Telephone: 650-498-5691; FAX: 650-498-5692
Liver transplantation for alcoholic liver disease: Current concepts and length of sobriety
Article first published online: 20 SEP 2004
Copyright © 2004 American Association for the Study of Liver Diseases
Supplement: Summary Report of a National Conference: Evolving Concepts in Liver Allocation in the MELD and PELD Era From the October 29, 2004 AASLD/ILTS Transplant Course
Volume 10, Issue Supplement 10S, pages S31–S38, October 2004
How to Cite
Lim, J. K. and Keeffe, E. B. (2004), Liver transplantation for alcoholic liver disease: Current concepts and length of sobriety. Liver Transpl, 10: S31–S38. doi: 10.1002/lt.20267
- Issue published online: 21 SEP 2004
- Article first published online: 20 SEP 2004
- 1The 1-year and 5-year actuarial survival rates following liver transplantation for patients with alcoholic liver disease are 82% and 68%, respectively, in the United States and 85% and 70%, respectively, in Europe. These survival rates are similar to the outcomes of patients who undergo transplantation for other types of chronic liver disease.
- 2Posttransplant improvements in health-related quality of life are similar in patients who undergo transplantation for alcoholic liver disease compared to those who undergo transplantation for other causes of end-stage liver disease.
- 3Approximately 20% of patients who undergo transplantation for alcoholic liver disease use alcohol posttransplant, with one-third of these individuals exhibiting repetitive or heavy drinking. Surprisingly, little evidence exists to document a significant detrimental effect on graft or patient survival associated with resumption of drinking.
- 4There are few reliable predictors of relapse in alcoholic patients after liver transplantation. Although not supported by all studies, abstinence of fewer than 6 months prior to transplantation may be a reasonable predictor of recidivism and is widely employed as a criterion for listing for liver transplantation. There are no good data to determine if some patients with sobriety fewer than 6 months might benefit from liver transplantation. (Liver Transpl 2004;10:S31–S38.)
Liver transplantation represents a well-accepted standard of care for end-stage liver disease and is associated with excellent long-term outcomes. One-year survival rates average 85%, and late graft loss from chronic rejection is uncommon.1 Alcoholic liver disease (ALD) is a leading cause of end-stage liver disease in the United States and results in 7.9 deaths per 100,000 population.2 ALD is a leading indication for liver transplantation, in part because short-term survival rates are comparable to patients who undergo transplantation for other conditions, and the rate of alcohol relapse is acceptably low.3–5 ALD is the second most common indication for liver transplantation,6 accounting for greater than 20% of all cases.7
Despite the rise of living donor programs and significant public efforts in recent years to increase organ donation, donor grafts remain in short supply. Currently, 5% to 15% of patients on the waiting list die before a liver becomes available.8 As of June 2004, nearly 17,500 individuals with end-stage liver disease were on the transplant waiting list.9 In 2003, 5,671 liver transplants were performed, including 321 grafts from living donors.10
This shortage has raised questions about the degree of priority that programs should give to patients with ALD, a condition some believe to be a predictable consequence of an individual's actions. Not all patients who may potentially benefit from liver transplantation can receive a graft, and therefore allocation of organs must be rationed in some manner. In many programs, this has involved systematic barriers to access for patients with ALD, often in the form of a 6-month or longer period of abstinence requirement, agreement to an alcohol abstinence contract, randomized testing of urine and blood, and mandated alcohol rehabilitation. Despite their frequent use, the United Network for Organ Sharing (UNOS) has never adopted these measures as a formal recommendation.11
Questions regarding transplantation for ALD
In the present article, we seek to address the following questions:
- 1Is liver transplantation effective in patients with ALD versus those with nonalcoholic liver disease (non-ALD)?
- 2Does liver transplantation result in a significant improvement in health-related quality of life (HRQOL)?
- 3How often do patients with ALD experience alcohol relapse following liver transplantation?
- 4Do alcohol relapses result in clinically significant differences in outcome after liver transplantation?
- 5Are there reliable ways to predict recidivism in patients prior to liver transplantation?
