Emergence of Resistant Variants in HCV Genotype 1
Drug resistance has been shown to emerge with different classes of DAA regimens. However, the reason why treatment fails in some patients remains unclear. Amino acid substitutions in HCV related proteins that confer resistance to DAAs can exist at low frequency prior to antiviral treatment with DAAs. Enrichment of variants during therapy has been reported, although monitoring changes in variant frequency using ultra-deep sequencing is not commonly performed. HCV is an error-prone RNA virus in which substitutions frequently occur throughout the HCV genome, and drug-resistant variants are sometimes present as minor populations in patients who have never been exposed to DAAs.
In this study, ultra-deep sequence analysis detected DCV-resistant variants in 5 (50%) patients before treatment. In recent Japanese studies, the prevalence of NS3/4A protease inhibitor- and NS5A inhibitor-resistant variants in HCV genotype 1b-infected patients was reported to be approximately 4.9% and 11–23%, respectively, by direct sequence analysis,.
Patients with no ASV-resistant variants but with NS5A L31M/S and a high frequency of Y93H variants (32.4% and 99.4%) resulted in the development of double resistance variants, indicating that pre-existence of a high frequency of Y93H variants might be associated with relapse or viral breakthrough with ASV and DCV combination treatment. However, Case 3 achieved SVR despite a pre-existing NS5A L31V variant and a high frequency (99.4%) Y93H variant. These results suggest that pre-existing DCV-resistant variants might be associated with viral breakthrough for DCV and ASV combination treatment; however, identifying pre-existing resistant variants by ultra-deep sequence seems to have limited utility in predicting the outcome of therapy.
Karino et al. reported a relationship between pre-existing drug-resistant variants by direct sequencing analysis and clinical antiviral responses to DCV and ASV combination treatment. McPhee et al. also reported that six of seven genotype 1a HCV-infected patients treated with ASV and DCV developed viral breakthrough even though no resistance variants were detected at baseline by population sequencing analysis. Ultra-deep sequence analysis may permit more detailed analysis of resistance variants.
In Case 9, although ASV-resistant variants had not been detected before treatment, the frequency of D168A had reached 100% by 16 weeks after cessation of treatment, indicating that the wild type amino acid had been completely replaced. Similarly, in Case 10, the frequency of the D168V variant had already reached 31.8% by week 10, and the frequency increased to 45.7% by week 16. In both patients, NS5A aa31 and 93 were predominantly replaced by DCV-resistant variants. In Case 10, NS3 aa168 had completely returned to wild type 16 weeks after cessation of the treatment, while NS5A aa31 was completely replaced by the DCV-resistant variant. We previously reported that TVR-resistant variants have reduced replication capacity and are easily replaced by wild type when TVR is not present. Karino et al. reported that DCV-resistant substitutions persisted through 48 weeks post-treatment, whereas ASV-resistant substitutions were no longer detectable by direct sequence analysis in viral breakthrough patients treated with DCV and ASV. Long-term follow-up of these variants by ultra-deep sequence analysis is required to fully understand their fitness vs wild type sequence. The analysis of a larger number of patients is now ongoing.
In conclusion, 10 patients with HCV genotype 1b infection were treated with ASV and DCV combination treatment. This treatment is expected to improve the SVR rate greatly, but viral breakthrough might develop in some patients with the emergence of ASV- and DCV-resistant variants. Patients with a high frequency of pre-existing DCV-resistant variants might be more susceptible to viral breakthrough during combination therapy, although it remains to be seen whether ultra-deep sequencing analysis of resistance variants prior to treatment can effectively predict treatment outcome.
Discussion
Drug resistance has been shown to emerge with different classes of DAA regimens. However, the reason why treatment fails in some patients remains unclear. Amino acid substitutions in HCV related proteins that confer resistance to DAAs can exist at low frequency prior to antiviral treatment with DAAs. Enrichment of variants during therapy has been reported, although monitoring changes in variant frequency using ultra-deep sequencing is not commonly performed. HCV is an error-prone RNA virus in which substitutions frequently occur throughout the HCV genome, and drug-resistant variants are sometimes present as minor populations in patients who have never been exposed to DAAs.
In this study, ultra-deep sequence analysis detected DCV-resistant variants in 5 (50%) patients before treatment. In recent Japanese studies, the prevalence of NS3/4A protease inhibitor- and NS5A inhibitor-resistant variants in HCV genotype 1b-infected patients was reported to be approximately 4.9% and 11–23%, respectively, by direct sequence analysis,.
Patients with no ASV-resistant variants but with NS5A L31M/S and a high frequency of Y93H variants (32.4% and 99.4%) resulted in the development of double resistance variants, indicating that pre-existence of a high frequency of Y93H variants might be associated with relapse or viral breakthrough with ASV and DCV combination treatment. However, Case 3 achieved SVR despite a pre-existing NS5A L31V variant and a high frequency (99.4%) Y93H variant. These results suggest that pre-existing DCV-resistant variants might be associated with viral breakthrough for DCV and ASV combination treatment; however, identifying pre-existing resistant variants by ultra-deep sequence seems to have limited utility in predicting the outcome of therapy.
Karino et al. reported a relationship between pre-existing drug-resistant variants by direct sequencing analysis and clinical antiviral responses to DCV and ASV combination treatment. McPhee et al. also reported that six of seven genotype 1a HCV-infected patients treated with ASV and DCV developed viral breakthrough even though no resistance variants were detected at baseline by population sequencing analysis. Ultra-deep sequence analysis may permit more detailed analysis of resistance variants.
In Case 9, although ASV-resistant variants had not been detected before treatment, the frequency of D168A had reached 100% by 16 weeks after cessation of treatment, indicating that the wild type amino acid had been completely replaced. Similarly, in Case 10, the frequency of the D168V variant had already reached 31.8% by week 10, and the frequency increased to 45.7% by week 16. In both patients, NS5A aa31 and 93 were predominantly replaced by DCV-resistant variants. In Case 10, NS3 aa168 had completely returned to wild type 16 weeks after cessation of the treatment, while NS5A aa31 was completely replaced by the DCV-resistant variant. We previously reported that TVR-resistant variants have reduced replication capacity and are easily replaced by wild type when TVR is not present. Karino et al. reported that DCV-resistant substitutions persisted through 48 weeks post-treatment, whereas ASV-resistant substitutions were no longer detectable by direct sequence analysis in viral breakthrough patients treated with DCV and ASV. Long-term follow-up of these variants by ultra-deep sequence analysis is required to fully understand their fitness vs wild type sequence. The analysis of a larger number of patients is now ongoing.
In conclusion, 10 patients with HCV genotype 1b infection were treated with ASV and DCV combination treatment. This treatment is expected to improve the SVR rate greatly, but viral breakthrough might develop in some patients with the emergence of ASV- and DCV-resistant variants. Patients with a high frequency of pre-existing DCV-resistant variants might be more susceptible to viral breakthrough during combination therapy, although it remains to be seen whether ultra-deep sequencing analysis of resistance variants prior to treatment can effectively predict treatment outcome.
Source...