Potential conflict of interest: A.J.T and J.G.M are coinventors of a patent related to the IL28B discovery. A.J.T has received research support from Merck, Roche and Gilead Sciences, has served as a consultant for Merck, Roche and Janssen-Cilag, and has served on speaker bureaus for Merck. J.G.M. is an employee of Gilead Sciences.
We have now entered a new era in hepatitis C virus (HCV) therapeutics. Direct-acting antiviral (DAA) therapies directly inhibit specific steps in the HCV viral life cycle, with targets including the HCV NS3 protease, the NS5B polymerase, and the NS5A phosphoprotein, as well as host cell proteins involved in HCV replication (e.g., cyclophilin inhibitors). All have potent antiviral effects in vitro. Unfortunately, monotherapy is associated with the rapid selection of resistant HCV variants and virological breakthrough. Therefore, the clinical efficacy of the first-generation DAAs will be reliant on combination with a pegylated interferon (pegIFN) and ribavirin (RBV) backbone. The first agents are the NS3 inhibitors boceprevir (BOC) and telaprevir (TVR), both recently approved by the Food and Drug Administration (FDA).1, 2
The relevance of IL28B polymorphism to DAA treatment regimens is not yet clear. The key questions for the field relate to IL28B genotype as a predictor of on-treatment virological response rates, overall sustained virological response (SVR) rates, the emergence of antiviral resistance, and necessary treatment duration in the context of first-generation triple therapy regimens, as well as future interferon-free regimens.
BOC plus pegIFN and RBV therapy increases rates of SVR compared to pegIFN and RBV alone in both treatment-naive (the SPRINT-2 study3) and pegIFN-experienced patients (the RESPOND-2 study,4 see below). The BOC treatment paradigm involves 4 weeks of lead-in phase pegIFN and RBV, before an additional 24-44 weeks of triple therapy with BOC plus pegIFN and RBV (BOC/PR) according to the rate of on-treatment virological suppression (response-guided therapy, RGT). The primary clinical utility of the lead-in phase of PR is risk stratification according to IFN responsiveness.3 Reduction in HCV RNA ≥ 1 log10 IU/mL at the end of the lead-in phase predicts for a higher rate of viral clearance and lower rate of BOC resistance.
Analysis of the SPRINT-2 data according to IL28B genotype (rs12979860) was recently presented.5 In all, 653 of 1,048 (62%) patients were recalled and consented to genetic testing (Table 1).5 BOC attenuated the association between IL28B genotype and treatment outcome, relative to the PR control arm. IL28B genotype remained an independent predictor of SVR, however (P < 0.0001).5 Among C/C patients, the rates of SVR were high regardless of treatment arm (80%, 82%, and 78% for BOC/PR48, BOC/RGT, and PR control, Table 1).5 Although BOC was not associated with an increased rate of SVR in C/C patients, 89% of C/C patients were HCV RNA undetectable at week 8 and eligible for short-duration therapy. This suggests that the major benefit of BOC in the C/C population might be to allow short treatment duration. BOC therapy was associated with a much greater increment in SVR in patients with the non-C/C IL28B genotypes, where BOC therapy lead to a 2-fold increase in SVR compared to control (Table 1). A lower percentage of the non-C/C patients (52%) had undetectable HCV RNA at week 8 of treatment and were eligible for short-duration therapy using the response-guided strategy. The clinical utility of the lead-in phase for identifying IFN responsiveness differed according to IL28B genotype. 97% of C/C patients in the SPRINT-2/RESPOND-2 studies achieved a 1 log reduction in HCV RNA by week 4. The lead-in phase was more useful for risk stratification in non-C/C: 68% of non-C/C patients in the BOC-treatment arms achieved a ≥ 1 log10 IU/mL reduction in HCV RNA at week 4.5 The SVR rate in these patients was 75%-82%.5 Among non-C/C patients with an HCV RNA reduction < 1 log10 IU/mL the SVR rate was only 24%-44%.