- 6Are 6- to 12-month mandated abstinence requirements justified by the evidence?
- 7What criteria should transplant programs use to select patients with ALD for liver transplantation?
Is Liver Transplantation Effective in Patients With ALD Versus Those With Non-ALD?
Current evidence supports ALD as an appropriate indication for liver transplantation, as highlighted by well-documented data suggesting that patients with ALD have similar graft and patient survival outcomes as patients who undergo transplantation for non-ALD5, 12 and incur no increase in cost or resource utilization.13, 14 The cumulative experience of multiple transplant centers have confirmed that the 1-year and 5-year actuarial survival rates following liver transplantation are similar for patients with ALD and non-ALD in both US and European liver registries. The overall 1-year and 5-year patient survival rates of patients with ALD were 82% and 68%, respectively, in the UNOS database, and 85% and 70%, respectively, in the European database.7, 8
Table 1 presents UNOS data on 1-year and 5-year patient survival following liver transplantation and is stratified by recipient diagnosis. Results show that outcomes in patients with ALD were comparable to those in patients with other diagnoses and superior to those in patients with chronic viral hepatitis. Further analysis of the UNOS data by Belle et al. revealed that the underlying causes of graft dysfunction and loss were also similar between patients with ALD and those with non-ALD, suggesting minimal differences in both graft and patient outcomes.5
|Diagnosis||1-Year Survival (%)||No. of Patients||5-Year Survival (%)||No. of Patients|
|Alcoholic liver disease||81.9||3,063||67.6||1,561|
|Chronic viral hepatitis||80.3||4,267||65.3||2,102|
|Primary biliary cirrhosis||85.8||1,726||79.4||860|
|Primary sclerosing cholangitis||87.0||1,601||76.2||776|
Data on long-term outcomes beyond 5 years remains limited, but early reports are promising. Actuarial patient and graft survival rates at 7 years reported from a University of Pittsburgh cohort of 123 patients with ALD were 63% and 59%, respectively.15 However, it appears that survival beyond 5 to 7 years is worse in patients with ALD due to significantly increased rates of cerebrovascular accidents, myocardial infarction, respiratory failure, infection, and de novo malignancies, believed to be the consequence of prolonged exposure to tobacco and alcohol in these patients. The rate of de novo oropharyngeal cancer and lung cancer was 25 and 4 times higher, respectively, in patients with ALD compared with patients with non-ALD and the general population using surveillance epidemiological end result data matched for age, sex, and length of follow-up.16, 17 Implementation of posttransplant screening programs to detect early de novo malignancies may further improve long-term outcomes in these patients.
Does Liver Transplantation Result in a Significant Improvement in HRQOL?
Significant improvement in HRQOL is observed following liver transplantation for end-stage liver disease of any cause18 and appears to be equally beneficial in patient with ALD and non-ALD.19–22 Posttransplant recidivism does not generally seem to diminish this improvement in quality of life19, 23; however, one study demonstrated that although overall scores were similar, patients who returned to daily alcohol ingestion of 3 units or greater experienced more sleep disturbances and required more regular use of benzodiazepines.19 Specific improvements were seen in employment, marital status, psychological health, and social activity following liver transplantation.21, 24 It is unlikely that these improvements in HRQOL will be sustained in long-term studies due to the documented increase in vascular and cancer morbidity and mortality in these patients.
How Often do Patients With ALD Experience Alcohol Relapse Following Liver Transplantation?
