Introduction

Hepatitis C virus (HCV) infection is a common cause of chronic hepatitis and hepatocarcinogenesis worldwide [1, 2]. As one of six HCV genotypes, genotype 2 accounts for up to 30% of chronic HCV infections in Japan. In treatment-naïve patients with HCV genotype 2, a combination of peg-interferon (PEG-IFN) plus ribavirin (RBV) results in sustained virological response (SVR) rates of 70–90% [3–5]. However, the combination treatment is poorly tolerated, and some patients do not respond.

Recently, a number of new direct-acting antivirals (DAA) have been developed that selectively target HCV proteins, including the nonstructural protein (NS)3/4A, a protease essential for cleaving the non-structural portion of the HCV polyprotein; NS5A, a phosphoprotein required for HCV replication; and NS5B, an RNA-dependent RNA polymerase required for synthesis of HCV RNA [6]. Sofosbuvir is an oral nucleotide analogue inhibitor of the HCV-specific NS5B polymerase. An IFN-free DAA therapy with sofosbuvir plus RBV has improved the SVR rate and shortened the duration of therapy for patients with chronic HCV genotype 2 infection [7–9]. In the phase 3 clinical trials, twelve weeks of therapy resulted in an SVR rate of 97% [10], and in 2015 the treatment was approved for treatment of HCV genotype 2 in Japan.

Two single nucleotide polymorphisms (SNPs) in the inosine triphosphatase (ITPA) locus (rs7270101 and rs1127354) have been found to be associated with treatment-related anemia in patients treated with PEG-IFN/RBV therapy [11, 12]. The rs7270101 A allele is fixed in the Japanese population, but patients with the rs1127354 CC genotype are prone to developing severe anemia during RBV therapy [12]. ITPA SNPs have also been shown to affect incidence of anemia in patients during treatment with PEG-IFN/RBV plus telaprevir or simeprevir triple therapy [13–15]. However, the effect of ITPA polymorphism on anemia and treatment response in IFN-free therapies, including patients treated with sofosbuvir plus RBV, has not been sufficiently evaluated.

In the present study, we assessed the effects of ITPA polymorphism on anemia, RBV dose reduction, and treatment response in HCV genotype 2-infected Japanese patients treated with sofosbuvir and RBV.

Methods

Patients

A total of 244 patients with chronic HCV genotype 2 infection treated with sofosbuvir and RBV combination treatment between June 2015 and February 2016 at Hiroshima University Hospital and hospitals belonging to the Hiroshima Liver Study Group were enrolled. Inclusion criteria for the study included remaining positive for genotype 2 HCV RNA for 6 months. All patients were administered 400 mg of sofosbuvir (Sovaldi®, Gilead Sciences Inc., USA) and 200–1000 mg of RBV (Copegus®, Chugai, or Rebetol®, MSD) for 12 weeks. The initial RBV dose was determined by body weight (600 mg for < 60 kg; 800 mg for 60-80 kg; and 1000 mg for >80 kg). End-of-treatment response (ETR) and SVR12 were defined as undetectable serum HCV RNA at the end of treatment and 12 weeks after the end of treatment, respectively.

Table 1 summarizes the baseline characteristics of the 244 patients. The study included 115 males and 129 females, aged 32 to 90 years (median, 70 years). All subjects gave written informed consent to participate in the study according to the process approved by the ethical committee of each hospital and conforming to the ethical guidelines of the 1975 Declaration of Helsinki.

Table 1 Patient characteristics
Full size table

Clinical and laboratory assessments

Clinical and laboratory assessments were performed at least once every month before, during, and after treatment. HCV RNA levels were measured using COBAS TaqMan HCV test (Roche Diagnostics, Tokyo, Japan). The detection limit for the assay was 1.2 log IU/mL. HCV genotype was determined by sequence determination of the 5′ non-structural region of the HCV genome, followed by phylogenetic analysis. FIB4 index were calculated as a surrogate marker of liver fibrosis [16]. FIB4 index = age (years) × aspartate aminotransferase (AST) [U/l]/(platelet count [109/l] × (ALT [U/l])1/2). We considered FIB4 index of ≥3.25 as severe fibrosis according to Sterling et al. [16].

Single-nucleotide polymorphism (SNP) genotyping

We genotyped each patient for two SNPs: rs8099917 in the IL28B locus associated with response to interferon therapy [17], and rs1127354 in the ITPA locus associated with RBV-induced anemia [11, 12]. Samples were genotyped using the Invader assay as described previously [12].

Statistical analysis

Differences in hemoglobin levels between patients treated with PEG-IFN/RBV therapy compared to sofosbuvir plus RBV therapy and between patients with and without the anemia-associated ITPA SNP genotype during sofosbuvir plus RBV therapy were analyzed by two-way repeated measures ANOVA followed by pairwise Mann–Whitney U tests with Bonferroni correction for multiple comparisons. The χ 2 and Mann–Whitney U tests were applied to detect significant associations. Multiple regression analysis was used to examine the association between treatment outcome and the values of other markers, using P < 0.1 as the criterion for inclusion in the multivariate model. All statistical tests were two-sided, and P < 0.05 was considered significant. All statistical analysis was performed using IBM SPSS Statistics 22 program (IBM SPSS, Inc., Chicago, IL, USA).

