Introduction

Hepatitis C virus (HCV) is one of the most common causes of chronic liver disease worldwide [18]. HCV genotype 1 is more abundant than other genotypes and HCV genotype 3 is the next most common type [16, 25]. In Iran, the most frequent subtypes of HCV have been shown to be 1a, 3a, and 1b [26].

A combination of pegIFN-α/RVB was used to treat Iranian patients with chronic hepatitis C virus infection for 24 or 48 weeks [17, 26]. On this regimen, patients infected with the HCV genotype 1 and 2/3 had SVR rates of approximately 50 % and 70 to 90 %, respectively [16]. Long-term treatment response is affected by many viral and host factors including age, sex, race, HCV genotype, fibrosis, HCV viral load and basal metabolic index (BMI) [8].

Different studies in CHC patients have revealed that three host genetic single nucleotide polymorphisms (SNPs) in rs12979860, rs8099917, and rs12980275 near the Interleukin-28B (IL28B) gene, located on chromosome 19, may be predictors of virologic response in patients with CHC virus infection treated with pegIFN-α/RVB [2, 4, 7, 13]. CHC patients with rs12979860 CC, rs8099917 TT, and rs12980275 AA genotypes were strongly associated with SVR after pegIFN-α/RVB combination therapy [2, 8, 23, 27].

A recent genome-wide association study (GWAS) and several reports reveal that the human leukocyte antigen (HLA)-DQB1*03:01 genotype frequencies are associated with spontaneous clearance of HCV infection and are potential predictors of HCV treatment efficacy [5].

The GWAS in African and European populations reported a significant association between the rs4273729 (located on chromosome 6 in a 100-kbp region consisting of HLA class II genes) and spontaneous clearance of HCV infection [10].

A study in China showed that HLA rs4273729 alleles can predict response to peg IFN-α/RBV therapy in patients with HCV genotype 1 [28]. Thus far, no studies about the association between the rs4273729 and treatment response in Iranian patients have been conducted. In this study, we were interested in whether the rs4273729 may predict treatment response in Iranian patients infected with HCV genotype 1 and 3.

The development of antivirals that act directly (DAAs) on HCV may diminish the importance of SNP genotyping in predicting SVR. Because HCV genotype 1 and 3 are the most prevalent viruses in Iranian patients, DAA therapy for the hepatitis C virus in Iran has not yet been approved; pegIFN-α/RVB combination therapy is the standard treatment. Therefore, IL28B (rs12979860, rs8099917 and rs12980275) genotyping and other SNPs may also predict how HCV responds to interferon [2, 11, 16, 17].

Several methods, such as direct DNA sequencing, real-time polymerase chain reaction (PCR), amplification refractory mutation system (ARMS)-PCR, PCR-restriction fragment length polymorphism (RFLP), and allele-specific PCR can be used for genotyping SNPs, such as IL28B and HLA alleles. The ARMS method is a reliable, simple, and rapid method for detection of any mutation involving single base changes or small deletions. This method is based on the use of sequence-specific PCR primers that allows amplification of test DNA only when the target allele is contained within the sample and will not amplify the nontarget allele [7].

This study has been designed to investigate how two IL28B SNPs (rs12979860 and rs12980275) and HLA rs4273729 correlate with rapid virologic response (RVR), complete early virologic response (cEVR), and sustained virologic response (SVR) to standard treatment in HCV-infected Iranian patients treated with combination therapy pegIFN-α/RVB.

Materials and methods

Study population

This study was performed at the Pasteur Institute of Iran (PII) from December 2013 to November 2015. The study was conducted according to the Declaration of Helsinki and relevant local regulations and was approved by the Ethical Committee of PII. Written informed consent was obtained from each patient.

In the current cross-sectional study a total of 190 patients with CHC infection were enrolled according to the following inclusion criteria: 1) desire to participate in this study, 2) of Iranian nationality, 3) being treatment-naive.

The exclusion criteria were other previous antiviral treatments for HCV infection, CHC patients with HIV or HBV, using immunosuppressive drugs, diabetes mellitus, hepatocellular carcinoma and liver cirrhosis.