A return to alcohol consumption occurs in a significant proportion of patients with ALD and represents a major concern of transplant physicians. Patients may establish a pattern of alcohol intake that severely injures the transplanted liver and also leads to decreased compliance with medical visits, laboratory testing, and use of immunosuppressive medications, thereby placing the patient at risk for graft dysfunction or loss. Defining rates of alcohol consumption, however, remains difficult and is largely dependent on inaccurate methods of detection. Self-reporting of alcohol use through phone or mail communication, or during routine clinic visits, is subject to significant reporting error. Biochemical markers such as blood alcohol levels, urine toxicology screens, and carbohydrate-deficient transferring are subject to the timing, quantity, and frequency of alcohol intake, and therefore are poorly sensitive in detecting alcohol use.25
When alcohol use is reported or detected following liver transplantation, the rate of consumption can vary widely from a single isolated drink (“slip”), to a pattern of repetitive intake of significant amounts of alcohol (“relapse”). When a definition of any alcohol use is applied, the rate of posttransplant alcohol consumption appears to be quite high, approximately 20% (range, 7%-95%) of patients at a follow-up of 21 to 83 months following transplantation.15, 21, 26–51 Table 2 presents summary data from 30 published studies that assessed rates of alcohol relapse based on an “any use” definition. The wide variation in rates may reflect, in part, the timing, frequency, and methods of assessing alcohol intake.
|Author||Reference||Year||No. of Patients||Mandated Abstinence||Follow-up||Relapse Rate||Study Design|
|Anand et al.||27||1997||39||None||25 months||13%||Retrospective|
|Bellamy et al.||15||2001||123||None||7 years||13%||Retrospective|
|Berlakovich et al.||21||1994||44||None||78 months||32%||Retrospective|
|Bird et al.||28||1990||18||None||4 mo–7 yrs||17%||Retrospective|
|DiMartini et al.||26||2001||34||None||1 year||38%||Retrospective|
|Doffoel et al.||50||1992||57||6 months||Not stated||33%||Retrospective|
|Everson et al.||29||1997||42||None||Not stated||17%||Retrospective|
|Fabrega et al.||30||1998||44||None||40 months||18%||Prospective|
|Foster et al.||31||1997||63||None||49 months||22%||Retrospective|
|Gerhardt et al.||32||1996||41||None||Not stated||49%||Retrospective|
|Gish et al.||33||1993||29||None||24 months||24%||Prospective|
|Gish et al.||34||2001||61||None||83 months||20%||Prospective|
|Gledhill et al.||22||1999||24||None||14 months||25%||Retrospective|
|Howard et al.||35||1994||20||None||34 months||95%||Retrospective|
|Iasi et al.||36||2003||66||None||14 months||15%||Retrospective|
|Karman et al.||56||2001||49||6 months||3 years||21%||Retrospective|
|Knechtle et al.||37||1993||32||None||Not stated||13%||Retrospective|
|Krom||51||1994||30||6 months||Not stated||13%||Retrospective|
|Kumar et al.||38||1990||52||None||25 months||12%||Retrospective|
|Lucey et al.||39||1997||50||None||63 months||34%||Retrospective|
|Mackie et al.||40||2001||46||None||22 months||46%||Retrospective|
|Osorio et al.||41||1994||43||6 months||21 months||19%||Retrospective|
|Pageaux et al.||42||2003||128||None||54 months||31%||Retrospective|
|Pereira et al.||43||2000||56||None||30 months||50%||Retrospective|
|Stefanini et al.||44||1997||18||6 months||Not stated||27%||Retrospective|
|Tang et al.||45||1998||56||None||Not stated||50%||Retrospective|
|Tome et al.||46||2002||68||3 months||38 months||10%||Retrospective|
|Tringali et al.||47||1996||58||None||49 months||22%||Retrospective|
|Yates et al.||48||1998||43||6 months||21 months||19%||Retrospective|
|Zibari et al.||49||1996||29||None||Not stated||7%||Retrospective|
Although there are some data to suggest that alcohol use may diminish over time following liver transplantation,21 it is striking that some patients resumed drinking very early in the posttransplant period. Approximately 15% of patients started drinking within the first 6 months, and isolated cases of drinking within the first 2 months have been described.22 Of greater concern are reports in which recurrent alcohol use is more directly quantified; these show that more than one-third of patients designated as “any use” drinkers were involved in repetitive drinking patterns (more than 10 episodes), while the remainder reported mild to moderate alcohol use.26 Frequent and systematic screening for recidivism may be useful in preventing and identifying relapses at an earlier point in time and thereby permit prompt management and medical therapy.
