Role of IL28B for chronic hepatitis C treatment toward personalized medicine

Authors

  • Kentaro Matsuura,

    1. Department of Virology, Liver Unit, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan
    2. Department of Gastroenterology and Metabolism, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan
    3. Infectious Disease and Immunogenetics Section, Department of Transfusion Medicine, Clinical Center, National Institutes of Health, Bethesda, Maryland, United States
    Search for more papers by this author
  • Tsunamasa Watanabe,

    1. Department of Virology, Liver Unit, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan
    Search for more papers by this author
  • Yasuhito Tanaka

    Corresponding author
    1. Department of Virology, Liver Unit, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan
    • Correspondence

      Professor Yasuhito Tanaka, Department of Virology, Liver Unit, Nagoya City University Graduate School of Medical Sciences, Kawasumi, Mizuho, Nagoya 467-8601, Japan. Email: ytanaka@med.nagoya-cu.ac.jp

    Search for more papers by this author

  • Conflicts of interests: Yasuhito Tanaka has research contracts with MSD (Merck Sharp & Dohme) and Chugai Pharmaceutical Co. Ltd. The remaining authors have no conflict of interests.
  • Financial support: This work was supported in part by a grant-in-aid from the Ministry of Health, Labour and Welfare of Japan, and a grant-in-aid from the Ministry of Education, Culture, Sports, Science and Technology.

Abstract

Genome-wide association studies recently revealed that certain interleukin-28B (IL28B) polymorphisms are strongly associated with responses to pegylated interferon (PEG-IFN) and ribavirin (RBV) therapy in patients chronically infected with hepatitis C virus (HCV) genotype 1, as well as with spontaneous clearance of HCV. Subsequent reports revealed that IL28B genotypes also affect treatment efficacy in chronic infection with other HCV genotypes. Furthermore, there have been several reports that implicate IL28B genotypes in inflammatory status, progression of fibrosis and adverse clinical outcomes in chronic hepatitis C (CHC). Therapy of CHC recently entered a new era with the deployment of direct-acting antivirals. These include nonstructural 3/4A protease inhibitors which have shown promise in combination with PEG-IFN/RBV in several clinical trials. IFN-free therapy is expected to be useful especially in IFN-resistant patients and may become the standard of care in the future. Several clinical trials have revealed an association between IL28B genotype and treatment efficacy in triple therapy or IFN-free regimens. On the other hand the mechanism of the effect of IL28B on HCV infection has not yet been elucidated. Recently, it was shown that the polymorphism of IFN-lambda 4 (IFNL4) is in high linkage disequilibrium with that of near IL28B, and more strongly associated with spontaneous or treatment-induced HCV clearance than IL28B genotypes, especially in individuals of African ancestry. This finding provides new insights into the genetic regulation of HCV clearance and its clinical management. IL28B genotyping will be also useful for personalized CHC treatment in the forthcoming era of direct-acting antivirals.

Introduction

Chronic hepatitis C virus (HCV) infection represents a significant health problem worldwide with approximately 170 million people infected.[1] Over 70% of individuals acutely infected with HCV go on to develop chronic infection and are at significant risk of progressive liver fibrosis and subsequent liver cirrhosis and hepatocellular carcinoma (HCC). Antiviral treatment has been shown to improve liver histology and decrease the incidence of HCC in chronic hepatitis C (CHC).[2, 3] Until 2011, the standard treatment for chronic HCV infection was weekly pegylated interferon (PEG-IFN) in combination with daily doses of ribavirin (RBV); however, less than 50% of patients infected with HCV genotype 1 treated in this way achieve a sustained virological response (SVR).[4, 5] In 2009, genome-wide association studies (GWAS), including our study of HCV infection,[6] showed that a single nucleotide polymorphism (SNP) near the interleukin-28B (IL28B) gene is strongly associated with response to PEG-IFN/RBV therapy for chronic HCV genotype 1 infection.[6-11] As a result, prediction of treatment outcome, especially nonresponsiveness to PEG-IFN/RBV, has been greatly improved by genotyping for the IL28B SNP, enabling personalized medicine to be developed for CHC. Newly developed treatments involving direct-acting antivirals (DAAs), including nonstructural (NS) 3/4A protease inhibitors have shown promising outcomes in combination with PEG-IFN/RBV in several clinical trials, wherein > 70% of patients infected with HCV genotype 1 achieved SVR.[12-14] Several clinical trials have revealed an association between IL28B genotype and treatment efficacy in triple therapy or IFN-free regimens. This review focuses on the role of IL28B in CHC treatment.

