Direct-acting antiviral treatment uptake and sustained virological response outcomes are not affected by alcohol use: A CANUHC analysis

Abstract

Background

Alcohol use and hepatitis C virus (HCV) are two leading causes of liver disease. Alcohol use is prevalent among the HCV-infected population and accelerates the progression of HCV-related liver disease. Despite barriers to care faced by HCV-infected patients who use alcohol, few studies have analyzed uptake of direct-acting antiviral (DAA) treatment.

Objective

We compared rates of treatment uptake and sustained virological response (SVR) between patients with and without alcohol use.

Methods

Prospective data were obtained from the Canadian Network Undertaking against Hepatitis C (CANUHC) cohort. Consenting patients assessed for DAA treatment between January 2016 and December 2019 were included. Demographic and clinical characteristics were compared between patients with and without alcohol use by means of t-tests, χ² tests, and Fisher’s Exact Tests. Univariate and multivariate analyses were used to determine predictors of SVR and treatment initiation.

Results

Current alcohol use was reported for 217 of 725 (30%) patients. The proportion of patients initiating DAA treatment did not vary by alcohol use status (82% versus 83%; p = 0.99). SVR rate was similar between patients with alcohol use and patients without alcohol use (92% versus 94%; p = 0.45). Univariate and multivariate analysis found no association between alcohol use and SVR or treatment initiation.

Conclusion

Patients engaged in HCV treatment have highly favourable treatment uptake and outcomes regardless of alcohol use. Public health interventions should be directed toward facilitating access to care for all patients irrespective of alcohol use. Research into high-level alcohol use and DAA outcomes is needed.

Introduction

Alcohol use and hepatitis C virus (HCV) infection are among the most common etiologies of chronic liver disease and liver cancer worldwide (1,2), with alcohol-related liver disease recently surpassing HCV as the most common indication for liver transplant in the United States (3,4). Excess alcohol consumption is prevalent among HCV-infected patients, as is HCV infection among patients with alcohol use disorder (5,6). Concurrent alcohol use among HCV-infected individuals is associated with increased mortality (7), and alcohol use accelerates the progression of HCV-related liver disease in a synergistic and dose-dependent manner (8–12). Even moderate alcohol consumption has been associated with fibrosis among HCV-infected patients (13–15). The histological features of fibrosis in this population resemble those of HCV infection alone, suggesting that the hepatotoxic effect of alcohol among HCV-infected patients is mediated by potentiating HCV-induced liver disease (16). HCV eradication among these patients may confer a particularly important benefit by interrupting this mechanism.

Despite this potential clinical benefit, HCV treatment programs have historically failed to engage patients with ongoing alcohol use. The efficacy of interferon-based regimens was diminished in patients with active alcohol use (17), and these regimens were typically contraindicated among patients with excess alcohol consumption (18,19). In the era of direct-acting antivirals (DAAs), efficacy data for patients with ongoing alcohol use is limited. A small number of retrospective studies that included alcohol use data demonstrated no difference in sustained virological response (SVR) rates by alcohol use (20–22), with one study demonstrating slightly decreased SVR only among heavy drinkers (23). Of three prospective studies (24–26), one study demonstrated reduced SVR among patients with alcohol use, largely as a result of loss to follow-up (26). Despite the generally promising efficacy data in this population, a number of retrospective cohorts have identified alcohol use as a negative predictor of treatment uptake (27–29), whereas other prospective cohorts have found no difference (30,31). To date, there have been no designated analyses of treatment uptake specifically among patients who use alcohol.

We sought to compare rates of DAA uptake and SVR between patients with and without alcohol use in a prospective, nationwide cohort of patients with chronic HCV infection.

Methods

The Canadian Network Undertaking against Hepatitis C (CANUHC) cohort contains prospectively collected demographic information and HCV DAA treatment information collected from 10 Canadian-based, publicly funded sites to which patients are referred for evaluation and treatment of HCV. All patients with an initial intake visit for assessment for DAA therapy from January 2016 to December 2019 who provided written, informed consent were enrolled. The cohort obtained research ethics board approval from the Health Research Ethics Board of Alberta–Community Health Committee (HREBA.CHC-16-0038_REN3), in addition to local ethics board approval for each participating site.

