Pangenotypic direct acting antivirals for the treatment of chronic hepatitis C virus infection: A systematic literature review and meta-analysis

Abstract

Background

Recent approval and adoption of pangenotypic direct acting antivirals (DAAs) necessitated a revision of the 2015 World Health Organization guidelines for the management of persons with hepatitis C virus (HCV) infection.

Methods

We searched MEDLINE, EMBASE, CENTRAL, and relevant conference proceedings to identify randomized and non-randomized trials, as well as prospective observational studies of DAAs. The proportions of persons with events were pooled for sustained virological response at 12 weeks post-treatment (SVR12), discontinuations due to adverse events (DAEs), serious adverse events (SAEs), and all-cause mortality. Analyses were stratified by HCV genotype and antiviral treatment experience, with subgroup analyses based on presence of cirrhosis and HIV-HCV coinfection.

Findings

The evidence base consisted of 238 publications describing 142 studies. In the overall analysis, which included all persons irrespective of treatment experience or comorbidities, the pooled proportion achieving SVR12 exceeded 0.94 for all pangenotypic regimens across genotypes 1, 2, and 4. Some heterogeneity may have led to lower SVR rates in persons with genotype 3 infection. High SVR12 (>0.90) was observed in persons with genotype 1 infection with cirrhosis, though evidence varied and was limited for genotypes 2–4. Evidence was sparse for persons with HIV–HCV coinfection. All regimens were associated with small proportions of persons with DAEs, SAEs, or all-cause mortality.

Interpretation

Based on this and other supporting evidence, the WHO issued updated guidelines with a conditional recommendation, based on moderate quality evidence, for the use of pangenotypic DAA regimens for persons with chronic HCV infection aged 18 years and older (July 2018).

Funding

This study was funded by the World Health Organization.

Research in context.

Evidence before this study

All-oral, direct-acting antiviral (DAA) regimens for chronic hepatitis C virus (HCV) infection are associated with higher rates of virological cure (sustained virological response) and are better tolerated than older, interferon-based antiviral therapies.

Added value of this study

To support the World Health Organization (WHO) guidelines for the care and treatment of persons with chronic HCV infection, we present a comprehensive summary of the evidence on the pangenotypic regimens sofosbuvir-velpatasvir, sofosbuvir-daclatasvir, and glecaprevir-pibrentasvir as well as sofosbuvir-ledipasvir. Pooled analyses of trials and observational studies demonstrate high proportions of persons achieving SVR12 across genotypes 1 through 4. Findings were generally consistent across subgroups, including treatment-naïve and treatment-experienced persons, persons with cirrhosis, and persons with HIV–HCV coinfection, though evidence in some subgroups was limited. The proportions of persons with serious adverse events, treatment discontinuation due to adverse events, and all-cause mortality, were very low across treatments.

Implications of all the available evidence

The interventions recommended in the WHO's July 2018 guidelines have demonstrated high efficacy across genotypes with favourable harms profiles. Widespread adoption of pangenotypic regimens for the treatment of chronic HCV infection will simplify administration and treatment. Simple oral administrations and short treatment durations will further enhance retention and the effectiveness of public health programmes. This new era of pangenotypic DAAs is a welcome addition to the global strategy to combat HCV infection.

Alt-text: Unlabelled box

1. Introduction

In 2017, the World Health Organization established a target to eliminate chronic hepatitis C virus (HCV) infection by the year 2030. Indeed, a recent mathematical model suggests that by focusing public health programs on preventing infection in persons who do not inject drugs, providing harm reduction services to persons who inject drugs, and expanding HCV diagnosis services and treatments to 90% of infected persons, the global elimination goal is achievable by 2032 [1]. Nonetheless, challenges persist. Political barriers will need to be overcome while securing funding from national and international public health sources. This is complicated by diminishing investments in global health funding and trends toward universal health coverage and away from disease-specific programming [2]. Yet, several countries, such as Brazil and Australia, have developed innovative approaches to funding HCV programs [3,4].

