Alcohol Use and Treatment of Hepatitis C Virus:

Alcohol Use and Treatment of Hepatitis C Virus: Results of a National Multicenter Study


Gastroenterology May 2006

VA-HCV-001 Study GroupBhupinder S. AnandCorresponding Author Informationemail address, Sue Currie‡, Eric Dieperink§, Edmund J. Bini∥, Hui Shen¶, Samuel B. Ho§, Teresa Wright‡

Department of Medicine, Michael E. DeBakey VA Medical Center, Houston, Texas

‡ Department of Medicine, San Francisco VA Medical Center, San Francisco, California

§ Department of Medicine, Minneapolis VA Medical Center, Minneapolis, Minnesota

∥ Department of Medicine, VA New York Harbor Healthcare System, Brooklyn, New York

¶ Department of Medicine, University of California, San Francisco, San Francisco, California

"Our findings show that physicians are reluctant to treat patients with a history of alcohol use, even when they are abstinent for 1 year or more. The only negative impact of alcohol use on treatment outcomes was a higher treatment discontinuation rate in recent drinkers. Patients who were able to complete treatment had similar ETR and SVR rates compared with nondrinkers, irrespective of the duration of abstinence. These findings were consistent among the subgroups of genotype 1 and black patients. Our observations are clinically important because they indicate that alcohol use should not be considered an exclusion criterion when evaluating patients for anti-HCV treatment, especially in view of the fact that nearly one third of patients with HCV infection have a history of recent alcohol use. The current attitude among physicians against offering treatment to patients who use alcohol should be reassessed. Moreover, greater emphasis should be placed on developing strategies to improve patient education and support to improve adherence to antiviral therapy."


ABSTRACT

Background & Aims: Patients with hepatitis C virus (HCV) infection who use alcohol have been excluded from clinical trials; therefore, outcomes with antiviral therapy are unknown. The aim of the study was to determine the impact of alcohol use on HCV treatment outcomes.

Methods: Subjects using alcohol were categorized as follows: no alcohol versus regular alcohol use, quantity consumed (none, <6 drinks/day, ≥6 drinks/day), CAGE score <2 or ≥2, and recent alcohol use (past 12 months). Patients were treated with interferon plus ribavirin.

Results: A total of 4061 subjects were enrolled, and 726 (18%) received treatment. Alcohol use (past and within 12 months) reduced treatment candidacy. Past alcohol use did not affect the end-of-treatment response, sustained virologic response (SVR), and treatment discontinuation rates. However, recent alcohol use resulted in higher treatment discontinuation (40% vs 26%; P = .0002) and tended to reduce the SVR (14% vs 20%; P = .06), but when patients who discontinued treatment were excluded from analysis, the trend in favor of nondrinkers for SVR disappeared (25% vs 23%). These findings were also consistent in subgroup analyses on race and genotype.

Conclusions: Eligibility for anti-HCV treatment was reduced in past and recent drinkers. Recent alcohol use was associated with increased treatment discontinuation and lower SVR. However, patients who use alcohol and completed the treatment had a response comparable to that of nondrinkers. Patients with a history of alcohol use should not be excluded from HCV therapy. Instead, additional support should be provided to these patients to ensure their ability to complete treatment.


Chronic hepatitis C is a common cause of end-stage liver disease, cirrhosis, and hepatocellular carcinoma and is the leading indication for liver transplantation.1 Left untreated, significant costs related to hepatitis C virus (HCV) infection in both direct medical care and indirect losses are projected.2, 3 At the same time, antiviral therapy continues to improve, with sustained virologic response (SVR) rates nearing 60% in randomized controlled trials.4, 5 However, as recent studies make clear, response to therapy in clinical populations is well below that seen in controlled trials.6, 7, 8 This is likely due to several factors, including patients not fully engaging in care; the impact of comorbid psychiatric, substance use, and medical problems; and difficulties in adhering to complicated treatment regimens.7, 9 Substance use disorders, particularly alcohol use, are a common reason for excluding patients from antiviral therapy. Typically, clinicians require complete abstinence from alcohol for at least 6 months before initiating antiviral treatment. However, few data exist in patients with hepatitis C regarding the impact of alcohol use on adherence to antiviral therapy or its influence on treatment outcomes.