Data were obtained from treatment forms collected over the course of treatment. Forms were completed at intake, pre-treatment, post-treatment, and annual follow-up visits. Information collected included demographic information; comorbidities; risk factors for HCV infection; HCV clinical data (HCV treatment history, genotype, viral load, fibrosis stage, and SVR-12 status); and other laboratory values such as hemoglobin, platelets, international normalized ratio, alanine aminotransferase (ALT), aspartate aminotransferase (AST), bilirubin, albumin, and glomerular filtration rate.

Self-reported alcohol use data were defined by the presence of any use and the number of drinks per day, week, and month. Excess alcohol use was defined as more than 3 alcoholic units per day, more than 20 alcoholic units per week, or the presence of binge drinking, defined as 5 or more alcoholic units for men (4 or more for women) on the same occasion (32). Demographic and baseline clinical data included gender, race, genotype, fibrosis stage, previous antiviral treatment status, and DAA initiation. HCV infection was defined by HCV RNA positivity. Fibrosis was determined by liver stiffness measures via transient elastography (FibroScan). DAA initiation was defined as having received any DAA prescription. SVR was determined by intent-to-treat analysis, where all patients who initiated DAA therapy were included, and treatment failure was assumed for all patients who were lost to follow-up with unknown SVR status.

Clinical and treatment data were described by means, standard deviations, and percentages; t-tests were used to compare continuous variables between patients with and without alcohol use, and categorical variables were compared using χ² tests or Fisher’s Exact Tests. Univariate analysis was used to determine clinical and demographic characteristics associated with SVR and treatment initiation. Multivariate analysis was then used to control for confounding. The level of significance was defined as p < 0.05 (two-tailed). IBM SPSS Statistics (Version 27; IBM Corp., Armonk, NY, USA) was used for statistical analysis.

Results

A total of 725 HCV-infected patients under assessment for DAA therapy were enrolled in CANUHC. The mean age of participants was 52.7 (SD 12.7) years, and the majority were male (66%) and White (78%).

Current alcohol use was reported for 217 (30%) patients, and 37% of patients reported either past or current use. Among the 217 patients with current alcohol use, 24 (11%) patients reported excessive use. The mean age of patients with current alcohol use was greater than that of patients without current alcohol use (54.5 vs. 51.2 y; p < 0.01). Alcohol use also varied by race (p < 0.01). Baseline demographic characteristics did not otherwise vary by alcohol use status (Table 1).

Table 1:

Baseline characteristics

Characteristic	n (%)*		p
	Current alcohol use (n = 217)	No current alcohol use (n = 508)
Male sex	149 (69)	327 (64)	0.26
Age, y, mean (SD)	54.5 (11.9)	51.2 (12.9)	<0.01
Race†			0.03
White	160 (80)	353 (78)
Black	11 (6)	10 (2)
East Indian	7 (4)	22 (5)
South East Asian	3 (2)	26 (6)
Indigenous	6 (3)	21 (5)
Other	7 (4)	22 (5)
Genotype‡			0.10
1a	96 (45)	239 (49)
1b	36 (17)	55 (11)
1 other subtype	5 (2)	9 (2)
2	30 (14)	42 (9)
3	42 (20)	124 (26)
4	3 (1)	12 (2)
6	2 (<1)	5 (1)
Fibrosis stage§			0.81
F1	80 (42)	178 (39)
F2	43 (23)	94 (21)
F3	25 (13)	64 (14)
F4	43 (23)	115 (25)
Antiviral treatment naive¶	83 (71)	201 (68)	0.52
Initiated a DAA regimen**	179 (82)	417 (82)	0.99
SVR (intent-to-treat)††	140 (94)	272 (92)	0.45

Note: Percentages may not total 100 because of rounding

^*Unless otherwise indicated

^†ns = 194 and 454 for current alcohol use and no current alcohol use, respectively

^‡ns = 214 and 486 for current alcohol use and no current alcohol use, respectively

§ ns = 191 and 451 for current alcohol use and no current alcohol use, respectively

¶ ns = 117 and 297 for current alcohol use and no current alcohol use, respectively

^**ns = 217 and 508 for current alcohol use and no current alcohol use, respectively

^††ns = 149 and 297 for current alcohol use and no current alcohol use, respectively

DAA = Direct-acting antiviral; SVR = Sustained virological response

Baseline laboratory measures were compared by alcohol use status (Table 2). Mean liver stiffness was numerically greater in patients with current alcohol use. However, this trend was not statistically significant (11.1 kPa versus 9.7 kPa; p = 0.06). No between-groups differences were observed with respect to HCV RNA level, liver function, or renal function.