Do Alcohol Relapses Result in Clinically Significant Differences in Outcome After Liver Transplantation?
Long-term, prospective studies evaluating the impact of alcohol relapses on outcomes are lacking, but existing data suggest that a return to moderate or even heavy use of alcohol has minimal detriment to graft function and patient survival.21, 29, 39, 52 Nevertheless, there is a reasonable concern that continued use of alcohol may promote hepatocellular injury and decrease compliance with protocol visits and immunosuppressive regimens, resulting in graft loss. In contrast, the UNOS database has shown that the rates of graft loss in patients with recurrent ALD (0.4/1,000 patient years) are similar or even lower than those in patients with non-ALD (1.2/1,000 patient years).5
In one retrospective study involving 53 patients who underwent liver transplantation for ALD, patients who experienced a relapse were compared with those who remained abstinent and were followed for a mean of 42 months. Surprisingly, the alcohol relapsers had a longer mean survival (54 vs. 44 months), experienced less acute (33% vs. 50%) and chronic (0% vs. 9%) rejection, experienced fewer bacterial infections (26% vs. 37%), and required repeat liver transplantation less often (0% vs. 15%) than non-alcohol relapsers.52 Although not measured directly, the similarity in rejection rates suggests that relapsers did not have significantly different compliance rates with immunosuppressive drugs.
Short-term outcome studies have consistently shown that patient survival rates are similar between relapsers and nonrelapsers. In one study of 132 patients, 82% of patients who returned to moderate drinking remained alive at 5 years, versus 75% of patients who remained abstinent.5 Even poor compliance with follow-up clinic visits and use of immunosuppressive drugs among some patients with ALD does not appear to diminish graft rejection rates compared to patients with non-ALD, with acute rejection observed in 41% of patients with ALD and 43.7% with non-ALD, and chronic rejection observed in 5.6% of patients with ALD and 6.2% with non-ALD.27 There are numerous anecdotal reports, however, of patients who return to heavy alcohol consumption and develop acute alcoholic hepatitis or other medical complications (e.g., pancreatitis, pneumonia, delirium tremens) with secondary rejection, graft loss, and death.29, 39 It is likely that long-term outcome studies will demonstrate poorer survival in patients who revert to heavy drinking, reflecting the previously documented increases in cardiovascular and cancer mortality.
The largest study evaluating the impact of alcohol relapse on outcome after transplantation followed 128 patients over a mean of 54 months and compared 3 different groups of patients: heavy drinkers, occasional drinkers, and those who remained abstinent.42 As seen in other studies, there were no significant differences in actuarial survival rates or rejection rates, but all rejection episodes observed in heavy drinkers were directly related to poor compliance with immunosuppressive drugs; 3 of 7 deaths observed among heavy drinkers were directly attributable to alcohol relapse.
One special setting warranting close attention is coexisting chronic hepatitis B or C virus infection, in which alcohol relapse may be particularly deleterious. Significant alcohol exposure may contribute to rapid progression of recurrent viral infection to advanced fibrosis and cirrhosis, and worsen graft and survival outcomes.53, 54
Are There Reliable Ways to Predict Recidivism in Patients Prior to Liver Transplantation?
There are few reliable predictors of relapse in alcoholic patients, whether or not they undergo liver transplantation.55 Recognizing the challenges in predicting alcohol consumption patterns in patients without ALD, it is not surprising to find that existing data evaluating pretransplant demographic, medical, and psychiatric variables in ALD have failed to identify any single factor of significant predictive value.
One of the earliest studies evaluating risk factors for relapse identified short periods of abstinence of less than 6 months prior to transplantation as the best predictor of posttransplant behavior, thus serving as the basis for implementation of a 6-month mandated abstinence policy in most programs.41 Limited by its small, retrospective design, some subsequent studies have disproved this finding and have suggested that pretransplant sobriety is not a reliable factor in predicting abstinence.
