurgical subspecialties and from all across the globe to deliver a book that is clearly laid out, readable, and well-​illustrated. I am sure it will find its place as a reference book on the bookshelves of neurosurgeons across the globe. Franco Servadei

The past four decades have seen advances on a scale and pace that have made the practice of neurosurgery almost unrecognizable in comparison to the specialty in which I trained. The dawning in the 1970s of microsurgery and cross-​sectional imaging, along with advances in pre-​and perioperative care, sparked technological and technical developments that reinvented concepts of what is possible several times over. Specialization became necessary to take full advantage of the new opportunities to advance patients’ outcomes and neurosurgeons were spurred to accumulate the dedicated, detailed knowledge, skills, and experience that drove the specialty forward and made it increasingly complex and diverse. The editors of the Oxford Textbook of Neurological Surgery have taken on a substantial challenge:  to provide, in a single, comprehensive volume, the knowledge that spans the practice of modern neurosurgery. Alongside this sheer breadth are integrated two critical perspectives: a rigorous understanding of basic science and the subtlety of neurosurgical operative practice. It succeeds in filling a gap in the neurosurgical literature: a single volume that provides a thorough review of neurological surgery for trainee and trained surgeons alike. Its approach is forward-​looking, highlighting the importance of basic and clinical research and evidence-​based medicine. Safe and successful neurosurgery will increasingly require deep anatomical knowledge, clinical judgement, and high levels of technical skill, tied to an ability to synthesize an ever-​growing and complex literature into reasoned practice. UK neurosurgery has a proud history. Modern neurological surgery dawned in Scotland (Glasgow) in 1879 and the Society of British Neurological Surgeons is one of the oldest national societies in the world. But its perspective has been and remains broad, engaging colleagues in the worldwide community of neurosurgeons that transcends shifting national, political boundaries. Reflecting this, although the book is undoubtedly grounded in UK neurosurgery, the editors have successfully drawn in some of the most prolific and experienced surgeons and insightful scientists from around the world. As a result, the Oxford Textbook of Neurological Surgery provides an invaluable companion to neurosurgeons, from their earliest years of training into subspecialty experience and consultant practice. It will become the definitive single volume textbook for those who treat surgical disorders of the nervous system in the United Kingdom and around the world. Sir Graham Teasdale

Contents

Symbols and abbrevations  xv List of contributors  xxi

11. Radiotherapy and radiosurgery for brain tumours  141 Susan Short

SECTION 1 Principles of neurological surgery 1. The history of neurosurgery  3 Eleni Maratos and Henry Marsh

12. Chemotherapy for brain tumours  149 Nicholas F. Brown, Daniel Krell, and Paul Mulholland

13. Surgical techniques in the management of intrinsic tumours  163 Shawn Hervey-​Jumper and Mitchel Berger

2. Clinical assessment  11 Peter Bodkin and Elizabeth Visser

3. Overview of neuroimaging  31 Tomasz Matys, Daniel. J. Scoffings, and Tilak Das

4. The operating theatre environment  45 Neil Kitchen and Jonathan Shapey

5. Perioperative care of the neurosurgical patient  57 Karol P. Budohoski, Alessandro Scudellari, Sylvia Karcheva, and Derek Duane

SECTION 3 Tumours and skull base—​extra-​axial and skull lesions 14. Meningiomas and haemangiopericytoma (HPC): s​ olitary fibrous tumour (SFT)  175 Harjus S. Birk, Seunggu J. Han, Ramez W. Kirollos, Thomas Santarius, and Michael W. McDermott

15. Chordomas and chondrosarcomas of the skull base  189 Rami O. Almefty and Ossama Al-​Mefty

SECTION 2 Tumours and skull base-​intrinsic tumour 6. Low-​grade glioma  75 Thomas Santarius, Lorenzo Bello, and Hugues Duffau

7. High-​grade gliomas and molecular biology of neurosurgical oncology  89 Stephen J. Price, Harry Bulstrode, and Richard Mair

8. Intracranial metastasis  107 Andrew Brodbelt and Rashed Zakaria

9. Primary central nervous system lymphoma  117 Boon Leong Quah, Thangaraj Munusamy, and Colin Watts

10. Glioneuronal and other epilepsy-​associated tumours  129 Matthias Simon and Alexander Grote

16. Dermoid and epidermoid cysts  197 Andrew McEvoy

17. Esthesioneuroblastoma  205 Georgios Klironomos, Lior Gonen, and Fred Gentili

18. Malignant skull base tumours  213 Benedict Panizza and Adel Helmy

19. Surgical management of tumours of the orbit  221 Joseph D. Chabot, S. Tonya Stefko, and Paul Gardner

20. Skull lesions  229 Giorgio Gioffre and Ivan Tmofeev

21. Surgical management of anterolateral skull base lesions  241 Michael D. Cusimano and Michael P. Meier

x

Contents

SECTION 4 Tumours and skull base—​CP angle lesions

SECTION 7 Tumours and skull base—​intraventricular

22. Schwannomas  255

32. Intraventricular tumours  385

Tiit Mathiesen, Petter Förander, and David Pettersson

23. Glomus tumours  271 Omar Pathmanaban and Andrew King

24. Surgical management of cerebellopontine angle and petrous lesions  281 Nicholas Hall, Yuval Sufaro, and Andrew H. Kaye

Paul Grundy and Vasileios Apostolopoulos

33. Colloid cyst  393 Asim Sheikh and Paul Chumas

34. Choroid plexus tumours  403 Jonathan Roth, Rina Dvir, and Shlomi Constantini

35. Surgical management of intraventricular lesions  411 Eduardo C. Ribas, Guilherme C. Ribas, and Ramez W. Kirollos

SECTION 5 Tumours and skull base—​sellar and suprasellar tumours 25. Pituitary tumours  299 Kanna Gnanalingham, Zsolt Zador, Tara Kearney, Federico Roncaroli, and H. Rao Gattamaneni

26. Craniopharyngioma and Rathke’s cleft cysts  321 Rudolf Fahlbusch, V. Gerganov, and H. Metwali

27. Surgical management of sellar and suprasellar tumours  339

SECTION 8 Tumours and skull base—​pineal 36. Pineal tumours  427 Mueez Waqar, Samantha Mills, Conor L. Mallucci, and Michael D. Jenkinson

37. Surgical management of pineal region lesions  439 Christoph M. Woernle, René L. Bernays, and Nicolas de Tribolet

Jayson A. Neil and William T. Couldwell

SECTION 6 Tumours and skull base—​posterior fossa

SECTION 9 Tumours and skull base—​tumour syndromes

28. Medulloblastoma  353

38. Neurophakomatoses  449

James Rutka

29. Ependymoma  361 Christopher Chandler

30. Haemangioblastoma  367 Ammar Natalwala and Donald MacArthur

31. Surgical approaches to posterior fossa tumours  375 Jacques J. Morcos, Osaama H. Khan, and Ashish H. Shah

Fay Greenway, Frances Elmslie, and Timothy Jones

39. Uncommon brain lesions  461 Yizhou Wan, Hani J. Marcus, and Thomas Santarius

Contents

SECTION 10 Neurotrauma and intensive care  40. Epidemiology of head injury and outcome after head injury  475 Nabeel Alshafai and Andrew Maas

41. Pathophysiology of traumatic brain injury  483 John K. Yue, Hansen Deng, Ethan A. Winkler, John F. Burke, Catherine G. Suen, and Geoffrey T. Manley

42. Intensive care management of head injury  497 Matthew A. Kirkman and Martin Smith

43. Surgical management of head injury  509 Hadie Adams, Angelos G. Kolias, Adel Helmy, Peter J.A. Hutchinson, and Randall M. Chesnut

44. Complications of head injury  521 Fardad T. Afshari, Antonio Belli, and Peter C. Whitfield

45. Concussion and sports-​related head injury  531 Mark Wilson

SECTION 11 Vascular neurosurgery 46. Normal cerebrovascular physiology and vascular anatomy  539 Diederik O. Bulters and Andrew Durnford

47. The pathophysiology of aneurysms  559 Federico Cagnazzo, Giuseppe Lanzino, and Neal F. Kassell

48. The pathophysiology of subarachnoid haemorrhage  567 Jason McMillen

49. Management of subarachnoid haemorrhage  575 Roberto Rodriguez Rubio, Brian P. Walcott, and Michael T. Lawton

50. Cerebral arteriovenous malformation and dural arteriovenous fistulae  591 Michael Morgan

51. Carotid artery disease and cerebral ischaemia  615 Kieron Sweeney, A. O’Hare, and Mohsen Javadpour

52. Extracranial-​intracranial bypass for cerebral ischaemia  627 Mathew R. Guilfoyle and Peter J. Kirkpatrick

53. Giant aneurysms and bypass surgery  633 Mario Teo, Omar Choudhri, and Michael Lawton

54. Spontaneous intracranial haematoma  643 Berk Orakcioglu and Andreas W. Unterberg

55. Cavernomata and angiographically occult lesions  651 Janneke van Beijnum and Hiren Patel

SECTION 12 Spinal surgery—​principles  56. Surgical principles in spinal surgery  661 Simon Thomson, Chris Derham, and Senthil Selvanathan

57. Spinal stability  669 Peter R. Loughenbury and Richard M. Hall

58. Spinal physiology  679 Sadaquate Khan, Kevin Tsang, and Lamia Nayeb

59. Medical pathologies of the spinal cord  687 Chris McGuigan, Karen O’Connell, Eavan McGovern, and Iain McGurgan

SECTION 13 Spinal surgery 60. Cervical spinal disease  697 Navin Furtado, Georgios Tsermoulas, and Adikarige Haritha Dulanka Silva

61. Thoracic spinal disease  711 Kieron Sweeney, Catherine Moran, and Ciaran Bolger

62. Lumbar spinal disease  719 Christopher G. Kellett and Matthew J. Crocker

63. Spinal tumours  733 John Brecknell and Boon Leong Quah

64. Vascular lesions of the spinal cord  751 Daniel Walsh

65. Spinal cerebrospinal fluid dynamics  761 Graham Flint

66. Scoliosis and spinal deformity  769 Nigel Gummerson

xi

xii

Contents

SECTION 14 Spinal trauma

80. Neurosurgical interventions for psychiatric disorders  929 Chris Bervoets, Bart Nuttin, and Loes Gabriëls

67. Managing spinal cord trauma  781 Saksith Smithason, Bryan S. Lee, and Edward C. Benzel

68. Cervical spine injuries  789 Calan Mathieson, Chris Barrett, and Likhith Alakandy

69. Thoracic and lumbar spine injuries  801 Bedansh Roy Chaudhary and Shiong Wen Low

70. Spinal cord injury rehabilitation  813 Fahim Anwar, Wail Ahmed, Tamara Tajsic, Damiano G. Barone, and Harry Mee

SECTION 17 Epilepsy 81. Classification of seizures and epilepsy  935 Andrew McEvoy, Tim Wehner, and Victoria Wykes

82. Epilepsy—​diagnosis and assessment  945 Richard Selway

83. Surgical management of epilepsy  957 Johannes Schramm

SECTION 15 Peripheral nerve surgery 71. Electrodiagnostics  827 Alan Forster and Robert Morris

72. Entrapment syndromes  839 Grainne Bourke and Mobin Syed

73. Supraclavicular brachial plexus and peripheral nerve injuries  847 Jonathan Perera and Marco Sinisi

74. Peripheral nerve tumours  855 Rikin Trivedi and Vincent Nga

SECTION 18 Paediatrics 84. Developmental disorders of the brain  971 Colin Ferrie, Daniel Warren, and Atul Tyagi

85. Spinal development and spinal dysraphism  983 Dominic Thompson

86. Craniofaciosynostosis: Syndromic and non-​syndromic craniofacial anomalies  993 Federico Di Rocco, Pierre-​Aurelien Beuriat, and Eric Arnaud

87. Special considerations in paediatric head and spinal trauma  999 Andrew Kay, Desiderio Rodrigues, Melanie Sharp, and Guirish Solanki

SECTION 16 Functional neurosurgery

88. Paediatric brain tumours  1009

75. Principles of deep brain stimulation  865

89. Paediatric hydrocephalus  1023

Erlick A. C. Pereira, Alexander L. Green, and Tipu Z. Aziz

76. Movement disorders  873 Keyoumars Ashkan and Ismail Ughratdar

77. Spasticity  885 John Goodden, Catherine Hernon, and Brian Scott

78. Pain pathophysiology and surgical management  895 Richard Mannion and Rokas Tamosauskas

79. Cranial nerve vascular compression syndromes  909 Marc Sindou and George Georgoulis

Jonathan Roth and Shlomi Constantini Matt Bailey, Chris Parks, and Conor L. Malucci

90. Paediatric neurovascular disorders  1037 Helen G. McCullagh,Tufail Patankar, Tony Goddard, and Atul Tyagi

91. Paediatric epilepsy  1047 Sophia Varadkar and Martin Tisdall

Contents

SECTION 19 CSF disorders

SECTION 20 Infection

92. Hydrocephalus and normal CSF dynamics  1057

97. Microbiology  1109

Alexander Gamble and Harold Rekate

93. Shunt technology and endoscopic ventricular surgery  1071 Ian K. Pople and William Singleton

94. Normal pressure hydrocephalus  1083

Walter A. Hall

98. Cranial infections  1117 Thangaraj Munusamy, Boon Hoe Tan, and Eugene Yang

99. Spinal infection  1129 Nicholas Haden and Edward White

Nicole C. Keong

95. Pseudotumour cerebri syndrome  1093 John D. Pickard and Nicholas Higgins

96. Arachnoid cysts  1101 Ruichong Ma and Stana Bojanic

Index  1137

xiii

Symbols and abbreviations Ω omega α alpha β beta δ gamma µg micrograms AAA AAICH

asleep-​awake-​asleep anticoagulation-​associated intracerebral haemorrhage AANS acute and chronic settings ABC aneurysmal bone cyst ABI auditory brainstem implants ABR auditory brainstem responses AC anterior and posterior commissure (also arachnoid cyst) ACA anterior cerebral artery ACD anterior cervical discectomy ACPP atypical choroid plexus papilloma ADC apparent diffusion coefficient ADI atlantodental interval ADL activities of daily living ADNFLE autosomal-​dominant nocturnal frontal lobe epilepsy ADPKD autosomal dominant polycystic kidney disease AED antiepileptic drug AF atrial fibrillation AGNIR Advisory Group on Non-​ionising Radiation AICA anterior inferior cerebellar artery AIDS acquired immunodeficiency syndrome AIP aryl hydrocarbon receptor-​interacting protein AIS Abbreviation Injury Score (also ASIA Impairment Scale) ALI acute lung injury ALIF anterior lumbar interbody fusion ALL acute lymphatic leukaemia ALL anterior longitudinal ligament AP anteroposterior ASA American Society of Anesthesiologists ASA anterior spinal artery ASIA American Spinal Injuries Association ASPECTS Alberta Stroke Programme Early CT Score AT anaplastic transformation ATA anterior temporal artery ATIII antithrombin III ATLS advance trauma life support ATRT atypical teratoid rhabdoid tumour AVF arteriovenous fistula

AVM AZ BAEP BBB BECTS BIS BMD BMI BOLD BP BRAT BTF CA CAA CAD CAMS CAPECTH CAR CAS CBF CBV CCA CCM CDC CDK CFAM CHF CHLA CISS CM CM CMAP CMRO2 CN CNS COPD COSS CPA CPC CPH CPP CPP CPT

arteriovenous malformation Annulus of Zinn brainstem auditory evoked potentials blood–​brain barrier benign epilepsy with centrotemporal spikes bispectral index bone mineral density body mass index blood oxygen level dependent blood pressure Barrow Ruptured Aneurysm Trial Brain Trauma Foundation cerebral autoregulation cerebral amyloid angiopathy coronary artery disease cerebrofacial metameric arteriovenous syndromes craniectomy-​associated progressive extra-​axial collections with treated hydrocephalus cerebral autoregulation carotid angioplasty and stenting cerebral blood flow cerebral blood volume common carotid artery cerebral cavernous malformations Centers for Disease Control cyclin dependent kinases cerebral function analysing monitor congestive heart failure Children’s Hospital Los Angeles Constructive Interference in Steady State cavernous malformations central myelin compound motor action potential cerebral metabolic rate of oxygen consumption cranial nerve central nervous system (also Congress of Neurological Surgeons) chronic obstructive pulmonary disease Carotid Occlusion Surgery Study cerebellopontine angle choroid plexus carcinoma choroid plexus hyperplasia cerebral perfusion pressure (also choroid plexus papilloma) choroid plexus papilloma choroid plexus tumour

xvi

Symbols and abbreviations

CRF corticotrophin factor CRH corticotrophin-​releasing hormone CRP C-​reactive protein CRPS complex regional pain syndrome CRW Cosman-​Roberts-​Wells CSF cerebrospinal fluid CSW cerebral salt wasting CSWS cerebral salt wasting syndrome CT computed tomographic CTA computed tomography angiography CTP cerebellar tonsillar prolapse CTS carpal tunnel syndrome CTV clinical target volume CUSA cavitron ultrasonic surgical aspirator DBS deep brain stimulation DEBS direct electrical brain stimulation DES drug-​eluting  stents DESD detrusor external sphincter dyssynergia DEXA dual-​energy X-​ray absorptiometry DI diabetes insipidus DIND delayed ischaemic neurological deficits DISH diffuse idiopathic skeletal hyperostosis DLGG diffuse low-​grade gliomas DNP dynamic nuclear polarization DNT dysembryoplastic neuroepithelial tumours DPG diffuse pontine glioma DRG dorsal root ganglion DRIFT Drainage, Irrigation and Fibrinolytic Therapy DSA digital subtraction angiography DSB double-​strand  break DTI diffusion tensor imaging DVA deep venous anomaly DVA developmental venous anomaly DVT deep vein thrombosis DWI diffusion-​weighted imaging DWMH deep white matter hyperintensities DXA dual X-​ray absorbitometery ECA external carotid artery ECG electrocardiogram ECoG electrocorticography EDF elongation derotation flexion EDL extensor digitorum longus EEA endoscopic endonasal approach EEG electroencephalography EGFR epidermal growth factor receptor EIEE early infantile onset epileptic encephalopathies ELISA enzyme-​linked immunosorbent assay EMA epithelial membrane antigen EMG electromyography ENT ear, nose, and throat EOIS early onset idiopathic scoliosis EOR extent of resection EORTC European Organization for Research and Treatment of Cancer EP evoked potential EPC epilepsia partialis continua ES ethmoid sinus ERG Electroretinogram

ES Ewing’s sarcoma ESO European Stroke Organisation ESR erythrocyte sedimentation rate ET essential tremor ETT endotracheal tube ETV endoscopic third ventriculostomy EVD external ventricular drain EZ epileptogenic zone FA fractional anisotropy FCD focal cortical dysplasias FCU flexor carpi ulnaris FD fibrous dysplasia FDA Food and Drug Administration FEF frontal eye field FES functional electric stimulation FFP fresh frozen plasma FGN French Glioma Network FIESTA fast imaging in steady state FIPA familial isolated pituitary adenoma FLAIR fluid attenuated inversion recovery FLE frontal lobe epilepsy FM Foramen of Monro FSH follicle-​stimulating hormone FSU functional spinal unit FTT failure to thrive FV flow velocities FZS fronto-​zygomatic  suture GABA gamma aminobutyric acid GAF Global Assessment of Function GCS Glasgow Coma Scale GCT germ cell tumours (also granular cell tumour) GFAP glial fibrillary acid protein GFR glomerular filtration rate GH growth hormone GHIH growth hormone-​inhibiting hormone GHRH growth hormone release hormone GI gastrointestinal GMFM Gross Motor Function Measure GSPN greater superficial petrosal nerve GTCS generalized tonic-​clonic seizures GTR gross total resection GTV gross tumour volume GW gliadel wafers H&E haematoxylin and eosin HBO hyperbaric oxygen HDDST high-​dose dexamethasone suppression tests HGG high-​grade  glioma HHT hereditary haemorrhagic telangiectasia HIF hypoxia inducible factor HIFU high-​intensity focused ultrasound HIV human immunodeficiency virus HLA human leukocyte antigen HMSN hereditary motor, and sensory neuropathy HO heterotrophic ossification HPA hypothalamic-​pituitary  axis HPC haemangiopericytoma HR homologous recombination

Symbols and abbreviations

HRQOL HS HS HSV HU IA IAC IAM IBE ICA ICA ICBP ICE ICH ICP ICU IFOF IJV ILAE IO IOF IOM IPG IPG IPSS IR IR ISAT ISCoS ISNCSCI

health-​related quality of life hippocampal sclerosis hypertonic saline herpes simplex virus Hounsfield units intra-​arterial internal auditory canal internal auditory meatus International Bureau for Epilepsy inferior cerebellar artery (also internal carotid artery) internal carotid arteries infraclavicular brachial plexus ifosfamide, carboplatin, and etoposide intracerebral haemorrhage intracranial pressure intensive care unit inferior fronto-​occipital fasciculus internal jugular vein International League Against Epilepsy inferior oblique inferior orbital fissure Intraoperative monitoring implantable pulse generator implanted battery-​operated pulse generators inferior petrosal vein sampling inferior rectus iterative reconstruction International Subarachnoid Aneurysm Trial International Spinal Cord Society International Standards for Neurological Classification of Spinal Cord Injury ITB intrathecal baclofen IVC inferior vena cava IVP intraventricular pressure JET Japanese EC-​IC Bypass Trial KOLT Kendrick object learning test kPa kilopascal KPS Karnofsky Performance Status LCH Langerhans cell histiocytosis LDD L’hermitte-​Duclos disease LDDST low-​dose dexamethasone suppression test LDL low-​density lipoprotein LDM limited dorsal myeloschisis LFS Li-​Fraumeni syndrome LG lacrimal gland LGG low-​grade  glioma LH luteinizing hormone LINAC linear accelerator LMN lower motor neuron LMWH low molecular weight heparin LOC level of consciousness LOH loss of heterozygosity LOIS late onset idiopathic scoliosis LOR line of response LOVA longstanding overt ventriculomegaly in adults LP levator palpebrae LP lumbar puncture LR lateral rectus

LR Lindegaard ratio LSO lumbar sacral orthosis LSR lateral spread responses MAC minimal alveolar concentration MAP mean arterial pressure MC Meckel cave MCA middle cerebral artery MCD malformation of cortical development MDT multidisciplinary team MEG magnetoencephalography MEP minimally endoscopic procedures MEP motor evoked potential MHC major histocompatibility complex MI myocardial infarction MIP maximum intensity projection MIT minimally invasive techniques MLF medial longitudinal fasciculus MM multiple myeloma MMD moya moya disease MMS moya moya syndrome MMSE mini-​mental state examination MND motor neurone disease MOG myelin oligodendrocyte glycoprotein MPBT malignant paediatric brain tumours MPNST malignant peripheral nerve sheath tumours MR magnetic resonance MR medial rectus MRA magnetic resonance angiography MRA MR-​angio MRC Medical Research Council MRI magnetic resonance imaging MRM magnetic resonance myelography MRN magnetic resonance neurography MRS magnetic resonance spectroscopy MRSA methicillin-​resistant Staphylococcus aureus MRV magnetic resonance venography MS multiple sclerosis MSH melanocyte-​stimulating hormone MTG middle temporal gyrus MTLE mesial temporal lobe epilepsy MTT mean transit time MUAP motor unit action potential MVA motor vehicle accidents MVC motor vehicle collisions MVD microsurgical vascular decompression Na+ sodium NAA N-​acetyl aspartate NASCIS National Acute Spinal Cord Injury Studies NCS nerve conduction studies NeuN neuronal nuclei NFPA non-​functioning pituitary adenomas NHEJ non-​homologous end-​joining NHL non-​Hodgkin’s lymphoma NICE National Institute of Clinical Excellence NIRS near infrared spectroscopy NLI neurological level of injury NMDA N-​methyl-​D-​aspartate NMS non-​motor symptoms

xvii

xviii

Symbols and abbreviations

NOAC NOS NPH NPSA NPUAP NSAID OC OCR OCT ODI OEF OLE OLF ON OPG OPLL OS OSA PA PAR PBI PBT PCA PCC PCNSL PCR PCV PD PDGFR PE PE PEDI PEEK PEEP PEG PET PFO PFS PI PICA PIH PIOL PLL PML PNET PONV PPTID PRH PSA PSO PT PTA PTH PTPR PTS PTSD PTT

new oral anticoagulants not otherwise specified normal pressure hydrocephalus National Patient Safety Agency National Pressure Ulcer Advisory Panel non-​steroidal anti-​inflammatory  drug optic canal optico-​carotid recesses optical coherence tomography Oswestry Disability Index oxygen extraction fraction occipital lobe epilepsy ossification of the ligamentum flavum optic nerve optic pathway glioma ossification of the posterior longitudinal ligament overall survival obstructive sleep apnoea pilocytic astrocytoma protease-​activated receptors penetrating brain injury paediatric brain tumour posterior cerebral artery prothrombin complex concentrate primary central nervous system lymphoma polymerase chain reaction procarbazine, CCNU, vincristine Parkinson’s disease (also proton density) platelet-​derived growth factor receptor preoperative embolization pulmonary embolism Pediatric Evaluation of Disability Inventory poly-​ether-​ether-​ketone positive end-​expiratory pressure percutaneous endoscopic gastrostomy positron emission tomography patent foramen ovale progression-​free survival pelvic incidence posterior inferior cerebellar artery prolactin-​inhibiting hormone primary intraocular lymphoma posterior longitudinal ligament progressive multifocal leukoencephalopathy primitive neuroectodermal tumour postoperative nausea and vomiting pineal parenchymal tumours of intermediate differentiation prolactin-​releasing hormone posterior spinal arteries pedicle subtraction osteotomy pelvic tilt (also prothrombin time) pure-​tone audiogram post-​traumatic hydrocephalus papillary tumour of the pineal region post-​traumatic seizures post-​traumatic stress disorder partial thromboplastin time

PTV planning target volume PVA poly-​vinyl-​alcohol QoL quality of life QOLIBRI Quality of Life after Brain Injury RA rheumatoid arthritis RANKL receptor activator of NF-​KB ligand RANO Response Assessment in Neuro-​Oncology RAPD relative afferent pupil defect RCC renal cell carcinoma RCN Rare Cancer Network RCT randomized clinical trials REM rapid eye movement REZ root entry zone RF rheumatoid factor RNFL retinal nerve fibre layer RT radiation therapy RT resistance in the tube RVAD rib-​vertebral-​angle difference SAH subarachnoid haemorrhage SARS sacral anterior nerve root stimulator SBP supraclavicular brachial plexus SCA superior cerebellar artery SCA superior cerebellar artery SCAVM spinal cord arteriovenous malformations SCI spinal cord injury SCID severe-​combined immunodeficiency disease SCM strap muscles SCO spindle cell oncocytoma SCPP spinal cord perfusion pressure SDAVF spinal dural arteriovenous fistula SDR selective dorsal rhizotomy SDS speech discrimination score SEA spinal epidural abscess SEEG stereoencephalography SEER Surveillance Epidemiology and End Results SEGA subependymal giant cell astrocytomas SEP Somatosensory evoked potential SESH spontaneous epidural spinal haemorrhage SFT solitary fibrous tumour SGCT subependymal giant cell tumour SIADH syndrome of inappropriate antidiuretic hormone SIVMS Scottish Intracranial Vascular Malformation  Study SLE systemic lupus erythematous SLF superior longitudinal fasciculus SMA supplementary motor area SNAP sensory nerve action potential SNO Society for NeuroOncology SNUC sinonasal undifferentiated carcinoma SO superior oblique SOF superior orbital fissure SOM spheno-​orbital meningiomas SOV superior ophthalmic vein SPECT single photon emission computed tomography SPES single pulse electrical stimulation SPN selective peripheral neurotomy SPV superior petrosal venous SR superior rectus

Symbols and abbreviations

SSFP SSI SSMA SSS SST STA STASCIS STG STIR STR SUDEP SUNCT SVA SVM SWI TBI TCD TCGA TF TFPI TIVA TLE TLSO TMG TMS TN TOF TORCH TREZ TRH

steady state free precession surgical site infection supplementary sensorimotor area superior sagittal sinus Short Synacthen Test superficial temporal artery (also superior thyroid artery) Surgical Treatment of Acute Spinal Cord Injury Study superior temporal gyrus short tau inversion recovery subtotal resection sudden unexplained death in epilepsy short-​lived, unilateral neuralgic headache with conjunctival injection and tearing sagittal vertical axis spinal vascular malformation susceptibility weighted imaging traumatic brain injury transcranial Doppler The Cancer Genome Atlas tissue factor tissue factor pathway inhibitor total intravenous anaesthesia temporal lobe epilepsy thoracolumbar sacral orthosis transmural pressure gradient tumefactive multiple sclerosis trigeminal neuralgia time of flight Toxoplasmosis, Other (syphilis, varicella-​zoster, parvovirus B19), Rubella, Cytomegalovirus (CMV), and Herpes infections trigeminal root entry zone thyroid-​releasing hormone (also thyrotropin-​ releasing hormone)

TSC tuberous sclerosis complex TSH thyroid-​stimulating hormone TT thrombin time TTM targeted temperature management TTP time to peak TZ transitional zone UH unfractionated heparin UMN upper motor neuron UMNL upper motor neurone lesions VA vertebral artery VAD ventricular access device VAE venous air embolism VASO vascular space occupancy VBA vertebrobasilar artery VDE velocity of diametric expansion VEGF vascular endothelial growth factor VEP visual evoked potential VGAM vein of Galen malformations VGPN vago-​glossopharyngeal neuralgia VHL Von Hippel-​Lindau VMAT volumetric modulated arc therapy VNS vagus nerve stimulation VP vancomycin powder VP ventriculoperitoneal VRE vancomycin-​resistant enterococci VS vestibular schwannoma (also ventral striatum) VTE venous thromboembolism vWF von Willebrand factor WBC white blood cell WBRT whole-​brain radiation therapy WFNS World Federation of Neurosurgical Societies WFSBP World Federation of Societies of Biological Psychiatry WHO World Health Organization YBOCS Yale-​Brown Obsessive-​Compulsive  Scale

xix

Contributors Hadie Adams  Cambridge University Hospitals

Damiano G. Barone  Department of Neurosurgery,

Grainne Bourke  Leeds General Infirmary,

Fardad T. Afshari  Cambridge University Hospitals

Chris Barrett  Consultant Neurosurgeon, Southern

John Brecknell  Queen’s Hospital, Romford,

Wail Ahmed  Consultant, Spinal Cord Injuries

Janneke van Beijnum  Salford Royal NHS

Andrew Brodbelt  The Walton Centre NHS

Likhith Alakandy  Consultant Neurosurgeon,

Antonio Belli  Reader in Neurotrauma, School of

NHS Foundation Trust, Cambridge, UK 43: Surgical management of head injury NHS Foundation Trust, Cambridge, UK 44: Complications of head injury

in Stoke Mandeville Hospital, Aylesbury, Buckinghamshire, UK 70: Spinal cord injury rehabilitation

Southern General Hospital, Glasgow, UK 68: Cervical spine injuries

Kieren Allinson  Department of Pathology,

Addenbrooke’s Hospital, Cambridge, UK

Ossama Al-​Mefty  UAMS Medical Center,

Little Rock, AK, USA; and Department of Neurosurgery, Brigham and Women’s Hospital, Boston MA, USA 15: Chordomas and chondrosarcomas of the skull base

Rami O. Almefty  Department of Neurosurgery,

Temple University, Philadelphia, PA, USA 15: Chordomas and chondrosarcomas of the skull base

Nabeel Alshafai  Assistant Professor,

Neurosurgery, Antwerp, Belgium 40: Epidemiology of head injury and outcome after head injury

Fahim Anwar  Consultant in Rehabilitation

Medicine, Department of Rehabilitation Medicine, Cambridge University Hospital NHS, Foundation Trust, Cambridge, UK 70: Spinal cord injury rehabilitation

Vasileios Apostolopoulos  University Hospital

Southampton NHS Foundation Trust, Southampton, UK 32: Intraventricular tumours

Eric Arnaud  Department of Pediatric

Neurosurgery, Necker-​Enfants Malades Hospital, Paris, France 86: Spinal development and spinal dysraphism

Keyoumars Ashkan  Kings College Hospital,

London, UK 76: Movement disorders

Tipu Aziz  The Nuffield Department of Clinical

Neurosciences, University of Oxford, Oxford, UK 75: Principles of deep brain stimulation

Matt Bailey  Consultant Neurosurgeon, Salford

Royal NHS Foundation Trust, Salford, UK 89: Paediatric hydrocephalus

Addenbrooke’s Hospital, Cambridge, UK 70: Spinal cord injury rehabilitation General Hospital, Glasgow, UK 68: Cervical spine injuries

Foundation Trust, Salford, UK 55: Cavernoma and angiographically occult lesions

Clinical and Experimental Medicine, College of Medical and Dental Sciences, University of Birmingham, Birmingham, UK 44: Complications of head injury

Lorenzo Bello  Associate Professor Neurosurgery,

University of Milan, Italy 6: Low-​grade  glioma

Edward Benzel  Chairman and Spinal Surgeon,

Cleveland Clinic, OH, USA 67: Managing spinal cord trauma

Mitchel Berger  Department of Neurological

Surgery, University of California, San Francisco, CA, USA 13: Surgical techniques in the management of intrinsic tumours

René Bernays  Department of Neurosurgery,

Klinik Hirslanden, Zurich, Switzerland 37: Surgical management of pineal region lesions

Chris Bervoets  Psychiatrist, UZ Leuven, Leuven,

Belgium 80: Neurosurgical interventions for psychiatric disorders

Pierre-​Aurelien Beuriat  Hopital Chu de Lyon,

Lyon, France 86: Spinal development and spinal dysraphism

Harjus Birk  Department of Neurological

Surgery, University of California, San Francisco, CA, USA 14: Meningiomas and haemangiopericytoma (HPC)—​solitary fibrous tumour (SFT)

Peter Bodkin  Aberdeen Royal Infirmary,

Scotland, UK 2: Clinical assessment

Stana Bojanic  Department of Neurosurgery,

Oxford University Hospitals NHS Foundation Trust, Oxford, UK 96: Arachnoid cysts

Ciaran Bolger  College of Surgeons, Dublin,

Ireland 61: Thoracic spinal disease

Clarendon Way, Leeds, UK 72: Entrapment syndromes London, UK 63: Spinal tumours

Foundation Trust, Liverpool, UK 8: Intracranial metastasis

Nicholas Brown  University College Hospitals,

London, UK 12: Chemotherapy for brain tumours

Karol P. Budohoski  Neurosurgical Specialist

Registrar, Cambridge University Hospital NHS Trust, Department of Neurosurgery, Addenbrooke’s Hospital, Cambridge University Hospitals, Cambridge, UK 5: Perioperative care of the neurosurgical patient

Harry Bulstrode  Clinical Lecturer in

Neurosurgery, Neurosurgery Division, Department of Clinical Neurosciences, Cambridge Biomedical Campus, Cambridge, UK 7: High-​grade gliomas and molecular biology of neurosurgical oncology

Diederik O. Bulters  Department of Neurosurgery,

Wessex Neurological Centre, Southampton General Hospital, Southampton, UK 46: Normal cerebrovascular physiology and vascular anatomy

John F. Burke  Resident Physician, Department of

Neurological Surgery, University of California, San Francisco, CA, USA 41: Pathophysiology of traumatic brain injury

Federico Cagnazzo  Università di Pisa,

Pisa, Italy 47: The pathophysiology of aneurysms

Joseph D. Chabot  Department of Neurological

Surgery, University of Pittsburgh, PA, USA 19: Surgical management of tumours of the orbit

Christopher Chandler  Consultant Adult and

Paediatric Neurosurgeon, King’s College Hospital, London, UK 29: Ependymoma

Bedansh Roy Chaudhary Cambridge

University Hospitals NHS Foundation Trust, Cambridge, UK 69: Thoracic and lumbar spine injuries

xxii

Contributors

Randall M. Chesnut  Professor, Neurological

Surgery; Professor, Orthopedics and Adjunct Professor, Global Health, Harborview Medical Center, Department of Neurosurgery, Seattle, WA, USA 43: Surgical management of head injury

Omar Choudhri  Assistant Professor,

Department of Neurosurgery, University of Pennsylvania, PA, USA 53: Giant aneurysms and bypass surgery

Paul Chumas  Consultant in Neurosurgery, Leeds

Teaching Hospitals, Leeds, UK 33: Colloid cyst

Shlomi Constantini  Director, Department

of Pediatric Neurosurgery, Dana Children’s Hospital, Tel Aviv Sourasky Medical Center, Sackler Faculty of Medicine, Tel Aviv University, Tel Aviv, Israel 34: Choroid plexus tumours; 88: Paediatric brain tumours

William T. Couldwell  Professor, Neurosurgery,

University of Utah, UT, USA 27: Surgical management of sellar and suprasellar tumours

Matt Crocker  Consultant Neurosurgeon, St

George’s Hospital, London, UK 62: Lumbar spinal disease

Michael D. Cusimano  Division of Neurosurgery,

University of Toronto, ON, Canada; Keenan Research Centre for Biomedical Science, The Li Ka Shing Knowledge Institute, St. Michael’s Hospital, Toronto, ON,  Canada 21: Surgical management of anterolateral skull base lesions

Tilak Das  Addenbrooke’s Hospital, Cambridge

University Hospitals, Cambridge, UK 3: Overview of neuroimaging

Hansen Deng  Resident Physician, Department of

Rina Dvir  Department of Pediatric

Hemato-​Oncology, Dana Children’s Hospital, Tel-​Aviv Medical Center, Tel-​Aviv University, Tel-​Aviv, Israel 34: Choroid plexus tumours

Frances Elmslie  South West Thames Regional

Genetics Service, St George’s University Hospitals NHS Foundation Trust, London, UK 38: Neurophakomatoses

Rudolf Fahlbusch  International Neuroscience

Institute (INI), Hannover, Germany 26: Craniopharyngioma and Rathke’s cleft cysts

Colin Ferrie  Department of Paediatric Neurology,

Leeds General Infirmary, Leeds, UK 84: Developmental disorders of the brain

Graham Flint  Consultant Neurosurgeon, Queen

Elizabeth Hospital, Birmingham, UK 65: Spinal cerebrospinal fluid dynamics

Petter Förander  Karolinska Institute, Stockholm,

Sweden 22: Schwannomas

Alan Forster  Consultant Clinical

Neurophysiologist, Aberdeen Royal Infirmary, Scotland, UK 71: Electrodiagnostics

Navin Furtado  Queen Elizabeth Hospital,

Birmingham, UK 60: Cervical spinal disease

Alexander Gamble  Bedford, New Hampshire,

USA 92: Hydrocephalus and normal CSF dynamics

Paul Gardner  Department of Neurological

Surgery, University of Pittsburgh, PA, USA 19: Surgical management of tumours of the orbit

H. Rao Gattamaneni  Consultant in Clinical

oncology, Christie Foundation Trust, Manchester, UK 25: Pituitary tumours

Neurological Surgery, University of Pittsburgh Medical Center, Pittsburgh, PA, USA 41: Pathophysiology of traumatic brain injury

Fred Gentili  Division of Neurosurgery, University

Chris Derham  Consultant Neurosurgeon, Leeds

George Georgoulis  Department of Neurosurgery,

General Infirmary, Leeds, UK 56: Surgical principles in spinal surgery

Federico Di Rocco  The Hôpital Necker Enfants

Malades, Paris, France 86: Spinal development and spinal dysraphism

Derek Duane  Consultant in Neuroanaesthesia

and Neurointensive Care, Cambridge University Hospital NHS Trust, Department of Anaesthesia, Addenbrooke’s Hospital, Cambridge, UK 5: Perioperative care of the neurosurgical patient

Hugues Duffau  Institute for Neurosciences of

Montpellie, France 6: Low-​grade  glioma

Andrew Durnford  Department of Neurosurgery,

Wessex Neurological Centre, Southampton General Hospital, Southampton, UK 46: Normal cerebrovascular physiology and vascular anatomy

of Toronto, ON, Canada 17: Esthesioneuroblastoma

General Hospital of Athens “Gennimatas”, Athens, Greece 79: Cranial nerve vascular compression syndromes

V. Gerganov  Consultant Neurosurgeon,

Department of Neurosurgery, International Neuroscience Institute, Hannover; and Associate Professor, Hannover Medical School, Germany 26: Craniopharyngioma and Rathke’s cleft cysts

Giorgio Gioffre  Treviso, Italy

20: Skull lesions

Kanna Gnanalingham  Salford Royal NHS

Foundation Trust, UK 25: Pituitary tumours

Tony Goddard  Consultant Interventional

Neuroradiologist, Leeds General Infirmary, Leeds, UK 90: Paediatric neurovascular disorders

Lior Gonen  Division of Neurosurgery,

Department of Surgery, University of Toronto, Canada. Department of Neurosurgery, Aurora Neuroscience Innovation Institute, Aurora St. Luke’s Medical Center, Milwaukee, Wisconsin 17: Esthesioneuroblastoma

John Goodden  Consultant Neurosurgeon

(Adult & Paediatric), Leeds General Infirmary, Leeds, UK 77: Spasticity

Alex Green  The Nuffield Department of

Clinical Neurosciences, University of Oxford, UK 75: Principles of deep brain stimulation

Fay Greenway  Atkinson Morley Department of

Neurosurgery, St George’s University Hospitals NHS Foundation Trust, London, UK 38: Neurophakomatoses

Alexander Grote  Department of Neurology,

University of Bonn, Bonn, Germany 10: Glioneuronal and other epilepsy-​associated tumours

Paul Grundy  Department of Neurosurgery,

Wessex Neurological Centre, University Hospital Southampton NHS Foundation trust, UK 32: Intraventricular tumours

Mathew Guilfoyle  Department of Neurosurgery,

Addenbrookes’s NHS Trust, Cambridge, UK 52: Extracranial-​intracranial bypass for cerebral ischaemia

Nigel Gummerson  Consultant in Orthopaedic

Trauma, Leeds General Infirmary, Leeds, UK 66: Scoliosis and spinal deformity

Nicholas Haden  Plymouth Hospitals NHS Trust,

Plymouth, UK 99: Spinal infection

Nicholas Hall  Surgeon, Royal Melbourne Hospital

Academic Centre, Parkville, VIC, Australia 24: Surgical management of cerebellopontine angle and petrous lesions

Richard M. Hall  Professor of Spinal Biomechanics,

School of Mechanical Engineering, University of Leeds, Leeds, UK 57: Spinal stability

Walter Hall  Professor of Neurosurgery, Upstate

Medical University, Syracuse, NY, USA 97: Microbiology

Seunggu J. Han  Department of Neurological

Surgery, University of California, San Francisco, CA, USA 14: Meningiomas and haemangiopericytoma (HPC)—​solitary fibrous tumour (SFT)

Adel Helmy  University Lecturer Neurosurgery,

University of Cambridge, UK/Honorary Consultant Neurosurgeon, Addenbrooke’s Hospital, Cambridge, UK 18: Malignant skull base tumours; 43: Surgical management of head injury

Catherine Hernon  Consultant in Plastic &

Reconstructive Surgery, Leeds General Infirmary, Leeds, UK 77: Spasticity

Contributors

Shawn Hervey-​Jumper  Associate Professor,

Neurosurgery, University of California San Francisco, USA 13: Surgical techniques in the management of intrinsic tumours

Nicholas Higgins  Department of Radiology,

Addenbrooke’s Hospital, Cambridge, UK 95: Pseudotumour cerebri syndrome

Peter J.A. Hutchinson  Professor of Neurosurgery,

University of Cambridge, Cambridge, UK 43: Surgical management of head injury

Mohsen Javadpour  Beaumont Hospital, Dublin,

Ireland 51: Carotid artery disease and cerebral ischaemia

Michael D. Jenkinson  Department of

Neurosurgery, University of Liverpool, Liverpool, UK 36: Pineal tumours

Timothy Jones  Department of Neurosurgery, St

George’s University Hospital Foundation Trust, London, UK 38: Neurophakomatoses

Sylvia Karcheva  Consultant in Neuroanaesthesia,

Cambridge University Hospital NHS Trust, Department of Anaesthesia, Addenbrooke’s Hospital, Cambridge University Hospitals, Cambridge, UK 5: Perioperative care of the neurosurgical patient

Neal Kassell  University of Virginia School of

Medicine, Charlottesville, VA, USA 47: The pathophysiology of aneurysms

Andrew Kay  Consultant Paediatric

Neurosurgery, Queen Elizabeth Hospital, Birmingham, UK 87: Special considerations in paediatric head and spinal trauma

Andrew H. Kaye  Surgeon, Royal Melbourne

Hospital Academic Centre, Parkville, VIC, Australia 24: Surgical management of cerebellopontine angle and petrous lesions

Tara Kearney  Salford Royal NHS Foundation

Trust, UK 25: Pituitary tumours

Chris Kellett  University College London Hospitals,

London, UK 62: Lumbar spinal disease

Nicole C. Keong  Consultant Neurosurgeon,

National Neuroscience Institute and Duke-NUS Medical School, Singapore 94: Normal pressure hydrocephalus

Osaama H. Khan  Department of Neurological

Surgery, Northwestern University 31: Surgical approaches to posterior fossa tumours

Sadaquate Khan  Department of Clinical

Neurosciences, Western General Hospital, Edinburgh, Scotland, UK 58: Spinal physiology

Andrew King  Manchester Skull Base Unit,

Department of Neurosurgery, Manchester Centre for Clinical Neurosciences, Salford Royal Hospital, University of Manchester, Manchester, UK 23: Glomus tumours

Matthew A. Kirkman  Specialty Registrar

in Neurosurgery and Honorary Fellow in Neurocritical Care; and The National Hospital for Neurology and Neurosurgery, University College London Hospitals NHS Foundation Trust, Queen Square, London, UK 42: Intensive care management of head injury

Peter Kirkpatrick  Department of Neurosurgery,

Addenbrookes’s NHS Trust, Cambridge, UK 52: Extracranial-​intracranial bypass for cerebral ischaemia

Ramez W. Kirollos  Consultant Neurosurgeon,

Addenbrooke’s Hospital, Cambridge, UK; Department of Neurological Surgery, University of California, San Francisco, CA, USA 14: Meningiomas and haemangiopericytoma (HPC)—​solitary fibrous tumour (SFT); 35: Surgical management of intraventricular lesions

Neil Kitchen  National Hospital for Neurology

and Neurosurgery, University College London Hospital NHS Foundation Trust, London, UK 4: The operating theatre environment

Georgios Klironomos  Division of Neurosurgery,

Department of Surgery, University of Toronto, Canada 17: Esthesioneuroblastoma

Angelos G. Kolias  Clinical Lecturer Neurosurgery,

Addenbrooke’s Hospital, Cambridge, UK 43: Surgical management of head injury

Daniel Krell  University College Hospitals,

London, UK 12: Chemotherapy for brain tumours

Giuseppe Lanzino  Mayo Clinic, Rochester,

MN, USA 47: The pathophysiology of aneurysms

Michael Lawton  Department of Neurosurgery,

Barrow Neurological Institute, Phoenix, AZ, USA 49: Management of subarachnoid haemorrhage; 53: Giant aneurysms and bypass surgery

Bryan Lee  Neurosurgeon, Barrow Neurological

Institute, Phoenix, AZ 67: Managing spinal cord trauma

Boon Leong Quah  Department of Neurosurgery,

Queen’s Hospital, Romford, London, UK and Division of Neurosurgery, Khoo Teck Puat Hospital, Singapore 9: Primary central nervous system lymphoma; 63: Spinal tumours

Shiong Wen Low  Consultant Neurosurgeon,

National University Hospital and Ng Teng Fong General Hospital, Singapore; and Associate Professor, Department of Surgery, National University of Singapore 69: Thoracic and lumbar spine injuries

Peter Loughenbury  Leeds Teaching Hospital NHS

Trust, Leeds, UK 57: Spinal stability

Ruichong Ma  Department of Neurosurgery,

Oxford University Hospitals NHS Foundation Trust, Oxford, UK 96: Arachnoid cysts

Andrew Maas  University of Antwerp and

Antwerp University Hospital, Department of Neurosurgery, Antwerp, Belgium 40: Epidemiology of head injury and outcome after head injury

Richard Mair  Clinical Lecturer in Neurosurgery,

Neurosurgery Division, Department of Clinical Neurosciences, Cambridge Biomedical Campus, Cambridge, UK 7: High-​grade gliomas and molecular biology of neurosurgical oncology

Conor L. Mallucci  Department of Paediatric

Neurosurgery, Alder Hey Children’s Hospital, Liverpool, UK 36: Pineal tumours; 89: Paediatric hydrocephalus

Geoffrey T. Manley  Professor and Vice Chair,

Department of Neurological Surgery, University of California San Francisco, San Francisco, CA, USA 41: Pathophysiology of traumatic brain injury

Richard Mannion  Department of Neurosurgery,

Addenbrooke’s Hospital, Cambridge, UK 78: Pain pathophysiology and surgical management

Hani Marcus  Faculty of Medicine, Department of

Surgery and Cancer, Imperial College London, London, UK 39: Uncommon brain lesions

Henry Marsh  Senior Consultant Neurosurgeon,

St George’s Hospital, London, UK 1: The History of Neurosurgery

Eleni Maratos  Consultant Neurosurgeon, King’s

College Hospital, London, UK 1: The History of Neurosurgery

Tiit Mathiesen  Department of Clinical

Neuroscience, Karolinska Institutet, Stockholm, Sweden; and University of Copenhagen & Rigshospitalet, Copenhagen, Denmark 22: Schwannomas

Calan Mathieson  Consultant Neurosurgeon,

Southern General Hospital, Glasgow, UK 68: Cervical spine injuries

Tomasz Matys  University Lecturer and Honorary

Consultant Neuroradiologist, Department of Radiology, University of Cambridge, Cambridge, UK 3: Overview of neuroimaging

Donald McArthur  Consultant Adult and

Paediatric Neurosurgeon, Nottingham University Hospitals NHS Trust, Nottingham, UK 30: Haemangioblastoma

Helen McCullagh  Leeds General Infirmary,

Leeds, UK 90: Paediatric neurovascular disorders

Michael McDermott  Department of Neurological

Surgery, University of California, San Francisco, USA 14: Meningiomas and haemangiopericytoma (HPC)—​solitary fibrous tumour (SFT)

Andrew McEvoy  National Hospital for Neurology

and Neurosurgery, University College London Hospitals NHS Foundation Trust, UK 16: Dermoid and epidermoid cysts; 82: Classification of seizures and epilepsy

xxiii

xxiv

Contributors

Eavan McGovern  University College Dublin

and St. Vincent’s University Hospital, Dublin, Ireland 59: Medical pathologies of the spinal cord

Chris McGuigan  University College Dublin

and St. Vincent’s University Hospital, Dublin, Ireland 59: Medical pathologies of the spinal cord

Iain McGurgan  University College Dublin

and St. Vincent’s University Hospital, Dublin, Ireland 59: Medical pathologies of the spinal cord

Jason McMillen  Consultant Neurosurgeon, Royal

Brisbane and Women’s Hospital, Brisbane, Queensland, Australia 48: The pathophysiology of subarachnoid haemorrhage

Harry Mee  Department of Rehabilitation

Medicine, Cambridge University Hospital NHS Foundation Trust, Cambridge, UK 70: Spinal cord injury rehabilitation

Michael P. Meier  Department of Surgery,

Division of Neurosurgery, St. Michael’s Hospital, University of Toronto, ON, Canada 21: Surgical management of anterolateral skull base lesions

H. Metwali  International Neuroscience Institute

(INI), Hannover, Germany 26: Craniopharyngioma and Rathke’s cleft cysts

Samantha Mills  Department of Neuroradiology,

The Walton Centre NHS Foundation Trust, Liverpool, UK 36: Pineal tumours

Catherine Moran  College of Surgeons, Dublin,

Ireland 61: Thoracic spinal disease

Jacques J. Morcos  Department of Neurological

Surgery, University of Miami Miller School of Medicine, FL, USA 31: Surgical approaches to posterior fossa tumours

Michael Morgan  Cerebrovascular Neurosurgeon,

Macquarie University Hospital, Sydney, NSW, Australia 50: Cerebral arteriovenous malformation and dural arteriovenous fistulae

Robert Morris  Consultant Neurosurgeon,

Addenbrooke’s Hospital, Cambridge, UK 71: Electrodiagnostics

Paul Mulholland  University College Hospitals,

London, UK 12: Chemotherapy for brain tumours

Thangaraj Munusamy  Department of

Neurosurgery, Addenbrooke’s Hospital, Cambridge, UK; and Division of Neurosurgery, Khoo Teck Puat Hospital, Singapore 9: Primary central nervous system lymphoma; 98: Cranial infections

Ammar Natalwala  Specialist Registrar in

Neurosurgery, Queen’s Medical Centre, Nottingham, UK 30: Haemangioblastoma

Lamia Nayeb  South Thames Foundation Trust,

London, UK 58: Spinal physiology

Jayson A. Neil  Midwest Neurosurgery Associates,

Kansas City, MO, USA 27: Surgical management of sellar and suprasellar tumours

Vincent Nga  Consultant Neurosurgeon, National

University Hospital Singapore, Singapore 74: Peripheral nerve tumours

Bart Nuttin  Neurosurgeon, UZ Leuven, Leuven,

Belgium 80: Neurosurgical interventions for psychiatric disorders

Karen O’Connell  University College Dublin

and St. Vincent’s University Hospital, Dublin, Ireland 59: Medical pathologies of the spinal cord

A. O’Hare  Beaumont Hospital, Dublin, Ireland

51: Carotid artery disease and cerebral ischaemia

Berk Orakcioglu  Ethianum, Heidelberg, Germany;

and Department of Neurosurgery, University of Heidelberg, Heidelberg, Germany 54: Spontaneous intracranial haematoma

Brian P. Walcott  Department of Neurosurgery,

University of Southern California, Los Angeles, USA 49: Management of subarachnoid haemorrhage

Benedict Panizza  University of Queensland,

Princess Alexandra Hospital, QLD, Australia 18: Malignant skull base tumours

Chris Parks  Department of Paediatric

Neurosurgery, Alder Hey Children’s Hospital, Liverpool, UK 89: Paediatric hydrocephalus

Tufail Patankar  Consultant Interventional

Neuroradiologist, Leeds General Infirmary, Leeds, UK 90: Paediatric neurovascular disorders

Hiren Patel  Salford Royal NHS Foundation Trust,

Salford, UK 55: Cavernoma and angiographically occult lesions

Omar Pathmanaban  Manchester Skull Base

Unit, Department of Neurosurgery, Manchester Centre for Clinical Neurosciences, Salford Royal Hospital, University of Manchester, Manchester, UK 23: Glomus tumours

Erlick Pereira  Senior Lecturer and Honorary

Consultant at St George’s, London, UK 75: Principles of deep brain stimulation

Jonathan Perera  Guy’s and St. Thomas’ NHS

Foundation Trust, London, UK 73: Supraclavicular brachial plexus and peripheral nerve injuries

David Pettersson  Karolinska Institute, Stockholm,

Sweden 22: Schwannomas

John D. Pickard  Division of Neurosurgery,

Department of Clinical Neurosciences, University of Cambridge, Cambridge, UK 95: Pseudotumour cerebri syndrome

Ian K. Pople  Department of Neurosurgery,

Southmead Hospital, Bristol, UK 93: Shunt technology and endoscopic ventricular surgery

Stephen Price  Addenbrooke’s Hospital,

Cambridge University Hospitals, Cambridge, UK 7: High-​grade gliomas and molecular biology of neurosurgical oncology

Harold Rekate  Hofstra University, Hempstead,

NY, USA 92: Hydrocephalus and normal CSF dynamics

Eduardo C. Ribas  Division of Neurosurgery,

University of São Paulo Medical School, and Albert Einstein Hospital, São Paulo, Brazil 35: Surgical management of intraventricular lesions

Guilherme C. Ribas  Professor of Surgery,

Department of Surgery, University of São Paulo Medical School, Department of Surgery and Neurosurgeon, Albert Einstein Hospital, São Paulo, Brazil 35: Surgical management of intraventricular lesions

Desiderio Rodrigues  Consultant Paediatric

Neurosurgeon, Queen Elizabeth Hospital, Birmingham, UK 87: Special considerations in paediatric head and spinal trauma

Roberto Rodriguez Rubio  Department of

Neurological Surgery, University of California, San Francisco, USA 49: Management of subarachnoid haemorrhage

Federico Roncaroli  Clinical Reader in

Neuropathology, University of Manchester, Manchester, UK 25: Pituitary tumours

Jonathan Roth  Department of Pediatric

Neurosurgery, Dana Children’s Hospital Tel-​Aviv Sourasky Medical Center, Tel Aviv University, Israel 34: Choroid plexus tumours; 88: Paediatric brain tumours

James Rutka  Neurosurgeon and Senior Scientist,

The Hospital for Sick Children (SickKids), Toronto, ON, Canada 28: Medulloblastoma

Thomas Santarius  Department of Neurosurgery,

Addenbrooke’s Hospital, Cambridge, Cambridge, UK 6: Low-​grade glioma; 14: Meningiomas and haemangiopericytoma (HPC)—​solitary fibrous tumour (SFT); 39: Uncommon brain lesions

Johannes Schramm  Department of Neurosurgery,

University of Bonn, Bonn, Germany 83: Surgical management of epilepsy

Daniel. J. Scoffings  Addenbrooke’s Hospital,

Cambridge University Hospitals, Cambridge, UK 3: Overview of neuroimaging

Brian Scott  Consultant Paediatric

Orthopaedic Surgeon, Leeds General Infirmary, Leeds, UK 77: Spasticity

Contributors

Alessandro Scudellari  Consultant in

Neuroanaesthesia, Addenbrooke’s Hospital, Cambridge University Hospitals, Cambridge, UK 5: Perioperative care of the neurosurgical patient

Senthil Selvanathan  Consultant Neurosurgeon,

Leeds Teaching Hospital NHS Trust, Leeds, UK 56: Surgical principles in spinal surgery

Richard Selway  King’s College Hospital NHS

Foundation Trust, London, UK 81: Diagnosis and assessment

Ashish H. Shah  Department of Neurological

Surgery, University of Miami Miller School of Medicine, FL, USA 31: Surgical approaches to posterior fossa tumours

Jonathan Shapey  The National Institute of

Neurology and Neurosurgery, London, UK 4: The operating theatre environment

Melanie Sharp  Consultant Paediatric

Neurosurgeon, Queen Elizabeth Hospital, Birmingham, UK 87: Special considerations in paediatric head and spinal trauma

Asim Sheikh  Consultant Neurosurgeon, Leeds

General Infirmary, Leeds, UK 33: Colloid cyst

Susan Short  Professor of Clinical Oncology and

Neuro-​Oncology, Leeds Institute of Cancer and Pathology, Leeds, UK 11: Radiotherapy and radiosurgery for brain tumours

Adikarige Haritha Dulanka Silva  Queen Elizabeth

Hospital, Birmingham, UK 60: Cervical spinal disease

Matthias Simon  Department of Neurology,

University of Bonn, Bonn, Germany 10: Glioneuronal and other epilepsy-​associated tumours

Marc Sindou  Department of Neurosurgery,

University of Lyon 1, Hôpital Neurologique P. Wertheimer, Groupement Hospitalier Lyon, Lyon, France 79: Cranial nerve vascular compression syndromes

William Singleton  Department of Neurosurgery,

Southmead Hospital, Bristol, UK 93: Shunt technology and endoscopic ventricular surgery

Marco Sinisi  Consultant Surgeon, Guy’s

and St Thomas’ NHS Foundation Trust, London, UK 73: Supraclavicular brachial plexus and peripheral nerve injuries

Martin Smith Martin  Consultant in

Neuroanaesthesia and Neurocritical Care, National Hospital for Neurology and Neurosurgery, University College London Hospitals, UK 42: Intensive care management of head injury

Saksith Smithason  Cleveland Clinic,

OH, USA 67: Managing spinal cord trauma

Guirish Solanki  Consultant Paediatric

Neurosurgeon, Queen Elizabeth Hospital, Birmingham, UK 87: Special considerations in paediatric head and spinal trauma

S. Tonya Stefko  Department of Ophthalmology,

University of Pittsburgh, PA, USA 19: Surgical management of tumours of the orbit

Catherine Suen  Department of Neurology,

University of Utah School of Medicine, UT, USA 41: Pathophysiology of traumatic brain injury

Yuval Sufaro  Skull Base Fellow, Royal Melbourne

Hospital, Australia 24: Surgical management of cerebellopontine angle and petrous lesions

Kieron Sweeney  Department of Paediatric

Neurosurgery, Children’s University Hospital, Dublin, Ireland 51: Carotid artery disease and cerebral ischaemia; 61: Thoracic spinal disease

Mobin Syed  Consultant Plastic surgeon Guys and

St Thomas Hospital, London [email protected]

Tamara Tajsic  Department of Neurosurgery,

Cambridge University Hospital NHS Foundation Trust, Cambridge UK 70: Spinal cord injury rehabilitation

Rokas Tamosauskas  Consultant in Anaesthesia

and Pain Medicine, Cambridge University Hospitals NHS Foundation Trust, Department of Pain Medicine, Addenbrooke’s Hospital, Cambridge, UK 78: Pain pathophysiology and surgical management

Boon Hoe Tan  Division of Neurosurgery, Khoo

Teck Puat Hospital, Singapore 98: Cranial infections

Mario Teo  Consultant Neurosurgeon,

Bristol Institute of Clinical Neuroscience, UK 53: Giant aneurysms and bypass surgery

Dominic Thompson  Consultant in Paediatric

Neurosurgery, Great Ormond Street Hospital, London, UK 85: Spinal development and spinal dysraphism

Simon Thomson  Leeds Teaching Hospital NHS

Trust, Leeds, UK 56: Surgical principles in spinal surgery

Martin Tisdall  Consultant Paediatric

Neurosurgeon, Great Ormond Street Hospital, London, UK 91: Paediatric epilepsy

Ivan Tmofeev  Department of Neurosurgery,

Addenbrookes’s NHS Trust, Cambridge, UK 20: Skull lesions

Nicolas de Tribolet  Professor of Neurosurgery,

University of Lausanne, Switzerland 37: Surgical management of pineal region lesions

Rikin Trivedi  Department of Neurosurgery,

Addenbrooke’s Hospital, Cambridge, UK 74: Peripheral nerve tumours

Kevin Tsang  Consultant Neurosurgeon, Imperial

College NHS Trust, London, UK 58: Spinal physiology

Georgios Tsermoulas  Queen Elizabeth Hospital,

Birmingham, UK 60: Cervical spinal disease

Atul Tyagi  Consultant Neurosurgeon (Adult &

Paediatric), Leeds General Infirmary, Leeds, UK 84: Developmental disorders of the brain; 90: Paediatric neurovascular disorders

Ismail Ughratdar  University Hospitals

Birmingham NHS Foundation Trust, Birmingham, UK 76: Movement disorders

Andreas W. Unterberg  Department of

Neurosurgery, University of Heidelberg, Heidelberg, Germany 54: Spontaneous intracranial haematoma

Sophia Varadkar  Consultant Paediatric

Neurologist, Great Ormond Street Hospital, London, UK 91: Paediatric epilepsy

Elizabeth Visser  Aberdeen Royal Infirmary,

Scotland, UK 2: Clinical assessment

Daniel Walsh  Consultant Neurosurgeon, Kings

College Hospital, London, UK 64: Vascular lesions of the spinal cord

Yizhou Wan  Resident Physician, Department

of Neurosurgery, Oxford University Hospitals, Oxford, UK 39: Uncommon brain lesions

Mueez Waqar  Department of Neurosurgery,

The Walton Centre NHS Foundation Trust, Liverpool, UK 36: Pineal tumours

Daniel Warren  Consultant Neuroradiologist,

Leeds General Infirmary, Leeds, UK 84: Developmental disorders of the brain

Colin Watts  Professor of Neurosurgery/Honorary Consultant Neurosurgeon, University of Birmingham

9: Primary central nervous system lymphoma

Tim Wehner  Formerly Department of Clinical

Neurophysiology, National Hospital for Neurology and Neurosurgery, Department of Clinical and Experimental Epilepsy, Institute of Neurology, University College London, London, UK 82: Classification of seizures and epilepsy

Edward White  Consultant Neurosurgeon,

University Hospitals Birmingham, UK 99: Spinal infection

Peter C. Whitfield  University Hospitals Plymouth,

Derriford Road, Plymouth, UK 44: Complications of head injury

Mark Wilson  St Mary’s Hospital, London, UK

45: Concussion and sports-​related head injury

Ethan A. Winkler  Resident Physician, Department

of Neurological Surgery, University of California, San Francisco, CA, USA 41: Pathophysiology of traumatic brain injury

xxv

xxvi

Contributors

Christoph M. Woernle  Department of

Neurosurgery, Klinik Hirslanden, Zurich, Switzerland 37: Surgical management of pineal region lesions

Victoria Wykes  Department of Neurosurgery,

National Hospital for Neurology and Neurosurgery, University College London, London, UK 82: Classification of seizures and epilepsy

Eugene Yang  Division of Neurosurgery, Khoo

Teck Puat Hospital, Singapore 98: Cranial infections

John K. Yue  Resident Physician, Department of

Neurological Surgery, University of California, San Francisco, CA, USA 41: Pathophysiology of traumatic brain injury

Zsolt Zador  NIHR Academic Clinical Lecturer and

Specialty Registrar in Neurosurgery, Department of Neurosurgery, Salford Royal NHS Foundation Trust, University of Manchester, Manchester, United Kingdom 25: Pituitary tumours

Rashed Zakaria  The Walton Centre NHS

Foundation Trust, Liverpool, UK 8: Intracranial metastasis

SECTION 1

Principles of neurological surgery 1. The history of neurosurgery  3 Eleni Maratos and Henry Marsh

4. The operating theatre environment  45 Neil Kitchen and Jonathan Shapey

2. Clinical assessment  11 Peter Bodkin and Elizabeth Visser

5. Perioperative care of the neurosurgical patient  57 Karol P. Budohoski, Alessandro Scudellari, Sylvia Karcheva, and Derek Duane

3. Overview of neuroimaging  31 Tomasz Matys, Daniel. J. Scoffings, and Tilak Das

1

The history of neurosurgery Eleni Maratos and Henry Marsh

Introduction The history of neurosurgery falls naturally into the premodern era, where it is essentially the history of surgery to the skull and of head injuries, and the modern era, where it is the history of surgery to the brain itself, made possible by cerebral localization theory, antisepsis, and anaesthesia, all of which developed in the nineteenth century. The first known neurosurgical procedures were skull trephines, seemingly carried out on both the living and the dead. It is unclear whether these were performed for therapeutic or ritualistic reasons. There are many trepanned skulls dating back thousands of years to the Neolithic era, and perhaps to even earlier, from sites all over the world.

Ancient Egypt The earliest neurosurgical writings can be traced to Ancient Egypt and have been preserved in the Edwin Smith Papyrus that dates from c.1600 BC. This is the first time that the management of head injuries was based on rational scientific method rather than magic. In a style that is highly reminiscent of modern case reports, the first ten cases focus on head wounds and also contain the earliest known reference to the brain itself (as opposed to the skull), which describes a ‘convoluted structure like ripples that happen in copper through smelting’. The cases vividly describe a methodical approach—​first ascertaining the depth of the injury and whether there is an underlying skull fracture or exposed brain, and then advocating different management strategies according to the findings. The papyrus has been translated and reproduced by the National Institutes of Health and can be studied online (https://​ceb.nlm.nih.gov). As well as the earliest known advice on head injury management, the Edwin Smith Papyrus also contains the first references to spinal immobilization in order to prevent further injury.

Ancient Greece and Byzantium Hippocrates (460–​370 BC) wrote extensively on head injuries in De capitis vulneribus (On Head Wounds). His management strategy was almost exclusively based on classifying the fracture rather than on

the clinical state of the patient. He described five categories of skull fracture, including contre-​coup injuries. On head injuries, he wrote ‘nullum capitis vulnus contemnendum est’ (no head injury should be considered trivial). His writings were not confined to head injury and he is also credited with the first description of a subarachnoid haemorrhage: .  .  .  when persons in good health are suddenly seized with pains in the head and straight away are laid down speechless and breath with stertor they die in seven days unless fever comes on.

He also described contralateral convulsions associated with brain injury; he can therefore perhaps be credited with being the first doctor to describe cerebral localization. The Alexandrian school introduced formal anatomy dissection in approximately 300 BC. It was at this time that Herophilus began to develop the anatomic nomenclature that we use today (see Fig. 1.1). He identified that nerves and tendons were indeed different structures, contrary to what the Egyptians had thought, and he also was the first to describe the anatomy of the ventricles and the venous sinuses. The confluence of the sinuses, of course, still bears his name ‘torcula (wine press) Herophili’. He also described the pen nib or ‘calamus scriptorius’ at the base of the fourth ventricle ‘αναγλυφη της χαλαμης’ and the choroid plexus (named after its resemblance to the vessels of the placenta). Celsus, who lived from 25 BC to 50 AD, was the first to describe inflammation (rubor, dolor, tumor) and also wrote about extradural haemorrhage, hydrocephalus, trigeminal neuralgia, and spinal fractures. He advocated craniotomy only as a last resort in head injury and described the technique of drilling holes and connecting them up with a hammer and chisel with a protective blade. In another early account of cerebral localization, he advised operating on the side with the greatest pain.

Galen 129–200 ad With Galen came the encephalocentric model of science and medicine, where the brain was recognized as the seat of intelligence and voluntary movement. He identified the pia (as different from the dura mater), the corpus callosum, the ventricles (alongside Herophilus), and the pineal and pituitary glands. He described the aqueduct of

4

Section 1  Principles of neurological surgery

Fig. 1.1  Herophilus of Chalcedon (c. 330–​260 bc), a Greek doctor who practised in Alexandria. Here he is seen depicted in discussion with Erasistratus of Ceos. Wellcome Collection

Sylvius and the foramen of Monro before the anatomists whose names they now possess. He made other significant contributions to neuroscience by identifying that transecting the spinal cord leads to loss of function below the level of the lesion, and he also described what we now call Brown-​Séquard syndrome. In addition, he noted that recurrent laryngeal nerve injuries led to hoarse ‘voices’ in his experimental dogs. He also advocated elevating depressed skull fractures and using irrigation to reduce the heat created by trephining. His teachings were accepted, largely unquestioned, for 1500 years. During mediaeval times (750–​1200 AD) most innovative medical writing came from the Islamic and Arabic worlds, which also kept the Hippocratic and Galenic teachings alive. It was during this time that traditional bedside teaching was established.

The beginning of the scientific era The sixteenth and seventeenth centuries were remarkable for the advances in all aspects of science and philosophy. Anatomical dissection of cadavers was already practised in mediaeval Europe before Vesalius, but he was more extensive and systematic in his dissections and was the first to challenge the teachings of Galen and Aristotle. William Harvey (1578–​1657) published his seminal work on the function of the heart and circulation of blood de motu Cordis et Sanguinis in Animalibus (1628). Willis (1621–​1675) applied this knowledge to his understanding of cerebral anatomy and showed that occluding parts of his eponymous circle did not compromise flow, thereby confirming their anastomotic nature. Surgery continued to develop in all areas, although neurosurgery remained confined to trauma. In a challenge to orthodoxy, Yonge (1646–​1721) stated that mortality was ‘not inevitable’ once the dura had been breached. Percival Pott not only worked on tuberculosis (TB) of the spine (which has made a recent resurgence in neurosurgical practice) but also made significant advances in our understanding of head injuries. He asserted that in head injury, trepanning could relieve the pressure from extravasated fluid, thus providing a rationale for the oldest neurosurgical operation. At the same time, Jean Louis Petit (1674–​1750) described the classic ‘lucid interval’ associated with an extradural haematoma as a brief loss of

consciousness followed by a gradual deterioration due to accumulation of blood compressing the brain. The importance of brain injury rather than head injury began to be recognized. Bell (Edinburgh 1749–​1806) described the loss of pupillary reaction to light and recognized this as an indication to perform a ‘rapid and prompt evacuation’ of an extradural haematoma. Bell is also credited with identifying that hydrocephalus can be associated with spina bifida. Although the majority of operations were still for trauma, one key exception was the successful extirpation by Morand (1697–​1773) of a temporal abscess following otitis media and mastoiditis. He describes exploring it with his finger and washing out the cavity before placing a silver tube which was slowly withdrawn—​a method which resonates today, where silver is once again finding a medical role for its bactericidal properties. Cotugno (1736–​1822) further characterized the cerebrospinal fluid (CSF) pathways and described the ‘nervous origins’ of sciatica for the first time. He also described hydrocephalus ex vacuo as being secondary to cerebral atrophy. The surgeon/​ anatomist John Hunter (1728–​ 1793) was a student of Pott’s and made major contributions to general surgery and neurosurgery. His extensive anatomy collection houses the Irish giant (Cushing’s original acromegalic) and can still be seen at the Hunterian Museum at the Royal College of Surgeons, London. However, during this time neurosurgical operations were still plagued by infection and the lack of anaesthetic agents meant that operations had to be rapid and not always accurate. The clinical skill of cerebral localization was also yet to be described, so surgery was confined to lesions that had external manifestations.

Early modern to nineteenth century The modern practice of neurosurgery as we know it today began in the nineteenth century. The parallel advances in anaesthesia and antisepsis allowed mortality rates to fall to acceptable levels and the ability to localize cerebral lesions by clinical examination alone allowed surgeons to tackle occult lesions, thereby dramatically broadening the scope of the specialty. These advances ultimately led to the creation of neurosurgery as a specialty in its own right.

Parallel advances Anaesthesia The introduction of ether in the 1840s allowed surgery to be performed in a pain-​free manner without the surgeon being rushed or restraints being used. The dentist William Morton persuaded Dr Warren to use ether to excise a submaxillary tumour in 1846 at the Massachusetts General Hospital. Shortly after this, in 1847, Marie Jean Pierre Flourens successfully used chloroform anaesthesia. It was not without risk, but the advantages to both surgeons and patients were clear. As anaesthesia became safer, so the range of pathologies that could be tackled grew and complex surgical approaches could be developed. Antisepsis Before antisepsis, neurosurgery was almost inevitably fatal due to ‘suppuration’. Louis Pasteur (1822–​1895) developed his germ theory and radically changed current opinion as to the origins of

CHAPTER 1  The history of neurosurgery

infection. He postulated that meat putrefaction and fermentation were not due to ‘spontaneous’ degeneration but to living microscopic organisms. Joseph Lister (1827–​1912), who was Professor of Surgery at the University of Glasgow, applied Pasteur’s germ theory to surgery and in particular to wound infections. He began using carbolic acid and commented that his wounds healed without pus. The mortality from amputations was dramatically cut from 45% to 15%. Meanwhile, in America William Keen (1837–​1932) adopted Lister’s principles in his operating theatre, applying them to the surgeon, the patient, and to the theatre environment; principles that are still being pursued today. He insisted all surfaces were cleaned with carbolic and carpets and furniture were removed. The surgeon’s hands were washed with soap, alcohol, and sublimate, and the patients were also prepped for the first time with a head shave, soap and water, ether, and wet sublimate dressings as well as mercuric chloride washings. Handwashing was also shown to radically reduce infections in obstetrics and gynaecology by Holmes Semmelweiss. Keen boiled his surgical instruments for 2 hours in a precursor to heat sterilization, which was introduced in 1891 by Ernst von Bergman. In 1883 Neuber began to use sterile gowns and caps, but it was not until William Stewart Halsted (1852–​1922) commissioned Goodyear to make some rubber gloves in 1890 to protect his nurse’s hands from the mercuric chloride that gloves were routinely used in surgery (see Fig. 1.2). Mikulicz subsequently introduced masks in 1897. Cushing himself stated in 1915 that ‘certainly infections cannot be attributed to the intervention of the devil but must be laid at the surgeon’s door’. He published his series of 130 cases with only a single infection—​an infection rate of less than 1%, which remains enviable today. A Centre for Disease Control and prevention (CDC) audit on craniotomy infection rates from 1992 to 2003 reported rates between 0.86% for low risk cases and 2.32% for the high risk. Cushing also reported a perfectly respectable 8.4% mortality in the

same series (Cushing, 1915). In the 1940s antibiotic prophylaxis was introduced to reduce postoperative infection. Cerebral localization In the 1860s the correlation of neurological symptoms and signs with the cerebral location of a lesion began to transform the scope of neurosurgery. The neurological examination is so ingrained in our current practice that it is difficult to imagine a time before cerebral localization. At the beginning of the nineteenth century, localization theory had been associated with the discredited pseudo-​science of phrenology. Paul Broca (1824–​1880) then showed that speech was localized in the brain’s left hemisphere and John Hughlings Jackson (1835–​1911) described what came to be known as the ‘Jacksonian march’ of focal motor activity ‘in which parts of the body are affected one after another’. He also correctly identified that a third cranial nerve palsy lateralizes the haematoma to the side of the fixed and dilated pupil. He directed Sir Jonathan Hutchinson (1828–​1913) to perform surgeries based on his localizations and ipsilateral pupillary dilatation carries the eponym ‘Hutchinson’s pupil’. It was during the same period that David Ferrier (1843–​1928) was using Faradic current stimulation in animals including primates to create one of the first cortical maps, including the correct location of the motor cortex. He credited Hughlings Jackson with predicting the outcome of his animal studies. Fritz and Hitzig were using similar cortical stimulation techniques to map the cortex in dogs during the same time period. Dr Alexander Hughes Bennett (1848–​1901) is credited with being the first neurologist to direct the removal of a tumour by cerebral localization alone. In 1885 he directed Sir Rickman Godlee (1849–​ 1925) to remove a motor strip tumour from a patient who had presented with contralateral focal motor seizures and subsequent progressive hemiparesis. The operation was successful, but the patient later died of infection, as was so often the case (Bennett and Godlee, 1965; Kerr et al., 2005).

Fig. 1.2  Photograph of Harvey Cushing and William Halsted in the operating room. Note the rubber gloves, made by Goodyear, that Halsted was the first to use routinely in surgery.

5

6

Section 1  Principles of neurological surgery

Haemostasis

William Macewen 1848–​1924

The development of effective neurosurgery required new methods to deal with the torrential haemorrhage often associated with craniotomy. William Bovie (1882–​1958) developed the method of electrocautery for haemostasis and it was first used by Cushing in 1926. The development of blood transfusion in the early years of the twentieth century was also of importance. Until then, craniotomies would sometimes be done as staged procedures, with an interval between the operations to allow the patient’s own haematopoeisis to replace the blood lost on opening the scalp and skull.

William Macewen (1848–​1924) was born in Port Bannatyne on the Isle of Bute, Scotland. He was Regius Professor of Surgery at Glasgow University from 1892 to 1924 and was knighted in 1909 when he became Surgeon to the King. His major contributions to surgery were in orthopaedics, establishing bone graft surgery, and setting up the Princess Louise Scottish Hospital for Limbless Soldiers and Sailors where he also designed the Erskine artificial limb. He is credited with performing the first brain tumour operation when he successfully removed a presumed meningioma from a 14-​year-​old girl in 1879. He did not require the then-​novel technique of cerebral localization as there was hyperostosis overlying the meningioma, which directed his surgery. She had indeed presented with cosmetic deformity, and subsequently had refractory seizures contralateral to the lesion, which provided Macewen with the indication to attempt the resection. He was a strong believer in sterile surgical techniques and she survived a further 8  years, and was able to work, before dying of other causes. The low mortality of his operations for cerebral abscess bears comparison with modern series.

Early pioneers These major technical advances in anaesthesia, antisepsis, cerebral localization, and haemostasis set the scene for the early pioneers of neurosurgery to establish neurosurgery as a specialty in its own right. Victor Horsley 1857–​1916 Sir Victor Horsley (1857–​1916) was a clinician, researcher, and surgeon, and was therefore ideally qualified to make significant advances in cerebral localization. He published works on the topography of the motor cortex and ‘the arrangement of the internal capsule’, as well as relating surface anatomy to underlying cortical features in ‘topographical relations of the cranium and surface of the cerebrum’. In 1888 he was the first to successfully remove a spinal cord tumour (localized by William Gowers, neurologist, 1845–​ 1915). He was not the first to remove a brain tumour—​that was William Macewen (1848–​1924)—​but he is credited with performing several pioneering operations, including craniosynostosis surgery for raised intracranial pressure and sectioning of the posterior root of the trigeminal nerve for neuralgia; he was also the first to operate on the pituitary. His technical advances included the development of antiseptic beeswax for bone bleeding (still used today) and the Horsley-​Clark stereotactic frame. He was a general surgeon himself, but was at the forefront of developing neurosurgery as a specialty and was given the first specifically neurosurgical appointment while at Queen Square. He died in the First World War from desert fever.

Fig. 1.3  Harvey Cushing, the father of modern neurosurgery.

William W. Keen 1837–​1924 The first American neurosurgeon was William Williams Keen Jr. He studied at Jefferson Medical College and was President of the Philadelphia School of Anatomy. As was commonplace at that time he toured Europe during his education, spending time in Paris and Berlin. He is credited with introducing the Gigli saw to America. He pioneered CSF drainage for hydrocephalus, describing the eponymous parietal burr hole still often used for ventriculoperitoneal shunt insertion 3 cm posterior and superior to the pinna. Although not the first to remove a brain tumour, he was noted for removing some large tumours successfully. Harvey Williams Cushing 1869–​1939 Harvey Williams Cushing is often seen as the father of modern neurosurgery (see Figs. 1.3 and 1.4). Arriving on the scene at just the right time when antisepsis, anaesthesia, and cerebral localization

CHAPTER 1  The history of neurosurgery

that exerted against the medulla’—​now known as the Cushing reflex. His travels then took him to Italy where he was so impressed with the blood pressure monitor designed by Scipione Riva-​Rocci that he brought it back to America with him. This era from 1901 to 1910 was a key decade for neurosurgery. Cushing recognized the importance of raised intracranial pressure and of meticulous surgical technique. He was responsible for the introduction of intraoperative blood pressure monitoring and following his own experience of a death on the operating table, he also introduced anaesthetic charts to record pulse and respirations during a procedure. With Percival Bailey (1892–​1973) and Louise Eisenhardt (1891–​ 1967) he coauthored the standard works on gliomas and meningiomas, and was responsible for pioneering work on the diagnosis and treatment of pituitary disorders. Walter Dandy 1886–​1946

Fig. 1.4  Harvey Cushing stands by a patient with gigantism.

Walter Edward Dandy studied medicine and subsequently spent his whole career at Johns Hopkins Hospital (see Fig. 1.5). He was of the same Halsted surgical dynasty as Cushing and was Cushing’s research assistant at the Hunterian laboratory that Cushing had established at Johns Hopkins. His work on the production and absorption of CSF acknowledged the importance of the choroid plexus and made advances in our understanding of hydrocephalus of infancy. His work on ‘pneumoventriculography’ (1918) radically changed neurosurgery for the next 60 years. In 1913 Luckett had described the radiological appearances of pneumocephalus but Dandy developed the technique by which air ventriculography (or pneumoencephalography as it was known) allowed the interpreter to infer the location of intra-​axial masses by the distortion they created in the ventricles. Along with cerebral angiography (discovered by

Reproduced from Shin P. Harvey Cushing’s Ghosts: Death and Hauntings in Modern Medicine. Yale J Biol Med. 2011;84(2):91–101

were now established, Cushing was well positioned to make his contribution to the specialty. He was trained in general surgery by Halsted himself, who was Professor of Surgery at the time, and who is credited for teaching Cushing his meticulous surgical technique. Another great influence on Cushing was Sir William Osler (1849–​1919), who encouraged him in the more philosophical aspects of medicine including the study of the history of medicine. Cushing was one of Osler’s ‘latchkeyers’, students who lived next door to Osler and were given the ‘latchkey’ to his extensive library. In those days it was common for American doctors to spend some time visiting their counterparts in Europe, and Osler encouraged Cushing to do the same. Cushing visited Horsley in London and by his own account was ‘a little disappointed’ in him. A meticulous surgeon himself, Cushing described Horsley’s technique as ‘execrable’ and commented that he must have many ‘septic wounds’. Indeed, at that time Horsley accepted mortality rates of 30–​50% from ‘brain fungation’ (i.e. infection). Cushing then travelled to France and Switzerland where he worked with Kocher in Bern. In contrast to what he writes about Horsley, he comments that Kocher’s technique was ‘detailed’ and ‘tedious’ with ‘absolute haemostasis’. Cushing’s seminal work on intracranial pressure began at this time in the lab of Kronecker. It was here that he made the discovery that ‘an increase in intracranial tension occasions a rise of BP which tends to find a level slightly above

Fig. 1.5  Walter Dandy, a student of Cushing’s and later his rival, made several key contributions to neurosurgery.

7

8

Section 1  Principles of neurological surgery

Moniz in 1927)  this became the mainstay of diagnostic imaging until the advent of the computed tomography (CT) scanner in the 1970s. Previously Cushing and others had used X-​rays to identify extra-​axial masses such as sellar lesions but until the technique of pneumoencephalography was introduced, intra-​axial lesions could only be diagnosed by the clinical skill of cerebral localization. He was the first surgeon to operate for a colloid cyst in 1921. He advocated total resection of an acoustic neuroma as opposed to Cushing’s more conservative subtotal removal, sectioning the fifth cranial nerve for trigeminal neuralgia and developed a transcallosal approach to the pineal. He also identified a ruptured intervertebral disc as a cause of cauda equina syndrome and bilateral sciatica in 1929.

Late twentieth century to today In recent decades, progress has largely been driven by advances in technology, with the sometimes slightly paradoxical effect of changing management away from neurosurgery, with developments such as interventional radiology for aneurysms and stereotactic radiosurgery for small vestibular schwannomas.

Technological advances Diagnostic imaging The development of diagnostic imaging techniques has revolutionized neurological diagnosis (and introduced the modern problem of the incidental finding). Following Dandy’s pneumoencephalography the next major advance in imaging techniques was cerebral angiography. This was developed by Antonio Caetano de Egas Moniz (1874–​ 1935) who was subsequently awarded the Nobel Prize in 1949 for the more dubious invention of frontal leucotomy for psychosis (and was also shot and rendered paraplegic by one of his patients). Myelography was introduced by Jean Athanase Sicard (1872–​1919) and allowed mass lesions of the spine to be identified. Parallel advances in X-​rays and computing allowed the development of the CT scanner by Godfrey Hounsfield (1919–​2004) at EMI. The first CT scan of a patient was in 1971 at the Atkinson Morley Hospital, London, confirming a right frontal lobe tumour. The development of positron emission tomography (PET) and magnetic resonance imaging (MRI) swiftly followed in the 1980s. Operative microscope Gunnar Holmgren in Stockholm in 1927 was one of the pioneers of the binocular microscope and its first clinical use was by otorhinolaryngologists for the management of otitis media. Several refinements were made to the apparatus, largely led by ears, nose, and throat (ENT) surgeons themselves. Perhaps surprisingly it was not until 1957 that the microscope was first used in neurosurgery. Theodor Kurze, apparently inspired by a film of the ENT surgeon William House using the operative microscope, used it for the removal of a neurilemoma in a 5-​year-​old. There are innumerable other advances in the field of neurosurgery. The adoption of the endoscope, for example, has transformed skull base surgery and the introduction of ventriculoperitoneal shunts has radically changed our management of hydrocephalus. These technologies are still, relatively speaking, in their infancy and will continue to develop throughout this century at least.

History of spinal neurosurgery Although this chapter mainly focuses on the history of cranial neurosurgery, the development of spinal surgery should not be overlooked. Once more, the Edwin Smith Papyrus gives us vivid descriptions of the way spinal trauma was managed in Ancient Egypt. The ability of the Ancient Egyptians, and in particular Imhotep (thought to be the world’s first physician) to tie in the clinical picture with the management plan and the prognosis was remarkable. Case 31 clearly describes a complete spinal cord injury secondary to a cervical vertebra dislocation—​ centuries before Brown-​ Séquard’s treatise on spinal localisation in 1858. The Hippocratic School in Ancient Greece described the first traction device, advocating correction of an acquired kyphosis following trauma. Hippocrates recognized compression of ‘spinal marrow’ by displaced vertebrae as the cause of paralysis. As in Ancient Egypt, no treatment was advised if the patient was paralysed. Celsus and Galen did not advocate surgery either, recognizing the poor outcomes associated with spinal cord damage, especially once urinary dysfunction had occurred. Celsus recognized that high cervical lesions were associated with quadriparesis and respiratory compromise while thoracic lesions led to paraparesis. Galen (131–​201 AD) coined the terms kyphosis, lordosis, and scoliosis. It is worth noting that he also was the first to describe tuberculous spinal disease (a few hundred years before Percival Pott 1713–​1788) and, having conducted experiments where he demonstrated that spinal cord transection led to loss of function below the level of the lesion, he also contributed to the conceptual advance of the brain as the source of voluntary action with signals being passed along the spinal cord. Paulus of Aegina (624–​ 690  AD) was the first to advocate surgery on the spine, removing bony fragments in injured patients with paralysis. Interestingly, in an early reference to ‘informed consent’ he describes his surgery as being carried out ‘after warning of the dangers’. Very little progress was made in the mediaeval period and the ‘dark ages’ in Europe. The seat of medicine was transferred to the Arabic world. Previously acquired knowledge from the Alexandrian and Hippocratic schools was preserved by translating the Latin documents into Arabic, but not advanced in any significant way. One notable exception is the Turkish physician Sabunuoglu (b.1385) who provides us with the first description of treating sciatica—​the first description of degenerative rather than traumatic spinal disease. It was not until 1764 that Cotugno postulated that neural compression was the source of sciatica, a model that was later added to by the French neurologists Lasegue, Dejerine, and Sicard. During the eighteenth and nineteenth centuries, the Age of Enlightenment in Europe, the foundations for modern spinal surgery were built. Percival Pott (1713–​1788) described his now eponymous tuberculous spinal disease and washed out a paravertebral abscess with some success. John Bell (1763–​1826) described the flaccid and spastic paralysis and sphincter disturbance associated with spinal cord injury and, with regards to spinal surgery, is quoted as saying ‘the cutting into a vertebra is a dream’ in 1799. Mr Cline Jr, a surgeon at St Thomas’ Hospital, London, is credited with the first ‘trepanation’ of the spine in 1814 on a paralysed man who was paraplegic after falling from a balcony. He performed a laminectomy and removed the facet joints to reduce the dislocation, but the patient died. There are reports from military history of surgeon Louis operating

CHAPTER 1  The history of neurosurgery

on Captain Villedon, who had sustained a spinal cord injury from a thoracic gunshot wound and reportedly regained some distal function postoperatively. Surgery remained controversial due to its high mortality and morbidity until the advent of antisepsis (Lister 1882–​1912 and Semmelweiss 1818–​1865) and safe general anaesthesia. Spinal cord localization mirrored cerebral localization with the early work being carried out by Blesius in 1666 (he described the grey-​white matter differentiation) and the anterior-​posterior spinal nerve roots. The decussation of the pyramids was described by Mistichelli in 1709 and Huber described the denticulate ligaments in 1739. The substantia gelatinosa was described by Rolando in 1809 and Brown-​ Séquard’s description of the decussation of the sensory tracts was published in 1846. Brown-​Séquard was a strong advocate of surgery for spinal cord injury. His contribution to spinal cord localization and his description of the spinal cord tracts is immortalized in his eponymous syndrome of hemisection of the cord. Further work on biomechanics of the spine was carried out in vivo and in vitro by Rauber, Weber, and Messerer in the 1800s. In the late 1800s, spinal surgery was still mostly confined to trauma and infection (TB). Almost two centuries after Cotugno described the neural compression in sciatica, Oppenheim and Kruse described the first surgery for disc herniation (1909). Mixter and Barr fully described the pathogenesis of disc herniation in 1933 and advocated a ‘transdural’ approach to the disc. In 1977 Caspar and Yasargil described the microsurgical intralaminar extradural approach that is practised today. Critical to these developments had been Roentgen’s discovery of X-​rays in 1895, radically improving the diagnosis of spinal fractures. Previously spinal fractures were thought to be almost invariably associated with neurological deficit (90%). Sudeck developed the systematic interpretation of X-​rays and established that most fractures in intact patients were missed and that therefore the true rate of paralysis was more like 15–​20%. The advent of imaging modalities such as CT (1970s) and MR (1980s) had an equally great impact on the development of surgery for degenerative spinal disease. Although a few key spinal operative ‘firsts’ are attributed to nineteenth-​century surgeons, it has been the twentieth century where advances in imaging and materials science has allowed the field of spinal surgery to expand into spinal fixation. Anterior, posterior, and lateral approaches to all regions of the spine were developed. The cervical spine has particular challenges due to its poor biomechanical strength and the proximity of important neural and vascular structures. During the last 50–​100  years, we have seen the progression from the first anterior cervical surgery of Bailey and Badgley in the 1950s who described the anterior cervical discectomy (ACD) without any graft, to the ACD and fusion (ACDF) advocated by Cloward in 1958 and then by Smith and Robinson. The first anterior cervical plate was designed by Orozco and Lovet and required bicortical screws, which are technically challenging to place. Technological and biomechanical advances allowed the introduction of the unicortical locking screw and then the dynamic load-​ sharing plates with variable angle screws which was introduced in 2000. Throughout this time, posterior approaches to the subaxial spine have also benefited from advances in materials science with stronger, lower profile screws and posterior cervical fixation has moved forward from the silver wire used in the first posterior c-​spine fixation in 1891 via interspinous wires (1942 by Rogers), sublaminar

wires (1970), facet wires, and eventually Roy-​Camille’s lateral mass screws in the 1980s. The history of spinal fixation warrants a chapter in itself. Key advances took forward the principles of open reduction and internal fixation used by orthopaedic surgeons for long bone fractures and applied them to the spine. In 1953 Holdsworth and Hardy described a system of plates and screws. Transpedicular screw fixation was described by King in 1944 but not implemented until 1959 by Boucher. In 1958 Harrington described his system of dorsal instrumentation to correct scoliosis—​initially due to polio and then applied to idiopathic. Ventral scoliosis surgery was first described in 1964 by Dwyer. The early focus in disc surgery was to preserve motion and avoid stiffness; this was challenged subsequently when the pathology of disc herniation was found to be secondary to instability and the drive towards fusion became popular. Contrary to this belief, others have focused on developing motion-​preserving disc replacements first proposed by Fernstrom in 1966 and patented (although never implanted) by Froning in 1975. The charité intervertebral disc is now the most commonly used artificial disc. Intradural spinal tumour surgery has progressed along similar lines to cranial surgery, with the first successful removal of an intradural tumour being credited to Horsley in 1887. Elsberg performed the first successful removal of an intramedullary tumour in 1907 and described a two-​stage technique for myelotomy and then tumour removal at a second sitting. The technique of pneumoencephalography developed by Dandy was also useful in the spine and, of course, the introduction of the operating microscope greatly improved the success of intradural and intramedullary resections. Technological advances continue at a rapid pace in spinal surgery. Recent advances in minimal access surgery, image guided screw placement, and endoscopic approaches have all developed in the last 30 years. The thoracoscopic discectomy and fusion of the 1990s has progressed to the treatment of scoliosis and corpectomies using minimal access. Laparoscopic discectomies also developed at a similar time. The father of endoscopy Desormenaux in 1853 could not have foreseen its current myriad applications. Spinal surgery continues to be a rapidly expanding field and the current population demographics mean there will be no shortage of demand for the treatment of spinal trauma, infection or, of course, the degenerative spine.

