Critical Care in Acute Liver Failure [1 ed.] 9781780842554, 9781780842578

Citation preview

Critical care in acute liver failure Editors Roger Williams Institute of Hepatology, London, UK Julia Wendon Institute of Liver Studies, King’s College Hospital, London, UK

Published by Future Medicine Ltd Future Medicine Ltd, Unitec House, 2 Albert Place, London N3 1QB, UK www.futuremedicine.com ISSN: 2047-332X ISBN: 978-1-78084-257-8 (print) ISBN: 978-1-78084-256-1 (epub) ISBN: 978-1-78084-255-4 (pdf) © 2013 Future Medicine Ltd 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 electronic, mechanical, photocopying, recording or otherwise without prior written permission of the copyright holder. British Library Cataloguing-in-Publication Data. A catalogue record for this book is available from the British Library. Although the author and publisher have made every effort to ensure accuracy of published drug doses and other medical information, they take no responsibility for errors, omissions, or for any outcomes related to the book contents and take no responsibility for the use of any products described within the book. No claims or endorsements are made for any marketed drug or putative therapeutic agent under clinical investigation. Any product mentioned in the book should be used in accordance with the prescribing information prepared by the manufacturers, and ultimate responsibility rests with the prescribing physician. Content Development Editor: Duc Hong Le Senior Manager, Production & Design: Karen Rowland Head of Production: Philip Chapman Managing Production Editor: Harriet Penny Production Editor: Georgia Patey Assistant Production Editors: Samantha Whitham & Gemma King Graphics & Design Manager: Hannah Morton

Contents Critical care in acute liver failure Roger Williams & Julia Wendon Definition, pattern of disease and prognosis William Bernal Histopathological basis of syndrome Alberto Quaglia & Bernard Portmann Worldwide differences in acute liver failure Shalimar, Subrat K Acharya & William M Lee Investigation and diagnostic pathways Heather Patton & Robert G Gish Disease-specific treatments Catherine Paugam-Burtz, Emmanuel Weiss & Richard Moreau General supportive management Alexander Wilmer & Frederik Nevens Management of encephalopathy and cerebral edema Fin Stolze Larsen & Peter Bjerring Inflammation, infection and renal failure Darren G Craig & Kenneth Simpson

3 7 19 33 49 67 75 89 101

Contents Continued Acute liver failure coagulation disturbances and need for ­treatment Andrew K Burroughs Selection and results of liver transplantation Sarah A Hughes & John O’Grady Acute liver failure in children Naresh P Shanmugam & Anil Dhawan Plasmapheresis and extracorporeal liver support Roger Williams & Julia Wendon Index

115 125 143 163 176

About the Editors Roger Williams Roger Williams is Director of the Institute of Hepatology, London, UK and of the Foundation for Liver Research. Before that, he had established, over a period of 30 years, the world-renowned Institute of Liver Studies at King’s College Hospital (London, UK). He is a Fellow of the Academy of Medical Sciences and is the recipient of numerous honorary fellowships, medals and prizes, including the American Society of Transplantation Senior Achievement Award in 2004, the Hans Popper Lifetime Achievement Award in 2008 and in 2011 the Distinguished Service Award of the International Liver Transplant Society. His main clinical and research interests are in acute liver failure, liver transplantation, complications of cirrhosis and management of viral hepatitis.

Julia Wendon Julia Wendon trained in internal medicine before specializing in liver intensive care and hepatology, and has been a consultant within the Institute of Liver Studies, King’s College Hospital (London, UK) since 1992. Her focus is in liver intensive care incorporating encephalopathy, hepatorenal failure, haemodynamic failure, sepsis and immune function, liver support systems, liver function assessment and management of acute liver failure.

2

2

© 2013 Future Medicine www.futuremedicine.com

Foreword Critical care in acute liver failure Roger Williams & Julia Wendon When I received the invitation to edit a book on acute liver failure (ALF) in this new e-learning series, I accepted at once after the initial groan as to yet another addition to my workload. If ever there was a condition where speed of obtaining information is paramount, it is ALF. Almost always striking down people in previously good health and with a rapid disease progression, there is often little time to intervene, and the appropriate intervention can truly represent the difference between survival and mortality. Our hope as editors is that the availability on the internet of a clearly formatted account of the subject will help in obtaining the best possible care for these desperately ill patients, with the opportunity to access information regardless of time or place. ALF is a subject too that has always been very close to my heart. I shall never forget those early days in the 1970s when we were flooded by an epidemic of fulminant hepatic failure due to hepatotoxicity af ter paracetamol/acetaminophen taken in suicidal overdose. Cases of fulminant hepatitis B in renal dialysis units were also causing great concern at that time and led to the opening of the first wholly dedicated intensive care facility for the treatment of ALF at King’s College Hospital (London, UK) in 1973. Hepatitis B remains the commonest cause of ALF in the Far East, and it is interesting to recollect that the first description in 1946 of fulminant fatal hepatitis by Lucké and Mallory in soldiers at the end of World War II was attributed to a presumed serum hepatitis virus. doi:10.2217/EBO.12.302

© 2013 Future Medicine

3

Williams & Wendon Not surprisingly, I turned to Julia Wendon to join me as co-editor of this book with her unrivalled knowledge of managing liver failure in all its different guises. The knowledge base encompassing 35 years of experience in over 3000 cases of ALF admitted to the Liver Intensive Treatment Unit at King’s has just been submitted for publication and some of the important findings are covered in the first chapter of this book. The observation that full restitution of liver structure is possible despite the severity of liver injury in ALF has for over 40 years spurred many, many attempts to develop a means of temporary liver support to provide extra time for the patient to recover from the liver injury. Exchange transfusion was followed by extracorporeal (animal) and human liver perfusion. Crosscirculation was tried and there is no better example of the harmful toxemia of liver failure – the healthy volunteer becoming ill as the ALF patient appeared to stabilize. The first efforts with adsorbent hemoperfusion were tantalizingly close to a successful device. Today we have more sophisticated adsorption/dialysis machines, although the holy grail of a fully functioning bioartificial device providing a clear survival benefit continues to elude us. The contributors of this volume were asked on the basis of their important contributions and expertise in ALF. We are indebted to them all and to the Future Science Group for making this book possible. Acknowledgements R Williams would like to extend his very special thanks to Enda O’Sullivan who has done so much of the hard work in collaborating with the contributors to get all the chapters finalized for this volume. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

4

www.futuremedicine.com

5

About the Author William Bernal William Bernal first worked at the Institute of Liver Studies at Kings College Hospital (London, UK) in 1995, and became a Consultant Intensivist at the Liver Intensive Therapy Unit there in 2002. He is currently Reader in Hepatology and Intensive Care Medicine. His clinical interests cover liver failure and transplantation, and research interests include prognostic modelling in acute liver failure and the pathogenesis of encephalopathy and multiple organ failure.

6

6

© 2013 Future Medicine www.futuremedicine.com

1

Chapter

Definition, pattern of disease and prognosis

Definition & pattern of disease

William Bernal 8

Prognostic assessment 

12

Changing outcomes?

14

This chapter discusses the current concepts of acute liver failure, focusing on the international use of definitions, the differing patterns of disease that occur, and their relation to etiology and outcome. It presents a framework for prognostic assessment and referral, with observations on the changing nature of the condition.

doi:10.2217/EBO.12.338

© 2013 Future Medicine

7

Bernal Hepatic encephalopathy: the occurrence of altered mental state or consciousness as a result of liver failure.

Definition & pattern of disease Definitions The term ‘fulminant hepatic failure’ was Coagulopathy: disease-related altered coagulability first coined in 1970 to describe a “potentially of the blood. reversible condition, the consequence of Presentation: the temporal evolution of physical signs severe liver injury, with an onset of or symptoms. encephalopathy within 8  weeks of the appearance of the first symptoms and in the absence of pre-existing liver disease” [1]. Key elements of this definition remain relevant today, particularly the recognition of the central importance of the development of encephalopathy, although classifications have evolved through the observations that prognosis and complications vary in relation to the rate of evolution of illness. More modern definitions therefore seek to quantify the interval between symptom onset and development of hepatic encephalopathy and, in some cases, the severity of liver injury by degree of coagulopathy. However, no consensus exists on the severity of coagulopathy or encephalopathy that mark the transition from liver injury to acute liver failure (ALF). A ‘fulminant’ presentation is now also recognized to occur in some specific forms of previously subclinical chronic liver disease, exemplified by some cases of Wilson’s disease or reactivation of hepatitis B virus infection [2]. A number of classification systems exist and include: n The ‘O’Grady’ terminology, which is commonly used in adults, subclassifies into three groups: hyperacute, acute, and subacute. These are dependent upon the interval between the development of jaundice and the onset of encephalopathy [3]; Bernuau and colleagues classify the condition into two categories of ‘fulminant’ and ‘subfulminant’ disease, with a jaundice to encephalopathy interval of less than 2 weeks and 2 weeks to 3 months, respectively [4];

n

Japan consensus criteria subdivide the condition into ‘fulminant hepatitis’ and ‘late-onset hepatic failure’ with onset of encephalopathy at less than 8 weeks and 8–24 weeks after disease onset, respectively. ‘Fulminant hepatitis’ is further subdivided into acute and subacute subtypes [5].

n

These classifications have direct clinical relevance as they may give clues as to the etiology of disease, its probable complications and overall prognosis. However, in young children, clinical encephalopathy may be absent or only occur late in the course of illness. Working pediatric definitions therefore do not depend on the presence of encephalopathy but only on coagulopathy due to liver injury. An accepted definition is that of “a multisystem disorder

8

www.futuremedicine.com

Definition, pattern of disease & prognosis in which severe impairment of liver function, with or without encephalopathy, occurs in association with hepatocellular necrosis in a patient with no recognized underlying chronic liver disease” [6]. Patterns of disease The central place of encephalopathy in all definitions of adult ALF reflects its key prognostic importance, with development indicating critically impaired liver function (Figure 1.1). Its clinical features cover a spectrum of neurological impairment that ranges from mild disorders of concentration to frank coma (Table 1.1). In patients with subacute presentations, even low-grade encephalopathy may indicate extremely poor prognosis, whereas in hyperacute disease survival with medical management may be good even in patients developing high grades. Adoption of a single threshold of encephalopathy severity to mark transition from liver injury to frank failure across all presentations and etiologic groups is thus somewhat simplistic. The etiology of disease and the clinical phenotype of illness that results are closely linked. Using the O’Grady classification, hyperacute liver failure typically results from acetaminophen or acute hepatitis A or E virus infection causing liver injury with a week or less between symptom onset and Figure 1.1. Nontransplanted survival in patients with acute liver injury and failure based on the severity of encephalopathy developed. 100

Nontransplant survival (%)

90

Nonacetaminophen Acetaminophen

80 70 60 50 40 30 20 10 0 >Grade 3

15 s or prothrombin activity 70% of the global population resides. A recent study by the WHO [16] assessed liver disease burden globally and identified that, worldwide, approximately 3.7 million people get affected by HEV with 70,000 figure mortality in 1 year. In AFLP, which has a strong genetic component, and may have geographical variation in prevalence, termination of pregnancy is required for improving prognosis. However, there is no rationale for actively terminating pregnancy in HEV endemic regions with the hope of improving the outcome of the patient. Genotypes 1 and 2 of HEV are prevalent in hyperendemic regions where the reservoir for HEV seems to be human and cause outbreaks, sporadic acute hepatitis, ALF and acute-on-chronic liver failure [3]. Genotype 3 and 4 are more prevalent in the USA, Europe and Japan where the reservoir seems to be represented by pigs and other mammalian and avian species. In these countries, zoonotic transmission leads to autochthonous

www.futuremedicine.com

41

Shalimar, Acharya & Lee Management: liver transplant improves survival in selected patients. Long-term transplant results are good. In absence of transplant (developing countries), prevent and treat infection and cerebral edema.

acute HEV. Genotype 3 and 4 are infrequently associated with severe liver disease and the majority of cases represent subclinical infection.

Outcome The etiology of ALF, which differs across the world as highlighted above, is a major determinant of outcome. In the West, the etiology is heterogeneous. Acetaminophen-induced ALF has a better outcome than most other etiologies, even after development of hepatic failure. The spontaneous survival of patients who develop encephalopathy is approximately 64% and exceeds that for most other forms of ALF, such as ALF due to idiosyncratic drug toxicity (where survival without transplant is seen only in ~20% of cases). However, when acetaminophen patients do progress, they do so very rapidly. In addition, because acetaminophen-induced ALF constitutes the bulk of all ALF cases; the total number of deaths due to acetaminophen toxicity exceeds all other diagnoses. Nearly a third of patients who develop encephalopathy die. These patients are often found unsuitable for transplantation due to evidence of associated substance abuse, suicidal intent or history of prior suicidal attempts. In cases of acetaminophen-induced hepatic injury, once ALF develops, the outcome for both types of overdose – suicidal or unintentional – is similar [4]. Both from the East and West and Japan, there are reports of ALF due to ATT. In studies from India, a mortality of 70% has been reported for ATTinduced ALF [14]. Studies from India also documented that more than 90% of ALF were due to hepatitis viruses indicating a homogeneous etiology [10,14]. Therefore, when survivors and nonsurvivors were compared, the etiology was not found to be different and the multivariate ana­lysis also could not establish etiology as an independent predictors of mortality. However, when HEV as a separate group was compared with other etiologies, such as ATT-induced ALF, the survival frequency among HEV was significantly superior to other etiologies [5]. Of course, the availability of liver transplantation greatly impacts on outcome with recent studies suggesting an overall survival in Western countries exceeding 85%. Outcome: ƒƒ West: etiology affects the outcome. ƒƒ India and Japan: etiology appears less important; >90% have a single etiology (viral hepatitis). ƒƒ Among viral ALF, HAV and HEV have better prognoses than HBV. HCV causes ALF very rarely if at all. ƒƒ ATT-induced ALF: high mortality.

42

Prognostic models The ability to predict which patients with ALF will recover with medical management alone and who will succumb without liver transplantation is of paramount important. Static variables that correlate with survival

www.futuremedicine.com

Worldwide differences in acute liver failure are age and etiology. Patients with HAV, HEV and acetaminophen toxicity, and AFLP have better survival rates than those with drug-induced, autoimmune, hepatitis B and cryptogenic ALF, where spontaneous survival rates are 7  days, grade 3 or 4 encephalopathy, presence of cerebral edema, prothrombin time ≥35 s, and creatinine ≥1.5 mg/dl. Presence of any three out of six clinical prognostic indicators was superior to Model for End Stage Liver Disease or KCH in identifying survivors and nonsurvivors [19]. ALF is a dynamic process in which variables determining prognosis at admission change over time, and thus, the clinical course varies accordingly. A new prognostic model from India, ALF early dynamic model [20] is based on four variables: arterial ammonia, serum bilirubin, international normalized ratio and hepatic encephalopathy grade >2. This model takes into account the values of these variables over 3 days. The performance of the ALF early dynamic model has been reported to be superior to the KCH criteria and the Model for End Stage Liver Disease score.

Management The general principles of care are the same across the world. All patients are managed in intensive care units. Organ support systems are used as

www.futuremedicine.com

43

Shalimar, Acharya & Lee required. Liver transplantation has improved the survival in these patients. In the USA, 25–29% of patients with ALF are subjected to liver transplantation [21]. In the largest published cohort of patients who underwent liver transplant for ALF across Europe [22], the patient survival at 1, 5 and 10 years was 74, 68 and 63%, respectively. The graft survival was 63, 57 and 50% over the same duration of follow-up. The survival rate has progressively improved over time. Similar results are reported from the USA and Japan. Liver transplantation is not widely available in developing countries; hence, the management is mainly supportive with organ support, prevention and treatment of infections, and control of cerebral edema. In low-endemic regions, sporadic cases of locally acquired HEV infection, mainly caused by genotypes 3 or 4 are reported. In a series of 14 patients from France who underwent solid organ transplantation and were on immunosuppression, all patients presented with unexplained elevation of liver enzymes and were detected to have acute HEV infection [23]. All patients were infected with genotype 3 HEV. Eight out of 14 patients developed chronic hepatitis, and chronic hepatitis may progress to cirrhosis in immunosuppressed patients. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

Summary. ƒƒ Considerable geographical and regional differences exist in defining characteristics, etiologies and prognostic markers of acute liver failure. ƒƒ In Western countries, drug-induced acute liver failure including acetaminophen accounts for most cases; in the East, viral hepatitis comprises the majority of cases. ƒƒ Outcomes for certain other viral infections (hepatitis A virus/hepatitis E virus), and acetaminophen are better than for other etiologies. ƒƒ Pregnant females are more prone to develop acute liver failure, but mortality is not increased by pregnancy. ƒƒ Different prognostic models are described across the world, dynamic models have been described recently. ƒƒ Management is conservative and includes organ support and liver transplantation in selected patients.

44

www.futuremedicine.com

Worldwide differences in acute liver failure References 1

Wlodzimirow KA, Eslami S, Abu-Hanna A, Nieuwoudt M, Chamuleau RA. Systematic review: acute liver failure: one disease, more than 40 definitions. Aliment Pharmacol. Ther. 35(11), 1245–1256 (2012).

2

Trey C, Davidson C. The management of fulminant hepatic failure. In: Progress in Liver Disease (Volume 3). Popper H, Schaffner F (Eds). 282–298 (1970).

3

Aggarwal R. Hepatitis E. Historical, contemporary and future perspectives. J. Gastroenterol. Hepatol. 26(1), 72–82 (2011).

4

5

6

7

Larson AM, Polson J, Fontana RJ et al. Acute Liver Failure Study Group. Acetaminopheninduced acute liver failure: results of a United States multicenter, prospective study. Hepatology 42(6), 1364–1372 (2005). Kumar R, Shalimar, Bhatia V et al. Antituberculosis therapy-induced acute liver failure: magnitude, profile, prognosis, and predictors of outcome. Hepatology 51(5), 1665–1674 (2010). Tandon BN, Bernauau J, O’Grady J et al. Recommendations of the International Association for the Study of the Liver Subcommittee on nomenclature of acute and subacute liver failure. J. Gastroenterol. Hepatol. 14(5), 403–404 (1999). Escorsell A, Mas A, de la Mata M; the Spanish Group for the Study of Acute Liver Failure. Acute liver failure in Spain: analysis of 267 cases. Liver Transpl. 13, 1389–1395 (2007).

8

Ostapowicz G, Lee WM. Acute hepatic failure: a Western perspective. J. Gastroenterol. Hepatol. 15(5), 480–488 (2000).

9

Forde KA, Reddy KR, Troxel AB, Sanders CM, Lee WM, Acute Liver Failure Study Group. Racial and ethnic differences in presentation, etiology, and outcomes of acute liver failure in the United States. Clin. Gastroenterol. Hepatol. 7(10), 1121–1126 (2009).

10 Acharya SK, Dasarathy S,

Kumer TL et al. Fulminant hepatitis in a tropical population: clinical course, cause, and early predictors of outcome. Hepatology 23(6), 1448–1455 (1996).

11 Blei AT. Brain edema in acute

liver failure. Crit. Care Clin. 24(1), 99–114 (2008).