All patients were treated with 180 mg/week of Peg-IFNα-2a and 1000-1200 mg/day of RBV as initial antiviral treatments. The patients with CHC genotypes 1 and 3 were treated for 48 and 24 weeks, respectively [6]. All the patients were HIV-Ab (ELISA Test, Bio-Rad, Hercules, California, USA) and hepatitis B surface antigen (HBsAg) negative (ELISA Test, Radim SPA, Pomezia-Rome, Italy) but anti-HCV (ELISA Test, Bio-Rad, Hercules, California, USA) and HCV RNA positive in serum.

In terms of treatment responses, RVR was defined as an undetectable HCV RNA at the fourth treatment session, complete EVR (undetectable HCV RNA in serum at the 12th treatment session), and SVR as the removal of the HCV RNA level at the 24th session (that is, treatment completion) [15].

Virologic testing

The Amplicor Monitor HCV 2.0 (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) was used to measure HCV RNA viral load in all subjects at baseline after the 4th, 12th, 24th and, 48th treatment sessions. The HCV genotype was determined using the AmpliSens® HCV-1/2/3-FEP PCR kit (InterLabService Ltd., Moscow, Russia) according to the manufacturer’s instructions, respectively.

HLA rs4273729 and IL28B genotyping

Peripheral blood mononuclear cells (PBMCs) from blood samples were isolated by density-gradient centrifugation using Ficoll (Ficoll-Paque PLUS, GE Healthcare) [34] and stored at -20 °C. Genomic DNA was extracted from 190 PBMC samples using a high pure PCR template preparation kit (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) according to the manufacturer’s instructions.

The genotyping of SNPs rs12980275 and rs12979860 was performed using the amplification refractory mutation system PCR (ARMS-PCR). The primers were designed by Primer Premier Software (ver 6.23 Primer, PREMIER Biosoft International, USA) (Table 1). IL28B amplification was conducted in a 25 µL reaction mixture containing 50 ng of genomic DNA, 10 pM of each primer (two forward and two reverse primers), 0.2 mM dNTP’s, 2× PCR buffer (Fermentas, Vilnius, Lithuania), Ex Taq DNA Polymerase (1.5 U) (Seoul, South Korea, Takara Korea Biomedical Inc), PCR buffer (10×) containing 1.5 mM MgCl2 and RNase-free water. The PCR conditions for rs12979860 were 95 °C for 10 min followed by 40 cycles of 95 °C for 45 s, 67 °C for 45 s, 72 °C for 45 s and 72 °C for 10 min. For rs12980275 the most appropriate PCR conditions were: 94 °C for 5 min followed by 30 cycles of 94 °C for 45 s, 53 °C for 45 s, 72 °C for 45 s and 72 °C for 10 min. The PCR products were detected by 2 % agarose gel electrophoresis.

Table 1 Primers used in the ARMS-PCR and PCR sequencing
Full size table

To confirm the ARMS-PCR results, the rs12979860 and rs12980275 SNPs were genotyped using the direct sequencing assay. The polymorphism of HLA rs4273729 was also determined by using the direct sequencing assay. The primers used for sequencing are shown in Table 1. PCR products were purified by High Pure PCR Product Purification Kit (Roche Diagnostics Deutschland GmbH, Mannheim, Germany) and were sequenced using an ABI Automated Sequencer (Applied Biosystems, Foster City, CA, USA). The raw sequencing data was analyzed using CLC Main Workbench software (CLC Bio, Aarhus, Denmark).

Statistical analysis

All calculations were performed using SPSS version 21 software (SPSS. Inc., Chicago, IL, USA). One-way ANOVA, the Pearson χ2 test, and the Kruskal–Wallis test were used to assess the differences of the categorical and continuous variables, respectively. Predictors for RVR, cEVR, and SVR were evaluated by univariate and multivariate logistic regression modeling analyses. The odds ratio (OR) with 95 % confidence interval (CI) was also calculated. Differences with a p-value of <0.05 were considered statistically significant. The linkage disequilibrium (LD) between rs12979860 and rs12980275 SNPs was calculated with DnaSP v5 (http://www.ub.edu/dnasp/). The Hardy-Weinberg equilibrium was evaluated using the DeFinetti program (http://ihg.gsf.de/cgi-bin/hw/hwa1.pl).