The future of neurosurgery The last 100 years have brought dramatic advances in the practice of neurosurgery. The future is likely to see a further move away from open cranial surgery to minimally invasive, endoscopic, and radiologically guided interventional techniques.

REFERENCES Bennett, A.H. & Godlee, R.J. (1965). Case of cerebral tumour. In: Wilkins R.H. (ed.) Neurosurgical Classics, pp. 361–​71. New York/​ London: Johnson Reprint Corp. Cushing, H. (1915). Concerning the results of operations for brain tumor. JAMA, 64, 189–​95. Kerr, P.B., Caputy, A.J., & Horwitz, N.H. (2005). A history of cerebral localization. Neurosurg Focus, 18(4), e1.

9

2

Clinical assessment Peter Bodkin and Elizabeth Visser

Introduction The practical process of clinical assessment in modern neurosurgery is slightly different to the traditional model. The ground rules of history taking and examination set down by William Osler and the other forefathers of clinical method rightly remain core to our approach, but we must also take into account the realities of the patient journey before the first encounter with a neurosurgeon. Whether by the bedside or in the outpatient setting, it would be rare nowadays for us not to have already been presented with pertinent background information—​and indeed, often we will have seen high-​quality imaging detailing the patient’s pathology with great anatomical detail. The referring physician will most likely have provided a potential list of differential diagnoses. In other words, the patient has usually been neatly packaged well before we ever get to meet them. Our job, however, is more than simply to supply a surgical fix to a given problem. We are not merely technicians. Although we must take into account the information given to us, we must never take it on face value. All sorts of errors in assessment and incorrect conclusions may have been made along the way. Likewise, imaging is very useful but should never be viewed outside the clinical context. An MRI scan will localize a lesion with more confidence than the best clinician, but technology cannot make us consider a diagnosis of acromegaly from a spongy handshake or common peroneal palsy from the sound of a slapping foot coming to the clinic door. As we gather and make sense of this information there is an equally, if not more, important process at work. By putting the patient’s symptoms into the context of their daily life we begin to develop a relationship with the patient, starting to establish trust and mutual understanding. The rapport built here will be the basis of how a patient measures the success or failure of our interventions. The traditional neurological clinical assessment described in most textbooks has other subtle differences to those required of the neurosurgeon. This chapter aims to cover that broad spectrum of clinical method for the practising neurosurgeon and those in training.

The neurosurgical history The main aim of the neurosurgical history is to gain sufficient information to estimate the anatomical location of the problem and to get

an impression of the pathological process at work. In particular, the time course and severity of the symptoms will be important guides as to whether surgical intervention is required and how quickly. As with any important task, being well prepared will provide a good first impression and will save time in the long run. Referral letters, clinical notes, imaging, and other investigations should have been carefully reviewed. Prior discussions with relevant team members may also be useful. In the clinic, calling the patient from the waiting room yourself can be very valuable. An impression of their social support may be gained by seeing who is with the patient and how attentive they are. Anaesthetic fitness can be crudely assessed by how long it takes to get up and into the clinic room or how out of breath they might be. It is also an opportune moment to make an assessment of gait and a note of walking aids, and so on. Thought should be given to the physical environment for the meeting. This should be arranged such that the patient and doctor are on as level a playing field as possible. One should be aware that being at a higher eye level or sitting behind a desk may have an intimidating effect and will detract from getting the most out of the encounter. When dealing with digital images, viewing platforms should already be opened with relevant images downloaded. Introductions should be clear, giving your name and position. Significant others should be welcomed and acknowledged but it should be made clear that the patient is the focus of discussions. How you open the consultation is important. One should start with open questions, ‘So what’s been the trouble?’, or ‘How can I help you?’ The referral letter or consultation request has a tendency to emphasize the symptoms that will lure you into seeing the patient in the first place. It is wise, therefore, to avoid saying things like, ‘So your doctor has asked me to see you about your facial pain?’ Assumptions can be misleading and may encourage patients to tell you what you are expecting to hear. Once you have encouraged the patient to tell their story, it may be necessary to fill in some gaps. The patient may have painted a picture but there could be large areas missing or fundamental details that are only sketchy. When it comes to the information that is going to influence your decision on treatment keep delving until you feel there is satisfactory detail for a conclusion to be drawn. In the course of this it is important to remember to find out the occupation of the patient and often handedness is pertinent. It is not enough to say a patient

12

Section 1  Principles of neurological surgery

is ‘retired’. A retired university professor will have quite different expectations in life compared to a retired ship builder. The history taking needs to cover time course, anatomical location of symptoms, variability, and character. One must refine questioning with the aim of localizing the lesion or getting clues as to the underlying pathological process. If a patient is not volunteering a symptom, one must consider symptoms that might be associated and ask about them directly. For example, when assessing a patient presenting with spatial disorientation due to a right parietal tumour one needs to remember to ask about problems with their visual fields in case of involvement of the optic radiation, and so on. The most vital part of the story is often the impact that the condition is having on the patient’s way of life. It may seem an unnecessary intrusion to pursue this but most patients are happy to let you know their problems and the knock-​on effects on their home, work, and family life. By establishing that you are not solely interested in dealing with their particular pathological entity but rather you want to help them get back to doing the things they enjoy, you turn a medical interview into a more meaningful conversation. The patient will understand that you are treating them and not just their tumour or slipped disc. Pick up cues. Have an ear for the incongruous. If what you hear doesn’t make sense, explore it. Don’t let it pass unmentioned. It is also useful to find out what has been done so far to address the problem: pain killers, physio, injections, visits to other physicians, and so on. Is the patient fit for an anaesthetic? Are there drugs that need to be stopped prior to surgery? Could an unhealthy lifestyle be contributing to the problem? Drawing the history to a conclusion, the patient should feel that their issues have been adequately addressed. ‘Is there anything else you’d like to discuss?’ is a useful way of allowing any additional information to be voiced. The patient will hopefully be in a more relaxed and open frame of mind by this stage of the interview and might reveal underlying motivations and concerns. Finally, it may be useful to agree on a brief summary and have a final effort to clarify any lingering grey areas or inconsistencies.

The neurosurgical examination Examination of the unconscious patient The approach to the patient who has altered conscious level is obviously limited by the inability of the patient to comply with given instructions and to provide verbal feedback. We are therefore restricted to rather crude and basic bedside tests (i.e. examination of pupillary response to light and the Glasgow Coma Score). Often high-​stake decisions are made on the basis of these assessments and it is therefore crucial that they are performed with utmost care, being mindful of possible confounding factors (Table 2.1). Impairments of pupillary response may be due to damage at a number of locations along the afferent and efferent pathways. Direct trauma to the orbit may cause rupture of the pupil sphincter muscles and produce a traumatic mydriasis. Traumatic optic neuropathy may be due to direct disruption due to penetrating injury or indirectly by shearing forces in blunt head trauma. Lesions in the region of the pretectal nuclei or Edinger–​Westphal nuclei in

Table 2.1  Confounding factors in assessment of GCS Glasgow Coma Score

Confounding factors

Eye opening (4 = spontaneous, 3 = to speech, 2 = to pain, 1 = do not open)

Orbital injuries Ecchymosis Photophobia

Verbal response (5 = oriented, 4 = confused, 3 = inappropriate words, 2 = incomprehensible sounds, 1 = no sound)

Non-​native speakers Maxillofacial injury Endotracheal intubation/​tracheostomy Deafness

Motor response (6 = obeys commands, 5 = localizes to pain, 4 = flexion/​withdrawal to pain, 3 = abnormal flexion to pain, 2 = extension to pain, 1 = no movement)

Spinal cord injury Upper limb fractures, casts, and fixation

Reprinted from The Lancet, Vol 304, Issue 7872, Graham Teasdale, Bryan Jennett, Assessment of Coma and Impaired Consciousness A Practical Scale, pp. 81–​4, Copyright (1974), with permission from Elsevier.

the rostral midbrain will also cause mydriasis. There are complex regulatory pathways from various sources including the ipsilateral hypothalamus that drives sympathetic pupillary tone and indeed poorly understood descending control from the cortex that results in ipsilateral or contralateral mydriasis or miosis following a seizure (Plum and Posner, 2007). Therefore, damage in many brain regions may result in pupillary abnormalities (Fig. 2.1). Compression of the parasympathetic fibres along the third cranial nerve by the herniating uncus is a common cause of pupillary abnormality in neurosurgical practice. The Glasgow Coma Score brings together diverse clinical features to provide a guide to global brain function. Of the three divisions (eye opening, verbal response, and motor response) the motor response is most likely to differentiate the severity of injury. Lesions above the red nucleus produce decorticate posturing (flexion of upper limbs and extension of lower, rubrospinal tract function) and those below produce decerebrate posturing (extension in upper and lower limbs, vestibulospinal tract function). Normal flexion (M4) constitutes flexion with supination; abnormal flexion (M3) constitutes flexion with pronation analogous to decorticate posturing and release of rubrospinal tract function as just described; and extension (M2) is analogous to decerebrate posturing and release of vestibulospinal tract function. It is best to apply painful stimuli to trigeminal territories in case of spinal cord injury causing peripheral numbness. Pressure over the supraorbital notch is sufficient. Rising blood pressure, falling heart rate, and altered respiratory pattern (Cushing’s triad) is a classical response to raised intracranial pressure but is usually a very late, agonal feature.

Examination of language and speech disorders It is important to appreciate any abnormalities of speech and language as this can impact on the history taking, neurological examination, and assessment of higher function and thus alter the outcome of the consultation in general. The ability to correctly identify disorders of speech can aid in localization of neurological pathology. In order for us to communicate through speech and language, hearing, understanding, voice production, articulation, consciousness, thought, and word finding must be intact. Language is a complex

CHAPTER 2  Clinical assessment

Diffuse effects of drugs, metabolic encephalopathy, etc,: small, reactive

Diencephalic small, reactive

Pretectal: large, ‘fixed’, hippus

Ill nerve (uneal): dilated, fixed

Midbrain: midposition, fixed

Pons: pinpoint

Fig. 2.1  Typical pupillary abnormalities associated with anatomical location of damage. Reproduced with permission from Kandel et al., Principles of Neural Science, Fifth Edition, McGraw Hill, New York, Copyright © 2000.

interaction of combinations of sounds, writing, and meaning often linked to a cultural background. To assess understanding start by engaging the patient in normal conversation and asking simple questions:  ‘What is your name? What is or was your occupation? How did you get here today?’ Ensure that the patient can hear you properly and enquire as to what their first language is. Establish if the patient is left-​or right-​handed. As a rule of thumb, 99% of right-​handed individuals have language dominance in their left hemisphere; 60% of left-​handed patients will be left hemisphere dominant; 20% bilateral; and 20% right hemisphere dominant. Now consider the different disorders of language defined here:

Further examinations of these disorders include assessment of spontaneous speech, fluency, naming, repetition, articulation, speech volume, reading, and writing, and will help to localize the pathology.

1. Aphasia—​this is defined as a disorder of spoken language. It is divided into subcategories as follows: 1.1 Non-​fluent aphasia (anterior, motor, or Broca’s); 1.2 Fluent aphasia (posterior, sensory, or Wernicke’s); 1.3 Conduction aphasia; 1.4 Transcortical aphasia (sensory and motor). . Alexia or dyslexia is a disorder of acquired reading ability. 2 . Agraphia or dysgraphia is defined as disorders of written 3 language. . Dysarthria is a disorder of articulation or speech production. 4 . Dysphonia is defined as an abnormality of noise production by 5 expired air over vibrating vocal cords.

2. To assess for alexia or dyslexia, ask the patient to read a sentence or obey a written command. . Agraphia or dysgraphia can be examined by asking the patient 3 to write a sentence; this can only be assessed if there is no motor disability. . Dysarthria is examined by asking the patient to repeat a phrase; for 4 example, ‘red lorry, yellow lorry’ requires intact lingual function and ‘baby hippopotamus’ requires intact labial function. Listen for slurring and rhythm of speech. Dysarthria can be described as spastic (caused by pseudobulbar palsy as in motor neuron disease), extrapyramidal (associated with Parkinsonian syndromes, often associated with dysphonia), cerebellar dysarthria (associated

1. Assessment for aphasia Assess spontaneous speech, fluency, and if the patient uses the wrong words (paraphasia). Ask the patient to name animals or words beginning with ‘F’ in a minute. This tests word finding ability. Also ask them to name familiar objects: a pen, a watch, a tie, and so on. Now ask the patient to repeat phrases. Table 2.2 summarize the findings on examination and localization.

13

14

Section 1  Principles of neurological surgery

Table 2.2  Assessment of aphasia (Clark, 2009; Fuller, 2013) Comprehension

Fluency

Naming

Repetition

Other features

Localization

Non-​fluent aphasia Intact

Non-​fluent

Impaired

Impaired

Right hemiplegia, depressed

Left frontal lobe (inferior and temporal insula), Broca’s area

Fluent aphasia

Impaired

Fluent

Can be intact

Impaired

Neologisms, meaningless speech, paranoid, could have a visual field defect

Posterior superior temporal lobe, Wernicke’s area

Conduction aphasia

Intact

Fluent

Impaired

Impaired

Depressed, cortical sensory loss right arm

Parietal operculum/​ arcuate fasciculus

Global aphasia

Impaired

Non-​fluent

Impaired

Impaired

Right hemiparesis worst in arm

Peri-​Sylvian, both Wernicke’s and Broca’s areas

Nominal aphasia

Non-​fluent

Impaired

Transcortical motor Intact aphasia

Fluent

Impaired

Intact

Halting, effortful speech

Left anterior superior frontal area

Transcortical sensory aphasia

Impaired

Fluent

Can be intact

Intact

Semantic paraphasia

Posterior temporo-​ occipital-​parietal area

Transcortical mixed aphasia

Impaired

Non-​fluent

Impaired

Intact

Angular gyrus

Both Wernicke’s and Broca’s areas

Data from Clarke, C; Howard, R; Rossor, M; Shorvon, S. (2009) Neurology: A Queen Square Textbook. Wiley-​Blackwell (p. 252–​4), Fuller, G. (2004) Neurological Examination Made Easy, 3e. Churchill Livingstone. (p. 17–​25).

with multiple sclerosis, alcohol intoxication, or inherited ataxia), and lower motor neuron dysarthria (this is caused by lesions affecting palatal movement causing nasal speech, tongue movements causing patients to struggle with the letters ‘T’ and ‘S’ and facial movement resulting in difficulty with the letters ‘B’, ‘P’, ‘M’, and ‘W’ and often involves the lower cranial nerves). 5. Finally listen to the volume of speech; if this is reduced it is described a dysphonic speech.

The lobar examination The examination of the functions of the individual lobes of the brain should be a familiar and fluent part of the neurosurgeon’s assessment. One should be aware of the somewhat arbitrary divisions between the lobes, however, and it may well be that one should test more than one lobe if the lesion is on or near a dividing sulcus. Anatomically, Yasargil’s seven lobe system is most satisfactory (frontal, central, parietal, occipital, temporal, insular, and limbic; see Ribas, 2010)  but here we will use frontal, parietal, temporal, occipital, and cerebellum. Frontal lobe As one might expect, as it is the largest lobe, frontal examination has the most components and complexity (Box 2.1). Anatomically, it is useful to consider four distinct regions—​the precentral gyrus, the dorsolateral cortex, the orbitofrontal cortex, and medial cortex (Fig. 2.2). The assessment of the function of the primary motor cortex requires testing for upper motor neurone signs on the opposite side. The patient may adopt postures typical of pyramidal weakness (i.e. flexors stronger than extensors in the upper limbs and vice versa in the lower limbs). Pronator drift is perhaps the archetypal neurosurgical test and is extremely useful in bringing out subtle weakness. Lying anterior to the primary motor cortex lies an area known laterally as the premotor cortex and medially as the supplementary motor area (SMA). The functions of these areas are complex but act

in conjunction with the primary motor cortex. The cortical area immediately anterior to the primary motor cortex (Brodmann’s Area 6) comprises of the lateral premotor area on the lateral aspect and the SMA on the medial and interhemispheric aspect. The lateral premotor area has reciprocal connections with the cerebellum and is involved with refinement of movements with external sensory cues. The SMA has reciprocal connections with the basal ganglia and is involved with initiation of movements from internal sensory cues. In contrast to the primary motor cortex homunculus (leg medial, upper limb, face, and tongue lateral) the SMA homunculus is arranged horizontally (leg posterior adjacent to the paracentral lobule and primary motor area for leg, upper limb, then face and tongue more anteriorly). The SMA has roles in postural stability in walking, initiating and sequencing movements, and coordination of both sides of the body. Frontal lobe ataxia causes a characteristic magnetic gait as if stuck to the floor (Brun’s apraxia). This has similarities to Parkinsonian gait, but does not have the lack of arm swing. It is part of the clinical triad of normal pressure hydrocephalus (disturbance of gait, continence, and cognition). This reflects the anatomical proximity of the micturition inhibitory area (just inferior to

Box 2.1  Scheme for frontal lobe assessment • Ask about handedness and assess speech • Observe behaviour—​abulia, inappropriate dress, verbal dysdecorum • Posture/​gait—​decorticate, ‘magnetic’ • Pyramidal weakness • Saccadic eye movements • Primitive reflexes • Look for urinary catheter • Anosmia and Foster Kennedy syndrome • Neuropsychological tests—​echopraxia, perseveration, conceptualization, working memory

CHAPTER 2  Clinical assessment

Frontal eye fields

Premotor cortex

Micturition inhibitory area

Supplementary motor area Primary motor cortex

Supplementary motor area

Primary motor cortex

Central sulcus

Prefrontal cortex

Prefrontal cortex

Broca’s area (in dominant hemisphere)

Limbic orbitofrontal cortex

Limbic anterior cingulate and orbitofrontal cortex

Sylvian fissure

Limbic orbitofrontal cortex

Prefrontal cortex Orbitofrontal olfactory area

Fig. 2.2  The frontal lobe. Neuroanatomy through Clinical Cases, 1st Edition by Blumenfeld (2002) Fig.19.11 p. 848. By permission of Oxford University Press, USA.

the SMA) and the more diffuse role of the frontal lobe in cognition. Frontal lobe dysfunction may also cause gegenhalten—​the resistance to passive movement.

The frontal eye fields (FEFs, Brodmann’s Area 8)  lie in the posterior part of the middle frontal gyrus and adjacent precentral sulcus. A mass lesion here may result in horizontal conjugate gaze deviation

15

16

Section 1  Principles of neurological surgery

towards the side of the lesion (Prévost or Vulpian sign). Seizures will result in looking away from the lesion. Saccadic eye movement should be tested by asking the patient to look between two fixed points (e.g. a fist and a finger) while keeping the head still. Damage to the FEF may result in impaired saccades away from the lesion. Re-​emergence of primitive reflexes or ‘frontal release signs’ may also be demonstrated. The grasp reflex is assessed by gently passing a finger across the palm. It is most reliably done by distracting the patient by asking them to count backwards from ten. The palmomental reflex is a brief contraction of the ipsilateral mentalis muscle in response to stroking the palm. The snouting reflex is seen when the lips purse in response to tapping the upper lip or the sucking reflex when pressing an object to the lips. Glabellar tap may also cause persistent blinking, whereas in the normal individual it attenuates. It should be remembered that these reflexes may well be seen in normal individuals, particularly in older people (25% of normal adults have the palmomental reflex; see Brazis et al., 2011). The prefrontal area contains a large volume of brain that has complex roles in control of behaviour. One may divide these into restraint (the restriction of behaviour to that which is socially and culturally acceptable), initiative (motivation to put thoughts into action), and order (to sequence tasks appropriately; see Blumenfeld, 2002). Simply observing and talking with the patient will give insights into this. Apparently quite contradictory behaviours may be encountered, some patients lacking any kind of ‘get up and go’ compared to others for whom it is difficult to stop talking, or who are overfamiliar or tactless. In general, abulia is more frequently seen in lesions of the dorsolateral convexity whereas disinhibition is often an orbitofrontal feature. There are many neuropsychological tests that are pertinent to examining the frontal lobes (Box 2.2). Parietal lobe The parietal lobe (Fig. 2.3) may be divided into the postcentral gyrus, posteriorly is the superior parietal lobule and inferior parietal lobule

(angular and supramarginal gyrus). Processing of somatic sensations and perceptions occurs in the postcentral gyrus (monomodal); the posterior parts are polymodal assimilating inputs from somatic, visual, and other sensory modalities mostly for the control of movement especially the hand and upper limb (Kolb and Whishaw, 2009). The left inferior parietal lobule has a role in language and is considered under that heading. Disorders of the parietal lobe will impair sensation. However, as much of somatic sensation is processed in the thalamus there will not be complete numbness but rather more subtle impairments. These may be tested looking for sensory extinction, astereoagnosia, dysgraphasthesia, and two-​point discrimination (Box 2.3). Lesions of the dominant parietal lobe have been associated with a collection of signs known as Gerstmann’s syndrome. Although rare in combination, it is still useful to have these four signs in mind for completeness of examination (Box 2.4). For the non-​dominant parietal lobe (Box 2.5) there is a preponderance for there to be contralateral neglect. This may be noted from the patient’s appearance with lack of grooming on one side. Other features may become apparent such as loss of geographical orientation (getting lost in familiar places), dressing apraxia, and anosognosia (lack of awareness of illness). Temporal lobe The temporal lobes have roles in processing auditory information (Heschl’s gyrus), visual information (inferotemporal cortex), emotion (amygdala), and memory and spatial navigation (hippocampus). Its role in language is important especially in the vocalization (dominant) and perception of (non-​dominant) emotion in language (emotional prosody). It should be recalled that Meyer’s loop passes over the roof of the temporal horn and lesions may, therefore, cause a ‘pie in the sky’ quadrantanopia. Apart from simple tests of memory, speech, and visual fields there are limited bedside tests for the temporal lobe. Neuropsychologists may perform dichotic listening tests where different recordings are presented to either ear of a pair of headphones to assess selective auditory attention or may carry out advanced tests of verbal and non-​verbal memory.

Box 2.2  Neuropsychological tests of frontal lobe function

Occipital lobe

Luria’s three-​step test This tests motor sequencing. Tell the patient you are going to show him a series of hand movements. Without verbal prompting show a sequence of fist, edge, palm, and repeat five times. Ask the patient to do the same. The patient may demonstrate perseveration with repetition of the same movement or be quite unable to do the sequence in order.

Vision and its interpretation is the primary function of the occipital lobe and is dealt with in that section.

Go/​no-​go test This tests ability to shift set. Hold two fingers out palm down—​‘put out one finger when I do this’, hold down one finger—​‘put out two fingers when I do this’. Do this several times. Then change the instructions. Place two fingers—​ ‘put out one finger when I do this’. Place one finger—​‘do nothing when I do this’. The inability to follow the second set of instructions implies a deficit. Verbal fluency Produce as many words beginning with a particular letter in one minute, proper nouns not allowed. Normally 12 or more. Tower of Hanoi A game moving discs between stacks aiming to achieve goal in as few moves as possible Wisconsin card sorting test This is a card matching test testing executive function. Stroop test Read out a list of coloured words (i.e. the word green spelt out in yellow text). Test of restraint.

Cerebellum The cerebellum coordinates smooth, planned motor actions by analysing extensive sensory inputs from the brain and spinal cord. The midline structures (the vermis and flocculonodular lobes) control coordination of trunk and eye movement. The cerebellar hemispheres maintain control of limb movement and aid motor planning. Because cerebellar outputs remain ipsilateral or cross twice, lesions of a cerebellar hemisphere will cause an ipsilateral deficit. Ataxia is a characteristic sign of cerebellar dysfunction. Movements are clumsy due to poor coordination between agonist and antagonist muscle groups. There is imprecision of trajectory though space (dysmetria) and also in timing of movements. Truncal ataxia may make even sitting up in bed very difficult. Gait may be wide-​based and staggering. Appendicular ataxia may be tested by asking the patient to repeat rapid alternating movements of the limbs (e.g. supinating/​ pronating one hand on the other). When abnormal this is known as dysdiadochokinesia. One can also ask the patient to outstretch their arm, then touch their nose, or to touch the examiner’s finger and back to their nose. An intention tremor may be also elicited. Holmes

CHAPTER 2  Clinical assessment

Postcentral sulcus and gyrus

Superior parietal lobule

Posterior paracentral lobule

Intraparietal sulcus

Top of the central sulcus

Cingulate sulcus (marginal branch)

Precuneus

Supramarginal gyrus

Angular gyrus

Parietooccipital sulcus

Calcarine sulcus

Subparietal sulcus

Fig. 2.3  Parietal lobe. This article was published in Gross Anatomy and General Organization of the Central Nervous System in Nolte’s The Human Brain, John Nolte, Copyright Elsevier (2009).

rebound test demonstrates overshooting. For example, with arms outstretched and eyes closed, the examiner pushes one arm downward. On release the patient’s arm shoots up higher than originally placed. There is ipsilateral hypotonia and a pendular knee jerk may be found. The cerebellum has a role in articulation and when damaged will cause dysarthria with laboured, slurred speech. Mutism is a well-​recognized problem following surgical resections in the midline, well documented in children following medulloblastoma resection. The anatomical basis is from disruption of a dentato-​ rubro-​thalamo-​cortical pathway such that bilateral dentate nucleus injury causes mutism. This explains why dominant hemisphere SMA injury, bilateral thalamotomies and mesencephalic strokes (red nucleus) can result in the same syndrome. Cerebellar mutism was originally, mistakenly, attributed to approaches that split the vermis given the close proximity of the dentate nuclei to the midline. Box 2.3  Examination of parietal lobe function (either hemisphere)

Nystagmus may also be present with the fast phase towards the abnormal side. Vertical nystagmus (e.g. downbeat nystagmus may be seen in Chiari malformation). Head tilt may occur in children with posterior fossa lesions. Cognitive-​affective symptoms are increasingly being recognized in cerebellar disorders. Box 2.4  Parietal lobe examination—​dominant hemisphere (Gerstmann’s syndrome) Dyscalculia Ask the patient to subtract 7 from 100 and continue subtracting 7 sequentially. Agraphia Ask the patient to write a simple sentence. Finger anomia and left-​right disorientation These two can be examined together by crossing your hands and asking, ‘Which finger am I wiggling?’ (Fig. 2.4) or alternatively asking the patient to touch their right ear with their left ring finger. Remember to also assess speech (see earlier).

Sensory extinction Ask patient to hold out arms with eyes shut. Touch either one or both sides of the corresponding part of the body and ask where he has been touched. Extinction occurs when the patient says that only one side is being touched when in fact it is both. Stereognosis With the patient’s eyes closed, place a familiar object in their hand and ask them to identify it. Coins of different denominations may be used. Graphesthesia Ask the patient to identify the number or letter that you trace on their palm. It should be agreed which way is up before starting. Two-​point discrimination Using callipers or a bent paper clip, ask the patient if they can feel one or two points. On the fingertips one should be able to recognize two separate points to about 2–​4 mm apart, on the palm 8–​15 mm. Visual fields Examine for homonymous inferior quadrantanopia. Constructional apraxia Ask patient to copy a 3D drawing.

Fig. 2.4  Testing finger anomia and left-​right disorientation: ‘Which finger am I wiggling?’ Correct answer: ‘Your left ring finger.’

17

18

Section 1  Principles of neurological surgery

Box 2.5  Parietal lobe examination—​non-​dominant hemisphere Unilateral spatial neglect Ask the patient to mark the middle of a horizontal line. Displacement of the centre towards the side of the brain lesion (generally right hemisphere) indicates neglect. Target cancellation tests can also be used. This is where a patient is asked to circle every ‘a’ on a page or something similar. Only one side of the page will be attended to. Crossed response test Ask patient to move the limb opposite the one that is touched (motor neglect). Dressing apraxia Ask the patient to take off a jumper or other item of clothing. Turn it inside out and ask the patient to put it back on the right way. Paper cutting Ask patient to use a pair of scissors to cut out a shape from a piece of paper.

The visual system examination This section will detail the examination of the eye and cranial nerves II, III, IV, and VI. A thorough history is necessary to understand the patient’s range of visual symptoms. These may include drooping eyelids, blurred vision, double vision, ‘seeing things’, visual loss (transient or persistent, partial, or complete), abnormal movements of the visualized world, eye pain, headache, or orbital pain. In this section we will provide an examination example that avoids missing the most important eye signs. The neuro-​ophthalmological examination should include: 1. Visual acuity Optic nerve, chiasmal, optic tract lesions, and ocular pathology can all influence visual acuity. This is assessed with corrected vision or refractive vision, like the patient’s own glasses or a pin hole. The standard is the Snellen chart that measures vision at a distance of 6 metres. This is expressed as a fraction

L

Visual field defects

R

(distance in metres from chart/​distance in metres at which letters should be seen). If the acuity is reduced but is correctable with refraction; this is due to an ocular defect. If the acuity is not correctable, then it signifies a problem in the visual pathway. If the patient is unable to see the largest print, the chart can be brought closer or assessed by finger counting, perception of hand movements, or perception of light. 2. Colour vision This is most useful for assessing the optic nerve function. It is assessed by using Ishihara plates and scored in each eye individually by the number of plates correctly identified. The speed of identifying the plate should also be considered when comparing the eyes. Remember that 8% of males and 0.5% of females may have X-​linked recessive congenital colour deficiency or dyschromatopsia. In these patients the loss of colour vision will be bilateral with normal visual acuity and fields. A cruder method of assessing colour vision is to ask the patient to look at a coloured target like a red cap and in the affected eye it may seem faded or ‘washed out’. . Visual field testing 3 Visual fields are tested to direct confrontation and each eye is assessed individually. The patient should fixate on the examiner’s nose and cover each eye in turn. This test can detect hemianopia, quadrantanopia, altitudinal, and central field defects. With the eye covered ask the patient if all the parts of the examiner’s face are clear, or if there are parts that are blurred or missing. Ask the patient to count fingers in each quadrant. Peripheral fields are best assessed with a white hat pin, whereas central fields and blind spots should be tested with a red hat pin. The latter is assessed against the examiner’s blind spot by moving the pin. Visual field defects are mapped out according to the patient’s description. Abnormalities are illustrated as follows (see Fig. 2.5) and can localize the defect in the visual pathway All patients with a suspected visual field Visual fields

L

R

1

Retina

5 1

Optic nerve Optic chiasma

3 2

Optic tract

3 4

Lateral geniculate body Optic radiation

4 5

6

Lower fibres in temporal lobe 5

6

Upper fibres in anterior pareital lobe All fibres in posterior pareital lobe

Occipital cortex

Fig. 2.5  Assessment of visual fields and localization of defects. From: Macleod’s Clinical Examination, Ninth edition fi­ gure 7.8 page 211, J. Munro and C.R.W. Edwards.

CHAPTER 2  Clinical assessment

Box 2.6 Parinaud’s syndrome • Impairment of upgaze • Large irregular pupils that do not react to light but can accommodate (light-​near dissociation) • Eyelid abnormalities—​lid retraction or ptosis • Impaired convergence • Nystagmus retractorius

defect should have formal perimetry for accurate localization and monitoring. Macular sparing with ischaemic occipital lobe lesions relates to the dual blood supply to the occipital pole from both the middle cerebral artery as well as the posterior cerebral artery. 4. Pupils The pupillary light reaction pathways include the optic nerve (afferent) and the parasympathetic component of the third cranial nerve (efferent). Accommodation arises from the frontal lobes (afferent) and the parasympathetic component of cranial nerve III (efferent). Inspect first for anisocoria, and then ask the patient to fixate on a distant target and then test the direct and indirect light reflexes, the pupillary reactions to accommodation and finally perform the swinging light test. 4.1 Size The examination of the pupils should first be conducted in room light, if the anisocoria is greater in light than in dark, the parasympathetic system is abnormal, and the larger pupil is abnormal. A  cranial nerve III paresis, Adie’s pupil, or damaged iris sphincter are examples of the parasympathetic pathway. Then examine for anisocoria in dim light, if this is greater in dim light, a

sympathetic dysfunction is present and the smaller pupil is abnormal. Horner’s syndrome is a manifestation of a sympathetic disorder. If the anisocoria is the same under both conditions, it is not indicative of a neurological problem. Conditions that impair light response but do not affect accommodation (light-​near dissociation) include Parinaud’s syndrome (Box 2.6), neuro-​syphilis, diabetes mellitus, Adie’s pupil, bilateral optic neuropathy, and aberrant regeneration of cranial nerve III. 4.2 Light response (direct and indirect) Shine a bright light in each eye, and observe the pupils for speed and magnitude of constriction (see Fig. 2.6). If the pupil does not react normally to light, examine the response by viewing a near target. Again, check for speed and magnitude of constriction to a near target and the speed of dilatation when looking at a distant target. 4.3 Swinging light test To perform this test the lights should be dimmed while the patient should fixate on a target in the distance. Swing the light from eye to eye for about a second at a time. Look for the initial movement in each pupil (normally constriction). To diagnose a relative afferent pupil defect (RAPD), one pupil should consistently dilate rather than constrict. The presence of a RAPD is indicative of optic neuropathy (Table 2.3). 5. Ophthalmoscopy For the purposes of this chapter we advise that ophthalmoscopy should be performed on all patients to examine for papilloedema that can suggest raised intracranial pressure and optic pallor suggestive of optic atrophy. Systematically examine the optic disc, the bloods vessels, and then the retinal background. 2

Pretectal nucleus

Action potentials from right eye reach both right and left pretectal nuclei.

Oculomotor nerves (III) Ciliary ganglia

1

3

Light is shone on right eye only. 4

The right and left sides of the Eddinger–Westphal nuclei generate action potentials through the right and left oculomotor nerves, causing both pupils to constrict.

Fig. 2.6  The light reflex. Authored by: OpenStax College. Provided by: Rice University. Project: Anatomy & Physiology. http://​cnx.org/​content/​col11496/​latest/​

The pretectal nuclei stimulate both sides of the Eddinger–Westphal nucleus even though the light was perceived only in the right eye.

19

20

Section 1  Principles of neurological surgery

Table 2.3  Relative afferent pupil defect (Pane, 2007) Normal pupil Light stimulus

Right pupil

Left pupil

None

•

•

Left

•

•

Right

•

•

Left relative afferent pupil defect Light stimulus

Right pupil

Left pupil

None

•

•

Left

•

•

Right

•

•

This article was published in The Neuro-​ophthalmology Survival Guide, A Pane, p 379, Copyright Elsevier (2007).

Table 2.4 summarizes some findings that can aid in the localization of lesions that present with visual symptoms using the techniques described so far. 6. Eye movements 6.1 Fixation, nystagmus, and saccades Disorders of fixation include nystagmus and saccadic intrusions. The examination will stem from a history of involuntary eye movements or oscillopsia. Congenital fixation disorders are usually asymptomatic and incidentally noticed. Nystagmus is the rhythmic oscillation of the eyes. Start the examination with eyes in the primary position, then different positions of gaze. Note the fast phase of the nystagmus and the direction or if it is pendular. Also note if the nystagmus is similar in both eyes and if there is a latent component or increased nystagmus with one eye covered. Different types of nystagmus can aid in localization of pathology as summarized in Table 2.5. Assessing how rapidly and accurately patients can fixate on an eccentric target by testing saccades can expose a subtle internuclear ophthalmoplegia or sixth cranial nerve palsy. It is tested by asking the patient to look at two different objects on either side of the patient’s head. The patient is asked to alternate gaze between the objects. Note saccadic initiation, velocity, and accuracy. This can be either too small (hypometric) or too large (hypermetric).

6.2 Eye movements and cranial nerves III, IV, and VI Eye movements include saccades controlled by the frontal lobe, pursuit (the slow movement that facilitate fixation on a moving object) that is controlled by the occipital lobe, the vestibulo-​ocular reflex (that allows compensation of eye position for movement of the body or head) controlled by the cerebellar vestibular nuclei and convergence (fixating on an object close to the face) that is controlled by the midbrain. Input from the different control centres have to be integrated to allow synchronous eye movement. The medial longitudinal fasciculus (MLF) in the midbrain runs between the nuclei of cranial nerves III and IV in the midbrain and VI in the pons. The eye muscles are the lateral rectus controlled by cranial nerve VI, the superior oblique controlled by cranial nerve IV, and all the rest are controlled by cranial nerve III. 6.2.1 Cranial nerve III palsy (oculomotor nerve) A complete palsy is a triad of ptosis, a large and unreactive pupil, and the eye position of ‘down and out’ (Fig. 2.7). Pupil sparing implies that the deeper nerve fibres have been affected most likely from ischaemia. The parasympathetic fibres lie circumferentially on the outside of the nerve and are more commonly affected by compressive causes, classically a posterior communicating artery aneurysm. Internuclear ophthalmoplegia is the result of a lesion of the MLF. The patient has dysconjugate eye movements; there is incomplete adduction of one eye and jerky nystagmus of the other eye on abduction when testing lateral gaze. It is described as left-​sided when there is failure of left adduction when looking to the right (Fig. 2.8). The horizontal conjugate gaze centre resides in the paramedian pontine reticular formation at the level of the fifth cranial nerve nucleus. On an attempt to move the eyes in a horizontal direction, the abducting eye is able to move (VI) but the signal to the third nerve is unable to pass through the MLF to the third nerve is nucleus and there is a failure of adduction. The resulting diplopia may underlie the nystagmus in the abducting eye as there is an attempt to overcome this. The vertical gaze centre resides in the rostral interstitial nucleus of the MLF (riMLF) at the level of the third nerve/​superior colliculus.

Table 2.4  Localization of lesions in the visual system (see Beck and Smith, 1988) Optic nerve

Optic chiasm

Optic tract

Temporal lobe

Parietal lobe

Occipital lobe

Visual acuity

Normal or reduced

Normal or reduced

Normal or reduced

Normal

Normal

Normal

Colour vision

Normal or reduced

Normal or reduced

Normal or reduced

Normal

Normal

Normal

Visual field

Central scotoma

Bitemporal

Homonymous incongruous

Homonymous superior

Homonymous inferior or complete

Homonymous exquisitely congruous

Relative afferent pupil defect

Present

Present or absent

Present or absent

Absent

Absent

Absent

Disc pallor

Present or absent

Present or absent

Present or absent

Absent

Absent

Absent

Beck & Smith, Neuro-​Ophthalmology: A Problem-​Oriented Approach, 1e, Little Brown & Co, USA, Copyright © 1987.

CHAPTER 2  Clinical assessment

Table 2.5  Types of nystagmus and neurological localization Type of nystagmus

Localization

Down beat

Craniocervical junction

Periodic alternating

Craniocervical junction

Gaze-​evoked

Vestibular, cerebellum

Up beat

Cerebellum, medulla

Seesaw

Diencephalon, mesencephalon

Torsional

Central vestibular

Convergence–​retraction

Dorsal midbrain

Rebound

Cerebellum

Beck & Smith, Neuro-​Ophthalmology: A Problem-​Oriented Approach, 1e, Little Brown & Co, USA, Copyright © 1987.

6.2.2 Cranial nerve IV palsy (Trochlear nerve) This presents with diplopia on looking down and head tilt away from the side of the palsy with no clear squint (Fig. 2.9). The pupil of the affected eye may lie slightly higher than the normal eye, but this is not obvious. On suspecting a fourth nerve palsy, it is useful to ask the patient to look at a horizontal object (e.g. the top of the door frame alternately with each eye). A trochlear palsy will produce two images angled towards the abnormal side (Fig. 2.10). 6.2.3 Cranial nerve VI palsy (Abducens nerve) This presents with diplopia with two images with horizontal separation (Fig. 2.11). There is a squint, diplopia away from the weak muscle. The pupil is normal. 7. Lid position Inspect for the eyelid position, contour, and measure the palpebral fissures keeping the eyes in the primary position. The upper lid usually covers 1 mm of the cornea, more than that indicates ptosis. Also examine for variable ptosis in different directions of gaze as seen in congenital ptosis and Duane’s syndrome. Fatigable ptosis (>2  mm after 2 minutes) on up gaze is suggestive of neuromuscular weakness such as myasthenia gravis. . Orbits 8 Finally, during inspection of the orbits one should look for proptosis, enophthalmos, or ocular injection. One should stand behind the patient to assess proptosis looking down over the forehead. A  Hertel exophthalmometer provides a more accurate measurement. It may be useful to auscultate over the closed eyelid for a bruit that could signify a carotid-​cavernous fistula.

R

L

Abnormal

command

Normal

Fig. 2.8  Left internuclear ophthalmoplegia. From Fuller, G. (2004) Neurological Examination Made Easy, 3e. Churchill Livingstone, p. 88, Fig 9.6.



command

Left IVth nerve palsy

Fig. 2.9  Left fourth nerve palsy. From Fuller, G. (2004) Neurological Examination Made Easy, 3e. Churchill Livingstone, p. 88, Fig 9.4.

Fig. 2.10  ‘Door frame’ test of patient with left fourth nerve palsy—​two images pointing to abnormal side.

R

L

command

Left IIIrd nerve palsy

Left VIth nerve palsy

Fig. 2.7  Left third nerve palsy.

Fig. 2.11  Left sixth nerve palsy.

From Fuller, G. (2004) Neurological Examination Made Easy, 3e, Churchill Livingstone, p. 88, Fig 9.4.

From Fuller, G. (2004) Neurological Examination Made Easy, 3e. Churchill Livingstone, p. 88, Fig 9.4.

21

22

Section 1  Principles of neurological surgery

Examination of the cranial nerves This section will include the examination of the cranial nerves not discussed in the section on eye examination. 1. Cranial nerve I  (olfaction). Ask if the patient has noticed any change in smell. Examine with two different smells (e.g. coffee or orange). Olfactory test kits are available. . Cranial nerve II (typically acuity, fields, and pupillary function). 2 A complete examination would also include fundoscopy, assessment of blind spot, and colour vision. . Cranial nerves III, IV, and VI (eye movements). 3 . Cranial nerve V (sensation to the face and muscles of mas4 tication). The trigeminal nerve has both motor and sensory components: the maxillary division is the nerve of the first branchial arch derivatives. The motor examination is performed by palpating the muscles of mastication when the patient clenches

the jaw and look for jaw movement when asked to open the mouth against resistance. Also tap the jaw jerk. The sensory examination is performed by testing all three divisions of the trigeminal nerve (ophthalmic, maxillary, and mandibular divisions) with light touch. Lastly, perform the corneal reflex (afferent limb Va, efferent limb VII) by gently touching the cornea with a small piece of cotton wool. The approach to the cornea should be lateral to the field of vision. An alternative method is to gently blow in the eye and assess response. Also note the normal conjunctival reaction of lacrimation and injection. 5. Cranial nerve VII (facial expression and taste to anterior two-​ thirds of the tongue). The facial nerve innervates the muscles of facial expression, stylohyoid, stapedius, and the posterior belly of the digastric muscle, and is the nerve of the second branchial arch (Fig. 2.12). When facial weakness occurs it is important to decide if the weakness is upper (UMN) or lower motor neuron

Stapedius muscle

Nervus intermedius Internal Greater Geniculate acoustic petrocal ganglion meatus nerve

Lacrimal glands

Superior salivatory nucleus Facial nucleus Nucleus solitarius Spinal trigeminal nucleus

Sphenopalatine ganglion

Chorda tympani Petrotympanic fissure Posterior auricular branch Taste, anterior 2/3 of tongue

Temporal branch Zygomatic branch

Submandibular ganglion

Buccal branch Mandibular branch

Sublingual gland

Cervical branch Submandibular gland

Fig. 2.12  Facial nerve—​motor, parasympathetic, and special sense innervation. Blumenfeld, Hal. (2002) Neuroanatomy through Clinical Cases. Sunderland, Massachusetts: Sinauer. Fig 12.10, ­chapter 12 p. 480.

Stylomastoid foramen Lingual nerve

CHAPTER 2  Clinical assessment

(LMN). The different clinical features are explained by the bilateral innervation from the nuclei of the facial nerve supplying the superior facial muscles. Firstly, look for asymmetry in facial movement, which can be very subtle: just a hint of incomplete eye closure, slow blinking, or delay in a grin. Then ask the patient to raise the eyebrows (frontalis, temporal branch), close the eyes tightly (orbicularis oculi, zygomatic branch), blow out the cheeks (buccinators) or whistle (orbicularis oris, both buccal branch), show your teeth (zygomaticus major, buccal or marginal mandibular branch), wrinkle the chin (mentalis, marginal mandibular branch), and wrinkle the neck (platysma, cervical branch). The severity of facial palsy is classified using the House–​ Brackmann scale (see Chapter 24 on the surgical management of cerebellopontine angle lesions). In UMN weakness the forehead is seen to be stronger than the lower face, also called sparing of the forehead. Unilateral weakness can be due to vascular lesions, demyelination, tumours, or brain stem lesions. Bilateral weakness can be due to pseudobulbar palsy or motor neuron disease. In LMN weakness, the lower face is seen to be as weak as the forehead. Bell’s palsy is a common cause of this and is usually worst within 12 hours from the onset and can be associated with pain around the mastoid, loss of taste, and hyperacusis. Other causes of LMN facial weakness are pontine lesions, cerebellopontine angle lesions, and parotid tumours. Causes of bilateral LMN facial weakness are disorders of the neuromuscular junction like myasthenia gravis, motor neuron disease, myopathies, neurosarcoidosis, or Guillain–​Barré syndrome. 6. Cranial nerve VIII (auditory and vestibular nerves). The auditory examination is conducted by testing each ear individually. Cover the ear on the opposite side of the ear being tested. Test the hearing by rubbing your fingers together or whispering. If there is reduced hearing, perform Rinne’s and Weber’s tests. For these tests you will need a 512 Hz tuning fork. Rinne’s test compares bone versus air conduction. The tuning fork is struck and the base is placed on the mastoid process (bone conduction). When the tone disappears, the tuning fork prongs are placed over the external auditory meatus (air conduction). The normal response is for the tone to reappear when the tuning fork is moved in front of the external auditory meatus (i.e. air conduction is better than bone conduction, AC>BC). This reflects the amplification provided by the tympanum and middle ear ossicles. Weber’s test involves striking the tuning fork and placing the base over the midline over the forehead. The tone should be heard equally in both ears. If it localizes to one side this can reflect a sensorineural hearing deficit on the contralateral side, or a conductive hearing loss on the ipsilateral side with an increased sensitivity in that ear. These tests can be difficult to interpret and the formal investigation is pure tone audiometry, and with sensorineural hearing loss imaging of the internal auditory canal. To examine vestibular function, examine the gait and look for nystagmus. Other tests of value are the Hallpike’s test used in patients with positional vertigo. This is performed with the patient sitting up, then turning the head to one side, lie the patient back quickly with their head extended, the examiner should support the patient’s head. Observe for nystagmus, associated delay, and also if it fatigues. Ask if the patient experiences vertigo. Repeat this on both sides. If the nystagmus

is fatigable and delayed, this is in keeping with peripheral vestibular disturbance such as benign paroxysmal positional vertigo. If not fatigable, this is caused by a central abnormality. Unterberger’s test is performed by asking the patient to stand facing the examiner with their arms stretched out in front of him. Ask the patient to close their eyes and march on the spot. Watch for any rotation. Patients turn to the side of vestibular pathology. Only perform this if the patient is able to stand safely. Romberg’s test is to assess proprioceptive loss, not specifically balance or vestibular function. 7. Cranial nerve IX and X (palatal movement, swallow, and gag reflex). The glossopharyngeal and vagus nerves are examined together. Firstly, listen to the voice quality for evidence of bulbar weakness, then ask the patient to cough. Inspect the position of the uvula and then ask the patient to say ‘Aah’. If there is weakness, the uvula will deviate away from the side of the lesion. The gag reflex is rarely performed and should not be performed if there is any doubt about their safety to swallow as there is potential for aspiration. Video fluoroscopy is safer in such patients. . Cranial nerve XI (power to sternocleidomastoid and trapezius). 8 The accessory nerve is examined by testing the power of the trapezius muscles by asking the patient to shrug their shoulders, noting any asymmetry in the initial movement and on inspection. Winging of the scapula may be seen with weakness of trapezius (although serratus anterior palsy from damage to the long thoracic nerve is more common). Test the power and bulk of the sternocleidomastoid muscles by asking the patient to rotate the head to the opposite side against resistance. . Cranial nerve XII (tongue power and movements). Examination 9 of the hypoglossal nerve should start with inspection of the tongue lying in the floor of the mouth. Note any wasting or fasciculations. When the patient is asked to protrude the tongue, it will deviate towards the weak side or the side of the unilateral lesion. The speed and fluidity of the movement will be affected in bilateral pyramidal lesions. Power is assessed by asking the patient to press against resistance from the examiner’s finger on the outside of the cheek by pressing the tongue firmly on the inside of the same cheek.

Examination of autonomic function and dysfunction An understanding of how the autonomic system is organized is required to interpret clinical signs that may be encountered during head and neck examination. Sympathetic fibres arise from the preganglionic cell bodies of the lateral horns of T1–​4, form the sympathetic chain and distribute postganglionic fibres via the inferior (to vertebral artery, C7–​8), middle (to inferior thyroid artery and C5–​6), and superior cervical ganglia (to carotid artery and C1–​4). This chain lies posteromedial to the carotid sheath and anterior to the longus muscles. The postganglionic fibres then ride along the external carotid to supply the ciliary ganglion and hence the dilator of the pupil, and the external carotid to supply the submandibular ganglion (submandibular and sublingual glands) and the otic ganglion (parotid gland). A  lesion to first, second, or third-​order sympathetics will result in Horner’s syndrome where there is pupil constriction, partial ptosis, and loss of hemifacial sweating (Fig. 2.13 and Box 2.7). The remainder of the sympathetic

23

24

Section 1  Principles of neurological surgery

Ipsilateral eye

Müller’s muscle

Postero-lateral hypothalamus Dilator pupil Internal carotid artery and carotid sympathetic piexus Vasomotor and sweat gland nerve fibres (ipsilateral face)

Brainstem

External carotid artery Superior cervical ganglion

First order neuron

Second order neuron Stellate ganglion

Cilispinal centre (C8-T2)

T1 Nerve root

Fig. 2.13  The sympathetic supply of the pupil and causes of Horner’s syndrome (Kong, 2007). Reproduced with permission from Kong, YX, Wright, G, Pesudovs, K, O’Day J, Wainer Z, Weisinger H, Horner syndrome, Clinical and Experimental Optometry, volume 90 issue 5, pp. 336–​44. Copyright ©2007 Optometrists Association of Australia and John Wiley & Sons.

supply to the head and neck travels with the internal carotid artery: the deep cervical nerve forms over the petrous segment of the carotid, continues into the middle fossa through the foramen lacerum, and joins the parasympathetic greater superficial petrosal nerve (GSPN) to form the Vidian nerve. The Vidian nerve passes through its own canal to supply autonomic fibres to the pterygopalatine ganglion and distribute parasympathetics with branches of the maxillary division of the trigeminal nerve. Parasympathetic fibres are carried by four cranial nerves:  III, VII, IX, and X. There are four equivalent brainstem parasympathetic nuclei: Edinger Westphal nucleus (III); superior salivatory nucleus (VII); inferior salivatory nucleus (IX); and the dorsal vagal nucleus (X). The vagus provides no cranial parasympathetic supply but provides supply to all the viscera of the thorax and abdomen as far as the midgut (two-​thirds of the transverse colon). The preganglionic parasympathetic fibres of the remaining three cranial nerves synapse in four ganglia:  ciliary ganglion (III via nerve to inferior oblique); pterygopalatine ganglion (VII via the GSPN); submandibular ganglion (VII via the chorda tympani); and otic ganglion (IX via lesser superficial petrosal nerve). Thus, the facial nerve has two parasympathetic branches that originate at the brainstem level within the nervus intermedius (Fig. 2.12). Crocodile tears syndrome is an uncommon consequence of damage to the facial nerve where there is misdirection of regenerating gustatory fibres to the lacrimal gland resulting in tearing when presented with food. The postganglionic parasympathetic fibres are then delivered to the target glands by a branch of the trigeminal nerve. Fibres

from the ciliary ganglion pass along the short ciliary nerves (branch of nasociliary nerve, Va) to control pupillary constriction and accommodation. Fibres from the pterygopalatine ganglion pass to the lacrimal gland via the zygomaticotemporal (Vb) and lacrimal (Va) nerves. The remainder of the fibres from the pterygopalatine ganglion supply the nose, palate, and nasopharynx sinuses via branches of the maxillary division of the trigeminal nerve. The fibres from the submandibular ganglion supply submandibular and sublingual glands via the lingual nerve (Vc). Finally, the fibres from the otic ganglion supply the parotid gland via the auriculotemporal nerve (Vc).

Examination of the peripheral nervous system: Upper and lower limbs The neurological examination of the extremities is based on inspection, assessment of tone, power, reflexes, coordination, and sensation. Having been directed by the history further localization will rely on key examination findings (e.g. confirmation of long tract signs in the patient with suspected myelopathy, correlating deficits in corresponding myotomes and dermatomes in radiculopathy, testing spinothalamic versus dorsal column function in possible syringomyelia, differentiating signs of peripheral neuropathy versus radiculopathy in brachialgia or sciatica, and so on). Understanding the differences between upper and lower motor neurone signs is of course crucial. With this in mind there should be a logical process of elimination, the order of which is open to personal choice and circumstances.

CHAPTER 2  Clinical assessment

Box 2.7  Differential diagnoses for Horner syndrome

Box 2.8 Gait abnormalities

First-​order neuron lesions Arnold–​Chairi malformation Basal meningitis (e.g. syphilis) Basal skull tumours Cerebrovascular accident/​lateral medullary syndrome Demyelinating disease (e.g. multiple sclerosis) Intrapontine haemorrhage Neck trauma Pituitary tumour Syringomyelia

• Hemiplegic—​spastic leg traces a semicircle due to fixed plantarflexion and extension at the knee • Diplegic—​narrow-​based, scissoring gait often due to cerebral palsy • Myelopathic—​broad-​based clumsy  gait • Equine—​high stepping gait seen in patients with foot drop due to L5 radiculopathy or common peroneal palsy • Myopathic—​waddling gait with pelvis dropping on alternating sides • Parkinsonian—​slow little steps marche à petits pas • Cerebellar—​veering uncontrolled gait • Magnetic—​feet seem to be stuck to the floor, seen in normal pressure hydrocephalus

Second-​order neuron lesions Pancoast tumour Birth trauma with injury to lower brachial plexus Cervical rib Aneurysm/​dissection of aorta, subclavian, or common carotid artery Central venous catheterization Trauma/​ surgical injury (radical neck dissection, thyroidectomy, carotid angiography, coronary artery bypass graft, upper spine chiropractic manipulation) Chest tube insertion Lymphadenopathy (e.g. Hodgkin’s disease, leukaemia, tuberculosis, mediastinal tumours) Mandibular tooth abscess Lesions of the middle ear (e.g. acute otitis media) Neuroblastoma Lumbar epidural anaesthesia Third-​order neuron lesions Internal carotid artery dissection Cluster/​migraine headaches Carotid artery thrombosis Carotid-​cavernous fistula Herpes zoster Orbital apex tumour Idiopathic Reproduced with permission from Kong, YX, Wright, G, Pesudovs, K, O'Day J, Wainer Z, Weisinger H, Horner Syndrome, Clinical and Experimental Optometry, volume 90 issue 5, pp. 336–​44. Copyright ©2007 Optometrists Association of Australia and John Wiley & Sons.

Inspection There should be an initial general inspection noting features such as abnormal posturing, or preferential use of one limb over another. Stigmata of related systemic disease such as rheumatoid arthritis may be apparent. Assessment of gait may be done at this point. Characteristic pathological gaits point to particular neuro-​ anatomical disorders (Box 2.8). Heel-​toe (tandem) walking and Romberg’s test are useful adjuncts. Classically, Romberg’s test is positive only with disorders of proprioception. Notably, this is not a test of cerebellar function as ataxia would be present with or without closing the eyes with cerebellar disorders. Inspection should continue with examination for atrophy of muscle groups, fasciculations, tremors, and scars such as previous carpal tunnel decompression. Tone and reflexes The resistance of muscle groups across a joint is a result of a complex system involving feedback from muscle spindles and Golgi tendon organs, monosynaptic reflexes with agonist and antagonist alpha motor fibres, inhibitory supraspinal control, and processing via the cerebellum. Tone can be decreased (hypotonia) or increased

(spasticity or rigidity). Spasticity is a velocity dependent increase in muscle tone in response to a passive stretch, with exaggerated tendon jerks, in association with other features of the upper motor neurone syndrome (Lance, 1980). At the end of range of movement there is often a characteristic ‘give’ (clasp knife). Rigidity is not velocity or force dependent and lacks any ‘give’. Spasticity is more often associated with corticospinal damage whereas rigidity is generally extrapyramidal in origin. Assessment may be by gentle internal and external rolling of the leg, brisk raising of the knee, or pronation/​ supination at the wrist. The deep tendon reflexes are tests of a monosynaptic arc. The afferent neurone is stimulated by activity of a Golgi tendon organ; the efferent neurone is an alpha motor neurone. The most frequently tested reflexes are the biceps (C5), supinator (C6), triceps (C7), knee (L3/​4), and ankle (S1) jerks. The patient should be in a relaxed posture. A tendon hammer should be used to briskly strike the tendon. If absent, reinforcement can be used by asking the patient to clench the teeth or by interlocking the fingers against resistance. Specific tests of upper neurone dysfunction include Hoffmann’s sign, the plantar response, and clonus. A positive Hoffmann’s sign is when quickly flicking the distal phalanx of the middle finger downwards causes flexion of the thumb. Clonus should be tested with the knee flexed, and anything more than five beats of the ankle is pathological. Power Muscle groups may be tested in several ways. One can ask the patient to maximally contract the muscle and then attempt to counteract it (isometric). Alternatively, one can attempt to counteract the muscle group through a range of motion (isotonic). The Medical Research Council (MRC) scale provides a reproducible grading system (see Table 2.6). The emphasis placed on examining different muscle groups will depend if one is looking for pyramidal weakness, long tract disease, radiculopathy, or a peripheral nerve problem. When

Table 2.6  MRC grading of power Grade 5 Grade 4 Grade 3 Grade 2 Grade 1 Grade 0

Normal power Submaximal movement against resistance Movement against gravity but not against resistance Movement with gravity eliminated Flicker of movement No movement

Used with the permission of the Medical Research Council.

25

Fig. 2.14  International standards for neurological classification of spinal cord injury workbook.

CHAPTER 2  Clinical assessment

testing the nerve roots or level of cord dysfunction, the American Spinal Injuries Association workbook (ASIA, 2015) provides an excellent, logical schema (Fig. 2.14). For each level, a particular action is specified to test individual nerve roots. Sensation As for the motor examination, sensation should be tested either in a radicular manner or according to peripheral nerve distribution (Fig. 2.15). In addition to peripheral localization, there are different modalities to be considered. Importantly, one should remember that discriminative light touch, proprioception, vibration, and two-​ point discrimination are conducted through large fast-​conducting fibres in the dorsal columns. Pinprick (superficial pain), and temperature are conducted through smaller, slower fibres in the spinothalamic tracts. Again, the ASIA chart is an invaluable guide as to the optimal points for testing sensation. An exact point on the skin is indicated where light touch and pin prick should be tested. It is vital to compare one side against the other as one proceeds. Areas of hypoaesthesia or hyperaesthesia should be carefully mapped out. Asking the patient to close their eyes may provide additional objectivity. On testing joint position sense, it is important to remember that the patient may be able to tell the direction of motion by how pressure is exerted on a digit. One should, therefore, ensure that

the thumb or the great toe are held side to side rather than above and below. Should proprioception be lost, one should test the next proximal joint. Examination techniques for some common peripheral neuropathies are detailed in Boxes 2.9, 2.10 and Table 2.7.

Putting it all together As one progresses with the history and examination the information should constantly be distilled, therefore refining the line of questioning and narrowing the focus of examination. Eventually, one has to draw these disparate pieces of information together, make sense of them, and attempt to identify the most likely anatomical source of the patient’s complaint and potential pathological processes that may have caused it. Once this point has been reached, it is necessary to decide on whether further investigations are required and what they might consist of. In the course of this decision-​making process, one needs to be rigorous in ensuring strictest objectivity. Doctors . . . are more interested in patients they can help, and diseases they can cure, than the ones they can't. There is, therefore, a tendency for them to make the diagnosis fit their skills or even . . . to have a vested interest in inventing illnesses which they can cure. (Handy, 1990)

Fig. 2.15  Cutaneous distribution of peripheral nerves (Harrison and Kasper, 2015).

27

28

Section 1  Principles of neurological surgery

Box 2.9  Examining for carpal tunnel syndrome • Atrophy of thenar eminence • Test opposition of thumb by making an O sign • Test abduction of thumb, lay back of hand down and point thumb upwards • Splitting of sensation of the ring finger—​more sensation on ulnar side • No loss of sensation in central palm as the palmar cutaneous branch runs above flexor retinaculum • Phalen’s sign (good predictive value) • Tinel’s sign (poor predictive value) • Absence of benediction sign (attempts to make fist, but thumb and index finger remain extended)—​this is due to more proximal compression of median nerve

Box 2.10  Examining for ulnar neuropathy at the elbow • Wasting of first dorsal interosseous muscle and hypothenar eminence • Wartenberg’s sign—​little finger remains abducted, has a tendency to catch on pockets • Test finger abduction • Ulnar paradox—​ulnar claw hand seen with ulnar nerve damage at the wrist, not present with more proximal lesion at the elbow • Froment’s test—​test of adductor pollicis, the only muscle supplied by the ulnar nerve in the thenar eminence. With hands supine ask patient to hold onto piece of paper. Flexion of the thumb on withdrawing the paper indicates positive test • Tender and enlarged ulnar nerve in the ulnar groove, positive Tinel’s test • Splitting of sensation of the ring finger—​more sensation on radial  side • Remember to rule out C8 radiculopathy, Pancoast’s tumour, and thoracic outlet syndromes (look for Horner’s syndrome to help differentiate)

Table 2.7  Clinical features differentiating L5 radiculopathy and common peroneal palsya,b

a

L5 radiculopathy

Common peroneal palsy

Pattern of weakness

Foot inversion (tibialis posterior, the foot invertor, supplied by L4 and L5 via tibial nerve, not the peroneal nerve) Hip abduction (gluteus medius and minimus supplied by L4/​L5/​S1 via superior gluteal nerve)

Foot eversion

Nerve root tension signs

On hip flexion

On ankle inversion

Tinel’s test at fibular neck

Negative

Positive

Always remember central causes for foot drop, especially a parafalcine meningioma pressing on the motor cortex. b For peripheral causes the frequencies are as follows: peroneal neuropathy (46%); lumbar radiculopathy (15%); sciatic nerve disorder (5%) ( Jeon, 2013).

In considering a patient for an operation there is always a balance to be struck. What are the risks of surgery versus the risk of doing nothing? A clear understanding of natural history is required just as much as the potential complications of a particular operation. The history and examination provide the foundation stones of patient management. What one does with this information must be guided by our knowledge of neurosurgical conditions. This is described in subsequent chapters.

FURTHER READING Blumenfeld, H. (2002). Neuroanatomy Through Clinical Cases. Sunderland, MA: Sinauer Associates. Brazis, P.W., Masdeu J.C., & Biller J. (2011). Localization in Clinical Neurology, 6th edition. Philadelphia PA: Wolters Kluwer/​Lippincott Williams & Wilkins Health. Fuller, G. (2013). Neurological Examination Made Easy, 5th edition. Edinburgh, UK: Churchill Livingstone. Kolb, B. & Whishaw, I.Q. (2009). Fundamentals of Human Neuropsychology. New York: Worth Publishers. Pane, A. (2007). The Neuro-​ophthalmology Survival Guide, 1st edition. London, UK: Mosby Elsevier, p. 379.

REFERENCES American Spinal Injuries Association (ASIA) (2015). International Standards for Neurological Classification of Spinal Cord Injury. https://​asia-​spinalinjury.org/​information/​downloads/​ Beck, R. & Smith, C. (1988). Neuro-​ophthalmology:  A Problem-​ orientated Approach, 1st edition. Boston, MA:  Little, Brown and Company, p. 14. Blumenfeld, H. (2002). Neuroanatomy Through Clinical Cases. Sunderland, MA: Sinauer Associates, p. 848. Brazis, P., Biller, J., & Masdeu, J. (2011). Localization in Clinical Neurology. Philadelphia PA: Wolters Kluwer/​Lippincott Williams & Wilkins Health, p. 535. Clarke, C. (2009). Neurology, 1st edition. Chichester, UK: Wiley, pp. 252–​4. Handy, C. (1990). Inside Organizations, 1st edition. London, UK: BBC Books. Harrison, T. & Kasper, D. (2015). Harrison’s Principles of Internal Medicine, 19th edition. New York [u.a.]: McGraw-​Hill Medical, p. 160. Jeon, C., Chung, N., Lee, Y., Son, K., & Kim, J. (2013). Assessment of hip abductor power in patients with foot drop. Spine, 38(3), 257–​63. Kolb, B. & Whishaw, I. (2009). Fundamentals of Human Neuropsychology. New York: Worth Publishers. Kong, Y.X., Wright, G., & Pesudovs, K., et al. (2007). Horner syndrome. Clin Exp Optom, 90(5), 336–​44. Lance, J.M. (1980). Pathophysiology of spasticity and clinical experience with baclofen. In:  Lance, J.W., Feldman, R.G., Young, R.R., Koella, W.P. (eds.) Spasticity: Disordered Motor Control, pp. 185–​204. Chicago, IL: Year Book Medical Publishers. Plum, F. & Posner, J. (2007). Plum and Posner’s Diagnosis of Stupor and Coma, 1st edition. Oxford, UK: Oxford University Press, p. 58. Ribas, G. (2010). The cerebral sulci and gyri. Neurosurg Focus, 28(2),  1–​24.

CHAPTER 2  Clinical assessment

RELATED LINKS TO EBRAIN An Approach to the Cognitively-​impaired Adult. https://​learning. ebrain.net/​course/​view.php?id=37 Examination of the Limbs. https://​learning.ebrain.net/​course/​view. php?id=49 Examining Patients in Coma. https://​learning.ebrain.net/​course/​view. php?id=44 Neurological Assessment:  Movement Disorders. https://​learning. ebrain.net/​course/​view.php?id=38 Neurological Assessment:  Weakness and Sensory Loss. https://​ learning.ebrain.net/​course/​view.php?id=36 Neurological Examination:  Cranial Nerves. https://​learning.ebrain. net/​course/​view.php?id=51

Neurological Examination:  Gait. https://​learning.ebrain.net/​course/​ view.php?id=47 Neurological Examination:  Speech. https://​learning.ebrain.net/​ course/​view.php?id=45 Neurological Examination:  Visual Function. https://​learning.ebrain. net/​course/​view.php?id=50 Neurological History:  General Approach and Common Pitfalls. https://​learning.ebrain.net/​course/​view.php?id=39 Neuromuscular Assessment. https://​learning.ebrain.net/​course/​view. php?id=41 Principles of Neurological Investigation. https://​learning.ebrain.net/​ course/​view.php?id=43

29

3

Overview of neuroimaging Tomasz Matys, Daniel. J. Scoffings, and Tilak Das

Introduction The development of computed tomography (CT) (Hounsfield, 1995)  and magnetic resonance imaging (MRI) (Mansfield and Maudsley, 1977; Lauterbur, 1989) in the last 40 years has revolutionized neuroimaging, overcoming the limitations of previously used methods that relied on the demonstration of displacement of intracranial vessels or ventricles, and instead allowing the direct visualization of intracranial pathology (Hoeffner et al., 2012). The first clinical CT examination was performed on a patient with a brain tumour and it could be argued that no other field of medicine has benefited more from the further development of cross-​ sectional imaging techniques than neuroscience. Refinement of MRI and nuclear medicine methods, with the introduction of positron emission tomography (PET) in particular, has allowed for imaging of selected aspects of central nervous system function, such as blood flow, tissue microenvironment, and metabolism. Future progress will no doubt include continuing improvement in the spatial and soft tissue resolution of anatomical imaging and extending the application of functional methods, but also hopefully clinical translation of the full scope of molecular imaging, allowing visualization of specific cellular and molecular events underlying the disease process in addition to its pathological and functional sequelae (Hoffman and Gambhir, 2007; Massoud and Gambhir, 2007). In this chapter we discuss current neuroradiology imaging modalities that are useful in neurosurgical practice. Due to space constraints, information regarding the underlying physical principles is limited to the basics, while more comprehensive descriptions can be found elsewhere (Allisy-​Roberts and Williams, 2007). We focus here on the general usefulness and limitations of neuroradiological methods rather than the imaging manifestations of individual disease processes, which are discussed elsewhere in the relevant chapters of this book. The interested reader can find further information in dedicated neuroradiology monographs (see Nadgir and Yousem, 2016).

Principles of imaging Radiographs and fluoroscopy Radiographs (‘plain X-​rays’) are produced by passing a collimated beam of X-​rays through a patient onto an image receptor, traditionally a cassette containing combination of a fluoroscopic screen and radiographic film, but in modern practice a digital detector is typically used. The tissues in the patient’s body attenuate the X-​ray beam to differing degrees so that the intensity of the X-​rays reaching the detector varies across its surface. In this way radiographs can distinguish between air (black), fat (dark grey), soft tissue and water (light grey), and bone (white). The radiographic image is also a two-​ dimensional representation of a three-​dimensional structure and it is typically necessary to obtain images in different (often orthogonal) projections to infer the position of a given abnormality within the patient. With the increasing availability of CT and MRI, the role of conventional radiographs in the assessment of patients with neurosurgical disease has reduced, but they retain a role in certain circumstances. The advantages of radiographs include their ease of acquisition, low cost, widespread availability, the ability to obtain images at the bedside, and also to obtain a degree of dynamic information by acquiring radiographs with the patient in different positions (e.g. standing, in flexion, and extension of the spine). Although they have high spatial resolution, a major disadvantage of radiographs is their limited soft tissue resolution. Indications for radiographs include assessment of the continuity of the extracranial components of ventricular shunts, the position and integrity of spinal implants and prostheses, and the follow-​up of patients with spinal fractures in whom loss of normal alignment or progression of vertebral body compression is suspected. Radiographs are less sensitive than CT in the detection of spinal fractures and have a very limited role in the primary assessment of suspected spinal injury. Flexion-​extension radiographs are often used in evaluation of potential instability in patients with degenerative or spondylolytic spondylolisthesis.

32

Section 1  Principles of neurological surgery

Fluoroscopy provides real-​time images of the area of anatomical interest by passing a continuous or pulsed X-​ray beam through the patient onto an image-​intensifier, most commonly using a C-​arm apparatus which can be rotated and positioned around the patient to provide the desired view. Uses of fluoroscopy include the identification of the correct level for spine surgery, monitoring the placement of spinal fixation devices or prostheses, and the performance of lumbar puncture in patients in whom bedside lumbar puncture (LP) has been unsuccessful due to obesity or degenerative changes.

Computed tomography CT basics and techniques Introduced into clinical practice in 1973, CT was the first diagnostic imaging modality to enable direct visualization of the brain (Wolpert, 2000). Similar to plain film radiography, CT is based on a differential attenuation of X-​rays by tissues, but uses a differently collimated beam sweeping around the patient in a circular fashion and reaching a large number of individual detectors; the image is then created using mathematical reconstruction algorithms. In brief, when an X-​ray photon travels along a particular line through the patient’s body, it experiences energy loss that depends on attenuation coefficients of tissues it encounters on its path. By measuring photon energy from multiple directions around the object, it is possible to calculate the attenuation coefficients for each voxel (volume element) within the examined volume. Attenuation coefficient values are then converted into more convenient-​to-​use radiodensity values (‘CT numbers’) calculated in relation to water and air, and expressed in Hounsfield units (HU). The scale of CT numbers ranges from –​1000 HU (air) to over 3000 HU (cortical bone), with 0 HU being the attenuation of distilled water. Typical values for different central nervous system (CNS) tissues are given in Table 3.1. The way in which the multidirectional information is obtained and the speed of acquisition have dramatically changed with technological advances, from a pencil-​like X-​ray beam translating along a slice in the original machine to a fan-​shaped beam and increasing number of detectors along the circumference of the gantry and across multiple rows (Ginat and Gupta, 2014). An important development in CT scanning technology was the advent of helical or spiral CT, where the data is obtained in continuous fashion while the patient moves smoothly through the gantry and the X-​ray tube rotates, continuously tracing a spiral trajectory around the body.

While this mode of acquisition allowed significant improvements in body CT scanning, it introduces unwanted ‘windmill artefacts’ (Barrett and Keat, 2004), and sequential CT may still be preferable for head imaging. In neuroradiological applications, helical CT does allow excellent quality CT angiography for the evaluation of neck and intracranial vessels, as well as isotropic resolution volumetric reconstructions that can be visualized in any arbitrary plane. To convert raw CT data into a useful image, the matrix of radiodensities is displayed as shades of grey. From the range of the Hounsfield scale, possible CT numbers span up to 4000 levels of radiodensity, yet the human observer can theoretically perceive approximately 720 shades of grey in optimal conditions (Kimpe and Tuytschaever, 2007) and this is much lower in practice. It is therefore not possible to map the entire dynamic range of CT numbers to greyscale at the same time, necessitating the use of a ‘window’ centred at a certain level and of particular width to display a suitable fragment of this range. Values of radiodensities within the window are displayed as shades of grey, while values below and above the window boundaries are displayed as black and white, respectively. Because most tissues of interest in the brain lie in the range of 0 to 100 HU, these are optimally depicted using a window centred at a level of around 40 HU and width of 80 HU (Fig. 3.1A). As the radiodensity of acute haemorrhage lies close to the upper limit of the standard window, haemorrhage appears bright and is readily apparent; it is however difficult to appreciate blood immediately adjacent to bone. Using ‘subdural windows’ with at a slightly higher level and a larger width (e.g. L/​W 70/200) makes acute blood better distinguishable from the adjacent bone (Fig. 3.1B). Setting the window level to negative values allows differentiation of fat from air. To interrogate bone detail, the window needs to be set to a higher level and width (e.g. L/​W 500/3000, Fig. 3.2). One of the most important parameters that affect image quality is the kind of mathematical reconstruction algorithm, also referred to as a reconstruction kernel or filter. Use of different kernels changes the balance between spatial resolution and image noise. A smooth kernel generates images with lower noise and better low-​contrast detectability more suited to brain examinations, at a cost of reduced spatial resolution. A sharp kernel generates images with high spatial

(a)

(b)

Table 3.1  Attenuation values of the main tissues/​materials in neuroimaging Material

Hounsfield units

Acute blood

56 to 76

Air

−1000

Bone

1000 to 3000

Calcification

140 to 200

Cerebrospinal fluid

0

Fat

−30 to −100

Grey matter

32 to 41

White matter (centrum semiovale)

23 to 34

This table was adapted from Neuroradiology: The Requisites, David Yousem, Robert Zimmerman, Robert Grossman, Copyright Elsevier (2010).

Fig. 3.1  Standard window (A, level/​width 40:80) provides good differentiation of the grey and white matter but high-​density haemorrhage adjacent to the skull is difficult to appreciate. Widening the window (B, ‘subdural window’ 70:200) makes the thin right sided subdural haemorrhage (arrow) more conspicuous.

CHAPTER 3  Overview of neuroimaging

(a)

(b)

administered intrathecally through a lumbar, cervical, cisternal, or ventricular approach (Nadgir and Yousem, 2016); this may require lower concentration compound approved for intrathecal use, especially in children. Advantages and disadvantages of CT

Fig. 3.2  Image of the skull seen on bone window (L/​W 500:3000) reconstructed using standard (A) and bone (B) kernel. Hairline fracture of the left squamous temporal bone is much more conspicuous using bone kernel reconstruction (arrow).

resolution better suited to the assessment of bony structures. Use of bone kernel is essential for detection of fractures (Fig. 3.2) and should be a part of any CT performed in the context of trauma, as well as sinonasal and temporal bone examinations. CT contrast media and risks Iodinated contrast media have historically been the contrast agent of choice for X-​ray and CT applications due to high attenuation of iodine compounds thanks to its high atomic number (Lusic and Grinstaff, 2013). In CT brain imaging, contrast media are used to detect areas of blood-​brain barrier breakdown, where contrast medium leaks into parenchyma, or to opacify blood vessels in CT angiography, including venography. The risks of iodinated contrast media include allergic reactions and contrast-​induced acute kidney injury. To minimize this, it is important to identify susceptible individuals (Box 3.1) and take appropriate preventative steps (Iodinated Contrast Media Guideline, Royal Australian and New Zealand College of Radiologists). The current standard is to use non-​ionic, low osmolarity compounds. The incidence of severe reactions with this type of media is 0.04% and for very serious reactions is 0.004% (Hunt et al., 2009). A typical adult dose of intravenous contrast medium for head CT is 50 ml of a compound containing 300 mg/​ml iodine. Larger amounts are used in angiographic applications. Contrast medium can also be

Box 3.1  Risk factors for contrast-​induced acute kidney injury (CI-​AKI) and allergic reaction to contrast media Risk factors for CI-​AKI Known kidney disease (including kidney transplant) Presence of diabetes Metformin (risk of lactic acidosis) Risk factors for allergic reaction to contrast media Previous allergic reaction to iodinated contrast media Previous significant allergic reactions to other substances or history of eczema History of asthma Use of beta blockers The Royal Australian and New Zealand College of Radiologists. Iodinated Contrast Media Guideline. Sydney: RANZCR; 2018.

CT plays an extremely important role in neuroimaging, particularly in the emergency and neurosurgical settings (Nadgir and Yousem, 2016). It is a rapid, easily accessible, and efficient modality for screening patients with major traumatic injury. It has high sensitivity to haemorrhage, particularly in the first 24 hours compared to MRI, and can demonstrate features of raised intracranial pressure and brain herniation, allowing for rapid decision-​making in early stages of traumatic injury. For the detection of subarachnoid haemorrhage (SAH), it remains the initial imaging study of choice in suspected cases and can be rapidly followed by CT angiography when SAH is identified. Similarly, it remains a useful technique for assessment of cerebral infarction and, in conjunction with CT angiography and CT perfusion, for guiding thrombolytic therapy in patients with suspected ischaemic stroke (Kidwell and Wintermark, 2010). For postoperative examinations in both brain and spine, CT is useful in evaluating for complications (such as haemorrhage, infarct, or hydrocephalus), and the position of ventricular drains or other adjuncts. Portable head CT scanners are now available allowing examination at a neurocritical care bed without the need for transporting an unstable patient. CT venography can often be used as an alternative to MR venography or help with troubleshooting in the case of confounding artefacts on MRI. CT is the best modality for assessing bone lesions and is essential for the evaluation of skull, skull base, facial, and spine fractures. It is also the technique of choice for primary evaluation of the temporal bones and paranasal sinuses. CT sensitivity for calcification is helpful with the diagnosis of calcium-​containing CNS tumours (e.g. meningioma, oligodendroglioma, craniopharyngioma), metabolic disorders (e.g. hyperparathyroidism), and congenital lesions (e.g. TORCH infections, tuberous sclerosis). Finally, CT myelography is a viable alternative to MR in spine imaging in patients with absolute contraindications to MR imaging. The use of radiation remains a drawback for CT technology. For a head CT, the average effective dose is approximately 2 millisievert (mSv), compared to an annual background radiation of 2–​ 5 mSv. Other neuroradiological examinations incur higher doses (Table 3.2). Possible effects of radiation can be divided into deterministic (predictable and depending on cumulative radiation dose, for example, cataract due to lens irradiation), and stochastic (chance-​like, for example, induction of malignancy) (Allisy-​Roberts and Williams, 2007). The risk of cataract should be taken into account, especially in children (Michel et al., 2012). In pregnancy, the dose to the fetus from head and neck CT is low and the risk of inducing childhood cancer is thought to be < 1 in 1 000 000. Examinations with direct exposure of the fetus (e.g. spinal CT) result in higher risk (1 in 1000–​10 000) and should be avoided (The Royal College of Radiologists, 2009). The administration of iodinated contrast agents carries a theoretical risk of fetal thyroid suppression and thyroid function should be checked in the first week after birth.

33

34

Section 1  Principles of neurological surgery

Table 3.2  Typical radiation dose in neuroradiological CT-​based examinations (Cohnen et al., 2006; Mettler et al., 2008) Examination

Average dose (range)

CT head

1.7 mSv

CT angiogram head

1.9 mSv

CT angiogram neck

2.8 mSv

CT perfusion cerebral

1.1–​5.0 mSv

Comprehensive stroke protocol

Up to 9.5 mSV

CT spine

6 mSV

Data from Cohnen, M et al., Radiation exposure of patients in comprehensive computed tomography of the head in acute stroke, AJNR. American Journal of Neuroradiology, volume 27, issue 8, pp. 1741–​5. 2006, and Mettler, F.A et al., Effective doses in radiology and diagnostic nuclear medicine: a catalog, Radiology, volume 248, issue 1, pp. 254–​63.  2008.

Advances in CT technology include attempts to reduce the necessary radiation dose while maintaining acceptable image quality, particularly important for paediatric imaging. The main way of reducing radiation exposure is adapting the dose by changing the tube voltage and current (and therefore X-​ray photon energy and beam intensity) according to the patient’s size, weight, and specific imaging application. A different method that allows scanning with a lower dose while maintaining image quality is replacement of the conventional reconstruction technique, known as filtered back projection, with iterative reconstruction (IR) which requires additional computation time but generates images with lower noise and higher spatial resolution, as well as reduced beam hardening and metal artefacts. Although the effects are more pronounced in body imaging, dose reductions have also been demonstrated for head imaging (Kilic et al., 2011; Mirro et al., 2016). The images appear subjectively different between the two reconstruction techniques and IR is therefore still under clinical evaluation.

Magnetic resonance imaging (MRI) MRI basics and techniques MRI takes advantage of the phenomenon of nuclear magnetic resonance, inherent to nuclei that have a magnetic moment due to an uneven number of protons or neutrons; in practice this mainly involves imaging of hydrogen nuclei (1H) containing a single proton. MRI is based on a radiofrequency signal that is emitted when protons aligned with a strong magnetic field are tipped out of alignment by an externally applied radiofrequency pulse and then return to equilibrium (Fig. 3.3). The emitted signals are read out in a spatially ordered fashion and reconstructed into an image reflecting the magnitude of signal in a given voxel. The type of excitation pulse, time at which the signal is read out, and interval between excitations can be varied resulting in a different contrast in the final image (Allisy-​ Roberts and Williams, 2007). Two principal MRI sequences provide image weighting that depends on T1 or T2 relaxation times of the tissue (Table 3.3); appearances of the brain on the most commonly used sequences are shown in Figure 3.4. Images weighted according to proton density (PD) can be obtained together with T2-​weighted images with no additional scanning time required; they provide good contrast between the grey and white matter, and are of value in certain situations, but are not universally used in neuroradiology. Water has high signal on T2-​weighted sequences, and because most pathological processes are associated with increased water content

(oedema), they appear hyperintense on T2 weighted images. Fat is hyperintense on T2-​weighted images but unlike water it shows high T1 signal. Besides fat, T1 hyperintensity can also be due to melanin, protein-​rich fluid, calcification, and gadolinium-​chelate contrast agents (see later). Haemorrhage has complex appearances on MRI with combination of signal intensities dependent on the degradation phase of blood products (Table 3.4). Signal in MRI can be produced using spin echo or gradient recalled echo sequences (Bitar et al., 2006; Allisy-​Roberts and Williams, 2007). Spin echo provides better signal-​to-​noise ratio and is generally preferable, but slower than gradient echo. Acquisition of large volumetric datasets, required for example by neurosurgical navigation systems, is therefore performed using a variant of the gradient echo sequence. An important feature of gradient echo sequences is the lack of compensation of magnetic field inhomogeneities, resulting in its sensitivity to local susceptibility effects. This is used in practice for detection of haemorrhage, as the local susceptibility effect caused by iron in haemosiderin results in a very low signal with ‘blooming’ on gradient echo sequences. Even better depiction of haemosiderin, for example, microhaemorrhages in the setting of traumatic brain injury, is achieved with susceptibility weighted imaging (SWI), a high-​ resolution 3D velocity-​compensated long echo time gradient echo sequence (Di Ieva et al., 2015). Gradient echo sequences are also used in other situations in which image contrast depends on susceptibility effects such as dynamic susceptibility contrast perfusion imaging and blood-​oxygenation level dependent methods. Image contrast in MRI can be varied using suppression techniques designed to null signal from a specific tissue. In the brain, this is used in fluid-​attenuated inversion recovery (FLAIR), which nulls the cerebrospinal fluid signal; it is routinely used with T2-​weighting to make parenchymal T2 hyperintensity more conspicuous (T2-​FLAIR, usually referred to simply as FLAIR). This should not be confused with a less common T1-​FLAIR sequence, which is used for example to improve grey-​white matter contrast that is otherwise reduced at higher magnetic field strengths. Fat suppression techniques increase conspicuity of fluid signal on T2-​weighted images or contrast enhancement on T1-​weighted images in situations where they could be masked by the presence of fat (such as detection of oedema in the vertebral column or paraspinal soft tissues, or contrast enhancement at the skull base, respectively). Fat suppression can be achieved by short tau inversion recovery (STIR) or spectral saturation. STIR is useful near bone-​containing structures (orbit, skull base, sinuses), metallic foreign bodies, and across large fields of view, for example, in spine imaging. STIR is however problematic when combined with contrast medium, when spectral fat saturation is used instead. Balanced steady state free precession (SSFP) sequences provide high resolution and excellent signal-​to-​noise ratio with a mixed T2/​T1-​ weighting. In neuroimaging, these sequences are usually performed with a heavy T2-​weighting delivering high contrast resolution between cerebrospinal fluid and contained structures, such as the cisternal segments of the cranial nerves and adjacent blood vessels. Balanced SSFP techniques such as fast imaging in steady state (FIESTA) or constructive interference in steady state are therefore useful in investigation of vascular loops as the cause underlying trigeminal neuralgia or hemifacial spasm, as well as in screening for cerebellopontine angle tumours. Other uses include the evaluation of inner ear structures, and identification of cerebrospinal fluid leaks (Saindane, 2015).

CHAPTER 3  Overview of neuroimaging

(a)

(b)

(c)

(d)

(e)

RF

Mz

t=0

Mxy

Mxy

Mz

63% 37%

t = 0 T2

Dephasing

t=0

T1 Longitudinal relaxation

Fig. 3.3  Due to having a charge and spin, protons display magnetic properties and behave like magnetic dipoles. Outside of a magnetic field, the dipoles are randomly distributed and the net magnetization is zero in any direction in space. When placed in a magnetic field, the magnetic dipoles align with the field, either parallel (‘spin up’) or antiparallel (‘spin down’) to it, and start to precess (wobble around their axes) with a frequency dependent on the strength of the magnetic field (Larmor frequency). (A) Because a small excess of dipoles (approximately three per million) point ‘spin up’ their longitudinal magnetization vectors add up creating net longitudinal magnetization Mz. Due to precession, each dipole also creates a transverse magnetization vector rotating in the transverse plane. However, because the dipoles precess in different phases, their transverse magnetization vectors cancel each other and the net transverse magnetization remains zero. (B) Application of an external radiofrequency (RF) pulse turns some dipoles spin down; a so-​called 90°RF pulse turns exactly half of the excess dipoles spin down, so now equal numbers of dipoles point in either direction and the net Mz vector is null. The RF pulse also causes synchronization of the phase of precession, so all the dipoles precess together (they achieve phase coherence) and their transverse magnetization adds up to produce a net transverse magnetization vector Mxy, rotating in the transverse plane. (C–​E) After the RF pulse ends, within a short time the protons return to their initial state by losing phase coherence (dephasing, resulting in a loss of transverse magnetization Mxy) and by returning to the spin up orientation (longitudinal relaxation, resulting in ‘regrowth’ of the longitudinal magnetization vector Mz). These processes happen exponentially with time constants T2 and T1, respectively—​T2 indicates the time after which dephasing is complete in 63% (so 37% phase coherence remains), and T1 corresponds to the time at which 63% of the longitudinal magnetization has recovered (bottom panels). T2 is always shorter and dephasing is complete (D) before longitudinal relaxation (E). T1 and T2 times depend on the type of the tissue and determine tissue contrast on T1-​and T2-​weighted sequences. Tissues with short T1 appear bright on T1-​weighted sequence, while long T2 results in low signal on T2-​weighted sequence (e.g. fat has high signal on T1-​and T2-​weighted sequences due to short T1 and relatively long T2).

