Liver Intensive Care Group of Europe www.licage.org

International Course on Anaesthesia and Critical Care for

Liver Transplantation

26 – 27 May 2005, The Sage Gateshead

SESSION SUMMARIES

Supported by an unconditional educational grant from Astellas Ltd.

© LICAGE 2005

LICAGE

i

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 the prior permission of the copyright owner.

ii

Contents (listed by session number) Page No. Session 1 – Chronic Liver Failure

Prof. Andrew Burroughs – Pathogenesis of Chronic Liver Disease 1

Dr Kosh Agarwal – Pathophysiology and management of end stage cirrhosis 2 Dr James Findlay – Cardio-respiratory physiology 6 Prof. James Neuberger – Assessment for liver transplantation 10

Session 2 - Surgical and donor considerations

Dr Jonathan Mackay – Optimising the graft 14

Mr Chris O’Sullivan – Operative techniques and hazards 19

Mr Nigel Heaton – Increasing the organ supply 23

Session 4 – Anaesthetic management

Prof. Mark Bellamy – the “good” patient 24

Prof. Susan Mandell – Crisis management 29

Dr Sue Mallett – Coagulation 34

Session 5 – Acute Liver Failure

Dr Julia Wendon – Overview 38

Dr Alistair Lee – Neurological complications of ALF 41

Prof. Michael Gropper – Supporting other organs 42

Dr Elizabeth Sizer – Liver support 46

Session 6 – Anaesthesia for ALF

Dr Christopher Snowden – Intraoperative care 49

Dr Elizabeth Sizer – Postoperative care 53

Session 7 – Paediatric transplantation

Dr James Bennett – Indications 54

Dr Moira O’Meara – Post operative and paediatric intensive care issues 60

Prof. Alastair Millar – Surgical aspects 64

Session 8 – Post-operative care Dr Kosh Agarwal – Immunosupression 67

Mr Derek Manas – Principles and early complications 71

Dr John O’Grady – Late complications and outcomes 76

iii

Contents (listed alphabetically by speaker surname) Page No. Dr Kosh Agarwal

Chronic Liver Failure – Pathophysiology and management of end stage cirrhosis 2 Post-operative care – Immunosupression 67

Prof. Mark Bellamy Anaesthetic management – The “good” patient 24

Dr James Bennett Paediatric transplantation – Indications 54

Prof. Andrew Burroughs Chronic Liver Failure – Pathogenesis of Chronic Liver Disease 1

Dr James Findlay Chronic Liver Failure – Cardio-respiratory physiology 6

Prof. Michael Gropper Acute liver failure – Supporting other organs 42

Mr Nigel Heaton Surgical and donor considerations – Increasing the organ supply 23

Dr Alistair Lee Acute liver failure – Neurological complications of ALF 41

Dr Jonathan Mackay Surgical and donor considerations – Optimising the graft 14

Dr Sue Mallett Anaesthetic management – Coagulation 34

Mr Derek Manas Post-operative care – Principles and early complications 71

Prof. Susan Mandell Anaesthetic management – Crisis management 29

Prof. Alastair Millar Paediatric transplantation – Surgical aspects 64

Prof. James Neuberger Chronic Liver Failure – Assessment for liver transplantation 10

Dr John O’Grady Post-operative care – Late complications and outcomes 76

Dr Moira O’Meara Paediatric transplantation – Post operative and paediatric intensive care issues 60

Mr Chris O’Sullivan Surgical and donor considerations – Operative techniques and hazards 19

Dr Elizabeth Sizer Acute liver failure – Liver support 46

Anaesthesia for ALF – Postoperative care 53

Dr Christopher Snowden Anaesthesia for ALF – Intraoperative care 49

Dr Julia Wendon Acute liver failure – Overview 38

1

Session 1 - Chronic Liver Failure Pathogenesis of Chronic Liver Disease Prof. Andrew Burroughs. Consultant Physician and Hepatologist, Professor of Hepatology, Royal Free Hospital, London

2

Session 1 - Chronic Liver Failure Pathophysiology and management of end stage cirrhosis Dr Kosh Agarwal. Consultant Physician / Hepatologist, Freeman Hospital, Newcastle The normal liver - The liver weighs 1400-1600 gm (2.5% of body weight). - Function: maintains the body’s metabolic homeostasis, processes dietary amino acids,

carbohydrates, lipids & vitamins, synthesizes serum proteins, and detoxifies & excretes waste products into bile. Increasing evidence for role of liver in immunity.

- Bile promotes dietary fat absorption through the detergent action of bile salts, and eliminates waste products, that are insufficiently water-soluble to be excreted into urine (bilirubin, cholesterol).

- Substances excreted in bile (e.g. bile salts) can be returned to the liver in the hepatic portal blood (entero-hepatic circulation).

- Liver receives blood (25% of CO) from: portal vein (60-70%) & hepatic artery (30-40%), to exit the liver by the hepatic veins.

- Bile exits the liver in bile canaliculi, then via the canals of Hering to bile ducts, and finally to the common hepatic duct. Myosin and actin filaments around canaliculi propel bile.

- The histologic unit is the hepatic ‘lobule’, oriented around the terminal hepatic venules. The functional unit is the ‘acinus’, oriented around the portal tracts (the source of blood supply).

- Hepatocytes are arranged in ‘plates’, radiating from the central vein. The ‘Limiting plates’ surround the portal tracts.

- Sinusoids are lined by discontinuous endothelium & Kupffer cells (attached to luminal face of endos), separated from hepatocytes by the Disse space, which contains Hepatic stellate cells (transform into collagen-producing myofibroblasts when there is inflammation and fibrosis).

Fig 1: Schematic diagram of hepatic lobule

3

Hepatic response to injury Necrosis: Councilman bodies (apoptosis), drop-out necrosis (hepatocytolysis). Distribution may be: focal, zonal (central, mid-, peripheral), confluent (bridging), submassive or massive. Ischemic injury—usually centrilobular. Degeneration: Damage from toxic or immunological insult. hydropic degeneration (= ballooning, feathery); accumulation of lipids (steatosis—foamy degeneration, microvesicular no nuclear displacement, macro- displaces) or pigments (bile, Fe, Cu,) due to failure of excretion. Inflammation (Hepatitis): inflammatory cells limited to portal tracts, or spill over into the parenchyma in areas of hepatocyte necrosis, crossing damaged limiting plates (interface hepatitis). Macros engulf cells, leaving inflammatory cells in normal-appearing liver. Regeneration: Mitosis, thickening of hepatocyte cords & disorganization of the parencyhma. Excess bile duct profiles in portal tract. The condition of reticulin framework determines restitution (normal) or disorganization of structure (cirrhosis). Fibrosis: The ‘healing’ response of the injured liver. Irreversible response to injury (above are reversible), occurs around portal tracts, central veins & within the space of Disse → cirrhosis. Pathogenesis of cirrhosis The central process in cirrhosis is progressive fibrosis coupled with nodule formation. - Normally: types I & III collagen are present in portal tracts & around central veins & occasional

type IV collagen bundles in the space of Disse. - In cirrhosis: types I & III collagen are deposited all through the lobules (in the space of Disse)

& the sinusoidal endothelial cells lose their fenestrations (= capillarization) & become impermeable to proteins (albumin, clotting factors).

- The source of collagen synthesis in the space of Disse is the hepatic stellate cells; stimulated by inflammatory & endogenous cytokines to turn into myofibroblast-like cells. Tonic contraction constricts sinusoids & increases vascular resistance.

- Within the confines of the fibrous septa, hepatocytes are stimulated to regenerate as expanding spherical nodules that have severely compromised blood flow. Biliary channels are also obliterated, obstructing bile flow.

- New vessels in the septa connect portal and terminal hepatic vessels, shunting blood around the parencyhma.

Clinical features of cirrhosis - Vary from clinically silent, to nonspecific symptoms (anorexia, weight loss, weakness)

[compensated], to overt hepatic failure (hepatic encephalopathy, generalized oedema, coagulopathy, multi-organ failure) [decompensated].

- Hepatic failure is usually precipitated by a metabolic load on the liver, such as gastrointestinal haemorrhage or infection.

Causes of death in cirrhotic patients 1. Progressive liver failure, sepsis. 2. Portal hypertension (ruptured oesophageal varices, ascites). 3. Hepatorenal syndrome/renal failure 4. Development of hepatocellular carcinoma (HCC). Physiological consequences of cirrhosis - Increased, cardiac output, arterial hypotension, decreased total systemic vascular resistance. - Increased total plasma cell volume - Increased activity of vasoconstrictor systems - Increased portal pressure - Porto-systemic shunting - Reduced splanchnic vascular resistance, splanchnic bed vasodilatation and pooling - Increased renal vascular resistance, decreased renal perfusion pressure

4

- Sodium retention due to increased activity of the sympathetic nervous system and of the renin-angiotensin-aldosterone system

- Dilutional hyponatraemia secondary to impairment of free water excretion due to hyperaldosteronism.

Portal hypertension Occurs in 90% of those with cirrhosis surviving 10 years, present in 50% of cirrhotics at diagnosis. Defined as hepatic portal venous gradient (HPVG) > 10mmHg. Due to fixed (mechanical) and (dynamic) increases in intrahepatic resistance in combination with splanchnic vasodilatation and increased inflow. Most relevant clinical consequence is oesophageal varices. Varices progress from small to large at a rate of 8%/year. Mortality from bleed approximately 30-40%. Pharmacological and endoscopic therapy to reduce HPVG decreases rebleeding. Ascites - Hepatic sinusoidal hypertension-drives fluid into space of Disse which is removed by hepatic

lymphatics. - Sodium & water retention because of secondary hyperaldosteronism. - Lymph overflow obstruction-thoracic duct normally empties 1L/d, may be 20L/d in ascites.

Leakage of protein-rich fluid into abdomen. - Intestinal fluid leakage-Increased perfusion pressure in intestinal capillaries, osmotic action of

protein-rich ascitic fluid pulls fluid out (oncotic/hydrostatic drive to ascites formation. - Spontaneous bacterial peritonitis (SBP). Diagnosis established by polymorphonuclear cell

count of >250cell/mm. Bacterial gut translocation may be an important precipitant. Hepatic Encephalopathy (HE) - A spectrum of changes ranging from mild confusion, associated with asterixus (liver flap,

tremors), rigidity, hyper-reflexia & rarely seizures, progressing to coma & death. - There are only minor morphologic changes in the brain, such as mild edema and an astrocytic

reaction although cerebral oedema may be more frequent. - Related to shunting of blood around the liver through portosystemic shunts. Key

neuromediators include: ammonia, glutamine, gamma-aminobutyric acid (GABA)-like substances, mercaptans & abnormal neurotransmitters.

Precipitating factors include gastrointestinal haemorrhage, constipation, infection, diuretic therapy. Coagulopathy - Coagulopathic tendency parallels the severity of hepatic failure, due to both coagulation &

platelet defects. - A low platelet count of <80,000/micro-L is common, accompanied by qualitative abnormalities

in platelet function. The cause may be hypersplenism, bone marrow depression or the consumption of platelets by intravascular coagulation (DIC).

- Reduced hepatic synthesis of coagulation factors (I, II, V, VII, IX & X), worsened by malabsorption of vitamin K, as well as inadequate hepatic clearance of the activated anticoagulant factors. Tendency toward GI bleed which will worsen hepatic failure.

Hepatorenal Syndrome (HRS) - Development of renal failure in patients with severe liver disease, in absence of intrinsic renal

causes. - Manifested by a drop in urinary output, rising urea and creatinine. Urine output is low with low

sodium & negative for proteins (contrary to acute tubular necrosis). - Related to intra-renal vasoconstriction of cortical vessels → decreased glomerular perfusion &

↓ GFR, in addition to decreased renal perfusion pressure. - Prognosis is poor with a mortality of 80-95%. Two types of HRS are seen: rapid deterioration

in renal function over short time period, survival usually less than 2 weeks (Type 1) and a more indolent progression over months (Type 2).

5

Principles of therapy - Management is supportive allowing time for liver reserve to (hopefully) recover. - Meticulous fluid management and attention to electrolyte disturbance. - Optimize oxygen delivery and effective organ perfusion. - Low threshold for antibiotic and antifungal therapy- often culture negative. - Synthetic somatostatin analogues, particularly glypressin are effective, coupled with early

therapeutic endoscopy. TIPSS is best utilized for oesophageal/gastric variceal bleeding unresponsive to therapeutic endoscopy.

- Proactive management rather than ‘nihilistic’ approach, with aggressive initial resuscitation, broad-spectrum antibiotic therapy, withdrawal/alleviation of precipitant and invasive monitoring to prevent multi-organ failure.

6

Session 1 - Chronic Liver Failure Cardio-respiratory physiology Dr James Y. Findlay BSc, FRCA. Consultant Anaesthetist, Mayo Clinic College of Medicine, MN, USA Assessment for Transplantation Liver failure results in distinct pathophysiological changes in the cardiovascular and respiratory systems. Some are common to all patients with liver failure; other syndromes only affect a proportion. The physiological changes will be briefly reviewed and the implications for transplant and pretransplant assessment discussed. One of the challenges in cardiopulmonary assessment of the liver transplantation candidate is that the symptoms often associated with cardiac and respiratory disease (dyspnea, fatigue, reduced exercise tolerance) are non-specific and often present in such patients as a result of their liver disease. Cardiovascular The typical cardiovascular changes of end stage liver disease (ESLD) are characterized by tachycardia, high cardiac output and low peripheral vascular resistance. Blood pressure is typically reduced. Within the systemic circulation there is vascular hyporeactivity. The underlying cause is still debated. An increase in circulating vasodilators has been shown in cirrhosis, as has a decreased responsiveness to vasoconstrictors. Autonomic reflexes are impaired, this may be related to the decreased post-junctional responses seen. The vasodilation leads to a decrease in effective blood volume, this in turn may drive activation of sympathetic nervous system, renin-angiotensin-aldosterone-system and endothelin system leading to the sodium and water retention typical of ESLD. This may also be a cause of the decreased renal perfusion and sometimes failure (hepatorenal syndrome) associated with ESLD. Thus both vasodilatory and vasoconstrictor mechanisms appear to be simultaneously at work(1). The vascular changes of ESLD are reversible with liver transplantation. In addition to the vascular changes in ESLD cardiac function is also affected. Despite the elevated cardiac output associated with ESLD there is accumulating evidence of impaired contractility, termed cirrhotic cardiomyopathy (CCM), an entity distinct from alcoholic cardiomyopathy. This may not be obvious secondary to the “autotreatment” of decreased vascular resistance, but is demonstrated by impaired response to stress and may be related to elevated troponin and BNP levels seen in some patients(2,3). In addition to the specific cardiovascular changes associated with ESLD liver transplant candidates are prone to the cardiovascular diseases afflicting the population in general. Of particular interest to clinicians assessing patients for transplantation are coronary artery disease and structural cardiac abnormalities. Coronary Artery Disease The longstanding belief that cirrhosis was rarely associated with coronary artery disease(CAD) has been refuted. A prevalence of 27% in liver transplant candidates over 50yrs is reported(4). Published data report a high mortality (30% three month, 50% three year) and morbidity (80%) for patients with known CAD undergoing liver transplantation(5). Although the data can be criticized the potential for poor outcomes has led to the search for a means of identifying CAD patients in the pretransplant evaluation. Dobutamine stress echocardiography (DSE) has been the most widely accepted screening test for both theoretical reasons (the high cardiac output low SVR stress may best reflect conditions encountered during LT) and clinical experience in risk stratification for vascular surgery. Published assessments of the utility of DSE in identifying CAD in liver transplantation candidates are difficult to interpret as a group due to differences in selection criteria and outcome measures. A near perfect predictive value has been reported(6), however other studies have found a high false positive rate (positive DSE but no significant CAD on coronary angiography) and a poor predictive value of positive DSE for cardiovascular events during liver transplantation(7-9). The

7

utility of other testing modalities (exercise, radionucleotide scanning) is even less well defined. In this setting recommendations are hard to make and justify. Currently our practice is to refer for DSE those patients who meet criteria under the ACC/AHA guidelines. If significant inducible ischemia is detected then we perform coronary angiography. Patients with significant CAD then proceed to revascularization (usually by interventional cardiology), if possible). Patients who have corrected CAD and no evidence of inducible ischemia are considered satisfactory candidates. Despite the published data it is our impression that such patients are not at greatly increased risk of perioperative morbidity or mortality. Questions that merit answering in this area are whether the risks of CAD in liver transplantation remain high, or whether recent advances in the management of CAD have altered this. It would also seem that the recent literature on beta blockade in such patients would be pertinent. Finally, even if the perioperative period is safer the long term outcome in such patients should be investigated. Structural heart disease There is a very scant literature on structural heart disease and OLT. Valvular and other cardiac lesions are either known about prior to OLT candidacy or are discovered during assessment either clinically or on the screening echocardiogram. What is clear from the literature is that patients with ESLD who undergo cardiopulmonary bypass have high mortality so an approach based on surgical correction prior to OLT is of limited applicability. There are case reports of successful concurrent liver transplantation and valvular surgery. Preliminary reports suggest that dynamic intracavitary gradients can be successfully managed during OLT. Case reports of 2 patients with HOCM successfully managed. In absence of clinical evidence individual cases may be reviewed with knowledge of physiology of lesion and expected physiology of OLT and individual decisions made. Pulmonary Alterations in gas exchange are common in patients with ESLD. Many patients, particularly those with an alcohol etiology for their failure are or have been smokers. There are, however, some distinctive changes associated with ESLD, and two specific syndromes of note, hepatopulmonary syndrome (HPS) and portopulmonary hypertension (PortoPH). Due to the presence of ascites and/or pleural effusion restrictive lung defects are common. On blood gas testing, the most common abnormality is a mild respiratory alkalosis. The pathophysiology of this is not clear. An increased A-a gradient is present in 35%, less have hypoxemia. Decreased DLCO is common (50%)(10). A finding of hypoxemia should prompt investigation into the cause. Hepatopulmonary syndrome This syndrome is characterized by the triad of chronic liver disease, arterial hypoxemia and intrapulmonary vascular dilations. Its reported prevalence is 4-20% of candidates, the incidence being a reflection of the definitions used as the disease represents a continuum(11). Dyspnea is the most common symptom, orthodeoxia and platypnea can be seen. Oximetry or arterial pO2 measurements show improvements in supine versus upright measurements; most hypoxemia is readily improved by 100% oxygen, but may not be in advanced HPS. Diagnosis is made by demonstrating intrapulmonary vascular dilation by either contrast echocardiography (delayed contrast appearance in the left heart) or by radiolabeled albumin macroaggregate scanning(12). The underlying pathophysiology appears to be dilation of the pulmonary vasculature which, associated with high cardiac output, results in incomplete oxygenation of some red blood cells during their transit through the lungs. Nitric oxide has been implicated as a possible causative agent. HPS is reversible post liver transplantation, although in the more severe cases may take several months to resolve(13,14). Some studies have suggested a higher perioperative mortality associated with greater severity of HPS(15) whilst more recent series do not support this(13).

8

The longer time to recovery with more severe HPS, however, supports the policy of hastening transplantation in those individuals with significant HPS. Portopulmonary hypertension Portopulmonary hypertension (portoPH) is pulmonary hypertension associated with portal hypertension. Elevated PA pressures are not uncommon in liver transplantation candidates (up to 20%)(16). The reasons for this include the high flow state secondary to elevated cardiac output; the elevated volume state and potential for cardiac dysfunction related to cirrhosis; the presence of pulmonary vasoconstriction/plexogenic arteriopathy. PortoPH refers to those patients with the vasoconstrictive process similar to that of primary pulmonary hypertension exhibiting the spectrum of medial hypertrophy, intimal fibrosis and plexogenic arteriopathy with or without in situ thrombosis /recanalization. This occurs in approximately 1- 4% of candidates(17). The cause is unknown. The presenting symptom may be dyspnea but often the patients are asymptomatic. Investigations that suggest the diagnosis include an accentuated respiratory alkalosis and wider than anticipated a-A gradient(18). The diagnosis is more usually suggested by a high estimated pulmonary arterial systolic pressure on echocardiography. Confirmation is by right heart catheterization demonstrating elevated pulmonary arterial pressures with pulmonary capillary wedge pressure < 15mmHg. Categorization based on the mean PAP is usual: mild, MPAP 25-35mmH; moderate, MPAP 35-50mmHg; severe, MPAP > 50mmHg. The importance of identifying portoPH lies in the outcome with transplantation. A review of published cases identified a peritransplant mortality of 100% in severe portoPH; in patients with moderate portoPH and PVR > 250 dyne.s/cm5 mortality was 50%. No mortality was reported for patients with mild disease, or with moderate disease and PVR < 250 dyne.s/cm5 .(19) Management of patients considered too high risk for transplantation involves treatment to lower PA pressures as for primary pulmonary hypertension. Once hemodynamics return to an acceptable range transplant can then be considered. Intravenous epoprostenol has been used to successfully lower PAP, however mortality from progression of the underlying liver disease was 60% prior to transplantation being accomplished (20). Other agents currently under assessment include sildenafil, inhaled prostaglandins and bosentan. Preoperative assessment There is no good published data on what constitutes an adequate preoperative cardiopulmonary assessment prior to liver transplantation. Our approach is to have standardized testing for all candidates in addition to the history and physical examination, with a defined course, when possible if abnormalities are detected. All candidates receive 12-lead ECG, arterial blood gases and a transthoracic echocardiogram with estimation of PA systolic pressure. Those who meet ACC/AHA guidelines undergo dobutamine stress echocardiography, if positive for inducible ischemia they are referred to cardiology for coronary angiography. If PA systolic pressure is estimated to be > 50mmHg right heart catheterization is performed, discovery of severe portoPH or moderate portoPH with PVR > 250 dyne.s/cm5 results in deferral of transplantation and referral for vasodilator therapy. Candidates with hypoxemia have supine and upright gases checked, if HPS is a consideration either a contrast echo or a radio-labeled albumin macroaggregate scan is performed. Resting echocardiography for PA pressure estimation is performed at least yearly for candidates on the waiting list as portoPH may develop rapidly. For this reason we are also reluctant to perform liver transplantation without placing a pulmonary artery catheter.

9

References 1. Moller S, Bendtsen B, Henriksen J. Splanchnic and systemic hemodynamic derangement in

decompensated cirrhosis. Canadian Journal of Gastroenterology 2001;15:94-106. 2. Henriksen J, Gotze J, Fuglsang S et al. Increased circulating probrain natriuretid peptide (pro-BNP) and

brain natriuretic peptide (BNP) in patients with cirrhosis: relation to cardiovascular dysfunction and severity of disease. Gut 2003;52:1151-517.

3. Myers P, Lee S. Cirrhotic cardiomyopathy and liver transplantation. Liver Transplantation 2000;6:S44-S52.

4. Carey W, Dumot J, Pimentel R et al. The prevalence of coronary artery disease in liver transplant candidates over age 50. Transplantation 1995;59:859-64.

5. Plotkin JS, Scott VL, Pinna A et al. Morbidity and mortality in patients with coronary artery diease undergoing orthotopic liver transplantation. Liver Transplantation and Surgery 1996;2:426-30.

6. Plotkin JS, Benitez RM, Kuo PC et al. Dobutamine stress echocardiography for preoperative cardiac risk stratification in patients undergoing orthotopic liver transplantation. Liver Transplantation and Surgery 1998;4:253-7.

7. Donovan CL, Marcovitz PA, Punch JD et al. Two-dimensional and dobutamine stress echocardiography in the preoperative assessment of patients with end-stage liver disease prior to orthotopic liver transplantation. Transplantation 1996;61:1180-8.

8. Williams K, Lewis JF, Davis G, Geiser EA. Dobutamine stress echocardiography in patients undergoing liver transplantation evaluation. Transplantation 2000;69:2354-6.

9. Findlay J, Keegan M, Pellikka P et al. Preoperative dobutamine stress echocardiography, intraoperative events and intraoperative myocardial injury in liver transplantation. Transplantation Proceedings 2005;in press.

10. Mohamed R, Freeman J, Guest P et al. Pulmonary gas exchange abnirmalities in liver transplant candidates. Liver Transplantation 2002;8:802-8.

11. Schenk P, Fuhrmann V, Madl C et al. Hepatopulmonary syndrome: prevalence and predicitive value of various cut offs for arterial oxygenation and their clinical consequences. Gut 2002;51:853-9.

12. Krowka MJ, Cortese DA. Hepatopulmonary syndrome. Current concepts in diagnostic and therapeutic considerations. Chest 1994;105:1528-37.

13. Taille C, Cadranel J, Bellocq A et al. Liver transplantation for hepatopulmonary syndrome: aten-year experinece in Paris, France. Transplantation 2003;79:1482-9.

14. Battaglia SE, Pretto JL, Irving LB et al. Resolution of gas exchange abnormalities and intrapulmonary shunting following liver transplantation. Hepatology 1997;25:1228-32.

15. Krowka M, Wiseman G, Burnett O et al. Hepatopulmonary syndrome: a prospective study of relationships between severity of liver disease, PaO2 response to 100% oxygen, and brain uptake after 99mTc MAA lung scanning. Chest 2000;118.

16. Castro M, Krowka MJ, Schroeder DR et al. Frequency and clinical implications of increased pulmonary artery pressures in liver transplant patients. Mayo Clinic Proceedings 1996;71:543-51.

17. Krowka MJ. Hepatopulmonary syndrome versus portopulmonary hypertension: distinctions and dilemmas. Hepatology 1997;25:1282-4.

18. Kuo PC, Plotkin JS, Jonson LB et al. Distinctive clinical features of portopulmonary hypertension. Chest 1997;112:980-6.

19. Krowka M, Plevak D, Findlay J et al. Pulmonary hemodynamics and perioperative cardiopulmonary-related mortality in patients undergoing liver transplantation. Liver Transplantation 2000;6:443-50.

20. Krowka M, Frantz R, McGoon M et al. Improvement in pulmonary hemodynamics during intravenous epoprostenol (prostacyclin): a study of 15 patients with moderate to severe portopulmonary hypertension. Hepatology 1999;30:641-8.

10

Session 1 - Chronic Liver Failure Assessment for liver transplantation Prof. James Neuberger. Consultant Physician, Queen Elizabeth Hospital, Birmingham and Honorary Professor of Medicine, University of Birmingham Assessment of patients for liver transplantation When a person with liver disease is considered for liver transplantation, the risks and benefits of transplantation need to be balanced against the risks and benefits of continuing medical or surgical therapy. Most centres will state the transplantation is indicated either when the quality of life, because of the liver disease, is unacceptable to the patient or when the anticipated survival, in the absence of transplantation, is less than one year. In practice, the latter criterion is not followed in those with chronic liver disease. Because of the relative shortage of donor livers, a recent meeting agreed that liver transplantation is indicated only when there is a greater than 50% 5 year expected survival after transplantation, with the patient having a quality of life acceptable to the patient. Thus, the assessment of the patient with liver transplantation will need to address the following questions:

• Does the patient need a transplant at this time • Will the patient survive the procedure • Will the patient meet the 50% 5 year survival criterion • Does the patient understand the implications of transplantation

Patients with chronic liver disease Assessment of prognosis: the MELD (Model for end-stage liver disease) score has largely replaced the Child-Pugh classification. The former has a better performance and relies on laboratory rather than subjective data. US data suggest that a patient with a MELD score <12 is likely to have a greater survival probability without a transplant; the MELD score does not well predict outcome. The MELD score needs adjustment for those with malignancies. Absolute contra-indications: There are relatively few absolute contra-indications for transplantation; these include:

• Extra-hepatic cancer • Metastatic liver disease (although good palliation can be achieved in some cases of

neuroendocrine tumours) • Advanced cardiac/pulmonary disease so the patient will not survive the procedure or will not

survive 1 year Relative contra-indications: There are many relative contra-indications for liver transplantation. Usually, one or two will not preclude a patient but the greatest practical difficulty is in establishing how many relative contra-indications make an absolute contra-indication.

• Age: although there is no absolute age above which transplantation is contra-indicated, age is itself a risk factor for poor survival.