12 Bernal W, Hall C, Karvellas CJ,

Auzinger G, Sizer E, Wendon J. Arterial ammonia and clinical risk factors for encephalopathy and intracranial hypertension in acute liver failure. Hepatology 46(6), 1844–1852 (2007).

13 Karvellas CJ, Pink F, McPhail

M et al. Predictors of bacteraemia and mortality in patients with acute liver failure. Intensive Care Med. 35(8), 1390–1396 (2009).

14 Bhatia V, Singhal A, Panda SK,

Acharya SK. A 20-year singlecenter experience with acute liver failure during pregnancy: is the prognosis really worse? Hepatology 48(5), 1577–1585 (2008).

15 Bose PD, Das BC, Kumar A,

www.futuremedicine.com

Gondal R, Kumar D, Kar P. High viral load and

deregulation of the progesterone receptor signaling pathway: association with hepatitis E-related poor pregnancy outcome. J. Hepatol. 54(6), 1107–1113 (2011). 16 Rein DB, Stevens GA, Theaker

J, Wittenborn JS, Wiersma ST. The global burden of hepatitis E virus genotypes 1 and 2 in 2005. Hepatology 55(4), 988–997 (2012).

17

Hoofnagle JH, Carithers RL, Shapiro C, Ascher N. Fulminant hepatic failure: summary of a workshop. Hepatology 21, 240–252 (1995).

18 O’Grady JG, Schalm SW,

Williams R. Acute liver failure: redefining the syndromes. Lancet 342, 273–275 (1993).

19 Dhiman RK, Seth AK, Jain S,

Chawla YK, Dilawari JB. Prognostic evaluation of early indicators in fulminant hepatic failure by multivariate ana­lysis. Dig. Dis. Sci. 43(6), 1311–1316 (1998).

20 Kumar R, Shalimar, Sharma H

et al. Prospective derivation and validation of early dynamic model for predicting outcome in patients with acute liver failure. Gut 61(7), 1068–1075 (2012).

21 Lee WM. Acute liver failure.

Semin. Respir. Crit. Care Med. 33(1), 36–45 (2012).

22 Germani G, Theocharidou E,

Adam R et al. Liver transplantation for acute liver failure in Europe: outcomes over 20 years from the ELTR database. J. Hepatol. 57(2), 288–296 (2012).

23 Kamar N, Selves J, Mansuy JM

et al. Hepatitis E virus and

45

Shalimar, Acharya & Lee chronic hepatitis in organtransplant recipients. N. Engl. J. Med. 358(8), 811–817 (2008). 24 Bernal W, Auzinger G,

Wendon J. Prognostic utility of the bilirubin lactate and etiology score. Clin. Gastroenterol. Hepatol. 7(2), 249 (2009).

25 Ichai P, Samuel D. Etiology

and prognosis of fulminant hepatitis in adults. Liver Transplant. 14(8), S67–S79 (2008).

46

26 Hadem J, Tacke F, Bruns T

et al. Acute Liver Failure Study Group Germany. Etiologies and outcomes of acute liver failure in Germany. Clin. Gastroenterol. Hepatol. 10(6), 664–669 (2012).

27 Gow PJ, Jones RM, Dobson JL,

Angus PW. Etiology and outcome of fulminant hepatic failure managed at an Australian liver transplant unit. J. Gastroenterol. Hepatol. 19(2), 154–159 (2004).

28 Khuroo MS, Kamili S.

Aetiology and prognostic factors in acute liver failure in India. J. Viral. Hepat. 10(3), 224–231 (2003).

29 Oketani M, Ido A, Tsubouchi

H. Changing etiologies and outcomes of acute liver failure: a perspective from Japan. J. Gastroenterol. Hepatol. 26(1), 65–71 (2011).

www.futuremedicine.com

47

About the Authors Heather Patton Heather Patton is a transplant hepatologist and Assistant Clinical Professor of Medicine at the University of California, San Diego (UCSD; CA, USA). Her clinical and research interests are in complications of cirrhosis and she is the Primary Investigator for the UCSD site of the North American Consortium for the Study of End Stage Liver Disease (NACSELD). She has developed an evidence-based protocol for the management of acute liver failure for the UC San Diego Health System.

Robert G Gish Robert G Gish, MD, is the Medical Director of the Center for Hepatobiliary Disease and Abdominal Transplantation (CHAT), the Chief of Clinical Hepatology, and a Professor of Clinical Medicine at the University of California, San Diego (CA, USA). His clinical research focuses on viral hepatitis, liver cancer, end-stage liver disease, liver transplant, bioartificial liver, and public policy related to liver transplantation and viral hepatitis. He has a quarter century of experience as a clinician and clinical investigator. He has published over 100 articles in peer-reviewed journals, and given thousands of lectures on liver disease in the USA and a dozen other countries.

48

48

© 2013 Future Medicine www.futuremedicine.com

Chapter

4 Investigation and diagnostic pathways

Investigations aimed at determining the etiology of acute liver failure 50 Diagnostic studies aimed at determining prognosis 59 Diagnostic studies aimed at monitoring clinical progress of the ALF patient 60

Heather Patton & Robert G Gish Once a diagnosis of acute liver failure (ALF) has been established (see Chapter 1), the next key steps in patient evaluation are to determine the etiology of ALF, when possible, and to assess the patient’s prognosis (in particular, the need for liver transplant). Concurrent with these diagnostic evaluations, the patient requires diligent monitoring for development of complications so that these may be identified and managed in a timely fashion. This chapter will present an approach to the diagnostic evaluation of the ALF patient organized according to those studies aimed at: determining the diagnosis; establishing a prognosis; and following progression and monitoring for complications. This chapter is designed to help facilitate a rapid and accurate assessment of these patients given that their condition may evolve quickly and, in cases where a treatable etiology can be identified, medical management may increase the likelihood of overall survival for those patients who are not liver transplant candidates, and of transplant-free survival where a transplant can be offered.

doi:10.2217/EBO.12.325

© 2013 Future Medicine

49

Patton & Gish As drug-induced liver injury is the predominant cause of acute liver failure (ALF) in the USA and western Europe, drug history in cases of ALF must include: acetaminophen-containing products, antibiotics, anticonvulsants, and nonsteroidal antiinflammatory agents taken within the past 6 months.

Investigations aimed at determining the etiology of acute liver failure The preponderant causes of acute liver failure (ALF) vary geographically, largely according to the prevalence of hepatotropic viruses and population patterns of medication use and abuse, as well as drug abuse. Thus, the resources dedicated to diagnostic evaluation may vary regionally according to the most common etiologies observed in the local population [1]. For example, in developing countries, viral etiologies are predominantly the causal agents of ALF whereas in the USA and western Europe, drug-induced liver injury accounts for the majority of cases (see Chapter 3). The attainment of a thorough history from patients with ALF may be challenging and, in some cases, impossible due to encephalopathy. As mental status changes can evolve quickly, medical personnel should interview patients as soon as possible so as not to miss the window of opportunity to obtain important historical details from the ALF patient. Family and other close contacts may serve as important sources of information in cases where patients have severe encephalopathy. The key aspects of history relevant to ALF are detailed below: Past medical history: relevant elements include history of autoimmune diseases, thrombotic disorders, psychiatric disease, cardiovascular disease, malignancy, pain syndromes and immunocompromised status;

n

Timing of onset of symptoms: time course from first symptoms, which may be nonspecific, such as fatigue or malaise, and the onset of jaundice and subsequently altered mental status;

n

Acetaminophen (paracetamol) product use: most ingestions resulting in ALF exceed 10 g/day with exceptional cases of severe injury reported in those with exposures as low as 3–4 g/day;

n

Review of all medication exposures from the last 3–6 months: the most common classes of drugs implicated in druginduced liver injury are anticonvulsants, Severe transaminitis (aspartate amino­ transferase/alanine aminotransferase antimicrobials and nonsteroidal anti>1000 U/l) without significant jaundice should raise inflammatory agents. Specific questions concern for acetaminophen overdose, early viral about over-the-counter, herbal/natural/ hepatitis and shock liver, while with a severe hepatitis supplemental agents should also be asked, accompanied by a dramatically elevated bilirubin, a recognizing that patients may not consider diagnosis of autoimmune hepatitis should be n

considered.

50

www.futuremedicine.com

Investigation & diagnostic pathways this history to be relevant and/or may be reluctant to share this history with healthcare providers; Alcohol and drug use: both remote and recent use with details regarding type, quantity, route and duration;

n

Travel history: particularly to regions endemic for hepatotropic viruses, such as Asia, the Middle East, Africa and Central and South America [101,102];

n

Risk factors for hepatitis A and E: ingestion of fecally contaminated food or drinking water, close household contact with infected family members or sexual partners, and travel to endemic locations;

n

Risk factors for hepatitis B, C and D, and HIV: injection drug use, sexual activity, needle stick exposures, snorted drugs with shared paraphernalia, and tattoos (especially when obtained in countries where unsterilized needles and knives are commonly reused in tattoo shops);

n

Exposure to mushrooms from the Amanita genus: most common in Europe though also reported in northern Africa, Asia, and parts of the USA (Pacific Northwest, Pennsylvania, New Jersey, Ohio and Texas). Typical symptoms include nausea, vomiting, abdominal pain and diarrhea occurring within hours to a day of ingestion.

n

The initial diagnostic studies in patients with ALF need to be fairly extensive so as to arrive at a determination of etiology and prognosis as quickly as possible. The following comprehensive diagnostic studies are recommended: Acetaminophen level: should be obtained as soon as possible with any measurable level raising the concern about toxicity;

n

Toxicology screen: for drugs of abuse and blood alcohol levels;

n

Viral hepatitis serologies: the most common hepatotropic virus infections associated with ALF should be routinely tested (see section on diagnostic evaluation of hepatotropic viruses);

n

Abdominal ultrasound with Doppler: to assess liver size, contour, vascular patency, and presence of intrahepatic mass or hemorrhage;

n

HIV-1, HIV-2: given potential impact on type and severity of liver disease as well as transplant candidacy;

n

Ceruloplasmin level and 24-h urine copper: additional studies are warranted if clinical suspicion for Wilson’s disease (WD) is high (see later);

n

www.futuremedicine.com

51

Patton & Gish Pregnancy test: in women of childbearing potential. Pregnancyassociated cases of ALF typically present in the third trimester; thus, patients will be aware and confirmatory testing not indicated;

n

Markers of autoimmune liver disease: autoantibodies and immunoglobulin levels;

n

Transjugular liver biopsy [2]: indicated if autoimmune hepatitis, metastatic liver disease, WD, herpes simplex hepatitis [3], or infiltrative disease (lymphoma, leukemia) are suspected.

n

Acetaminophen (ACM)-induced hepatotoxicity is the predominant cause of ALF in the USA and UK. ACM overdose may either be intentional (suicide attempt) or due to the inadvertent consumption of supratherapeutic quantities of analgesic agents. ACM hepatotoxicity is the exemplar of the hyperacute variant of ALF. Salient diagnostic features of ACM-induced hepatotoxicity include: Very high serum aminotransferase levels (aspartate aminotransferase [AST] and alanine aminotransferase [ALT], typically >1000 IU/l) with low bilirubin levels [4]. Aminotransferase levels in excess of 3500 IU/l are highly suggestive of ACM;

n

Most commonly, after intentional major overdose, patients will have detectable levels of ACM while with unintentional overdose, it is more likely that intake may have been staggered, resulting in ACM levels that are low or even undetectable. However, it is important to be aware that even with intentional overdose some cases will present after a delay when ACM will be undetectable;

n

Rapidly progressive course with early onset of multiorgan failure;

n

Pancreatitis;

n

Features of unintentional overdose include use of multiple ACM‑containing products and use of narcotic ACM preparations [5];

n

Features of intentional overdose include larger total doses and concomitant consumption of alcohol and/or other drugs.

n

Cases of occult ACM overdose (undetectable ACM level in a patient unaware of having taken an overdose and/or unable to provide a history of ACM use due to severe hepatic encephalopathy) may be identified through an assay that detects acetaminophen-containing protein adducts that are released into the plasma by dying hepatocytes [6]. Importantly, these adducts are not found in cases of ACM use not associated with

52

www.futuremedicine.com

Investigation & diagnostic pathways hepatotoxicity. Presently, the ACM adduct assay is not routinely available for clinical use. Thus, clinicians need to rely on adjunctive evidence of ACM overdose as reviewed earlier. In the developing world, acute viral hepatitis is the predominant cause of ALF. Hepatitis E is the primary cause of ALF in India, Pakistan, China, and Southeast Asia [7]. The most commonly observed viral causes for ALF (hepatitis A, B and E) should be tested for in all patients presenting with ALF with the following serologies: n Anti-hepatitis A virus (HAV) IgM; Hepatitis B virus (HBV) surface antigen (HBsAg); if positive, reflex to HBV DNA, HBeAg and anti-HBe;

n

Anti-HBV core (HBc) IgM; if positive, reflex to HBV DNA quant;

n

Anti-hepatitis E virus (HEV) IgM; if positive, a subsequent order needs to be placed for blood and stool PCR testing for HEV to confirm infection.

n

Historically, patients with ALF from acute hepatitis B infection and those with ALF from an acute exacerbation of chronic hepatitis B infection could not be distinguished from each other unless historical or histological evidence of chronicity was available. However, data from the Acute Liver Failure Study Group (ALFSG) published this year reported that those with acute HBV infection had much higher IgM anti-HBc titers, lower viral loads (HBV DNA) and better prognosis in comparison to those with acute exacerbation of chronic HBV infection [8]. While felt to be quite rare, cases of ALF from acute hepatitis C virus (HCV) infection have been reported in Japan and India; thus, testing for anti-HCV and HCV RNA may be considered. Other uncommon but documented viral causes of ALF that may be tested in specific clinical settings include the following (shown in parentheses are serological markers of acute rather than chronic/historical infection): n Herpes simplex virus (HSV) 1 and 2 antibody (IgM antibody-positive or HSV DNA-positive with negative IgG antibody); Anti-varicella zoster virus (VZV; fourfold rise in antibody titer over 10–14 days);

n

Human herpesvirus 6 (HHV-6; IgM antibody-positive with negative IgG antibody or PCR-positive with negative IgG antibody);

n

Epstein–Barr virus (EBV; IgM of viral capsid antigen [VCA] and/or IgM early antigen with negative antibody to nuclear antigen [NA], or PCRpositive with negative IgG to VCA and NA);

n

www.futuremedicine.com

53

Patton & Gish Cytomegalovirus (CMV; PCR-positive with negative IgG antibody or IgM antibody >1:8 with negative IgG antibody);

n

Parvovirus B19 (PCR-positive with negative IgG antibody or IgM antibody-positive with negative IgG antibody).

n

Herpes virus infection is most commonly reported in patients who are immunosuppressed or during pregnancy (particularly the third trimester) but has also been reported very rarely in healthy individuals. Approximately one quarter of patients with autoimmune hepatitis (AIH) present acutely with jaundice, a small subset of whom progress to ALF. Features suggestive of AIH include young age, female gender, Caucasian race, negative viral serologies, history-negative for drug and alcohol use, and high transaminases (ALT and AST often elevated to ≥1000 IU/l) with a high total bilirubin level (often ≥20  mg/dl). Recommended studies for evaluation of autoimmune hepatitis include: Anti-nuclear antibody (ANA);

n

F-actin (smooth muscle) antibody;

n

Anti-LKM-1 antibody, especially in children and young adults;

n

Quantitative immunoglobulins.

n

Note that autoantibodies may be absent in up to 30%; thus, liver biopsy should be strongly considered if there is any degree of clinical suspicion for AIH (Figure 4.1) [9]. Histological features of AIH presenting with ALF from the USA ALF registry were: massive hepatic necrosis (present in 42% of sections), presence of lymphoid follicles (32%), a plasma cell-enriched inflammatory infiltrate (63%), and central perivenulitis (65%) [9]. Ischemic hepatitis (‘shock liver’) results from arterial hypoxemia and/or impaired hepatic perfusion. While this is not an unusual diagnosis among critically ill patients, it is an uncommon cause for liver failure (4.4% of patients enrolled in the AFLSG) [10]. Echocardiogram should be obtained in cases where ischemic hepatitis is suspected as patients with congestive hepatopathy are at increased susceptibility for ischemic hepatitis. Typical clinical features of ischemic hepatitis include: n Rapid and significant transaminitis (ALT may reach or exceed 1000 IU/ml); Elevated lactate dehydrogenase (LDH) levels;

n

Markers of muscle necrosis;

n

Acute kidney injury.

n

54

www.futuremedicine.com

www.futuremedicine.com B

C

Patient was a 19-year-old male who presented with 1 month of jaundice and fatigue. His laboratory evaluation was significant for: aspartate ­aminotransferase 1454 U/l, alanine aminotransferase 1244, total bilirubin 25 mg/dl, international normalized ratio 1.7, antinuclear antibody > 1:640 and IgG 3551 mg/dl. (A) Perivenulitis (40x); (B) interface activity and plasma cells (40x); and (C) parenchymal collapse and ballooning (10x). Images reproduced with permission of Michael Peterson, Department of Pathology, University of California, San Diego.

A

Figure 4.1. Histological features of severe acute autoimmune hepatitis.

Investigation & diagnostic pathways

55

Patton & Gish Episodes may be precipitated by cardiac events, such as congestive heart failure, myocardial infarction, cardiopulmonary arrest and cardiac arrhythmias, as well as non-cardiac events, such as sepsis and seizures. Typically, there is early recovery of hepatic function. While WD is a rare cause for ALF both in the USA and Europe (2–3% in the USA ALFSG) [11], early recognition is important as this is universally fatal without liver transplantation [12]. Suspicion for WD should be heightened in the following: Age 5–40 years;

n

Coombs-negative hemolytic anemia with high indirect-reacting bilirubin;

n

Rapidly progressive acute renal failure (due to renal tubular damage from copper);

n

Decreased uric acid level;

n

High total bilirubin (mg/dl) to alkaline phosphatase (IU/l) ratio (>2.0);

n

Normal or very low alkaline phosphatase.

n

Note that while ceruloplasmin is recommended as the screening test for WD, levels may be normal in 15% of cases, may be elevated as an acute phase reactant, and may be decreased in up to 50% of patients with ALF from causes other than WD. Thus, in cases of suspected WD, other diagnostic studies to confirm diagnosis may include: Kayser–Fleischer rings (Figure 4.2) may be observed in approximately 50% of WD patients with an acute presentation;

n

Urinary copper levels;

n

Transjugular liver biopsy with copper quantification;

n

Acute Budd–Chiari syndrome (BCS; hepatic vein thrombosis) is a rare cause of ALF that classically presents with: Abdominal pain;

n

Ascites;

n

Significant hepatomegaly.

n

Abdominal ultrasound with Doppler is the initial diagnostic test of choice. If nondiagnostic, other imaging studies such as computed tomography, magnetic resonance imaging, or venography may be obtained (Figure 4.3). Transjugular liver biopsy demonstrates central hepatic congestion, necrosis, portal central bridging and interhepatocyte hemorrhage. Note that

56

www.futuremedicine.com

Investigation & diagnostic pathways percutaneous liver biopsy should not be Figure 4.2. Kayser–Fleischer ring: copper attempted in this setting as intractable deposition in Descemet’s membrane of the peritoneal hemorrhage may ensue due to cornea. hepatic outflow obstruction. Once diagnosed with BCS, further investigations are required to identify an underlying hypercoagulable state (polycythemia, genetic disorders of coagulation, paroxysmal nocturnal hemoglobinuria, malignancy) as precipitant for the acute thrombosis as this may impact transplant candidacy. A rare complication seen in late pregnancy (typically third trimester), ALF poses threat to both mother and fetus. However, prompt recognition of acute fatty liver of pregnancy and HELLP (hemolysis, elevated liver enzymes, low platelets) syndrome offers the opportunity for good outcomes with rapid delivery. Diagnostic features of these pregnancy-related causes of ALF include:

These rings can be either dark brown, golden, or reddish-green, are 1–3 mm wide, and appear at the corneal limbus. With rare exceptions, they are diagnostic of Wilson’s disease. Reproduced from [103].