See editorial on page 1912.


Alcohol abuse and HCV frequently coexist in the same patient. HCV infection is detected in 16% to more than 40% of alcoholic patients with or without liver disease.10, 11, 12, 13 Several studies have shown increased histologic liver damage in chronic alcoholic patients with HCV infection, in the form of higher rate of fibrosis progression and development of cirrhosis compared with HCV infection in nondrinking subjects.14, 15, 16, 17, 18, 19 However, the effect of any amount of alcohol use on antiviral treatment outcomes is much less clear. Several studies suggest that cumulative lifetime alcohol use worsens antiviral treatment outcomes.20, 21, 22, 23 In one of these studies, Ohnishi et al suggested that prolonged abstinence from alcohol improves interferon treatment outcomes.20 In an Italian study, Loguercio et al showed that the SVR decreased as alcohol use increased.24 These studies have several limitations, including small sample size, retrospective data analysis, and inclusion of patients treated with interferon monotherapy. None of these studies assessed the effect of alcohol use on treatment compliance and treatment outcomes with interferon and ribavirin combination therapy. The present study was performed to determine prospectively the effect of various levels of alcohol intake on treatment outcomes in patients with HCV infection.

Materials and Methods



This was a multicenter study to evaluate the epidemiology, natural history, and treatment response in unselected patients with HCV infection. The criteria for patient enrollment in the screening phase included a positive antibody to HCV serologic test and patient consent. The study was approved by the institutional review boards at each participating site.

Screening Phase


At enrollment, the patients were screened to determine whether or not they were suitable candidates for treatment with interferon and ribavirin. Additionally, patient counseling on HCV disease and risk factor assessments were performed. The inclusion criteria for treatment were male or female patients 18 years of age or older, a positive HCV RNA level by qualitative or quantitative polymerase chain reaction, and no previous HCV treatment with interferon and ribavirin combination therapy. The treatment exclusion criteria included patients with active hepatitis B virus infection and other forms of liver disease such as hemochromatosis, Wilson’s disease, and α1-antitrypsin deficiency; hemoglobinopathies; evidence of advanced liver disease or hepatocellular carcinoma; preexisting uncontrolled severe depression or other psychiatric diseases; evidence of ischemia or significant cardiac disease; or other comorbidities such as renal disease.


All patients were counseled to stop drinking and were provided with detailed counseling about the risks associated with continued use of drugs and alcohol and the factors associated with virus transmission. Patients with ongoing substance or alcohol use and those who declined to quit their substance use were not considered for treatment. Practitioners, in conjunction with patients, discussed treatment candidacy eligibility based on the risk factors, laboratory results, and standardized treatment inclusion and exclusion criteria. Patients who were determined to be treatment candidates and agreed to proceed with the recommended therapy entered the treatment phase of the protocol. In patients not selected for therapy, the caregivers were required to state why a patient was not considered a treatment candidate.

Alcohol Questionnaire


The patients completed a questionnaire on alcohol use during the screening phase of the study (see Appendix). Information collected included the type of alcoholic beverage usually used (beer, wine, mixed drinks), the highest number of drinks consumed on a regular basis, the total duration of alcohol use, and the number of drinks usually consumed during the past 12 months. In addition, the CAGE score was determined for each patient. The CAGE score is a validated screening instrument designed to assess alcohol dependence or abuse.25 The total CAGE score ranges from 0 to 4, based on responses to the following 4 questions: have you ever felt you should cut down your drinking, have people ever annoyed you by criticizing your drinking, have you ever felt bad or guilty about your drinking, and have you ever had a drink first thing in the morning to steady your nerves or to get rid of a hangover (an eye opener).