• HIV: the advent of effective treatment (HAART) for those with HIV infection has allowed a re-evaluation of transplantation in those who are HIV infected. Although there are no long-term data, centres will consider transplantation in those with low or undetectable levels of HIV RNA, a CD4 count >200/mlk and no AIDS-defining condition.

• Malnutrition: many patients with chronic liver disease have malnutrition; there are many contributing factors, including anorexia, use of inappropriate dietary restrictions, malabsorption and loss of protein. There is a close correlation between markers of nutritional assessment and degree of liver damage (assessed by MELD or Child score). There is no great evidence that malnutrition is, in itself, a significant risk factor for survival. However,

11

most units would consider severe protein malnutrition to be a risk factor. Conversely, morbid obesity is a relative risk factor.

• Vascular thromboses: portal vein thrombosis is not usually a contra-indication but extensive mesenteric thrombosis may preclude transplantation, even if vascular grafts are used.

• Cholangiocarcinoma: with the exception of some intra-hepatic cholangiocarcinomas discovered incidentally, cholangiocarcinoma is an absolute contra-indication, because of the early spread of these cancers, along the nerves and lymphatics. However, the Mayo Clinic has suggested that a small proportion may become suitable transplant candidates following Brachytherapy and chemotherapy.

• Diabetes: our studies have suggested that insulin-requiring diabetes (but not diet or tablet controlled disease) is associated with an increased risk of death post-transplantation.

• Renal impairment: renal failure is an adverse risk factor for survival; some of those with intrinsic renal disease (such as polycystic disease or glomerulonephritis) may benefit from a combined liver and kidney transplant.

• Past history of malignancy: depending on the time and type of malignancy, a previous history may contra-indicate transplantation.

• Drug misuse and past psychiatric history: although not in itself, an absolute contra-indication, past or current drug misuse is associated with non-compliance and a return to alcohol excess. A past history of psychiatric disease requires careful evaluation. The criterion for transplantation rests with whether the psychiatric illness can be successfully managed and allow the patient to lead a quality of life acceptable to the patient.

• Smoking: will add to the problems peri- and post-transplantation but is not a contra-indication.

• Previous upper abdominal surgery: several studies have suggested that previous surgery around the liver or stomach will add to the risks of the procedure.

• Active sepsis will contra-indicate transplantation but once this is treated, the patient can be safely grafted.

Indications and contra-indications

• Hepatopulmonary syndrome (HPS) and Pulmonary Hypertension (PPH): these may be indications for transplantation if not too advanced.

• PPH: It is important to differentiate true pulmonary hypertension from a hyperdynamic circulation: if mean PA pressure is less than 25mm Hg, then there is no PPH; if mean PA pressure is between 25-35mm Hg and PVR 150-250 dynes.s.cm-5, then the patient has mild PPH and no additional treatment is required. If mean PA >35mm Hg and PVR >250 dynes.s.cm-5, then the patient has moderate/severe PPH and is at risk of not surviving the procedure. Treatments that may be effective include prostacyclin, Bosentan and nitric oxide.

• HPS: this is suggested by orthodeoxia and confirmed by a bubble echo or MUGA scan. • Osteopenia: osteopenia is associated with chronic cholestatic liver disease; severe

osteoporosis may impair respiratory function so as to contra-indicate transplantation. Controversial indications:

• Alcohol: the professional and public debate about the place of liver transplantation for those with alcoholic liver disease (ALD) has meant that the focus of assessment has been lost. Transplantation is indicated in those where there is a low likelihood of the patient returning to alcohol consumption and will comply with follow-up. Most centres state (but do not always enforce) a rule that all candidates must be abstinent for 6 months, even though there is ample evidence that the length of abstinence does not predict subsequent compliance. A period of abstinence is needed to exclude those who will improve with abstinence to such a degree that transplantation is not needed and also to identify what measures can be put in place to reduce to a minimum the risk that the patient will return to drinking. The role of transplantation in those with alcoholic hepatitis is uncertain.

• Hepatocellular carcinoma (HCC): the Milan criteria are used to define those HCCs that are associated with a low likelihood of recurrence. These criteria include a single tumour less than 5cm on ultrasound or fewer than 3 nodules measuring less than 3 cm in diameter. More recently, the San Francisco group have suggested that these criteria are too limiting and it is possible to expand the criteria to include a single lesion <6.5cm in diameter, 3 or fewer lesions none greater than 3cms in diameter or a total tumour diameter of <8cm. In all cases, extra-hepatic spread or vascular invasion will contra-indicate transplantation. Of course, it is not the size that is important but the biological behaviour of the cancer that matters: the

12

degree of differentiation is a good marker but most centres avoid pre-transplant biopsy as this carries the risk of seeding the tumour. Many centres are using treatment regimes such as radio-frequency ablation and trans-arterial chemo-embolisation to treat or down-size tumours. There is no convincing data to suggest that these treatments will alter the natural history post-transplant.

• Poor quality of life: there are many reasons for a person with chronic liver disease to have a poor quality of life. Lethargy, associated with chronic liver disease, may be disabling but there are many other causes that will need excluding, such as co-existing depression, drug toxicity, electrolyte disturbance and so on. In our experience, lethargy does not always improve after transplantation. In contrast, transplantation is an effective treatment for intractable itching but all other modalities, including rifampicin, naltrexone and albumin dialysis (MARS), need trying first.

Assessment The following is the assessment as done in Birmingham (other centres will vary)

• Full medical and social history • Standard tests: FBC, liver tests, renal function tests, blood sugar • AFP for all with cirrhosis: use of CEA, CA125 and Ca19-9 for detection of

cholangiocarcinoma in those with PSC is uncertain • Virology: Hepatitis A, B, C (those who are not immune to A and B should be offered

immunisation); CMV (to consider need for prophylaxis/pre-emptive treatment for CMV disease), EBV status for children

• Auto-antibodies and immunoglobulins • Chest X-ray • Standard lung function tests • Arterial blood gases (standing, breathing air): if low PO2, repeat lying and standing to look for

orthodeoxia and on 100% O2 • Ultra-sound (for space occupying lesions, vessel patency) • Ultra-sound of neck vessels if previous cannulation • ECG • Isotope perfusion scan/coronary angiography if signs/symptoms or investigations suggestive

of ischemic heart disease • Echocardiogram (to measure LV size, assess pulmonary hypertension) – evidence of

pulmonary hypertension will require formal measurement of pulmonary artery pressures and dynamics.

• Colonoscopy for those with a history of colitis • For those with HCC: usually CT or MR imaging of abdomen and chest is indicated

Fulminant Hepatic Failure The Kings Criteria remain the gold standard for identifying those with a poor prognosis. Additional use of blood lactate will add precision to these prognostic markers. Criteria for transplantation include

• Aetiology of liver failure • Interval between onset of jaundice and encephalopathy • Serology: creatinine, bilirubin, prothrombin time (and INR), lactate • Evidence of progressive organ failure (such as haemodynamic instability, evidence of raised

intra-cerebral pressure). Full assessment requires an understanding of the indications and contra-indications for listing. There is often a narrow window between establishing the need for transplantation and the onset of complications (such as haemodynamic instability, sepsis or intra-cerebral hypertension) that render the procedure futile.

13

References 1. Carithers RL. Liver Transplantation. Liver Transplantation 2000;6:122-135. 2. Devlin J, O’Grady J. Indications for referral and assessment in adult liver transplantation: a clinical

guideline. Gut 1999;45 (suppl 6):VI1-VI22. 3. Farnswoth N, Fagan SP, Berger DH, Awad SS. Child-Turcotte-Pugh versus MELD score as a predictor

of outcome after elective and emergent surgery in cirrhotic patients. Am J Surg 2004:188:580-83. 4. Hoeper MM, Krowka MJ, Strassburg CP. Portopulmonary hypertension and hepatopulmonary syndrome.

Lancet 2004:363:1461-1468. 5. Neuberger J. Developments in liver transplantation. Gut 2004;53:759-68. 6. Neuberger J, Schulz KH, Day C, Fleig W, Berlakovich GA., Pageaux GP et al. Transplantation for

alcoholic liver disease. J Hepatol 2002;36:130-137. 7. Schwartz M. Liver transplantation in patients with hepatocellular carcinoma. Liver Transpl 2004;10(suppl

1):S81-S85 8. Yao FY, Bass NM, Ascher NL, Roberts JP. Liver transplantation for hepatocellular carcinoma. Liver

Transplantation 2004:10:621-30.

14

Session 2 - Surgical and donor considerations Optimising the graft Dr Jonathan Mackay. Consultant in Cardiothoracic Anaesthesia and Intensive Care, Papworth Hospital, Cambridge Pathophysiology and management of potential heart beating organ donors

Results of organ transplantation continue to improve though demand for organs outstrips supply. Organs for transplantation come from three sources

1. Heart beating donors following diagnosis of brain death 2. Non-heart beating donors following cardio-respiratory arrest 3. Living related organ donors

Early transplantation relied on organs from non-heart beating and living related donors. The introduction of direct brain stem testing in the 1970s’ changed the emphasis to utilising organs from so-called “heart beating” donors. Improvements in diagnosis and management of severe brain injuries together with a reduction in numbers of severe head injuries have resulted in fewer patients fulfilling brain stem testing criteria and renewed interest in “non-heart beating” and living related donors. Although numbers are declining, heart-beating donors remain the most common source of transplantable organs and a very precious resource. I aim to cover the following three areas in my 25-minute presentation.

1. Pathophysiology of brain stem death and implications for potentially transplantable organs 2. General principles and potential conflicts in donor optimisation 3. Specific strategies to improve liver graft function

1. Pathophysiology of brain stem death and implications for potentially transplantable organs Pathophysiological changes after brain stem death jeopardise the function of potentially transplantable organs. Haemodynamic changes may cause endothelial shear stress, vasoconstriction, ischaemia, endothelial activation and an inflammatory response. Although the longer cold ischaemia times of grafts taken from brain dead donors are probably the primary cause of poorer outcomes compared to living donors, it is increasingly believed that the enhanced inflammatory status after brain death is an important secondary contributory factor. The clinical course of a ventilated but otherwise unsupported brain stem dead patient is short with asystolic cardiac arrest generally occurring within 72 hours. However cardiac and other body functions have been maintained for many days in fully supported brain dead patients.

Cardiovascular

There are usually two distinct phases, characterised by sympathetic overactivity and underactivity respectively. The first phase is not seen in all patients.

1) Hyperdynamic phase: Sympathetic overactivity causes a transient catecholamine surge, (particularly adrenaline and noradrenaline) which increases heart rate, blood pressure, cardiac output and systemic vascular resistance. The catecholamine storm adversely affects the delicate balance between myocardial oxygen supply and demand.

2) Cardiovascular collapse phase: Hypotension results from loss of sympathetic tone,

profound vasodilatation and myocardial depression. Hypovolaemia secondary to diabetes insipidus frequently contributes to hypotension in unsupported patients with brain stem death. Subendocardial myocardial ischaemia and ventricular dysfunction are common even in previously healthy hearts. Most changes are usually temporary and reversible.

15

Endocrine

Reduced production of anti-diuretic hormone (ADH) causes diabetes insipidus (DI), which occurs in up to 65% of organ donors. It is characterised by diuresis, hypovolaemia, plasma hyperosmolality and hypernatraemia. Reductions in cortisol production are unrelated to the degree of hypotension but may impair donor stress response. Anterior pituitary failure is less common and function may be preserved due to an extradural blood supply. Failure causes significant reductions in the circulating levels of tri-iodothyronine (T3) and thyroxine (T4). ↓ T3 may contribute to cardiovascular deterioration. Insulin secretion is reduced and contributes to the development of hyperglycaemia. This may be aggravated by the administration of large volumes of glucose containing fluids, if used to treat hypernatraemia and increased levels of catecholamines. Untreated hyperglycaemia leads to increased extracellular osmolality, metabolic acidosis, osmotic diuresis and cardiovascular instability.

Pulmonary

Rapid haemodynamic changes including transient pulmonary hypertension may disrupt capillary integrity and cause severe lung injury. Pulmonary dysfunction is common in the organ donor and may be due to the development of pneumonia, aspiration of gastric contents, neurogenic pulmonary oedema or pulmonary trauma.

Liver & Coagulopathy

Liver viability is impaired after brain death though the exact mechanisms of reduced graft quality are still largely unknown. Histological changes include central venous congestion and leukocyte infiltration. Hepatocellular enzymes may be elevated and bile production reduced. Prothombin time is frequently elevated secondary to release of tissue thromboplastin by ischaemic or necrotic brain. Disseminated intravascular coagulation is common.

Hypothermia

Hypothalamic failure after brain stem death results in impairment of temperature regulation. Heat production is reduced because of a fall in metabolic rate and loss of muscular activity. This is associated with an increase in heat loss because of peripheral vasodilatation.

Table 1: Pathophysiology of Brain Stem Death 2. General principles and potential conflicts in donor optimisation Resuscitation and monitoring of the organ donor Following the certification of death by brain stem testing and the lack of objection to donation, there is a change in the emphasis of care. Therapy previously aimed at preserving brain function is now directed at optimising transplantable organ function. There should be no decrease in patient dependency because the need for therapeutic intervention and support of relatives continues. Adequate time must be allowed for confirmation of brain stem death but unnecessary delays should be avoided to minimise the risk of deterioration of the donor. High quality critical care including chest physiotherapy, aseptic precautions and antibiotics may be required. The patient has usually been rendered slightly hypovolaemic by brain protecting therapies. The optimal volume status is controversial and there is disparity between the interests of lungs and kidneys. Although hypovolaemia should be corrected, it is important to avoid excessive volume replacement particularly in potential lung and heart donors. The primary goals are maintenance of adequate tissue perfusion and preservation of organ viability. Key considerations include myocardial dysfunction, potentially reversible lung injury, invasive infections, hypernatraemia together with trends in hepatocellular enzymes and creatinine. Optimal haemodynamic management requires invasive monitoring. Because of the order in which the great vessels are ligated during the donor operation, any newly placed arterial cannula should be inserted into the left radial or brachial artery. Equally a new central venous or pulmonary artery catheter should be inserted into the right internal jugular or subclavian veins. Intravenous fluid administration must be carefully monitored as organs, particularly the lungs, which may have been damaged during the period of sympathetic hyperactivity, are susceptible to volume overload and capillary leakage.

16

Echocardiography is useful to exclude major structural abnormalities of the heart and to measure left ventricular ejection fraction. Some transplant units will only request insertion of a pulmonary artery catheter in those patients estimated to have a left ventricular ejection fraction below 45%. Transthoracic echocardiography (TTE) may be technically difficult and better images may be obtained with transoesophageal echocardiography (TOE). Transient regional wall motion abnormalities are common and systolic inward motion and thickening may improve with haemodynamic optimisation. Assessment of right ventricular size and function is important and frequently challenging.

Cardiovascular

The goals of haemodynamic management are to optimise cardiac output maintaining normal preload and afterload. Where possible, the use of high dose β adrenoreceptor agonists, other inotropes and vasopressors, which increase myocardial oxygen demand and deplete myocardial high-energy phosphates should be avoided. Choice of inotropic support varies between transplant units and may be guided by data from pulmonary artery catheterisation but:

• High dose adrenaline may result in detrimental vasoconstriction in donor organs. • The vasodilator effects of dobutamine may lead to undesirable hypotension and

tachycardia. • Vasopressin is less likely to cause metabolic acidosis or pulmonary hypertension

and may be more appropriate than noradrenaline for the cardiovascular collapse phase.

Endocrine

Hormone replacement therapy may reduce inotrope requirements and should be considered in all organ donors. Debate continues over the value of T3 replacement. Practical difficulties identifying the subgroup of patients with decreased free T3 have led to most transplant units empirically commencing T3 infusions in all potential organ donors. High dose methylprednisolone 15 mg.kg-1 is commonly given as part of the hormone package to diminish the inflammatory response.

Pulmonary

If the lungs are to be transplanted, the FiO2 should be kept at or below 0.4 to minimise the risks of oxygen toxicity. A modest level of positive end expiratory pressure (PEEP) (i.e. <5 cmH2O) will prevent alveolar collapse. Strict asepsis should be continued during physiotherapy and tracheal toilet. Suitable goals for respiratory support are:

• Maintain normocapnia (PaCO2 ~ 5.0-5.5 kPa). • Ventilate with lowest FiO2 to maintain PaO2 of >10.0 kPa. • PEEP > 5 cmH2O may reduce cardiac output and is rarely required. • Avoid high inspiratory pressures

Renal

Hypotension is associated with acute tubular necrosis and failure of transplanted kidneys. Although low dose dopamine is now unfashionable in the general critical care setting, there is some evidence that donor pre-conditioning with dopamine improves initial graft function after kidney transplantation.

Haematology

Haemoglobin concentration should be maintained over 9 gm.dl-1. Deranged coagulation should be treated with fresh frozen plasma and platelets.

Microbiology

Daily endotracheal, urine and blood cultures. Antibiotics for presumed or proven infection.

Temperature

Hypothermia should be anticipated and heat loss prevented by using:

• Warmed intravenous fluids and warming blankets • Heated and humidified inspired gases • Increased ambient temperature

Table 2: General support strategies of the potential heart beating organ donor

17

The appropriate haemodynamic goals and principles of haemodynamic management for potential adult heart donors are given in Tables 3 and 4. Patients who do not achieve these haemodynamic goals may still be considered for donation of other organs. Therapies used to correct the common endocrine disturbances are shown in Table 5.

Mean Arterial Pressure 60-80 mmHg

Preload Central venous pressure (CVP) ~ 4-8 mmHg for potential lung donors Pulmonary artery occlusion pressure ~ 10-15 mmHg

Heart Rate 60 – 100 beats.min-1

Rhythm Sinus rhythm is desirable

Cardiac Output Cardiac Index > 2.1 l.min-1.m-2

Table 3: Appropriate haemodynamic goals for potential adult heart donors. (Higher CVP ~ 10-12 mmHg, lower mean arterial pressure and cardiac index are permissible if only abdominal organs being considered for transplantation.)

Clinical status Haemodynamic management

↓ Mean Arterial Pressure ↑ Cardiac Output Preload optimisation and Vasopressor to ↑ afterload

↓ Mean Arterial Pressure ↓ Cardiac Output Preload optimisation, Vasopressor to ↑ afterload and Inotrope to ↑ contractility

↑ Mean Arterial Pressure ↓ Cardiac Output Preload optimisation and Vasodilator to ↓ afterload ± Inotrope to ↑ contractility

Table 4: Principles of critical care management for adult heart donation Clinical problem

Management

Diabetes Insipidus

Maintain • Na+ ≤ 155 mmol.l-1 with 5% dextrose • Urine output about 1 - 2 ml.kg-1.h -1 with vasopressin (pitressin) 1 U bolus and 0.5-4.0 U.h-1 infusion. Higher doses of vasopressin may cause coronary, renal or splanchnic vasoconstriction. If vasopressin fails to control diuresis, intermittent desmopressin (DDAVP), which is highly selective for the V2 receptor found in the renal collecting ducts, may occasionally be required.

Hyperglycaemia Insulin infusion to maintain plasma glucose 4-9 mmol.l-1 Maintain K+ >4.0 mmol.l-1

Hypothyroidism Tri-iodothyronine (T3) 4 µg bolus then infusion at 3 µg.h-1

Table 5: Common endocrine problems seen in potential heart beating donors 3. Strategies to improve liver graft function A) Liver protection Primary liver dysfunction in the first week is generally attributable to preoperative or intra-operative problems. Ensuring high quality organ protection is the cornerstone of transplantation. Most centres rely on cold intermittent perfusion for protection of livers from beating heart donors. University of Wisconsin (UW), Histidine-Tryptophan-Ketoglutarate (HTK) and Celsior solution are three of the most commonly used solutions for liver preservation. Key differences in composition

18

(electrolytes, osmotic agents and buffers) will be illustrated and results of clinical studies summarised. Ischaemia leads to rapid breakdown of ATP and alters mechanisms for handling the breakdown products of ATP causing free radical damage on reperfusion. Cold storage aims to minimise ATP depletion as metabolism decreases by 7% for every 0C fall in temperature. Cooling therefore slows but doesn’t entirely eliminate cell dysfunction. Prolonged cold ischaemia is an important factor affecting graft outcome. Although the liver tolerates longer cold ischaemia times than the cardio-thoracic organs, sinusoidal cells may show signs of degeneration after just 4 hours of cold storage. Reperfusion injury: The sinusoidal lining cells of the liver are particularly vulnerable to the effects preservation reperfusion injury. Strategies of potential value in attempting to reduce ischaemia and reperfusion injury include anti-inflammatory pre-treatment of donors, ischaemic preconditioning and use of various flush solutions before reperfusion. Continuous (as opposed to intermittent) liver perfusion has a greater potential importance in controlled non-heart beating donations. Reperfusion and resuscitation of liver ATP stores can be undertaken before cold storage. B) Donor glycaemic control: Hyperglycaemia is common in patients with brain death. It has recently been shown that strict glucose control (4.4 – 6.1 mmol l-1) improves survival in critically ill patients by reducing septic deaths. However, the hypothesis that similar tight glycaemic control in the brain dead donor improves graft survival has not been tested. Most intensivists aim to maintain glucose levels 4 –9 mmol l-1 with insulin infusions. C) Donor nutrition: The importance of nutritional support in multi-organ donors is unclear. Animal studies have demonstrated significant improvements of hepatic lining cell viability and reduced reperfusion injuries with enteral & parenteral nutrition respectively. Demonstration of benefits of nutrition in humans has been harder to establish. Gonzalez failed to identify an independent effect donor nutritional status on postoperative graft function. Most transplant units initiate and continue enteral nutrition as tolerated until shortly before the multi-organ donor operation. However, feeding should probably be discontinued 4 hours before change of endotracheal tube if for instance the lungs are being considered for transplantation and a larger endotracheal tube is required for bronchoscopy. D) Donor hypernatraemia: Donor hypernatraemia >155-mmol l-1 is associated with hepatic dysfunction after transplantation. The mechanism is uncertain but may be related to accumulation of idiogenic osmolytes or intracellular solutes that subsequently cause intracellular swelling following transplantation.

19

Session 2 - Surgical and donor considerations Operative techniques and hazards Mr Chris O’Sullivan. Consultant in HPB and liver transplantation surgery, Freeman Hospital, Newcastle Orthotopic Liver Transplantation Standard Technique (end-to end cavocavostomy); Piggy-Back (end-to-side cavocavostomy); Modified Piggy-Back (side-to-side cavocavostomy) The standard technique (ST) of orthotopic liver transplantation (OLT), as described by Starzl in 1963, involves resection of the recipient inferior vena cava (IVC) from above the renal veins to the diaphragm as part of the hepatectomy and then interposition of the donor’s vena cava attached to the graft as an end-to-end cavocavostomy. The cross-clamping of the IVC and portal vein during the anhepatic phase of the standard technique results in a decreased filling pressure, decreased cardiac output, decreased renal perfusion pressure, increased infrahepatic caval pressures, and moderately decreased systemic arterial pressure. Successful application of veno-venous bypass (VVB), as introduced by Shaw in 1984, allowed for improved haemodynamic stability during the anhepatic phase and allowed decompression of the splanchnic venous system. Other advantages reported from VVB include reduced requirements of blood products, preservation of renal function, and improved 30-day survival. At its introduction the advantages of VVB were considered such that many adopted it as a routine procedure during OLT. However VVB is associated with significant complications such as hypothermia, coagulopathies, vein thrombosis and air embolism and a longer operating time and despite extensive use of VVB in OLT its use during the anhepatic phase continues to be controversial. In a recent survey of North American Transplant Centres the routine use of VVB has significantly decreased from 91% in 1987 to 42% in 1997. Some centres use VVB routinely, others do not use it at all, and some employ a selective policy depending on the pre-operative renal function, diagnosis, or IVC trial-clamping. Calne in 1968 was the first to describe the Piggy-Back technique (PBT), with preservation of the recipient IVC during hepatectomy and anastomosis of the donor suprahepatic IVC to the recipients hepatic veins or cava (end-to-side cavocavostomy). The advantages this technique were maintenance of the caval flow during explantation, leading to reduced fluid resusitation requirements, better maintenace of core body temperature and improved cardiac hemodynamic stability. Furthermore it did not require dissection of the retrocaval compartment, thus reducing retroperitoneal blood loss. The technique also had the advantage of resolving technical challenges such as different recipient and donor vena cava sizes, particularly in paediatric, reduced size, or living donor liver transplantation. Indeed the PBT was originally intended for children receiving reduced size grafts and in selected adults such as those receiving grafts from a significantly smaller donor but has gained much greater acceptance in recent years in adult OLT particularly in Europe. As originally described, the PBT was frequently performed using transient clamping of the IVC, either because of difficulty in disconnecting the native liver from the retro-hepatic IVC or to fashion the upper caval anastomosis. As a consequence the procedure was occasionally performed with VVB. To avoid the need for VVB, Belghiti, described a modification of the PBT (MPBT) in which caval flow was preserved during the entire OLT procedure. The modification made preservation of caval flow possible by performing either a side-to-side cavocavostomy while side clamping the IVC and having repaired donor upper and lower caval orifices on the back table or performing an end-to-side cavocavostomy using the recipient middle and left hepatic vein cuff with partial clamping of the IVC. Despite the advantages of performing OLT with recipient caval preservation and without VVB, caval complications such as obstruction or stenosis appear to occur more frequently than with the ST end-to-end cavocavostomy. Caval anastomotic complications after OLT are mostly due to technical problems and usually require reoperation or retransplantation with recognised mortality of as high as 60% in late diagnosed cases.

20

In a consecutive series of 1000 patients undergoing ST with end-to-end cavocavostomy, the incidence of caval complications was 1.7%. The incidence of caval stenosis after MPBT with side-to-side cavocavostomy was reported as 14.2% in one series and 5.2% in another. With the PBT the incidence has been reported as high as 7.8%. Caval obstruction in the PBT is usually due to either kinking or torsion of the middle and left hepatic vein stump, which has been left too long to facilitate an easier anastomosis. Stenosis tends to occur as a result of a size mismatch between the donor suprahepatic IVC and recipient hepatic vein stump, leading to a significant pressure gradient across the anastomosis. Both conditions result in outflow obstruction, with liver congestion, ascites and impaired renal function (Budd-Chiari syndrome). Outflow obstruction/stenosis can be successfully treated following the PBT using balloon angioplasty, with metallic Wallstent placement useful in refractory cases. Obstruction can be treated surgically in PBT cases by anastomosing the graft infrahepatic cava in an end-to-side fashion to the recipient retrohepatic recipient cava. In the MPBT outflow obstruction can occur if an adequate window has not been cut in the recipient IVC prior to the side-to-side cavocavostomy and the donor graft is small and rotates usually to the right side occluding the venous return. The bile duct anastomosis in liver transplantation Surgical technique and complications Biliary tract complications are frequently still referred to as the ‘Achilles heel’ of liver transplantation with an overall incidence of 8-15%, and a recognised mortality of 10%. Although the causes of post-OLT complications include occult hepatic artery thrombosis, extended cold ischaemia time of the graft, other injuries to the biliary epithelium such as chronic rejection, the majority of biliary complications are directly related to surgical technique. One of the principle technical advances of OLT in the early years has been the standardisation of the biliary reconstruction. The gallbladder of the donor is now routinely removed and procedures such as cholecystoduodenostomy or cholecystojejunostomy have been abandoned. Biliary reconstruction during OLT is generally performed by end-to-end or occasionally side-to-side choledocho-choledochostomy (CC), or a Roux-en-Y choledochojejunostomy (CRY). End-to-end and side-to-side CC have been shown to be equally effective. Where the recipient bile ducts are healthy and of suitable calibre then the anastomotic technique of choice is the CC. CRY is preferred in cases of pre existing biliary tract disease (i.e. primary sclerosing cholangitis) or previous biliary tract surgery. Until a few years ago the majority of biliary reconstructions following OLT were performed with the placement of a T Tube across the anastomosis to act as a splint, to test the anastomosis prior to removal of the T Tube and to act as a potential route for radiological intervention in the case of a biliary stricture or leakage. Two prospective randomised trials have shown that there was either no difference or an increase in biliary complications associated with T Tube insertion and that T Tubes did not provide safer access to the biliary tree compared to other types of intervention. Therefore the routine use of T Tubes has been largely abandoned. Most biliary complications occur in the first 3 months after OLT, with most leaks occurring in the first month. Biliary strictures tend to develop later. Biliary leaks and strictures represent 70% of all biliary complications regardless whether the reconstruction is CC or CRY. In almost 90% of cases the biliary obstruction is located at the site of the anastomosis. About 50% of leaks are located at the site of anastomosis, followed by leakage at the site of T Tube outlet. Leaks may also appear due to aberrant bile ducts or as a result of hepatic artery thrombosis. Over 75% of all biliary tract related deaths are associated with biliary leaks. Although there is no apparent difference in the incidence of leaks between CC and CRY the leak-related mortality is considerable and significantly higher in CRY, presumably due to leakage of enteric content. Hepatic anastomotic strictures, occurring within two years of OLT, in patients who have a CC reconstruction can be successfully treated with repeated endoscopic balloon dilatation and stent placement. There are now several studies with long term follow up, which support this approach. Failed endoscopic therapies should undergo surgical revision of the anastomosis with conversion to a CRY.