Predictors of response to IFN-based therapy

Various viral and host factors have been identified as significant determinants of the outcome of IFN-based treatments. Viral genotype and baseline viral load are well-known predictors of response to therapy. Other viral factors include amino acid substitutions at positions 70 and 91 in the HCV core region[15] and in the IFN sensitivity-determining region in NS5A[16] in patients infected with HCV genotype 1. Several host factors related to failure of treatment-induced viral clearance include older age, insulin resistance, advanced fibrosis and hepatic steatosis.[17, 18] Ethnicity is also a factor in treatment outcome. The proportion of African American patients achieving SVR on treatment with PEG-IFN/RBV is lower than Caucasian patients,[19-21] indicating that host genetic factors can be an important determinant of treatment outcome.

Analysis of candidate genes has revealed an association between several host genes and spontaneous or treatment-induced clearance of HCV. These include type I IFN receptor-1 (IFNAR1), mitogen-activated protein kinase-activated protein kinase 3 (MAPKAPK3),[22, 23] killer cell immunoglobulin-like receptor (KIR2DL3) and its human leukocyte antigen C group1 (HLA-C1) ligand,[24] and HLA type.[25, 26]

The association between IL28B and response to PEG-IFN/RBV

The success of the Human Genome Project accelerated studies on genetic factors involved in different outcomes of HCV infection. Significant breakthroughs in identifying phenotype-associated SNPs followed when the GWAS approach was established. Compared with the traditional gene candidate approach, GWAS can identify functionally important polymorphisms in genes that have no predicted role in disease pathogenesis. In 2009, four independent groups simultaneously published the results of GWAS to assess the role of genetic variation in response to PEG-IFN/RBV for CHC patients.[6-8, 27] All four revealed a strong association between genetic polymorphism near the IL28B locus on chromosome 19 and treatment-induced HCV clearance (Table 1). Ge et al. and Suppiah et al. studied genetic variants associated with SVR on treatment with PEG-IFN/RBV in individuals infected with HCV genotype 1.[7, 8] Ge et al. studied patients from the IDEAL trial,[17] a large randomized, controlled trial involving Caucasians, African Americans, and Hispanics in North America (n = 1137). The CC genotype at rs12979860 showed a twofold greater rate of achievement of SVR in Europeans and Hispanics, and a threefold higher rate of SVR in African Americans relative to non-CC genotype. Suppiah et al. analyzed Caucasians consisting of 293 Australian individuals infected with HCV genotype 1 and also validated their findings in an independent replication cohort consisting of 555 Europeans from the UK, Germany, Italy, and Australia. They showed that rs8099917 was the polymorphism most strongly associated with SVR. Tanaka et al. studied host factors associated with null virological response (NVR) on treatment with PEG-IFN/RBV in 142 Japanese CHC patients infected with HCV genotype 1, and an independent replication cohort of another 172 Japanese. They found that rs8099917 showed the most significant associations (P = 2.68 × 10−32, odds ratio [OR] = 27.1).[6] Rauch et al. investigated 465 Caucasians infected with HCV genotypes 1, 2, 3, or 4.[27] Strong predictive value of the IL28B polymorphism was observed in genotype 1 and 4 patients, but not in genotypes 2 and 3 infection. The earlier studies document that rs12979860 or rs8099917 are the polymorphisms most significantly associated with response to therapy. These SNPs are in strong linkage disequilibrium except in patients of African ancestry; they are in partial linkage disequilibrium in Caucasian,[7, 27] but in near-complete linkage disequilibrium in East Asian.

Table 1. GWAS of associations between favorable IL28B genotype and SVR in PEG-IFN/RBV therapy
StudyRef.PopulationGWAS/replication sizeSignificant SNPOdds ratioP valueHCV genotype
  1. Favorable IL28B genotype means rs8099917 TT or rs12979860 CC genotype.
  2. GWAS, genome-wide association study; IL28B, interleukin-28B; PEG-IFN, pegylated interferon; RBV, ribavirin; Ref., reference number; SNP, single nucleotide polymorphism; SVR, sustained virological response.
Ge et al.[7]European American, African American, Hispanic1137/noners129798603.101.37 × 10−281
Suppiah et al.[8]Australian, European293/555rs80999171.987.06 × 10−81
Tanaka et al.[6]Japanese142/172rs809991712.101.18 × 10−181
Rauch et al.[27]European465/noners80999175.193.11 × 10−81–4

