The selection of HCV variants that are resistant to active-site protease inhibitors by amino acid substitutions in the HCV protease domain of NS3 was demonstrated in vitro and in clinical trials of DAA monotherapy.
22–
28 Several amino acid positions are associated with resistance, yet others may act as compensatory mutations that restore fitness to resistant isolates. summarizes the available data to date on the emergence of HCV resistance to BOC and other NS3/4A protease inhibitors.
| Table 1Mutations in the HCV NS3/4A protease inducing resistance to HCV protease inhibitors |
The emergence of compound-specific HCV resistance is rapid in vivo, and can even occur within the first 2 weeks of exposure to a given DAA;
17,
29–
31 however, some resistant strains may show reduced fitness.
32 HCV is a highly variable virus with a high mutation rate and a large population size that circulates as a swarm of closely related variants that can be rapidly selected.
33 Minority-resistant HCV variants to new DAAs (not detectable by direct population sequencing) can be hidden within the complex genetic pool of the virions that circulate in a single infected patient.
34 Using clonal sequence analysis, Susser et al found that minority resistance mutations can be selected at six positions within the HCV NS3 protease during BOC therapy, although 2 weeks after the end of treatment with 400 mg BOC twice or three times daily, the frequency of resistant variants declines and the number of wild-type viruses increases to 95%.
35 However, it remains to be determined if such low-frequency resistant variants can compromise subsequent treatment options in the case of treatment failure, because the rapid selection of low-frequency resistant variants was observed during retreatment.
29The available data on HCV resistance have been reviewed recently.
12,
36 A major feature of resistance to HCV protease inhibitors is cross-resistance. The most relevant resistance mutations are substitutions in residues R155 and A156, which confer high levels of resistance to BOC and TPV and cross-resistance to most NS3 protease inhibitors. Other mutations at V36, T54, and V170 are associated with low levels of resistance to both TPV and BOC. It is important to note that (i) some TPV-resistant variants remain detectable for up to 4 years after cessation of treatment,
37 and (ii) that late relapse may occur 24–36 weeks after TPV + peg-IFN-α + RBV therapy.
31HCV-resistant mutations can also emerge rapidly when BOC is used in monotherapy or is combined with peg- IFN-α.
35 After BOC monotherapy, high frequencies of resistant variants were detected by clonal sequencing of HCV quasi-species in some patients at their 1-year follow-up. Moreover, resistant mutations are rapidly selected during retreatment with BOC + peg-IFN-α in some patients.
29,
38 In the SPRINT-1 trial emergence of resistant viruses, assessed by population sequencing, was detected early on viral breakthrough, mainly in mutations V36M, T54S, and R155K (>25% of samples), as well as T54A, V55A, R155T, A156S, V158I, and V170A (5%–25% of samples), and V36A, V36L, and I170T (<5% of samples).
20 In addition, more than 25% of patients with viral breakthroughs carried cross-resistant mutations of both BOC and TPV.
20 Data from the follow-up study, P05063 (NCT00689390), indicate that at least one resistant mutation persists for more than 1 year in patients who did not achieve SVR in previous BOC trials (n = 174).
39 The most common resistance mutations were R155K (64%), T54S (54%), V36M (54%), and T54A (22%) during follow-up. Furthermore, the overall reversion to the wild-type virus was seen in 59% of the patients over a 2-year period, but T54S- and R155K-carrying viruses reverted more slowly. The authors reported no late relapse in patients who achieved SVR (n = 290) in this follow-up cohort.
39Because HCV isolates are widely variable, and infections with HCV genotypes other than genotype 1 account for a large number of chronic carriers worldwide,
40 defining the variability of HCV NS3/4A protease in natural isolates will be an important step in determining the potential selection of naturally resistant strains, as in the case of human immunodeficiency virus (HIV).
41 A relevant polymorphism of NS3/4 proteases between HCV subtypes was found in sites associated either with resistance or with compensatory mutations after an analysis of more than 350 worldwide viral isolates (genotypes 1–6).
42 For instance, V170A (which confers low levels of resistance to BOC) was present in 184/275 HCV genotype 1 isolates, and D168V/A was an amino acid signature in HCV genotype 3, which explains the reduced sensitivity of genotype 3 viruses to ciluprevir, and potentially to other protease inhibitors.
42 The different genetic barriers to resistance between HCV subtypes 1a and 1b illustrate the relevance of the variation between the genotypes and subtypes. The genetic barrier refers to the number of nucleotide substitutions required for the virus to acquire resistance to the drug. For BOC and TPV, the differences in genetic barriers include higher viral breakthrough rates and the selection of resistant variants observed in patients infected with subtype 1a, compared to those with subtype 1b. The resistance mutation, R155K, emerges through a single nucleotide substitution in subtype 1a viruses, but requires two different substitutions in subtype 1b viruses.
43Later studies involving a large number of patients also detected variants associated with resistance to PI in 5.5% and 1.4% of patients from the United States, Switzerland, and Germany who were infected with subtypes 1a or 1b, respectively, including the V36L/M and R115K variants associated with low or high-level resistance.
44 In HCV genotype 1-infected patients, 0.9% and 0.7% of viruses carried the V36M or the R155K variants, respectively, and patients with the R155K virus appeared to have slower viral load declines during TVR + peg-IFN-α + RBV treatment than those with wt viruses. Finally, in patients from Australia, Switzerland, and the UK, the prevalence of single resistance mutations to NS3 protease inhibitors can account for up to 4.4% of viral isolates, and the frequencies for single or combined resistance mutations to NS3 protease and/or NS5b polymerase inhibitors can be found in up to 21.5%, 44.4%, or 41.8% for subtypes 1a, 1b, or 3a, respectively.
45 Although the overall frequency of single resistance mutations is low in all of these studies, naturally occurring polymorphisms that confer resistance to DAAs could eventually compromise the treatment response of DAA-based regimes.
46In summary, the absence of response to triple BOC + peg- IFN-α + RBV therapy is associated with the selection of viral resistant mutations. Because several second-generation NS3 protease inhibitors are in advanced clinical development, the selection of viral resistance may compromise future therapeutic options involving DAAs of the same class, and therefore, should be avoided whenever possible.