Jaundice, coagulopathy and thrombocytopenia;

n

Hypertension and proteinuria (features of pre-eclampsia) are frequently present;

n

Hepatic steatosis (may be identified on ultrasound);

n

Intrahepatic hemorrhage, hepatic rupture may rarely occur with these.

n

While there are cases in which a causative etiology cannot be established (so-called seronegative disease or indeterminate cause), this is a diagnosis of exclusion. Causes implicated in patients with seronegative disease include toxic exposures, autoimmune disease, and viral infections yet to be identified. Seronegative disease has a tendency to follow a subacute course and to have particularly low rates of spontaneous survival with supportive medical care [1]. Review of indeterminate cases by the US ALFSG identified several cases with ACM-protein adducts as well as several others with histology consistent with AIH, demonstrating the potential usefulness of M30 antigen: the M30 antigen is a caspase the ACM-protein adduct assay and cleaved neoepitope of cytokeratin 18, indicative of apoptotic hepatocyte cell death. This transjugular liver biopsies in these biomarker is detected with an ELISA assay and, in seronegative cases. Finally, while the combination with other clinical variables, may diagnostic laboratories and imaging perform well as an indicator of prognosis in ALF [23,26].

www.futuremedicine.com

57

58 B

C

On presentation, her aspartate aminotransferase was 1592 U/l, alanine aminotransferase 1244 U/l, total bilirubin 2.7 mg/dl and INR 1.8. Initial ultrasound with Doppler demonstrated absent flow in the hepatic veins as well as thrombosis of the main portal vein and concern for a liver mass. Further evaluation was carried out with MRI: (A) T2-weighted image shows increased signal in the periphery of the liver consistent with edema and relative sparing of the caudate lobe; (B) portal venous phase image shows heterogeneous enhancement of the liver with better enhancement of the caudate/central liver compared with the periphery (referred to as the ‘fan pattern’ of enhancement); and (C) delayed phase image still demonstrates heterogeneous enhancement of the liver, but the caudate lobe now looks lower in signal than portions of the periphery. The reversal of the enhancement pattern with the passage of time is referred to as the ‘flip-flop’ pattern of enhancement. The mass seen on ultrasound was demonstrated to be a regenerative nodule on MRI. Images reproduced with permission of Cynthia Santillan, Department of Radiology, University of California, San Diego.

A

Figure 4.3. MRI images from a 41‑year-old woman who presented with 4 weeks of abdominal distention.

Patton & Gish

www.futuremedicine.com

Investigation & diagnostic pathways discussed are often critical at arriving at a determination for the cause of ALF, there are limitations in the performance characteristics of all of these tests; some of the potential problems with these diagnostics are shown in Box 4.1.

Box 4.1. Diagnostic pitfalls in acute liver failure. ƒƒ Nodular contour to liver on imaging (may be seen in ALF and not indicative of cirrhosis in this setting) ƒƒ Undetectable acetaminophen level when there is a delay in presentation after ACM ingestion (may be detected with ACM protein adduct assay) ƒƒ Absent skin lesions in suspected herpes simplex virus infection (may be present in only 50% of cases) ƒƒ Low ceruloplasmin levels in cases not due to WD or normal ceruloplasmin levels in cases with WD ƒƒ Negative autoantibodies in suspected AIH (up to 30%)

Diagnostic studies aimed at determining prognosis There is perhaps no greater responsibility for a transplant hepatologist than to determine whether the ALF patient will require liver transplantation. The consequences of an inaccurate judgment mean that either a precious resource is consumed unnecessarily and an individual is left requiring lifelong immunosuppressant ALF: Acute liver failure; ACM: Acetaminophen; therapy or that a (typically) young and AIH: Autoimmune hepatitis; WD: Wilson’s disease. otherwise healthy young person dies without having had the opportunity for a life-saving intervention. There are no laboratory or clinical variables or prognostic models that have been proven to predict with 100% accuracy the need for liver transplant among patients with ALF. However, as reviewed in Chapter 10, there are several laboratory measures and models that have been devised towards this end. The recommended diagnostic studies to determine prognosis are: n Arterial blood gas (ABG; arterial pH) [13]; Arterial ammonia level [14,15];

n

Arterial lactate [16];

n

Serum phosphorus level [17,18];

n

Factor V level [19];

n

a-fetoprotein [20].

n

ABG should be obtained early following fluid resuscitation in order to determine arterial pH and lactate, both of which are components of the King’s College Hospital (KCH) criteria [13], widely used to determine need for liver transplant listing in the UK and the USA. Other variables included in the KCH Hepatomegaly in ALF may be seen with acute criteria are routinely obtained on hospital viral hepatitis, acute Budd–Chiari syndrome, congestive heart failure and malignant infiltration of admission and during routine clinical the liver.

www.futuremedicine.com

59

Patton & Gish follow-up (INR, bilirubin and creatinine). Other criteria applied frequently in northern Europe are the Clichy–Villejuif [21] which require determination of coagulation Factor V concentration. In addition to these Thromboelastogram: provides a global assessment of scoring systems, so-called supplemental coagulation incorporating the cumulative effect of the markers of prognosis include serum lactate interactions at various levels between plasma components (clotting proteins) and cellular level [16] , phosphate [17] , serum components (platelets, red and white blood cells, and Gc Globulin [22] and a-fetoprotein [20]. While microparticles) of coagulation. This functional assay not routinely monitored in patients with allows for dynamic assessment of different stages of chronic liver disease, there is a body of clot formation (initial clot formation, rate of fibrin literature supporting the prognostic formation and cross-linking, maximal clot strength, and fibrinolysis) [24,25]. significance of arterial ammonia in patients with ALF [14,15]. Recent data from the US ALF Registry proposed new criteria, the ALFSG index, comprised of coma grade, INR, levels of bilirubin and phosphorus, and measurements of the apoptosis biomarker M30 [23]. Note that while transjugular liver biopsy has been proposed to facilitate diagnosis of the underlying etiology of ALF in certain clinical scenarios, liver histology is not felt to provide an accurate estimate of prognosis (e.g., due to parenchymal collapse, estimation of percentage hepatocyte necrosis is prone to error). Gc  globulin: a multifunctional protein synthesized in the liver, the main physiologic activity of which is felt to be actin binding and actin scavenging, levels of which are profoundly reduced in ALF [27].

Diagnostic studies aimed at monitoring clinical progress of the ALF patient The potential complications of ALF include coagulopathy with bleeding, cerebral edema with intracranial hypertension (ICH), infection, hypoglycemia, hypo- and hyper-phosphatemia and acute kidney injury/renal failure. There have been two recent publications evaluating thromboelastogram (TEG) in patients with ALF, recognizing that prothrombin time alone reflects only one aspect of the coagulation process that is affected by liver failure [24,25]. Further data are needed to determine if use of TEG in this patient population allows for reduced blood product transfusions with equivalent patient outcomes with respect to bleeding and thrombotic complications. Development of cerebral edema and ICH are recognized to be the most serious complications of ALF given the potential for uncal herniation. There is no consensus regarding the use of intracranial pressure monitors; and practices vary widely among centers in their use (although the clinical usefulness of this monitoring is readily apparent). Other means of monitoring for evidence of Diagnostic criteria for acetaminophen-induced ICH are listed in Table 4.1; note that conhepatotoxicity include any ingestion >4  g, detectable acetaminophen level, and aspartate tinuous EEG monitoring, transcranial aminotransferase >1000 IU/l [5].

60

www.futuremedicine.com

Investigation & diagnostic pathways Doppler ultrasound, and Table 4.1. Recommended studies to monitor intraocular ultrasonography complications in the acute liver failure patient. are not widely available and Complication Test(s) their clinical utility in ALF Bleeding Serial INR, fibrinogen and platelet count remains to be determined. A measurement baseline head CT should be Thromboelastogram† considered with repeat Cerebral Serial pupillary and neurological examinations imaging for patients who edema/ Assessment for non-convulsant seizure activity progress to Grade III–IV intracranial (intermittent or continuous EEG monitoring†) encephalopathy. If antibiot- hypertension Jugular bulb oxygen saturation† Near-infrared spectrophotometry† ics are not given prophylactiTranscranial Doppler ultrasonography† cally (and data have not yet Intraocular ultrasonography† demonstrated survival benIntracranial pressure monitors efit with prophylactic antibiComputed tomography brain scan otics), surveillance cultures Infection Serial surveillance cultures and chest are recommended. Blood radiography glucose levels should be routinely monitored for hypo- Hypoglycemia Serial measurements of blood glucose glycemia. Renal failure is Acute kidney Serial creatinine levels common in ALF and may injury/renal Measurement of urine output Neutrophil gelatinase-associated lipocalin† contribute to mortality; failure Cystatin C† thus, careful monitoring of † renal function and manage- These items are not considered standard of care and/or may not yet have data demonstrating utility in the acute liver failure patient ment of hemodynamic population. derangement should be INR: International normalized ratio. undertaken. At present, there are no data in patients with ALF on biomarkers of acute kidney injury, such as cystatin C and neutrophil gelatinase-associated lipocalin, that may facilitate a more rapid detection of acute kidney injury, although both appear promising in patients with cirrhosis and those undergoing liver transplant. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the ­production of this manuscript.

www.futuremedicine.com

Neutrophil gelatinase-associated lipocalin (NGAL): a small protein expressed in the renal tubules, renal expression of Neutrophil gelatinaseassociated lipocalin is dramatically increased in kidney injury from a variety of causes, and NGAL is released into both urine and plasma. NGAL levels rise within 2 h of the insult, making NGAL attractive as an early biomarker of kidney injury (with data supporting its use in AKI in the setting of cirrhosis and liver transplantation) [28,29].

61

Patton & Gish Summary. ƒƒ As mental status changes can evolve quickly in acute liver failure (ALF), medical personnel should interview patients as soon as possible so as not to miss the window of opportunity to obtain important historical details from the ALF patient; family and other close contacts may serve as important sources of information in cases where patients have severe encephalopathy. ƒƒ While it is prudent to cast a wide diagnostic net early in ALF so as to arrive at an etiologic diagnosis as quickly as possible, an understanding of predominant causes in a given geographic region can help to tailor diagnostic resources based on most likely cause. ƒƒ Drug-induced liver injury (predominantly from acetaminophen) is the most common cause of ALF in the USA and western Europe; however, it is recommended to test for acute infection with the hepatotropic viruses most commonly implicated in ALF (hepatitis A, E and B viruses) in all patients. ƒƒ Imaging studies may be diagnostic for Budd–Chiari Syndrome and helpful for diagnosis of complications such as intrahepatic hemorrhage or rupture in pregnancy-associated ALF but may be falsely interpreted as showing cirrhosis in the setting of ALF (nodular contour may be seen related to parenchymal collapse). ƒƒ Transjugular liver biopsy may be indicated when there is clinical concern for diagnosis of autoimmune hepatitis, herpes hepatitis, or infiltrative liver disease (lymphoma) or in cases of seronegative disease. ƒƒ In addition to standard laboratory tests obtained routinely upon hospital admission (comprehensive metabolic panel, complete blood cell count, and prothrombin time/INR), diagnostic studies obtained to help assess prognosis in ALF may include (following fluid resuscitation) an arterial blood gas, arterial levels of ammonia and lactate, serum phosphorus, factor V level and a-fetoprotein. ƒƒ Diagnostic studies important in monitoring for complications include an in-depth assessment of coagulation profile (such as with a thromboelastogram), serial neurologic examinations (as well as more sophisticated testing for cerebral edema, where available), surveillance cultures and close monitoring of glucose and renal function (urine output and serum creatinine pending development of more sensitive indicators of acute kidney injury in this population).

References 1

Bernal W, Auzinger G, Dhawan A, Wendon J. Acute liver failure. Lancet 376(9736), 190–201 (2010).

2

Singhal A, Vadlamudi S, Stokes K et al. Liver histology as predictor of outcome in patients with acute liver failure. Transpl. Int. 25(6), 658–662 (2012).

3

Moldovan B, Mentha G, Majno P et al. Demographics and outcomes of severe herpes simplex virus hepatitis: a registry-based

62

study. J. Hepatol. 55(6), 1222–1226 (2011).

study. Hepatology 42(6), 1364–1372 (2005).

4

Khandelwal N, James LP, Sanders C, Larson AM, Lee WM, Acute Liver Failure Study Group. Unrecognized acetaminophen toxicity as a cause of indeterminate acute liver failure. Hepatology 53(2), 567–576 (2011).

5

Larson AM, Polson J, Fontana RJ et al. Acetaminopheninduced acute liver failure: results of a United States multicenter, prospective

6

Davern TJ 2nd, James LP, Hinson JA et al. Measurement of serum acetaminophenprotein adducts in patients with acute liver failure. Gastroenterology 130(3), 687–694 (2006).

7

Dalton HR, Bendall R, Ijaz S, Banks M. Hepatitis E: an emerging infection in developed countries. Lancet Infect. Dis. 8(11), 698–709 (2008).

www.futuremedicine.com

Investigation & diagnostic pathways 8

9

Dao DY, Hynan LS, Yuan HJ et al. Two distinct subtypes of hepatitis B virus-related acute liver failure are separable by quantitative serum immuno­globulin M anti-hepatitis B core antibody and hepatitis B virus DNA levels. Hepatology 55(3), 676–684 (2012). Stravitz RT, Lefkowitch JH, Fontana RJ et al. Autoimmune acute liver failure: proposed clinical and histological criteria. Hepatology 53(2), 517–526 (2011).

hyperammonemia is associated with complications and poor outcomes in patients with acute liver failure. Clin. Gastroenterol. Hepatol. 10(8), 925–931 (2012). 16 Bernal W, Donaldson N,

Wyncoll D, Wendon J. Blood lactate as an early predictor of outcome in paracetamolinduced acute liver failure: a cohort study. Lancet 359(9306), 558–563 (2002).

17

10 Taylor RM, Tujios S,

Jinjuvadia K et al. Short and long-term outcomes in patients with acute liver failure due to ischemic hepatitis. Dig. Dis. Sci. 57(3), 777–785 (2012).

18 Schmidt LE, Dalhoff K. Serum

phosphate is an early predictor of outcome in severe acetaminopheninduced hepatotoxicity. Hepatology 36(3), 659–665 (2002).

11 Ostapowicz G, Fontana RJ,

Schiodt FV et al. Results of a prospective study of acute liver failure at 17 tertiary care centers in the United States. Ann. Intern. Med. 137(12), 947–954 (2002).

19 Pereira LM, Langley PG,

Hayllar KM, Tredger JM, Williams R. Coagulation Factor V and VIII/V ratio as predictors of outcome in paracetamol induced fulminant hepatic failure: relation to other prognostic indicators. Gut 33, 98–102 (1992).

12 Korman JD, Volenberg I,

Balko J et al. Screening for Wilson disease in acute liver failure: a comparison of currently available diagnostic tests. Hepatology 48(4), 1167–1174 (2008).

13 O’Grady JG, Alexander GJ,

Hayllar KM, Williams R. Early indicators of prognosis in fulminant hepatic failure. Gastroenterology 97, 439–445 (1989).

14 Bhatia V, Singh R, Acharya

SK. Predictive value of arterial ammonia for complications and outcome in acute liver failure. Gut 55(1), 98–104 (2006).

15 Kumar R, Shalimar, Sharma H

Baquerizo A, Anselmo D, Shackleton C et al. Phosphorus as an early predictive factor in patients with acute liver failure. Transplantation 75(12), 2007–2014 (2003).

20 Schiodt FV, Ostapowicz G,

Murray N et al. a-fetoprotein and prognosis in acute liver failure. Liver Transpl. 12(12), 1776–1781 (2006).

21 Ichai P, Samuel D. Etiology

and prognosis of fulminant hepatitis in adults. Liver Transpl. 14(Suppl. 2), S67– S79 (2008).

22 Schiodt FV, Rossaro L,

et al. Persistent

www.futuremedicine.com

Stravitz RT et al. Gc-globulin and prognosis in acute liver

failure. Liver Transpl. 11(10), 1223–1227 (2005). 23 Rutherford A, King LY, Hynan

LS et al. Development of an accurate index for predicting outcomes of patients with acute liver failure. Gastroenterology 143(5), 1237–1243 (2012).

24 Agarwal B, Wright G, Gatt A

et al. Evaluation of coagulation abnormalities in acute liver failure. J. Hepatol. 57(4), 780–786 (2012).

25 Stravitz RT, Lisman T, Luketic

VA et al. Minimal effects of acute liver injury/acute liver failure on hemostasis as assessed by thromboelastography. J. Hepatol. 56(1), 129–136 (2012).

26 Bantel H, Lugering A,

Heidemann J et al. Detection of apoptotic caspase activation in sera from patients with chronic HCV infection is associated with fibrotic liver injury. Hepatology 40(5), 1078–1087 (2004).

27 Schiodt FV. Gc-globulin in

liver disease. Danish Med. Bull. 55(3), 131–146 (2008).

28 Portal AJ, McPhail MJ, Bruce

M et al. Neutrophil gelatinase-associated lipocalin predicts acute kidney injury in patients undergoing liver transplantation. Liver Transpl. 16(11), 1257–1266 (2010).

29 Verna EC, Brown RS, Farrand

E et al. Urinary neutrophil gelatinase-associated lipocalin predicts mortality and identifies acute kidney injury in cirrhosis. Dig. Dis. Sci. 57(9), 2362–2370 (2012).