Other clinical variables that have been identified as possible risk factors for alcohol relapse include family history of alcoholism25, 53; history of drug abuse31; absence of life insurance policy, reflecting stability of relationships and employment31; prior alcohol rehabilitation, likely suggestive of more severe addictive disorder25; and daily alcohol consumption.15 One Northern California study suggested that noncompliance with regular clinic visits, medication instructions, and an alcohol abstinence contract predicted recidivism with a hazard ratio of 20.9.34
Karman et al. evaluated 49 patients with ALD who underwent liver transplantation and identified 5 major risk factors for failure to maintain abstinence: single marital status, reflecting suboptimal social support systems; history of suicidal ideation; history of alcohol-related hospitalization; absence of prior alcohol rehabilitation prior to transplantation; or failure to accept further alcohol rehabilitation.56 This highlights the need for thorough evaluation of underlying psychiatric and substance-use problems that must be addressed prior to consideration for liver transplantation.
Many other medical and psychiatric variables were not found to be predictive of alcohol relapse, including age, sex, education, unemployment, coinfection with hepatitis C virus, history of substance abuse, tobacco use, socioeconomic status, prior drunk driving convictions, and residential stability.25, 40, 51 Most importantly, repeated studies demonstrated that specific variables relating to addiction history, including the quantity, frequency, and duration of alcohol abuse, and length of sobriety, were not helpful predictors of relapse. Specifically, abstinence of fewer than 6 months failed to identify individuals who were more likely to remain abstinent following liver transplantation in several studies.25, 40, 51 Clearly, an improved, evidence-based understanding of specific factors that accurately predict recidivism is needed to advise transplant centers in improving patient selection procedures and targeting interventions for individuals at greatest risk.
Are 6- to 12-Month Abstinence Requirements Justified by the Evidence?
Most transplant programs require patients with ALD to demonstrate a long-term commitment to alcohol abstinence prior to consideration for liver transplantation, typically for 6 months or more.57 This provides physicians further opportunity to assess the patient's ability to maintain sobriety and comply with the rigorous demands of a transplant protocol, and it permits some patients to recover from acute alcoholic hepatitis with medical management and obviate the need for liver transplantation. Approximately 75% of centers will additionally ask patients with ALD to sign a binding contract to enter formal alcohol rehabilitation.58 Active alcohol use is considered an absolute contraindication by more than 80% of centers, and most would either delay or remove patients from transplant consideration if they resumed alcohol use while on the waiting list.58
Recognizing the absence of convincing data that mandated periods of abstinence may reliably predict which individuals will relapse following liver transplantation, current policies restricting access to patients with end-stage ALD raises important ethical concerns. Strict enforcement of a 6-month rule will certainly punish some individuals unfairly, including those who may die or experience serious complications from liver decompensation prior to achieving the 6-month eligibility criteria to receive a donor graft.Decisions on transplant eligibility should be made on an individual basis, with careful consideration of the balance of medical, surgical, and psychosocial factors. Short-term and long-term graft and patient survival outcomes in patients with ALD are equal to those in patients with non-ALD; therefore, patients with ALD should receive consideration for listing. Detailed evaluation of one's medical history, substance-use profile, psychiatric assessment, and social environment assists in making this decision. Optimal timing of liver transplantation in ALD varies widely, and therefore specific guidelines and selection criteria that encompass these factors will be important in reaching an appropriate listing decision. Amid increasing public pressure to allocate scarce organ resources away from individuals with “self-afflicted” conditions,8, 59 transplant physicians need to justify through longitudinal, prospective trials that liver transplantation for ALD truly represents best medical practice and an efficient use of resources.
Until further data becomes available, application of a 6-month rule in most cases provides a safe and prudent means of ensuring that an individual with ALD is an appropriate candidate for liver transplantation. A minority of patients will, in fact, demonstrate sufficient improvement following prolonged abstinence that liver transplantation is no longer required. During this bridge period, alcohol-use disorders should be treated aggressively and other comorbid medical and liver diagnoses identified and managed appropriately.
What Criteria Should Transplant Programs Use to Select Patients With ALD for Liver Transplantation?