Results

Clinical and demographic characteristics

Of the 190 patients, 93 (48.9 %), 26 (13.7 %) and 71 (37.4 %) patients were infected with HCV 1a, 1b, and 3a genotypes, respectively. The mean age of patients was 44.6 ± 9.9 years (range 24-61) (Table 2). Of the 190 subjects, 101 (53.2 %), 150 (78.9 %), and 125 (65.8 %) achieved RVR, cEVR, and SVR, respectively. The response rates were higher in HCV subtype 3a in comparison to HCV subtype 1a and 1b (Table 3).

Table 2 Demographic and Baseline Clinical Characteristics of Iranian patients with chronic hepatitis C virus infection
Full size table
Table 3 Demographic features and factors associated with rapid virological response (RVR), complete early virological response (cEVR) and sustained virologic response (SVR) in patients with chronic HCV infection
Full size table

Frequency of IL28B and HLA genotypes

The frequency of IL28B rs1297960 (Fig. 1A), rs12980275 (Fig. 1B), and HLA rs4273729 (Fig. 1C) in CHC patients is shown in Figure 1. The genotypes rs1297960 CT, rs12980275 AA, and rs4273729 GG were most frequently observed in CHC patients. The allelic distributions of the rs12979860, rs12980275, and rs4273729 in CHC patients were consistent with Hardy-Weinberg equilibrium. The SNP, rs12979680, was in strong LD with rs12980275 (D’ = 0.886, r2 = 0.748).

Fig. 1
figure 1

Distribution of IL28B and HLA: (A) Percentage distribution of IL28B rs12979860 genotypes (B) Percentage distribution of IL28B rs12980275 genotypes (C) Percentage distribution of HLA rs4273729 genotypes

Full size image

Association between HLA & IL28B SNPs and HCV genotype and baseline viral load

The correlation between IL28B SNPs (rs12979860, rs12980275) and HLA rs4273729 and viral load and HCV genotype in CHC patients were evaluated. No significant association was observed between viral load and IL28B rs12980275 (p = 0.488); and HLA rs4273729 (p = 0.697) genotypes, but there was association between IL28B rs12979860 and viral load (p = 0.028). In the present study, only the HLA rs4273729 (p = 0.002) SNP was associated with HCV genotype while, there was no association between IL28B rs12980275 (p = 0.057) and rs12979860 (p = 0.637) genotypes.

Relationship between IL28B and HLA genotypes and treatment outcomes

The factors associated with treatment responses are represented in Table 3. RVR and SVR rates were significantly higher in patients with the rs12979860 CC genotype, compared to in those with CT and TT genotypes. In rs12980275 SNPs, the rates of RVR, cEVR, and SVR were significantly higher in patients with the AA genotype compared to those with the AG and GG allele. For patients with the HLA rs4273729 GG genotype, a higher RVR, cEVR, and SVR rate was observed in comparison to the CC and GC genotypes.

Of the 190 patients, 59 had rs12979860 CC, rs12980275 AA, and HLA rs4273729 GG. Patients with co-expression of the common homozygote CC genotype at rs12979860, AA genotype at rs1290275 and GG genotype at rs4273729 showed a better response to Peg-IFNα-2a and RBV treatments compared to other genotypes.

The positive predictive value of the combination of IL28B rs12979860 CC genotype, IL28B rs12980275 AA genotype and HLA rs4273729 GG genotype for achieving RVR, cEVR and SVR was 82.6 %, 88.7 % and 89.2 %, respectively.

Factors associated with RVR, cEVR, and SVR

We analyzed the factors associated with RVR, cEVR, and SVR using multivariate logistic regression analysis. Multivariate analysis revealed that patients with rs12980275 AA (OR 58.34, 95 % CI 6.45–527.39, p < 0.001), rs12979860 CC (OR 12.10, 95 % CI 1.22–119.20, p = 0.033), and HLA rs4273729 GG (OR 4.82, 95 % CI 2.87-8.10, p < 0.001) genotypes were significantly associated with a higher RVR rate and cEVR was associated with rs12980275 AA (OR 0.08, 95 % CI 0.01–0.41, p = 0.003) and HLA rs4273729 GG (OR 4.77, 95 % CI 2.67-8.53, p < 0.001) genotypes (Table 4).