Table 3.3  Signal intensity of different substances and tissues on  T1-​and T2-​weighted MRI sequences Substance

T1-​weighted sequence

T2-​weighted sequence

Water/​Cerebrospinal  fluid

↓

↑

Fat

↑

↑

Air

↓

↓

Cortical bone

↓

↓

Red bone marrow

=/​↑

↑

Yellow bone marrow

↑

↑

Diffusion-​weighted imaging (DWI) is sensitive to the movement of water molecules. DWI weighting is usually achieved by applying special diffusion-​ encoding gradients along the three principal spatial directions, and combining the signal into a trace image. Because the DWI trace image has both diffusion weighting and T2-​weighting, apparent hyperintensity can be due to restricted diffusion, or ‘shine-​through’ phenomenon if the underlying area is T2-​hyperintense. It is therefore essential to review the apparent diffusion coefficient map (ADC), which eliminates the effect of T2 weighting and demonstrates restricted diffusion as low signal intensity. Restricted diffusion may reflect a reduction in the size of the extracellular space that can be a result of cell swelling (e.g. in ischaemic stroke) or high cellularity (e.g. in lymphoma or

35

36

Section 1  Principles of neurological surgery

(a)

(b)

(c)

Fig. 3.4  Appearances of the brain on routine MRI sequences. (A) On T1-​weighted image grey matter is darker (‘greyer’) than white matter. (B) On T2-​weighted image the white matter is darker than grey matter and fluid is hyperintense. (C) T2-​FLAIR show similar contrast between the grey and white matter as T2-​weighted images but the signal from CSF is nulled.

medulloblastoma), as well as in lesions such as cholesteatoma or epidermoid cyst. In neurosurgical practice, the most important application of DWI is differentiation of pyogenic abscess from other ring-​enhancing lesions, with the contents of the abscess demonstrating markedly restricted diffusion. Applying diffusion gradients in at least six different directions underlies diffusion tensor imaging (DTI) which enables calculation of the predominant direction of diffusion in a given voxel. Because the movement of water is relatively unrestricted along the white matter tracts, DTI allows calculation of nerve fibres’ direction and reconstruction of white matter tracts (DTI tractography), useful in neurosurgical planning (Waldman et al., 2009; Saindane, 2015). DTI also enables estimation of several measures of diffusion anisotropy with promising results in imaging and prognostication in traumatic brain injury (Hulkower et al., 2013) and high-​grade gliomas (Price et al., 2007; Yan et al., 2016). Functional MRI (fMRI) is based on blood oxygen level dependent (BOLD) effect resulting from different magnetic properties of oxyhaemoglobin (diamagnetic) and deoxyhaemoglobin (paramagnetic). Areas of the cortex with increased neuronal activity demonstrate higher oxygen consumption, but also increased perfusion due to neurovascular coupling; the effect of increased perfusion predominates, and the net result is a relative increase in MRI signal. Different activation paradigms allow the individual mapping of important cortical functions such as sensorimotor cortex and language localization (Stippich and Blatow, 2007). Table 3.4  Signal intensity of haemorrhage of different stages on T1-​and T2-​weighted MRI sequences Haemorrhage stage

Haemoglobin product

T1-​weighted sequence

T2-​weighted sequence

Hyperacute (1 month)

Haemosiderin

Contrast media in MRI The majority of the contrast media used in MRI are paramagnetic agents based on gadolinium compounds. The pharmacokinetics of these agents are similar to those of iodinated contrast media with rapid passage from the vascular compartment to the interstitial compartment; an exception is gadofosveset, a blood pool agent that binds to albumin and remains in the vascular compartment. Gadolinium compounds do not pass through an intact blood-​brain barrier so contrast enhancement is considered reflective of blood-​brain barrier breakdown, apart from circumventricular organs, which have an incomplete blood-​brain barrier and demonstrate physiological contrast enhancement that should not be mistaken for abnormality (Horsburgh and Massoud, 2013). Contrast enhancement is more conspicuous on MRI than on CT, and is invaluable in characterization of tumours, vascular disease, inflammation, and infection, both in the brain and in the spine. Gadolinium compounds are generally safe but there are potential side effects associated with the release of unbound gadolinium from its chelates and accumulation in tissues. Nephrogenic systemic fibrosis is a rare but disabling dermopathy resembling scleroderma and eosinophilic fasciitis, first observed in patients on dialysis undergoing gadolinium contrast-​enhanced MRI. Most cases were associated with the use of less stable linear agents that are now classified as high risk (Table 3.5). These compounds require renal function monitoring and are contraindicated in patients with glomerular filtration rate (GFR) below 30 ml/​min/​1.73 m2, in neonates, and in the perioperative liver transplantation period; breastfeeding should Table 3.5  Classification of gadolinium-​based MRI contrast agents according to the risk of nephrogenic systemic fibrosis High risk

Medium risk

Low risk

Gadodiamide (Omniscan)

Gadoxetic acid (Primovist)

Gadoteric acid (Dotarem)

↓

Gadoversatamide (OptiMARK)

Gadobenic acid (MultiHance)

Gadoteridol (ProHance)

↑

↑

Gadopentetic acid (Magnevist)

Gadofosveset (Vasovist)

Gadobutrol (Gadovist)

↓

↓↓

Data from Drug Safety Update Jan 2010, vol 3 issue 6: 3. https://​www.gov.uk

CHAPTER 3  Overview of neuroimaging

be discontinued for at least 24 hours after high-​risk agent administration. Similar precautions are recommended but not mandatory for medium-​and low-​risk agents. Recently it has been recognized that repeated administration of gadolinium-​based contrast agents leads to accumulation of gadolinium in the brain, with the highest concentrations detected in the dentate nucleus and globus pallidus (Kanda et al., 2015b; McDonald et  al., 2015). Accumulation of gadolinium is seen in patients receiving the less stable linear, but not macrocyclic compounds (Kanda et al., 2015b), and occurs even in the presence of normal renal function (Kanda et al., 2015a; McDonald et al., 2015). The clinical significance of these findings is however currently unknown. General application of MRI in neurosurgery Structural MRI has several advantages over CT and is the investigation of choice in the assessment of intracranial and spinal pathology with the exception of a few specific situations. In structural imaging, MRI provides superior tissue contrast, greater sensitivity to parenchymal abnormalities and contrast enhancement, better demonstration of anatomy, and allows imaging in any plane without using ionizing radiation; further information can be gained by addition of functional sequences. Disadvantages of MRI include long scanning time making it problematic in medically unstable patients and more prone to motion artefacts; therefore CT, with its established value in demonstration of intracranial haemorrhage and fractures, remains the investigation of choice in emergency situations, especially in trauma. CT is also better in demonstrating cortical bone and calcification, and remains more easily available and less costly. Use of MRI can be limited by several safety issues related to the static magnetic

Positron emission tomography (PET) PET is a nuclear medicine technique that uses tracers labelled with positron-​emitting isotopes. The most commonly used radionuclide is fluorine-​18 (18F) due to its relatively long half-​life that approaches two hours and is sufficiently long to allow transportation from an external cyclotron facility to the imaging centre. Short-​lived isotopes require an on-​site cyclotron (11C, 13N, 15O) or are produced in a generator (68Ga, 82Rb). In the target organ, positrons emitted by the radionuclides travel a very short distance before interacting with electrons in neighbouring atoms and undergoing annihilation with the emission of two high energy photons that travel in opposite directions, which are detected by a ring array of solid-​state scintillator detectors. The image is reconstructed based on the coregistration of pulses between opposite pairs of detectors, allowing placement of the annihilation event along a specific line of response (Fig. 3.5). Time of flight PET (TOF-​PET) also measures tiny differences in time of arrival of the two photons to the opposite detectors, allowing more precise placement of the annihilation event to a segment along the line of response. Data needs to be corrected for the attenuation of photons inside the patient; this is achieved by using a rotating radiation source in standalone PET or direct information on attenuation

(b)

LO

R

(a)

field, radiofrequency pulses, and gradient fields. MRI cannot generally be performed in patients with pacemakers, ferromagnetic aneurysm clips, implants, or foreign bodies. Care must be taken when imaging patients with shunt valves, cochlear implants, and heart valves, as well as in pregnancy; departmental practices vary. Finally, claustrophobia may preclude many patients from undergoing the MRI investigation or limit the scan time.

γ 511 kEV

+

18F

–

γ 511 kEV

+ Positron

– Electron

γ Photon

Fig. 3.5  (A) A positron-​emitting isotope (most commonly 18F) releases positrons which collide with electrons in the target organ resulting in a series of annihilation events. Each annihilation event results in the emission of two high energy (511 keV) photons travelling in a straight line in opposite directions (at 180° to each other) at the speed of light. The two photons reach a ring of detectors placed around the gantry at the same time (if the annihilation happens in the centre of the gantry) or within a very short time window (if annihilation happens off-​centre). The detection of such a coincidence event allows localization of the source of photons somewhere along the line of response (LOR) joining the pair of detectors which registered the photons. In time of flight (TOF) PET the tiny difference in time it takes for each photon to reach the detector for off-​centre events is also taken into account, allowing the localization of the annihilation event more precisely along the LOR. (B) As photons produced by a series of annihilation events travel along different lines of response, combining information from multiple detected coincidence events allows localization of the radiation source.

37

38

Section 1  Principles of neurological surgery

obtained from a CT scan in PET-​CT hybrid scanners. Data correction in MRI-​PET is more challenging and requires calculation of attenuation values from anatomical MRI images. PET is extremely sensitive, being able to detect picomolar quantities of tracer labelled with a positron emitter. In theory, any physiologically occurring substance, drug, or receptor ligand can be labelled, making PET the most promising technique for functional and molecular imaging. The most widely used tracer is 18F-​ fluorodeoxyglucose (18F-​FDG), the uptake of which reflects the rate of glucose metabolism. Use of 18F-​FDG in the brain is problematic due to high rate of physiologic glucose metabolism; its potential applications include imaging of high-​grade brain tumours, anaplastic transformation of low-​grade gliomas, differentiation of radiation necrosis from tumour progression, differentiation of lymphoma from toxoplasmosis in immunocompromised patients, as well as detection of focal cortical dysplasia in epilepsy investigation. The armamentarium of PET tracers and range of potential applications in brain imaging is expanding; these are summarized in Table 3.6. Most PET scanners in clinical use currently are either standalone machines, or PET-​ CT hybrids. Increasingly, PET-​ MRI hybrid scanners are becoming available. In view of the superior tissue contrast and resolution of MRI in the brain, integrated PET-​MRI has potential advantages in neuroimaging. So far, it has been possible to combine sequential PET and MRI images acquired sequentially on separate scanners but hybrid PET-​MRI imaging greatly facilitates such fusion eliminating the need for subsequent coregistration. More importantly, simultaneous acquisition of PET and MRI images could be of value when measuring parameters that undergo temporal changes, such as cerebral perfusion or hypoxia (Catana et al., 2012). These methods are currently largely limited to neuroimaging research, and their clinical usefulness remains to be established.

Other nuclear medicine techniques Most nuclear medicine techniques use tracers labelled with radioligands emitting gamma radiation. The most commonly used radionuclide is 99mTc (technetium) due to a combination of optimal half-​life, ease of production, and ease of incorporation into radiopharmaceuticals. Other isotopes used in neuroimaging include 111In (indium) and 133Xe (xenon). Emitted gamma rays can be detected using a planar gamma camera or rotating gamma camera in single-​photon emission computed tomography (SPECT). Applications of SPECT in neuroradiology include cerebral perfusion measurements with 99mTc-​HMPAO, 99mTc-​ECD, and 133Xe, diagnosis of dementia, identifying seizure focus in epilepsy, dopamine transporter imaging in diagnosis of parkinsonian syndromes, and brain tumour evaluation (McArthur et  al., 2011). Planar gamma camera images are obtained in radionuclide cisternography, usually with 111In-​DTPA, performed for investigation of hydrocephalus or CSF leaks.

Myelography Although MRI is the preferred method for examination of most spinal abnormalities, myelography remains indicated in patients who cannot undergo MRI (e.g. because of a pacemaker) and it also has a role in the detection of CSF leak sites in patients with spontaneous intracranial hypotension (Kranz et al., 2016). Iodinated contrast medium is injected into the spinal subarachnoid space, most commonly after lumbar puncture but sometimes after a lateral puncture at C1–​2 if lumbar puncture cannot be performed. The puncture is typically done using fluoroscopic guidance and the injection of contrast is monitored by intermitted fluoroscopy to ensure that inadvertent epidural or subdural injection has not occurred. The contrast is then moved along the spinal canal to the appropriate location(s)

Table 3.6  Examples of PET tracers used in neuroimaging PET tracer

Mechanism of action and applications

11

Transport into the cell via amino acid transporter increased in malignant tumours—​characterization of tumour extent, biopsy guidance, treatment planning, response assessment (radiation necrosis, recurrence, pseudoresponse), prognostication; assessment of striatal dopamine pathway (18F-​DOPA)

18

Uptake correlated with activity of thymidine kinase-​1 in proliferating cells—​assessment of tumour grade, treatment response, prognostication

18

Accumulation in hypoxic cells—​assessment of hypoxia in brain tumours with potential relevance to tumour progression and resistance to treatment

11

Binding to translocator protein (TSPO) expressed in activated microglia and cancer cell lines—​assessment of microglial activation/​inflammation, assessment of tumour grade

68 68

Somatostatin analogues with high uptake in certain intracranial tumours—​diagnosis of meningioma, pituitary adenoma, hemangioblastoma, medulloblastoma, PNET

11

C-​PiB (Pittsburgh compound) F-​florbetaben, 18F-​florbetapir 18 F-​flutemetamol

Binding to beta-​amyloid plaques—​diagnosis of Alzheimer disease

11

Binding to tau protein/​neurofibrillary tangles—​diagnosis of Alzheimer disease and tauopathies

C-​methionine, 18F-​FET

F-​FLT F-​FMISO, 18F-​FAZA C-​PK11195, 18F-​GE180 Ga-​DOTA-​TOC Ga-​DOTA-​TATE

18

C-​PBB3 F-​FDDNP

18 11

C-​flumaeznil C-​diprenorphine C-​SCH23390 11 C-​raclopride

Benzodiazepine receptor ligand—​receptor binding studies Opioid receptor ligand—​receptor binding studies Dopamine D1 receptor ligand—​receptor binding studies Dopamine D2 receptor ligand—​receptor binding studies

15

Cerebral blood flow, oxygen metabolism Protein synthesis rate Plasma volume pH

11 11

O2 and [15O] H2O C-​leucine, 11C-​tyrosine 11 C-​albumin 11 CO2, 11C-​DMO 11

CHAPTER 3  Overview of neuroimaging

by tilting the examination table and radiographs are obtained with the patient supine, prone, and in oblique positions. Myelography is typically then followed by a CT examination of the spine, which provides greater anatomical information than can be obtained from radiographs alone and allows excellent depiction of bony anatomy and spinal nerve roots on the same image. Contraindications to myelography include allergy to iodinated contrast media, seizure disorders, bleeding disorders, or patients taking anticoagulants or antiplatelet agents, patients with suspected raised intracranial pressure, and uncooperative patients. Complications are uncommon with modern water-​soluble iodinated contrast agents but include postdural puncture headache, seizures, bleeding at the puncture site, nerve root injury, and arachnoiditis (Sandow and Donnal, 2005).

Ultrasound Ultrasound (US) is sound that has a frequency greater than the upper threshold of human hearing, approximately 20  kHz. Ultrasound used for medical imaging typically has a frequency between 2 MHz and 18  MHz. When transmitted from an ultrasound probe into the body, an US beam undergoes reflection and refraction when it passes through changes in tissue density and compressibility. The US probe can detect the returned echoes and determine their depth of origin within the patient based on the time since transmission and the speed of sound in tissue. The intensity of the returned echoes is affected by the reflectivity of the tissue and the degree to which it attenuates the US beam. It is thus possible to construct a two-​ dimensional greyscale image in which the position and brightness of each pixel corresponds to the site of origin and strength of an echo arising within the patient, so-​called B-​mode imaging (Allisy-​ Roberts and Williams, 2007). Bone and calcified tissues are both highly reflective and attenuating, preventing through transmission of sound, so appear as a bright white interface beyond which there is an ‘acoustic shadow’ in which no information is obtained. Conversely fluid is of low reflectivity and attenuates the beam weakly, thus appearing as an anechoic area beyond which there is increased through transmission of sound relative to adjacent tissues and resulting in posterior acoustic enhancement. The attenuation of the US beam by bone limits the utility of B-​mode imaging for assessing the brain in adults and children. Transfontanellar US scanning of neonates and infants can be used to assess the size of the ventricles, detect intraventricular and parenchymal haemorrhage in preterm neonates, and evaluate the brain parenchyma in cases of hypoxic ischaemic encephalopathy. Ultrasound of the spine can also be performed in neonates for assessment of suspected cases of dysraphism to assess the position of the conus, presence of diastematomyelia, or associated intraspinal masses. When US waves are reflected by a moving interface their frequency is changed by the Doppler effect, resulting in an increase in frequency if the interface is moving towards the transducer and a reduction in frequency for a reflector that is moving away. The velocity of reflector motion affects the amount of change in frequency. This effect is exploited by pulsed-​wave Doppler US to obtain waveforms from blood vessels that show the velocity and direction of blood flow with respect to the transducer. In colour Doppler imaging the information regarding direction and

velocity of blood flow is overlaid on a greyscale B-​mode image as a colour-​coded map. Doppler US can be used to detect the presence of carotid artery stenosis in patients with carotid bruits, transient ischaemic attacks, or ischaemic strokes. The degree of carotid stenosis is generally assigned to a percentage range based on the increase in blood flow velocity through area of stenosis. It is also possible to assess the blood flow of major intracranial arteries by transcranial Doppler (TCD). This requires the use of low frequency (2–​3 MHz) transducers that can penetrate through the relatively thin bone of the pterion or by scanning through the foramen magnum. TCD can be used to perform bedside monitoring of patients with aneurysmal subarachnoid haemorrhage to look for increases in blood flow velocity that can indicate vasospasm (Marshall et al., 2010). Some institutions also use intraoperative colour Doppler US for the assessment of CSF flow at the foramen magnum in cases of foramen magnum decompression for Chiari I malformation.

Vascular imaging Digital subtraction angiography Digital subtraction angiography (DSA) (Wolpert, 1999) remains the reference standard for imaging of vascular disorders of the brain and spine, although with improvements in CT angiography and MR angiography, the indications for DSA as the first-​line investigation have decreased. DSA is usually performed under local anaesthesia, although sedation or general anaesthesia may be necessary in uncooperative patients, for interventional neuroradiology procedures and, depending on operator preference, in diagnostic spinal angiography. The usual route of access to the arterial system is through the common femoral artery; where this is not possible then transradial or transbrachial arterial approaches may be used. After puncture with a needle, a soft-​tipped guidewire is threaded through the needle into the artery and the needle is then removed and exchanged for a polyethylene catheter. A variety of catheter shapes are available, the selection partly depending on operator preference but also on the degree of arterial tortuosity. Under fluoroscopic guidance the catheter can be positioned in the internal and external carotid arteries and the vertebral arteries as necessitated by the indication for the DSA. Neuroangiography is typically performed using biplane angiography units that have two X-​ray tubes and image intensifiers or, increasingly often, flat panel digital detectors. This allows acquisition of images simultaneously from two different projections, reducing the length of the examination and the volume of contrast material that needs to be injected. Images are acquired using pulsed X-​rays immediately before (so-​called mask images) and then during the injection of contrast material into the artery. By subtracting mask images from those obtained during the injection of contrast, images are produced that eliminate bone and soft tissue detail and show only the contrast within the blood vessels. If the patient moves during image acquisition the mask images will not be correctly registered with the postcontrast images and they may need to be aligned manually. It is also possible to obtain three-​ dimensional images by rotating the C-​arm around the patient’s head during the injection of contrast (rotational 3D DSA). Possible complications of cerebral DSA include those common to all invasive angiographic procedures such as puncture site haematoma (approximately 5%), arterial dissection, and contrast reactions. Manipulation of catheters in the aortic arch and neck arteries

39

40

Section 1  Principles of neurological surgery

also carries a small risk of neurologic deficits, which are most often transient (approximately 2.5%) but can rarely be permanent (approximately 0.2%). The risk of neurological complications depends on the operator’s experience and is increased in older patients and longer procedures. CT angiography and CT venography CT angiography (CTA) involves the rapid acquisition of a thin section data set of the area of clinical interest during the first pass of an injected bolus of iodinated contrast material. Acquisition of images during the arterial phase can be achieved by bolus tracking or test bolus methods. Bolus tracking involves triggering the scan acquisition when the attenuation within a chosen vessel, most often the internal carotid artery, reaches a predetermined level after the start of contrast injection. The test bolus method determines the optimum time to start scan acquisition by using a separate injection of a small volume of contrast medium. Although image interpretation always requires review of the axially acquired source images, reconstruction of the 3D data using multiplanar reformats, maximum intensity projection (MIP), and volume-​rendered images can facilitate the identification of abnormalities and enhance the appreciation of their relationship to surrounding vessels. Advantages of CTA over DSA include its non-​invasive nature, more widespread availability, and the ease and speed of acquisition. Disadvantages of conventional CTA include its poor temporal resolution—​images are essentially a ‘snapshot’ that do not provide dynamic information, reduced spatial resolution relative to DSA, and potential obscuration of small vascular abnormalities adjacent to bone. Bone-​subtracted CTA images can be obtained by acquiring a mask of CT data before injecting contrast and subtracting it from the postcontrast images in a manner analogous to DSA, or by using dual energy X-​ray acquisition (although this requires a dedicated dual energy CT scanner). CT venography is similar in principle to CTA except that the images are acquired when the majority of the injected contrast is in the venous phase. The timing of image acquisition is less critical and can generally be achieved with a fixed scan delay of 40 seconds after the start of contrast injection. MR angiography (MRA) and venography (MRV) There are several different methods by which angiographic images can be obtained with MRI. Non-​contrast methods rely on signal differences between flowing blood and surrounding stationary tissues (Lim and Koktzoglou, 2015). Time-​of-​flight magnetic resonance angiography (TOF-​MRA) depends on flow-​related enhancement, where stationary tissues saturated by repeated radiofrequency pulses display low signal, while non-​saturated blood flowing into the imaged slice is hyperintense. Phase-​ contrast angiography (PC-​MRA) uses two bipolar gradients inducing a phase shift in blood flowing along the gradient direction, which is proportional to the flow velocity. By changing the gradient strength, velocity encoding sensitivity (Venc) can be tailored to optimally demonstrate arterial or venous flow. Both TOF-​MRA and PC-​MRA can be performed using two (2D) or three-​dimensional (3D) acquisition. Contrast-​enhanced MRA (CE-​MRA) is similar in principle to CTA, whereby the vessel visualization depends on intravenously injected contrast agent. In practice, 3D TOF-​MRA is the main technique used to image intracranial vasculature, with sensitivity and

specificity in aneurysm detection similar to CTA when performed on a three-​tesla system (Lim and Koktzoglou, 2015). Large aneurysms or aneurysm with slow flow are better demonstrated with CE-​ MRA, which can also be performed as time-​resolved examination (Saindane, 2015) for evaluation of arteriovenous malformations or dural fistulas. In detection of carotid artery stenosis, TOF-​MRA is less sensitive and specific than CE-​MRA (Gough, 2011) and tends to overestimate the degree of stenosis. MRV is most often performed using a phase-​contrast method, as time of flight can be confounded by hyperintense appearances of a subacute thrombus that can mimic normal flow-​related enhancement (Nadgir and Yousem, 2016). MRV can also be performed as contrast-​enhanced examination (Sadigh et al., 2016). Other vascular techniques As mentioned earlier, high sensitivity of SWI to even small quantities of blood products renders it highly useful for detection of intracranial haemorrhage, for example, microhaemorrhages in cerebral amyloid angiopathy or traumatic brain injury. In addition, due to its sensitivity to paramagnetic properties of deoxyhaemoglobin, SWI allows excellent demonstration of venous structures (Di Ieva et al., 2015). SWI has been shown to be useful in demonstrating venous drainage of arteriovenous malformations (AVMs) and evaluation of the extent of their neurosurgical resection, as well as depiction of micro-​AVMs (Mossa-​Basha et al., 2012; Saindane, 2015). In addition to evaluation of the vessel lumen, there has been increasing emphasis on the assessment of the vessel wall. In carotid artery imaging, the main focus is on the evaluation of atherosclerotic plaque, in particular its stability and propensity to rupture (Young et al., 2011). Intracranial vessel wall imaging can be useful in evaluation of intracranial atherosclerosis, moyamoya disease, vasculitis, reversible cerebral vasoconstriction syndrome, and intracranial arterial dissection (Mandell et al., 2017).

Imaging methods to measure cerebral blood flow and metabolism Imaging of cerebral perfusion and metabolism has a wide variety of research and clinical applications including assessment of acute stroke, hypoxic brain injury, vasospasm, characterization of cerebrovascular reserve, and evaluation of traumatic brain injury (Coles, 2006; Dani and Warach, 2014). Major applications are found in neuro-​oncology where different methods are combined to derive information on tumour type and grading, differentiation from non-​neoplastic lesions, biopsy guidance, response assessment, and differentiation of chemotherapy and radiotherapy effects from tumour progression (Herholz et al., 2007; Waldman et al., 2009; Kim et al., 2016). Regional differences in blood flow and metabolism also provide additional diagnostic information in dementia and neurodegenerative disease (Schuff, 2013; Ishii, 2014). Space constraints dictate only a brief description of the available modalities; more comprehensive analysis can be found in the references provided, as well as dedicated volumes (Gillard et al., 2010; Barker et al., 2013). Measuring cerebral blood flow The imaging methods used to measure cerebral perfusion can be based on inert diffusible tracers or non-​ diffusible intravascular tracers (Coles, 2006). The former rely on the Fick principle stating that

CHAPTER 3  Overview of neuroimaging

uptake of any substance by the target organ depends on its arteriovenous concentration gradient and the organ perfusion. Techniques that use diffusible traces include xenon-​enhanced CT with non-​ radioactive 131Xe, and nuclear medicine methods such as SPECT with 99mTc-​HMPAO, 99mTc-​ECD or radioactive 133Xe isotope, and PET with 15O2. Quantitative 15O2 PET and xenon CT are considered the gold standard for measuring cerebral perfusion, but are not routinely used in clinical imaging (Thompson et al., 2010). Magnetically tagged water molecules in MR perfusion using arterial spin labelling also behave like a diffusible tracer, and this approach is likely to be increasingly applied in a clinical setting (Grade et al., 2015). Techniques that require intravascular injection of a non-​diffusible tracer are more widely available and are increasingly becoming part of routine imaging, especially in stroke and neuro-​oncological imaging. The basic principles underlying measurement of perfusion parameters with these methods by CT or MRI are similar and rely on rapid serial imaging during the first pass of an injected bolus of contrast material (Konstas et al., 2009a). In simple terms, the resulting changes in attenuation (CT) and signal intensity (MRI) in the brain depend on the combination of the inherent perfusion characteristics of the tissue and the inflow of contrast from the arteries supplying the voxels of interest (arterial input function), assuming that the contrast agent remains within the vascular space. The area under the contrast concentration curve and the width of the contrast bolus provide estimates of cerebral blood volume (CBV) and mean transit time (MTT), respectively, and the cerebral blood flow (CBF) can be calculated from the central volume theorem (CBF  =  CBV/​MTT). CT perfusion (CTP) is now widely used in the evaluation of the extent of core infarction and ischaemic penumbra in acute stroke (Konstas et al., 2009a; Konstas et al., 2009b) and in diagnosis and management of vasospasm following subarachnoid haemorrhage (Mir et  al., 2014). The greatest advantage of CTP is the linear relationship between the concentration of iodine and tissue attenuation, and direct measurement of arterial input and venous output functions, allowing for direct calculation of vascular parameters. A disadvantage of perfusion CT compared with MRI has been the use of ionizing radiation, and its limited craniocaudal coverage, now largely overcome with larger detector array CT scanners and shuttling table techniques. Perfusion MRI techniques are based either on measuring susceptibility effects of gadolinium contrast agent causing a reduction of T2* signal (dynamic susceptibility contrast, DSC), or its relaxivity effects reflected by an increase in T1 signal (dynamic contrast enhancement, DCE). Quantitative assessment of absolute perfusion parameters is more difficult than for CTP and requires mathematical and pharmacokinetic modelling with several assumptions about the behaviour of the injected tracer, and correction accounting for contrast leakage into extravascular space (Jackson, 2004). To simplify comparisons, perfusion parameters derived from DSC can be expressed as relative values, usually normalized against contralateral white matter. DCE allows calculation of several parameters that characterize the microvascular environment, including permeability. The most commonly used indices include estimates of blood flow (F), permeability surface area product per unit mass of tissue (PS), the volume transfer constant (Ktrans), volumes of the extravascular extracellular space (Ve) and blood plasma (Vb), as well as semiquantitative indices derived from the shape of the enhancement curve (Thompson et al., 2010; Griffith and Jain, 2015).

An emerging technique of cerebral blood volume quantification is vascular space occupancy imaging (VASO), which uses T1 difference between the blood and tissue to separate the two compartments. VASO MRI can detect changes in CBV with high temporal resolution without the use of injected contrast medium, and provides absolute CBV measurements when also performed with intravenous contrast agent (Lu and Uh, 2013). Recently there has been renewed interest in measuring brain perfusion using intravoxel incoherent motion imaging (IVIM) which is based on the observation that signal in diffusion weighted MRI depends not only on diffusion itself, but also on blood flow within the capillary bed (pseudo diffusion). By using a range of diffusion gradient strengths and bi-​exponential fitting or kurtosis models, the effects of capillary flow and true diffusion can be separated, providing an estimation of perfusion (Federau et al., 2014). It is also worth mentioning here the methods used to evaluate brain oxygenation, which depends on perfusion and oxygen metabolism. The gold standard technique is 15O2 PET which may be supplemented with [15O]H2O and [15O]CO administration (triple oxygen PET), providing absolute values of cerebral metabolic rate of oxygen (CRMO2) and oxygen extraction fraction, in addition to cerebral perfusion indices. Emerging MRI-​based methods include quantitative BOLD, phase-​and susceptibility-​based imaging, and intravascular T2-​based approaches (Christen et al., 2013). Measuring cerebral metabolism The two major imaging modalities currently used for assessment of cerebral metabolism are PET and magnetic resonance spectroscopy (MRS). As described earlier, depending on the tracer PET allows estimation of local or regional glucose metabolism (18FDG-​PET) and oxygen metabolism (15O2, triple oxygen). In brain tumours, PET allows estimation of amino acid transport (11C-​methionine, 18F-​FET), protein synthesis (11C-​methionine and 11C-​leucine) and DNA synthesis (18F-​FLT), as well as tissue-​level hypoxia (18F-​FMISO and 18F-​FAZA). MRS is based on chemical shift, a specific change in resonance frequency that arises as a result of electron shielding when an atom of interest is chemically bound within a compound. MRS allows probing of the biochemical milieu in a defined voxel of interest (single voxel spectroscopy) or as 2D or 3D magnetic spectroscopic imaging (CSI), mapping the concentration of a given compound across the brain. MRS is generally a challenging technique due to low sensitivity and low signal-​to-​noise ratio, and requires higher strength and precise shimming of the magnetic field. The most commonly used form of MRS due to relative abundance of the target nucleus is proton spectroscopy (1H-​MRS). In the brain, 1H-​MRS detects a spectrum of several metabolites considered markers of specific metabolic aspects, in particular N-​acetyl aspartate (NAA, a marker of neuronal integrity), choline (Cho, a marker of membrane turnover), creatine, and myoinositol (markers of cellular energy status). Pathologic conditions cause changes in the spectral pattern, or appearance of abnormal metabolite peaks such as lactate (Lac, marker of metabolic acidosis and cell death), or lipids (marker of tissue breakdown); for example, high-​grade gliomas demonstrate a reduction in NAA and elevated Cho, accompanied by lipid and lactate peaks (Herholz et al., 2007). More technically challenging 31P-​ MRS allows estimation of energy metabolism by detecting spectra from phosphorus-​containing molecules of ATP, phosphocreatine, and inorganic phosphate, as well as of the neuronal membrane

41

42

Section 1  Principles of neurological surgery

metabolism by detecting phosphomonoesters and phosphodiesters (Herholz et al., 2007). Recently there has been intense interest in dynamic nuclear polarization (DNP or hyperpolarization), a technique aiming to address the inherent problem of low sensitivity in spectroscopy by increasing the proportion of nuclei contributing to magnetic resonance signal. Normally, magnetic moments of nuclei polarized in a magnetic field align either parallel (spin up) or antiparallel to the field (spin down); net magnetic resonance signal is attributed to the tiny difference between the number of nuclei pointing each way (with just a few more nuclei per million pointing spin up). DNP is performed by extreme cooling of a sample containing nuclei of interest and free radicals as a source of unpaired electrons in a solid state, and subjecting it to a strong magnetic field. In these conditions, electron spins become highly polarized, and this polarization is partly transferred to nuclear spins by microwave irradiation, resulting in a more than 10 000-​fold increase in signal-​to-​noise ratio (Ardenkjaer-​Larsen et al., 2003). Such a treatment cannot of course be performed in vivo, but can be used to hyperpolarize a sample containing MRS-​detectable nuclei, which can be quickly warmed up, injected into a patient, and imaged with increased signal-​to-​noise ratio. Systems capable of DNP in humans are now available, and there is intense research effort in this area. The most common metabolite currently used in DNP is [1-​13C]pyruvate, which after injection is converted into [1-​13C]lactate by lactate dehydrogenase (LDH), [1-​13C]alanine by alanine transaminase, and 13CO2 by pyruvate dehydrogenase (PDH), the latter further converted into [13C] bicarbonate. Conversion of pyruvate into lactate is increased in malignant tumours, even in the presence of sufficient oxygen (Warburg effect), which is leveraged in cancer diagnosis and assessment of treatment response (Kim et  al., 2016). Other promising markers which proved potentially useful in animal models include [1,4-​13C]fumarate for detection of cellular necrosis, [13C]bicarbonate for in vivo measurement of pH, and [5-​13C]glutamine for assessment of tumour metabolism, while [1-​13C]urea can be used as a perfusion marker. It is also possible to inject and detect several hyperpolarized substrates simultaneously, provided that their chemical shift is sufficiently different to distinguish them spectroscopically (Brindle et al., 2011).

Future directions Current conventional radiological methods allow demonstration of anatomical and macroscopic pathological features of disease with excellent sensitivity and resolution. Functional imaging methods that have enabled insight into variety of physiological and pathological phenomena as described earlier, are an important first step beyond anatomical imaging towards more precise characterization of the disease process. These methods however remain generally non-​specific, demonstrating changes that are common endpoints of many possible pathological pathways. One avenue towards more specific information is radiomics—​conversion of images into higher dimensional data and subsequent feature extraction and data mining in a manner analogous to genomic, proteomic, or metabolomic approaches, with the aim to obtain imaging signatures corresponding to particular features of disease (Gillies et al., 2016). Radiomic features include both semantic measures (e.g. shape, size, or vascularity), and mathematically derived agnostic descriptors (such as histogram skewness or kurtosis, fractal, and

texture analysis). Radiomics provides useful additional information for diagnosis, treatment selection, and prognostication, but its inherent limitation is the same as of the original imaging method used to obtain the source data. The ultimate goal would be to obtain information on the cellular and molecular events that precede and underlie the observed anatomical and functional sequelae by using probes specific to a particular molecular target or biochemical pathway. This would lead to better understanding of abnormal biology, enable earlier detection, and provide precise diagnosis and disease staging. Molecular imaging could also provide better endpoint biomarkers in the drug discovery process, allowing rapid assessment of treatment effect on molecular level even before phenotypical response occurs (Massoud and Gambhir, 2007). Arguably, the most important application of such methods will be in neuro-​oncology, where there are already several examples of clinical applications of detecting specific receptors, or protein products specific to certain enzyme isoforms (Kim et al., 2016). It is likely that PET and MRI will be used for detection of imaging probes due to limitations of optical, photoacoustic, or ultrasound techniques for intracranial imaging, although these could be used intraoperatively or during endoscopy. Nanoparticles are emerging as an ideal imaging agent (Thakor and Gambhir, 2013) that can be targeted to a specific tissue and labelled for detection by a chosen assay, or even multiple methods, for example, for preoperative imaging by MRI, and intraoperative localization using photoacoustic and optical imaging (Kircher et  al., 2012). Concurrent loading of nanoparticles with a therapeutic agent enables combination of diagnostics and therapy—​theranostics (Kelkar and Reineke, 2011; Thakor and Gambhir, 2013), in which imaging is paired to the targeted delivery of a chemotherapeutic, gene-​silencing agent; or radioactive isotope; or to radiosensitization, photodynamic, or photothermal interventions. In the future, imaging will undoubtedly be the key element of precision medicine, enabling precise diagnosis and individualized therapy tailored for the individual patient and targeted to a particular pathological entity (Kim et al., 2016).

REFERENCES Allisy-​Roberts, P. & Williams, J.R. (2007). Farr’s Physics for Medical Imaging, illustrated edition. Philadelphia, PA:  Elsevier Health Sciences. Ardenkjaer-​Larsen, J.H., Fridlund, B., Gram, A., et al. (2003). Increase in signal-​to-​noise ratio of > 10,000 times in liquid-​state NMR, Proc Natl Acad Sci U S A, 100(18), 10158–​63. Barker, P.B., Golay, X., & Zaharchuk, G. (2013). Clinical Perfusion MRI Techniques and Applications. Cambridge, UK:  Cambridge University Press. Barrett, J.F. & Keat, N. (2004). Artifacts in CT: recognition and avoidance. Radiographics, 24(6), 1679–​91. Bitar, R., Leung, G., Perng, R., et al. (2006). MR pulse sequences: what every radiologist wants to know but is afraid to ask. Radiographics, 26(2), 513–​37. Brindle, K.M., Bohndiek, S.E., Gallagher, F.A., & Kettunen, M.I. (2011). Tumor imaging using hyperpolarized 13C magnetic resonance spectroscopy. Magn Reson Med, 66(2), 505–​19. Catana, C., Drzezga, A., Heiss, W.D., & Rosen, B.R. (2012). PET/​MRI for neurologic applications. J Nucl Med, 53(12), 1916–​25.

CHAPTER 3  Overview of neuroimaging

Christen, T., Bolar, D.S., & Zaharchuk, G. (2013). Imaging brain oxygenation with MRI using blood oxygenation approaches: methods, validation, and clinical applications. AJNR, 34(6), 1113–​23. Cohnen, M., Wittsack, H.J., Assadi, S., et  al. (2006). Radiation exposure of patients in comprehensive computed tomography of the head in acute stroke. AJNR, 27(8), 1741–​5. Coles, J.P. (2006). Imaging of cerebral blood flow and metabolism. Curr Opin Anaesthesiol, 19(5), 473–​80. Dani, K.A. & Warach, S. (2014). Metabolic imaging of ischemic stroke: the present and future. AJNR, 35(6 Suppl), S37–​43. Di Ieva, A., Lam, T., Alcaide-​Leon, P., Bharatha, A., Montanera, W., & Cusimano, M.D. (2015). Magnetic resonance susceptibility weighted imaging in neurosurgery: current applications and future perspectives. J Neurosurg, 123(6), 1463–​75. Federau, C., O’Brien, K., Meuli, R., Hagmann, P., & Maeder, P. (2014). Measuring brain perfusion with intravoxel incoherent motion (IVIM): initial clinical experience. J Magn Reson Imaging, 39(3), 624–​32. Gillard, J.H., Waldman, A.D., & Barker, P.B. (2010). Clinical MR Neuroimaging: Physiological and Functional Techniques. Cambridge, UK: Cambridge University Press. Gillies, R.J., Kinahan, P.E., & Hricak, H. (2016). Radiomics: images are more than pictures, they are data. Radiology, 278(2), 563–​77. Ginat, D.T. & Gupta, R. (2014). Advances in computed tomography imaging technology. Annu Rev Biomed Eng, 16, 431–​53. Gough, M.J. (2011). Preprocedural imaging strategies in symptomatic carotid artery stenosis. J Vasc Surg, 54(4), 1215–​18. Grade, M., Hernandez Tamames, J.A., Pizzini, F.B., Achten, E., Golay, X., & Smits, M. (2015). A neuroradiologist’s guide to arterial spin labeling MRI in clinical practice. Neuroradiology, 57(12), 1181–​202. Griffith, B. & Jain, R. (2015). Perfusion imaging in neuro-​ oncology:  basic techniques and clinical applications. Radiol Clin North Am, 53(3), 497–​511. Herholz, K., Coope, D., & Jackson, A. (2007). Metabolic and molecular imaging in neuro-​oncology. Lancet Neurol, 6(8), 711–​24. Hoeffner, E.G., Mukherji, S.K., Srinivasan, A., & Quint, D.J. (2012). Neuroradiology back to the future: brain imaging, AJNR, 33(1),  5–​11. Hoffman, J.M. & Gambhir, S.S. (2007). Molecular imaging:  the vision and opportunity for radiology in the future. Radiology, 244(1),  39–​47. Horsburgh, A. & Massoud, T.F. (2013). The circumventricular organs of the brain: conspicuity on clinical 3 T MRI and a review of functional anatomy. Surg Radiol Anat, 35(4),  343–​9. Hounsfield, G.N. (1995). Computerized transverse axial scanning (tomography):  part I.  Description of system. 1973. Br J Radiol, 68(815), H166–​72. Hulkower, M.B., Poliak, D.B., Rosenbaum, S.B., Zimmerman, M.E., & Lipton, M.L. (2013). A decade of DTI in traumatic brain injury: 10 years and 100 articles later. AJNR, 34(11), 2064–​74. Hunt, C.H., Hartman, R.P., & 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. Ishii, K. (2014). PET approaches for diagnosis of dementia. AJNR, 35(11), 2030–​8. Jackson, A. (2004). Analysis of dynamic contrast enhanced MRI. Br J Radiol, 77 Spec No 2, S154–​66. Kanda, T., Fukusato, T., Matsuda, M., et al. (2015a). Gadolinium-​based contrast agent accumulates in the brain even in subjects without severe renal dysfunction: evaluation of autopsy brain specimens with inductively coupled plasma mass spectroscopy. Radiology, 276(1), 228–​32.

Kanda, T., Osawa, M., Oba, H., et  al. (2015b). High signal intensity in dentate nucleus on unenhanced T1-​weighted MR images: association with linear versus macrocyclic gadolinium chelate administration. Radiology, 275(3),  803–​9. Kelkar, S.S. & Reineke, T.M. (2011). Theranostics: combining imaging and therapy. Bioconjugate Chem, 22(10), 1879–​903. Kidwell, C.S. & Wintermark, M. (2010). The role of CT and MRI in the emergency evaluation of persons with suspected stroke. Curr Neurol Neurosci Rep, 10(1),  21–​8. Kilic, K., Erbas, G., Guryildirim, M., Arac, M., Ilgit, E., & Coskun, B. (2011). Lowering the dose in head CT using adaptive statistical iterative reconstruction. AJNR, 32(9), 1578–​82. Kim, M.M., Parolia, A., Dunphy, M.P., & Venneti, S. (2016). Non-​ invasive metabolic imaging of brain tumours in the era of precision medicine. Nat Rev Clin Oncol, 13(12), 725–​39. Kimpe, T. & Tuytschaever, T. (2007). Increasing the number of gray shades in medical display systems—​how much is enough? J Digi Imaging, 20(4), 422–​32. Kircher, M.F., de la Zerda, A., Jokerst, J.V., et al. (2012). A brain tumor molecular imaging strategy using a new triple-​ modality MRI-​ photoacoustic-​Raman nanoparticle. Nat Med, 18(5), 829–​34. Konstas, A.A., Goldmakher, G.V., Lee, T.Y., & Lev, M.H. (2009a). Theoretic basis and technical implementations of CT perfusion in acute ischemic stroke, part 1: theoretic basis. AJNR, 30(4),  662–​8. Konstas, A.A., Goldmakher, G.V., Lee, T.Y., & Lev, M.H. (2009b). Theoretic basis and technical implementations of CT perfusion in acute ischemic stroke, part  2:  technical implementations. AJNR, 30(5), 885–​92. Kranz, P.G., Luetmer, P.H., Diehn, F.E., Amrhein, T.J., Tanpitukpongse, T.P., & Gray, L. (2016). Myelographic techniques for the detection of spinal CSF leaks in spontaneous intracranial hypotension. AJR Am J Roentgenol, 206(1),  8–​19. Lauterbur, P.C. (1989). Image formation by induced local interactions:  examples employing nuclear magnetic resonance. Clin Orthop Relat Res, (244), 3–​6. Lim, R.P. & Koktzoglou, I. (2015). Noncontrast magnetic resonance angiography: concepts and clinical applications. Radiol Clin North Am, 53(3), 457–​76. Lu, H. & Uh, J. (2013). Vascular space occupancy (VASO) imaging of cerebral blood volume. In:  Barker, P.B., Golay, X., Zaharchuk, G., et al. (eds) Clinical Perfusion MRI Techniques and Applications, pp. 89–​102. Cambridge, UK: Cambridge University Press. Lusic, H. & Grinstaff, M.W. (2013). X-​ray-​computed tomography contrast agents. Chem Rev, 113(3), 1641–​66. Mandell, D.M., et  al. & Vessel Wall Imaging Study Group of the American Society of Neuroradiology (2017). Intracranial vessel wall MRI:  principles and expert consensus recommendations of the American Society of Neuroradiology. AJNR Am J Neuroradiol, 38(2), 218–​29. Mansfield, P. & Maudsley, A.A. (1977). Medical imaging by NMR. Br J Radiol, 50(591), 188–​94. Marshall, S.A., Nyquist, P., & Ziai, W.C. (2010). The role of transcranial Doppler ultrasonography in the diagnosis and management of vasospasm after aneurysmal subarachnoid hemorrhage. Neurosurg Clin N Am, 21(2), 291–​303. Massoud, T.F. & Gambhir, S.S. (2007). Integrating noninvasive molecular imaging into molecular medicine:  an evolving paradigm, Trends Mol Med, 13(5), 183–​91. McArthur, C., Jampana, R., Patterson, J., & Hadley, D. (2011). Applications of cerebral SPECT. Clin Radiol, 66(7), 651–​61.

43

44

Section 1  Principles of neurological surgery

McDonald, R.J., McDonald, J.S., Kallmes, D.F., et  al. (2015). Intracranial gadolinium deposition after contrast-​enhanced MR imaging. Radiology, 275(3), 772–​82. Mettler, F.A., Huda, W., Yoshizumi, T.T., & Mahesh, M. (2008). Effective doses in radiology and diagnostic nuclear medicine:  a catalog. Radiology, 248(1), 254–​63. Michel, M., Jacob, S., Roger, G., et al. (2012). Eye lens radiation exposure and repeated head CT scans: a problem to keep in mind. Eur J Radiol, 81(8), 1896–​900. Mir, D.I., Gupta, A., Dunning, A., et al. (2014). CT perfusion for detection of delayed cerebral ischemia in aneurysmal subarachnoid hemorrhage: a systematic review and meta-​analysis. AJNR, 35(5), 866–​71. Mirro, A.E., Brady, S.L., & Kaufman, R.A. (2016). Full dose-​reduction potential of statistical iterative reconstruction for head CT protocols in a predominantly pediatric population. AJNR, 37(7), 1199–​205. Mossa-​Basha, M., Chen, J., & Gandhi, D. (2012). Imaging of cerebral arteriovenous malformations and dural arteriovenous fistulas. Neurosurg Clin N Am, 23(1),  27–​42. Nadgir, R. & Yousem, D.M. (2016). Neuroradiology: The Requisites, 4th edition. Philadelphia, PA: Elsevier. Price, S.J., Jena, R., Burnet, N.G., Carpenter, T.A., Pickard, J.D., & Gillard, J.H. (2007). Predicting patterns of glioma recurrence using diffusion tensor imaging. Eur Radiol, 17(7), 1675–​84. Sadigh, G., Mullins, M.E., & Saindane, A.M. (2016). Diagnostic performance of MRI sequences for evaluation of dural venous sinus thrombosis. AJR Am J Roentgenol, 206(6), 1298–​306. Saindane, A.M. (2015). Recent advances in brain and spine imaging. Radiol Clin North Am, 53(3), 477–​96. Sandow, B.A. & Donnal, J.F. (2005). Myelography complications and current practice patterns. AJR Am J Roentgenol, 185(3), 768–​71. Schuff, N. (2013). MR perfusion imaging in neurodegenerative disease. In:  Barker, P.B., Golay, X., Zaharchuk, G., et  al. (eds) Clinical Perfusion MRI: Techniques and Applications, pp. 164–​178. Cambridge, UK: Cambridge University Press. Stippich, C. & Blatow, M. (2007). Clinical Functional MRI: Presurgical Functional Neuroimaging. Berlin, Germany/​New York: Springer. Thakor, A.S. & Gambhir, S.S. (2013). Nanooncology:  the future of cancer diagnosis and therapy. CA Cancer J Clin, 63(6), 395–​418. The Royal Australian and New Zealand College of Radiologists (2016). RANZCR Iodinated Contrast Guidelines. https://​www.ranzcr.com/​ search/​ranzcr-​iodinated-​contrast-​guidelines

The Royal College of Radiologists (2009). Protection of Pregnant Patients during Diagnostic Medical Exposures to Ionising Radiation. https:// ​ w ww.rcr.ac.uk/ ​ p rotection- ​ p regnant- ​ p atients-​ d uring-​ diagnostic-​medical-​exposures-​ionising-​radiation Thompson, G., Mills, S.J., Stivaros, S.M., & Jackson, A. (2010). Imaging of brain tumors: perfusion/​permeability, Neuroimaging Clin N Am, 20(3), 337–​53. Waldman, A.D., et  al. & National Cancer Research Institute Brain Tumour Imaging Subgroup (2009). Quantitative imaging biomarkers in neuro-​oncology. Nat Rev Clin Oncol, 6(8), 445–​54. Wolpert, S.M. (1999). Neuroradiology classics. AJNR, 20(9), 1752–​3. Wolpert, S.M. (2000). Neuroradiology classics. AJNR, 21(3),  605–​6. Yan, J.L., van der Hoorn, A., Larkin, T.J., Boonzaier, N.R., Matys, T., & Price, S.J. (2016). Extent of resection of peritumoral diffusion tensor imaging-​detected abnormality as a predictor of survival in adult glioblastoma patients. J Neurosurg, 126(1),  1–​8. Young, V.E., Sadat, U., & Gillard, J.H. (2011). Noninvasive carotid artery imaging with a focus on the vulnerable plaque. Neuroimaging Clinics of North America, 21(2), 391–​405, xi–​xii.

RELATED LINKS TO EBRAIN CT Based Imaging Techniques. https://learning.ebrain.net/course/ view.php?id=692 Digital Subtraction Angiography. https://learning.ebrain.net/course/ view.php?id=693 Intraoperative Imaging Techniques. https://learning.ebrain.net/ course/view.php?id=89 MR Based Imaging Techniques. https://learning.ebrain.net/course/ view.php?id=1029 Nuclear Medicine Based Imaging Techniques. https://learning.ebrain. net/course/view.php?id=701 Principles of Neurological Investigation. https://learning.ebrain.net/ course/view.php?id=43 Spinal Myelography. https://learning.ebrain.net/course/view. php?id=702 Ultrasound and Doppler Imaging. https://learning.ebrain.net/course/ view.php?id=703 X-ray Development, Physics, Uses, Risks and Regulation. https:// learning.ebrain.net/course/view.php?id=704

4

The operating theatre environment Neil Kitchen and Jonathan Shapey

The neurosurgical operating theatre Over the last century the neurosurgical operating theatre has evolved with significant improvements in lighting, equipment, air flow, and sterility. Further developments with intraoperative imaging and integrated navigation systems are now beginning to emerge. (Fig. 4.1). Modern standard theatre complexes are divided into zones in order to minimize bacterial contamination. The operating room is an aseptic zone, the anaesthetic room is a clean zone and other areas are dirty zones. Standard theatres have a unidirectional air flow system to reduce the risk of airborne contamination with a minimum of 20 air changes per hour. Air is filtered before it enters the operating theatre and travels from the cleanest areas to the more contaminated ones, but the air flow is disrupted if theatre doors are left open. Smoke tests may be used to assess how air flows through operating theatres. Ultraclean theatres with laminar flow hoods have been developed for use during orthopaedic implant surgery but their use in neurosurgery is problematic because of the difficulties of positioning equipment, especially the operating microscope. The temperature and humidity of operating theatres are also regulated. An average temperature of 20–​22oC is comfortable for most theatre teams but warming blankets and the infusion of warmed intravenous fluids must be used to prevent patient hypothermia. Special consideration is needed to ensure and maintain sterility during instrumented spine and implant surgery. This may include minimizing the number and entry of theatre personnel and using a ‘no-​touch technique’ skin preparation, equipment sterilization, and intraoperative antibiotic prophylaxis to minimize the risks of infection are considered in detail in Chapter 94. In 2009 the World Health Organization (WHO) published a landmark study that found that implementing a systemic process of checks can reduced safety incidents by up to a third (Haynes et al., 2009). The National Patient Safety Agency’s (NPSA) launched the ‘five steps to safer surgery’ initiative in 2010 and all NHS healthcare organizations are now obliged to use the WHO Surgical Safety Checklist (Fig. 4.2). The aim of the initiative is to reduce harm associated with perioperative care and to support a change in culture within the theatre environment leading to better communication throughout the team. The five steps are:  (1) preoperative team briefing; (2) sign in; (3) time out; (4) sign out; and (5) postoperative team briefing.

Surgical patient positioning The central piece of equipment in the operating theatre is the operating table. Modern neurosurgical operating tables are radiolucent with interchangeable head and leg sections and are set on an offset column base that allows the table top to be moved cranially or caudally and tilted in all planes. Head-​up positioning is used for most neurosurgical procedures as it reduces intracranial pressure and venous haemorrhage. Excessive head-​up positioning however should normally be avoided because of the risks of air embolus, subdural postoperative pneumocephalus, and watershed infarction. Skull fixation is indicated in most microsurgical procedures to ensure that the head position is fixed in an optimal position with consideration given to the resulting angulation of the skull base, the desired surgical corridor, and the need to minimize brain retraction. Skull fixation is especially important when intraoperative image guidance systems are used, and it is also required for many cervical spine operations. Skull fixation is achieved by using a pin-​fixation device such as the three-​point fixation Mayfield clamp. Tension rings of 20, 40, 60, and 80 lbs are marked on the rocker pin with 60 lbs of tension providing secure fixation in an adult. When positioning the pins it is important to avoid the squamous part of the temporal bone, the supraorbital ridge, implanted shunt devices, and large frontal sinuses. Skull fixation should be used with care in paediatric patients and is generally best avoided as it can result in skull perforation and extradural haemorrhage. Various surgical positions are used in neurosurgery and each surgeon will develop their preferred position for each operation. Common surgical positions include the supine, prone, park-​bench, and sitting positions (Fig. 4.3). Close collaboration with the anaesthetist is essential in ensuring the patient is placed in the optimal surgical position but every member of the surgical team must remain vigilant to the potential complications of improper positioning. Bony prominences are at risk of pressure damage during surgery and must be well padded, while the patient’s eyes must also be covered to protect them from pressure and to avoid injury from alcoholic cleaning solutions. Peripheral nerve injuries may be caused by pressure or traction palsies with the ulnar and common peroneal nerves most at risk, especially when operating in the lateral (park-​bench) position. In the sitting position there is an increased

46

Section 1  Principles of neurological surgery

(a)

(b)

(c)

(d)

Fig. 4.1  Neurosurgical operating theatres. (A) Victor Horsley in the operating room at Queen Square Hospital, 1906; (B) 1930s operating theatre; (C) The modern neurosurgical operating theatre; (D) intraoperative MRI suite.

risk of air embolism (evidenced most rapidly by a rise in end tidal CO2), and there is a small risk of blindness associated with surgery performed in the prone position as a result of increased intraocular pressure and reduced optic nerve perfusion pressure. Improper skull fixation can also cause scalp necrosis or local haematomas.

Positioning equipment in theatre to optimize surgical performance The surgeon needs to consider the position of theatre personnel during a procedure. The surgical assistant should not be positioned between the surgeon and the scrub nurse, excess unused equipment should not be in theatre. Drills, ultrasonic aspirators, and bipolars must be positioned so the surgeon can use them with their dominant hand and without wires crossing the operative field. Image-​guidance cameras should be positioned in such a way that the microscope and theatre staff do not obstruct the line of site to the patient. Sound levels in theatre should not be at a level where the surgeon, anaesthetist, and other theatre staff cannot communicate effectively. Hypoglycaemia, coffee, and a full bladder may all affect a surgeon’s ability to conduct microsurgery. Surgeon comfort is not a luxury but a necessity for safe surgery. Some retractor systems and chairs will allow a surgeon to rest their elbows and wrists on a firm surface and this is very useful in minimizing fatigue. Some surgical positions,

notably the sitting position require that the surgeon operates with arms elevated in a way that makes fatigue problematic. Long operations may be best performed by surgical teams so that rests can be taken as required.

Theatre equipment The neurosurgeon needs to be familiar with many types of surgical equipment and to cover all these is beyond the scope of this chapter, but the following equipment merits special consideration.

Operating microscope There remains considerable controversy regarding who invented the first true microscope, but it is thought to have been developed in the late sixteenth century. It was Carl Zeiss, a German machinist working with the physicist Ernst Abbé who revolutionized lens making and microscope manufacturing. By the early twentieth century, the microscope was an integral part of laboratory medical research but was not used in the operating room until 1921 when Carl Nylén, a Swedish otolaryngologist, used a microscope to operate on a patient with chronic otitis media. Zeiss introduced their first series operating microscope in 1953 and in 1957 Theodore Kurze became the first neurosurgeon to use the microscope on a patient, removing a neurilemmoma of the seventh nerve in a 5-​year-​old

SURGEON REVIEWS: WHAT ARE THE CRITICAL OR UNEXPECTED STEPS, OPERATIVE DURATION, ANTICIPATED BLOOD LOSS?

DOES PATIENT HAVE A:

IS ESSENTIAL IMAGING DISPLAYED? YES NOT APPLICABLE

HAS ANTIBIOTIC PROPHYLAXIS BEEN GIVEN WITHIN THE LAST 60 MINUTES? YES NOT APPLICABLE

NURSING TEAM REVIEWS: HAS STERILITY (INCLUDING INDICATOR RESULTS) BEEN CONFIRMED? ARE THERE EQUIPMENT ISSUES OR ANY CONCERNS?

Reproduced with permission from the World Health Organization.

Fig. 4.2  WHO surgical safety checklist.

THIS CHECKLIST IS NOT INTENDED TO BE COMPREHENSIVE. ADDITIONS AND MODFICATIONS TO FIT LOCAL PRACTICE ARE ENCOURAGED.

RISK OF >500ML BLOOD LOSS (7ML/KG IN CHILDREN? NO YES, AND ADEQUATE INTRAVENOUS ACCESS AND FLUIDS PLANNED

DIFFICULT AIRWAY/ASPIRATION RISK? NO YES, AND EQUIPMENT/ASSISTANCE AVAILABLE

KNOWN ALLERGY? NO YES

ANAESTHESIA TEAM REVIEWS: ARE THERE ANY PATIENT-SPECIFIC CONCERNS?

ANTICIPATED CRITICAL EVENTS

PULSE OXIMETER ON PATIENT AND FUNCTIONING

SITE MARKED/NOT APPLICABLE

ANAESTHESIA SAFETY CHECK COMPLETED

CONFIRM ALL TEAM MEMBERS HAVE INTRODUCED THEMSELVES BY NAME AND ROLE

TIME OUT

Before skin incision

SURGEON, ANAESTHESIA PROFESSIONAL AND NURSE VERBALLY CONFIRM • PATIENT • SITE • PROCEDURE

PATIENT HAS CONFIRMED • IDENTITY • SITE • PROCEDURE • CONSENT

SING IN

Before induction of anaesthesia

SURGEON, ANAESTHESIA PROFESSIONAL AND NURSE REVIEW THE KEY CONCERNS FOR RECOVERY AND MANAGEMENT OF THIS PATIENT

WHETHER THERE ARE ANY EQUIPMENT PROBLEMS TO BE ADDRESSED

HOW THE SPECIMEN IS LABELLED (INCLUDING PATIENT NAME)

THAT INSTRUMENT, SPONGE AND NEEDLE COUNTS ARE CORRECT (OR NOT APPLICABLE)

THE NAME OF THE PROCEDURE RECORDED

NURSE VERBALLY CONFIRMS WITH THE TEAM:

SIGN OUT

Before patient leaves operating room

S S C (F E)

48

Section 1  Principles of neurological surgery

(a)

(b)

(c)

(d)

Fig. 4.3  Commonly patient positions for neurosurgery. (A) Supine position: pterional craniotomy (pictured); frontal/​subfrontal/​supraorbital craniotomy; parietal craniotomy; temporal craniotomy (shoulder roll placed under ipsilateral shoulder); transsphenoidal surgery. (B) Prone position: midline suboccipital craniectomy; spinal surgery (pictured). (C) lateral/​park-​bench/​three-​quarter prone position: retrosigmoid/​lateral suboccipital craniectomy (pictured); spinal surgery; (D) sitting position: suboccipital craniectomy; cervical spine surgery.

child. The first microneurovascular case was performed in 1960 by Raymond Donaghy. He also established a microsurgical training laboratory in Vermont where a young Gazi Yasgaril was sent to learn this emerging technique. More than anyone else, Yasgaril publicized the advantages of microscopic neurosurgery and made the operating microscope an integral part of modern neurosurgery (Kriss and Kriss, 1998). The modern microscope enables the operating surgeon to see three dimensional images at high magnification. The main components of a modern microscope are the objective lens, the binocular tube, the light source, a motorized zoom system, and the suspension system. The design of the operating microscope ensures that highquality lighting runs exactly parallel and very close to the optical path. The two eye fields are also close together, permitting binocular vision can be achieved at the bottom of a deep, narrow wound. The more a microscope is zoomed in the less depth of field the surgeon has so the surgeon should attempt to keep the microscope as close as possible to the patient. However, a good working distance between the microscope and the wound is required to allow instruments to be used without clashing with each other or the microscope. Modern microscopes have electric variable objective lenses with focal lengths

from 200 to 500 mm and the focal length is virtually identical to the working distance. Eyepieces are used to re-​magnify the intermediate image 10-​fold and can be used to compensate for ametropia. In order to allow free movement of the microscope preoperative balancing is essential, and drapes must not restrict microscope movements. Modern microscopes are also commonly fitted with two additional light sources; an infrared 800 nm light allows visualization of blood flow using intraoperative fluorescence and blue 400 nm light is used for fluorescence-​guided tumour surgery.

Surgical endoscope Neuroendoscopy was first performed in 1910 when urologist Victor Lespinasse used a paediatric cystoscope to treat paediatric hydrocephalus by attempting endoscopic coagulation of the choroid plexus (Abd-​El-​Barr and Cohen, 2013). Following this, Dandy and Mixter attempted endoscopic fenestration of the third ventricle for the treatment of hydrocephalus in the 1920s but true advances in neuroendoscopy came in the 1970s with technological developments in optics and electronics (Liu et al., 2004). The amount of intraoperative illumination and vision under the microscope is determined by the surgical approach and may be

CHAPTER 4  The operating theatre environment

significantly limited in cases where the surgical corridor is long and narrow. Modern endoscopes are able to bring increased light intensity into the surgical field and provide an extended viewing angle with better magnification in close-​up positions. Other potential advantages of endoscopic surgery include minimal tissue disruption and brain retraction, improved cosmetic results, shorter hospital stay, and reduced surgical morbidity. Two main types of endoscopes are available; namely, fibreoptic flexible and rigid rod lens systems. Both types of endoscope use a xenon light source that is transmitted via a fibreoptic cable to the endoscope. Fibreoptic flexible endoscopes transmit the image through a group of tightly packed fibreoptic threads and are smaller and more malleable than rigid solid lens endoscopes. Rigid endoscopes are more expensive than fibreoptic endoscopes, but they provide superior image resolution and light transmission through a series of lenses. Some endoscopes have a working channel that allows the insertion of compatible instruments such as microforceps, endoscopic scissors, monopolar diathermy, or Fogarty balloon catheters. Irrigation channels allow the constant irrigation of fluid to clear debris from the lens and bleeding from the surgical field. Endoscopes come with a variety of lens angles (0°, 30°, 70°, and 120°) to allow visualization of hidden parts of the surgical field although intervention is usually only performed using 0° and 30° scopes. Neuroendoscopic procedures were initially limited to the ventricles of the brain but the endoscope is now used in treating a wide spectrum of neurosurgical pathology. H.D. Jho pioneered the endonasal endoscopic approach for pituitary surgery (Jho and Alfieri, 2001)  and its application to other skull base pathologies continues to expand (Paluzzi et  al., 2012). The endoscope is also utilized in spinal surgery, particularly for thoracic procedures. It is important to select an appropriate endoscope with the desired characteristics for the procedure being performed. For instance, a non-​channelled fibreoptic scope may be appropriate for a diagnostic ventriculostomy but a rigid solid multichannelled lens endoscope is preferable when performing endoscopic endonasal surgery. Despite endoscopy being a rapidly expanding field endoscopic surgery is still not practised routinely by all neurosurgeons and there is consequently a steep learning curve. Appropriate training is essential and collaboration with consultants from other specialties, such as otolaryngology, who are experienced in endoscopy, may shorten the difficult learning phase and reduce operative time and surgical morbidity.

Fluorescent agents Indocyanine green (ICG) Indocyanine green (ICG) is the commonest fluorescent agent used for intraoperative angiography. It has a peak spectral absorption at around 800 nm following intraoperative intravenous injection and due to its tight binding to plasma proteins, becomes confined to the vascular system, permitting excellent intraoperative angiography. 5-​aminolevulinic acid (5-​ALA) Fluorescence-​guided tumour surgery is performed following the oral administration of 5-​aminolevulinic acid (5-​ALA) approximately 2–​4 hours before anaesthesia. Once administered patients should avoid direct sunlight due to the risk of developing a photosensitivity

skin reaction. 5-​ALA is a natural precursor of protoporphyrin IX (PpIX) in the haem biosynthesis pathway and excess 5-​ALA provides selective and abundant accumulation of PpIX in malignant glioma cells but only slight or no accumulation in the normal brain. 5-​ALA in itself is not fluorescent but PpIX is a highly fluorescent substance with maximal absorption at 440  nm, thus enabling the surgeon to differentiate between pathological tumour and normal brain intraoperatively. Fluorescein Fluorescein is a synthetic organic compound available as a dark orange/​red powder. It is soluble in water and is widely used as a fluorescent tracer for many applications. In neurosurgery, intrathecal fluorescein administration has become an important tool in the diagnosis and localization of cerebrospinal fluid (CSF) leak and is used intraoperatively to assist in localizing the site of skull base defects. Current recommendations of fluorescein administration suggest removing 10 ml of the patient’s CSF and replacing it with an equivalent injection of a 10 ml dilute solution of fluorescein into the intrathecal space. The dilute solution is bright green in colour and does not usually require visual augmentation; however, a blue light filter can assist in localization of fluorescein-​coloured CSF at the skull base in equivocal cases.

Drills Modern high-​speed surgical bone drills may be pneumatically or electrically powered. The rotational speed of a drill is measured in revolutions per minute (rpm) and pneumatic drills can generate up to 100 000 rpm. Electric drills have adjustable speeds from 200 to 75 000 rpm and are smaller, lighter, and more easily manoeuvrable, although they do not provide the same level of power as pneumatic drills. Various drill bits may be attached to the drill depending on the surgical need. Table 4.1 illustrates some of the commonly used drill attachments and how they are used in cranial and spinal neurosurgery.

Retractors The brain retractor was first introduced by Harvey Cushing and Victor Horsley in the early twentieth century and has since been a key neurosurgical instrument. Brain spatulas are either rectangular or tapered and are available in a variety of sizes. Articulated arms, such as the Yasargil Leyla retractor, were designed to stabilize the spatula while in use and various retractor systems have since been developed to provide a solid anchoring platform for retractors (Dujovny et al., 2010). Inappropriate or prolonged brain retraction has the potential to cause cortical contusions, vessel injury, brain swelling, and neurological deficit and the incidence of brain retraction injury has been observed to be as high as 10% in some series (Spetzler et al., 1992). Other surgeons advocate the avoidance of retractors altogether—​ and certainly careful patient positioning, so that gravity assists rather than hinders the surgeon, draining cerebral spinal fluid, mannitol infusion, skilled anaesthetics, and surgical patience all contribute to make excessive retraction unnecessary. Judicious and careful use of brain retraction can minimize retraction related complications and such techniques include the placing of surgical linteens under the

49

50

Section 1  Principles of neurological surgery

Table 4.1  Common drill bits and attachments used in cranial and spinal neurosurgery Match stick/​match head -​Side cutting burr, blunt tip improves safety when drilling down

Footed attachment

Cutting burr -​Rapid circumferential bone removal

Straight short/​large bore attachment

Diamond burr -​Provides extra control when drilling due to slow bone removal -​Generates heat for ‘hot-​drilling’ -​Can be used safely to drill bone over venous sinuses (e.g. translabyrinthine approach)

Straight variable exposure attachment (10 mm adjustment)

Acorn -​Rapid vertical bone removal -​Often used for creating ‘mini’ burr holes as part of turning a craniotomy flap

Angled short/​large bore attachment

Medtronic, Midas Rex.

retractors, avoiding firm retraction, and the use of silicon coated retractor blades.

Diathermy Diathermy is used for cutting tissues and haemostasis. Heat is generated by the passage of high-​frequency alternating current through body tissues. Currents of up to 500 mA are safe at frequencies of 400 kHz to 10 MHz and locally concentrated high-​density currents can generate temperatures of up to 1000°C. There is no stimulation of neuromuscular tissue at frequencies above 50 kHz. There are two distinct types of diathermy. Monopolar diathermy Monopolar diathermy uses a high-​power unit (400  W) to generate high-​ frequency alternating current that passes from a tipped electrode (high current density), via a ground plate (with a low current density) back to the generator. The voltage used changes the mode of diathermy. Cutting diathermy uses continuous low voltage output (500–​1000 V) to generate high local temperature that causes superficial tissue disruption and water vaporization but minimal vessel coagulation and haemostasis. The alternative coagulation setting generates pulsed high voltage output (up to 6000 V) of high-​frequency current at short intervals resulting in tissue water vaporization and vessel coagulation. Many diathermy machines will also allow a blend of cutting and coagulation. Monopolar diathermy does not work in a wet field and can cause heat to spread, damaging the surrounding (1–​2 cm) tissues. Incorrect placement of the patient electrode plate is the most common cause of accidental diathermy burns. To minimize the risk of complications the patient electrode plate should have a surface area of at least 70 cm2 and be placed with good contact on dry, shaved skin away

from bony prominences, metal prostheses (e.g. hip replacements) and scar tissue. Monopolar diathermy should be used with caution in patients with a history of arrhythmias and cannot be used in patients with a cardiac pacemaker due to the risk of causing cardiac arrest. There are case reports of alcoholic skin prep catching fire if not left to dry before diathermy is used. Bipolar diathermy Bipolar diathermy uses lower power (50  W) to generate high-​ frequency alternating current that passes between the two tips of diathermy forceps. It uses a 1 MHz waveform for both cutting and coagulation and because it only uses 140 V it can be used in patients with pacemakers and defibrillators. There is no need to use a patient electrode plate and bipolar diathermy can be used effectively in a wet environment. Forceps vary in length and tip sharpness. Coagulated, charred tissue can stick to the forceps and they must be kept clean to avoid this risk. Modifications available include irrigating bipolar forceps and cooler tipped forceps that make this risk less likely.

Ultrasonic aspirators Ultrasonic aspirators are vibrating surgical instruments which utilize ultrasound as a physical energy to fragment (cavitate), emulsify, and remove unwanted tissue. Examples of ultrasonic aspirators used in neurosurgery include the Cavitron Ultrasonic Surgical Aspirator (CUSA) (Integra) and Sonopet (Stryker). The hollow titanium tip vibrates along its longitudinal axis generating an ultrasound frequency of 20–​60 kHz and all models allow the surgeon to specify the level of vibration, irrigation, and suction in order to achieve maximal control during tissue dissection. The CUSA includes a supplementary Tissue Select function that further regulates the relative

CHAPTER 4  The operating theatre environment

tissue fragmentation rate and some aspirators (e.g. Sonopet) permit fine bone dissection by coupling longitudinal vibration with torsional vibration. Ultrasonic surgical aspirators provide ultrasonic cavitation, irrigation and aspiration in a single hand-​held unit which helps to limit instrument traffic within the surgical field and improve the surgeon’s view. The selective fragmentation of target neural tissues also reduces the risk of vessel injury, but one must be mindful of causing collateral damage from thermal injury and the risk of developing indistinct ‘false’ tissue planes.

Neuronavigation Imaging techniques have revolutionized the practice of neurosurgery and have become essential in guiding diagnosis, preoperative planning, and intraoperative navigation. In particular, image-​guided surgery has become an indispensable modern neurosurgical technique and is applicable to numerous neurosurgical procedures including neuro-​oncology, vascular, and functional neurosurgery. The aim of stereotactic surgery is to register the anatomical or imaging space to the surgical or physical space in which the patient is located and in which surgery is to take place. Image guidance allows for smaller more precisely positioned incisions and accurate localization of lesions. Stereotactic neurosurgery may be frame-​based or frameless.

Haemostatic agents Achieving and maintaining haemostasis in neurosurgery is critical to the outcome. Various haemostatic agents have been developed and target different elements of the clotting cascade (Fig. 4.4 and Table 4.2) (Grant, 2007). Some of these agents may swell in the postoperative period and care should be taken to avoid leaving haemostatic agents in places where swelling will cause neurological compression. This is especially relevant in the spine. Some haemostatic agents may result in MR artefacts making postoperative assessment of residual tumour difficult. Problematic bleeding can also be controlled by changing the patient position, tranexamic acid, haemostatic clips, or muscle patches. Bees wax was developed for use in neurosurgery by Victor Horsley, and is now widely used to stop bleeding from bone edges.

Cryoprecipitate & fibrin sealants (e.g. TisseelVH, Hemaseel APR)

VIIa

XIIa XI

Collagen (e.g. Avitene)

Extrinsic pathway (tissue factor-mediated)

Intrinsic pathway (contact system) XII

Vasoconstriction

The concept of stereotactic localization was first applied in 1908 in the device of Victor Horsley and Robert Clark although the apparatus was only developed for work on small to medium sized experimental animals (Fig. 4.5A). The first successful cranial application of stereotactic surgery in humans is credited to the team of Ernest Speigel and Henry Wycis in 1947 but it was Lars Leksell, a Swedish neurosurgeon, who developed the arc-​centred stereotactic frame system following a visit to Wycis in Philadelphia in 1949. Leksell’s frame consisted of a rectangular base ring attached to the skull

Vascular injury

Topical haemostatic agents (e.g. Gelfoam, Surgicel, Oxycel)

Mechanical pressure

Frame-​based stereotactic neurosurgery

VII

XIa IX

Tissue factor (Thromboplastin)

IXa

Recombinant factor VIIa (e.g. Novoseven)

VIIIa, Ca++ Haemostasis X

Xa

X Va

Final common pathway (tissue factor-mediated)

Prothrombin (II)

Thrombin (e.g. FloSeal)

Thrombin (Iia)

Fibrinogen (I)

Fibrinmonomer (Ia) XIIIa Fibrin polyment (Fibrin clot) Fibrinolysis

Aprotinin Clot lysis

Fig. 4.4  Haemostatic adjuncts and their mechanism of action. Adapted from Surgery, Volume 142, issue 4, Gerald A. Grant, Update on hemostasis: neurosurgery, pp. S55–​S60, Copyright (2007), with permission from Elsevier.

51

52

Section 1  Principles of neurological surgery

Table 4.2  Haemostatic adjuncts and their mechanism of action Haemostatic adjunct

Trade name

Active component

Topical haemostatic agents

Gelfoam

Gelatine sponge

Surgical

Oxidized regenerated cellulose

Oxycel

Oxidized cellulose

Collagen

Avitene

Microcrystalline collagen (reacts with platelets)

Fibrin sealants

Tisseel Haemaseel APR

Fibrinogen/​factor XIII concentrate (cryoprecipitate) and thrombin mixture

Thrombin

FloSeal

Gelatine matrix and thrombin mixture

Aprotinin

Fibrinolytic inhibitor of plasmin (inhibits fibrinolysis)

Adapted from Surgery, Volume 142, issue 4, Gerald A. Grant, Update on hemostasis: neurosurgery, pp. S55–​S60, Copyright (2007), with permission from Elsevier.

with four pins and used three polar coordinates (angle, depth, and anterior–​posterior location). The three coordinates always indicate the centre of the arc, representing the target, and the arc-​centred device allows maximum flexibility in choosing the probe entry point and trajectory. The frame has been modified over the ensuing years, but remarkably remains very similar in function and appearance to the original 1949 device (Fig. 4.5B). Another commonly used frame system is the Cosman-​Roberts-​ Wells (CRW) Precision™Arc System (Fig. 4.5C). It has an N-​shaped ‘picket-​fence’ localizer ring and the lesion is placed at the centre of a stereotactic sphere with a fixed radius that allows free selection of an unlimited number of entry points. The CRW Precision™ Arc System also has a phantom base that can be used to confirm the target before applying the settings to the patient. Computed tomography (CT) and MRI compatible models of both Leksell and CRW Precision™ Arc Systems are available. The advent of high-​resolution MR imaging has transformed stereotactic surgical planning. Stereotaxy is no longer reliant on anatomical atlases and is usually performed on a computer workstation that allows the planning of targets and trajectories on a computer-​generated 3D model. Frame-​based stereotaxy has the highest degree of precision and, although largely superseded in general intracranial practice by frameless stereotactic techniques, continues to be used for cases where target placement is critical including, for example, functional lesioning and electrode or catheter placement, biopsy of small or deep lesions, and gamma knife stereotactic radiosurgery.

Frameless stereotactic neurosurgery Frameless stereotaxy involves the use of bony landmarks, facial features, or fiducial markers to coregister the patient’s skull relative to a 3D volumetric image obtained prior to surgery. Fiducials are adhesive markers, visible on both CT and MRI, and are placed on a patient’s head, covering as much of the geometry of the head as possible but centred on the target lesion. For optical registration, the patient’s head is fixed using a pin-​fixation device and infrared reflective spheres arranged in a predetermined geometrical configuration and anchored to the same fixation device. An infrared camera detects light reflected from the spheres.

Registration is performed by the surgeon and computer software correlates the imaging studies with the defined anatomical landmarks of the patient’s head. Registration remains accurate so long as the patient’s head remains in same position relative to the reflectors. Registration accuracy can be enhanced by collecting as many registration points as possible, by using fiducials and by spreading anatomical registration points around the head as much as possible. Electromagnetic (EM) registration involves generating an electromagnetic field around the patient’s head in order to triangulate the position of instruments and intraoperative tools. EM navigation does not require the head to be fixed and is especially useful when the head cannot be fixed easily (e.g. shunt or paediatric surgery). EM systems track the tip of the surgical instrument and can therefore be used to track the tip of a flexible endoscope. A registration error is a computer-​generated number that denotes the degree of deviation from the ideal image-​to-​patient correlation. Acceptable registration accuracy depends on the size of the target lesion. Volumetric registration enables the surgeon to view images in three orthogonal planes and once registered, instruments fitted with reflective spheres may be used by the surgeon to relate the position of the instrument to the images in real time. Additional image sequences such as functional imaging and tractography can also be merged with the volumetric data enabling the surgeon to view these intraoperatively. Image-​guidance systems have also been adapted for use in spinal instrumentation surgery and can be linked to operating microscopes so that the focal point of the microscope is used as the image-​guidance pointer and so that image-​guidance data can be overlaid in the operators view down the microscope. Frameless image-​guidance systems are a significant technological advance and an essential piece of modern neurosurgical equipment however one must be aware of their limitations. Inaccuracies in registration may occur as a result of intraoperative brain shift following loss of cerebrospinal fluid or tumour resection, because of movement of the patient, the reference frame, or fiducials, and because of a loss of line of sight between the camera and infrared reflectors. In cases where submillimetre accuracy is required and a fixed trajectory is acceptable, a frame-​based system is preferable.

Intraoperative imaging X-​ray Mobile X-​ray image intensifiers are commonly used in the neurosurgical operating theatre for a variety of applications, including localization of the correct anatomical level during spinal surgery, imaging during spinal instrumentation, and for localizing other bony anatomy such as skull base foraminae or the pituitary fossa. Mobile image intensifiers typically consist of a single unit with a low intensity X-​ray source and an image intensifier positioned directly opposite each other on a mount. The mount, often referred to as a C-​arm, may be rotated and translated around the patient as required. Mobile image intensifiers can provide good spatial resolution of bony anatomy and are quick and cheap to use, although the image quality is dependent on operator experience and the patient’s body habitus. Imaging during spinal instrumentation helps to ensure that the desired anatomical alignment is achieved and assists in the accurate

CHAPTER 4  The operating theatre environment

(a)

(b)

(c)

Fig. 4.5  Stereotactic frames: (A) Horsley-​Clarke stereotactic frame (1908); (B) Leksell stereotactic frame; (C) CRW Precision™Arc System.

placement of implanted metalwork, such as pedicle screws. A radiolucent operating table should be used if fluoroscopic imaging is to be used intraoperatively to permit anterior–​posterior views in

addition to lateral views. Ensuring accurate screw positioning from 2D imaging can be difficult, particularly where there is abnormal anatomy or where complex instrumentation is used. Newer imaging

53

54

Section 1  Principles of neurological surgery

techniques are being introduced to overcome this difficulty including dedicated 3D equipment such as ‘3D C-​arms’ or the ‘O-​arm’, and neuronavigation techniques incorporating registered fluoroscopic images with preoperative 3D CT scans.

Ultrasound Intraoperative ultrasound is a relatively inexpensive technique that has the potential to provide surgeons with fast, real-​time imaging. Ultrasound imaging relies on the use of high-​frequency sound waves to image tissues. Sound is defined as mechanical energy travelling through matter as longitudinal waves producing alternating zones of compression and rarefaction. Intraoperative ultrasound uses a transducer to transmit ultrasonic waves in the range of 5–​ 10 MHz into tissues. Where there are interfaces of different tissue density, energy is reflected back by the tissue towards the transducer as ‘echoes’. The amount of energy reflected depends on a tissue’s acoustic impedance, which is proportional to the density of the tissue propagating the sound wave. Echo vibrations are then converted to electrical pulses and used to produce representative images. Greyscale or Brightness (B-​mode) is the mainstay of anatomical ultrasound imaging and provides 2D greyscale images of tissues based on their relative echogenicity. Doppler ultrasound can also be incorporated into intraoperative ultrasound. The doppler effect describes a change in the frequency of sound when a sound wave encounters a moving interface, such as blood within an artery and blood flow may be mapped in colour on real-​time ultrasound images. Intraoperative ultrasound is very operator dependent and image interpretation is associated with a learning curve as most surgeons are unfamiliar with this imaging modality. Ultrasound is unable to penetrate through bone and so cannot be used for planning a craniotomy and it has limited penetration through to deeper structures meaning that careful case selection is required. Intraoperative ultrasound has been used in treating a variety of pathologies but is most frequently used during glioma surgery. Some navigation machines are now also capable of coregistering the ultrasound pictures with a reconstructed plane from the pre-​op MR scan. Intraoperative doppler ultrasound has a role in neurovascular surgery and some surgeons also use ultrasound to demonstrate adequate CSF flow following foramen magnum decompression.

Intraoperative MRI (iMRI) Neurosurgeons have become increasingly dependent on image guidance to perform optimal, safe, and efficient surgery. Routine neuronavigation systems do not allow the neurosurgeon to adjust for intraoperative brain shift and this consideration led to the development of intraoperative MRI systems (Fig. 4.1D) (Hall and Truwit, 2008). The principal feature of iMRI systems is the ability to acquire contemporaneous imaging during surgery to inform surgeons on the extent of resection achieved in order to guide further surgical decision making and most iMRI systems also allow the new images to be registered onto the neuronavigation system. iMRI systems may be vertically or horizontally open. In vertically open systems the patient remains within a fixed position and the surgeon works within the confines of the scanner. All surgical instruments must be MR compatible and various levels of shielding of other equipment must be maintained. Images can be easily and quickly acquired at any point during the procedure without moving

the patient or the magnet however vertically open systems are typically low field scanners (0.3–​0.5 T) so the standard of the images obtained are not of the same quality as those generated by standard diagnostic imaging techniques. Horizontally open iMRI systems include a high-​field magnet (1.5–​ 3.0 T) located within an integrated operating theatre (Fig. 4.1D). The table can slide or pivot out of a marked magnetic field line (5-​Gauss line), allowing surgery to be performed in the usual manner using standard instruments and equipment until a scan is required. New high-​field horizontally open iMRI scanners produce higher quality images allowing greater versatility and acquisition of functional MRI, MR angiography, MR venography, MR spectroscopy, and tractography. The disadvantages of this system include the significant cost required to build the intraoperative suite and the increased time spent intraoperatively manoeuvring the patient in and out of the scanner.

Intraoperative neurophysiological monitoring Intraoperative neurophysiological monitoring offers a real-​time, continuous assessment of the nervous system integrity and is considered essential for various types of neurosurgical procedure (Gonzalez et al., 2009). It is vital to consult with an expert neurophysiologist before the procedure to ensure that the most appropriate modality is selected and results must always be interpreted in the context of the surgical procedure being performed. For this reason, it is helpful to have a neurophysiologist in the operating theatre during surgery. Intraoperative monitoring can be significantly affected by anaesthetic agents so close collaboration with the anaesthetist is also essential.

Sensory evoked potentials The visual, auditory, and somatosensory systems can be stimulated and evoked potentials can record the biofeedback response of the nervous system to these external stimuli. The recordings give information on the sensory pathway from the peripheral nerve to the level of the cortex.

Somatosensory evoked potentials (SSEP) Somatosensory evoked potentials (SSEPs) are produced by short (0.1–​1.0 msec) electrical stimulation (1–​5 Hz) of a peripheral nerve and the generated responses are recorded over a proximal part of the nerve or sensory pathway. Recording electrodes are placed in named, reproducible locations. Each location is designated by a letter; N (negative deflection) or P (positive deflection) and a number that corresponds to the latency of wave. The median or tibial nerve is usually stimulated and SSEPs are monitored in terms of latency and amplitude. The SSEP waveform is often quite variable and should therefore be monitored before surgery begins and throughout the procedure. To separate the SSEP signal from background noise a technique called signal averaging is used to combine a large number of SSEP signals. The time required to perform signal averaging means that SSEP changes may not be seen until a few minutes after the neurological injury has occurred. A decrease of about 50% in amplitude or a 10% increase in latency of the SSEP waveform is considered to be significant. Inhalational anaesthetic agents, fluctuations in blood pressure,

CHAPTER 4  The operating theatre environment

hypothermia, and oxygenation can all affect SSEPs to varying degrees so close consultation with the anaesthetist is needed beforehand to ensure stable anaesthesia is achieved. In thoracic spinal surgery, the upper limb SSEP signal can be used as a control. Continuous SSEP monitoring of the dorsal columns is commonly used during spinal cord and complex spine surgery and is considered essential for scoliosis surgery (Thirumala et al., 2014). Cortical mapping with SSEPs can also be used in epilepsy surgery or sensorimotor and insular cortex tumour surgery.

Monitoring of pudendal MEPs and sacral EMG recording from the anal sphincter is used when performing selective dorsal rhizotomy for the treatment of spasticity and helps identify nerve rootlets that can be selectively sectioned. Unlike SSEPs, MEPs are not usually monitored continuously as the patient may move following a stimulation. The surgeon and neurophysiologist also need to work together to ensure frequent MEPs are obtained without disrupting the surgery.

Visual evoked potentials (VEP) and brainstem auditory evoked potentials (BAEP)

A recording electrode is placed inside a needle and inserted into a muscle to record the MUAP of that muscle. Signals are then converted into sound to demonstrate muscle activity. Evoked EMG uses electrical stimulation to identify specific nerves whereas free-​running EMG is used to identify any interruptions to the nerve’s normal electrical discharge. Spontaneous benign discharges produce short non-​ sustained discharges (likened to the sound of ‘popcorn’) and indicate proximity of the stimulus to the nerve whereas a neurotonic discharge is prolonged and indicates ongoing nerve injury. Facial nerve monitoring is the commonest type of intraoperative EMG used in cranial neurosurgery whereby recording electrodes are placed in the orbicularis oris and orbicularis oculis muscles. Muscle activity is recorded either during a surgeon’s intentional stimulation of the nerve or by inadvertent nerve stimulation during manipulation, mechanical, or thermal injury. Optimal facial nerve EMG recordings are obtained when neuromuscular blockade is avoided but facial muscles appear relatively resistant to the effects of muscle relaxants. Hartman’s solution irrigation should be used instead of saline irrigation to avoid stimulation of the facial nerve. EMGs are also used to monitor muscle group activity during selective dorsal root rhizotomy for spasticity and spinal root monitoring is considered a key component of multimodality spinal monitoring used during spinal instrumentation.

Intraoperative visual evoked potentials (VEPs) and brainstem auditory evoked potentials (BAEPs) are less commonly used. VEPs can be evoked by patterned or unpatterned stimuli and may be used during surgery around the optic chiasm or optic tract however results under anaesthesia are not reliable. BAEPs record a waveform composed of seven discrete waves after an auditory stimulus is delivered to the external auditory canal. Each wave corresponds to a specific locus of the auditory system from the organ of Corti to the auditory cortex. BAEPs may be used during hearing preservation vestibular schwannoma surgery or during microvascular decompression for hemifacial spasm.

Motor evoked potentials Motor evoked potentials (MEPs) are generated by stimulating the motor cortex, spinal cord, or a peripheral nerve with an electrical or magnetic stimulus. Transcranial magnetic stimulation is preferred in preoperative mapping because it is less invasive but transcranial electric stimulation using corkscrew or needle electrodes is usually performed intraoperatively. Neurogenic MEPs are recorded directly from the spinal cord or from peripheral nerves. Neurogenic MEPs generate two types of wave: direct propagation of the stimulus along the corticospinal tract generates direct waves (D waves), while indirect waves (I waves) are caused by activation of adjacent cortex. Myogenic MEPs record the muscle response (M response) of a specified muscle by using electromyography (EMG) recording. The M response reflects the motor unit action potential (MUAP) and represents the summated action potentials of the muscle fibres of one motor unit. Inhalation anaesthetics have a significant effect on MEPs, in particular I wave responses. D waves are not affected by inhalation anaesthetics, but the α-​motor neurone is blocked at higher doses, even when stimulated by transcranial electric stimulation. Intravenous anaesthetics only have a mild effect on MEPs and total IV anaesthetic using propofol and etomidate is typically preferred when using MEPs as they are less suppressive than gases at similar depths of anaesthesia. However, excessively high doses of propofol may be similarly suppressive to inhalation agents. Benzodiazepines are usually avoided because they eliminate MEP responses. Neuromuscular blocking agents do not affect neurophysiological monitoring and can actually enhance I wave recordings but MUAPs cannot be recorded when the neuromuscular transmission of acetylcholine is completely blocked. The anaesthetist observes the M-​responses to evaluate the degree of neuromuscular blockade. MEPs can monitor the dorsolateral and ventral spinal cord tracts during spinal cord and complex spine surgery and complements SSEP monitoring of the dorsal columns (Schwartz et  al., 2007).

Electromyography

Intraoperative cortical mapping (brain mapping) Intraoperative brain mapping is typically used to locate the motor, sensory, or speech cortex during epilepsy or tumour surgery in and around eloquent areas. Phase reversal SSEPs may be utilized in anaesthetised patients to localize the primary sensory or motor cortex whereby a strip grid is placed on the surface of the brain perpendicular to the anticipated orientation of the central sulcus. A phase reversal of the N20/​P20 peak indicates that those electrodes straddle the central sulcus. In awake patients, direct cortical stimulation may be used to map the motor cortex. For speech mapping in the temporal lobe, a recording electrode strip is placed on the brain surface and the cortex is then stimulated using a bipolar cortical stimulator. The stimulating current is increased in small 2 mA increments, up to a maximum of 10 mA, while observing for any after discharges (akin to focal seizures). Once the patient’s threshold for after discharges has been established, patients are asked to name objects shown on picture cards while the cortex continues to be stimulated and any paraphasic errors or speech arrest are noted. The aforementioned steps are then repeated so that the patient’s speech area may be mapped out on the brain’s surface. Awake craniotomy is usually required for brain mapping, especially for the speech areas. The key to a successful awake procedure is patient preparation. Good patient cooperation is required, and careful patient selection

55

56

Section 1  Principles of neurological surgery

is needed to ensure the patient fully understands what will happen during surgery. Various anaesthetic and surgical techniques have been described in the context of performing awake craniotomy but a successful procedure mandates anticipation of specific problems and clinical vigilance (Erickson and Cole, 2012). Fully awake surgery is performed after the infiltration of local anaesthetic into the scalp or following regional nerve block of the scalp. The asleep-​awake-​asleep (AAA) technique uses general anaesthesia, with or without the use of an airway, during the opening and closing portions with emergence of patients from anaesthesia in the interim. However, the more commonly advocated anaesthesia technique for the opening and closing portions is termed monitored anaesthesia care (also called conscious sedation). The same medications used in the AAA technique are given (propofol and fentanyl) but they are given in pulses and at lower doses with the goal of providing a smooth transition to alertness and obviating the difficulties of airway intervention (Erickson and Cole, 2012).

FURTHER READING Abd-​El-​Barr, M.M. & Cohen, A.R. (2013). The origin and evolution of neuroendoscopy. Childs Nerv Syst, 29, 727–​37. Gonzalez, A.A., Jeyanandarajsan, D., Hansen, C., Zada, G. &, Hsieh, P.C. (2009). Intraoperative neurophysiological monitoring during spine surgery: a review. Neurosurg Focus, 27(4), E6. Hall, W.A. & Truwit, C.L. (2008). Intraoperative MR-​guided neurosurgery. J Magn Reson Imaging, 27, 368–​75. Haynes, A.B., et  al.; Safe Surgery Saves Lives Study Group (2009). A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med, 360(5),  491–​9. Kriss, T.C. & Kriss, V.M. (1998). History of the operating microscope: from magnifying glass to microneurosurgery. Neurosurgery, 42(4), 899–​907.

REFERENCES Abd-​El-​Barr, M.M. & Cohen, A.R. (2013). The origin and evolution of neuroendoscopy. Childs Nerv Syst, 29, 727–​37. Dujovny, M., Ibe, O., Perlin, A., & Ryder, T. (2010). Brain retractor systems. Neurol Res, 32(7), 675–​83. Erickson, K.M. & Cole, D.J. (2012). Anesthetic considerations for awake craniotomy for epilepsy and functional neurosurgery. Anesthesiol Clin, 30(2), 241–​68. Gonzalez, A.A., Jeyanandarajan, D., Hansen, C., Zada, G., & Hsieh, P.C. (2009). Intraoperative neurophysiological monitoring during spine surgery: a review. Neurosurg Focus, 27(4), E6. Grant, A.G. (2007). Update on hemostasis:  neurosurgery. Surgery, 142(4 Suppl), S55–​60.

Hall, W.A. & Truwit, C.L. (2008). Intraoperative MR-​guided neurosurgery. J Magn Reson Imaging, 27, 368–​75. Haynes A.B., et  al.; Safe Surgery Saves Lives Study Group. (2009). A surgical safety checklist to reduce morbidity and mortality in a global population. N Engl J Med, 360(5),  491–​9. Jho, H.D. & Alfieri, A. (2001). Endoscopic endonasal pituitary surgery: evolution of surgical technique and equipment. Minim Invasive Neurosurg, 44(1),  1–​12. Kriss, T.C. & Kriss, V.M. (1998). History of the operating microscope: from magnifying glass to microneurosurgery. Neurosurgery, 42(4), 899–​907. Liu, C.Y., Wang, M.Y., & Apuzzo, M.L.J. (2004). The physics of image formation in the neuroendoscope. Childs Nerv Syst, 20, 777–​82. Paluzzi, A., Gardner, P., Fernandez-​Miranda, J.C., & Snyderman C. (2012). The expanding role of endoscopic skull base surgery. Br J Neurosurg, 26(5), 649–​61. Schwartz, D.M., Auerbach, J.D., Dormans, J.P., et  al. (2007). Neurophysiological detection of impending spinal cord injury during scoliosis surgery. J Bone Joint Surg Am, 89(11), 2440–​9. Spetzler, R.F., Daspit, C.P., & Pappas, C.T. (1992). The combined supra-​ and infratentorial approach for lesions of the petrous and clival regions: experience with 46 cases. J Neurosurg, 76, 588–​99. Thirumala, P.D., Bodily, L., Tint, D., et  al. (2014). Somatosensory-​ evoked potential monitoring during instrumented scoliosis corrective procedures: validity revisited. Spine J, 14(8), 1572–​80.