21

Arterial reconstruction and revascularisation of the graft Hepatic artery thrombosis, arterial reconstruction, arterial conduits In contrast, to the native liver, which in most cases, is adequately perfused by the portal vein and from collaterals from the reteroperitoneal diaphragm, the arterial blood supply to the graft is of great importance, especially in the early postoperative period. This is because potential collateral supply to the graft is cut at the time of hepatectomy and the transplanted liver is thus particularly dependent on the arterial blood flow. Hepatic artery thrombosis (HAT) is still one of the commonest technical complications occuring in 2-12% of cases after OLT, leading to rapid graft loss requiring urgent retransplantation (up to 75% of cases), or, if it occurs later, biliary strictures, bilomas, recurring septic episodes and impaired liver function. The incidence is much higher in children (10-40%) because of smaller calibre donor and/or recipient arteries and greater fluctuations in coagulation factor concentrations and haemocrits in children. The main reasons for HAT in adults is technical failure, followed by preservation injury-induced vasoconstriction and fulminant or uncontrolled allograft rejection with intimal thickness and increased vascular resistance. The clinical presentation of HAT varies from mild elevation of serum transaminases and bilirubin levels to fulminant hepatic necrosis. This variation is to a large degree time dependant. Two main forms of HAT are recognised (1) an acute presentation usually characterised by massive injury to hepatocytes and biliary epithelium leading to biliary leaks, sepsis and graft failure and (2) a delayed presentation generally associated with a milder clinical course. Overall about a third of patients with HAT are asymptomatic, a third are not immediately life threatening but present with biliary tract ischaemia and a third cause rapid graft failure and require immediate retransplantation with an overall mortality of 50%. Up to 70% of patients with HAT will come to retransplantation. Urgent revascularisation using an arterial conduit has been shown to be effective in avoiding retransplantation in a significant number of patients, if HAT is recognised early. Arterial reconstruction is dependent on the anatomy and size of the donor and recipient hepatic arteries, recognition of anatomical variants and consideration to the inflow through the recipient celiac artery. End-to-end anastomosis of the recipient’s common hepatic artery with the donor celiac trunk is the most commonly used technique for arterial reconstruction. A variation on this is to anastomose the donor coeliac trunk with a small aortic patch to the recipient’s common hepatic artery at the division of the hepatic and gastroduodenal artery. This procedure allows for a wide anastomosis and a double inflow via the common hepatic artery as well as via the mesenteric artery. However only 55% of the population (and therefore donor organs) have conventional anatomy where the common hepatic artery, a branch of the celiac trunk, supplies the liver through its right and left hepatic branches. Other variants include:

a) accessory left hepatic artery from the left gastric artery b) accessory right hepatic artery from the superior mesenteric artery c) accessory left hepatic artery from the left gastric artery and an accessory right hepatic

artery from the superior mesenteric artery d) common hepatic artery originating as a branch from the superior mesenteric artery e) common hepatic artery arising from the aorta directly

Accessory donor hepatic arteries are often reconstructed by re-implantation at the insertion of the splenic or gastroduodenal artery on the back-table. There does not appear to be an increase in the complication rate (occlusions, stenosis of the hepatic artery) following this reconstruction. The success of the reconstructed artery depends not only on the quality of the arterial anastomosis but also on the inflow in the preanastomotic arterial segment. It has been reported that in about 12% of adults and 50% of children the common hepatic artery is not suitable for anastomosis due either to hypersplenism or stenosis of the coeliac artery (often due to compression of the coeliac trunk by the median arcade ligament).

22

Arterial conduits using donor iliac artery have provided an effective and reliable method of revascularisation in patients at higher risk of arterial thrombosis, particularly small children, and those undergoing retransplantation for HAT. Their use is indicated whenever there is:

a) doubt about the quality of the inflow provided by the native hepatic artery due to coeliac artery stenosis or hypersplenism

b) small, multiple, or anomalous recipient hepatic arteries c) friable or attenuated native hepatic arteries such as during retransplantation d) small recipients in whom the recipient artery is less than 3mm in diameter e) or for more complex procedures including reduced, split, living related, or auxillary liver

transplantation. The incidence of HAT when the iliac arteries are used as conduits is similar to that in the standard direct arterial anastomosis with 1-year and 5-year patency rates of 88 and 85% respectively. The infrarenal site for the iliac conduit is safe and access is usually good. The theorectical advantage of placing a supracoeliac aortic conduit is that it is shorter and therefore has a lower vascular resistance. The construction of this conduit however must usually be performed during the anhepatic phase, as space is limited when the graft has been implanted. The use of the splenic artery for arterial inflow has been described, but it involves ligation of the distal splenic artery in patients who have often got significant splenomegaly resulting in splenic infarction. The use of cryopreserved vascular grafts has been complicated with aneurysmal dilatations and stenoses. The cryopreservation process may cause both direct endothelial damage and immunologic response to the altered tissues of the vascular wall. Cadaveric arteries can be stored for up to 14 days in a lymphocyte culture medium (Terasaki solution) without increase in the rate of HAT. Conduits can also be used for emergency revascularisation to rescue allografts after early HAT and in cases of postoperative mycotic aneurysm formation. Portal vein The portal vein (PV) anastomosis is an end-to-end anastomosis of the donor and recipient portal vein with a ‘growth factor’ (one third to one half of the vein diameter) at both ends of the running sutures to allow additional spacing of the anastomotic site. Both the donor and recipient PV should be kept short to avoid kinking of the PV after revascularisation. Postoperative stenosis and thrombosis is a rare complication. Preoperative thrombosis of the PV, present in 10% of cirrhotic patients, is still regarded as a relative but rarely as an absolute contraindication for liver transplantation. In the case of partial or total PV thrombosis reconstruction of the donor PV with recipient superior mesenteric vein (SMV) or one of its larger branches, or with a collateral vein has been described. Arterialisation of the PV has also been used to provide inflow. If the PV thrombosis is fresh or short, thrombectomy can be performed with if necessary resection of a small part of the PV and anastomosis to the SMV. When the thrombosis involves the confluence of the SMV and splenic vein interposition of donor iliac common vein beyond the pancreas to bridge the defect and to avoid an extra-anatomical low-pressure venous conduit around the pancreas can be performed. However, if the thrombosis reaches the branches of the mesenteric veins, reconstruction of the PV may be technically impossible and OLT is then not an option.

23

Session 2 - Surgical and donor considerations Increasing the organ supply Mr Nigel Heaton. Consultant Surgeon, King’s College Hospital, London

24

Session 4 - Anaesthetic management The “good” patient Prof. Mark Bellamy. Professor of Critical Care Medicine, St James’ University Hospital, Leeds Since the first successful orthotopic liver transplant by Starzl in 19631, liver transplantation has become the treatment of choice for end stage chronic liver disease and some cases of acute hepatic failure2 3. Overall one-year survival rates now exceed 80% while improvements in immuno-suppression have resulted in prolonged graft survival and 5 year survival rates. This compares favourably with the medical treatment of acute liver failure which in some patient groups yields only 5 - 20% survival4. Liver transplantation is equally applicable to young and old patients with good success rates in the over-70 age group, provided patients are selected without limiting comorbidity5. Indications Indications for liver transplantation are: chronic irreversible liver failure, acute hepatic failure and primary hepatic malignancy. Suitability for transplantation depends more on severity of hepatic dysfunction than diagnosis. Specific prognostic criteria are used in to select patients with acute liver failure6 and chronic liver disease7. Conditions associated with poor outcome represent relative contra-indication. Transplantation for primary hepatobiliary malignancy is considered in the absence of metastasis but results poor (25 - 40% 2-year survival). Other indicators of poor prognosis include impaired renal function, hyponatraemia, muscle wasting, impaired cardiopulmonary function and severe pulmonary hypertension. Mild to moderate pulmonary hypertension (mean <35 mmHg) is not a contraindication to transplantation8. Preoperative assessment Preoperative assessment should be performed jointly by hepatologist, surgeon and anaesthetist prior to listing for transplantation. This requires an understanding of the pathophysiology of liver failure and likely complications. Assessment is aimed at evaluating both hepatic dysfunction and presence of concomitant disease. History and examination often reveal poor exercise tolerance, ascites and pleural effusions. Stigmata of chronic liver disease suggest poor hepatic synthetic function. Exercise ECG, echocardiogram and gated gamma scintigraphy (MUGA scan) help exclude potentially deleterious ischaemic heart disease and cardiomyopathies. Dobutamine stress-testing, dipyridamole-thallium scanning or coronary angiography may be necessary 9. Chest radiograph and respiratory function tests help exclude specific pathologies as well as pleural effusions. Up to 10% of cirrhotics suffer some pulmonary hypertension, often in association with portal hypertension. The hepatopulmonary syndrome, with associated hypoxia, is not a contraindication to liver transplantation as it may resolve following successful grafting in a high proportion of cases. Portal hypertension may trigger ascites formation. Ascites is often associated with water retention and consequent low serum sodium. If below 122 mmol/l this should be carefully corrected pre-operatively (days or weeks) to reduce the risk of pulmonary oedema and central pontine myelinolysis associated with acute peri-operative sodium shifts. Pre-operative correction of hypoalbuminaemia or coagulopathy is unnecessary except where there is a specific indication (e.g. active bleeding), as there is little if any relationship between this and intraoperative blood loss10. However, increased perioperative blood loss seems associated with a higher complication and mortality rate11. Sedative premedication is best avoided in patients with hepatic encephalopathy. Benzodiazepine premedication seems safe in other elective cases. Intraoperative management General principles Surgery falls into three phases. Phase I involves dissection of the structures around the liver and porta hepatis, achieving mobilisation. Stage II, anhepatic phase, involves division of hepatic

25

artery, portal vein and bile duct. The vena cava is cross-clamped both at the diaphragm and immediately below the liver. Some centres employ veno-venous bypass to shunt blood from lower cava and portal vein back to the heart via a centrifugal pump12. Systemic anticoagulation is not required. While this may reduce gut oedema, the evidence for improved renal function is equivocal13. The liver together with its included portion of vena cava is then removed, and new liver inserted. Caval and portal anastomoses are fashioned. An alternative technique involves caval preservation with partial side-clamping, followed by reimplantation involving a cavocavostomy. Phase III, the reperfusion phase, begins when blood is allowed to re-enter the portal vein, and caval clamps are removed. Finally, the hepatic artery is anastomosed and biliary system reconstructed. Monitoring and induction of anaesthesia Prior to induction of anaesthesia, ECG and direct arterial pressure monitoring are commenced. Invasive cardiovascular monitoring may be established pre- or post- induction of anaesthesia, depending on the condition of the patient. Radial arterial cannula, triple lumen central venous line and other haemodynamic monitors are inserted. Alternatives include fibre-optic oximetric pulmonary artery flotation catheter, Picco, LiDCO and trans-oesophageal echo-cardiography. Temperature is monitored either from the PA catheter or from a central temperature probe. Trans-oesophageal echocardiography provides continous information on venticular wall motion and embolic phenomena, but is not available in all centres. Patients with acute liver failure are at risk of raised intra-cranial pressure (ICP) and are transferred from intensive care with intra-cranial pressure monitoring in situ. A large bore nasogastric tube and urinary catheter are also required. Full blood count, clotting, electrolytes and blood gases are monitored hourly or as clinically indicated. These guide subsequent ventilation, Ca2+/K+ supplementation, glucose and bicarbonate infusions and blood replacement therapy. Thrombo-elastography is widely used as a rapid and sensitive means to gauge clotting factor requirements14 15. Prior to induction of anaesthesia, an intravenous infusion is commmenced. Many combinations of drugs have been used for induction of anaesthesia, the main requirement being cardiovascular stability and rapid tracheal intubation. Some centres routinely employ a traditional "rapid sequence" technique. Increasing interest in early extubation has lead in recent years to the use of shorter-acting drug combinations, including midazolam, propofol and remifentanil. The patient is positioned with one or both arms abducted to a maximum of 80o (hyperextension has been associated with brachial plexus injury). A hot air warming overblanket maintains normothermia16. The lower chest and abdomen are exposed for surgical access. After intubation IPPV is established with a suitable volatile agent, eg isoflurane 0.5%-1.5% end-tidal concentration in oxygen-enriched air. Many centres also employ desflurane or sevoflurane. Nitrous oxide is avoided to reduce cardiovascular depression, gut distension and bubble formation if on venovenous bypass. Total Intravenous anaesthesia (TIVA) using propofol has been used successfully but may be associated with cardiac depression following reperfusion17. Both fentanyl alfentanil and remifentanil infusion have been used with good effect, though there is a tendency to cumulation and prolonged post-operative ventilation with fentanyl. The choice of muscle relaxant depends on clearance, volume of distribution and clinical preference. Most relaxants are reasonable and safe provided neuromuscular blockade is monitored to prevent delayed recovery. At standard infusion rates atracurium provides reliable relaxation and reversal after prolonged use. Pancuronium has an increased volume of distribution in liver disease, while both pancuronium and vecuronium show reduced clearance18 19. Their use requires caution. Low-dose dopamine is routine in some centres although there is little evidence to suggest any significant benefit on renal function. Appropriate crystalloid is infused to maintain hydration. Hypoglycaemia is uncommon in elective cases but may be severe in acute liver failure. Dextrose 10 - 50% may be used in such cases. Though elective cases may have a hyperdynamic circulation, this is seldom a problem before reperfusion. By contrast, patients with acute liver failure generally require vasopressors such as noradrenaline or angiotensin II, particularly where cardiac reserve is poor.

26

IV Access and Rapid Infusion System Large bore cannulae are required for rapid transfusion of blood and fluids. This may be achieved by the use of 2 separate 8.5 Fr. catheters sited either in the antecubital fossa or internal jugular vein. Alternatively an 18 Fr. percutaneous bypass cannula may be inserted into an internal jugular vein. This serves both as venous access and the return line for veno-venous bypass, avoiding the need for axillary vein cut down for bypass return. The rapid infusion system (Hemonetics, Midvale, Utah) allows transfusion of warmed fluids (37 - 38oC) at rates of up to 1500mls /min20. Alternatives include the “Level-1” infusion system. Fresh frozen plasma (FFP), colloid and electrolyte solutions can all be used with packed cells to achieve a haematocrit in the region of 0.28 to 0.32. FFP infusion is started early either as a separate infusion or mixed in the rapid infusion reservoir. The rate of FFP infusion is adjusted according to TEG appearance. Specific considerations Each of the three phases of liver transplantation poses particular problems for the anaesthetist. Phase I The abdomen is opened, usually via an inverse T or "Mercedes" incision. Dissection may be made difficult by varices in the abdominal wall, and by adhesions within the abdominal cavity. Steady and at times rapid haemorrhage may ensue. Cardiovascular effects of ascitic decompresion may be marked. Hypokalaemia may necessitate potassium chloride infusion during phase I. Hypocalcaemia develops because of citrate accumulation, and is treated by slow intravenous infusion. Ionised calcium should be maintained between 0.84-1.4 mmol/l to prevent coagulopathy and cardiac depression. Hypomagnesaemia also occurs21. A cell saver is used to scavenge lost red blood cells and prepare them for re-infusion22. Contraindications include malignant aetiology and infected ascitic fluid. Scavenged blood is heparinised as it enters a reservoir, washed with normal saline, centrifuged and returned via the rapid infusion system. Bypass (where used) Cross clamping of the inferior vena cave, portal vein and hepatic artery is associated with: reduced venous return (up to 50%), necessitating large volume infusion prior to and during caval cross-clamping; possible need for vasopressors; high systemic venous pressure resulting in reduced renal and splanchnic blood flow; increased portal pressure, resulting in gut oedema and increase bleeding; and relative volume overload on caval declamping. Early use of bypass circuits by Starlz in 1984 resulted in improved cardiovascular stabiltiy but required systemic heparinisation and was abandoned due to bleeding. The introduction of heparin bonded circuits has resulted in improved outcome. Many centres now routinely use veno-venous bypass. Blood is taken from the femoral vein and returned to the axillary or internal jugular vein. A flow 20% of cardiac output is readily achieved through the bypass circuit. Adding portal bypass increases flow to about 40% of cardiac output (3 - 4 litres/min). Haemodynamic stability is improved, and the need for acute volume loading reduced by around 3l23. Although complications of bypass are uncommon there is potential for air embolism, clot formation, fragmentation of red cells, and failure of the system. Phase II During the anhepatic phase (phase II), physiological derangement progresses. There is no production of clotting factors, worsening coagulopathy with fibrinogen deficiency, and derangement of the t-PA/ alpha-2 antiplasmin ratio with the onset of fibrinolysis24 25. Anti-thrombin III levels fall. There is absent citrate or lactate metabolism, reduced gluconeogenesis and glucose uptake, and a worsening acidosis. Protection from gastro-intestinal tract antigens is lost.

27

In addition there is the development of thrombocytopenia due to massive transfusions and platelet consumption often in patients who are already thrombocytopenic. Fibrinolysis can be a major problem which may be prevented or attenuated by antifibrinolytic agents, such as aprotinin26 or tranexamic acid27. The role of other serine esterase inhibitors (such as alpha-2 antiplasmin) are currently being investigated28. Methyl prednisolone 10mg/kg is given as a slow intravenous infusion prior to reperfusion, and protects against graft dysfunction and renal injury29. Phase III Reperfusion of 1500 grams of preserved tissue releases a massive load of cold, hyperkalaemic, acidotic fluid which goes directly to the heart. Immediately before reperfusion 10 mmols calcium chloride may be given to protect the myocardium from the potassium surge. Acute reperfusion serum K+ levels have been measured at up to 10 mmols/l. If arterial pressure falls excessively, small increments of epinephrine (25-50µg) are given. A rising CVP may contribute to venous congestion in the graft. Post-reperfusion syndrome (PRS) is defined as a reduction in mean arterial pressure of 30% occuring within 5 minutes of graft reperfusion and persisting at least one minute30. This may be cytokine mediated31 and is associated with changes in arachidonic acid metabolism. Activation of complement also plays a major role32. Hypotension associated with a low systemic vascular resistance (SVR) may persist an hour or more. Most patients tolerate this well, but those with pre existing cardiac disease cope less effectively and may require vasopressor or inotropic support. After the initial plasma potassium rise, the functioning graft avidly takes up potassium. It is often necessary to provide 60 - 120 mmol KCl in the hours following reperfusion. Lactate metabolism and production of bicarbonate are restored. Metabolic acidosis begins to correct. Primary non function of the graft (up to 10% of cases) 33 is indicated by failure to correct metabolic acidosis, continuing coagulopathy and absence of bile production. Most cases require urgent re-transplantation. Post operative care At the end of the surgical procedure, most transplant recipients require 6 hours ventilation on the ICU. Some patients are suitable for early on-table extubation34 35. The sickest patients require many days of intensive care management. ICU management of liver transplant recipients follows the same principles as their anaesthesia. Sedation is continued usually with propofol and alfentanil infusions. Muscle relaxation is no longer required except in unusual circumstances. Insulin infusion for tight glycaemic control is likely to be beneficial. K+ supplementation may continue for 24 - 36 hours. Coagulation tests and full blood count guide further transfusion of blood and blood products. The haematocrit is maintained between 0.26 and 0.32. Higher levels are associated with an increased incidence of hepatic artery thrombosis and graft failure36 37. In patients with known hypercoagulable states (Budd- Chiari, Protein C and S deficiency) anticoagulation with heparin is required and haematocrit kept at the lower end of the range. Immunosuppression is commenced early in the postoperative period. Summary Transplantation is now established therapy for end stage liver disease and offers a 1 year survival of 82% with 5 year survival of 72%38. Improvements in technology, surgical and anaesthetic management have overcome many of the early problems associated with liver transplantation. References 1. Starzl TE, Marchioro TL, von Kaulla KN et al. Surgery Gynecology and Obstetrics 1963;117:659-676. 2. Castells A, Salmeron JM, Navasa M et al. Liver transplantation for acute liver failure: analysis of

applicability. Gastroenterology 1993;105(2):532-8. 3. Rakela J, Perkins JD, Gross JB et al. Acute hepatic failure: the emerging role of orthotopic liver

transplantation. Mayo Clin Proc. 1989; 64(4): 424-8. 4. O'Grady JG, Alexander GJM, Hayllar KM, Williams R. Early indicators of prognosis in fulminant hepatic

failure. Gastroenterology 1989;97(2):439-45.

28

5. Rudich S, Busuttil: Transplant Proc 31:523, 1999. 6. O'Grady JG, Schalm SW, Williams R. Acute liver failure: redefining the syndromes. Lancet 1993 (Jul

31);342(8866):273-5. 7. Freeman RB, Harper A, Edwards EB. Excellent Liver Transplant Survival Rates Under the MELD/PELD

System. Transplant Proc. 2005 Mar;37(2):585-8. 8. M. Castro, M.J. Krowka, D.R. Schroeder et al.. Frequency and clinical implications of increased

pulmonary artery pressures in liver transplant patients. Mayo Clin Proc 1996; 71: 543. 9. Edwards ND, Reilly CS. Detection of perioperative myocardial ischaemia. Br J Anaesth 1994; 72(1): 104-

15. 10. Calne RY, Williams R, Rolles K. World J Surg 1986;10:422. 11. E. Mor, L. Jennings, T.A. Gonwa et al.. Surg Gynecol Obstet 1993; 176: 219. 12. Shaw BW Jr, Martin DJ, Marquez JM et al. Venous bypass in clinical liver transplantation. Annals of

Surgery 1984; 200(4): 524-534. 13. L. Grande, A. Rimola, E. Cugat et al.. Effect of venovenous bypass on perioperative renal function in

liver transplantation: Results of a randomized, controlled trial. Hepatology 1996; 23; 1418 – 28. 14. Mallett SV, Cox DJA . Thrombelastography. Brit J Anaesthesia 1992; 69: 307-313. 15. Kang YG, Martin DJ, Marquez JM et al. Intra-operative changes in blood coagulation and

thromboelastographic monitoring in liver transplantation. Anesthesia and Analgesia 1985; 64: 888-896. 16. Russel SH, Freeman JW. Prevention of hypothermia during orthotopic liver transplantation: comparison

of three different intra-operative warming methods In: Slooff MJH and Wierda JMKH, eds. Proceedings of the 7th. Meeting of the Liver Intensive Care Group of Europe. The Hague: CIP Data Koninklijke Bibliotheek;1994:23.

17. Webster NR, Bellamy MC, Sadek SA, Lodge JPA. Right ventricular function in orthotopic liver transplantation: A comparison of two anaesthetic techniques. British Journal of Anaesthesia 1994;72(4):418-421.

18 Bell CF, Hunter JM, Jones RS, Utting JE. Use of atracurium and vecuronium in patients with oesophageal varices. Br J Anaesth 1985;57(2):160-8.

19 Miller RD. Pharmacokinetics of atracurium and other non-depolarizing neuromuscular blocking agents in normal patients and those with renal or hepatic dysfunction. Br J Anaesth 1986;58 Suppl 1:11S-13S.

20 Sassano JJ,. The rapid infusion system. In: Winter PM, Kang YG (eds). Hepatic transplantation: anesthetic and perioperative management. New York, Praeger, 1986; 120-134.

21 Bennett MW, Webster NR, Sadek SA. Alterations in plasma magnesium concentrations during liver transplantation. Transplantation 1993;56(4):859-61.

22 Lindop MJ, Farman JV, Smith MF 1983. The Cambridge-King's College Hospital Series. In: Calne RY (ed) Liver Transplantation. Grune and Stratton, New York. 128-129.

23 Cheema SP, Hughes A, Webster NR, Bellamy MC. Cardiac function during orthotopic liver transplantation with venovenous bypass. Anaesthesia. 1995 Sep;50(9):776-8.

24 Dzik WH, Arkin CF, Jenkins RL, Stump DC. Fibrinolysis during liver transplantation in humans: role of tissue-type plasminogen activator. Blood 1988; 71: 1090-1095.

25 Cheema SP, Webster NR, Dunn F, Bellamy MC. Mediators of fibrinolysis in orthotopic liver transplantation. Clin Transplant. 1996 Feb;10(1 Pt 1):24-7.

26 Mallett SV, Cox D, Burroughs AK, Rolles R. Aprotinin and reduction of blood loss and transfusion requirements in orthotopic liver transplantation. Lancet 1990: ii: 886-887.

27 Yassen KA, Bellamy MC, Sadek SA, Webster NR. Tranexamic acid reduces blood loss during orthotopic liver transplantation. Clinical Transplantation 1993;7:453-458.

28 Bellamy MC, Mullane D, O’Beirne HA. Is tranexamic acid as effective as alpha 2 antiplasmin at correcting coagulopathy in liver transplantation? British Journal of Anaesthesia 1998; 81: 282P.

29 Bellamy MC, Turner S, Dhamarajah S, Bosomworth M. Effect of perioperative steroids on renal function during liver transplantation. British Journal of Anaesthesia 2004; 93(1): 169P.

30 Aggarwal S, Kang Y, Freeman JA et al. Postreperfusion syndrome: cardiovascular collapse following hepatic reperfusion during liver transplantation. Transplant Proc 1987;19 (Suppl. 3):54-55.

31 Bellamy MC, Galley HF, Webster NR. Changes in inflammatory mediators during orthotopic liver transplantation. British Journal of Anaesthesia 1997; 79: 338-341.

32 Bellamy MC, Gedney JA, Buglass H, Gooi JH. Complement membrane attack complex and hemodynamic changes during human orthotopic liver transplantation. Liver Transpl. 2004 Feb;10(2):273-8.

33 Shaw BW, Gordon RD, Iwatsuki S et al. Retransplantation of the liver. Semin Liver Dis 1985;5:394. 34 Mandell MS, Lezotte D, Kam I, Zamudio S. Reduced use of intensive care after liver transplantation:

influence of early extubation. Liver Transpl. 2002 Aug;8(8):676-81. 35 Mandell MS, Lezotte D, Kam I, Zamudio S. Reduced use of intensive care after liver transplantation:

patient attributes that determine early transfer to surgical wards. Liver Transpl. 2002 Aug;8(8):682-7. 36 Buckels JA, Tisone G, Gunson BK, McMaster P. Low haematocrit reduces hepatic artery thrombosis

after liver transplantation. Transplant Proc 1989;21(1 Pt 2):2460-1. 37 Tisone G, Gunson BK, Buckels JA, McMaster P. Raised hematocrit--a contributing factor to hepatic

artery thrombosis following liver transplantation. Transplantation 1988;46(1):162-163. 38 European Liver Transplant Registry 1998 – 2003. http://www.eltr.org.