Categorization of Alcohol Use


Alcohol use was categorized in 4 different ways. In category 1, the patients were divided into nondrinkers and drinkers. Nondrinkers included patients who either did not drink at all or drank only occasionally, whereas drinkers included patients who consumed alcohol on a regular basis, which was defined as at least one drink a day. In category 2, the patients were divided into 3 subgroups: nondrinkers (as defined previously), moderate drinkers (<6 drinks a day), and heavy drinkers (≥6 drinks/day). In category 3, the patients were divided into 2 groups based on the CAGE score (CAGE score <2 or ≥2). In category 4, the patients were divided into 2 groups based on current alcohol use patterns (active drinking within the past 12 months and not actively drinking). Thus, categories 1, 2, and 3 were based on the alcohol use patterns of patients in the past (beyond 1 year), while category 4 assessed drinking status within the past 12 months at the time of baseline enrollment. All patients who were actively drinking were advised to stop alcohol use, and those who agreed were considered for therapy.

Treatment Phase


Patients selected for treatment were started on interferon alfa 2b 3 million units 3 times a day and ribavirin 1000–1200 mg/day (1000 mg for weight ≤75 kg and 1200 mg for weight ≥75 kg) for 24 weeks (genotypes 2 and 3) and 48 weeks (all other genotypes). Patients were assessed for efficacy and safety of treatment by regular follow-up visits scheduled at week 1–2 and every 4 weeks after initiation of therapy. At each follow-up visit, the patients were questioned about drug-related adverse effects, and blood tests were performed for hematologic and biochemical assessment. The study coordinator was required to fill out a form indicating compliance with therapy at weeks 12, 24, and 48 during the treatment phase of the study. The primary efficacy end point of treatment was SVR, which was defined as undetectable plasma HCV RNA (<100 copies/mL by qualitative polymerase chain reaction assay; Amplicor Roche Diagnostics, Nutley, NJ) 24 weeks after the end of treatment. Viral response was analyzed on an intention-to-treat basis, and lack of HCV RNA data was counted as viral nonresponse. The secondary end points were the end-of-treatment response (ETR) at treatment week 24 (for genotypes 2 and 3) or week 48 (genotype 1) and early treatment discontinuation (defined as stopping treatment with interferon and ribavirin before the scheduled end of therapy).

Statistical Analysis


Statistical analysis was conducted using SAS version 8.2 for Windows (SAS Institute, Inc, Cary, NC). Prevalence and 95% confidence intervals (CIs) were calculated based on binomial distributions. Summary statistics included frequency tables for categorical variables and medians and interquartile ranges for continuous variables. Continuous variables such as age and cumulative time of infection were categorized for contingency analyses. Bivariate analyses examining associations between antibody to HCV status and demographic, risk exposure, and alcohol-related variables were conducted using the χ2 test of association, and odds ratios (ORs) and 95% CIs were calculated. Mantel–Haenszel χ2 calculations were conducted to test for associations when 3 alcohol dose variables were used, such as nondrinkers, light drinkers (≤2 drinks/day), and moderate/heavy drinkers (>2 drinks/day). Independent associations between the dependent and independent variables were assessed using unconditional logistic regression analyses. Similar Mantel–Haenszel χ2 analysis was conducted on 2 subgroups, persons with genotype 1 and black patients, to test for associations between recent drinkers and nondrinkers in these groups on SVR and early treatment discontinuation. Severity of liver disease at baseline (Table 1), categorized by regular alcohol use of drinkers versus nondrinkers who were considered for treatment, was measured using the medium interquartile range due to the nonnormal distribution of the laboratory values. The nonparametric test, Wilcoxon–Mann–Whitney test, was then used to compare associations between the 2 groups accordingly.

Results


Twenty-four Veterans Administration hospitals throughout the United States participated in the study. The study was performed from 1999 to 2001. A total of 4462 patients with a positive HCV antibody test were enrolled at baseline. Of these, 4061 provided complete data on the alcohol use questionnaire. All patients were assessed for treatment with interferon alfa 2b and ribavirin based on the inclusion and exclusion criteria described previously. A total of 986 patients (24.2%) were found to be suitable candidates for treatment, and 726 (17.9%) were started on antiviral therapy. The common reasons for not being a treatment candidate were ongoing or recent substance abuse (20.2%), active psychiatric disease (18.3%), and comorbid medical disease (17.9%). The remaining patients who were considered treatment candidates but were not treated included those who declined or deferred treatment, those who were enrolled in other clinical research trials, and those who were lost to follow-up. In addition, 28 patients with combined human immunodeficiency virus and HCV infections were excluded from the 726 patients analyzed in the present study.

