A retrospective study of a prospective cohort was conducted at San Cecilio University Hospital (Granada, Spain) from 1991 until 2015 (Figure 1). A total of 680 patients with CHC were treated with 941 anti-HCV treatments (historical cohort). Of these patients, 478 had genotype 1/4 HCV, of whom 139 received two or more anti-HCV treatments. A total of 375 treatments were applied. The study cohort was selected by applying the following inclusion criteria: patients with genotype 1/4 HCV, two or more anti-HCV treatments, and blood samples stored at -80 °C. A total of 72 patients with stored blood samples, having received a total of 201 treatments, were included in the study (Study cohort; Figure 1).

Open in New Tab   Full Size Figure   Download Figure

Figure 1 Historical and study chronic hepatitis C patient cohort from 1991 to 2015 in the San Cecilio University Hospital. ¹First-generation DAAs (n = 13): peg-IFN-α + RBV + telaprevir (n = 7) and peg-IFN-α + RBV + boceprevir (n = 6); ²Second-generation DAAs (n = 17): peg-IFN-α + RBV + simeprevir (n = 2), simeprevir + sofosbuvir + RBV (n = 2), simeprevir + sofosbuvir (n = 1), sofosbuvir + daclatasvir (n = 1), ledipasvir + sofosbuvir + RBV (n = 4), ledipasvir + sofosbuvir (n = 4) and ombitasvir/paritaprevir/ritonavir + dasabuvir (n = 3). DAAs: Direct-acting antivirals NVR: No virological response; IFN: Interferon; peg-IFN + RBV: Pegylated-interferon + ribavirin; RBV: Ribavirin; SVR: Sustained virological response.

A total of 48/72 patients (66.7%) achieved SVR after several treatments. Of the 201 treatments, 153 resulted in no virological response (NVR), and 48 led to SVR. The outcomes of treatments resulting in NVR were defined as follows: relapse (R), treatment with serum HCV RNA undetectable in patients at 12 wk but detectable subsequently; partial virological response (pVR), treatment with a low viral load > 2log in patients at 12 wk; null response (NR), treatment with a low viral load < 2log in patients at 12 wk; and, unknown, i.e., treatments for patients whose reason for NR was not stated in their medical history.

The study protocol conformed to the ethical guidelines of the 1975 Helsinki Declaration, and all patients involved in the study were informed verbally and in writing of the characteristics of the study and provided their consent to participate by signing an informed consent form. This study was approved by the Ethics Committee of the San Cecilio University Hospital.

HCV genomes were analysed in serum samples obtained before the patients were subjected to treatments involving different antiviral therapies. The sequences of viruses from baseline serum samples associated with 201 treatments were amplified; samples that failed long RT-PCR amplification were excluded (n = 16). The amplified HCV genome region corresponded to a fragment from gene NS5A (nucleotides 6966 to 7164 in the reference HCV genome sequence, GenBank accession number AF009606), including the PKR binding domain (PKRBD) and the interferon-sensitivity-determining region (ISDR). These regions include structural motifs related to antiviral treatment.

Viral RNA was extracted from 300 μL of serum using the PrepitoViral DNA/RNA 300 Kit (Chemagen Technologie, PerkinElmer, Madrid, Spain). Two microlitres of the RNA solution were subjected to reverse transcription using a qScript Flex cDNA Synthesis Kit (Quanta Biosciences, Gaithersburg, MD, United States). PKRBD sequences were amplified via RT-PCR using SYBR Green Supermix Low ROX (Quanta Biosciences). The primers used for reverse transcription and first and second round amplifications of the NS5A region (nt 6966-7164) are shown in Table 1.

RT was performed at 45 °C for 30 min and terminated at 85 °C for 5 min, followed by the first-round PCR over 35 cycles, with each cycle consisting of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s and extension at 72 °C for 70 s. The second-round PCR was performed under the same conditions. The sequences of the amplified fragments were determined via direct sequencing. The second-round PCR product was purified using the Wizard SV Gel and the PCR Clean-Up System (Promega, Promega Biotech Ibérica, Madrid, Spain) and resuspended in 20 μL of water. Next, 2-5 μL of purified PCR product was sequenced in a 3130 XL Genetic Analyzer automatic sequencer (Applied Biosystems, Foster City, CA, United States).

All of the nucleotide sequences obtained in the present study were aligned. To detect amino acid substitutions, sequences were aligned with reference strands, HCV-J (GenBank accession number D90208) for genotype 1b and HCV-1 (GenBank accession number M62321) for genotype 1a, and translated into amino acid sequences with Mega 5.0 software (Molecular Evolutionary Genetics Analysis)[26].

