Professor M. R. Lucey, University of Wisconsin-Madison School of Medicine, H6/516 CSC, 600 Highland Avenue, Madison, WI 53792, USA. E-mail: email@example.com
Alcoholic liver disease, including acute alcoholic hepatitis and alcoholic cirrhosis, is a major cause of morbidity and mortality in the Western world. Abstinence remains the cornerstone of management of all forms of alcoholic liver disease. Recent research, which has elucidated the mechanisms of alcohol-induced liver injury, offers the prospect of advances in the management of alcoholic liver disease. We review the most recent data on the efficacy of treatment of acute alcoholic injury, including nutritional support, corticosteroids, anti-inflammatory agents and antioxidants, and agents that are directed against the progression to fibrosis, such as colchicines, propylthiouracil and antioxidants. Although these therapies offer a tantalizing glimpse into a future that may include therapies that directly alter the process of injury and repair in the liver, none has been shown consistently to improve the course of alcoholic liver damage. Consequently, liver transplantation remains an ultimate option for selected patients with liver failure due to chronic alcoholic liver damage.
Alcohol intake remains the most important cause of cirrhosis in the Western world.1 The causal association between alcohol intake and the development of alcoholic liver disease has been well demonstrated.2,3 Nevertheless, it remains a conundrum that only a small proportion of heavy drinkers develop the most advanced form of the disease (liver cirrhosis).3 It is presumed, therefore, that other factors, such as gender,3,4 genetic background5–7 and additional environmental influences, particularly chronic viral infection,8 play a role in the genesis of alcoholic liver disease. The effects of both the duration and quantity of alcohol consumption on liver disease have also been studied extensively.3,9 In this review, we concentrate on therapies for alcoholic liver disease and attempt to link therapeutic options to recent advances in our understanding of the pathogenesis. We use the terms alcohol and ethanol interchangeably.
What is alcoholic liver disease?
The spectrum of liver damage caused by alcohol is not uniform. For descriptive purposes, three main histological stages are described as if they constitute separate and definitive lesions: steatosis, acute alcoholic hepatitis and cirrhosis.10 In reality, these entities overlap, and it is difficult to find them isolated in their pure histopathological form. Steatosis is a predictable histological abnormality which develops in many heavy drinkers. It results from the redox imbalance generated by the metabolism of ethanol to acetate. Alcoholic steatosis completely reverses within several weeks of discontinuation of alcohol intake.11 Acute alcoholic hepatitis is characterized by hepatocellular injury with associated inflammation and fibrosis. Like steatosis, acute alcoholic hepatitis usually improves with abstinence. When alcohol use continues unabated, inflammation triggers fibrogenesis and, over time, collagen is deposited in a characteristic perivenular and pericellular distribution. Approximately 40% of patients with this lesion (zone 3 fibrosis extending in a lattice-like peri-hepatocyte network) will develop cirrhosis within 5 years.12 Severe acute alcoholic hepatitis has a poor outcome with standard supportive management. For example, the mortality rate of patients with severe alcoholic hepatitis in two prospective studies was 35% and 46%, respectively.13,14 The addition of acute renal failure worsens the prognosis further.15
Mechanisms of alcohol-related liver injury
Acetaldehyde is the most important metabolite of ethanol leading to liver damage.16 The immune response to acetaldehyde delivery, as well as the hypermetabolic state due to the increased oxygen consumption secondary to an altered NADH/NAD ratio (NAD, nicotinamide adenine dinucleotide), may be important steps in the genesis of more severe liver damage.
Oxidative stress is defined as an imbalance favouring pro-oxidants and disfavouring antioxidants, leading to tissue damage. We know that the administration of alcohol to healthy humans leads to the generation of reactive oxygen species (products of lipid peroxidation), and that acute alcoholic hepatitis is a condition of marked oxidative stress.17 In alcoholic liver disease, oxidative stress is caused by pro-oxidant formation, inadequate intake of antioxidants, antioxidant depletion and alcohol-mediated inhibition of glutathione synthesis. The pro-oxidants involved in alcohol-induced liver disease are reactive oxygen species and reactive nitrogen species. The sources of these antioxidants include mitochondria and cytochrome P450 2E1 in hepatocytes, and NAD(p)H oxidases and the inducible form of nitric oxide synthase (iNOS) in inflammatory cells. Recent studies have shown that iNOS knockout mice are protected completely against oxidative stress caused by ethanol.18 The mechanisms of cellular injury due to oxidative stress are complex. Potential mechanisms include chemical modification of biological molecules and further stimulation of the host immune response, including the release of cytokines, with resultant inflammatory and fibrotic responses, direct activation of stellate cells and, finally, inhibition of the synthesis of S-adenosyl methionine.