Prior to 2014, HCV treatment centred on the use of interferon-based regimens with generally low cure rates, long durations of therapy, and substantial toxicity. The introduction of highly effective and well-tolerated short course oral direct-acting antiviral (DAA) therapy without interferon that can cure HCV infection with high rates of sustained virological response (SVR) within weeks transformed the treatment landscape. In 2016, the World Health Organization (WHO) updated its guidelines for the screening, care, and treatment of persons with HCV infection to recommend DAA-based regimens in place of IFN-based regimens. Since the publication of the 2016 guidelines, DAA regimens that do not require ribavirin have continued to improve and several pangenotypic regimens, which successfully resolve HCV infection in over 85% of treated individuals across all six major genotypes, were approved by regulatory bodies including the US Food and Drug Administration (FDA) and the European Medicines Agency (EMA). Successful resolution of chronic HCV infection is defined by the WHO, the European Association for the Study of the Liver (EASL), the US Centres for Disease Control and Prevention (CDC), and the American Association for the Study of Liver Diseases (AASLD) as SVR at 12 weeks following the end of treatment (SVR12). This outcome is associated with very low rates of virological relapse. Currently, the pangenotypic DAAs glecaprevir-pibrentasvir (8 week course), sofosbuvir-daclatasvir (12 week course), and sofosbuvir-velpatasvir (12 week course) are approved in most markets for the treatment of HCV-infected persons without cirrhosis.

The emergence of pangenotypic regimens presents new opportunities for the public health response to HCV infection, with simplified procurement, an omission of resource-intensive genotyping, and no need for frequent laboratory monitoring. Thus, the objective of this review was to identify and synthesize the evidence for the efficacy and safety of DAA regimens in adults with chronic HCV infection. Here, we present a subsection of the systematic literature review commissioned to support the WHO's Guidelines Development Group (GDG) in formulating the updated July 2018 guidelines [5]. The complete technical report has been published previously [6].

2. Methods

This systematic literature review and meta-analysis was conducted in accordance with PRISMA guidelines [7,8].

2.1. Search strategy and selection criteria

Systematic searches were conducted in MEDLINE, EMBASE, and the Cochrane Register of Controlled Trials (CENTRAL) to identify randomized controlled trials, non-randomized trials, and prospective observational studies of adults with chronic HCV infection published in English from March 2015 to July 2017. Conference proceedings from Digestive Diseases Week (DDW), the AASLD, and EASL were hand-searched. Studies included in a previous systematic literature review, commissioned by the WHO to support the April 2016 guidelines, were assessed for eligibility in the current review [9].

As the updated WHO guidelines recommend treatment with pangenotypic regimens sofosbuvir-velpatasvir, sofosbuvir-daclatasvir, glecaprevir-pibrentasvir, these regimens are the focus of this paper. We also describe the evidence for sofosbuvir-ledipasvir, a non-pangenotypic regimen commonly used in regions where only a single genotype is dominant. Given their high prevalence, this review focuses on persons with genotype 1–4 infection.

2.2. Data extraction and outcomes

All titles and abstracts, as well as the full text publications of included abstracts, were screened independently and in duplicate by two reviewers (MZ, AS). Data extraction of study characteristics and outcomes of included studies were performed independently and in duplicate by at least two reviewers (MZ, AS, RM, DZ). Outcomes extracted included the proportion of persons achieving SVR12 as well as the proportion of persons with serious adverse events (SAEs), discontinuations due to adverse events (DAEs), and all-cause mortality.

2.3. Data analysis

For each outcome, untransformed proportions of persons with events of interest were pooled to generate a point estimate with 95% confidence interval using the DerSimonian-Laird method (binary random effects model). A 0.5 correction was applied to zero-count cells [10]. Analyses were performed in OpenMeta[analyst] based on the Metafor package [11,12].