Table 2:

Baseline laboratory measures

Measure	Mean (SD)		p
	Current alcohol use	No current alcohol use
HCV RNA, IU/mL	1.86 × 106 (3.7 × 10 6)	1.51 × 106 (2.57 × 106)	0.50
Hemoglobin, g/L	145 (24)	137 (29)	0.01
Platelets, 109/L	204 (73)	207 (81)	0.66
ALT, IU/L	79 (83)	75 (111)	0.64
AST, IU/L	65 (66)	61 (63)	0.43
Bilirubin, mmol/L	12 (9)	11 (8)	0.10
Liver stiffness, kPa	11.1 (10.1)	9.7 (7.6)	0.06
eGFR, mL/mon/ 1.73 m2	105 (146)	100 (96)	0.62

HCV = Hepatitis C virus; ALT = Alanine aminotransferase; AST = Aspartate aminotransferase; eGFR = Estimated glomerular filtration rate

Comorbidities and social determinants of health are presented in Table 3 and Table 4. There were no between-groups differences in comorbidities, including HIV co-infection, hepatocellular carcinoma, diabetes, and psychiatric illness. The proportion of patients who were employed (37% versus 33%; p = 0.33) and had a history of incarceration (22% versus 23%; p = 0.64) did not vary by alcohol status. Patients with current alcohol use were less likely to report use of injection drugs (54% versus 65%; p = 0.02) or other recreational drugs (31% versus 42%; p < 0.01).

Table 3:

Concurrent comorbidities

Comorbidity	n/N (%)		p
	Current alcohol use	No current alcohol use
Hepatocellular carcinoma	1/216 (<1)	7/501 (1)	0.28
HIV co-infection	7/49 (14)	20/132 (15)	0.21
Chronic renal disease	2/217 (<1)	2/508 (<1)	0.38
Diabetes	17/217 (8)	35/508 (7)	0.65
Psychiatric illness	42/217 (19)	114/508 (22)	0.35

Table 4:

Social determinants of health

Social determinant	n/N (%)		p
	Current alcohol use	No current alcohol use
Employed currently	69/186 (37)	126/382 (33)	0.33
Incarceration history	47/217 (22)	118/508 (23)	0.64
IDU history	81/150 (54)	193/295 (65)	0.02
Recreational drug use history	67/217 (31)	231/508 (46)	<0.01
Excess alcohol use currently	24/217 (11)	N/A

IDU = Injection drug use; N/A = Not applicable

DAA treatment was initiated in 596 of 725 (82%) included participants, and SVR was achieved by 412 of 446 patients (92%) who had completed therapy and had HCV RNA bloodwork drawn 12 or more weeks after completion of treatment. The proportion of patients initiating DAA therapy did not differ by alcohol use (82% versus 83%; p = 0.99). Patients with alcohol use had similar rates of SVR as patients without current alcohol use (92% versus 94%; p = 0.45).

Univariate and multivariate analysis were conducted to assess predictors of SVR (Table 5). Current alcohol use did not predict SVR in univariate (OR 1.43, 95% CI 0.65–3.15) or multivariate analyses (OR 1.27, 95% CI 0.53–3.03). Cirrhosis, as determined by FibroScan score of F4, was negatively associated with SVR in univariate analysis (OR = 0.40, 95% CI 0.19–0.83). This association remained significant in multivariate analysis (OR = 0.39, 95% CI 0.17–0.90). Age, gender, and HCV genotype did not predict SVR in univariate or multivariate analyses.