29

Session 4 - Anaesthetic management Crisis management Prof. Susan Mandell. Professor of Anaesthesiology, University of Colorado, USA INTRAOPERATIVE CRISIS MANAGEMENT HYPERKALEMIA Prevention of renal failure and hyperkalemia (Preoperative) Many of the advanced treatments for ascites also improve renal function and consequently potassium secretion by increasing circulation in the afferent arteriole of the kidney. In preparation for liver transplantation, there are a number of interventions that can reduce intraoperative risk from complications of renal insufficiency and failure(1). These include the placement of a transjugular intrahepatic portosystemic shunt(2) and paracentesis combined with albumin infusions(3). In patients with aggressive hepatorenal syndrome, vasoconstrictors with volume expansion have been used in order to increase glomerular filtration rate. The vasoconstrictors reduce the capacitance of the splanchnic circulation so that volume loading improves circulation to the kidney rather than pooling in the abdomen. Most outcomes have been measured using telripressin, a vasopressin analogue not yet available in the United States(4). Administration of noradrenalin with albumin infusion has also been effective(5). Prevention of hyperkalemia (Intraoperative) Serum potassium can increase to dangerous levels during liver transplantation following the administration of blood products to patients with impaired renal function or those who have received diuretics that block aldosterone receptors. Factors that promote an extracellular shift of potassium including metabolic acidosis and insulin resistance magnify the mechanisms leading to hyperkalemia. Irradiation of red blood cells also increases the risk of hyperkalemia by disruption in the cell membrane that favors the efflux of potassium out of the cell (6). Removing potassium from transfusion products or reducing the patient’s serum potassium are two pre-emptive approaches in intraoperative planning that can reduce the risk of life-threatening hyperkalemia. Some institution order units of red blood cells for liver transplant recipients that have been stored for less than 14 day to minimize the natural extracellular flux of potassium into the surrounding plasma and consequently infuse less potassium. .Other institutions have extended the use of blood scavenging devises by washing all red blood cells prior to transfusion. Most of these devises reduce the concentration of potassium in the blood retrieved from the surgical site by washing the salvaged blood. Limited experience suggests that this is an effective way to prevent hyperkalemia in patients during liver transplant surgery(7). The only theoretical limitation to the use of prewashed cells is the significant reduction in 2, 3 diphosphogyleride levels that occurs(8). Potassium absorption filters, developed for massive transfusion in pediatric patients may be of use during liver transplantation. The filter, made of sodium polystyrene sulfonate has the capacity to absorb 95% of effluxed potassium in three units of irradiated red blood cells with minimal damage to the red cell membrane(9). Limitation of this technique in adult patients with cirrhosis include the limited absorption capacity of each filter (approximately 4 units of 400cc) and the fact that sodium is released in equimolar amounts for the potassium exchanged(9). A standard way to reduce serum potassium concentration is by hemodialysis. Some institutions have chosen to provide dialysis care in the operating room as a means of controlling serum electrolytes and intravascular volume during surgery. Currently there is a limited experience using intraoperative dialysis and the relative risks and benefits compared to other techniques are not known. Two approaches to hemodialysis are intermittent and continuous renal replacement therapy. The advantages of intermittent hemodialysis are the rapid control of both intravascular volume and electrolytes, while continuous renal replacement therapy is hemodynamically better tolerated by some critically ill patients(10). In outcome studies, no advantage of one technique over the other has been shown in critically ill patients admitted in the ICU to date(11).

30

Emergency treatment of hyperkalemia In cases where hyperkalemia is unanticipated and arises in the operating room, emergency treatment is necessary to prevent cardiac arrest. The administration of insulin and glucose is a standard approach to decreases serum potassium. Insulin regulates transmembrane flux of potassium and thus is capable of shifting extracellular potassium into the intracellular space. This has been shown to be an effective way to control sudden and unexpected hyperkalemia during liver transplantation despite the presence of insulin resistance in liver disease(12,13). The use of calcium salts in the acute treatment of hyperkalemia-induced heart block is controversial as there are no controlled trials. Experience is primarily derived from cardiac surgery and emergency medicine, both of which have never shown a survival benefit even with temporary normalization of the electrocardiogram(14). There is however suggestion of a survival benefit using calcium infusions as opposed to intermittent boluses(15). If calcium salts are used to treat hyperkalemia during liver transplantation, the chloride preparation offers significantly better bioavailability as the gluconate preparation needs to be activated by hepatic metabolism (16). Administration of sodium bicarbonate should decrease serum potassium by decreasing proton concentration in the extracellular fluid. This would activate a sodium dependent exchange to move hydrogen ions out and sodium into the cell. Subsequently sodium is actively exchanged for potassium. A significant reduction in potassium has not been observed in the absence of acidosis(17). The use of sodium bicarbonate is however associated with complications including hypernatremia, volume expansion and a fall ionized serum calcium. Increased serum potassium is commonly treated with cation-exchange resin, however this treatment modality works too slowly and is impractical as a intervention for the acute hyperkalemia that arises during liver transplantation. POSTREPERFUSION SYNDROME A fall in the mean arterial blood pressure of greater than 30% that lasts for more than one minute following reperfusion of the liver graft is defined as postreperfusion syndrome (PRS) (18). There has been considerable interest in uncovering the etiology of PRS with the intent of predicting its occurrence and consequently preventing or optimizing its treatment since this life-threatening syndrome can affect up to 30% of transplant recipients(19). The hemodynamic response to occlusion of the inferior vena cava has the best predictive value for the occurrence of PRS(19). Investigators hypothesize that patients who suffer from PRS are unable to mount a baroreceptor mediated sympathetic adrenergic response to hemodynamic instability, having less vascular adaptability(19). Failure of baroreceptor-mediated reflexes likely results from autonomic impairment in patients with greater severity of liver disease(20). Studies using extracorporeal hepatic support systems showed that there was a causal relationship between the condition of donor hepatocytes and hemodynamic stability of the patient receiving treatment, supporting a hypothesis that donor liver function plays a role in expression of PRS(21). This may explain some of the random cases of PRS that are not predicted by the response to occlusion of the vena cava and may also account for cases of severe and protracted PRS in sicker patients. Some studies suggest that preservation of the inferior vena cava during the anhepatic stage is associated with a decreased incidence of PRS(22). There is insufficient evidence to determine the effects of the piggyback technique on the prevalence of PRS; further the mechanism by which this surgical technique would specifically ameliorate PRS is unclear. Veno-venous bypass is another means of facilitating venous return to the heart during the anhepatic stage; however unlike the piggyback technique there is a general consensus that veno-venous bypass does not influence the prevalence or severity of PRS(23). The literature suggests that the propensity to develop PRS can be predicted by the severity of patient illness and the function of the donor graft. A single center study found that the prophylactic use of Aprotinin ameliorates PRS(24). This observation is supported by evidence which shows an increased conversion of high molecular weight kininogen to the potent

31

vasodilator, kalikrein during reperfusion(25). Otherwise the treatment of PRS is aimed at supporting perfusion with the use of vasoconstrictors to treat the drop in systemic vascular resistance. Case reports or small studies have found chronotropic agents such as atropine, the α-adrenergic agent phenylephrine and combined α and ß-adrenergic agents to all have some benefit in treating PRS. The administration of volume to patients with PRS should be guided by the cardiac filling pressures, recognizing that resolution of PRS will lead to an increase in systemic vascular resistance and consequently relative intravascular volume. PORTO-PULMONARY HYPERTENSION What is portopulmonary hypertension? Previously, pulmonary artery hypertension was classified as primary or secondary depending on the presence or absence of identifiable risk factors and co morbid conditions(26). The new Evian classification system groups pulmonary arterial hypertension according to similarities in pathophysiology, clinical presentation and therapeutic options(27). In the Evian classification, PPHTN is now included in a diverse collection of causes of pulmonary artery hypertension that share pathological similarities in the small pulmonary muscular arterioles(27). Although the Evian classification does not specify, all disease categories are limited to chronic and progressive conditions, excluding acute causes such as adult respiratory disease syndrome. The unique hyperdynamic features of liver disease caused investigators to further stipulate that a diagnosis of PPHTN be accompanied by the presence of portal hypertension and elevation in pulmonary vascular resistance(28). Current outcomes data Many have questioned whether the cutoff values used to define the wide category of pulmonary artery hypertension diseases are valid to use as hemodynamic markers for PPHTN considering the effects of the hyperdynamic circulation.(29). Consequently, there is no current consensus on the definition of PPHTN. Further, there are multiple outcome reports that include patients in who the diagnosis of PPHTN is unclear. Thus, the diversity in reported outcomes in patients diagnosed with PPHTN may result in part from the variability in diagnostic accuracy. This inaccuracy masks the relative risks and benefit of specific interventions, including liver transplant surgery. Despite these limitations, most would agree that preoperative diagnosis is a key step in planning and thus preventing intraoperative crises. Investigators have used mPAP as a marker of severity of disease and implied surgical risk, reporting a zero, fifty and one hundred percent perioperative mortality rates for PPHTN when the mPAP is < 35 mm Hg,, 35-50 mm Hg and > 50 mm Hg respectively(30, 31). Other reports seem to confirm an increase mortality risk when the mPAP exceeds 35 mm Hg (32), but there is no agreement on the inherent perioperative risks to patients when the mPAP exceeds 35 mm Hg. Rather, the results of a multi center database of 66 patients showed no significant difference in mPAP between the 23 patients who survived liver transplantation (44.82 ± 13.65 mm Hg) as opposed to the 13 those who did not (44.08 ± 8.87 mm Hg) (32). Investigators cited the severity of pulmonary hypertension as the reason for denying 30 of 66 patients transplantation. These patients had a slightly but significantly higher mPAP (52.93 ± 11.18 mm Hg), showing that there is a considerable difference in opinion of who physicians think are candidates for transplantation. That, 7 of 11 patients (63%) with a mPAP > 50 mm Hg survived liver transplantation and were discharged home, supported previous findings that a higher mPAP does not imply imminent and unacceptable risk for transplantation(33). The lack of concordance in outcome studies shows that there is not a simple relationship between the single hemodynamic variable mPAP and mortality risk in a population of patients with PPHTN across institutional boarders. Either mPAP is simply not a surrogate marker for severity of disease or it is a relatively insensitive tool that must be modified by the inclusion of other hemodynamic variables or patient characteristics. Further, the influence of style or choice of anesthesia management has never been justly considered for its potential role in perioperative morbidity and mortality of PPHTN patients. This observation underscores the fact that to date, there is no consensus on what constitutes best practice for uncomplicated liver transplant patients little lone patients as complex as those with PPHTN.

32

It is reasonable to hypothesize that PPHTN should improve once portosystemic shunting is eliminated. There are reports of improvement(34) persistence(35) and de novo development of PPHTN following liver transplantation(36-37). It is unclear why some patients improve while others do not and why transplant patients would develop pulmonary hypertension, as a new diagnosis following transplantation. Selecting transplant candidates consequently becomes more complicated by the fact that investigators have not yet uncovered any patient variables that could predict PPHTN outcome in those who survive liver transplantation. The implementation of Model for End Stage Liver Disease in the United States has caused physicians to become very conservative when choosing patients with PPHTN; predicated on the principle of maximizing transplant benefit while reducing deaths on the waiting list. Thus, even an estimated survival rate of 60% is considered insufficient by many institutions to undertake transplantation in patients who have a mPAP > 35 mm Hg. Institutions are therefore seeing fewer complications and accompanying intraoperative crises related to PPHTN because they are choosing patients whose perioperative mortality risk is no greater than other transplant patients. Many have in fact questioned whether patients with a mPAP between 25 to 35 mm Hg have PPHTN at all and whether the 35 mm Hg cutoff simply reflects a higher normal value in patients with hyperdynamic circulation due to liver disease. Therapeutic interventions There are few treatment options for patients with PPHTN. Most treatment has focused on the systematic reduction of pulmonary artery pressures and vascular impedance by chronic administration of vasodilators. Investigators have shown that continuous intravenous epoprostenol infusion reduces pulmonary artery pressure and resistance, improving right ventricular function in PPHTN patients over six months of chronic therapy(38). Some investigators suggest all patients with PPHTN should receive preoperative epoprostenol therapy, hypothesizing that treatment will improve transplant survival(39, 40). However, only 21% of the patients enrolled in the multicenter database were treated preoperatively with epoprostenol(32). That survival in the 4 of the 5 treated patients did not correlate with any other hemodynamic variable, suggests there may be an independent positive effect of the drug. Events associated with clamping and unclamping of the vena cava and portal vein can cause significant hemodynamic instability and place the PPHTN patient at risk of right heart failure. Abrupt increases in pulmonary vascular resistance(21) combined with sudden expansion of central filling volumes can precipitate acute right ventricular dilation with obliteration of the left ventricular cavity and dynamic occlusion of the right coronary system. This leads to right ventricular ischemia, a critical drop in left ventricular stroke volume and death. Treatment is aimed at improving coronary artery perfusion and reducing impedance to right ventricular ejection. Nitric oxide has been effective in reducing pulmonary vascular resistance in some patients with PPHTN. The efficacy of nitric oxide in treating precipitous increases in pulmonary resistance during liver transplantation is unknown. The minimal adverse effects associated with treatment makes nitric oxide a reasonable choice. Other treatment aims to increase coronary perfusion pressure through the use of vasoconstrictors and to support myocardial contractility with the use of inotropes that reduce pulmonary vascular resistance such as the phophodiesterase-3 inhibitor milrinone. References 1. Moore K, Wong F, Gines P, et al, The management of ascites in cirrhosis: Report on the consensus

conference of the international ascites club. Hepatology 2003; 38:258-66. 2. Rosado B, Kamath P, Transjugular intrahepatic portosystemic shunts: an update. Liver Transpl.

2003;9:207-17. 3. Saadeh S, Davis G,. Management of ascites in patients with end-stage liver disease. Rev Gastroenterol

Disord. 2004;4:175-85. 4. Ortega R, Gines P, Uriz J, et al, Terlipressin therapy with and without albumin for patients with

hepatorenal syndrome. Results of a prospective, non-randomized study. Hepatology 2002; 36:941-48. 5. Duvoux C, Zanditenas D, Hezode C, et al, Effects of noradrenalin and albumin in patients with type 1

hepatorenal syndrome. Hepatology 2002; 36:374-80. 6. Thorp J, Plapp F, Cohen G, et al, Hyperkalemia after irradiation of packed red blood cells: possible

effects with intravascular fetal transfusion. Am J Obstet Gynecol 1990;163;607-09.

33

7. Knichwitz G, Zahl M, Van Aken H, et al, Intraoperative washing of long-stored packed red blood cells by using an autotransfusion devise prevents hyperkalemia. Anesth Analg 2002;95:324-25.

8. Weisbach V, Riego W, Strasser E, et al, The in vitro quality of washed, prestorage leucocyte-depleted red blood cell concentrates. Vox Sang 2004;87:19-26.

9. Inaba S, Nibu K, Tankano H, et al, Potassium-absorption filter for RBC transfusion: a Phase III clinical trial. Transfusion 200;40:1469-74.

10. Waldrop J, Ciraulo D, Milner T, et al, A comparison of continuous renal replacement therapy to intermittent dialysis in the management of renal insufficiency in the acutely ill surgical patient. Am Surg 2005;71:36-39.

11. Kellum J, Angus D, Johnson J, et al, Continuous versus intermittent renal replacement therapy: a meta-analysis. Intensive Care Med 2002 ;28 :29-37.

12. Li Q, Zhou M, Wang Y, et al, Effect of insulin on hyperkalemia during anhepatic stage of liver transplantation. World J Gastroenterol 2004;15:2427-29.

13. Takats D, Using calcium salts for hyperkalemia. Nephrol Dial Transplant 2004;19:1333-34. 14. Meroney W, Herndon R, The management of acute renal insufficiency. JAMA 1954;155:877-83. 15. Davey M, Caldicott D. Calcium salts in management of hyperkalaemia. Emerg Med J. 2002;19:92-3. 16. Blumberg A, Weidmann P, Ferrari P, et al, Effect of prolonged bicarbonate administration on plasma

potassium in terminal renal failure. Kidney Int 1992;41:369-74. 17. Aggarwal S, Kang Y, Freeman J, et al, Postreperfusion syndrome: cardiovascular collapse following

hepatic reperfusion during liver transplantation. Transpl Proc 1887;19:54-55. 18. Garutti Martinez I, Olmedilia L, Perez-Pena J, et al, Response to clamping of the inferior vena cava as a

factor for predicting postreperfusion syndrome during liver transplantation. Anesth Analg 1997;84:259-59.

19. Aggarwal S, Kang Y, Freeman J, et al Postreperfusion syndrome: hypotension after reperfusion of the transplanted liver. J Crit Care 1993;8:154-60.

20. Perez-Pena J, Rincon D, Banares R, et al, Autonomic neuropathy is associated with hemodynamic instability during human liver transplantation. Transplant Proc. 2003;35:1866-68.

21. Ricciardi R, Foley D, Quarfordt S, et al, Donor hepatic function. A factor in postreperfusion syndrome. J Gastrointest Surg 2002;6:248-54.

22. Acosta F, Rodriguez M, Sansnao T, et al, Influence of surgical technique on postreperfusion syndrome during liver transplantation. Transplant Proc 1999 ; 31 :2380-81.

23. Jugan E, Albaladejo P, Jayais P, et al, The failure of venovenous bypass to prevent graft liver postreperfusion syndrome. Transplantation 1992;54:81-84.

24. Molenaar I, Begliomini B, Martinelli G, et al, Reduced need for vasopressor in patients receiving Aprotinin during orthotopic liver transplantation. Anesthesiology 2001;94:433-38.

25. Scholz T, Backman L, Mathisen O, et al, Activation of the plasma contact system and hemodynamic changes after graft revascularization in liver transplantation. Transplantation 1995;60:36-40.

26. Rich S, Dantzer D, Ayres S, et al, Primary pulmonary hypertension. A national prospective study. Ann Intern Med 1987;107:216-28.

27. Simonneau G, Galie N, Rubin L, et al, Clinical classification of pulmonary hypertension. J Am Coll Cardiol 2004 ; 43 : 5S-12S.

28. Mandell M. Krowka M. National database for hepatopulmonary syndrome and portopulmonary hypertension in liver transplant candidates/recipients. Anesthesiology 1997;87:450-51.

29. Mandell M, Critical care issues: Portopulmonary hypertension. Liver Transpl 200;6:S36-S43. 30. Krowka MJ, Plevak DJ, Findlay JY, Rosen CB, Wiesner RH, Krom RA. Pulmonary hemodynamics and

perioperative cardiopulmonary-related mortality in patients with portopulmonary hypertension undergoing liver transplantation. Liver Transpl. 2000;6:443-50.

31. Ramsay M, Simpson B, Nguyen A, Ramsay K, East C, Klintmalm G Severe pulmonary hypertension in liver transplant candidates. Liver Transpl Surg. 1997;3:494-500.

32. Krowka MJ, Mandell MS, Ramsay MA, Kawut SM, Fallon MB, Manzarbeitia C, et al. Hepatopulmonary syndrome and portopulmonary hypertension: a report of the multicenter liver transplant database. Liver Transpl. 2004;10:174-82.

33. Starkel P, Vera A, Gunson B, Mutimer D Outcome of liver transplantation for patients with pulmonary hypertension. Liver Transpl. 2002 ;8:382-88.

34. Koneru B, Ahmed S, Weisse A, Grant G, McKim K. Resolution of pulmonary hypertension of cirrhosis after liver transplantation. Transplantation. 1994 ;58:1133-35.

35. Rafanan A, Maurer J, Mehta A, Schilz R. Progressive portopulmonary hypertension after liver transplantation treated with epoprostenol. Chest. 2000;118:1497-1500.

36. Koneru B, Fisher A, Wilson D, Klein K, delaTorre A, Seguel J. De novo diagnosis of portopulmonary hypertension following liver transplantation. Am J Transplant. 2002;2:883-86.

37. Mandell M, Groves B, Duke J. Progressive plexogenic pulmonary hypertension following liver transplantation. Transplantation. 1995;59:1488-90.

38. Kuo P, Johnson L, Plotkin J, Howell C, Bartlett S, Rubin L. Continuous intravenous infusion of epoprostenol for the treatment of portopulmonary hypertension. Transplantation. 1997;63:604-6.

39. Plotkin J, Kuo P, Rubin L, Gaine S, Howell C, Laurin J, et al. Successful use of chronic epoprostenol as a bridge to liver transplantation in severe portopulmonary hypertension. Transplantation. 1998;65:457-59.

40. Krowka M. Editorial: Pulmonary hypertension, (high) risk of orthotopic liver transplantation, and some lessons from "primary" pulmonary hypertension. Liver Transpl. 2002 ;8:389-90.

34

Session 4 - Anaesthetic management Coagulation Dr Sue Mallett. Consultant Anaesthetist, Royal Free Hospital, London It is well established that there is a direct relationship between transfusion requirements and patient outcome.1 Transfusion of more than 10 units RBC significantly increases postoperative morbidity and mortality in liver graft recipients.2 During the early 1980’s, mean transfusion was of the order of 40 units per case, with transfusion of over 100 units not uncommon. Improvements in operative management, surgical techniques and graft preservation over the last two decades have widely contributed to a significant reduction in transfusion, to the extent that by 2004, the average transfusion during OLT was reported as 3-4 units, with up to one third of patients receiving no RBC transfusion at all.3 However, blood losses remain highly variable and massive transfusion remains a problem in a proportion of OLT cases. Coagulopathy is still a significant contributory factor towards the total amount of blood lost and components transfused during OLT. Liver transplantation & coagulopathy Multiple mechanisms underlie the coagulation abnormalities encountered in OLT. Pre-existing disorders of coagulation: Both the aetiology of the liver disease and its severity are important factors in determining the extent and nature of pre-existing coagulation disorders. The liver is the major site of synthesis of most blood coagulation factors and inhibitory proteins and also plays a central role in the clearance of activated coagulation factors. Hepatocellular, as opposed to cholestatic disease, is associated with more severely deranged coagulation.4

Fibrinogen synthesis is abnormal in cirrhosis with dysfibrinogenaemia. Thrombocytopenia related to hypersplenism and impaired platelet function is common in ESLD and cirrhosis. Enhanced fibrinolysis is common in cirrhosis, related to an imbalance between high levels of tPA relative to PAI and /or decreased levels of alpha 2 anti-plasmin. Low grade DIC with markers of increased activation of the coagulation and fibrinolytic system has also been described. Intra-operative disturbances of coagulation: The dissection phase may be associated with heavy blood loss from the transection of large collaterals in portal hypertension or vascularised adhesions. Haemodilution will worsen intrinsic haemostatic deficiencies. During the anhepatic period there is a marked increase in tPA activity due to the lack of hepatic clearance. Some patients with PBC have been shown to develop hypercoagulability during the anhepatic phase. Reperfusion is associated with profound changes in coagulation, much of which is transient and corrects spontaneously with the onset of graft function. A severe “explosive” fibrinolysis has been described in up to 20% of patients following reperfusion. Heparin like activity may contribute to coagulopathy following reperfusion. This is most likely to be due to the release of endogenous heparin/ heparan from the damaged endothelium of the ischaemic graft, as it appears to be more profound and of longer duration in patients with marginally functional grafts. Predicting bleeding Many studies have attempted to identify the factors associated with high blood loss (HBL) during OLT. Multivariant analysis of patients with HBL (> 12 units RBC) disclosed three associated independent variables: low Hb (<10g/dl), High FDP concentrations (>24mmol/l) and previous abdominal surgery.5 When combined these resulted in a high specificity (98%) but low sensitivity. Massicotte et al. found the starting INR value to be the most sensitive variable to predict transfusion, but it had a low specificity. They identified the surgeon’s and anaesthetist’s experience and attitude as highly relevant factors in determining transfusion requirements.3 Clotting data were not independent predictors of HBL in many studies.6 It is well known that blood losses and transfusion requirements vary considerably from centre to centre, however they may not be directly comparable for a number of reasons, including non-homogenous patient selection, different surgical conditions and different intra-operative management. 7

Coagulation monitoring The aim of perioperative coagulation monitoring is to detect clinically significant coagulopathy in order to guide therapy. Conventional coagulation tests: (PT, aPTT, TT, Fibrinogen, platelet count) have several disadvantages: Firstly, they are static tests, performed on citrated plasma, looking at single end

35

points in the coagulation process. Secondly, turn around times (in the absence of point of care testing) lead to long delays in result reporting, resulting in information that may be of no relevance to the current situation. Finally, they give no information about the stability of the clot or its strength, both of which are highly relevant to the assessment of perioperative haemostasis. Thromboelastography (TEG): The uniqueness and severity of the coagulation dysfunction during OLT requires a method of monitoring that provides a comprehensive overview of the interaction of the various cellular and protein elements involved in producing clot. The TEG is a whole blood visco-elastic test of clot tensile strength. It enables a global analysis of the whole coagulation process from fibrin formation, through to eventual clot retraction and/or lysis, enabling a visual assessment and quantification of overall clot strength and stability. It gives information about coagulation factors, platelet number and function, platelet and fibrinogen cross linkage and any fibrinolytic activity.8 Its particular value in OLT was first described by Kang and his colleagues in Pittsburgh. Another advantage of the technique is the ability to assess the efficacy of pharmacological intervention in vitro prior to administrating drugs to the patient (eg. anti-fibrinolytics). The native TEG is extremely sensitive to the presence of heparin and a straight line TEG at reperfusion is not uncommon. Running simultaneous heparinase and native TEGs will differentiate any heparin effect from other concurrent coagulation defects. The addition of activators such as Kaolin or tissue factor (TF) accelerate the rate at which the TEG trace is generated. The ROTEM is based on similar principles and gives broadly comparable information. Both the TEG and ROTEM now also have the facility to estimate functional fibrinogen. Management of coagulopathy in OLT Evaluating the efficacy of any intervention (blood component support or pharmacological) in reducing blood loss in OLT is extremely difficult and complicated by a number of variables including the heterogeneity of the recipient population, different transfusion protocols (triggers and monitoring techniques) and different outcome variables (blood loss versus transfusion). Blood Component Therapy: Reviewing the literature numerous different guidelines have been developed with different transfusion thresholds and different blood component administration schemes, however, the appropriateness of the different regimes been not been evaluated by prospective randomised trials. The minimal platelet count indicating transfusion varies from 30 to 100,000. FFP is given on a variety of premises including clinical grounds, on the basis of maintaining a fixed ratio to the number of RBC transfused, PT time and TEG parameters. Fibrinogen is generally administered when the value is less than 1.0 g/dl. In a recent retrospective review of transfusion practice in 8 different transplant centres significant heterogeneity in transfusion therapy was found.9 Antifibrinolytics: The use of antifibrinolytics has been advocated during OLT because of the hyperfibrinolytic state and resulting coagulopathy that can develop, particularly around the time of graft reperfusion. However, many questions remain, to which there are as yet no definitive answers: that is which is the best antifibrinolytic to use, should they be used prophylactically or for treatment only, in what dose (low versus high) and whether they should be used routinely or only for high risk patient groups. Establishing efficacy is difficult as many of the earlier studies were methodologically flawed (historical controls and not randomised), and most of these studies were not stratified for disease aetiology or severity. 1: Synthetic Antifibrinolytic agents: Epsilon Amino Caproic Acid (EACA) and Tranexamic Acid (TA): inhibit the binding between plasmin and fibrin. The use of EACA to treat fibrinolysis identified on the TEG was first described by Kang’s group in 1983. However they did not recommend routine prophylactic administration because of the possibility of pro-thrombotic complications. In a study by Kaspar et al, it was found that continuous low dose tranexamic acid (2mg/Kg/hr), whilst reducing the incidence of fibrinolysis, did not result in a significant reduction in transfusion requirements. High dose (10mg/Kg/hr) TA was found to reduce blood loss by half. 2: Serine - Protease Inhibitors: Aprotinin is a naturally occurring proteolytic enzyme derived from bovine lung. It is a potent anti-fibrinolytic, the mechanism of which appears to be dose related. At high doses (200 KIU) aprotinin reduces tPA synthesis through its inhibition of kallikrein, but at low dose (50 KIU) its anti-fibrinolytic effect is almost entirely due to plasmin inhibition. The effect of Aprotinin in reducing blood loss in OLT (2x106KIU followed by an infusion of 500,000 KIU/hr) was

36

first reported by Neuhaus et al in 1989. Whilst these findings were confirmed by others, historical control groups were used in all these initial studies. Subsequently, there has been much discussion as to if and when its use is indicated and the optimal dose required. A randomised controlled trial (RCT) of low dose infusion (200,000 KIU/hr) versus control demonstrated reduced FFP usage and a reduced incidence of fibrinolysis, but no difference in RBC use.10 Soilleux et al randomised patients to high or low dose aprotinin regimes but found no difference between the groups in any blood component use.11 Porte et al in a multi-centre trial with 137 patients randomised to receive large or regular dose aprotinin or placebo, showed a significant reduction in intraoperative blood loss in the aprotinin treated patients (60% and 40% respectively).12 Other RCT have not shown similar benefits.13 There are some concerns about adverse effects, including anaphylaxis and the potential (but unproven) risk of increasing thromboembolic events. Although many centres do use aprotinin routinely, there still is a need for further studies to determine which subpopulations of patients would truly benefit from its use. Other drugs that have been shown to have some benefit in reducing blood loss include DDAVP, conjugated oestrogens and Anti-Thrombin III. Activated Recombinant Factor VII: rFVIIa accelerates thrombin generation through TF pathway activation, leading to an enhancement of coagulation at the site of vessel injury, without systemic activation of the coagulation cascades. This occurs after complexing with TF or in interaction with activated platelets. The PT time can be shortened considerably by administration of rFVIIa (80mls/Kg) and this is attributed to the direct activation of factor X by VIIa. TEG analysis following administration of rFVIIa shows that the velocity of clot formation normalises (R and K times shorten) but also that there is a significant increase in the α angle, indicating that it not only affects the speed of clot formation but also the tensile properties of the clot. In a pilot study of rFVIIa in OLT (single dose of 80 mcg/Kg, 10 minutes prior to start of surgery) it was found that transfusion requirements were lower in the study patients than in the matched controls ( RBC 3 (0-5) versus 9 (4-40) p=0.002, FFP 1 (0-7) versus 8 (2-35) p=0.004. Blood loss was 3.5l (1.4 -5.3) versus 9.8l (3.7 -35) p= 0.004.14 One of the six study patients developed HAT on day 1 post-operation. Plasma levels of tissue activated thrombin (TAT) and coagulation factors 1+2 increased strongly after reperfusion in study patients in contrast with controls, indicating that more thrombin was being formed. Pavese et al 15 reported its use in 4 patients with fulminant hepatic failure (FHF). In all patients the coagulation defect was corrected by rFVIIa but thrombotic complications were reported in 2 of the 4 patients (myocardial infarction (MI) and portal vein thrombosis) and the implication that rFVIIa was involved cannot be excluded. Similar complications have been reported in a multicentre trial of its use in liver resection (2 cases of pulmonary emboli and 2 MI versus non in placebo group). However, the numbers involved were too small to determine if this occurred by chance or as a result of treatment. Clearly, further studies are required to define its safety in this setting and also to determine whether its place is as a prophylactic intervention or whether it should be limited to emergency rescue therapy in cases of uncontrolled blood loss during OLT. References 1 Massicotte L, Sassine MP, Lenis S et al. Survival rate changes with transfusion of blood products during

liver transplantation. Can J Anesth 2005;52:148-155. 2 Cacciarelli TV, Keefe EB, Moore DH et al. Effect of intraoperative blood transfusion on patient outcome

in hepatic transplantation. Arch Surg 1999;134:25-9. 3 Massicotte L, Sassine MP, Lenis S, Andre R. Transfusion Predictors in Liver Transplant. Anesth Analg

2004;98:1245-1251. 4 Segal H, Cottam S, Potter D, Hunt BJ. Coagulation and Fibrinolysis in Primary Biliary Cirhosis compared

with other liver disease and during orthotopic liver transplantation. Hepatology 1997;25: 683-688. 5 Steib A, Freys G, Lehmann C, Meyer C, Mahoudeau G. Intraoperative blood losses and transfusion

requirements during adult liver transplantation remain difficult to predict. Can J Anaesth 2001;48: 1075-1079.