The primary analysis included all persons irrespective of treatment experience and comorbidities (all-comer analysis). Analyses were additionally stratified by treatment experience (treatment-experienced and treatment-naïve) where persons were considered treatment-experienced if they had received any prior HCV intervention, including interferon-based regimens and/or DAAs, or were classified as non-responders. Separate analyses, specified a priori, present the evidence for subgroups of persons with cirrhosis and for persons with HIV–HCV coinfection. We present analyses which include evidence from all eligible study designs.

2.4. Critical appraisal and grade

The validity of randomized trials was assessed by the Risk of Bias instrument, endorsed by the Cochrane collaboration [13]. Studies of other designs, including ‘single-arm’ trials, cohort studies, and observational studies, were evaluated using the Tool to Assess the Risk of Bias in Cohort Studies, developed by the CLARITY group at McMaster University [14].

The Grading Recommendations of Assessment, Development and Evaluation (GRADE) approach was used to assign a rating of high, moderate, low, or very low, to reflect the certainty of evidence for each outcome [15], [16], [17], [18], [19], [20]. The traditional approach was modified to suit the unique clinical and methodological characteristics of HCV research in the context of this review and to reflect the analysis approach, which combined both trial and non-trial evidence (Appendix A).

2.5. Role of the funding source

The study sponsor (World Health Organization) assisted in the conception and design of this review. However, the corresponding author had full access to all data in the study and had final responsibility for the decision to submit for publication.

3. Results

A summary of the complete review process is illustrated in Fig. 1. From the initial 4200 publications identified from the systematic searches of the bibliographic databases, 3553 publications were excluded at the abstract-screening phase with a further 458 publications excluded during full-text screening. From the review of conference proceedings and additional supplementary hand searching, 28 publications were identified and included. An additional 21 publications were retrieved from the 2016 WHO review. Thus, the complete evidence base consisted of 238 publications describing 142 unique studies. Here we summarize the evidence from 63 studies (109 publications [21], [22], [23], [24], [25], [26], [27], [28], [29], [30], [31], [32], [33], [34], [35], [36], [37], [38], [39], [40], [41], [42], [43], [44], [45], [46], [47], [48], [49], [50], [51], [52], [53], [54], [55], [56], [57], [58], [59], [60], [61], [62], [63], [64], [65], [66], [67], [68], [69], [70], [71], [72], [73], [74], [75], [76], [77], [78], [79], [80], [81], [82], [83], [84], [85], [86], [87], [88], [89], [90], [91], [92], [93], [94], [95], [96], [97], [98], [99], [100], [101], [102], [103], [104], [105], [106], [107], [108], [109], [110], [111], [112], [113], [114], [115], [116], [117], [118], [119], [120], [121], [122], [123], [124], [125], [126], [127], [128], [129]) reporting efficacy or effectiveness outcomes for the pangenotypic regimens as well as sofosbuvir-ledipasvir.

Fig. 1.

Study selection.

Outcomes of interest in persons treated with sofosbuvir-velpatasvir, sofosbuvir-daclatasvir, glecaprevir-pibrentasvir, and sofosbuvir-ledipasvir were reported in 10, 20, 10, and 40 studies, respectively. This evidence consisted of 51 studies classified as either a randomized or other trial design (e.g. single arm trial) and 23 observational cohort or database studies. Available evidence varied by subgroup, with 61 studies reporting outcomes for persons in the all-comer analysis, 22 studies with evidence for persons with cirrhosis, and six studies for persons with HIV-HCV co-infection. Most studies consisted of mixed samples of treatment-naïve and treatment-experienced persons (n = 52, 70.3%), with 13 (17.6%) studies enrolling only persons who had no prior treatment experience, six (8.1%) studies enrolling only persons with treatment experience, and three (4.1%) studies where treatment history was unclear or not reported. Most studies were multinational with several countries represented, including the USA (n = 31, 41.9%), France (n = 13, 17.6%), Canada (n = 10, 13.5%), and Japan (n = 9, 12.2%). The majority of studies (63.5%) reported receiving at least some industry funding. Study characteristics and outcomes were only available from a conference abstract in 15 cases.

