Evaluating the generalizability of clinical trials of direct-acting antivirals for HIV–hepatitis C virus coinfection

HIV coinfection accelerates hepatitis C virus (HCV) progression and interferon-based therapies have shown relatively poor efficacy and tolerability in the HIV–HCV-coinfected population. The introduction of direct-acting antivirals (DAAs) for the treatment of HCV represented a paradigm shift leading to cure of chronic HCV infection, and was held to be a promising alternative for patients with HIV–HCV coinfection. Clinical trials have demonstrated the efficacy and tolerability of various DAAs in patients with HIV–HCV coinfection, bringing the added benefits of simplified dosing and shorter treatment times. However, there continues to be significant concern about the generalizability of DAA clinical trial results to the highly heterogeneous population of patients with HIV–HCV coinfection. To illustrate this point, Saeed et al.[1] compared the eligibility criteria from clinical trials of various DAAs to the characteristics of a representative HIV–HCV-coinfected population, the Canadian Coinfection Cohort. Cohort participants with active HCV infection (N = 874) were subdivided by HCV genotype to match those of the five trial populations. A large proportion of the Cohort participants had a history of drug use and poverty. The majority were receiving antiretroviral therapy (ART; 86%) and had undetectable HIV loads (78%). The median duration of HCV infection was approximately 22 years; 13% had cirrhosis and 15% had advanced liver disease. With the goal of examining how many of the Canadian Coinfection Cohort participants would have been included in the DAA clinical trials, the authors compared the characteristics of the Cohort against the eligibility criteria of five trials: NCT01479868 (simeprivir), PHOTON-1 (sofosbuvir), TURQUOISE-I (ombitasvir, paritaprevir/ritonavir/dasabuvir), ION-4 (sofosbuvir/ledipasvir), and ALLY-2 (daclatasvir/sofosbuvir). For four of the five trials, only 6–10% of the Canadian Coinfection Cohort would have been eligible for inclusion, primarily because of restrictions on the use of specific ART agents (would have excluded 63–79% of the cohort) and active drug use (would have excluded 53–55% of the cohort). The authors commented that even if individuals had switched ART to meet trial eligibility, 74–77% of the cohort would continue to have been excluded because of active drug use. Approximately, 15% of the cohort would have been ineligible for the DAA trials because of HIV loads or CD4+ cell counts that did not meet prespecified thresholds. Few of the cohort participants would have been excluded based on safety-related criteria, such as anemia or renal or liver function. The authors concluded that the eligibility criteria used in DAA clinical trials exclude important subpopulations of HIV–HCV-coinfected patients – specifically, active drug users – and call into question whether the results of the trials are applicable to the broader population of coinfected patients. The concerns of Saeed et al. regarding the applicability of the DAA clinical trial results to the larger HIV–HCV-coinfected population are widely recognized; in the past year alone, several prospective cohort studies in the United States and Europe have reported on the real-world safety and efficacy of DAAs in coinfected individuals, many with advanced liver disease [2–5]. More observational studies are needed to accurately evaluate the efficacy and safety of DAAs in representative HIV–HCV-coinfected populations. Acknowledgements Conflicts of interest There are no conflicts of interest.