Endotoxin is another important element in the pathogenesis of alcohol-induced liver damage.19 Circulating endotoxin levels increase after alcohol intake, and endotoxins may trigger both cytokine release and oxidative stress.
Fibrogenesis within the liver is a consequence of the activation of collagen-producing stellate cells. The pathway leading to the activation of stellate cells is not certain, and may include a number of simultaneous processes, such as direct injury from reactive oxygen species or through intermediate steps which include the expression of interleukins (IL), such as tumour necrosis factor (TNF), IL1, IL6, IL8 and the transforming growth factors.20 The profile of cytokines released in response to alcohol, lipid peroxidation, circulating endotoxin and Kupffer cell activation is influenced by genetic polymorphisms.5,21 These observations are of potential importance in the discovery of therapies for alcoholic liver disease by antioxidant agents, or by blocking the release of cytokines and growth factors precipitating collagen deposition (Figure 1).
Abstinence and prognosis
The prognosis of alcoholic liver disease is strikingly related to abstinence. The 5-year survival of patients with clinically compensated alcoholic cirrhosis is about 90%, but declines to 70% if the patient continues to drink. When a patient with decompensated cirrhosis continues drinking, the chance of living 5 years is 30% at best.12,22 Moreover, abstinence plays an important role in the reversibility of steatosis, acute alcoholic hepatitis, lipid peroxidation, inflammation and collagen deposition. Indeed, retrospective longitudinal observational studies have suggested that ‘pure’ alcoholic steatosis reverses completely with abstinence.23 Consequently, abstinence from alcohol is a cornerstone of many different approaches to therapy. In the following, we review the different therapeutic options available for both acute and chronic alcoholic liver disease.
Therapy for alcoholism
Specialists in addiction medicine consider alcoholism to be a chronic disorder of resolution and relapse.24 Consequently, addiction specialists have tended to consider control of the drinking behaviour as a more realistic goal than an absolutist formulation of complete abstinence. This point of view is well described by Fuller: ‘in alcoholism, as in diabetes, a substantial partial response is better than no response’;25 in other words, euglycaemia 80% of the time is better than euglycaemia 20% of the time. Both brief and extended psychological interventions and pharmacological therapies directed at controlling drinking have been found to have beneficial effects on harmful drinking, when benefit is defined as a qualitative reduction in alcohol consumption. Unfortunately, as many as 70% of subjects in these studies return to harmful drinking within 6 months of completing a treatment protocol, and even roughly 40% of alcoholic subjects who establish 2 years of complete abstinence will relapse.26
The effect of structured protocols designed to limit alcohol intake (whether psychological or pharmacological) on the histological features of alcoholic liver disease has not been studied. Nevertheless, typical studies of treatment for alcoholic liver disease (as opposed to studies of treatment of alcoholism) take complete abstinence from alcohol as their starting point. In this review, unless stated otherwise, all studies of treatment for alcoholic liver disease to which we refer were undertaken in the setting of abstinence from alcohol, and we will not reiterate the need for abstinence.
Acute alcoholic hepatitis
The clinical syndrome of acute alcoholic hepatitis includes a wide spectrum of liver damage, from the asymptomatic patient with mild inflammation on liver biopsy to the severely ill patient with fever, cholestasis, coagulopathy and leucocytosis. In the majority of mild cases, avoiding alcohol intake alone allows the clinical picture to resolve.27,28 We review the numerous therapies investigated in an attempt to reverse the process of acute alcoholic hepatitis. These include nutritional supplements, corticosteroids, TNF modulation, propylthiouracil and antioxidants (Figure 2).