3.1. SVR12 in the all-comer population

In the all-comer analysis, which included persons irrespective of cirrhosis status or comorbidities, high SVR12 rates were observed across both treatments and genotypes, commonly with narrow confidence intervals (Fig. 2). Point estimates of pooled SVR12 proportions exceeded 0.94 for all treatments for persons with genotypes 1, 2, or 4 infection, with the exception of persons with genotype 2 infection treated with sofosbuvir-ledipasvir (0.86; 95% CI: 0.65, 1.00; N = 53). This estimate was from a single non-randomized comparative trial of sofosbuvir-ledipasvir administered for either 8 (n = 27) or 12 (n = 26) weeks, where 74% of persons treated for 8 weeks achieved SVR12 [54]. Outcomes in persons with genotype 3 HCV infection were typically characterized by wider confidence intervals, exceeding 0.05 for the pooled SVR12 proportions for sofosbuvir-velpatasvir (0.89; 95% CI: 0.85, 0.93; N = 776), sofosbuvir-daclatasvir (0.89; 95% CI: 0.85, 0.94; N = 895) and sofosbuvir-ledipasvir (0.65; 95% CI: 0.51, 0.79; N = 42). The number of persons included in the analyses by treatment varied from a high of 13 327 persons with genotype 1 infection across 47 study arms for sofosbuvir-ledipasvir to a single arm of 16 persons with genotype 4 infection treated with glecaprevir-pibrentasvir.

Fig. 2.

Illustration of findings from the all-comer analyses, stratified by genotype, all treatment experience.

In the analyses stratified by treatment experience (Table B1), the proportions of persons with genotypes 1 or 3 infection achieving SVR12 were consistent with the primary analyses. Limited or no evidence was available for persons with genotypes 2 or 4 infection, with the exception of sofosbuvir-ledipasvir in persons with genotype 4 infection which was consistent with the primary analysis.

3.2. SVR12 in persons with cirrhosis

Most evidence for persons with cirrhosis came from studies of genotype 1 infection (Fig. 3). In this subpopulation, pooled SVR12 estimates exceeded 90% for all treatments, though the width of confidence intervals varied. Small pooled sample sizes were available for persons with genotype 2 or 4 infection, though the proportion of persons achieving SVR12 remained high.

Fig. 3.

Illustration of findings from the analyses of persons with cirrhosis, stratified by genotype, all treatment experience.

The proportion of persons achieving SVR12 was consistent across treatment-experience subgroups (Table B2) for sofosbuvir-ledipasvir in genotype 1 infection (treatment-experienced: 0.97 [95% CI: 0.95, 1.00], N = 175; treatment-naïve: 0.97 [95% CI: 0.93, 1.00], N = 78). For persons with genotype 3 infection, the proportion treated with sofosbuvir-velpatasvir achieving SVR12 was high in both the treatment-experienced and treatment-naïve stratifications (treatment-experienced: 0.90 [95% CI: 0.83, 0.97], N = 69; treatment-naïve: 0.97 [95% CI: 0.92, 1.00], N = 120). However, evidence by treatment experience was otherwise limited or unavailable.

3.3. SVR12 in persons with HIV–HCV coinfection

Limited evidence was identified for persons with HIV–HCV coinfection (Fig. 4). Outcomes for persons with HCV genotype 1 infection treated with sofosbuvir-ledipasvir were the best represented in this subgroup, with a high pooled proportion achieving SVR12 (0.96; 95% CI: 0.93, 0.98; n = 383). Outcomes for sofosbuvir-velpatasvir were available from the ASTRAL-5 trial, where high SVR12 rates were observed across genotypes, though with small sample sizes. Given the limited evidence identified for this subgroup, analyses were not stratified by treatment-experience.

Fig. 4.

Illustration of findings from the analyses of persons with HIV–HCV coinfection, stratified by genotype, all treatment-experience.