63

Patton & Gish Websites 101 WHO. International travel

and health interactive. http://apps.who.int/ithmap

64

102 Centers for disease control

and prevention. Hepatitis E FAQs for health professionals. www.cdc.gov/hepatitis/HEV/ HEVfaq.htm

103 Connexions. Fred H, van Dijk

H. Images of memorable cases: case 9. http://cnx.org/content/ m15007/1.3

www.futuremedicine.com

65

About the Authors Catherine Paugam-Burtz University Paris Diderot, Sorbonne Paris Cité, F-75018, Paris, France; INSERM U773, CRB3, F-75018, Paris, France; Critical Care & Anesthesiology Department, AP-HP, Hôpital Beaujon, Hôpitaux Universitaires Paris Nord Val de Seine, F-75018, Paris, France

Emmanuel Weiss INSERM U773, CRB3, F-75018, Paris, France; Critical Care & Anesthesiology Department, AP-HP, Hôpital Beaujon, Hôpitaux Universitaires Paris Nord Val de Seine, F-75018, Paris, France

Richard Moreau Univ Paris Diderot, Sorbonne Paris Cité, F-75018, Paris, France; INSERM U773, CRB3, F-75018, Paris, France; Service d’Hépatologie, AP-HP, Hôpital Beaujon, Hôpitaux Universitaires Paris Nord Val de Seine, F-75018, Paris, France

66

66

© 2013 Future Medicine www.futuremedicine.com

Chapter

5 Disease-specific treatments

Acetaminophen poisoning68 NAC in nonacetaminopheninduced liver failure 68 ALF related to viral diseases

69

ALF related to autoimmune hepatitis 

70

Wilson’s disease

70

Hypoxic hepatitis

70

Budd–Chiari syndrome

70

Pregnancy-specific ALF

71

Lymphoma

71

Catherine Paugam-Burtz, Emmanuel Weiss & Richard Moreau Specific treatments for acute liver failure (ALF) are scarce and dedicated to a few causes of ALF [1]. Consequently, the diagnostic approach should be driven by the early recognition of these few etiologies. Moreover, all these treatments share a common feature: their administration has to be performed as early as possible in the evolution of the liver injury to provide a clinical benefit. Otherwise, they may have little effect and liver transplantation remains frequently the only therapeutic option for ALF (Table 5.1).

doi:10.2217/EBO.12.369

© 2013 Future Medicine

67

Paugam-Burtz, Weiss & Moreau N-acetylcysteine: is the paradigm for medical therapy of acute liver failure. N-acetylcysteine is an antioxidant treatment that scavenges reactive oxygen species.

Acetaminophen poisoning In therapeutic conditions, glucuronidation or sulfation to nontoxic metabolites accounts for 90% of the metabolism of acetaminophen. A small part of the drug is metabolized by cytochrome to a hepatotoxic metabolite. This metabolite is secondarily detoxified by interaction with glutathione. In case of acetaminophen overdose, hepatic gluthatione stores can be depleted leading to the increase of the hepatotoxic metabolite concentration. N-acetylcysteine (NAC) restores hepatic glutathione [2]. NAC also acts as a reactive oxygen species scavenger [3].

From a clinical point of view, NAC has been shown to reduce liver injury and probably to improve prognosis of acetaminophen overdose [4]. It also has other beneficial effects, such as improving systemic hemodynamics and oxygen use and decreasing cerebral edema [5]. NAC has to be administered for all cases of suspected acetaminophen-induced acute liver failure (ALF), whether related to overuse or therapeutic administration and risk factors. Intravenously, the recommended scheme is a loading dose of 150 mg/kg over a period of 15–60 min, followed by an infusion of 12.5 mg/kg/h over a 4-h period, and finally an infusion of 6.25 mg/kg/h over a 16-h period [2]. NAC can also be administered orally. NAC treatment should be introduced as soon as possible; however, this treatment remains of value even in delayed administration after overdose [2]. The most common adverse effects of intravenous NAC are allergic reactions, such as rash, pruritus, urticaria, bronchospasm, tachycardia, and hypotension. Vomiting and flushing are also commonly reported [2]. Flushing does not require specific treatment. Other effects might require antihistamines or bronchodilatators [2]. These adverse effects are moderate.

NAC in nonacetaminophen-induced liver failure As mentioned earlier, NAC has been shown to have extrahepatic beneficial effects for ALF patients. Moreover, experimental studies in mice have shown that NAC attenuates cerebral complications of nonacetaminopheninduced ALF mechanisms, suggesting that NAC would have a beneficial effect in nonacetaminophen-induced ALF [6]. N-acetylcysteine restores hepatic gluthation stores and is an oxygen species scavenger. It reduces liver injury related to acetaminophen overuse.

68

In a prospective, randomized, double blind, placebo-controlled trial, 173 patients with ALF without clinical or historical evidence of acetaminophen overdose were

www.futuremedicine.com

Disease-specific treatments randomized to receive NAC or placebo infusion for 72  h [7]. Overall survival at 3  weeks (primary outcome) was 70% for patients given NAC and 66% for patients given placebo (p  =  0.28). Transplant-free sur vival (secondar y outcome) was significantly better for NAC patients (40%) than for placebo (27%; p = 0.043). Subgroup analysis show that the benefits of transplant-free survival were significant for the patients with encephalopathy grades 1–2 (52 vs 30% for placebo; p  =  0.01). Noteworthy, patients with advanced coma grades did not benefit from NAC. Based on these results, NAC treatment is frequently considered for nonacetaminophen-induced ALF and tends to be extensively used in ALF whether acetaminophen-induced or not [4].

Table 5.1. Specific treatments for acute liver failure. Etiology

Treatment

Acetaminophen-related

Acetylcysteine

Nonacetaminophenrelated

Acetylcysteine?

HSV or VZV related

Acyclovir

HBV related

Lamivudine

Primary Budd–Chiari syndrome

Anticoagulation TIPS if anticoagulation failure

Pregnancy-specific

Delivery

Lymphoma

Chemotherapy

HBV: Hepatitis B virus; HSV: Herpes simplex virus; TIPS: Transjugular intrahepatic portosystemic shunt; VZV: Varicella Zoster virus.

ALF related to viral diseases Viral hepatitis Antiviral therapy is ineffective in the treatment of severe ALF related to hepatitis A virus. For acute severe hepatitis B virus (HBV)-related fulminant liver failure, early (before overt encephalopathy) antiviral treatment particularly with lamivudine, may prevent further deterioration of liver function and improve survival without transplantation [8]. Other antiviral therapies, such as entecavir or tenofovir, might also be utilizable [9]. Ribavirin can be used for the treatment of acute hepatitis E. Whether ribavirin would prevent progression to liver failure in patients with severe acute hepatitis E remains to be determined and liver transplantation is often the only treatment option in such cases [10]. Viral diseases Herpes simplex virus (HSV)-related hepatitis is rare but severe. Delayed diagnosis without antiviral therapy significantly contributes to the unfavorable outcome [11]. Since results of diagnostic tests might not be readily available, antiviral therapy should be pre-emptively administered with acyclovir in cases of suspicion of HSV or in cases of ALF of unknown origin without When acute liver failure related to herpes or waiting virological confirmation [11,12]. varicella zoster virus is suspected, antiviral Despite this early pre-emptive treatment, treatment should be started before diagnosis confirmation.

www.futuremedicine.com

69

Paugam-Burtz, Weiss & Moreau There are specific therapies for acute liver failure related to either HBV, herpes simplex virus, varicella zoster virus, primary Budd–Chiari syndrome, pregnancy or lymphomas.

emergency liver transplantation frequently remains the only therapeutic option [13]. Acyclovir should also be administered in case of ALF related to varicella zoster virus.

ALF related to autoimmune hepatitis Autoimmune hepatitis may be revealed on investigation of an ALF patient at initial presentation. Whether corticosteroids might prevent liver transplantation in these conditions, is unclear especially in the more severe presentations [14]. Moreover, infectious side effects of these treatments must be considered [15]. Wilson’s disease Liver transplantation remains the preferred treatment for fulminant hepatic failure related to Wilson’s disease. However, in moderate cases before encephalopathy onset, early administration of d-penicillamine could be associated with survival without transplantation [16]. Hypoxic hepatitis As far as today, we cannot consider that there is any specific treatment for ALF related to hypoxic hepatitis. The main treatment of this condition is supportive care of the underlying organ failure. However, these treatments can be considered as etiologic treatment since they aim at increasing the oxygen delivery to hepatic cells [17]. Whether drugs aiming at reducing the consequences of ischemia reperfusion injuries might be effective, is unknown. Budd–Chiari syndrome Budd–Chiari syndrome (BCS) can be divided into ‘secondary’ BCS when related to compression or invasion by a lesion originating outside the veins (e.g., benign or malignant tumor, abscess, cyst); and ‘primary’ BCS when related to a primarily venous luminal disorder (thrombosis or phlebitis) [18]. Treatment of secondary BCS is mainly based on the treatment of the underlying cause [18]. Regarding primary BCS, the following therapeutic strategy is used in all patients [19]. First, they receive anticoagulation with low-molecular-weight heparin, rapidly shifted to vitamin K antagonists targeting an international normalized ratio of 2:3. Monitoring of platelet count is required because of the high incidence of heparin-induced thrombocytopenia in BCS patients. Second, when possible, the underlying prothrombotic conditions are treated. Third, a short-length stenosis of the inferior vena cava or of a major hepatic vein should be searched for. When such a stenosis is found, the stenosis is managed with percutaneous

70

www.futuremedicine.com

Disease-specific treatments Causes of acute liver failure that deserve a angioplasty and stenting. General care specific treatment should be promptly inc lud es treatm ent of as c i tes, recognized. In these cases, early administration of the gastrointestinal bleeding, infections, renal disease specific treatments, especially before overt f ailure and encephalopathy as encephalopathy might help to avoid liver recommended for other patients with acute transplantation. or chronic liver disease [19]. Transjugular intrahepatic portosystemic shunt (TIPS) insertion should be considered if the patient fails to improve after the above steps. TIPS insertion is feasible in over 90% of the cases in experienced hands. Liver transplantation should be considered in patients with a technical failure to insert TIPS or a lack of improvement after TIPS [19]. This therapeutic strategy has allowed the achievement of 5-year survival rates in the order of 90% [19,20].

Pregnancy-specific ALF These may be related to pre-eclampsia, the HELLP syndrome (hemolysis, elevated liver enzyme levels, and a low platelet count) and acute fatty liver of pregnancy. Liver abnormalities generally resolve promptly after delivery [21]. Lymphoma Hepatic involvement by hematologic malignancies although frequent, rarely causes severe hepatic dysfunction. Furthermore, ALF as a manifestation of a hematologic malignancy is extremely uncommon, although some cases have been reported in the literature [22,23]. Transjugular liver biopsy may be useful to diagnose the nature of liver infiltration [24]. Occasional cases of improvement of liver function by chemotherapy have been reported [22]. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

Summary. ƒƒ There are a relatively low number of causes of acute liver failure that deserve a specific treatment. However, these treatments have to be promptly administered. ƒƒ N-acetylcysteine should be given in all cases of acetaminophen-induced acute liver failure and may be considered in nonacetaminophen-related acute liver failure especially in early stage of the disease (before advanced coma grades). 

www.futuremedicine.com

71

Paugam-Burtz, Weiss & Moreau References 1

Ichai P, Samuel D. Liver transplantation for fulminant hepatitis. Gastroenterol. Clin. Biol. 33, 51–60 (2009).

2

Heard KJ. Acetylcysteine for acetaminophen poisoning. N. Engl. J. Med. 359, 285–292 (2008).

3

Downs I, Liu J, Aw TY, Adegboyega PA, Ajuebor MN. The ROS scavenger, NAC, regulates hepatic Va14iNKT cells signaling during Fas mAb-dependent fulminant liver failure. PLoS ONE 7, e38051 (2012).

4

Stravitz T, Kramer D. Management of acute liver failure. Nat. Rev. Gastroenterol. Hepatol. 6, 552–553 (2009).

5

Harrison PM, Wendon JA, Gimson AE, Alexander GJ, Williams R. Improvement by acetylcysteine of hemodynamics and oxygen transport in fulminant hepatic failure. N. Engl. J. Med. 324, 1852–1857 (1991).

6

Bemeur C, Qu H, Desjardins P, Butterworth RF. IL-1 or TNF receptor gene deletion delays onset of encephalopathy and attenuates brain edema in experimental acute liver failure. Neurochem. Int. 56, 213–215 (2010).

7

8

Lee WM, Hynan LS, Rossaro L et al. Intravenous N-acetylcysteine improves transplant-free survival in early stage nonacetaminophen acute liver failure. Gastroenterology 137, 856–864, 64 e1 (2009). Tillmann HL, Hadem J, Leifeld L et al. Safety and efficacy of lamivudine in patients with severe acute or fulminant

72

hepatitis B, a multicenter experience. J. Viral. Hepat. 13, 256–263 (2006). 9

16 Durand F, Bernuau J, Giostra E

et al. Wilson’s disease with severe hepatic insufficiency: beneficial effects of early administration of D-penicillamine. Gut 48, 849–852 (2001).

Tillmann HL, Zachou K, Dalekos GN. Management of severe acute to fulminant hepatitis B: to treat or not to treat or when to treat? Liver Int. 32, 544–553 (2012).

17

10 Wedemeyer H, Pischke S,

18 DeLeve LD, Valla DC, Garcia-

Manns MP. Pathogenesis and treatment of hepatitis E virus infection. Gastroenterology 142, 1388–1397 e1 (2012).

11 Riediger C, Sauer P,

Matevossian E, Muller MW, Buchler P, Friess H. Herpes simplex virus sepsis and acute liver failure. Clin. Transplant. 23(Suppl. 21), S37–S41 (2009). Sebagh M et al. Herpes simplex virus-associated acute liver failure: a difficult diagnosis with a poor prognosis. Liver Transplant. 11, 1550–1555 (2005). Lakeman FD et al. Detection and diagnosis of herpes simplex virus infection in adults with acute liver failure. Liver Transplant. 14, 1498–1504 (2008).

14 Potts JR, Verma S. Optimizing

management in autoimmune hepatitis with liver failure at initial presentation. World J. Gastroenterol. 17, 2070–2075 (2011).

15 Ichai P, Duclos-Vallee JC,

Guettier C et al. Usefulness of corticosteroids for the treatment of severe and fulminant forms of autoimmune hepatitis. Liver Transplant. 13, 996–1003 (2007).

Tsao G. American Association for the study liver diseases. Vascular disorders of the liver. Hepatology 49, 1729–1764 (2009).

19 Plessier A, Rautou PE, Valla

DC. Management of hepatic vascular diseases. J. Hepatol. 56(Suppl. 1), S25–S38 (2012).

20 Darwish Murad S, Plessier A,

Hernandez-Guerra M et al. EN-Vie (European Network for Vascular Disorders of the Liver). Etiology, management, and outcome of the Budd– Chiari syndrome. Ann. Intern. Med. 151(3), 167–175 (2009).

12 Ichai P, Roque Afonso AM,

13 Levitsky J, Duddempudi AT,

Henrion J. Hypoxic hepatitis. Liver Int. 32, 1039–1052 (2012).

21 Hay JE. Liver disease in

pregnancy. Hepatology 47, 1067–1076 (2008).

22 Shehab TM, Kaminski MS, Lok

AS. Acute liver failure due to hepatic involvement by hematologic malignancy. Dig. Dis. Sci. 42, 1400–1405 (1997).

23 Armstrong LA, Toro DH,

Martinez-Souss J, Chinea B, Conde-Sterling D. Anaplastic T-cell lymphoma presenting as fatal acute liver failure. P. R. Health Sci. J. 24, 343–346 (2005).

24 Brousse N, Solal-Celigny P,

Degott C, Lebrec D. Hepatic biopsy by transvenous approach in hemopathies with coagulation disorders. Presse Med. 12, 2439–2441 (1983).

www.futuremedicine.com

73

About the Authors Alexander Wilmer Alexander Wilmer is Associate Clinical Professor at the Faculty of Medicine of the Catholic University of Leuven (Belgium) and Head of Clinic of the Medical Intensive Care Units of the University Hospitals Leuven (Belgium). His clinical research is focused mainly on clinical studies of the management of liver failure in the intensive care unit and nutritional support strategies in the critically ill. He is author or coauthor of numerous articles in medical journals, reviews and book chapters.

Frederik Nevens Frederik Nevens is Professor of Medicine and Director of the Research Laboratory of Hepatology at the KU Leuven, Belgium. He is Head of the Department of Hepatology at the University Hospitals Leuven. His translational and clinical research encompasses a variety of liver disease-related topics including the complications of liver disease and liver transplantation. He is or has been a board member of several national and international scientific societies. He is author or coauthor of more than 300 articles in medical journals and books.

74

74

© 2013 Future Medicine www.futuremedicine.com

Chapter

6 General supportive management

General supportive treatment

76

Specific treatment

83

Alexander Wilmer & Frederik Nevens The medical therapy of acute liver failure consists of disease-specific treatment (Chapter 5), general supportive management and the prevention/management of specific complications of the syndrome (Chapters 7–9). As acute liver failure is an orphan disorder, large clinical trials are not available and the management of acute liver failure to a large extent is based on observational studies along with clinical experience.

doi:10.2217/EBO.12.307

© 2013 Future Medicine

75

Wilmer & Nevens Critical care of patients with acute liver failure (ALF) is the key to their survival, and improvements in supportive intensive care unit (ICU) management together with liver transplantation (LT) has substantially improved their survival (Chapter 1). The incidence of death due to hemorrhage has decreased from 25 to 1.5. Even mild encephalopathy can indicate a life-threatening situation in the following hours. Cerebral edema can appear suddenly and will compromise the safety of the transfer. Therefore, in case of an evolving hepatic The hemodynamic alterations in ALF mimic the encephalopathy there is an indication for systemic hemodynamic changes seen in intubation and sedation prior to the transfer, patients with cirrhosis and portal hypertension: low systemic vascular resistance and renal artery as the transfer itself can be a risk factor for vasoconstriction.

76

www.futuremedicine.com

General supportive management increased intracranial pressure. Owing to the increased risk of hypoglycemia during and after the transfer, intravenous (iv.) glucose in a continuous drip should be given. Glycemic levels need to be controlled regularly, aiming at levels of ±140 mg/dl.

Box 6.1. Decisions to be made at presentation of acute liver failure. ƒƒ Early diagnosis of acute liver failure, as evolution is highly unpredictable ƒƒ Decision to administer N-acetylcysteine ƒƒ Transfer to a specialized unit in due time ƒƒ If this patient is a candidate for emergency liver transplantation

Invasive monitoring The first step when the patient arrives at the ICU is, in addition to routine laboratory testing, to start continuous monitoring of several parameters such as urinary output every 2 h, ECG monitoring , oxygen saturation, and frequent measurement of arterial blood gas and lactate (Figure 6.1). Ammonia levels should also be followed closely, with arterial samples to be preferential over venous samples. If not already available, chest radiography, baseline ECG, an electroencephalogram and liver echography should be obtained. An arterial catheter is required, but the administration of clotting factors such as fresh frozen plasma (FFP; 1–2 units) or blood platelets if the platelet count is 2 require intubation and mechanical ventilation. Appropriate monitoring should be instituted depending on the type of ventilatory support chosen. A nasogastric tube should be placed after intubation and the gastric residual monitored according to local practice. Nonconvulsive seizure activity may occur in ALF and more frequently in patients with advanced stages of HE. In cases of documented convulsive or nonconvulsive seizure activity, continuous electroencephalogram monitoring is recommended. In the rare patient developing important abdominal distention due to ascites or accumulation of intestinal gas, intra-abdominal pressure monitoring via the bladder method every 6 h is desirable. Prevention of infections ALF induces a state of immune deficiency. Infections are extremely frequent and one of the complications that directly affects

www.futuremedicine.com

The patient with ALF can evolve from mild to grade 3 or 4 encephalopathy with severe intracranial hypertension in a matter of only a few hours. Close and frequent neurologic evaluation is strictly necessary.