The selection process for listing a patient for liver transplantation is a lengthy, arduous process that seeks to optimize the allocation of a limited supply of donor grafts to a minority of patients who meet basic requirements of medical necessity and do not have medical or psychosocial issues that would prevent a successful transplant outcome. Although not standardized, minimal listing criteria include the following conditions: a Child-Turcotte-Pugh score greater than or equal to 7, an estimated likelihood of surviving 1 year without transplant less than 90%, a single episode of spontaneous bacterial peritonitis, or the appearance of stage II encephalopathy in the setting of liver failure. If a patient meets one of these criteria, and does not have any absolute contraindications to transplantation (e.g., active substance abuse, extrahepatic malignancy), he or she must undergo a series of tests that assess the preoperative status of the heart, lungs, and kidneys, followed by a psychosocial evaluation that addresses underlying social environment, financial capacity, psychiatric illness, and chemical dependency. Once these evaluations have demonstrated an individual's eligibility and capacity to undergo liver transplantation, allocation of organs is now based on the Model for End-Stage Liver Disease regression equation that incorporates serum bilirubin, serum creatinine, and international normalized ratio into a score that predicts 3-month survival.60
Listing procedures for individuals with ALD are complicated by concern for the potential for alcohol relapse following transplantation and the accompanying risk of graft loss. Data presented elsewhere in this paper suggest that graft and patient survival rates are equal to those of patients with non-ALD, rates of alcohol relapse are small—and do not appear to impair outcomes even when present—and liver transplantation may significantly improve functional outcomes and HRQOL. The necessity of a 6-month mandated abstinence period remains in question but appears to have a useful role in patient monitoring and is certainly in common use across US centers. Many programs additionally use prediction tools to guide this decision, including Vaillant's Prognostic Factors for Long-Term Sobriety, which defines 4 factors that predict success in maintaining sobriety (e.g., compulsory supervision, substitute dependence, new relationships, and inspirational group membership), and the Michigan Alcoholism Prognosis Scale, which evaluates an individual's acceptance of alcoholism and the consequences of the alcohol-use disorder.61, 62 A decision must ultimately be made following careful evaluation by a psychiatrist, who can help diagnose and treat alcohol and chemical dependency and comorbid psychiatric disease.
As our expertise in preoperative and postoperative liver transplantation management continues to evolve, the number of conditions representing contraindications to transplantation will shrink. Current evidence suggests that ALD represents an excellent indication for liver transplantation, and that short-term graft, patient survival, and functional outcomes are equal to those of patients with non-ALD. Liver transplantation in this setting is limited only by a continuing shortage of available organs and plausible concerns regarding the eligibility of some patients due to coexisting medical or psychosocial problems. Living donor liver transplantation represents a promising option for patients with end-stage ALD but raises difficult questions concerning the optimal timing of transplantation. In conclusion, additional research is needed to further define the natural history of ALD, determine long-term outcomes following liver transplantation for ALD, and develop novel approaches to treating alcohol dependency in these patients.
- 71997 Annual report of the US Scientific Registry for Organ Transplantation and the Organ Procurement and Transplantation Network: Transplant data 1988-1997. Bureau of Health Resources and Service Administration, US Dept. of Health and Human Services, Rockville, MD, 1998.
- 9United Network for Organ Sharing (UNOS). Available at: http://www.unos.org. Updated June 26, 2004.
- 10Organ Procurement and Transplantation Network (OPTN). Available at: http://optn.org. Updated June 26, 2004.
- 47Assessment and follow-up of alcohol-dependent liver transplantation patients: A clinical cohort. Gen Hosp Psychiatry 1996; 18( Suppl): S70–S77., , , .
- 50Results of liver transplantation in 75 French patients with alcoholic cirrhosis: Comparison with a non-alcoholic cirrhotic group [Abstract]. Hepatology 1992; 16: 50A., , , , , , et al.
- 51Liver transplantation and alcohol: Who should get transplants? Hepatology 1994; 20( Suppl): S28–S32..
- 57Liver transplantation for alcoholic liver disease. Clin Liver Dis 1998; 2: 839850., .