RELATED LINKS TO EBRAIN Advanced Microsurgical Skills. https://​learning.ebrain.net/​course/​ view.php?id=93 Image-​guided Neurosurgery. https://​learning.ebrain.net/​course/​view. php?id=92 Intraoperative Imaging Techniques. https://​learning.ebrain.net/​ course/​view.php?id=89 Intraoperative Neurosurgical Electrophysiological Monitoring. https://​learning.ebrain.net/​course/​view.php?id=489 Preparing the Operative Site. https://​learning.ebrain.net/​course/​view. php?id=96 Principles of Awake Craniotomy. https://​learning.ebrain.net/​course/​ view.php?id=496 Surgery in Children. https://​learning.ebrain.net/​course/​view. php?id=73 The Modern Theatre Environment. https://​learning.ebrain.net/​course/​ view.php?id=97 The Operating Microscope. https://​learning.ebrain.net/​course/​view. php?id=94 Use of an Endoscope. https://​learning.ebrain.net/​course/​view. php?id=91 Use of Standard Neurosurgical Equipment. https://​learning.ebrain. net/​course/​view.php?id=70

5

Perioperative care of the neurosurgical patient Karol P. Budohoski, Alessandro Scudellari, Sylvia Karcheva, and Derek Duane

Introduction Multidisciplinary involvement in the care and planning of patients for neurosurgical procedures is now a common and necessary requirement to limit complications and promote better neurological outcomes. In the perioperative period, failings in clinical management can be devastating and cause severe disability or death (Wacker and Staender, 2014). The shared goal of the neurosurgeon along with the neuroanaesthetist acting as perioperative physician, is to facilitate optimization of the patient’s clinical and psychological readiness for the operative intervention and subsequent recovery while preventing avoidable harm. To enable this, an established perioperative pathway is followed that includes performing a preoperative evaluation, completing appropriate investigations, communicating risk, complying with patient safety measures, promoting teamwork, and pursuing specific intra-​and postoperative management strategies. These processes will be discussed in detail in the following sections, while highlighting their importance in the preservation of normal neurological function.

Preoperative anaesthetic evaluation The purpose of the preoperative assessment is to optimize a patient’s comorbidities and formulate a perioperative management plan. This review should follow the time-​honoured tradition of history taking, examination, deciding on necessary investigations, risk assessment analysis, information sharing, and seeking informed consent. When assessing patients preoperatively, use is made of the American Society of Anesthesiologists (ASA) physical status classification system. This is one accepted method used for the stratification of a patient’s pre-​existing health status (Table 5.1).

Airway Airway difficulty resulting in hypoxia and hypercarbia is the single most important cause of anaesthesia related morbidity and mortality. Therefore, it is vital to conduct a comprehensive examination of the airway preoperatively. Many tests to predict a difficult airway assess parameters such as mouth opening with grading of

pharyngeal structures, neck mobility, and anatomical oropharyngeal measurements but none has the desired specificity and positive predictive value to be consistently useful. In the neurosurgical population, patients with cervical spine disease leading to limited movement, instability requiring a cervical collar or traction weights and those with acromegaly can all cause potential airway difficulty for the anaesthetist. To ensure adequate preoperative hydration and to help avoid gastric contents aspiration during intubation, recent guidelines encourage anaesthetists to allow adults and children to drink clear fluids up to 2 hours before elective surgery, while solid food should be prohibited for 6 hours (Smith et al., 2011).

Respiratory system Optimizing pulmonary function is important to ensure the patient can be adequately oxygenated and ventilated intraoperatively, thus permitting respiratory manipulation of cerebrovascular physiology. Patients with neurological disorders have a higher risk of suffering aspiration pneumonia and neurogenic pulmonary oedema, both of which can adversely affect neurological outcome and survival. Patients with a smoking habit should stop at least 6–​8 weeks prior to surgery allowing for improvement in lung function and a better perioperative outcome. Acute exacerbations of infective or inflammatory airway diseases will need treatment and deferment of anaesthesia for up to 6 weeks to permit airway sensitivity to resolve. Chronic pulmonary diseases should be assessed clinically and where there is evidence of worsening impairment, further imaging and functional testing will sometimes be necessary to help clarify the risk of complications which can include postoperative ventilation. Patients who suffer from obstructive sleep apnoea especially those with acromegaly or Cushing’s disease are often very sensitive to opioid analgesics and sedative agents. These patients can develop impaired respiratory gas exchange, atelectasis, and pulmonary infections without the routine use of continuous positive airway pressure prior to surgery.

Cardiovascular system The prevalence of heart disease is common in neurosurgical patients and a careful assessment of the functional state of the cardiovascular system is necessary to reliably predict perioperative and long-​term

58

Section 1  Principles of neurological surgery

Table 5.1  American Society of Anesthesiologists (ASA) physical status classification system ASA grade

Definition

I

A healthy patient with no systemic disease

II

Mild to moderate systemic disease

III

Severe systemic disease imposing functional limitation on patient

IV

Severe systemic disease which is a constant threat to life

V

Moribund patient who is not expected to survive with or without the operation

VI

A declared brain-​dead patient whose organs are being removed for donor purposes For emergency cases the suffix ‘E’ is used

ASA Physical Status Classification System is reprinted with permission of the American Society of Anesthesiologists, 1061 American Lane, Schaumburg, Illinois 60173–​4973.

cardiac complications (Fleisher et  al., 2014). In patients with the usual risk factors and a history of shortness of breath, fatigue, or chest pain while performing daily living activities, usually have poor cardiovascular reserve requiring further investigation. The revised cardiac risk index (Table 5.2) using six independent variables is helpful in assessing the neurosurgical patient’s risk of cardiac complications and mortality (Ford et al., 2010). Poorly controlled hypertension (systolic >180 mmHg and mean arterial pressure (MAP) over 110  mmHg) increases cerebrovascular resistance and causes a right shift of the cerebral blood flow autoregulatory curve resulting in poor tolerance to acute hypotension. If surgery can be postponed, weeks to months of treatment may be needed to adequately control the blood pressure otherwise perioperative management by intravenous agents can be instituted for emergency surgery albeit with increased risk of complications. Patients who have had a recent myocardial infarction or coronary stent intervention are at a higher risk of further myocardial damage if operated on within six months. Heart failure, in particular, is a significant risk for perioperative morbidity and mortality and requires optimal treatment prior to surgery in line with international guidelines and local cardiology advice to ensure a good outcome (Yancy et al., 2013). Patients with a newly suspected heart valve lesion need echocardiography evaluation to quantify its severity. Some patients may need a valvular intervention before their elective neurosurgical procedure

Table 5.2  Revised Cardiac Risk Index Revised Cardiac Risk Index

to reduce their perioperative cardiac risk. In general, knowing the severity of the valvular heart disease allows choice of the appropriate anaesthetic technique as well as deciding on the best perioperative monitoring and level of postoperative care for the patient. Cardiac dysrhythmias and conduction disorders occur more frequently with increasing age such that all patients over the age of 55 years should have a routine electrocardiograph prior to surgery. The presence of a dysrhythmia in the preoperative setting should prompt investigations into the underlying cause. Atrial fibrillation (AF) is the most common tachydysrhythmia. Patients with rate-​ controlled AF who are clinically stable generally do not require special evaluation in the perioperative period other than adjustment of anticoagulation. Patients who develop conduction abnormalities leading to bradydysrhythmias, or episodes of supraventricular or ventricular tachycardias require referral to a cardiologist for further evaluation and possible cardioversion to sinus rhythm or insertion of a permanent transvenous pacing device. Those patients who present for neurosurgery with implanted cardiac pacemakers or defibrillators will need their device checked for any signs of malfunction after surgery if electrocoagulation is used during the procedure.

Neurological system A preoperative record of the patient’s neurological state using the Glasgow Coma Scale (GCS) score should be made to identify pre-​ existing deficits. Pupil size and reaction to light should be noted. Assessment of cranial nerve integrity is important to establish problems with bulbar function along with gag and cough reflexes. Examination of sensory and motor function will establish the presence of any anomalies. This baseline neurological examination is essential to allow comparison with a patient’s postoperative state for detection of any possible deterioration in function.

Other medical conditions Other medical disorders affecting the endocrine, renal, hepatic, neuromuscular, and metabolic systems will need to be optimized prior to neurosurgery. Electrolytes, (especially sodium and potassium) and water disturbances resulting from conditions such as cerebral salt wasting (CSW) syndrome, or inappropriate antidiuretic hormone secretion (SIADH) and diabetes insipidus (DI) will need appropriate management preoperatively to ensure anaesthesia can be administered safely. Hyperglycaemia, a commonly encountered problem in neurosurgical patients mainly due to corticosteroid treatment, has well documented adverse effects on injured brain tissue (Kramer et al., 2012). Insulin therapy instituted preoperatively according to local guidelines may be necessary to achieve adequate glycaemic control.

1

High risk type of surgery

Prehospital medications and allergy status

2

Ischaemic heart disease (includes any of the following: history of MI, positive exercise test, current chest pain, nitrate use, or electrocardio­ graphy with Q waves)

3

Congestive heart failure

4

History of cerebrovascular disease

5

Preoperative treatment with insulin

6

Preoperative serum creatinine >2.0 mg/​dl

The continuation of a patient’s prehospital medications is usual up to the day of surgery. However, depending on neurosurgical advice, specific recommendations in relation to corticosteroids, and anticonvulsants are often made. Most antihypertensive agents and analgesic drugs for chronic pain should also be continued. Diabetic patients can sometimes omit oral hypoglycaemic agents but to optimize their blood glucose they may need an insulin sliding-​scale regimen preoperatively. Anticoagulant and antiplatelet agents should be stopped at the appropriate time to allow clotting and platelet function to return to normal. This decision should be informed by haematology

Reproduced with permission from Thomas H. Lee et al., Derivation and Prospective Validation of a Simple Index for Prediction of Cardiac Risk of Major Noncardiac Surgery, Circulation, Volume 100, Issue 1, Copyright © 1999 Wolters Kluwer Health, Inc.

CHAPTER 5  Perioperative care of the neurosurgical patient

or cardiology advice and based on a consideration of procedure urgency, along with the risks associated with the potential for haemorrhage and thrombosis. A patient’s history of drug allergies should be documented prior to anaesthesia. Any record of anaphylaxis, facial swelling, bronchospasm, hypotension, or profound cutaneous erythema related to a particular agent are all highly significant and warrant avoidance. Drug allergies are often readily documented to antibiotics, intravenous contrast, and anaesthetic agents but enquiry should also be made about reactions due to latex, iodine, chlorhexidine, adhesive tapes, and other substances used during the surgical procedure.

Preoperative investigations Guidelines from both the ASA and the National Institute for Health and Care excellence (NICE) do not advocate routine preoperative investigations but rather tests based on the type of procedure (minor to complex major) and the functional impairment imposed by comorbidities. Biochemical, haematological, and electrophysiological testing along with imaging evaluate the degree of pathology and dysfunction of a particular organ system thus providing information on physiological reserve and the potential response to anaesthetic agents. A risk factor assessment based on a patient’s history and examination should indicate the need to check the haemoglobin level preoperatively. When significant blood loss is anticipated, blood group type and cross-​matching will be required. Establishing a history of familial bleeding disorders or other risk factors for coagulopathy necessitates preoperative coagulation testing. However, there is little evidence that abnormal tests are predictive of perioperative haemorrhage (Seicean et al., 2012). Comorbidities affecting the renal and hepatic systems either directly or indirectly require testing for sodium, potassium, creatinine, urea, estimated glomerular filtration rate, and liver enzyme levels. In patients with convulsive disorders, assessment of magnesium, calcium, and anticonvulsant drug levels may be appropriate. When endocrine abnormalities are suspected, baseline investigation of the hypothalamic-​pituitary-​end-​organ axis should be established. Other blood investigations should be guided by the severity of a patient’s comorbidities and the associated functional impairment. An electrocardiogram is useful as a baseline investigation in patients with signs and symptoms of ischaemic heart disease. It may highlight hypertensive heart disease, rhythm abnormalities, or occult ischaemia. Further investigation of the cardiovascular status of the patient using a combination of exercise stress testing, echocardiography, nuclear imaging studies, and coronary angiography is usually indicated with findings of exertional dyspnoea and chest pain. Preoperative chest radiographs and pulmonary function testing are necessary only in patients whose symptoms and underlying lung disease is progressive.

Intraoperative anaesthetic management Induction and maintenance of anaesthesia The basic principles of neuroanaesthesia for cranial neurosurgery are essentially twofold; (1) the provision of a relaxed brain to enable good

operative surgical conditions, and (2)  the maintenance of cerebral perfusion pressure (CPP) and cerebral oxygenation. Controversy remains regarding the best practice to enable these overall goals such that interventions are often empirical rather than evidence-​based. Venous access Once a patient is deemed fit for surgery their pathway from induction to emergence follows specific anaesthetic interventions. An inhalational induction is more usual in children and uncooperative or needle phobic adult patients but more commonly, an intravenous induction is used after initial venous access has been obtained. In general, both arterial and venous access can be achieved under local or general anaesthesia. Depending on the type of surgery and the severity of the patient’s comorbidities, a central venous catheter is sometimes indicated. Type of induction and drugs used Intravenous induction of anaesthesia is commonly performed using short-​acting agents such as propofol, thiopentone, etomidate, or ketamine (see Table 5.3 and Fig. 5.1). Each agent has their own benefits and drawbacks but there is little to choose between them. In general, propofol and thiopentone are fast-​acting but cause significant hypotension. Etomidate and ketamine cause less haemodynamic instability but the former agent can suppress adrenocortical activity for many hours and the later drug can cause postoperative hallucinations and nightmares. If intravenous access is not initially possible, there is concern over possible airway control or if the patient is a child, then an inhalational induction with a volatile agent and a mixture of oxygen and air is performed (see Table 5.4). This form of induction generally takes longer, is associated with excitatory phenomena, and if airway issues arise there is no immediate intravenous access to allow the administration of neuromuscular blocking agents. Airway management After induction of anaesthesia, intubation or the insertion of a supraglottic airway device is facilitated using a non-​depolarizing neuromuscular blocking agent such as atracurium or rocuronium. Depending on the dose, these agents can take 1–​3 minutes to reach peak effect and they are more commonly implicated in causing histamine release or rarely anaphylaxis. Securing the airway with a plain or reinforced cuffed endotracheal tube (ETT) is commonplace for most neurosurgical procedures. The ETT should be taped with care to avoid inadvertent dislodgement, especially for patients in the prone position. If a difficult airway is predicted from the scoring systems used in the preoperative assessment and the patient has cervical spine pathology, then a fibreoptic-​assisted awake or asleep intubation may be required. An unpredicted difficult airway is no more common than in the general surgical population and the standard management algorithms for this clinical scenario should be followed. After intubation, patients are ventilated using an air/​oxygen mixture of gases and an inhalational agent is added if required. In emergency situations, a rapid sequence (‘crash’) induction is commonly performed. This involves using 100% oxygen for up to 5 minutes to preoxygenate the patient, the administration of a short-​ acting intravenous sedative agent (e.g. propofol or thiopentone) and simultaneous application of cricoid pressure to help avoid aspiration of the contents of a presumed full stomach. Once the patient is unconscious, a rapid administration of a depolarizing neuromuscular

59

60

Section 1  Principles of neurological surgery

Table 5.3  Effects of intravenous anaesthetic agents on cerebrovascular, respiratory, and cardiovascular parameters Target receptor

ICP

CBF

CMRO2

CO2 reactivity

Auto-​ regulation

Respiratory depression

Cardiovascular depression

Indication

Propofol

GABA

↓↓

↓

↓↓

→

→

++

++

Induction and maintenance of anaesthesia

Thiopental

GABA

↓↓

↓

↓↓

→

→

++

+

Induction of anaesthesia; Deep barbiturate coma

Etomidate

GABA

↓↓

↓

↓↓

→

→

++

-​/​+

Induction of anaesthesia (cardiovascularly unstable patients)

Ketamine

NMDA

↑/​↓

↑

↑

→

→

-​

-​

Hypovolemic trauma patient

Dexmedetomidine

α2 adrenoceptor

↓

↓

↓

→

→

-​

Concentration dependent

Functional neurosurgery

Benzodiazepine

GABA

↓

↓

↓

→

→

+

+

Sedative premedication; ICU sedation

↓ decrease: ↑ increase: → unchanged a) Reproduced with permission from Srejic, U., ‘Volatile Anesthetic Agents’, pp. 62–​6, in R. Adapa, D. Duane, A. Gelb, & A. Gupta (Eds.), Gupta and Gelb's Essentials of Neuroanesthesia and Neurointensive Care, 2018 © Cambridge University Press. b) This article was adapted from Cottrell and Young's Neuroanesthesia, James Cottrell William Young, Effects of Anesthetic Agents and other Drugs on Cerebral Blood Flow, Metabolism and Intracranial Pressure, pp. 78–​95, Copyright Elsevier (2010).

blocking agent (e.g. suxamethonium) or a higher dose of a non-​ depolarizing agent is given. The time to adequate intubating conditions is usually 90-​seconds following this regimen. Monitoring Standard physiological parameters are monitored during neuroanaesthesia. These include:  heart rate; three or five lead

electrocardiography; pulse oximetry; non-​invasive or invasive blood pressure; respiratory rate; end tidal carbon dioxide; nasopharyngeal temperature; and the concentrations of oxygen and inhalational agent. Ventilation parameters (i.e. tidal volume, peak inspiratory pressure, and positive end-​expiratory pressure (PEEP)) are adjusted to achieve an arterial oxygen level of greater than 13 kPa and a carbon dioxide level of between 4.5 and 5 kPa. A urinary catheter is

N2O Hypocapnia Total CMRO2

Ketamine (limbic system)

5 ml/ 100g/ min

Maximal metabolic suppression (isoelectric EEG)

Benzodiazepines opioids

Barbiturates propofol etomidate

Decreased CBF

Ketamine (total CBF)

Inhalational anaesthetics ↑ MAC leads to ↑ CBF

Normal CBF 50 ml (100 g brain)–1 min–1

Increased CBF

Fig. 5.1  Effects of intravenous anaesthetic agents on cerebrovascular, respiratory, and cardiovascular parameters.

CHAPTER 5  Perioperative care of the neurosurgical patient

Table 5.4  The effects of inhalational anaesthetics on cerebrovascular parameters ICP

CBF

CMRO2

CO2 reactivity

Autoregulation

Cardiovascular depression

Sevoflurane

→ or ↑

↑ or →

↓↓

Dose related ↓

Dose related ↓

+

Desflurane

↑ or →

↑ or ↓

↓↓

Dose related ↓

Dose related ↓

+

Isoflurane

→ or ↑

↑ or →

↓↓

Dose related ↓

Dose related ↓

+

Reproduced with permission from Srejic, U., “Volatile Anesthetic Agents”, pp. 62–​6, in R. Adapa, D. Duane, A. Gelb, & A. Gupta (Eds.), Gupta and Gelb's Essentials of Neuroanesthesia and Neurointensive Care, 2018 © Cambridge University Press.

sometimes inserted to measure renal output during protracted surgery, for anticipated significant blood loss or where hyperosmolar therapy is used. Neuromuscular junction monitoring is essential in all cases where muscle paralysis is maintained throughout anaesthesia. In certain circumstances, electroencephalography (EEG), electrophysiological monitoring of cranial nerves, and the measurement of motor or sensory evoked potentials can help prevent postoperative neurological deficits. Processed EEG-​based depth of anaesthesia monitoring using the bispectral index or spectral entropy is recommended as an option in patients undergoing total intravenous anaesthesia (TIVA) with muscle relaxants, deep anaesthesia without the use of muscle relaxants and those at a higher risk of unintended awareness (Escallier et al., 2014). Positioning After induction of anaesthesia, potent stimuli resulting from laryngoscopy, insertion of a bladder or central vascular catheter, the application of skull pins, and positioning can cause a hypertensive response that must be obtunded to avoid an unexpected rise in intracranial pressure. This complication may also arise in patients positioned without a 15-​to 30-​degree head elevation, when the neck is flexed or turned resulting in obstruction to cerebral venous outflow, when PEEP greater than 10 cm H2O is used and in the prone position which obstructs adequate ventilation. Hyperflexion of the neck should also be avoided as it may result in tongue and airway oedema thus preventing safe extubation at the end of surgery. The sitting position is well known for its complications of hypotension, venous air embolism (VAE), macroglossia, pneumocephalus, and nerve injuries. Despite this, if a strict team protocol is followed the position is generally well tolerated and excellent operative conditions are ensured. In prone positioning, it is the responsibility of both surgical and anaesthetic teams to adopt strategies to minimize the risk of organ system pressure-​injuries, myocutaneous necrosis, peripheral neuropathies, and ophthalmic complications (DePasse et al., 2015). Maintenance of anaesthesia General anaesthesia comprises a triad of hypnosis, analgesia, and muscle relaxation. In general, intravenous anaesthetic agents couple the reduction in cerebral metabolic rate of oxygen consumption (CMRO2) and cerebral blood flow (CBF), preserve cerebral autoregulation (CAR) and carbon dioxide reactivity while reportedly inhibiting excitotoxicity and free radical-​mediated lipid peroxidation. However, they can lower the MAP and CPP with

resulting potentially harmful effects on cerebral oxygenation and intracranial pressure (ICP). In comparison, inhalational agents, to a varying degree, suppress CMRO2 in a dose-​dependent manner but cause cerebral vasodilatation (desflurane ~ isoflurane > sevoflurane) which may promote an increase in CBF and hence ICP unless compensated by hypocapnia induced vasoconstriction. These agents also cause a dose-​dependent decrease in CPP, CAR, and carbon dioxide reactivity. Despite the difference in the effects of both intravenous and inhalational agents on cerebral haemodynamics, there is not yet a consensus as to the best anaesthetic technique for neurosurgical procedures (Chui et  al., 2014). However, in patients with known intracranial hypertension, intravenous-​ maintained anaesthesia would seem to have the theoretical advantage while inhalational agents should be used by those more experienced. In procedures where electrophysiological monitoring of the brain and spinal cord are undertaken, TIVA, rather than inhalational agents, has only a minimal influence on measured potential. During operations on the vertebral column and spinal cord, the anaesthetic considerations should be the same as for cranial interventions.

Intraoperative events and anaesthetic emergence Respiratory and cardiovascular The ability to control arterial oxygen and carbon dioxide levels by altering the patient’s ventilation is essential in the management of intracranial hypertension. Important causes that can prevent adequate ventilation and require immediate remedy include: (1) inadvertent endobronchial intubation; (2)  sputum blocking or kinking of the ETT; (3)  bronchospasm; (4)  faulty prone positioning; and (5)  an unrecognized pneumothorax after central venous access or tunnelling during a cerebrospinal fluid (CSF) diversion procedure. Intraoperative arterial hypotension (systolic blood pressure Na+ >150  mmol/​litre) are common in patients carbon dioxide levels, along with temperature and glucose, within after subarachnoid haemorrhage, traumatic brain injury, intracra- normal range. nial tumours, and pituitary surgery. Patients with these conditions Many patients are the victims of accidents resulting in polytrauma usually benefit from measuring urine output and serum sodium for and may already be intubated and ventilated when received in the

63

64

Section 1  Principles of neurological surgery

operating room. Life-​threatening injuries are managed as a priority for cardiopulmonary stability before neurosurgical intervention. Every attempt should be made to assess and document GCS score, pupil size, reactivity, pre-​existing neurological deficits, axial spine stability, and any other available relevant medico-​social history. Aside from routine biochemical and haematological tests, analysis of clotting function and the availability of blood products must be ensured prior to any intervention. Induction of anaesthesia is usually with propofol or thiopentone, as both significantly reduce CMRO2 and therefore ICP. However, most anaesthetic agents, except ketamine and etomidate, cause a degree of hypotension which can be offset with vasopressors. Suxamethonium, a depolarizing neuromuscular blocking agent, is often used to facilitate airway management as part of a rapid sequence induction. Neither depolarizing or non-​depolarizing agents have significant effects on ICP. Intubation can be difficult in neuro-​ trauma patients who require a cervical collar and manual in-​line stabilization for possible cervical spine injury. It is essential to have a backup plan for a failed intubation to avoid any rise in ICP due to hypoxia and hypercapnia. During surgery, standard monitoring is augmented with measurements of temperature, urine output, and neuromuscular function. Invasive arterial blood pressure monitoring is essential and it allows regular blood gas measurements. Large-​bore peripheral venous access is sufficient in most cases while central venous access, unless essential, should not delay surgery. Total intravenous anaesthesia using infusions of propofol and remifentanil is a widely used technique for the maintenance of anaesthesia in brain injured patients. Inhalational agents can be used but in low concentrations (0.5 cm in children or >1.5 cm in adults) • Two or more cutaneous/​ subcutaneous neurofibromas or one plexiform neurofibroma • Axillary or groin freckling • Optic pathway glioma • Two or more Lisch nodules (iris hamartomas seen on slit lamp examination) • Bony dysplasia (sphenoid wing dysplasia, bowing of long bone pseudarthrosis) • First-​degree relative with NF1 Genetic analysis may clarify the diagnosis in young children who do not meet the diagnostic criteria and is useful for affected individuals to aid reproductive decision-​making.

Case history Case 38.1 A 24-​year-​old woman presents with extensive cutaneous neurofibromas and 3 months of weight loss, with L3 dermatomal numbness, grade 3/​5 power in her knee and brisk reflexes. MRI identified a retroperitoneal mass invading the L3 vertebral body with loss of height, canal impingement, and thecal compression. She underwent successful L1-​L5 laminectomy decompression and pedicle screw fixation. Histology confirmed malignant peripheral nerve sheath tumour (Fig. 38.1). Case 38.2 A 5-​year-​old girl presents with delayed milestones, multiple café au lait spots and increasing head circumference. MRI identified chiasmal and optic nerve lesions (Fig. 38.2).

Screening Regular clinical review in childhood under the care of a paediatrician with referral to a specialist NF1 clinic when indicated is recommended. Surveillance brain imaging is not advocated. Genetic testing of the NF1 gene should be made available for affected patients as well as prenatal diagnostic genetic testing when requested (Ferner et al., 2007).

450

Section 9  Tumours and skull base—tumour syndromes

(a)

(b)

(c)

Fig. 38.1  (A) Sagittal T2-​weighted MRI and (B) axial T2-​weighted MRI showing retroperitoneal mass invading L3 vertebral body with loss of height, canal impingement, and thecal compression. (C) Postoperative CT reconstruction.

Clinical features Cutaneous manifestations are typically the first sign of NF1. Café au lait spots, light brown patches on the trunk with well demarcated edges, are often present at birth, increasing in number during the first few years of life. The frequency and age of onset of NF1 clinical manifestations are shown in Table 38.1. Neurofibromas Neurofibromas are benign peripheral nerve sheath tumours. They may be focal cutaneous or subcutaneous, diffuse, or nodular plexiform lesions. Typically indolent, they follow a benign course, developing in the late teens and early twenties. They may cause cosmetic problems and can be removed, with risk of recurrence and hypertrophic scarring. Subcutaneous neurofibromas, evident on palpation of skin, may be tender or cause paraesthesia in the distribution of the affected nerve. They rarely undergo malignant transformation. Plexiform neurofibromas grow along the nerve length and may involve multiple nerve branches and plexi. Multiple lesions can develop on nerve trunks and can be nodular, resulting in thickened, potentially large, masses which infiltrate soft tissue and cause bone (a)

hypertrophy. When multiple spinal nerve roots are involved, patients may present with myelopathy secondary to cord compression. Central nervous system (CNS) tumours Pilocytic astrocytoma (WHO grade 1) is the most common paediatric brain tumour in the context of NF1, affecting up to 15% of patients. The majority are optic pathway gliomas (OPG) (Listernick et  al., 2007), with 20% occurring in the brainstem. They typically follow a more indolent course than sporadic counterparts. OPG can involve the optic nerve, chiasm, or hypothalamus. They remain asymptomatic in the majority of patients but can cause visual loss, proptosis, endocrine disturbance, and are more often symptomatic in children. Gliomas of the brainstem or cerebellum occur more frequently in NF1 patients than the general population, and of those, are more frequent in NF1 patients treated with radiotherapy for OPG in childhood (Kleinerman, 2009). Hyperintense lesions on T2 MRI (unidentified bright objects or UBOs) are seen in more than half of NF1 patients, most commonly in children aged 8–​16 years. They are pathognomonic of the condition and most likely caused by aberrant myelination or gliosis. (b)

Fig. 38.2  (A) Axial T2-​weighted MRI showing thickened, coiled optic nerves; high signal lesion (unidentified bright object) posterior to fourth ventricle. (B) Axial T1-​weighted MRI of the orbit showing thickened, coiled optic nerves, left greater than right.

CHAPTER 38 Neurophakomatoses

Table 38.1  Frequency and age of onset of major clinical manifestations of NF1 Clinical manifestation

Frequency (%)

Age of onset

Café au lait patches

>99

Birth to 12 y

Skin-​fold freckling

85

3 y to adolescence

Lisch nodules

90–​95

>3 y

Cutaneous neurofibromas

>99

>7 y (usually late adolescence)

Plexiform neurofibromas

30 (visible) –​50 (on imaging)

Birth to 18 y

Disfiguring facial plexiform neurofibromas

3–​5

Birth to 5 y

Malignant peripheral nerve sheath tumours

2–​5 (8–​13% lifetime risk)

5–​75 y

Scoliosis

10

Birth to 18 y

Scoliosis requiring surgery

5

Birth to 18 y

Pseudarthrosis of tibia

2

Birth to 3 y

Renal artery stenosis

2

Lifelong

Phaeochromocytoma

2

>10 y

Severe cognitive impairment (IQ 90%) in seizure frequency at one-​year follow-​up. Surgery should be considered for resection of tumours causing obstructive hydrocephalus, raised intracranial pressure, radiographic tumour progression, or development of new focal neurological deficits.

Controversy The discovery of upregulation of the mTOR pathway upregulation in TSC-​associated tumours presents new possibilities for treatment strategies. Sirolimus (aka rapamycin) is an immunosuppressant which binds to and inhibits the ability of mTOR to phosphorylate downstream targets. It has been approved as an immunosuppressant drug since 2001 and has been found to have application in the treatment of TSC (Jozwiak et al., 2006). Everolimus, a derivative of sirolimus, is effective in diminishing the volume of lesions in patients with renal angiomyolipomas, subependymal giant cell tumours (Franz et al., 2006) and sporadic lymphangioleiomyomatosis. It was approved by the FDA in the USA in 2010 for the treatment of TSC-​associated SGCTs not amenable to surgical resection (Krueger et al., 2010). In 2013, the results of a double blind, placebo-​controlled trial of everolimus to treat TSC-​associated subependymal giant cell astrocytomas were published—​the EXIST-​1 trial (Franz et al., 2013). These also showed at least 50% reduction in the volume of the tumour as compared with placebo (Franz et al., 2013). Surgery may no longer be the treatment of choice for tumours associated with TSC.

Von Hippel-​Lindau (VHL) VHL has an estimated incidence of 1 in 36 000 live births and prevalence of 1 in 39 000–​53 000 (Melmon and Rosen, 1964; Maher et al., 2011). It has autosomal dominant inheritance due to a mutation in the VHL tumour suppressor gene on chromosome 3p25 (Latif et al., 1993). The protein product regulates cell proliferation through the HIF (hypoxia inducible factor) pathway. 20% of cases arise from de novo mutations. VHL has marked phenotypic variability and over 90% penetrance by the age of 65 (Maher et al., 1990). It is a multisystem disorder characterized by the development of multiple haemangioblastomas (HGBs) of the central nervous system, retina, and other visceral tumours. Unlike other neurocutaneous syndromes, VHL is not typically associated with skin findings. Haemangioblastomas are composed of vascular and stromal cells.

Diagnostic criteria See Box 38.4 for the diagnostic criteria for VHL. In uncertain cases, genetic testing may be employed and is useful for management of the wider family. MRI brain imaging, renal

Box 38.4  Diagnostic criteria for VHL Confirmed family history The presence of one single VHL tumour is sufficient for diagnosis (e.g. retinal or CNS HGB, clear cell RCC, phaeochromocytoma, pancreatic endocrine tumour, or endolymphatic sac tumour). No family history All of the tumours typically found in VHL disease can occur as a sporadic event. Therefore, diagnosis requires the presence of two tumours: either two CNS HGBs or one CNS HGB and a visceral tumour (with exception of epididymal and renal cysts which are frequent in general population). Reprinted from The American Journal of Medicine, Volume 36, issue 4, Kenneth L. Melmon, Saul W. Rosen, Lindau’s disease: Review of the literature and study of a large kindred, pp. 595–​617, Copyright (1964), with permission from Elsevier.

ultrasound, abdominal CT, and urinary amine estimations are required to complete the evaluation.

Illustrative case Case 38.6 A 33-​year-​old man presented with headaches, dizziness, and unsteadiness. Imaging identified a posterior fossa lesion with local mass effect (Fig. 38.6). He underwent surgical resection; histopathology confirmed haemangioblastoma and serum VHL mutation. He also underwent cryotherapy for retinal angioblastoma, and nephrectomy for renal cell carcinoma (RCC). His pancreatic and epididymal cysts are under observation, together with ongoing CNS surveillance.

Screening One-​third of CNS HGBs are VHL-​associated. VHL-​HGBs tend to present 10  years earlier than sporadic lesions. Patients with a new diagnosis of CNS HGB should have a thorough family history and clinical review for VHL-​associated manifestations, and whole neuro-​axis imaging. If VHL is suspected, the patient should be referred to a specialist centre for further evaluation. Once the diagnosis is confirmed surveillance screening can be commenced. Screening may include annual ophthalmological assessment for retinal angioma beginning in early childhood, cranial MRI for haemangioblastoma and abdominal MRI for renal cell carcinoma and pancreatic tumours beginning in adolescence, and annual blood pressure monitoring and urinary studies for phaeochromocytoma (Maher, 2004).

Clinical features CNS HGBs occur in 60–​80% of VHL patients (Richard et al., 2000; Wanebo et al., 2003) most frequently in the cerebellum, spinal canal and, less frequently, brainstem and nerve roots. Retinal HGBs occur in more than 50% of patients, are often the earliest manifestation of the disease and can cause sudden visual loss. Other associated lesions include visceral cysts (renal, pancreatic and epididymal cystadenomas which are common but rarely compromise organ function), RCC, adrenal and extra-​adrenal phaeochromocytomas, non-​functioning pancreatic endocrine cancers, endolymphatic sac tumours and, occasionally, head and neck paragangliomas. RCC is the most common malignancy and cause of death, followed by CNS HGB (Maher et al., 1990).

CHAPTER 38 Neurophakomatoses

(a)

(b)

Fig. 38.6  (A) Contrast-​enhanced, T1-​weighted MRI showing solid enhancing HGB in right cerebellar hemisphere; (B) T2-​weighted MRI showing solid lesion, associated flow voids and perilesional oedema, and local mass effect.

Management VHL carries a lifetime risk of developing multiple organ tumours. Patients should be managed by a dedicated specialist VHL team. Surveillance should be undertaken to allow early detection and treatment of lesions. Complete surgical resection of accessible, symptomatic CNS HGBs is the mainstay of treatment. One study of 80 patients showed no recurrence after 61 months of follow-​up (Jagannathan et al., 2008); others quote a recurrence rate of 17% at 53 months (Conway et al., 2001). Factors which influence the decision to proceed with surgery include neurological symptoms, evidence of growth on surveillance imaging (such that resection would be more difficult), or the presence of an enlarging cyst or syrinx (Lonser et al., 2003; Weil et al., 2003). Preoperative work up should include MRI of the whole neuro-​ axis, investigation for phaeochromocytoma due to the risk of perioperative hypertensive crisis during anaesthesia, and consideration of preoperative embolization of the tumour. When planning surgery for posterior fossa HGBs, consider external ventricular drainage and release of CSF, opening the cistern magna to further reduce CSF pressure, ultrasound-​guided drainage of peritumoural or intratumoural cysts, intravenous mannitol, and/​ or furosemide and the use of transdural ultrasound to confirm the adequacy of bone removal for optimum exposure. The principle of resection of cystic HGBs is to remove the mural nodule in order to prevent recurrence of the cyst. The cyst wall itself does not need to be removed unless there is evidence of tumour within the cyst wall on MRI, or under direct vision at the time of surgery (Jagannathan et al., 2008). The standard principles of microsurgery should be applied to dissecting the tumour (i.e. define the plane between tumour margin and cerebellum circumferentially) and progressively dissect deeper tumour margins in layers. Vessels require coagulation and sharp division as they enter or leave the tumour capsule which is assisted by non-​stick bipolar forceps. The tumour softens as the feeding supply is interrupted, and gentle retraction of the capsule becomes possible to enhance exposure and resection of the tumour. VEGF is one of the downstream targets of the VHL protein. Sunitinib, a VEGF receptor inhibitor, is licensed for use in advanced metastatic renal cell carcinoma and has been tested for

use in VHL-​associated RCC. Its effect on CNS HGBs has not been proven, but research is ongoing (Jonasch et al., 2011). Stereotactic radiosurgery (SRS) may be considered for the management of surgically unresectable lesions. Few studies have yet determined the long-​ term safety and efficacy of the treatment (Asthagiri et al., 2010).

Controversy SRS can be used to treat VHL-​associated HGBs, less than 3cm3 volume, tumours incompletely excised, or unresectable. It may minimize the number of surgical resections in an individual with multiple HGBs. Tumour control rates post-​SRS range from 83% at 5 years, 61% at 10 years, 51% at 15 years (20 VHL patients with 44 HGBs) (Asthagiri et al., 2010) to 93% at 5 years, and 82% at 10 years (80 VHL patients with 335 HGBs) (Kano et al., 2015). Surgery, if without complication, is curative. Neurological status remains stable or improves following surgical resection (98% of patient improved at 3 months post-​op cerebellar HGB resection (Lonser et al., 2003; Weil et al., 2003; Jagannathan et al., 2008). Data from a retrospective review of 80 VHL patients showed new tumour development after SRS in 41% of patients at 67 months follow-​up (rates of 7% at 1 year, 21% at 3 years, 43% at 5 years, 84% at 10 years) (Kano et al., 2015). This is comparable to new tumour formation in VHL without SRS (Wanebo et al., 2003; Ammerman et al., 2006). As yet there are no reports of secondary malignancy or malignant transformation, but this should be considered. In a study of 186 patients with HGBs treated with SRS, only 13 (7%) developed adverse radiation effects, with one patient dying due to these side effects (Kano et al., 2015). However, this data was not defined as VHL or sporadic HGB patients. Patients with VHL are at risk of developing multiple HGBs and repeated radiotherapy is not always advisable.

Ataxia telangiectasia (AT) AT is no longer classified as one of the neurophakomatoses. It has historically been included in neurosurgical texts due to its case report associations with some CNS tumours and with

457

458

Section 9  Tumours and skull base—tumour syndromes

angiographically occult capillary malformations. Unlike the other neurophakomatoses it is not associated with cutaneous or ocular manifestations, and it exhibits autosomal recessive inheritance. It is of most relevance to neurologists due to it causing cerebellar degeneration and progressive ataxia.

FURTHER READING Evans, D.G.R. (2009). Neurofibromatosis type 2 (NF2): a clinical and molecular review. Orphanet J Rare Dis, 4, 16. Evans, D.G.R., Baser, M.E., O’Reilly, B., et al. (2005). Management of the patient and family with neurofibromatosis 2: a consensus conference statement. Br J Neurosurg, 19(1),  5–​12. Ferner, R.E., Huson, S.M., Thomas, N., et al., (2007). Guidelines for the diagnosis and management of individuals with neurofibromatosis 1. J Medical Genet, 44(2),  81–​8. Fisher, M.J., Loguidice, M., Gutmann, D.H., et al. (2012). Visual outcomes in children with neurofibromatosis type 1-​associated optic pathway glioma following chemotherapy:  a multicenter retrospective analysis. Neuro Oncol, 14(6),  790–​7. Franz, D.N., Belousova, E., Sparagana, S., et  al. (2013). Efficacy and safety of everolimus for subependymal giant cell astrocytomas associated with tuberous sclerosis complex (EXIST-​1): a multicentre, randomised, placebo-​controlled phase 3 trial. Lancet, 381(9861), 125–​32. Krueger, D.A. & Northrup, H. (2013). Tuberous sclerosis complex surveillance and management: recommendations of the 2012 international tuberous sclerosis complex consensus conference. Pediatr Neurol, 49(4), 255–​65. Maher, E.R., Neumann, H.P., & Richard, S. (2011). von Hippel-​Lindau disease:  a clinical and scientific review. Eur J Hum Genet, 19(6), 617–​23. Northrup, H. & Krueger, D.A. (2013). Tuberous sclerosis complex diagnostic criteria update: recommendations of the 2012 International Tuberous Sclerosis Complex Consensus Conference. Pediatr Neurol, 49(4), 243–​54.

REFERENCES Ammerman, J.M., Lonser, R.R., Dambrosia, J., Butman, J.A., & Oldfield, E.H. (2006). Long-​term natural history of hemangioblastomas in patients with von Hippel-​Lindau disease:  implications for treatment. J Neurosurg, 105(2), 248–​55. Asthagiri, A.R., Mehta, G.U., Zach, L., et al. (2010). Prospective evaluation of radiosurgery for hemangioblastomas in von Hippel-​Lindau disease. Neuro Oncol, 12(1),  80–​6. Baser, M.E., Evans, D.G.R., Jackler, R.K., Sujansky, E., & Rubenstein, A. (2000). Neurofibromatosis 2, radiosurgery and malignant nervous system tumours. Br J Cancer, 82(4), 998. Baser, M.E., Friedman, J.M., Joe, H., et al. (2011). Empirical development of improved diagnostic criteria for neurofibromatosis 2. Genet Med, 13(6), 576–​81. Conway, J.E., Chou, D., Clatterbuck, R.E., Brem, H., Long, D.M., & Rigamonti, D. (2001). Hemangioblastomas of the central nervous system in von Hippel-​ Lindau syndrome and sporadic disease. Neurosurgery, 48(1),  55–​63. Curatolo, P., Bombardieri, R., & Cerminara, C. (2006). Current management for epilepsy in tuberous sclerosis complex. Curr Opin Neurol, 19(2), 119–​23. Curatolo, P., Bombardieri, R., & Jozwiak, S. (2008). Tuberous sclerosis. Lancet, 372(9639), 657–​68.

Dodgshun, A.J., Elder, J.E., Hansford, J.R., & Sullivan, M.J. (2015). Long-​term visual outcome after chemotherapy for optic pathway glioma in children:  site and age are strongly predictive. Cancer, 121(23), 4190–​6. Evans, D.G., Huson, S.M., Donnai, D., et al. (1992). A clinical study of type 2 neurofibromatosis. Q J Med, 84(304), 603–​18. Evans, D.G.R., Baser, M.E., McGaughran, J., Sharif, S., Howard, E., & Moran, A. (2002). Malignant peripheral nerve sheath tumours in neurofibromatosis 1. J Med Genet, 39(5), 311–​14. Evans, D.G.R., Baser, M.E., O’Reilly, B., et al. (2005). Management of the patient and family with neurofibromatosis 2: a consensus conference statement. Br J Neurosurg, 19(1),  5–​12. Evans, D.G.R., Birch, J.M., Ramsden, R.T., Sharif, S., & Baser, M.E. (2006). Malignant transformation and new primary tumours after therapeutic radiation for benign disease: substantial risks in certain tumour prone syndromes. J Med Genet, 43(4), 289–​94. Evans, D.G.R. (2009). Neurofibromatosis type 2 (NF2): a clinical and molecular review. Orphanet J Rare Dis, 4, 16. Evans, D.G., King, A.T., Bowers, N.L., et al. (2018). Identifying the deficiencies of current diagnostic criteria for neurofibromatosis 2 using databases of 2777 individuals with molecular testing. Genet Med. doi: 10.1038/​s41436-​018-​0384-​y [Epub ahead of  print] Fallah, A., Guyatt, G.H., Carter Snead, O., et al. (2015). Resective epilepsy surgery for tuberous sclerosis in children:  determining predictors of seizure outcomes in a multicenter retrospective cohort study. Neurosurg, 77(4), 517–​24. Ferner, R.E., Huson, S.M., Thomas, N., et al., (2007). Guidelines for the diagnosis and management of individuals with neurofibromatosis 1. J Medical Genet, 44(2),  81–​8. Fisher, M.J., Loguidice, M., Gutmann, D.H., et al. (2012). Visual outcomes in children with neurofibromatosis type 1-​associated optic pathway glioma following chemotherapy:  a multicenter retrospective analysis. Neuro Oncol, 14(6),  790–​7. Franz, D.N., Leonard, J., Tudor, C., et al. (2006). Rapamycin causes regression of astrocytomas in tuberous sclerosis complex. Ann Neurol, 59(3),  490–​8. Franz, D.N., Belousova, E., Sparagana, S., et  al. (2013). Efficacy and safety of everolimus for subependymal giant cell astrocytomas associated with tuberous sclerosis complex (EXIST-​1): a multicentre, randomised, placebo-​controlled phase 3 trial. Lancet, 381(9861), 125–​32. Fryer, A.E., Chalmers, A., Connor, J.M., et al. (1987). Evidence that the gene for tuberous sclerosis is on chromosome 9. Lancet, 1, 659–​62. Golfinos, J.G., Hill, T.C., Rokosh, R., et al. (2016). A matched cohort comparison of clinical outcomes following microsurgical resection or stereotactic radiosurgery for patients with small-​and medium-​ sized vestibular schwannomas. J Neurosurg, 125(6), 1472–​82. Gottfried, O.N., Viskochil, D.H., & Couldwell, W.T. (2010). Neurofibromatosis type 1 and tumorigenesis:  molecular mechanisms and therapeutic implications. Neurosurg Focus, 28(1), E8. Jagannathan, J., Lonser, R.R., Smith, R., DeVroom, H.L., & Oldfield, E.H. (2008). Surgical management of cerebellar hemangioblastomas in patients with von Hippel-​ Lindau disease. J Neurosurg, 108, 210–​22. Jansen, F.E., van Huffelen, A.C., Algra, A., & van Nieuwenhuizen, O. (2007). Epilepsy surgery in tuberous sclerosis: a systematic review. Epilepsia, 48(8), 1477–​84. Jett, K. & Friedman, J.M. (2010). Clinical and genetic aspects of neurofibromatosis 1. Genet Med, 12(1),  1–​11. Jonasch, E., McCutcheon, I.E., Waguespack, S.G., et al. (2011). Pilot trial of sunitinib therapy in patients with von Hippel-​Lindau disease. Ann Oncol, 22(12), 2661–​6.

CHAPTER 38 Neurophakomatoses

Jozwiak, J., Jozwiak, S., & Oldak, M. (2006). Molecular activity of sirolimus and its possible application in tuberous sclerosis treatment. Med Res Rev, 26(2), 160–​80. Kandt, R.S., Haines, J.L., Smith, M., et al. (1992). Linkage of an important gene locus for tuberous sclerosis to a chromosome 16 marker for polycystic kidney disease. Nat Genet, 2(1),  37–​41. Kano, H., Shuto, T., Iwai, Y., et al. (2015). Stereotactic radiosurgery for intracranial hemangioblastomas: a retrospective international outcome study. J Neurosurg, 122(June),  1–​10. Kleinerman, R.A. (2009). Radiation-​sensitive genetically susceptible pediatric sub-​populations. Pediatr Radiol, 39(Suppl 1), S27–​S31. Krueger, D., Care, M.M., Holland, K., et  al. (2010). Everolimus for subependymal giant-​cell astrocytomas in tuberous sclerosis. N Engl J Med, 363(19), 1801–​11. Krueger, D.A. & Northrup, H. (2013). Tuberous sclerosis complex surveillance and management: recommendations of the 2012 international tuberous sclerosis complex consensus conference. Pediatr Neurol, 49(4), 255–​65. Latif, F., Tory, K., Gnarra, J., et  al. (1993). Identification of the von Hippel-​Lindau disease tumor suppressor gene. Science (New York, N.Y.), 260(5112), 1317–​20. Lee, A.G. (2007). Neuroophthalmological management of optic pathway gliomas. Neurosurg Focus, 23(5), E1. Listernick, R., Ferner, R.E., Liu, G.T., & Gutmann, D.H. (2007). Optic pathway gliomas in neurofibromatosis-​1: controversies and recommendations. Ann Neurol, 61(3), 189–​98. Lonser, R.R., Weil, R.J., Wanebo, J.E., DeVroom, H.L., & Oldfield, E.H. (2003). Surgical management of spinal cord hemangioblastomas in patients with von Hippel-​Lindau disease. J Neurosurg, 98(1), 106–​16. Maher, E.R., Yates, J.R., Harries, R., et al. (1990). Clinical features and natural history of von Hippel-​Lindau disease. Q J Med, 77(283), 1151–​63. Maher, E.R. (2004). Von Hippel-​Lindau disease. Curr Mol Med, 4(8), 833–​42. Maher, E.R., Neumann, H.P., & Richard, S. (2011). von Hippel-​Lindau disease:  a clinical and scientific review. Eur J Hum Genet, 19(6), 617–​23. Mahoney, D.H., Cohen, M.E., Friedman, H.S., et  al. (2000). Carboplatin is effective therapy for young children with progressive optic pathway tumors:  a pediatric oncology group phase II study. Neuro Oncol, 2(4), 213–​20. Mallory, G.W., Pollock, B.E., Foote, R.L., Carlson, M.L., Driscoll, C.L., & Link, M.J. (2014). Stereotactic radiosurgery for neurofibromatosis 2—​associated vestibular schwannomas. Neurosurg, 74(3), 292–​301. Mathieu, D., Kondziolka, D., Flickinger, J.C., et al. (2007). Stereotactic radiosurgery for vestibular schwannomas in patients with neurofibromatosis type 2: an analysis of tumor control, complications, and hearing preservation rates. Neurosurg, 60(3),  460–​8. Melmon, K.L. & Rosen, S.W. (1964). Lindau’s disease:  review of the literature and study of a large kindred. Am J Med, 36(4), 595–​617. Mohyuddin A, Wallace, A., Wu, C., et  al. (2002). Molecular genetic analysis of the NF2 gene in young patients with unilateral vestibular schwannomas. J Med Genet, 39(5), 315–​22. National Institutes of Health Consensus (1988). Neurofibromatosis. Conference statement. National Institutes of Health Consensus Development Conference. Arch Neurol, 45(5),  575–​8. NCT01158651, N.C.T.C. (2010). Study of RAD001 (Everolimus) for Children with NF1 and Chemotherapy-​ refractory Radiographic Progressive Low-​ grade Gliomas. https://​www.clinicaltrials.gov/​ show/​NCT01158651

NCT01767792, N.C.T.C. (2013). Phase 2 Study of Bevacizumab in Children and Young Adults With NF 2 and Progressive Vestibular Schwannomas. https://​www.clinicaltrials.gov/​show/​NCT01767792 Northrup, H. & Krueger, D.A. (2013). Tuberous sclerosis complex diagnostic criteria update: recommendations of the 2012 International Tuberous Sclerosis Complex Consensus Conference. Pediatr Neurol, 49(4), 243–​54. Nunes, F.P., Merker, V.L., Jennings, D., et  al. (2013). Bevacizumab treatment for meningiomas in NF2:  a retrospective analysis of 15 patients. PLoS One, 8(3), e59941. O’Callaghan, F.J., Shiell, A.W., Osborne, J.P., & Martyn, C.N. (1998). Prevalence of tuberous sclerosis estimated by capture-​recapture analysis. Lancet, 351(9114), 1490. Parry, D.M., Eldridge, R., Kaiser-​Kupfer, M.I., Bouzas, E.A., Pikus, A., & Patronas, N. (1994). Neurofibromatosis 2 (NF2): clinical characteristics of 63 affected individuals and clinical evidence for heterogeneity. Am J Med Genet, 52(4), 450–​61. Phi, J.H., Kim, D.G., Chung, H.T., Lee, J., Paek, S.H., & Jung, H.W. (2009). Radiosurgical treatment of vestibular schwannomas in patients with neurofibromatosis type 2:  tumor control and hearing preservation. Cancer, 115(2),  390–​8. Plotkin, S.R., Merker, V.L., Halpin, C., et al. (2012). Bevacizumab for progressive vestibular schwannoma in neurofibromatosis type 2: a retrospective review of 31 patients. Otol Neurotol, 33(6), 1046–​52. Plowman, P.N. & Evans, D.G.R. (2000). Stereotactic radiosurgery XI. Acoustic neuroma therapy and radiation oncogenesis. BrJ Neurosurg, 14(2),  93–​5. Richard, S., Marsot-​Dupuch, K., Giraud, S., Béroud, C., & Resche, F. (2000). Central nervous system hemangioblastomas, endolymphatic sac tumors, and von Hippel-​Lindau disease. Neurosurg Rev, 23(1), 1–​22; discussion  23–​4. Rosser, T. & Packer, R.J. (2002). Intracranial neoplasms in children with neurofibromatosis 1. J Child Neurol, 17(8), 630–​7; discussion 646–​51. Rouleau, G., Merel, P., Lutchman, M., et  al. (1993). Alteration in a new gene encoding a putative membrane-​organizing protein causes neuro-​fibromatosis type 2. Nature, 363(6429), 515–​21. Rowe, J., Grainger, A.C., Walton, L., & Radatz, M. (2007). Safety of radiosurgery applied to conditions with abnormal tumor suppressor genes. Neurosurgery, 60(5),  860–​3. Rowe, J.G., Radatz, M.W., Walton, L., Soanes, T., Rodgers, J., & Kemeny, A.A. (2003). Clinical experience with gamma knife stereotactic radiosurgery in the management of vestibular schwannomas secondary to type 2 neurofibromatosis. J Neurol Neurosurg Psychiatry, 74(9), 1288–​93. Sampson, J.R., Scahill, S.J., Stephenson, J.B., Mann, L., & Connor, J.M. (1989). Genetic aspects of tuberous sclerosis in the west of Scotland. J Med Genet, 26(1),  28–​31. Schwartz, R.A., Fernández, G., Kotulska, K., & Jóźwiak, S. (2007). Tuberous sclerosis complex:  advances in diagnosis, genetics, and management. J Am Acad Dermatol, 57(2), 189–​202. Sharif, S., Ferner, R., Birch, J.M., et al. (2006). Second primary tumors in neurofibromatosis 1 patients treated for optic glioma: substantial risks after radiotherapy. J Clin Oncol, 24(16), 2570–​5. Sharif, S., Moran, A., Huson, S.M., et al. (2007). Women with neurofibromatosis 1 are at a moderately increased risk of developing breast cancer and should be considered for early screening. J Med Genet, 44(8),  481–​4. Sharma, M.S., Singh, R., Kale, S.S., Agrawal, D., Sharma, B.S., & Mahapatra, A.K. (2010). Tumor control and hearing preservation

459

460

Section 9  Tumours and skull base—tumour syndromes

after gamma knife radiosurgery for vestibular schwannomas in neurofibromatosis type 2. J Neurooncol, 98(2), 265–​70. Subach, B.R., Kondziolka, D., Lunsford, L.D., Bissonette, D.J., Flickinger, J.C., & Maitz, A.H. (1999). Stereotactic radiosurgery in the management of acoustic neuromas associated with neurofibromatosis type 2. J Neurosurg, 90(5), 815–​22. Tee, A.R., Fingar, D.C., Manning, B.D., Kwiatkowski, D.J., Cantley, L.C., & Blenis, J. (2002). Tuberous sclerosis complex-​1 and -​2 gene products function together to inhibit mammalian target of rapamycin (mTOR)-​mediated downstream signaling. Proc Natl Acad Sci U S A, 99(21), 13571–​6. Trofatter, J.A., MacCollin, M.M., Rutter, J.L., et  al. (1993). A novel moesin-​, ezrin-​, radixin-​like gene is a candidate for the neurofibromatosis 2 tumor suppressor. Cell, 72(5), 791–​800.

Viskochil, D., Buchberg, A.M., Xu, G., et al. (1990). Deletions and a translocation interrupt a cloned gene at the neurofibromatosis type 1 locus. Cell, 62(1), 187–​92. Walker, L., Thompson, D., Easton, D., et al. (2006). A prospective study of neurofibromatosis type 1 cancer incidence in the UK. Br J Cancer, 95(2),  233–​8. Wanebo, J.E., Lonser, R.R., Glenn, G.M., & Oldfield, E.H. (2003). The natural history of hemangioblastomas of the central nervous system in patients with von Hippel-​Lindau disease. J Neurosurg, 98(1),  82–​94. Weil, R.J., Lonser, R.R., DeVroom, H.L., Wanebo, J.E., & Oldfield, E.H. (2003). Surgical management of brainstem hemangioblastomas in patients with von Hippel-​Lindau disease. J Neurosurg, 98(1), 95–​105.

39

Uncommon brain lesions Yizhou Wan, Hani J. Marcus, and Thomas Santarius

Introduction In this chapter we discuss neurosurgical tumours, which are uncommon and have not been covered elsewhere. Due to their rarity this chapter reviews what is known about these conditions and primarily discusses their diagnosis. The description of management is derived largely from published case series.

Choristoma Choristoma is a mass of heterotopic tissue (i.e. ‘normal’ tissue) present in an abnormal location (Lee and Roland, 2013). The presence of histologically normal tissue differentiates these lesions from hamartomas, one example of which is dysplastic gangliocytoma of the cerebellum (L’hermitte-​ Duclos disease). Epidermoid and dermoid tumours are the most well-​ known intracranial choristomas and are the subject of a separate chapter. Several rarer choristomas have been found in the nervous system, such as intracranial pancreatic (Heller et al., 2010), salivary gland (Hintz et  al., 2014), and neuromuscular tissue masses (Hebert-​Blouin et al., 2012). Salivary gland choristomas located in the sella have been well-​described (Hintz et  al., 2014). Other common locations include major peripheral nerves, such as the brachial plexus and sciatic nerve for neuromuscular choristomas (Hebert-​Blouin et al., 2013). Neural choristomas are exceedingly rare and the most commonly found in the head and neck region, especially in the nose (nasal gliomas). Other areas include the trunk and back, where they usually present as isolated asymptomatic masses (Downing et al., 1997). Patients can present from early childhood onwards and the mean age of presentation for neural choristomas is approximately 10 years of age (Hebert-​Blouin et al., 2012).

Pathology Histology of choristomas depends on their tissue of origin. Neural choristomas may comprise of a disorganized population of glial cells in a fibrous background, alternatively they may represent mature neural tissue resembling cerebral or cerebellar cortex (Downing et  al., 1997). Neuromuscular choristomas (NMC) are composed

of an admixture of mature striated and smooth muscle cells with nerve fibres, typically the sciatic and brachial nerves, although involvement of cranial nerves such as the optic, maxillary, oculomotor, facial and vestibulocochlear nerves have also been described (Hebert-​Blouin et al., 2012).

Management Clinical presentation depends on anatomical location. For example, salivary gland choristomas in the sella can cause hormone deficits whereas suprasellar extension can cause headaches or visual disturbances (Hebert-​Blouin et  al., 2012). NMCs may cause pain, weakness, and sensory changes when they produce a mass effect along the course of a peripheral nerve or result in sensory or motor deficits. Intracranially they are commonly found in the acoustic meatus or cerebellopontine angle, mimicking vestibular schwannomas on imaging (Boyaci et  al., 2011). Radiological findings of NMC are non-​specific and may be indistinguishable from schwannomas (Boyaci et al., 2011). Preoperative imaging is important to identify any potential connections between neural choristomas and the central nervous system (Downing et al., 1997). Malignant transformation of neural choristomas has never been reported (Newman et al., 1986).

Prognosis The rarity of these tumours means that their natural history is not well elucidated. It can generally be thought of as a ‘slow-​growing’ lesion and treatment can be conservative (Boyaci et al., 2011). Some authors have advocated complete resection (Awasthi et al., 1991).

Hamartoma Hamartoma is a tumour made of tissue components normally found in that anatomical site, but the tissue architecture is disorganized. Cysts and pseudocysts will be described in other chapters, this section describes other lesions found in the central nervous system including, dysplastic gangliocytoma of the cerebellum (L’hermitte-​ Duclos disease), cortical glioneuronal hamartomas, and hypothalamic hamartomas. Common to all these lesions are that they may represent developmental malformations, but may be mistaken for more common tumour or cystic lesions.

462

Section 9  Tumours and skull base—tumour syndromes

Epidemiology Since the first description in 1920 more than 230 cases of L’hermitte-​ Duclos disease (LDD) have been described in the literature. It can present from birth to the sixth decade with no gender predilection. It has been suggested that there is a genetic association between LDD and Cowden syndrome. Cowden syndrome is an autosomal dominant hamartoma-​neoplasia complex presenting with neoplasia and hamartomas of the breast, thyroid, genitourinary tract, and endometrium (Nowak and Trost, 2002). Among the CNS manifestations of tuberous sclerosis complex (TSC), an important diagnostic criterion is the presence of cortical glioneuronal hamartomas (tubers). TSC is an autosomal dominant multisystem disorder caused by mutations in genes encoding tuberculin and harmaline, proteins involved in cell proliferation and differentiation. They cause benign non-​invasive lesions in any organ system (Curatolo and Maria, 2013). It occurs at a birth rate of 1 in 6000 (Kwiatkowski and Manning, 2005). Hypothalamic hamartomas have a prevalence of 1 in 50 000 to 1 in 100 000 (Weissenberger et al., 2001).

Fig. 39.1  L’hermitte-​Duclos disease. Abnormal tissue involves the cerebellar cortex, appearing hyperintense on T2 with widening of cerebellar folia.

Pathology LDD is characterized by regional thickening of the cerebellar folia, caused by dysplastic replacement of cerebellar Purkinje and granular cells with hamartomatous overgrowth of hypertrophic ganglion cells (Wei et al., 2014). Microscopically there are abnormal ganglion cells in the granular cell layer with thickening and hypermyelination of the molecular cell layer, and absence of the Purkinje cell layer (Ozeren et al., 2014). There may also be the presence of calcification and ectatic vessels on histopathology but no pleomorphism or mitoses (Wei et al., 2014). In TSC cortical tubers are characterized by proliferation of glial and neuronal cells and the loss of the six-​layer cortical structure. Within tubers a variety of abnormal cells are found, predominantly, bizarrely shaped astrocytes, giant cells, and dysplastic neurons. A  further characteristic finding is that of subependymal nodules, hamartomas growths which can protrude into the wall of the lateral ventricle near the foramen of Monroe. These can slowly grow into subependymal giant-​cell astrocytomas (SEGAs), the most common cerebral tumours found in TSC (Curatolo and Maria, 2013). Hypothalamic hamartomas are developmental disorders affecting the hypothalamic area between the infundibulum and the mammillary bodies (Striano et al., 2012).

Management Patients with LDD are initially asymptomatic, but as the tumour slowly increases in size patients usually present with vague neurological complaints related to unilateral cerebellar dysfunction for as long as 15 years (Ozeren et al., 2014). Occasionally patients can present with acute decompensation from occlusive hydrocephalus (Nowak and Trost, 2002). MRI is fairly characteristic and can be used to diagnose LDD, especially in the context of Cowden’s disease. Enlarged cerebellar folia lose their secondary folding and lead to asymmetrical widening of the cerebellar hemispheres. The lesions are non-​enhancing and display a characteristic pattern of isointense bands within a discrete lesion of hypointensity on T2-​weighted imaging (Fig. 39.1). This indicates widened gyri and displaced sulci structure in expanded cerebellar cortex (Nowak and Trost, 2002).

It may prove to be difficult for symptomatic LDD to be resected because the lesions blend into normal cerebellar parenchyma. Intraoperative MRI or/​and intraoperative histological demonstration of a transition zone on histopathology maybe helpful. The clinical manifestations of TSC are protean and related to the systemic effects of the condition. From a neurological perspective, the management focuses on controlling epilepsy. This generally occurs in the first year of life, evolving from focal seizures to infantile spasms. Subependymal giant-​cell tumours are treated with mTOR pathway inhibitors or surgically (Franz et al., 2006). Classically the hallmarks of hypothalamic hamartomas are gelastic seizures, developmental delay, and precocious puberty. MRI usually demonstrates a lesion with a vertical plane of attachment within the third ventricle and/​or a horizontal plane of attachment on the underside of the hypothalamus (Fig. 39.2). These lesions appear to have an intrinsic epileptogenicity (Maixner, 2006). Surgical resection has been reported to improve symptoms and seizure control (Palmini et al., 2002). Complete resection can be challenging and because it is difficult to predict the severity of a patient’s epileptic course, the optimal timing of surgery is unclear (Maixner, 2006).

Prognosis Due to the progressive growth of LDD lesions surgical resection is usually required but recurrence is common due to the difficulties in achieving a complete resection (Ozeren et al., 2014). In TSC renal complications such as angiomyolipomas, renal cysts and renal cell carcinoma are the most frequent cause of the death (Curatolo and Maria, 2013). Harvey et  al. reported in their case series of 29 hypothalamic hamartoma patients that 15 patients achieved seizure-​freedom and seven patients achieved 90% reduction in seizures over a follow-​up of one to six years (mean 2.5 years). Complications of surgery included, injury to the optic tract, meningitis, communicating hydrocephalus and ischaemic infarction to the thalamus (Harvey et al., 2003).

CHAPTER 39  Uncommon brain lesions

Fig. 39.2  Delalande classification of hypothalamic hamartomas: type I has a horizontal orientation; type II has a vertical orientation; type III has both a horizontal and vertical component; and type IV is a giant hamartoma. Adapted from Endoscopic Treatment of Hypothalamic Hamartomas. J Korean Neurosurg Soc. 2017; 60(3): 294–300. doi: 10.3340/jkns.2017.0101.005, which is published under a Creative Commons License.

Pituicytoma, granular cell tumour (GCT), and spindle cell oncocytoma (SCO)

fascicles of spindled-​to-​epithelioid cells with mitochondrial accumulation seen on ultrastructural analysis (Zygourakis et al., 2015).

Management

The 2007 WHO Classification of Tumours of the Central Nervous System describes a spectrum of benign (WHO grade I) non-​ The most common presenting symptom of patients with non-​ adenomatous pituitary tumours including pituicytoma, GCT, and adenomatous pituitary tumours is visual disturbance (e.g. bitemporal hemianopsia) (Covington et al., 2011). Other symptoms SCO (Louis et al., 2007; Covington et al., 2011). include headache fatigue, and those related to hypopituitarism (e.g. Epidemiology decreased libido). Diabetes insipidus is rare. Several features on MRI can aid diagnosis (Fig. 39.3) (Covington Non-​adenomatous pituitary tumours are rare, with a recent meta-​ analysis identifying 112 cases in the literature (Covington et al., 2011). et al., 2011): pituicytoma may be considered if the lesion is purely The average age at diagnosis is between 50 and 60 years, with a slight intrasellar and clearly separate from the pituitary gland; GCT male preponderance for pituicytoma, and female preponderance for may be considered if the lesion is suprasellar or mixed supra-​and GCT and SCO (Covington et al., 2011; Pirayesh Islamian et al., 2012). intrasellar, and clearly separate from the pituitary gland; and SCO may be considered if the lesion is mixed supra-​and intrasellar, but Pathology the mass cannot be separated for the pituitary gland. Compared to Pituicytoma arises from pituicytes, specialized glial cells of the pos- GCT and SCO, Pituicytomas more frequently demonstrate homoterior lobe and the stalk of the pituitary gland (Zygourakis et  al., genous rather than heterogenous enhancement on T1-weighted im2015). Histologically, pituicytoma is characterized by elongated, bi- ages. (Covington et al., 2011). In practice, these lesions may be difficult to distinguish from pipolar spindle cells (Brat et al., 2007). tuitary adenoma, and transsphenoidal surgical resection is generGCT arises from granular pituicytes, and are composed of polally considered safe and effective (Zygourakis et al., 2015). However, ygonal eosinophilic granular cells that are strongly PAS positive because these lesions are firm and may be highly vascular, some due to abundant intracytoplasmic lysosomes (Orning et al., 2013). SCO also arises from pituicytes and are characterized by interlacing authors have advocated the use of preoperative embolization in (a)

(b)

(c)

Fig. 39.3  Granular cell tumour (GCT) of the pituitary. (A) MRI (T1 with contrast) showing a GCT of the pituitary stalk. In this case the tumour is separate from the pituitary gland and the diagnosis of should be considered. (B) Transsphenoidal endoscopic approach—​intraoperative photograph demonstrating the separation of the tumour from the pituitary gland as seen on the MRI. (C) Granular cell tumour of the pituitary—​in this case not distinguishable from a much more common pituitary adenoma or the much rarer primary non-​adenomatous pituitary gland tumours (i.e. pituicytoma and spindle cell oncocytoma, SCO).

463

464

Section 9  Tumours and skull base—tumour syndromes

selected cases (Wolfe et al., 2008). The role of adjuvant therapy remains unclear (Zygourakis et al., 2015).

Prognosis Non-​adenomatous pituitary tumours are generally benign and surgical resection may be curative.

the most commonly used term to describe them. Glomus jugulare consists of one or more bodies, found in the aventitia of the dome of the jugular bulb. Glomus jugulare tumours can be found anywhere along the course of the internal jugular vein.

Epidemiology

Gangliocytomas are usually considered alongside gangliogliomas as a spectrum of intra-​axial tumours originating in the grey matter. Gangliocytomas differ from gangliogliomas by their noted absence of a glial component and consist of abnormal mature ganglion cells (WHO grade I).

The incidence of chemodectoma is less than 1 in 3000 (Gad et al., 2014). They are ten times more common in people living at high-​ altitudes (Athanasiou et al., 2007). This may be related to hypoxia-​ induced hyperplasia of the catecholamine-​ secreting chief cells (Knight et  al., 2006). The mean age of onset is 45  years (Kotelis et al., 2009). Glomus jugulare tumours are rare, comprising 0.6% of tumours in the head and neck. They are two to three times more common in women than they are in men (Lee et al., 2002).

Epidemiology

Pathology

They account for 0.1–​0.5% of all brain tumours and occur in children and young adults (Koeller and Henry, 2001).

Chemodectomas are genetically heterogeneous. They can occur sporadically or as part of a neoplastic syndrome, such as multiple endocrine neoplasia II, neurofibromatosis I, and von Hippel-​Lindau disease (Knight et  al., 2006). Four different genetic susceptibility loci have been mapped in chemodectomas. They demonstrate autosomal dominant inheritance with age and hypoxia-​related penetrance. Mutations in three of these genes encode for subunits of the enzyme succinate dehydrogenase complex (Knight et al., 2006). Histology of chemodectoma demonstrates mostly chief cells, arranged in a characteristic pseudoalveolar pattern known to as ‘cell balls’ (Zellballen) (Athanasiou et al., 2007). Histologically glomus jugulare tumours are composed of epithelioid or chief cells in a hypervascular stroma, surrounded by a thin capsular layer (Guild, 1953).

Gangliocytoma

Pathology They are found in decreasing incidence in the temporal lobe, cerebellum, parieto-​occipital region, frontal lobe, hypothalamus adjacent to the third ventricle and spinal cord (Koeller and Henry, 2001). Intraoperative smear histology shows ganglion cells with vesicular nuclei and prominent nucleoli. One can differentiate between gangliocytoma and infiltrating glioma with entrapped neurons by the presence of variable-​sized neoplastic ganglion cells which form clusters and have abnormal branching of apical dendrites.

Management They present as a result of mass effect or seizures. There is possibly a long history of symptoms and it is important to consider their diagnosis because of their favourable prognosis postresection. Non-​ enhanced CT demonstrates hyperattenuation and no vasogenic oedema. MRI appearances show low signal intensity on T1-​weighted images and mixed signal intensity on T2-​weighted images (Shin et al., 2002). These features make differentiation between gangliogliomas and gangliocytomas challenging.

Management

Patients with chemodectomas typically present with a lateral neck mass, with or without involvement of the lower cranial nerves (8 to 12) (Knight et al., 2006). Glomus jugulare tumours are slow-​growing and there may be significant latency prior to diagnosis. Signs and symptoms include unilateral deafness, tinnitus, hoarseness, dysphagia, headache, and Prognosis lower cranial nerve palsies (Watkins et al., 1994). In 10–​20% of cases Malignant transformation is uncommon, it is more likely to be seen they can be multifocal and although generally benign, local invasion with gangliogliomas and is due to the glial/​astrocytic component and metastasis to liver, lung, lymph nodes, and spleen have been reshowing increased cellularity, mitosis, and necrosis (Adesina and ported. Preoperative workup should include MRI and CT to identify Rauch, 2010). extent of bone involvement. In contrast with phaeochromocytomas and extra-​adrenal sympathetic paragangliomas, head and neck paragangliomas rarely proParaganglioma duce significant amounts of catecholamine (Chen et al., 2010). This makes preoperative optimization of these patients less challenging Tumours derived from chromaffin cells (themselves derived from from the neuroendocrine perspective. neural crest cells) are known as paragangliomas regardless of their Workup with imaging is crucial for surgical success as location anatomical location. Intra-​ adrenal paragangliomas are called renders chemodectomas technically challenging and biopsy is phaeochromocytomas. The most common location for extra-​ hazardous. Tumour enlargement typically splays the internal and adrenal parasympathetic paragangliomas is in the head and neck. external carotid arteries but does not lead to narrowing of their The most common location here is the carotid body at the bifurca- diameters (Wieneke and Smith, 2009). The differential diagnoses tion of the internal jugular and external jugular veins, where they include vagal schwannomas and neurofibromas, glomus vagale are known as chemodectomas. Intracranially, paragangliomas occur tumours, and lymph node masses. Lymph nodes are not typically in the region of the jugular foramen and glomus jugulare tumour is hypervascular, glomus vagale tumours occur more rostrally, and

CHAPTER 39  Uncommon brain lesions

(a)

(b)

Fig. 39.4  Chemodectoma. A 59-​year-​old lady presented with right-​sided otalgia. (A) T1W Axial MRI with gadolinium demonstrates splaying of the internal and external carotid arteries which differentiate paragangliomas from other lesions such as schwannomas and neurofibromas. (B )The lesion was resected with no complications after preoperative embolization demonstrated, extensive vascular supply from hypertrophic branches of the right ascending pharyngeal artery. ECA, external carotid artery; ICA, internal carotid artery; IJV, internal jugular vein.

vagal masses tend to displace both the carotid vessels together rather than to splay them (Fig. 39.4). Larger tumours as much as 10 cm in diameter are more likely to encase the carotid artery. In these circumstances, neoadjuvant radiotherapy can be used to down-​stage the tumours and improve resectability. Intraoperatively, internal carotid artery reconstruction with end-​to-​end anastomosis may be performed (Gad et al., 2014). Historically, approaches to management of glomus jugulare tumours have been microsurgical. The median tumour doubling time has been reported to be four years or longer, but aggressive tumours exist and older patients may cope less well with its symptoms (Kemeny, 2009). Resection approaches depend on size and location and may include combined otological and neurosurgical approaches (Watkins et al., 1994). Stereotactic radiosurgery’s obvious advantages in terms of minimizing radiation damage to critical structures at the skull base have to be balanced against its limitations. Relative contraindications of radiosurgery include large tumours, typically a maximal of 3–​ 4 cm field size and lack of radiological boundaries at the skull base. There have been no randomized control trials to directly compare radiosurgery against microsurgery but a meta-​analysis of 19 retrospective studies demonstrated that radiosurgery is efficacious with 95% of a combined sample of 335 patients achieving clinical control (Guss et al., 2011).

Prognosis The mortality and morbidity rate associated with resection of chemodectoma has been quoted at 1% and 33%, respectively (Gad et  al., 2014). Cranial nerve injury, wound haematomas, and ischaemic stroke can occur. Although there is a low risk of distant metastasis, follow-​up for possible lymph node metastases is indicated. Residue disease may also be controlled with radiotherapy.

Glomus jugulare tumours are typically indolent and, as the management goal is to provide long-​term tumour control with low morbidity, a judicious use of stereotactic radiotherapy and modern microsurgery make achieving this goal possible (Fayad et al., 2009).

Langerhans cell histiocytosis Langerhans cell histiocytosis (LCH), formerly known as histiocytosis X, is a heterogeneous group of tumours, typically affecting bone, characterized by a clonal proliferation of pathological cells with the characteristics of Langerhans cells (Abla et  al., 2010). Langerhans cells are dendritic cells (antigen-​presenting cells) that arise from the bone marrow and usually reside in the lymph nodes and skin. The term ‘eosinophilic granuloma’ has been used to describe a single LCH lesion, ‘Hand-​Schüller-​Christian disease’ to describe multiple LCH lesions within one system, and ‘Abt-​Letterer-​Siwe disease’ to describe multiple LCH lesions within multiple systems (Arico et al., 2003).

Epidemiology LCH typically occurs in children, with an estimated incidence of 4–​ 4.5 cases per million per year (Chaudhary et al., 2013).

Pathology The aetiology of LCH remains unclear. There are data to suggest that germ line mutations may predispose some patients to LCH. Approximately 1% of patients with LCH have an affected relative. Moreover, twin studies have found a considerably higher rate of concordance in monozygotic versus dizygotic twins (92% and 10%, respectively) (Abla et al., 2010). There are also data to suggest that viruses such as HHV6 may result in LCH, however most studies have failed to confirm this viral association (Abla et al., 2010).

465

466

Section 9  Tumours and skull base—tumour syndromes

Histologically the diagnosis of LCH rests on the presence of Langerhans cells within lesions, with accompanying macrophages, T-​lymphocytes, eosinophils, and multinucleated giant cells. These Langerhans cells may be determined by the demonstration of CD1a on immunohistochemistry, or Birbeck granules on electron microscopy (Abla et  al., 2010). The latter may now be more easily identified by immunohistochemical staining of langerin (CD207), an endocytic receptor which induces the formation of Birbeck granules (Valladeau et al., 2000). Lesions associated with LCH are most commonly located within bone, particularly the skull. Within the central nervous system three patterns of lesions have been described (Grois et al., 2010): 1. Well-​circumscribed granulomas within the brain’s connective tissue with a predilection for the circumventricular organs. . Non-​granulomatous neurodegenerative lesions mainly affecting 2 the brainstem and cerebellum. . Granulomas in the pituitary stalk that invade the hypothalamus 3 and show diffuse infiltration of the surrounding brain.

Management The clinical symptoms and signs depend on the site and nature of LCH lesions. A single skull lesion is the most frequent presentation of LCH. Involvement of the hypothalamic-​pituitary axis can result in diabetes insipidus in up to 25% of all patients (Grois et al., 2010). Non-​granulomatous neurodegenerative lesions result in a highly variable presentation, ranging from mild abnormalities of the reflexes and subtle cognitive deficits, to pronounced ataxia or behavioural disturbance. CT and MRI are indicated to investigate skull and brain lesions, respectively (Grois et al., 2010). CT of the skull may demonstrate lesions within the calvarium, with a characteristic bevelled edge. MRI of the brain may show enlargement of the pituitary stalk with potential extension to the hypothalamus (Fig. 39.5). MRI may also reveal radiological degeneration with symmetrical hyperintense signal changes on T2-weighted images and hypo-​or hyperintense signals on T1-weighted images in the cerebellum, and sometimes changes in the basal ganglia and brainstem. The evidence guiding the decision-​making for LCH is limited. In unifocal skull lesions a simple curettage or even biopsy can provide diagnostic tissue and usually results in cure (Abla et al., 2010). The treatment of central nervous system lesions resulting in diabetes insipidus may include steroids or chemotherapy (Grois et al., 2010). Surgical resection and radiotherapy have also been used in lesions located in regions other than the hypothalamic-​ pituitary axis.

Prognosis The prognosis of LCH is highly variable. Patients with lesions limited to a single system generally have a high chance of spontaneous remission and a favourable outcome. Patients with multiple systems affected may have a very much worse prognosis; those in whom high-​risk organs (liver, spleen, or bone marrow) are involved, who fail to respond to therapy in the first 6 weeks, have a mortality rate as high as 35%.

Fig. 39.5  Langerhans cell histiocytosis (LCH). A 14-​year old boy presented with a lump on the forehead and chronic constipation. Subsequently diabetes insipidus and growth hormone deficiency were diagnosed and the MRI (T1 with gadolinium contrast) revealed a thickened and enhancing pituitary stalk and an enhancing lytic skull lesion. A biopsy of the skull lesion yielded a diagnosis of LCH. Further investigations uncovered involvement of the gut and chemotherapy according to LCH4 protocol was instituted.

Sarcoidosis Sarcoidosis is a multisystem inflammatory granulomatous disease, typically affecting the lungs (Joseph and Scolding, 2007). The involvement of the central nervous system, termed neurosarcoidosis, is a rare but feared complication of the disease.

Epidemiology The prevalence of sarcoidosis is approximately 40 per 100  000 (Joseph and Scolding, 2007). Less than 10% of patients with sarcoidosis exhibit symptoms and signs of neurological disease, though autopsy studies have suggested a significant proportion with subclinical disease. It is usually diagnosed between 20 and 40 years of age, and certain racial groups, such as Northern Europeans and West Africans, are at considerably higher risk of developing the disease.

Pathology The aetiology of sarcoidosis remains unclear. Genetic studies have suggested that class I human leukocyte antigen (HLA)-​B8 antigens and HLA class II antigens encoded by HLA-​DRB1 and DQB1 alleles may predispose some patients to the disease (Vargas and Stern, 2010). Several environmental triggers have also been put forward including mould, metals, wood burning stoves, photocopier toner, and infectious agents such as mycobacterium tuberculosis and propionibacterium acnes. Histologically sarcoidosis is characterized by non-​caseating granuloma. Neurosarcoidosis most commonly affects the leptomeninges

CHAPTER 39  Uncommon brain lesions

at the base of the brain. Extension to the brain parenchyma may occur, and is usually associated with surrounding astrocyte activation and gliosis.

Management Neurosarcoidosis presents most frequently with cranial nerve palsies, seen in 50–​75% of patients. Such palsies are most commonly the result of basal meningitis, but may also be due to nerve granulomas, or raised intracranial pressure. Seizures are another common manifestation of neurosarcoidosis, often reflecting underlying mass lesions, and occur in about 10% of patients (Joseph and Scolding, 2007). An enhanced MRI of the may show leptomeningeal enhancement with gadolinium. Multiple white matter lesions are seen in approximately 40% of patients (Joseph and Scolding, 2007). Large mass lesions may also be demonstrated, and can mimic neoplasia. Lumbar puncture may reveal raised protein (>2 g/​litre), mild lymphocytosis, and possibly oligoclonal bands. Systemic investigations can also provide evidence of sarcoidosis including blood tests (elevated serum ACE, ESR, and calcium), and a chest X-​ray or CT thorax (hilar and mediastinal adenopathy). The first-​line therapy for sarcoidosis are corticosteroids (Joseph and Scolding, 2007; Vargas and Stern, 2010). Long-​term corticosteroid use is associated with considerable morbidity, particularly in patients with diabetes mellitus, hypertension, or osteoporosis. To this end, steroid-​sparing immunosuppressive agents, including methotrexate and hydroxychloroquine, have also been used with varying success to manage sarcoidosis, though these too also carry risks of complications. In patients with drug-​resistant sarcoidosis, case series have reported success with radiotherapy (Bruns et  al., 2004). The role of surgery is limited to biopsy in equivocal cases, and the treatment of complications such as hydrocephalus.

tumefactive demyelinating lesions are indistinguishable from other MS lesions. Active lesions consist of areas of demyelination with reactive astrocytes and myelin-​ containing foamy macrophages (Hardy and Chataway, 2013). Macroscopically, lesions are usually well circumscribed, with a predilection to the frontal and parietal lobes (Altintas et al., 2012).

Management The presentation of TMS often resembles that of more sinister space occupying lesions, including symptoms and signs of raised intracranial pressure, seizures, and focal neurological deficits. Headache and vomiting are particularly common in children (Altintas et al., 2012). An enhanced MRI brain is indicated to investigate TMS. Lesions up to 12 cm in diameter have been reported, but most range between 2 and 6 cm (Lucchinetti et al., 2008). There is usually associated cerebral oedema, though often less than that observed with malignancy. The majority of lesions enhance with gadolinium, and demonstrate a ring appearance (usually closed, but open ring is highly specific for TMS) (Fig. 39.6), which is thought to reflect an advancing area of active inflammation away from a more chronic non-​enhancing core (Hardy and Chataway, 2013). An unenhanced CT head may be helpful in distinguishing TMS from malignancy; regions that are hyperintense on an enhanced MRI brain, but hypodense on an unenhanced CT head, are more likely to represent TMS (Kim et al., 2009). Lumbar puncture may reveal unmatched (present in CSF but not in serum) oligoclonal bands. Systemic investigations are also important to exclude other pathologies such as vasculitis, granuloma, infection, or malignancy. These include blood tests and CT of chest, abdomen, and pelvis (Hardy and Chataway, 2013). The treatment is with high-​dose corticosteroids (Altintas et  al., 2012). In patients in whom corticosteroids are ineffective, plasma

Prognosis Spontaneous remission of neurosarcoidosis has been reported but is very rare; most patients require long-​term follow with therapy provided when disease exacerbation occurs.

Tumefactive multiple sclerosis Multiple sclerosis (MS) is an autoimmune inflammatory disease affecting the central nervous system, characterized by lesions of demyelination and axonal loss. Tumefactive multiple sclerosis (TMS) describes the presence of MS lesions greater than 2 cm in diameter, often solitary, which may be mistaken for other space occupying lesions such as brain tumours on imaging (Hardy and Chataway, 2013).

Epidemiology The prevalence of TMS has been estimated to be 1–​2 per 1000 cases of MS (Hardy and Chataway, 2013). It is usually diagnosed between the ages of 20 and 40 years.

Pathology The aetiology of TMS remains unknown. Tumefactive demyelinating lesions have been reported in diseases other than MS, including viral infections such as HIV, and as a side effect of treatment with drugs such as tacrolimus. Histologically,

Fig. 39.6  Tumefactive MS plaque. An ‘open ring sign’ is relatively specific for multiple sclerosis. The enhancing component is thought to represent advancing front of demyelination, thus favouring the white matter side of the lesion. The open part of the ring will therefore usually point towards the grey matter. Stereotactic biopsy confirmed a diagnosis of MS.

467

468

Section 9  Tumours and skull base—tumour syndromes

fossa and this is where these tumours occur most commonly (Goldgeier et al., 1984).

Epidemiology The incidence of melanocytoma is approximately 1 per 10 million, and the incidence of primary CNS melanoma 1 per 20 million. The peak incidence is in the fifth decade, and there is a slight female preponderance (Liubinas et al., 2010).

Pathology Histologically, the diagnosis of primary melanocytic neoplasms of the CNS rests on the identification of melanin in tumour cells or associated macrophages (Liubinas et al., 2010). Melanocytomas are low-​grade lesions that do not invade the surrounding brain, although intermediate-​ grade melanocytomas are also recognized (Brat et  al., 1999; Navas et  al., 2009). Primary CNS melanoma is similar to melanoma arising in other sites, and may feature invasion, haemorrhage, or necrosis. Macroscopically, melanocytomas and primary CNS melanomas are solitary lesions that are generally extra-​axial. Fig. 39.7  Melanocytoma. A middle-​aged woman presented with headache. Dermatological and ophthalmological were unremarkable. An unenhanced CT brain showed a mixed density lesion with abnormal calcification, suggesting it was longstanding. A staging CT did not demonstrate lesions elsewhere, and she underwent a complete surgical resection. Histopathology confirmed melanocytoma and she remains well 5 years postoperatively.

exchange appears to be a reasonable second line agent. The anti-​ CD20 B cell monoclonal antibody rituximab has also been used with some success (Fan et al., 2012). Disease-​modifying therapy such as interferon beta or glatiramer acetate may be also be beneficial in reducing the risk of recurrence, but this must be balanced against the risk of adverse events. The role of surgery is limited to biopsy in cases of diagnostic uncertainty, and the treatment of complications such as hydrocephalus.

Prognosis Overall, the prognosis of patients with TMS does not appear to be worse than other patients with MS in general. Indeed, in the largest cohort of patients with TMS, the long-​term prognosis was better than in matched controls with typical MS (Lucchinetti et al., 2008).

Primary CNS melanoma and melanocytoma Primary melanocytic neoplasms of the CNS are a range of tumours characterized by a clonal proliferation of melanocytes of the leptomeninges, which are derived from the neural crest during early embryogenesis (Liubinas et  al., 2010). These neoplasms include diffuse leptomeningeal melanocytosis (DML) or melanomatosis, melanocytoma, and primary malignant melanoma; the present discussion is limited to melanocytoma and primary malignant melanoma. As the highest concentration of meningeal melanocytes is in the spinal cord and posterior

Management The presentation of melanocytomas and primary CNS melanoma is that of other space occupying lesions, including symptoms and signs of raised intracranial pressure, seizures, and focal neurological deficits. Dermatological and ophthalmological examinations as well as CT staging must be performed to rule out a distant primary source. Neuroimaging is indicated in suspected lesions. A CT brain often shows iso-​or hyperdense lesions with or without abnormal calcification (Fig. 39.7), which enhance with contrast. An MRI brain may demonstrate lesions that are, unusually for tumours, hyperintense on T1-​weighted images, hypointense on T2-​weighted images, and enhance homogenously with gadolinium. The treatment of melanocytoma and primary CNS melanoma is complete surgical resection. Radiotherapy may be indicated in incompletely resected melanocytoma, but malignant melanoma is relatively radioresistant. The role of chemotherapy is unclear, but it has been used as adjuvant therapy and in patients with recurrent disease (Cornejo et al., 2013).

Prognosis Melanocytomas are generally benign and surgical resection curative. Primary CNS melanoma, however, is associated with a poor prognosis. Primary CNS melanoma may be prone to local recurrence or systemic spread, albeit less so than metastatic melanoma, with an average survival time approximately 6  years (Cornejo et al., 2013).

Intracranial lipomas Epidemiology These tumours are extremely rare with an incidence estimated at 0.06–​0.46% of all intracranial tumours (Truwit and Barkovich, 1990). Older studies may however underestimate the true incidence due to

CHAPTER 39  Uncommon brain lesions

(a)

(b)

(c)

Fig. 39.8  Intracranial lipoma. (A) CT demonstrates a homogenously hypodense lesion in the lateral and third ventricle, approximately –​100 Hounsfield units. (B) T1-​weighted MRI showing a hyperintense mass lesion. (C) T2-​weighted MRI also showing a hyperintense mass lesion.

increasingly widespread use of imaging. There does not appear to be any age or sex-​differential in their distribution (Eghwrudjakpor et al., 1992). The youngest case reported is in a 3-​day old child and the oldest in a 91-​year-​old woman (Cascino et al., 1958; Yalcin and Fragoyannis, 1966).

Pathology Intracranial lipomas were first described by Rokitansky in 1856 following accidental discovery after an autopsy (Rokitansky, 1856). Merrill Sosman who was appointed roentgenologist to the Brigham Hospital in 1922 by Harvey Cushing made the first diagnosis in a living patient using pneumoencephalography (Sosman, 1946). These lesions are thought to be congenital malformations caused by maldifferentiation of the meninx primitiva during the development of the subarachnoid cisterns (Verga, 1929). They may interfere with cortical development near their location (Yildiz et al., 2006). Most commonly they are associated with corpus collosum dysgenesis or agenesis (Macpherson et  al., 1987). Other reported abnormalities include, saccular or fusiform aneurysms, arteriovenous malformations and sagittal sinus and falcine sinus fenestrations (Futami et al., 1992; Sasaki et al., 1996). Macroscopically they do not exert a mass effect so nerves and blood vessels can run through them (Yildiz et al., 2006).

Prognosis Lipomas are slowly progressing lesions so prognosis is good even with conservative management. Surgery may be considered for patients with medication-​refractory seizures but tendency to adhere to surrounding neural tissue make radical resection hazardous (Yilmaz et al., 2006).

Controversies A timely and accurate diagnosis is a guiding principle in the management of patients with brain tumours (Omuro et  al., 2006). Making the diagnosis of uncommon brain lesions is inherently challenging, but this chapter has highlighted several elements that are important when investigating such lesions (Box 39.1). Considering the rarity of these lesions, their treatment is also invariably contentious. Unsurprisingly, there are no randomized controlled trials to

Box 39.1  Elements to consider when managing uncommon brain lesions History

Onset in young adults History of subtle and transient neurological symptoms Personal or family history of autoimmune or inflammatory disease

Examination

Skin rashes Slit lamp findings

Management In 45% of cases, lipomas are found in the interhemispheric region, in the pericallosal area. They are also frequently located in the dorsal mesencephalic area. Rarely, they have reported in the Sylvian fissure where they are associated with epilepsy (Yildiz et al., 2006). See Fig. 39.8. Most commonly they are incidental lesions that appear hypodense on CT, hyperintense on T1-weighted images, and hyperintense on T2-weighted images. It is important to differentiate them from dermoids which frequently appear in the midline (Yildiz et al., 2006).

Investigations Avoid empirical steroids unless significant mass effect MRI brain and spinal cord Lumbar puncture if not contra-​indicated Anti-​HIV and syphilis serology Whole-​body CT scan Biopsy

469

470

Section 9  Tumours and skull base—tumour syndromes

guide surgical decision-​making, and recommendations are therefore mostly based on case series.