An association between race and spontaneous HCV clearance has been reported.[28-30] Moreover, significant differences between ethnicities in response to PEG-IFN/RBV therapy were reported; the SVR rate was approximately 20–28% in African Americans and 40–52% in Caucasian patients with HCV genotype 1,[19-21] and 57% versus 82% in those with genotype 2/3.[31] The frequency of the IL28B genotype favorable to treatment varies by ethnicity, being > 80% in certain Asian populations, 35–55% in Caucasians and < 20% in patients of African ancestry. This variation explains, in part, the inferior response rates in African Americans as compared with Caucasians and the increased response rates in Asians as compared with Caucasians.[7, 9] However, it has been reported that IL28B genotype and ethnic background were independent pretreatment predictors for SVR in the IDEAL study:[32] IL28B genotype (CC vs non-CC at rs12979860: OR = 5.2, P < 0.0001) and ethnic background (Caucasian vs African American: OR = 2.8, P < 0.0001; Hispanic vs African American: OR = 2.1, P = 0.0041). Therefore, IL28B polymorphisms did not account for all of the ethnic differences in response to treatment.

Following the earlier mentioned GWAS, many studies have confirmed the impact of IL28B on response to treatment. Thompson et al. reported that the IL28B genotype also affected early viral kinetics during PEG-IFN/RBV therapy in patients infected with HCV genotype 1. Patients with a favorable IL28B genotype achieved a higher rate of rapid virological response (RVR). Even if they did not achieve RVR, a favorable IL28B genotype was also strongly associated with SVR. In contrast, the IL28B genotype was not associated with SVR in patients who experienced RVR.[32] These findings indicate that the IL28B genotype is useful as an on-treatment predictor of SVR in patients not experiencing RVR. In the IDEAL study cohort, SVR rates in patients with advanced liver fibrosis (METAVIR F3-4) were considerably lower, namely 41% for patients with CC, 22% for CT, and only 11% for TT at rs12979860.[32] Thus, liver fibrosis is also an important predictive factor of treatment efficacy in addition to the IL28B genotype. The IL28B genotype is also associated with the outcome of PEG-IFN/RBV therapy for patients co-infected with HCV genotype 1 and human immunodeficiency virus (HIV) as well as in HCV monoinfected patients.[33] In patients who underwent liver transplantation, IL28B genotypes of both donor and recipient were associated with treatment efficacy.[34, 35]

IL28B and treatment efficacy in patients with HCV genotype non-1

We summarized previous reports on the effect of IL28B genotype on treatment efficacy in patients infected with HCV genotype non-1 (Table 2). Rauch et al. reported that there were no significant associations between IL28B genotype and response to PEG-IFN/RBV in patients infected with HCV genotype 2 or 3 in their GWAS study (OR = 1.58; P = 0.18).[27] Mangia et al. noted that IL28B genotype was associated with SVR in patients with genotype 2 or 3 especially in those who did not experience RVR in PEG-IFN/RBV for 24 weeks: SVR rates were 87%, 67%, and 29% in patients with CC, CT, and TT at rs12979860, respectively (P = 0.0002).[36] Sakamoto et al. examined the relationship between IL28B genotype and response to therapy in Japanese patients infected with HCV genotype 2 who were treated with PEG-IFN/RBV for also 24 weeks. They showed that patients infected with genotype 2b had significantly lower RVR rates than those infected with genotype 2a. Moreover, both RVR and SVR were significantly associated with a favorable IL28B genotype in patients infected with genotype HCV 2b.[37] Other investigators showed that a favorable IL28B genotype was associated with RVR but not SVR in patients infected with HCV genotype 2 or 3.[38, 39] Taken together, these data suggest that the effect of IL28B genotype on SVR is weaker in patients infected with genotype 2 or 3 than genotype 1. With regard to HCV genotype 4, the IL28B genotype correlates with response to PEG-IFN/RBV therapy as well as it does for genotype 1.[27, 40-42, 45] There are very few reports on associations in patients infected with HCV genotype 5 or 6. Antaki et al. reported that the IL28B genotype did not predict response to treatment in a small study of patients infected with HCV genotype 5 (n = 49).[43] Seto et al. noted that the SVR rate was higher in patients with a favorable IL28B genotype than in those with an unfavorable genotype (96.2% vs 62.5%, P = 0.014) in their analysis of a total of 60 patients infected with HCV genotype 6.[44]

