Impact of alcohol abstinence on survival after hepatic resection for hepatocellular carcinoma in patients with alcohol-related liver disease

Background This study aimed to evaluate the prognostic impact of alcohol abstinence on survival after hepatic resection for hepatocellular carcinoma (HCC) in patients with alcohol-related liver disease (ALD). Patients and methods In total, 92 patients with ALD-HCC who underwent initial and curative hepatic resection were identified, including 56 and 36 patients with and without alcohol abstinence, respectively. Results The 3-, 5-, and 7-year recurrence-free survival (RFS) were 46%, 43%, and 37% in the abstinence group, and 61%, 36%, and 36% in the non-abstinence group, respectively (p = 0.71). The 3-, 5-, and 7-year overall survival (OS) were 91%, 76%, and 66% in the abstinence group, and 87%, 57%, and 44% in the non-abstinence group, respectively (p = 0.023). Multivariate analysis revealed that non-abstinence was an independent prognostic factor for OS (P = 0.026). The incidence rate of liver-related death including HCC-specific death, liver failure, and renal failure in cirrhosis (hepatorenal syndrome) between the non-abstinence and abstinence groups were 41.7% vs. 19.6% (p = 0.032). Worsening of the Child–Pugh grade at intrahepatic recurrence was more frequently observed in the non-abstinence (33.3%) than that in the abstinence group (6.5%) (p = 0.039). Conclusions Alcohol abstinence might improve the long-term survival of patients with ALD-HCC undergoing hepatic resection.


Introduction
Hepatocellular carcinoma (HCC) is one of the most common cancers worldwide [1,2]. Its occurrence is closely related to the presence of chronic liver disease, mostly as a result of hepatitis B or C virus (HBV or HCV) infection. However, the proportion of patients with HCC who are negative for hepatitis B surface antigen (HBs Ag) and anti-HCV antibody has recently increased [3][4][5]. Without HBV or HCV infection, lifestyles such as dietary habit, smoking, physical activity, and alcohol consumption have been reported to increase the risk of HCC development [6][7][8]. Alcohol abuse is one of the causes of HCC, reported to be responsible for approximately 30%-45% increase of HCC incidence in Western countries [5,9,10].
Alcohol-related liver disease (ALD) is manifested by alcohol-induced liver inflammation leading to fibrosis and carcinogenesis progression [11,12]. The most effective treatment for patients with ALD is the achievement and maintenance of alcohol abstinence due to its limited effective medical treatment [13]. Previous reports indicated that alcohol abstinence would improve the long-term prognosis in patients with ALD including alcoholic steatohepatitis and alcoholic cirrhosis [14,15]. In patients undergoing liver transplantation for ALD, alcohol recidivism is reported to impair long-term survival after liver transplantation [16,17]. However, in patients undergoing hepatic resection for HCC associated with ALD (ALD-HCC), to the best of our knowledge, no report has evaluated the impact of alcohol abstinence on the long-term survival postoperatively.
Therefore, this study aimed to evaluate the prognostic impact of alcohol abstinence on survival outcome after hepatic resection for ALD-HCC.

Patients
A total of 92 patients with ALD-HCC who underwent initial and curative hepatic resection at the Osaka City University Hospital from January 1994 to December 2018 were identified. ALD was diagnosed according to chronic heavy alcohol use (alcohol consumption of >60 g/ day for >5 years) with the following histological features: steatosis, hepatocellular injury with ballooning, lobular inflammation, fibrosis or cirrhosis [18], and exclusion of other causes of liver disease, such as HCV or HBV infection, autoimmune hepatitis, primary biliary cirrhosis, and Budd-Chiari syndrome [19]. Curative hepatic resection was defined as complete resection of a recognizable tumor and the histological absence of tumor cells along the parenchymal transection line. This study was approved by the guidelines of our institutional ethics committee (no. 3815). This study was conducted in accordance with the Declaration of Helsinki and reported in line with the Strengthening the Reporting of Cohort Studies in Surgery 2019 guidelines [20].

