Hepatitis C virus (HCV) protease variability and anti-HCV protease inhibitor resistance in HIV/HCV-coinfected patients

Authors


Dr Pascale Trimoulet, Laboratoire de Virologie, Hôpital Pellegrin Tripode, Place Amélie Raba-Léon, EA 2968, Bordeaux University Hospital, Bordeaux Cedex 33076, France. Tel: +33 5 56 79 55 10; fax: +33 5 56 79 56 73; e-mail: pascale.trimoulet@chu-bordeaux.fr

Abstract

Objectives

Data on the natural selection of isolates harbouring mutations within the NS3 protease, conferring resistance to hepatitis C virus (HCV) protease inhibitors (PIs), are limited for HIV/HCV-coinfected patients. The aim of this study was to describe the natural prevalence of mutations conferring resistance to HCV PIs in HIV/HCV-coinfected patients compared with HCV-monoinfected patients.

Methods

The natural prevalences of HCV PI resistance mutations in 120 sequences from HIV/HCV-coinfected patients (58 genotype 1a, 18 genotype 1b and 44 genotype 4) and 501 sequences from HCV-monoinfected patients (476 genotype 1 and 25 genotype 4), retrieved from GenBank as a control group, were compared.

Results

Of 76 sequences from HIV/HCV genotype 1-coinfected patients, six (7.9%) showed amino acid substitutions associated with HCV PI resistance (V36L, n=1; V36M, n=2; T54S, n=2; R155K, n=1). In 31 of 476 (6.5%) HCV genotype 1 sequences retrieved from the GenBank database, HCV PI resistance mutations were found. The difference was not statistically significant (P=0.6). All of the sequences from HIV/HCV genotype 4-coinfected patients and those retrieved from the GenBank database had amino acid changes at position 36 (V36L).

Conclusion

Our study suggests that the natural prevalence of strains resistant to HCV PIs does not differ between HCV-monoinfected and HIV/HCV-coinfected patients. Further studies on larger cohorts are needed to confirm these findings and to evaluate the impact of these mutations in clinical practice.

Introduction

It is hoped that specifically targeted antiviral therapies for hepatitis C virus (HCV) (STAT-C) will greatly improve the therapeutic management of individuals chronically infected with HCV genotype 1 or 4. In particular, new protease inhibitors (PIs) blocking the NS3 protease-dependent cleavage of the HCV polyprotein have recently been tested in clinical trials, and available data for telaprevir and boceprevir are encouraging [1–3].

The high level of HCV variability and diversity is an ongoing challenge for STAT-C. The natural presence of resistant variants at baseline offers the potential for their rapid selection during treatment. Numerous drug resistance substitutions have been shown to develop in vitro (Q41, F43, T54, R109, S138, R155, A156, D168 and V170) [4] and in patients treated with HCV PIs (V36, T54, V55, Q80, R155, A156, V158, D168 and V170) [3–5].

One-third of HIV-infected patients in the USA and in Europe are coinfected with HCV through common routes of transmission. The combination of pegylated interferon (PEG-IFN) plus ribavirin for 48 weeks results in a sustained virological response in 35% of HIV/HCV genotype 1 or 4-coinfected patients [6]. Approaches using HCV PIs may be of interest, in view of the high rate of resistance to standard HCV treatment and the faster progression of HCV-related liver diseases in HIV-coinfected patients. The selection pressure exerted by humoral and cellular immune responses on HCV in HIV-coinfected patients is different from that observed in HCV-monoinfected patients [7]. Consequently, previous data concerning NS3 protease natural polymorphism in HCV-monoinfected patients may not be relevant in HIV/HCV-coinfected patients [8,9].

In the light of these observations, the aim of the study was to describe the natural prevalence of mutations conferring resistance to HCV PIs in HIV/HCV-coinfected patients compared with HCV-monoinfected patients.