The rationale for nutritional supplementation is related to the fact that most patients with acute alcoholic hepatitis have some degree of malnutrition. Furthermore, the mortality as a result of acute alcoholic hepatitis is correlated closely with the severity of protein-calorie malnutrition. Thus, in the study by Mendenhall et al. of men with acute alcoholic hepatitis in the Veterans Administration system, the 30-day mortality was increased from 2% in patients with mild malnutrition to 52% in those with severe malnutrition.29
Malnourishment of patients with alcoholic liver disease is of multi-factorial origin, due to a combination of poor intake of nutrients and decreased intestinal assimilation and hepatic storage. The mechanisms by which malnutrition enhances liver damage and worsens mortality in patients with alcoholic liver disease are unknown.30 It has been suggested that protein-calorie deficiency could enhance the toxicity of alcohol, in part, through the influence of nutritional status on the integrity of the immune system and the capacity to respond to infection.30 An additional explanation implicates the dependence of hepatic regeneration, which is the process of recovery of the damaged liver, on protein synthesis.30
Nutritional therapy for alcoholic liver disease was the subject of a recent review in Alimentary Pharmacology and Therapeutics.31 The value of treatment with parenteral/enteral supplementation in patients with acute alcoholic hepatitis was studied in 12 prospective clinical trials up to the early 1990s, which were reviewed by Schenker and Halff in 1993.30 Their predominant conclusion was that nutritional support improves nutritional status and abnormal liver tests, but does not decrease early mortality, in patients with acute alcoholic hepatitis. Nevertheless, a striking effect on mortality was found by Cabréet al., who reported that nutritional support was associated with an improvement in serum albumin and an important reduction in mortality in the group of patients treated.32 More recently, a randomized study was designed by the same group to compare the efficacy and safety of prednisolone and total enteral nutrition in the treatment of severe acute alcoholic hepatitis.33 The results of this study suggested that total enteral nutrition was as effective as corticosteroids in the treatment of this condition. Early deaths were more frequent with total enteral nutrition, whereas late deaths were more frequent with corticosteroids. Unfortunately, this trial is difficult to interpret as a result of the dubious standing of corticosteroids as standard therapy for acute alcoholic hepatitis (see below). It remains uncertain whether a combination of nutrients and corticosteroids would be more efficacious than either alone, or when compared with no additional treatment. Thus, although nutritional supplements are reasonable for any severely malnourished patient, their efficacy in the management of acute alcoholic hepatitis must be considered to be ‘unproven’.
The rationale underlying the use of corticosteroids in acute alcoholic hepatitis is to ameliorate the characteristic inflammatory response which accompanies this disorder. These agents have well-known effects on the immune response, and reduce cytokine production, suppress the formation of acetaldehyde adducts and inhibit the production of collagen.
Corticosteroids have been studied in several randomized double-blind trials,13,34,35 and have been subject to a number of meta-analyses.36,37 Despite this intense interest, neither the source studies nor the meta-analyses have reached a consensus conclusion regarding the efficacy of corticosteroids in the treatment of acute alcoholic hepatitis. The most recent study of Mathurin et al. re-analysed the original data from three randomized trials that included patients with severe illness (102 patients were treated with placebo and 113 were treated with steroids).36 The 28-day survival was higher in the corticosteroid group, so that, for every five patients treated, one death was avoided. This benefit was consistent during the first year of treatment, but disappeared at 2 years of follow-up. Despite the apparent benefit of corticosteroids in severe illness, many clinicians remain sceptical about their use. One explanation for this phenomenon is that many potential candidates for corticosteroid therapy have one or more contraindications to treatment, such as coincidental bacterial infections, gastrointestinal bleeding or renal failure. An additional reason may be the conflicting data from trials, including meta-analyses, which have failed to demonstrate a benefit of corticosteroids.37
Anabolic-androgenic steroids have not been shown to have lasting benefit in acute alcoholic hepatitis. Mendenhall et al. compared the effects of prednisolone, oxandrolone and placebo in a controlled clinical trial, without demonstrating a sustained survival benefit to the recipients of oxandrolone.38,39
Oxpentifylline (pentoxifylline) is a phosphodiesterase inhibitor used in the treatment of intermittent claudication. In addition to improving the responses of red blood cells to deforming stress, oxpentifylline (pentoxifylline) inhibits the output of TNF-α by modulating the transcription of the TNF-α gene.40 Recently, Akriviadis et al. reported a randomized, placebo-controlled clinical trial of oxpentifylline (pentoxifylline) involving 101 patients with severe alcoholic hepatitis (Maddrey discriminant factor ≥ 32) who entered a 4-week double-blind randomized trial of oxpentifylline (pentoxifylline) (400 mg orally three times daily) vs. placebo.14 Twelve (24.5%) of the 49 patients who received oxpentifylline (pentoxifylline) and 24 (46.1%) of the 52 patients who received placebo died during the initial hospitalization. Hepatorenal syndrome was the cause of death in six of 12 cases (50%) in the treated group and 22 of 24 cases (91.7%) in the placebo group. Thus, the benefit of oxpentifylline (pentoxifylline) in treating acute alcoholic hepatitis appears to be related to a significant decrease in the risk of developing hepatorenal syndrome. Although these data provide evidence of a dramatic benefit, they await the validation of independent randomized controlled clinical trials.
