Josep Quer, Josep Gregori, Francisco Rodríguez-Frias, Maria Buti, Antonio Madejon, Sofia Perez-del-Pulgar, Damir Garcia-Cehic, Rosario Casillas, Maria Blasi, Maria Homs, David Tabernero, Miguel Alvarez-Tejado, Jose Manuel Muñoz, Maria Cubero, Andrea Caballero, Jose Antonio delCampo, Esteban Domingo, Irene Belmonte, Leonardo Nieto, Sabela Lens, Paloma Muñoz-de-Rueda, Paloma Sanz-Cameno, Silvia Sauleda, Marta Bes, Jordi Gomez, Carlos Briones, Celia Perales, Julie Sheldon, Lluis Castells, Lluis Viladomiu, Javier Salmeron, Angela Ruiz-Extremera, Rosa Quiles-Pérez, Ricardo Moreno-Otero, Rosario López-Rodríguez, Helena Allende, Manuel Romero-Gómez, Jaume Guardia, Rafael Esteban, Javier Garcia-Samaniego, Xavier Forns, Juan Ignacio Esteban. 2016. High-Resolution Hepatitis C Virus Subtyping Using NS5B Deep Sequencing and Phylogeny, an Alternative to Current Methods. Journal of Clinical Microbiology 54, 7, DOI: 10.1128/JCM.01007-16
Hepatitis C virus (HCV) is classified into seven major genotypes and 67 subtypes. Recent studies have shown that in HCV genotype 1-infected patients, response rates to regimens containing direct-acting antivirals (DAAs) are subtype dependent. Currently available genotyping methods have limited subtyping accuracy. We have evaluated the performance of a deep-sequencing-based HCV subtyping assay, developed for the 454/GS-Junior platform, in comparison with those of two commercial assays (Versant HCV genotype 2.0 and Abbott Real-time HCV Genotype II) and using direct NS5B sequencing as a gold standard (direct sequencing), in 114 clinical specimens previously tested by first-generation hybridization assay (82 genotype 1 and 32 with uninterpretable results). Phylogenetic analysis of deep-sequencing reads matched subtype 1 calling by population Sanger sequencing (69% 1b, 31% 1a) in 81 specimens and identified a mixed-subtype infection (1b/3a/1a) in one sample. Similarly, among the 32 previously indeterminate specimens, identical genotype and subtype results were obtained by direct and deep sequencing in all but four samples with dual infection. In contrast, both Versant HCV Genotype 2.0 and Abbott Real-time HCV Genotype II failed subtype 1 calling in 13 (16%) samples each and were unable to identify the HCV genotype and/or subtype in more than half of the non-genotype 1 samples. We concluded that deep sequencing is more efficient for HCV subtyping than currently available methods and allows qualitative identification of mixed infections and may be more helpful with respect to informing treatment strategies with new DAA-containing regimens across all HCV subtypes.