Resistance-associated substitutions among HCV1b virus populations in patients who failed DAA based regimens

Background: Resistance-associated substitutions (RAS) are usually determined after treatment failure of direct-acting antiviral (DAA) drugs, before a new DAA regimen is started. We investigated the presence of RAS in HCV quasispecies after therapy failure and compared them with baseline sequences to assess the accumulation of RAS and their role in treatment failure. Materials and methods: Frequent longitudinal serum sampling from four HCV1b chronic infected patients was performed. Deep sequencing of NS3 and NS5B regions, by Ion Personal Genome Machine Sequencer, was applied. Sample-specific consensus sequences were generated from error corrected and adapter cleaned reads using reference guided de novo assembly with Vicuna and the Viral Finishing and Annotation Toolkit. HCV subtyping was performed by manual phylogenetic analysis using PhyML, and by typing tools COMET and Oxford. RAS were analyzed aligning HCV1b reference sequences and subsequently by Geno2pheno tool. Results: All patients were DAA-naïve, not HIV/HBV co-infected. First line therapy for patients 1 and 2 consisted of telaprevir containing regimen, while patient 3 was first treated with simeprevir + daclatasvir and patient 4 with boceprevir in combination with peg-interferon/ribavirin. After failure, all of them were treated with simeprevir + sofosbuvir. Patient 3 had further failure to the last therapy. At breakthrough in majority rule consensus sequence of patient 1, we found NS354A RAS, which reduces susceptibility to telaprevir and boceprevir. At baseline this patient’s virus population had NS3 substitution 122G as majority variant, which is not a RAS but located at a hot spot position related to simeprevir resistance (NS3122R). Concerning patient 3, the breakthrough viral population harboured NS3168V RAS (simeprevir, asunaprevir, paritaprevir and grazoprevir resistances) as majority variant. This RAS was not detected at baseline and could no longer be detected as majority variant nine months after breakthrough. Additionally, in patient 3, NS5B159F sofosbuvir RAS and the NS5B316N fitness-associated substitution were detected in the majority rule consensus sequence at all time points, these pre-existing RAS may have contributed to treatment failure. For two of the patients, only a baseline sample was available. In patient 2, the majority rule consensus sequence of NS3 contained the NS3170I substitution on a hot spot position related to boceprevir, simeprevir and telaprevir resistances (NS3170A/T). In patient 4, the majority population harboured NS3174T substitution on a hot spot position related to boceprevir and telaprevir resistances (NS3174F/S). Conclusions: Our data illustrate the importance of resistance testing at baseline and before retreatment after DAA therapy failure. As recently reported in published data, the V amino acidic substitution on 168 position is generally less fit in the absence of selective Protease Inhibitors pressure. Indeed, reversal of treatment acquired NS3168V RAS after treatment failure in one patient suggests a fitness cost for this mutation. This reversion to DAA-sensitive virus in just a few months raises concern about time of testing and retreatment.status: publishe