Methods

Plasma samples for HCV protease analysis were obtained from 120 HIV/HCV-coinfected patients (58 genotype 1a, 18 genotype 1b and 44 genotype 4) included in the Aquitaine cohort [10]. Patients were recruited from the Department of Infectious Diseases, Pellegrin Hospital (Bordeaux, France). For inclusion in the study, patients had to be positive for serum HCV RNA, harbour HCV genotype 1a, 1b or 4, and be naïve to any novel or investigational anti-HCV drug. The NS3 protease domain (amino acids 1027–1207) was amplified by reverse-transcriptase nested polymerase chain reaction (PCR). PCR products were directly sequenced and multiple alignment of nucleotides and deduced amino acid sequences was inferred using Clustal_W, version 1.74 (Conway Institute UCD, Dublin, Ireland).

We retrieved 501 available sequences (476 genotype 1 and 25 genotype 4) from the GenBank database as a control group. These sequences were chosen from HCV-monoinfected patients only and to concern exclusively the NS3 protease domain. Phylogenetic criteria were used to exclude very closely related sequences (that is, cases of clonal sequences from the same patient and time-point) from the data set.

Fisher's exact test was used to compare the frequencies of mutations at positions 36, 54, 155, 156 and 170, which are known to confer resistance to HCV PIs [3,5], in the sequences obtained from HIV/HCV-coinfected individuals and the GenBank control group. Patients' characteristics were compared according to the presence of HCV PI resistance mutations using a Fisher's exact test for qualitative variables and a Wilcoxon–Mann–Whitney test for quantitative variables. Distributions are described as medians [with 25th and 75th percentiles, and interquartile range (IQR)]. All statistical tests were two-sided. Statistical analyses were performed using sas 9.1 (SAS Institute Inc., Cary, NC, USA).

Results

At the time of HCV protease analysis, the median age of the HIV/HCV-coinfected patients was 47 years (IQR 45–49 years). Eighty patients were male. One hundred and ten patients had received antiretroviral therapy for at least 6 months. The median HIV load was 40 HIV-1 RNA copies/mL (range 20–560 800 copies/mL) and the median CD4 cell count was 474 cells/μL (range 3–1671 cells/μL). Eighty-two patients had never been treated for their chronic hepatitis C, whereas 38 were relapsers or nonresponders to previous anti-HCV treatment.

Of 76 sequences from HIV/HCV genotype 1-coinfected patients, six (7.9%) showed amino acid substitutions associated with HCV PI resistance. Three patients showed a mutation at position 36 known to confer low-level resistance to HCV PIs: V36L in one patient and V36M in the other two. Three patients carried mutations conferring intermediate or high levels of resistance to HCV PIs: R155K and T54S in one and two patients, respectively. In 31 (6.5%) of 476 HCV genotype 1 sequences retrieved from the GenBank database, HCV PI resistance mutations were found. Amino acid mutations detected in the sequences were as follows: V36L in six sequences, V36M in six, T54S in 11, R155K in five, V170A in one, and T54S+R155K in two. The proportion of patients with HCV PI resistance mutations was not significantly different between HIV/HCV-coinfected and HCV-monoinfected patients (P=0.6). None of the HCV genotype 1-infected patients (either monoinfected or HIV-coinfected) showed a substitution at position A156, known to confer the highest level of resistance to telaprevir or boceprevir. The clinical and virological characteristics of the 76 genotype 1 HIV/HCV-coinfected patients are presented in Table 1. Patients' characteristics did not differ between the group of six HIV/HCV-coinfected patients harbouring HCV protease mutations and those without known HCV PI resistance mutations (Table 2).

Table 1.   Clinical and virological characteristics of the 76 HIV/hepatitis C virus genotype 1-coinfected patients
Patient characteristic 
  1. IQR, interquartile range; HCV, hepatitis C virus.

Age (years) [median (IQR)] (n=76)46 (44–48)
Female gender [n/total (%)]23/76 (30.3)
Log10 HCV RNA (IU/mL) [median (IQR)] (n=59)6.47 (6.17–6.74)
Genotype 1a [n/total (%)]58/76 (76.3)
Genotype 1b [n/total (%)]18/76 (23.7)
Previous anti-HCV treatment [n/total (%)]32/76 (42.1)
Cirrhosis [n/total (%)]17/72 (23.6)
HIV RNA<50 copies/mL [n/total (%)]53/74 (69.7)
CD4 count (cells/μL) [median (IQR)] (n=74)481 (272–662)
Antiretroviral therapy [n/total (%)]72/76 (94.7)
HIV protease inhibitor treatment [n/total (%)]55/76 (72.4)
Table 2.   Patient characteristics according to the presence of HCV protease inhibitor resistance mutations
 Nonmutated NS3 proteaseMutated NS3 protease* (n=6)P
  • IQR, interquartile range; HCV, hepatitis C virus.