Table 2. Associations between favorable IL28B genotype and SVR in PEG-IFN/RBV-treated patients infected with HCV genotype non-1
StudyRef.PopulationSNPNumber of patientsOdds ratioP valueHCV genotype
  1. Favorable IL28B genotype means rs8099917 TT or rs12979860 CC genotype.
  2. IL28B, interleukin-28B; N.A., not available; N.S., not significant; PEG-IFN, pegylated interferon; RBV, ribavirin; Ref., reference number; SNP, single nucleotide polymorphism; SVR, sustained virological response.
Rauch et al.[27]Europeanrs80999172301.580.182, 3
Mangia et al.[36]Europeanrs129798602681.800.00462, 3
Sakamoto et al.[37]Japanesers80999171293.010.0132
Yu et al.[38]Taiwanrs80999174821.370.502
Moghaddam et al.[39]North Europeanrs80999172810.91N.S.3
Asselah et al.[40]Egyptian, European, Sub-Saharan Africanrs12979860823.320.00084
Antaki et al.[41]Syriars80999171824.17< 0.00014
Abdo et al.[42]Saudi Arabiars129798601291.50.0084
Antaki et al.[43]Syriars8099917491.040.95
Seto et al.[44]Hong Kongrs809991760N.A.0.0146

IL28B and spontaneous clearance of HCV

Spontaneous clearance of HCV occurs in approximately 20–30% of patients following acute infection. Host factors have been suggested to have a significant role in HCV clearance or persistence.[29, 46, 47] Data are accumulating regarding the significance of IL28B polymorphisms not only in response to therapy but also in spontaneous clearance of acute HCV infection (Table 3). GWAS on spontaneous clearance of HCV has been carried out by Rauch et al.[27] A case–control study was designed for 347 individuals with spontaneous HCV clearance, 567 with CHC, and 448 with HCV/HIV co-infection. The significant SNP was also found to be rs8099917 (combined P = 6.07 × 10−9, OR = 2.31) in this study. The effect on HIV co-infection was similar to that of HCV monoinfection (P = 8.25 × 10−5, OR = 2.16; P = 1.96 × 10−5, OR = 2.49, respectively). Recently, another group reported the results of GWAS on spontaneous resolution of HCV infection in a larger number of patients (919 persons with spontaneous clearance and 1482 with persistent infection) from multiple cohorts. They showed that IL28B (rs12979860, OR = 0.45, P = 2.17 × 10−30) and HLA class II (rs4273729, OR = 0.59, P = 1.71 × 10−16) were independently associated with spontaneous resolution of HCV infection.[48] Thomas et al. performed a candidate gene study to determine whether rs12979860 is also associated with spontaneous clearance of HCV infection.[9] That study included 388 individuals with spontaneous HCV clearance and 620 with persistent HCV infection in a cohort consisting of HCV and HIV/HCV co-infected patients. A strong association of rs12979860 with spontaneous recovery was found in both European and African American individuals (OR = 2.6 and 3.1, respect ively). Grebely et al. reported that rs8099917 TT was a factor that independently predicted spontaneous clearance in an Australian population (OR = 3.78, P = 0.044).[49] Moreover, they showed that participants who had jaundice and resulted in spontaneous clearance were more frequently in patients with rs8099917 TT than with non-TT genotypes (32% vs 5%, P = 0.047). This suggests a stronger immune response during the acute phase of HCV infection among patients with the rs8099917 TT genotype, resulting in a higher frequency of spontaneous clearance. However, IL28B genotypes did not affect the response to treatment during recent HCV infection. Tillmann et al. also reported that spontaneous viral clearance and jaundice during acute HCV infection was more common in patients with a favorable IL28B genotype.[50] Recently, an analysis of nine prospective international cohorts evaluating outcomes following acute HCV infection reported that spontaneous clearance occurred in 173 (25%) of 632 acute HCV infections during 1 year follow-up and that female gender, favorable IL28B genotype and HCV genotype 1 were independent predictors thereof.[51] In addition, for individuals with spontaneous clearance, the median time to clearance was 16.5 weeks, with two-thirds clearing within the first 6 months of infection. These findings provide guidance in clinical decision-making for the treatment of acute HCV infection. With regard to treatment strategy for acute HCV infection in consideration of IL28B genotype, Grebely et al.[52] and Mangia et al.[53] recommended early therapeutic intervention in non-jaundiced patients with an unfavorable IL28B genotype because of their low likelihood of spontaneous HCV clearance.

