Elbasvir and grazoprevir for hepatitis C virus genotype 1 infection in people with recent injecting drug use (DARLO‐C): An open‐label, single‐arm, phase 4, multicentre trial

Abstract Background and Aims Direct‐acting antiviral therapy for hepatitis C virus (HCV) is effective, but few prospective studies among people with ongoing injecting drug use exist. This study evaluated the efficacy of elbasvir/grazoprevir in people with HCV genotype 1/4 (G1/4) infection and recent injecting drug use. An exploratory aim evaluated the feasibility of fingerstick point‐of‐care HCV RNA testing prior to and following treatment. Methods DARLO‐C (http://clinicaltrials.gov: NCT02940691) is an open‐label phase 4 trial. Participants were recruited between May 2017 and March 2018 from two drug treatment clinics, two hospital clinics, and one community clinic in Australia. Inclusion criteria included recent injection drug use (previous 6 months) and HCV G1/4 infection. Exclusion criteria included prior HCV treatment and decompensated liver disease. Participants received elbasvir/grazoprevir once‐daily for 12 weeks. The primary endpoint was undetectable HCV RNA 12 weeks post‐treatment (SVR). Fingerstick whole‐blood samples were tested using the Xpert HCV Viral Load Fingerstick (Xpert HCV VL Fingerstick) assay and compared to the Aptima HCV Quant Dx Assay on plasma samples. Results Of a planned 150 participants, 32 were enrolled due to slower than anticipated recruitment [median age 46 years, 10 (31%) female, 29 (91%) G1a]. Eighteen (56%) were receiving opioid agonist therapy and 29 (91%) injected in the previous month. Twenty‐six (81%) of 32 completed treatment (lost to follow‐up, n = 5; incarceration, n = 1). There were no virological failures. Twenty‐four (75%, 95% CI 59%‐91%) of 32 achieved SVR. Two participants who completed treatment did not have SVR (loss to follow‐up, n = 1; refused test, n = 1). Among paired samples (n = 36), sensitivity of the Xpert HCV VL Fingerstick assay for HCV RNA detection was 100.0% (95% CI 75.3%‐100.0%) and specificity was 95.7% (95% CI 78.1%‐99.9%). Conclusion Elbasvir/grazoprevir is effective among people with HCV G1 with recent injecting drug use. Implementation of point‐of‐care HCV RNA testing was feasible, but the high error rate requires investigation.

Conclusion: Elbasvir/grazoprevir is effective among people with HCV G1 with recent injecting drug use. Implementation of point-of-care HCV RNA testing was feasible, but the high error rate requires investigation.  4 As such, it will be difficult to eliminate HCV infection without strategies focused on addressing HCV prevention, testing, linkage to care, and treatment among PWID.
Direct-acting antiviral therapy is effective in people receiving opioid agonist therapy (OAT) and in people with recent injecting drug use, even in "real-world" settings. 5 In the COSTAR study, among people receiving grazoprevir and elbasvir with HCV genotypes 1, 4, or 6 on stable OAT who had never previously received HCV treatment, 6 the intention-to-treat (ITT) SVR was 91%. 6 However, only 25% had injected drugs within the previous 6 months. 7 Although many studies have been published on HCV treatment outcomes among people with recent drug use, studies have been limited by being performed at single centres and the heterogeneous study populations considered, particularly with respect to the inclusion of people with recent injecting drug use. Further data on HCV treatment outcomes among people with recent injecting drug use are needed.
Among PWID, poor venous access has been noted as a major barrier to HCV testing and treatment. [8][9][10][11] PWID with poor venous access may perceive blood collection as distressing due to perceived stigmatisation by healthcare workers (and phlebotomists), inexperience of phlebotomists in collecting blood samples from PWID, and poor access to experienced phlebotomists at services preferred by PWIDs. 12 Fingerstick whole-blood collection is acceptable to PWID and is preferred to venepuncture. 10,13,14 The availability of a fingerstick point-of-care HCV RNA test may provide one strategy to address barriers to venous access among PWID. 15 Participants having previously received any HCV treatment and those with decompensated liver disease were excluded. In the original study protocol, participants required resistance testing to be performed at screening, and participants with any one of the following HCV resistance-associated NS5A substitutions were excluded: M28L/T/V, Q30H/L/R, L31M, or Y93C/H/N/S. However, given that resistance testing was not broadly available in Australia and not required by national treatment guidelines, 18 the study protocol was revised in December 2017 to make resistance testing optional. Full eligibility criteria are provided in the study protocol (Supporting Information).

| Procedures
Participants with HCV genotype 1 received a fixed-dose combination tablet that contained 50 mg of elbasvir and 100 mg of grazoprevir orally once-daily for 12 weeks (funded through the Australian govern-  [19][20][21] Assessments during treatment included physical examinations, measurements of HCV RNA levels (performed at local laboratories), and standard laboratory testing (liver function tests, full blood count, and biochemistry). All adverse events were recorded and graded according to the Medical Dictionary for Regulatory Activities,

MedDRA.
As part of an exploratory study to assess the feasibility of

| Outcomes
The primary efficacy endpoint was the proportion of participants with a SVR12, defined as an HCV RNA level below the limit of quantification 12 weeks after the end of treatment in all participants who received at least one dose of study medication (ITT population). When HCV RNA had not been assessed at SVR12, the result of the next available HCV RNA assessment was used to calculate SVR. Participants with no result at or following the SVR12 visit were considered to not have had an SVR. In addition, a post hoc modified ITT analysis was performed excluding participants with a missing SVR12 test. Secondary endpoints included treatment completion, treatment adherence, end of treatment response (ETR, negative HCV RNA at the end of treatment), severe adverse events, and treatment discontinuations because of adverse events.