Alcohol abstinence
All patients were advised to abstain from alcohol. Alcohol intake was evaluated through patients' self-report and family member interviews by primary doctors. Alcohol abstinence was defined as drinking occasionally, a monthly alcohol intake of less than once with <20 g/day of alcohol intake [14]. Of 92 enrolled patients, 56 achieved persistent abstinence from alcohol postoperatively without relapse (abstinence group) and 36 patients did not abstain from alcohol postoperatively (non-abstinence group).

Patient follow-up
All patients were followed up 1 month postoperatively and 3 months thereafter. Follow-up evaluations were as follows: physical examination, liver function tests, HCC-specific tumor marker, chest radiographs to examine for pulmonary metastases, and ultrasonography, dynamic computed tomography, or magnetic resonance imaging to examine for recurrence in the remnant liver or other abdominal organs. For patients with recurrence, appropriate therapeutic treatment was adopted such as re-hepatic resection, radiofrequency ablation, or percutaneous ethanol injection therapy defined as curative treatment, and transcatheter arterial chemoembolization or other treatment alternatives, as noncurative treatment [21]. Overall death was categorized into liver-related death and other causes of death. Liver-related death included HCC-specific, liver failure, and renal failure in cirrhosis (hepatorenal syndrome) [22]. The median follow-up period was 45.5 (interquartile range, 23.4-76.3) months in the overall cohort. Of the 92 patients, 29 (31.5%) dropped out during the follow-up; the median follow-up period was 41.8 (interquartile range, 17-65.4) months for the patients who dropped out.
Gamma-glutamyl transpeptidase (GGT) and liver fibrosis indices including aspartate aminotransferase (AST) to platelet ratio index (APRI) and fibrosis index based on four factors (FIB-4) index were also evaluated for available patients postoperatively at our outpatient. Liver fibrosis indices were calculated using the following formulas: APRI =

Histology
The guidelines of the Liver Cancer Study Group of Japan [23] were used to evaluate histological tumor classifications and degree of the background liver. The grade (active hepatitis severity) and stage (degree of hepatic fibrosis) of non-cancerous hepatic tissue were determined by scoring based on the histologic activity index [24,25].

Statistical analysis
Categorical variables were compared using Fisher's exact test. Continuous data were compared using the Mann-Whitney U test. Rates of recurrence-free survival (RFS) and overall survival (OS) were evaluated using the Kaplan-Meier method. Differences between curves were evaluated using the log-rank test. RFS and OS were evaluated by univariate and multivariate analyses using the Cox proportional hazards model. Variables potentially associated with recurrence were selected based on previous study results or on our own clinical experience, including age (≤65 or >65 years), gender, body mass index (BMI) (≤25 or >25 kg/m 2 ), serum GGT (<65 or ≥65 U/L), ALT activity (≤30 or >30 U/L), albumin concentration (>3.5 or ≤3.5 g/dL), prothrombin time activity percentage (>70 or ≤70%), total bilirubin level (≤1.0 or >1.0 mg/dL), serum alpha-fetoprotein (≤20 or >20 ng/mL), Child-Pugh grade, tumor size (>3.0 or ≤3.0 cm), degree of tumor differentiation (poorly or well/moderate), microscopic portal invasion, multiple tumors, liver cirrhosis, grading score (0-2 or 3-4), and major hepatic resection (≥2 section). Variables with a P-value of <0.1 in univariate analysis were entered into the multivariate analysis. The threshold for statistical significance was set at P < 0.05. Statistical analyses were performed using the SPSS software v21.0 (IBM Corp, Armonk, NY) and EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan).

Patient characteristics
Clinicopathological characteristics of 92 patients are shown in Table 1. The mean age was 68 (range, 41-82) years, and patients comprised of 89 men and 3 women. The number of patients with BMI of >25 kg/m 2 were significantly higher in the non-abstinence than that in the abstinence group (p = 0.016). Proportions of >3-cm tumor size and major hepatic resection were significantly higher in the abstinence group than that in the non-abstinence group (p = 0.028 and p < 0.001, respectively).

Time course of postoperative serum GGT level and liver fibrosis indices
Although the median GGT levels 1 month after hepatic resection were comparable between the abstinence and non-abstinence groups (74.5 U/L and 73.5 U/L, respectively, p = 0.91), those of the nonabstinence group were significantly higher than of the abstinence group (77 U/L and 51 U/L, respectively, p = 0.031) at 24 months postoperatively. The median FIB-4 index and APRI 1 month postoperatively in the abstinence and non-abstinence groups were 2.9 and 2.6 (p = 0.84) and 0.70 and 0.58 (p = 0.28), and at 24 months postoperatively were 2.8 and 3.5 (p = 0.43) and 0.65 and 0.70 (p = 0.33), respectively (Table 4).