Patient Demographics


For the group as a whole, the mean (±SD) age of the patients was 50.3 (±7.6) years. There were 701 men (96.6%) and 25 women (3.4%). The racial profile of the patients consisted of 452 (62.5%) non-Hispanic white subjects, 166 (22.9%) non-Hispanic black subjects, and 78 (10.7%) nonblack Latino subjects; the remaining patients belonged to other racial groups such as Asian and American Indian. The duration of the infection was 25.8 (±8.6) years. The age at infection was 23.3 (±9.0) years. The mean weight was 91.9 (±18) kg (lower quartile, 79.4; upper quartile, 102.4), and the mean body mass index was 29.3 (±5.2) kg/m2 (lower quartile, 25.8; upper quartile, 32.1). The most common source of infection was intravenous drug use, seen in 60% of patients. The majority of patients were Vietnam-era veterans (76.8%). At liver histology, advanced fibrosis in the form of bridging fibrosis (stage 3) and cirrhosis (stage 4) was seen in 37.3% of patients. Genotype 1 was present in the majority of patients (74.8%). There was no statistical difference between patients who were offered and accepted the treatment and those who were offered and did not accept treatment with respect to any of the demographic features described.

Demographics and Treatment Candidacy


There were only minor differences in the demographic characteristics of drinkers compared with nondrinkers who were accepted for treatment (Table 2). Drinkers were more often men and were more likely to have a history of injection drug use (P = .024), incarceration (P < .001), and cocaine use (P < .001) compared with nondrinkers. Other risk factors, service in Vietnam, duration of infection, age at infection, and racial profile were similar in the 2 groups.

Status of Liver Disease and Treatment Candidacy


The baseline liver test results in drinkers and nondrinkers who were accepted for treatment are shown in Table 2. There were no significant differences in any of the liver test results and platelet counts between the 2 groups. Moreover, liver fibrosis assessed by staging the severity from 0 (no fibrosis) to 4 (cirrhosis) showed no difference between the 2 groups.

Alcohol Use and Treatment Candidacy


Alcohol use had a negative impact on a patient being considered for antiviral therapy, as judged by all 4 categories of alcohol use (Table 3). The most significant effect was observed in patients with a history of recent alcohol use. However, alcohol use within the past year did not automatically disqualify a patient from treatment. For example, 20% of patients who admitted to recent drinking were enrolled in the treatment phase of the study (vs 27% of nondrinkers). When the amount of daily alcohol use was assessed, patients who admitted to heavy drinking (≥6 drinks/day) in the past were just as likely to be treated (23%) compared with those drinking <6 drinks/day (24%).

Alcohol Use and Response to Antiviral Therapy


The effect of alcohol use on the response to treatment is shown in Table 4. At the end of the treatment phase of the study, 218 patients (30%) were HCV polymerase chain reaction negative. The ETR in heavy alcohol users in the past (≥6 drinks/day) was not significantly different compared with nondrinkers and moderate drinkers (31% vs 32% vs 27%, respectively). Similarly, recent alcohol use did not impact the ETR rates compared with those not drinking within the past 12 months (28% vs 31%). After discontinuation of therapy, 12% of patients had a virologic relapse, resulting in an overall SVR rate of 18% (133/726 patients). The SVR was not significantly different between heavy drinkers, moderate drinkers, and nondrinkers (19% vs 15% vs 20%, respectively) (Table 3). The SVR rate tended to be lower in patients drinking within the past 12 months compared with nondrinkers, but the difference failed to reach statistical significance (14% vs 20%; P = .06).

Alcohol Use and Adherence to Antiviral Treatment


Treatment adherence, as defined by the patient’s ability to complete antiviral treatment, is shown in Table 4. For the group as a whole, 215 of the 726 patients (30%) started on treatment failed to complete the full course of therapy. A history of alcohol use did not have a significant impact on treatment completion rates except in recent drinkers. In this group, the early discontinuation rate was significantly higher compared with nondrinkers (40% vs 26%; P < .0002). The treatment discontinuation rate for current drinkers who had high CAGE scores was not significantly different from that for patients with lower CAGE scores (42% vs 38%; P = .55). Moreover, the treatment discontinuation rate was not influenced by the amount of alcohol used in the past; there was no difference in the early discontinuation rate between moderate and heavy drinkers (33% vs 29%, respectively).