| Statistical analysis
The primary aim of this study was to evaluate the efficacy of elbasvir and grazoprevir for 12 weeks in participants infected with HCV genotype 1 with recent injecting drug use (previous 6 months).
A total of 150 participants were planned for enrolment and evaluation as the ITT population. Assuming an overall SVR of 90% (135 of 150), the 95% confidence interval (95% CI) around this estimate would be 84% to 94%. However, due to slower than anticipated enrolment as a result of the availability of pan-genotypic regimens in Australia leading to reduced prescribing of the elbasvir and grazoprevir regimen, recruitment was closed prematurely (n = 32).
Assuming an overall SVR of 90%, the 95% CI around this estimate (calculated using Clopper-Pearson binomial confidence intervals) would be expected to be 75% to 98%.
We used the Clopper-Pearson method to calculate point estimates and two-sided 95% exact confidence intervals for the proportion with SVR overall, as well as according to HCV genotype, and various subgroups. Factors hypothesized to be associated with SVR were age (stratified by median), gender, current OAT at baseline, recent (previous month) injecting drug use at baseline, and frequency of injecting at baseline (none, less than daily, daily, or greater). 5,26 The sensitivity and specificity of the Xpert HCV Viral Load Fingerstick assay for detection of HCV RNA in plasma samples collected via venepuncture and capillary whole-blood samples collected by fingerstick was assessed using both detectable and quantifiable thresholds (limit of quantification >100 IU/mL Xpert HCV VL Fingerstick) compared to Aptima HCV Quant Dx assay in plasma as the reference standard (limit of quantification >10 IU/mL). Any discordant results were included in all calculations of sensitivity and specificity. All data are reported in log 10 units.
For all analyses, statistically significant differences were assessed at a 0.05 level; P-values were two-sided. All analyses were performed using Stata v12.0 (StataCorp, College Station, Texas).

| Role of the funding source
The study was funded by a research grant from Merck/MSD. The funder had no role in the study design, data collection, analysis, interpretation of the results, the writing of the report, or the decision to submit the report for publication. J.G., E.C., and G.D. had access to the raw data. The sponsor (The Kirby Institute, UNSW Sydney) designed the study, collected the data, managed study samples, monitored study conduct, and performed the statistical analysis. J.G. and G.D. were responsible for the decision to submit for publication.

| Participant characteristics
Of 36 participants screened, 32 were enrolled and received at least one dose of study medication (ITT population, Figure 1, Table 1). Most participants (91%, 29 of 32) had genotype 1a. The median age was 46 years, 31% (10 of 32) were female, and 6% (2 of 32) had cirrhosis.  In ITT analysis, 24 (75%, 95% CI: 59%, 91%) of 32 had an ETR and 24 (75%, 95% CI: 59%, 91%) of 32 had an SVR. Among participants who completed treatment but did not have an SVR (n = 2), reasons for not achieving an SVR included an inability to obtain a blood sample post-treatment and refusing to have an SVR test. There were no virologic relapses. In a modified ITT analysis (excluding people without an SVR test), SVR was 100% (24 of 24 with available testing).
No reinfections were observed (eight person-years of follow-up). The proportion with SVR stratified by key characteristics is shown in Table S1.

| Safety
A total of five (16%) of 32 participants experienced a serious adverse event that required hospitalisation. These events were considered not to be related to the study drugs and included drug-induced psychosis, laceration of arm, schizoaffective disorder, suicidal ideation, and psychosis. There were no deaths. The overall proportion with SVR of 75% is lower than the weighted mean SVR of 87% in a systematic review of direct-acting antiviral (DAA) therapy among people who recently injected drugs. 5 In this systematic review, a meta-regression analysis was also performed 5 and "real-world" observational studies were associated with a lower SVR compared to clinical trials. The lower SVR was explained by a higher proportion of participants lost to follow-up in "real-world" observational studies compared to clinical trials. 5 This is consistent with the results in the current study demonstrating that only 81% of participants completed treatment and lost to follow-up was the primary reason for not achieving an SVR. In fact, there were no cases of virological failure in this study. The lower treatment completion and SVR observed in this study is consistent with a "real-world" study of DAA treatment among people who have recently injected drugs in Sydney, Australia. 27 In this study by Read et al, one-third of people elected to receive daily or weekly enhanced adherence support, resulting in equivalent follow-up for SVR testing and SVR to those who did not receive enhanced adherence support. 27 It is interesting that those who received enhanced treatment support in this study were more often homeless, identified as Aboriginal, had a mental health diagnosis, and ≥daily injecting drug use. 27 Among those who discontinued therapy early, two-thirds (4 of 6) discontinued after week 8 of therapy. Shorter durations of HCV therapy (eg, <8 weeks) could be explored for populations who might be at higher risk of early discontinuation, such as people with ongoing injecting drug use. Collectively, these data suggest that strategies should be explored to improve retention and facilitate HCV treatment completion, particularly among marginalized PWIDs who may require enhanced support during therapy.
Injecting and non-injecting drug use was stable prior to and during HCV therapy, consistent with results from studies of interferonbased [28][29][30] and DAA-based therapies. 6