PloS one 2018 03 2213(3) e0194396 doi 10.1371/journal.pone.0194396
Direct-acting antiviral (DAA) drug performances depend on the viral genotype. So international recommendations give typing of the virus a prerequisite for treatment choice and patient management. Commercially available HCV genotyping kits are scarce and this analysis is often in-house using tedious PCRs and Sanger sequencing, leading to a lack of standardization. A newly commercialized HCV genotyping assay based on real-time PCR has been developed by Roche Diagnostics (Mannheim, Germany). We compared this new assay with our in-house PCRs -sequencing technique on 101 regular samples and 81 LiPA failures or low viral load samples. No genotype or 1a/1b subtype mismatch was observed. Two samples were misidentified at the subtype level without clinical impact. Three genotype 1b and two genotype 1a samples with low viral load could not be subtyped. Nevertheless, 13 (13%) samples from the regular panel and 35 (43%) from the more difficult-to-type panels failed to give results on first pass with the Roche kit. Failures were mostly associated with genotype 3 subtype a, with genotype 4 subtype non-a, or with viral loads <200 IU/mL (p = 0.0061). The workflow allowed a non-specialized technician to obtain results in less than 4 hours whereas 2 to 3 days and experienced staff were required with the in-house assay. In conclusion, the Roche cobas® HCV GT kit is easy and rapid to use and provides reliable results. The high rate of uninterpretable results particularly for low viral load samples and less frequent genotypes, and the absence of subtyping for non-genotype 1 could require sending complex samples to a specialized laboratory.