Table 4 Baseline factors predictive of rapid virologic response, complete early virologic response and sustained virologic response by multivariate logistic regression analysis
Full size table

In patients with SVR, multivariate analysis showed that a young age (<40 years) (OR 44.30, 95 % CI 2.64–743.55, p = 0.008), HCV genotype (OR 0.04, 95 % CI 0.01–0.628, p = 0.032), IL28B rs12979860 CC (OR 10.35, 95 % CI 1.93–55.65, p < 0.001), rs12980275 AA (OR 6.90, 95 % CI 2.20–21.80, p < 0.001), HLA rs4273729 GG (OR 7.32, 95 % CI 4.01-13.33, p < 0.001), RVR (OR 8.46, 95 % CI 2.78–25.79, p < 0.001), and cEVR (OR 111.46, 95 % CI 1.84–67.40, p = 0.024) responses were significant predictors of SVR (Table 4).

Discussion

In this study, we evaluated the viral (HCV genotype and HCV viral load) and host (liver enzymes, age, gender, HLA SNP and IL28B SNPs) factors influencing the treatment outcome of pegIFN-α/RVB and sought to evaluate SNPs with RVR, cEVR and SVR in Iranian patients with chronic hepatitis C genotype 1 and 3 infections.

Several studies have shown that HLA class II genotypes play a key role in regulating immune response to HCV infection and are associated with spontaneous viral resolution [5, 28]. In the current study, the relationship between rs4273729 in HLA class II region and treatment outcomes in Iranian patients with CHC virus infection is evaluated.

Few studies have addressed the relationship between rs4273729 and treatment response. The results of the current report demonstrated that HLA rs4273729 GG genotype is associated with increased RVR, cEVR, and SVR in Iranian CHC patients. Xu et al (2016), demonstrated that the rs4273729 GG genotype achieved significant SVR and that the rate of SVR was lower in patients with the CC genotype [28], consistent with the current study. However, Franco et al (2012) showed that in patients with HCV/HIV co-infection, HLA rs4273729 genotypes did not significantly affect the chances of achieving a SVR. Co-infection might have complicated the treatment process [10]. The MHC region has several SNPs; however, we detected only HLA rs4273729 SNP in the class II region. This represents one of the limitations of this study.

Several studies have shown that patients with a high CD4+ T-cell count were the most likely to achieve a SVR. Therefore, HLA rs4273729 genotypes could affect CD4+ T-cell response in CHC patients [12, 20]. However, more studies must be done to evaluate the efficacy of HLA rs4273729.

In the present study, only IL28B rs12979860 genotypes were associated with higher levels of baseline HCV RNA, therefore these results are consistent with other studies [2, 21]. Also, only HLA rs4273729 genotypes were associated with HCV genotype. Inconsistent with other reports, there was no association between IL28B rs12980275 and rs12979860 genotypes [6].

We demonstrated higher SVR rates in patients that achieved RVR (80.0 %) and cEVR (90.4 %), compared to patients who did not achieve RVR (20.0 %) and non-cEVR (9.6 %). This finding is consistent with previous reports of higher SVR in HCV patients who achieved RVR and cEVR [2, 4, 19, 23].

Several studies have reported that rs12979860 and rs12980275 genotypes are related to virologic response. In this study, the frequency of RVR was higher in patients with rs12979860 CC (59.4 %) than in those with rs12979860 CT (38.6 %) and rs12979860 TT (2.0 %) genotypes. Other reports indicated a higher antiviral response at the second and fourth treatment [2, 9, 29]. Furthermore, RVR was higher in patients with the rs12980275 AA (90.1 %) genotype. This finding is consistent with Covolo et al (2014), who demonstrated a high rate of RVR (55.0 %) in patients with AA genotype, compared to non-AA genotypes patients (28.0 %) [3].