Table 3. Associations between favorable IL28B genotype and spontaneous clearance of HCV
StudyRef.PopulationSNPNumber of patientsOdds ratioP valueHCV genotype
  1. Favorable IL28B genotype means rs8099917 TT or rs12979860 CC genotype.
  2. GWAS, genome-wide association study; HCV, hepatitis C virus; IL28B, interleukin-28B; N.A., not available; Ref., reference number.
GWAS
Rauch et al.[27]Europeanrs809991713622.316.07 × 10−91–4
Duggal et al.[48]European, African, othersrs1297986014822.222.17 × 10−30N.A.
Candidate gene study
Thomas et al.[9]European, Africanrs1297986010083.03< 1.00 × 10−121
Grebely et al.[49]Australianrs80999171633.784.4 × 10−21–4
Tillmann et al.[50]German womenrs12979860190N.A.< 1.00 × 10−31b
Grebely et al.[51]Caucasian, Aboriginal, Asian, Blackrs129798606322.26< 1.00 × 10−31–4, 6

IL28B genotype and disease progression

Fabris et al. reported that patients with an unfavorable IL28B genotype were at increased risk of severe liver fibrosis.[54] In contrast, a favorable IL28B genotype has been shown to be associated with higher inflammatory activity and progression of fibrosis in several reports. Abe et al. analyzed the effect of IL28B genotype on histological findings in 364 Japanese CHC patients. Inflammation was more active and fibrotic progression was more severe in patients with a favorable IL28B genotype.[55] Barreiro et al. analyzed the impact of IL28B genotype on the risk of developing cirrhosis in HIV/HCV co-infected patients receiving antiretroviral therapy. In patients with a favorable IL28B genotype, cirrhosis was more frequent and mean alanine aminotranferase[56] level was higher than in patients with unfavorable IL28B genotypes, suggesting that favorable IL28B carriers may experience a more rapid progression of HCV-related liver fibrosis as a result of increased liver inflammation.[57] Bochud et al. also reported data consistent with this notion especially in patients monoinfected with HCV genotype non-1.[58] However, Marabita et al. reported that the IL28B genotype was not associated with progression of fibrosis in patients whose dates of infection were known.[59] Recently, Noureddin et al. examined the effect of IL28B genotype on fibrotic progression and clinical outcomes in large cohorts. In their baseline cross-sectional analysis of 1483 individuals, patients with CC at rs12979860 had significantly higher portal inflammation and ALT levels (P < 0.05) at baseline liver biopsy. However, in the paired liver biopsy analysis (median time between biopsies, 4 years), there was no difference in the frequency of fibrotic progression between CC and non-CC genotypes in 276 individuals. In addition, they showed that patients with the CC genotype were twice as likely to develop adverse clinical outcomes than non-CC genotypes (32% vs 16%, P = 0.007).[56] On the other hand, the impact of IL28B genotype on hepatocarcinogenesis is controversial.[54, 58] Fabris et al. showed that carriage of the T allele at rs12979860 was associated with an increased risk of developing HCC.[54] In contrast, Akuta et al. reported that IL28B genotype did not influence hepatocarcinogenesis over a long-term follow-up period in 515 patients who had not received antiviral therapy.[60] Other investigators have also failed to find any association between IL28B genotype and the development of HCC.[61, 62] Recently, Asahina et al. showed the association between IL28B genotype and HCC risk in a large-scale (n = 792), long-term cohort of IFN-treated patients, indicating that rs8099917 non-TT is significantly associated with HCC development particularly in patients infected with HCV genotype 1 who were treated with PEG-IFN/RBV combination therapy. Interestingly, they also demonstrated that a decrease in ALT and α-fetoprotein levels after IFN therapy is less in non-TT patients among non-SVR, resulting in a higher incidence of HCC.[63]

IL28B genotype and response to regimens including DAA

The HCV genome is translated into one polyprotein that is subsequently cleaved by viral and cellular proteases and processed into 10 structural and non-structural proteins. DAA therapies directly inhibit specific steps in the HCV viral life cycle, with targets including NS3/4A protease, NS5B polymerase, and NS5A phosphoprotein that are essential for viral replication. To date, the first-generation protease inhibitors, telaprevir and boceprevir, have been approved and various clinical trials of new DAAs are ongoing.