3.4. Harms

Analyses on harms data were conducted irrespective of genotype, treatment experience, and comorbidities (Fig. 5). Across the regimens studied, and based on large pooled sample sizes, the proportions of persons with DAEs, SAEs, or all-cause mortality were low. Pooled proportions of persons reporting a DAE or all-cause mortality did not exceed 0.01 and were characterized by tight confidence intervals. However, certainty around the evidence for SAE outcomes varied by regimen, with the widest confidence interval observed for sofosbuvir-daclatasvir (0.03; 95% CI: 0.01, 0.05; N = 1875).

Fig. 5.

Illustration of findings from the analyses of harms, all treatment-experience.

4. Discussion

Guidelines issued by the WHO are developed through an iterative process to address an area of uncertainty and unmet guidance need. These processes are explicit and transparent, involving consultation with several multidisciplinary stakeholders to ensure benefits and barriers are evaluated systematically and comprehensively. Importantly, the evidence used to develop these guidelines is publicly available. This review is one component of the evidence that was considered in the development of the 2018 guideline on the care and treatment of persons diagnosed with chronic HCV infection. Here we summarize SVR12 and harms outcomes for three pangenotypic DAA regimens (sofosbuvir-velpatasvir, sofosbuvir-daclatasvir, and glecaprevir-pibrentasvir) as well as sofosbuvir-ledipasvir, which were of particular relevance to decision-makers. Evidence for sofosbuvir-ledipasvir was considered as countries where HCV is largely isolated to a single genotype have had success managing care with this non-pangenotypic regimen. Across treatments, outcomes for persons with genotype 1 infection were the best represented, reflecting the high prevalence of this strain in the United States and Europe [130]. In the overall population analyses, all treatments were associated with high rates of SVR12 across genotypes 1 through 4. These findings were generally consistent across stratifications by treatment experience. While the rates of SVR12 were again high across treatments for persons with cirrhosis and genotype 1 infection, relatively limited evidence was available for genotype 2, 3, or 4 infection. The sparse evidence available for persons with HIV–HCV coinfection, both with respect to genotypes and treatments more generally, is likely a reflection that outcomes in these persons are similar to mono-infected persons. This observation has been acknowledged by international guidelines [131]. However, consideration of this sub-population is still warranted given evidence suggesting that persons with HIV–HCV coinfection are at risk of HCV treatment failure for factors such as ongoing illicit drug use and mental illness [132]. Analyses of harms suggest consistency across treatments for the very low proportions of persons with DAEs, SAEs, or all-cause mortality. This is a marked difference from previous generations of antiviral therapy for chronic HCV infection. For example, a 2013 review reported that the percentage of persons with SAEs varied from 4.7% for persons managed with pegylated interferon α−2b with ribavirin to 16% for 24 weeks of telaprevir-based triple therapy or 48 weeks of telaprevir dual therapy. Similarly, the percentage of persons with DAEs varied from 6.6% for dual therapy pegylated interferon α−2a with ribavirin to 15% for 24 weeks of telaprevir-based triple therapy [133].

The evidence described in this review was considered and applied to update current treatment guidelines for adults with chronic HCV infection. Based on this and other supporting evidence, the WHO issued a conditional recommendation, based on moderate quality of evidence, that pangenotypic DAA regimens be used for the treatment of persons with chronic HCV infection aged 18 years and above. In this recommendation, pangenotypic was defined as leading to SVR in over 85% of persons treated across all six major HCV genotypes. At the time this guidance was issued, the pangenotypic regimens available to adults without cirrhosis included sofosbuvir-velpatasvir (12 week course), sofosbuvir-daclatasvir (12 week course), and glecaprevir-pibrentasvir (8 week course). For adults with compensated cirrhosis, available regimens included sofosbuvir-velpatasvir (12 week course), glecaprevir-pibrentasvir (12 week course), sofosbuvir-daclatasvir (24 week course or 12 week course in regions where the genotype 3 prevalence is known to be less than 5%). A treatment duration of 24 weeks was recommended for sofosbuvir-daclatasvir given the lower SVR rates in persons with genotype 3 infection. For adults with or without compensated cirrhosis, glecaprevir-pibrentasvir should be used for 16 weeks for persons with genotype 3 infection who have previously received interferon and/or ribavirin.