6 Findlay JR, Rettke SR. Poor prediction of blood transfusion requirements in adult liver transplantation from preoperative variables. J Clin Anesth 2000;12:319-23.

7 Ozier Y, Albu A. Liver Transplant Surgery and Transfusion. International Anaesthesiology Clinics 2004; 42: 147-162.

8 Kang Y. Transfusion based on clinical coagulation monitoring does reduce haemorrhage during liver transplantation. Liver Transpl Surg 1997;3:655-659.

37

9 Ozier Y, Pessione F, Samain E et al. Institutional variability in transfusion practice for liver transplantation. Anesth Analg 2003;97:671-9.

10 Marcel R, Stegall WC, Sit CT et al: Continuous small-dose aprotinin controls fibrinolysis during orthotopic liver transplantation. Anesth Analg 1996;82:1122-25.

11 Soilleux H,Gillon MC, Mirand A et al. Comparative effects of small & large aprotinin doses on bleeding during Orthotopic Liver Transplantation. Anesth Analg 1995;80: 349-352.

12 Porte RJ, Molenaar IQ, Begliomini B et al. Aprotinin for reducing blood loss and transfusion requirements in orthotopic liver transplantation: A multicentre randomised double-blind study. Lancet 2000;355:1303-1309.

13 Garcia-Huete L, Domenech P, Sabate A et al. The prophylactic effect of aprotinin on intraoperative bleeding in liver transplantation: a randomised clinical study. Hepatology 1997;26:1143-8.

14 Hendriks HGD, Meijer K, DeWolf JThM et al. Reduced transfusion requirements by recombinant factor VIIa in Orthotopic Liver Transplantation, a pilot study. Transplantation 2001;71:402-405.

15 Pavese P, Bonadona A, Beaubien J et al. FVIIa corrects the coagulopathy of fulminant hepatic failure but may be associated with thrombosis: a report of four cases. Can J Anaesth 2005;52:26-29.

38

Session 5 - Acute liver failure Overview Dr Julia Wendon. Senior Lecturer and Honorary Consultant Physician, Institute of Liver Studies, King’s College Hospital, London 1. Acute liver failure In terms of acute liver failure it is important to note the definition, which inherently excludes anyone with pre-existing liver disease. It also requires the development of encephalopathy and is thence sub-divided into hyper-acute, acute and sub-acute as defined by O’Grady. The aetiology of acute liver failure varies depending on the country however, worldwide viral hepatitis remains the commonest cause. Hepatitis B and also A can cause acute liver failure and increasingly hepatitis E is recognisable. A large cohort of patients that is contributed to by so-called sero-negative hepatitis (that is all other investigations are negative) and those caused by drug induced liver failure. Other important, but less common diagnosis to consider are veno-inclusive disease contributing to the possibility of hepatic vein occlusion with a Budd Chiari syndrome, acute Wilson’s disease, fatty liver of pregnancy and pre-eclamptic liver rupture and a malignant infiltration. Regardless of aetiology, the patients show a characteristic pattern of multiple organ failure though the instance of encephalopathy and cerebral oedema is considerably higher in those with hyper-acute and acute liver failure than those with sub-acute who more frequently present with ascites and may be wrongly diagnosed as patients with chronic liver disease. Cardiovascularly, these patients’ characteristically show a hyper-dynamic circulation with elevated cardiac output and relative under filling of their central blood volume. Hypotension in adequate volume resuscitation usually requires therapy with Noradrenaline. In general intensive care, it has been recognised that patients who are requiring Noradrenaline have a high instance of adrenal suppression and the addition of Hydrocortisone 300mg is beneficial. This data has been looked at in patients with acute liver failure and similar data has been shown with respect to the instance of adrenal insufficiency. Respiratory function is frequently compromised by the virtue of encephalopathy resulting in aspiration and subsequent infection; general immune depression resulting in increased instance of respiratory related infection. Complications such as ARDS and acute lung injury are indeed seen. There is a moderate instance of pancreatitis associated with acute liver failure and this should be sought and excluded, intestinal function is frequently preserved and enteral nutrition in these patients should be encouraged from an early time point in their illness. In a proportion of patients who require intubation and ventilation, gastric varices may occur and feeding can then frequently be instituted by insertion of a post-pyloric tube. Standard enteral nutrition should be undertaken and protein restriction is inappropriate. The pattern of liver function test abnormalities is determined by the aetiology of the liver disease and appropriate screening investigations as to aetiology should be undertaken in all patients as appropriate. Coagulopathy is an almost inevitable finding in patients with acute liver failure and this, of importance with respect to prognosis. The prophylactic repletion with coagulation factors has not been shown to affect outcome and therefore should not be routinely undertaken. In some patients however, with significant bleeding and a pattern of DIC and severe thrombocytopenia, coagulation support may be required. Such coagulation support however, should be discussed with transplant hepatologist and transplant surgeons so as not to occlude decisions relating to the decision to transplant. Intra-abdominal hypertension is something that should be considered, as with all critically ill patients and should be regularly measured. Intra-abdominal pressures of >20 are associated with significant organ dysfunction with splenic hypo-perfusion. Renal failure is a common occurrence in patients with acute liver failure recurring in up to 40% of patients. It is normally multi-factorial in aetiology relating to dehydration and a prerenal component in addition to potential drug related nephro-toxicity and a component of hepato-renal failure. Initial therapy should relate to appropriate volume loading. However, in those patients who remain resolutely oliguric, early institution of renal replacement therapy is appropriate. There is no proven benefit for renal dose Dopamine, Frusemide or purported renal protective agents.

39

Coagulation of renal replacement circuits can often be difficult and the standard is frequently that of Epoprostenol at 2.5-5ng/kg/minute rather than systemic anti-coagulation with Heparin. Continuous modes of renal replacement therapy should be undertaken in patients with acute liver failure as intermittent therapies may predispose to neurological and cardiovascular deterioration. Neurological deterioration with the onset of encephalopathy and subsequent deterioration to a Grade IV coma, which may be complicated by cerebral oedema and severe intra-cranial hypertension, is frequently seen. The aetiology of encephalopathy is multi-factorial but at the present time ammonia is thought to be of significance particularly with regard to brain ammonia levels. Other groups, Last et al, have shown that high serum ammonia levels are seen in patients who subsequently coned compared to those who died of septic multiple organ failure or who survived. The management of neurological complications should relate to appropriate intubation and ventilation with protection of the airway in patients who deteriorate to Grade III-IV coma, then the consideration for intra-cranial pressure bolt monitoring. This allows appropriate multi-modallatory monitoring along with and delineation of the presence or absence of auto regulation. Temperature control has been shown to be a benefit in decreasing intra-cranial pressure in a small cohort of patients, and standard therapy for surges in intra-cranial pressure remains that of Mannitol. Appropriate sedation is also important and first line therapy would normally be a mixture of an opiate and Propofol. Stratification of disease process is of significance as is their biochemistry and haematology results with respect to decisions pertaining to transplantation. The most commonly used transplant criteria are either those of O’Grady et al or the Paris criteria. The criteria of O’Grady et al separate patients into acetaminophen induced and non-acetaminophen induced. Acetaminophen transplant criteria are those of a pH of <7.3 after 24 hours post overdose and following appropriate volume resuscitation or the finding within a 24 hour period of an INR >6.5, Grade III-IV encephalopathy and renal failure with creatinine >300. Non-acetaminophen patients would be considered for transplant if they had an isolated INR of >6.5, were acidotic following volume resuscitation, pH of <7.3 or had 3 of the following 5 criteria, that is: aetiology which was drug induced or sero negative hepatitis, INR >3.5, bilirubin >300, jaundiced to encephalopathy time >7 days and age <10 or >40. The Paris criteria use a mixture of encephalopathy with a factor 5 level of <20% being a determinant for transplantation or in patients who are >30 years of age, a factor 5 level of <30%. There have been some studies out of North America suggesting that biopsy of the liver showing >50% necrosis of hepatocytes is a good indicator for a need of a transplantation but it should be noted that this is subjected to sampling error. Similarly a decreasing liver volume, certainly to <1ℓ as assessed by CT, volumetric scanning is indicative of the need for liver transplantation in acute liver failure. We have also recently examined the role of lactate as a prognostic indicator in acute liver failure and have noted that a lactate of >3.5 following volume resuscitation is of significant prognostic importance. 2. Chronic liver failure Patients with chronic liver disease frequently present to the intensive care unit with an episode of decompensation. Diagnostic difficulties can arise separating decompensative chronic liver disease from a sub-acute liver failure. The commonest cause is of decompensation with the development of encephalopathy. This may be precipitated by a variety of insults, such as, surgical or percutaneous shunting for varices, sepsis, dehydration, over-diuresis, electrolyte imbalance, variceal bleeding, hepatocellular carcinoma or alcoholic hepatitis. The cause of encephalopathy should always be sought and appropriately treated and management is dependent on appropriate care of the airway and ensuring that possible precipitants, including that of constipation are appropriately treated. It is exceptionally rare to see cerebral oedema in association with chronic liver disease. The other common precipitant for admission to ITU is that of a variceal bleed and if this is of significant volume or is associated with encephalopathy, it is appropriate to intubate and ventilate the patient prior to an endoscopic procedure being undertaken. Terlipressin initially 2mg 4 hourly has been shown to be efficacious as endoscopy in the management of oesophageal varices (TESTS study), however it has not been shown to be a benefit in gastric varices. The normal practice would be to administer Terlipressin whilst organizing endoscopy. Appropriate management for oesophageal varices and varices at the gastrooesophageal junction would be

40

that of banding therapy and if haemostasis has been achieved, Terlipressin is normally discontinued. Secondary prophylactic therapy with betablockers would be undertaken once the patient was haemodynamically stable. In terms of gastric varices, these are considerably more challenging in terms of therapeutic intervention, they may be glued with a mixture of serona acrylate or may rarely have endo-loops applied and then glued. The development of endoscopic failure, that is haemodynamic instability and bleeding following 2 episodes of endoscopy, should result in a further alternative being considered such as TIPPS shunting. Another common precipitant for decompensation is that of acute alcoholic hepatitis, these patients are usually significantly jaundiced with only a moderate transaminase, often <200. They are often encephalopathic, have significant ascites and have an enlarged tender liver, which is often bright with fat on ultrasound scanning. Following screening for infection and appropriate antibiotic therapy (such patients characteristically show neutrophilia) therapy may be introduced in the form of steroids, normally 30-40mg of Prednisolone and Pentoxifylline may also be added as this has been shown to decrease the incidence of hepato-renal failure in this cohort of patients in a single centre study. Spontaneous bacterial peritonitis is another disease, which frequently results in decompensation and should always be considered. In any patient with liver disease who is deteriorating, a diagnostic ascitic tap should be undertaken. The ascites should be sent for protein, albumin and a white cell count in addition to culture. If the white cell count is >250pmns/mm3 then a diagnosis of bacterial peritonitis has been made and appropriate antibiotics should be started. Renal failure is a frequent occurrence in patients who are deteriorating, characteristically this is hepato-renal failure but other aetiology should also be sought, such as that of IgA nephropathy, acute tubular necrosis, glomerula nephritis associated with viral disease and tubular disorders. The development of hepato-renal failure can be divided into Type I and Type II, Type I being more rapidly progressive with a doubling of creatinine over 2 weeks. Again as previously, appropriate volume loading is paramount and recent studies have shown that Terlipressin may be of significant benefit in this cohort of patients. Similarly one small study has suggested that TIPPS may be of benefit but this was only undertaken in a small cohort of patients, all of whom had a Child’s Pugh score of <12. Patients with decompensated chronic liver disease require aggressive initial support in intensive care. The cohort with variceal bleeding can be expected to do well, the development of multiple organ failure however, is associated with a poor outcome and recent studies from Germany have looked at patients with a SOFA score of >9 at 24 hours and shown a mortality in excess of 90%. It is of note that the development of extracorporeal liver support may indeed change this poor outcome in an appropriate sub-group of patients.

41

Session 5 - Acute liver failure Neurological complications of ALF Dr Alistair Lee. Consultant in Anaesthesia and Critical Care, Royal Infirmary, Edinburgh

42

Session 5 - Acute liver failure Supporting other organs Prof. Michael Gropper. Professor of Anaesthesia and Physiology, and Director of Critical Care Medicine, University of California Medical Centre, USA Patients with acute liver failure (ALF) present significant challenges in management, both pre- and postoperatively. This review will address the specific management of cardiovascular, respiratory, and renal dysfunction in patients with FHF. Cardiovascular management The most common cardiovascular complication in patients with ALF is shock. As liver failure results in hyper-dynamic, distributive shock, it can be difficult to identify those patients with superimposed infection and sepsis syndrome, in addition to ALF. Patients with both disorders will present with a hemodynamic profile characterized by normal to high cardiac output, low systemic vascular resistance, and low to normal preload. The etiology of this abnormality is not clear, but likely involves a number of factors, including endotoxemia, abnormal nitric oxide (NO) metabolism, and elevated circulating levels of tumor necrosis factor (TNF). Management of shock in ALF is based on establishing adequate preload, and then titrating vasopressors as needed to ensure organ perfusion. The use of pulmonary artery catheters (PAC) remains controversial. Although often used, there is no evidence of improved outcomes(1, 2), and in at least one study, was associated with increased mortality(3). Placement of a central venous pressure catheter (CVP) should provide adequate information to optimize preload, and in those patients where preload is uncertain, transthoracic echocardiography can demonstrate ventricular filling. Patients with ALF should have an arterial line placed, both for blood pressure monitoring and for measurement of acid/base status and oxygenation. Optimization of blood pressure can be achieved with a number of inotropes and vasopressors, both catecholamine and non-catecholamine. The following table lists commonly used vasopressors, along with common dosing. All should be delivered via a CVP catheter due to the risk of extravasation. Agent Typical Dosing Dopamine 3-25 mcg/kg/min Epinephrine 1-10 mcg/min Norepinephrine 1-30 mcg/min Phenylephrine 40-200 mcg/min Vasopressin 0.01-0.04 units/min

There is little evidence to determine the best pressor for patients in shock. Almost all the evidence currently available was generated in patients with septic shock. Martin et al demonstrated that norepinephrine was superior to dopamine for resuscitation of patients with shock (4). Recently, vasopressin and terlipressin have been widely used as vasopressors in shock. Vasopressin causes vasoconstriction in a catecholamine receptor independent manner, and retains efficacy in the setting of severe acidosis. However, there are no data suggesting a survival benefit with the usage of vasopressin. Determining adequate oxygen delivery is controversial. Perhaps the best measure of oxygen delivery is end-organ function. Measurement of mental status, urine output, and cardiac function are reliable measures of oxygen delivery. Recently, measurement of mixed venous oxygen saturation (SvO2) and central venous oxygen saturation (ScvO2) have been used to measure adequacy of oxygen delivery, especially in patients with severe sepsis. Rivers et al used measurement of ScvO2 as part of a strategy of early goal directed therapy for patients with severe sepsis (5). Using this technique of early resuscitation, Rivers was able to demonstrate a substantial reduction in in-hospital mortality. Although this method has not been confirmed in other studies, it is likely that early, goal-oriented therapy will result in improved clinical outcomes. At UCSF, we have instituted a sepsis resuscitation team, that focuses on early identification of and management of patients with severe sepsis.

43

Management of acute renal failure Acute renal failure (ARF), and in particular hepatorenal syndrome, are common consequences of liver failure. Although there are effective mechanical means to support renal function, the attributable mortality of ARF is actually greater than that for acute respiratory failure, and is as high as 50%. Both intermittent hemodialysis (IHD) and continuous renal replacement therapy (CRRT) are used to support these patients. Acute renal failure is defined as an increase in creatinine by 0.5 mg/dl occurring over two weeks or less(6). Before attributing the cause of ARF to ALF, the clinician should rule out two common causes of ARF: prerenal azotemia and postrenal obstruction. Patients with ALF and ascites are particularly susceptible to prerenal azotemia, as these patients have circulatory abnormalities such as renal artery vasoconstriction, hypotension, and decreased renal blood flow. Acute tubular necrosis (ATN) can occur from renal ischemia, or from toxicity from radiocontrast agents, aminoglycoside antibiotics, cyclosporine, tacrolimus, and other pharmacologic agents. Determining the exact cause of ARF in patients with ALF can be difficult, but laboratory testing is essential. Table 2 shows common urine indices in ARF. Condition Urine Osmolality Sodium Concentration FeNa Prerenal > 500 mOsm/kg < 20 mmol/L < 1% ATN < 350 mOsm/kg > 40 mmol/L > 1% Obstructive < 350 mOsm/kg > 40 mmol/L > 1% Hepatorenal < 1%

Hepatorenal syndrome (HRS) is renal failure occurring in the setting of liver failure, and is a type of prerenal failure. The accepted definition of hepatorenal failure is defined by the International Ascites Club(7), and is as follows:

1. Severe cirrhosis 2. Serum creatinine > 1.5 mg/dL 3. Absence of shock, infection, or treatment with nephrotoxins 4. Absence of gastrointestinal fluid loss 5. No improvement in serum creatinine despite optimization of intravascular volume. 6. Proteinuria below 0.5 mg/d 7. No evidence of renal obstructive disease

Management of HRS begins with treating other causes of renal failure. If HRS is suspected, then the focus is on optimizing renal perfusion with adequate preload, and if necessary, vasoconstrictors. One retrospective study suggested that terlipressin (a vasopressin analog) is useful for treating HRS(8). Low dose dopamine has no role in treatment of HRS. The multicenter ANZICS study demonstrated in 328 patients that dopamine had no efficacy in changing the course of ARF. (9). A recent meta analysis confirmed that dopamine can increase urine output, but has no effect on mortality or resolution of ARF(10). This data is in spite of the fact that in general, patients with nonoliguric renal failure have a better prognosis than those with oliguric renal failure(11). Renal replacement therapy for ARF is indicated for patients with signs and symptoms of volume overload, uremia, hyperkalemia, or acidosis. Intermittent hemodialysis (IHD) is effective, but is dependent on adequate blood pressure to allow volume removal. The dose of IHD is controversial, but Schiffl demonstrated in a prospective study that survival was higher with daily IHD when compared to alternate day HD(12). Many patients with ALF and ARF are treated with continuous renal replacement therapy (CRRT). Advantages of CRRT include the ability to remove large volumes of fluid in hemodynamically unstable patients due to the continuous nature of the dialysis. A disadvantage is the high nursing staffing ratio required for CRRT, as patients must have one-to-one nursing. There are no prospective, randomized trials of CRRT versus IHD for management of ARF. We currently are participating in such a trial in the US.

44

There are a number of modes of CRRT, which are summarized in the following table: IHD SLEDD SCUF CVVH CVVHD(F*) Blood flow (ml/min) 250-400 100-200 <100 200-300 100-200 Filtrate (L/day) 0-4 0-4 0-4 24-96 24-48* Clearance mechanism Diffusion Diffusion Convection Convection Both* Urea clearance (ml/min) 180-240 75-90 1.7 17-67 30-60 Duration (hours) 3-4 8-12 >24 >24 >24

IHD=intermittent hemodialysis, SLEDD=slow low efficiency dialysis, SCUF=slow continuous ultrafiltration, CVVH=continuous venovenous hemofiltration, CVVHD=continuous venovenous hemodialysis (hemodiafiltration). Modified from Mehta, Semin Dial, 1993. Each mode has advantages and disadvantages for management of individual patients. We await the results of the VA/NIH trial to see if outcomes such as mortality and recovery of renal function can be demonstrated. Management of respiratory failure Respiratory failure is a common complication of ALF. The liver failure population presents a particular clinical challenge because of co-existing clinical entities, such as hepatopulmonary syndrome. In addition, hepatic encephalopathy often leads to the need for endotracheal intubation for airway protection and prevention of aspiration pneumonitis. This discussion will focus on management of patients with hepatopulmonary syndrome (HPS), acute lung injury (ALI) and the acute respiratory distress syndrome (ARDS). Hepatopulmonary syndrome is defined as gas exchange abnormalities resulting in hypoxemia, and intrapulmonary vascular dilatation in the presence of liver disease(13). Clinically, these patients develop severe hypoxemia as a consequence of intrapulmonary shunting. HPS should be differentiated from portopulmonary hypertension (PPH), where pulmonary vascular resistance and pulmonary artery pressures are elevated. Patients with PPH are at increased risk of death with liver transplantation. Outpatient management is with oral vasodilators such as sildenafil, or with continuous infusion of prostacyclin. Acute management consists of close hemodymamic monitoring, with use of inhaled nitric oxide and intravenous prostacyclin. Inspired oxygen concentration should be kept high, as oxygen remains the most potent pulmonary vasodilator available. Hepatopulmonary syndrome is diagnosed using contrast-enhanced echocardiography, which will demonstrate intrapulmonary vasodilation. Microbubbles are used as contrast material, and they rapidly transverse the pulmonary circulation to appear in the left atrium. Management is largely supportive, as no therapy other than liver transplantation has been demonstrated to be efficacious. Indomethacin, almitrine, tamoxifen, somatostatin, and methylene blue dye have all been used with variable success(14). ALI and ARDS are characterized by hypoxemia, poor pulmonary compliance, bilateral infiltrates on chest radiograph, and the need for mechanical ventilation. Causative factors include pneumonia, severe sepsis, aspiration pneumonitis, and trauma. The 1990 European-American Consensus Conference on ALI/ARDS created definitions for these syndromes(15). With these definitions, clinical trials have been able to enroll homogeneous patients with ALI/ARDS. The definitions are listed below:

1. Acute Lung Injury a. Acute onset of respiratory failure b. Bilateral infiltrates on chest radiograph c. Absence of left atrial hypertension d. PaO2/FiO2 <300mmHg

2. Acute Respiratory Distress Syndrome a. Acute onset of respiratory failure b. Bilateral infiltrates on chest radiograph c. Absence of left atrial hypertension d. PaO2/FiO2 <200mmHg

45

The optimal ventilatory management of patients with ALI/ARDS has been tested in a prospective, randomized trial using a “protective” ventilation strategy with lower tidal volumes. Patients with ALI ventilated with 6 ml/kg tidal volume had 25% lower mortality than those ventilated with 12 ml/kg(16). Patients in the low tidal volume group also had fewer days of mechanical ventilation, and also fewer organ system failures. The mechanism for these effects is thought to be decreased inflammation from decreased lung stretch (volutrauma), whereby overdistention of normal areas of lung results in cytokine release into the alveolar space and systemic circulation. At this time, patients with ALI/ARDS should be managed using the following principles:

1. Assist Control ventilation using tidal volume of 6 ml/kg (ideal body weight). 2. Weaning from mechanical ventilation using a nurse and therapist-driven weaning

protocol(17). 3. Daily interruption of sedation(18) to minimize duration of mechanical ventilation. 4. Positioning the head of the bed at 70 degrees to minimize the risk of ventilator-associated

pneumonia(19). References 1 Richard C, Warszawski J, Anguel N, et al. Early use of the pulmonary artery catheter and outcomes in

patients with shock and acute respiratory distress syndrome: a randomized controlled trial. Jama 2003;290(20):2713-20.

2 Bernard GR, Sopko G, Cerra F, et al. Pulmonary artery catheterization and clinical outcomes: National Heart, Lung, and Blood Institute and Food and Drug Administration Workshop Report. Consensus Statement. Jama 2000;283(19):2568-72.

3 Connors AF, Jr., Speroff T, Dawson NV, et al. The effectiveness of right heart catheterization in the initial care of critically ill patients. SUPPORT Investigators. Jama 1996;276(11):889-97.

4 Martin C, Viviand X, Leone M, Thirion X. Effect of norepinephrine on the outcome of septic shock. Crit Care Med 2000;28(8):2758-65.

5 Rivers E, Nguyen B, Havstad S, et al. Early goal-directed therapy in the treatment of severe sepsis and septic shock. N Engl J Med 2001;345(19):1368-77.

6 Singri N, Ahya SN, Levin ML. Acute renal failure. Jama 2003;289(6):747-51. 7 Arroyo V, Gines P, Gerbes AL, et al. Definition and diagnostic criteria of refractory ascites and

hepatorenal syndrome in cirrhosis. International Ascites Club. Hepatology 1996;23(1):164-76. 8 Moreau R, Durand F, Poynard T, et al. Terlipressin in patients with cirrhosis and type 1 hepatorenal

syndrome: a retrospective multicenter study. Gastroenterology 2002;122(4):923-30. 9 Bellomo R, Chapman M, Finfer S, Hickling K, Myburgh J. Low-dose dopamine in patients with early renal

dysfunction: a placebo-controlled randomised trial. Australian and New Zealand Intensive Care Society (ANZICS) Clinical Trials Group. Lancet 2000;356(9248):2139-43.