Table 1. IL28B Genotype Is Associated With Virological Outcome in Treatment-Naive Patients Who Receive Boceprevir (SPRINT-2). The Association Between IL28B Genotype and Treatment Response Is Attenuated in IFN-Experienced Patients (RESPOND-2). IL28B Genotype Is Associated With Virological Decline During the Lead-in Phase of pegIFN Plus RBV Therapy.
SVR rates for each treatment arm according to IL28B genotype (rs12979860, a bi-allelic SNP with three possible genotypes – C/C, previously associated with good IFN response; C/T and T/T, previously associated with poor IFN response 40). Treatment groups: i) BOC/PR48 (lead-in period of pegIFN alfa-2b and RBV (PR) for 4 weeks before boceprevir (BOC) plus PR for 44 weeks; ii) BOC/RGT (lead-in period before BOC/PR for a further 24 weeks; those with a detectable HCV RNA level between weeks 8 and 24 then received placebo plus P/R for an additional 20 weeks [RGT = response-guided therapy]); iii) PR (lead-in before PR for a further 44 weeks).
Decision point for short versus long treatment duration with RGT.
Data on the risk of selection of BOC-resistant variants according to IL28B genotype was not presented, but given the strong relationship between the selection of resistant variants and IFN responsiveness,6 IL28B genotype can be assumed to predict antiviral resistance.
An analysis of the association between IL28B genotype and genotype 1 HCV treatment outcome in the TVR phase III registration study of treatment-naive patients (the ADVANCE study7) has been performed. This was a retrospective analysis of IL28B genotype tested in deidentified leftover specimens, available for 454/1,088 (42%) participants. Only Caucasian patients were included in the analysis. Detailed clinical and demographic data, including quantitative serum HCV RNA measures and liver fibrosis stage, were not accessible, and formal statistical analysis was not performed. Despite these limitations, the data were informative. TVR-based therapy increased the SVR rate in the setting of all IL28B genotypes (Table 2). As for BOC, the major benefit of TVR-based therapy was in patients with the poor response IL28B genotypes, where SVR rates were more than double that of the control group. TVR-based therapy was also associated with a numerical increase in the SVR rates in the good response C/C patients. The addition of TVR therefore attenuated the association between IL28B genotype and pegIFN treatment outcome, although patients with the good response C/C genotype still had higher rates of SVR in each treatment arm (Table 2). IL28B genotype was associated with higher rates of extended rapid virological response (eRVR) (Table 2). eRVR is used to identify patients suitable for short-duration therapy (24 weeks), and the data therefore suggest that most C/C patients will be eligible, compared to only approximately half of patients with the poor response genotypes.
Table 2. Data From ADVANCE.7 IL28B Genotype Is Associated With Virological Response in Treatment-Naive Patients Who Receive Telaprevir Therapy.
ADVANCE IL28B Cohort N = 454/1088 (42%)
N = 140
N = 153
N = 161
SVR rates for each treatment arm according to IL28B genotype (rs12979860) Treatment groups: i) T12 PR24/48 BOC (telaprevir for 12 weeks plus pegIFN alfa-2b and RBV (PR) for 24 or 48 weeks total according to the achievement of an eRVR); ii) T8 PR24/48 BOC (telaprevir for 8 weeks plus PR for 24 or 48 weeks total according to the achievement of an eRVR); iii) PR48 (PR for 48 weeks). eRVR = extended rapid virological response (defined as an undetectable HCV RNA [<25 IU/mL] at weeks 4 and 12 of treatment).
SVR n (%)
48/68 (71% )
RVR/eRVR n (%)
CC – RVR
CT – RVR
TT – RVR
As for BOC, the association between IL28B genotype and rapid on-treatment viral suppression, as well as SVR, suggests that IL28B genotype may be relevant to the emergence of TVR-resistant HCV variants. Resistant variants have been observed in most patients who failed to attain SVR in the TVR phase 3 studies.8 As yet, no analysis of the association between IL28B genotype and the emergence of resistant variants has been presented.
Other DAAs in Development.