FURTHER READING Abla, O., Egeler, R.M., & Weitzman, S. (2010). Langerhans cell histiocytosis:  current concepts and treatments. Cancer Treat Rev, 36,  354–​9. Adesina, A. & Rauch, R. (2010). Ganglioglioma and gangliocytoma. In:  Adesina, A.M., Tihan, T., Fuller, C.E., & Poussaint, T.Y. (eds.) Atlas of Pediatric Brain Tumors, pp. 185–​94. New  York, NY: Springer. Gad, A., Sayed, A., Elwan, H., et al. (2014). Carotid body tumors: a review of 25 years experience in diagnosis and management of 56 tumors. Ann Vasc Dis, 7,  292–​9. Hardy, T.A. & Chataway, J. (2013). Tumefactive demyelination: an approach to diagnosis and management. J Neurol Neurosurg Psychiatry, 84, 1047–​53. Louis, D.N., Perry, A., Reifenberger, G., et al. (2016). Acta Neuropathol, 131, 803.

REFERENCES Abla, O., Egeler, R.M., & Weitzman, S. (2010). Langerhans cell histiocytosis:  current concepts and treatments. Cancer Treat Rev, 36,  354–​9. Adesina, A. & Rauch, R. (2010). Ganglioglioma and gangliocytoma. In:  Adesina, A.M., Tihan, T., Fuller, C.E., & Poussaint, T.Y. (eds.) Atlas of Pediatric Brain Tumors, pp. 185–​94. New York, NY: Springer. Altintas, A., Petek, B., Isik, N., et al. (2012). Clinical and radiological characteristics of tumefactive demyelinating lesions:  follow-​ up study. Mult Scler, 18, 1448–​53. Arico, M., Girschikofsky, M., Genereau, T., et al. (2003). Langerhans cell histiocytosis in adults. Report from the International Registry of the Histiocyte Society. Eur J Cancer, 39, 2341–​8. Athanasiou, A., Liappis, C.D., Rapidis, A.D., Fassolis, A., Stavrianos, S.D., & Kokkalis, G. (2007). Carotid body tumor: review of the literature and report of a case with a rare sensorineural symptomatology. J Oral Maxillofac Surg, 65, 1388–​93. Awasthi, D., Kline, D.G., & Beckman, E.N. (1991). Neuromuscular hamartoma (benign triton tumor) of the brachial-​plexus—​case report. J Neurosurg, 75,  795–​7. Boyaci, S., Moray, M., Aksoy, K., & Sav, A. (2011). Intraocular neuromuscular choristoma:  a case report and literature review. Neurosurgery, 68, E551–​5. Brat, D.J., Giannini, C., Scheithauer, B.W., & Burger, P.C. (1999). Primary melanocytic neoplasms of the central nervous systems. Am J Surg Pathol, 23, 745–​54. Brat, D.J., Scheithauer, B.W., Fuller, G.N., & Tihan, T. (2007). Newly codified glial neoplasms of the 2007 WHO Classification of Tumours of the Central Nervous System:  angiocentric glioma, pilomyxoid astrocytoma and pituicytoma. Brain Pathology, 17, 319–​24. Bruns, F., Pruemer, B., Haverkamp, U., & Fischedick, A.R. (2004). Neurosarcoidosis:  an unusual indication for radiotherapy. Br J Radiol, 77,  777–​9. Cascino, J. P., Lake, F., Jackson, C., & Kaplan, A. (1958). Lipoma of the corpus callosum. J Int Coll Surg, 29,  171–​4. Chaudhary, V., Bano, S., Aggarwal, R., et al. (2013). Neuroimaging of Langerhans cell histiocytosis: a radiological review. Jpn J Radiol, 31, 786–​96.

Chen, H., Sippel, R.S., O’Dorisio, M.S., Vinik, A.I., Lloyd, R.V., & Pacak, K. (2010). The North American Neuroendocrine Tumor Society Consensus guideline for the diagnosis and management of neuroendocrine tumors pheochromocytoma, paraganglioma, and medullary thyroid cancer. Pancreas, 39, 775–​83. Cornejo, K.M., Hutchinson, L., Cosar, E.F., et al. (2013). Is it a primary or metastatic melanocytic neoplasm of the central nervous system? A molecular based approach. Pathol Int, 63, 559–​64. Covington, M.F., Chin, S.S., & Osborn, A.G. (2011). Pituicytoma, spindle cell oncocytoma, and granular cell tumor: clarification and meta-​analysis of the world literature since 1893. Am J Neuroradiol, 32, 2067–​72. Curatolo, P. & Maria, B.L. (2013). Tuberous sclerosis. Handb Clin Neurol, 111, 323–​31. Downing, M.T., Hamoudi, A.B., & Besner, G.E. (1997). Brain heterotopia: choristoma of the back. Pediatr Surg Int, 12,  183–​5. Eghwrudjakpor, P.P.O., Kurisaka M., Fukuoka, M., & Mori, K. (1992). Intracranial lipomas:  current perspectives in their diagnosis and treatment. Br J Neurosurg, 6, 139–​44. Fan, X., Mahta, A., DE Jager, P. L., & Kesari, S. (2012). Rituximab for tumefactive inflammatory demyelination: a case report. Clin Neurol Neurosurg, 114, 1326–​8. Fayad, J.N., Schwartz, M.S., & Brackmann, D.E. (2009). Treatment of recurrent and residual glomus jugulare tumors. Skull Base, 19(1),  92–​8. Franz, D.N., Leonard, J., Tudor, C., et al. (2006). Rapamycin causes regression of astrocytomas in tuberous sclerosis complex. Ann Neurol, 59(3),  490–​8. Futami, K., Kimura, A., & Yamashita, J. (1992). Intracranial lipoma associated with cerebral saccular aneurysm: case report. J Neurosurg, 77,  640–​2. Gad, A., Sayed, A., Elwan, H., et al. (2014). Carotid body tumors: a review of 25 years’ experience in diagnosis and management of 56 tumors. Ann Vasc Dis, 7,  292–​9. Goldgeier, M.H., Klein, L.E., Klein-​Angerer, S., Moellmann, G., & Nordlund, J.J. (1984). The distribution of melanocytes in the leptomeninges of the human brain. J Invest Dermatol, 82,  235–​8. Grois, N., Fahrner, B., Arceci, R.J., et al. (2010). Central nervous system disease in Langerhans cell histiocytosis. J Pediatr, 156, 873–​81, 881 e1. Guild, S.R. (1953). The glomus jugulare, a nonchromaffin paraganglion, in man. Ann Otol Rhinol Laryngol, 62, 1045–​71; concld. Guss, Z.D., Batra, S., Limb, C.J., et al. (2011). Radiosurgery of glomus jugulare tumors: a meta-​analysis. Int J Radiat Oncol Biol Phys, 81, e497–​502. Hardy, T.A. & Chataway, J. (2013). Tumefactive demyelination: an approach to diagnosis and management. J Neurol Neurosurg Psychiatry, 84, 1047–​53. Harvey, A.S., Freeman, J.L., Berkovic, S.F., & Rosenfeld, J.V. (2003). Transcallosal resection of hypothalamic hamartomas in patients with intractable epilepsy. Epileptic Disord, 5, 257–​65. Hebert-​ Blouin, M.N., Amrami, K.K., & Spinner, R.J. (2013). Addendum: evidence supports a ‘no-​touch’ approach to neuromuscular choristoma. J Neurosurg, 119,  252–​4. Hebert-​Blouin, M.N., Scheithauer, B.W., Amrami, K.K., Durham, S.R., & Spinner, R.J. (2012). Fibromatosis: a potential sequela of neuromuscular choristoma. J Neurosurg, 116, 399–​408. Heller, R.S., Tsugu, H., Nabeshima, K., & Madsen, O.D. (2010). Intracranial ectopic pancreatic tissue. Islets, 2,  65–​71. Hintz, E.B., Yeaney, G.A., Buchberger, G.K., & Vates, G.E. (2014). Intracranial salivary gland choristoma within optic nerve dural

CHAPTER 39  Uncommon brain lesions

sheath: case report and review of the literature. World Neurosurgery, 81(5–​6),  842. Joseph, F.G. & Scolding, N.J. (2007). Sarcoidosis of the nervous system. Pract Neurol, 7, 234–​44. Kemeny, A.A. (2009). Contemporary management of jugular paragangliomas (glomus tumours): microsurgery and radiosurgery. Acta Neurochir (Wien), 151(5), 419–​21. Kim, D.S., Na, D.G., Kim, K.H., et  al. (2009). Distinguishing tumefactive demyelinating lesions from glioma or central nervous system lymphoma: added value of unenhanced CT compared with conventional contrast-​enhanced MR imaging. Radiology, 251, 467–​75. Knight, T.T., Jr., Gonzalez, J.A., Rary, J.M., & Rush, D.S. (2006). Current concepts for the surgical management of carotid body tumor. Am J Surg, 191, 104–​10. Koeller, K.K. & Henry, J.M. (2001). From the archives of the AFIP—​superficial gliomas:  radiologic-​pathologic correlation. Radiographics, 21, 1533–​56. Kotelis, D., Rizos, T., Geisbusch, P., et al. (2009). Late outcome after surgical management of carotid body tumors from a 20-​year single-​ center experience. Langenbecks Arch Surg, 394, 339–​44. Kwiatkowski, D.J. & Manning, B.D. (2005). Tuberous sclerosis: a GAP at the crossroads of multiple signaling pathways. Hum Mol Genet, 14 Spec No. 2, R251–​8. Lee, J.H., Barich, F., Karnell, L.H., et al. (2002). National Cancer Data Base report on malignant paragangliomas of the head and neck. Cancer, 94,  730–​7. Lee, K. & Roland, P. (2013). Heterotopias, teratoma, and choristoma. In:  Kountakis, S. (ed.) Encyclopedia of Otolaryngology, Head and Neck Surgery. Berlin Heidelberg, Germany: Springer. Liubinas, S.V., Maartens, N., & Drummond, K.J. (2010). Primary melanocytic neoplasms of the central nervous system. J Clin Neurosci, 17, 1227–​32. Louis, D.N., Ohgaki, H., Wiestler, O.D., et al. (2007). The 2007 WHO classification of tumours of the central nervous system. Acta Neuropathol, 114, 97–​109. Lucchinetti, C.F., Gavrilova, R.H., Metz, I., et al. (2008). Clinical and radiographic spectrum of pathologically confirmed tumefactive multiple sclerosis. Brain, 131, 1759–​75. Macpherson, R.I., Holgate, R.C., Gudeman, S.K. (1987). Midline central nervous system lipomas in children. J Can Assoc Radiol, 38, 264–​70. Maixner, W. (2006). Hypothalamic hamartomas—​ clinical, neuropathological and surgical aspects. Childs Nerv Syst, 22, 867–​73. Navas, M., Pascual, J. M., Fraga, J., et  al. (2009). Intracranial intermediate-​grade meningeal melanocytoma with increased cellular proliferative index: an illustrative case associated with a nevus of Ota. J Neurooncol, 95, 105–​15. Newman, N.J., Miller, N.R., & Green, W.R. (1986). Ectopic brain in the orbit. Ophthalmology, 93, 268–​72. Nowak, D.A. & Trost, H.A. (2002). Lhermitte-​Duclos disease (dysplastic cerebellar gangliocytoma):  a malformation, hamartoma or neoplasm? Acta Neurol Scand, 105, 137–​45. Omuro, A.M., Leite, C.C., Mokhtari, K., & Delattre, J.Y. (2006). Pitfalls in the diagnosis of brain tumours. Lancet Neurol, 5, 937–​48. Orning, J.L., Trembath, D.G., Zanation, A.M., & Germanwala, A.V. (2013). Endoscopic endonasal approach for resection of infundibular granular cell tumor: case report and literature review. J Case Rep Med, 2, 235775.

Ozeren, E., Gurses, L., Sorar, M., Er, U., Onder, E., & Arikok, A.T. (2014). L’hermitte-​Duclos disease in an elderly patient: a case report and review of the literature. Asian J Neurosurg, 9, 246. Palmini, A., Chandler, C., Andermann, F., et al. (2002). Resection of the lesion in patients with hypothalamic hamartomas and catastrophic epilepsy. Neurology, 58, 1338–​47. Pirayesh Islamian, A., Buslei, R., Saeger, W., & Fahlbusch, R. (2012). Pituicytoma:  overview of treatment strategies and outcome. Pituitary, 15, 227–​36. Rokitansky, C. (1856). Lehrbuch der Pathologischen Anatomie, Vol 2. Vienna, Austria: Wilhelm Braumuller, pp. 468–​8. Sasaki, H., Yoshida, K., Wakamoto, H., Otani, M., & Toya, S. (1996). Lipomas of the frontal lobe. Clin Neurol Neurosurg, 98,  27–​31. Shin, J. H., Lee, H.K., Khang, S.K., et al. (2002). Neuronal tumors of the central nervous system: radiologic findings and pathologic correlation. Radiographics, 22, 1177–​89. Sosman, M.C. (1946). Discussion of Echternacht AP, Campbell JA. Midline anomalies of the brain. Their diagnosis by pneumoencephalography. Radiology, 46, 119–​31. Striano, S., Santulli, L., Ianniciello, M., Ferretti, M., Romanelli, P., & Striano, P. (2012). The gelastic seizures-​hypothalamic hamartoma syndrome: facts, hypotheses, and perspectives. Epilepsy Behav, 24,  7–​13. Truwit, C.L. & Barkovich, A.J. (1990). Pathogenesis of intracranial lipoma: an MR study in 42 patients. AJNR Am J Neuroradiol, 11(4), 665–​74. Valladeau, J., Ravel, O., Dezutter-​ Dambuyant, C., et  al. (2000). Langerin, a novel C-​type lectin specific to Langerhans cells, is an endocytic receptor that induces the formation of Birbeck granules. Immunity, 12,  71–​81. Vargas, D.L. & Stern, B.J. (2010). Neurosarcoidosis:  diagnosis and management. Semin Respir Crit Care Med, 31, 419–​27. Verga, P. (1929). Lipoma ed osteolipomi della pia madre. Tumori, 15, 321. Watkins, L.D., Mendoza, N., Cheesman, A.D., & Symon, L. (1994). Glomus jugulare tumours:  a review of 61 cases. Acta Neurochir (Wien), 130,  66–​70. Wei, G., Zhang, W., Li, Q., et al. (2014). Magnetic resonance characteristics of adult-​onset Lhermitte-​Duclos disease:  an indicator for active cancer surveillance? Mol Clin Oncol, 2, 415–​20. Weissenberger, A.A., Dell, M.L., Liow, K., et al. (2001). Aggression and psychiatric comorbidity in children with hypothalamic hamartomas and their unaffected siblings. J Am Acad Child Adolesc Psychiatry, 40, 696–​703. Wieneke, J.A. & Smith, A. (2009). Paraganglioma: carotid body tumor. Head Neck Pathol, 3,  303–​6. Wolfe, S.Q., Bruce, J., & Morcos, J.J. (2008). Pituicytoma: case report. Neurosurgery, 63, E173–​4; discussion E174. Yalcin, S. & Fragoyannis, S. (1966). Intracranial lipoma. Case report. J Neurosurg, 24(5),  895–​7. Yildiz, H., Hakyemez, B., Koroglu, M., Yesildag, A., & Baykal, B. (2006). Intracranial lipomas:  importance of localization. Neuroradiology, 48(1),  1–​7. Yilmaz, N., Unal, O., Kiymaz, N., Yilmaz, C., & Etlik, O. (2006). Intracranial lipomas—​ a clinical study. Clin Neurol Neurosurg, 108(4),  363–​8. Zygourakis, C.C., Rolston, J.D., Lee, H.S., Partow, C., Kunwar, S., & Aghi, M.K. (2015). Pituicytomas and spindle cell oncocytomas: modern case series from the University of California, San Francisco. Pituitary, 18,  150–​8.

471

SECTION 10

Neurotrauma and intensive care

40. Epidemiology of head injury and outcome after head injury  475 Nabeel Alshafai and Andrew Maas

43. Surgical management of head injury  509 Hadie Adams, Angelos G. Kolias, Adel Helmy, Peter J.A. Hutchinson, and Randall M. Chesnut

41. Pathophysiology of traumatic brain injury  483 John K. Yue, Hansen Deng, Ethan A. Winkler, John F. Burke, Catherine G. Suen, and Geoffrey T. Manley

44. Complications of head injury  521 Fardad T. Afshari, Antonio Belli, and Peter C. Whitfield

42. Intensive care management of head injury  497 Matthew A. Kirkman and Martin Smith

45. Concussion and sports-related head injury  531 Mark Wilson

40

Epidemiology of head injury and outcome after head injury Nabeel Alshafai and Andrew Maas

Introduction Traumatic brain injury (TBI) constitutes a major cause of death and disability. In low-​and middle-​income countries the incidence of TBI is increasing, subsequent to increased motorization. It is, however, particularly vulnerable road users (pedestrians, cyclists) where the risk is greatest. In higher-​income countries, children, young adults, and elderly patients have the highest rate of TBI and a substantial increase in incidence of TBI in elderly patients as a result of falls has been noted. Nevertheless, TBI is a disease that affects the population of all ages and is referred to as ‘a silent epidemic’. Knowledge of the epidemiology of TBI is essential to inform healthcare planning and to target prevention campaigns appropriately. In this chapter we will summarize global TBI epidemiological perspectives and reflect on the burden that TBI imposes on health economics and society. We will review current classification systems, outcome measures, and prognostic models for TBI.

Definition Many different definitions have been, and continue to be, used to describe TBI. This may cause confusion and ambiguity concerning an accurate diagnosis, leading to difficulties in comparing patient cohorts. In practice, it may be difficult to distinguish between symptoms of a mild TBI and the effects of alcohol intoxication or drug abuse. In elderly patients, a fall with some bruising to the scalp may be secondary to a primary cardiac event and it may be difficult to differentiate between a temporary loss of consciousness or dizziness due to the cardiac event and a mild TBI. Furthermore, some patients may present late and the diagnosis then needs to be established on basis of the case history, introducing a subjective element. A great need exists for objective parameters to confirm or exclude a diagnosis of TBI. Various new tools, including biomarkers, studies of eye movements, balance, are showing potential, but still need to be validated in large cohorts. WHO defines TBI as an acute injury to the brain resulting from mechanical energy to the head from external physical (Carroll et al., 2004) force excluding injuries relating to illicit drug, alcohol, or substance, medication, or caused by other treatment or injuries (Andelic et al., 2012).

TBI has also been defined as: ‘an alteration in brain function, or other evidence of brain pathology, caused by an external force’ (Menon et al., 2010). TBI has replaced the old term ‘head injury’ as it captures the importance of the brain injury (Roozenbeek et al., 2013), in addition to injuries purely affecting the scalp, skull, and face.

Aetiology Motorized vehicles have been described as the most persistent killers. Preventive measures have been shown to be highly effective in reducing the number of TBI related to road traffic incidents, and combined with improved trauma care saves lives (Table 40.1). Legislation, seat belts, air bags, and helmet use have all contributed to increased safety. Prevention campaigns should target vulnerable road users who are unaware of traffic risks (e.g. the frail, or children) as well as vehicle drivers. While road traffic incidents remain the most common cause of TBI in low and middle-​income countries, in Europe, falls are now a more frequent cause of TBI (Peeters et al., 2015). Table 40.1 provides a summary overview of the most common causes of TBI. In the United States fire arm injuries exceeded road traffic incidents for the first time in 1990. In combat situations, blast injuries due to improvised explosive devices are a frequent cause of TBI. Table 40.2 lists methods of reducing the impact of TBI.

Incidence When we analyse the current epidemiological incidence studies, there appear to be a number of limitations, namely, deficient standardized epidemiologic reporting, the lack of age standardized incidence rates, the lack of data specific to mild and moderate TBI and the lack of data from lower-​and middle-​income countries. These challenges need to be addressed in future studies. An important point to bear in mind is poly-​trauma and repeated head injuries in the same patient (first time vs. recurrent head injury) which is often overlooked and is crucial to understand the late consequences of such recurrent injuries. In a systematic review of 28 European studies published over 24 years, an aggregate incidence rate of hospital admissions was 262 cases per 100  000 people annually (Peeters et  al., 2015). A recent population-​based study in New Zealand shows a much

476

Section 10  Neurotrauma and intensive care

Table 40.1  Common causes of TBI Main causes of TBI

Road traffic incidents (44–​80%) Falls$ (12–​55%) Violence@ (7–​17%) Fire arm (>50% suicide) (Aarabi et al., 2001) Work and sports (4%) Others $

More recently, the leading cause in northern Europe, United States, and Australia specially in young children and elderly (>75-​year-​old) @ Higher risk population: male, young adults, non-​white (56% African-​American), unemployed at the time of the injury, unmarried, history of illegal substance use, history of law enforcement encounters, poverty, and low educational level.

higher incidence rate of 790/​100 000 (Feigin et al., 2013). The discrepancy between the high incidence rate reported in this study, with complete case ascertainment, and the lower rates based upon hospital admission data indicate that the latter approach grossly underestimates the total burden of TBI. Differences in reported incidence rates between countries are unlikely to reflect real differences, but rather result from differences in case ascertainment, the use of variable definitions and in some cases a focus on subpopulations.

Table 40.3  Mortality rate after TBI across the world Region

Mortality incidence/​100 000

References

USA 17.1 (2010) West Virginia 23.6 (1989–​1999)

CDC, 2015, Rutland-Brown et al., 2006 Adekoya and Majumder, 2004

EU -​ Austria -​ Denmark -​  France -​ Germany -​  Italy -​  Norway -​  Sweden

10.5–11.2 12.6 7.3 11.5 8.3 7.6 10.4 9.2–9.5

Peeters et al., 2015, Majdan et al., 2016 Majdan et al., 2016 Majdan et al., 2016 Masson et al., 2001 Majdan et al., 2016 Majdan et al., 2016 Sundstrøm et al., 2007 Sundstrøm et al., 2007; Majdan et al., 2016

Brazil China

5.1–26.2 12.99

De Almeida et al., 2016; Koizumi et al., 2000 Cheng et al., 2017, Jiang et al., 2019

Prevalence is the total volume of living people with TBI (existing and new cases) at a given point in time (point prevalence) or in a given period (period prevalence). In a meta-​analysis in 2013 of 15 prevalence studies (Frost et al., 2013) (25 134 adults), Frost et al. found that 12% had experienced a TBI with loss of consciousness (male twice as much risk as female). McKinlay et al. looked at the paediatric age group (higher risk) and showed that over 30% of people had experienced at least one TBI before they are 25 years of age (McKinlay et al., 2008) These numbers appear high compared to estimates reported by the Center for Disease Control (CDC) in the United States, that 5.3 million people are currently living with TBI in the United States today.

et al., 2015; Majdan et al., 2017). Table 40.3 presents an overview of mortality rates reported in the United States and Europe. Despite a rise in the incidence of TBI worldwide, some studies claim a decrease in mortality rate from TBI secondary to implementation and adherence to evidence-​based guidelines and recommendations in the developed countries (Lu et  al., 2005; Gerber et al., 2013). However, improved survival was not found when comparing mortality rates reported in observational studies over the past 30 years (Rosenfeld et al., 2012). Another important point is that ‘acute mortality’ is not the only concern in TBI: TBI patients have been found to have a substantially elevated long-​term risk for premature mortality. This includes: suicide, subsequent injuries, and assaults, even after adjustment for socio demographic and familial factors six months after injury (Fazel et al., 2014). There is a reduced average life expectancy of nine years following inpatient rehabilitation for TBI (McMillan et al., 2011). This varies according to age and severity. It has been found that after experiencing TBI, the annual mortality remains increased, up to sevenfold, for at least 13 years (McMillan et al., 2011).

Mortality

Classification of TBI

For Europe, an average mortality rate of 10.5–11.7/100 000 population has been derived, with a large range identified varying from 3.0 to 21.8/100 000, mainly determined by case mix (Peeters

Separate child restraint law

Various approaches to the classification of TBI can be taken, these include classification by mechanism, clinical severity, burden of injury, the extent of structural damage, initial prognostic risk, and expected prognosis. Table 40.4 presents these different domains, which may be used for classification. We suggest that appropriate classification of TBI should not be restricted to a single domain but should incorporate a multidimensional perspective. Currently, no such multidimensional integrated system for classification exists, but this goal is being pursued in ongoing large-​scale observational studies (Maas et al., 2017).

Firmly enforced speed limits

Mechanism of injury

Compulsory helmets for motorcycle riders

Different mechanisms cause different types of injuries. In closed TBI, acceleration and deceleration forces may cause a typical picture of diffuse axonal injury in patients injured during road traffic incidents. A  fall, in which the head hits a hard surface would typically result in focal contusional brain injury. Penetrating injuries caused by high velocity missiles, typically cause extensive damage along the missile track, while

Prevalence

Table 40.2  Reducing the IMPACT of TBI Reducing the IMPACT of TBI

Strict limits in blood alcohol concentration The use of seatbelts

Clinical care improved with faster communication Better trauma systems Increased availability of CT scanning Improved intensive care unit (ICU) facilities and care Others

CHAPTER 40  Epidemiology of head injury and outcome after head injury

Table 40.4  Different approaches to classification of TBI Mechanism of injury

Clinical severity

• Closed • Penetrating • Crush • Blast • Combination

• GCS 3–​8: Severe • GCS 9–​13: Moderate • GCS 14–​15: Mild

Injury burden (AIS and ISS)*

Structural damage: CT Classification

1: Minor 2: Moderate 3: Serious 4: Severe 5: Critical 6: Virtually unsurvivable

• Diffuse injury I • Diffuse injury II • Diffuse injury III • Diffuse injury IV • Evacuated mass lesion • Non-​evacuated mass lesion

Prognostic risk CRASH and IMPACT prognostic models * AIS per body region: external (skin), head/​neck (including brain injury), thorax, abdomen/​ pelvic contents, spine, and extremities ISS: Injury Severity Score (Baker et al., 1974)

the amount of kinetic energy transferred results in substantial distant damage. In crush injuries, dynamic forces are to a large extent absorbed by the skull and not all energy is transferred to the brain. As a consequence, skull injuries may be extensive while the brain injury may be less severe. In blast injuries, several mechanisms combine in damaging the brain. These include primary injury (due to pressure waves), secondary injury (collision with debris and other objects), tertiary (being thrown down by the blast and hitting the head against an object) or quaternary (heat and toxic injury).

Clinical severity There is a wide spectrum of clinical presentation in TBI ranging from a hit to the head with symptoms of disorientation or

alteration of consciousness that quickly resolve with or without amnesia, to complete loss of consciousness and coma. The Glasgow Coma Scale (Teasdale and Jennett, 1974) (GCS) has been adopted worldwide as instrument for assessing the level of consciousness. In GCS assessments the responses observed in three components (eye, motor, verbal scale) are recorded. For purposes of classification and research, a total sum score may be derived, ranging from 3 to 15; however, on an individual basis the importance of separating out the individual domains of the GCS cannot be overemphasized. The GCS scale is made up of entirely distinct components (E, V, M) and each ‘score’ is qualitative and discontinuous (i.e. a change from M2 to M3 is in no way related to a change from M4 to M5, let alone E3 to E4). It should be noted that especially early after injury the GCS may change over time, following resuscitation or as part of early recovery. Accurate assessments of one or more of the components of the GCS may be confounded by prior alcohol or substance use, prehospital use of sedation, paralysis, and intubation (Balestreri et al., 2004; Stocchetti et al., 2004). Indeed, the literature review revealed over 25 different TBI severity classification systems. Notwithstanding the limitations just mentioned, based on the GCS sum score, TBI is classified as mild, moderate, or severe (Fig. 40.1). In addition to the GCS, multiple clinical approaches to the classification of TBI severity have been proposed. Many of these address specific subpopulations, but none have been shown to be convincingly superior to the GCS. We should recognize, however, that the GCS does not provide a granular classification for mild TBI, as it is not intended to capture more subtle changes such as ‘an alteration in brain function’. Subtle and transient symptoms are reported, which may overlap with symptoms due to either post-​traumatic stress disorders (PTSD) and/​or postconcussion syndromes. This is an area of ambiguity at the present time and further research is required to distinguish the different diagnoses and the implications on patient outcome (Maas et al., 2017).

Decision level 1 Abnormalities no

yes Decision level 2 Mass

CT class I

yes

no

Decision level requiring clinical information Evacuation

Decision level 2a Cistems Normal CT class II

yes

Absent/Compressed Decision level 2b Shift

CT class V

no

yes

CT Class III

CT Class IV

no CT class VI

Fig. 40.1  Marshall CT classification presented in a tree structure. Prediction of outcome in traumatic brain injury with CT characteristics: A comparison between the CT classification and combinations of CT predictors. Maas AIR, Hukkelhoven CWPM, Marshall LF, Steyerberg EW. Neurosurgery 2005, 57, 6, 1173–​82.

477

478

Section 10  Neurotrauma and intensive care

Extracranial injuries (injury burden) TBI is associated with extracranial injuries in about 35% of cases. The severity of injuries in six body regions is universally scored according to The Abbreviation Injury Score (AIS). In the AIS, severity is scored for six body regions: external (skin), head/​neck (including brain injury), thorax, abdomen/​pelvic contents, spine, and extremities. The severity score can range from 1 to 5 (1 is minor; 2 is moderate; 3 is serious; 4 is severe; 5 is critical; and 6 is likely unsurvivable). The Injury Severity Score (Baker et al., 1974) aims to summarize the total burden of injury by adding the quadratic scores of the three body regions with the highest score.

The extent of structural damage (CT classification) The extent of structural damage is best assessed by neuroimaging. While MR examinations are more sensitive than computed tomography (CT), CT remains the primary procedure for diagnostic imaging because of its sensitivity for detecting intracranial hematoma and the speed, availability, and safety of examination. The extent of structural damage on the CT is commonly classified according to the Marshall Classification (Fig. 40.1) (Marshall et  al., 1991). This provides a descriptive system which primarily differentiates between presence and absence of visible structural damage and differentiates between diffuse injuries (and a subclassification of these) and patients with mass lesions. Limitations of the Marshall Classification are that the scale is not ordinal, and that is does identify signs of raised intracranial pressure in patients with diffuse injury but not in those with mass lesions. Although the Marshall CT Classification was not developed from a perspective of prognosis, many studies have shown a clear association between the CT classification and outcome. The Rotterdam and Helsinki CT scores represent classification systems developed from a prognostic perspective (Maas et al., 2005; Raj et al., 2014).

Prognostic risk A different approach to classification is by baseline prognosis. Many prognostic models for predicting risk of mortality or unfavourable outcome in patients with TBI have been developed. The two largest prognostic models, CRASH and IMPACT, have been extensively validated externally, show good performance, and aim to predict mortality and unfavourable outcome (https://​ www.crash2.lshtm.ac.uk and https://​www.tbi-​impact.org) (Perel et al., 2008; Steyerberg et al., 2008; Lingsma et al., 2013; Han et al., 2014; Majdan et al., 2014). The IMPACT models were developed on patients with moderate and severe TBI; the CRASH models also included patients with milder injuries. They were developed with data available upon admission, prior to provision of in-​hospital care. They are therefore ideally suited for a baseline calculation of prognostic risk. Both approaches confirm that the largest amount of prognostic information is contained in a core set of three predictors:  age, motor score, and pupillary reactivity. The IMPACT study group further evaluated the additional benefit of factors such as structural imaging (CT characteristics such as mass lesion, cerebral oedema, traumatic subarachnoid haemorrhage), secondary insults (hypoxia, hypotension) and laboratory data (glucose, haemoglobin). Additional predictive power was noted in more complex models however the model could only explain

35% of variance in outcome. Many other factors have been suggested as having prognostic importance, although not investigated within multivariate models to the rigour found in the CRASH and IMPACT models, including MRI burden of injury, comorbidities, Injury Severity Score, time to craniotomy of more than 4 hours, raised ICP, and autoregulatory indices. Although useful for more reliably informing relatives of the anticipated outcome, it should be recognized that a prediction model is a population-​based estimate and yields no more than a risk estimate with a substantial element of uncertainty, which can be expressed in the confidence interval. For this reason, the greatest application of prognostic analysis is not at the level of the individual patient, but rather at the ‘group’ level for quantifying and stratifying the severity of brain injury, as a reference for evaluating quality of care and for stratification and covariate adjustment in clinical trials (Maas et al., 2017). It is worthy to add that recent studies are showing the superiority in early MR imaging as a prognostic tool in TBI, but there is still uncertainty about its discriminative predictive value and which acute phase lesion patterns correlate with long-​term outcome (Haghbayan et al., 2014). Although a significant body of evidence has indicated that MRI is superior to CT in detecting most types of traumatic lesions, CT is currently used as a routine while the use of MRI is sporadic in the acute phase (Haghbayan et al., 2014). There are several studies emphasizing the value of biomarkers as a tool for prognostications such as S100β protein and ApoE4 allele: S100 S100β protein is a promising biomedical marker for the diagnosis, monitoring, and prognosis of TBI severity. Preoperative estimation of serum S100β in patients with TBI could be used as a prognostic predictor for postoperative survival and neurological outcome (Goyal et al., 2013) as well as being responsive to secondary insults. Measuring levels of S100β is useful in evaluating the severity of TBI and in the determination of long-​term prognosis in patients presenting with moderate and severe injury (Mercier et al., 2013). ApoE4 A current meta-​analysis indicated that the ApoE4 allele might be associated with a poor prognosis in patients with severe TBI, but it may be used as a biomarker in predicting the prognosis of patients with TBI (Zeng et al., 2014). APoE4 allele presence influences recovery rate from severe TBI independent of other covariates (Alexander et al., 2007) and may also relate to the risk of subsequent cognitive decline.

Assessment of outcome: The Glasgow Outcome Scale (GOS) and beyond Accurate characterization of outcome after TBI is essential, both from a clinical and research perspective. Historically, most TBI studies have used the Glasgow Outcome Scale or the extended version of the GOS to provide an overall global assessment of functional outcome after TBI (Table 40.5). The GOS was initially described in 1975 as a global assessment of function following TBI (Jennett and Bond, 1975). While appropriate for more severe TBI, it was recognized that the scale was relatively insensitive at the upper levels and for this reason the Glasgow

CHAPTER 40  Epidemiology of head injury and outcome after head injury

Table 40.5  Glasgow Outcome Scale (GOS) and Glasgow Outcome Scale Extended (GOSE) Scale

GOS

Scale

GOSE

1

Dead

1

Dead

2

Vegetative

2

Vegetative

3

Severe disability

3

Lower severe disability

(Conscious but dependent)

4

Upper severe disability

Moderate disability

5

Lower moderate disability

(Independent but disabled)

6

Upper moderate disability

Good recovery

7

Lower good recovery

(Can resume normal activities)

8

Upper good recovery

4 5

GOS: Reprinted from The Lancet, Volume 305, issue 7905, Bryan Jennett, Michael Bond, Assessment of Outcome After Severe Brain Damage: A Practical Scale, pp. 480–​4, Copyright (1975), with permission from Elsevier. GOSE: Reproduced with permission from B Jennett, J Snoek, MR Bond, N Brooks, Disability after severe head injury: observations on the use of the Glasgow Outcome Scale, Journal of Neurology, Neurosurgery & Psychiatry, Volume 44, Issue 4, pp. 285–​93, Copyright © 1981 BMJ Publishing Group Ltd.

Outcome Scale Extended (GOSE) was introduced. Alterations in major roles, independence in living and participation in social and leisure activities are assessed by the investigators and used to summarize the effects of diverse changes caused by injury. It should be recognized, however, that the GOSE remains a global approach to classifying outcome and that ceiling effects exist. TBI may cause problems in many domains: a wide variety of neuropsychiatric disturbances can be associated with functional impairments and low quality of life. Neuropsychological sequelae include: mood disturbances (Bombardier et al., 2010; Tanev et al., 2014); cognitive impairment; personality changes; and social and family effects. An overlap may exist between sequelae of mild TBI and symptoms consistent with PTSD. The overlap of cognitive disruption between mild TBI and PTSD reflect the complex interplay of neurological, psychological, and physical factors in veterans with mild TBI and/​or PTSD, and highlights the need for specialized evaluation and management (Benge et al., 2009; Brenner, 2011). It has long been thought that quality of life measures are not appropriate for TBI as the neurocognitive sequelae might prevent accurate rating of the patient’s perception of quality of life. However, this has been shown to be false. Patient-​reported outcome measures in the sense of quality of life assessments are highly relevant to the field of TBI. Quality of life (QOL) measures may be generic; for example, the Short-​Form-​36 (SF36) or disease specific. A disease-​ specific measure QOL by a multidisciplinary working party and published as the Quality of Life after Brain Injury (QOLIBRI) scale. This scale provides a profile of health-​related quality of life (HRQOL) in domains typically affected by brain injury. However, for more global assessments a summary measure may be preferable and as such the QOLIBRI-​OS presenting a short six-​item assessment was proposed. QOLIBRI was found to have moderate to strong relationships with QOLIBRI-​OS, GOSE, Short-​Form-​36, Hospital Anxiety and Depression scale (Ware et al., 1993; Truelle et  al., 2010; von Steinbüchel et  al., 2010a; von Steinbüchel et  al., 2010b). An important aspect is to capture the patient’s own perspective on QOL (Dijkers, 2004). To date, outcome assessments have

tended to be unidimensional, focusing on single domains. An alternative approach might be to combine these in multidimensional assessment scores. Much further work will however be required to determine which instruments are best suited for specific situations and to determine the weighting of different factors. Historical data on outcome differentiated by severity of injury, show that patients with severe TBI have a mortality rate of 36%, 5% are classified as vegetative, 15% have severe disability, 15–​20% moderately disability, and 25% good recovery. Patients with moderate injury have been reported to have a mortality of 7%, vegetative classification of 1%, severe disability rate of 7%, moderate disability rate of 25%, and good recovery rate of 60%. Patients with mild injury have a very low mortality with a substantial majority designated as good recovery but nevertheless a high frequency of ongoing symptoms including headache, visual disturbance, dizziness, fatigue, and memory, concentration, and higher executive function impacting on family and social life and employment. With advances in treatment there is evidence that six-​month outcome has improved in moderate and severely injured patients treated in specialist intensive care (increased in moderate disability and good recovery from 40.4% to 59.9%) (Patel et al., 2002).

Health economic cost The economic cost of TBI management is tremendous and is a huge burden on society although, unfortunately, there are a paucity of studies that quantify this cost. Due to the long-​term economic impact and burden of disability to family, work, and society, the average lifetime cost per person for TBI in the United States is estimated to be US$396 000. In the United States the Centre for Disease Control and Prevention estimates direct and indirect costs of TBI to be more than US$58 billion in total. In most studies, the costs of non-​hospitalized TBI were not considered at all. Feigin et al. performed a population-​based TBI study and found out that 30% of cases were from non-​hospitalized patients (Feigin et al., 2013). Although the cost of TBI is known to increase by severity (see Table 40.6), the unexpectedly large number of mild TBI in any population (95% of all TBI cases) means that the total cost of treating these cases is nearly three times that of moderate/​severe TBI. The economic burden and overall outcome can be reduced by targeting both high cost injuries and population-​based programmes aimed at reducing the incidence of lower cost mild injuries.

Conclusion We have only uncovered the tip of the iceberg in TBI. There is a pressing need for further epidemiological studies with standardized criteria involving, specifically, lower income countries. This will have Table 40.6  The cost of TBI is known to increase by severity especially with those requiring hospitalization and higher demand on rehabilitation services TBI Grade

Mild

Direct medical 21 160 (Farhad cost/​person et al., 2013) to (US$) 35 954

Moderate

Severe

25 271 (Farhad et al., 2013) to 81 153 (McGregor and Pentland, 1997)

57 637 (Farhad et al., 2013) to >100 000 (Faul et al., 2007)

479

480

Section 10  Neurotrauma and intensive care

a significant impact in relocating resources and better economic plans to reduce the healthcare burden on societies in the long run. One critical area in TBI research that requires attention is the effects of risk factors and predictors of TBI outcome. Currently, a few prospective multicentre studies are undergoing. The two largest studies are the Transforming Research and Clinical Knowledge in TBI study (TRACK-​TBI) to enrol 3000 patients and the Collaborative European Neuro-​Trauma Effectiveness Research in TBI study (CENTER-​TBI) (Maas et al., 2015) to enrol 5000 patients with controls, over the next five years, using an extensive standardized set of variables.

FURTHER READING Butcher, I., McHugh, G.S., Lu, J., et  al. (2007). Prognostic value of cause of injury in traumatic brain injury: results from the IMPACT study. J Neurotrauma, 24,  281–​6. IMPACT. International Mission for Prognosis and Analysis of Clinical Trials in TBI. http://​www.tbi-​impact.org/​ Perel, P., Arango, M., Clayton, T., et al. (2008). Predicting outcome after traumatic brain injury: practical prognostic models based on large cohort of international patients. BMJ, 336(7641),  425–​9.

REFERENCES Aarabi, B., Alden, T., & Chestnut, R. (2001). Management and prognosis of penetrating brain injury—​guidelines. J Trauma, 51(suppl), S1–​86. Adekoya, N. & Majumder, R.R. (2004). Fatal traumatic brain injury, West Virginia, 1989–​1998. Public Health Rep, 9–​119(5), 486–​92. Alexander, S., Kerr, M.E., Kim, Y., Kamboh, M.I., Beers, S.R., & Conley, Y.P. (2007). Apolipoprotein E4 allele presence and functional outcome after severe traumatic brain injury. J Neurotrauma, 24(5),  790–​7. De Almeida, C.E.R., de Sousa Filho, J.L., Dourado, J.C., Gontijo, P.A., Dellaretti, M.A., & Costa, B.S. (2016). Traumatic brain injury epidemiology in Brazil. World Neurosurgery, 87,  540–​7. Andelic, N., Anke, A., Skandsen, T., et al. (2012). Incidence of hospital-​ admitted severe traumatic brain injury and in-​hospital fatality in Norway: a national cohort study. Neuroepidemiology, 38(4), 259–​67. Baker, S.P., O’Neill, B., Haddon, W. Jr., & Long, W.B. (1974). The Injury Severity Score: a method for describing patients with multiple injuries and evaluating emergency care. J Trauma, 14, 187–​96. Balestreri, M., Czosnyka, M., Chatfield, D.A., et al. (2004). Predictive value of Glasgow Coma Scale after brain trauma: change in trend over the past ten years. J Neurol Neurosurg Psychiatry, 75,  161–​2. Benge, J.F., Pastorek, N.J., & Thornton, G.M. (2009). Postconcussive symptoms in OEF-​OIF veterans:  factor structure and impact of posttraumatic stress. Rehabil Psychol, 54(3),  270–​8. Bombardier, C.H., Fann, J.R., Temkin, N.R., Esselman, P.C., Barber, J., & Dikmen, S.S. (2010). Rates of major depressive disorder and clinical outcomes following traumatic brain injury. JAMA, 303(19), 1938–​45. Brenner, L.A. (2011). Neuropsychological and neuroimaging findings in traumatic brain injury and post-​traumatic stress disorder. Dialogues Clin Neurosci, 13(3), 311–​23. Carroll, L.J., Cassidy, J.D., Holm, L., et al. (2004). Methodological issues and research recommendations for mild traumatic brain injury: the who collaborating centre task force on mild traumatic brain injury. J Rehabil Med, 43, Suppl(2), 113–​25.

Centers for Disease Control and Prevention. Report to Congress on traumatic brain injury in the United States: epidemiology and rehabilitation. 2015. https://www.cdc.gov/traumaticbraininjury/pdf/tbi_ report_to_congress_epi_and_rehab-a.pdf (accessed 7 august, 2019). Cheng P., Yin P., Ning P., et al. (2017). Trends in traumatic brain injury mortality in China, 2006–2013: a population-based longitudinal study. PLoS Med 14, e1002332. Dijkers, M.P. (2004). Quality of life after traumatic brain injury: a review of research approaches and findings. Arch Phys Med Rehabil, 85(4 Suppl 2), S21–​35. Farhad, K., Khan, H.M., Ji, A.B., Yacoub, H.A., Qureshi, A.I., & Souayah, N. (2013). Trends in outcomes and hospitalization costs for traumatic brain injury in adult patients in the United States. J Neurotrauma, 30(2),  84–​90. Faul, M., Wald, M.M., Rutland-​Brown, W., Sullivent, E.E., & Sattin, R.W. (2007). Using a cost-​benefit analysis to estimate outcomes of a clinical treatment guideline: testing the Brain Trauma Foundation guidelines for the treatment of severe traumatic brain injury. J Trauma, 63(6), 1271–​8. Fazel, S., Wolf, A., Pillas, D., Lichtenstein, P., & Långström, N. (2014). Suicide, fatal injuries, and other causes of premature mortality in patients with traumatic brain injury a 41-​year Swedish population study. JAMA Psychiatry, 71(3), 326–​33. Feigin, V.L., Theadom, A., Barker-​Collo, S., et al. (2013). Incidence of traumatic brain injury in New Zealand: a population-​based study. Lancet Neurol, 12(1),  53–​64. Frost, R.B., Farrer, T.J., Primosch, M., & Hedges, D.W. (2013). Prevalence of traumatic brain injury in the general adult population: a meta-​analysis. Neuroepidemiology, 40(3),  154–​9. Gerber, L.M., Chiu, Y.L., Carney, N., Härtl, R., & Ghajar, J. (2013). Marked reduction in mortality in patients with severe traumatic brain injury. J Neurosurg, 119(6), 1583–​90. Goyal, A., Failla, M.D., Niyonkuru, C., et al. (2013). S100b as a prognostic biomarker in outcome prediction for patients with severe traumatic brain injury. J Neurotrauma, 30(11), 946–​57. Haghbayan, H., Boutin, A., Laflamme, M., et al. (2014). The prognostic value of magnetic resonance imaging in moderate and severe traumatic brain injury: a systematic review and meta-​analysis. Intensive Care Medicine, 40(1), S202. Han, J., King, N.K., Neilson, S.J., Gandhi, M.P., & Ng, I. (2014). External validation of the CRASH and IMPACT prognostic models in severe traumatic brain injury. J Neurotrauma, 31(13), 1146–​52. Jiang J.Y., Gao G.Y., Feng J., et al. (2019). Traumatic brain injury in China. Lancet Neurol, 18, 286–295. Jennett, B. & Bond, M. (1975). Assessment of outcome after severe brain damage. Lancet (London, England), 1(7905),  480–​4. Koizumi, M.S., Lebrão, M.L., Mello-​Jorge, M.H., & Primerano, V. (2000). [Morbidity and mortality due to traumatic brain injury in Sao Paulo City, Brazil, 1997]. Arq Neuropsiquiatr, 58(1),  81–​9. Lingsma, H., Andriessen, T.M., Haitsema, I., et al. (2013). Prognosis in moderate and severe traumatic brain injury. J Trauma Acute Care Surg, 74(2), 639–​46. Lu, J., Marmarou, A., Choi, S., et al. (2005). Mortality from traumatic brain injury. Acta Neurochir Suppl, 95,  281–​5. Maas, A.I., Hukkelhoven, C.W., Marshall, L.F., & Steyerberg, E.W. (2005). Prediction of outcome in traumatic brain injury with computed tomographic characteristics: a comparison between the computed tomographic classification and combinations of computed tomographic predictors. Neurosurgery, 57(6), 1173–​82.

CHAPTER 40  Epidemiology of head injury and outcome after head injury

Maas, A.I.R., Menon, D.K., Adelson, P.D., et  al. (2017). Traumatic brain injury: integrated approaches to improve prevention, clinical care, and research. Lancet Neurol, 16(12), 987–​1048. Majdan, M., Lingsma, H.F., Nieboer, D., Mauritz, W., Rusnak, M., & Steyerberg, E.W. (2014). Performance of IMPACT, CRASH and Nijmegen models in predicting six month outcome of patients with severe or moderate TBI: an external validation study. Scand J Trauma Resusc Emerg Med, 22, 68. Majdan M, Plancikova D, Brazinova A, et al. (2016) Epidemiology of traumatic brain injuries in Europe: a cross-sectional analysis based on hospital discharge statistics and death certificates in 2012. Lancet Public Health, 1:e76–83. Marshall, L.L.F., Marshall, S.B., Klauber, M.R., & Van Berkum Clark, M. (1991). A new classification of head injury based on computerized tomography. J Neurosurg, 75(1), s14–​s20. Masson, F., Thicoipe, M., Aye, P., et  al. (2001). Epidemiology of severe brain injuries: a prospective population-​based study. J Trauma, 51(3),  481–​9. McGregor, K. & Pentland, B. (1997). Head injury rehabilitation in the U.K.: an economic perspective. Soc Sci Med, 45(2), 295–​303. McKinlay, A., Grace, R.C., Horwood, L.J., Fergusson, D.M., Ridder, E.M., & MacFarlane, M.R. (2008). Prevalence of traumatic brain injury among children, adolescents and young adults: prospective evidence from a birth cohort. Brain Inj, 22(2), 175–​81. McMillan, T.M., Teasdale, G.M., Weir, C.J., et al. (2011). Death after head injury: the 13-​year outcome of a case control study. J Neurol Neurosurg Psychiatry, 82(8),  931–​5. Menon, D.K., Schwab, K., Wright, D.W., et al. (2010). Position statement: definition of traumatic brain injury. Arch Phys Med Rehabil, 91(11), 1637–​40. Mercier, E., Boutin, A., Lauzier, F., et al. (2013). Predictive value of S-​100beta protein for prognosis in patients with moderate and severe traumatic brain injury: systematic review and meta-​analysis. BMJ, 346, f1757. Patel, H.C., Menon, D.K., Tebbs, S., Hawker, R., Hutchinson, P.J., & Kirkpatrick, P.J. (2002). Specialist neurocritical care and outcome from head injury. Intensive Care Med, 28(5), 547–​53. Peeters, W., van den Brande, R., Polinder, S., et al. (2015). Epidemiology of traumatic brain injury in Europe. Acta Neurochir (Wien), 157(10), 1683–​96. Perel, P., Arango, M., Clayton, T., et al. (2008). Predicting outcome after traumatic brain injury: practical prognostic models based on large cohort of international patients. BMJ, 336(7641),  425–​9. Raj, R., Siironen, J., Skrifvars, M.B., Hernesniemi, J., & Kivisaari, R. (2014). Predicting outcome in traumatic brain injury: development of a novel computerized tomography classification system (Helsinki computerized tomography score). Neurosurgery, 75(6), 632–​47. Roozenbeek, B., Maas, A.I.R., & Menon, D.K. (2013). Changing patterns in the epidemiology of traumatic brain injury. Nat Rev Neurol, 9(4),  231–​6.

Rosenfeld, J.V., Maas, A.I., Bragge, P., Morganti-​Kossmann, M.C., Manley, G.T., & Gruen, R.L. (2012). Early management of severe traumatic brain injury. Lancet, 380(9847), 1088–​98. Rutland-​Brown, W., Langlois, J.A., Thomas, K.E., & Xi, Y.L. (2006). Incidence of traumatic brain injury in the United States, 2003. J Head Trauma Rehabil, 21(6),  544–​8. Servadei, F., Ciucci, G., Piazza, G., & Bianchedi, G. (1988). A prospective clinical and epidemiological study of head injuries in northern Italy:  the Comune of Ravenna. Italian J Neurol Sci, 9(5), 449–​57. Steyerberg, E.W., Mushkudiani, N., Perel, P., et al. (2008). Predicting outcome after traumatic brain injury:  development and international validation of prognostic scores based on admission characteristics. PLoS Med, 5(8), e165. Stocchetti, N., Pagan, F., Calappi, E., et al. (2004). Inaccurate early assessment of neurological severity in head injury. J Neurotrauma, 21(9), 1131–​40. Sundstrøm, T., Sollid, S., Wentzel-​Larsen, T., & Wester, K. (2007). Head injury mortality in the Nordic countries. J Neurotrauma, 24(1), 147–​53. Tanev, K.S., Pentel, K.Z., Kredlow, M.A., & Charney, M.E. (2014). PTSD and TBI co-​morbidity: scope, clinical presentation and treatment options. Brain Inj, 28(3), 261–​70. Teasdale, G. & Jennett, B. (1974). Assessment of coma and impaired consciousness. A practical scale. Lancet, 2(7872),  81–​4. Truelle, J.-​L., Koskinen, S., Hawthorne, G., et al. (2010). Quality of life after traumatic brain injury: the clinical use of the QOLIBRI, a novel disease-​specific instrument. Brain Inj, 24(11), 1272–​91. Von Steinbüchel, N., Wilson, L., Gibbons, H., et al. (2010a). Quality of Life After Brain Injury (QOLIBRI): scale development and metric properties. J Neurotrauma, 27(7), 1167–​85. Von Steinbüchel, N., Wilson, L., Gibbons, H., et al. (2010b). Quality of Life After Brain Injury (QOLIBRI): scale validity and correlates of quality of life. J Neurotrauma, 27(7), 1157–​65. Ware JE, Jr., Snow KK, Kosinski M, et  al. (1993). SF-​36 Health Survey: Manual and Interpretation Guide. Boston, MA: The Health Institute, New England Medical Center. Zeng, S., Jiang, J., & Wang, Z.-​ H. (2014). Prognostic value of apolipoprotein E epsilon4 allele in patients with traumatic brain injury:  a meta-​analysis and meta-​regression. Genetic Testing and Molecular Biomarkers, 18(3), 202–​10.

RELATED LINKS TO EBRAIN Epidemiology of Head Injury. https://learning.ebrain.net/course/view. php?id=1082 Outcome after Head Injury. https://learning.ebrain.net/course/view. php?id=1081

481

41

Pathophysiology of traumatic brain injury John K. Yue, Hansen Deng, Ethan A. Winkler, John F. Burke, Catherine G. Suen, and Geoffrey T. Manley

Pathophysiology of primary brain injury Head trauma is classified into two pathophysiological stages that differ in mechanism and necessitate distinct clinical treatments. Primary trauma is due to a mechanical load that translates into deformation of cerebral tissue (e.g. neurons, glia, axons, and blood vessels), which then initiate cellular responses that lead to disturbances in autoregulation and metabolism. The first signs of trauma range from rapid, indiscrete depolarization of cortical neurons as exemplified in concussion, to focal lesions due to blunt impact, to severe damage such as diffuse shearing of axonal integrity, to acute vascular disruptions in the form of haematomas. Secondary injuries occur across a more protracted phase and include cerebral hypoxia, hypotension, swelling, and raised intracranial pressure (ICP). In a 2007 meta-​analysis of 9205 moderate to severe TBI patients as part of the International Mission for Prognosis and Analysis of Clinical Trials in TBI (IMPACT) study, McHugh and colleagues reported that hypoxia, hypotension, and hypothermia are common on admission with observed prevalence of 20%, 18%, and 10% respectively (McHugh et al., 2007). Secondary insults, particularly when in combination during acute injury and hospitalization, correlate with poor outcome, and require imperative attention, stabilization, and surgical and/​or medical management.

Biomechanics of closed head injury Closed head injuries are cases when the cranium and dura mater remain intact after impact, with patients presenting with a range of neurologic signs and symptoms. Mechanical loading to 5% gelatine in the human skull caused focal lesions to areas of indentation or fracture, and diffuse shearing throughout the subcortex occurred during rotation (Holbourn, 1944). Extensive studies in animal models, human cadavers, computational simulations, and advanced imaging have since characterized traumatic brain deformation, but direct human measurements have traditionally been challenging to obtain as part of clinical studies.

The primary phase of trauma from TBI is defined through the nature of the mechanical load, the type of induced motion, and the duration and velocity of impact. Forces along different axes produce unique modes of stress upon the head (e.g. elongation, compression, bending, shear, and torsion) (Ommaya et al., 2002). Resultant injuries in TBI cannot be attributed to any single appellation. External force generally translates into an impact or impulsive load on the head, and in the real world can include both. Head motion can be translational or rotational, depending on the alignment of two axes: the axis of force application, and the axis between the centre of gravity of the head and the atlanto-​occipital joint. Whether the head is fixed or free to extend/​flex by the neck is an important determinant for injury mechanism.

Impact loading Impact loading occurs from collision typically with a short duration of 3–​7 ms measured in low velocity falls. Impact can result in a skull fracture and associated extradural haematoma as well as focal cerebral lesions. Contusions known as coup (directly below impact) or contre-​coup (diametrically opposite to impact) are observed. High positive pressure is observed at the coup site, and the transmission of the force vector through the brain parenchyma generates a slapping effect to the contre-​coup site. At the cellular level, high negative pressure at the contre-​coup site, the development and subsequent collapse of cavitation bubbles known as contre cavitations, along with the brain parenchyma bouncing against inner posterior skull are associated with the contre-​coup lesions. Elastic rebounding of the skull results in significant spikes in cerebrospinal fluid (CSF) pressure and worsen tissue damage. Intracerebral haematomas are frequently associated with contusions. In the early stage after trauma, a contusion is haemorrhagic and more severe at the crest of gyrus than in the sulcus. It is associated with swelling that subsides with time though discoloration remains. Autopsy findings show higher incidence of coup and/​or contre-​coup contusions in the frontotemporal and basal regions of brain, irrespective of the primary impact site. Lagrangian strain tensor fields during occipital deceleration show anatomic constraints near the frontal base of

484

Section 10  Neurotrauma and intensive care

skull that must be breached in order for the brain to compress against the occipital bone, thus predisposing these regions for injury. The contre-​coup lesion is associated with temporal and occipital impacts. Cerebral rotation during impact is also thought to be a significant contributing factor to contre-​coup lesions, although definitive evidence on convulsional gliding have yet to confirm this theory.

Impulse loading Impulse loading occurs due to inertial forces during translational or rotational motion. Within normal physiological circumstances, CSF pressure provides a considerable dampening effect on brain displacement during motion, and the lack of CSF is shown to significantly increase convolutional gliding and shear strain (Pudenz and Shelden, 1946). Brain displacement lags behind that of the skull and the affixed dura mater due to the inertial force of cerebral mass, and convolutional gliding occurs in varying degrees depending on the region of brain, inducing diffuse damage to white matter tracts. Relative to the region of the skull base near the sella and suprasellar space, the brain parenchyma is more mobile. White matter is stiffer than grey matter, thus more strain is distributed at the interface. Vascular, neural, and dural elements (e.g. distal internal carotid arteries, optic nerves, olfactory tracts, oculomotor nerves, and pituitary stalk) that tether brain to the base of skull experience are most susceptible to diffuse axonal injury (DAI) due to inertial effect. The splenium of the corpus callosum and dorsolateral brainstem can experience acute DAI due to a similar trajectory to that of the skull base, relative to the more compliant cerebral cortex. Translational and rotational motions occur simultaneously to varying degrees. Purely translational motion of the head and neck generally does not produce acute subdural haematomas (SDH), diffuse petechial haemorrhages, or injuries to the cervical cord at linear acceleration of up to 1000 g. When linear acceleration is combined with angular motion of the head and neck, the aforementioned injuries are generated. During angular acceleration, there is a higher incidence of deep-​seated haematomas in the basal ganglia in addition to DAI. Rotational insults, often seen in vehicular accidents, induce shear strain as axons and vessels are stretched or compressed beyond their physiologic limits and often result in acute SDH from tearing of subdural bridging veins. When a single inertial load combines with a minor terminal impact load as in severe extension/​flexion whiplash trauma, SDH, gliding contusions, and spinal cord injuries are prevalent. During angular acceleration, the rotational axis is at either the occipital condyles or the base of the neck, amplifying the rotational load and shearing potential. This correlates with sudden vehicular impacts, when translational motion is applied at the seat to induce inertial force onto the head and neck. Rear loading occurs when an applied force from behind causes the head to rotate back at the base of the neck, while front loading occurs when an applied force from the front causes the head to rotate forward.

Static or quasistatic loading A less common type of mechanical load is static or quasistatic loading, in which gradual compression occurs with negligible velocity and acceleration—​mechanistically similar to a closing elevator door. A steady load results in skull fractures and cerebral injuries that are deeper than cortical contusions from an impact load. In

comparison to blunt impact trauma with relatively shorter duration and higher velocity, energy from crushing trauma tends to be transmitted to the foraminae and hiatus of the middle cranial fossa, causing in damage to the associated cranial nerves, sympathetic nerves, and intima of blood vessels.

TBI classification by morphology Traditional approaches to TBI classification focus on the extent of injury (focal vs. diffuse) and the anatomical location of injury with respect to the meninges (extradural/​extra-​axial vs. intracerebral/​ intra-​axial). The following section proceeds via the latter classification scheme, beginning with the extra-​axial layer closest to the skull and proceeding inward to intracerebral tissue.

Extra-​or epidural haematoma (EDH) EDH overlays the dura mater and typically results from a fracture of the thin squamous part of the temporal bone which lacerates the middle meningeal artery. The resulting haematoma strips the dura from the inner table of the skull forming an ovoid mass that compresses the adjacent brain. The haematoma is constrained by the periosteum, which passes through the cranial sutures so these haematomas do not cross suture lines. Venous EDH is less common and results from the disruption of the dural venous sinuses at the vertex, posterior fossa, or the anterior aspects of the cranial fossa. Although less common, EDH may also be found in the frontal region, parietal region, or the posterior fossa in 25% of cases. EDHs are far less common than subdural haematomas and account for approximately 2% of brain injuries. Although EDHs occur in all age groups, they are far more common in patients below 50  years of age and particularly in paediatric patients where they are primarily due to meningeal and diploic vein haemorrhage and the dura mater is tightly adherent to the paediatric skull. The relatively lower venous pressure and the strong dural adhesion to the skull are responsible for the variation and occasional subacute presentation of EDH in children. Radiographic progression of EDHs can be classified into three categories on computed tomography (CT) imaging:  type I  (acute), type II (subacute) and type III (chronic) occurring at presentation in 58%, 31%, and 11% of cases, respectively (Zimmerman and Bilaniuk, 1982). Type I is characterized as a dense haematoma with low density ‘swirl’ indicative of bleeding (i.e. a hypodense zone of active bleeding from the torn vessel into a dense organizing clot). The rise in pressure eventually produces tamponade of the bleeding site and progresses to type II—​a homogenous, hyperdense, and organized clot. Type III is characterized by a low density collection due to blood resorption by perivascular tissue, along with a contrast-​enhanced membrane consisting of neovascularity and granulation tissue. The classic presentation of EDH, seen in 15–​20% cases, is a patient who presents following a head trauma with transient loss of consciousness, progressing to a lucid interval, before deteriorating neurologically to a coma. The lucid interval represents recovery from the concussion, during which the patient may complain of severe headache on the ipsilateral side of the lesion accompanied by nausea, vomiting, and lethargy with subsequent decompensation into unconsciousness due to increased ICP secondary to the expanding EDH. The lucid interval is not specific to an EDH and was

CHAPTER 41  Pathophysiology of traumatic brain injury

originally described in cases of acute SDH. Neurological deterioration after EDH may be accompanied by contralateral hemiparesis, ipsilateral oculomotor nerve paresis, decerebrate rigidity, arterial hypertension, cardiac arrhythmias, respiratory disturbances, and if uncorrected, apnoea and death.

Subdural haematoma (SDH) SDH results from tearing of bridging veins that cross the subdural space to communicate with the venous sinuses or from disruption of superficial pial arteries on the brain surface. SDH is not limited by dural attachment and extends over the cortical surface in the typical biconcave pattern over the tentorium cerebelli or into the interhemispheric fissure. SDH is classified as acute, subacute, and chronic. Acute SDH presents as a crescent-​shaped hyperdense collection. Subacute SDH is isodense with symptomatic improvement and typically presents between 7 and 21 days. Chronic SDH (CSDH, >21 days postinjury) is hypodense and may not present symptomatically until the collection has expanded significantly and signs of compression appear (Sambasivan 1997). Acute SDH is associated with stereotypic motor disorders, impaired oculomotor reflexes, and following transtentorial herniation of the uncus of the temporal lobe, unilaterally fixed and dilated pupils. Arterial bleeds are associated with larger clots near the Sylvian fissure. Postoperative complications include reaccumulation of the haematoma and infection at the accessed site (e.g. osteomyelitis, meningitis, ventriculitis). Acute traumatic SDH has high reported mortality rates of 22% to 66%. While a lesion surgically evacuated within four hours to reduce the risk of disability, conservative methods may be considered for patients with shallow haematomas in a good clinical state. The Brain Trauma Foundation guidelines recommend specific criteria based on thickness, midline shift, and GCS for intervention, however, these must be viewed in the context of the individual patient, their premorbid state, the length of history and any associated injuries (Chesnut, 1997). Historically, patients undergoing decompression within four hours show a 30% mortality rate, compared with 90% for patients surgically evacuated after four hours (Seelig et al., 1981). More recently series with rapid surgical intervention show a lower overall mortality rate between 8% and 10% (Bajsarowicz et al., 2015). Other predictors include extent of neurologic deficit, sex, and postoperative ICP. Rapid diagnosis and time to surgical evacuation remain of highest prognostic importance after acute SDH. The issue of decompressive craniectomy at the time of subdural haematoma evacuation is currently being investigated in a randomized control trial (RESCUE-​ASDH). Currently, the decision to remove the bone flap at the time of surgery (primary decompressive craniectomy) is based on the injury burden on the brain (e.g. the presence of large contusions), the severity of injury assessed clinically and by mechanism of injury and most importantly the degree of intraoperative swelling at the time of surgery.

with cerebral atrophy. The initial small volume haematoma may be asymptomatic and the initial insult is often forgotten. In a subset of patients, an inflammatory neomembrane forms and potentiates ongoing haemorrhage and swelling of the enclosed haematoma by breakdown of blood products and the development of an osmotic gradient across the neomembrane. The chronicity of this process leads to a clinical presentation several weeks after the initial injury, with an SDH hypodense to brain. The haematoma may nonetheless show areas of acute haemorrhage from repeated haemorrhages, presumably from the inflammatory neomembrane. Clinical features of CSDH are headache, hemiparesis, speech disturbance (dominant hemisphere) or behavioural disturbances (e.g. emotional outbursts, lack of concentration, manic and depressive states), and ultimately coma if large and untreated. Bilateral CSDH are more likely to progress to coma rapidly and consequently are treated at a smaller absolute volume. CSDH occur most commonly on the cerebral convexity but can occur in the interhemispheric fissure, over the tentorium cerebelli and rarely the posterior fossa. Treatment can be with burr hole drainage (one or two to aid lavage of the haematoma), by mini-​craniotomy or via closed drainage systems. The risks of intervention include infection of the subdural space, seizures, and recurrence. Proliferating microcapillaries in the capsule of SDH are implicated in bleed recurrence due to marked mitotic potential of endothelial cells and vascular permeability. There is level 1 evidence that recurrence can be reduced by the placement of a soft drain into the subdural space for 48 hours (Santarius et al., 2009).

Subarachnoid haemorrhage (SAH) SAH is a frequent finding in closed head injuries as a result of direct damage to cortical vessels. It correlates with poorer outcome and more severe injury, although in most cases it appears be a reflection of a greater degree of violence at injury rather than being directly responsible for secondary injury. Nevertheless, in a subset of patients traumatic SAH may also contribute to secondary insults (e.g. cerebral swelling, hemodynamically significant vasospasm, and disturbances of metabolism and autoregulation due to reduced adaptive thresholds of the brain to additional physiological shifts). Post-​ traumatic vasospasm can be observed as early as two days postinjury and reaches maximum intensity between five to seven days (Weber et al., 1990). Cortical SAH is associated with progression of adjacent cerebral contusions, suggestive of subarachnoid bleeding in the vault as an early sign of cortical microbleeding and significant intraparenchymal brain damage. In TBI patients matched by GCS, age, sex, and the presence of intracranial mass lesions after blunt head injuries, traumatic SAH patients spend more time in the intensive care unit (ICU), are less likely to be discharged home, and are 1.5 times more likely to die during acute hospitalization. In penetrating TBI there is a significant correlation between SAH and poor outcome.

Chronic subdural haematoma

Intraventricular haemorrhage (IVH)

Chronic subdural haematoma (CSDH) can be thought of as a distinct pathological entity as it often presents in a specific demographic (elderly patients or those with cerebral atrophy) as a specific clinical syndrome. The presumed pathology is a minor head injury that leads to a small haematoma from tearing of the stretched bridging veins that span the subdural space and are therefore unsupported in those

IVH is found in 1.5–​3.0% of all head trauma and predominantly in severe TBI. Damage to the septum pellucidum, choroid plexus, and subependymal veins in the fornix are seen in post-​mortem exams in patients with primary IVH. The incidence of primary IVH, referring to isolated bleeding confined to the ventricular system, is low. An estimated 21% of patients with IVH attain functional recovery

485

486

Section 10  Neurotrauma and intensive care

(Glasgow Outcome Scale (GOS) moderate disability or good recovery) due to presence of concomitant intracranial injuries. In infrequent cases of isolated IVH, more improvement in GOS is possible. In general, the small sample sizes make it difficult to interpret the specific prognostic significance of IVH as opposed to the associated TBI.

Cerebral contusions Cerebral contusions are focal or multifocal and located in cortical or subcortical regions from direct impact or an acceleration/​deceleration injury. Contusions appear as a bruise to the surface from damaged small arteries, veins, or capillaries. Herniating contusions occur when tissue is displaced from one cranial compartment to another typically along the margin of the falx cerebri, the tentorium, or the foramen magnum leading to compression of the herniating tissue. Intermediary contusions are subcortical lesions affecting the corpus callosum, basal ganglia, hypothalamus, and brainstem. Cortical impact murine models demonstrate that cerebral oedema peaks at 24 hours coincident with a marked reduction of cerebral blood flow (CBF) to contused cortex. CBF normalizes after day seven post-​trauma, during which focal areas of hyperaemia appear adjacent to regions of low flow. Studies of the human brain after non-​ penetrating head injury show contusions being more severe in the frontal and temporal lobes and in patients who do not experience lucid intervals (Adams et al., 1985). Cellular analyses of excised contusions from emergency craniotomies corroborate necrosis as the predominant finding and apoptosis as part of the secondary injury cascade. Tumour necrosis factor increases immediately and rapidly in CSF during acute neuronal degeneration and gradually decreases within a few hours. Neutrophils and macrophages accumulate with elevated chemokines to sustain the cerebral inflammatory cascade. ICP can become elevated due to cerebral swelling and delayed haematoma formation in the pericontusional area, thus requiring ICP monitoring regardless of the need for surgical intervention. Unilateral lesions portray better prognosis while bilateral lesions can be delayed in clinical presentation. This secondary deterioration in clinical state can be delayed as long as 10 days following injury and requires careful counselling of patients and their relatives on discharge. Hyponatraemia can exacerbate this risk and serum Na must be monitored before considering discharge, as the syndrome of inappropriate antidiuretic hormone secretion can coexist in the context of TBI.

Diffuse axonal injury (DAI) DAI is caused by angular acceleration leading to damage of axonal integrity. Diffuse brain injury is seen in up to 50% of TBIs and spans a range of outcomes from concussion to coma and vegetative state. Classically, DAI is defined as diffuse damage in the cerebral hemispheres, corpus callosum, brainstem and cerebellum (Gennarelli et al., 1982). Long-​tract structures (axons and blood vessels) are especially at risk. This is graded on MR as: Grade I (parasagittal white matter of the cerebral hemisphere) Grade II (Grade I plus focal lesion in the corpus callosum), and Grade III (Grade II plus focal lesion in the cerebral peduncle) (Table 41.1). Grade I and II typically show marked improvement in GCS within 2 weeks while Grade III requires ~2 months for recovery, suggestive of prolonged loss of consciousness or coma with DAI involving the brain stem. Strictly, a definitive diagnosis of DAI is established by immunostaining for β-​amyloid precursor protein (β-​APP) at autopsy and identifying axonal retraction balls in the deep white

Table 41.1  Diffuse axonal injury grading DAI grade

Pathology

I

Grey-​white matter interface (commonly parasagittal white matter of frontal lobes, periventricular temporal lobes)

II

Focal lesions in corpus callosum (commonly posterior body and splenium) in addition to Grade I lesions

III

Brainstem (commonly rostral midbrain, cerebellar peduncles, medial lemnisci, and corticospinal tracts) in addition to Grade I and II lesions

GRE: T2-​weighted gradient-​recalled-​echo (GRE) magnetic resonance (MR) technique SW: Susceptibility-​weighted (SW) magnetic resonance (MR) technique. Data from Tong KA, et al, Hemorrhagic shearing lesions in children and adolescents with posttraumatic diffuse axonal injury: improved detection and initial results, Radiology, Volume 227, Issue 7, pp. 332–​9, 2003

matter. In vivo diagnosis is challenging as it may present microscopically and in combination with one or more type of brain injuries including: contusion, anoxia, and intracerebral haemorrhage. Although there is ample data implicating DAI in moderate or severe TBI, its involvement in mTBI is difficult to prove histologically as patients rarely die from their injuries. Minor degrees of axonal injury may provide the pathophysiologic basis for deficits experienced during and after mTBI/​concussion. Alternatively, functional and metabolic disturbances may be responsible with little structural axonal injury (Vespa et al., 2005).

Mild TBI/​Concussion The United States (US) Centers for Disease Control and Prevention (CDC) defines mild TBI (mTBI)/​concussion as a transient neurological disturbance caused by rapid linear and/​or rotational acceleration and deceleration forces, resulting in a disruption in cerebral structural or vascular physiology. The American Congress of Rehabilitation Medicine (ACRM) defines mTBI by one or more of the following: loss of consciousness (LOC), loss of memory immediately before or after the incident, alteration in mental state, or focal neurologic deficit that may or may not be transient. The U.S. Department of Veterans Affairs and Department of Defense (VA/​DOD) provide similar clinical guidelines while specifying that presence of intracranial lesions classify the patient with, at minimum, moderate TBI. An estimated 75% of TBI patients are classified as ‘mild’, which is likely an underestimate as most of these patients do not present to the emergency department. Current classifications of concussion centre on the impairment of neurological function, the disturbance of vision, memory, or equilibrium, alteration of consciousness (AOC), and/​or LOC. The diagnosis of concussion is largely appropriate for the non-​penetrating head trauma that results in one or more of the following:  confusion/​disorientation, LOC less than 30 minutes or post-​traumatic amnesia (PTA) less than 24 hours in duration, transient focal neurological deficits, and/​or seizure, with a GCS of 13–​ 15 upon acute medical evaluation; however, multiple concussion grading scales currently exist without definitive consensus. The lack of evidence in imaging studies, perhaps due to lack of sensitivity, is a challenge to the clinical diagnosis of mTBI/​concussion. Recent efforts in utilization of magnetic resonance imaging (MRI), including MR Spectroscopy, for TBI diagnosis has shown promise in identifying a subset of patients with structural changes post-​mTBI. The pathophysiology of mTBI/​concussion begins with a sudden change in momentum of the head, causing mechanical stretching or compression of the partially tethered brain against the cranial vault,

CHAPTER 41  Pathophysiology of traumatic brain injury

initiating diffuse neuronal firing, discharge, and axonal shearing. Neurocognitive deficits tend to occur with minimal detectable pathology and can resolve over time. However, even patients with GCS score 13–​15 present with acute and subacute glucose metabolic disturbances on positron emission topography (PET) and require careful assessment for postconcussive risks (Giza and Hovda, 2001). Cerebral microdialysis immediately following mTBI/​ concussion contains elevated levels of excitatory amino acids and ionic fluxes throughout hemispheric regions, the hippocampus, and the brainstem. Specifically, there is a sharp increase of extracellular K+ through voltage-​gated channels. Studies using electroencephalogram (EEG) demonstrated that postconcussive EEG was excitatory and epileptiform in nature comparable to generalized seizures. Likewise, during this period the sensory evoked potential was completely lost (Shaw, 2002). In early stages after mTBI/​concussion, increased glucose utilization and hyperglycolysis sustain the indiscriminate neuronal depolarization and firing, but can also restore the disturbed membrane potential. Later, diffuse neuronal suppression (e.g. spreading depression) takes effect. Spreading depression is characterized by near-​complete breakdown of ion gradients, loss of electrical activity, and vasoconstriction, which can trigger infarct growth and initiate cascades leading to apoptotic cell death (Dreier, 2011). Physiological disturbances persist after mTBI/​ concussion and increase susceptibility to further injury. After the initial hypermetabolism, hypometabolism and reduction in CBF may persist for up to 4 weeks. Additional excitation (e.g. direct cortical stimulation or recurrent concussion) prolongs cellular recovery and induced cell death due to inability to match increased metabolic demand in part due to decoupling between neuronal activation and CBF after concussion. Dependent on the severity of impact, intracellular accumulation of Ca2+ also remains elevated, which impairs mitochondrial metabolism and initiates programmed cell death. In the surviving cells, reduction in cellular plasticity is associated with neurotransmission changes, particularly in subunits of N-​methyl-​ D-​aspartate (NMDA) receptors.

Penetrating brain injury (PBI) Compared with blunt TBI, PBI is less prevalent and associated with worse prognosis, occurring when a non-​blunt projectile breaches the cranium and dura mater. CSF leaks are common and may require surgical intervention. When a projectile also induces an exit wound, it is termed a perforating brain injury. The projectile crushes soft brain tissue in its path and generates bone fragments at impact, which can require surgical debridement. The shape and kinetic energy of projectile E = 1/​2mv2 (E = energy, m = mass, v = velocity) are associated with the degree of tissue damage. A high velocity projectile generates waves of compression and re-​ expansion (cavitation wave) and inflicts focal shearing damage, parenchymal contusions, and haematomas. CT provides clear characterization of in-​driven bone fragments, missile trajectory, extent of tissue damage, haematomas, and mass effects. Cerebral angiography is recommended due to the high risk of vascular injury. High risk factors include a penetrating track crossing the ventricle, involving both hemispheres, crossing the geographical centre of the brain or associated vascular injury. Seizure control following PBI is a key part of clinical management.

Cerebral metabolism and autoregulation The brain is responsible for the greatest overall resting energy expenditure by organ at approximately 22%, compared with the liver 21%, the heart 9%, and the kidneys 8%. The brain is 2% of overall weight while accounting for 20% of basal cardiac output and 20% of resting O2 consumption. The cerebral metabolic rate of oxygen (CMRO2) remains relatively constant in healthy adults at approximately 3.5 ml O2/​100 g/​min and may be as high as 5.2 ml O2/​100 g min in children aged 3–​12 years. Glucose is the primary source of ATP in the brain and approximately 50% of energy generated from cerebral metabolism is utilized for synaptic activity (i.e. neurotransmitter production, release, and uptake); 25% is for maintaining for electrochemical gradients, and the remaining 25% is for molecular transport, biosynthesis, and other processes. Aerobic metabolism is the primary process in the brain through which glucose is efficiently converted to 36–​38 molecules of ATP by glycolysis, the citric acid cycle, and aerobic oxidation. Glucose metabolism is disturbed after TBI. There is a significant increase in anaerobic glycolytic turnover and elevated level of extracellular lactate in cerebral circulation as neurons and astrocytes convert glucose to two molecules of ATP and two molecules of lactate. Hyperglycolysis contributes to prolonged elevated lactate/​glucose ratio, CSF lactic acidosis, and compromised mitochondrial function via calcium-​ mediated interference. The duration of hyperglycolysis and lactate accumulation may reflect the extent of injury and worse prognosis. See Figure 41.1. CMRO2, cerebral metabolic rate of glucose (CMRG), and CBF are common measurements utilized in neurophysiology. In healthy adults, CMRO2 is 3.3 ml/​100 g/​min, CMRG is 5.5 mg/​100 g/​min, and CBF is 54 ± 12 ml/​100 g/​min. In comatose patients with TBI, CMRO2 is reduced to between 1.2 and 2.3 ml/​100 g/​min and is associated with GCS score and recovery outcome. Storage of glucose and oxygen by neurons and astrocytes is limited, therefore continuous CBF is essential to ensure survival. Arteries and arterioles 30–​300 um in diameter can alter their resistance to control cerebral circulation. There is a linear relationship between CMRO2 and CBF consistent with the notion of diffusion-​limited oxygen delivery. Flow-​ metabolism coupling is the process by which blood flow changes in relation to the metabolic demands of cerebral tissue, largely influenced by the changes in tissue CO2. PaO2 does not affect CBF except in patients with drastically reduced PaO2 below 50 mmHg. Cerebral metabolism declines after TBI with corresponding decreases in CBF.

Pressure and viscosity The Hagen-​Poiseuille law of laminar flow in a cylindrical tube is utilized to describe cerebral blood flow in relation to cerebral perfusion pressure (CPP), vessel diameter, and blood viscosity. CBF= k[CPP × d(4)]/​(8 × l × v), where k is a constant, d is artery diameter, l is artery length, and v is blood viscosity. CBF is maintained at 50 ml/​100 g/​min in the adult to facilitate constant metabolic supply to the brain, provided that CPP is in the range of 60–​160  mmHg (Cipolla, 2009). CPP is calculated as the difference between mean arterial pressure (MAP) and ICP. CBF is also maintained via autoregulation (Fig. 41.2). Cerebral vasculature possesses myogenic capacity to dilate or contract to counteract deviations in wall tension (that reflect the CPP) to maintain

487

Section 10  Neurotrauma and intensive care

constant flow. Outside the normal CPP 60–​160 mmHg, vessels become either maximally dilated or experience forced dilation (i.e. ‘pressure breakthrough’). Autoregulation is impaired or absent in the majority of severe TBI patients at some point in their clinical course. When autoregulation is lost, the brain becomes vulnerable to systemic pressure disturbances leading to secondary insults (e.g. ischaemia from reduced CBF or swelling from excess CBF). Severe TBI patients are more prone to secondary injury from low CPP. Thus, flow-​metabolism and autoregulation are distinct mechanisms allowing the cerebral circulation to meet local metabolic demand and to stabilize against systemic pressure disturbances. Changes in whole blood viscosity, largely determined by haematocrit and serum fibrinogen, affect CBF and induce an autoregulatory response under normal physiological circumstances. An increase in viscosity decreases metabolic supply to the brain and causes arterial dilation, while a decrease in viscosity increases metabolic supply and causes arterial constriction. Following acute cerebral infarction, increased haematocrit and fibrinogen are associated with reduced CBF, and that changes in blood viscosity may have important hemodynamic effects on cerebral circulation as well during acute stress.

Glucose metabolism Glucose

2 ATP

2 Lactate

2 NADH

2 NAD+ 2 CO2 2 NADH

Hyperglycolysis after TBI

2 NAD+

2 ADP

2 Pyruvate

2 Acetyl-CoA Oxidative phosphorylation

+

6 NAD 2 FAD 2 ADP

Krebs cycle

6 NADH 2 FADH2 2 ATP

CO2 reactivity

Fig. 41.1  Glucose metabolism can be divided into three distinct processes. Glycolysis is the conversion of glucose into pyruvate. This generates two ATP molecules (net) and does not require oxygen if coupled with conversion of pyruvate into lactate (anaerobic respiration). The Kreb’s cycle is a circular biochemical pathway that consumes the carbon skeletons from acetyl-​coenzyme A and generates reducing equivalents (nicotinamide adenine dinucleotide (NAD)-​H and flavin adenine dinucleotide (FAD)-​H), carbon dioxide, and two ATP molecules. The final step of oxidative phosphorylation utilizes the reducing equivalents generated in the preceding steps to generate a hydrogen ion gradient across the mitochondrial membrane which drives a proton ATPase that generates an additional 34 ATP molecules, giving a total of 38 ATP molecules per molecule of glucose (aerobic respiration).

Maximal dilation

CO2 reactivity is the process by which partial pressure of arterial CO2 (PaCO2) affects CBF and the cerebral vasculature. Within the normal range of PaCO2 between 20 and 60 mmHg, every 1 mmHg fluctuation results in 2% to 3% change in CBF. Hypercarbia (hypoventilation) results in vasodilation and increased CBF, while hypocarbia (hyperventilation) has the inverse effect. This is mediated through changes in pH of the perivascular space via carbonic anhydrase, and this forms the basis of the acetazolamide challenge. The normal response to acetazolamide administration (e.g. 1 g intravenous) is vasodilation and augmentation of CBF to 30–​60% over

Maximal constiction

...Reactive...

100

CBF (ml/min/100 g)

488

Maximal constriction (R0 = R1 = R2) C0 B0

50

LL0 LL1

A0 A1 0 0

LL2

B1 B2

UL1

C2

C1

UL0 R1

R2

UL2

Normocapnia

R0

Hypocapnia (mild) Hypocapnia (severe)

A2 50 60

100

D

150

200

CPP (mmHg)

Fig. 41.2  Cerebrovascular autoregulation (A, autoregulation) is the mechanism by which a constant blood flow (B, blood flow) is maintained in the face of variations in systemic blood pressure. Over a range of blood pressures CBF is constant, however above (UL, upper limit) and below (LL, lower limit) this range the vessels are maximally vasoconstricted or vasodilated (respectively) and changes in blood pressure are transmitted as changes in CBF. This risks ischaemia below LL and exacerbates intracranial hypertension above UL. Therapeutic hyperventilation (C0–​2) in order to reduce intracranial pressure leads to vasoconstriction (R0–​2). This has two potentially damaging effects. Firstly, CBF is reduced risking ischaemia. This can be mitigated by using a measure of intracranial oxygenation. Secondly, the autoregulatory range is reduced, making it more difficult to target an appropriate CPP. This can be mitigated by using a metric of autoregulation such as pulse reactivity index (PRx). Reproduced with permission from Meng, Lingzhong; Gelb, Adrian W., Regulation of Cerebral Autoregulation by Carbon Dioxide, Anesthesiology, Volume 122, Issue 1, pp. 196–​205, Copyright © 2015 Wolters Kluwer Health, Inc.

CHAPTER 41  Pathophysiology of traumatic brain injury

10–​15 minutes. A failure to vasodilate in response to acetazolamide implies maximal vasodilation, usually resulting from chronic ischaemia. Hyperaemia and metabolic acidosis in CSF are associated with the acute phase of TBI, typically in the first 24 hours, and patients with persistent loss of CO2 reactivity risk high mortality or severe neurological compromise. While CO2 reactivity causes changes in both CBF and AVDO2, autoregulation affects CBF with AVDO2 maintained relatively constant. Hyperventilation and control of PaCO2 are important therapeutic interventions in the management of raised ICP, however this requires cerebral oxygenation monitoring to ensure that excessive arterial vasoconstriction does not inflict ischaemia.

Secondary injury after severe head trauma Complex neurochemical pathways after TBI disrupt cerebrovascular circulation and inflict secondary injuries that may last from minutes to weeks, and treatment of secondary injuries is essential to maximizing patient outcome. Though not all these mechanisms are fully understood, the International Mission for Prognosis and Clinical Trials in TBI (IMPACT) database identified hypoxia and hypotension to be present in 20% and 18% of TBI patients, respectively (Murray et al., 2007). Repeated investigations have consistently identified five clinical variables that are correlated with morbidity and poor outcome: arterial hypotension, reduced CPP, elevated ICP, hypoxemia, and pyrexia.

Glutamate-​mediated excitotoxicity Mitochondria play an essential role in cerebral energy metabolism, calcium homeostasis, and reactive oxygen species generation. As early as one-​hour post TBI, mitochondrial oxidative phosphorylation can fail, resulting in reduction of ATP production, failed energy-​dependent membrane ion pumps, and disrupted homeostasis in neural cellular metabolism. Depressed ATP production increases the rate of glycolysis, elevating lactate production, and potential accumulation in the CSF and extracellular fluid of the injured brain. High lactate levels can further contribute to neuronal dysfunction by inducing acidosis, membrane damage, altered blood brain barrier permeability, and cerebral oedema. Activation of intracellular digestive enzymes (i.e. peroxidases, proteases, phospholipases) and caspases lead to structural damage at the subcellular level, DNA degradation, and neuronal death (i.e. necrosis and apoptosis). The relationship between excessive Ca2+ influx and glutamate-​ triggered injury is well supported, resulting in uncoupling of mitochondrial electron transfer from ATP synthesis and overactivation of calpains. Glutamate-​mediated excitotoxicity activates NMDA receptors, triggering neuronal depolarization with an unchecked influx of calcium into mitochondria. Mitochondrial dysfunction becomes the main cause of energy failure of damaged tissue, as ion pumps are unable to restore membrane potentials. Experiments indicate that Ca2+ loading at L-​type voltage-​sensitive channels is non-​toxic, but Ca2+ loading at NMDA receptors is neurotoxic. The production of toxic reactive oxygen species (i.e. nitric oxide, superoxide, and hydrogen peroxide) is also contingent on NMDA receptor activation, therefore NMDA receptors are key drivers of excitotoxicity after head injury. These cellular changes lead to

cerebral oedema, elevated ICP, vascular compression, and ultimately herniation (Lucas and Newhouse, 1957). When assessing causes and treatments for ischaemia and abnormal oxygen measurements such as cerebral perfusion pressure and jugular venous oxygen saturation, mitochondrial dysfunction should be considered, as maintenance of adequate perfusion may not improve clinical outcome due to underlying mitochondrial impairment. A  pragmatic definition of mitochondrial dysfunction is an inability to efficiently carry out oxidative phosphorylation of ADP in the presence of adequate oxygen and metabolic substrate. Elevated extracellular glutamate is important in mediating both traumatic and ischaemic brain injury. Interstitial glutamate is shown to increase by several mechanisms: BBB damage and extravasation of glutamate to region of impact, membrane damage, and microporation, upregulation of complexin I  and complexin II enhancing exocytosis, and glutamate transporter impairment. Beta-​lactam antibiotics (e.g. ceftriaxone) are potent stimulators of glutamate transporter (GLT1), which is the astroglial protein responsible for inactivating synaptic glutamate. Beta-​lactams increase brain expression of GLT1 and its functional activity, which decreases glutamate neurotoxicity to provide neuroprotective potential in ischaemic injury and neuronal degeneration (Rothstein et al., 2005). Calcium-​induced mitochondrial matrix swelling and ruptured outer mitochondrial membranes uncouple the respiratory chain and further impair ATP production. The unregulated influx of calcium and membrane rupturing also induces the opening of the mitochondrial permeability transition pore (mPTP), causing an apoptotic cascade. Cyclosporin A, an immunosuppressant investigated for its ability to inhibit mPTP opening in TBI patients, has demonstrated neuroprotective effects against mitochondrial swelling, membrane breakdown, and ionic homeostatic imbalance (Okonkwo et al., 1999).

Perturbations in cerebral metabolism TBI damages mitochondrial functionality to propagate neurodegeneration and limit neuroregeneration. Severe TBI induces a sudden shift from aerobic metabolism to hyperglycolysis/​anaerobic metabolism with high glucose turnover, contributing to cerebral acidosis. Even when sufficient oxygen is present, hyperglycolysis has been demonstrated in 56% of patients through findings using PET within 1 week following TBI. Acidosis shifts the haemoglobin oxygen-​dissociation curve to the right and may increase O2 supply to injured brain tissue. Elevated CSF lactate level after TBI typically normalizes within 12 hours of injury, except following cerebral ischaemia and diffuse cerebral swelling. The hypothesized role of lactate in injured brain energetics is controversial, and has changed from metabolic waste to supplemental fuel, neuroprotection, and signalling molecule in cortical neurons. Intracerebral microdialysis catheters (Fig. 41.3) allow for biochemical analysis of changes in extracellular neurotransmitters and metabolites during the secondary stage of TBI. The microdialysis measured lactate/​pyruvate ratio is a marker of anaerobic metabolism and correlates with outcome after TBI. (Timofeev et al., 2011). Increase in the lactate/​ pyruvate ratio can be due to cerebral ischaemia or mitochondrial dysfunction. In patients with elevated cerebral lactate/​pyruvate ratio (LPR), microdialysis demonstrates that exogenous lactate therapy significantly increases cerebral glucose levels. This correlates with previous evidence that lactate is directly oxidized to pyruvate in the

489

490

Section 10  Neurotrauma and intensive care

Microdialysis probe Dialysate

Perfusate

Perfusate (Drug or calibrator) Dialysate (analyte or drug)

Dialysis membrane

Fig. 41.3  Microdialysis is a technique for sampling the brain extracellular space. An invasive probe is inserted into the brain substance and perfused with fluid (perfusate) at ultralow flow rates (typically 0.3 µl/​minute). A semipermeable membrane at the tip of the catheter allows substances to diffuse into the catheter and are recovered in the exiting fluid (microdialysate). For clinical purposes, the commonest analytes are glucose, lactate, pyruvate, glutamate (an excitotoxic neurotransmitter) and glycerol (a measure of cellular membrane breakdown). The ratio between lactate and pyruvate (LPR, lactate-​pyruvate ratio) provides a measure of the balance between anaerobic and aerobic respiration (Fig. 41.1). An LPR more than 25 is regarded as pathological and correlates with poor clinical outcome in TBI.

mitochondria of neurons and astrocyte to contribute to cerebral aerobic metabolism (Lazaridis and Andrews, 2014). Increased activity of the pentose phosphate pathway (PPP) after severe TBI contributes to elevated cerebral glucose uptake exceeding CMRO2 and lactate production (Fig. 41.4). Infusion of isotype glucose after TBI increased PPP activity by 19.6% compared to 6.9% in controls. The PPP products (NADPH, ribose 5-​phosphate, and erythrose 4-​phosphate) that upregulate fatty acid synthesis, DNA repair and replication, and amino acid and neurotransmitter production are processes that are stimulated after brain injury (Dusick et  al., 2007). NADPH is used to produce reduced forms of glutathione and thioredoxin, cofactors for glutathione peroxidases, and peroxiredoxins that combat oxidative stress after acute injury. The PPP does not use oxygen as substrate and does not produce ATP, but rather serves as an antioxidant that contributes to the shift in glucose metabolism from glycolysis to PPP in ischaemic brain tissue. In TBI patients, 30% patients are found to have CMRO2 values profoundly reduced under non-​ischaemic conditions from normal value of 3.3 ml/​100 g/​min to 1.2–​2.3 ml/​100 g/​min. While CMRO2 after TBI correlate with GCS and is a strong predictor for neurologic outcome, reduced CMRO2 does not necessarily result from ischaemia nor directly cause patient decline. Causes of patient decline include the severity of primary and secondary injuries and the level of consciousness. Studies of severe TBI with mean GCS of 6 report a 25% incidence of reduced oxidative metabolism and persistent metabolic crisis (elevated LPR >40) despite absence of ischaemia, which suggest LPR as an indicator of widespread mitochondrial dysfunction causing metabolic depression following TBI, further impairing brain capacity to meet metabolic demands in coping with secondary insults. While very low CMRO2 indicates irreversible

cell injury, LPR may be utilized as a sensitive marker of reversible mitochondrial dysfunction in regions with no observable tissue damage. Published studies using 15O PET demonstrate a metabolism threshold of 37.6 μmol/​100 ml/​min resulting in irreversible injury.