10 Friedrich JO, Adhikari N, Herridge MS, Beyene J. Meta-analysis: low-dose dopamine increases urine output but does not prevent renal dysfunction or death. Ann Intern Med 2005;142(7):510-24.

11 Thadhani R, Pascual M, Bonventre JV. Acute renal failure. N Engl J Med 1996;334(22):1448-60. 12 Schiffl H, Lang SM, Fischer R. Daily hemodialysis and the outcome of acute renal failure. N Engl J Med

2002;346(5):305-10. 13 Lange PA, Stoller JK. The hepatopulmonary syndrome. Ann Intern Med 1995;122(7):521-9. 14 Herve P, Lebrec D, Brenot F, et al. Pulmonary vascular disorders in portal hypertension. Eur Respir J

1998;11(5):1153-66. 15 Bernard GR, Artigas A, Brigham KL, et al. Report of the American-European Consensus conference on

acute respiratory distress syndrome: definitions, mechanisms, relevant outcomes, and clinical trial coordination. Consensus Committee. J Crit Care 1994;9(1):72-81.

16 Ventilation with lower tidal volumes as compared with traditional tidal volumes for acute lung injury and the acute respiratory distress syndrome. The Acute Respiratory Distress Syndrome Network. N Engl J Med 2000;342(18):1301-8.

17 Ely EW, Baker AM, Dunagan DP, et al. Effect on the duration of mechanical ventilation of identifying patients capable of breathing spontaneously. N Engl J Med 1996;335(25):1864-9.

18 Kress JP, Pohlman AS, O'Connor MF, Hall JB. Daily interruption of sedative infusions in critically ill patients undergoing mechanical ventilation. N Engl J Med 2000;342(20):1471-7.

19 Kollef MH. The prevention of ventilator-associated pneumonia. N Engl J Med 1999;340(8):627-34.

46

Session 5 - Acute liver failure Liver support Dr Elizabeth Sizer. Consultant in Liver Intensive Care and Anaesthesia, King’s College Hospital, London Liver support is a system of therapies, which aim to improve outcome, provide an environment for regeneration and possible spontaneous recovery or failing that support the patient and bridge until the liver transplantation is an option. Such therapies may be applied to patients with acute liver failure (hyper-acute, acute and sub-acute), acute on chronic decompensation of liver disease and patients with graft dysfunction following liver transplant. It is important to note that the requirements of all of these groups is somewhat different and it is essential that such groups should not be mixed together when assessing the efficacy of such therapeutic manoeuvres. Ideally any liver support systems should be easily available, optimise or prevent deterioration of condition and be cost effective with a very low level of side effects. The systems available to us at the present to support the failing liver are liver transplant itself, either as a whole liver graft, a split liver graft, an auxiliary graft and in the paediatric population increasingly the role of hepatocyte transplantation is being developed and clinically applied. In addition, there are bio-artificial livers utilising cell based therapies and dialysis methods incorporating in addition, plasmapheresis. Within the bioartificial system consideration should be given to extra-corporeal liver perfusion. This has been investigated for many years, and recent reports by Horslen et al (Transplantation 2000: 70 (10:1472)), human and porcine livers were continuously perfused until either transplantation or withdrawal of support for the patient. Of 14 patients 9 were successfully bridged to transplantation and it appeared that there was improvement with respect to stabilisation of intra-cranial pressure and cerebral perfusion pressure. Ammonia also fell over 12 hours, as did bilirubin. Such systems must however be recognised as being very expensive, both in terms resource and time. A recent systematic review of the literature (Xenotrransplantation 2002: 9 (5:309)) examined 198 patients who had been exposed to such liver perfusion. Long term survival was 28% and was similar to that of standard of care. Independent predictors of positive outcome appeared to be in age <20, level of encephalopathy Grade 3 or 4, a perfusion time of >10 hours and use of human or baboon livers. The role of charcoal has also been in the clinical arena for many years. The first reports by Gimson et al in the Lancet in 1982 was that of an uncontrolled study of acute liver failure incorporating 75 patients. There was significant survival benefit in the treated patients 65 vs 15% in the historical controls. A further controlled study was then undertaken (O’Grady et al, Gastroenterology 1988: 94; 1186). This was a controlled study of charcoal haemo-perfusion and in this study no survival benefit was seen in the active limb. Further studies have subsequently been undertaken using haemo-diabsorption with the biologic DT system. Using this system there has been no convincing data to support the use in either acute liver failure or acute on chronic liver disease. Hybrid systems have utilised a mixture of charcoal and cellular mechanisms. The Berlin extra-corporeal liver support system utilises a very elegant 3D capillary structure containing hepatocytes. In the original reports this was porcine hepatocytes but more recent work is being undertaken with human hepatocytes. Provisional work with this system suggests it to be safe and provide haemodynamic stability but controlled trials are lacking. (Busse et al, Archives Surgery 1999: 384; 588). Another bio-artificial liver device of Circe utilises a 50g hepatocytes on collagen-coated micro-carriers. It is re-incorporated into an extra-corporeal circuit where the patient’s plasma is separated and then the plasma run over charcoal columns prior to being exposed to the poor sign hepatocytes. This system is applied for 6-7 hours and in original studies was shown to provide significant improvement in Glasgow Coma Score with a significant fall in intra-cranial pressure and improvement in cerebral perfusion pressure. (Arkadopoulos et al, International Journal of Artificial Organs, 1998: 21; 12; 781). This system also demonstrated significant falls in blood ammonia, bilirubin and creatinine. Other work with this system was reported by Didier Samuel (Transplantation 2002; 73 (2); 257). In this study 13 patients with acute liver failure were entered into the study, 10 were subsequently treated. There was a

47

significant decrease in bilirubin and an improvement in Glasgow Coma Scale that was related to volume of plasma exchange. Five patients had bleeding complications. A randomised control trial has been published in abstract form in regards of this liver support system some 147 patients with acute liver failure and 27 patients with primary graft non-function were entered into the trial. The 30 day survival with or without liver transplant was similar for patients receiving standard medical care and those in the liver support group. Within the acute liver failure group only, there was a benefit in the bio-artificial liver system of 59% and 73% survival. There also to be significant benefit in terms of survival within the sub-group analysis for acetaminophen induced acute liver failure and those patients who developed encephalopathy within 2 weeks of jaundice. (Stephens et al, American Association for the Study of the Liver, 2001). Another bio-artificial device is that of the Elad system of Vitagen. This incorporates hepatocytes in the form of hepatoblastoma cells (C3A cell line) that are grown to confluence in the extra-capillary compartment of a hollow fibre filter. The initial studies incorporated veno-venous access only with no oxygenator and the system was shown to be bio-compatible but to have little if any affect on measured clinical parameters. (Ellis et al, Hepatology, 1996: 24; 1446). A more sophisticated Elad circuit was thence developed, this incorporated plasma separation and an oxygenator system prior to the patient’s plasma being exposed to the hepatocytes. Another bio-artificial liver system is that of the porcine bio-artificial liver system of Amsterdam. This incorporates an elegant radial 3D flow system and animal data with this system is highly suggestive of efficacy. It has been reported in a case series of some 7 patients with evidence of bio-compatibility and possible haemodynamic stability. (van der Kerkhove, International Journal of Artificial Organs, 2002: 25 (10); 950). In the liver support systems that do not utilise biological components, there are therapies such as albumin dialysis and plasmapheresis. Albumin dialysis has been advocated for the removal of both water soluble and albumin bound toxins providing the albumin dialysis is incorporated with standard dialysis or haemofiltration. Mitzner et al (Liver Transplantation 2000; 6: 277) reported 13 patients with hepato-renal syndrome treated with either MARS therapy (8) or haemo-diofiltration (5). The MARS group of patients showed a significant decrease in creatinine, bilirubin and improvement in prothrombin time. There also appeared to be a survival benefit. There have also been several subsequent studies utilising albumin dialysis (MARS) in acute on chronic liver disease. Majority of these are not controlled but there is a persistent message of improved biochemistry, several studies have shown improved amino acid profiles and a suggestion of haemodynamic stability and neurological improvement. In more recent controlled trial published in Hepatology by Heemann (Hepatology 2002; 36: 949) looked at 24 patients with acute liver injury on the background of cirrhosis. Patients randomised to standard medical therapy with MARS or standard medical therapy alone. In the MARS group there was a reduction in bile acids and bilirubin with improved encephalopathy, but no differences were seen in albumin or prothrombin time. The control group had worsening in biochemistry and encephalopathy. At 30 days survival was significantly greater in the MARS treated group, 11/12 vs 6/12 in standard medical therapy. At 6 months survival however, this significance was lost with 6/12 of the MARS patients surviving and of 4/11 in the control group surviving. The role of plasmapheresis has been suggested as being significantly beneficial by the Copenhagen Group (Clemenson et al, Hepatology 1999, 29: 327). This group utilised membrane filtration with a pore size of 0.65mm and exchanged 15% of body weight on 3 consecutive occasions. They have noted improvement in encephalopathy with improved cerebral blood flow, improved cerebral perfusion pressure with no change in intra-cranial pressure. Mean arterial pressure improved significantly with an improvement in systemic vascular resistance. A multi-centre trial is at present being undertaken examining mortality. A recent systematic review published in JAMA (2003: 289 (2); 217) examined liver support systems. Over 528 references only 2 were randomised controlled trials in 483 patients. Ten trials were in acute or acute on chronic liver disease and 2 were in true acute liver failure. Overall the support systems did not appear to have an affect on mortality compared with standard of

48

care. The regression analysis suggested that the affect was dependent on the type of liver failure with the greatest chance of support being in that on acute on chronic liver disease. The present systems provide us with a method of controlling and improving haemodynamic variables and biochemistry. As yet there is little evidence to support the use of these systems in acute liver failure with respect to mortality benefit and two small studies suggest benefit in terms of mortality at 30 days in patients with acute on chronic liver disease (albumin dialysis). It behoves us as clinicians to undertake appropriately designed clinical trials examining the different sub-groups of patients with liver disease and developing a modular system of support that incorporates both bio-artificial and non-bio-artificial systems as the clinical situation demands.

49

Session 6 - Anaesthesia for ALF Intraoperative care Dr Christopher Snowden B.Med Sci (Hons), FRCA, MD. Consultant Anaesthetist, Freeman Hospital, Newcastle Perioperative Considerations for transplantation in Acute Liver Failure In contrast, to the intensive care management of the patient with acute liver failure (ALF), the perioperative anaesthetic management of emergency liver transplantation for acute liver failure has received limited attention in the medical literature. Although many of the principles regarding recipient management can be transferred from the intensive care unit (ICU) setting into the intraoperative period, the liver transplantation procedure (including transfer to theatre) creates specific stresses that require anaesthetic management. Ultimately, this will allow a smooth transition throughout the procedure and into the recovery phase. Patient population The paucity of literature concerning the anaesthetic management of ALF patients may in some part be explained by the fact that they make up only 7(USA)1 to 10%(UK)2 of liver transplantation recipients. Where the time from listing to organ procurement is extended, the population becomes self-selecting, in that existing ICU measures are not able to support the sickest patients. This often leaves the most physiologically adaptable patients to receive the available organs. Hence, it is not surprising that the current outcome of patients with ALF who make it to transplantation is excellent and in some series (especially paracetamol poisoning) rivals that of elective chronic liver transplantation3. These outcomes may change in the near future for two reasons; a) If artificial liver support and bridging therapies improve, even the sickest patients will be

supported to transplantation, which may lead to more severe perioperative problems. b) Recent series in the UK have been dominated by the presence of high percentage of

paracetamol poisoning. This may change in the UK secondary to the changes in the prescribing laws concerning paracetamol 4 .

Currently, the predominance of paracetamol poisoning in adult patients with ALF who undergo emergency liver transplantation means the overall population are a younger age group with minimal evidence of chronic liver disease. In addition, there is less incidence of other co-morbid disease including ischaemic heart disease or chronic pulmonary conditions that may increase perioperative risk. In contrast, the high prevalence of preoperative multiorgan failure creates specific practical and physiological challenges for the transplant anaesthetist. Preoperative theatre preparation Preoperative management of the FHF patient in the ICU is particularly relevant to the transplant anaesthetist and communication between clinicians is important from an early stage prior to and following listing. The ICU management of these patients has been covered in previous talks but various practical aspects are particularly relevant. An example checklist would include; • Invasive access/monitoring - vascular access and position related to theatre requirements,

ICP bolt, access for continuous veno-venous haemofiltration (CVVH) • Ventilation parameters - required to maintain adequate oxygenation and PCO2 levels, relation

to theatre ventilation modes, presence of permeability pulmonary oedema • Stability issues - cardiovascular and ICP stability, response to therapy, relevance to

intraoperative expectations • Renal support – use of CVVH, overall fluid balance • Sedation and paralysis • Coagulation issues – adequate preoperative correction, pre-ordered blood products Transfer to theatre is a potentially destabilising period and must be performed safely. CVVH should be discontinued in ICU but may be recommenced in the theatre. All infusions should be continued to ensure stability and nearly completed syringes changed in the ICU. Ventilation is optimally provided by a portable ventilation system, given the variation in manual ventilation with the inherent risk of hypercapnia and intracranial hyperperfusion. Positioning regarding ICP

50

pressures must also be adhered to including 15% head raised with neutral position. Many units prefer not to paralyse their patients on the ICU but paralysis prior to theatre should be provided for transfer. Total hepatectomy – bridge to transplantation? Where incipient cardiovascular or neurological collapse secondary to liver failure seems imminent and where a donor organ is not yet available, the possibility of elective total hepatectomy with portocaval shunting followed by liver transplantation should be discussed 5,6. It has been stated that "toxic liver syndrome" could be treated by means of this strategy 5,6. In most reports, this procedure has demonstrated a stabilising effect on the neurological status in patients with ALF 7,8. In contrast, the effect on cardiovascular stability has been variable 9,10. This procedure provokes some difficult ethical issues and in our institution, hepatectomy has only been used in extreme cases where a donor organ is available and harvest is imminent. Liver transplantation procedure The type of surgical procedure for emergency liver transplantation will depend on regional surgical experience and expertise. No published study has demonstrated an advantage of any one surgical technique for ALF, although in the author’s opinion, a conventional technique without veno-venous bypass is likely to be more cardiovascularly and neurologically challenging than other techniques. Our institution usually performs a standard orthotopic liver transplant supplemented with veno-venous bypass. The primary considerations for the anaesthetist involved in emergency transplantation for patients with ALF are those of:

• Cerebrovascular stability (closely linked to cardiovascular stability) • Fluid Balance • Acceptance of requirement for extended postoperative recovery

Cerebrovascular stability Patients with FHF and encephalopathy have impaired cerebral blood pressure and metabolic autoregulation and variations in mean arterial pressure will be reflected by changes to cerebral blood flow. It follows that cardiovascular stability during transplantation, is of paramount importance to the maintenance of cerebral perfusion pressure (CPP). Unfortunately, intraoperative changes in ICP are often haemodynamically silent and in our institution, intraoperative ICP monitoring is the norm. Changes in ICP are often temporally predictable during emergency transplantation. An early study, 8demonstrated that peaks of ICP occurred in the dissection, anhepatic and early reperfusion phases. However, in more recent reports, the changes in ICP seem to occur more consistently in the reperfusion and dissection phase with either a stabilisation or more often a reduction in pressure during the anhepatic phase. The differences may relate to use of VVB, with a requirement to increase central venous pressure to maintain venous return during the anhepatic period. The temporal changes during the transplantation procedure have been attributed to various mechanisms including inflammatory substances released from the failing liver 11, de novo cytokine production from the new liver, cerebral hyperperfusion secondary to an increase in venous return and cerebral ischaemia due to systemic hypotension. Although, temporal ICP rises during the procedure are somewhat predictable, the cerebral response of an individual patient is variable. In an early report12,all patients maintained a higher ICP throughout the procedure when compared with preoperative ICU values. Detry et al 13 suggested that those patients who developed preoperative rises in intracranial pressure, may be at greater risk of intraoperative changes in intracranial pressure presumably representing the reduced brain compliance in these patients. However, this finding has not been universally accepted. Some individual variation may be explained by the complex relationship between CBF, CPP and cerebrovascular resistance in patients with abnormal autoregulation. An increase in ICP in a non-compliant brain may be caused by increased CBF or a reduction in CPP causing ischaemia. Many patients with acute liver failure have evidence of cerebral “luxury” perfusion and cerebral hyperaemia secondary to reduced cerebrovascular resistance. It has been demonstrated that ICP surges in FHF are likely to be due to an increase in CBF14. Intraoperative measurements of the ICP, CMRO2 and CBF during transplantation in patients with FHF have also demonstrated that ICP is more usually related to rises in CBF than ischaemia induced by reduction in CPP, secondary to systemic hypotension15. Therefore, although a threshold CPP must be maintained,

51

relative hypertension during and after reperfusion, may be detrimental in terms of increasing the risk of increased microvascular pressure, cerebral hyperperfusion and ultimately cerebral oedema. More recently, the use of moderate hypothermia to control changes in ICP has been applied to patients with FHF11, 16. Jalan et al17 have shown that moderate hypothermia abolished ICP variability throughout the transplantation procedure in patients with the most difficult preoperative control. Other evidence, suggests that hypothermia may reinstate cerebral autoregulation and reduce cerebral hyperperfusion. The significance of these changes has not been demonstrated in an outcome study, but intuitively hypothermia would seem a reasonable therapeutic strategy where ICP control is troublesome. Where hypothermia can be used as an important baseline strategy for reducing surges in ICP, other manipulations may also be important. Variations in ICP during the early dissection phase can be reduced by expeditious hepatic artery/portal vein clamping. The development of an anhepatic state also promotes a reduction in the requirement for vasoconstrictors and inotropes 18. Furthermore, fluid removal via continuous veno-venous haemofiltration (CVVH) and mild hyperventilation in anticipation of increased CO2 production, may attenuate increases in intracranial pressures at reperfusion. Perioperative fluid balance The initiation of CVVH in ICU has become a standard procedure in patients with FHF to allow fluid management, ICP control and coagulation control. The decision to continue CVVH in theatre has been previously hampered by the accuracy and reliability of fluid removal of the CVVH machines. Some centres described the use of haemofiltration filters in parallel to the VVB circuit in a modified CVVH system19. However, advances in CVVH technology have made it more attractive to continue standard CVVH during the transplant, although optimal intraoperative regimes have not been defined20. In addition to rapid fluid removal enabling the transfusion of significant blood products, CVVH may also be useful in controlling anhepatic acidosis and pre-reperfusion hyperkalaemia. Realistic expectations of delayed recovery It is important to be realistic about the recovery of the patient with established FHF even after seemingly successful liver transplantation. Preoperative severe organ dysfunction, delayed new liver function, continuing raised intracranial pressure, pulmonary oedema and persistent renal failure are all reasons why recovery will be delayed. This requires continual supportive ICU care. Summary Patients with ALF make up only 7-10% of liver transplant recipients but create unique perioperative challenges for anaesthetists. Advances in supportive ICU care have improved the likelihood of patients surviving until a donor organ is procured, but where premorbid organ failure and imminent demise seems likely, “bridging” therapies including total hepatectomy must be considered. If emergency liver transplantation is performed, cerebrovascular stability remains a priority. The use of moderate hypothermia seems to abolish ICP variability and may constitute an important therapeutic strategy in high risk patients. However, this has not been subjected to a rigorous clinical trial. In addition, the early initiation of the anhepatic state will improve the patients’ condition both in terms of cardiovascular and cerebrovascular stability. Surges in cerebral blood flow are often more relevant to increases in ICP, whereas CPP is usually maintained even during periods of systemic hypotension due to the reduction in cerebrovascular resistance seen in acute liver failure patients. Intraoperative fluid balance and early post-reperfusion hepatic function are important to intraoperative success, but realistic acceptance of delayed postoperative recovery is also important. References

1 Hoofnagle JH, Carithers RL, Jr., Shapiro C, Ascher N. Fulminant hepatic failure: summary of a

workshop. Hepatology 1995; 21: 240-52. 2 Bernal W, Wendon J. Liver transplantation in adults with acute liver failure. J Hepatol 2004; 40: 192-7. 3 Farmer DG et al. Liver transplantation for fulminant hepatic failure: experience with more than 200

patients over a 17-year period. Ann Surg 2003; 237: 666-75. 4 Bernal W. Changing patterns of causation and the use of transplantation in the United kingdom. Semin

Liver Dis 2003; 23: 227-37.

52

5 Husberg BS et al. A totally failing liver may be more harmful than no liver at all: three cases of total hepatic devascularization in preparation for emergency liver transplantation. Transplant Proc 1991; 23: 1533-5.

6 Ringe B, Lubbe N, Kuse E, Frei U, Pichlmayr R. Total hepatectomy and liver transplantation as two-stage procedure. Ann Surg 1993; 218: 3-9.

7 Ejlersen E et al. Hepatectomy corrects cerebral hyperperfusion in fulminant hepatic failure. Transplant Proc 1994; 26: 1794-5.

8 Rozga J et al. Control of cerebral oedema by total hepatectomy and extracorporeal liver support in fulminant hepatic failure. Lancet 1993; 342: 898-9.

9 Bustamante M et al. Intensive care during prolonged anhepatic state after total hepatectomy and porto-caval shunt (two-stage procedure) in surgical complications of liver transplantation. Hepatogastroenterology 2000; 47: 1343-6.

10 Dominguez FE, Lange K, Lange R, Eigler FW. Relevance of two-stage total hepatectomy and liver transplantation in acute liver failure and severe liver trauma. Transpl Int 2001; 14: 184-90.

11 Jalan R, Pollok A, Shah SH, Madhavan K, Simpson KJ. Liver derived pro-inflammatory cytokines may be important in producing intracranial hypertension in acute liver failure. J Hepatol 2002; 37: 536-8.

12 Lidofsky SD et al. Intracranial pressure monitoring and liver transplantation for fulminant hepatic failure. Hepatology 1992; 16: 1-7.

13 Detry O et al. Intracranial pressure during liver transplantation for fulminant hepatic failure. Transplantation 1999; 67: 767-70.

14 Jalan R, Olde Damink SW, Hayes PC, Deutz NE, Lee A. Pathogenesis of intracranial hypertension in acute liver failure: inflammation, ammonia and cerebral blood flow. J Hepatol 2004; 41: 613-20.

15 Ardizzone G et al. Cerebral hemodynamic and metabolic changes in patients with fulminant hepatic failure during liver transplantation. Transplant Proc 2004; 36: 3060-4.

16 Roberts DR, Manas D. Induced hypothermia in the management of cerebral oedema secondary to fulminant liver failure. Clin Transplant 1999; 13: 545-7.

17 Jalan R et al. Moderate hypothermia prevents cerebral hyperemia and increase in intracranial pressure in patients undergoing liver transplantation for acute liver failure. Transplantation 2003; 75: 2034-9.

18 Noun R et al. Liver devascularisation improves the hyperkinetic syndrome in patients with fulminant and subfulminant hepatic failure. Transplant Proc 1995; 27: 1256-7.

19 Tobias MD, Jobes CS, Aukburg SJ. Hemofiltration in parallel to the venovenous bypass circuit for oliguric hypervolemia during liver transplantation. Anesthesiology 1999; 90: 909-11.

20 Blackwell MM, Chavin KD, Sistino JJ. Perioperative perfusion strategies for optimal fluid management in liver transplant recipients with renal insufficiency. Perfusion 2003; 18: 55-60.

53

Session 6 - Anaesthesia for ALF Postoperative care Dr Elizabeth Sizer. Consultant in Liver Intensive Care and Anaesthesia, King’s College Hospital, London

54

Session 7 - Paediatric Transplantation Indications Dr James Bennett. Consultant Anaesthetist, Birmingham Childrens Hospital, Brimingham Paediatric Anaesthesia – assessment and intraoperative management Aims The subject is vast, so I, like the other speakers, will necessarily be focused on certain areas. I aim to discuss the particular difficulties of anaesthesia for paediatric transplantation, discussing the major differences between children and adults, some of the specific indications, the changing indications and new techniques. Background Historically the first liver transplant was performed in a child in 1963. The child was a 3 yr old male with biliary atresia. The operation was a failure and the child died during the transplant. In 1967 the first successful liver transplantation was carried out in an 18month old child with a malignant liver tumour. The operation was a success and she died 400 days later with disseminated disease. The initial problems were sepsis and bleeding. Over the next 12 years the 1-year mortality never fell below 50%. It was not until the advent of cyclosporine in 1979 and split (1988)/reduced graft techniques became available in the 1980s that the results ceased to be poor1. Current status Results are now excellent. Survival in many centers is around 90% at 1 year. This success however has precipitated a severe shortage of whole grafts as more children are taken on for liver transplantation. The waiting list morbidity and mortality remains high2,3. The shortage of grafts has led to innovative surgical techniques such as living related transplantation and auxiliary transplantation to expand the donor pool. The success of transplantation has led to a widening of indications with a consequent further pressure on organ availability. It is crucial that we allocate resources to those who will benefit. What are the differences between paediatrics and adult anaesthesia? The most obvious difference is in size: a range of 1.8kg to 80kg is the BCH experience, with most transplants being performed on children 7-10kg. It is not surprising that there is a huge difference in the physiology between a small for dates neonate and adolescent child. There is an obvious fiddle factor. Interventions in small infants are technically more challenging and have a higher morbidity. There is also a much wider range of indications for transplant in paediatrics as well as more associated pathology. Most paediatric transplant programmes have evolved from adult ones, the equipment and drugs available are limited in paediatrics. Perhaps we should not forget the other important factors such as needlephobia, separation anxiety, anxious parents, need for play, lack of compliance and stress. Physiological differences Paediatric airway: To allow for growth the infant trachea has soft incomplete rings of cartilage, and the narrowest part of the airway is circular and at the cricoid level. Thus we tend to use uncuffed endotracheal tubes and calculate the size and length of tube from standard calculations or charts. The tongue is relatively large as is the occiput making the larynx more difficult to visualize. The distance between vocal cords and carina is short making bronchial intubation and

55

accidental extubation an ever-present risk. For this reason most transplant centers choose a nasal tracheal system for post operative ventilation. It must be stressed that this poses its own risks in the presence of coagulopathy.4 Haemodynamics Premature neonates may revert to fetal circulation under conditions of stress, acidosis and hypoxia. Otherwise the ventricles tend to be somewhat stiff and thus require optimal filling and high rate for optimal function. Thus factors which lead to a relative bradycardia can lead to a dramatic fall in cardiac output. Infants can maintain a phenomenal cardiac output. It is also important to remember the circulating blood volume is relatively greater in neonates at 80ml/kg. Thus a 3kg baby has a circulating blood volume of 240ml! Respiratory function Immature lungs tend to be stiff and noncompliant. The surrounding structure such as ribs and cartilage are soft to allow for growth. The ribs tend to angle laterally not downwards as in adults, this means that children rely a lot on diaphragmatic function. The result is a need to breathe rapidly and a tendency to dynamic collapse of airways and basal atelectasis in the presence of increased intra-abdominal pressure. Added to this the Oxygen requirements are high due to high metabolic rate. The net result is that a child is prone to arterial desaturation. Respiratory distress syndrome occurs in 10% of premature neonates and there is good evidence that as the child grows he is still prone to complications. Other causes of arterial desaturation in children are diaphragmatic splinting from ascites/hepatosplenomegaly, effusion, infection, shunting and hepato-pulmonary syndrome. Definitions Premature neonate: A baby born before 36 weeks gestation Neonate: A baby up to 28 days of life Infant: Up to the first year of life. Child: Up to the 16th year. The European paediatric liver transplant registry tends to analyse data of children under 2 and above 2 years, because of the impact of Biliary atresia on transplantation. , (roughly 70% in most series).5

56

Indications for liver transplantation Aims of liver transplantation6 50% survival at 5 years Improvement in quality of life Avoid end organ damage Anaesthetic assessment This tends to be carried out over 1 week for elective liver transplants. The anesthetist has a vital role in the multidisciplinary team. The aim is to assess/exclude co-morbidity, give an explanation of anaesthesia, PICU and analgesia to the child and family. In my institution the usual blood tests FBC, INR, Urea and Electrolytes, and liver function tests are performed. Additionally ECG, Echo and Chest Xray are performed as routine. The interpretation of many of these results can be difficult in children with end stage liver disease. For example most of these children are relatively overloaded with high cardiac output- so functional tricuspid regurgitation and left ventricular hypertrophy are not uncommonly seen on echo. Thus liaison with cardiology is very helpful. I often chase old anaesthetic records from referring hospitals or locally to see if intubation is straightforward. Remember the ex-prem infant may have a subglottic stenosis which will have major implications on peri and post operative management. For our patients who have had previous central lines we gain a neck ultrasound to assess patency of the great veins. If there is loss of 2 or more great veins we then proceed to angiography or magnetic resonance venography. It is crucial to identify such problems prior to embarking upon a transplant operation. It must be remembered that these children are sick and that many of the problems will not correct without a transplant. So a crucial piece of advice is DON’T FORGET COMMON SENSE. It must also be remembered that this is a hugely stressful time for the family and child and they can receive a huge amount of encouragement from the multidisciplinary team.