Quantitative variables are expressed as the means ± SD, and qualitative variables are expressed as absolute values with percentages. The χ² test was used for qualitative variables. Comparisons of continuous variables with a non-normal distribution were performed using the Mann-Whitney U test or Kruskal-Wallis test. The degree of association between dependent and independent variables was determined by calculating the crude odds ratio and its 95%CI (bivariate analysis). Interaction studies were performed using logistic regression. Synergy studies were conducted in accordance with the method of Andersson et al[27]. The criterion for statistical significance was P ≤ 0.05. All statistical analyses were performed using SPSS 15.0 for Windows software (SPSS Inc, Chicago, IL, United States).

The characteristics of the historical cohort (n = 680 patients) are shown in Table S1. After application of the inclusion criteria, 72 patients were selected [45/72 (63.5%) were men], with a total of 201 treatments (72 first treatments and 129 retreatments; Figure 1). The number of treatments received was distributed as follows: 2 treatments (n = 35, 48.6%), 3 treatments (n = 24, 33.3%), 4 treatments (n = 9, 12.5%), 5 treatments (n = 3, 4.2%) or 8 treatments (n = 1, 1.4%). The following anti-HCV therapies were provided: IFN-α alone (24%, n = 49), IFN-α + RBV (37%, n = 75), peg-IFN-α + RBV (23%, n = 47) and DAAs (16%, n = 30) (including first-generation DAAs: peg-IFN-α + RBV + telaprevir 24% and peg-IFN-α + RBV + boceprevir, 20%; and second-generation DAAs: peg-IFN-α + RBV + simeprevir, 7%; simeprevir + sofosbuvir + RBV, 7%; simeprevir + sofosbuvir, 3%; sofosbuvir + daclatasvir, 3%; ledipasvir + sofosbuvir + RBV, 13%; ledipasvir + sofosbuvir, 13% and ombitasvir/paritaprevir/ritonavir + dasabuvir, 10%).

One patient was treated with eight different therapies (one treatment with IFN-α (Wellferon; the Wellcome Foundation Ltd, part of GlaxoSmithKline PLC, Brentford, United Kingdom), three with IFN-α + RBV, two with peg-IFN-α + RBV, one with peg-IFN-α + RBV + telaprevir and one with simeprevir + sofosbuvir + RBV) and finally achieved SVR. Of the 72 patients, 66.7% (n = 48) achieved SVR, while 33.3% (n = 24) did not respond to any of the treatments received. In the latter scenario, the clinical decision in almost all cases was not to perform a retreatment until a new treatment became available. Regarding the 201 total treatments in our cohort, 24% (48 treatments) achieved SVR, while 76% (153 treatments) showed NVR. The types of NVR in the 153 treatments were as follows: R, 21% (n = 32); pVR, 8% (n = 13); NR, 34% (n = 52); and unknown 37% (n = 56).

We then analysed how the prior type of SVR or NVR affected subsequent treatment responses (Table 2), and we found that previously treated patients showing R were 6.09 times more likely to respond to a new treatment than were patients with a previous NR (P = 0.006). In our study cohort, we also analysed the response achieved according to the type of therapy employed (Table 3). As expected, retreatments based on therapies introduced before the use of DAAs were significantly less likely to achieve SVR than were those based on DAAs.

When we distinguished between first- and second-generation DAA treatments, as expected, logistic regression analysis showed that first-generation DAA treatments were significantly less likely to achieve SVR than second-generation DAA treatments (OR = 0.16, 95%CI: 0.026-0.97, P = 0.04). Regarding this analysis, it should be noted that the patients recruited from the historical cohort who responded to the initial treatment (treatment-naïve patients) were not included in the study cohort, which accounts for the low rates of SVR observed in the study. To avoid bias in the analysis, the first 72 treatments were removed from the analysis as these all involved NVR according to the application of the inclusion criteria, regardless of the treatment received.

The change from one type of therapy to another in retreatments increased the probability of achieving SVR. In our cohort, of the total of 129 retreatments, 82 were based on a change to a more effective therapy, while 47 involved no change in therapy. In neither case was the difference in responses statistically significant (data not shown). It should be noted that the 2 patients who did not respond to treatment with second-generation DAAs (relapsers) previously showed NVR to treatment with first-generation DAAs.

The genetic evolution of the HCV NS5A-PKRBD and NS5A-ISDR regions in the course of the different treatments received was investigated using 201 baseline samples from each treatment, of which 16 were subsequently excluded due to non-amplification (total sample, n = 185). Figure 2 shows the number of mutations found in each region for the 185 amplified samples. The mean number of mutations in the PKRBD region (min-max = 1-15) compared to the number in the reference sequence HCV-J and HCV-1 was 5.19 ± 2.6; in the ISDR (min-max = 0-10), this number was 1.85 ± 2.1.

Open in New Tab   Full Size Figure   Download Figure

Figure 2 Number of baseline mutations by type of treatment in the interferon-sensitivity-determining region (A) and PKR binding domain (B) of NS5A. PKRBD: PKR binding domain; ISDR: Interferon-sensitivity-determining region.