Table 5:

Univariate and multivariate analysis of predictors of SVR

Characteristic	Univariate analysis, AOR (95% CI)	Multivariate analysis OR (95% CI)
Current alcohol use	1.43 (0.65–3.15)	1.27 (0.53–3.03)
Male	0.59 (0.20–1.68)	0.69 (0.27–1.76)
Race
Caucasian	0.65 (0.08–5.05)	0.30 (0.03–2.94)
Black	0.22 (0.02–2.31)	0.13 (0.01–1.89)
Asian	0.60 (0.03–10.50)	0.24 (0.02–4.88)
Hispanic	0.90 (0.05–11.67)	0.38 (0.02–9.12)
Indigenous (Ref.)	1.00	1.00
Cirrhosis by FibroScan*	0.40 (0.19–0.83)	0.39 (0.17–0.90)
Age	0.99 (0.97–1.02)	0.99 (0.97–1.03)
Genotype
1a	1.51 (0.23–7.57)	2.56 (0.45–11.01)
1b	0.58 (0.12–3.12)	1.07 (0.17–6.68)
2	1.78 (0.24–12.15)	3.15 (0.35–11.98)
3	0.74 (0.16–3.78)	1.02 (0.16–6.31)
Other (Ref.)	1.00	1.00

^*Defined as >12.5 kPa

SVR = Sustained virological response; Ref. = Reference

Demographic characteristics were generally similar among those with current excess alcohol use: 25 of 217 (12%) were current alcohol users (mean age 52.4 [10.5] y, 76% male); mean liver stiffness was 10.1 (6.3) kP, and 26% had cirrhosis. A larger proportion were White (81%), and ALT (90 [75] IU/ml) and AST (78 [56] IU/mL) were numerically higher among those currently consuming excess alcohol. DAA treatment initiation (83% versus 87%; p = 0.63) and SVR (93% versus 91%; p = 0.78) were similar between those consuming excess alcohol and those who were not. Regression analysis was not conducted, given the small sample size.

Discussion

Our study, representing HCV-infected patients from across Canada, is among the first and largest real-world cohort analyses to evaluate the effectiveness and uptake of DAAs on the basis of alcohol use status. We demonstrated that although alcohol use is prevalent among patients living with HCV, those who report current alcohol consumption are equally likely to initiate DAA treatment and to achieve SVR.

Many of the dedicated studies on DAA efficacy among patients using alcohol consist of single-centre analyses involving clinics that are directed toward at-risk populations, hence limiting the generalizability of existing evidence to the broader HCV-infected population (20,24,31). One retrospective study with a sample size of 29,037 identified no difference in SVR by no alcohol use, low-level use, and unhealthy use while on HCV DAA treatment (33). This study included only US military veterans, and the results may not be generalizable to other demographic groups living with HCV. Our large real-world cohort addresses this research gap by obtaining data from all patients considered for HCV treatment in centres across Canada. With this in mind, we believe that our analysis is generalizable to the broader HCV-infected population. Our primary finding, that DAA therapies are highly curative irrespective of previous or ongoing alcohol use, provides robust support to other single-site and multicentre cohort analyses (22,25,30,33). The consistency of findings across studies despite differing settings, population characteristics, and definitions of alcohol use is noteworthy.

Taken altogether, the equally favourable outcomes of patients using alcohol contributes strong evidence in favour of increasing access to DAA therapy for HCV-infected patients irrespective of alcohol use. In 2016, 25 US states maintained Medicaid reimbursement policies that include alcohol- and drug-based restrictions from accessing DAAs (34). Such restrictions contribute to the large proportion of patients denied coverage for DAAs in the United States, by both private and public insurers (35). In Canada, provincial drug coverage programs do not include restrictions to DAAs on the basis of alcohol use.

Given that the efficacy and tolerability of DAA therapies is now firmly established, the ongoing burden and incidence of HCV has motivated a shift in research priorities toward accessibility of testing and treatment (36). Likewise, our study demonstrates similarly high DAA efficacy and treatment uptake among those patients who are evidently already engaged in health care through participating CANUHC clinics. Therefore, our study provides important evidence that the remaining challenge in reducing the burden of HCV among those who consume alcohol does not require improvements in cure rates or follow-up but rather the facilitation of initial access to care. In Canada, patients with alcohol or substance misuse are more likely to face barriers in accessing health care, including mental health needs, housing status, and other complex social factors (37). More important, patients without access to care were inherently excluded from our study. These challenges necessitate further qualitative research and targeted public health interventions such as integrative models of care (38), especially given the significant proportion of undiagnosed HCV infection among marginalized individuals (39).