Effect of Different Amounts of Alcohol Use on Treatment Outcomes in Recent Drinkers


Because recent alcohol use may have a greater impact on treatment outcomes, patients drinking in the past 12 months were divided into 3 groups: nondrinkers, light (social) drinkers (≤2 drinks/day), and moderate/heavy drinkers (>2 drinks/day). The results are shown in Table 5. There was a significant negative effect on the treatment discontinuation rates between abstinent subjects compared with light and moderate/heavy drinkers (P < .001). There was no significant difference on the ETR response between the 3 groups. There was a trend toward a better SVR rate based on the dose of alcohol intake: 20%, 16%, and 13% in nondrinkers, light drinkers, and moderate/heavy drinkers, respectively. However, these differences failed to reach statistical significance. When patients who discontinued treatment early were excluded from the analysis, the trend in favor of nondrinkers for SVR disappeared compared with those who were drinking in the past year (Table 6). The most common reason for treatment discontinuation was the development of adverse events. Other reasons included a patient’s decision to no longer continue treatment (unrelated to adverse events), an inability to continue treatment due to incarceration, and patients who were lost to follow-up. There was no difference in the treatment discontinuation rates between nondrinkers, light drinkers, and heavy drinkers with respect to any of the previously described factors (Table 7).

Picture 1.png

Picture 2.png

Alcohol Use, Race, and Genotype


The interaction between the patients’ race and genotype with alcohol use was analyzed. For the whole group, black subjects were more frequently infected with genotype 1 (87% vs 65%; P < .001) and were less likely to have an SVR (8.4% vs 21.6%; P < .001) compared with nonblack subjects. Black subjects with genotype 1 had a lower SVR rate than nonblack subjects (6.1% vs 14.1%; P = .004), but there was no difference in the SVR rate for genotypes 2 and 3 (nonblack subjects vs black subjects, 50.0% vs 36.5%, respectively; P = .47). The effect of race and genotype on treatment outcomes is shown in Table 8. The ETR and SVR rates were better in non–genotype 1 and in nonblack patients, but there was no significant difference in the treatment outcomes (ETR, SVR, and early discontinuation of therapy) between drinkers and nondrinkers based on all 4 categories of alcohol use. The impact of viral level (divided into <1 million IU/mL and ≥1 million IU/mL) in patients with genotype 1 on treatment response was assessed. Patients with genotype 1 and viral level <1 million IU/mL had better ETR (24% vs 16%; P < .05) and SVR (18% vs 9%; P < .009) compared with those with a viral level >1 million IU/mL.

Discussion


Despite the frequent presence of HCV infection in subjects with chronic alcoholism, the impact of alcohol use on treatment outcomes with antiviral agents is poorly understood. Substance use disorders, particularly alcohol use, are a common reason for excluding patients from antiviral therapy.26 Indeed, all the large multicenter registration trials designed to assess the efficacy of interferon alone or combined with ribavirin for the treatment of HCV infection either completely excluded patients with a history of alcohol abuse or placed a restriction on prior alcohol use to a minimum of 6 months of abstinence. These requirements severely limited the inclusion of such subjects in well-conducted clinical trials. The current data on the effect of alcohol use on HCV treatment outcomes therefore are based on mostly retrospective studies performed on a small number of subjects.20, 21, 22, 23, 24


This is the first prospective study to assess in a large cohort of patients the impact of past and recent alcohol use on antiviral treatment candidacy, adherence, and outcomes. Our study provided several interesting results. Alcohol use was extremely common in patients treated with antiviral therapy, with 80% (584/726) drinking on a regular basis in the past and 27% (194/726) in the past year. The reasons for this decline in alcohol use in the past year are not clear but may be related to more frequent patient interaction with medical practitioners because of their HCV infection.