Loss of autoregulation and pressure reactivity index As previously mentioned, vascular autoregulation is a myogenic mechanism where wall tension directly results in vascular smooth muscle constriction. An increase in MAP leads to vasoconstriction within 5 to 15 seconds to decrease cerebral blood volume (CBV) and ICP, while a decrease in MAP leads to vasodilation and increase in ICP. Autoregulation may be disrupted on the days following injury with the magnitude and risk of disruption depending on the severity of injury. Quantifying vascular reactivity is clinically important after TBI when cerebral vasculature can become unreactive. Slow wave fluctuations in MAP and ICP lasting 30 seconds to several minutes are common in ventilated patients after head injury. These changes are continuously monitored and their relationship is calculated as pressure reactivity index (PRx) between –​1 and +1 (Czosnyka et al., 1997). A negative correlation indicates good vasoreactivity (ICP declines as a result of arterial vasoconstriction when MAP increases) and autoregulatory protection, but a positive correlation indicates the loss of autoregulation and worse prognosis. In large series of patients, a PRx of less than 0.3 correlates with good outcome on dichotomized GOS. PRx is frequently compromised on the first day after TBI, and the loss of autoregulation in the first 48 hours has strong indication for additional secondary injury including cerebral hypoxia and further ischaemic insults. Preserved autoregulation is a crucial protective

CHAPTER 41  Pathophysiology of traumatic brain injury

Glucose-6-phosphate NADP+

Pentose phosphate pathway

NADPH 6-phosphogluconolactone H2O

Oxidative phase

H+ 6-phosphogluconate NADP+

Ribulose-5-phosphate

Xylulose-5-phosphate

NADPH CO2

Ribose-5-phosphate

Sedoheptulose-7-phosphate + Glyceraldehyde-3-phosphate Non-oxidative phase Erythrose-4-phosphate + Fructose-6-phosphate

Fructose-6-phosphate + Glyceraldehyde-3-phosphate

Glycolysis

Fig. 41.4  Pentose phosphate pathway. The pentose phosphate pathway (PPP) is an alternative pathway for glucose utilization and is found to be upregulated following TBI. It provides several biochemical substrates that are potentially protective, and this is an actively explored research avenue.

mechanism against transient MAP derangements. In addition to being a strong predictor of outcome, PRx is an important clinical tool in guiding posttraumatic treatment and CPP management. PRx changes dynamically with CPP making it possible to define the CPP threshold at which PRx is minimized to maximize autoregulatory protection: this has been termed ‘CPP optimal’ (CPPopt). Continuous PRx monitoring in TBI patients determines a real-​time CPPopt dependent on the severity of injury, and deviation from the CPP less than CPPopt appears to increase risk of mortality while CPP more than CPPopt appears to discriminate between good and poor outcome on dichotomized GOS. It is yet to be determined whether targeting CPPopt is a valid therapeutic target.

Raised intracranial pressure The Monro-​Kellie hypothesis explains that under normal conditions the total volume of the intracranial cavity remains constant and is formed from three main components: blood (CBV, both arterial and venous), fluid (CSF) and cerebral parenchyma. An increase in one compartment normally results in compensation in another through translocation of CSF and venous blood to maintain constant ICP. When compensatory mechanisms are exhausted, an exponential increase in ICP occurs. Patients with ICP below 20–​25 mmHg have significantly better outcomes. Raised intracranial pressure (>20–​ 25  mmHg) can induce ischaemia and hypoxia by reducing CPP

(MAP–​ICP); as CBF is directly proportional to CPP, the presence of hypotension, elevated ICP, and hypoxia can severely limit cerebral supply after TBI. Oedema is common after TBI and can be attributed to three distinct mechanisms:  vasogenic, cytotoxic, and osmotic. Structural damage due to breakdown of BBB endothelium causes intravascular flow of protein-​rich exudate into the brain interstitium, increasing extracellular volume without cell swelling. Second, intracellular volume can accumulate without BBB disruption. Ion influxes and increased membrane permeability of neural cells cause cytotoxic oedema and cellular swelling. Cytotoxic oedema is the most common brain oedema and results in decreased interstitial volume. Third, necrotic tissue is hyperosmolar, causing osmotic-​gradient driven fluid accumulation in the cell. Not surprisingly, oedema causes increased ICP and secondary ischaemic events that further exacerbate structural integrity and tissue damage. In the management of post-​ traumatic cerebral oedema, hyperosmolar agents such as mannitol or hypertonic saline be administered to lower ICP. However, it is important to note that once the BBB is compromised, mannitol may exacerbate vasogenic cerebral oedema. Bolus mannitol can initiate more than 10% ICP reduction among 86% of patients with intact autoregulation, whereas ICP reduction was observed in only 35% of patients with impaired autoregulation (Muizelaar et al., 1984). When autoregulation is lost,

491

492

Section 10  Neurotrauma and intensive care

osmotic agents may increase CBF without compensatory vasoconstriction and thus the desired reduction in ICP may not be achieved. Other potential mechanisms of mannitol action include acting as a free radical scavenger, improving microvascular flow by dehydrating endothelial cells and reducing haematocrit as well the osmotic load. Despite intensive therapeutic measures, intractable ICP is the primary cause of 46% of TBI deaths.

Cerebral blood flow disturbances In an uninjured brain, changes in CPP between 60 and 160 mmHg have only minor effects on CBF. Following trauma and autoregulatory impairment, CBF becomes increasingly dependent on CPP and minor fluctuations can instigate further ischaemic injury. As previously mentioned, CPP = MAP−​ICP, and CBF = CPP/​CVR, where CVR is cerebrovascular resistance. Secondary elevation in ICP may worsen CPP and decrease CBF. CBV is defined as millilitres of blood per 100 g cerebral tissue, and can be calculated based on measures of CBF using stable xenon and mean transit time after injection of contrast material. CBF is reflective of CBV under normal conditions, but acutely after trauma there is a dissociation between CBF and CBV due to physiological disturbances. Because metabolism decreases after severe head injury, determining AVDO2 is necessary for accurate interpretation of CBF and diagnosing hypo-​or hyperperfusion. CBF is generally low during the first 6 hours after injury and increase significantly in the first 24 hours. A low CBF is generally not associated with a high AVDO2, and is more likely a sign of low oxidative metabolism rather than ischaemia. High CBV does not always correlate with cerebral swelling in severe TBI; other factors besides hyperaemia contribute to cerebral swelling and refractory hypertensive ICP. In patients without acute hyperaemia, CBF consistently correlates with functional recovery (i.e. the lowest CBFs exist in patients with the most severe disabilities). More recent studies that combine oxygen 15-​ labelled positron emission tomography (15O PET) with fluorine 18-​labelled fluoromisonidazole (18F FMISO) further characterize macrovascular and microvascular ischaemia in early TBI (Veenith et  al., 2016). Comparison of ischaemic brain volume (IBV) and hypoxic brain volume (HBV) demonstrates that the diffusion hypoxia barrier in HBV is associated with microvascular collapse, whereas IBV represents macrovascular ischaemia and irreversible damage. Traumatic contusions show reduced CBF and CMRO2, but similar CBV and oxygen extraction fraction (OEF) as controls. In pericontusional regions with no identifiable tissue dysfunction, CMRO2 is lower, but CBF, CBV, and OEF are similar to the controls. Classic reasoning dictates that with reduced CMRO2, OEF increase to compensate. However, recent studies using DTI showed that surrounding contusions there is a frequent rim of low diffusion coefficient consistent with cytotoxic oedema, which may explain the minimal change in OEF because of widespread microvascular failure and selective neuronal loss.

Hyperaemia Hyperaemia, also known as vascular engorgement or luxury perfusion, results from disruption of flow-​metabolism coupling when CBF exceeds CMRO2 requirements. Increased CBV often indicates posttraumatic hyperaemia which in conjunction with elevated ICP portends poor recovery. While acute hyperaemia is frequently associated with hypertensive ICP (>20  mmHg), subnormal flow

indicates intact flow-​metabolism coupling rather than ischaemia. Hyperaemic ICP patients are often younger, present with low mean GCS (≤6), risk effacement of basal cisterns and refractory hypertension, and demonstrate gross loss of autoregulation. Of note, CBF is functionally coupled to CMRG and not CMRO2, thereby luxury perfusion may be an appropriate metabolic response to hyperglycolysis after injury. Historically, hyperventilation was initiated to decrease CBF and subsequently ICP, however this approach has been tempered by the risk of inducing ischaemia. After brain injury, CBF may no longer accurately reflect CBV due to disruption in normal physiology, and in patients with acute SDH, elevated ICP, and ischaemia, CBV is half of normal value. Therefore, multiple aetiologies may contribute to intracranial hypertension and cerebral swelling after severe head trauma.

Cerebral ischaemia Ischaemic injury consists of the infarct core and the penumbra. The core undergoes hypoxia-​induced loss of cellular homeostasis and necrosis, while the penumbra’s apoptotic processes can be targeted for therapeutic intervention. Cerebral ischaemia occurs when CBF does not meet cerebral metabolic demands (e.g. low CBF and high AVDO2). Assuming flow-​metabolism coupling is intact, jugular bulb oximetry is frequently utilized in the ICU to estimate the adequacy of CBF from its inverse relationship to AVDO2. While normal AVDO2 ranges from 4 to 9  ml 100  ml-​1, low CBF and ischaemic state increase oxygen extraction and increase AVDO2. Transcranial doppler non-​invasively measures the blood flow velocity of basal cerebral arteries in order to estimate CBF, with the caveat that haemorrhage can cause vasospasm which increases flow velocities irrespective of CBF. Ischaemia is often seen in acute SDH and diffuse cerebral swelling and inadequate metabolic supply to the brain after TBI is predictive of high morbidity and disability. PET has demonstrated that previous definitions of hyperaemia and ischaemia based on CBF and oxygen are inaccurate, and maintaining oxygen delivery may not be sufficient to prevent secondary injury. Mitochondrial dysfunction after TBI decreases oxygen demand, and the transition to hyperglycolysis increases lactate accumulation. Thus, decreased CBF can be misinterpreted as ischaemia, and the maintenance of adequate perfusion (CPP >70 mmHg) may not result in improvement due to impaired mitochondrial ATP synthesis. As mentioned earlier, in normal physiology, AVDO2 remains relatively constant while CBF changes in accordance to CMRO2/​ CMRG. When autoregulation is lost in TBI and the ability of CBF to meet metabolic demands is impaired, AVDO2 may increase to improve O2 extraction in the cerebral circulation. AVDO2 is capable of increasing from baseline value 6.7 ml/​100 ml to a maximum 13  ml/​100  ml. The normal CBF threshold at which AVDO2 compensation occurs is 18 ml/​100 g/​min, and can rise to 20 ml/​100 g/​ min following TBI due to decreased oxidative metabolism. 15O PET shows that diffusion barrier ischaemia (increased oxygen diffusion gradient from vasculature to interstitium) is due to microvascular disturbances such a perivascular oedema surrounding traumatic lesions causing decreased partial pressure of oxygen in brain tissue (PbO2). In addition to caspases and other proteins directly causing neuronal death, three distinct mitogen-​activated protein kinase (MAPK) inflammatory pathways contribute to ischaemic injury: extracellular signal-​regulated kinase (ERK) pathway, c-​Jun-​N-​terminal protein

CHAPTER 41  Pathophysiology of traumatic brain injury

kinase (JNK) pathway, and the p38 pathway. Emerging research on the use of biomarkers has yielded several markers of importance to neuronal, glial, and axonal damage. Neuron-​specific enolase is a glycolytic enzyme predominantly in neurons. S100B is a calcium binding protein predominantly in astroglia. Myelin basic protein is in white matter and a marker for axonal damage. Recent studies have found these to be important biomarkers of pathological processes involved in cerebral ischaemia and neuroinflammation following TBI and reflecting the time course of secondary injury.

Cortical spreading depression and epilepsy Cortical spreading depression (CSD) is a pervasive wave of ionic homeostasis failure that transiently interrupts cortical function. It is characterized by EEG depression, pial vasodilation, and a slow negative potential shift following severe head trauma and ischaemia (Leao, 1947). Spontaneous peri-​infarct depolarizations (PID) propagate to normal cerebral tissues and take on the characteristics of CSD. Electrocorticographic (ECoG) recordings show spontaneous depressions accompanied by stereotyped CSDs spreading across the cortical mantle at 3.3 mm/​min; ECoG background activity recovers spontaneously or 2 to 5 hours later (Fabricius, 2006). Propagating depolarization waves in neuronal and glial cells cause an increase in oxidative metabolism, compensated by increased CBF and metabolic activity; hence CSD does not cause neuronal injury in the normal brain when flow-​metabolism is intact. Haemodynamic disruption and flow-​metabolism decoupling are frequent phenomena after head injury, thus CSD may incur additional energy demand that cannot be metabolically compensated for by the injured brain. CSD alters BBB permeability via matrix metalloproteinase (MMP) activity. MMPs consists of neutral proteases implicated in wide range of processes: the opening of the BBB; immune cell infiltration of neural tissue; shedding of cytokine and cytokine receptors; direct cellular damage; and neuronal development and plasticity. CSD increases MMP-​9 activity, which leads to the breakdown of BBB and oedema formation. Neuronal expression of MMP-​9 and MMP-​2 is implicated in neurotoxicity after ischaemia and trauma, increasing the susceptibility to cerebral haemorrhage. CSD occurs more frequently in the younger acutely injured patients, and may be attenuated with glutamate receptor inhibitors, specifically NMDA receptor antagonists. CSD and PID contribute to the increase of final infarct volume in a step-​wise fashion as part of additional secondary insults and stress-​induced changes in the penumbra of primary lesions.

Seizures Acute seizures (both convulsive and non-​convulsive) are well documented secondary insults that may be due to neurochemical disarray and can exacerbate tissue damage and worsen clinical recovery. EEG monitoring demonstrates that the majority of seizures do not have overt motor manifestations and require continuous monitoring of seizure activity. ECoG recordings can distinguish the hallmark slow potential change of CSD from seizures; while CSD and seizures frequently occur together, there is no correlation between the two phenomena in regard to time, space, and intensity (Fabricius, 2006). Following moderate to severe TBI, seizures occur in more than 20% of patients in the first week; peak incidence is bimodal with an early peak at 29 hours and later peak at 140 hours, with mean duration of 2.8 minutes and a tendency to cluster. At these respective peaks for seizure activity, ICP is secondarily elevated, with the ictal ICP being

higher than interictal ICP. Elevated ICP and the risk for swelling due to seizure may be attributed to prolonged increases in CBF, extracellular oedema, and glutamate accumulation in the interstitium. Epileptiform activity is known to further stimulate glycolysis and neurochemical changes that may further stress neuronal integrity. Microdialysis LPR is higher for TBI patients with seizures (compared to without), and during the ictal periods (compared to interictal periods). Increased LPR is associated with prolonged metabolic stress, therefore seizure control is a therapeutic target for minimizing metabolic stress postinjury. Pericontusional regions, in particular in the frontotemporal region, are electrically active and capable of epileptiform activity. Patients with severe TBI are predisposed to seizures. Strong risk factors include the presence of a brain contusion and/​or subdural haematoma, and early evacuations of mass lesions may be associated with reduced seizure activity (Annegers et al., 1998). Focal seizures with secondary generalization occur in 78% of reported seizures after severe TBI.

Beta-​amyloid deposition and apolipoprotein E interaction TBI is shown to trigger pathological production and accumulation of amyloid-​(Aβ) peptides in the cerebral cortex, particularly localized to contusional and pericontusional regions (Roberts et  al., 1991). Cleavage of amyloid precursor protein (APP) produces apolipoprotein E (ApoE) and Aβ, and the polymorphism of the ApoE gene determines the risk for plaque deposition: ApoE ε4 carries greater risk, while ApoE ε3 (more frequent) and ε2 (rare) both reduce the risk for Aβ burden. Of the three isoforms, ApoE ε4 binds the least avidly to cytoskeletal proteins that would promote neurite growth and neuroprotective effect. Rather, ApoE ε4 binds more avidly to Aβ and promotes aggregation into amyloid fibrils. Studies using intracranial microdialysis measured higher levels of Aβ in patients with DAI compared to focal injuries, and thus Aβ accumulation may also depend on the injury type. Long-​term axonal degeneration and intra-​axonal Aβ deposition in the brain parenchyma is likely due to continued neurodegeneration and secondary insults. In vitro studies support evidence of ApoE-​Aβ interactions where TBI patients with ApoE ε4 allele have more Aβ deposition in the hippocampus and frontal cortex. In the acute phase postinjury, decreased ELISA measurements of Aβ concentrations and apolipoprotein E (ApoE) in the CSF occur. Their levels gradually increase over time to correlate with improved neurologic outcome. There is reduced Aβ production immediately following neuronal trauma, and the decreased ApoE level is possibly due to increased utilization of ApoE-​lipid complexes for repair as part of a coordinated response to neuronal injury. ApoE can be neuroprotective against cerebral ischaemia, and intraventricular infusion of ApoE reduced neuronal damage after ischaemia in animal models via clearance of lipid and cholesterol debris. ApoE facilitates anti-​inflammatory effects through downregulation of microglia and cytokine release (Kay et al., 2003). Aβ plaques from TBI are similar to the hallmark pathology found in early stages of Alzheimer’s disease. Long-​term follow-​up of patients with TBI also correlated with incidence of Alzheimer’s disease, with ApoE isotype as a key predictor. Notably, follow-​up 6 months after TBI reported 57% of patients possessing the ApoE ε4 allele to have unfavourable outcome (GOS: dead, vegetative state, or severe disability) compared to 27% of patients with the ApoE ε2 or ε3 allele.

493

494

Section 10  Neurotrauma and intensive care

Future directions Successful implementation of the TBI Common Data Elements (CDEs), along with demonstration of their utility in generating novel classification approaches with robust statistical power, informed the refinement of the NIH/​NINDS CDEs, version 2 (Hicks et al., 2013), and generated the impetus for the concurrent 22-​nation Collaborating European NeuroTrauma Effectiveness Research in TBI (CENTER-​TBI) study in the EU and 11-​site TRACK-​TBI study in the US (Manley and Maas 2013). These had the goals of enrolling close to 10 000 combined patients using the CDEs in order to provide novel multidimensional approaches to TBI classification, establish evidentiary benchmarks for quality of care, and generate large, well-​curated TBI data repositories to enable and ensure legacy research opportunities.

Conclusion While risk reduction of primary neurological trauma is mainly preventative in nature, understanding the sequelae of TBI in order to clinically manage and/​or pre-​empt secondary injuries is paramount in current neurosurgical and neurocritical care. Preventing potential complications in TBI patients who ‘talk and deteriorate’ emphasizes the crucial concept that a primary injury can readily induce pathophysiologic conditions for compounded secondary damage. Finally, TBI is recognized more as a disease process rather than a single event; the inherent heterogeneity of injury types and severities highlight the critical need to converge upon a classification scheme using a consensus definition of pathoanatomic injury—​measurable through validated clinical, neuroimaging, and serum biomarkers similar to current standard approaches in heart disease, cancer, and diabetes. Large-​scale international TBI trials using validated common data standards to accelerate advances in classification and prognosis are currently underway.

FURTHER READING Chesnut, R.M. (1997). Guidelines for the management of severe head injury. In:  Vincent, J.L. (ed.) Yearbook of Intensive Care and Emergency Medicine, pp. 50–​67. Berlin, Heidelberg: Springer. Lazaridis, C. & Andrews, C.M. (2014). Brain tissue oxygenation, lactate-​ pyruvate ratio, and cerebrovascular pressure reactivity monitoring in severe traumatic brain injury: systematic review and viewpoint. Neurocrit Care, 21(2), 345–​55. McHugh, G.S., Engel, D.C., Butcher, I., et al. (2007). Prognostic value of secondary insults in traumatic brain injury:  results from the IMPACT study. J Neurotrauma, 24(2), 287–​93. Ommaya, A.K., Goldsmith, W., & Thibault, L. (2002). Biomechanics and neuropathology of adult and paediatric head injury. Br J Neurosurg, 16(3), 220–​42. Veenith, T.V., Carter, E.L., Geeraerts, T., et al. (2016). Pathophysiologic mechanisms of cerebral ischemia and diffusion hypoxia in traumatic brain injury. JAMA Neurol, 73(5), 542–​50.

REFERENCES Adams, J.H., Doyle, D., Grahma, D.I., et  al. (1985). The contusion index: a reappraisal in human and experimental non-​missile head injury. Neuropathol Appl Neurobiol, 11(4), 299–​308.

Annegers, J.F., Hauser, W.A., Coan, S.P., & Rocca, W.A. (1998). A population-​based study of seizures after traumatic brain injuries. N Engl J Med, 338(1),  20–​4. Bajsarowicz, P., Prakash, I., Lamoureux, J., et al. (2015). Nonsurgical acute traumatic subdural hematoma: what is the risk? J Neurosurg, 123(5), 1176–​83. Chesnut, R.M. (1997). Guidelines for the management of severe head injury. In:  Vincent, J.L. (ed.) Yearbook of Intensive Care and Emergency Medicine, pp. 50–​67. Berlin, Heidelberg: Springer. Cipolla, M.J. (2009). The cerebral circulation. Colloquium Series on Integrated Systems Physiology: From Molecule to Function, 1(1),  1–​59. Czosnyka, M., Smielewski, P., Kirkpatrick, P., Laing, R.J., Menon, D., & Pickard, J.D. (1997). Continuous assessment of the cerebral vasomotor reactivity in head injury. Neurosurgery, 41(1), 11–​17; discussion  17–​19. Dreier, J.P. (2011). The role of spreading depression, spreading depolarization and spreading ischemia in neurological disease. Nat Med, 17(4), 439–​47. Dusick, J.R., Glenn, T.C., Lee, W.N., et al. (2007). Increased pentose phosphate pathway flux after clinical traumatic brain injury: a [1,2–​ 13C2] glucose labeling study in humans. J Cereb Blood Flow Metab, 27(9), 1593–​602. Fabricius, M. (2006). Cortical spreading depression and peri-​infarct depolarization in acutely injured human cerebral cortex. Brain, 129(3), 778–​90. Gennarelli, T.A., Thibault, L.E., Adams, J.H., Graham, D.I., Thompson, C.J., & Marcincin, R.P. (1982). Diffuse axonal injury and traumatic coma in the primate. Ann Neurol, 12(6), 564–​74. Giza, C.C. & Hovda, D.A. (2001). The neurometabolic cascade of concussion. J Athl Train, 36(3), 228–​35. Hicks, R., Giacino, J., Harrison-​Felix, C., Manley, G.T., Valadka, A.B., & Wilde, E.A. (2013). Progress in developing common data elements for traumatic brain injury research: version two—​the end of the beginning. J Neurotrauma, 30(22), 1852–​61. Holbourn, A.H. (1944). Mechanics of head injuries. Lancet, 243 (6293), 483. Kay, A.D., Petzold, A., Kerr, M., Keir, G., Thompson, E., & Nicoll, J.A. (2003). Alterations in cerebrospinal fluid apolipoprotein E and amyloid β-​protein after traumatic brain injury. J Neurotrauma, 20(10), 943–​52. Lazaridis, C. & Andrews, C.M. (2014). Brain tissue oxygenation, lactate-​ pyruvate ratio, and cerebrovascular pressure reactivity monitoring in severe traumatic brain injury: systematic review and viewpoint. Neurocrit Care, 21(2), 345–​55. Leao, A.A. (1947). Further observations on the spreading depression of activity in the cerebral cortex. J Neurophysiol, 10(6), 409–​14. Lucas, D.R. & Newhouse, J.P. (1957). The toxic effect of sodium L-​ glutamate on the inner layers of the retina. Arch Ophthalmol, 58(2), 193–​201. Manley, G.T. & Maas, A.I. (2013). Traumatic brain injury:  an international knowledge-​based approach. JAMA, 310(5), 473–​74. McHugh, G.S., Engel, D.C., Butcher, I., et al. (2007). Prognostic value of secondary insults in traumatic brain injury:  results from the IMPACT study. J Neurotrauma, 24(2), 287–​93. Muizelaar, J.P., Lutz, H.A., Becker, D.P. (1984). Effect of mannitol on ICP and CBF and correlation with pressure autoregulation in severely head-​injured patients. J Neurosurg, 61(4),  700–​6. Murray, G.D., Butcher, I., McHugh, G.S., et al. (2007). Multivariable prognostic analysis in traumatic brain injury:  results from the IMPACT study. J Neurotrauma, 24(2), 329–​37. Okonkwo, D.O., Buki, A., Siman, R., & Povlishock, J.T. (1999). Cyclosporin A  limits calcium-​induced axonal damage following traumatic brain injury. Neuroreport, 10(2),  353–​8.

CHAPTER 41  Pathophysiology of traumatic brain injury

Ommaya, A.K., Goldsmith, W., & Thibault, L. (2002). Biomechanics and neuropathology of adult and paediatric head injury. Br J Neurosurg, 16(3), 220–​42. Pudenz, R.H. & Shelden, C.H. (1946). The lucite calvarium; a method for direct observation of the brain; cranial trauma and brain movement. J Neurosurg, 3(6), 487–​505. Roberts, G.W., Gentleman, S.M., Lynch, A., & Graham, D.I. (1991). βA4 amyloid protein deposition in brain after head trauma. Lancet, 338(8780), 1422–​23. Rothstein, J.D., Patel, S., Regan, M.R., et al. (2005). β-​lactam antibiotics offer neuroprotection by increasing glutamate transporter expression. Nature, 433(7021),  73–​7. Sambasivan, M. (1997). An overview of chronic subdural hematoma: experience with 2300 cases. Surg Neurol, 47(5), 418–​22. Santarius, T., Kirkpatrick, P.J., Ganesan, D., et al. (2009). Use of drains versus no drains after burr-​hole evacuation of chronic subdural haematoma: a randomised controlled trial. Lancet, 374(9695), 1067–​73. Seelig, J.M., Becker, D.P., Miller, J.D., Greenberg, R.P., Ward, J.D., & Choi, S.C. (1981). Traumatic acute subdural hematoma. N Engl J Med, 304(25), 1511–​18. Shaw, N.A. (2002). The neurophysiology of concussion. Prog Neurobiol, 67(4), 281–​344.

Timofeev I, Carpenter KL, Nortje J, et  al. (2011). Cerebral extracellular chemistry and outcome following traumatic brain injury:  a microdialysis study of 223 patients. Brain, 134(Pt 2), 484–​94. Veenith, T.V., Carter, E.L., Geeraerts, T., et al. (2016). Pathophysiologic mechanisms of cerebral ischemia and diffusion hypoxia in traumatic brain injury. JAMA Neurol, 73(5), 542–​50. Vespa, P., Bergsneider, M., Hattori, N., et al. (2005). Metabolic crisis without brain ischemia is common after traumatic brain injury: a combined microdialysis and positron emission tomography study. J Cereb Blood Flow Metab, 25(6), 763–​74. Weber, M., Grolimund, P., & Seiler, R.W. (1990). Evaluation of posttraumatic cerebral blood flow velocities by transcranial doppler ultrasonography. Neurosurgery, 27(1), 106–​12. Zimmerman, R.A. & Bilaniuk, L.T. (1982). Computed tomographic staging of traumatic epidural bleeding. Radiology, 144(4), 809–​12.

RELATED LINK TO EBRAIN Neuropathology of Head Injury. https://learning.ebrain.net/ course/view.php?id=1088

495

42

Intensive care management of head injury Matthew A. Kirkman and Martin Smith

Introduction Traumatic brain injury (TBI) is a leading cause of death and disability in adults under 40 years of age. Outcome is determined by the severity of the primary injury and several mediators of secondary injury, including intracranial hypertension, systemic hypotension, and hypoxia, pyrexia, and hypo-​and hyperglycaemia. The monitoring, prevention, and management of factors that lead to secondary brain injury are the cornerstones of the intensive care management of TBI.

Early management of head injury The initial approach to head injury focuses on prompt and systematic assessment and management in line with standard prehospital and Advanced Trauma Life Support protocols (Fig. 42.1). Resuscitation and early management is a crucial stage at which mortality and morbidity after severe TBI can be influenced (Boer et  al., 2012). The prevention or immediate correction of hypoxaemia and hypotension, rapid diagnosis and evacuation of an expanding intracranial haematoma, and treatment of raised intracranial pressure (ICP) are key determinants of outcome. Minimizing time to stabilization and definitive treatment is of the essence. Although airway management and ventilation are crucial interventions in comatose TBI patients and those with a deteriorating conscious level, the role of prehospital endotracheal intubation and who should perform it remains controversial (Boer et al., 2012). Unless the patient is alert and cooperative, the cervical spine should be immobilized while awaiting imaging and/​or formal clinical assessment. Life-​threatening extracranial injuries should be stabilized prior to transfer.

Management of polytrauma in the context of head injury TBI is often associated with extracranial injuries, including spinal trauma (see Chapter 64), which can complicate brain-​directed therapy. Timely and appropriate intervention to stabilize or treat extracranial injuries is critical to prevent prolonged haemorrhagic shock, systemic inflammatory response syndrome, and multiple organ failure.

A particular challenge in patients with multiple injuries is balancing the competing interests of hypotensive resuscitation and its’ potential to limit ongoing blood loss and speed time to resuscitation with the risk of worsening associated brain injury. It is recommended that systolic blood pressure be maintained more than 90 mmHg at all times in patients with TBI. However, if a lower blood pressure is required during the treatment of life-​threatening haemorrhage, the duration of hypotension should be as short as possible and other physiological variables optimized to maximize cerebral oxygen delivery. In particular, hypocarbia must be avoided. ICP monitoring is recommended during multiple surgical procedures and/​or prolonged sedation for extracranial injuries when the opportunity for clinical assessment is limited (Stocchetti et al., 2014).

General principles for the intensive care management of head injury All patients with severe TBI should be managed in an intensive care unit (ICU) offering immediate access to multidisciplinary clinical neuroscience teams and other relevant specialties, supported by appropriate imaging and investigational facilities. Patients with mild or moderate TBI may also require admission to the ICU for a variety of indications (Box 42.1). The intensive care management of TBI is complex and involves meticulous general intensive care support to optimize systemic physiology, and interventions targeted to the injured brain (Table 42.1). In addition to the continuous monitoring and assessment of cardiorespiratory functions common to all critically ill patients, several techniques are available for global and regional brain monitoring (see next).

Cardiorespiratory function A single episode of hypoxaemia (PaO2 20–25 mmHg?

Initial options • Elevate head of the bed • Ensure the neck is in a neutral position • Avoid unnecessary pressure on the neck (e.g. from collar or endotracheal tube ties) • Optimise ventilation to precent hypercarbia • Sedation • Analgesia • Paralysis • Surgical removal of mass lesions (e.g. subdural or extradural haematoma)

ICP remains >20–25 mmHg?

• • • • • •

Further treatment options Ventriculostomy Inotropes Mannitol Hypertonic saline Loop diuretics Therapeutic hypothermia

ICP remains >20–25 mmHg?

Further treatment options • Barbiturates • Decompressive craniectomy • Decompressive laparotomy/thoractomy

Fig. 42.2  Management algorithm for raised intracranial pressure.

CHAPTER 42  Intensive care management of head injury

serum osmolality (>320  mOsm/​litre), neurological complications, and acute kidney injury. HS may have a more profound and long-​lasting effect on ICP compared with mannitol, but outcome benefits over mannitol have not been demonstrated (Diringer, 2013). Clinical outcome studies are inconsistent in terms of dose, concentration, frequency, and mode (bolus or continuous infusion) of HS administration, and the optimal osmolar load to lower elevated ICP is undefined. HS has multiple theoretical beneficial effects on the injured brain in addition to its osmotic action, including vasoregulatory, immunological, and neurochemical actions; it also expands intravascular volume, potentially augmenting CPP. Although previously believed to be only a by-​product of anaerobic metabolism, there is some evidence that lactate can act as a preferential fuel in the injured brain. Preliminary clinical data suggest that hypertonic lactate solutions spare cerebral glucose, improve cerebral energy metabolism, and effectively reduce elevated ICP (Bouzat and Oddo, 2014). Further data are required before its routine use can be recommended Barbiturates A Cochrane review has confirmed that the use of barbiturates in patients with TBI does not reduce mortality, and increases the risk of hypotension by 25% (Roberts and Sydenham, 2012). The adverse cardiovascular effects of barbiturates are likely to offset any benefit on CPP of ICP reduction. Barbiturates continue to be used in the treatment of refractory intracranial hypertension when first-​and second-​tier therapies have failed to control ICP, although in some centres there is a preference for decompressive craniectomy over barbiturates in this situation. Surgery The role of decompressive craniectomy, CSF drainage, and other neurosurgical interventions in the management of intracranial hypertension are discussed in Chapter 43. Evidence for intracranial pressure monitoring and management It is well recognized that intracranial hypertension is detrimental, but there is little evidence that monitoring and managing ICP improves outcomes. A  meta-​analysis of 14 studies including 24  792 patients with severe TBI concluded that ICP monitoring-​guided management of intracranial hypertension had no significant mortality benefit overall compared to treatment without ICP monitoring, although mortality was lower in patients who underwent ICP monitoring in studies published after 2012 (Yuan et al., 2015). There has been only one randomized study evaluating the utility of ICP monitoring in TBI. The Benchmark Evidence from South American Trials: Treatment of Intracranial Pressure (BEST-​TRIP) compared treatment guided by ICP monitoring with care based on imaging and clinical examination in the absence of ICP monitoring, and found no difference in three-​or six-​month outcomes (Chesnut et al., 2012). Patients in the ICP monitoring group received significantly fewer days of ICP-​directed treatment (hyperventilation, HS/​mannitol, and barbiturates), although the length of ICU stay was similar. This is in contrast to other (observational) studies, which have demonstrated an increased burden of brain-​directed therapy in patients in whom treatment is guided by ICP monitoring,

BEST-​TRIP has been criticized on several counts, but its findings reinforce the notion that assessment of ICP is integral to managing severe TBI. The evaluation and diagnosis of intracranial hypertension, whether through monitoring ICP directly, or indirectly through assessment of clinical and imaging variables, was fundamental to patient management in both groups.

Cerebral perfusion pressure Guidelines recommend that CPP should be maintained between 60 and 70 mmHg (Carney et al., 2017), through manipulation of MAP with fluid resuscitation and vasopressors/​inotropes or treatment of raised ICP, with evidence that both lower and higher values are associated with deleterious outcomes. CPP more than 70 mmHg is associated with increased risk of ALI related to overly aggressive fluid resuscitation or vasopressor therapy, and CPP less than 50 mmHg with cerebral hypoperfusion. An alternative approach, the Lund concept, targets a lower CPP (50 mmHg) to minimize increases in intracapillary hydrostatic pressure and intracerebral water content in an attempt to avoid secondary increases in ICP (Grände, 2006), but it does not have a strong evidence base and is not universally accepted. Rather than a single threshold for CPP, it is likely that ‘optimal’ values can be identified at the individual level using multimodality neuromonitoring (Kirkman and Smith, 2014).

Multimodality neuromonitoring Monitoring multiple physiological variables is more accurate in the prediction of cerebral hypoxia-​ischaemia than measurement of one variable alone (Bouzat et al., 2015) (Table 42.3). In particular, ICP monitoring should be regarded as one part of a multimodality neuromonitoring strategy rather than a monitoring modality in isolation (Kirkman and Smith, 2012).

Cerebrovascular reactivity Cerebrovascular reactivity, a key component of cerebral autoregulation (CA), is often impaired after TBI, rendering the brain more susceptible to secondary ischaemic insults. Monitoring and correlation of spontaneous slow waves in ABP and ICP allows calculation of a pressure reactivity index (PRx) which can be used as a continuous bedside assessment of CA. A negative value for PRx, when ABP is inversely correlated with ICP, indicates normal CA, and a positive value a non-​reactive cerebrovascular circulation. PRx may be used to guide therapy and identify optimal CPP after TBI (Fig. 42.3) (Aries et al., 2012). An oxygen reactivity index (ORx) has also been described as the moving correlation between brain tissue oxygenation (PbtO2) and ABP, as have non-​invasive alternatives using the correlation between ABP and TCD-​derived blood flow velocity and several near infrared spectroscopy (NIRS)-​derived variables (Kirkman and Smith, 2012).

Cerebral blood flow monitoring TCD is a non-​invasive technique allowing real-​time assessment of cerebral haemodynamics at the bedside. Although it measures relative changes in CBF rather than absolute values, it can be used to detect impaired CBF and assess cerebrovascular reactivity after TBI. Quantitative measurement of absolute regional CBF is possible

503

504

Section 10  Neurotrauma and intensive care

Table 42.3  Neuromonitoring techniques used in traumatic brain injury Technique

Monitored variable(s)

Indications in TBI

Intracranial pressure Intraparenchymal or • ICP • To allow CPP optimization subdural microsensor • CPP through derivation of PRx and • Autoregulatory indices ORx Ventricular catheter

Cerebral blood flow Transcranial Doppler ultrasonography

• As above

• As above, plus: • To allow therapeutic drainage of CSF

• Blood flow velocity • May detect cerebral • Pulsatility index hypoperfusion and facilitate early • Autoregulatory indices goal-​directed therapy

Cerebral oxygenation Jugular venous • Jugular venous oxygen • To evaluate the adequacy of oximetry saturation cerebral perfusion and oxygen • Arterio-​venous oxygen delivery content difference • To supplement ICP/​CPP-​guided therapy • To assist titration of medical and surgical therapies to guide ICP/​ CPP therapy

Disadvantages

• Relatively easy insertion • Low procedural complication rate • Low infection risk

• In vivo calibration not possible • Measures localized pressure • Small zero-​drift with time

• Measures global ICP • Therapeutic CSF drainage • In vivo calibration possible

• More difficult insertion • Risk of procedure-​related haemorrhage • Risk of catheter-​related ventriculitis

• Noninvasive • Can be used on an intermittent or continuous basis • Good temporal resolution

• Measures relative (not absolute) CBF • Operator dependent • Failure rate in 5–​10% of patients (absent acoustic window) • Little evidence base in TBI

• Global assessment of balance • Non-​quantitative assessment of between CBF and metabolism cerebral perfusion • Insensitive to regional ischaemia • Risk of extracranial contamination of samples

Brain tissue pO2

• Brain tissue oxygen partial pressure • Oxygen reactivity

Near infrared spectroscopy (cerebral oximetry)

• Regional cerebral • As for jugular venous oxygen saturation oximetry, plus: • Autoregulatory indices • Detection of intracranial haematoma

• Noninvasive • Real time • Multisite measurement

• rScO2-​derived ‘ischaemic’ threshold not defined • Readings affected by the presence of intracranial haematoma and ‘contamination’ of signals by extracranial tissue • Little evidence base in TBI

Cerebral microdialysis

• Glucose • Lactate, pyruvate, and LPR • Glycerol • Glutamate • Multiple biomarkers for research purposes

• Assessment of cerebral glucose metabolism • Detection of hypoxia/​ ischaemia • Assessment of non-​ischaemic causes of cellular bioenergetic dysfunction

• Focal measure • Thresholds for abnormality unclear • Not continuous • Labour-​intensive

• Seizures • To detect convulsive and non-​ • Diagnosis-​specific EEG convulsive seizures patterns • To diagnose characteristic • Some evidence that EEG patterns to facilitate detection of SDs might prognostication be possible

• Noninvasive • Detection of non-​convulsive seizures • Correlates with cerebral ischaemic and metabolic changes

• Skilled interpretation required • Affected by anaesthetic and sedative agents

• Cortical SDs

• Currently the only method to identify SDs accurately

• Highly invasive • No evidence that treatment of SDs improves outcome • Little evidence base in TBI and thus currently remains a research tool

Electrophysiology EEG

ECoG

• As for jugular venous oximetry, plus: • To allow CPP optimization through derivation of ORx

Advantages

• Regional assessment of • Minimally invasive balance between CBF and • Measures oxygenation within a small metabolism region of interest • Continuous • One hour ‘run-​in’ period limits • Ischaemic ‘thresholds’ defined intraoperative applications

• To detect cerebral ischaemia, hypoxia, cellular energy failure, and glucose deprivation • To predict the development of secondary injury before it is detectable clinically or by other monitoring modalities • To optimize CPP • To assist titration of medical therapies such as systemic glucose control

• To detect spreading cortical depolarizations

using a thermal diffusion flowmetry probe, but clinical data using this technique are limited.

assess the balance between cerebral oxygen delivery and utilization, and the adequacy of cerebral perfusion.

Cerebral oxygenation monitoring

Jugular venous oximetry

Maintenance of ICP and CPP within established thresholds for normality does not guarantee against brain tissue hypoxia (Oddo et al., 2011). Cerebral oxygenation monitoring is therefore often used to

Jugular venous oxygen saturation (SvjO2) monitoring was the first bedside monitor of cerebral oxygenation, and formed the basis of our understanding of cerebral oxygenation changes after brain injury.

CHAPTER 42  Intensive care management of head injury

– cpp .. ICP mmHg

100 50 0 – prx 1 0.5 0 –0.5 –1

1/5 21:30

1/5 22:00

1/5 22:30

1/5 23:00

1/5 23:30

2/5 00:00

2/5 00:30

2/5 01:

PRX

0.05 0

–0.05

Time (%)

= 120

20 10 0

= 120

CPP (mmHg)

Fig. 42.3  Screenshot from ICM+ software panel (Cambridge Enterprise, Cambridge, UK) with 4 h trend charts. Determination of the PRx can guide therapy and allow identify optimal CPP at the individual level (CPPopt). First chart: CPP and ICP; second chart: PRx; third chart: PRx/​CPP plot for evaluating CPPopt; and fourth chart: percentage of the 4-​h time period spent within a given CPP interval. The ICM+® brain monitoring software developed at the University of Cambridge was used to generate this image. Reproduced with permission from Celeste Dias, Maria João Silva, Eduarda Pereira et al., Optimal Cerebral Perfusion Pressure Management at Bedside: A Single-​Center Pilot Study, Neurocritical Care, Volume 23, Issue 1, pp. 92–​102, Copyright © 2015 Springer Nature.

Maintenance of SvjO2 more than 50% after TBI is recommended in BTF guidelines (Carney et al., 2017), but no interventional trials have confirmed a direct benefit of SvjO2-​directed therapy on outcome and in any case SvjO2 monitoring is being superseded by other modalities. Brain tissue oxygen partial pressure PbtO2 monitoring is considered the ‘gold standard’ bedside monitor of cerebral oxygenation, and recommended after severe TBI (Le Roux et al., 2014). It is a complex and dynamic variable resulting from interaction of all factors affecting cerebral oxygen delivery and demand (oxygen metabolism), the relative proportion of arterial or venous vessels in the region of interest, and tissue oxygen diffusion gradients. It is therefore a biomarker of cellular function rather than simply a monitor of hypoxia/​ischaemia. PbtO2 is a regional monitor and the probe is typically placed in tissue immediately surrounding a hematoma or contusion to monitor ‘at risk’ brain regions. Normal brain PbtO2 is reported to lie between 2.7 and 4.7  kPa (20–​35  mmHg), and the ischaemic threshold usually defined as 1.33–​2.0 kPa (10–​15 mmHg). However, PbtO2 values are best considered within a range rather than as a precise threshold and ischaemia defined by both duration and depth of hypoxia. Guidelines recommend interventions to maintain PbtO2 more than 2.7  kPa (>20 mmHg) (Le Roux et al., 2014). Observational studies suggest a potential benefit of supplementing ICP/​ CPP-​ guided management with PbtO2-​directed

therapy after TBI (Nangunoori et al., 2012). In addition to systemic blood pressure, PbtO2 is affected by several other factors including PaO2, PaCO2, and haemoglobin concentration, and which intervention or combination of interventions should be used to reverse brain tissue hypoxia is contentious. In fact, it appears that the responsiveness of brain tissue hypoxia to a given intervention is the prognostic factor, with reversal of hypoxia being associated with reduced mortality (Nangunoori et al., 2012). RCT data are urgently needed to confirm any benefits of PbtO2 monitoring, and the results of the Brain Tissue Oxygen Monitoring in Traumatic Brain Injury (BOOST2) study are awaited (https://​clinicaltrials.gov/​ct2/​show/​ NCT00974259). Near infrared spectroscopy Near infrared spectroscopy-​based cerebral oximetry provides continuous and non-​invasive monitoring of regional cerebral oxygen saturation (rScO2), with high temporal and spatial resolution, and the possibility simultaneous measurement over multiple regions of interest. The ‘normal’ range of rScO2 is reported as 60–​75% but there is substantial intra-​and interindividual variability, and rScO2 values are best used as a trend monitor. There has been limited investigation of the utility of NIRS after TBI where its application is confounded by the optical complexity of the injured brain (Ghosh et al., 2012). NIRS-​derived variables have been investigated as a non-​ invasive assessment of cerebral autoregulatory status, but its role in the broader monitoring of TBI is undefined.

505

506

Section 10  Neurotrauma and intensive care

Cerebral microdialysis Cerebral microdialysis (CMD) allows bedside analysis of biochemical substances in brain tissue extracellular fluid. The catheter is usually placed in ‘at risk’ tissue so that biochemical changes in the area of brain most vulnerable to secondary insults can be monitored. Glucose, lactate, pyruvate, lactate:pyruvate ratio (LPR), glycerol and glutamate are the variables most commonly measured clinically, and each is a marker of a particular cellular process associated with glucose metabolism, hypoxia/​ischaemia, or cellular energy failure. CMD measures the supply of substrate and its cellular metabolism, so it is a monitor not only of cerebral ischaemia but also of non-​ischaemic causes of cellular energy dysfunction. It may identify cerebral compromise at the cellular level before it is detectable clinically or by other monitored variables. Elevated lactate:pyruvate ratio (LPR) more than 40 combined with low brain glucose (25 ml) and/​or radiological signs of raised intracranial pressure (ICP) underwent emergency cranial surgery (Compagnone et al., 2007). Acute subdural haematomas were the most frequent evacuated mass lesions, contributing to approximately two-​third of all emergency cases, followed by extradural haematoma and contusions/​intracerebral haematomas.

Intracranial pressure monitoring Severe TBI patients in need of neurointensive care in the acute phase are often sedated and paralysed to facilitate endotracheal intubation and for control of intracranial pressure. This limits the utility of neurological examinations, and therefore computed tomography (CT) scanning and monitoring of intracranial pressure (ICP), remain the primary methods to guide treatment in a neuro-​intensive care setting (Le Roux et  al., 2014; Brain Trauma Foundation, 2016). Control of ICP remains the central tenet of management of moderate-​ severe TBI to maintain adequate

perfusion and oxygenation of the brain and to avoid or mitigate secondary insults (Valadka and Robertson, 2007; Brain Trauma Foundation, 2016). The Brain Trauma Foundation (BTF) guidelines recommends management of severe TBI patients using ICP monitoring to reduce in-​hospital and 2-​week postinjury mortality (Brain Trauma Foundation, 2016). ICP monitoring is routinely used in the developed world, and observational studies have demonstrated repeatedly that ICP crises can lead to adverse outcomes (Forsyth et al., 2015). This is discussed further in Chapter 42 (Intensive Care Management of Head Injury). Ultimately, if left untreated, the pressure within the intracranial compartment can rise above a critical threshold, causing damage to neuronal structures, and eventually herniation, and brain death. It important to be aware of intracranial pressure gradients within the brain parenchyma, especially in patients with space-​occupying lesions. This group is prone to develop ICP gradients, since the supporting dural folds, namely the falx cerebri, and the tentorium cerebelli divide the intracranial cavity into compartments that normally protect against excessive movement. Pressure differences between the cerebral hemispheres have been described greater than 10 mmHg in the context of acute subdural haematoma (Chambers et al., 1998). This has underpinned the argument for intraventricular probes for measurement of ICP as, with a patent ventricular system, the pressure can equilibrate between the CSF spaces to a degree (Brain Trauma Foundation, 2016). In practice, the risks associated with prolonged ventricular drainage (both infectious and procedural) limit this approach in many units in favour of parenchymal probes. Advanced cerebral monitoring combines ICP monitoring with other cerebral monitoring techniques, including cerebral autoregulation monitoring, microdialysis monitoring of brain chemistry, brain tissue oxygen (PbrO2) monitoring, and jugular bulb monitoring of arteriovenous oxygen content difference (AVDO2) (Valadka and Robertson, 2007; Le Roux et al., 2014; Brain Trauma Foundation, 2016; Makarenko et  al., 2016). These monitoring methods allow for individualized management of secondary cerebral insults by assessing brain metabolic needs and the response to therapeutic intervention. Some cranial access devices allow for multiple catheters and sensors to be transmitted into the brain parenchyma to facilitate continuous multimodality monitoring at the bedside (Hutchinson et al., 2000).

510

Section 10  Neurotrauma and intensive care

Ventriculostomy The role of CSF drainage and the optimal draining techniques (continues vs. intermittent) in TBI patients has not been firmly established (Nwachuku et  al., 2014; Brain Trauma Foundation, 2016). An external ventricular drain (EVD), or ventriculostomy, is a rapid method of controlling ICP by reducing the intracranial CSF volume, even in the absence of hydrocephalus (Kerr et al., 2001; Timofeev et al., 2009; Brain Trauma Foundation, 2016). Its use has been described as a secondary ICP-​lowering manoeuvre as part of protocol-​ driven management of elevated ICP. Physiological improvements have been reported following CSF drainage, however in a significant proportion of patient’s ICP fails to be controlled if not used alongside other ICP-​lowering treatments (Timofeev et  al., 2009). Ventriculostomies are also associated with complications, include the risk of EVD-​related infections, haemorrhages during insertions, and subsequent shunt-​dependency. An EVD connected to a pressure transducer can measure ICP in a closed position (see section on ICP monitoring), and drain CSF when opened. There is no consensus on the optimal drainage technique in TBI patients (Nwachuku et  al., 2014; Brain Trauma Foundation, 2016). Practice patterns seem to vary significantly, and shown to be influenced by physician and departmental preferences and resources. Management of CSF drainage in TBI patients also appears to be related to the patient’s age (Shore et al., 2004; Nwachuku et al., 2014). Continuous CSF drainage has been more commonly described in the paediatric TBI population. It is important to note that while intermitted drainage can allow for real-​time visualization of ICP, the patient can be exposed to detrimental elevations of ICP between drainage periods. Continuous drainage allows for more stable ICP control, however, this risks overdrainage with subsequent collapse of the ventricles. In this circumstance, the EVD can no longer be reliably employed as a method of gauging ICP as it relies on a continuous column of fluid between the site of the transducer and the intraventricular cavity.

Craniotomy for mass lesions Surgical intervention is aimed at mitigating injuries to the brain caused by (expanding) haematomas and their sequelae (Bullock et al., 2006a; Bullock et al., 2006b; Bullock et al., 2006c). An initial assessment of neurologic deficits, pupil abnormalities, degree of midline shift, hematoma volume, and the presence/​severity of associated trauma is required to determine the necessity for emergent cranial surgery. For neurosurgeons one of the most difficult decisions to make is whether or not moderate-​sized mass lesions should be evacuated or observed. As surgical interventions might not always be necessary, there is a risk with conservative management of neurological deterioration with possible secondary insults to the brain that may negatively impact the patient’s outcome. Current guidelines and recommendations are available, but are limited to expert consensus with limited quality of evidence (Bullock et  al., 2006a; Bullock et al., 2006b; Bullock et al., 2006c). It is worth emphasizing that although guidelines necessarily incorporate specific metrics that guide surgery, the whole clinical picture incorporating the clinical state (GCS, pupillary function) of the patient, the trajectory

of deterioration, other causes of neurological compromise (such as metabolic, drug induced, hypoxia or hypotension), associated injuries, and comorbidities all influence the speed and manner of intervention (Bullock et  al., 2006a; Bullock et  al., 2006b; Bullock et al., 2006c).

Extradural haematomas Extra (or epi-​) dural hematomas (EDH) usually develop in young adults following traffic-​related accidents, falls, and assaults (Bullock et  al., 2006b). In TBI patients the incidence of surgical and non-​ surgical EDH cases has been estimated to be between 2.7 to 4% (Bullock et al., 2006b) EDH are thought to result from a direct blow to head and are usually found on the same side impacted by the blow. Typically, the source of bleeding is arterial, following a fracture in the region of the pterion with subsequent tearing of the middle meningeal artery and hematoma formation in the middle cranial fossa. Extradural hematomas may occur in any anatomical location including in the frontal, occipital, and vertex regions and can be associated with the anterior ethmoidal artery, transverse or sigmoid sinuses, and superior sagittal sinus, respectively. EDHs originating from venous sources are thought to expand more slowly compared to their arterial counterparts (Nalbach et  al., 2012). EDH specific mortality has been described to be around 10% in adult patients (Bullock et al., 2006b). The role of surgery is to prevent irreversible brain injury or death caused by hematoma expansion, raised ICP, and herniation of the brain. Patients presenting with (progressive) focal neurologic signs or symptoms and/​or hematoma growth, have to be considered as an emergency case. Evidence and expert-​based recommendations for evacuations of EDH recommend surgery for all adult patients with a hematoma volume more than 30 cm3 (>30 ml) regardless of the GCS score, and comatose patients (GCS 20 ml) with either midline shift of more than 5 mm and/​or cisternal compression on CT scan (Bullock et al., 2006a). Patients with diffuse injuries developing medically refractory post-​traumatic cerebral oedema and intracranial hypertension may be considered for a bifrontal DC (Bullock et al., 2006a). DCs may also be considered for patients with refractory intracranial hypertension and diffuse injuries with clinical and radiographic evidence for transtentorial herniation. Evacuation of a traumatic ICH in the posterior fossa is recommended when there is evidence of neurologic dysfunction/​deterioration and significant mass effect on the basal cisterns, fourth ventricle, or signs of obstructive hydrocephalus (Bullock et al., 2006d). Intensive monitoring and serial imaging are appropriate for patients with no focal neurologic deficits, and non-​significant mass effect on imaging.

Decompressive craniectomy Management of refractory raised ICP following severe TBI consist of medical and surgical treatments (Stocchetti et al., 2008; American College of Surgeons, 2015). DC is generally considered a surgical treatment for patients with diffuse brain swelling or expanding contusions/​haematomas refractory to medical treatment and impending herniation (Bohman and Schuster, 2013; Kolias et al., 2013). In recent years, the role of DC has been discussed as a primary treatment in the acute phase, leaving out the bone flap after evacuation of a mass lesion, or as a second-​or third-​tier therapeutic measure for diffuse brain injury and oedema, commonly named secondary or protocol-​driven DC. The expansion of a swollen brain outside the skull, can potentially lead to a reduction in ICP and risk of herniation. The physiological improvements described in severe TBI patients after DC, include improvement in brain tissue oxygenation, cerebral perfusion, and neurochemistry (Yamakami and Yamaura, 1993; Stiefel et al., 2004; Ho et al., 2008; Jaeger et al., 2010; Bor-​Seng-​ Shu et al., 2012; Lazaridis and Czosnyka, 2012). The risk of complications should also be considered as early or delayed complications can occur after DC (Kolias et al., 2013). Expansion of (contralateral) mass lesions, wound infections, and healing problems, subdural or subgaleal collections, hydrocephalus, syndrome of the trephined, and complications related to the subsequent cranioplasty have been recognized as DC-​related complications (Flint et  al., 2008; Stiver, 2009; Nalbach et al., 2012). As the risk of severe disability and death in severe TBI remains relatively high, several trials have explored the use of DC to improve patient outcomes ( Protocol 14PRT/​6944; Hutchinson et al., 2006; Cooper et al., 2011). However, defining the indications, timing, techniques, and optimal outcome measures for DC has proven to be difficult, and good quality evidence linking efficacy to outcome is lacking (Kolias et al., 2013).

Decompressive craniectomy methods Decompressive craniectomy is an umbrella term for a group of procedures in which part of the skull is removed. In severe TBI the most frequently described DC procedures in adults are bifrontal DC and unilateral frontotemporoparietal craniectomy, also termed hemi(spheric)-​craniectomy or unilateral DC (Kjellberg and Prieto, 1971; Guerra et al., 1999; Kolias et al., 2013). For unilateral pathologies

511

512

Section 10  Neurotrauma and intensive care

(a)

(b)

(c)

Fig. 43.1  Unilateral decompressive craniectomy. (A) The dotted line represents the usual skin incision made during unilateral decompressive craniectomy. To preserve adequate vascular supply, the length of the incision (distance B) should not exceed its width (distance A). (B) A myocutaneous flap is reflected. The dotted line represents the usual extent of the craniectomy. (C) The dotted line on the dura mater represents our preferred method for opening the dura. The dura is opened in a C-​shaped fashion with its base along the sphenoid ridge. The dural incision is kept 5–​10 mm away from the craniectomy edges to minimize the risk of injury to the protruding brain. Reproduced with permission from I. Timofeev, T. Santarius, A.G. Kolias et al., Decompressive craniectomy—​operative technique and perioperative care, Advances and Technical Standards in Neurosurgery, Volume 38, pp. 115–​36, Copyright © 2012 Springer Nature.

with midline shift and (potential) swelling (e.g. aSDH with parenchymal injuries), a hemi-​craniectomy can be useful (Fig. 43.1). Evidence and expert-​based recommendations for adequate hemi-​ craniectomies suggest that the bone flap should be large with a minimum anteroposterior diameter of 11–​12 cm (Li et al., 2012; Tagliaferri et al., 2012) in order to achieve an adequate reduction of ICP and also reduce the risk of transcalvarial herniation that is associated with parenchymal injuries and cortical venous occlusion at the bone edge (von Holst et al., 2012; Li et al., 2013). Bifrontal DC is a treatment option for diffuse (bihemispheric) injuries with medically refractory intracranial hypertension (Fig. 43.2) (Kolias et al., 2013). A bifrontal DC extends from the floor of the anterior cranial fossa to the coronal suture posteriorly and to the temporal floor bilaterally. A widely opened dura mater is required to allow the brain to sufficiently expand. Different techniques have been described for the dura (left open with onlay of (a)

(b)

haemostatic material, pericranium, or temporalis fascia, or closure with dural expansion grafts) (Whitfield et  al., 2001; Timofeev and Hutchinson, 2006; Guresir et al., 2011) and sagittal sinus sectioning or sparing (Bohman and Schuster, 2013). Studies have also described bilateral hemi-​craniectomies as an approach for patients with diffuse injuries, although an improvement over the bifrontal DC approach has not been investigated (Guerra et al., 1999). For patients with temporal lesions or oedema causing brainstem compression, extension of the DC to the floor of the middle cranial fossa is essential.

Evidence base for DC in TBI The DECRA trial failed to find an improvement in functional outcome by performing early bifrontal DC over medical management for patients with diffuse TBI (Cooper et al., 2011). The study showed that patients treated with DC had shorter duration of ventilation (c)

Fig. 43.2  Bifrontal decompressive craniectomy. (A) The dotted line represents the usual skin incision for bifrontal decompressive craniectomy, which should be kept behind the hairline. (B) A bicoronal myocutaneous flap is reflected anteriorly. The dotted line on the skull represents the usual extent of the craniectomy. Subtemporal decompression is optional. (C) The bone flap has been removed. The dotted line on the dura mater represents our preferred method for opening the dura. The dura is opened on either side of the midline in a C-​shaped fashion with its base along the superior sagittal sinus. Division of the superior sagittal sinus anteriorly and of the falx (red line) is optional. Reproduced with permission from I. Timofeev, T. Santarius, A.G. Kolias et al., Decompressive craniectomy—​operative technique and perioperative care, Advances and Technical Standards in Neurosurgery, Volume 38, pp. 115–​36, Copyright © 2012 Springer Nature.

CHAPTER 43  Surgical management of head injury

and length of stay in the intensive care unit. The RESCUEicp trial aimed to examine the clinical and cost effectiveness of secondary DC (unilateral or bifrontal DC) for severe TBI patients with refractory intracranial hypertension as a last-​tier therapy. 408 patients with raised and refractory ICP (threshold 25 mmHg >1–​12 hours despite standard medical therapy) were randomised to ongoing medical therapy (plus optional barbituartes) or secondary decompressive craniectomy (Hutchinson et  al., 2006). The results showed that decompressive craniectomy resulted in a marked reduction in mortality, increase in vegetative state, increase in lower (dependent) and upper (independent at home) with similar rates of moderate disability and good recovery. Outcome improved between 6 and 12 months with a significant proportion of patients in the surgical arm being upper severe disability or better. In contrast to the aforementioned trials, the RESCUE-​ASDH trial is a randomised trial comparing primary unilateral DC to craniotomy (bone flap out versus bone flap replaced) for patients with aSDH (Protocol 14PRT/​ 6944). Information from these studies will define the role of secondary and primary DC in future TBI treatment guidelines. On the basis of the current available evidence, neurosurgeons and neuro-​ intensivists must weigh the potential risks and benefits faced by their individual patients when deciding to perform a DC.

Surgical management of depressed skull fractures Skull fractures most frequently involve the parietal bone, followed by the temporal, occipital, and frontal bones (Cooper and Golfinos, 2000). TBI patients usually present with linear fractures, and less frequently with depressed and skull base fractures. The force of the trauma to the skull required to cause fractures is significant, therefore patients are at significant risk of underlying brain injury. Evidence and expert-​based recommendations for skull fractures recommend elevation and washout for patients with open skull fractures depressed more than the thickness of the cranium or more than 5 mm below the adjacent inner table (Bullock et al., 2006e). The rationale is reducing the risk of infection for these cases with early surgery, especially in the presence of dural tears, pneumocephalus, frontal sinus involvement, or contaminated wounds. Emergent surgery is also indicated for fractures with an underlying (expanding) hematoma. Elevation of the fracture will also improve cosmesis for cases with significant displacement of the bone. Reconstruction can usually be achieved by using the bone fragments; if this is not feasible, implants can be used to cover the skull defect (Marbacher et al., 2008). Antibiotics are usually administered to patients with open skull fractures; currently routine prophylaxis for all skull fractures is not supported by the available evidence (Ali and Ghosh, 2002; Bullock et al., 2006e; Ratilal et al., 2015).

Management of skull base fractures Skull base fractures can involve the anterior, middle, and posterior cranial fossae, or a combination thereof. This section will focus on fractures of the petrous temporal bone and frontal sinus fractures.

Petrous temporal bone fractures are traditionally classified according to the orientation of the fracture line in relation to the long axis of the petrous temporal bone (i.e. line from petrous base to apex). Longitudinal fractures are parallel to the long axis of the petrous bone and their most frequent complications are ossicular injury, tympanic membrane rupture, and conductive hearing loss. Transverse fractures are perpendicular to the long axis of the petrous bone and they are thought to carry a higher risk of sensorineural hearing loss and facial nerve injury compared to longitudinal ones (Zayas et al., 2011). In reality, many fractures are mixed and therefore, a new classification system based on whether the otic capsule is involved or not may be clinically more meaningful (Dahiya et al., 1999). The so-​called otic capsule–​violating fractures involve the labyrinth (i.e. cochlea, vestibule, or semicircular canals) and patients were found to be twice as likely to develop facial paralysis, four times more likely to have a CSF leak, and seven times more likely to develop profound hearing loss (Dahiya et al., 1999). The onset of palsy (immediate versus delayed) and the degree of palsy are factors that are typically taken into account when managing patients with petrous fractures and facial nerve palsy. However, in reality, it is often difficult to ascertain the timing of onset and the severity of paresis, as many patients have impaired consciousness, are sedated and ventilated, or are undergoing treatment for life-​threatening injuries (Nash et  al., 2010). Steroids are often used for patients with facial nerve palsy but a recent systematic review identified only low-​quality evidence with lack of controlled studies (Nash et al., 2010). Surgical decompression of the nerve in its canal is usually reserved for patients with immediate onset of serious facial nerve palsy that does not improve with steroids. The timing of surgery remains controversial (Nash et al., 2010). In terms of natural history, of 189 patients managed with observation, who were included in a recent systematic review, 66% had an outcome equivalent to House-​Brackmann (HB) I, one-​quarter had HB II-​V, and only two patients had an HB VI score (Nash et al., 2010). The frontal sinuses are involved in up to 15% of facial fractures (Echo et al., 2010). Overall, recent years have seen a shift towards conservative management, especially with the advent of endoscopic techniques. Fractures involving the anterior wall do not usually require intervention, unless they are open or depressed. Fractures involving the posterior wall carry a higher risk of CSF leak. The indications for intervention for posterior wall fractures are controversial but in general persistent CSF rhinorrhoea or significant displacement of the posterior wall are accepted indications for intervention (Echo et al., 2010). This typically requires a bicoronal incision, repair of any dural defects if present, and cranialization of the sinus. Cranialization of the sinus aims to reduce the risk of mucocoele formation and intracranial infection; it involves removal of its posterior wall, stripping of the mucosa, and packing the frontonasal ducts. A pericranial flap positioned across the anterior cranial fossa floor, reinforced with a dural sealant, can help prevent a CSF leak. The LeFort type of fractures are usually managed by the maxillofacial surgeons; LeFort I fractures involve the maxilla and pterygoid plates; LeFort II have a pyramidal shape and involve the orbital floor, medial orbital wall, and nasofrontal suture; LeFort III involves the zygomatic arches, frontozygomatic and nasofrontal sutures, and orbital floors, resulting in craniofacial dislocation.

513

514

Section 10  Neurotrauma and intensive care

Surgical management of penetrating head trauma All traumatic brain injuries that are not caused by a blunt mechanism are defined as penetrating brain injury (PBI) (Esposito and Walker, 2009). The incidence of PBI is lower than closed TBI, but PBI has been associated with more adverse outcomes. Most PBIs are caused by gunshot wounds to the head (GSWH), which are the typically lethal, as more than 90% of GSWH have a fatal outcome (Maas et al., 2008). The incidence of PBI is increasing, especially in the United States, due to the use of firearms in violence-​related injuries. In addition, worldwide, armed conflicts are causing more brain injuries (No authors, 2001; Maas et al., 2008). Since the majority of PBIs are caused by missiles or projectiles, and understanding of ballistics is required. Primary injury from GSWH is determined by the ballistic properties (kinetic energy, mass, velocity, shape, and so on.) of the projectile and any secondary projectiles, such as bone or metallic fragments (Ordog et al., 1984). Projectile and possible secondary metal or bone fragments travel through brain parenchyma creating a permanent track of injury. Higher velocity projectiles will cause temporary cavitation, which is a kinetic energy and velocity-​related phenomenon (Aarabi, et al., 2001). This results from the transmission of the kinetic energy of the projectile to the surrounding tissue as the projectile slows down. Subsequently, this temporary cavity collapses and expands in wave-​like pattern causing further shearing forces to surrounding tissues resulting in haemorrhages and neuronal injuries (Ordog et al., 1984; No authors, 2001). A non-​contrast CT of the head is the recommended modality in the radiological evaluation of PBI cases, since it is quick and provides identification of bone and metallic fragments, characterization of the projectile trajectory, evaluation of the extent of brain injury, detection of intracranial haemorrhages, and possible mass effect (Aarabi, et al., 2001; Tsuei et al., 2005). A formal digital subtraction angiogram is recommended, in particular when the projectile trajectory is near the Sylvian fissure, the supraclinoid carotid artery, the vertebrobasilar vessels, the cavernous sinus region, the major dural venous sinuses, or if a delayed hematoma or SAH develops (No authors, 2001; Offiah and Twigg, 2009). The common vascular complications following PBI include traumatic intracranial aneurysms, arteriovenous fistulas (AVFs), SAH, and vasospasm (No authors, 2001). Surgical management of PBI entails a superficial debridement and watertight dural closure as a standard of care, but in patients with small entry wounds, simple wound closure may be considered (Maas et al., 2008). In the case of significant mass effect, necrotic brain tissue and bone fragments should be removed, but routine bone or projectile removal in the eloquent areas is not recommended (No authors, 2001). Additionally, any significant intracerebral haemorrhages with significant mass effect should be evacuated. PBI patients are at high risk of CNS infection since contaminated foreign bodies are driven into the brain tissue along the projectile track, hence prophylactic broad-​spectrum antibiotics are recommended, and should be started as soon as possible (No authors, 2001; Maas et al., 2008).

Surgical management of CSF leaks In TBI patients it is estimated that CSF leaks occur in approximately 2% of patients (Mendizabal et al., 1992; Brodie and Thompson, 1997;

Friedman et al., 2001). Fractures carrying a high risk of CSF leaks are those involving the frontal or ethmoidal sinuses, and the petrous temporal bone (Mendizabal et al., 1992; Ommaya, 1996). The majority of CSF leaks are self-​limiting and resolve spontaneously within days (Chandler, 1983; McGuirt and Stool, 1995; Brodie and Thompson, 1997). Surgical interventions are aimed at reducing the symptoms and risk of infection in cases with persistent CSF leaks or fistulae. Studies have reported infection rates between 7% and 30% for TBI patients with CSF leaks, with each day of leakage increasing the risk of ascending intracranial infection (Leech and Paterson, 1973; Spetzler and Wilson, 1978; Brodie and Thompson, 1997; Savva et al., 2003; Nyquist et al., 2013; Mathias et al., 2016). Lumbar drainage has been recognized as a treatment option for TBI patient with CSF leaks, in order to relieve CSF pressure aiding spontaneous and surgical repairs, while also facilitating a route to administer intrathecal fluorescein for diagnostic purposes (Shapiro and Scully, 1992; Mathias et al., 2016). However, lumbar drainage in the presence of a patent CSF fistula risks entraining intracranial air with a consequent risk of intracranial infection. The relative importance of antibiotic prophylaxis as well as the selection of the high-​risk candidates for surgical intervention remain to be defined (Rimmer et al., 2014). However, the Centre for Disease Control and Prevention recommends either the pneumococcal conjugate vaccine (PCV13) or the pneumococcal polysaccharide vaccine (PPSV23) for patients with CSF leaks (CDC, 2015b). In the United Kingdom, the National Institute of Clinical Excellence recommends pneumococcal vaccination for all skull base fractures.

Investigation and management of carotid or vertebral dissection Post-​traumatic dissection of the internal carotid (ICA) or vertebral artery (VA) is most prevalent in young TBI patients (Rubinstein et  al., 2005). Mechanisms include motor vehicle accident, falls, hangings or sport accidents, and may lead to a stroke in an acute or delayed fashion from acute vessel occlusion or thromboembolism (Geddes et al., 2016). The mechanism of injury can be direct compression of the arteries against bony structures owing to flexion, extension and rotational head movements or acceleration/​deceleration that results in shearing forces to the vessel endothelium. The VA is commonly dissected at the C1–​C2 level by rotational forces or directly due to a fracture or dislocation of facet joints or transverse processes, whereas the ICA can be compressed between the spine and mandible (Blacker and Wijdicks, 2004). Clinically, dissections can present as TIA, stroke, or SAH. There is often a silent period between the injury and onset of neurological symptoms. Patients with polytrauma with impaired consciousness due to the head injury itself or the use of sedatives, can make the clinical diagnosis difficult (Blacker and Wijdicks, 2004). Prognosis is variable, as dissections can vary from being asymptomatic to causing profound neurological deficits and death (Mohan, 2014). Dissections are initiated with a tear on the innermost intimal layer or the media layer of the blood vessel wall. Extravasation of arterial blood under pressure can extend the dissection between intima and tunica media creating luminal narrowing or occlusion. A subintimal dissection may result in an intramural hematoma with consequent luminal stenosis. Dissection within the intracranial cavity may also

CHAPTER 43  Surgical management of head injury

lead to aneurysmal dilatation and haemorrhage into subarachnoid space. Rupture of an artery with subsequent encapsulation of the extravascular hematoma may or may not produce luminal narrowing (Yamaura, 1994). Distal embolization occurs due to platelet aggregation stimulated by exposed surfaces or dislodged thrombus. It may also cause reduced distal flow due to occlusion of lumen due to thrombosis or mural haematoma (Mohan, 2014). Prompt recognition and diagnosis is paramount. Cerebral ischaemia related to ICA most commonly affects the middle cerebral artery territory, whereas VA dissections affect the brainstem, cerebellum, and occipital lobes. Infarctions or haemorrhages can be visualized with CT imaging, however CT angiogram (CTA) or catheter angiography alone can demonstrate luminal stenosis, fusiform dilatations, occlusions, intimal flaps, proximal beading, double lumen signs, or a kinked appearance of vessels (Kitanaka et al., 1994; Eastman et al., 2006; Geddes et al., 2016). Magnetic resonance imaging angiogram is generally not considered as accurate as CTA or catheter angiography. Intracranial lesions with SAH can be treated endovascularly or microsurgically (Halbach et al., 1993; Pham et al., 2011), however, in modern practice surgical management is rarely warranted. Dissecting aneurysms may be fusiform and cover a long segment that precludes microsurgical clipping or endovascular stenting, and in this circumstance may be candidates for occlusion of the VA proximal to the basilar artery. Balloon test occlusion of the dominant VA is recommended since some patients might not tolerate total occlusion (Mohan, 2014). Other surgical techniques described include a combination of VA clipping with vascular bypass, resection accompanied with autogenous interposition vein graft, usage of fusiform-​ type clipping, and wrapping (Arimura and Iihara, 2016). Extracranial dissections are managed medically, and this includes antiplatelet therapy or anticoagulation for a total of 6 months (CADISS trial investigators, 2015).

Surgical management of post-​traumatic hydrocephalus Post-​traumatic hydrocephalus (PTH) is a complication of TBI and studies have reported clinical improvement after permanent CSF diversion (Guyot and Michael, 2000). Patients who suffered TBI and under active follow-​up are usually monitored for PTH, which can present as worsening neurological status or lack of improvement associated with pressure related headaches or the normal pressure hydrocephalus syndrome. Late CSF leak is also suspicious for PTH. It is important to recognize and treat PTH, since it could both impact morbidity and mortality if left untreated (Paoletti et al., 1983; Tribl and Oder, 2000; Low et al., 2013). The incidence of PTH has been reported to range between 0.7% and 51.4% (Bontke and Boake, 1991; Guyot and Michael, 2000; De Bonis et al., 2010). However, it is often difficult to determine whether ventriculomegaly observed post-​TBI is related to atrophy or hydrocephalus; computerized CSF infusion studies have been reported to be useful in distinguishing between the two different processes (Marmarou et al., 1996; Czosnyka et al., 2005). Selection of patient benefiting from permanent CSF diversion is important, since shunting is also associated with significant complications. No clear guidelines exist for PTH treatment, however adjustable or flow-​regulated ventriculo-​peritoneal shunts are

most commonly described as the preferred choice of shunting to reduce the risk of overdrainage. PTH has been reported as a relative contraindication to endoscopic third ventriculostomy (Singh et al., 2008), however this notion has been challenged by others (De Bonis et al., 2013). It is difficult to predict the response of CSF diversion in PTH, since these patients often have comorbidities and significant underlying brain injury. Studies have also suggested that DC is a risk factor for hydrocephalus (De Bonis et al., 2010; Kolias et al., 2013) whereas others do not support this hypothesis (Rahme et al., 2010). Hydrocephalus has been described in TBI patients undergoing DC, ranging in case series between 0% and 88.2% (Ding et al., 2014). It is thought that CSF malabsorption or obstructed flow are the cause of post-​DC hydrocephalus. However, the current case series are limited by their design and heterogeneity of criteria used to diagnose hydrocephalus.

Cranioplasty Cranioplasty is the surgical reconstruction of a bone defect after a previous operation, usually DC, or due to skull injury (Fig. 43.3). A cranioplasty is recommended for both protection of the underlying brain that is left vulnerable to damage with a skull defect and also for restoring the skull contour that might have psychological and social consequences for the patient (Kolias et  al., 2013). The cranioplasty can also facilitate neurological rehabilitation and may also improve neurological function, as observed in patients with syndrome of the trephined (Segal et al., 1994; Di Stefano et al., 2012; Bender et al., 2013; Honeybul et al., 2013). However, this procedure is associated with challenging complications; the most common are wound-​healing problems and implant related infections (Gooch et al., 2009; Honeybul and Ho, 2011; Wachter et al., 2013). The wide range of techniques, graft materials (autologous bone, metal or synthetic) and timing of reconstruction (1–​12 months) discussed in the literature reflects the lack of consensus and good quality evidence (Liang et al., 2007; Beauchamp et al., 2010; Yadla et al., 2011; Glover et al., 2012; Schuss et al., 2012; Bender et al., 2013; Klinger et al., 2014). There main controversy in relation to timing revolves around whether performing the cranioplasty early increases the risk of infections or not and whether it can independently improve neurological outcome. A recent systematic review of available studies (all retrospective), including 528 patients, concluded that cranioplasty may improve neurological outcome, and earlier cranioplasty may enhance this effect ( Liang et al., 2007; Beauchamp et al., 2010; Yadla et al., 2011; Glover et al., 2012; Schuss et al., 2012; Bender et al., 2013; Malcolm et al., 2018). In terms of materials, the patient’s own bone is the most commonly used one, as it fulfils many of the requirements of an ideal material (i.e. it has a low cost and ideal contour and additionally is biocompatible, strong, and radiolucent). However, a few reports in recent years have shown a high failure rate due to resorption or infection; this was confirmed by a recent single-​centre RCT of autologous cranioplasty versus titanium cranioplasty which found that of the 31 patients with an autologous cranioplasty, 7 patients (22%) had complete bone resorption (Honeybul et al., 2017). Acrylic cranioplasty, in the form of polymethylmethacrylate, is cheap, malleable, and lightweight. However, the fragility of the plates that can lead to fractures and the risks associated with exothermic reaction are concerns (Aydin et  al., 2011). Hydroxyapatite has

515

516

Section 10  Neurotrauma and intensive care

(a)

(b)

Fig. 43.3  Cranioplasty. (A) Postoperative computed tomography (CT) three-​dimensional (3D) reconstructions of the head demonstrating a large calvarial defect following unilateral decompressive craniectomy for an evacuation of an acute subdural haematoma with significant intraoperative swelling. (B) Postcranioplasty CT 3D reconstructions following repair of the bony defect with customized PEEK prosthesis. Preoperative CTs were used for production of a tailored flap.

become popular in recent years, as it offers good osseointegration. Nevertheless, its cost remains high and there have been reports that the material can fracture following trauma. Titanium remains a widely used and relatively cheap option but leads to artefact on imaging. A recently published single-​centre RCT of hydroxyapatite versus titanium found that the reoperation rate at 6-​months (26.9% vs. 29.2% respectively) was similar, with a lower number of infections for hydroxyapatite (7.7% vs. 20.8%) but higher number of epidural haematomas (34.6% vs. 8.3%) (Lindner et al., 2017). Several other materials or combinations of materials for cranioplasty exist; detailed description of all options is beyond the scope of this chapter. Key steps of the surgical technique include careful dissection of the scalp from the dura, closure of any dural tears, exposure of bone margins circumferentially, and secure fixation of the autologous bone or artificial implant. In cases of an infection, the plate is usually removed and the wound debrided; it is important to wait until the underlying infection is clear, a process which can take up to 1 year, prior to re-​inserting a plate in order to minimize the risk of further implant infection.

REFERENCES [No authors] (2001). Part 1: guidelines for the management of penetrating brain injury. Introduction and methodology. J Trauma, 51(2 Suppl),  S3–​6. Aarabi, B., Alden, T.D., Hunter Downs III, J., et al. (2001). Neuroimaging in the Management of Penetrating Brain Injury. J Trauma Acute Care Surg, 51(2), S7–​S11. Ali, B. & Ghosh, A. (2002). Antibiotics in compound depressed skull fractures. EMJ, 19(6),  552–​3. American College of Surgeons (2015). TQIP Best Practices in the Managment of Traumatic Brain Injury. https://​www.facs.org/​ quality-​programs/​trauma/​tqip/​best-​practice Arimura, K. & Iihara, K. (2016). Surgical management of intracranial artery dissection. Neurol Med Chr (Tokyo), 56(9), 517–​23.

Aydin, S., Kucukyuruk, B., Abuzayed, B., Aydin, S., & Sanus, G.Z. (2011). Cranioplasty: review of materials and techniques. J Neurosci Rural Pract, 2(2),  162–​7. Beauchamp, K.M., Kashuk, J., Moore, E.E., et al. (2010). Cranioplasty after postinjury decompressive craniectomy:  is timing of the essence? J Trauma, 69(2),  270–​4. Bender, A., Heulin, S., Röhrer, S., et  al. (2013). Early cranioplasty may improve outcome in neurological patients with decompressive craniectomy. Brain Inj, 27(9), 1073–​9. Blacker, D.J. & Wijdicks, E.F. (2004). Clinical characteristics and mechanisms of stroke after polytrauma. Mayo Clin Proc, 79(5),  630–​5. Bohman, L.E. & Schuster, J.M. (2013). Decompressive craniectomy for management of traumatic brain injury: an update. Curr Neurol Neurosci Rep, 13(11), 392. Bontke, C.F. & Boake, C. (1991). Traumatic brain injury rehabilitation. Neurosurg Clin N Am, 2(2), 473–​82. Bor-​Seng-​Shu, E., Figueiredo, E.G., Amorim, R.L.O., et  al. (2012). Decompressive craniectomy: a meta-​analysis of influences on intracranial pressure and cerebral perfusion pressure in the treatment of traumatic brain injury. J Neurosurg, 117(3), 589–​96. Brain Trauma Foundation (2016). Guidelines for the Management of Severe Traumatic Brain Injury, 4th edition. https://​braintrauma.org/​ uploads/​03/​12/​Guidelines_​for_​Management_​of_​Severe_​TBI_​4th_​ Edition.pdf Brodie, H.A. & Thompson, T.C. (1997). Management of complications from 820 temporal bone fractures. Am J Otology, 18(2), 188–​97. Bullock, M.R., Chesnut, R., Ghajar, J., et  al. (2006a). Surgical management of traumatic parenchymal lesions. Neurosurgery, 58(3), S2–​25–​S2–​46. Bullock, M.R., Chesnut, R., Ghajar, J., et al. (2006b). Surgical management of acute epidural hematomas. Neurosurgery, 58(3 Suppl), S7–​ 15; discussion Si–​iv. Bullock, M.R., Chesnut, R., Ghajar, J., et al. (2006c). Surgical management of acute subdural hematomas. Neurosurgery, 58(3 Suppl), S16–​ 24; discussion Si–​iv. Bullock, M.R., Chesnut, R., Ghajar, J., et  al. (2006d). Surgical management of posterior fossa mass lesions. Neurosurgery, 58(3 Suppl), S47–​55; discussion  Si–​iv.

CHAPTER 43  Surgical management of head injury

Bullock, M.R., Chesnut, R., Ghajar, J., et al. (2006e). Surgical management of depressed cranial fractures. Neurosurgery, 58(3 Suppl), S56–​ 60; discussion Si–​iv. CADISS Trial Investigators (2015). Antiplatelet treatment compared with anticoagulation treatment for cervical artery dissection (CADISS): a randomised trial. Lancet Neurol, 14(4),  361–​7. Centers for Disease Control and Prevention (CDC) (2015a). Report to Congress on Traumatic Brain Injury in the United States: Epidemiology and Rehabilitation. National Center for Injury Prevention and Control; Division of Unintentional Injury Prevention. Atlanta, GA. https:// ​ w ww.cdc.gov/​ t raumaticbraininjury/​ p df/​ t bi_​ report_​ to_​ congress_​epi_​and_​rehab-​a.pdf Centers for Disease Control and Prevention (CDC) (2015b). Vaccines and Immunizations. Centers for Disease Control and Prevention. Atlanta, GA. http://​www.cdc.gov/​vaccines/​vpd-​vac/​pneumo/​vacc-​ in-​short.htm Chambers, I.R., Kane, P.J., Signorini, D.F., Jenkins, A., & Mendelow, A.D. (1998). Bilateral ICP monitoring: its importance in detecting the severity of secondary insults. In: Marmarou, A., Bullock, R., Avezaat, C., et al. (eds.) Intracranial Pressure and Neuromonitoring in Brain Injury:  Proceedings of the Tenth International ICP Symposium, pp. 42–​3, Williamsburg, VA, 25–​29 May 1997. Vienna, Austria: Springer Vienna. Chandler, J.R. (1983). Traumatic cerebrospinal fluid leakage. Otolaryngol Clin N Am, 16(3), 623–​32. Compagnone, C., Murray, G.D., Teasdale, G.M., et al. (2007). The management of patients with intradural post-​traumatic mass lesions: a multicenter survey of current approaches to surgical management in 729 patients coordinated by the European Brain Injury Consortium. Neurosurgery, 61(1 Suppl), 232–​40; discussion 40–​1. Cooper, D.J., Rosenfeld, J.V., Murray, L., et al. (2011). Decompressive craniectomy in diffuse traumatic brain injury. N Engl J Med, 364(16), 1493–​502. Cooper, P.R. & Golfinos, J.G. (2000). Head Injury, 4th edition. New York, NY: McGraw-​Hill, Medical Pub. Division, p. 598. Czosnyka, Z., Czosnyka, M., Owler, B., et al. (2005). Clinical testing of CSF circulation in hydrocephalus. Acta Neurochir Suppl, 95, 247–​51. Dahiya, R., Keller, J.D., Litofsky, N.S., Bankey, P.E., Bonassar, L.J., & Megerian, C.A. (1999). Temporal bone fractures:  otic capsule sparing versus otic capsule violating clinical and radiographic considerations. J Trauma, 47(6), 1079–​83. De Bonis, P., Pompucci, A., Mangiola, A., Rigante, L., & Anile, C. (2010). Post-​traumatic hydrocephalus after decompressive craniectomy: an underestimated risk factor. J Neurotrauma, 27(11), 1965–​70. De Bonis, P., Tamburrini, G., Mangiola, A., et  al. (2013). Post-​ traumatic hydrocephalus is a contraindication for endoscopic third-​ ventriculostomy: isn’t it? Clin Neurol Neurosurg, 115(1),  9–​12. Di Stefano, C., Sturiale, C., Trentini, P., et  al. (2012). Unexpected neuropsychological improvement after cranioplasty:  a case series study. Br J Neurosurg, 26(6), 827–​31. Ding, J., Guo, Y., & Tian, H. (2014). The influence of decompressive craniectomy on the development of hydrocephalus:  a review. Arquivos de neuro-​psiquiatria, 72(9), 715–​20. Eastman, A.L., Chason, D.P., Perez, C.L., McAnulty, A.L., & Minei, J.P. (2006). Computed tomographic angiography for the diagnosis of blunt cervical vascular injury: is it ready for primetime? J Trauma, 60(5), 925–​9; discussion 9. Echo, A., Troy, J.S., & Hollier, L.H. Jr. (2010). Frontal sinus fractures. Semin Plast Surg, 24(4), 375–​82. Esposito, D.P. & Walker, J.B. (2009). Contemporary management of penetrating brain injury. Neurosurgery Quarterly, 19(4), 249–​54.

Esposito, T.J., Reed, R.L., 2nd, Gamelli, R.L., & Luchette, F.A. (2005). Neurosurgical coverage:  essential, desired, or irrelevant for good patient care and trauma center status. Ann Surg, 242(3), 364–​70; discussion  70–​4. Flint, A.C., Manley, G.T., Gean, A.D., Hemphill, J.C., 3rd, & Rosenthal, G. (2008). Post-​operative expansion of hemorrhagic contusions after unilateral decompressive hemicraniectomy in severe traumatic brain injury. J Neurotrauma, 25(5), 503–​12. Forsyth, R.J., Raper, J., & Todhunter, E. (2015). Routine intracranial pressure monitoring in acute coma. Cochrane Database Syst Rev, 11, CD002043. Friedman, A.J., Ebersold, J.M., & Quast, M.L. (2001). Post-​traumatic cerebrospinal fluid leakage. World J Surg, 25(8), 1062–​6. Geddes, A.E., Burlew, C.C., Wagenaar, A.E., et al. (2016). Expanded screening criteria for blunt cerebrovascular injury: a bigger impact than anticipated. Am J Surg, 212(6), 1167–​74. Gennarelli, T.A. & Thibault, L.E. (1982). Biomechanics of acute subdural hematoma. J Trauma Acute Care Surg, 22(8),  680–​6. Glover, L.E., Tajiri, N., Lau, T., Kaneko, Y., Van Loveren, H., & Borlongan, C.V. (2012). Immediate, but not delayed, microsurgical skull reconstruction exacerbates brain damage in experimental traumatic brain injury model. PLoS One, 7(3), e33646. Gooch, M.R., Gin, G.E., Kenning, T.J., & German, J.W. (2009). Complications of cranioplasty following decompressive craniectomy: analysis of 62 cases. Neurosurg Focus, 26(6), E9. Gregson, B.A., Rowan, E.N., Mitchell, P.M., et al. (2012). Surgical trial in traumatic intracerebral hemorrhage (STITCH(Trauma)):  study protocol for a randomized controlled trial. Trials, 13, 193. Guerra, W.K., Gaab, M.R., Dietz, H., Mueller, J.U., Piek, J., & Fritsch, M.J. (1999). Surgical decompression for traumatic brain swelling: indications and results. J Neurosurg, 90(2), 187–​96. Guresir, E., Vatter, H., Schuss, P., et al. (2011). Rapid closure technique in decompressive craniectomy. J Neurosurg, 114(4), 954–​60. Guyot, L.L. & Michael, D.B. (2000). Post-​traumatic hydrocephalus. Neurol Res, 22(1),  25–​8. Halbach, V.V., Higashida, R.T., Dowd, C.F., et al. (1993). Endovascular treatment of vertebral artery dissections and pseudoaneurysms. J Neurosurg, 79(2), 183–​91. Haselsberger, K., Pucher, R., & Auer, L.M. (1988). Prognosis after acute subdural or epidural haemorrhage. Acta Neurochir (Wien), 90(3–​4), 111–​16. Hatashita, S., Koga, N., Hosaka, Y., & Takagi, S. (1993). Acute subdural hematoma: severity of injury, surgical intervention, and mortality. Neurol Med Chir (Tokyo), 33(1),  13–​18. Ho, C.L., Wang, C.M., Lee, K.K., Ng, I., & Ang, B.T. (2008). Cerebral oxygenation, vascular reactivity, and neurochemistry following decompressive craniectomy for severe traumatic brain injury. J Neurosurg, 108(5),  943–​9. Honeybul, S. & Ho, K.M. (2011). Long-​ term complications of decompressive craniectomy for head injury. J Neurotrauma, 28(6), 929–​35. Honeybul, S., Janzen, C., Kruger, K., & Ho, K.M. (2013). The impact of cranioplasty on neurological function. Br J Neurosurg, 27(5), 636–​41. Honeybul, S., Morrison, D.A., Ho, K.M., Lind, C.R., & Geelhoed, E. (2017). A randomized controlled trial comparing autologous cranioplasty with custom-​made titanium cranioplasty. J Neurosurg, 126(1),  81–​90. Hutchinson, P.J., Corteen, E., Czosnyka, M., et  al. (2006). Decompressive craniectomy in traumatic brain injury: the randomized multicenter RESCUEicp study (www.RESCUEicp.com). Acta Neurochir Suppl, 96,  17–​20.