77

Wilmer & Nevens Figure 6.1. Diagnostic procedures, monitoring and standard care for acute liver failure at admission to the intensive care unit. Initial laboratory tests

Routine monitoring

Routine biochemistries Blood gas + lactate Arterial ammonia Cultures (respiratory, blood, urine) Factor V Blood group

ECG Oxygen saturation Arterial pressure Respiratory rate Urine output every 1–2 h Clinical neurological status every 1–2 h

Toxico-screening on urine and if indicated dosage of acetaminophen

If mechanically ventilated Ventilation parameters Gastric residuals

If candidate for LT

If abdominal distension

HLA typing CMV, HIV, hepatitis B and C status Length and weight

Intra-abdominal pressure every 4–6 h

Other tests if not already available Chest x-ray ECG EEG Liver echography

Standard care Quiet surroundings Head of bed >30° Head in neutral position If candidate for LT: strict isolation Glucose infusions: glycemic target ± 140 mg/dl Stress ulcer prophylaxis Lactulose N-acetylcysteine if encephalopathy grade 200 mM/l [9,10]. Several procedures decrease the risk for intracranial hypertension and should be part of the general management of ALF, for instance manipulation of the patient should be restricted to a minimum, the patient should be cared for in a quiet surrounding, and the head position of the bed must be elevated to >30° and kept in a neutral position (Figure  6.2). Spontaneous hyperventilation and spontaneous hypothermia up to 35°C can be allowed. Lactulose should be administered cautiously to avoid aspiration, diarrhea and gaseous distention of a degree that may interfere with LT. Standard doses are 30–60 ml orally or via nasogastric tube every 6 h aiming at 2–4 soft stools per day. Monitoring for neurological signs of cerebral edema should be instituted at two-hourly intervals and more frequently if the patient is deteriorating. Appropriate treatment needs to be started as soon as possible (Chapter 7). The value of intracranial pressure monitoring remains controversial [11].

www.futuremedicine.com

79

Wilmer & Nevens Figure 6.2. General support of established organ dysfunction associated with acute liver failure. Encephalopathy

Renal dysfunction

If encephalopathy grade >2: intubate If encephalopathy grade >3: try to maintain sodium levels between 145–155 mmol/l If EEG shows nonconvulsive epilepsy: start continuous EEG monitoring, if available, and treat according to local practice

Maintain MAP >70 mmHg Try to avoid positive fluid balances In cases of fluid overload, a trial with furosemide frip can be considered If indication for CRRT is present: choose CVVH Restrict CVVH with citrate to experienced centers

Respiratory failure

Hypotension

Utmost care at the time of intubation to prevent aspiration Short-acting medications, such as remifentanil and propofol, work well in ALF Ventilator settings aiming at ‘best PEEP’ PEEP preferably 65 mmHg Cristalloids are acceptable Albumin is preferred Avoid starches If needed: norepinephrine is the preferred vasoactive agent Terlipressine may be added safely Consider more invasive hemodynamic monitoring in case of severe hemodynamic instability In case of persistent hypotension: a trial with hydrocortisone can be considered

CRRT: Continuous renal replacement therapy; CVVH: Continuous veno–venous hemofiltration; EEG: Electroencephalogram; MAP: Mean arterial pressure; PEEP: Positive end-expiratory pressure.

Nutrition ALF is a catabolic state. Nutritional support by enteral route is recommended without restriction of proteins (80–100 g/day) [12]. Caloric targets in this population are unclear, but more than 600–800 kcal per day during the first week of ICU stay seems unnecessary. In patients in ICU there seems to be no benefit of supplemental parenteral nutrition during this time period [13]. Vitamin deficiency frequently appears to be present in patients with ALF and therefore vitamin supplementation is necessary [14]. Respiratory management Encephalopathy provokes a loss of the gag reflex and effective cough. In case of encephalopathy grade 3 or signs of cerebral edema, intubation and ventilation If the need for renal replacement therapy are necessary. Intubation should be done arises, continuous veno–venous hemofiltration very carefully to prevent aspiration and, in is the preferred modality, owing to lower swings in cases of encephalopathy with agitation, a intercranial pressure and more stable hemodynamics compared with intermittent hemodialysis .

80

www.futuremedicine.com

General supportive management As the INR is a very important prognostic short-acting muscle relaxant is best used factor, use clotting factors only for active after induction of loss of consciousness bleeding or invasive procedures. (Figure  6.2). For adequate sedation, medications with a short half-life such as propofol and remifentanil, or fentanyl are preferred. Remifentanil, in this setting, is particularly convenient owing to its metabolism by tissue and plasma esterases independent of liver or kidney function. Very high levels of positive end-expiratory pressure (PEEP; >15  mmHg) will increase intracranial pressure. The lowest PEEP that achieves adequate oxygenation should be used, to avoid a possible increase of intercranial pressure (ICP) [15]. Arterial PaCO2 increases ICP in patients with ALF [16]. The initial PaCO2 goal should be approximately 35 mmHg by adjusting minute ventilation and this can later be adapted in relation to the degree of ICP elevation. Hyperventilation to levels below 30 mmHg are effective in reducing ICP but not for more than approximately 48 h. Higher PEEP levels (>10 cm H2O) have been reported to reduce liver blood flow and hepatic oxygen delivery.

Strategies in case of renal impairment Besides acute tubular necrosis (urinary sodium level >20 mM/l with an active sediment), which occurs early and is due to the intrinsically nephrotoxic effects of some etiologies of ALF (acetaminophen or toxins from Amanita mushroom), mild functional renal failure is frequently found late in the course of ALF, mimicking the systemic hemodynamics of hepatorenal syndrome of patients with end-stage cirrhosis with a low systemic vascular resistance and renal artery vasoconstriction [17]. In this condition, the urinary sodium level is 65 mmHg at all times. Renal function is often ionized serum calcium [23,24]. At the first signs impaired in ALF and mean arterial pressures of renal dysfunction, mean arterial pressure >65  mmHg may help preserve kidney blood flow autoregulation.

www.futuremedicine.com

81

Wilmer & Nevens (MAP) needs to be increased to levels >70–75 mmHg in an attempt to preserve the autoregulation capacity of renal blood flow. Hyponatremia is a common finding in patients with ALF and requires extra attention because a rapid increase in serum sodium concentration can cause brain damage through osmotic myelinolysis. In patients with encephalopathy stage >2, maintenance of hypernatremia between 145 and 155 mM/l has been reported to reduce the risk and severity of intracranial hypertension [25]. Cardiovascular tone & its management Patients with ALF are relatively hypotensive with a normal to high cardiac output and vasodilatation with a low systemic vascular resistance, resembling the hemodynamics of end-stage cirrhosis or septic shock [26]. Due to mental alterations, patients have poor oral intake and therefore require iv. fluid with glucose and saline, aiming at a total daily fluid intake of 25–35 ml/kg under stable conditions. A low MAP will provoke cerebral hypoperfusion and worsen encephalopathy. In cases of hypotension, the patients need volume expansion (Figure 6.2). Cristalloids are acceptable, albumin is preferred and starches are not recommended. The aim is to keep the MAP >65 mmHg at all times. In case of persistent hypotension, with a fall to 7) FFP transfusion to maintain INR between 5 and 7 is advised. In cases of severe, persistent bleeding, rescue therapy with recombinant activated Factor VII may be considered [36]. Temporary correction of coagulopathy may be expected for a period of 2–6 h, although thrombotic complications are possible [37]. The incidence of upper gastrointestinal bleeding in patients with ALF is decreased by gastric acid suppression [38,39] and stress ulcer prophylaxis is recommended.

Specific treatment N-acetylcysteine NAC is the antidote for patients with acetaminophen-related liver injury and ALF (Chapter 5). Treatment with NAC is also beneficial in patients with other etiologies of ALF, either by improving systemic hemodynamics and tissue oxygen delivery [40–45] or via other mechanisms [46,47]. In patients with early-stage non-acetaminophen-related ALF, NAC improves transplant-free survival. Among those with encephalopathy grades 1–2 at admission, 52% of the patients who received NAC survived without LT versus 30% who received placebo [48]. Patients with advanced coma grades did not benefit in this study. The dosage and duration of the NAC treatment in this first randomized study was the same as for acetaminophen overdose: an initial loading dose of 150 mg/kg/h over 1 h followed by 12.5 mg/kg/h for 4 h and finally a continuous infusion of 6.25 mg/kg NAC for the remaining 67 h. Nausea and vomiting were the only symptoms seen more frequently with NAC therapy. These favorable results, however, were not observed in pediatric patients with non-acetaminophen-related ALF. Indeed, the 1-year LT-free survival was significantly lower with NAC [49]. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

www.futuremedicine.com .

83

Wilmer & Nevens Summary. ƒƒ General supportive management at intensive care units is the key to the survival of patients with acute liver failure (ALF). ƒƒ Even mild encephalopathy can indicate a life-threatening situation developing in the next few hours. Therefore, even at the first symptoms of mental alternations, the patient with ALF needs close monitoring and a referral to a specialized liver unit should be considered. ƒƒ The administration of clotting factors must be considered with utmost reserve, as it will influence the international normalized ratio, one of the most important factors of prognosis. ƒƒ Patients with ALF can be infected without obvious clinical or biochemical abnormalities, and in cases of sudden change in hemodynamics and the degree of encephalopathy, antibiotics need to be started on clinical suspicion. ƒƒ The risk of intracranial pressure significantly increases at a level of arterial ammonia >200 mM/l. In case of encephalopathy grade 3 or signs of cerebral edema, ventilation is necessary. ƒƒ Mild functional renal failure is frequently found late in the course of ALF and in this condition mean arterial pressure needs to be increased to >70–75 mmHg. Serum sodium increases intracranial pressure and should be maintained between 145 and 155 mM/l. ƒƒ To keep mean arterial pressure >65 mmHg, the following steps are necessary: daily fluid intake of 25–35 ml/kg, if hypotensive in spite of adequate volume substitution; norepinephrine (and, in case of failure, terlipressin can be added); and finally, with persistent need of vasopressors, a trial dose of hydrocortisone. ƒƒ In ALF, hemostasis is mostly preserved unless the platelet count is very low. Stress ulcer prophylaxis is recommended. ƒƒ N-acetylcysteine has been shown to be beneficial for non-acetaminophen-related ALF in adults if the grade of encephalopathy is 150 µM), sepsis, need for inotropic support, or hyponatremia often develop cerebral edema and high ICP. In such cases, ICP measurement is an essential component of the monitoring systems in the intensive care unit. It is only moderately invasive, and the complication rate has decreased with the development of new devices and knowledge of how to prevent such complications.

96

www.futuremedicine.com

Management of encephalopathy & cerebral edema In daily practice many interventions are based upon experimental observations and/or uncontrolled studies of patients with ALF. At first glance, most of these therapies work by reducing brain blood volume rather than by decreasing the amount of water within the brain tissue. Although this may be appropriate in some patients, it raises the concern of a critically restricted capillary flow that fail to provide adequate oxygen, glucose and other substrates to the brain. So far, clinical experience and reports indicate that cerebral edema can be prevented, and managed by the maintenance of high plasma tonicity and removal of ammonia by cRRT. There are some indications that liver assist systems may have a role in increasing ammonia clearance, in modulating the immune system and improving even transplant-free survival. In the future, development of treatments that specifically counteract cerebral edema formation will probably be based on the identification of the endogenous factor that dilates the cerebral arterioles and of the ion channels in the BBB that control cerebral water in- and out-flux. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

Summary. ƒƒ Hyperammonemia and high brain concentrations of organic osmolytes are of central importance in the pathophysiology of encephalopathy and brain edema. ƒƒ Systemic inflammation and recurrent infections together with hyperammonemia accelerates development of brain edema. ƒƒ Keeping a high tonicity is central to avoid high intracranial pressure. Early use of continuous renal replacement therapy is of central importance. ƒƒ Induction of hypothermia reverses high intracranial pressure, but its efficacy as a prophylactive intervention remains unsettled.

References 1

Bernal W, Auzinger G, Sizer E, Wendon J. Intensive care management of acute liver failure. Semin. Liver Dis. 28, 188–200 (2008).

2

www.futuremedicine.com

Norenberg MD. Oxidative and nitrosative stress in ammonia neurotoxicity. Hepatology 37, 245–248 (2003).

3

Zwingmann C, Chatauret N, Leibfritz D, Butterworth RF. Selective increase of brain lactate synthesis in experimental acute liver

97

Larsen & Bjerring failure: results of a [H-C] nuclear magnetic resonance study. Hepatology 37(2), 420–428 (2003). 4

Tofteng F, Larsen FS. Monitoring extracellular concentrations of lactate, glutamate, and glycerol by in vivo microdialysis in the brain during liver transplantation in acute liver failure. Liver Transpl. 8, 302–305 (2002).

5

Cordoba J, Gottstein J, Blei AT. Glutamine, myo-inositol, and organic brain osmolytes after portocaval anastomosis in the rat: implications for ammonia-induced brain edema. Hepatology 24(4), 919–923 (1996).

6

Larsen FS. Optimal management of patients with fulminant hepatic failure: targeting the brain. Hepatology 39, 299–301 (2004).

7

Larsen FS. Cerebral circulation in liver failure. Semin. Liver Dis. 16(3), 281–292 (1996).

8

Pedersen HR, Ring-Larsen H, Olsen NV, Larsen FS. Hyperammonemia acts synergistically with lipopolysaccharide in inducing changes in cerebral hemodynamics in rats anaesthetised with pentobarbital. J. Hepatol. 47, 245–252 (2007).

98

9

Keays R, Potter D, O’Grady J, Peachey T, Alexander G, Williams R. Intracranial and cerebral perfusion pressure changes before, during and immediately after orthotopic liver transplantation for fulminant hepatic failure. Q. J. Med. 79(289), 425–433 (1991).

Howard J, Williams R. The systemic inflammatory response syndrome in acute liver failure. Hepatology 32 (Pt 1), 734–739 (2000). 14 Murphy N, Auzinger G, Bernel

W, Wendon J. The effect of hypertonic sodium chloride on intracranial pressure in patients with acute liver failure. Hepatology 39, 464–470 (2004).

10 Tofteng F, Hauerberg J,

Hansen BA, Pedersen CB, Jorgensen L, Larsen FS. Persistent arterial hyperammonemia increases the concentration of glutamine and alanine in the brain and correlates with intracranial pressure in patients with fulminant hepatic failure. J. Cereb. Blood Flow Metab. 26, 21–27 (2006).

15 Clemmesen JO, Kondrup J,

Nielsen LB, Larsen FS, Ott P. Effects of high-volume plasmapheresis on ammonia, urea, and amino acids in patients with acute liver failure. Am. J. Gastroenterol. 96(4), 1217–1223 (2001).

16 Tofteng F, Larsen FS. The

effect of indomethacin on intracranial pressure, cerebral perfusion and extracellular lactate and glutamate concentrations in patients with fulminant hepatic failure. J. Cereb. Blood Flow Metab. 24, 798–804 (2004).

11 Bernal W, Auzinger G,

Dhawan A, Wendon J. Acute liver failure. Lancet 376(9736), 190–201 (2010).

12 Eefsen M, Dethloff T,

Frederiksen HJ, Hauerberg J, Hansen BA, Larsen FS. Comparison of terlipressin and noradrenalin on cerebral perfusion, intracranial pressure and cerebral extracellular concentrations of lactate and pyruvate in patients with acute liver failure in need of inotropic support. J. Hepatol. 47, 381–386 (2007).

13 Rolando N, Wade J, Davalos

M, Wendon J, Philpott-

17

Jalan R, O’Damink SW, Deutz NE, Lee A, Hayes PC. Moderate hypothermia for uncontrolled intracranial hypertension in acute liver failure. Lancet 354, 1164–1168 (1999).

18 Kurtovic J, Riordan SM,

Williams R. Fulminant hepatic failure. Curr. Treat. Options Gastroenterol. 8(6), 503–518 (2005).

www.futuremedicine.com

99

About the Authors Darren G Craig Darren G Craig is a Clinical Research Fellow at the Scottish Liver Transplantation Unit, Royal Infirmary, Edinburgh, UK

Kenneth Simpson Kenneth Simpson is a Senior Lecturer of Hepatology at the University of Edinburgh, and a Hon Consultant Physician at the Scottish Liver Transplantation Unit, Royal Infirmary, Edinburgh, UK

100

100

© 2013 Future Medicine www.futuremedicine.com

Chapter

8 Inflammation, infection and renal failure

Systemic inflammatory response syndrome

102

Cell death & ALF

103

Sterile inflammation & the double edged sword 106 Monocytes & macrophages

107

NK & NKT cells

108

Neutrophils

108

Infection

108

Renal failure

109

Proposed diagnostic criteria for the presence of AKI 109 Management of AKI in ALF & future directions 110

Darren G Craig & Kenneth Simpson Acute liver failure (ALF) is characterized by overwhelming innate immune activation. This results in increased cytokine production and expansion of cellular components of the innate immune system. The systemic inflammatory response syndrome is a characteristic feature of ALF, and is an important prognostic marker. Intense research is being undertaken to understand the pathophysiology of the systemic inflammatory response syndrome following liver injury. The heightened immune response in ALF is accompanied by a compensatory anti-inflammatory response, which may explain the high incidence of both bacterial and fungal infections in ALF. Close vigilance for infection should be maintained with a low threshold for commencing antimicrobial therapy. Renal failure is common in ALF and is multifactorial. Prevention of acute kidney injury requires optimization of fluid status, blood pressure and stopping nephrotoxins. Continuous renal replacement therapy is preferable to intermittent hemodialysis.

doi:10.2217/EBO.12.377

© 2013 Future Medicine

101

Craig & Simpson Acute liver failure (ALF) is a devastating clinical syndrome of diverse etiologies that is characterized by hepatic encephalopathy (HE) marked peripheral vasodilatation and multiple organ dysfunction syndrome. The clinical changes seen in ALF closely resemble those of patients with septic shock, and are associated with an overwhelming innate immune response characterized by excessive production of both pro- and anti-inflammatory cytokines, and mobilization of the cellular components of the immune system. Unravelling the links between the initial hepatotoxic injury, ongoing and subsequent liver injury, and the downstream systemic immune responses is an area of intense research in ALF as a means of devising alternative treatment strategies other than emergency orthotopic liver transplantation (OLT).

Acute liver failure (ALF) is associated with intense activation of the innate immune system, with elevations of both pro- and antiinflammatory cytokines, and mobilization of cellular components to aid tissue repair.

