Patterns of Resistance-associated Substitutions in Patients With Chronic HCV Infection Following Treatment with Direct-acting Antivirals

Abstract

BACKGROUND & AIMS Little is known about substitutions that mediate resistance of HCV to direct-acting antivirals (DAAs), due to the small number of patients with treatment failure in approval studies. It is important to identify resistance patterns to select effective salvage treatments.
METHODS We performed a comprehensive analysis for resistance-associated substitutions (RASs) in HCV genes (NS3, NS5A, NS5B) targeted by DAAs. We compared NS3, NS5A, and NS5B sequences from 626 patients in Europe with DAA failure with sequences from 2322 DAA-naïve patients, infected with HCV genotypes 1-4. We considered RASs to be relevant if they were associated with DAA failure in patients or conferred a greater than 2-fold changed in susceptibility compared with a reference strain in in vitro replicon assays. Data were collected on pretreatment status, DAA regimen, the treatment initiation date and duration, and virologic response. Patients who received at least 4 weeks antiviral treatment were included in the analysis.
RESULTS RASs in NS3 associated with simeprevir or paritaprevir failure include R155K and D168E/V. In addition, several RASs were specifically associated with failure of simeprevir (Q80K/R in patients with genotype 1a or 4) or paritaprevir (Y56H in combination with D168V in patients with genotype 1b). Y93H in NS5A was the RAS most frequently associated with failure of daclatasvir, ledipasvir, or ombitasvir in patients with genotype 1b infection, and L31M was associated with failure of daclatasvir or ledipasvir, but not ombitasvir. RASs in NS5A were heterogeneous among patients with HCV genotype 1a or genotype 4 infections. In patients with HCV genotype 3, Y93H was associated with resistance to daclatasvir, but no RASs were associated with ledipasvir failure, pointing to a limited efficacy of ledipasvir in patients with genotype 3. Among patients failed by sofosbuvir-containing regimens, L159F was enriched in patients with genotype 1b (together with C316N) or genotype 3 infection, whereas the RAS S282T was rarely observed.
CONCLUSIONS We compared RASs in NS3, NS5A, and NS5B among patients failed by DAA therapy. Theses varied with the HCV genotype and subtype, and the different drug classes. These findings might be used to select salvage therapies.