In treatment-naïve patients, the SPRINT-2[64] and ADVANCE trials[14] for boceprevir and telaprevir, respectively, showed that the IL28B SNP: rs12979860 affected treatment outcome. The SVR rates in SPRINT-2 and ADVANCE were higher in patients with CC (80%, 90%) compared with CT (71%, 71%) or TT (59%, 73%) (Table 4).[13, 66] On the other hand, in pretreated patients, the RESPOND-2[65] and REALIZE[12] trials for boceprevir and telaprevir, respectively, showed that the previous response to PEG-IFN/RBV strongly affected SVR; thus the SVR rate increased from null response to partial response and then relapse to previous therapy.[13, 67] The IL28B genotype was not significantly associated with SVR in those who relapsed or in partial responders, whereas the SVR rate tended to be higher in prior null responders with a favorable IL28B genotype than in those with unfavorable genotypes.[71] Bota et al. performed a meta-analysis and discerned a role for IL28B polymorphisms as predictors of SVR in patients treated with triple therapy. They selected five studies (1641 cases) of which the regimens of four were telaprevir/PEG-IFN/RBV, and the 5th was boceprevir/PEG-IFN/RBV. The SVR rate was significantly higher in patients with CC at rs12979860 than in those with non-CC (OR = 3.92, P < 0.0001). Moreover, higher SVR rates were seen in patients with CC regardless of therapeutic status (treatment-naïve patients: OR = 3.99, P < 0.0001; treatment-experienced patients: OR = 2.15, P = 0.001).[72]

Table 4. Associations between IL28B genotype and SVR in DAA plus PEG-IFN/RBV therapy
DrugRef.Study populationTreatment armrs12979860 genotype
CCCTTT
  1. BOC, boceprevir; DAA, direct-acting antiviral; FDV, Faldaprevir; IL28B, interleukin-28B; PEG-IFN, pegylated interferon; RBV, ribavirin; Ref., reference number; SMV, simeprevir; SVR, sustained virological response; TVR, telaprevir.
Boceprevir[13, 64]SPRINT-2 (treatment-naïve)BOC PR4844/55 (80%)82/115 (71%)26/44 (59%)
[13, 65]RESPOND-2 (treatment-experienced)BOC PR4817/22 (77%)48/66 (73%)13/18 (72%)
Telaprevir[14, 66]ADVANCE (treatment-naïve)TVR1245/50 (90%)48/68 (71%)16/22 (73%)
[12, 67]REALIZE (treatment-experienced): overallAll TVR60/76 (79%)160/266 (60%)49/80 (61%)
prior relapseAll TVR51/58 (88%)100/117 (85%)29/34 (85%)
prior partial responseAll TVR5/8 (63%)33/57 (58%)10/14 (71%)
prior null responseAll TVR4/10 (40%)27/92 (29%)10/32 (31%)
[68]Japanese (treatment-naïve and experienced)All TVR31/42 (84%)CT+TT 10/34 (32.3%)
Simeprevir (TMC-435)[69]PILLAR (treatment-naïve)SMV 75 mg26/31 (84%)50/64 (78%)7/14 (50%)
SMV 150 mg34/35 (97%)48/60 (80%)8/12 (67%)
Faldaprevir (BI-201335)[70]SILEN-C1 (treatment-naïve)FDV 240 mg22/22 (100%)CT+TT 34/48 (71%)

In addition to IL28B genotype, several factors influencing responses to triple therapy have been identified. The REALIZE study showed that the severity of liver fibrosis was a predictive factor for SVR in telaprevir/PEG-IFN/RBV therapy: the SVR rate was 74% in those with F0-F2 fibrosis, 66% in those with F3, and 47% in those with F4.[12] Akuta et al. showed that the SVR rate was 84% irrespective of substitution of core aa70 in patients with TT at rs8099917, whereas in those with non-TT, the SVR rate was 50% for patients with the wild-type core aa70 and 12% in those with non-wild type.[68] Combining these factors with IL28B genotyping might improve the prediction of responsiveness to triple therapy.