Cause of death
Overall death was confirmed in 19 patients (52.8%) in the nonabstinence group and 13 (23.2%) in the abstinence group (p = 0.0066). Incidence rates of liver-related death including HCC-specific death, liver failure, and renal failure in cirrhosis (hepatorenal syndrome) between the non-abstinence and abstinence groups were 41.7% vs. 19.6% (p = 0.032). Incidence rates of other causes of death between the two groups were 11.1% vs. 3.6% (p = 0.21) (

Discussion
Multivariate analysis revealed that non-abstinence was an independent prognostic factor for OS with an approximately 2.2-fold increased risk, although RFS showed no significant difference between patients with and without alcohol abstinence. Worsening of Child-Pugh grade at intrahepatic recurrence was less frequently observed in the abstinence group. Hepatic resection tended to be more frequently performed for the treatment of HCC recurrence in the abstinence than that in the nonabstinence group. The incidence rate of liver-related death including HCC-specific death, liver failure, and renal failure in cirrhosis (hepatorenal syndrome) was significantly lower in the abstinence than that in the non-abstinence group. To the best of our knowledge, this is the first report clarifying the impact of alcohol abstinence on survival outcome after hepatic resection in patients with ALD-HCC.
Chronic alcohol consumption promotes hepatocyte injury and liver inflammation leading to the progression of liver fibrosis and hepatocarcinogenesis [11,12]. In patients with ALD, previous reports demonstrated that alcohol abstinence would reduce the increased serum levels of liver fibrosis markers and provide survival benefit of liver function preservation [26,27]. In this study, alcohol abstinence improved the OS after hepatic resection of ALD-HCC. Focusing on the cause of death, patients without abstinence showed the twofold increased risk for liver-related death compared with those with abstinence. At the time of recurrence, patients in the non-abstinence group experienced more frequent deterioration of liver function with worsening Child-Pugh grade and were less likely to receive hepatic resection for intrahepatic recurrence. Hepatic resection for HCC recurrence was reportedly associated with long-term survival [28]. These indicated that alcohol abstinence would prevent the worsening of liver function leading to not only reduced risk of liver failure progression but also increased chance of re-hepatic resection for HCC recurrence, resulting in improved prognosis postoperatively. Meanwhile, liver fibrosis indices including the FIB-4 index and APRI showed no significant differences postoperatively between patients with and without abstinence. As previously described, serum fibrosis markers were initially designed for patients with hepatitis C, and the accuracy as an index to evaluate the degree of fibrosis would differ in each serum fibrosis markers [21,26]. Clinical practice guidelines for the ALD management by the European Association for the study of the liver indicated that APRI would be of limited use in diagnosing fibrosis in patients with ALD [29]. Conversely, in this study, the serum GGT level at 24 months postoperatively was lower in the abstinence than that in the non-abstinence group. Previous reports demonstrated that the serum GGT level is the most frequently used markers to detect previous alcohol consumption, which is closely correlated with liver conditions and most useful for the ALD diagnosis [29,30]. Therefore, the serum GGT level might be useful as a biological marker of liver condition after hepatic resection in patients with ALD-HCC.
Excessive alcohol consumption would also be associated with increased risk for chronic disease progression including cardiovascular disease, cerebrovascular disease, metabolic and endocrine disease, and cancer incidence [31,32]. In this study, patients in the non-abstinence group tended to have threefold increased incidence rate of other Table 4 Time course of postoperative serum GGT level, FIB-4 index, and APRI.   causes of death compared with the abstinence group, such as respiratory failure, myocardial infarction, and aortic dissection. Therefore, reducing the risk for these chronic diseases induced by alcohol abstinence might contribute to improve OS. Previous reports showed several risk factors predicting poor OS after hepatic resection of HCC, such as tumor size, vascular invasion, and tumor multiplicity as tumor-related factors, and liver cirrhosis and background liver function as liver-related factors [33][34][35]. In this study, no other tumor-or liver-related factors were identified as prognostic factors for OS. A possible reason for this is the small number of overall death. As previously described, patients with HCC who were negative for HBs Ag and anti-HCV antibody had preserved background liver function and lower frequency of liver cirrhosis [36] and showed