57

Cardiac assessment: Biliary atresia is associated with ASD,VSD and anomalous pulmonary venous drainage.7 Alagilles syndrome with branch pulmonary artery stenosis and pulmonary hypertension.8 Retransplants are associated with cardiomyopathy. Echocardiography is an essential investigation as it gives a functional assessment as well as structural assessment of cardiac anatomy. It is particularly useful in Hepatopulmonary syndrome where a contrast echo with agitated saline picks up the condition rapidly. Cardiac catheterization is less used now that we have high resolution Echo but is still useful particularly if an intervention is planned. Patent foramen ovale, ASD and VSD can all be closed angiographically with much less insult to a child with end stage liver disease What do we do with a chance finding of a PFO? Respiratory assessment Arterial hypoxaemia is not uncommon and is commonly caused by atelectasis, right to left shunting, pleural effusion or interstitial disease. Hepatopulmonary syndrome is an association of arterial hypoxaemia, cirrhosis and right to left shunt of blood through the lungs. It is reversible in adults and children after correction of cirrhosis by liver transplantation. It is commoner in children and is now considered an indication for transplantation. Diagnosis is clinical and by exclusion of other causes of hypoxia, to which is added radionucleide(Technecium 99 labelled micro albumin) scanning and contrast echo. Most children approach normal saturations within 3 months of transplantation.9 Respiratory distress occurs in 10% of premature infants. Our experience is poor with liver and small bowel transplantation. Cardiac assessment Alagilles syndrome is an association of cholestasis, triangular facies, hemivertebrae and branch/distal pulmonary artery stenosis and pulmonary hypertension. It is well assessed by echo. We have tried interventional cardiological procedures with little success. There is a suggestion that they tolerate intra-operative haemodynamics better than other children with pulmonary hypertension. However severe pulmonary hypertension has dismal results. Congenital cardiac disease is relatively common in biliary atresia. In most the cardiac function is good. A balance must be found between repair and the child declining into liver failure. Renal disease Possibly the most difficult to assess because of the huge number of factors that affect it in a child undergoing liver transplantation. The need for post-op renal replacement has a strong influence on survival. However the best predictor of renal failure is poor pre-op renal function, sepsis, blood loss and retransplantation.10 Hyponatraemia is a reasonably common finding. We have a minimum of 125mmol/l and in the past have used haemofiltration to raise the serum sodium. There is however little evidence that this is of benefit. Vascular access TPN administration is both a cause of cholestasis and loss of vascular access. 11 Loss of vascular access is an indication for intestinal transplantation +/- liver transplantation. At risk groups require careful placement and management of central catheters. Vascular ultrasound is particularly useful both in the pre-operative assessment and in the placement of lines at the time of transplantation.12, 13 Intra-operative monitoring Children require the same high quality monitoring as adults receive. However the equipment such as capnography is less accurate with high respiratory rates and low tidal volumes, and saturation probes are less reliable so vigilance is necessary. We also use core/peripheral temperature

58

monitoring as children are prone to hypothermia as well as iatrogenic warming from warming devices. Cardiac output monitoring is less good in paediatrics. We use a combination of TOE and PICCO reserving PA catheters for older kids. We have had some knotted PA catheters and there would appear to be more complications in the paediatric age group. PICCO is one of a number of new dilution cardiac output monitors. Its place in paediatrics is yet to be established. It requires a femoral arterial trace and a central catheter. It may have advantages in the small infant. Our experience of TOE would suggest it is easy to use and interpret even in the presence of varices. It has a particular role in intestinal transplantation where venous pressures may be impossible to use. It also gives an excellent guide to structure and function, and the dynamic relationship between filling and cardiac output14. Availble evidence and our experience suggest that it results in more aggressive transfusion of fluid. Equipment Perhaps it is the area of equipment where the biggest difference lies in paediatrics. A level 1 infusor has little value for a 5kg infant. Similarly measuring hourly urine output in a 2kg neonate with wide bore tubing is not appropriate. This leads to the need for some equipment that is used infrequently, which is problematical in these days of rigorous cost cutting. Small infants are particularly prone to hypothermia so we use a warming mattress and a Bair Hugger, the mattress is particularly useful during line placement and positioning. Organisational issues There are various models of care associated with anaesthesia and intensive care support for paediatric liver transplantation. They each have their advantages and disadvantages. A balance needs to be kept between exposure and experience on one hand and excessively onerous rotas on the other. We have an established rota of 4 paediatric anaesthetists, other centers have transplant anaesthetists who support paediatric transplantation. With time I suspect that external pressure will regulate against the “occasional paediatric anaesthetist” however there is probably very little evidence for this.15 The future of consultant working patterns and training are likely to lead to huge changes in practice. Widening indications The success of liver transplantation has moved it from an experimental procedure to a life saving and improving therapy for an ever increasing number of indications. This is placing an extra burden on the limited number of cadaveric donors. Ever more elegant surgical innovations have struggled to keep pace with demand. Over the next few years as a transplant community we are faced with achieving a balance between benefiting children with chances of a good outcome and treating the very sick who pose the highest risk. References: 1 Strong, R.W. Liver transplantation:current status and future prospects. J.R.Coll.Surg.Edinb.,46,Feb

2001,1-8 2 Vilca-Melendez,H and Heaton N.D. Paediatric liver transplantation: the surgical view. Postgraduate

Medical Journal 2004;80:571-576 3 Carter,B,A. History of Paediatric Liver Transplantation. E-medicine www.emedicine.com 4 Sumner,E and Hatch,D,J. Textbook Of Paediatric Anaesthetic Practice. Chapter 11, p 255-274

59

5 European liver transplant registry. www.eltr.org. 6 Liver transplantation in children. Donor organ use-protocols and guidelines for children undergoing

cadaveric liver transplantation. Nov 2002. www.uktransplant.org.uk. 7 Zukin, D,D. et al. Extrahepatic biliary atresia associated with cyanotic congenital heart disease. Three

case reports and a review.Clin Pediatr 1981 Jan ;20(1):64-66. 8 Choudry,D et al. The Alagille’s Syndrome and its anaesthetic considerations. Paediatric Anaesthesia

1998;8:79-82. 9 Van Obbergh,L.J et al. Hepatopulmonary syndrome and liver transplantation:a review of the

perioperative management of seven paediatric cases. Paediatric Anaesthesia 1998;8:59-64. 10 Perez, G,O. et al. Dialysis, haemofiltration and other extracorporeal techniques in the treatment of the

renal complications of liver diseasein The Kidney in Liver Disease, Hanley & Belfus, INC.,Philadelphia. 11 Puntis J.Nutritional support at home and in the community. Arch Dis Child 2001;84:295-298. 12 National institute for clinical excellence technology appraisal guidance 49. www.nice.org.uk. 13 Selvaggi, G et al . Analysis of vascular access in intestinal transplant recipients using the “Miami

classification” from the VIII international small bowel transplant symposium. Transplantation awaiting publication.

14 Suriani,R,J et al. Intraoperative transoesophageal echocardiography during liver transplantation. Journal of Cardiothoracic and Vascular Anaesthesia Vol. 10;1996;699-707.

15 McNicol, R. Paediatric anaesthesia –who should do it? The view of the specialist hospital. Anaesthesia 1997;52;513-516.

60

Session 7 - Paediatric transplantation Post operative and paediatric intensive care issues Dr Moira O’Meara. Consultant Paediatric Anaesthetist, St James’s University Hospital, Leeds When a child is transferred to Paediatric Intensive Care (PICU) after liver transplantation it is the responsibility of this team to continue the process of achieving a warm, well perfused, well oxygenated pain free medium for the new liver to work in. Providing there is a good functioning graft there will be normalisation of metabolic and physiological variables and the recipient will recover. Ventilation/ Sedation The process of liver transplantation has been progressively refined and consequently PICU duration of stay is getting shorter. The median duration of stay for first elective paediatric liver transplant in UK in 2003 was two days. Patients are returning from theatre in better physiological condition and the requirement for post-operative ventilation is less. Ganschow et al 1 examined trends in post-operative ventilation times in the Hamburg unit and showed a progressive reduction from a mean of 5.2 days ventilation between 1991 and 1994 to 1.2 days between 1996 and 1998. Concomitant with this change, these authors commented on a reduced incidence of pulmonary complications and earlier discharge from PICU. Despite this, so far, for children, a short period of observation in PICU whilst ventilated remains normal practice. Children can be difficult to sedate for short periods because only a limited range of drugs are recognised for this use in PICUs, particularly when the oral route is avoided in the early post-operative period. Morphine (0-40mcg/kg/hour) and Midazolam (0-6mcg/kg/min) infusions are normally used. When a patient returns from a theatre environment where they have received short acting agents, which are not dependent upon the liver or kidneys for metabolism, such as Remifentanil, Atracurium and Desflurane, it is frustrating to see them turned into an ICU patient. With these factors in mind, and with the benefit of working along side an adult unit that has extubated it patients immediately after liver transplantation for some time 2 we have established a practice of immediate extubation of paediatric liver transplant recipients by anaesthetist unless there are specific contra-indications 3. The decision to “immediately extubate” is made shortly before wound closure. There must be hemodynamic stability, no evidence of ongoing hemorrhage, no persistent inotrope requirement, evidence of early graft function (falling lactate, normal Thromboelastograph), adequate oxygenation at an FiO2 of 0.4 or less, and no requirement for high ventilatory pressures. Since adoption of this policy in June 2002, 46 cadaveric transplants have been performed in 40 patients: 26 of 34 (76%) elective transplants and 4 of 12 urgent transplants were extubated immediately after surgery. Eight of 14 (57%) children weighing less than 10kg were successfully extubated and reduced/split liver grafts were not contraindications to immediate extubation. Furthermore, hepatopulmonary syndrome (3/3 extubated), cystic fibrosis with pulmonary disease (3/3 extubated), retransplantation (3/7 extubated) and super-urgent cases (4/12 extubated) did not preclude successful immediate extubation, although careful respiratory management with supplementary oxygen, judicious chest physiotherapy, and adequate analgesia is mandatory. In acute liver failure, none of the four immediately extubated patients had severe encephalopathy (grade 3-4) or evidence of raised intracranial pressure prior to transplant and only one was ventilated preoperatively. In the elective group 3 extubated children experienced respiratory problems: 2 teenagers required facial CPAP on day 3 post operatively for pulmonary oedema and an eleven year old girl with hepatopulmonary syndrome required reintubation 4 hours after transplant for respiratory distress. It was subsequently confirmed that she had developed a transfusion related acute lung injury (TRALI), which recovered after seven days of ventilation. There were no deaths in this elective group. In the super urgent group there were no adverse events in the 4 patients who were immediately extubated. In our experience, the criteria we applied for immediate extubation proved effective since none of the immediately extubated patients developed primary non-function or had to return to theatre to control bleeding or treat vascular thrombosis. Three children in this series had primary non-function and underwent successful retransplantation but they did not satisfy criteria for immediate extubation and remained intubated until after retransplantation.

61

Only one previous study has documented the feasibility of immediate tracheal extubation of pediatric liver transplant recipients. Ulukaya et al (4) reported that 12 of 40 children (30%) were extubated immediately after OLT during a 5-year period with no adverse outcomes. There may be physiologic and clinical advantages to immediate extubation, as well as being more comfortable for patients. Positive pressure ventilation reduces cardiac output, increases right ventricular workload, induces temporary backflow into the inferior vena cava and hepatic veins, and reduces splanchnic blood flow. These effects are more pronounced when positive end-expiratory pressure is used. In spontaneously breathing patients, venous return is improved and cardiac output and hepatic blood flow are increased. Thus, after immediate extubation, venous return from the graft is almost certainly improved. In this context, it is interesting to note that mean peak ALT concentrations after elective transplantation were significantly lower in the extubated group. Finally, there may be economic advantages to immediate extubation. Although we could not demonstrate a shorter postoperative PICU stay in our cohort of elective patients compared to national data from units not employing a policy of immediate extubation, we were obviously cautious about early discharge from PICU after immediate extubation and earlier discharge to the ward may well have been possible. TISS-28 scores 5 were lower in the immediately extubated group indicating a less intense nursing requirement which may reduce PICU costs. Analgesia Regardless of the timing of extubation, patients require effective analgesia in the post-operative period. The most popular form of analgesia remains Morphine, either as a simple infusion (0-40 mcg/kg/hour), patient controlled analgesia, or nurse controlled analgesia. Some authors suggest that there may be less need for opiate analgesia after liver transplantation 6,7.The reasons for this are unclear. Shelly and Park 8 looked at the disposition of morphine in two children post operatively and observed that both patients metabolised Morphine normally but one patient who had renal failure accumulated the active metabolites. Patients who have been extubated can express pain and exhibit recognisable signs of pain. We retrospectively reviewed the case notes of 19 children who had been immediately extubated after liver transplant to assess their analgesic requirements in the first 24 hours after transplant as judged by nurses trained in pain management. 9 patients received a Morphine infusion at 0-40mcg/kg/hour and 10 patients received a patient controlled analgesia with a background of up to 20mcg/kg and 10mcg/kg boluses with a five-minute lock out. The infusion group required a mean dose of 23.3mcg/kg /hour (range 18.3-43) and the PCA group a mean dose of 35.6mcg/kg/hour (range 20-58). These results indicate an analgesic requirement in the expected range for a laparotomy but further prospective work is needed. There are no published reports of use of epidural analgesia in children undergoing liver transplantation. There is, however, one case report of the use of caudal morphine, which suggested that it facilitated immediate extubation and reduced requirement for systemic analgesia (9). Paracetamol We prescribe regular rectal or nasogastric paracetamol (20mg/kg) six hourly as a co-analgesic to allow morphine sparing. Others centres are more cautious and use a maximum daily dose of 60mg /kg. There is limited information about the safety of paracetamol after liver transplantation although Park JM et al 10 have suggested that there may be transiently altered metabolism of Paracetamol post operatively (at least10 days) as a result of impaired glucuronidation and sulfation and enhanced N-acetyl-p-benzoquinone imine (NAPQI) formation which is a toxic metabolite. We do not prescribe analgesic doses of NSAIDs during the nil by mouth phase. Fluids Several protocols have suggested that this is difficult to get right and I would agree. Our current practice is to give 100% maintenance including the fluid in drugs. Other centres employ a policy of 2/3 maintenance. Ascitic and other drain losses are half replaced with a mixture of HAS 4.5% and dextrose 5% provided CVP (2-5 extubated patients, 5-10 intubated patients) and urine out put (>1 ml/kg) are acceptable.

62

Hyperglycaemia occurs commonly following liver transplantation 11.The transplanted liver is completely denervated with lack of glycoregulation and insulin resistance of glucose metabolism. The recipient has also received high doses of steroids and undergone a stress response and may have received a large transfusion of glucose rich blood. Tacrolimus also promotes hyperglycaemia, so there are many reasons why hyperglycaemia occurs. Children, including infants only rarely require a greater concentration than 5% dextrose as maintenance. It is our practice to start an insulin infusion if blood glucose remains greater than 12mmol/l several hours after the procedure. We wish to avoid polyuria associated with spilling of glucose into urine, which results in intravascular dehydration. There is also a higher incidence of acute rejection in hypergylcaemic patients. Hypothetically early extubation, or at least early discontinuation of sedative agents aids optimal fluid balance because there is less need to treat the hypotensive side effects of sedation and IPPV. A moving awake patient has activation of their muscle pumps and greater lymphatic drainage and so is less likely to redistribute fluid to third spaces. Hypernatraemia is also relatively common, partly because blood and blood products are sodium rich. Hypokalaemia is expected and a sign of a functioning liver and magnesium suphate (0.5mmol/kg/day) is usually required, especially for children who have been receiving diuretics. Tacrolimus also promotes renal excretion of Magnesium. Nasogastric or oral feeding can usually commence from day 3. Ranitidine (1mg/kg IV TDS) or Omeprazole will protect from gastric erosions prior to this Immunosuppression and Infection Children are usually immunosuppressed with Tacrolimus and steroids or in retransplants or children with renal problems MMF or IL2 antibodies are used. Post operative infections are commonly described in the post operative period despite antibacterial prophylaxis with Cefuroxime and Amoxil and Metronidazole (when there is a roux) and anti fungal prophylaxis (Fluconazole or Ambisome depending on risk). CMV prophylaxis is usually started on day 2 when there is a CMV positive donor and CMV negative recipient. Small children are usually CMV negative. When infection is suspected all potential sites must be cultured. Bacterial infections are most common and are often related to central lines (Strep faecalis, Strep viridans, Pseudomonas Aeroginosa and Staph Aureus). Fungal infections (Candida Albicans and Aspergillus) are more common in ALF with acute hepatic necrosis pre transplant. Young, small patients with long PICU stays and prolonged ventilation are most at risk of infection 12. Prophylaxis against pneumocystis carinii is usually started on day 5. Non infectious complications of immunosuppression include hypertension, which can be treated with Nifedine or Atenolol. When all is not well In a small percentage (4 to 7%) of cases the patient may return from theatre with an on going metabolic acidosis, still requiring inotropes, oliguric and coagulopathic and hypoglycaemic. In these situations we have to consider the possibility of Primary Non Function: absence of metabolic and synthetic function of the liver. The causes of primary non-function are varied including hyperacute rejection, a preservation injury to donor liver, adverse conditions in donor, some recipient factors. Urgent retransplantation is usually indicated and the patient must be supported physiologically until a second donor is available. Monitoring Cardiac output measurement Whatever the cause of dysfunction the more information we have about patient, the easier it is to manage them. We can make a reasonable assessment of cardiovascular status based on HR, BP, peripheries and core/peripheral temperature and urine output. However there are disadvantages to starting inotropes or loading with fluid with out direct information about cardiac out put. Although PA catheters may be inserted into children as small as 10 kg there are complications associated with their use. Alternatives include the PICCO, Trans-esophageal Echocardiography (TOE) and LiDCO. We use LiDCO. The LiDCO monitor measures the area under the curve of arterial waveform and uses this to calculate stroke volume, which is multiplied by heart rate to calculate cardiac output. It further calibrates the value using a Lithium dilution technique. Using one of these techniques helps to identify whether more filling, greater cardiac work or tightening of systemic vascular resistance (for example with Norepinephrine) is required. TOE remains the gold standard but it is expensive and requires expertise to be used. . LiDCO

63

has the advantages of requiring only an arterial line and a peripheral cannula or central line and gives continuous values. There is evidence that N Acetyl Cysteine improves physiology after reperfusion in liver transplantation 13. In our centre we maintain an infusion of 100mg/kg/24hours from induction of anaesthesia until AST starts to fall in the post-operative period. Hepatic artery thrombosis may occur in up to 10% of paediatric liver transplants: in the early post operative phase there are initially no collateral pathways and compensation between the portal vein and the hepatic vein may be poor initially. Early HAT usually requires retransplantation. In children less than 5 years HAT is less in split or reduced grafts than in small whole graft. Preventative measures include ensuring good flow (low viscosity with Hb 8-10 g/dl, use of anticoagulants such as aspirin, enoxaparin). Early diagnosis of stenotic or thrombotic problems using Doppler ultrasound of the liver and angiography may indicate requirement for either thrombectomy and hepatic artery re-anastomosis or stenting to occur thus avoiding retransplantation. Later complications of HAT include biliary leaks, strictures and hepatic abcesses. Renal complications may arise as result of PNF or HAT or simply secondary to insults in the perioperative period. This may resolve with careful fluid management and judicious use of diuretics or CVVH may be required. Tacrolimus and cyclosporin are nephrotoxic and may have to be withheld or substituted with alternative agents. Bartosh 14 noted that mortality was high in children who required dialysis after liver transplantation. Neurological complications are relatively common and children have a reduced seizure threshold post transplant. This may be secondary to pyrexia, metabolic derangements, intra cranial bleeding, hypoxia, hypocalcaemia and hypomagnesaemia. Treatment should be directed towards stopping the fit (Lorazepam), preventing further fits (Phenytoin) and treating the underlying cause. In summary, children who have received liver transplants present multi-system challenges in the post operative period but despite this it is possible to achieve very good outcomes. References: 1 Ganschow R, Nolkemper D, Helmke K et al. Intensive care management after pediatric liver

transplantation: a single-centre experience. Pediatric Transplantation 2000;4(4): 273-9. 2 Buglass H, Harding RJ, Milligan L et al. A prospective randomised trial of early extubation in liver

transplantation: an interim analysis. British Journal of Anaesthesia 2003; 90 (6): 830P. 3 O’Meara ME, Whiteley SM, Sellors JM. Immediate extubation of children following liver transplantation is

safe and may be beneficial. Transplantation; in press. 4 Ulukaya S, Arikan C, Aydogdu S, Ayanoglu HO, Tokat H. Immediate tracheal extubation of pediatric liver

transplant recipients in the operating room. Pediatr Transpl 2003;7: 381-384. 5 Miranda DR, de Rijk A, Schaufeli W. Simplified Therapeutic Intervention Scoring System: the TISS-28

items. Results from a multicenter study. Crit Care Med 1996; 24:64-73. 6 Donovan KL, Janicki Pk, Striepe VI et al. Decreased patient analgesic requirements after liver

transplantation and associated neuropeptide levels. Transplantation 1997; 27; 63(10): 1423-9. 7 Moretti EW, Robertson KM, Tuttle-Newhall JE at al. Orthotopic liver transplant patients require less post

operative morphine than do patients undergoing hepatic resection. J Clin Anesth 2002; 14(6):416-20. 8 Shelly MP, Cory EP, Park GR. Pharmacokinetics of Morphine in two children before and after liver

transplantation. British Journal Anaesthesia 1986; 58(11): 1218-23. 9 Kim TW, Harbott M The use of caudal morphine for pediatric liver transplantation. Anesth Analg 2005;

100(2): 602-3. 10 Park JM, Lin YS, Calamia JC et al Transiently altered acetaminophen metabolism after liver

transplantation. Clin Pharmacol Ther 2003; 73(6):545-53. 11 Shangraw RE, Hexem JG, Glucose and potassium metabolic responses to insulin during liver

transplantation. Liver Transpl Surg 1996; 2(6):443-54. 12 Bouchu JC, Stamm D, Boillot O et al: Postoperative infectious complications in paediatric liver

transplantation: a study of 48 transplants. Paed Anaes 2001; Jan 11(1): 93-98. 13 Thies JC, Teklote j, Clauer U et al. The efficacy of N-acetyl cysteine as a hepatoprotective agent in liver

transplantation. Transpl Int 1998; 11suppl1:S390-2. 14 Bartosh SM, Alonso EM, Whitington PF. Renal outcomes in pediatric liver transplantation. Clinical

Transplantation 1997; 11: 354-60.

64

Session 7 - Paediatric transplantation Surgical aspects Prof. Alastair Millar. Consultant Paediatric Hepatobiliary and Transplant Surgeon, Birmingham Children’s Hospital, Birmingham Some surgical aspects of paediatric liver transplantation Paediatric liver disease was to a great extent the driving force behind the development of human liver transplantation, with biliary atresia being the most frequent indication. The longest surviving patient with a transplant received her graft in 1969. The techniques developed by Starzl and Calne involved removing the whole liver and replacing it, including the vena cava in the orthotopic position. After the reintroduction of cyclosporine increasing success was achieved but paediatric waiting list mortalities in the late ‘70s and early ‘80s remained about 50% due to lack of size matched donors. The concept of an auxiliary transplantation had been developed as early as the mid 1950s in the laboratory. This was initially heterotopic but, from a technical point of view, particularly with regard to hepatic venous outflow, the graft was unsatisfactory, although it did have the advantage that the native liver was not removed. Orthotopic partial auxiliary transplantation (APOLT) has been a relatively new development used specifically for acute liver failure and in some instances of metabolic liver disease. The types of grafts available are cadaveric donors, living donors including the so-called domino living donor and xenografts, which have not been successfully introduced in humans. Due to the shortage of paediatric donors and after experimental animal models had been developed by Bax in Belgium in the 1970s, the concept of a reduced liver graft was introduced into clinical practice. This was an adult or larger donor liver, usually with the right lobe removed and grafted in place with the vasculature and common hepatic and even common bile duct intact. This addressed the problem of shortage of donors for paediatric recipients but reduced the total pool of donors. The next step was to split the liver into two functioning units, usually using a left lateral segment for the paediatric recipient and the right lobe for a smaller adult donor. Driven particularly in Japan by cultural and legal factors, which did not accept a brain dead, heart-beating donor, living related transplant was developed, pioneered by Russell Strong in Brisbane and Broelsh and Whittington in Chicago. There has now been widespread use of living donors over the last 15 years and this has been extended to adult living donor transplantation in the last decade. It has been conclusively shown that the long-term outcome of these grafts do not differ whether they are whole grafts, reduced size or split grafts. Small size and age is no exclusion. The smallest transplant has been done in infants down to under 2 kg in weight. However, there are specific issues with regard to small size and weight and I will address some of these later. The problem of size mismatch in the paediatric population as opposed to the adult where the “small for size” is a current hot topic, is rather that the size of the graft is too big for the patient and various techniques have been developed to deal with this. Particularly where liver and small bowel are grafted together, a planned abdominal expansion can be done using skin expanders and in the situation where the graft cannot be accommodated, staged abdominal closure or prosthetic patch closure is indicated. Although the concept has been around for some time, abdominal compartment syndrome is clearly recognised as having significant morbidity and even mortality in the transplant setting. In small infants particularly where the main arterial pressure can be as low as 45-55 mmHg, an abdomen closed under any pressure could compromise organ perfusion and particularly where a segmental graft has been placed in the abdomen with a large anterior-posterior diameter, any posterior displacement of the graft against the rigid posterior abdominal wall could compromise blood flow through small vessels at low pressure. Indications for transplant can be broadly divided into chronic liver disease, cirrhosis, metabolic diseases, acute liver failure and neoplastic disease. There have been significant changes in the pattern of indications over the last 20 years. Initially chronic liver disease was dominant with more than 60% of patients having biliary atresia in the many reported series. Currently the proportion of chronic liver disease patients listed for transplant has been reduced with biliary

65

atresia representing less than 25% of our patients in Birmingham and a new group of patients - infants with intestinal failure associated liver disease (IFALD), now nearly the biggest single group in the last year. Metabolic disease and acute liver failure remain at about 10%. Isolated liver transplant for intestinal failure associated liver disease is used in specific categories where it is expected that full bowel adaptation would take place in infants with short bowel but early onset liver disease, fuelled by the necessity for TPN and recurrent episodes of sepsis and bacterial translocation, has resulted in irreversible liver damage and portal hypertension. It has been shown that by giving these children an isolated liver transplant along with appropriate gastrointestinal management promoting bowel adaptation, many cases can be weaned from parenteral nutrition. In acute liver failure, transplantation is life saving. What has been difficult is to judge when liver disease and its consequences will lead to irreversible encephalopathy. Various indications have been identified in children, which differ from the adult King’s criteria. Currently indications for listing are: encephalopathy and decrease in coagulation factors below 20%, or the new King’s criteria put forward recently which includes an INR of >4, a total serum bilirubin of >235 mm/l, age <2 years and WBC >9000. Any one of these carries a specificity of 100% and any 4 has a sensitivity of 100%. There are currently few absolute contraindications for liver transplantation: brain death and full blown AIDS would be regarded as absolute contraindications and relative contraindications are sepsis, expected severe neurological deficit, HIV infection and severe cardiopulmonary disorder. In the last 20 years in Birmingham 174 patients have been referred with acute liver failure of which half were listed. Of the 87 who were listed, 73 were transplanted, 2 recovered and 12 died. Of those transplanted, overall 68% remain alive. Unfortunately there is still little in the way of alternative therapy for acute liver failure although extracoporeal perfusion, e.g. the MARS system has been effective in avoiding transplantation. Recently, auxiliary partial transplantation has been offered as an alternative but has not had a generally enthusiastic reception. Auxiliary partial orthotropic liver transplantation (APOLT) has not been widely used in children although there are some units, which advocate its use. However, on reviewing the literature there have only been relatively few patients under 15 years, the main reasons being prolonged ischaemia, longer duration of surgery, higher transfusion requirements, more bacteraemia, more central nervous system complications and a greater need for re-transplantation with ultimately reduced survival. Living related transplantation in the acute liver failure setting has not been widely accepted but may be life saving in those centres advocating its use. However, it does put the donors under enormous pressure in difficult circumstances. Living donor transplantation does, of course, have significant advantages in that the donor is in optimal condition, it is usually an elective operation with a short ischaemic time, haemostasis is complete prior to transplantation and the long term outcome is excellent. Disadvantages, of course, are the healthy donor is at risk for complications and mortality. There is a significant step up in technical expertise required with regard to not only the procurement of the donor liver but also with its implantation and the use of microvascular techniques. Ensuring immediate post-operative daily or even hourly monitoring of the graft with Doppler ultrasound is an essential part of the management and any suspicion of under perfusion or occlusion of graft vasculature is taken very seriously with usually immediate response in the form of re-exploration or release of abdominal tension. Currently the survival with transplantation is excellent such that a recent series of more than 100 patients had a 100% one year survival with liver transplantation excluding those patients with intestinal failure liver disease which are a special high risk category. Paediatric liver transplantation has now become a victim of its own success and as traffic surveillance and child accident prevention strategies become more advanced, paediatric donorship will be something of the past and increasingly paediatric liver transplantation will have to survive using full resources of brain dead, heart beating livers in the form of split liver and even monosegmental grafts as well as expanding living related transplantation.