Our results contrast with certain studies of high-risk HCV sub-populations. Two cohorts, including HIV–HCV co-infected patients and patients on opioid substitution therapy (23,26), reported reduced SVR rates among patients with excessive alcohol use, which was mainly attributable to loss to follow-up. In contrast, we demonstrate no difference in missing data or attrition in our cohort, as evidenced by the equally high intent-to-treat SVR rates. Similarly, regarding treatment uptake, other Canadian studies of persons who inject drugs and HIV–HCV co-infected patients have reported that alcohol use was associated with reduced uptake (27,40). This may suggest that alcohol use acts as a barrier to treatment initiation or follow-up only in the context of excessive consumption (23,26,27). Because our study included only 25 patients with excessive usage, we did not have an adequate sample size to determine the effect of excessive versus moderate alcohol use on treatment uptake or follow-up. These conflicting results may also reflect the specific sub-populations of these studies, each of which focused on patients with intravenous drug use, co-infection, or opioid substitution therapy. Alcohol use may have had a compounding impact on these groups, who already face unique challenges in engaging with care, given that marginalization and injection drug use are independent negative predictors of treatment initiation (41). However, it is noteworthy that among those in our cohort using substances with potential for abuse, there was little overlap between alcohol users and those using other recreational drugs.

Of the patients in our cohort, 30% reported current alcohol use. Given that even moderate alcohol consumption can exacerbate HCV-induced fibrosis (13–15), the prevalence of alcohol use reported in our study suggests this may be a key contributing factor to HCV-related liver disease in Canada. There was a trend toward increased fibrosis among patients with alcohol use despite a low proportion of excessive use (12%). This observation may lend further support to the hepatotoxic effects of moderate alcohol use among patients with HCV infection.

We did not analyze measures of adherence among patients who use alcohol. Previous prospective studies describing self-reported DAA adherence suggest that alcohol use is associated with missed doses. Despite this, overall adherence rates were high and not associated with SVR (42,43), which aligns with our finding of consistently high intent-to-treat SVR rates and low rate of loss to follow-up irrespective of alcohol consumption. Taken together, there is no evidence that reduced adherence rates associated with alcohol use are detrimental to SVR rates.

Certain limitations must be acknowledged. Although our use of real-world data improves the generalizability of our findings, it may also introduce sampling bias into our data. Patients registered in the CANUHC cohort are more engaged in the health care system, and so our findings may not reflect HCV-infected patients in Canada who face barriers in access to health care, many of whom may consume alcohol excessively (37). Our use of observational data precludes determining causality. Although we considered both any alcohol consumption and excess alcohol use at the time of treatment, more robust quantitative data related to alcohol may have provided additional insight into the effect, if any, of excessive use on HCV treatment uptake and SVR.

Conclusion

In a large cohort of HCV centres across Canada, HCV-infected patients achieved high rates of DAA uptake and SVR rates, and neither treatment initiation nor SVR rate varied by reported alcohol use. Alcohol use among HCV-infected patients is common. Our findings indicate that patients with alcohol use who are engaged in HCV treatment have highly favourable treatment outcomes and should be offered treatment irrespective of alcohol consumption. Removal of existing barriers to DAA access faced by patients who use alcohol is needed. Doing so is essential to reduce the ongoing public health burden of HCV in Canada and, ultimately, represents a necessary step toward achieving HCV eradication.

Funding Statement

Funding from Gilead and Merck in support of CANUHC is gratefully acknowledged.

Ethics Approval:

The cohort obtained research ethics board approval from the Health Research Ethics Board of Alberta–Community Health Committee (HREBA.CHC-16-0038_REN3), in addition to local ethics board approval for each participating site.

Informed Consent:

Informed consent was obtained from the patients.

Registry and Registration No. of the Study/Trial:

N/A

Funding:

Funding from Gilead and Merck in support of CANUHC is gratefully acknowledged.

Disclosures:

CC is an advisor and speaker with Merck, Gilead, and AbbVie.

Peer Review:

This article has been peer reviewed.