Systemic inflammatory response syndrome Systemic inflammatory response syndrome (SIRS) is the clinical manifestation of an abnormal generalized inflammatory reaction that may be precipitated by infection or tissue injury (i.e., sterile inflammation). The presence of the SIRS has been shown to be strongly associated with an adverse prognosis, irrespective of the presence of infection in several large ALF studies. In a study of 887 ALF patients admitted to a single center, Rolando et al. demonstrated the presence of at least one SIRS component in 504 (56.8%) of patients [1]. In the 353 noninfected patients in this study, a SIRS on admission was associated with deeper HE and greater mortality compared with noninfected patients without a SIRS. Similarly, Vaquero et al. and the US ALF Study Group demonstrated that increased numbers of SIRS components at admission predicted an increased likelihood of progression to deeper HE in a prospective study of 227 ALF patients [2]. The SIRS develops rapidly following acetaminophen (APAP) overdose; in a time course analysis of single time point APAP overdoses, the cumulative incidence of the SIRS was 67.5% by 48-h postoverdose, with a mortality rate in the SIRS group of 44.4%, compared with 7.7% for those patients not mounting a SIRS response. By 96-h postoverdose, the cumulative incidence of the SIRS had risen to 70%, with a mortality rate of 30% in this group; compared with a mortality rate of 0% in Systemic inflammatory response syndrome: the 30 patients without the SIRS [3]. SIRS two or more of: temperature >38 °C or 90 beats/min; respiratory rate >20 APAP overdose in patients that died or breaths/min or PaCO2 38 or 90 beats/min −− Respiratory rate >20 breaths/min or PaCO2 2- to 3-fold) from baseline

Less than 0.5 ml/kg/h for more than 12 h

3

Increase in serum creatinine to more than 300% (>3-fold) from baseline (or serum creatinine of ≥4.0 mg/dl [≥354 μmol/l] with an acute increase of at least 0.5 mg/dl [44 μmol/l])

Less than 0.3 ml/kg/h for 24 h or anuria for 12 h

www.futuremedicine.com

109

Craig & Simpson Box 8.3. Pathophysiological contributors to acute kidney injury in acute liver failure. ƒƒ Sepsis ƒƒ Systemic inflammatory response syndrome ƒƒ Hypovolemic renal hypoperfusion ƒƒ Drug-induced nephrotoxicity (glycoside antibiotics, contrast agents and acetaminophen) ƒƒ Cadmium release from necrotic liver ƒƒ Intra-abdominal hypertension

strongly associated with the presence of infection, and is virtually universal among ALF patients developing fungal infections (Box 8.3) [15].

Traditionally, AKI in ALF was thought to be mediated through hemodynamic changes and renal hypoperfusion, mirroring those seen in chronic liver disease. However, the SIRS has recently been shown to independently predict AKI in non-APAPinduced ALF [18], suggesting that the AKI of ALF may more accurately mirror the hemodynamic changes seen in acute sepsis. Patients with ALF develop a hyperdynamic circulation with marked renal vasoconstriction, increased plasma renin activity and reduced renal prostaglandin excretion. As a result, even modest reductions in mean arterial pressure can result in profound reductions in renal blood flow. Cytokinemia, reduced glomerular filtration pressure and potentially DAMP-induced tubular apoptosis may all contribute to the development of AKI in ALF.

The markedly increased incidence of AKI in APAP patients supports a direct nephrotoxic effect of APAP independent of liver injury [18]. This may reflect a specific acute tubular injury due to NAPQI production, endoplasmic reticular injury, or through tubular apoptosis driven by a Bcl-xL-dependent pathway. APAP decreases Bcl-xL protein in apoptotic tubular epithelium in vitro, and Bcl-xL overexpression has been shown to protect against APAP- and TNF-a-induced apoptosis, raising the prospect of targeting Bcl-xL therapeutically in the future [19]. Recently, cyclophilin A, a proinflammatory DAMP, was identified from the urine of patients with APAP-induced acute liver injury, but not in control groups [20]. This suggests that cyclophilin A could play a particular role in the pathogenesis of APAP-induced renal injury, and may have a role as a future renal biomarker in these patients.

Management of AKI in ALF & future directions Renal dysfunction is frequently secondary to hypovolemia, sepsis or nephrotoxins. Management should therefore focus on the prevention of renal failure through maintaining adequate systemic blood pressure, prompt identification and treatment of infections, and judicious use of contrast agents, since once established, the prognosis is considerably poorer [21]. Hypotension has been identified as an independent predictor of AKI in patients with ALF and patients therefore require aggressive

110

www.futuremedicine.com

Inflammation, infection & renal failure fluid resuscitation with either crystalloid or colloid and subsequent vasopressor therapy to maintain mean arterial pressure >75 mmHg [17]. There is no convincing evidence to support the use of low-dose dopamine in these patients. Clinical indications for introducing renal replacement therapy (RRT) and targets for therapy are poorly defined. In general, early introduction of RRT is recommended via continuous, rather than intermittent, methods of extracorporeal support to minimize circulatory instability and rises in intracerebral pressure. Replacement fluids/ dialysate without lactate may allow improved control of acidosis. Anticoagulation for dialysis requires careful monitoring and, occasionally, use of anticoagulant free or regional anticoagulation may be preferred. The majority of patients requiring RRT fully recover renal function either by the time of hospital discharge or following liver transplantation [22]. Predictors of complete renal recovery following APAP injury include: female gender; lower model for end-stage liver disease scores at day 3; patients with admission hypotension; and patients with lower grades of AKI [23]. Future studies should examine whether renal biomarkers, such as serum or urinary cystatin C, neutrophil gelatinase-associated lipocalin, IL-18 and kidney injury molecule-1 could be used to rapidly identify those liver injury patients at high risk of developing downstream AKI. Financial & competing interests disclosure The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

Summary. ƒƒ Acute liver failure is characterized by overwhelming innate immune activation. The innate immune system is a ‘double-edged sword’ and may cause collateral tissue damage when activated. ƒƒ Development of the systemic inflammatory response syndrome is an important clinical landmark, as it may precede multiple organ dysfunction syndrome. ƒƒ Acute liver failure patients are highly susceptible to bacterial and fungal infections. ƒƒ Clinicians should have a low threshold for initiating antimicrobial therapy in acute liver failure patients, but prophylactic antibiotics have not been proven to improve overall outcome. ƒƒ Renal failure is common in acute liver failure and is multifactorial. ƒƒ Prevention of renal failure is important; stopping nephrotoxins, limiting the use of intravenous contrast agents, and maintaining adequate circulating volume may all assist in this.

www.futuremedicine.com

111

Craig & Simpson References 1

2

3

4

5

6

7

Rolando N, Wade J, Davalos M, Wendon J, PhilpottHoward J, Williams R. The systemic inflammatory response syndrome in acute liver failure. Hepatology 32(4 Pt 1), 734–739 (2000).

8

Vaquero J, Polson J, Chung C et al. Infection and the progression of hepatic encephalopathy in acute liver failure. Gastroenterology 125(3), 755–764 (2003). Craig DG, Reid TW, Martin KG, Davidson JS, Hayes PC, Simpson KJ. The systemic inflammatory response syndrome and sequential organ failure assessment scores are effective triage markers following paracetamol (acetaminophen) overdose. Aliment. Pharmacol. Ther. 34(2), 219–228 (2011).

9

16 Karvellas CJ, Pink F, McPhail

M et al. Predictors of bacteraemia and mortality in patients with acute liver failure. Intensive Care Med. 35(8), 1390–1396 (2009).

17

10 Holt MP, Cheng L, Ju C.

Identification and characterization of infiltrating macrophages in acetaminophen-induced liver injury. J. Leukoc. Biol. 84(6), 1410–1421 (2008).

Rutherford AE, Hynan LS, Borges CBS et al. Serum apoptosis markers in acute liver failure: a pilot study. Clin. Gastroenterol. Hepatol. 5(12), 1477–1483 (2007).

12 Wu Z, Han M, Chen T, Yan W,

112

et al. Fungal infection: a common, unrecognized complication of acute liver failure. J. Hepatol. 12(1), 1–9 (1991).

Imaeda AB, Watanabe A, Sohail MA et al. Acetaminophen-induced hepatotoxicity in mice is dependent on Tlr9 and the Nalp3 inflammasome. J. Clin. Invest. 119(2), 305–314 (2009).

11 Possamai LA, Antoniades CG,

Ni HM, Bockus A, Boggess N, Jaeschke H, Ding WX. Activation of autophagy

15 Rolando N, Harvey F, Brahm J

Antoine DJ, Jenkins RE, Dear JW et al. Molecular forms of HMGB1 and keratin-18 as mechanistic biomarkers for mode of cell death and prognosis during clinical acetaminophen hepatotoxicity. J. Hepatol. 56, 1070–1079 (2012).

Antoniades CG, Berry PA, Davies ET et al. Reduced monocyte HLA-DR expression: a novel biomarker of disease severity and outcome in acetaminophen-induced acute liver failure. Hepatology 44(1), 34–43 (2006).

Bechmann LP, Jochum C, Kocabayoglu P et al. Cytokeratin 18-based modification of the MELD score improves prediction of spontaneous survival after acute liver injury. J. Hepatol. 53(4), 639–647 (2010).

controversies? Hepatology 48(3), 699–701 (2008).

protects against acetaminophen-induced hepatotoxicity. Hepatology 55(1), 222–232 (2012).

Anstee QM et al. Role of monocytes and macrophages in experimental and human acute liver failure. World J. Gastroenterol. 16(15), 1811–1819 (2010). Ning Q. Acute liver failure: mechanisms of immunemediated liver injury. Liver Int. 30(6), 782–794 (2010).

13 Liu ZX, Han D, Gunawan B,

Kaplowitz N. Neutrophil depletion protects against murine acetaminophen hepatotoxicity. Hepatology 43(6), 1220–1230 (2006).

14 Jaeschke H. Innate immunity

and acetaminophen-induced liver injury: why so many

Lee WM, Stravitz RT, Larson AM. Introduction to the revised American Association for the Study of Liver Diseases position paper on acute liver failure 2011. Hepatology 55(3), 965–967 (2012).

18 Leithead JA, Ferguson JW,

Bates CM et al. The systemic inflammatory response syndrome is predictive of renal dysfunction in patients with non-paracetamolinduced acute liver failure. Gut 58(3), 443–449 (2009).

19 Lorz C, Justo P, Sanz AB, Egido

J, Ortiz A. Role of Bcl-xL in paracetamol-induced tubular epithelial cell death. Kidney Int. 67(2), 592–601 (2005).

20 Dear JW, Simpson KJ, Nicolai

MP et al. Cyclophilin A is a damage-associated molecular pattern molecule that mediates acetaminopheninduced liver injury. J. Immunol. 187(6), 3347–3352 (2011).

21 Moore K. Renal failure in

acute liver failure. Eur. J. Gastroenterol. Hepatol. 11(9), 967–975 (1999).

22 Leithead JA, Ferguson JW,

Bates CM, Davidson JS,

www.futuremedicine.com

Inflammation, infection & renal failure Simpson KJ, Hayes PC. Chronic kidney disease after liver transplantation for acute liver failure is not associated with perioperative renal dysfunction.

Am. J. Transplant. 11(9), 1905–1915 (2011). 23 O’Riordan A, Brummell Z,

www.futuremedicine.com

Sizer E et al. Acute kidney injury in patients admitted to

a liver intensive therapy unit with paracetamol-induced hepatotoxicity. Nephrol. Dial. Transplant. 26(11), 3501–3508 (2011).

113

About the Author Andrew K Burroughs Andrew Burroughs is Consultant Physician and Hepatologist at the Royal Free Sheila Sherlock Liver Centre, Royal Free Hospital, London and Professor of Hepatology at the Institute of Liver and Digestive Health at University College London (UK). His interests are complications of cirrhosis, hepatocellular carcinoma,liver transplantation, primary biliary cirrhosis and meta analysis. He has published over 500 peer-reviewed papers.He was elected Fellow of the Academy of Medical Sciences UK in 2010, and is ciurrently Vice President (Hepatology) of the British Society of Gastroenterology.

114

114

© 2013 Future Medicine www.futuremedicine.com

Chapter

9 Acute liver failure coagulation disturbances and need for treatment

Abnormalities of tests of hemostasis in ALF 117 Coagulation factor concentrations

117

Thromboelastography

117

Thrombin generation

118

Microparticles

118

Fibrinogen & clot formation dynamics

118

Endogenous heparinoids 120 Use of prophylactic transfusion of blood products or hemostatic agents

120

Monitoring of coagulation

121

Current needs

121

Andrew K Burroughs Acute liver failure (ALF) is associated with grossly deranged routine tests of coagulation, such as the international normalized ratio. This is perceived to relate to an increased risk of bleeding. However, recent studies suggest that as in cirrhosis, thrombin generation and thus clot formation are preserved so that the risk of bleeding is small. Indeed spontaneous bleeding is rare. Evaluation of the components of the hemostatic system demonstrates mostly preserved coagulation. Current recommendations do not suggest that prophylactic correction of abnormal coagulation tests should be used. These measures are only recommended for the placement of intracranial bolts. More research is needed to inform clinical practice on the use of hemostatic agents in ALF.

doi:10.2217/EBO.12.405

© 2013 Future Medicine

115

Burroughs Coagulopathy is always present in acute liver failure (ALF). Coagulation indices, such as prothrombin time (PT), international normalized ratio (INR), and factor V concentrations are used as prognostic indices and are part of algorithms for listing for liver transplantation [1,2]. The markedly abnormal coagulation profiles have been considered to indicate ‘autoanticoagulation’ and, therefore, a considerable bleeding risk. However, spontaneous bleeding, except from the stomach is rare, suggesting that hemostasis is far less comprised. The need for invasive monitoring has led to using fresh frozen plasma or cryoprecipitate or plasma protein concentrates, platelet transfusions, and fibrinogen preparations, as prophylactic measures to prevent bleeding. However, there is little evidence for this practice. Recently the finding that thrombin generation is normal in cirrhosis, because of a similar and ‘balanced’ deficiency in pro- and anti-coagulation factors, has led to a re-examination of the coagulopathy of ALF.

The abnormal coagulation tests in acute liver failure (ALF), particularly INR, are not associated with an increased bleeding risk.

The coagulation abnormalities in ALF are more severe than in cirrhosis, so traditionally the bleeding risk has been considered to be worse [1]. Indeed more prophylactic hemostatic drugs are used in managing ALF compared with cirrhosis and its complications. However, spontaneous bleeding is currently very rare, the most common source being from the stomach (but prevented by H2 receptor antagonists) and probably related to sepsis (again prophylactic antibacterials are frequently used and this risk has become less). Indeed in the 1970s, a third of patients died from bleeding [1] and today bleeding is a very rare cause of death. Despite recent recommendations to avoid prophylactic correction of coagulation in ALF [2], this is often still carried out outside of specialized liver centers. However, PT or INR, or indeed Factor V concentrations are used as prognostic markers in ALF, and form part of the decision making to list for liver transplantation, so unnecessary correction of coagulation can hinder this process [1,2]. In addition, due to the rarity of spontaneous bleeding, it has been thought that hemostasis may not be so disturbed, similar to the new paradigm of hemostasis in cirrhosis in which there is conserved thrombin generation and, therefore, clot formation, because pro- and anticoagulant factors are similarly ‘reduced’ [3] , or there can be even a prothrombotic state [4] . Thus, there has been a recent interest in evaluating the coagulation abnormalities in ALF, demonstrating that not all patients Prophylactic correction of abnormal are hypocoagulable [5]. coagulation tests in ALF is not indicated except

prior to insertion of intracranial bolts.

116

www.futuremedicine.com

Acute liver failure coagulation disturbances & need for treatment Abnormalities of tests of hemostasis in ALF The PT and INR are greatly prolonged particularly as factors V and VII have the shortest half-lifes of factors synthesized by the liver. Activated partial thromboplastin time is also raised and platelets reduced; fibrinogen concentrations remain normal in most cases [1]. Coagulation factor concentrations As in cirrhosis both procoagulants (Factor II, V, VII, IX, X, XI and XII) and fibrinogen and anticoagulants (protein C, S and antithrombin III) are reduced [5], but factor VIII and Von Willebrand factor (VWF) are greatly elevated, which are procoagulant as they enhance platelet function (Figure 9.1). Thromboelastography There has been little evidence to demonstrate an association of prolongation of PT with bleeding [1]. Thus, it is interesting that in two Figure 9.1. The ratio of procoagulant and anticoagulant factors in 20 patients with acute liver failure at admission to intensive therapy unit.

ATIII

PC

PS

FVIII

Fibrinogen

FXII

FXI

FX

FVII

FIX

FV

0

VWF

1

FII

Fold change in anti- and procoagulants in ALF within first 48 h (multiples of normal values)

2

-1 Procoagulants

Anticoagulants

AT: Antithrombin; F: Factor; PC: Protein C; PS: Protein S; VWF: Von Willebrand factor. Reprinted with permission from Elsevier [5].

www.futuremedicine.com

117

Burroughs recent studies [5,6] thromboelastographic (TEG) changes are minor and often show preserved and normal hemostasis. In the study from my center [5] , the TEG traces were normal in 45%, hypercoagulable in 35% and only hypocoagulable in 20%.

Platelet function, despite reduced platelet counts, is well preserved in ALF.

Thrombin generation This was studied recently [5]. Thrombin generation was evaluated by using the Calibrated Automated Thrombography method [7]. A modified thrombin generation assay in the presence of Protac® (Pentapharm, Switzerland) was used to assess the impact of ALF on the protein C pathway [4]. There was no correlation between PT or INR and the thrombin generation parameters. The thrombin generation was within the normal range, but reduced in terms of capacity compared with normal individuals. In the presence of Protac, thrombin generation was increased compared with controls, reflecting relative protein C deficiency in ALF. Thus, there is a reduced thrombin generation potential overall, but there is increased speed in thrombin generation (once the initial amount of thrombin to activate factors VIII, IX and XI is formed). Thrombin is generated rapidly, enhanced by the increased circulating factor VIII seen in ALF (similar to cirrhosis), and there is reduced thrombin inactivation, as there is resistance to activated protein C (the addition of Protac to ALF plasma did not change the time to first generation of thrombin). Microparticles ALF is an acute inflammatory condition, so it is not surprising that increased concentrations of microparticles are seen, as they are increased in other acute conditions, or where inflammation or tissue injury is present, such as acute coronary syndromes and cancer. Microparticles are small cell fragments in blood derived mostly from monocytes and platelets as well as erythrocytes [8]. They have procoagulant activity as the membrane components contain phospholipids, and they contain tissue factor if derived from monocytes [9]. This may contribute to the preserved coagulation seen in ALF. Indeed the TEG traces [5] show preserved platelet parameters (Figure 9.2), despite the lower platelet counts, suggesting that platelet aggregability and procoagulant function is increased relative to platelet number, compared with normal in ALF, as it is in cirrhosis [10]. Fibrinogen & clot formation dynamics Platelet function in ALF is also preserved due to fibrinogen concentrations [5], which are relatively well maintained in ALF. Fibrinogen binds to the platelet

118

www.futuremedicine.com

Healthy controls

Healthy controls

ALF patients

ALF patients F

***

0

100

200

300

400

500

0

50

100

150

Healthy controls

Healthy controls

ALF patients

*

ALF patients

***

C

G

0

2

4

6

8

0

20

40

60

80

100

Healthy controls

Healthy controls

ALF patients

*

ALF patients

**

D

H

Peak (nM)

B

*p ≤ 0.05; ***p ≤ 0.001. ALF: Acute liver failure; ETP: Endogenous thrombin potential; PT: Prothrombin time. Reprinted with permission from Elsevier [5].

0

2

4

6

8

10

0

500

1000

1500

2000

E

ETP (nM/min)

PT (min)

ETP% (inhibition with Protax®)

Slope (velocity index; nM/min)

Lag-time (min) Microparticles (PPL assay) Clotting time (s)

www.futuremedicine.com 0

10

20

30

40

50

0

100

200

300

400

500

Microparticles (activity assay; nM)

A 2500

Healthy controls

Healthy controls

ALF patients

***

ALF patients

***

Figure 9.2. Thrombin generation using thromboelastographic parameters and microparticle concentrations in patients with acute liver failure compared with healthy controls.