Results

Treatment efficacy

All patients achieved ETR following 12 weeks of sofosbuvir plus RBV therapy, but serum HCV RNA relapse occurred in 13 patients after the end of therapy. Overall, the SVR12 rate was 94.7% (231 out of 244 patients) based on intent-to-treat analysis. SVR12 rates were 95.3% (143 out of 150 patients) and 93.5% (87 out of 93 patients) for genotypes 2a and 2b, respectively.

Reduction of hemoglobin levels by peg-interferon plus ribavirin therapy compared to sofosbuvir plus ribavirin therapy

We first compared the dynamics of hemoglobin levels between PEG-IFN/RBV and sofosbuvir plus RBV therapies. To overcome bias due to the different distributions of covariates among patients from the two groups, one-to-one matches were created using propensity score analysis [18, 19]. Variables entered in the propensity model were age, sex, body weight, ITPA genotypes (CC or CA/AA), and hemoglobin levels. The model was then used to obtain a one to-one match by using the nearest-neighbor matching method [20, 21] in 473 PEG-IFN/RBV-treated and 244 sofosbuvir plus RBV-treated patients, resulting in a sample size of 93 patients per cohort (Table 2). Reduction of hemoglobin levels was significantly lower in sofosbuvir plus RBV treated patients compared to PEG-IFN/RBV treated patients (two-way ANOVA; treatment, P < 0.001; time, P < 0.001; interaction treatment × time, P < 0.001) (Fig. 1).

Table 2 Characteristics of propensity score-matched patient treated with PEG-IFN plus RBV or sofosbuvir plus RBV
Full size table
Fig. 1
figure 1

Reduction of hemoglobin levels during peg-interferon plus ribavirin (PEG-IFN/RBV) and sofosbuvir plus ribavirin (sofosbuvir/RBV) therapies. *P = 0.003. **P < 0.001

Full size image

Reduction of hemoglobin levels during sofosbuvir and ribavirin combination treatment with respect to ITPA SNP genotype

Decrease in hemoglobin levels during sofosbuvir plus RBV therapy was analyzed by ITPA genotype. A rapid decrease in hemoglobin levels during the first four weeks was observed in genotype CC patients (Fig. 2a). Reduction of hemoglobin levels was significantly greater in genotype CC patients than CA/AA patients at each time point from the 2nd week to the 12th week (two-way ANOVA; ITPA genotype, P = 0.001; time, P < 0.001; interaction ITPA genotype × time, P < 0.001). Frequencies of patients with minimum hemoglobin levels >12, 10–12, <10 g/dL during the therapy were 21.3, 53.3, and 25.4% and 61.7, 34.0, and 4.3% in the ITPA CC and CA/AA groups, respectively (P < 0.001) (Fig. 2b).

Fig. 2
figure 2

Reduction of hemoglobin levels by ITPA polymorphism during sofosbuvir and ribavirin combination treatment. a Time course of hemoglobin levels by ITPA rs1127354 genotype. Patients were grouped by rs1127354 genotype (CC or CA/AA). b Frequencies of patients with minimum hemoglobin levels >12, 10–12, and <10 g/dL during the treatment with respect to ITPA genotype. *P < 0.001

Full size image

Effect of ITPA genotype on total dosage and outcome of the therapy

In response to hemoglobin decline, RBV doses were reduced as necessary according to treatment guidelines and symptoms of anemia in each patient. Reduction of RBV doses began earlier, and the frequency of the patients with RBV dose reduction was significantly higher in ITPA genotype CC patients compared to CA/AA patients (P = 0.001) (Fig. 3a). Consequently, the amount of RBV administered relative to the amount planned for each patient was significantly lower in the ITPA CC group compared to the CA/AA group (84.7 and 100%, respectively; P = 0.018) (Fig. 3b). Although ITPA genotype was associated with the decline of hemoglobin and RBV dose reduction, SVR12 rates did not differ significantly between patients in the ITPA CC and CA/AA groups (94.7% [180 out of 190 patients] and 94.4% [51 out of 54 patients], respectively; P = 0.933) (Fig. 4).