Acute alcoholic hepatitis is a potentially important indication for the use of direct anti-TNF-α agents, although the timing of their administration in relation to the last use of alcohol and the provision of appropriate prophylaxis against infection may be critical practical issues. Up to now, only two human studies utilizing direct anti-TNF therapy in acute alcoholic hepatitis have been reported. In one, the monoclonal antibody anti-TNF-α (infliximab) was compared with placebo in patients with alcoholic hepatitis receiving corticosteroids.41 Although the histology did not improve, the 28-day Maddrey score improved significantly in the group receiving anti-TNF. Unfortunately, mortality data were not reported and, in their absence, the interpretation of the Maddrey score as a surrogate end-point is questionable. In another recent study, open-label uncontrolled use of infliximab was described in 12 patients with acute alcoholic hepatitis. The authors reported that more than 80% of patients survived for 15 months from the initiation of therapy. Although TNF-α messenger RNA expression in the liver did not change, the expression of IL-8, a cytokine regulated mainly by TNF-α, was almost absent on day 28.42 Both studies provide support for randomized controlled trials of appropriate power to determine whether anti-TNF-α treatment is efficacious in acute alcoholic hepatitis.
Therapies for acute alcoholic hepatitis and/or alcoholic cirrhosis
Despite the evidence that alcoholic liver disease is associated with enhanced oxidative stress, this research has not been translated into clinical therapy.17,18 Studies published in abstract form have failed to show that different kinds of antioxidants (when compared with corticosteroids or placebo) confer any benefit in the treatment of acute alcoholic hepatitis.43,44
Turning to the use of antioxidants in the setting of cirrhosis, there have been several trials using silymarin (milk thistle) as antioxidant. In a double-blind, prospective, randomized clinical trial involving 170 patients with cirrhosis (91 patients with alcoholic cirrhosis), 87 patients were treated with 140 mg silymarin three times daily and 83 patients received placebo. The mean observation period was 41 months. Fifty of 87 (58%) silymarin-treated patients, as opposed to 32 of 83 (39%) placebo-treated patients, survived for 4 years or longer. This difference, which was significant at the P < 0.05 level, suggested a benefit in the silymarin group.45 In contrast, Pares et al. found no benefit of silymarin administered to patients with alcoholic cirrhosis, despite a greater number of patients and 5 years of follow-up.46
Another goal of antioxidant therapy involves the enhancement of the cell's own antioxidant capacity, which is the rationale for external supplementation of S-adenosyl methionine. A randomized, double-blind trial of S-adenosyl methionine or placebo was performed in 123 alcoholic cirrhotic patients. The overall mortality/liver transplantation at the end of the trial decreased from 30% in the placebo group to 16% in the treated group, although this difference was not statistically significant.47 In summary, there remains a lack of compelling evidence that antioxidants, such as silymarin, or membrane stabilizers, such as S-adenosyl methionine, exert a substantial salutary effect in patients with cirrhosis due to alcohol-mediated liver injury.