With respect to the association between the number of mutations and the response to antiviral treatment, the treatments that resulted in SVR were associated with significantly more baseline mutations in both regions than were the treatments resulting in NVR (SVR vs NVR; PKRBD: 5.82 ± 3 vs 4.86 ± 2 mutations, P = 0.045; ISDR: 2.65 ± 2 vs 1.51 ± 1.7 mutations, P = 0.005). When we distinguished patient outcomes by type of NVR (R, pVR, NR), significant differences were found in the ISDR region (P = 0.029; Table 4), with R patients presenting fewer mutations than SVR patients (P = 0.025). In the ISDR region, and applying the criteria recommended by Enomoto et al[19], we found that treatments associated with 4 or more mutations in the baseline sample showed an SVR rate of 65%, compared to 31% for those associated with 0-3 baseline mutations (P = 0.002; Figure 3).

Open in New Tab   Full Size Figure   Download Figure

Figure 3 Relationship between sustained virological response and number of baseline mutations in the interferon-sensitivity-determining region of NS5A. The number of baseline mutations in the interferon-sensitivity-determining region is based on the criteria of Enomoto et al[19]. Comparisons between groups were made using χ² tests. SVR: Sustained virological response; NVR: No virological response.

With respect to changes in the viral population during the time between one treatment and another and its relationship with the response to antiviral therapy, our analysis shows that when there was no change in the number of mutations both in the PKRBD and in the ISDR, or when this number decreased, between one treatment and another, the patient tended to be unresponsive to antiviral therapy (PKRBD, P = 0.02; ISDR, P = 0.001) (Table 5). From a clinical standpoint, we studied the viral variability of the ISDR region of the HCV between one treatment and another to determine how this variability was associated with the response or non-response to antiviral therapies (i.e., its use as a diagnostic test), and the following diagnostic values were obtained: 33% sensitivity, 90% specificity, 68% positive predictive value and 70% negative predictive value.

Viral load is another variable that can be considered a predictor of the treatment response. In our cohort study, logistic regression analysis showed that treatments for patients with a baseline viral load ≤ 600000 IU/mL were 3.3 times more likely to result in SVR (OR = 3.3, 95%CI: 1.4-7.7, P = 0.005). However, the analysis also showed that the variation in response according to baseline viral load was influenced by the change in the number of mutations between one treatment and another (Figure 4). Thus, when the treatments were associated with a low baseline viral load (≤ 600000 IU/mL), a response was usually achieved, whereas for treatments associated with a baseline viral load > 600000 IU/mL, depending on the genetic variability of the virus, a response may or may not be achieved; when the number of mutations had increased since the previous treatment, there was a 71% probability of SVR (P ≤ 0.0001).

Open in New Tab   Full Size Figure   Download Figure

Figure 4 Interaction between baseline viral load and viral variability in the interferon-sensitivity-determining region. A: Viral load ≤ 600000 IU/mL; B: Viral load > 600000 IU/mL. Comparisons between groups were made with the χ² test. SVR: Sustained virological response; NVR: No virological response.

Using logistic regression, we then sought to corroborate these results for the interaction between the two variables (P = 0.023) (Table 6). Treatments in which the patient had a baseline viral load of > 600000 IU/mL were 9.21 times more likely to achieve SVR if there had been an increase in the number of mutations in the ISDR region since the earlier treatment (P = 0.001). However, in those treatments in which the patient’s baseline viral load was < 600000 IU/mL, the increase in the number of mutations did not affect the response (P = 0.75) (Table 6). No synergy was detected between the two variables [relative excess risk due to interaction (RERI) = -6.052; proportion attributable to the interaction (PA) = -0.74; synergy index, S = 0.54].

When antiviral treatments are used against hepatitis C virus (HCV), some patients are non-responders (NR). Our aim in this study was to analyse the change of mutations in the NS5A-interferon-sensitivity-determining region and PKR binding domain regions in patients with chronic hepatitis C genotype 1 and who showed NR to one or more treatment cycles over time. We also analysed the relationship between this change and sustained virologic response (SVR) to a new treatment.

To our knowledge, this is the first study providing data about how the genetic variability of HCV evolves in patients who undergo various treatments over a period of two decades.

This study represents a significant contribution to knowledge of the genetic diversity of the NS5A protein from HCV and its relationship to the treatment response in patients for whom clinical treatment is problematic and who achieve very low response rates.

This report shows that the genetic diversity of the HCV genome, during decades of infection, continues to be an independent predictor of SVR, regardless of the treatment applied. No previous studies with these particular characteristics have been performed.

This manuscript presents a single-centre, retrospective, observational study of the genetic evolution of HCV genotype 1 in patients who received antiviral therapies but were NR to prior therapies. The major strength of this manuscript is its relevance to contemporary practice, and the results could be applied to clinical decisions regarding therapeutic strategies to treat non-responders who receive novel DAAs.