Fig. 3
figure 3

Ribavirin dose reduction by ITPA polymorphism. a Time until ribavirin dose reduction by ITPA rs1127354 genotype. The difference in the number of weeks until the first ribavirin dose reduction grouped by ITPA genotype (CC or CA/AA) was determined by the log-rank test. b Differences in the ratio of the administered dose to the planned dose by ITPA genotype. Medians are shown as thick horizontal bars. Boxes cover the interquartile range, and tails show the minimum and maximum values. Circles indicate outliers of 1.5–3.0 IQR higher than 75th percentile or 1.5–3.0 IQR lower than 25th percentile. *P < 0.05

Full size image
Fig. 4
figure 4

Virological response for sofosbuvir plus ribavirin therapy grouped by ITPA genotype. ETR, end of treatment response. SVR12, sustained virological response at 12 weeks after the end of therapy

Full size image

Predictive factors associated with SVR12

Significant univariate predictors for SVR12 included FIB4 index and body weight (Table 3). Multivariate logistic regression analysis identified FIB4 index <3.25 (odds ratio [OR], 9.388 for ≥3.25; P = 0.005) and low body weight (OR, 1.059 for high body weight; P = 0.017) as independent predictors for SVR12. Patients with FIB4 index ≥3.25 were likely to fail to respond to sofosbuvir plus RBV therapy compared to those with FIB4 index <3.25 (89.8% [97 out of 108 patients] and 98.5% [134 out of 136 patients], respectively; P = 0.003).

Table 3 Univariate and multivariate analyses of factors associated with SVR12
Full size table

Discussion

This study examined the effect of the ITPA SNP genotype on anemia and treatment response for sofosbuvir and RBV combination treatment in HCV genotype 2-infected patients.

Patients treated with sofosbuvir plus RBV treatment experienced less reduction in hemoglobin levels than patients treated with PEG-IFN plus RBV therapy, suggesting that the addition of PEG-IFN exacerbates the risk of anemia during treatment with RBV. Given the milder effect on hemoglobin levels and shorter duration of therapy, treatment-related anemia seems to be more manageable in sofosbuvir plus RBV therapy than in PEG-IFN/RBV treatment. In fact, no patient discontinued therapy due to anemia in the present study.

However, patients with the ITPA CC genotype were more vulnerable to anemia and experienced significantly faster hemoglobin decline than patients with ITPA CA or AA genotypes during sofosbuvir plus RBV therapy. Therefore, ITPA CC patients required an earlier reduction of the dose of RBV, and RBV dose adherence was significantly lower than in patients with ITPA CA or AA genotypes. Although ITPA genotype was associated with treatment-related anemia and adherence of RBV doses in sofosbuvir plus RBV treatment, no effect on treatment response was observed. Conversely, in PEG-IFN/RBV treatment, RBV is correlated with HCV relapse after the end of treatment in a dose-dependent manner [22], and ITPA genotype is associated with the outcome of treatment [12, 23]. It is possible that the difference of the effect of ITPA genotype on treatment response between PEG-IFN/RBV and sofosbuvir plus RBV therapies is due to the high SVR rate in sofosbuvir plus RBV treatment. Further analysis is needed to clarify the effect on treatment response of ITPA genotype during sofosbuvir plus RBV treatment using a larger number of patients.

The present study identified FIB4 index as an independent predictor for SVR12. SVR12 was significantly lower in patients with FIB4 index ≥3.25 than in those with FIB4 index <3.25 (89.8 and 98.5%, respectively; P = 0.003). Progression of liver fibrosis is strongly associated with IFN-based therapy [22, 24]. In contrast, the effect of liver fibrosis on IFN-free DAA treatment is controversial [25, 26]. Although treatment response was similar between patients with chronic hepatitis and cirrhosis (97.1 and 94.1%, respectively) in the phase 3 clinical trial of sofosbuvir plus RBV therapy for HCV genotype 2-infected patients in Japan [9], the SVR12 rate was significantly lower in patients with progression of liver fibrosis in the present real-world study. To clarify the effect of liver fibrosis on IFN-free treatment, further analysis including other DAA treatments is needed.

Low body weight was also identified as an independent predictor for SVR12. Although initial RBV dose was adjusted by body weight, all patients were administered 400 mg of per day of sofosbuvir irrespective of body weight. Since it was reported that serum RBV concentration was related to the treatment response for sofosbuvir plus RBV [27], it would be interesting to analyze the relationship between body weight and concentrations of sofosbuvir and RBV on response to treatment.

In the present study, only one patient (0.4%) discontinued the treatment due to an adverse event. Although serum HCV RNA became negative at four weeks in this patient, the treatment was discontinued at 6 weeks due to rash, and serum HCV RNA relapsed after cessation of the treatment. Although hemoglobin levels were reduced due to RBV, no patient needed to discontinue the therapy due to anemia. Therefore, the safety of sofosbuvir plus RBV therapy seems to be relatively high.

In conclusion, sofosbuvir plus RBV therapy is associated with an increased risk of anemia. Patients with ITPA CC genotype experienced significantly faster hemoglobin decline and were more vulnerable to anemia than patients with ITPA CA/AA genotype during the therapy. Although the ITPA CC genotype was related to low adherence to RBV, no significant effect on treatment response was observed. Patients with progression of liver fibrosis and greater body weight are more likely to fail to respond to the treatment.