517

518

Section 10  Neurotrauma and intensive care

Hutchinson, P.J., Hutchinson, D.B., Barr, R.H., Burgess, F., Kirkpatrick, P.J., & Pickard, J.D. (2000). A new cranial access device for cerebral monitoring. Br J Neurosurg, 14(1),  46–​8. Jaeger, M., Dengl, M., Meixensberger, J., & Schuhmann, M.U. (2010). Effects of cerebrovascular pressure reactivity-​guided optimization of cerebral perfusion pressure on brain tissue oxygenation after traumatic brain injury. Crit Care Med, 38(5), 1343–​7. Kerr, M.E., Weber, B.B., Sereika, S.M., Wilberger, J., & Marion, D.W. (2001). Dose response to cerebrospinal fluid drainage on cerebral perfusion in traumatic brain-​injured adults. Neurosurg Focus, 11(4),  1–​7. Kitanaka, C., Tanaka, J., Kuwahara, M., et al. (1994). Nonsurgical treatment of unruptured intracranial vertebral artery dissection with serial follow-​up angiography. J Neurosurg, 80(4), 667–​74. Kjellberg, R.N. & Prieto, A.J. (1971). Bifrontal decompressive craniotomy for massive cerebral edema. J Neurosurg, 34(4), 488–​93. Klinger, D.R., Madden, C., Beshay, J., White, J., Gambrell, K., & Rickert, K. (2014). Autologous and acrylic cranioplasty: a review of 10 years and 258 cases. World Neurosurg, 82(3), e525–​e30. Koc, R.K., Akdemir, H., Oktem, I.S., Meral, M., & Menku, A. (1997). Acute subdural hematoma:  outcome and outcome prediction. Neurosurg Rev, 20(4), 239–​44. Kolias, A.G., Belli, A., Li, L.M., et al. (2012). Primary decompressive craniectomy for acute subdural haematomas:  results of an international survey. Acta Neurochir (Wien), 154(9), 1563–​5. Kolias, A.G., Kirkpatrick, P.J., & Hutchinson, P.J. (2013). Decompressive craniectomy: past, present and future. Nat Rev Neurol, 9(7), 405–​15. Lawrence, T., Helmy, A., Bouamra, O., Woodford, M., Lecky, F., & Hutchinson, P.J. (2016). Traumatic brain injury in England and Wales: prospective audit of epidemiology, complications and standardised mortality. BMJ Open, 6(11), e012197. Lazaridis, C. & Czosnyka, M. (2012). Cerebral blood flow, brain tissue oxygen, and metabolic effects of decompressive craniectomy. Neurocrit Care, 16(3), 478–​84. Le Roux, P., Menon, D.K., Citerio, G., et al. (2014). The International Multidisciplinary Consensus Conference on Multimodality Monitoring in Neurocritical Care:  evidentiary tables:  a statement for healthcare professionals from the Neurocritical Care Society and the European Society of Intensive Care Medicine. Neurocrit Care, 21 Suppl 2, S297–​361. Leech, P.J. & Paterson, A. (1973). Conservative and operative management for cerebrospinal-​fluid leakage after closed head injury. Lancet, 1(7811), 1013–​16. Li, L.M., Kolias, A.G., Guilfoyle, M.R., et al. (2012). Outcome following evacuation of acute subdural haematomas: a comparison of craniotomy with decompressive craniectomy. Acta Neurochir (Wien), 154(9), 1555–​61. Li, X., von Holst, H., & Kleiven, S. (2013). Decompressive craniectomy causes a significant strain increase in axonal fiber tracts. J Clin Neurosci, 20(4), 509–​13. Liang, W., Xiaofeng, Y., Weiguo, L., et al. (2007). Cranioplasty of large cranial defect at an early stage after decompressive craniectomy performed for severe head trauma. J Craniofac Surg, 18(3), 526–​32. Lindner, D., Schlothofer-​Schumann, K., Kern, B.C., Marx, O., Muns, A., & Meixensberger, J. (2017). Cranioplasty using custom-​made hydroxyapatite versus titanium:  a randomized clinical trial. J Neurosurg, 126(1), 175–​83. Low, C.Y., Low, Y.Y., Lee, K.K., Chan, S.P., & Ang, B.T. (2013). Post-​ traumatic hydrocephalus after ventricular shunt placement in a Singaporean neurosurgical unit. J Clin Neurosci, 20(6), 867–​72. Maas, A.I., Stocchetti, N., & Bullock, R. (2008). Moderate and severe traumatic brain injury in adults. Lancet Neurol, 7(8), 728–​41.

Makarenko, S., Griesdale, D.E., Gooderham, P., & Sekhon, M.S. (2016). Multimodal neuromonitoring for traumatic brain injury: a shift towards individualized therapy. J Clin Neurosci, 26,  8–​13. Malcolm, J.G., Rindler, R.S., Chu, J.K., et al. (2018). Early cranioplasty is associated with greater neurological improvement:  a systematic review and meta-​analysis. Neurosurgery, 82(3), 278–​88. Marbacher, S., Andres, R.H., Fathi, A.R., & Fandino, J. (2008). Primary reconstruction of open depressed skull fractures with titanium mesh. J Craniofac Surg, 19(2),  490–​5. Marmarou, A., Foda, M.A., Bandoh, K., et al. (1996). Posttraumatic ventriculomegaly: hydrocephalus or atrophy? A new approach for diagnosis using CSF dynamics. J Neurosurg, 85(6), 1026–​35. Mathias, T., Levy, J., Fatakia, A., & McCoul, E.D. (2016). Contemporary approach to the diagnosis and management of cerebrospinal fluid rhinorrhea. Ochsner J, 16(2), 136–​42. Maxeiner, H. & Wolff, M. (2007). Pure subdural hematomas: a postmortem analysis of their form and bleeding points. Neurosurgery, 61(1 Suppl), 267–​72; discussion 72–​3. McGuirt, W.F., Jr. & Stool, S.E. (1995). Cerebrospinal fluid fistula: the identification and management in pediatric temporal bone fractures. Laryngoscope, 105(4 Pt 1), 359–​64. Mendelow, A.D., Gregson, B.A., Rowan, E.N., et al. (2015). Early surgery versus initial conservative treatment in patients with traumatic intracerebral hemorrhage (STITCH[Trauma]): the first randomized trial. J Neurotrauma, 32(17), 1312–​23. Mendizabal, G.R., Moreno, B.C., & Flores, C.C. (1992). Cerebrospinal fluid fistula:  frequency in head injuries. Revue de laryngologie—​ Otologie—​Rhinologie, 113(5),  423–​5. Mohan, I.V. (2014). Current optimal assessment and management of carotid and vertebral spontaneous and traumatic dissection. Angiology, 65(4), 274–​83. Nalbach, S.V., Ropper, A.E., Dunn, I.F., & Gormley, W.B. (2012). Craniectomy-​ associated progressive extra-​ axial collections with treated hydrocephalus (CAPECTH): redefining a common complication of decompressive craniectomy. J Clin Neurosci, 19(9), 1222–​7. Nash, J.J., Friedland, D.R., Boorsma, K.J., & Rhee, J.S. (2010). Management and outcomes of facial paralysis from intratemporal blunt trauma: a systematic review. Laryngoscope, 120 Suppl 4, S214. Nguyen, R., Fiest, K.M., McChesney, J., et al. (2016). The international incidence of traumatic brain injury: a systematic review and meta-​ analysis. Can J Neurol Sci, 43(6), 774–​85. Nwachuku, E.L., Puccio, A.M., Fetzick, A., et al. (2014). Intermittent versus continuous cerebrospinal fluid drainage management in adult severe traumatic brain injury: assessment of intracranial pressure burden. Neurocrit Care, 20(1),  49–​53. Nyquist, G.G., Rosen, M.R., Friedel, M., et  al. (2013). Management of cerebrospinal fluid leaks of the anterior and lateral skull base. J Neurol Surg B Skull Base, 74(Suppl 1), A046. Offiah, C. & Twigg, S. (2009). Imaging assessment of penetrating craniocerebral and spinal trauma. Clin Radiol, 64(12), 1146–​57. Ommaya, A. (1996). Neurosurgery, 2nd edition. New  York, NY: McGraw-​Hill. Ordog, G.J., Wasserberger, J., & Balasubramanium, S. (1984). Wound ballistics: theory and practice. Ann Emerg Med, 13(12), 1113–​22. Paoletti, P., Pezzotta, S., & Spanu, G. (1983). Diagnosis and treatment of post-​traumatic hydrocephalus. J Neurosurg Sci, 27(3),  171–​5. Pham, M.H., Rahme, R.J., Arnaout, O., et  al. (2011). Endovascular stenting of extracranial carotid and vertebral artery dissections:  a systematic review of the literature. Neurosurgery, 68(4), 856–​66; discussion 66. Protocol 14PRT/​6944. Randomised Evaluation of Surgery with Craniectomy for patients Undergoing Evacuation of Acute Subdural

CHAPTER 43  Surgical management of head injury

Haematoma (RESCUE-​ASDH)—​ISRCTN873705452015. http://​ www.thelancet.com/​doi/​story/​10.1016/​html.2015.08.14.2280 Rahme, R., Weil, A.G., Sabbagh, M., Moumdjian, R., Bouthillier, A., & Bojanowski, M.W. (2010). Decompressive craniectomy is not an independent risk factor for communicating hydrocephalus in pa tients with increased intracranial pressure. Neurosurgery, 67(3), 675–​8; discussion 8. Ratilal, B.O., Costa, J., Pappamikail, L., & Sampaio, C. (2015). Antibiotic prophylaxis for preventing meningitis in patients with basilar skull fractures. Cochrane Database Syst Rev, 4, CD004884. Rimmer, J., Belk, C., Lund, V., Swift, A., & White, P. (2014). Immunisations and antibiotics in patients with anterior skull base cerebrospinal fluid leaks. J Laryngol Otol, 128(7),  626–​9. Rubinstein, S.M., Peerdeman, S.M., van Tulder, M.W., Riphagen, I., & Haldeman, S. (2005). A systematic review of the risk factors for cervical artery dissection. Stroke, 36(7), 1575–​80. Ryan, C.G., Thompson, R.E., Temkin, N.R., Crane, P.K., Ellenbogen, R.G., & Elmore, J.G. (2012). Acute traumatic subdural hematoma: current mortality and functional outcomes in adult patients at a Level I trauma center. J Trauma Acute Care Surg, 73(5), 1348–​54. Savva, A., Taylor, M.J., & Beatty, C.W. (2003). Management of cerebrospinal fluid leaks involving the temporal bone: report on 92 patients. Laryngoscope, 113(1),  50–​6. Schuss, P., Vatter, H., Marquardt, G., et al. (2012). Cranioplasty after decompressive craniectomy:  the effect of timing on postoperative complications. J Neurotrauma, 29(6), 1090–​5. Segal, D.H., Oppenheim, J.S., & Murovic, J.A. (1994). Neurological recovery after cranioplasty. Neurosurgery, 34(4), 729–​31. Servadei, F., Nasi, M.T., Cremonini, A.M., Giuliani, G., Cenni, P., & Nanni, A. (1998). Importance of a reliable admission Glasgow Coma Scale score for determining the need for evacuation of posttraumatic subdural hematomas: a prospective study of 65 patients. J Trauma, 44(5), 868–​73. Shapiro, S.A. & Scully, T. (1992). Closed continuous drainage of cerebrospinal fluid via a lumbar subarachnoid catheter for treatment or prevention of cranial/​spinal cerebrospinal fluid fistula. Neurosurgery, 30(2),  241–​5. Shore, P.M., Thomas, N.J., Clark, R.S., et al. (2004). Continuous versus intermittent cerebrospinal fluid drainage after severe traumatic brain injury in children: effect on biochemical markers. J Neurotrauma, 21(9), 1113–​22. Singh, I., Haris, M., Husain, M., Husain, N., Rastogi, M., & Gupta, R.K. (2008). Role of endoscopic third ventriculostomy in patients with communicating hydrocephalus: an evaluation by MR ventriculography. Neurosurg Rev, 31(3), 319–​25. Spetzler, R.F. & Wilson, C.B. (1978). Management of recurrent CSF rhinorrhea of the middle and posterior fossa. J Neurosurg, 49(3),  393–​7. Stiefel, M.F., Heuer, G.G., Smith, M.J., et al. (2004). Cerebral oxygenation following decompressive hemicraniectomy for the treatment of refractory intracranial hypertension. J Neurosurg, 101(2),  241–​7. Stiver SI (2009). Complications of decompressive craniectomy for traumatic brain injury. Neurosurg Focus, 26(6), E7. Stocchetti, N., Zanaboni, C., Colombo, A., et  al. (2008). Refractory intracranial hypertension and ‘second-​tier’ therapies in traumatic brain injury. Intensive Care Med, 34(3),  461–​7. Tagliaferri, F., Compagnone, C., Korsic, M., Servadei, F., & Kraus, J. (2006). A systematic review of brain injury epidemiology in Europe. Acta Neurochir (Wien), 148(3), 255–​68; discussion 68. Tagliaferri, F., Zani, G., Iaccarino, C., et  al. (2012). Decompressive craniectomies, facts and fiction: a retrospective analysis of 526 cases. Acta Neurochir (Wien), 154(5), 919–​26.

Timofeev, I., Dahyot-​ Fizelier, C., Keong, N., et  al. (2009). Ventriculostomy for control of raised ICP in acute traumatic brain injury. In: Steiger, H.J. (ed.) Acta Neurochirurgica Supplements, pp. 99–​104. Vienna, Austria: Springer Vienna. Timofeev, I. & Hutchinson, P.J. (2006). Outcome after surgical decompression of severe traumatic brain injury. Injury, 37(12), 1125–​32. Tribl, G. & Oder, W. (2000). Outcome after shunt implantation in severe head injury with post-​traumatic hydrocephalus. Brain Inj, 14(4), 345–​54. Tsuei, Y.-​S., Sun, M.-​H., Lee, H.-​D., et  al. (2005). Civilian gunshot wounds to the brain. J Chinese Med Assoc, 68(3), 126–​30. Valadka, A.B. & Robertson, C.S. (2007). Surgery of cerebral trauma and associated critical care. Neurosurgery, 61(1 Suppl), 203–​20; discussion  20–​1. von Holst, H., Li, X., & Kleiven, S. (2012). Increased strain levels and water content in brain tissue after decompressive craniotomy. Acta Neurochir (Wien), 154(9), 1583–​93. Wachter, D., Reineke, K., Behm, T., & Rohde, V. (2013). Cranioplasty after decompressive hemicraniectomy: underestimated surgery-​associated complications? Clin Neurol Neurosurg, 115(8), 1293–​7. Whitfield, P.C., Patel, H., Hutchinson, P.J., et  al. (2001). Bifrontal decompressive craniectomy in the management of posttraumatic intracranial hypertension. Br J Neurosurg, 15(6),  500–​7. Yadla, S., Campbell, P.G., Chitale, R., Maltenfort, M.G., Jabbour, P., & Sharan, A.D. (2011). Effect of early surgery, material, and method of flap preservation on cranioplasty infections: a systematic review. Neurosurgery, 68(4), 1124–​30. Yamakami, I. & Yamaura, A. (1993). Effects of decompressive craniectomy on regional cerebral blood flow in severe head trauma patients. Neurol Med Chir, 33(9), 616–​20. Yamaura, A. (1994). Nontraumatic intracranial arterial dissection: natural history, diagnosis, and treatment. Contemporary Neurosurgery, 16(5), 1. Zayas, J.O., Feliciano, Y.Z., Hadley, C.R., Gomez, A.A., & Vidal, J.A. (2011). Temporal bone trauma and the role of multidetector CT in the emergency department. RadioGraphics, 31(6), 1741–​55. Zumkeller, M., Behrmann, R., Heissler, H.E., & Dietz, H. (1996). Computed tomographic criteria and survival rate for patients with acute subdural hematoma. Neurosurgery, 39(4), 708–​12; discussion  12–​3.

RELATED LINKS TO EBRAIN Carotid and Vertebral Artery Dissection. https://learning.ebrain.net/ course/view.php?id=126 Evacuation of a Chronic Subdural Haematoma. https://learning. ebrain.net/course/view.php?id=1100 Insertion of an Intracranial Pressure Monitor. https://learning.ebrain. net/course/view.php?id=1103 Management of Depressed Skull Fractures. https://learning.ebrain. net/course/view.php?id=1102 Performing a Decompressive Craniectomy for Traumatic Brain Injury. https://learning.ebrain.net/course/view.php?id=1107 Performing a Cranioplasty. https://learning.ebrain.net/course/view. php?id=1108 Performing a Trauma Craniotomy. https://learning.ebrain.net/course/ view.php?id=1105 Surgical Management of Head Injuries. https://learning.ebrain.net/ course/view.php?id=1099

519

44

Complications of head injury Fardad T. Afshari, Antonio Belli, and Peter C. Whitfield

Introduction Complications following head injury can be classified according to the timing of the complication—​early (within 1 week) or late (beyond 1 week). Complications can also be considered according to the underlying pathological process (e.g. cerebrospinal fluid (CSF) leak, vascular complications, epilepsy, CSF leak, and hydrocephalus). In addition, many rehabilitation needs are commonly identified following head injury. In the early phase, these focus around medical care (e.g. chest physiotherapy, avoidance of contractures, nutrition, and communication). In the later phases, patient-​centred goal-​directed therapy is of increasing importance. In this chapter we describe several early complications. Consideration is then given to delayed complications although it is recognized that classification by timing is an inexact science and there is considerable overlap between early and late complications.

Early complications of head injury Vascular injuries Traumatic brain injury may be associated with vascular injuries. Most of these occur as early sequalae directly related to the transmission of operant forces caused by the trauma. Rarely a vascular injury can occur as a result of penetrating injury (e.g. low velocity bullet) but is more commonly associated with bony injuries to the cervical spine and/​or skull base. Traumatic subarachnoid haemorrhage Traumatic subarachnoid haemorrhage is the most common cause of subarachnoid blood on CT head. One-​third of patients with moderate to severe head injury present with evidence of subarachnoid blood on CT head (Eisenberg et al., 1990; Servadei et al., 2002). The volume of traumatic subarachnoid haemorrhage on initial CT is associated with a worse outcome (Paiva et al., 2010). Although vasospasm can occur following traumatic subarachnoid haemorrhage, this complication is not as common as vasospasm following aneurysmal subarachnoid haemorrhage. A  systematic review found no evidence to suggest nimodipine improves outcomes in patients with traumatic subarachnoid haemorrhage (Vergouwen et al., 2006).

Carotid and vertebral artery dissection Carotid dissection is the most frequent vascular injury and usually occurs in the extracranial or skull base segments at points of tethering or fracture (Fig. 44.1). Similarly, vertebral artery dissection occurs in the V2 and V3 segments where the artery enters and exits the foramen transversarium. Trauma to the carotid or vertebral arteries can cause an intimal tear resulting in a dissection with an associated false lumen. Expansion of this false tract can lead to occlusion of the parent vessel causing downstream infarction. Localized trauma to a vessel can lead to formation of mural thrombus and partial or total occlusion in the absence of dissection. A thrombus in a partially occluded vessel can cause distal emboli typically affecting watershed areas of the brain. Rarely a traumatic injury may cause a pseudo-​ aneurysm that can cause a spontaneous post-​traumatic intracranial haemorrhage. CT angiogram, MR angiogram and digital subtraction cerebral angiography are the most sensitive studies in the investigation of vascular injuries in traumatic brain injury. Antiplatelet therapy is the main mode of treatment for dissection injuries, although consideration must be given to concomitant intracranial or extracranial haemorrhagic injuries. Endovascular treatment is an option in some cases; however, most traumatic dissections tend to heal spontaneously in a matter of weeks. Wherever possible, the management of a traumatic dissection should involve neurosurgery, stroke medicine, and interventional radiology in a multidisciplinary fashion. Arteriovenous fistulae Although rare, arteriovenous fistulae can form following traumatic brain injury. These may present in the first week or may have a delayed appearance. The most common location for fistula formation is between the cavernous portion of carotid artery and the cavernous venous sinus. The incidence of carotico-​cavernous fistulae (CCF) following head trauma is 0.2–​0.3% (Fabian et al., 1999). These direct fistulae are true connections between a ruptured intracavernous carotid artery and the surrounding venous channels within the cavernous sinus and are therefore high flow and likely to be symptomatic (see Chapter 49). Clinically this condition can present with the triad of pulsatile proptosis, conjunctival injection, and bruit. The management of such fistulae is predominantly endovascular and aims to preserve ocular function and prevent cortical venous reflux and its haemorrhagic sequelae. Indirect CCF do not typically occur

522

Section 10  Neurotrauma and intensive care

(a)

(b)

Fig. 44.1  Bilateral carotid dissection. (A) CT angiogram showing bilateral tapering of internal carotid indicative of dissection. (B) T2-​weighted MRI demonstrating high signal within carotids. Normally the major vessels demonstrate flow voids (low signal such as in the basilar artery), however this is lost with carotid occlusion.

acutely following trauma, although, if identified in a patient with trauma, they can be managed less acutely and sometimes spontaneously regress.

Cranial nerve injuries Cranial nerve injuries are common following head injury. Although they are almost always present as an early post-​traumatic complication, it is not until the rehabilitation phase that the effects of a palsy become prominent. Cranial nerve palsies may result from traumatic disruption of cranial nerves, particularly at the interface with the exiting skull foraminae. They may also occur consequential to the effects of raised ICP (e.g. third nerve palsy, sixth nerve palsy, and seventh nerve palsy). Damage to the olfactory nerve is common, with or without anterior skull base fractures, as the delicate olfactory nerves are easily injured as they pass through the cribriform plate. The olfactory nerves can slowly recover in some cases, but the sensory loss is usually permanent; in addition, loss of smell can affect the sensation of taste due to overlapping perception of taste and olfaction. Optic nerve transection or compression due to facial and orbital bone fractures typically presents with unilateral visual loss. A relative afferent pupillary defect is normally present. Visual impairment after trauma always warrants urgent radiological investigation and an ophthalmological consultation. Diplopia following head injury is common and can occur due to damage to oculomotor, trochlear, or abducent nerves and should be differentiated from non-​neural, mechanical mechanisms of diplopia such as an orbital wall fracture or damage to the trochlear sling within the lateral orbit. Temporal bone fractures can cause a range of deficits, particularly if they involve the otic capsule. Historically, these were divided into longitudinal fractures (along the long access of the petrous temporal bone) or transverse fractures. Transverse fractures being less common but more likely to cause sensorineural hearing loss and facial palsy. The involvement of the otic capsule is now recognized

as the pre-​eminent factor that influences cranial neuropathy. The deficits include facial nerve palsy, hearing loss, and CSF leak. Facial nerve palsies can recover over time, but the cornea must be protected if exposed. Evidence to justify the use of corticosteroids in the treatment of traumatic facial nerve palsy is weak. Early involvement of speech and language therapy and physiotherapy is recommended. Facial nerve reanimation techniques can be considered for permanent disabling deficits. Hearing loss is a common presentation following head injury associated with temporal and petrous bone fractures. Hearing deficits may be conductive or sensory in nature. Conductive hearing loss is due to occlusion of the auditory meatus with blood, damage to tympanic membrane, or disruption of the ossicular chain. The latter may be amenable to surgical treatment and should be investigated with a high resolution fine-​cut CT scan. Otological input is advised to guide further investigation and treatment. Sensorineural deafness can occur due to damage to the cochlear structures or the cochlear nerve itself. This can arise from fracture of the petrous temporal bone or secondary to nerve ischaemia caused by raised ICP. Lower cranial nerve palsies can occur individually or in combination. These are usually secondary to occipital bone fractures extending to the jugular foramen. Different syndromes associated with lower cranial nerve palsies have been documented in the context of both tumour and trauma and are summarized in Table 44.1.

Depressed skull fractures Traumatic brain injury is commonly associated with depressed skull fractures. Depressed fractures should be identified on the initial CT imaging and can be well visualized with volumetric reconstructions. Depressed skull fractures may be associated with underlying brain contusions, extradural, or subdural haematoma. In addition, they may be associated with a dural breach, increasing the risk of delayed infection, seizure, meningitis, or empyema. Absolute indications for elevation of a depressed skull fracture include treatment of an

CHAPTER 44  Complications of head injury

Table 44.1  Jugular foramen syndromes Syndrome

IX

X

XI

XII

Sympathetic

Villaret

Yes

Yes

Yes

Yes

Yes

Collet-​Sicard

Yes

Yes

Yes

Yes

No

Vernet

Yes

Yes

Yes

No

No

Schmidt

No

Yes

Yes

No

No

Tapia

No

Yes

No

Yes

Yes/​No

Jackson

No

Yes

Yes

Yes

No

open fracture that necessitates debridement of bone fragments and, in some cases, dural repair. Pericranium or temporalis fascia is the preferred material for dural repair. Relative indications for repair include significant fragment depression (+/​-​haematoma) causing mass effect, depression greater than the width of the bone and cosmetic defects. Any depressed fractures overlying the venous sinuses should be managed cautiously due to the risk of catastrophic haemorrhage following fracture elevation.

breach of the dura. Patients with pneumocephalus are at high risk of CSF leak and should be closely monitored and managed as mentioned. Although small volumes of air within the cranium do not pose any danger, large amounts of air particularly when under pressure can act as a mass lesion causing midline shift. Pneumocephalus can cause a reduced level of consciousness, seizures, or focal neurological signs. The appearance of a large amount of air antero-​ superior to both frontal lobes in the supine patient on axial CT scan is termed ‘mount Fuji sign’ owing to double peak appearance of the compressed frontal lobes (Michel, 2004). At times this can manifest as a tension pneumocephalus mandating emergency treatment. Pneumocephalus is normally managed with high-​flow oxygen to reduce the partial pressure of nitrogen and encourage absorption of air from the intracranial space. Rarely, in cases of severe midline shift and tension pneumocephalus, urgent burr holes may be required to allow air egress from the subdural space.

Late complications of head injury

Early CSF leak

Seizures

CSF rhinorrhoea and otorrhoea are common complications of skull base fractures and indicate breach of the dura. CSF rhinorrhoea can occur from complex ethmoidal fractures breaching the anterior fossa dura. CSF rhinorrhoea can also occur as a result of CSF egress via the aerated portion of the temporal bone, into the middle ear then through the Eustachian tube into the nasopharynx. CSF otorrhea can occur from a temporal bone fracture that is associated with a perforated ear drum. Up to 90% of these CSF leaks resolve spontaneously within 2 weeks. The remainder of the patients may continue to have persistent leak or develop delayed CSF leak. Antibiotic prophylaxis is not indicated, but antibiotic treatment should be instigated without delay if signs of meningitis develop. Antipneumococcal vaccination is also recommended in all skull base fractures or proven or suspected CSF leakage (Phang et al., 2016). For those patients with persistent CSF leak beyond 2 weeks, there is some contention as to the best subsequent treatment. CSF drainage by lumbar drain is very effective for minimizing the flow through the site of leak and therefore promoting healing. However, if the fistula is large, this may entrain air into the intracranial cavity, and caution must be used in any patient with presumed raised intracranial pressure to avoid tonsillar descent. Although CSF rhinorrhoea or otorrhoea may be clinically evident, there are occasions that CSF leak post injury maybe difficult to distinguish from nasal discharge. In such cases, manoeuvres such as leaning the head forward may accentuate the CSF leak and further suggest CSF rhinorrhoea. Alternatively, if a CSF sample can be collected, it can be tested for beta-​transferrin which is found at high levels in CSF. Glucose testing is unlikely to be helpful as nasal secretions may also contain glucose (Baker et al., 2005). Figure 44.2 shows an algorithm for the management of a suspected CSF leak.

Although seizures can present early after a traumatic brain injury, late seizures are of prime concern to recovering patients. Overall, seizures can occur in up to 30% of severe traumatic brain injury patients. In the early phase, seizures resulting in increased metabolic demand, excitotoxicity and raised intracranial pressure, and may cause severe secondary brain damage. Such issues also exist in the late phase although antiepileptics usually afford a degree of control over these processes. The risk of seizure post head injury depends on several factors. The classical work of Bryan Jennett (pre-​CT) defined the following: severity of injury, duration of PTA, presence of depressed skull fracture, and intracerebral haematoma/​contusion. Patients who sustain a seizure within the first week are an increased risk of late epilepsy. Annegers et al. have defined the risk according to severity of injury, albeit also using data predominantly acquired pre-​CT. Severe injury is characterized by brain contusion, intracranial haematoma, loss of consciousness, or post-​traumatic amnesia for more than 24 hours. The 5-​year cumulative probability of post-​traumatic seizure is 10% and 30-​year cumulative probability is 16.7%. Christensen’s et al. study in children and young adults showed the risk of epilepsy was increased after a mild brain injury (RR 2.22, 95% CI 2.07–​2.38), severe brain injury (7.40, 6.16–​8.89), and skull fracture (2.17, 1.73–​ 2.71). Overall, the risk of developing late epilepsy reduces with time, but nevertheless for patients at greatest risk, this remains above the baseline risk of the general population for 10–​15  years (Jennett, 1973; Jennett, 1975; Annegers et al., 1998; Christensen et al., 2009). There is currently class I evidence available that use of antiepileptics can lead to reduction in seizure frequency in early phase following head injury (first 7  days) but has no effect on the incidence of late onset seizures (Temkin et  al., 1990). The use of prophylactic antiepileptics has no effect on mortality or neurological disability when used in the early or late phases post head injury (Schierhout and Roberts, 2001). Traditionally phenytoin has been used as a first-​line agent for seizure control in head injury patients. Recent prospective randomized studies have demonstrated that levetiracetam and phenytoin are

Pneumocephalus Pneumocephalus (air within the cranial cavity) can occur following head trauma in association with calvarial or skull base fractures. The incidence of pneumocephalus secondary to head injury is between 0.5% and 1%. Air within the subarachnoid space indicates a

523

524

Section 10  Neurotrauma and intensive care

Laboratory and radiological confirmation of CSF leak

Presence of features needing surgical management: → Absolute: ♦ Delayed post-traumatic CSF rhinorrhea ♦ Intracranial pathology requiring acute intervention ♦ Open fractures → Relative: ♦ Anatomy of the skull fracture suggests that spontaneous closure would be difficult, e.g. in a large (>1 cm) fracture, comminuted fractures, depressed skull base fracture, fracture accompanied by complications (e.g. cranial nerve deficits), or tension pneumocephalus

No

Yes

Non surgical management: Monitor patient condition for first 7 days: → Bed rest → Head elevation → Strict CSF rhinorrhea precautions (avoid nose blowing, Valsalva manoeuvres, the use of straws) → Avoid high blood pressures CSF leak persist

Successful resolution of CSF leak

CSF diversion using lumbar or ventricular drain or repeated LP for not more than 14 days after initial head injury Successful resolution of CSF leak

Surgical management

CSF leak persists

Does the fracture have the following features: → Concomitant intracranial injury → Large fractures (Schlosem) → Brain herniation (Schlosem) → Open fractures (Schlosem) → Fracture involving the posterior wall of the frontal sinus

For frontal sinus fractures involving the posterior wall

Yes

No

Does the fracture have the following features: → Fractures with more than one fracture fragment → Fracture with more than one table width’s of displacement at any fracture line → Open fracture of the frontal sinus → Presence of any intracranial pathology requiring emergency surgical management

Intracranial

Extracranial

Is the dural defect simple and identifiable

Endoscopic transnasal surgery

Yes

No

Cranialization of the frontal sinus

Does the fracture involve the frontonasal outflow tract? Yes

No

Obliteration of the frontal sinus

Reconstruction of the frontal sinus

Extradural intracranial approach

Intradural intracranial approach

Failed extracranial surgery

Successful resolution of CSF leak

Fig. 44.2  Algorithm for investigation of CSF leaks. Reproduced with permission from See Yung Phang, Kathrin Whitehouse, Lucy Lee, et al., Management of CSF leak in base of skull fractures in adults, British Journal of Neurosurgery, Volume 30, Issue 6, pp. 596–​604, Copyright © 2016 The Neurosurgical Foundation, reprinted by permission of Taylor & Francis Ltd, www.tandfonline.com on behalf of The Neurosurgical Foundation.

CHAPTER 44  Complications of head injury

both effective as seizure prophylaxis agents in head injury with no significant difference in seizure control rates (Szaflarski et al., 2010). Levetiracetam has therefore become the favoured option given that treatment is effective, with a low risk of symptomatic side effects, no impact on cognition, and is relatively low cost (Eddy et al., 2011). Patients with head injury and seizures should be informed about risks of swimming and working on open machinery. They should be advised to refrain from driving and to inform the appropriate driving regulatory agency. The decision to stop antiepileptic medication in the longer term, once a patient has been seizure free for a period of time, is a difficult one. The risk of seizures is likely to decline as time progresses from the initial injury, however no specific guidelines exist. A decision on stopping antiepileptics needs to balance the complications and risks of prolonged therapy against the risk of a further seizure that can risk injury and lead to further limitations in driving and independence.

Hydrocephalus Hydrocephalus can occur both in acute or late phase following head injury. Acute hydrocephalus is rare following head injury and tends to be associated with intraventricular haemorrhage. In some units, ventricular drainage is performed as an adjunct in the management of raised ICP. This is normally a temporary measure and is not usually indicative of hydrocephalus. Late phase hydrocephalus following head injury appears to be a separate pathophysiological entity. The incidence of post-​traumatic brain injury hydrocephalus is variable depending on the criteria used in the studies. Kammersgaard et  al. reported post-​traumatic hydrocephalus in 14.2% of patients with majority of patients presenting during the rehabilitation (Kammersgaard et  al., 2013). Many patients present with a plateau in their recovery, often in association with the triad of symptoms associated with idiopathic normal pressure hydrocephalus (gait disturbance, cognitive decline, and urinary urgency or incontinence). The exact mechanism underlying late phase hydrocephalus after head injury is not fully understood. Imaging normally demonstrates communicating hydrocephalus and lumbar puncture may demonstrate normal opening pressures. Patients with delayed hydrocephalus can however present with reduced level of consciousness and drowsiness even following a long period of recovery. Patients with hydrocephalus benefit from CSF diversion procedures such as a ventriculoperitoneal shunt. Uncertainty concerning the diagnosis is however a confounding issue: the differential diagnosis of ex-​vacuo dilatation of the ventricles following post-​traumatic cerebral atrophy should be considered, as, in this case, CSF diversion will not help and may risk overdrainage with the risk of subdural collections. ICP monitoring studies and CSF infusion studies may be helpful adjuncts in the investigation of these patients.

Pituitary dysfunction In recent years pituitary dysfunction has been increasingly recognized as a previously overlooked delayed complication of traumatic brain injury. Traumatic brain injury may be associated with pituitary dysfunction, which may contribute to long-​term physical, cognitive, and psychological disability. Pituitary dysfunction can be broadly divided into anterior or posterior pituitary dysfunction and can be detected either in early or delayed phase (>6 months) following head injury. In a study by Bondanelli, anterior pituitary dysfunction was

reported to be 37.5%, 57.1%, and 59.3% in mild, moderate, and severe traumatic brain injury patients, respectively but in a small cohort of patients (Bondanelli et al., 2004). Other studies have reported rate of anterior pituitary dysfunction at 20–​25% following traumatic brain injury (Agha et al., 2007; Alavi et al., 2016). The most common alterations in anterior pituitary hormones appear to be deficiency of gonadotrophins and growth hormone, followed by ACTH and thyrotropin. Symptoms of pituitary dysfunction, such as fatigue, depression, and sexual dysfunction, overlap with postconcussional symptoms; therefore, post-​ traumatic pituitary insufficiency can remain undiagnosed unless specifically investigated with specific endocrinology tests. Tests in the acute phase do not accurately predict long-​term dysfunction and it is important to monitor symptomatic patients in the first year after injury. The notable exception is cortisol insufficiency, which can appear very early on and is potentially a life-​threatening emergency: requiring immediate hydrocortisone replacement. Posterior pituitary dysfunction can also lead to major complications following head injury. Abnormalities of this axis can occur in both the early and late phases. The syndrome of inappropriate antidiuretic hormone secretion (SIADH), diabetes insipidus, and cerebral salt wasting are three conditions that can lead to electrolyte imbalance following head injury. Clinical assessment of patient fluid status and measurement of plasma and urine sodium and osmolality are crucial in delineating the cause of electrolyte imbalance in these patients. SIADH is the most common posterior pituitary/​hypothalamic dysfunction syndrome and is reported to affect 33% of patients with head injury. Uncontrolled release of antidiuretic hormone (ADH) can lead to excessive fluid resorption resulting in serum hyponatraemia and reduced plasma osmolality. The patient is also in a state of relative fluid overload. This should be differentiated from hyponatraemia secondary to cerebral salt wasting: this is related to brain natriuretic peptide causing excessive sodium excretion in conjunction with a reduction in total body water. Diabetes insipidus is associated with impaired release of ADH leading to excess diuresis and dehydration in association with hypernatraemia. Replacement therapy with DDAVP nasal spray may be required. The British Neurotrauma Group has published guidance for the screening and management of pituitary dysfunction following traumatic brain injury in adults (Tan et al., 2017).

Delayed CSF leak Although the majority of CSF leaks post head injury resolve within 2 weeks of injury, there are cases where persistent otorrhoea or rhinorrhoea continue and require surgical intervention (Phang et al., 2016). Fractures of the posterior wall of the frontal sinus are one of the most common injuries where persistent CSF leaks occur. In addition to persistent CSF leak since the time of injury, delayed CSF leakage sometimes occurs years following the injury, which may be due to re-​opening of the fistula or coexistent hydrocephalus. Thin-​slice cranial CT scan with 3D reconstruction provides the necessary anatomical detail to delineate the site of CSF leak. Due to the serious infective complications associated with CSF leaks, a systematic approach to the management of patients with skull base fractures is advisable. An algorithm for management is provided in Figure 44.3.

525

526

Section 10  Neurotrauma and intensive care

Clinical features of BSF present after head trauma All patients need HRCT HRCT identifies skull base fracture

if signs of CSF leak present

No fracture identified on HRCT

No sign of CSF leak

if signs of CSF leak present

Collect fluid for β2 - Tr/βTP testing

β2 - Tr/βTP positive

No sign of CSF leak

Collect fluid for β2 - Tr/βTP testing and review HRCT

β2 - Tr/βTP negative β2 - Tr/βTP positive

β2 - Tr/βTP negative

Dural lesion present

CTc & MRIc +/– MRI with intrathecal GdDTPA

If either one is positive

Both are negative

Dural lesion present

Follow up in 3 months in outpatient clinic

Negative findings then discharge

No dural lesion

Advice and discharge

Positive findings in clinic: → Meningitis → Headaches → Salty or sweet taste → Cerebral abscess

Fig. 44.3  Algorithm for the investigation and treatment of a skull base fracture (HRCT, high resolution CT). Reproduced with permission from See Yung Phang, Kathrin Whitehouse, Lucy Lee, et al., Management of CSF leak in base of skull fractures in adults, British Journal of Neurosurgery, Volume 30, Issue 6, pp. 596–​604, Copyright © 2016 The Neurosurgical Foundation, reprinted by permission of Taylor & Francis Ltd, www.tandfonline.com on behalf of The Neurosurgical Foundation.

Formal skull base repair of troublesome CSF leaks is challenging. At the anterior skull base, a bicoronal flap can be raised and the pericranium delivered through an anterior ‘letterbox’ at the inferior extent of the craniotomy in order to provide a vascularized flap to cover the site of leak either extradurally or intradurally. This is a significant surgical undertaking and risks anosmia. Alternatively, endonasal endoscopic repair with a nasoseptal flap provides a much less invasive approach providing there is accurate localization of the leak. This approach may be augmented with the use of soft tissue (such as fascia lata graft or a vascularized nasoseptal flap), tissue glues, and lumbar CSF drainage with good effect. The leak can be identified using fine-​cut bone kernel CT scan, contrast-​enhanced CSF studies on CT/​MRI (iodinated contrast or gadolinium delivered via the lumbar cistern), or fluorescein instillation into the lumbar cistern to allow intraoperative localization. Lateral skull base leaks can be more challenging and may require surgical approaches such as middle cranial fossa craniotomy, transmastoid or combined procedures (Phang et al., 2016). Ultimately, blind sac closure (obliteration of the Eustachian tube) can address the final common pathway for these leaks but at the cost of hearing.

Infections post CSF leak and skull fracture Skull fractures or persistent CSF leaks following head injury can increase the risk of subsequent meningitis or delayed subdural empyema. Patients with penetrating trauma or gunshot wounds are at highest risk due to contamination and breach of the dura. Management of patients with contaminated wounds requires debridement, washout, and dural repair. Intravenous antibiotics are indicated for such contaminated or infected open fractures. Skull base fractures associated with CSF leak pose a high risk of meningitis and delayed infection, and require specific attention. The incidence of meningitis post severe head injury ranges from 0.2 to 17.8% (Helling et al., 1988). The use of prophylactic antibiotics in patients with CSF leak is not supported by current literature. A Cochrane review of randomized controlled trials concluded that there is no evidence for prophylactic antibiotic use in patients with skull base fractures, whether there is evidence of CSF leakage or not (Ratilal et  al., 2015). Pneumococcal vaccination is recommended for all skull base fractures given that Streptococcus pneumoniae is a common pathogen, although the evidence based for this treatment

CHAPTER 44  Complications of head injury

is poor. In persistent cases of CSF rhinorrhoea or otorrhoea, repairing the defect is advocated to prevent delayed risk of infection.

Growing skull fractures Skull fractures in paediatric population (particularly below age 3), when associated with dural tear, can progress to enlarging skull defect over time and are termed growing skull fractures. Interposition of a fold of arachnoid between the edges of the fracture transmits CSF pulsation into this space and can lead to progressive expansion of the defect with subsequent herniation of the neural elements. The herniated brain and the formation of cystic arachnoidal elements is called a leptomeningeal cyst. The incidence of growing skull fracture is reported to be between 0.05% to 1.6% (Muhonen et al., 1995). This normally occurs in the weeks to months following a fracture. Therefore, monitoring skull fractures in the paediatric population, particularly below age 3, is recommended. Treatment of leptomeningeal cysts involves craniotomy and exposure of the dural tear. Following repair of the dural tear, the bone flap is replaced to cover any bony defect. In advanced cases, extensive reconstructive surgery or ventriculoperitoneal shunting is required to allow correction of brain herniation and the bony defect.

Chronic subdural haematoma Chronic subdural haematomas are generally considered to develop as a delayed complication of a minor traumatic brain injury. In clinical practice many patients have no recollection of trauma so other mechanisms of pathogenesis may exist. The incidence of chronic subdural is 5 per 100 000 in general population with increasing incidence linked to an ageing population (Santarius and Hutchinson, 2004; Whitehouse et  al., 2016). Chronic subdural haematomas classically occur following a long delay after head injury. This delay can range from weeks to months. An initial minor head injury, which may go unnoticed by the patient, can lead to tears in bridging veins leading to low flow bleeding and acute subdural haematoma. Liquefaction of the haematoma over time leads to progressive compression of the underlying brain. Patients may present with headaches, reduced level of consciousness, seizures, or focal neurological deficits. Symptomatic chronic subdural haematomas can be managed surgically using burr hole evacuation. Cases complicated by recurrence or significant membranes may be managed with mini-​craniotomy to allow better access to vascularized membranes that may occur over the clot. The role of steroids in management of chronic subdural haematoma remains controversial with no reported randomized trials reported in the literature. Randomized trials assessing the efficacy of steroids in chronic subdural haematoma are currently underway

Postcraniectomy complications Following craniectomy, several specific problems can occur that require special consideration. Focal neurological deficits related to the area of brain below the craniectomy defect are termed the ‘syndrome of the trephine’. This may relate to venous congestion due to kinking of the draining veins at the edge of the craniectomy or due to deranged cerebral blood regulation. These deficits may reverse following reconstitution of the integrity of the skull cavity. Hydrocephalus in the context of a craniectomy can be difficult to assess as with an open craniectomy defect, the pressure that normally accumulates within a closed skull to drive the passive resorption of

CSF, is absent. Ventriculoperitoneal shunting in the presence of a craniectomy defect is susceptible to overdrainage and sinking of the craniectomy flap. In this case, if the skull integrity can be reconstituted with cranioplasty this can sometimes aid CSF resorption, and if there are further concerns about hydrocephalus, this is a better juncture to consider CSF diversion.

Post-​traumatic encephalopathy and dementia pugilistica Post-​traumatic encephalopathy can develop many years after a traumatic brain injury (TBI). Evidence from contact sports such as boxing and rugby suggests that repeated head trauma can lead to neuropathological changes including accumulation of beta-​ amyloid, neurofibrillary tangles, cerebellar degeneration, gliosis, and cortical atrophy. Patients may present with a wide range of symptoms including affective disorders, gait disturbance, cognitive decline, and parkinsonism. This chronic traumatic encephalopathy is sometimes referred to as punch drunk syndrome, or dementia pugilistica (McKee et al., 2009). This syndrome has been reported in association with a single severe injury or repeated concussion over a long period of time, such as sustained by contact sport players, and combat personnel. The pathophysiology is complex as it does not occur in every instance, and may reflect a genetic predisposition towards tau deposition in vulnerable individuals. There are currently no therapies available for this condition.

Neurorehabilitation following head injury Rehabilitation is the process of facilitating the recovery of the patient postinjury and aims at maximizing quality of life and the patient’s ability to participate in activities of daily living, work, and leisure. Traumatic brain injury is a heterogeneous condition and the consequences of head injury are wide depending on the location of the injury and severity of head injury. The neurorehabilitation service should be delivered by a multidisciplinary team involving medical, nursing, occupational therapy, speech and language therapy, neuropsychology, physiotherapy, social work, and psychiatry. The process of rehabilitation starts with assessment of the patient to determine the nature of cognitive, emotional, and behavioural impairment, and to identify the functional consequences. A patient-​ centred goal setting process is then undertaken identifying short-​ term and long-​term goals. This is a dynamic process depending on patient’s changing needs (Holliday et al., 2007). Neuro-​rehabilitation has both acute and postacute and community components. Acute phase rehabilitation places a significant focus on the medical condition of the patient. Physiotherapy in the acute setting aims at optimizing chest function and oxygenation. This involves attention to positioning of the patient, chest physiotherapy, manual chest techniques of sputum expectoration, regular suctioning with pre-​and post-​treatment hyperoxygenation. Speech and language therapy, in the acute phase is normally focussed upon difficulties with swallowing. The incidence of swallowing disorders following TBI is unknown. Post-​ traumatic impairment in the level of consciousness and tracheostomy can lead

527

528

Section 10  Neurotrauma and intensive care

to dysphagia. Many of these patients are at increased risk of aspiration, particularly if there is any coexistent gastroparesis. Early assessment of swallowing allows an appropriate route of feeding to be established, preventing unnecessary weight loss and malnutrition. Once swallowing reflexes resume an appropriately textured diet can commence. Occupational therapy in the acute phase focuses on reduction of impairment and the prevention of complications to maximize long-​ term function. This involves evaluation of cognitive, motor, sensory skills of the patient allowing appropriate and meaningful goals to be established. Cognitive assessment establishes the level of awareness in this group of patients and facilitates their interaction with the environment. Motor and sensory assessment allows appropriate management and provision of equipment such as splints, seating, wheelchairs and various specialized positioning aids and orthoses to allow maximal function and prevention of long-​term complications of muscle disuse (Whitfield, 2009). Acute neuropsychological input aims at dealing with large array of behavioural and psychological issues caused by TBI. Neuropsychologists develop individualized behavioural management programmes and make recommendations to health professionals and staff about modifying approaches to provide the patient with the optimum environment for rehabilitation (Wilson and Zangwill, 2003). In addition, neuropsychological input can provide education to patients’ families allowing a better understanding of the patient’s behaviour and therefore enabling better coping and adaptation strategies to maximize the function of the patient. Following acute rehabilitation of the patient, the long-​term and community phases of rehabilitation ensue. This phase aims at maximizing the function of the patient on the path of recovery and is composed of similar components seen in acute phase albeit with a greater focus on promoting recovery of function, adaptation to dysfunction, and insight into psychological components with development of long-​term compensatory strategies. The interaction of the patient with their environment and others is a crucial consideration in development of rehabilitation goals during this phase of care. Physiotherapy at this stage aims at optimizing and maximizing limb function. Spasticity following head trauma is common and can have detrimental effects on patient’s limb function. A long period of immobility may also have led to fixed deformities which further limit the functional recovery of the patient. Active and passive stretching of muscles aims to prevent contracture formation. Prolonged stretching of muscles is achieved by casting to reduce spasticity (Conine et al., 1990). Further control of spasticity can be achieved by use antispasmodics such as baclofen or use of botulinum toxin. In addition, physiotherapy aims to enhance proprioceptive input by moving the patient in different positions to allow better efferent control of muscles (Allum et al., 1998). Speech and language therapy plays an important role at this stage of rehabilitation as it aims to maximize communication and interaction of the patient with others. Both verbal and non-​verbal communication is impaired frequently post head injury. Although in many cases this is due to linguistic processes of language generation, in others difficulty is due to cognitive communication disorders. Patients with cognitive deficits such as impaired concentration and attention exhibit poor communication and language skills. The speech and language therapist in conjunction can identify the communication aspects that are most disabling and employ techniques

aiming at resolving issues. In refractory cases, compensatory and adaptive techniques are used to assist the patient and their communication skills: technological aids may be helpful in these cases (Evans et al., 2009). Occupational therapy during this phase focuses on functional performance and improving independence to allow patient independence in as many daily activities as possible. This is achieved by relearning basic skills such as activities of daily living including personal hygiene, dressing, meal preparation, and shopping. In addition, occupational therapy provides aids and adaptive equipment, such as specialized seating and wheelchairs to support independent living. Restorative or adaptation techniques developed by occupational therapists involve tailoring individual treatment programmes to introduce activities that the patient normally enjoys performing and to optimize their functional capacity in living independently and in regaining their role in society. Occupational therapy therefore plays an extremely important role in transition of care from hospital to community to allow ongoing holistic recovery of the patient (Evans et al., 2009). Cognitive and neuropsychological rehabilitation is pivotal to this stage of recovery. Patients with TBI may suffer from impairments of memory, attention, executive functioning/​planning, language, or perception. Various training models for improving memory, attention, executive functioning, and visuospatial function are currently under investigation and being applied that have the potential to improve our approach to cognitive rehabilitation of patients post head injury (Whyte et al., 2009).

FURTHER READING Kammersgaard, L.P., Linnemann, M., & Tibaek, M. (2013). Hydrocephalus following severe traumatic brain injury in adults. Incidence, timing, and clinical predictors during rehabilitation. NeuroRehabilitation, 33, 473–​80. Phang, S.Y., Whitehouse, K., Lee, L., Khalil, H., McArdle, P., & Whitfield, P.C. (2016). Management of CSF leak in base of skull fractures in adults. Br J Neurosurg, 30, 596–​604. Santarius, T. & Hutchinson, P.J. (2004). Chronic subdural haematoma: time to rationalize treatment? Br J Neurosurg, 18, 328–​32. Schierhout, G. & Roberts, I. (2001). Anti-​epileptic drugs for preventing seizures following acute traumatic brain injury. Cochrane Database Syst Rev, 4, CD000173. Wilson, B.A. & Zangwill, O.L. (2003). The future of neuropsychological rehabilitation. In:  Wilson, B. (ed.) Neuropsychological Rehabilitation: Theory and Practice, pp. 293–​301. Exton, PA: Swets & Zeitlinger Publishers.

REFERENCES Agha, A., Phillips, J., & Thompson, C. J. 2007. Hypopituitarism following traumatic brain injury (TBI). Br J Neurosurg, 21, 210–​16. Alavi, S.A., Tan, C.L., Menon, D.K., Simpson, H.L., & Hutchinson, P.J. (2016). Incidence of pituitary dysfunction following traumatic brain injury: a prospective study from a regional neurosurgical centre. Br J Neurosurg, 30,  302–​6. Allum, J.H., Bloem, B.R., Carpenter, M.G., Hulliger, M., & Hadders-​ Algra, M. (1998). Proprioceptive control of posture: a review of new concepts. Gait Posture, 8, 214–​42.

CHAPTER 44  Complications of head injury

Annegers, J.F., Hauser, W.A., Coan, S.P., & Rocca, W.A. (1998). A population-​based study of seizures after traumatic brain injuries. N Engl J Med, 338,  20–​4. Baker, E.H., Wood, D.M., Brennan, A.L., Baines, D.L., & Philips, B.J. (2005). New insights into the glucose oxidase stick test for cerebrospinal fluid rhinorrhoea. Emerg Med J, 22,  556–​7. Bondanelli, M., De Marinis, L., Ambrosio, M.R., et  al. (2004). Occurrence of pituitary dysfunction following traumatic brain injury. J Neurotrauma, 21, 685–​96. Christensen, J., Pedersen, M.G., Pedersen, C.B., Sidenius, P., Olsen, J., & Vestergaard, M. (2009). Long-​term risk of epilepsy after traumatic brain injury in children and young adults: a population-​based cohort study. Lancet, 373, 1105–​10. Conine, T.A., Sullivan, T., Mackie, T., & Goodman, M. (1990). Effect of serial casting for the prevention of equinus in patients with acute head injury. Arch Phys Med Rehabil, 71, 310–​12. Downey, B., Jackson, T., Fewings, J., et al. (2009). Acute rehabilitation of the head-injured patient. In: Whitfield, P.C., Thomas, E.O., Summers, F., Whyte, M., & Hutchinson, P.J. (eds.) Head Injury: A Multidisciplinary Approach, pp. 233–61. Cambridge, UK: Cambridge University Press. Eddy, C.M., Rickards, H.E., & Cavanna, A.E. (2011). The cognitive impact of antiepileptic drugs. Ther Adv Neurol Disord, 4, 385–​407. Eisenberg, H.M., Gary, H.E., Jr., Aldrich, E.F., et al. (1990). Initial CT findings in 753 patients with severe head injury. A report from the NIH Traumatic Coma Data Bank. J Neurosurg, 73, 688–​98. Evans, J.J., Whyte, M., Summers, F., et al. (2009). Post-acute and community rehabilitation of the head-injured patient . In: Whitfield, P.C., Thomas, E.O., Summers, F., Whyte, M., & Hutchinson, P.J. (eds.) Head Injury: A Multidisciplinary Approach, pp. 233–61. Cambridge, UK: Cambridge University Press. Fabian, T.S., Woody, J.D., Ciraulo, D.L., et al. (1999). Posttraumatic carotid cavernous fistula:  frequency analysis of signs, symptoms, and disability outcomes after angiographic embolization. J Trauma, 47, 275–​81. Helling, T.S., Evans, L.L., Fowler, D.L., Hays, L.V., & Kennedy, F.R. (1988). Infectious complications in patients with severe head injury. J Trauma, 28(11), 1575–​7. Holliday, R.C., Ballinger, C., & Playford, E.D. (2007). Goal setting in neurological rehabilitation:  patients’ perspectives. Disabil Rehabil, 29, 389–​94. Jennett, B. (1973). Epilepsy after non-​missile head injuries. Scott Med J, 18,  8–​13. Jennett, B. (1975). Epilepsy After Non-​missile Head Injuries. London, UK: William Heinemann. Kammersgaard, L.P., Linnemann, M., & Tibaek, M. (2013). Hydrocephalus following severe traumatic brain injury in adults. Incidence, timing, and clinical predictors during rehabilitation. NeuroRehabilitation, 33, 473–​80. McKee, A.C., Cantu, R.C., Nowinski, C.J., et al. (2009). Chronic traumatic encephalopathy in athletes: progressive tauopathy after repetitive head injury. J Neuropathol Exp Neurol, 68, 709–​35. Michel, S.J. (2004). The Mount Fuji sign. Radiology, 232, 449–​50. Muhonen, M.G., Piper, J.G., & Menezes, A.H. (1995). Pathogenesis and treatment of growing skull fractures. Surg Neurol, 43, 367–​72; discussion  372–​3.

Paiva, W.S., De Andrade, A.F., De Amorim, R.L., et  al. (2010). The prognosis of the traumatic subarachnoid hemorrhage: a prospective report of 121 patients. Int Surg, 95,  172–​6. Phang, S.Y., Whitehouse, K., Lee, L., Khalil, H., McArdle, P., & Whitfield, P.C. (2016). Management of CSF leak in base of skull fractures in adults. Br J Neurosurg, 30, 596–​604. Ratilal, B.O., Costa, J., Pappamikail, L., & Sampaio, C. (2015). Antibiotic prophylaxis for preventing meningitis in patients with basilar skull fractures. Cochrane Database Syst Rev, 4, CD004884. Santarius, T. & Hutchinson, P.J. (2004). Chronic subdural haematoma: time to rationalize treatment? Br J Neurosurg, 18, 328–​32. Schierhout, G. & Roberts, I. (2001). Anti-​epileptic drugs for preventing seizures following acute traumatic brain injury. Cochrane Database Syst Rev, 4, CD000173. Servadei, F., Murray, G.D., Teasdale, G.M., et  al. (2002). Traumatic subarachnoid hemorrhage: demographic and clinical study of 750 patients from the European brain injury consortium survey of head injuries. Neurosurgery, 50, 261–​7; discussion 267–​9. Szaflarski, J.P., Sangha, K.S., Lindsell, C.J., & Shutter, L.A. (2010). Prospective, randomized, single-​blinded comparative trial of intravenous levetiracetam versus phenytoin for seizure prophylaxis. Neurocrit Care, 12, 165–​72. Tan, C.L., Alavi, S.A., Baldeweg, S.E., et al. (2017). The screening and management of pituitary dysfunction following traumatic brain injury in adults:  British Neurotrauma Group guidance. J Neurol Neurosurg Psychiatry, 88(11), 971–​81. Temkin, N.R., Dikmen, S.S., Wilensky, A.J., Keihm, J., Chabal, S., & Winn, H.R. (1990). A randomized, double-​blind study of phenytoin for the prevention of post-​traumatic seizures. N Engl J Med, 323, 497–​502. Vergouwen, M.D., Vermeulen, M., & Roos, Y.B. (2006). Effect of nimodipine on outcome in patients with traumatic subarachnoid haemorrhage: a systematic review. Lancet Neurol, 5, 1029–​32. Whitehouse, K.J., Jeyaretna, D.S., Enki, D.G., & Whitfield, P.C. (2016). Head injury in the elderly: what are the outcomes of neurosurgical care? World Neurosurg, 94, 493–​500. Whyte, M., Summers, F., Herbert, C., et al. (2009). Neuropsychology and head injury. In: Whitfield, P.C., Thomas, E.O., Summers, F., Whyte, M., & Hutchinson, P.J. (eds.) Head Injury: A Multidisciplinary Approach, pp. 233–61. Cambridge, UK: Cambridge University Press. Wilson, B.A. & Zangwill, O.L. (2003). The future of neuropsychological rehabilitation. In:  Wilson, B. (ed.) Neuropsychological Rehabilitation: Theory and Practice, pp. 293–​301. Exton, PA: Swets & Zeitlinger Publishers.

RELATED LINKS TO EBRAIN Cranial nerve and vascular complications of head injury. https:// learning.ebrain.net/course/view.php?id=1109 CSF leak and intracranial infection following head injury. https:// learning.ebrain.net/course/view.php?id=1115 Hydrocephalus, seizures and pituitary dysfunction following head injury. https://learning.ebrain.net/course/view.php?id=1113 Neuropathology of Head Injury. https://learning.ebrain.net/course/ view.php?id=1088 Neuropsycology and Head Injury. https://learning.ebrain.net/course/ view.php?id=1116

529

45

Concussion and sports-​related head injury Mark Wilson

Introduction Interest in concussion and sports-​related injury has intensified in recent years for three main reasons: (1) it is a preventable form of brain injury; (2) there is increasing evidence that repeated injury can result in long-​term neurocognitive loss; and (3) as a result there are potential medicolegal costs to organizations that, possibly inadvertently, allow this form of brain injury to occur within their sport. The long-​term effects of boxing resulting in dementia pugilistica have been appreciated for some time, however the results of repeated mild head injury in other sports is now under focus. The definition of concussion has changed over time. Traditionally sports concussion occurred if there was immediate and full loss of consciousness (LOC) of variable duration accompanied by post-​traumatic amnesia at least for the event. The requirements of LOC were dropped and the Congress of Neurological Surgeons defined concussion (and it has subsequently been applied to sports concussion) as: A clinical syndrome characterized by immediate and transient impairment of neural function, such as alteration of consciousness, disturbance of vision, equilibrium etc due to mechanical forces. (Gurdjian and Volis, 1966)

Others have defined concussion as: A biomechanically induced brain injury characterized by the absence of gross anatomic lesions. (Signoretti et al., 2011) Concussion is recognized as a clinical syndrome of biomechanically induced alteration of brain function typically affecting memory and orientation, which may involve loss of consciousness (LOC). (The American Association of Neurology)

Concussion can have short, medium, and long-​term sequelae. In the immediate period an assessment and decision on return to play is required. In the medium term, managing postconcussion syndrome and the risk of second impact syndrome should be recognized, with the longer-​term risks managed at an organizational or union level. Recently there has been a call to ‘retire’ the term concussion, which, as can be seen from the evolving and different definitions, is confusing. Replacing it with a term that reflects it as being on the traumatic brain injury spectrum (e.g. mild traumatic brain injury),

a term that is commonly used interchangeably with concussion anyway, is probably most appropriate (Sharp and Jenkins, 2015). However, for the purposes of this chapter, the term concussion will continue to be used.

Postconcussion syndrome Common symptoms that occur after mild traumatic brain injury (TBI) include headache, dizziness, fatigue, irritability, reduced concentration, sleep disturbance, memory impairment, anxiety, heightened sensitivity to noise and light, blurred vision, and depression. The majority of these symptoms resolve within 3 months, although in approximately one-​third of patients they persist beyond 6 months.

Impact brain apnoea Although rarely considered, early descriptions of concussion specifically mention respiratory changes and apnoea occurring (Koch and Filehne, 1874). John Hughling Jackson (1835–​1911) knew that concussion if severe enough could be fatal through cardiovascular and respiratory failure and at autopsy, the brain in such cases could be normal/​near normal. The animal literature reports that apnoea after relatively mild brain injury is common, and the greater the impact the longer the apnoea (Atkinson et al., 1998). These phenomena occur with very little or no visible structural defect on imaging or at post-​mortem. Although not the focus of this chapter, this phenomenon should receive greater focus as it may account for a significant proportion of TBI-​related death.

History and sport-​related injury The Persian physician Razes (854–​925 AD) used the term concussion to describe an abnormal physiological state of the brain, separate from severe traumatic brain injury. In the thirteenth century Lanfrancus subsequently separated commotio cerebri (transient disruption due to brain shaking) from contusio cerebri (brain injury with bruising). In the early twentieth century, many accepted the theory that explained cerebral concussion and contusions were caused by deformation of the shape of the skull causing an anaemia of the brain.

532

Section 10  Neurotrauma and intensive care

In more modern times, concussion has become commonly associated with sport. In 1893 a shoemaker fabricated a leather helmet for an American football player who had had repeated head injuries. In 1905 the American College Football season suffered 18 deaths and 159 serious injuries which came to the attention of President Roosevelt. He ushered in rules to make the game safer. The use of leather helmets increased and in 1935 these became mandatory at the high school level and in 1939, mandatory at the college level. Between 1931 and 1940 the number of football fatalities dropped from 33 to 11. In 1950s the outer plastic rigid shell of the helmet was invented, however this enabled the head to be used as a battering ram—​the protective equipment becoming a weapon. With this, fatalities increased (244 between 1961 and 1970) 75% of which were from cerebral haemorrhages. Alteration of rules and increased awareness of this between 1987 and 2008 has resulted in direct football fatalities for all levels of play in the single digits per year (Mueller and Cantu, 2011). Following a high-​profile brain injury, Washington and many other states now have the Zackery Lystedt Law (named after the individual who had the TBI). In this the key provisions are: (1) education for parents, athletes, and coaches; (2) immediate removal from play during game or practice after a suspected concussion with no return to play until (3); (3) written clearance by a concussion expert for return to play; and (4) uniformity of rules for all schools who use public land (Ellenbogen, 2014). Sport-​related brain injury was historically the domain of neurosurgeons. For example, in 1941 Walter Dandy patented the first protective baseball cap insert liner (Fox, 1984). However, recent decades have seen intense interest from emergency medicine, sports physicians, and other health professionals, each bringing important aspects of diagnosis and management to this field. Despite this, the regular management of all forms of TBI by neurosurgeons provides a unique experience into the patterns of parenchymal injury and their likely sequelae. Neurosurgeons have a role to lead and be accountable in this field. The association of concussion with sport has resulted in a definition specific to sport-​ related concussion. The International Symposia on Concussion in Sport define sports concussion as:  A complex pathophysiological process affecting the brain, induced by traumatic forces. Several common features include: 1. Can be caused by direct blow to head, face, neck, or elsewhere with impulsive force transmitted. . Usually results in rapid onset of short-​lived functional neuro2 logical impairment. . While neuropathological changes can occur, acute symptoms are 3 neurological functional impairment rather than structural. . Can be graded and may or may not involve LOC. 4 . Grossly normal neuroimaging studies. 5 . Occasionally symptoms are prolonged. 6 There is nothing specific that separates sport-​related concussion from concussion due to other mechanisms (e.g. falling off a bike) hence again this separate definition and classification only adds more confusion. Of course, some sports can result in considerably more injury than mild TBI or concussion. For example, the forces involved in motorsport are more likely to result in immediately life-​threatening injuries.

The pathophysiology of concussion Aetiology of head injury in sport The results of brain injury in sport have to reflect a combination of laws of physics with the specifics of neuroanatomy (and the function of the areas affected). The head can undergo sudden acceleration/​ deceleration events causing shearing between intracranial structures. This occurs at high velocity such as in motor-​racing and horse riding in which there is external force applied to the individual. Slower velocities result where the head hits the floor or other players during tackles in contact sports such as rugby, American football and soccer, and similarly causes shearing forces. While these may often be milder, they are more likely to recur resulting in cumulative injuries. In contrast to acceleration/​deceleration of the head, other sports have a propensity to cause focal injuries where objects strike a (relatively) stationary head. This can occur with hard balls for example in golf and cricket. The type of injury can be quite different. Concussion causes a temporary disturbance of proper functioning of the brain. If allowed to return to play, he/​she may appear to play normally (especially if a regular player) with relative preservation of cerebellar and motor reflex actions. However, questioning may reveal confusion or disorders of memory. John Hughling Jackson described an intermediate or ‘middle’ centre which would correlate with a midbrain-​diencephalic region including the upward projections of the reticular activating system and thalamocortical projections that maintain arousal and alertness. It is likely that LOC occurs with transverse shearing forces (produced by angular accelerating rather than linear forces) specifically disrupting this area. British Neurologists Derek Denny-​ Brown (1901–​ 1981) and W. Ritchie Russell (1903–​1980) demonstrated that acceleration/​deceleration induced concussions in cats and monkeys caused stunning without macroscopic or microscopic lesions. Holding the head rigid when the blow was delivered reduced concussion (Denny-​Brown and Ritchie Russell, 1940). Shortly after this, it was demonstrated that nerve cell loss and chromatolysis within the brainstem reticular activating system occurred in primate concussion (Gurdjian, 1975). Sabina Strich (a British neurologist) characterized extensive white matter axonal injury in those severely concussed (Strich lesions). These patients survived because of improved immediate critical care. It became apparent that neuronal disruption not previously appreciated was occurring. Bryan Jennet (1926–​2008) understood that this diffuse concussional injuries could have cumulative effects such as those that occur in boxing. Ayub Ommaya tested Holbourn’s hypothesis that annular (rotational) acceleration was more damaging through shearing forces than linear acceleration and impact forces which caused less harmful compressive and tensile (pulling) forces. Non-​impact isolated translational linear acceleration had a much greater threshold to produce concussion and resulted in predominantly focal lesions (Ommaya and Gennarelli, 1974). In the 1950s and 1960s, Richard Schneider’s group extensively investigated concussion and patterns of injury in monkeys visualizing the brain through a transparent clear lexan calvarium surgically replacing skull. He suggested that instantaneous respiratory or cardiac arrest may be secondary to compression of the vertebral arteries

CHAPTER 45  Concussion and sports-related head injury

between the occipital condyles and C1; or direct transaxial transmission of force to the skull, cervicomedullary junction and high cord, or shear force damage caused by the tethering of the upper cervical cord by the dentate ligaments (Schneider et al., 1970).

Functional vs. structural injury Traditionally concussion has been thought of as a ‘functional’ injury because of the lack of macroscopic pathology demonstrated on imaging such as CT scanning. However, there is increasing evidence that there is microstructural injury underlying the functional impairment. While not detectable on CT, injury can often be seen using advanced MRI techniques such as susceptibility weighting and diffusion tensor imaging. Advanced MRI techniques taken before and after a single season of college hockey suggest the presence of scattered cerebral microhaemorrhages and white matter injury in those who experienced concussion (Helmer et al., 2014; Pasternak et al., 2014; Sasaki et al., 2014). At a cellular level, biomechanical injury results in potassium efflux, sodium and calcium influx, depolarization (and hence indiscriminate glutamate release), and a diffuse ‘spreading depression-​like’ state that may account for acute postconcussive symptoms (Katayama et al., 1992). To correct this cellular ionic imbalance, adenosine triphosphate (ATP)-​requiring membrane ion pumps work excessively causing hyperglycolysis, depleting ATP, and increasing ADP. In the setting of reduced cerebral blood flow this leads to metabolic insufficiency. The energy crisis results in free radical generation and more damage or possibly underlie susceptibility to further injury. It has been suggested that following the hyperglycolysis period, a state of impaired metabolism occurs lasting 7 to 10 days and is associated with some postconcussive memory symptoms (Giza and Hovda, 2014). At a cytoskeletal level, microstructures such as dendritic arbours and astrocytic processes can be disrupted and calcium influx can result in neurofilament side arm phosphorylation and collapse breaking down axons. Axonal stretch disrupts microtublues that alters bidirectional axonal transport altering synaptic neurotransmitter release. Diffuse axonal injury has axonal disconnection and microhaemorrhages resulting in ‘moderate’ and ‘severe’ TBI. Stretching and bending of axons as may occur in mild TBI or concussion can disrupt axonal microstructure. Axolemmal permeability increases after experimental TBI (Pettus et al., 1994). Fluid percussion models injuring the white matter tracts of the corpus callosum demonstrate unmyelinated axons are more vulnerable to injury (Reeves et al., 2005). This, in addition to anatomical factors altered shearing forces may explain greater susceptibility of some brain areas to injury. There is evidence that TBI results in altered glutamate (N-​Methyl-​D-​aspartate) receptor subunit composition altering neurotransmission, calcium influx and both intra and extracellular cell signalling (Atkins et  al., 2009). This could have functional (e.g. memory) consequences. There is a dropout of hilar GABAergic interneurons following lateral fluid percussion injury in rats (Zanier et al., 2003). All of these micropathologies could explain slowed cognition. Inflammatory changes also occur in mild TBI with up-​regulation of cytokines and inflammatory genes as well as microglial infiltration. The latter is associated with damage to the substantia nigra with implications specifically for TBI-​related parkinsonism (Hutson et al., 2011). The location of these injuries clearly influences

functional symptomatology (hippocampus → memory; amygdala → fear/​anxiety).

Markers of concussion The chronic sequelae of repeated concussions are classically seen in boxers, but they are becoming attributed to other contact sports as well. Aggregated tau protein is considered the pathological hallmark of ‘chronic traumatic encephalopathy’. While human studies indicating a dose–​response correlation are lacking, there is little doubt that boxers develop hippocampal atrophy and ventriculomegaly, strongly implying that cell death has occurred. This appears to be accelerated in elderly TBI. N-​acetylaspartate (NAA), a brain-​specific compound may be surrogate marker of post-​traumatic biochemical damage, and monitor concussion recovery being detectable through proton magnetic resonance spectroscopy. NAA levels are reduced in the frontal lobes of concussed adults, returning to normal over 30 days except where second injury occurs delaying return for 45 days (Vagnozzi et al., 2008).

Genetic vulnerability Channelopathies, such as the CACNA1A mutation which is associated with familial hemiplegic migraine is also associated with an exaggerated response to mild TBI (Kors et al., 2001) and may underlie different postconcussive response to injury, or even susceptibility to second impact syndrome. In animal studies, transgenic animals (3×TG-​ApoE4) show increased tau protein accumulation after TBI (Tran et al., 2011). The role of ApoE4 in TBI and Alzheimer’s is gradually becoming apparent (Wilson and Montgomery, 2007).

Imaging concussion Standard MRI techniques can falsely reassure that there is no structural injury. Gradient Echo and susceptibility weighted imaging can demonstrate small microbleeds (Fig. 45.1). Diffusion tensor imaging can demonstrate specific white matter tract disruption. While these have previously been mainly research techniques, they should now be considered clinically for those with persisting concussion symptoms. Positron emission tomography imaging can be used to demonstrate B-​amyloid and tau pathology which may be of clinical use in the future.

Symptoms of concussion The classification of concussion has evolved since the first mild, moderate and severe grades proposed by Thorndike in 1948 (Thorndike, 1948). Many others including Schneider (Schneider and Kriss, 1969)  and Kelly (Kelly et  al., 1991)  have proposed other systems with the most recent being a revised version by Cantu (Table 45.1) (Cantu, 2001). Grading concussion on the duration of LOC and amnesia, does not truly reflect the nature or severity of the underlying injury, though on average on a population basis, longer unconsciousness probably reflects greater injury.

Management of concussion Figures vary, but concussion usually resolves for most patients within a few weeks of injury.

533

534

Section 10  Neurotrauma and intensive care

T1

FLAIR

SWI

Fig. 45.1  A microbleed is clearly identified on susceptibility weighted MRI, but not seen on standard T1 of fluid attenuated inversion recovery.

Immediate management Removal from play Prior to removal from play, a sports club should undertake preseason planning (specific to sport and venue), assessment and education such that removal (or not removal) from play decisions are not made without forethought. Planning comprises immediate and long-​term plans incorporating emergency management of immediately life-​threatening head injuries. Assessment of individuals requires a comprehensive history of previous concussions and pre-​ existing conditions that can modify concussion (e.g. migraine, depression, and anxiety). Sideline assessment The most universal assessment tool is the Sport Concussion Assessment Tool, or SCAT (3rd edition) (http://​bjsm.bmj.com/​content/​47/​5/​259.full.pdf). This is the accepted tool for many organizations such as FIFA and the Olympic committee. It is suitable for those over 13 (there is a Child SCAT for those under). Within the initial assessment, if the player lost consciousness, has balance or incoordination issues, is disorientated or confused, has memory loss, or looks blank or vacant, then they should be withdrawn from play and not return for the day. There is a Glasgow Coma Scale (GCS) and Maddocks score component. Beyond this is a symptom evaluation including cognitive assessment (orientation, memory, and concentration), neck, balance, and coordination examinations. Return to play At the 4th International Conference on Concussion in Sport (2012), it was unanimously agreed that no return to play on the day of Table 45.1  Revised classification of concussion by Cantu Grade I

No loss of consciousness, posttraumatic amnesia/​ postconcussion signs or symptoms 40 years) Body distribution • Focal (e.g. writer’s cramp, blepharospasm) • Segmental (involving two or more contiguous regions, craniocervical dystonia) • Multifocal (two or more non-​contiguous regions) • Generalized (+/​–​leg involvement) • Hemidystonia Temporal pattern • Disease course • Static • Progressive • Variability • Persistent • Action specific • Diurnal • Paroxysmal Associated features Isolated dystonia or combined with another movement disorder • Isolated dystonia • Combined dystonia Occurrence of other neurological or systemic manifestations • List of co-​occurring neurological manifestations Axis II. Aetiology • Nervous system pathology • Evidence of degeneration • Evidence of structural (often static) lesions • No evidence of degeneration or structural lesion • Inherited or acquired • Inherited • Autosomal dominant • Autosomal recessive • X-​linked recessive • Mitochondrial Acquired • Perinatal brain injury • Infection • Drug • Toxic • Vascular • Neoplastic • Brain injury • Psychogenic Idiopathic • Sporadic • Familial Reproduced with permission from Alberto Albanese, Kailash Bhatia, Susan B. Bressman, et al., Phenomenology and classification of dystonia: A consensus update, Movement Disorders, Volume 28, Issue 7, Copyright © 2013 John Wiley and Sons.

Epidemiology There is estimated to be approximately 70 000 adults and children in the United Kingdom who suffer with some form of dystonia, which equates to a prevalence of 1 in 900. Though difficult to quantify accurately, according to minimum prevalence estimates, primary dystonia should be considered the third most frequent movement disorders after ET and Parkinson’s disease.

Genetics Genetic forms of the disease are important to recognize clinically and also provide valuable information about possible pathogenic mechanisms within the wider disorder. Hereditary dystonia is classified either by the gene underlying the condition or by reference to an ever-​expanding list of dystonia loci, of which there are currently over 20. The DYT loci system were assigned in chronological order based on appearance of reports in the medical literature and in theory once the underlying genetic cause was known, the locus was supposed to be removed. However, in practice this has not occurred and thus dystonia loci and gene names appear interchangeably (e.g. DYT1/​TOR1A) (Charlesworth et al., 2013). DYT1/​TOR1A is responsible for early onset dystonia and is inherited in an autosomal dominant fashion with reduced penetrance. The offspring of an affected individual or an asymptomatic individual known to have a TOR1A mutation have a 50% chance of inheriting the disease-​causing mutation and 30–​40% chance of developing symptoms.

Pathogenesis The underlying pathogenesis for dystonia remains unclear. Interestingly in DYT1 dystonia most of the studies report no overt neurodegeneration or cell loss. One study has described ubiquitin and torsin A-​positive inclusions within the brainstem. Functional MRI studies have implicated the pedunculopontine nucleus as part of dystonia ‘metabolic network’. In secondary dystonias the pathology is linked to basal ganglia structures, though several studies point to a cerebellar outflow problem. While progress has been made to pin down the exact pathophysiology much work remains to be done to refine these hypotheses.

Diagnosis Though diagnosing dystonia can be difficult due to its variability in presentation, wide aetiology, and overlap with other movement disorders, it is a clinical diagnosis and thus best assessed by a neurologist with a special interest in movement disorders. Family history is important as up to 44% of patients have relatives with similar movement disorders. A careful drug history to rule out secondary causes is also vital. Blood tests including liver function, ceruloplasmin, copper studies may be appropriate. Neuroimaging is useful to exclude structural causes and where appropriate to study the anatomy for potential deep brain stimulation. Genetic screening for DYT and other mutations and genetic counselling are vital for patients with early onset primary dystonia or a relative affected with condition. Assessment for dystonia is usually performed using one of several validates rating scales:

879

880

Section 16  Functional neurosurgery

• BFM-​DRS (Burke-​Fahn-​Marsden Dystonia Rating Scale) is a 120-​point rating scale used to test the severity of dystonia in nine body regions. This scale takes into account the severity and frequency of the dystonic movements. A higher score means greater impairment. • CDIP-​58 (Cervical Dystonia Impact Profile) includes 58 evaluative items forming eight scales. High scores indicate worse health. • TWSTRS (Toronto Western Spasmodic Torticollis Rating Scale) is another commonly used rating scale for cervical dystonia. This exam consists of three sections: severity (range 0–​30), disability (range 0–​30), and pain (range 0–​40).

Medical therapy Depending on the severity of symptoms, a multidisciplinary approach to the management of dystonia is required comprising a combination of medications, allied healthcare (physiotherapy, psychological therapy, speech, and occupational therapy) and where appropriate surgery. The use of rehabilitative physiotherapy for focal dystonias can be very effective with an aim to restore and maintain as much independence as possible. Cognitive behavioural therapy, autogenic training and counselling have an important role in the management of depression and anxiety which are often associated with the condition. Muscle spasms and abnormal posturing can cause significant pain and may require referral to pain management programmes. Where speech difficulties are present, referral to a speech and language therapist should be offered. Medications are commonly utilized to provide symptomatic relief, though there is no one drug that is the definitive medication of choice and often a combination of several drugs is required to enable effective management. • Levodopa—​in primary dystonia, a trial of levodopa is vital to rule out dopa-​responsive dystonia which can be managed in the long term on levodopa. • Trihexyphenidyl—​is usually the first line of management, though anticholinergic side effects need to be monitored. • Baclofen is useful but can cause drowsiness, drooling, and droopiness of the trunk. • Benzodiazepines—​ limited by respiratory depression and addiction. • Clonidine—​can cause somnolence. • Gabapentin—​may have additional benefit in stabilizing mood and sleep quality. • Botulinum toxin A injection—​especially useful for focal dystonia (e.g. blepharospasm, writer’s cramp). It can be injected EMG/​ ultrasound assisted. Used off-​licence in children, it is often used to treat pain and spasm in specific muscle groups. • Intrathecal baclofen—​this can prove effective for regional and total body dystonia with less risks of somnolence associated with oral baclofen. Other support that may be considered includes dietary support, occupational therapy input, podiatry (especially for gait), genetic counselling, and social support.

Surgical management Surgical approaches to movement disorders are an important consideration in the management of medically refractory cases. Increasingly, given their potential to avoid long-​term complications of the disorder (e.g. deformity and contractures in dystonia), surgery is being used at earlier stages in the disease process. Contemporary functional neurosurgical techniques trace their origins to Dr Irving Cooper when he inadvertently ligated the anterior choroidal artery in a patient with PD resulting in serendipitous improvement in tremor and rigidity (Cooper, 1953). This was largely superseded by specific lesioning techniques which conferred beneficial results. However, concerns about irreversibility and complications associated with ablation, especially when done bilaterally, has resulted in high-​frequency DBS becoming the gold standard surgical procedure for movement disorders.

Lesioning This procedure involves acquiring stereotactic images and then passing an insulated radiofrequency electrode to basal ganglia nuclei coagulating a small area within the target nucleus. The effects are usually immediate. These techniques have largely been superseded by DBS due to concerns over irreversibility. Its use has dwindled over the decades; however, this ablative procedure is still a useful procedure in the armamentarium of a functional neurosurgeon. It has the advantage of being cheap, having a low infection profile, is relatively safe, and avoids repeated visits for adjustment of stimulation parameters or battery changes. The recent introduction of the high-​intensity focused ultrasound (HIFU) technology, allowing incision-​less lesioning of target nuclei under the guidance of real time MRI thermometry, is likely to further improve the safety profile of lesioning surgery.

Stimulation DBS involves identification of basal ganglia target nuclei utilizing high-​resolution imaging, especially MRI, and subsequent implantation of permanent electrode leads using stereotactic techniques. The procedure is usually performed under local anaesthetic to allow for neuro-​physiological and clinical confirmation of optimal electrode placement. The leads are then connected subcutaneously via extension wires to an implantable pulse generator (IPG) which is implanted in the chest/​abdominal wall. The IPG serves both as a battery and a programming device, amenable to adjustment externally to alter the electrical output from the leads.

Transplantation Transplantations and restorative therapies for PD have a long history, dating to the 1970s, but are still only in use as part of clinical trials. In a randomized study of 40 patients with severe PD receiving either a transplant of human embryonic dopamine-​neurons or a sham surgery, it was shown that the transplanted cells survived and resulted in some clinical benefit in younger but not older patients. Furthermore, clinical benefit and graft viability were sustained up to 4  years after transplantation and the imaging changes reliably correlated with clinical outcome over the post-​transplantation time

CHAPTER 76  Movement disorders

course. In contrast, in another prospective 24-​month double-​blind placebo-​controlled trial of fetal nigral transplantation in 34 patients with advanced PD, a slight treatment effect in patients with milder disease was observed. Some 56% of the transplanted patients however developed off-​phase dyskinesia (Olanow et  al., 2003). These dyskinesias were also described by other groups and can persist up to 11 years. Currently neural transplantation remains as an experimental therapy. Further studies are necessary to obtain objective evidence of benefit, define target patient population, determine the optimal cells for transplant and the exact brain target area for transplantation.

Surgical outcomes Parkinson’s disease There is good long-​term evidence on effectiveness of DBS to improve motor symptoms in advanced PD with significant benefits in tremor, rigidity, and bradykinesia. Typically, improvements in general motor state as measured by UPDRS-​III of 50%, reduction in off-​periods during the day by 70%, and prolongation of periods with good mobility of approximately 20% are expected (Alamri et al., 2015). Overall, tremor and rigidity show better long-​term improvement than bradykinesia. Axial symptoms that respond to the Levodopa usually also improve. Furthermore, medication-​related complications like motor fluctuations and dyskinesias decrease by 60–​70% after STN DBS. Despite these benefits, the natural neurodegenerative course of PD does not appear to be altered by DBS and, thus in the long term, worsening of axial symptoms and akinesia should be expected. Cognitive decline is also unchanged reflecting the natural course of the disease. Nevertheless, a consistent marked motor improvement especially of tremor, rigidity, motor fluctuations, and dyskinesias have been reported in studies up to 5–​10 years after STN DBS. Levodopa equivalent dose of parkinsonian daily medications can be reduced by 40–​60% which is maintained after 5 years. Recently it has been shown that motor improvements after STN DBS is not limited to patients with advanced PD. The randomized, prospective multicentre EARLYSTIM trial demonstrated favourable outcomes in the neurostimulation group. Given these results, especially early at the onset of disabling motor fluctuations, DBS has been advocated at an earlier stage in the disease process than the current practice. However, long-​term outcome in this cohort is awaited (Schuepbach et al., 2013). Non-​motor symptoms (NMS) are common in PD patients and can have dramatic impact on patients’ quality of life. Though DBS literature has primarily focused on motor symptoms, evidence indicates that STN DBS can improve NMS in selected patients (Witt et al., 2008).

Essential tremor The short-​and long-​ term benefits of DBS have been repeatedly demonstrated with significant resolution of upper extremity tremor reported in up to 90% of patients. Several studies have compared the clinical outcomes and complications related to DBS and thalamotomy in ET. Both methods produce near complete resolution of contra-​lateral upper extremity tremor in approximately 60–​ 90% of patients (Zhang et al., 2010). Long-​term benefits are similar

between both methods, with significant lasting improvement seen in the majority of patients although with both therapies, over time there may be a reduction in benefit in some patients. The primary difference between the two therapies is an increase in serious neurological complications associated with thalamotomy compared to DBS. Thalamotomy is currently only considered in selected patients with medication resistant disabling tremor who would not be good candidates for DBS either for medical issues precluding tolerance of the DBS hardware, known need for repeated MRI scanning, inability to comply with follow-​up appointments for DBS programming and costs. In rare cases, an ET patient may not be able to tolerate DBS or thalamotomy due to increased surgical risk secondary to other medical conditions. In these cases, radiosurgical thalamotomy may be considered.

Dystonia Ablative stereotactic surgery to treat dystonia has been performed in both the thalamus and the GPi. Outcomes have been difficult to assess due to lack of standardization of inclusion criteria, assessments, and surgical technique. Nevertheless, results of thalamotomy for primary and secondary dystonia have shown 25–​100% improvement in 10/​20 idiopathic cases and 25–​100% improvement in 68% of 29 patients with secondary dystonia. However, significant side effects (15% hemiparesis and up to 56% dysarthria/​dysphonia) have limited the use of this procedure. Buzaco has reported almost the same efficacy of pallidotomy compared to thalamotomy but less frequent complications with subsequent reports, confirming the efficacy of pallidotomy in alleviating the dystono-​dyskinetic symptoms in patients with primary or secondary dystonia. No long-​lasting complication is documented in any of these reports (Speelman et al., 2010). The GPi is currently the target of choice for DBS surgery in dystonia, though thalamic and subthalamic nucleus stimulation have also been utilized. In primary generalized dystonia, a significant improvement of the motor and disability scores of the BFMRS has been reported (range 40–​80%). Similar results have been described in patients with segmental dystonia. With a follow-​up as long as 10 years, the pallidal DBS therapeutic effects appear stable. Patients with secondary dystonia represent a heterogeneous population and though it is accepted that, in general, secondary dystonia does not respond to DBS as well as primary dystonia, several reports showed satisfactory clinical outcome in many secondary dystonia, such as tardive dystonia. The discrepancy in the post DBS outcomes in this patient population when using validated movement disorder scales versus those measuring the quality of life/​patient-​reported outcomes, has encouraged the increasing use of the latter.

Controversies DBS has been firmly established as an effective intervention for movement disorders for nearly three decades and thus controversies chiefly relate to differences in surgical technique rather than the efficacy: • General vs. local anaesthesia—​good outcome from DBS relies on accurate lead position within a very small, deep target. An awake patient allows for intraoperative macrostimulation through the

881

882

Section 16  Functional neurosurgery

DBS lead to assess the optimal position for symptom relief and thresholds for avoiding side effects, and thus allows for immediate re-​positioning of the electrode to a more effective target. However, an awake patient is required to be cognitively intact to perform mental tasks, cooperative and remain still for prolonged periods while in the ‘off-​state’. These latter concerns are removed with general anaesthesia; however, lead positioning is then reliant only on imaging targeting methods and intraoperative microelectrode recording. There has been no randomized controlled trial comparing outcomes for both techniques. • Microelectrode recordings (MER)—​ is an electrophysiological technique of verifying the borders of a target nucleus using its firing pattern and characteristic signals. Recording is performed by inserting very fine tipped electrodes. However, MER necessitates expertise, expense, and additional operating time to an already prolonged procedure. There is also evidence that MER increases the risk of intracerebral haemorrhage. With improvements in surgical technique allowing for accurate image guided ‘direct’ targeting of the nuclei, the continued routine use of MER has been questioned by some surgeons. • STN vs. GPi—​The debate over whether to use GPi or STN DBS as the target for PD has stemmed from various studies that have shown varying degrees in outcome. Overall opinion over the past 10 years have been in favour of STN DBS for the majority of DBS candidates, while favouring GPi stimulation in the older patients, those with predominant dyskinesia, or those at risk of neuropsychiatric disturbance. More recent studies have shown comparable outcomes with both targets. • Duodopa vs. DBS—​Maintaining steady plasma concentrations of oral levodopa presents problems due to gastric emptying which results in fluctuating motor symptoms. With a reliable pharmacological profile and a portable delivery system, the use of intrajejunal L-​dopa/​carbidopa gel (Duodopa) has become an option in patients with motor fluctuations. Studies comparing Duodopa with DBS are limited, though one comparing the two interventions showed that the DBS group had significantly more improvement in dyskinesias duration and disability compared to duodopa (Merola et al., 2011). Although duodopa appears to be a promising alternative for PD patients, particularly in whom DBS is contraindicated, the large complication rates related to PEG tube insertion cannot be dismissed. The issue of costs is also an important one with duodopa being significantly more expensive than DBS.

FURTHER READING Alamri, A., Ughratdar, I., Samuel, M., & Ashkan, K. (2015). Deep brain stimulation of the subthalamic nucleus in Parkinson’s disease 2003–​2013: where are we another 10 years on? Br J Neurosurg, 29(3), 319–​28. Albanese, A., Bhatia, K., Bressman, S.B., et al. (2013). Phenomenology and classification of dystonia:  a consensus update. Mov Disord, 28,  863–​73 Calabresi, P., Picconi, B., Tozzi, A., Ghiglieri, V., & Di Filippo, M. (2014). Direct and indirect pathways of basal ganglia: a critical reappraisal. Nat Neurosci, 17, 1022–​30. Chaudhuri, K.R., Healy, D.G., Schapira, A.H., & National Institute for Clinical Excellence (2006). Non-​motor symptoms of Parkinson’s disease: diagnosis and management. Lancet Neurol, 5,  235–​45

Schuepbach, W.M., Rau, J., Knudsen, K., et al. (2013). Neurostimulation for Parkinson’s disease with early motor complications. N Engl J Med, 368, 610–​22.

REFERENCES Alamri, A., Ughratdar, I., Samuel, M., & Ashkan, K. (2015). Deep brain stimulation of the subthalamic nucleus in Parkinson’s disease 2003–​2013: where are we another 10 years on? Br J Neurosurg, 29(3), 319–​28. Albanese, A., Bhatia, K., Bressman, S.B., et al. (2013). Phenomenology and classification of dystonia: a consensus update. Mov Disord, 28, 863–​73. Benito-​Leon, J. & Louis, E.D. (2006). Essential tremor: emerging views of a common disorder. Nat Clin Pract Neurol, 2, 666–​78; quiz 2p following 691. Bertoni, J.M., Arlette, J.P., Fernandez, H.H., et al. (2010). Increased melanoma risk in Parkinson disease: a prospective clinicopathological study. Arch Neurol, 67, 347–​52. Calabresi, P., Picconi, B., Tozzi, A., Ghiglieri, V., & Di Filippo, M. (2014). Direct and indirect pathways of basal ganglia: a critical reappraisal. Nat Neurosci, 17, 1022–​30. Charlesworth, G., Bhatia, K.P., & Wood, N.W. (2013). The genetics of dystonia: new twists in an old tale. Brain, 136, 2017–​37. Chaudhuri, K.R., Healy, D.G., Schapira, A.H., & National Institute for Clinical Excellence (2006). Non-​ motor symptoms of Parkinson’s disease: diagnosis and management. Lancet Neurol, 5, 235–​45. Cooper, I.S. (1953). Anterior chorodial artery ligation for involuntary movements. Science, 118, 193. Dauer, W. & Przedborski, S. (2003). Parkinson’s disease: mechanisms and models. Neuron, 39, 889–​909. Deng, H., Le, W., & Jankovic, J. (2007). Genetics of essential tremor. Brain, 130, 1456–​64. Dorsey, E.R., Constantinescu, R., Thompson, J.P., et  al. (2007). Projected number of people with Parkinson disease in the most populous nations, 2005 through 2030. Neurology, 68,  384–​6. Goetz, C.G. & Pal, G. (2014). Initial management of Parkinson’s disease. BMJ, 349, g6258. Helmich, R.C., Toni, I., Deuschl, G., & Bloem, B.R. (2013). The pathophysiology of essential tremor and Parkinson’s tremor. Curr Neurol Neurosci Rep, 13, 378. Hughes, A.J., Daniel, S.E., Kilford, L., & Lees, A.J. (1992). Accuracy of clinical diagnosis of idiopathic Parkinson’s disease:  a clinico-​ pathological study of 100 cases. J Neurol Neurosurg Psychiatry, 55,  181–​4. Klein, C. & Fahn, S. (2013). Translation of Oppenheim’s 1911 paper on dystonia. Mov Disord, 28, 851–​62. Little, S., Pogosyan, A., Kuhn, A.A., & Brown, P. (2012). Beta band stability over time correlates with Parkinsonian rigidity and bradykinesia. Exp Neurol, 236,  383–​8. Liu, B. & Dluzen, D.E. (2007). Oestrogen and nigrostriatal dopaminergic neurodegeneration:  animal models and clinical reports of Parkinson’s disease. Clin Exp Pharmacol Physiol, 34, 555–​65. Lorenz, D., Poremba, C., Papengut, F., Schreiber, S., & Deuschl, G. (2011). The psychosocial burden of essential tremor in an outpatient-​ and a community-​based cohort. Eur J Neurol, 18,  972–​9. Merola, A., Zibetti, M., Angrisano, S., Rizzi, L., Lanotte, M., & Lopiano, L. (2011). Comparison of subthalamic nucleus deep brain stimulation and Duodopa in the treatment of advanced Parkinson’s disease. Mov Disord, 26, 664–​70.