66

References: Surgical Techniques Surgical Techniques and Management pp265-280: Jean de Ville de Goyet, in Pediatric Solid Organ Transplantation Eds.AH Tejani, WE Harmon, RN Fine Munsgaard Copenhagen 2000. Living related liver transplantation Otte, J.B.: Is it right to develop living related livr transplantation? Do reduced and split livers not suffice to cover needs?. Transpl. Int. 1995; 8:69. Parental experience with living related liver transplantation Editorial, Pediatr Transplantation 2004; 8:317-21 Emre S et al: Living related liver transplantation for acute liver failure in children. Liver Transpl Surg. 1999; 5(3);161-65. Sugawara Y, Makuuchi M: Review: Advances in adult living donor liver transplantation: A review based on reports from the 10th anniversary of the adult-to-adult living donor transplantation meting in Tokyo. Liver Transplantation 2004; 10:6;715-20. Acute liver failure in children Dhawan A et al: Approaches to acute liver failure in children. Pediatr Transplant 2004; 8(6); 584-88. Split liver transplantation Malago, M. et al: Split liver transplantation: Future use of scarce donor organs. World J Surg 2002: 26;275-82. Auxiliary liver transplantation Boudjema, K. et al.: Auxiliary Liver Transplantation and Bioartificial Bridging Procedures in Treatment of Acute Liver Failure. World J Surg 2002: 26; 264-74. Azoulay D. et al.: Auxiliary partial orthotopic versus standard whole liver transplantation for acute liver failure: a reappraisal from a single center by a case-control study. Ann Surg 2002: 234(6); 723-31. Abdominal Compartment Syndrome Hunter JD, Damani Z, Review Article: Intra-abdominal hypertension and the abdominal compartment syndrome. Anaesthesia 2004: 59;899-907. Handschin AE et al: Abdominal Compartment Syndrome after Liver Transplantation. Liver Transplantation 2005: 11(1); 98-100. Beck R et al: Abdominal compartment syndrome in children. Pediatr Crit Care Med 2001: 2; 51-55.

67

Session 8 - Post-operative care Immunosupression Dr Kosh Agarwal. Consultant Physician / Hepatologist, Freeman Hospital, Newcastle Although advances in other fields have made important contributions, developments in immunosuppressive therapy have been paramount to transplantation. Solid organ transplantation can only be deemed successful if rejection of the allograft is prevented. The introduction of cyclosporine in the late 1970s was a landmark event, ushering in the current era of transplantation as an accepted, routine, indication for end-stage organ failure. The development of immunosuppressive regimes has heralded improving short and medium survival. The 1 and 5 year liver allograft and patient survival rates are approximately 87% to 70% respectively. However, long-term graft survival remains problematic, with chronic rejection and recurrent disease concerns being tempered by increasing morbidity and mortality secondary to long-term exposure to immunosuppression. Whilst a ‘functional tolerant state’ is achievable; from an immunological perspective, many issues remain unclarified with regard to the induction of ‘true’ tolerance to the liver allograft. Overview of alloimmune responses: a central role for T cells The CD4 T cell is crucial in the initiation and co-ordination of the immune response to the allograft after recognition of foreign antigen1. CD4 T cells recognise antigens through direct and indirect pathways, become activated and undergo clonal proliferation. Recipient T cells can recognise intact foreign Major Histocompatability Complex (MHC) encoded on donor cells (direct), which may contribute to hyperacute mechanisms of rejection. In the indirect pathway, alloantigens are presented to lymphocytes by antigen presenting cells (APC) (Fig 1), with these antigens being encoded by highly polymorphic loci within the MHC. This occurs in conjunction with ‘accessory’ molecules, mainly within secondary lymphoid organs. APCs ingest foreign proteins and load them onto MHC molecules to present and interact with the T cell receptor and generate ‘signal 1‘, transduced through the CD3 complex. Other signals are required, ‘signal 2’ results from binding of co-stimulatory molecules on T cells (CD28 receptor) with their ligands present on APCs (CD80/CD86). Signals 1 and 2 activate signal transduction pathways, which in turn activate transcription factors that lead to clonal expansion and the expression of molecules such as IL-2, CD154 and CD25. These events are required for naïve T cell activation and are primarily mediated by calcineurin dependent pathways. ‘Signal 3’ of T cell activation results from autocrine and paracrine cytokine mediated signalling, mainly through IL-2 and other cytokines. Generation of effector cells, such as activated macrophages, CD8 (cytotoxic) T cells and B cells occurs via the cytokine milieu facilitated by alloactivated CD4 T cells. This cascade results in alloantibody production, antigen-specific cell lysis and delayed-type hypersensitivity responses, which can lead to graft destruction. Increasing evidence suggest that tolerance is a dynamic process, that must be reinforced. Transplantation tolerance involves several mechanisms, including deletion, anergy, and suppressor or regulatory cells. Whilst mechanisms of tolerance can be classified as central or peripheral, there is undoubtedly overlap. Central mechanisms mainly refer to deletion, which occurs primarily within the thymus. Peripheral mechanisms refer to anergy or suppressor/regulatory cells, either by failure of signal 2 or via populations of immunoregulatory T cells

68

Figure 1: Model of T-cell regulation by CD28 and CTLA-4. T-cell activation is initiated by the interaction of T-cell receptor with antigenic peptides presented by the MHC on antigen presenting cells. Secondary signals are critical to development of immunity or tolerance. B7 molecules are expressed on antigen presenting cells and tightly regulate T-cell activation by delivering signals through CD28 and inhibitory signals through CTLA-4. PTK=protein tyrosine kinase. SHP-2=SH 2 domain containing protein tyrosine phosphatase. PIK’-k=phosphatidylinositol 3’ kinase. Modified from1. Immunosuppression agents (Table 1 adapted from 2) Immunosuppressive drugs exert a spectrum of effects with clinical relevance. Firstly, by targeting T cell activation, cytokine production and clonal expansion, the immune response to the allograft is abrogated and rejection minimised. However, as a consequence of immunosuppression the donor is more susceptible to infection or cancer. Lastly, significant non-immune toxicity can accrue from exposure to long-term immunosuppression. This toxicity is agent specific. Currently the most clinically relevant non-immune toxicity is the nephrotoxicity of the calcineurin inhibitors, although increasing cardiovascular morbidity is also an issue. Immunosuppressive agents can be classified into induction (peri-operative) and maintenance therapies and are usually used in combination. Another classification is categorising agents into pharmacological or biological – consisting of monoclonal/polyclonal anti-lymphocyte antibodies and anti-cytokine receptor antibodies.

69

Glucocorticoids Small molecule drugs Immunophilin-binding drugs Calcineurin-inhibitors Cyclophilin-binding drugs (cyclosporin, Neoral) [signal 1] FKBP12-binding drugs (tacrolimus, Prograf) [signal 1] Transduction via IL-2R, Rapamycin (sirolimus, everolimus) [signal 3] Inhibitors of nucleotide synthesis Purine synthesis (IMPDH) inhibitors (mycophenolate mofetil, Cellcept) Antimetabolites (azathioprine, 6-mercatopurine) Protein drugs Depleting antibodies (against T cells, B cells or both) Polyclonal antibody: antithymocyte globulin (ATG) [signal 1-3] Mouse monoclonal anti-CD3 antibody (OKT3) [signal 1] Non-depleting antibodies and fusion proteins Humanised/chimeric monoclonal CD25 antibody (daclizumab. basiliximab) [signal 3] Humanised monoclonal anti-CD52 antibody (Campath 1-H) Table 1: Immunosuppressive drugs Immunosuppression trends There has been a gradual evolution in immunosuppression regimes. Initial regimes utilised corticosteroids and azathioprine. Patient mortality was high with significant toxic long-term side effects of bone marrow toxicity and chronic steroid excess (diabetes, osteoporosis, hypertension). Cyclosporin, first used in 1978, revolutionised outcomes with reduced rejection and incremental improvement in patient and graft survival. Regimes involving cyclosprin, azathioprine and prednisone were adopted, with concerns at this time being mainly related to variable intestinal absorption of cyclosporin. In retrospect, even with therapeutic level monitoring, relatively high doses of cyclosporin were used. Thus significant side effects of cyclosporin were appreciated early, mainly nephrotoxcity, but also neurotoxicity, diabetogenicity, hyperlipidaemia and increased susceptibility to opportunistic infection. The introduction of tacrolimus, in the early 1990s, was a further step in immunosuppression. Initially used by Starzl in 1989, tacrolimus had a ten-fold greater in vivo potency than cyclosporin. Whilst acting in a similar mechanism (ie inhibiting calcineurin but via another immunophilin) to cyclosporin, absorption was more predictable. A more stable, microemulsified, preparation of cyclosporin, Neoral, has been compared to tacrolimus in several trials. One of these, the TMC trial3, is a well designed large trial enrolling all patients transplanted in the UK, with a clear study protocol and well defined end-points. Over 600 patients were enrolled with tacrolimus being superior in all major end-points of graft and patient survival and incidence of rejection at 12 months. Tacrolimus was found to be more diabetogenic. Whilst the relative advantage of one CNI over another remains somewhat questionable, the majority of transplant units have adopted tacrolimus based immunosuppression regimes. Additionally there has been a shift from azathioprine to mycophenolate mofetil4 usage. Induction therapies have evolved with a shift from polyclonal based ATG to monoclonal-based IL-2 receptor antagonists. The prevailing immunosuppression ethos to prevent any rejection has been refined, aiming towards a degree of ‘immune engagement’. This concept has become more attractive, particularly with the significant long-term toxicities of immunosuppression and increased incidence of neoplasia. Furthermore, immunosuppression requirements have become more flexible depending on the aetiology of disease and the particular profile of the donor. With hepatitis C the most common indication for liver transplantation in the US, most studies have focused on steroid free5 or early withdrawal of steroid regimes, by utilising induction therapies interfering or depleting CD4 T cells. Furthermore, preliminary evidence suggests cyclosporin may have an antiviral effect on

70

the HCV replicon and thus may decrease HCV recurrence post-transplant6. These modifications may have a significant impact on recurrent hepatitis C post transplant. The biological properties of other immunosuppressants may lend to their utilisation in specific clinical circumstances. For example, rapamycin may have anti-neoplastic effects that may make it attractive for use in hepatocellular cancer. The future An improved understanding of the immunobiology of transplantation will clearly provide further targets to modulate the immune response. Intriguing data from the Pittsburgh group may suggest avenues for future immunosuppressive regimes7, to induce a state of donor tolerance and allow minimal or even no immmunosuppression. Elucidating the predominant mechanisms in chronic rejection will additionally provide insight. Individualising immunosuppression via pharmacogenomics to minimise toxicities holds promise. However, only large, well-designed trials with clear endpoints will clarify the benefits (and risks) of specific immunosuppressive regimes. The ‘holy grail’ in transplantation remains the induction of donor-specific tolerance to the allograft whilst maintaining general immunocompetence. References 1. Denton MD et al. Immunosuppressive strategies in transplantation. Lancet 199;353:1083-91. 2. Halloran PF. Immunosuppressive drugs for kidney transplantation. NEJM 2004;351:2715-29. 3. O’Grady J et al. Tacrolimus versus microemulsified Ciclosporin in liver transplantation: the TMC

randomised controlled trial. Lancet 2002;360:1119-25. 4. Shapiro R et al. Immunosuppression: evolution in practice and trends, 1993-2003. Am J Transplant

2005;5:874-86. 5. Boillot O et al. Corticosteroid-free immunosuppression with tacrolimus following induction with

daclizumab: a large randomised study. Liver Transpl 2005;11:61-7. 6. Nakagawa A et al Specific inhibition of hepatitis C replication by cyclosporin A. Biochem Biophys Res

Commun 2004;313:42-7. 7. Starzl TE et al. Tolerogenic immunosuppression for organ transplantation. Lancet 2003;361:1502-10.

71

Session 8 - Post-operative care Principles and early complications Mr Derek Manas. Consultant Hepatobiliary and Liver Transplantation Surgeon, Surgical Director of Transplant Services, Freeman Hospital, Newcastle Complications of Liver Transplantation (LT) can be mechanical or non-mechanical. Postoperative mechanical complications are often related to problems with the Hepatic Artery, Portal Vein, IVC and Biliary tree. Non-mechanical complications include allograft non-function, rejection, viral reinfection by the host, recurrent hepatic malignancy, and posttransplantation lymphoproliferative disease (PTLD). Primary Non-Function (PNF) The development of PNF is a devastating complication that in its worst form resembles fulminant hepatic failure and requires urgent retransplantation. PNF can result from an unstable donor, pre-existing disease in the donor, inadequate or overly long preservation, an imperfect recipient operation, or a perioperative immunologic reaction. These factors can occur separately or in combination. PNF rates of 5% to 10% have been reported. A number of donor-specific factors have been identified which predict the development of initial poor function (IPF) or PNF in the immediate post-operative period.

• Advanced Age ( > 40 years) • Elevated Bilirubin > 20 mmol/l, • ITU stay > 3 days • Hepatic steatosis > 50% in donor graft

(However, grafts with 30-50% steatosis, are acceptable, provided the cold ischemic time of the graft is short.) In a majority of the cases of PNF, the liver produces little or no bile after reperfusion; the preexisting coagulopathy worsens (or occurs de novo), and the lactate level fails to decrease or even increases. Occasionally, the liver function is good or fair during the first 24 hours or so, only to deteriorate rapidly after. Postoperatively, the patient is either comatose or extremely agitated, and the bile output is minimal (if a T-tube is present), with mucous, greenish bile. The urine output usually decreases, with a concomitant increase in blood urea and creatinine. The coagulation parameters are abnormal; the liver enzymes are very high, and the bilirubin increases rapidly. If the situation does not improve within 24 to 36 hours, the patient's only chance for survival lies with emergency re-transplantation. Recently, repeated sessions of plasmapheresis have been used with some success in buying time to allow the liver to return to normal (unpublished material). The morbidity and mortality of this complication are high. Survival following re-transplantation for PNF is only half of that seen in the general liver transplant population. Overall, it is the most lethal of all possible complications of liver transplantation.

• Attentive selection of the donor and careful management, • perfect harvesting technique, • optimal preservation, and • uncomplicated recipient operation are each an essential factor in minimizing the

incidence of PNF. Prostaglandins have also been demonstrated to have cytoprotective effects in a number of experimental models of acute liver injury. In 3 studies, the routine use of PGE has been associated with PNF rates of less than 5%. Moreover, in a prospective randomized trial of PGE1 (n=78) versus placebo (n=82), the PGE group had improved renal function and a shorter ICU and total hospitalization stay. There was, however, no difference in the incidence of PNF or acute rejection.

72

The incidence of adult graft loss from PNF within the first week following LT in our centre is less than 2%. The reasons for the reduction in IPN and PNF may include improved donor management, a policy of keeping the cold ischemic time short and caution in the use of steatotic livers. Technical Complications Hepatic artery thrombosis (HAT) HAT is one of the most common and most disastrous arterial complications. It occurs in 4-12% of adult LT patients and in 9-42% of paediatric LT patients. Causes include; allograft rejection, hepatic artery kinking due to vascular redundancy, underlying HA Stenosis(HAS), and technical problems at the anastomosis. Doppler ultrasonography (US) has a sensitivity of approximately 90% in detecting HAT. Doppler findings include: absent flow at the porta hepatis with absent intrahepatic arterial flow and absent flow in the donor hepatic artery with abnormal intrahepatic flow. Collateralized intrahepatic flow can be differentiated from normal flow by its tardus parvus Doppler waveform morphology. Tardus parvus can be diagnosed when the intrahepatic arterial resistive index (RI) is less than 0.5 and the systolic acceleration time (SAT) is 0.08 seconds or longer. Without aggressive diagnosis and treatment, HAT is associated with a mortality rate of greater than 80%. Unfortunately, percutaneous thrombolytic therapy of HAT is problematic because it often occurs soon after LT. Thrombolysis and angioplasty of underlying HAS have been successful in a few patients; however, HAT usually requires urgent retransplantation for patient survival. Hepatic artery stenosis (HAS) HAS occurs in approximately 11-13% of LTs. HAS can be caused by clamp injury, intimal injury due to perfusion catheters, anastomotic ischemia due to a disrupted vasa vasorum, and rejection. The consequences of hemodynamically significant HAS (>50% stenosis) are essentially identical to those of HAT. Although a few stenoses have been corrected with balloon angioplasty, HAS, usually requires operative vascular reconstruction or re-transplantation. Doppler US is used to screen for HAS. If the surgical anastomosis can be interrogated directly, stenosis is diagnosed if peak systolic velocity exceeds 200-300 cm/s. In most patients, direct Doppler evaluation of hepatic artery anastomosis is not possible because the donor-recipient arterial anastomosis is tortuous, and because it is usually obscured by overlying bowel gas. However, similar to HAT, hemodynamically significant HAS results in intrahepatic arterial tardus parvus waveform morphology. By using a RI of less than 0.5 in conjunction with a SAT 0.08 seconds or longer yields a sensitivity of 45% for the detection of HAS. The sensitivity increases to 70% if 2 of the following 3 criteria are abnormal: (1) RI less than 0.5, (2) SAT 0.08 second or longer, and (3) peak systolic velocity greater than 200 cm/s at the anastomotic site. Because the specificity of Doppler is only 64% in detecting marked arterial disease (ie, HAT or hemodynamically significant HAS), angiography usually is required to confirm the diagnosis. At the time of diagnostic angiography, a decision can be made regarding the effectiveness of percutaneous correction of documented HAS. Pseudoaneurysm Extrahepatic pseudoaneurysms occur after LT in approximately 2% of patients. The most common location of pseudoaneurysm formation is at the donor-recipient arterial anastomosis. A less common site is found at the ligation of the donor gastroduodenal artery. Pseudoaneurysms can be caused by infection or technical failure. Because of the potential for rupture and life-threatening hemorrhage, pseudoaneurysms need to be treated promptly. To ensure maximum arterial flow, most extrahepatic lesions are repaired surgically.

73

Intrahepatic pseudoaneurysms usually result from percutaneous biopsy, biliary procedures, or infection. Most intrahepatic lesions can be treated by transcatheter embolization or stent placement. Biliary complications Biliary complications occur in approximately 13-19% of adults following LT. With adult LTs, a primary, end-to-end choledochocholedochostomy is performed after donor cholecystectomy. In most cases of LDLT, a biliary-enteric hepatojejunostomy is used to reconstitute bile drainage. T tubes no longer are used for either type of biliary anastomosis; therefore, T-tube complications are no longer a problem. Bile duct strictures can be anastomotic or nonanastomotic. Some anastomotic strictures result from technical difficulties or are caused by fibrosis and scarring that can be associated with bile leaks. Anastomotic stenosis can be related to marginal blood supply of the cut ends of the donor and recipient ducts; stricturing results from anastomotic ischemia. Nonanastomotic strictures can be caused by HAS, HAT, prolonged cold ischemia time, rejection, cytomegalovirus infection, intraductal sludge and stone formation, and recurrent primary sclerosing cholangitis in the allograft. Nonischemic strictures can develop over time and present with deterioration of liver function tests and bouts of cholangitis. Dominant biliary strictures can be detected using US, CT, MRI, cholangiography (MRCP), percutaneous transhepatic cholangiography (PTC), and endoscopic retrograde cholangiopancreatography (ERCP). The choices for mechanical revision of a bile duct stricture include PTC, ERCP, or repeat operation. Ischemic bile duct strictures can present with rapid onset of allograft dysfunction and sepsis. Unlike the native biliary tree that is supplied by collaterals from the gastroduodenal artery, the harvested donor bile duct and the anastomotic end of the recipient duct are solely dependent on the hepatic artery. Compromise of the hepatic artery by HAS or HAT results in abrupt ischemia and necrosis of the biliary epithelium. This causes strictures, disruption of the ducts, bile leaks, bilomas, infected bilomas, and abscesses. Timely angiographic correction of HAS or HAT occasionally can result in graft salvage. Percutaneous drainage of associated fluid collections is often only a temporizing measure because severe underlying biliary tree damage has usually occurred. Patients with sepsis who have severe biliary and hepatic damage from HAS or HAT require retransplantation for ultimate survival. Complications involving the portal vein Complications involving the portal vein occur in approximately 1-13% of LT recipients. Portal vein stenosis (PVS) usually occurs at the donor-recipient anastomosis. Portal vein thrombosis (PVT) often involves the main extrahepatic segment. Causes of PVS and PVT include surgical difficulties, thrombus formation from the portal venous bypass cannula, excessive vessel redundancy, and hypercoagulability. PVS can lead to PVT. PVT and hemodynamically significant PVS cause graft dysfunction and portal hypertension. US, CT, MRI, and magnetic resonance venography (MRV) can be used to detect both PVS and PVT.

• An abrupt 3- to 4-fold increase in velocity occurs within the portal vein on spectral Doppler US and aliasing on color Doppler US reflects turbulent flow associated with PVS. (PVS can be successfully corrected in many patients by using percutaneous transluminal balloon angioplasty.)

Hemodynamically significant PVS must be distinguished from portal vein pseudostenosis, which results when the recipient portal vein is somewhat larger than the donor portal vein. The

74

difference in caliber causes increased velocity and turbulence at the anastomosis that are not physiologically significant. If portal vein narrowing is associated with a velocity increase of less than 3-4 fold on spectral Doppler analysis, the narrowing likely is pseudostenosis. PVT also can be diagnosed noninvasively (US, CT, MRI).

• Percutaneous intervention can be successful in restoring venous patency. Techniques include infusion of thrombolytics (eg, rTPA), angioplasty of associated PVS (if present), mechanical thrombectomy, and catheter embolization of competing collaterals.

• Surgical options in PVT include thrombectomy, segmental PV resection, placement of a jump graft, or creation of a portosystemic shunt. PVT can necessitate retransplantation.

• Occasionally, PVT is detected in patients with normal allograft function and without portal hypertension. In these patients, sufficient hepatopetal collateralization has developed to maintain adequate venous inflow.

Complications involving the IVC and associated hepatic vein Posttransplantation IVC stenosis or thrombosis is an uncommon complication that occurs in approximately 1-4% of cases. IVC stenosis and thrombosis tend to occur at the superior and inferior caval anastomoses. Causes include:

• Technical problems, • IVC compression due to a fluid collection or hematoma, and • mass effect from hepatic regeneration.

Patients with suprahepatic lesions tend to present with signs and symptoms of Budd-Chiari syndrome; (pleural effusion, hepatic enlargement, and ascites). IVC obstruction can cause lower-extremity edema, which can be a prominent feature of infrahepatic stenosis. Clinically, leg swelling helps distinguish a suprahepatic caval complication from a portal vein problem, both of which are associated with features of portal hypertension.

US is the imaging study of choice for screening for IVC complications. A stenosis is detected as a gray-scale narrowing with a 3- to 4-fold increase in velocity on spectral Doppler analysis and associated color Doppler aliasing. Indirect findings of suprahepatic IVC stenosis or thrombosis include distension of the hepatic veins with dampening of the hepatic venous spectral Doppler waveform and loss of its usual phasicity. Hemodynamically significant IVC stenosis can be differentiated from pseudostenosis by the presence of features of Budd-Chiari syndrome. Stenosis and thrombosis of the IVC can also be detected by using CT scans. Inferior venacavographic results can confirm stenosis and thrombosis. Pressure gradient measurements can distinguish physiologically significant lesions from pseudostenoses. Balloon angioplasty and stent placement can be used to correct an IVC stenosis. Isolated complications of the hepatic vein are rare. Strictures at the suprahepatic caval anastomosis can involve the confluence of the hepatic vein. Like caval strictures, lesions of this vein are amenable to percutaneous transjugular angioplasty and stent placement. Postop collections Postoperative fluid collections are common and can be both supradiaphragmatic and infradiaphragmatic. In most patients, small parahepatic hematomas are present after surgery. Small hematomas are frequently noted on postoperative sonograms obtained within 24 hours to 1-2 weeks of transplantation. Hematomas tend to be located in the gallbladder fossa and the hepatorenal space. A large volume of ascites can form after liver transplantation. Bilomas are often associated with serious postoperative complications. They can reflect dehiscence of a choledochocholedochostomy or biliary-enteric anastomosis. Bile leaks require

75

interventional (image-guided aspiration) or surgical correction. Infected bilomas require percutaneous 8F catheter drainage, which is often performed in conjunction with other biliary and arterial interventions if underlying bile duct or vascular complications are present. In patients with irreversible graft damage, percutaneous control of bilomas, particularly infected ones, can be a temporizing measure while retransplantation is pending. In the setting of postoperative fever and sepsis, some hematomas must be aspirated to determine if superinfection is present. Catheter placement in an uninfected hematoma is avoided because semisolid subacute blood is not amenable to catheter drainage and because superinfection (by leaving an indwelling catheter) of an otherwise sterile collection is to be avoided. Hepatic lymphatics are completely disrupted in the recipient. In the early period after transplantation, periportal fluid can originate in an accumulation of lymph, which is often reabsorbed by the peritoneum. Postoperative pyogenic abscesses in the abdomen or retroperitoneum can be treated percutaneously. Acute rejection Acute (cellular) hepatic allograft rejection can occur in as many as 40% of patients during the first 3 months posttransplant. Acute rejection normally occurs 7-14 days after operation but can occur earlier or much later.

• Hyperacute rejection of the liver, comparable to that observed in kidney transplantation, is controversial and difficult to diagnose.

• Early accelerated rejection certainly occurs. Liver biopsy may be required to distinguish between rejection and viral infection. Rejection is most commonly manifested by malaise, fever, graft enlargement, and diminished graft function, with a rise in bilirubin and transaminase levels. Graft biopsy should be performed, if safe, to document rejection. Rejection episodes are managed sequentially by pulse steroids, ATG, and/or the use of CellCept, tacrolimus switch (if patient was on cyclosporine), or the addition of Rapamycin. Retransplantation is the last resort when therapy fails and the patient develops hepatic failure.

76

Session 8 - Post-operative care Late complications and outcomes Dr John O’Grady. Consultant Hepatologist, King’s College Hospital, London