Alcohol use had a significant negative impact on treatment candidacy. Compared with nondrinkers, physicians were less inclined to offer treatment to patients with a history of alcohol use despite being abstinent for at least 1 year before therapy. As expected, recent drinkers were even less likely to be offered therapy. The overall ETR and SVR rates were 30% (218/726) and 18% (133/726), respectively. These rates are lower than those reported in registration trials, in which the ETR and SVR rates ranged between 50% and 61% and between 36% and 47%, respectively.27, 28, 29, 30 The virologic response rates in the present study cannot be compared with the large registration trials because the latter included highly selected patients, including a very low enrollment of black Americans, who respond less well to antiviral therapy.31, 32 Other investigators have also reported response rates that were well below those observed in the registration trials.6, 7, 8


Alcohol use did not have any impact on the ETR, SVR, or treatment discontinuation rates in patients who had stopped drinking at least 1 year prior. By contrast, in recent drinkers, the treatment discontinuation rate was higher (P = .0002) and the SVR rate was lower but not significantly (P = .06) compared with nondrinkers. However, when patients who had failed to complete the full course of therapy were excluded from the analysis, there was no difference in the SVR rate between recent drinkers and nondrinkers, indicating that alcohol use did not interfere with the efficacy of antiviral therapy. In a recent study, alcohol use among human immunodeficiency virus–infected persons was associated with lower adherence to antiretroviral therapy compared with nondrinkers.33 Therefore, rather than excluding alcohol users from antiviral therapies, strategies should be developed to improve treatment adherence in this patient group. Fortunately, simple and effective methods exist, such as brief behavioral counseling interventions for reducing alcohol use.34 These techniques can be expanded to include treatment adherence and compliance.


The 2002 National Institutes of Health Consensus Conference on HCV made a special mention of the lack of data on the impact of light (social) alcohol use, defined as the consumption of ≤2 drinks/day.35 In the present study, we specifically analyzed the influence of light alcohol use on HCV treatment outcomes. Light drinking had no effect on the ETR (35% vs 31%) and SVR (16% vs 20%) rates compared with nondrinkers, but the treatment discontinuation rate was higher compared with nondrinkers (35% vs 26%; P < .0001). In addition, we did not find any interaction between viral genotype and patients’ racial background with alcohol use and treatment outcomes.


Our study had several limitations. A precise estimate of alcohol use in a clinical setting is fraught with problems, such as subject compliance and accuracy of recall. However, we believe the present data provide a fairly accurate estimate of the past and recent drinking habits of the study subjects. Alcohol use was categorized in several ways, including the CAGE score, which is easy to administer in a clinical setting; a score of ≥2 is associated with a sensitivity and specificity of 74% and 91%, respectively, for alcohol abuse or dependence.25 When selecting patients for therapy, there was no fixed period of abstinence before initiation of treatment and the decision was left to the treating physicians. Therefore, patients with differing periods of abstinence before therapy were included in the study. During treatment, although patients were counseled against drug and alcohol use at each follow-up visit, no objective assessment was made of continuing alcohol use. The study medications consisted of interferon alfa and ribavirin, and our findings may not reflect the results obtained with pegylated interferons, the current standard of care for HCV treatment. Although the investigators were advised to follow the Veterans Administration treatment guidelines, personal bias in selecting patients for treatment cannot be completely excluded. However, we do not expect this to be any different from what occurs in routine clinical practice and in that respect is more representative of patient selection in “the real world,” outside of controlled clinical trials. Finally, being a study based on US veterans, our cohort was 95% male, which may have introduced gender-based inconsistency in the study results.


In conclusion, alcohol use is extremely common in patients with HCV being considered for treatment at Veterans Administration facilities. Our findings show that physicians are reluctant to treat patients with a history of alcohol use, even when they are abstinent for 1 year or more. The only negative impact of alcohol use on treatment outcomes was a higher treatment discontinuation rate in recent drinkers. Patients who were able to complete treatment had similar ETR and SVR rates compared with nondrinkers, irrespective of the duration of abstinence. These findings were consistent among the subgroups of genotype 1 and black patients. Our observations are clinically important because they indicate that alcohol use should not be considered an exclusion criterion when evaluating patients for anti-HCV treatment, especially in view of the fact that nearly one third of patients with HCV infection have a history of recent alcohol use. The current attitude among physicians against offering treatment to patients who use alcohol should be reassessed. Moreover, greater emphasis should be placed on developing strategies to improve patient education and support to improve adherence to antiviral therapy.