lower risk for overall death than those of patients with hepatitis B virus-and hepatitis C virus-related HCC [3]. Our previous report also indicated the lower risk for overall death in patients negative for HBs Ag and anti-HCV antibody than patients with HCV-related HCC [37]. In this study, >90% of patients in both non-abstinence and abstinence groups had a favorable preoperative liver function with the proportion of Child-Pugh grade A, and only 32 patients died during the study period. The low incidence rate of overall death might have obscured the prognostic impact of tumor-or liver-related factors. Therefore, further accumulation of patients will reveal other prognostic factors such as tumor-or liver-related factors for survival in patients with ALD-HCC.
In terms of HCC recurrence, alcohol abstinence did not affect the RFS rate in this study. Alcohol intake has been definitely known to be involved with hepatocarcinogenesis through the liver fibrosis progression [38]. Alcohol abstinence would be expected to reduce the hepatocarcinogenic potential in patients with ALD. However, a previous report demonstrated that former drinkers who stopped 1-10 years previously still had increased risk for HCC development, whereas >10 years duration of abstinence achieved significant reduction of HCC development in patients with ALD [38]. In the present study, the median follow-up time after surgery was 45.5 months. Therefore, longer follow-up duration might clarify the favorable impact of alcohol abstinence on the risk of HCC recurrence.
In this study, proportions of >3-cm tumor size and major hepatic resection were higher in the abstinence than that in the non-abstinence group. Although all patients were advised to abstain from alcohol by primary doctor, patients with small tumor size and minor hepatic resection were less likely to achieve alcohol abstinence even after the experience of hepatic resection for HCC. Generally, lifestyle improvements, including those related to alcohol-consumption habit, require both catalysts and a strong will. Receiving a diagnosis of a large tumor and undergoing major hepatic resection would more strongly change the willingness for abstinence compared with receiving a diagnosis of a small tumor and undergoing minor hepatic resection. The change of consciousness in patients with large tumors and in those undergoing major hepatic resection might have led to the higher rate of abstinence. Therefore, the beneficial information in better survival provided by alcohol abstinence in this study may help motivate patients with ALD-HCC to abstain from alcohol after hepatic resection.
This study had several limitations that should be considered. First, because this was a retrospective study and performed in a single center, potential bias in patient enrollment may exist. Second, the total number of patients was relatively small in both groups. Third, approximately 30% of all patients dropped out during the follow-up period in the present study. However, the median follow-up periods were comparable between the patients who dropped out and those who did not drop out; therefore, the influence of patient drop-out during follow-up on longterm outcomes might be limited in the present study. Fourth, this study did not evaluate the impact of other lifestyle-related factors on the prognosis, such as body weight changes, diabetes mellitus control, and smoking status postoperatively. However, current results were based on rather long-term follow-up after hepatic resection. Thus, results of the present analysis could provide important information for the treatment of ALD-HCC. Fifth, data on alcohol consumption were retrieved from patients' self-report or information obtained from family of patients. Such information can be biased because the true amount of alcohol consumption is often concealed. However, despite these limitations, a beneficial effect of abstinence during follow-up could be confirmed in this study.
In conclusion, alcohol abstinence might improve the long-term survival in patients with ALD-HCC undergoing hepatic resection. The current result would be important information for the management of patients with ALD-HCC.

Ethical approval
This study was approved by the institutional ethics committee of Osaka City University Graduate School of Medicine (no. 3815).

Consent
We gain fully informed written consent and get our institutional ethics committee approval (no. 3815).We were documented it in the paper.

Author contribution
Daisuke Shirai and Hiroji Shinkawa made the conception and design of this study. Authors other than Daisuke Shirai contributed to the collection, analysis, and interpretation of the data. Daisuke Shirai wrote the draft manuscript, and other authors performed the critical revision of the manuscript. All authors gave final approval of the version to be published. Daisuke Shirai has overall responsibility and guarantees the scientific integrity.