  • *

    Having ≥1 mutation from the following list: V36L, V36M, T54S and R155K.

  • Numbers of patients with complete data:

  • n=70.

  • n=53.

  • §

    n=68.

Age (years) [median (IQR)]46 (44–48)47 (45–49)0.60
Female gender [n/total (%)]22/70 (31.4)1/6 (16.7)0.66
Log10 HCV RNA (IU/mL) [median (IQR)]6.48 (6.22–6.73)6.28 (6.11–6.99)0.79
Genotype 1a [n/total (%)]54/70 (77.1)4/6 (66.7)0.60
Genotype 1b [n/total (%)]16/70 (22.9)2/6 (33.3)
Previous anti-HCV treatment [n/total (%)]28/70 (40.0)4/6 (66.7)0.23
Cirrhosis [n/total (%)]16/66 (21.2)1/6 (16.7)1.00
HIV RNA<50 copies/mL [n/total (%)]49/68 (72.0)4/6 (66.7)1.00
CD4 (cells/μL) [median (IQR)]467 (287–655)§588 (117–709)0.91
Antiretroviral therapy [n/total (%)]66/70 (94.3)6/6 (100.0)1.00
HIV protease inhibitor treatment [n/total (%)]50/70 (71.4)5/6 (83.3)1.00

All of the sequences from HIV/HCV genotype 4-coinfected patients and those retrieved from the GenBank database had amino acid changes at position 36 (V36L) shown to confer decreased susceptibility to telaprevir.

Finally, the NS3 catalytic triad (H57, D81 and S139) was highly conserved among the 120 sequences from HIV/HCV-coinfected patients.

Discussion

We found no significant difference in natural polymorphisms at positions associated with HCV PI resistance between HCV-monoinfected and HIV/HCV-coinfected patients.

Our results are in accordance with those of a study by Halfon et al. in a small group of patients. They did not find any difference in observed mutation rates between HCV-monoinfected and HIV/HCV-coinfected patients (19% and 18%, respectively) at positions associated with HCV PI resistance [9]. In contrast, Morsica et al. found a higher prevalence of HCV PI resistance mutations in 37 sequences obtained from coinfected patients in comparison with 250 sequences from HCV-monoinfected patients retrieved from the GenBank database (16.2% and 0.8%, respectively) [8]. In our study, which included a large number of patients, previous HCV treatment did not seem to influence the prevalence of HCV PI resistance mutations.

No patient showed substitutions at position A156, which are known to confer the highest level of resistance to telaprevir or boceprevir. The role of other mutations is difficult to predict, but the possibility that they may have an impact on the virological response to treatment cannot be excluded and needs to be investigated. Indeed, HCV strains with naturally occurring mutations that may confer resistance to HCV PIs show reduced fitness and are generally sensitive to interferon and/or interferon plus ribavirin therapy regimens. The role of these mutations in long-term therapy and the likelihood of viral breakthroughs are still to be determined, in particular in patients who are nonresponders to previous interferon-based therapy or relapsers on this therapy. The preservation of the NS3 catalytic triad, as observed in our study, is probably attributable to functional constraints on the protease. Its structural and chemical integrity is required to process the HCV polyprotein.

All sequences from genotype 4-infected patients contained mutation V36L, which is known to confer decreased susceptibility to telaprevir [11]. Large clinical trials to better document the efficacy of STAT-C in patients infected with genotype 4 are required.

Our study on sequences from 120 HIV/HCV-coinfected patients suggests that the natural prevalence of strains resistant to HCV PIs does not differ between HCV-monoinfected and HIV/HCV-coinfected patients. Further studies on larger cohorts are needed to confirm these findings and to evaluate the impact of these mutations in clinical practice.

Acknowledgements

We thank Janssen-Cilag for their support.

Ancillary