The rationale underlying the use of propylthiouracil in the treatment of alcoholic liver disease rests on the observation that the most severe alcohol-induced damage is often in the perivenular area (zone 3), thereby resembling ischaemic injury. The putative mechanism of propylthiouracil to produce a benefit in this setting is by reducing hepatic oxygen consumption by hepatocytes. An initial trial by Halle et al. found no benefit of propylthiouracil in the treatment of alcoholic hepatitis.48 Some years later, Orrego et al. observed, in a larger cohort, that the administration of propylthiouracil reduced the mortality due to alcoholic liver disease.49 This randomized placebo-controlled clinical trial was remarkable for the care taken to monitor alcohol consumption by the study participants and thereby control for the greatest confounding factor complicating the interpretation of investigations of treatment for alcoholic liver disease. Despite such admirable study design and the demonstration of a statistically impressive benefit in propylthiouracil recipients compared with controls, propylthiouracil has not gained currency as a treatment for alcoholic liver disease. Moreover, a recent meta-analysis, including data from 710 patients with several forms of alcoholic liver disease, could not demonstrate any benefit of propylthiouracil on mortality, liver-related mortality, liver complications or liver histology.50
Colchicine is an inhibitor of collagen synthesis. Unfortunately, results regarding the use of colchicine in the treatment of liver cirrhosis are conflicting. The most potent data in support of its use come from the study by Kershenobich et al.51 In their study of colchicine for the treatment of all forms of cirrhosis, the 5-year survival rate was 74% in the colchicine-treated group compared with 34% in the placebo group. The 10-year survival rate was twice as great for the colchicine-treated group. This study was confounded by the high dropout rate. Furthermore, a recent meta-analysis conducted by Rambaldi and Gluud, combining the results of 14 randomized clinical trials involving 1150 patients, demonstrated no significant beneficial effects of colchicine on mortality, liver-related mortality, complications of liver failure, liver biochemistry, liver histology or alcohol consumption.52 Conversely, colchicine was associated with a significantly increased risk of adverse events.
Encouraging results have been obtained with some ‘super’ nutrients in the treatment of alcoholic liver disease. Phosphatidylcholine, purified from polyunsaturated lecithin, was discovered to oppose ethanol-induced fibrosis by decreasing the activation of stellate cells to transitional cells, and possibly also by stimulating collagenase activity.53 Recent evidence has suggested that phosphatidylcholine may have an effect on alcohol-induced liver damage by decreasing the activities of cytochrome P450 2E1, an oxidative enzyme implicated in alcohol metabolism, and by modulating TNF-α.54 Recently, a randomized, prospective, double-blind, clinical trial comparing phosphatidylcholine with placebo was conducted in 789 male alcoholics at 20 Veterans Affairs Medical Centers.55 All subjects demonstrated hepatic fibrosis that did not meet the criteria for cirrhosis on liver biopsy prior to the initiation of the study. Treatment was maintained for 2 years. No histological benefit was observed in the treated group compared with the placebo group.
In 1988, Starzlet al. reported a series of patients with alcoholic liver disease who underwent liver transplantation and who achieved reasonable survival.56 Since then, alcoholic liver disease has become the first or second most common cause of liver transplantation in the USA and Europe. Alcoholic relapse poses many problems, which frustrate both the pre-transplant selection and clinical management of patients after liver transplant. Between 20 and 50% of patients who receive a liver transplant because of end-stage alcoholic liver disease acknowledge some alcohol use in the first 5 years after liver transplantation, whereas 10–15% will resume heavy drinking.57,58 Despite the potential harmful effects of alcoholic relapse on the health of the allograft and recipient, the 5- and 10-year patient survival rates after liver transplant are at least as good as those for recipients transplanted for other indications.58
We have reviewed, in a separate publication, the evaluation of patients with alcoholic liver disease for liver transplantation and the determination of the optimal time for transplantation.59 The ideal time for the transplantation of patients with alcoholic liver disease is based on a balance between the outcome from medical management and the outcome of liver transplantation. The minimal criteria, published in 1997, for placement on the transplant waiting list was Child–Pugh class B.60 Poynard et al. studied the efficacy of liver transplantation for alcoholic cirrhosis using matched and simulated controls, and found a significant benefit in patients with advanced cirrhosis (Child class C; 11–15 points on the Child–Pugh scoring system) only.61 For patients with medium risk (Child class B; 8–10 points on the Child–Pugh scoring system), there was no significant survival difference in comparison with controls. These data suggest that liver transplantation should be confined to patients with advanced alcoholic liver failure.
A required interval of abstinence from alcohol, usually set at 6 months, is a typical component of the assessment for liver transplantation of patients with alcoholic liver disease. One justification of the period of abstinence before liver transplant has been to allow patients with acute alcoholic hepatitis the opportunity to recover with medical management. Up to now, patients with recent alcohol use have been precluded from liver transplantation. We lack good data to determine whether some of these patients would benefit from transplantation.
Santiago Tome received a grant from Carlos III Institute-Spanish Health Department (BAE 02/5024).