Alcohol Use Relapse Following Liver Transplantation for Alcoholic Liver Disease

Background

Behavioral and mental disorders related to alcohol addiction, called alcohol use disorders, affect 10% of European population; one of the consequences of alcohol use disorders is alcoholic liver disease (ALD) that may lead to cirrhosis and is the most common indication for liver transplantation (LTx) in Slovakia [1–3]. Slovakia ranks fourth in Europe for mortality rate from liver cirrhosis, with ALD being the most prevalent cause [4]. According to the 2018 Global Report on Alcohol and Health, consumption of pure alcohol per capita per year in Slovakia exceeds 11 L (almost double the average of 6.4 L). There are no officially accepted governmental policies to combat alcohol use disorders. Despite conflicting evidence behind the “6-months rule”, Slovak patients with a chronic form of ALD would only be eligible for LTx after at least 6 months of abstinence from alcohol consumption documented by the addictologist [5–7].

The management of patients referred with decompensated chronic liver disease for consideration of LTx to our center followed standard protocol. We attempt to confirm the etiology and stability of the decompensated state of the liver by the usual spectrum of evaluations: from history taking to hepatic venous pressure gradient and (transjugular) liver biopsy. We attempt to determine if the patient is critical with the diagnosis of ALD. Then we review the data related to the patient’s existing abstinence interval and refer the patient to a dedicated addictologist. This expert directly cooperates with our transplant center, and his or her task is to confirm etiology, sobriety, duration of abstinence interval of at least 6 months, and outlook concerning the long-term adherence to abstinence. The only exception is severe acute alcoholic hepatitis not responding to standard treatment in patients fulfilling the criteria of the Lille protocol.

The precise proportion of patients with alcohol use relapse after LTx (a form of alcohol use disorders) is not known; however, according to the literature, the prevalence ranges from 7% to 95% [5,8–10]. One of the reasons for this wide variation lies in differing definitions of the term “recurrence” and “relapse.” Post-LTx alcohol use relapse may lead to numerous consequences, such as liver graft injury (prevalence up to 17%) [11], liver graft loss (up to 32%) [12], and reduced 10-year survival (45% versus 85% in non-relapsers) [11]. The most common alcohol consumption patterns following LTx are occasional drinking (alcohol slip-ups), i.e., consumption of a limited amount of alcohol, followed by re-established abstinence, and harmful drinking, i.e., consumption of 4 or more drinks per day, or drinking for 4 or more successive days (Table 1) [13].

Risk factors for alcohol use relapse are not well established. The most commonly cited risk factors in the literature are (younger) age, cigarette smoking, mental disorders, non-compliance with regular follow-up, loss of social status (e.g., divorce, bereavement), and unemployment [14]. Our patients are routinely followed according to their clinical status and usually, if stable, on a 3-month follow-up schedule.

In this study, we aimed to determine the proportion of ALD patients who received a transplant who relapsed to alcohol drinking after LTx, and the risk factors for alcohol use relapse. We also compared clinical outcomes according to the presence or absence of alcohol use relapse.

Material and Methods

ETHICAL APPROVAL:

All procedures involving human participants were approved according to the ethical standards of the institutional research committee, including the 1964 Helsinki Declaration and its later amendments or comparable ethical standards. The reported clinical and research activities were consistent with the Principles of the Declaration of Istanbul as outlined in the Declaration of Istanbul on Organ Trafficking and Transplant Tourism.

Results

From the 196 LTx in 191 patients performed during the observation period of 115 months, 87 patients (45.5%) underwent LTx due to non-alcohol use disorders etiology and we excluded 15 patients by the predefined criteria of death during the first month post LTx. Our final cohort included 89 out of 191 patients (46.6%) for the final analysis (Figure 1). Patients transplanted due to ALD had a mean age of 55.35 years (IQR 48.79 to 59.87), and 24.7% were female. The mean MELD score was 16.0 points (IQR 13.79 to 19.00 points), and the mean Child-Pugh score was 10.0 points (IQR 9 to 11 points). Summary statistics and study group characteristics are displayed in Table 2.

We observed alcohol use relapse in 23 out of 89 patients (26%) transplanted for ALD. We did not find any post-transplantation alcohol use in the 153 patients transplanted for non-ALD liver disease (P=0.0001). In 12 patients (52%) the pattern of alcohol use relapse was harmful drinking, and in 11 patients (48%) it was occasional drinking (P=non significance), (Figure 1).

In order to determine possible risk factors in our cohort, we compared the risk factor prevalence between relapsers and non-relapsers (Table 3). Patients who relapsed to alcohol use were non-significantly younger (52.57 [AR 37.6] versus 56.15 [AR 47.57], P=0.11), more often smoked cigarettes (13 [56.52%] versus 9 [13.64%], P=0.0001), have lost their social status by divorce or bereavement (13 [56.5%] versus 7 [10.6%], P<0.0001), and were non-compliant with regular appointments (9 [39.1%] versus 2 [3%], P<0.0001). In contrast, we did not observe differences between relapsers and non-relapsers concerning sex, unemployment, premature retirement due to invalidity (IP), or the age-related retirement (OAP). Included study patients had no documented mental disorders.

We observed graft injury in 17 patients (73.9%) with alcohol use relapse as compared to 8 patients (12.1%) without relapse (P<0.0001).