Acute liver failure coagulation disturbances & need for treatment

119

Burroughs via GPIIb/IIIa receptors, and induces aggregation of platelets. Fibrinogen is also a fibrin precursor, which increases the tensile strength of the clot. However, in our study in ALF [5], platelets contributed 75% of the clot strength, more so than in healthy controls where it was 55%, and fibrinogen 45% [11]. Thus, there is relative ‘functional’ hypofibrinogenemia in ALF. This is compensated for by a reduced fibrinolytic potential in ALF due to elevated concentrations of plasminogen activator inhibitor 1 [3,5], so that the fibrinolytic system is in balance providing there is no severe sepsis. This is a significant difference from cirrhosis, indicating that each disturbance of coagulation needs to be evaluated separately according to etiology of liver injury. This reduced fibrinolysis and absence of hyperfibrinogenemia may also explain the relatively preserved coagulation in ALF.

Endogenous heparinoids A further abnormality of hemostasis is the finding of increased concentrations of endogenous heparinoids [5], which probably represent the release of heparin sulfate, normally bound to the surface of the endothelium and which is released when the endothelium is damaged. This has been shown in another cohort of patients [12], and may represent a hemostatic mechanism to prevent microthrombosis; theoretically via the binding of antithrombin, there is inhibition of thrombin (and factor X activity) limiting the conversion of fibrinogen to fibrin, and thus delaying clot formation. Use of prophylactic transfusion of blood products or hemostatic agents There is no evidence for the routine use of blood products or other hemostatic agents either to correct an abnormal INR, which is not recommended [2], nor before an invasive procedure, with the exception of intracranial bolts. Plasma protein concentrates are now available and safe, and have far less volume and greater concentration of coagulation factors than fresh frozen plasma, but there are no specific data in ALF. Cryoprecipitate contains fibrinogen, but fibrinogen can also be given on its own. In the absence of sepsis, central internal jugular lines, arterial lines and access lines for hemofiltration can all be placed without need of prophylactic cover. Platelet transfusions are commonly given if the platelet count is 20,000/ml or less, but again the evidence base for this does not exist in ALF. However, intracranial bolt insertion does have a reported fatal bleeding rate, which varies depending on whether there is extradural or intradural placement. Currently, most centers only place bolts extradurally. The reported fatal bleeding rate in the USA ranges from 3 to 10% [13]. In the USA, recombinant factor VII (rFVIIa) is given to correct INR prior to placing an intracranial

120

www.futuremedicine.com

Acute liver failure coagulation disturbances & need for treatment bolt [2], but there is no evidence that it prevents bleeding, and in other clinical settings it may be associated with thrombosis. The use of FFP is hampered by the large volumes needed to correct the INR in this clinical setting, to the extent that volume overload would be frequent. Plasma protein concentrates have been used. Spontaneous bleeding from the stomach was more frequent in the past, but the routine use of H2 antagonists has greatly reduced this complication, so that gastrointestinal bleeding is now rare [2]. Proton pump inhibitors are also effective, but are associated with more Clostridium difficile infections. Capillary and mucosal bleeding does occur, but is rare [4]. This is treated with fresh frozen plasma and platelets in the first instance, and then correction of coagulation is guided by monitoring of coagulation components in plasma.

Monitoring of coagulation Routine tests are not satisfactory as they do reflect the risk of bleeding. Global tests of coagulation, such as thromboelastography, appear to reflect risk of bleeding and thrombosis better than routine tests, but there is little data to confirm this belief, and the global tests are not universally available. Separate but rapid measurement of the components of clot formation and dissolution would be ideal. Current needs More research on coagulation, and prospective registries in ALF of procedures, complications and administration of hemostatic agents [5,13,14], is needed to answer the questions regarding if and when prophylactic measures are needed to prevent bleeding, and the best regimen to use. Initial studies suggest hemostasis is far better preserved than the clinical interpretation of the prolonged INR suggests [5]. A first step is to limit prophylactic measures to H2 antagonists to prevent acid induced stress related gastric bleeding, and to prophylactic transfusion of clotting factors (such as fresh-frozen plasma, plasma protein concentrates) or recombinant factor VII, before insertion of intracranial bolts. Other administration should be guided by global tests of hemostasis. The effects of superadded sepsis need to be studied. Financial & competing interests disclosure The author has no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties. No writing assistance was utilized in the production of this manuscript.

www.futuremedicine.com

121

Burroughs Summary. ƒƒ In acute liver failure, recent evaluation of the components of the coagulation system using thrombin generation assays and thromboelastography suggests coagulation is preserved in most patients. Spontaneous bleeding is rare. ƒƒ Prophylactic correction of an abnormal INR, whether associated with or without invasive procedures, is not necessary, nor recommended, except before insertion of intracranial bolts.

References 1

Munoz SJ, Stravitz RT, Gabriel DA. Coagulopathy of acute liver failure. Clin. Liver Dis. 13, 95–107 (2009).

2

Stravitz RT, Kramer AH, Davern T et al. Intensive care of patients with acute liver failure: recommendations of the US Acute Liver Failure Study Group. Crit. Care Med. 35, 2498–2508 (2007).

7

Hemker HC, Giesen P, Al Dieri R et al. Calibrated automated thrombin generation measurement in clotting plasma. Pathophysiol. Haemost. Thromb. 33, 4–15 (2003).

11 Gottumukkala VN, Sharma SK,

3

Lisman T, Cladwell SH, Burroughs AK et al. Coagulation in liver disease study group. J. Hepatol. 53, 362–371 (2010).

8

12 Senzolo M, Agarwal S,

4

Gatt A, Riddell A, Calvaruso V, Tuddenham EG, Makris M, Burroughs AK. Enhanced thrombin generation in patients with cirrhosis: induced coagulopathy. J. Thromb. Haemost. 8, 1994–2000 (2010).

Montoro-Garcia S, Shantsila E, Marin F, Blann A, Lip GY. Circulating microparticles: new insights into the biochemical basis of microparticle release and activity. Basic Res. Cardiol. 106, 911–923 (2011).

9

Aleman MM, Gardiner C, Harrison P, Wolberg AS. Differential contributions of monocyte and platelet derived microparticles towards thrombin generation and fibrin formation and stability. J. Thromb Haemost. 9, 2251–2261 (2011).

13 Munoz SJ, Rajender KV, Lee

5

6

Agarwal B, Wright G, Gatt A et al. Evaluation of coagulation abnormalities in acute liver failure. J. Hepatol. 57(4), 780–786 (2012). Stravitz RT, Lisman T, Luketic VA et al. Minimal effects of acute liver injury/acute liver

122

failure on haemostasis as assessed by thromboelastrography. J. Hepatol. 56, 129 (2012).

10 Lisman T, Bongers TN,

Adelmeijer J et al. Elevated levels of von Willebrand

factor in cirrhosis support platelet adhesion despite reduced functional capacity. Hepatology 44, 53–61 (2006). Philip J. Assessing platelet and fibrinogen contribution to clot strength using modified thromboelastography in pregnant women. Anesth. Analg. 89, 1453–1455 (1999). Zappoli P, Vibhakorn S, Mallett S, Burroughs AK. Heparin-like effect contributes to the coagulopathy in patients with acute liver failure. Liv. Int. 29, 54–59 (2009). W. The coagulopathy of acute liver failure and implications for intracranial pressure monitoring. Neurocrit. Care 9, 103–107 (2008).

14 Lisman T, Leebeek FW.

Hemostatic alterations in liver disease: a review on pathophysiology, clinical consequences and treatment. Dig. Surg. 24, 250–255 (2007).

www.futuremedicine.com

123

About the Authors Sarah A Hughes Sarah A Hughes graduated from the University of Cambridge (UK) in 1999. She is currently a Clinical Research Fellow in the Institute of Liver Studies, King’s College Hospital (London, UK). She has clinical interests in both acute and chronic liver diseases, and liver transplantation. Her research is focused on the subject of viral hepatitis, in particular hepatitis D virus.

John O’Grady Professor John O’Grady graduated from the National University of Ireland (Galway) in 1978. He joined the Liver Unit at King’s College Hospital (London, UK), in 1984 where he currently works as a Consultant Hepatologist with an interest in liver transplantation and hepatology. He was President of the British Association for the Study of the Liver (BASL) from 2007 to 2009. He is also Deputy Editor of the American Journal of Transplantation. He has published over 270 papers in his areas of interest.

124

124

© 2013 Future Medicine www.futuremedicine.com

Chapter

10 Selection and results of liver transplantation

Factors influencing prognosis

126

Prognostic models

127

Graft allocation

135

Results of liver transplantation for ALF 136 Auxiliary liver transplantation

137

Sarah A Hughes & John O’Grady Improvements in medical management and intensive therapy of patients with acute liver failure (ALF) have led to an improved overall prognosis. Mortality in those with severe ALF, however, remains high. Liver transplantation may be life-saving in this group, but is not suitable in all cases. The procedure carries with it its own mortality and morbidity, including the sequelae of long-term immunosuppression. Moreover, in the current era of global donor organ shortage, unnecessary liver transplantation in a patient with ALF who would otherwise have survived represents a missed opportunity for a patient awaiting liver transplantation for the complications of chronic liver disease. The challenge for the clinician, therefore, is to determine which of those patients with ALF are in need of, and, at the same time, will derive benefit from, liver transplantation, and who will survive with medical therapy alone.

doi:10.2217/EBO.12.381

© 2013 Future Medicine

125

Hughes & O’Grady Factors influencing prognosis Making a clinical judgment on the need for emergency liver transplantation in acute liver failure (ALF) requires an understanding of the factors that influence prognosis, and to what extent these predict death or survival. Many such factors have been identified, but in reality relatively few are routinely utilized for this purpose. Given the rapidity with which the ALF syndrome can evolve, and the need to maximize the time for a graft to become available in those who are listed for transplantation, the ideal prognostic marker would be objective, simple to measure, widely applicable and would have prognostic significance early in the course of the illness. Laboratory parameters Laboratory variables shown to correlate with prognosis in ALF include coagulation factors, in particular, prothrombin time (PT) or international normalized ratio (INR), serum bilirubin, serum creatinine, arterial pH and serum lactate [1–3]. Also of predictive value, but less widely used, are alternative coagulation factors, in particular, nonvitamin K-dependent factors V and VII [4–6], a-fetoprotein [4,7], arterial ammonia [8], serum phosphate [9,10] , ketone body ratio [11] and levels of Gc globulin, a liver‑derived component of the actin-scavenging system [12]. Clinical factors The etiology of the disease is an important determinant of outcome. Prognosis is much better in ALF that develops in relation to acetaminophen toxicity, pregnancy, hepatic ischemia or hepatitis A, where transplant-free survival is in excess of 50%, compared with idiosyncratic drug reactions, seronegative hepatitis, autoimmune hepatitis or Wilson’s disease, where spontaneous survival is generally less than 30% [13,14]. Encephalopathy grade on admission is also clearly linked to outcome, with those in grade 3–4 coma having poorer survival than those with grade 1–2 coma, regardless of the etiology of the liver failure. Prognosis is worse when grade 4 encephalopathy is complicated by cerebral edema, and worse still when there is coexistent renal failure [15]. Cerebral edema occurs less frequently in the current era, owing to improvements in the medical management of encephalopathy and measures to prevent and control intracranial hypertension, but still accounts for up to 25% of all deaths in ALF [16]. The third key clinical determinant of outcome is age, perhaps owing to the decline in ability to achieve hepatic regeneration The three most important clinical factors with increasing age. Age greater than predicting poor prognosis in acute liver failure 40 years is linked to a poorer prognosis [17]. are: etiology of liver failure, encephalopathy grade and increasing age.

126

www.futuremedicine.com

Selection & results of liver transplantation Radiological & histological indices It is recognized that serial collapse of the liver is a harbinger of poor prognosis in ALF. The ratio of computed tomography-derived liver volume to standardized liver volume at the time of diagnosis has been shown to predict the outcome of ALF with 76.5% sensitivity and 92.3% specificity [18]. Assessment of the role of transjugular liver biopsy has shown that the percentage of hepatic necrosis in the specimen is of discriminatory prognostic value [19]. The practicalities of implementing these methods as routine markers to evaluate prognosis in a critically ill population of patients render them of doubtful utility.

Prognostic models The feasibility of liver transplantation as a treatment for ALF was established during the 1980s. Although it is now widely accepted practice to undertake transplantation in this condition, no controlled trials have ever been performed to evaluate its efficacy or appropriate indications. A number of prognostic models have therefore been proposed, derived from the analysis of data from retrospective cohorts. The aspiration of these models is to establish simply applied criteria, which enable the clinician to list for emergency transplant all those who will die without this intervention, while sparing those who will survive spontaneously from unnecessary transplantation. The accuracy of such criteria to predict poor prognosis can be validated using various indices, in particular positive predictive value (PPV), negative predictive value (NPV), predictive accuracy, sensitivity and specificity. The limitations of existing prognostic models have been that although fulfillment of criteria is highly predictive of death, lack of fulfillment does not guarantee survival. Setting the criteria too high will increase the PPV of the model, Prognostic model: a group of conditions that, but with an inevitable reduction in NPV, when satisfied, predict outcome. equating to potentially avoidable deaths in Positive predictive value (PPV): the proportion of patients who may have benefited from patients fulfilling poor prognosis criteria that die. transplantation. The corollary, if criteria Negative predictive value (NPV): the proportion of are set too low, is the burden of unnecessary patients not fulfilling poor prognosis criteria that transplantation for an individual who may survive. have survived spontaneously, and a wasted Predictive accuracy: the overall proportion of cases organ from the limited donor pool in which the outcome is correctly predicted. (Figure  10.1). There are a number of Sensitivity: the proportion of fatal cases fulfilling reasons for the imperfections in these criteria. models, not least of which is the diversity Specificity: the proportion of survivors not fulfilling of the ALF syndrome, particularly in relation criteria.

www.futuremedicine.com

127

Hughes & O’Grady Figure 10.1. Schematic representation of the effect of setting a cutoff value for selection criteria. All those in need transplanted More unnecessary transplants

Missed deaths Fewer avoidable transplants

Criteria cut-off NPV PPV Sensitivity Specificity It should be noted that the relationship between the variable and performance indicators is not necessarily linear. PPV: Positive predictive value; NPV: Negative predictive value.

to disease etiology. The apparent paradox that those with hyper-ALF due to acetaminophen toxicity, who develop a worse coagulopathy and more cerebral edema, have a better outcome than those with seronegative sub-ALF, who are less coagulopathic, and rarely develop cerebral edema but are more jaundiced, is not something that is easily overcome in defining a one-size-fits-all set of selection criteria to enable the crucial decision as to whether to list a patient for transplantation. King’s College Hospital criteria The King’s College Hospital (KCH) criteria were proposed in 1989, and were the first to separate prognostic criteria by etiology of ALF, defining distinct criteria for those with acetaminophen-induced ALF from those with nonacetaminophen-induced causes [1]. The criteria, which include for acetaminophen-related ALF, arterial pH, PT or INR, creatinine and encephalopathy grade, and for non-acetaminophen etiologies, PT or INR, bilirubin, jaundice to encephalopathy time, age and etiology, are shown in Table  10.1. They form the basis of registration for emergency liver transplantation in the UK, and are in widespread use worldwide. The criteria were originally derived by multivariate analysis of prognostic factors Selection criteria: factors shown to predict in a retrospective cohort of 588 patients, outcome in acute liver failure, which can be used as a basis for selecting patients who need liver treated with medical management between transplantation.

128

www.futuremedicine.com

Acetaminophen induced

King’s College Hospital criteria (1989)

www.futuremedicine.com

Acetaminophen

Mixed

Acetaminophen

Arterial lactate (2002)

BiLE score (2008)

MELD (2007)

MELD >33 after onset of encephalopathy

Bilirubin level (mmol/l)/100 plus lactate (mmol/l) plus etiology (indeterminate; Budd–Chiari = 4; acetaminophen = 2; other = 0); score >6.9 poor prognosis

>3.0 mmol/l postresuscitation

>3.5 mmol/l on admission

Confusion, plus ƒƒ Factor V 300 µM/l (3.4 mg/dl) ƒƒ Grade 3/4 encephalopathy PT >100/INR >6.5, or any 3 of: ƒƒ NANB/drug/halothane etiology ƒƒ Jaundice to encephalopathy >7 days ƒƒ Age 40 years ƒƒ PT >50/INR >3.5 ƒƒ Bilirubin >300 µM/l (17.4 mg/dl)

Criteria for poor prognosis

Sensitivity 60%; PPV 65%; Specificity 69%; NPV 63%

Sensitivity 79% Specificity 84%

Sensitivity 67% Specificity 95% Sensitivity 76% Specificity 97%

PPV 90% NPV 95%

PPV 98% NPV 82%

PPV 84% NPV 86%

Performance

[32]

[35]

[3]

[3]

[28]

[1]

Ref.

ALFSG: Acute Liver Failure Study Group; ALT: Alanine aminotransferase; AST: Admission aspartate aminotransferase; AUROC: Area under receiving operator curve ; BiLE: Bilirubin, lactate and etiology; INR: International normalised ratio; MELD: Model for end stage liver disease; NANB: Non-A, non-B hepatitis; NPV: Negative predictive value; PPV: Positive predictive value; PT: Prothrombin time.

Viral hepatitis

Clichy (1991)

Nonacetaminophen induced

Etiology

Model (year)

Table 10.1. Proposed prognostic models for the prediction of poor outcome in acute liver failure.

Selection & results of liver transplantation

129

130

Mixed

Mixed

Viral hepatitis

Acetaminophen

CK-18/M65-MELD (2010)

ALFSG-index (2012)

Dhiman (2007)

MALD (2012)

Amatoxin

Amatoxin

Ganzert (2005)

Escudie (2007)

Predictive accuracy 78%

[42]

[41]

[40]

[39]

[19]

[37]

ALFSG: Acute Liver Failure Study Group; ALT: Alanine aminotransferase; AST: Admission aspartate aminotransferase; AUROC: Area under receiving operator curve ; BiLE: Bilirubin, lactate and etiology; INR: International normalised ratio; MELD: Model for end stage liver disease; NANB: Non-A, non-B hepatitis; NPV: Negative predictive value; PPV: Positive predictive value; PT: Prothrombin time.

Time from ingestion to diarrhea 70%

Admission: AST, ALT, INR, creatinine (see reference for details)

Any 3 of: ƒƒ Age >50 years ƒƒ Jaundice to encephalopathy >7 days ƒƒ Grade 3/4 encephalopathy ƒƒ Cerebral edema ƒƒ PT >35 s ƒƒ Creatinine >1.5 mg/dl (133 µM/l)

Coma grade ≥2, bilirubin, INR, phosphate (≥3.7 mg/dl), log10M-30

M-MELD >53.5 (see reference for formula)

Criteria for poor prognosis

Pregnancy-related Pregnancy related Lactate >2.8 mg/dl plus encephalopathy liver failure (2010)

Liver biopsy (1993) Mixed

Etiology

Model (year)

Table 10.1. Proposed prognostic models for the prediction of poor outcome in acute liver failure.

