Abstract Body

Recent development of direct-acting antiviral agents (DAA) has dramatically improved the effectiveness and tolerability of treatments for hepatitis C virus (HCV), resulting in >95% sustained virologic response (SVR) rates. However, cases of treatment failure have been associated with the emergence of resistance-associated substitutions (RAS). To better guide clinical decision-making, we developed and validated a near-whole-genome, HCV genotype (GT)-independent sequencing strategy on the Illumina MiSeq next-generation sequencing (NGS) platform.

HCV GT1-6 samples from treatment-naïve HCV-infected individuals as well as DAA-treated persons who did not achieve SVR were included. Viral RNA was extracted on a Biomerieux easyMag and underwent nested reverse-transcription-PCR. Libraries prepared by Nextera XT were sequenced on the MiSeq. NGS data were processed by an in-house pipeline that incorporates HCV reference sequence selection and an iterative mapping process for paired-end reads. Nucleotide consensus sequences were aligned to appropriate FDA reference strain sequences for downstream identification of RAS. Sequences were compared to data obtained from a previously validated in-house assay optimized for HCV GT1. A minimal threshold for minority species detection was estimated from the coefficient of variation of minor species quantification.

Roughly 90% sequencing success rates, defined as achieving >100-fold NGS read coverage across NS3, NS5a and NS5b, was observed for most genotypes in samples with HCV RNA >5 log10 IU/mL (Table 1). The genotype-independent HCV method showed >99.8% nucleotide concordance with the GT1-optimized method in NS3, NS5a, and NS5b using a 20% mixture-calling threshold. The assay demonstrated near-perfect precision and reproducibility at detecting variants above 2% prevalence and showed no systematic bias in amplifying specific RAS. An absolute lower limit of 0.2% for reproducible minor species detection was estimated but warrants a more conservative threshold given MiSeq error rates (0.5%) and limitations of PCR.

This study highlights the performance of a freely available, near whole-genome NGS assay and bioinformatic pipeline for genotype-independent HCV genotyping and RAS detection. The method demonstrated similar performance to a validated GT1 assay and can now be extended to other HCV genotypes. This method has been implemented clinically and has been used to deliver ~3000 resistance test reports to physicians across Canada.