CHAPTER 76  Movement disorders

Olanow, C.W., Goetz, C.G., Kordower, J.H., et  al. (2003). A double-​ blind controlled trial of bilateral fetal nigral transplantation in Parkinson’s disease. Ann Neurol, 54, 403–​14. Riederer, P. & Laux, G. (2011). MAO-​inhibitors in Parkinson’s Disease. Exp Neurobiol, 20,  1–​17. Schrag, A. & Quinn, N. (2000). Dyskinesias and motor fluctuations in Parkinson’s disease. A community-​based study. Brain, 123 (Pt 11), 2297–​305. Schuepbach, W.M., Rau, J., Knudsen, K., et al. (2013). Neurostimulation for Parkinson’s disease with early motor complications. N Engl J Med, 368, 610–​22. Speelman, J.D., Contarino, M.F., Schuurman, P.R., Tijssen, M.A., & De Bie, R.M. (2010). Deep brain stimulation for dystonia: patient selection and outcomes. Eur J Neurol, 17 Suppl 1, 102–​6. Stocchi, F. & Olanow, C.W. (2003). Neuroprotection in Parkinson’s disease: clinical trials. Ann Neurol, 53 Suppl 3, S87–​97; discussion S97–​9. Vlaar, A.M., Van Kroonenburgh, M.J., Kessels, A.G., & Weber, W.E. (2007). Meta-​analysis of the literature on diagnostic accuracy of SPECT in parkinsonian syndromes. BMC Neurol, 7, 27. Weintraub, D., Siderowf, A.D., Potenza, M.N., et al. (2006). Association of dopamine agonist use with impulse control disorders in Parkinson disease. Arch Neurol, 63, 969–​73. Witt, K., Daniels, C., Reiff, J., et al. (2008). Neuropsychological and psychiatric changes after deep brain stimulation for Parkinson’s

disease:  a randomised, multicentre study. Lancet Neurol, 7, 605–​14. Zesiewicz, T.A., Elble, R., Louis, E.D., et  al. (2005). Practice parameter: therapies for essential tremor: report of the Quality Standards Subcommittee of the American Academy of Neurology. Neurology, 64, 2008–​20. Zhang, K., Bhatia, S., Oh, M.Y., Cohen, D., Angle, C., & Whiting, D. (2010). Long-​term results of thalamic deep brain stimulation for essential tremor. J Neurosurg, 112, 1271–​6.

RELATED LINKS TO EBRAIN Classification and Epidemiology of Movement Disorders. https:// learning.ebrain.net/course/view.php?id=327 Diagnosis and Treatment of Essential Tremor. https://learning.ebrain. net/course/view.php?id=315 Dystonia: Classification and Genetics. https://learning.ebrain.net/ course/view.php?id=310 Parkinson’s Disease: Definition, Pathogenesis and Pathology. https:// learning.ebrain.net/course/view.php?id=326 Parkinson’s Disease: Surgical Treatment. https://learning.ebrain.net/ course/view.php?id=323

883

77

Spasticity John Goodden, Catherine Hernon, and Brian Scott

Introduction This chapter reviews spasticity causes and treatments in adults and children. Due to the diversity of this subject, and the age range covered, it should be noted that some treatments covered in this chapter are only applicable for a limited group of patients. Spasticity is the term used to describe an involuntary, velocity-​ dependent increase in muscle tone leading to increased resistance to movement. It arises as a result of upper motor neurone lesions (UMNL), which lead to loss or diminution of descending inhibitory influence on the alpha and gamma motor neurones. This is demonstrated by changes within the reflex arc, with loss of inhibitory input, leading to an exaggerated reflex. Depending upon the cause of the UMNL, the effects of spasticity may be seen in the upper or lower limbs. Traditional descriptions of spasticity have used terms such as hemiplegia, diplegia, and quadriplegia to describe the distribution of the spasticity and the limbs involved. However, neurorehabilitation experts recommend that these terms are not used as they falsely suggest that the other limbs are unaffected, when they actually may be mildly affected. It is therefore preferred to state the mainly affected limbs (e.g. ‘spasticity mainly affecting the legs’ rather than ‘diplegia’).

Pathophysiological basis of spasticity The pathophysiological changes in the brain and spinal cord that lead to spasticity are incompletely understood. However, it is thought to arise mainly from injury to the reticulospinal and corticospinal/​pyramidal tracts and the spinal interneurons. The injury leads to a reduction in the inhibitory input to the spinal motor circuits with a consequent exaggeration of the reflex arc. Enhanced H-​M ratios and F-​wave amplitudes on neurophysiology testing provide evidence of this increase in motor neurone excitability. Within the spinal cord, changes are seen in the excitability of the motor neurones, the interneurons, and reflex arc pathways. In the healthy subject, the 1a inhibitory interneurons mediate the reflex arc, receiving descending input from the descending corticospinal pathways as part of this action. Spasticity is seen when the interneurons are damaged, as well as when the descending input is reduced.

In the affected muscles, shortening and contracture commonly develop over time, but muscle spindle sensitivity is usually unaltered.

Clinical aspects of spasticity—​aetiology The causes of spasticity are congenital (cerebral palsy, spinal cord tethering/​ dysraphism) or acquired (stroke, trauma, tumour, demyelinating conditions, progressive neurodegenerative diseases, infections, and hydrocephalus and syringomyelia). Since spasticity arises from an UMNL, it would not be expected to result from a spinal injury below the conus medullaris. If the patient has suffered an acute UMNL, muscle tone is usually low with a flaccid paresis initially, before progressing to spasticity. Cerebral palsy is the commonest cause of paediatric spasticity and multiple sclerosis (MS) is the commonest in adults. Cerebral palsy has an incidence of 1 in 500 live births. It is a non-​ progressive condition, with a variety of clinical manifestations, including spasticity, dystonia, dyskinesia, and ataxia. While the brain lesion is static, the effects on the musculoskeletal system are often progressive because the muscles grow more slowly than the adjacent bones and also because of an imbalance of forces across the joints. These have deleterious effects on the shape of the developing bones and joints. Multiple sclerosis is an immune-​mediated, progressive degenerative condition with a variety of courses including relapsing-​remitting and gradually/​rapidly progressive. It occurs principally in people of a Northern-​European origin, has a peak onset around age 30, and a female:male ratio of up to 3:1. The UK incidence has recently been estimated at 9.64 per 100 000 population, with a prevalence in England and Wales of 100–​140 per 100 000 population, increasing to 170 in Northern Ireland, 190 in Scotland, 295 in Shetland, and 402 in Orkney.

Investigation, assessment, and diagnosis of spasticity Patients presenting with spasticity need referral to and workup by a skilled physician, with a thorough assessment. This includes a detailed history and examination, baseline blood tests (including FBC/​

886

Section 16  Functional neurosurgery

UE) and blood/​urine cultures to look for infection. Family history is particularly important for children when considering neurodegenerative conditions. Specialist investigations such as cerebrospinal fluid (CSF) analysis, MRI scanning, and EMG/​EEG may also be required. When a patient presents with a sudden exacerbation of their symptoms, the clinician needs to again undertake a detailed history and examination, blood tests, and urine cultures. Common factors causing deterioration include infection (especially urine or ear), medication changes, pain, pressure sores, constipation, occult fracture, foot/​orthotic problems, DVT, stress, fatigue, and even cold weather. A careful multidisciplinary team (MDT) assessment is required, including evaluation of which muscles are overactive and how the spasticity affects the patient throughout their daily life including self-​care, mobility, sleep, and employment. A  variety of scoring systems have been developed to document the extent of spasticity and functional limitation. The modified Ashworth Scale is the most commonly used tool to grade the extent of spasticity (Platz et al., 2005)  (Table 77.1), but others such as the Tardieu score are also used. In addition, documentation of range of movement, muscle power, quality of life, function, and gait (if relevant) are required (Haley et al., 1992; Steinbok, 2001; Platz et al., 2005; Haugh et al., 2006; O’Brien and Park, 2006; Engsberg et al., 2007; Harvey et al., 2007: Khan, 2007; Delhaas et al., 2008; Gorton et al., 2009; Akerstedt et al., 2010, Gilmore et al., 2010; Ou et al., 2010; Dasenbrock et al., 2011; Reynolds et al., 2011; Tedroff et al., 2014). These assessments should be performed ahead of any planned intervention so the effects of the intervention can be evaluated. For children with cerebral palsy, measures such as the Gross Motor Function Classification System (GMFCS) and Gross Motor Function Measure (GMFM) are useful tools to document and summarize a child’s functional capabilities. The GMFCS is graded from 1 (independent) to 5 (wheelchair-​bound) (see Fig. 77.1 and Table 77.2). The GMFM is a standardized observational instrument that has been designed and validated to measure change over time, helping to summarize this with a single score that can then be compared to chart their progress. Table 77.1  The modified Ashworth Scale Score

Description

Examination

0

No increase

Moves freely at any velocity

1

Slight increase

Catch and release or minimal resistance at end range

1+

Slight increase

Catch and release with minimal resistance through remainder of range of motion

2

More marked increase

Moderate resistance throughout range, but joint is easily moved

3

Considerable increase

High tone is present—​joint is difficult to move

4

Severe increase

Affected joints are rigid-​cannot be moved from flexed or extended position

Reproduced with permission from Bohannon, Richard W.; Smith, Melissa B, Interrater Reliability of a Modified Ashworth Scale of Muscle Spasticity, Physical Therapy, Volume 67, Issue 2, pp.206–207, Copyright © 1987 by permission of Oxford University Press on behalf of the American Physical Therapy Association.

Medical management of spasticity (including rehabilitation, physiotherapy, and occupational therapy) The MDT managing spasticity should include a rehabilitation physician, physiotherapists, occupational therapy, and a variety of surgeons (neurosurgery, plastic surgery, orthopaedic surgery) (Park and Albright, 2006; Steinbok, 2006; Abbott, 2007; Ronan and Gold, 2007). It is important to carefully balance the control of spasticity against any side effects of doing so. Spasticity may be providing a positive impact in the patient’s life as well as a negative one. For example, the spastic stiffness in a patient’s legs may counteract the underlying weakness and allow them to walk; removing this spasticity may render them immobile. Functional help can be provided via communication aids, splints, customized walking frames, and customized wheelchairs. The aim is to protect function and avoid/​reduce the harmful effects of the spasticity and disability as a whole. Regular physiotherapy is central to this, to maintain flexibility and avoid the progression of muscle/​ tendon contracture formation. Occupational therapy and orthotic reviews are necessary to protect skin, and optimize splints and supports, maximizing function, and preventing pressure sore development. Finally, good bladder care is also vital to prevent recurrent infections (urinary tract infection is very common cause of spasticity exacerbation). A variety of medical interventions can be undertaken to help limit the deleterious effects of spasticity. These treatments aim to improve tone, alleviate spasms, control gastrointestinal effects, and control bladder function. Commonly used antispasticity medications include baclofen, dantrolene, tizanidine, and diazepam (Papavasiliou, 2009; Quality Standards Subcommittee of the American Academy of Neurology et al., 2010; Novak et al., 2013). Their side effects are varied but can include drowsiness, weakness, fatigue, sedation, and memory impairment. For these reasons they can be unpopular, especially if patients/​carers read up about them prior to taking them. Baclofen is generally considered the mainstay of spasticity treatment, especially in cerebral palsy and spinal cord lesions. It acts via the pre-​and postsynaptic GABAB receptors, reducing excitatory transmission of the α-​motor neuron and decreasing nociception. It has good gastrointestinal (GI)-​tract absorption but poor transfer across the blood–​brain barrier. Therefore, high doses may be required before clinical effects are seen, creating a higher risk of side effects—​including sedation, behavioural changes, confusion, ataxia, urinary frequency, and insomnia. Baclofen is started at low dose and gradually increased until benefits are maximized or side effects become troublesome. If it is being discontinued, this must be done gradually due to the risk of seizures, rebound hyperspasticity, or hallucinations. Dantrolene is used for treatment of spasticity, cramps, and spasms. It acts by reducing depolarization-​induced calcium influx into sarcoplasmic reticulum of striated muscle. Its GI absorption is incomplete, slow but consistent. Side effects include muscle weakness (making mobility more difficult), sedation, and hepatitis.

CHAPTER 77 Spasticity

GMFCS E & R between 6th and 12th birthday: Descriptors and illustrations GMFCS Level I

Children walk at home, school, outdoors and in the community. They can climb stairs without the use of a railing. Children perform gross motor skills such as running and jumping, but speed, balance and coordination are limited.

GMFCS Level II

Children walk in most settings and climb stairs holding onto a railing. They may experience difficulty walking long distances and balancing on uneven terrain, inclines, in crowded areas or confined spaces. Children may walk with physical assistance, a handheld mobility device or used wheeled mobility over long distances. Children have only minimal ability to perform gross motor skills such as running and jumping.

GMFCS Level III

Children walk using a hand-held mobility device in most indoor settings. They may climb stairs holding onto a railing with supervision or assistance. Children use wheeled mobility when traveling long distances and may self-propel for shorter distances.

GMFCS Level IV

Children use methods of mobility that require physical assistance or powered mobility in most settings. They may walk for short distances at home with physical assistance or use powered mobility or a body support walker when positioned. At school, outdoors and in the community children are transported in a manual wheelchair or use powered mobility.

GMFCS Level V

Children are transported in a manual wheelchair in all settings. Children are limited in their ability to maintain antigravity head and trunk postures and control leg and arm movements.

Fig. 77.1  The Gross Motor Function Classification System.

887

888

Section 16  Functional neurosurgery

Table 77.2  The Gross Motor Function Classification System GMFCS level

General descriptor

1

Walks without limitations

2

Catch walks with limitations

3

Walks using a hand-​held mobility device

4

Self-​mobility with limitations; may use powered mobility

5

Transported in a manual wheelchair

Reproduced with permission from Robert Palisano, Peter Rosenbaum, Stephen Walter, et al., Development and reliability of a system to classify gross motor function in children with cerebral palsy, Developmental Medicine & Child Neurology, Volume 39, Issue 4, pp. 214–​23. Copyright © 2008 John Wiley and Sons.

Tizanidine is a central α2-​adrenergic agonist. It is commonly used for spasticity in MS, stroke, and spinal cord injury. It is short-​ acting, with peak effects 1–​2 h after administration; it wears off by 3–​6 h. As a result, it is generally reserved for particular activities and times when relief of spasticity is most important. Its pharmacokinetics differ between different preparations and also when taken with/​without food, being better when taken under fasting conditions (>3 h after food). As with other agents, its dose can be gradually increased over several days, and withdrawal should be similarly gradual. Diazepam acts via the GABAA receptors to increase the postsynaptic inhibition of α-​motor neurons, treating spasticity and muscle spasms. It is most useful in patients with complete spinal cord injuries. Doses should be increased or decreased incrementally. When used for more than one week, problems with drug-​dependency can occur. Its side effects include weakness, sedation, and poor stamina. Abruptly stopping may cause depression, seizures, and an acute withdrawal syndrome. Botulinum toxin A  therapy (Criswell et  al., 2006; Steinbok, 2006; Love et  al., 2010; Quality Standards Subcommittee of the American Academy of Neurology et  al., 2010; Williams et al., 2012; Novak et al., 2013; Cusick et al., 2015) is a neurotoxin, that permanently binds to the presynaptic terminal at the neuromuscular junction causing inhibition of acetylcholine release, functionally denervating the treated muscle. Administration is via direct injection, often with ultrasound guidance to improve accuracy and duration of effect. The effects are seen within 3–​7  days and it lasts up to 16 weeks; wearing off as new neuromuscular junctions grow from sprouting nerve rootlets. If used too frequently, it can become less effective due to antibody formation. Care must be taken over the dose; the benefits can be limited if too little is used, but overdose can also occur, leading to systemic effects and widespread muscle paresis. It is licensed for use in upper limb spasticity in adults, bladder instability & overactive bladder, and cervical dystonia. It is widely used for upper and lower limb spasticity in children but is unlicensed for these purposes. It is best used in combination with stretching, serial casting, muscle re-​education and strengthening, and orthotic adjustments.

Surgical procedures for spasticity Several surgical strategies are employed to help combat the adverse effects of spasticity. These involve several different disciplines. Neurosurgical techniques include intrathecal baclofen (ITB), selective dorsal rhizotomy (SDR), percutaneous radiofrequency foraminal rhizotomy, and myelotomy. In addition to these, selective peripheral neurotomy (SPN) is offered by plastic surgeons. Orthopaedic surgery also has a major role to play due to the complex bone and joint abnormalities that can develop due to the spasticity. Constraints of space limit us to a focus on ITB, SDR, peripheral neurotomy, and an outline of orthopaedic surgery.

Intrathecal baclofen ITB therapy is a well-​established treatment for adults and children to treat both spasticity and dystonia. It was reported by Penn and Kroin in 1985 (Penn and Kroin, 1985). They postulated and then proved that direct administration of baclofen into the spinal CSF would achieve the same benefits of oral baclofen with fewer side effects due to direct administration and smaller dosage. This is because the GABAB receptors in the spinal cord are in the surface layers, allowing the baclofen to work directly. However, in the brain, the GABAB receptors are in the deeper layers, meaning there are fewer unwanted central effects. When considering ITB therapy, a test dose is normally performed first (Albright and Ferson, 2006; Delhaas et al., 2008; Phillips et al., 2015). This is done under general or local anaesthetic, with either direct injection of 50–​100 µg of baclofen via lumbar puncture, or insertion of a lumbar catheter to allow serial test doses or a test-​ infusion to be undertaken. The aims of the test dose are to assess the benefits of the treatment and to examine for possible side effects. Patients with dystonia, however, may not experience much benefit from the brief test dose because they tend to get more benefit from the sustained effects of prolonged infusion. ITB therapy is delivered via an externally-​adjustable programmable pump system, which contains a drug reservoir (10–​40 ml), a battery, and the drug delivery mechanism. For the surgery, the patient is placed in the lateral position, right-​side-​up. The pump is implanted in a pocket in the right anterior abdominal wall. Left-​sided ITB implants are avoided as gastrostomies are usually inserted on this side. The pump pocket is either subcutaneous, (superficial to the rectus fascia), or subfascial, (deep to the fascia of the rectus and external oblique muscles). The subfascial pocket is particularly useful for smaller, thinner children (Vender et al., 2006; Motta et al., 2007; Fjelstad et al., 2009; Ammar et al., 2012; Motta and Antonello, 2014). Once the pocket has been made, a small incision is made in the lumbar region and the intrathecal catheter is inserted via a lumbar puncture with a Toohey needle. There is debate in the literature about the catheter tip position with authors divided over whether a higher tip position is required for upper limb spasticity/​dystonia or not (Grabb et al., 1999; Albright and Ferson, 2006; Dziurzynski et al., 2006; McCall and MacDonald, 2006; Motta et  al., 2007; Brennan and Whittle, 2008; Sivakumar et  al., 2010; Ughratdar et  al., 2012; Varhabhatla and Zuo, 2012). Some studies have suggested that CSF

CHAPTER 77 Spasticity

pulsatile flow is less in the thoracic and lumbar spine, with a consequent reduction in baclofen diffusion for catheters placed in this location (Bernards, 2006; Heetla et al., 2014). The tip position can be guided with X-​ray imaging during surgery. In some cases, catheters may be inserted into the ventricle to allow intraventricular baclofen infusion (Albright, 2011; Rocque and Albright, 2011; Bollo et  al., 2012; Turner et al., 2012). Once the catheter is inserted to the desired position, the catheter is tunnelled through to the abdominal pocket, secured to the pump and the system is then implanted (Fig. 77.2). After implantation, the pump is activated and the system can start delivering baclofen immediately. The dose is gradually titrated upwards, usually with increments of 10% every 1–​2 weeks. The ITB dose regime can be tailored with flexible dose programming delivering different amounts of baclofen at different times of the day. The pump reservoir needs to be emptied and refilled every 2–​6 months depending on the infusion rate and the drug concentration. The system will also need surgery to replace the pump when the battery is nearing end of life. Risks and unwanted effects of ITB therapy include the risks of the surgery itself, and risks associated with pump system failure (a)

(c)

of problems with the refill procedures (Albright and Ferson, 2006; Vender et  al., 2006; Motta et  al., 2007; Protopapas et  al., 2007; Brennan and Whittle, 2008; Fjelstad et al., 2009; Heetla et al., 2014; Motta and Antonello, 2014). • Drug underdose/​ acute withdrawal:  Like any mechanical device, an ITB pump can fail, leading to acute baclofen withdrawal syndrome—​a life-​threatening condition resulting in severe rebound spasticity, rhabdomyolysis, renal failure, and death. This can also occur if the wrong drug concentration is used when refilling the system. • Drug overdose can occur due to errors in programming or pump refilling. When refilling the pump, it is vital to ensure that the needle is inserted into the pump itself and has not skimmed alongside the pump, allowing the refill to be delivered to the pump pocket (a pocket-​fill) rather than the pump. A pocket-​fill is a life-​ threatening situation that requires immediate aspiration of drug from the pocket and may require supportive ventilation on intensive care unit (ICU) for the patient until the baclofen wears off. (b)

(d)

Fig. 77.2  Intraoperative photographs to illustrate insertion of an intrathecal baclofen pump. (A) A subfascial pocket is made via a right upper quadrant incision. (B) The lumbar catheter is inserted via a mid-​lumbar lumbar puncture. (C) The catheter is connected to the pump. (D) The pump inserted with the catheter coiled behind.

889

890

Section 16  Functional neurosurgery

Selective dorsal rhizotomy (SDR) SDR is a technique for permanently reducing/​removing abnormal tone in the legs. It is predominantly indicated for children rather than adults. It involves dissecting and transecting a portion of the lumbar sensory nerve roots to disrupt the abnormal reflex feedback loop. The procedure is done for cerebral palsy, where the brain injury is stable and not progressive. If SDR is performed for a progressive neurodegenerative condition, the benefits will be short-​lived. It is also best performed when the only motor disorder is spasticity—​if dystonia is present it will usually worsen after SDR. The first descriptions of dorsal rhizotomy were in 1898, with Sherrington’s experiments in a cat model of cerebral palsy (Sherrington, 1898). Foerster then reported the first human subjects in 1913 (Foerster, 1913). However, these two series involved complete transection of lumbar sensory roots, causing numbness and proprioceptive deficits, and hence increased mobility difficulty. Gros (Gros et al., 1967) and then Fasano (Fasano et al., 1978) investigated this further, before Peacock (Peacock and Eastman, 1981; Peacock and Arens, 1982) heralded the modern era of selective dorsal partial rhizotomy, using neurophysiology to select the most abnormal roots prior to sectioning them. The major developments since then were clarification of patient selection criteria (Peacock et al., 1987) and refining the surgery to a two-​level laminectomy (Park et al., 1993), then single-​level laminectomy at L1 (Park and Johnston, 2006). Careful MDT assessment is required when considering SDR. The typical patient for SDR is a child with confirmed cerebral palsy, and dynamic spasticity predominantly affecting the legs that interferes with their mobility. They are usually GMFCS level 2–​3, mobile with aids, and need adequate strength in their core muscles and leg muscles.

(a)

(c)

SDR is not usually recommended for patients with dystonia, and those with significant upper limb involvement. Its role for the more disabled GMFCS 4–​5 group is under debate. The surgery is performed via an L1 laminectomy over the conus. The L1–​S1 dorsal (sensory) roots are separated from the motor roots. Each root is divided into three rootlets, tested with intraoperative neurophysiology to find the most abnormally active rootlets and then two out of every three rootlets are divided (Fig. 77.3). Therefore, approximately two-​thirds of each sensory root is divided from L1 to S1. Anal sphincter monitoring is commonly used to help reduce risk of incontinence. Once the spasticity is removed/​reduced by SDR, the underlying muscle weakness is revealed. This requires intensive physiotherapy to build strength. After SDR, patients require 3 weeks’ intensive daily physiotherapy, prior to discharge. After this, ongoing regular community physiotherapy is required and patients are encouraged to remain active. Risks and unwanted effects of SDR include the standard risks of any neurosurgical intradural spinal procedure, and of course this includes risks of paralysis, numbness, and incontinence. However, the use of intraoperative neurophysiology helps reduce this risk to a minimum (Hays et al., 1998; Mittal et al., 2001; Steinbok et al., 2009; Turner, 2009; Fukuhara et al., 2011). SDR outcomes have been widely studied and reported in the literature, confirming long-​term benefits from this procedure both for spasticity control and also reducing rates of orthopaedic surgery requirement too (Bolster et al., 2013; Dudley et al., 2013; Hurvitz et al., 2013; Langerak et al., 2013; Steinbok, 2013; Tedroff et al., 2014; Josenby et al., 2015; Steinbok, 2015; Vermeulen and Becher, 2015). The percentage of rootlets cut in SDR is linked to the degree of

(b)

(d)

Fig. 77.3  Intraoperative photographs to illustrate the technique of selective dorsal rhizotomy via a single-​level laminectomy over the conus. (A) The right-​sided dorsal roots of L2-​S1 have been separated from the ventral roots and placed into a green silastic sling. (B) The first rootlet of L2 is tested with bipolar stimulation via custom-​made nerve hooks. (C) The second rootlet of L2 is tested. (D) The rootlet delivering the most abnormal response is divided.

CHAPTER 77 Spasticity

spasticity control, with less than 50% rootlet section associated with increased return of spasticity.

Upper limb spasticity and selective peripheral neurotomy (SPN) Upper limb spasticity is characterized by a combination of spastic and flaccid paralysis. Typically, there is spasticity of the flexor pronator muscles combined with flaccidity of the opposite extensor supinator muscles. This results in an adducted, internally rotated shoulder, a flexed elbow, a pronated forearm, a flexed and ulnarly deviated wrist, and a clasped hand with a thumb-​in-​palm deformity. The flexor pronator tonicity increases with passive stretch and is constant, except during sleep or under anaesthesia. The hypertonic muscles cocontract when the antagonists are used affecting the grasp and release ability. MDT management aims to maintain function and range of movement. Non-​operative measures of physiotherapy, splinting, tone reducing medication, and botulinum toxin should be used fully before considering surgical intervention. The surgical options include rebalancing (tendon transfers), contracture release (myofascial release, musculotendinous lengthening, joint releases), joint stabilizing procedures, and tone reducing procedures, otherwise known as SPN or selective denervation (Zancolli, 2003). SPN is only used in cases where excessive hypertonia is impeding function. Botulinum toxin treatment mimics the potential effects of SPN on selected nerves and should be used as a preliminary treatment before considering SPN. Local anaesthetic blocks in the outpatient setting can also be used to assess the outcome of potential neurotomy. SPN can be used in isolation or in combination with tendon transfers, releasing and stabilizing procedures. The surgery involves sectioning of both afferent and efferent pathways of the stretch reflex at the level of the neuromuscular junction. The afferent pathway, corresponding to proprioception of the muscle concerned, is lost and sectioning of the efferent pathway induces paralysis. The aim is to reduce excessive hypertonicity without suppressing useful muscular tone or impairing residual motor and sensory functions. Neurotomy should never involve a mixed nerve trunk of sensory and motor fibres as this can cause deafferentation pain. The responsible motor branches are isolated using intraoperative nerve stimulation together with the operating microscope for visualization. Variable portions (50–​80%) of the isolated motor branches are resected and cauterized to prevent regrowth of fibres. The extent of partial section is dependent upon the response to electrical stimulation (Sindou et al., 2007). SPN outcome depends upon accurately identifying the hypertonic muscles preoperatively and precisely partially denervating these muscles. Recurrence of spasticity can occur when the amount of sectioning is insufficient, however the operation can be repeated.

Orthopaedic procedures Spastic cerebral palsy is one of the commonest conditions seen in children’s orthopaedic practice.

Initially the muscles are functionally short because of spasm or contraction but with time the muscle/​tendon unit becomes structurally short from contracture. This also happens to joints—​after prolonged movement restriction the joint capsule becomes contracted resulting in abnormal, usually dysfunctional, postures such as fixed knee flexion. Joint instability culminating in dislocation is relatively common in children with spastic cerebral palsy, for instance grossly overpronated (flat) feet and hip subluxation and dislocation (Ounpuu et al., 2002; Pirpiris et al., 2003; Soo et al., 2006; Ross and Engsberg, 2007; Robin et al., 2009; Shore et al., 2012; Silva et al., 2012). Antispasticity management and physiotherapy are well-​ established treatments during the precontracture phase but once contractures have developed orthopaedic surgery is needed to restore more normal anatomy and, hopefully, function. There are two main groups of children: those who can walk and those who cannot. The management of the latter group concentrates on prevention of hip dislocation, maintaining feet in a position to wear shoes comfortably and maintaining a comfortable seating position. The management of the former is to maintain or restore standing and walking function. The role of orthopaedic surgery in walking children has changed following more widespread availability of SDR in the United Kingdom currently offered to those walkers with a GMFCS score of 2–​3. Formerly multilevel soft tissue releases (MLSTR) was the mainstay of treatment in walking children when function was plateauing or declining as they grew taller and heavier (Gough et al., 2004; Graham et al., 2005; Khan, 2007; Gorton et al., 2009; Akerstedt et al., 2010; Harvey et al., 2012). This would be for both contraction and contracture. Typically, the child would present with a crouch or jumper’s gait (hips adducted, hips and knees flexed, ankles in calcaneus or equinus), which is an energy-​sapping way to stand and walk. Children would undergo orthopaedic surgery between the ages of about 7 and 13 years but management is highly individualized. Typically, hip flexors (sartorius and psoas), knee flexors (gracilis, semitendinosus, semimembranosus medially, and sometimes biceps femoris laterally), and ankle plantarflexors (gastrocnemius or Achilles tendon) are lengthened in one session. Minimal splintage (usually a lightweight below-​knee cast for four weeks) is used and early intensive rehabilitation is fundamental. Children should be assessed in a multidisciplinary clinic and if SDR is done first the child has a further orthopaedic assessment afterwards. Often some areas of tightness will have improved so that any further orthopaedic surgery is less than would otherwise have been the case. The optimal timing of staged orthopaedic surgery is not defined except to say that if further surgery is required if it is best performed within a few months of SDR to facilitate rehabilitation. Although there is little written, selective percutaneous myofascial lengthening (‘percs’) is popular method of MLSTR in the United States having advantages of smaller skin incisions and more rapid rehabilitation as well as causing less muscle scarring.

Controversy At present, the main source of controversy in spasticity management concerns the role of SDR for children with spasticity. Concern has

891

892

Section 16  Functional neurosurgery

been voiced because it is irreversible and can make function worse. Rigorous patient selection is therefore key. Today, most authors accept the role of SDR for children with moderate levels of disability (GMFCS levels 2 and 3)—​for whom it can improve function and mobility. Even for this group, there remains controversy. The main critique of the SDR literature is that studies are usually single-​centre, small volume, and not case-​controlled. This surgery is low-​volume and highly specialist which makes larger, case-​control studies a challenge. Some authors also cite a lack of long-​term outcome data, despite there being an increasing number of reports of up to 26-​year follow-​up (Langerak et al., 2011; Bolster et al., 2013; Dudley et al., 2013; Hurvitz et al., 2013; Langerak et al., 2013; Steinbok, 2013; Tedroff et al., 2014; Josenby et al., 2015; Steinbok, 2015; Vermeulen and Becher, 2015). Others cite a supposed high risk of paralysis or incontinence, which has not been proven in the literature. However, some also advocate SDR for more disabled and even wheelchair-​bound children (GMFCS levels 4 and 5) as a replacement for ITB therapy (Gump et al., 2013; Ailon et al., 2015; Steinbok, 2015; Vermeulen and Becher, 2015). For this group, the indications and goals of treatment are different because these children will often not achieve independent walking. They are also more likely to have mixed spasticity and dystonia, which can make the outcome less certain. Some advocate offering SDR instead of ITB for GMFCS level 4–​5 children to help with personal hygiene and ease of movement, because SDR means that they will not require the regular hospital reviews and pump refills/​replacements over the subsequent years. This has not been proven to be the case, however, and is certainly an area for future clinical trial research. For this group SDR can also remove/reduce their ability to stand for transfers with associated negative consequences for quality of life. In addition, there is interest in whether SDR could have a role in genetic conditions (Kai et al., 2014), acquired injuries (Langerak et al., 2014; Reynolds et al., 2014) (e.g. spinal trauma or postinfectious), as well as patients with hemiplegia (Oki et al., 2010; Gump et al., 2013).

Conclusions Spasticity treatment requires carefully planned multidisciplinary assessment and treatment. Treatment strategies should be tailored around the patient’s functional and physical needs, with input from allied health professionals with rehabilitation, medical, and surgical disciplines. Regular reviews are required as part of this to ensure that treatments remain suited to the individual. The treatment goals include improving tone, alleviating spasms, controlling gastrointestinal effects, and controlling bladder function, with the aim of protecting function and minimizing the deleterious effects of spasticity and their condition as a whole.

FURTHER READING Ailon, T., Beauchamp, R., Miller, S., et al. (2015). Long-​term outcome after selective dorsal rhizotomy in children with spastic cerebral palsy. Childs Nerv Syst, 31, 415–​23. Fairhurst, C. (2011). Cerebral palsy: the whys and hows. Arch Dis Child Educ Pract Ed, 97(4), 122–​31. Harvey, A., Rosenbaum, P., Hanna, S., Yousefi-​ Nooraie, R., & Graham, K.H. (2012). Longitudinal changes in mobility following

single-​event multilevel surgery in ambulatory children with cerebral palsy. J Rehabil Med, 44, 137–​43. Novak, I., McIntyre, S., Morgan, C., et al. (2013). A systematic review of interventions for children with cerebral palsy:  state of the evidence. Dev Med Child Neurol, 55, 885–​910. Sindou, M.P., Simon, F., Mertens, P., & Decq, P. (2007). Selective peripheral neurotomy (SPN) for spasticity in childhood. Childs Nerv Syst, 23, 957–​70.

REFERENCES Abbott, R. (2007). Editorial on ‘The management of childhood hypertonia’. Childs Nerv Syst, 23, 937–​41. Ailon, T., Beauchamp, R., Miller, S., et al. (2015). Long-​term outcome after selective dorsal rhizotomy in children with spastic cerebral palsy. Childs Nerv Syst, 31, 415–​23. Akerstedt, A., Risto, O., Odman, P., & Oberg, B. (2010). Evaluation of single event multilevel surgery and rehabilitation in children and youth with cerebral palsy-​-​A 2-​year follow-​up study. Disability & Rehabilitation, 32,  530–​9. Albright, A.L. (2011). Technique for insertion of intraventricular baclofen catheters. J Neurosurg Pediatr, 8,  394–​5. Albright, A.L. & Ferson, S.S. (2006). Intrathecal baclofen therapy in children. Neurosurg Focus, 21, e3. Ammar, A., Ughratdar, I., Sivakumar, G., & Vloeberghs, M.H. (2012). Intrathecal baclofen therapy-​how we do it. J Neurosurg Pediatr, 10, 439–​44. Bernards, C.M. (2006). Cerebrospinal fluid and spinal cord distribution of baclofen and bupivacaine during slow intrathecal infusion in pigs. Anesthesiology, 105, 169–​78. Bollo, R.J., Gooch, J.L., & Walker, M.L. 2012. Stereotactic endoscopic placement of third ventricle catheter for long-​term infusion of baclofen in patients with secondary generalized dystonia. J Neurosurg Pediatr, 10,  30–​3. Bolster, E.A.M., Van Schie, P.E.M., Becher, J.G., Van Ouwerkerk, W.J.R., Strijers, R.L.M., & Vermeulen, R.J. (2013). Long-​term effect of selective dorsal rhizotomy on gross motor function in ambulant children with spastic bilateral cerebral palsy, compared with reference centiles. Dev Med Child Neurol, 55, 610–​16. Brennan, P.M. & Whittle, I.R. (2008). Intrathecal baclofen therapy for neurological disorders: a sound knowledge base but many challenges remain. Brit J Neurosurg, 22, 508–​19. Criswell, S.R., Crowner, B.E., & Racette, B.A. (2006). The use of botulinum toxin therapy for lower-​extremity spasticity in children with cerebral palsy. Neurosurg Focus, 21, e1. Cusick, A., Lannin, N., & Kinnear, B.Z. (2015). Upper limb spasticity management for patients who have received botulinum toxin A  injection:  Australian therapy practice. Aust Occup Ther J, 62,  27–​40. Dasenbrock, H.H., Pendleton, C., McGirt, M.J., et al. (2011). ‘Fulfilling the chief of his duties as a physician’:  Harvey Cushing, selective dorsal rhizotomy and elective spine surgery for quality of life. J Neurosurg Spine, 14,  421–​7. Delhaas, E.M., Beersen, N., Redekop, W.K., & Klazinga, N.S. (2008). Long-​term outcomes of continuous intrathecal baclofen infusion for treatment of spasticity: a prospective multicenter follow-​up study. Neuromodulation, 11, 227–​36. Dudley, R.W.R., Parolin, M., Gagnon, B., et  al. (2013). Long-​term functional benefits of selective dorsal rhizotomy for spastic cerebral palsy. J Neurosurg Pediatr, 12, 142–​50.

CHAPTER 77 Spasticity

Dziurzynski, K., Mcleish, D., Ward, M., & Iskandar, B.J. (2006). Placement of baclofen pumps through the foramen magnum and upper cervical spine. Childs Nerv Syst, 22,  270–​3. Engsberg, J.R., Ross, S.A., Collins, D.R., & Park, T.S. (2007). Predicting functional change from preintervention measures in selective dorsal rhizotomy. J Neurosurg, 106,  282–​7. Fasano, V.A., Broggi, G., Barolat-​Romana, G., & Sguazzi, A. (1978). Surgical treatment of spasticity in cerebral palsy. Childs Brain, 4, 289–​305. Fjelstad, A.-​B., Hommelstad, J., & Sorteberg, A. (2009). Infections related to intrathecal baclofen therapy in children and adults: frequency and risk factors. J Neurosurg Pediatr, 4, 487–​93. Foerster, O. (1913). On the indications and results of the excision of posterior spinal nerve roots in men. Surg Gynecol Obstet, 16, 163–​474. Fukuhara, T., Nakatsu, D., Namba, Y., & Yamadori, I. (2011). Histological evidence of intraoperative monitoring efficacy in selective dorsal rhizotomy. Childs Nerv Syst, 27(9), 1453–​8. Gilmore, R., Sakzewski, L., & Boyd, R. (2010). Upper limb activity measures for 5‐to 16‐year‐old children with congenital hemiplegia: a systematic review. Dev Med Child Neurol, 52(1),  14–​21. Gorton, G.E., Abel, M.F., Oeffinger, D.J., et al. (2009). A prospective cohort study of the effects of lower extremity orthopaedic surgery on outcome measures in ambulatory children with cerebral palsy. J Pediatr Orthop, 29,  903–​9. Gough, M., Eve, L.C., Robinson, R.O., & Shortland, A.P. (2004). Short-​ term outcome of multilevel surgical intervention in spastic diplegic cerebral palsy compared with the natural history. Dev Med Child Neurol, 46,  91–​7. Grabb, P.A., Guin-​Renfroe, S., & Meythaler, J.M. (1999). Midthoracic catheter tip placement for intrathecal baclofen administration in children with quadriparetic spasticity. Neurosurgery, 45, 833–​6; discussion  836–​7. Graham, H.K., Baker, R., Dobson, F., & Morris, M.E. (2005). Multilevel orthopaedic surgery in group IV spastic hemiplegia. JBJS, 87, 548–​55. Gros, C., Ouaknine, G., & Vlahovitch, B., & Frèrebeau, P. (1967). La radicotomie sélective postérieure dans le traitement neuro-​ chirurgical de lhypertonie pyramidale. Neurochirurgie, 13, 505–​18. Gump, W.C., Mutchnick, I.S., & Moriarty, T.M. (2013). Selective dorsal rhizotomy for spasticity not associated with cerebral palsy: reconsideration of surgical inclusion criteria. Neurosurg Focus, 35, E6. Haley, S.M, Andrellos, P.J., Coster, W., Haltiwanger, J.T., & Ludlow, L.H. (1992). Pediatric Evaluation of Disability Inventory (PEDI): Development, standardization and administration manual. Boston, MA: New England Medical Centre Hospitals. Harvey, A., Graham, H.K., Morris, M.E., Baker, R., & Wolfe, R. (2007). The Functional Mobility Scale:  ability to detect change following single event multilevel surgery. Dev Med Child Neurol, 49,  603–​7. Harvey, A., Rosenbaum, P., Hanna, S., Yousefi-​Nooraie, R., & Graham, K.H. (2012). Longitudinal changes in mobility following single-​ event multilevel surgery in ambulatory children with cerebral palsy. J Rehabil Med, 44, 137–​43. Haugh, A.B., Pandyan, A.D., & Johnson, G.R. (2006). A systematic review of the Tardieu Scale for the measurement of spasticity. Disability & Rehabilitation, 28, 899–​907. Hays, R.M., McLaughlin, J.F., Bjornson, K.F., Stephens, K., Roberts, T.S., & Price, R. 1998. Electrophysiological monitoring during selective dorsal rhizotomy, and spasticity and GMFM performance. Dev Med Child Neurol, 40,  233–​8. Heetla, H.W., Staal, M.J., Proost, J.H., & Van Laar, T. (2014). Clinical relevance of pharmacological and physiological data in intrathecal baclofen therapy. Arch Phys Med Rehabil, 95, 2199–​206.

Hurvitz, E.A., Marciniak, C.M., Daunter, A.K., et al. (2013). Functional outcomes of childhood dorsal rhizotomy in adults and adolescents with cerebral palsy. J Neurosurg Pediatr, 11,  380–​8. Josenby, A.L., Wagner, P., Jarnlo, G.-​B., Westbom, L., & Nordmark, E. (2015). Functional performance in self-​care and mobility after selective dorsal rhizotomy: a 10-​year practice-​based follow-​up study. Dev Med Child Neurol, 57, 286–​93. Kai, M., Yongjie, L., & Ping, Z. (2014). Long-​term results of selective dorsal rhizotomy for hereditary spastic paraparesis. J Clin Neuroscience, 21, 116–​20. Khan, M.A. (2007). Outcome of single-​event multilevel surgery in untreated cerebral palsy in a developing country. JBJS, 89, 1088–​91. Langerak, N.G., Du Toit, J., Burger, M., Cotton, M.F., Springer, P.E., & Laughton, B. (2014). Spastic diplegia in children with HIV encephalopathy: first description of gait and physical status. Dev Med Child Neurol, 56, 686–​94. Langerak, N.G., Hillier, S.L., Verkoeijen, P.P., Peter, J.C., Fieggen, A.G., & Vaughan, C.L. (2011). Level of activity and participation in adults with spastic diplegia 17–​26 years after selective dorsal rhizotomy. J Rehabil Med, 43,  330–​7. Langerak, N.G., Vaughan, C.L., Peter, J.C., Fieggen, A.G., & Peacock, W.J. (2013). Letter to the editor:  long-​term outcomes of dorsal rhizotomy. J Neurosurg Pediatr, 12,  664–​5. Love, S.C., Novak, I., Kentish, M., et al. (2010). Botulinum toxin assessment, intervention and after-​care for lower limb spasticity in children with cerebral palsy: international consensus statement. Eur J Neurol, 17 Suppl 2, 9–​37. McCall, T.D. & Macdonald, J.D. (2006). Cervical catheter tip placement for intrathecal baclofen administration. Neurosurgery, 59, 634–​40; discussion 634–​40. Mittal, S., Farmer, J.P., Poulin, C., & Silver, K. (2001). Reliability of intraoperative electrophysiological monitoring in selective posterior rhizotomy. J Neurosurg, 95,  67–​75. Motta, F. & Antonello, C.E. (2014). Analysis of complications in 430 consecutive pediatric patients treated with intrathecal baclofen therapy: 14-​year experience. J Neurosurg Pediatr, 13,  301–​6. Motta, F., Buonaguro, V., & Stignani, C. (2007). The use of intrathecal baclofen pump implants in children and adolescents:  safety and complications in 200 consecutive cases. J Neurosurg, 107,  32–​5. Novak, I., McIntyre, S., Morgan, C., et al. (2013). A systematic review of interventions for children with cerebral palsy:  state of the evidence. Dev Med Child Neurol, 55, 885–​910. O’Brien, D.F. & Park, T.S. (2006). A review of orthopedic surgeries after selective dorsal rhizotomy. Neurosurg Focus, 21, e2. Oki, A., Oberg, W., Siebert, B., Plante, D., Walker, M.L., & Gooch, J. L. (2010). Selective dorsal rhizotomy in children with spastic hemiparesis. J Neurosurg Pediatr, 6,  353–​8. Ou, C., Kent, S., Miller, S., & Steinbok, P. (2010). Selective dorsal rhizotomy in children: comparison of outcomes after single-​level versus multi-​ level laminectomy technique. Can J Neurosci Nurs, 32,  17–​24. Ounpuu, S., Deluca, P., Davis, R., & Romness, M. (2002). Long-​term effects of femoral derotation osteotomies: an evaluation using three-​ dimensional gait analysis. J Pediatr Orthop, 22, 139–​45. Papavasiliou, A.S. (2009). Management of motor problems in cerebral palsy: a critical update for the clinician. EJPN, 13, 387–​96. Park, T.S. & Albright, L. (2006). Treatment of spasticity. Neurosurg Focus, 21, E7. Park, T.S., Gaffney, P.E., Kaufman, B.A., & Molleston, M.C. (1993). Selective lumbosacral dorsal rhizotomy immediately caudal to the conus medullaris for cerebral palsy spasticity. Neurosurgery, 33, 929–​33; discussion  933–​4.

893

894

Section 16  Functional neurosurgery

Park, T.S. & Johnston, J.M. (2006). Surgical techniques of selective dorsal rhizotomy for spastic cerebral palsy. Technical note. Neurosurg Focus, 21, e7. Peacock, W.J. & Arens, L.J. (1982). Selective posterior rhizotomy for the relief of spasticity in cerebral palsy. SA Med J, 62, 119–​24. Peacock, W. J., Arens, L. J., & Berman, B. 1987. Cerebral palsy spasticity. Selective posterior rhizotomy. Pediat Neurosci, 13,  61–​6. Peacock, W.J. & Eastman, R. (1981). The neurosurgical management of spasticity. S Afr Med J, 60(22), 849–​50. Penn, R.D. & Kroin, J.S. (1985). Continuous intrathecal baclofen for severe spasticity. Lancet, 326,  125–​7. Phillips, M.M., Miljkovic, N., Ramos-​ Lamboy, M., et  al. (2015). Clinical experience with continuous intrathecal baclofen trials prior to pump implantation. PM R, 7(10), 1052–​8. Pirpiris, M., Trivett, A., Baker, R., Rodda, J., Nattrass, G.R., & Graham, H.K. (2003). Femoral derotation osteotomy in spastic diplegia. Proximal or distal? JBJS, 85, 265–​72. Platz, T., Eickhof, C., Nuyens, G., & Vuadens, P. (2005). Clinical scales for the assessment of spasticity, associated phenomena, and function: a systematic review of the literature. Disability & Rehabilitation, 27,  7–​18. Protopapas, M.G., Bundock, E., Westmoreland, S., Nero, C., Graham, W.A., & Nesathurai, S. (2007). The complications of scar formation associated with intrathecal pump placement. Arch Phys Med Rehabil, 88, 389–​90. Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society, Delgado, M.R., Hirtz, D., et al. (2010). Practice parameter: pharmacologic treatment of spasticity in children and adolescents with cerebral palsy (an evidence-​based review): report of the Quality Standards Subcommittee of the American Academy of Neurology and the Practice Committee of the Child Neurology Society. Neurology, 74, 336–​43. Reynolds, M.R., Ray, W.Z., Strom, R.G., Blackburn, S.L., Lee, A., & Park, T.S. (2011). Clinical outcomes after selective dorsal rhizotomy in an adult population. World Neurosurg, 75(1),  138–​4. Reynolds, R.M., Morton, R.P., Walker, M.L., Massagli, T.L., & Browd, S.R. (2014). Role of dorsal rhizotomy in spinal cord injury-​induced spasticity. J Neurosurg Pediatr, 14, 266–​70. Robin, J., Graham, H.K., Baker, R., et al. (2009). A classification system for hip disease in cerebral palsy. Dev Med Child Neurol, 51, 183–​92. Rocque, B.G. & Albright, A.L. (2011). Intraventricular versus intrathecal baclofen for secondarydystonia: a comparison of complications. Neurosurgery, 70(2 0 0), 10.1227/​NEU.0b013e31823f5cd9. Ronan, S. & Gold, J.T. (2007). Nonoperative management of spasticity in children. Childs Nerv Syst, 23, 943–​56. Ross, S.A. & Engsberg, J.R. (2007). Relationships between spasticity, strength, gait, and the GMFM-​66 in persons with spastic diplegia cerebral palsy. Arch Phys Med Rehabil, 88, 1114–​20. Sherrington, C.S. (1898). Experiments in examination of the peripheral distribution of the fibres of the posterior roots of some spinal nerves. Part II. Phil Trans R Soc Lond B, 190, 45–​186. Shore, B., Spence, D., & Graham, H. (2012). The role for hip surveillance in children with cerebral palsy. Curr Rev Musculoskel Med, 5, 126–​34. Silva, S., Nowicki, P., Caird, M.S., et al. (2012). A comparison of hip dislocation rates and hip containment procedures after selective dorsal rhizotomy versus intrathecal baclofen pump insertion in nonambulatory cerebral palsy patients. J Pediatr Orthop, 32,  853–​6. Sindou, M.P., Simon, F., Mertens, P., & Decq, P. (2007). Selective peripheral neurotomy (SPN) for spasticity in childhood. Childs Nerv Syst, 23, 957–​70. Sivakumar, G., Yap, Y., Tsegaye, M., & Vloeberghs, M. (2010). Intrathecal baclofen therapy for spasticity of cerebral origin—​does

the position of the intrathecal catheter matter? Childs Nerv Syst, 26, 1097–​102. Soo, B., Howard, J.J., Boyd, R.N., et  al. (2006). Hip displacement in cerebral palsy. JBJS, 88,  121–​9. Steinbok, P. (2001). Outcomes after selective dorsal rhizotomy for spastic cerebral palsy. Childs Nerv Syst, 17,  1–​18. Steinbok, P. (2006). Selection of treatment modalities in children with spastic cerebral palsy. Neurosurg Focus, 21, e4. Steinbok, P. (2013). Long-​term effect of selective dorsal rhizotomy. Dev Med Child Neurol, 55(7), 586. Steinbok, P. (2015). Outcomes after selective dorsal rhizotomy. Dev Med Child Neurol, 57, 214–​15. Steinbok, P., Tidemann, A.J., Miller, S., Mortenson, P., & Bowen-​ Roberts, T. (2009). Electrophysiologically guided versus non-​ electrophysiologically guided selective dorsal rhizotomy for spastic cerebral palsy: a comparison of outcomes. Childs Nerv Syst, 25, 1091–​6. Tedroff, K., Löwing, K., & Aström, E. (2014). A prospective cohort study investigating gross motor function, pain, and health-​related quality of life 17 years after selective dorsal rhizotomy in cerebral palsy. Dev Med Child Neurol, 57(5), 484–​90. Turner, M., Nguyen, H.S., & Cohen-​Gadol, A.A. (2012). Intraventri­ cular baclofen as an alternative to intrathecal baclofen for intractable spasticity or dystonia: outcomes and technical considerations. J Neurosurg Pediatr, 10, 315–​19. Turner, R.P. (2009). Neurophysiologic intraoperative monitoring during selective dorsal rhizotomy. J Clin Neurophys, 26,  82–​4. Ughratdar, I., Muquit, S., Ingale, H., Moussa, A., Ammar, A., & Vloeberghs, M. (2012). Cervical implantation of intrathecal baclofen pump catheter in children with severe scoliosis. J Neurosurg Pediatr, 10,  34–​8. Varhabhatla, N.C., & Zuo, Z. (2012). Rising complication rates after intrathecal catheter and pump placement in the pediatric population: analysis of national data between 1997 and 2006. Pain Physician, 15,  65–​74. Vender, J.R., Hester, S., Waller, J.L., Rekito, A., & Lee, M.R. (2006). Identification and management of intrathecal baclofen pump complications: a comparison of pediatric and adult patients. J Neurosurg, 104,  9–​15. Vermeulen, R.J. & Becher, J.G. (2015). Long‐term outcome in selective dorsal rhizotomy in spastic cerebral palsy:  differentiation in mobility levels is needed. Dev Med Child Neurol, 57(5),  408–​9. Williams, G., Olver, J., De Graaff, S., & Singer, B.J. (2012). The use of botulinum toxin type A  in the management of adult-​onset focal spasticity:  a survey of Australian allied health professionals. Aust Occup Ther J, 59, 257–​64. Zancolli, E.A. (2003). Surgical management of the hand in infantile spastic hemiplegia. Hand Clinics, 19, 609–​29.

RELATED LINKS TO EBRAIN The Management of Posture and Mobility. https://learning.ebrain.net/ course/view.php?id=786 Management of Spasticity in Post Stroke Syndrome. https://learning. ebrain.net/course/view.php?id=549 Management of Spasticity and Associated Features. https://learning. ebrain.net/course/view.php?id=787 Management of Spasticity in Children. https://learning.ebrain.net/ course/view.php?id=548 Managing Late Deterioration in Neurological Disability. https:// learning.ebrain.net/course/view.php?id=539 Orthotics and the Avoidance and Management of Contractures. https://learning.ebrain.net/course/view.php?id=546

78

Pain pathophysiology and surgical management Richard Mannion and Rokas Tamosauskas

neurons with the capacity to detect noxious stimuli (e.g. heat, chemical, mechanical) are the small, unmyelinated afferent fibres, with cell bodies in the dorsal root ganglia, that innervate peripheral strucTransduction and transmission of sensory input from our envirtures such as skin, muscles, and viscera (Woller et al., 2017). Their onment to the brain starts first in primary sensory neurons (first central axons run within the dorsal root to the dorsal root entry zone order neurons). This group are anatomically and functionally heterwhere they enter Lissauer’s tract in the dorsolateral white matter fuogenous, their cell bodies located in the dorsal root and trigeminal niculus of the spinal cord. ganglia and their axons divided into a peripheral projection towards As well as the usual neurotransmitters such as glutamate, nocitheir innervation targets and a central projection into the spinal cord ceptors preferentially express several peptides that play a role in (Fig. 78.1). neurochemical transmission, such as substance P and calcitonin Primary sensory neurons can be divided into the following: large gene-​related peptide (CGRP). Following their discovery, early exmyelinated Aβ fibre afferents responsible for conveying low-​ citement that these could present new therapeutic targets for pain threshold mechanical sensation and vibration; thinly myelinated management have not been realized, as the roles of peptide transfibre Aδ afferents, with smaller cell bodies responsible for fast pain mitters are complex, but they do play an important role in sen(e.g. pinprick); and small, unmyelinated C-​fibre afferents involved sory transmission and have a distinct and longer lasting action on in pain and temperature sensation. All these cells have peripheral postsynaptic neurons in the dorsal horn, that works in conjunction terminals spread widely throughout their innervation tissue as ‘free with fast transmitters such as glutamate. In addition, these peptides nerve endings’, and transduction through highly evolved, complex can be found in the peripheral terminals of nociceptors, and anteroprocesses allows the preferential detection of peripheral stimuli that grade action potentials from the cell body to these terminals leads is strictly regulated. Nociceptors are anatomically, functionally, and to the release of peptides, that results in vasodilation, plasma exneurochemically distinct, and respond to certain stimuli through travasation and recruitment of immune cells, producing neurogenic specific receptor-​mediated mechanisms (Woolf and Ma, 2007). This inflammation. results in predictable stimulus-​response relationship within primary Most are polymodal (i.e. they can detect more than one type sensory neurons that allows reliable and reproducible coding of senof stimulus modality). Typically they have a higher activation sory stimuli from our environment to the brain. It is important to threshold for detection of mechanical stimuli than Aβ fibres note, however, that these relationships are not ‘hard wired’; they are (though this nociceptive threshold may vary between tissues, e.g. continuously and actively maintained in the physiological state, and cornea vs. skin). A large number of mechanically sensitive proteins can therefore be altered in disease states, whereby primary sensory including mechano-​gated ion channels have been identified across neurons can change phenotype and consequently their function in different functions and species (Delmas et  al., 2011; Delmas and relation to coding for peripheral stimuli, that is, transduction and Coste, 2013). transmission. Indeed, they can even start to fire spontaneously in the Thermosensitivity is thought to be mediated via a family of reabsence of a peripheral stimulus—​one of the recognized mechanceptor proteins called TRPs (transient receptor proteins; Mickle isms of spontaneous (i.e. stimulus-​independent) neuropathic pain et al., 2015). The TRPV1 receptor is expressed by approximately 40% (Costigan et al., 2009). of primary sensory neurons, many of which are small unmyelinated C-​fibre neurons, and the ion channel opens in a temperature-​ Pain transduction and transmission sensitive manner. TRPV1 is thought to be involved in heat detecIn the physiological state, the sensation of physical pain results from tion, and interestingly altered TRPV1 expression in sensory neurons the detection of peripheral stimuli causing potential or actual tissue has been observed experimentally in different disease models. There damage: this is nociception. Nociceptors, that is, primary sensory are several other related proteins that have been implicated cold

Pain pathways—​anatomy and physiology

896

Section 16  Functional neurosurgery

(a)

(b)

Primary somatosensory cortex

Primary somatosensory cortex

Third-order neuron

Third-order neuron

Thalamus

Second-order neuron

Second-order neuron

Dorsal column nuclei Modulla oblongata Decussation of medial lemniscus Dorsal columns Spinal cord

Lissauer’s tract

First-order neuron (afferent)

Anterolateral quadrant

Proprioceptors or mechanoreceptors Dorsal column–medial lemniscal pathway

Nociceptors or thermoreceptors

First-order neuron (afferent)

Spinothalamic tract

Fig. 78.1  (A) Primary sensory neurons have a single axon that leaves the cell body and divides into a long peripheral axon and a shorter centrally projecting one. (B) Cartoon showing major sensory pathways.

detection, innocuous warm detection, and noxious heat sensation in normal and diseased states (Fig. 78.2). Different voltage gated ion channels are preferentially expressed by nociceptors versus non-​nociceptive primary sensory neurons, the best characterized of which are the NaV ion channels (1.7, 1.8, 1.9). These channels are critical to normal sensory transduction and transmission in nociceptors, and certain mutations in the genes coding for these channels are known to cause inherited chronic pain states (Cummins et  al., 2007). Of note, sodium channel function changes in chronic pain states such as inflammation and following nerve injury, such that receptors are sensitized (Liu and Wood, 2011). Other ion channels implicated in nociception are the acid-​ sensitive ion channels, also preferentially expressed by nociceptor neurons (Dubé et  al., 2009). These ion channels open at low pH, which are the conditions found in inflamed tissue where nociceptors are more sensitive and are activated at lower thresholds (Fig. 78.3).

There is a subpopulation of nociceptors which remain silent in the physiological state, not responding to any peripheral stimuli, even noxious ones, but are activated following tissue or nerve injury, when they begin to respond to mechanical and thermal stimuli. These neurons may play a role in altered pain responses during disease states such as inflammation (see Fig. 78.4).

Primary relay—​the dorsal horn Once a nociceptor is activated, action potentials pass along the axon and through the dorsal root ganglion (or trigeminal ganglion) to access the spinal cord. At the spinal levels, these fibres pass though the dorsal roots to the dorsal root entry zone (DREZ). Here, there is anatomical separation of large myelinated A-​fibre axons and small, unmyelinated C fibres. A-​fibre central axons pass into the ipsilateral dorsal columns to run rostrally. As they enter the white matter of the spinal cord from the DREZ, they provide a collateral axon that enters

CHAPTER 78  Pain pathophysiology and surgical management

Classical Inflammatory mediators

Cytokines

Neurotrophins

Neuropeptides

Second messenger systems Adrenergic mediators

TRP receptor Ion channel for mechanical for thermal transduction transduction

Voltage-gated Na+ channel (TTX-resistant)

Fig. 78.2  Immunohistological TTX—​resistant sodium channel expression in the dorsal root ganglion, demonstrated with immunohistochemical staining against peripherin, a protein found in larger A-​fibre afferents. Note the positive cell bodies are largely the small nociceptor neurons. Reproduced from Xi-​Yao Gu et al., Dexmedetomidine inhibits Tetrodotoxin-​resistant Nav1.8 sodium channel activity through Gi/​o-​dependent pathway in rat dorsal root ganglion neurons, Molecular Brain, Volume 8, Issue 15, Copyright © licensee BioMed Central. 2015.

the grey matter of the dorsal horn of the spinal cord. This axon plays a key role in sensory gating. C fibres enter the DREZ and either pass through, to access the dorsal horn, or run a few segments rostrally or caudally in Lissauer’s tract, prior to termination in the dorsal horn. These fibres are known to enter and terminate in the superficial laminae of the dorsal horn, also known as the substantia gelatinosa. Here they synapse onto the dendrites of neurons whose cell bodies lie within the dorsal horn. The majority of these cells are local interneurons involved in local anterograde and retrograde relay circuits. Some are excitatory neurons while others are inhibitory. Therefore, the superficial laminae of the dorsal horn contain a wide range of excitatory and inhibitory neurotransmitters and neuropeptide transmitters that forms a complex gating system, controlling how information is relayed from our environment (internal and external) to the brain. In addition,

there are some neurons with cell bodies found in superficial dorsal horn and the deeper dorsal horn layers, that themselves have a projecting axon which crosses the spinal cord segmentally, anterior to the central canal to run in the spinothalamic tract (STT) tract (see Box 78.1).

Wide dynamic range (WDR) cells These neurons with cell bodies located in the dorsal horn of the spinal cord, predominantly the deeper laminae, receive and respond to input from Aβ fibres, Aδ fibres and C fibres. They can respond in a graded fashion to innocuous and noxious stimuli. They are also known to have large receptive fields, whereas cells in the superficial laminae have much smaller receptive fields. It is thought that localization of a noxious stimulus is therefore encoded by populations or networks of neurons.

100 μm

Nav 1.8

Peripherin

Nav 1.8/Peripherin

50 μm

Nav 1.8/Peripherin

Fig. 78.3  Original model proposed by Melzack and Wall (1965). L—​large/​S —​small sensory neurons. T—​Transmission neurons. By exciting SG (substantia gelatinosa) neurons, this allow presynaptic modulation of sensory input to projection neurons via excitatory and inhibitory mechanisms, with further control superimposed from descending pathways Reproduced with permission from Lorne Mendell, Constructing and deconstructing the gate theory of pain, Pain, Volume 155, Issue 2, pp. 210–​16, Copyright © 2014 Wolters Kluwer Health, Inc.

897

898

Section 16  Functional neurosurgery

Central Control

Box 78.1 Gate theory

Gate Control System L –

+

+ INPUT

SG

T

– S

–

Action System

+

Fig. 78.4  Model of the sensory ending of a nociceptor showing ion channels for transduction of thermal and mechanical stimuli and action potential generation and metabotropic receptors subserving chemosensitivity.

While there are complex neural arrangements involving finely balanced processes of excitatory and inhibitory transmission through anterograde projection and retrograde feedback, in the physiological state at baseline, a fixed peripheral stimulus will typically produce a predictable response. However, our experience of pain when presented with a noxious stimulus can be highly varied. It was recognized long before we developed an understanding pain anatomy and physiology that our experience of pain very much depended on circumstances (i.e. pain perception was contextual). Henry Beecher recognized in 1956 that soldiers could suffer extensive battlefield injuries yet require little or no analgesia, whereas similar injuries in civilian casualties produced much more pain (Beecher, 1956). Much of the modulation is now known to occur within the dorsal horn of the spinal cord, the first sensory relay.

Spinal gating—​modulation of C-​fibre input Through mono-​and poly-​synaptic pathways in the dorsal horn, there is convergence of primary sensory input on to projection neurons. This allows different inputs to modulate transmission of others. Therefore, non-​nociceptive inputs can inhibit nociceptive input and the appreciation of pain; for example, rubbing an injured area reduced the perception of pain. Attempts to explain observations like this led to publication of the gate control theory by Melzack and Wall in 1965 (see Box 78.1). There are several different mechanisms within the spinal cord by which sensory transmission to the brain is modulated, either increased or decreased, involving local excitatory and inhibitory interneurons, and descending pathways from the brain. Direct electrical stimulation of the periaqueductal grey (PAG) produces a profound analgesia that is opioid-​mediated. The PAG projects to the rostroventral medulla (RVM), with includes the raphe nuclei of the medulla, rich in serotonin. In turn, this region projects to the dorsal horn of the spinal cord where they also influence transmission from primary sensory neurons to the central nervous system. These spino-​bulbo-​spinal pathways are thought to form a critical loop in ‘regulating the gain’ in sensory processing within the primary relay. It has been suggested that one mechanism by which acupuncture may impart analgesia may be through the activation

‘The Gate Theory of Pain’, published by Melzack and Wall in Science in 1965, provided a mechanism to explain how nociceptive input could be coded by the sensory system at the first sensory relay in the dorsal horn of the spinal cord. It focused on recently discovered mechanisms of presynaptic control of synaptic transmission from large and small sensory neurons, which was proposed to ‘gate’ incoming information, depending on the balance between these inputs (see Mendell (2014) for full review). It also incorporated other mechanisms, such as descending control pathways from the brain and the convergence of nociceptive and non-​ nociceptive input onto the same neurons in the spinal cord, involved in transmission to the brain. While it was later found that not all the predictions were accurate, the Gate Theory provided a catalyst for further research and understanding within the burgeoning field of pain neurobiology. It helped to change the accepted view of pain mechanisms, with a move away from fixed ‘labelled line’ pain pathways to the realization of both adaptive and maladaptive processes mediating physiological and pathological pain states. See Figure 78.4.

of descending pathways, a phenomenon known as diffuse noxious inhibitory control (DNIC). Further, these modulatory mechanisms are not fixed; nerve or tissue injury can alter the interplay and impact on how sensory information is relayed to the brain (D’Mello and Dickenson, 2008).

Ascending pathways Dorsal horn projection neurons are responsible for relaying sensory information from nociceptors to the brain. This happens via a number of pathways. Some are responsible for projection to thalamus and subsequently neocortex, thought to be critical in producing a conscious response to noxious stimuli. Others are responsible for producing more complex behaviours such as avoidance (away from the noxious stimulus), autonomic responses (e.g. sympathetic activation) and emotional aspects (e.g. fear, anger, sadness). Therefore, as sensory information passes from primary sensory neurons to higher centres, there is divergence of inputs to different targets. This brings new levels of complexity to the organization of the sensory system, which hampers our understanding of the significance of each component (relative to another), and makes the process much more difficult to study scientifically. Interestingly, none of the ‘pain pathways’ described next are specific to noxious input; all contain cells that respond the low-​threshold input, noxious input and ‘wide dynamic range’.

The spinothalamic tract (STT) This has been loosely named the ascending ‘pain pathway’ of the spinal cord, but this is an oversimplification. The STT runs contralaterally in the ventral and ventrolateral white matter funiculi. It has two components, lateral and ventral (anterior). The lateral STT is more heavily implicated in pain and temperature sensation, while the ventral STT is implicated in crude touch and firm pressure sensation. The STT is organized somatotopically; in the cervical spinal cord, cervical fibres are most medial and lumbosacral fibres lateral. The primary target is the ventroposterolateral (VPL) nucleus of the thalamus, part of the thalamocortical sensory coding system.

CHAPTER 78  Pain pathophysiology and surgical management

Spinoreticular tract (SRT) This pathway also consists of axons projecting from dorsal horn projection neurons in lamina V but also deeper laminae extending into ventral horn (VII, VIII) and medial grey matter (X). Like the STT, the SRT is also located in the ventral quadrant of the spinal cord white matter, close to the lateral STT, and it projects to a number of targets including reticular formation (located in the midbrain, pons, and medulla), thalamus (lateral and dorsal reticular nuclei) and the nucleus reticularis gigantocellularis, as well as others targets.

Spinomesencephalic tract This pathway is located ventral to the lateral STT, with projection neutron cell bodies located in superficial (I) and deep laminae of dorsal (IV, V) and ventral (VII, VIII) horns. It projects to the parabrachial nuclei and PAG; note that this region has formed a target for deep brain stimulation as a treatment of chronic pain—​see next). The parabrachial limb has been implicated in emotional response to noxious stimuli, as the parabrachial region, located at the junction of midbrain and pons, part of the lateral reticular formation, and has onward projections to the amygdala as well as thalamus, hypothalamus. Is thought to contain a ‘hedonic hotpot’ involved in liking certain stimuli, and in addition to pain receives inputs from the gustatory portion of the solitary nucleus and is implicated in taste.

Pain pathways—​pathophysiology In the physiological state, nociceptors detect the presence of noxious stimuli and alert us to their presence. While pain is generally an unpleasant experience, it is a crucial component of our integrity and survival, by protecting us from harmful stimuli (e.g. a hand on a hot plate leading to a rapid withdrawal response), or limiting the extent of damage (swollen ankle forcing us to limit weight-​bearing and allow healing). The absence of nociception leads inevitably to serious and potentially life-​threatening harm; rare individuals with congenital insensitivity to physiological pain have lives that feature repetitive injury and early death; other examples are the Charcot joint in diabetic patients, or the well-​recognized eye complications that can occur in patients with combined facial and trigeminal palsy. It is important to recognize that the function in sensory pathways involving primary and secondary sensory projection neurons, interneurons of the spinal cord, brainstem and thalamus, and the descending innervation they receive, are not fixed, like a light-​switch, in a passive, predetermined, hard-​wired system. The phenotype of any individual neuronal cell, and therefore its function, is actively maintained by the local chemical and electrical environment. Therefore, processes that result in a change to the local neuronal environment can lead to a change in neuronal phenotype that has functional implications for the sensory system as a whole (Woolf and Salter, 2000). To appreciate this, it is useful to consider two disease states: inflammation, and nerve injury (e.g. from trauma, compression, tumour invasion).

Inflammation In inflamed tissue, typically non-​painful stimuli (e.g. light pressure) become painful (pain hypersensitivity). As just referred to,

this example of increased or abnormal pain is likely to be evolutionarily advantageous, by protecting an injured area from further harm and allowing healing. Unfortunately, however, there are many disease states where inflammation can become autonomous and serve no useful purpose (e.g. a rheumatoid joint). Under these circumstances, chronic pain becomes the disease. There is now much greater appreciation of interactions between the nervous system and immune system, resulting in reciprocal modulation of function, in both physiological and disease states (McMahon et al., 2015). Whatever the trigger—​tissue injury or auto-​immune processes—​ inflammation is mediated by vasodilatation, plasma extravasation, tissue swelling, the recruitment of immune cells and the release of inflammatory mediators. Thermo-​and chemosensitivity of nociceptors means that a change in the local environment, with an increase in inflammatory cytokines (e.g. IL-​1, IL-​6, TNFα), elevation in local temperature, and reduction in pH lead to altered activity patterns and increased pain. This is not simply an increase in stimulus resulting in an increased response; inflammatory cytokines can alter the processes of sensory transmission from peripheral tissue to the spinal cord, via post-​translational modifications to surface receptors and intracellular signalling proteins, and transcriptional changes within the cell bodies of nociceptors. These changes of protein expression and phenotypic identity lead to an alteration in function, such that the same response results in increased gain and therefore more action potential activation. Wherever one looks at primary sensory neurons—​the peripheral terminal, the axonal membrane, the cell body, or the central terminals in the spinal cord—​inflammation produces changes in function, largely mediated by protein modification, itself driven by altered chemical environment and/​or activity-​ dependent mechanisms. The presence of inflammatory cytokines in peripheral tissue has been shown to correlate with increased pain levels and alterations in the relationship between stimulus and response. This results in phenomena such as hyperalgesia (increased pain from a noxious stimulus) and allodynia (pain from an innocuous stimulus). Furthermore, these changes can be long-​lasting. Understanding these changes gives us insight into how some of the features of chronic pain states, which typically feature abnormal pain behaviour (unlike ‘physiological (nociceptive) pain’) may be mediated.

Nerve injury There are many causes of peripheral nerve injury in addition to trauma or compression (e.g. from a disc prolapse), including systemic diseases causing polyneuropathy, tumour invasion, and inflammation itself. Classical lumbar or cervical radiculopathy is now understood as more than just a compression neuropathy, with the recognition of complex interplays between mechanical compression and an associated inflammatory response. Several inflammatory markers have been identified in neuropathic pain models and clinical studies, which potentially offer new pharmacological treatment targets and further revise the role of surgery for these conditions. Of course, decompressive surgery remains an important, evidence-​based treatment in the management of some patients with spinal cord or radicular compression, particularly patients with bowel or bladder impairment, or evolving neurological deficit. Unfortunately, most patients with neuropathic pain do not have nerve compression that can be treated surgically, and symptom

899

900

Section 16  Functional neurosurgery

Table 78.1  Examples of events observed in primary sensory neurons or their central terminals in disease states, with postulated clinical manifestations Event

Disease state

Symptom

Peripheral sensitization

Inflammation

Pain hypersensitivity (hyperalgesia, allodynia)

Wallerian degeneration/​ Peripheral nerve injury demyelination/​axonal death

Numbness/​reduced sensation

Ephaptic transmission

Peripheral nerve injury

Spreading pain (increased receptive fields)

Ectopic activation—​ C fibres

Inflammation, nerve injury

Spontaneous burning pain (stimulus-​independent)

Cell death—​primary sensory neuron

Peripheral nerve injury

Numbness/​reduced sensation

Cell death—​dorsal horn Peripheral nerve injury interneurons (inhibitory)

Spontaneous pain, spreading pain (increased receptive fields)

Central sensitization

Pain hypersensitivity (hyperalgesia, allodynia)

Inflammation, nerve injury

control can be extremely challenging, arguably one of the most difficult clinical problems to manage. Further, this is a huge societal problem. These patients classically have a mix of negative (sensory reduction or loss) and positive symptoms (hyperalgesia, allodynia, paraesthesia/​ dysaesthesia, spontaneous pain, hyperpathia). The effects of peripheral nerve injury on primary sensory neurons are widespread (Costigan et  al., 2009), and critically do not confine themselves to the injured neurons, but also involve adjacent, uninjured cells and fibres. Table 78.1 details some of the pathophysiological changes that occur following peripheral nerve injury and their likely clinical manifestations.

Spinal mechanisms of chronic pain: Sensitization and disinhibition The effects of peripheral inflammation and nerve injury on primary sensory neuron function results in changes in central nervous system sensory processing (D’Mello and Dickenson, 2008). The underlying causes are multifactorial, including activity-​dependent mechanisms as well as changes in protein function (transcriptional and post-​translational modifications), themselves resulting in changes in cellular activity and behaviour. These changes can be long-​lasting, and can drive postsynaptic changes in activity, protein expression, and even cell death dorsal horn neurons. Changes with the spinal cord can be considered under the headings of sensitization and disinhibition. Following nerve injury and inflammation, chronic pain often extends spatially beyond the usual dermatomal area, resulting in a zone of secondary hyperalgesia, which can become independent of the initial injury. These symptoms cannot be explained by changes in the peripheral nervous system, but instead reflect changes in spinal and supraspinal networks. Nociceptive mechanisms, whereby only physiological high-​threshold stimuli caused pain, can change such that the system now results in pathological low-​threshold pain hypersensitivity. One important mechanism responsible for this change is central sensitization (Latremoliere and Woolf, 2009; see Box 78.2). Other long-​lasting consequences of nerve injury include neuronal

Box 78.2 Central sensitization Under physiological circumstances, a fixed stimulus results in a predictable response. However, following tissue or nerve injury, where activity in primary nociceptors increases, this can lead to an increased response centrally, to the same stimulus peripherally. This is central sensitization, defined by the International Association for the Study of Pain (IASP) as the ‘increased responsiveness of nociceptive neurons in the central nervous system to their normal or subthreshold afferent input’. Activity-​dependent mechanisms of central sensitization include homosynaptic long-​term potentiation—​ with exaggeration of nociceptor responsiveness—​ and heterosynaptic potentiation—​recruiting low-​threshold Aβ fibre inputs into the pain pathway, and these mechanisms may be driven and sustained by ectopic activity in the injured nerve. Central sensitization is produced by the activation of several protein kinases by the neurotransmitter glutamate and various neuropeptide transmitters, which lead to post-​translational and transcriptional changes in postsynaptic receptors (e.g. the N-​ methyl-​ D-​ aspartate (NMDA) receptor). That central sensitization is a major driver of neuropathic pain is supported by the action of drugs that reduce central excitability, including gabapentinoids (e.g. gabapentin and pregabalin), tricyclic antidepressants (e.g. amitriptyline), SNRIs (e.g. duloxetine), and NMDA antagonists (e.g. ketamine). Central sensitization has become a key focus for the development of new analgesics, and is the rationale behind pre-​emptive analgesia (i.e. to prevent increased primary nociceptor input to the CNS).

phenotypic changes due to altered gene transcription and post-​ translational modification of membrane proteins, and decreasing GABA-​ergic and glycinergic tone in the dorsal horn. This ‘disinhibition’ is mediated by mechanisms such as reduced descending inhibitory control, loss of GABA-​ergic or glycinergic interneurons through cell death, reduced GABA or GABA-​synthesizing enzyme (e.g. glutamate decarboxylase), and altered properties of GABA receptors, glycinergic receptors, and cation-​chloride cotransporters. This further contributes to central hyperexcitability and augmentation of pain pathways. An emerging role in sensory transmission and pain processing in disease states has been identified for microglial (macrophage-​like) cells, within the CNS (McMahon and Malcangio, 2009). Following nerve or tissue injury, there is a rapid morphological and functional change in microglia, which has a profound impact on sensory processing (Calvo and Bennett, 2012). These neuroimmune interactions within the CNS have identified potential targets for development of novel pain management strategies for chronic pain conditions (Scholz and Woolf, 2007). Thus, changes in phenotype and function of primary sensory neurons, spinal interneurons, and projection neurons, as well as non-​neuronal cells, can lead to profound alterations in the functional spinal relay and establishing chronic pain states that can extend well beyond the period of primary tissue injury.

Supraspinal mechanisms Pathophysiological changes in the sensory system following peripheral tissue or nerve injury are not confined to primary sensory neurons or the spinal cord. Structural changes, changes in excitatory and inhibitory transmitters and in functional connectivity have all been detected in the brains of chronic pain patients. The challenge remains how to relate these changes to the cognitive, sensory, and emotional pain experience, understanding how much of this is cause versus effect, and whether these changes are reversible.

CHAPTER 78  Pain pathophysiology and surgical management

Surgical procedures for management of chronic pain There is a long history of ablative surgical procedures for the management of chronic pain (Loeser, 2006), though they predate a modern understanding of pain neurobiology and the likely underlying mechanisms that drive chronic pain. Of course, there is still some way to go before we fully understand these mechanisms and can identify them in individual patients, and even further from being able to offer evidence-​based, effective management (von Hehn et al., 2012; Vardeh et al., 2016). Many of the current ‘strategies’ for managing chronic pain patients is empirically based, and we are currently unable to predict who is likely to respond to different pain interventions (e.g. with pharmacology, injection therapies, acupuncture, physical therapy, TENS, and others) nor explain the reason as to why some do respond, while others do not. Understanding sensory pathophysiology, acknowledging the profound alterations that occur in primary sensory neurons, the spinal cord, and higher centres in many ways questions the rationale for surgical management of chronic pain with, particularly for ablative/​irreversible procedures which arguably might cause more maladaptive sensory plasticity than they solve. Perhaps this explains why many of these procedures only result in short-​lived improvements in chronic pain, and only then in some patients, and have a recognized incidence of exacerbating neuropathic symptoms. Many of the destructive procedures have fallen from favour in recent years in most neurosurgical centres, and are now often reserved for terminal cancer patients with a short prognosis. It is also important to acknowledge the placebo effect when considering the utility and efficacy of any surgical procedure, particularly when only short term follow up is utilized (see Colagiuri et al. (2015) for review of placebo effect and underlying mechanisms, widely recognized in the field of surgery). In carefully selected patients and in centres with high expertise, some of these procedures are still practised and considered useful (see Sindou et al., 1990 for review).

DREZ lesions These were first attempted in the 1970s by Marc Sindou in Lyon and gained popularity thereafter. The procedure is still used, although much less frequently, for central or deafferentation pain syndromes such as brachial plexus injury, phantom limb pain, traumatic spinal cord injury, and cancer pain, as well as spasticity. Sensory axons of different diameters are mixed in the dorsal root, but organize just before they reach the spinal cord. The root divides into rootlets and the small fibres lie laterally, prior to entering Lissauer’s tract, while the larger fibres enter more medially, nearer to the dorsal columns. The procedure is done through a laminectomy or hemilaminotomy. The pia over the DREZ is opened and microsurgical lesions are made, either with the use of diathermy or radiofrequency ablation. Complications include corticospinal tract or dorsal column injury, continence issues, and worsening or persistent neuropathic pain. The procedure is now generally considered only as a last resort for patients who have failed all other less invasive management strategies.

Cordotomy The lateral spinothalamic tract (STT) carries signals of pain, temperature, and touch from the contralateral side, therefore a destruction or

interruption of the tract should lead to the analgesia and loss of sensations to heat and cold on the contralateral side. Open cordotomy was described in 1911, but the procedure was associated with significant morbidity and mortality. The development of percutaneous techniques has made the procedure less invasive and cordotomy is a recognized component of the algorithm for management of malignant unilateral pain. Computed tomography (CT)-​guided techniques have allowed more precise needle positioning and lesioning. Endoscopic cordotomy techniques also have been explored. How does it work? The lateral STT tract is composed of fibres decussating from the contralateral side of the spinal cord. At each spinal level, a second order neural axons travel on the ipsilateral side of the spinal cord for two to five levels, before decussating to the contralateral anterolateral spinal cord where they join lateral STT. At the C1–​2 level, the lateral STT has clear somatotopic organization: the anteromedial part has fibres from the upper extremity, shoulder, anterior chest, whereas posterolateral aspect contains the fibres from lower extremity, lower back, and the pelvis (Fig. 78.5). Several techniques of percutaneous cordotomy have been described. A  radiofrequency (RF) cannula is inserted via the C1–​2 foramen into the anterolateral part of the spinal cord under local anaesthesia with either fluoroscopic (classical technique) or CT guidance. The aim is to penetrate the dura and enter the anterolateral spinal cord perpendicular to the sagittal plane, just anterior to the dentate ligament. Sensory stimulation should produce clear change in thermal sensations (heat or cold) with associated tingling on the contralateral side, ideally, but not necessarily overlapping the area of pain. Motor stimulation should produce ipsilateral contractions of the neck and shoulder muscles due to the stimulation of the adjacent corticospinal tract. Thermal sensation changes with stimulation on the contralateral side is considered a reliable predictor of good analgesic effect. Then, a thermal radiofrequency lesion is carried out in an incremental fashion, with continuous monitoring of motor function on the ipsilateral side. After lesioning, there should be a marked difference in sensation to pinprick between the sides. CT guided anterior transdiscal approach has been described. Some have also advocated laser lesioning with a lateral endoscopic technique, for even better precision and accuracy. Indications for percutaneous cordotomy Cooperative patients who suffer unilateral nociceptive pain below C5 due to malignancy, despite appropriate palliative pharmacological and interventional management, with adequate lung function and life expectancy from 6 months to less than 1 year, can be considered for cordotomy. The best-​studied groups of patients are those with malignant mesothelioma, Pancoast tumours, and metastatic lung cancer. There is some low-​level evidence that cordotomy is not effective for neuropathic pain. Evidence of efficacy Most publications are retrospective, with some prospective case series. A recent systematic review looked at percutaneous cordotomy in the treatment of mesothelioma-​related pain. In 160 patients, complete or partial pain relief was achieved immediately after the treatment in 94% of cases. Recurrence of pain is widely recognized in patients who live beyond a year.

901

Section 16  Functional neurosurgery

pin

Dorsal

Dentate ligament

Ce

902

r reb

d lui al F

os

A C B

Ventral Fig. 78.5  Schematic representation of a cross-​section of the spinal cord at C2. Reproduced with permission from Alireza Feizerfan, and JHL Antrobus, Role of percutaneous cervical cordotomy in cancer pain management, Continuing Education in Anaesthesia Critical Care & Pain, Volume 14, Issue 1, pp. 23–​6, Copyright © 2014.

Complications In experienced hands with carefully selected patient, major complications are very rare ( arm), dysaesthesia, mirror pain in the ipsilateral side, and dyspnoea due to intercostal/​diaphragmatic muscle weakness. General complications such as chest infection, orthostatic hypotension, urinary retention, incontinence, sexual dysfunction, confusion, and death have also been described. Lesions affecting nearby anterior spinocerebellar, reticulospinal and corticospinal tracts can produce ataxia, or disturbed autonomic and motor control.

Bilateral cordotomy This has been used in very small numbers for bilateral cancer pain, or development of a new pain following unilateral cordotomy. There is very scant efficacy data, however, it is known that bilateral cordotomy has a much higher complication rate including ataxia, motor weakness, dysaesthesia, incontinence, respiratory depression, and death.

Practical issues Following the arrival of long-​term oral and transdermal opiate formulations (Box 78.3) and intrathecal drug delivery systems, the number of cordotomy procedures has been steadily declining. This is despite the growing number of malignant mesothelioma cases (anticipated to peak around 2020) and often suboptimal pain relief with pharmacological and interventional therapies. While this can be a useful method of pain relief in some patients, it may become extinct as fewer are performed and physicians become more reluctant to offer destructive and irreversible treatments.

Mesencephalic tractotomy Like thalamotomy, this procedure has been used to treat poststroke pain and cancer pain though not in high numbers and there are few surgeons with experience. Mesencephalotomy is used for patients

Box 78.3  Opiates—​kings of pain control? Opioid receptors are the proteins that control nociceptive, hedonic, emotional, as well as autonomic, neuroendocrine, and immune responses. Gene cloning identified three receptor genes only, encoding mu, delta, and kappa receptors. Receptor-​deficient knock-​out mice exhibit enhanced pain sensitivity. In models of acute pain, mu receptors modulate mechanical, chemical, and supraspinally controlled thermal nociception, while kappa receptors modulate spinally mediated thermal nociception and visceral pain. There is strong evidence for a role of delta receptors in reducing hyperalgesia in situations of inflammatory and neuropathic pain. Endogenous opioid peptides. Met-​and leu-​enkephalins act on the spinal opioid receptors in lamina I and II of the dorsal horn, spinal trigeminal nucleus and periaqueductal grey (PAG). They are also present in the gastrointestinal (GI) tract and adrenal medulla. Dynorphins have actions on multiple spinal and supraspinal sites, but are considered to be weak analgesics. Multiple complex actions of beta-​endorphins are largely limited to the central nervous system sites. Spinal actions of opioids occur in tetrasynapse composed of the central terminal of primary afferent neurons, projection neurons, astrocytes, and microglia in the spinal cord dorsal horn. Opioids: 1 hyperpolarize primary afferent membrane by closing Ca ++ channels and opening K + channels, thus reducing glutamate release from nociceptive afferents (presynaptic effect); nociceptive input from the periphery is reduced; 2 reduce intracellular Ca ++ and increase K + postsynaptically thus quenching several mechanisms that would otherwise lead to hyperexcitability of second order neurons. Supraspinal sites of opioid analgesia There are multiple mechanisms and action sites in the brain. Opioids interact with 5-​HT, GABA-​ergic, and possibly other systems of central descending regulation. They move rostroventral medial medulla (RVM) output towards descending inhibition. Opioids also play an important role in regulating pain experience. Peripheral action of opioid analgesia. Opioid receptors are also synthesized in the dorsal root ganglion (DRG) and transported down the axon into the peripheral sites. Endogenous opioid peptide synthesis can be unregulated in the DRG in inflammatory models.

CHAPTER 78  Pain pathophysiology and surgical management

with intractable pain involving the head, neck, shoulder, and arm, can be considered as a supraspinal cordotomy. The outcomes however have been mixed, and complication such as oculomotor dysfunction are widely recognized. It is rarely carried out these days, and other less invasive management strategies are usually preferred.

Spinal cord stimulation Spinal cord stimulation (SCS) refers to direct application of electrical current to the spinal cord with the aim of achieving analgesia. Empirically, therapeutic use of electricity had been attempted at least as early as 63 BC by the ancient Greeks, who used torpedo fish as the source of electrical current in treating gout. As a modern method of chronic pain relief SCS has been in use since 1966, when Shealy applied an intrathecal monopole electrode connected to the external cardiac pulse generator for the first time and successfully palliated pain in a patient with advanced lung cancer. Shealy thought that direct electrical stimulation of large primary afferents would ‘close the gate’ to nociceptive signals propagating via small Aδ-​and C fibres (see Box 78.1 on Gate Theory). With the development of fully implantable spinal cord stimulation systems, SCS has become an established method of chronic pain relief (Song et al., 2014). National Institute of Clinical Excellence (NICE) in the United Kingdom recommends SCS for treatment of chronic peripheral neuropathic pain, and the FDA has approved SCS for the management of chronic pain of the trunk and extremities.

How does it work? SCS implantable system consists of two main parts—​ electrode lead(s) placed in the epidural space and an implantable pulse generator (IPG) positioned in the subcutaneous pocket (Fig. 78.6). Electrodes can be selectively programmed to act as cathode, neutral, or anode, thus creating an electrical current and electromagnetic field which directly affect surrounding spinal structures.

Neurons under a cathode becomes less negatively charged and will depolarize, with resulting bidirectional action potential propagation. The anode creates localized axon hyperpolarization as the membrane becomes more negatively charged. By using different cathode-​anode combinations along the lead, it is possible to create an electrical field that achieves paraesthesia in the effective dermatome(s) and concurrent analgesia. Holsheimer used computerized models to explain that stimulation efficacy is determined by several factors that include: the cathode position in relation to physiological midline; electrode polarity; cerebrospinal fluid (CSF) layer thickness under the cathode; axonal diameter and myelination; and axonal orientation within the dorsal columns, dorsal roots, and dorsal horn (anisotropism). More recently, electrical parameters of the stimulation pulse wave have been modified to allow high frequency stimulation, burst pattern stimulation, high current density stimulation, and subthreshold stimulation. Some of these novel simulation techniques do not result paraesthesiae, but still are able to achieve significant analgesic effect. Mechanisms of action in paraesthesia-​based, low frequency SCS have been investigated since the early 1970s. Data on SCS mechanisms have largely come from animal studies, and translation of preclinical data to clinical practice is challenging. While a definitive single mechanism has not yet been determined, several important factors are thought to be involved. First, activation of large diameter Aβ fibres (activating more of them appears to improve outcome), leads to inhibition of dorsal horn WDR neurons. Second, SCS induces release of neurotransmitters, some of which are inhibitory, such as GABA, but also 5-​HT (serotonin), glycine, adenosine, and acetylcholine, both pre-​and postsynaptically in the dorsal horn. Third, supraspinal pathway stimulation activates descending pain inhibition system resulting in a release of 5-​HT and noradrenaline at the synapses with WDR neurons. These mechanisms have some clinical validation:  it is known that paraesthesia (presumably due Aβ-​fibre activation) must be experienced by the patient and must overlap the patient’s painful areas to result in pain relief from low frequency SCS. The necessity of overlap indicates convergence of Aβ terminal projections in the hyperexcitable regions of the dorsal horn. Also, intrathecal administration of subclinical doses of baclofen, a GABAB agonist, potentiates the effects of previously ineffective low frequency SCS, thus implicating GABA in the pain relief from traditional, paraesthesia-​based  SCS. Recently, 10 kHz high frequency (HF10), paraesthesia-​free SCS has been demonstrated to be statistically and clinically superior to traditional low frequency, paraesthesia-​based SCS in the relief of chronic intractable back and leg pain. In contrast to low frequency SCS, patients receiving high frequency stimulation do not experience stimulation-​induced sensations, and the Aβ-​mediated paraesthesia-​pain overlap from HF10 therapy targets appears uncorrelated to pain relief. This suggests that Aβ fibres may not play as significant a role in kHz HF-​SCS mechanisms, and there is some evidence from preclinical studies that HF10 therapy engages different mechanisms to traditional low frequency SCS.

Indications and contraindications Fig. 78.6  AP view of two percutaneous cylindrical spinal cord stimulation (SCS) electrode leads placed in the lower thoracic epidural space in a in a patient with left radicular leg and bilateral low back pain.

SCS generally works very well for peripheral neuropathic pain, the most common indication being postsurgical radicular extremity pain. Its effect in treating central neuropathic pain is less

903

904

Section 16  Functional neurosurgery

well established. Until recently, the results on low back pain have been mixed. With the introduction of high frequency stimulation and multiple contact electrode leads, it is becoming clear that some forms of low back pain appear to be responsive to SCS. Common indications for a SCS treatment are presented as follows: • Chronic neuropathic pain in postlaminectomy syndrome • Complex regional pain syndrome • Peripheral neuropathies including post-​traumatic and diabetic neuropathy • Pain due to ischaemic peripheral vascular disease • Intractable angina • Sacral stimulation for pelvic pain Contraindications to SCS therapy include inability manage stimulation; cardiac pacemaker/​ICD/​intrathecal pump in situ; local or systemic infection; immunosuppression; coagulopathy. A  relative contraindication is pathology otherwise amenable to surgery (nerve root or spinal cord compression, spinal instability, canal stenosis), though even this has been challenged, for example, recurrent disc prolapse, where revision surgery is known to have higher complications.

Clinical evidence for SCS The first favourable RCT on SCS vs. reoperation for failed back pain syndrome (FBSS) patients was published by North et al. in 2005. In 2007, a prospective multicentre RCT (PROCESS) compared SCS therapy with conventional medical management (CMM) in 100 patients with FBSS. 48% patients achieved more than 50% pain control in SCS group versus 9% in a CMM group in 6 months (intention-​ to-​treat analysis), followed by 34% and 7% at 12 months and 37% versus 2% at 24 months (P = 0.003). Importantly, SCS group patients reported better function with Oswestry Disability Index and quality of life with EQ-​5-​D questionnaire (P