In the search for possible risk factors associated with liver graft injury in patients with alcohol use relapse we used univariate analysis and identified the following risk factors: younger age, harmful pattern of drinking, cigarette smoking, loss of social status, and non-compliance with regular appointments (Table 4).

In the multivariate logistic model displayed in Table 5A, we revealed 3 independent risk factors for alcohol relapse: days post liver transplantation (OR=1.0008; 95%CI 1.0001–1.0014, P=0.015), cigarette smoking (OR=5.92, 95%CI 1.6665–21.0308; P=0.006) and loss of social status (OR=7.61; 95%CI 2.0859–27.7745, P=0.0021).

In the multivariate logistic model displayed in Table 5B, we revealed 2 independent risk factors of liver graft injury: occasional drinking (OR=12.69, 95%CI 3.025–53.22, P=0.0005), and harmful drinking (OR=36.25, 95%CI 6.69–196.18, P<0.0001).

We did not find a significant difference in mortality during follow-up between patients with or without alcohol use relapse: 2 patients (9%) versus 3 patients (5%), P=0.83.

Comparison of overall survival between patients with and without alcohol use relapse is displayed in Figure 2. Kaplan-Meier survival curves revealed no difference in overall survival according to alcohol use relapse after LTx for ALD patients (P=0.93), (Figure 2).

Discussion

The most important finding of this single-center retrospective study covering 10 years of our LTx program was the alcohol use relapse rate of 26% seen in patients transplanted for ALD. This rate compares well with other studies [5,8–10,16]. However, some studies found alcohol use also in patients transplanted for other indications than ALD. This finding might be explained by the different inclinations of the ALD and non-ALD cohorts to drink, as was shown in the study by Abosh et al. that only 2% of patients who received a liver transplant for non-ALD drank alcohol after LTx [17]. However, we cannot exclude the possibility that our methodology caused the differences found, as we focused primarily on patients transplanted for ALD. The relatively high prevalence of alcohol use relapse in our cohort among other things adds to the arguments against the clinical usefulness of the 6-month rule [5,6,18,19]. We have adopted this rule for 2 historical reasons: as a part of the protocol from our mentoring institution and because it harmonized with the atmosphere in the professional and lay community at that time. In light of our results, however, we will have to reconsider. The most probable next step will be to scrutinize the feasibility in this context of the Lille model approach which we already adopted for the patients with acute alcoholic hepatitis [20].

The impact of alcohol use relapse on liver graft injury over time (73.9% versus 12.1%, P<0.0001) concurs with the available data [21], but the non-significant impact on mortality does not. According to available literature, the 1-year survival rate in patients after LTx due to ALD is at least as good as in other etiologies and reaches 93% [1,22–24]. However, with the alcohol use relapse, especially to harmful drinking, the mortality rate increases [25,26]. This discrepancy might be explained by a rather small number of patients and relatively short median follow-up time [11,12] as was the case in our study The median time (in days from LTx) at which we determined the alcohol use relapse was longer in patients with relapse as compared to patients without relapse (1228.0 days versus 2595.0 days, P=0,0006); in fact, the time from LTx was an independent predictor of alcohol relapse. The design of our study allowed us to evaluate only the cumulative rate of alcohol relapse, but not the temporal trends in more detail. However, patients that relapsed had a median time post LTx of over 7 years, which might suggest that the follow-up program of psychological and social support should be long-term, possibly indefinite. The evolution of liver graft injury over time has been addressed in several studies [15,27,28].

Of potential risk factors for alcohol relapse listed in the literature [5,23,25], several associations also appeared in our study: cigarette smoking, loss of social status, and non-adherence with regular follow-up All of these risk factors will be used at our center immediately as warning signals and we will implement them into the daily clinical practice. Also, we will address them in a prospective study. These risk factors could serve as a risk-stratification measure helping us to provide patients at risk with more intensive management by the addictology experts and social support team [18,25,29–34]. Several studies have suggested that younger age and female sex is related to increased alcohol relapse rates [18,21,32,35]. Our patients with alcohol use relapse were almost 3.5 years younger, but the difference was not statistically significant. Although other studies have found the opposite results [5,6,25], we consider younger age, especially in connection with other risk factors, as a reason for increased attention and focused care. Although unemployment could be a significant risk factor for alcohol use relapse [6,7,10,18], we were not able to analyze it as such.

Limitations of our study were the retrospective design, the single-center setting, and the relatively low number of patients. However, a single-center report provides results with unified selection criteria for LTx and unified diagnostic and management protocols. The obvious limitations of our study were the lack of liver biopsies in all patients with alcohol use relapse and the inability to determine the exact time of first relapse.

To the best of our knowledge, this is the first study of its kind in Slovakia, and the second in the published literature from the Central European region [35].

Conclusions

We documented alcohol use relapse in 26% of patients transplanted for ALD; this relatively high rate was comparable with data from the literature. We identified several risk factors associated with alcohol use relapse: cigarette smoking, loss of social status, and non-compliance with regular follow-ups. We will use these risk factors as a trigger for intervention in our center. In our study, alcohol use relapse was associated with liver graft injury but not with mortality, probably due to the limited follow-up time and sample size of our study. These study findings allow us to move forward in methods for data collection, risk stratification of waitlisted patients with ALD, as well as with a more intensive and directed post-transplantation psychological and social support.