Thus far, several reports have appeared on the effects of the IL28B genotype on treatment efficacy of next-generation DAA plus PEG-IFN/RBV therapy (Table 4). The PILLAR trial investigated the efficacy of two different doses of simeprevir together with PEG-IFN/RBV in treatment-naïve patients infected with HCV genotype 1. The SVR rate with simeprevir at 75 mg was 83.9%, 78.1%, and 50.0%, and with 150 mg 97.1%, 80%, and 66.7% in patients with CC, CT, and TT at rs12979860, respectively. Viral breakthrough was seen exclusively in the non-CC genotype.[69] The SILEN-C1 trial investigated efficacy of faldaprevir combined with PEG-IFN/RBV in treatment-naïve patients infected with HCV genotype 1. In the subgroup treated with once-daily faldaprevir at 240 mg and PEG-IFN/RBV, the SVR rate was 100% (22/22) in patients with CC at rs12979860 and 71% (34/48) in non-CC.[70] On the other hand, in patients who had failed previous PEG-IFN/RBV, a phase 2b study of vaniprevir achieved SVR rates that were not significantly different regardless of IL28B genotype.[73] Thus, next-generation DAA plus PEG-IFN/RBV therapy will likely weaken the effect of IL28B polymorphism. However, the IL28B genotype will remain relevant to treatment efficacy especially in treatment-naïve patients.

Furthermore, Lok et al. demonstrated that the combination of daclatasvir (NS5A inhibitor) and asunaprevir (NS3/4A protease inhibitor) with PEG-IFN/RBV was effective for patients infected with HCV genotype 1 who had had null response to prior PEG-IFN/RBV therapy: the SVR rate was 90%.[74] In robust treatment such as this quadruple therapy, the IL28B genotype might indeed not be associated with treatment outcome.

IFN-free therapy is expected to become the standard of care in future and is clearly required especially in IFN-resistant patients. Chayama et al. demonstrated that 9 of 10 patients infected with HCV genotype 1b who had failed to respond to prior PEG-IFN/RBV therapy experienced SVR on an IFN-free regimen containing daclatasvir (NA5A inhibitor) and asunaprevir (NS3/4A protease inhibitor).[75] This suggests that combination therapy with potent DAAs might obscure the influence of IL28B polymorphisms on treatment efficacy. However, it has been reported that IL28B polymorphisms may affect viral kinetics even in the context of IFN-free regimens in the case of a combination of mericitabine (NS5B polymerase inhibitor) and danoprevir (NS3/4A protease inhibitor).[76] Moreover, in a phase 2b, randomized, open-label trial of faldaprevir (NS3/4A protease inhibitor) and deleobuvir (NS5B polymerase inhibitor), the SVR rates tended to be higher in patients with CC at rs12979860 than in those with non-CC.[77] This suggests that innate immunity may still be important and IL28B genotype may affect treatment efficacy in certain IFN-free regimens. Larger cohort sizes will be required to confirm such associations.

Association between IFN-λ and HCV infection

IL28B encodes IFN-λ3, which belongs to the type III IFN-λ family consisting of IL29/IFN-λ1, IL28A/IFN-λ2, and IL28B. Signaling by IFN-λ is initiated through a membrane receptor distinct from receptors for type I IFNs composed of heterodimers of an IL28RA/IFN-λR subunit and an IL10R2 subunit.[78, 79] Type I and III IFNs induce transcription of IFN-stimulated genes (ISGs) by activating the Janus kinase-signal transducer and activator of transcription pathway through different cell surface receptors[78, 79] in order to mediate their potent antiviral effects. There have been several reports about the profile of ISG expression in liver or peripheral blood mononuclear cells (PBMCs) so far. It has been reported that high-level expression of intrahepatic ISGs affected poor response to PEG-IFN/RBV therapy.[80, 81] Moreover, recent studies have revealed an association between IL28B genotype and expression levels of intrahepatic ISGs.[82, 83] In addition, the innate immune system: Toll-like receptor 3 and retinoic acid-inducible gene I signaling pathways of IFN-β induction has an essential role in host antiviral defense against HCV infection. Asahina et al. showed that the intrahepatic genes expressions involving innate immunity were strongly associated with IL28B genotype and response to PEG-IFN/RBV.[84, 85] With regard to IL28 expression in PBMCs, Suppiah et al. and we have shown to be higher in patients with a favorable IL28B genotype.[6, 8] Asahina et al. showed that the induction of several ISGs in PBMCs after the initial administration of PEG-IFN/RBV tended to be stronger in SVR than in NVR group, but the difference was not statistically significant.[84] Similarly, most of other investigators have indicated less marked association between the expression of ISGs in PBMCs and treatment outcomes, or IL28B genotype in comparison with in liver of the same patients.[80, 86] Thus, although there are several reports about the association between ISGs in liver or PBMCs and IL28B genotype or response to IFN therapy, the biological pathways linking IL28B genetic variants to spontaneous and/or treatment-induced HCV clearance remain unknown. However, recent reports suggest some possible scenarios. Using primary human hepatocytes or chimpanzee, Thomas et al. found that type III but not type I IFNs are primarily induced after HCV infection and that their degree of induction is closely correlated with the levels of ISGs.[87] These results strongly suggest that hepatic IFN-λ production may have important roles and could be a principal driver of ISG induction in response to HCV infection. On the other hand, in chronically HCV-infected chimeric mouse model that have the characteristic of immunodeficiency, larger amounts of IFN-λs on HCV-infected human hepatocytes were produced in liver with a favorable IL28B genotype on treatment with IFN-α.[88] However, no significant differences in HCV-RNA reduction related to IL28B variants were observed because of the lack of intrinsic immune cells in the model. In contrast, Zhang et al. and Yoshio et al. reported that a certain subset of dendritic cells (DCs) within human PBMCs could recognize HCV and produce large amounts of IFN-λs.[89, 90] The ability of production of IFN-λ3 was superior in subjects with a favorable IL28B genotype.[90] Moreover, IFN-α directly affected DC function and significantly increased IFN-λ production.[89] Based on these findings, it is tempting to speculate that exogenous IFN-α would increase IFN-λ production by DCs and/or HCV-infected hepatocytes during IFN-α therapy, and this could provide a potential explanation as to why IL28B genetic variants predict the outcome of IFN-α therapy (Fig. 1).

