Pro: The Abstinence Period Should Be the Same for All Patients Undergoing Evaluation for Transplant

Watch a video presentation of this article

Abbreviations

HPSS

Hopkins Psychosocial Scoring System

HRAR

High‐Risk Alcoholism Relapse

MAPS

Michigan Alcohol Prognosis Scale

Key Points

• Alcohol relapse rates are high (20%‐30%), and relapse leads to lower long‐term survival rates.

• We lack reliable predictors of alcohol relapse.

• Custom abstinence periods have potential to worsen socioeconomic disparities related to liver transplantation.

Given the 6‐month mortality rate of patients with severe alcoholic hepatitis refractory to medical therapy is greater than 70%,1 there has been significant interest in offering these patients early liver transplantation in an effort to improve outcomes. To date, three pilot studies have been published in this population.2, 3, 4 These are landmark trials that have generated significant discussion regarding the importance of the duration of alcohol abstinence prior to liver transplantation, despite all three trials’ limitations of short follow‐up periods, small size, and lack of generalizability. Each of these protocols required patients to “commit to lifelong abstinence” but did not require any specified length of pretransplant abstinence.

From 1998 to 2009, alcohol‐related liver disease accounted for 20% of all liver transplants.5 Although not a national standard, most centers required 6 months of alcohol abstinence prior to transplant over this period. Mortality for all patients transplanted with alcohol‐related liver disease under this “6‐month rule” is comparable with patients with other causes of liver disease.6 Retrospective studies comparing patients with evidence of alcoholic hepatitis (clinically or based on explanted liver histology) have also shown similar posttransplant survival to patients without alcoholic hepatitis.7, 8 Prospective data for liver transplant survival in patients with alcoholic hepatitis are available from the early transplantation pilot studies. In these trials, patients with alcoholic hepatitis who underwent transplantation had comparable short‐term survival when compared with patients with alcohol‐related liver disease without alcoholic hepatitis.4 These comparable posttransplant outcomes have sparked discussion regarding the importance of a standard pretransplant abstinence requirement.

Allowing custom abstinence periods prior to liver transplantation is problematic for several reasons: (1) We cannot successfully predict which patients will suffer from alcohol relapse; (2) the percentage of patients who relapse early remains high (around 20% to 30%); (3) alcohol relapse leads to worse graft outcome and poorer survival; and (4) custom abstinence periods pose a threat to justice, one of the Organ Procurement and Transplantation Network’s primary ethical principles.9 We argue that, given the complex, unresolved issues, abstinence periods should remain standardized among all patients seeking liver transplantation until more objective and reproducible methods of pretransplantation relapse assessment are validated.

Many groups have attempted to define predictors of alcohol relapse.10, 11 Factors that seem intuitively likely to be relevant to sustained abstinence, such as duration of alcohol abstinence, family history of alcohol addiction in a first‐degree relative, history of mental illness, engagement in polysubstance abuse, lack of social support, employment status, and degree of insight, have all been evaluated with inconsistent conclusions.10, 12 (Table 1) Some studies have included the ability to undergo pretransplantation counseling as a factor, given the clinical perception that this may be a helpful intervention. Although findings on the impact of pretransplant counseling have been mixed,13 a lack of pretransplant abstinence period de facto precludes counseling from occurring, leaving the inciting problem (alcohol addiction) unaddressed. Attempts to develop validated clinical scoring systems that incorporate predictive factors (Hopkins Psychosocial Scoring System, High‐Risk Alcoholism Relapse [HRAR] score, among others4, 14) have unfortunately not resulted in a consensus system because of disparate conclusions regarding the influence of each predictor.4, 15, 16

Table 1.