Hughes & O’Grady

www.futuremedicine.com

Selection & results of liver transplantation 1973 and 1985. The criteria were validated in a further retrospective cohort of patients from 1986 to 1987, comprising 121 patients with ALF due to acetaminophen and 54 with ALF due to other etiologies. Many groups have assessed the performance of the KCH criteria in their own ALF cohorts (Table  10.2) [6,19–25]. Interpretation of these, mostly retrospective, reports is made difficult by the small numbers incorporated, the effect of era on the outcome of critically ill patients, and the heterogeneity of the cohorts. For example, reports vary as to whether those who underwent transplant are included in the analysis; the natural history of the illness is censored by intervening with liver transplantation, rendering it difficult to accurately evaluate transplant-free survival. Two recent meta-analyses of the performance of the KCH criteria have been published. The first, addressing non-acetaminophen-induced ALF, included 1105 patients in 18 studies and found an overall sensitivity of 68% and specificity of 82% [26]. Specificity was highest in those with higher-grade encephalopathy and where the criteria were applied dynamically. Sensitivity was lower in studies published after 2005 than before 1995. The second systematic review, addressing acetaminophen-induced ALF, included 1960 patients in 14 studies and found a sensitivity of 58% and specificity of 95% [27]. Clichy criteria The Clichy criteria were defined in 1986 and were modeled from a cohort of 115 patients with fulminant hepatitis B. Prognostic indicators were Factor V levels, age, a-fetoprotein concentration and absence of hepatitis B virus surface antigen [4]. The model incorporated a function of age, Factor V levels and encephalopathy grade (Table  10.1) [28]. The criteria were extended for application to other etiologies of liver failure and are in use across northern Europe. The utility of these criteria are limited in some areas by restricted access to a reliable factor V assay. The assessment of the performance of these criteria in one series of 120 consecutive patients showed a PPV of 75% and NPV of 58%, increasing to 87% and 67%, respectively, when children were included in the analysis [24]. Alternatives & improvements to current prognostic models In the process of assessing the performance of existing prognostic models, and seeking to improve their sensitivity and NPV in particular, many other refinements have been suggested. Bernal and colleagues identified arterial lactate as having prognostic value in identifying patients likely to die without transplantation [3]. A lactate threshold of 3.5 mmol/l early after admission, and 3.0 mmol/l post-resuscitation had a sensitivity, specificity,

www.futuremedicine.com

131

Hughes & O’Grady positive likelihood ratio and negative likelihood ratio of 67%, 95%, 13 and 0.35, and 76%, 97%, 30 and 0.24, respectively. Combining lactate values at the two time points had similar predictive ability to the KCH criteria, but allowed earlier identification, and therefore listing, of nonsurvivors. Adding post-resuscitation lactate to the KCH criteria increased its sensitivity to 91%. Other groups have examined the performance of lactate in alternative Table 10.2. Assessment of the performance of the King’s College criteria in other cohorts of patients with acute liver failure. Study (year)

Etiology

Donaldson et al. (1993)

Non-acetaminophen 46

91

82

Acetaminophen

100

100

Pauwels et al. (1993)

Non-acetaminophen 81

79

86

[20]

Izumi et al. (1996)

Non-acetaminophen 17

93

67

[6]

Acetaminophen

69

Shakil et al. (2000)

Non-acetaminophen 144

34 /93

Acetaminophen

33

49 /45

99 /94

Schmidt and Dalhoff (2002)

Acetaminophen

106

67

97

[9]

Bernal et al. (2002)

Acetaminophen

99

76

95

[3]

Zaman et al. (2006)

Acetaminophen

72

73

100

[31]

Dhiman et al. (2007)

Non-acetaminophen 144

47

89

[23]

Yantorno et al. (2007)

Non-acetaminophen 120

73

79

[24]

Schmidt and Larsen (2007)

Acetaminophen

67

79

[32]

Bechmann et al. (2010)

Non-acetaminophen 59

60

75

[34]

47

83

[25]

Cholongitas et al. Acetaminophen (2012)

Patients Sensitivity Specificity Ref. (n) (%) (%) 15

81

124

125

[19]

96

†

§

‡

¶

100†/90‡ §

[22]

¶

Criteria are as defined in Table 1. † Based on INR criteria. ‡ Based on “3 of 5” criteria. § Based on pH criteria. ¶ Based on non-pH criteria.

132

www.futuremedicine.com

Selection & results of liver transplantation prognostic models, and have not found it to be superior to the existing criteria [29]. The UK Transplant Agency incorporated lactate criteria into minor modifications of its categories for ‘super-urgent’ listing for acetaminophen-induced liver failure in 2008 [30]. Criteria for listing for nonacetaminophen-induced liver failure, based on those published by O’Grady in 1989 [1], have remained largely unchanged for over 20 years.

Key determinants of a poor outcome following liver transplantation for acute liver failure include: ƒƒ Recipient age >50 years ƒƒ Donor age >60 years ƒƒ ABO incompatibility ƒƒ Reduced-size grafts ƒƒ BMI ≥30 kg/m2 ƒƒ Serum creatinine >2.0 mg/dl ƒƒ History of organ support

The Model for End-Stage Liver Disease (MELD) score has been adapted for assessment of survival in those with chronic liver disease, and forms the basis of registration for elective liver transplantation in the USA. It incorporates the variables serum bilirubin, serum creatinine and INR, all of which feature in the KCH criteria. It seemed natural, then, to assess its performance in predicting outcome in those with ALF. Several groups have found MELD to perform as well as or better than the KCH or Clichy criteria [24,31], but others have not found it to be superior to the KCH criteria for acetaminophen-induced liver failure [32]. It has been acknowledged that MELD alone is insufficient in providing accurate survival probabilities [33], which is perhaps not surprising, since its component variables do not take into account some of the most important clinical factors influencing survival in ALF, for example, age and etiology, as discussed. Combining MELD with other factors therefore seems logical. An example is the substitution of a cell-death marker, cytokeratin-18, for bilirubin within the MELD, which was shown in a prospective cohort of 68 patients to outperform MELD and the KCH criteria in predicting outcome [34]. Other proposed models include the Bilirubin, Lactate and Etiology (BiLE) score, found by retrospective analysis of 102  patients to predict poor outcome better than bilirubin, lactate, MELD or the Simplified Acute Physiology Score III (SAPS-III), but did not conclusively outperform the KCH criteria [35]. A group from India identified six key clinical prognostic indicators: age ≥50  years, jaundice to encephalopathy interval >7  days, grade  3/4 encephalopathy, presence of cerebral edema, PT ≥35  s and creatinine ≥1.5 mg/dl. In their cohort of 144 patients, all of whom had acute viral hepatitis, fulfilling any three of these six factors predicted adverse outcome better than MELD or the KCH criteria [23]. The inclusion of a late complication of ALF, such as cerebral edema, in this study will naturally have increased PPV, but falls short of the aspiration to identify factors that discriminate nonsurvivors at an earlier stage in the course of the illness. Future models

www.futuremedicine.com

133

Hughes & O’Grady are likely to take into account the evolution of the illness, and to use criteria that are applied dynamically. Most recently, a model termed the ALF study group (ALFSG) index has been proposed, which incorporates coma grade, bilirubin, INR, phosphate and log10 -M30, a cleavage product of cytokeratin-18 caspase, and a marker of cell death [36]. The model uses values collected from day 1 of admission to a specialist unit and, based on the area under the receiver operating curve, predicted adverse outcome better than the KCH criteria or MELD. Although the model satisfies the need to improve on the sensitivity of the KCH criteria (sensitivity 85.6%), it is at the expense of specificity, particularly in the non-acetaminophen group where specificity was 59.2%. A recent study by Remien and colleagues has elegantly applied mathematical modeling to evaluate outcome in acetaminophen-induced liver failure, terming it the model for acetaminophen-induced liver disease (MALD) [37]. They base their model on the fact that size of overdose and time to administration of N-acetylcysteine are important determinants of prognosis, and use mathematical modeling of admission aspartate aminotransferase, alanine aminotransferase and INR, in conjunction with knowledge of the pattern of acetaminophen-induced liver injury to estimate overdose amount and timing. From that they derive an estimate of outcome, which, with the inclusion of creatinine, predicts survival with 91% specificity, 100% sensitivity, 67% PPV and 100% NPV. This approach is clearly attractive in its use of the very earliest available laboratory values, and warrants further analysis in a prospective multicenter study. Well-validated scores of organ dysfunction in critically ill patients have been applied to cohorts of patients with ALF to judge their prognostic performance. Examples include the sequential organ failure assessment (SOFA) score [25], systemic inflammatory response syndrome (SIRS) score [29] and the acute physiology and chronic health evaluation II (Apache-II) score [38]. Although all were shown to have prognostic value in the setting of ALF, none have been conclusively proven to perform better than existing criteria and, as with other proposed models, require further validation in large prospective cohorts before being considered as the basis for registration for emergency liver transplantation. Disease-specific prognostic models There are etiologies of ALF for which the KCH criteria were not validated, and separate studies have provided guidance on prognostic criteria in these more rare conditions. A retrospective analysis of 54  patients with pregnancy-related liver disease found a transplant-free survival of 80%. A

134

www.futuremedicine.com

Selection & results of liver transplantation total of 13% died and 7% underwent transplantation [39]. The KCH criteria did not accurately predict survival, instead an admission lactate of ≥2.8 mg/ dl had a sensitivity of 73% and specificity of 75% in predicting death or liver transplantation. These figures increased to 90% and 86%, respectively, with the addition of encephalopathy, concluding that increased lactate in the presence of encephalopathy was the best predictor of poor outcome in this group [39]. A study of 29 patients with hepatitis A-related ALF from the USA found that a prognostic model incorporating alanine aminotransferase 2.0 mg/dl, intubation status and need for vasopressor agents best predicted outcome, with an area under the receiver operating curve of 0.899, which was significantly better than the KCH criteria or MELD score [40]. Regarding patients with ALF caused by ingestion of mushrooms containing amatoxins, a series of 198 patients reported that a prothrombin index 106 µM/l from day 3 postingestion onwards had 100% sensitivity and 98% specificity for the prediction of death [41]. A second series of 27  patients with amanita poisoning found that an interval between ingestion and the onset of diarrhea of less than 8 h was an earlier predictor of fatal outcome, with a predictive accuracy of 78% [42].

Graft allocation Worldwide, approximately 45–50% of those meeting the definition of ALF are transplanted, although when acetaminophen toxicity is the etiology, this figure falls to less than 10%. The remainder survive spontaneously, have contraindications to transplantation, or die before an organ becomes available. Contraindications can include being too sick in relation to the severity of liver failure and associated organ dysfunction, coexisting illnesses conferring a limited life expectancy, for example, malignancy, and psychiatric or substance abuse issues deemed to be complex enough to preclude transplantation. The latter may be contentious, has ethical implications and requires a multidisciplinary team evaluation on an individual case basis [43]. The feasibility of emergency transplantation for ALF is predicated on the timely availability of a suitable graft. In both the USA and Europe, the majority will receive a graft within 48 h of being listed. The need for speed has influenced the type of grafts used, and ABO-incompatible grafts, steatotic livers, non-heart-beating deceased donors and reduced size grafts have all been used in this setting. The ideal scenario would be to individually match recipient and graft, and offer the best quality graft to the sickest patient. In reality, the risk of delaying in the hope that a better graft becomes available results in the acceptance of more marginal grafts. In areas such as Asia, where access to deceased donor transplantation is limited, there is

www.futuremedicine.com

135

Hughes & O’Grady Characteristics of optimal recipients for auxiliary grafts: ƒƒ Young (50 years, ABO incompatibility, donor age >60 years and reduced size graft as being associated with post-transplant death or graft loss [45]. The United Network for Organ Sharing data found BMI ≥30 kg/m2, serum creatinine >2.0 mg/dl, recipient age >50 years and a history of life support to be the factors most strongly associated with poor outcome [47]. From these four factors a scoring system was devised, with low-risk patients having a 5-year survival of 81%, compared with 42% in high-risk

136

www.futuremedicine.com

Selection & results of liver transplantation patients. In one single-center study, multivariate analysis of data from 204 patients transplanted over a 17-year period found that serum creatinine was the single-most important determinant of survival post-liver transplant [48].

Auxiliary liver transplantation In some situations where resolution of the insult causing liver failure would lead to regeneration of the native liver, auxiliary liver transplantation may be used to bridge the gap to recovery. In this procedure, a partial graft is transplanted heterotopically or orthotopically, leaving all or part of the native liver in situ. A right lobe graft is often preferred in an adult recipient, as it offers a larger liver volume. Splitting a deceased donor liver for use in auxiliary transplantation requires a good-quality graft. The operation is technically challenging and requires surgical skill and experience. Auxiliary transplantation is an attractive option, since it potentially avoids the need for longer-term immunosuppression and its various side effects and complications. After an appropriate time period, slow withdrawal of immunosuppression will lead to atrophy of the graft, with the hope that the native liver has regenerated adequately to support normal hepatic function. The success of this will depend on there being a critical mass of surviving native hepatocytes. Histologically, the majority of native livers will regenerate, with particularly high rates in cases of diffuse liver injury, typified by the liver injury caused by acetaminophen toxicity, where, unless there is massive cell loss, hepatocyte proliferation is at its peak in the first week post transplant [49]. In seronegative sub-ALF, the liver injury is more map-like, and histological regeneration is rather more unpredictable [49]. Clinically, outcomes after auxiliary transplantation are best in younger recipients (6.5, creatinine >300 µmol/l and hyperphosphatemia. Over the years, several prognostic markers have been proposed to predict outcomes in nonacetaminophen ALF in children, of which INR and Factor V concentrations remain the best indicators. An INR ≥4, bilirubin ≥235 µmol/l, age 9 × 109/l are associated with a poor outcome without liver transplantation [26]. Bhaduri and Mieli-Vergani have shown that the maximum INR reached during the course of illness was the most sensitive predictor of the outcome, with 73% of children with an INR 4 [27]. In children, a Factor V concentration of less than 25% of normal suggests a poor outcome. In adult ALF, persistently raised serum

152

www.futuremedicine.com

Acute liver failure in children lactate levels, despite adequate fluid resuscitation, have been shown to be a poor prognostic indicator [28]. More recently, the PALF study group validated King’s College Hospital (KCH) criteria in predicting mortality in pediatric ALF [29]. It was found that the sensitivity of applying the KCH criteria to a pediatric cohort was significantly lower (61 vs 91%; p = 0.002), but specificity did not differ significantly. Using appropriate statistical calculation, it was found that grade 2–4 encephalopathy with jaundice; grade 2–4 encephalopathy with INR >4.02; and grade 0–1 encephalopathy with prothrombin time >32.5 s and total bilirubin >2.02 mg/dl were associated with increased mortality.

Management Once ALF is suspected, contact with a specialist center should be made to establish a management plan and an early transfer. General measures All children with ALF should be closely monitored in a quiet setting. Vital parameters such as oxygen saturation, pulse, blood pressure, neurologic observations, should be done on a regular basis. Prophylactic broadspectrum antibiotics and antifungals should be started in all children, and acyclovir should be added in infants and neonates. Children with encephalopathy or an INR >4 (regardless of encephalopathy) should be admitted to an intensive care unit for continuous monitoring. Hypoglycemia should be avoided by use of intravenous glucose infusion or by ensuring adequate enteral intake. The view that protein restriction will limit the possibility of HE has now been disregarded, and adequate calories should be provided. Oral or nasogastric feeding is usually well tolerated. Prophylactic histamine type 2 blockers or proton pump inhibitors should be started in all patients requiring mechanical ventilation [30]. N-acetylcysteine N-acetylcysteine (NAC) is being increasingly used in nonacetaminopheninduced ALF as it enhances circulation and improves oxygen delivery. A prospective, double-blind trial in adults with nonacetaminophen-induced ALF showed NAC usage to be associated with significant improvement in transplant-free survival inpatients with stage I–II coma, indicating the necessity for an early initiation of treatment [31]. A more recent study in nonacetaminophen-induced pediatric ALF by the PALF study group showed that NAC did not improve 1-year survival in these children [32]. However, while awaiting further data, we continue to use this drug in our institution.

www.futuremedicine.com

153

Shanmugam & Dhawan Airway & ventilation Elective intubation and mechanical ventilation should be considered for patients with grade 3 or 4 encephalopathy, or when patients in grade 1 or 2 encephalopathy require sedation due to agitation. Apart from providing a secure airway, mechanical ventilation helps in reducing sudden variations of intracranial pressure (ICP). Sedation should be maintained with a combination of an opiate such as morphine or fentanyl, and a hypnotic such as midazolam. Peak end-expiratory pressure above 8 cm of water should be avoided because it may increase ICP. Fluid management & renal failure Intravenous fluids should be restricted to two-thirds of usual maintenance, with the aim of decreasing the possibility of development of cerebral edema. Due to fluid shift into interstitial space, these patients frequently have varying degrees of intravascular volume depletion and require fluid resuscitation. The Ultrasonic Cardiac Output Monitor (USCOM), which is a noninvasive method to measure cardiac parameters, helps in decision making regarding appropriate fluid regimens/inotrope even in small infants. In the presence of persistent hypotension despite adequate intravascular filling, noradrenaline is the inotropic agent of choice. Hypotension should be treated with adequate filling before administering inotropic agents. In hypotension refractory to inotropes, hydrocortisone should be considered. Renal replacement therapy should be considered when the urine output is less than 1  ml/kg/h to prevent acidosis and volume overload. Continuous filtration or dialysis systems are associated with less hemodynamic instability than intermittent hemodialysis. Neurologic complications Encephalopathy is not always recognisable in children and usually is a late feature. The most serious complications of ALF is cerebral edema with resultant intracranial hypertension along with hepatic encephalopathy. Clinical features of raised ICP include systemic hypertension, bradycardia, hypertonia, hyper-reflexia and, in extreme cases, decerebrate or decorticate posturing. EEG changes occur very early in HE, even before the onset of psychological or biochemical disturbances. Ammonia-lowering measures, such as dietary protein restriction, bowel decontamination or lactulose, are of limited or no value in rapidly advancing encephalopathy. Mannitol is an osmotic diuretic commonly used to treat intracranial hypertension. A rapid bolus of 0.5 g/kg as a 20% solution over a 15-min period is recommended, and the dose can be repeated if the serum osmolarity is less than 320 mOsm/l. In ventilated patients, the PaCO2 should

154

www.futuremedicine.com

Acute liver failure in children be kept between 4 and 4.5 kPa. Cerebral perfusion pressure (mean arterial blood pressure minus ICP) should be maintained at more than 50 mmHg and may require use of inotropic agents to increase mean arterial blood pressure. Studies have shown use of sodium thiopental, mild cerebral hypothermia (32–35°C) and induction of hypernatremia (serum sodium: >145 mmol/l) improves cerebral perfusion. Coagulopathy Coagulopathy is corrected only if the patient is already listed for transplant or prior to an invasive procedure, as it is a dynamic indicator of disease activity. To correct coagulopathy, fresh frozen plasma is given at a dose of 10 ml/kg and cryoprecipitate at 5 ml/kg (if fibrinogen is