The approach taken for this review was in accordance with WHO guideline methods and published standards. Standard literature search techniques were supplemented by contact with primary researchers, including drug manufacturers, in order to identify relevant unpublished data. Analyses were based on both randomized trials, non-randomized trials, and observational studies, with analyses stratified according to study type. In this paper we presented only analyses based on all eligible study designs. Moreover, several stratifications were planned a priori, including prior treatment experience, presence of cirrhosis, and HIV–HCV coinfection. The quality of evidence was evaluated using GRADE criteria and this approach was modified in consultation with a GRADE methodologist to suit the clinical research context of HCV.

Despite these strengths, there are limitations to this review. The scope did not encompass the complete landscape of HCV regimens and rather focused on newer, more effective pangenotypic DAAs. For example, outcomes for persons treated with ribavirin-containing regimens were not addressed here. However, some ribavirin-containing regimens were described in the full report, such as for the treatment of persons with cirrhosis and genotype 2 or 3 infection. This decision was based on the treatment burden associated with ribavirin, such as treatment-related side effects and a need for frequent laboratory monitoring; in addition, the non-ribavirin regimens summarized in this article demonstrate high efficacy. Importantly, analyses were conducted based on pooled proportions of persons with the pre-specified outcomes, irrespective of whether persons were enrolled in a multi-arm or single-arm trial or described in an observational cohort. Therefore, we cannot infer relative effects between interventions as we cannot disentangle study- and treatment-effects [134]. This limitation reflects the evidence landscape of HCV infection, where randomized trials comparing multiple active interventions are generally not feasible given the high efficacy of available interventions. However, the SVR outcome is an objective means by which to define efficacy and spontaneous SVR without antiviral therapy is very rare. For example, no patients (0/116) treated with placebo in the randomized ASTRAL-1 trial achieved SVR [118]. Important evidence may also come from study designs other than those considered here, such as retrospective studies, though these were out of scope in the current review. Finally, some unexplained statistical heterogeneity was present, which may be related to age, country, comorbidities, the use of generic drugs, treatment doses, the use of DAAs or interferon-based regimens in persons with prior treatment-experience, or treatment durations. Despite this, SVR rates were generally consistent across the included studies within each subgroup stratification. Future analyses may explore the sources of heterogeneity using patient-level, rather than study-level, evidence. Despite the strong efficacy outcomes demonstrated in the literature to date, the rapid evolution of treatments for HCV infection warrants an increased focus and scrutiny on emerging pangenotypic regimens, including the study of long-term treatment efficacy, harms, and uptake by front-line clinicians.

The benefits of curing HCV infections are far-reaching [135]. In response to emerging treatments, the WHO commissioned a systematic literature review to inform treatment guidelines for the management of persons with HCV infection. Based on the evidence, the WHO recommended the use of pangenotypic regimens to support the global campaign to eradicate HCV infection. With an oral administration, a feasibility advantage of bypassing genotype testing, and favourable tolerability profiles, the widespread adoption of pangenotypic regimens translates to a powerful and effective response to the public health threat posed by HCV infection.

Declaration of competing interest

Mr. Zoratti reports that he is a shareholder of Zoratti HEOR Consulting Inc. which was contracted to conduct this study. Ms. Siddiqua reports personal fees from Zoratti HEOR Consulting Inc. during the conduct of the study. Ms. Morassut reports personal fees from Zoratti HEOR Consulting Inc. during the conduct of the study. Ms. Zeraatkar reports personal fees from Zoratti HEOR Consulting Inc. during the conduct of the study. Dr. Chou reports serving as methodologist for the World Health Organization hepatitis C guideline, during the conduct of the study and grants from Agency for Healthcare Research and Quality outside the submitted work on hepatitis C screening and treatment. Mr. Druyts reports that he is a shareholder of Pharmalytics Consulting Group Inc. (“Pharmalytics Group”), registered in the province of British Columbia, Canada, which provides consulting services to the healthcare and pharmaceutical industries. No other author has anything to disclose.