Predictive Variables for Posttransplant Alcohol Relapse

Factor	Associated	Not Associated
Less than 6 months of abstinence	Raakow et al. (1995) Hillebrand et al. (1997) Pfitzmann et al. (2007)* De Gottardi et al. (2007)* Dew et al. (2008)* Tandon et al. (2009)* Karim et al. (2010)* DiMartini et al. (2010) (any relapse)*	Gish et al. (1993, 2001) Gerhardt et al. (1996) Anand et al. (1997) Coffman et al. (1997) Foster et al. (1997) Tang et al. (1998) DiMartini et al. (2001) Mackie et al. (2001) Jauhar et al. (2004) Kelly et al. (2006)* DiMartini et al. (2010) (heavy/moderate)*
Polysubstance abuse	Foster et al. (1997) Hillebrand et al. (1997) DiMartini et al. (2001)	Coffman et al. (1997) Jauhar et al. (2004) Kelly et al. (2006)* Karim et al. (2010)* DiMartini et al. (2010)*
Social instability/lack of social support	Gish et al. (1993, 2001) Lucey et al. (1992) Zibari et al. (1996) Coffman et al. (1997) Foster et al. (1997) Pageaux et al. (1999) Perney et al. (2005)* Pfitzmann et al. (2007)* Dew et al. (2008)* Rodrigue et al. (2013)*	Mackie et al. (2001) Karim et al. (2010)*
Family history of alcoholism in first‐degree relative	Foster et al. (1997) DiMartini et al. (2001) Jauhar et al. (2004) Dew et al. (2008)* DiMartini et al. (2010) (any relapse)* Deruytter et al. (2013)*	Mackie et al. (2001) DiMartini et al. (2010) (heavy/moderate)*
Younger age	Foster et al. (1997) Pageaux et al. (1999) Perney et al. (2005)	Mackie et al. (2001) Pfitzmann et al. (2007)* Dew et al. (2008)*
Poor previous treatment response	Gish et al. (1993, 2001) DiMartini et al. (2001)	Jauhar et al. (2004)
Lack of insight	Kelly et al. (2006)* Hartl et al. (2011)*	Berlakovich et al. (1994) Lucey et al. (1992) Raakow et al. (1995) Zibari et al. (1996) Foster et al. (1997) Campbell et al. (1998) Pageaux et al. (1999)
Contract	Gish et al. (1993, 2001)	Gerhardt et al. (1996)
Noncompliance	Gish et al. (1993, 2001) Zibari et al. (1996)
Coexisting mental disorders	Gish et al. (1993, 2001) Tripp et al. (1996) Coffman et al. (1997) Kelly et al. (2006)* Egawa et al. (2014)*	DiMartini et al. (2001) Jauhar et al. (2004) Tandon et al. (2009)* Karim et al. (2010)*
Clinical scoring systems (MAPS, HPSS, HRAR)	De Gottardi et al. (2007) (HRAR)*	Lucey et al. (1992) (MAPS) Coffman et al. (1997) (MAPS) Lee et al. (2017) (HPSS)* DiMartini et al. (2000) (HRAR)* Im et al. (2016) (HRAR)*
Length of abstinence before liver transplantation (months)	Perney et al. (2005) (heavy/severe)* Gedaly et al. (2008) (<12 months)* Hartl et al. (2011) (<3 months)*	Perney et al. (2005) (any relapse)*

Adapted with permission from Alcohol and Alcoholism.10 Copyright 2006, Medical Council on Alcohol and Oxford University Press, to include all studies to date.

^*Designates newer study added to McCallum et al.’s10 original table.

Relapse rates have been variable among studies of liver transplantation for alcohol‐related liver disease, with reported 1‐year relapse rates ranging from 8% to 22%.17 Although there is not yet a universal standard, most contemporary studies try to identify rates of meaningful relapse, with both quantity and duration of alcohol use incorporated to separate patients who consume minimal amounts of alcohol from those who experience substantial relapse. In these studies, the rate of meaningful relapse is approximately 15% to 20%.17, 18 Even among the super‐selected patients (2%‐6% transplanted of all patients evaluated) in the three recent studies of transplantation in acute alcoholic hepatitis, relapse rates within 1 to 2 years ranged from 11% to 23%.2, 3, 4 Furthermore, studies of alcohol relapse are universally hampered by patients who may not report because of fear of judgment, guilt, or decreased access to care.19 There are several promising biomarkers of occult alcohol use; however, few studies have incorporated these in evaluation of posttransplant relapse to date.20

Although older studies questioned the relationship between alcohol relapse and survival,18 newer data show alcohol relapse leads to increased graft failure and poorer survival in a dose‐dependent fashion.12, 21, 22 Pfitzmann et al.,12 for example, demonstrated a 10‐year survival gap of 20% versus 81% in patients who resumed hazardous drinking versus abstinent patients. As stated earlier, prior studies that did not show this relationship were limited by short follow‐up and imprecise definitions of alcohol relapse so that patients with minimal recurrent drinking were grouped with patients with meaningful relapse, leading to inflation of the survival of the “relapsers.” One recent study, in fact, reported an accelerated rate of cirrhosis in patients who had meaningful relapse after transplant, with a median of only 5 years between alcohol relapse and evidence of recurrent cirrhosis on biopsy.21

The most concerning element of custom abstinence periods is the potential they have to worsen socioeconomic disparities among chosen transplant candidates, directly threatening equity in organ allocation. Volk et al.23 demonstrated the significant intracenter and intercenter variability in factors considered during transplant selection committee meetings under current guidelines (Fig. 1). They noted that socioeconomic status of patients correlated with rejection for social barriers. Another study found that, among patients found to be ineligible for transplant, African American patients were more likely to have psychosocial barriers noted as a reason for exclusion than their Caucasian counterparts (39% versus 17%; P = 0.04).24 Removing one of the few objective factors in the psychosocial evaluation of patients with alcohol‐related liver disease could potentially worsen transplant access disparities given the incentives in place to compel institutions to accept patients who will reimburse centers well. This is especially dubious in the absence of a more successful objective, overarching system to predict alcohol relapse in the early transplant period. The data from the three early transplantation trials reflect this potential, with one study excluding patients with mental illness2 and rates of private insurance among patients as high as 82%,4 compared with a national average of 64%.25

Figure 1.

Image compares and contrasts the subjective factors assessed during the transplant selection process and how they correlate with long‐term survival versus the objective means of organ allocation (Model for End‐Stage Liver Disease sodium [MELD‐Na]) and how it associates with short‐term outcomes.

Given these concerns, we feel that a shift to custom abstinence periods is hazardous and premature for the aforementioned key reasons. We recommend maintaining standardized abstinence periods until accurate predictors of relapse are more firmly established and outcomes data for early transplantation are more generalizable.