Only the AA genotype of rs12980275 was predictive of cEVR. Interestingly, we did not find a significant association between cEVR and rs12979860 CC genotype patients, which contradicts other findings [4, 22].

We demonstrated that patients with the rs12979860 CC (52.0 %) genotype respond better to treatment. Sarrazin et al (2011), showed that 45.9 % of patients with the rs12979860 CC genotype achieve SVR. This may therefore be applied as a prognostic factor for the virologic response. Other studies confirm the association between rs12979860 CC genotype and SVR rate, indicating that a lower frequency of rs12979860 CC genotype may be influenced by a lower rate of SVR [21].

We also illustrated that rs12980275 AA was a remarkable predictor for SVR in Iranian patients. In this study, after deleting the data about IL28B rs12979860, rs12980275 powerfully predicted SVR. RVR and cEVR responses are considered important predictors of SVR. Yu et al (2007), reported that RVR and cEVR are independent predictive factors for HCV treatment [29]. We also demonstrated that RVR and cEVR were independent predictor factors of SVR.

Of the 190 patients, 59 had rs12979860 CC, rs12980275 AA, and HLA rs4273729 GG. Co-expression of the rs12979860 CC, rs1290275 AA, and rs4273729 GG was assessed. Patients with three favorable alleles exhibited a better chance of achieving SVR compared to other genotypes. Again, these results confirmed the remarkable impact that IL28B SNPs and HLA SNP have on treatment response. Shaikh et al (2015), reported that rs1290275 AA and rs12979860 CT predict a better chance of achieving SVR [23, 28].

In the current study, IL28B rs12980275 (AA genotype), which was in strong LD with rs12979860, plays an important role in the prediction of RVR, cEVR and SVR in CHC patients with pegIFN-α/RVB therapy. Our results are also in line with several larger studies [32, 33].

Other baseline predictors of SVR include the HCV genotype, pretreatment HCV viral load, gender, age, race, BMI, lowest baseline ALT levels, insulin resistance, and presence of fibrosis or cirrhosis on liver biopsy [14, 15, 24, 31]. In the present study, young age, HCV genotype, IL28B SNPs, and HLA rs4273729 predicted SVR. These results are consistent with the idea that IL28B SNPs are related to the immune response following pegIFN-α/RVB therapy [24, 28, 31]. A strong association has been reported between IL28B SNPs and SVR; the mechanism, however, remains poorly elucidated. Recent studies demonstrated that IL28B cytokine inhibits HCV replication through the JAK-STAT pathway [30]. As a result, the IL28B genotype appears related to differential expression of interferon stimulated genes (ISG) in CHC patients.

The strong relationship between IL28B SNPs, HLA rs4273729, and treatment response in CHC virus infection demonstrates that fully utilizing routinely collected information might aid prognosis and treatment protocol adjustment.

IL28B SNPs and HLA rs4273729 genotyping should not be the only factor considered in deciding on a treatment plan [1, 15]. Moreover, shortening HCV treatment time according to clinical parameters such as ALT level, HCV genotype, and HCV viral load in combination with these IL28B SNPs is possible, but there is inadequate information to make recommendations [1]. Therefore, IL28B (particularly rs12980275) and HLA genotyping is strong enough to predict RVR, cEVR, and SVR response.

Although we demonstrated that IL28B SNPs and HLA rs4273729 are associated with treatment outcomes in Iranian CHC patients, we noted the study limitations, including lack of subjects with naturally cleared HCV, small sample size, and a lack of information related to cirrhosis and fibrosis.

In conclusion, this study confirms for the first time that rs12980275 and HLA rs4273729 polymorphisms are an independent predictor of RVR, cEVR, and SVR in the Iranian population. The favorable alleles of rs12980275, rs12979860, and HLA rs4273729 were valuable in predicting RVR and SVR in patients treated with pegIFN-α/RVB, while rs12980275 and HLA rs4273729 in patients with cEVR was considered a predictive factor. Furthermore, we recommended genotyping for rs12979860, rs12980275, and HLA rs4273729 before pegIFN-α/RVB therapy, and also determination of RVR and cEVR responses during treatment as appropriate methods for patients with CHC.