Figure 1.

Potential role of interleukin-28B (IL28B) single nucleotide polymorphism (SNP) in the response to interferon (IFN)-α therapy. IFN-α upregulates hepatic IFN-stimulated genes (ISGs). According to in vivo models of chronic hepatitis C virus (HCV) infection,[88] exogenous IFN-α would increase IFN-λ production by HCV-infected hepatocytes during IFN-α therapy. The amounts of IFN-λs produced on HCV-infected human hepatocytes were larger in liver with a favorable IL28B genotype. Dendritic cells (DCs) also produce large amounts of IFN-λ, following an immune response against HCV infection in the liver environment. The ability of IFN-λ3 production by DCs was superior in subjects with a favorable IL28B genotype.[90]

Recently, Olsson et al. performed RNA sequencing in primary human hepatocytes activated with synthetic double-stranded RNA to mimic HCV infection. They discovered that a new transiently induced region that harbors a dinucleotide variant ss469415590 (TT or ΔG) was strongly associated with HCV clearance. The ss469415590 polymorphism is located upstream of IL28B and is in high-linkage disequilibrium with rs12979860. The ss469415590 ΔG allele is a frameshift variant that creates a novel gene, designated IFNL4, encoding the type III IFN-λ4 protein, which is fairly similar to IFN-λ3. Interestingly, compared with rs12979860, ss469415590 is more strongly associated with spontaneous and treatment-induced HCV clearance in individuals of African ancestry, whereas it did not improve prediction among Caucasians and Asians. This can be explained by a lower level of linkage disequilibrium between the two polymorphisms in African Americans (r2 = 0.71) compared with Caucasians (r2 = 0.92) and Asians (r2 = 1.00).[91] Bibert et al. also noted that this polymorphism improved prediction of treatment-induced HCV clearance in patients infected with HCV genotype 1/4 or 2/3. In addition, they determined that induction of IL28B and IFN-γ-inducible protein 10 messenger RNA relies on ss469415590 but not rs12979860 in PBMCs.[92] Their findings provide new insights into the genetic regulation of HCV clearance and have implications for its clinical management.

Conclusion

Application of GWAS technology has revealed an unexpected role of IL28B in HCV infection. This finding could provide a strong rationale for developing novel therapeutic strategies for HCV infection as well as furthering basic studies on IFN-λs. The IL28B genotype could assist clinical decision-making for the treatment of acute HCV infection. In the context of PEG-IFN/RBV therapy for CHC, IL28B genotypes are strongly associated with treatment efficacy in patients infected with HCV genotype 1 or 4, with some effects on other HCV genotypes. IL28B genotyping is also useful for pretreatment prediction of the outcome of DAA plus PEG-IFN/RBV therapy, especially in treatment-naïve patients. Moreover, the IL28B genotype may affect responses to IFN-free regimens. Future more aggressive treatments, such as quadruple therapy or potent DAA combinations might obscure the influence of IL28B, but IL28B genotyping will remain useful for making decisions on suitable regimens and treatment duration in patients in the forthcoming era of DAAs. The mechanisms by which IFN-λs are active against HCV infection must be elucidated through the functional analyses of IFN-λs in future.

Ancillary