Data concerning the relevance of IL28B genotype to treatment outcomes with the next wave of DAAs are emerging, including NS3 protease inhibitors (TMC4359, 10), NS5A inhibitors (BMS-79005211, 12), NS5B polymerase inhibitors (nucleos(t)ide inhibitors : mericitabine = RG712813; GS-797714; nonnucleos(t)ide inhibitors: ANA59815), and cyclophilin inhibitors (alisporivir16), all used in combination with pegIFN and RBV (available data summarized in Table 3). The more potent treatment regimens attenuate the association between IL28B and treatment response to a greater extent than BOC or TVR, although most studies have observed a small numerical difference in response rates. Again, the major benefit for most of the DAAs is in patients carrying the poor response IL28B genotypes. This is a rapidly moving field and many of these agents are now entering phase 3 development.
Table 3. Summary of Recent Data Presenting the Relationship Between IL28B Genotype (Rs12979860) and Virological Responses for DAA Currently in Development for the Treatment of Genotype 1 HCV (All Data Presented in Abstract Form Only)
The relevance of IL28B polymorphism to IFN-free treatment regimens remains unclear. The INFORM-1 study was the first study to explore combination therapy with two DAAs. Patients were treated with 2 weeks of the nucleoside analog mericitabine (RG7128) plus the NS3 protease inhibitor danoprevir (RG-7227).17 In a recent analysis of on-treatment viral kinetics stratified by IL28B genotype, the slope of viral decline was steeper in patients with the good response IL28B genotype.18 The difference in viral kinetics was not statistically significant in this small cohort, but the data suggested that IL28B may remain relevant even in the absence of IFN treatment, perhaps consistent with viral suppression leading to differential immune “restoration” according to IL28B genotype. This is biologically plausible given the recognized role for the NS3 protease in targeting intracellular IFN signaling pathways through cleavage of TRIF (TLR3) and IPS-1 (RIG-I).19-22 Interim data from the SOUND-C2 study23 also suggested that IL28B genotype may remain relevant to IFN-free treatment regimens, while also pointing to critical roles for HCV subtype and RBV (Table 3). Patients were treated with the NS3 protease inhibitor BI201335 and the NS5B nonnucleoside inhibitor BI207127 with or without RBV. In patients receiving triple therapy, the rate of HCV RNA undetectability at week 12 was higher in HCV-1A patients who carried the good response IL28B genotype compared to the poor response genotypes (Table 3). In contrast, week 12 response was high in all HCV-1B patients. HCV-1A is emerging as a “hard-to-cure” genotype for IFN-free regimens, likely reflecting a lower barrier to NS3 protease inhibitor resistance, but also perhaps a different sensitivity of these variants to endogenous IFN signaling, compared to subtype HCV-1B.11, 24 This interesting issue will require further evaluation. Among patients who received RBV-free therapy, week 12 response was also higher in good-response IL28B patients, suggesting that the role of RBV may be particularly important for poor-response IL28B patients.
The relationship between IL28B genotype and novel immunomodulatory therapies is also being investigated. IL28B genotype was strongly associated with week 12 complete early virological response (cEVR) in genotype 1/4 HCV patients treated with pegIFN-γ plus RBV in the EMERGE study.25, 26 In patients carrying the poor response IL28B genotypes, week 12 responses were higher with pegIFN-γ plus RBV than pegIFN-α plus RBV. IL28B genotype was also associated with the antiviral efficacy and week 12 virological response to the TLR-9 ligand IMO-2125 in phase 2 development.27 IL28B polymorphism may therefore be relevant to any immune modulator where the downstream signaling pathways converge on the intrahepatic IFN-stimulated gene (ISG) response.28-30
GI-5005, a yeast-based immunotherapy in phase 2 development, is designed to stimulate NS3- and core-specific cytotoxic T-cell responses.31 Used in combination with pegIFN and RBV, GI-5005 was observed to increase SVR rate and augment HCV-specific T-cell responses, with the greatest incremental benefit observed in patients carrying the poor responder IL28B genotype, suggesting GI-5005 may differentially augment adaptive immune responses in the setting of the poor-response IL28B patients.32, 33 This will require prospective confirmation.
IFN-Experienced Patient Populations
The RESPOND-2 study investigated the effectiveness of BOC triple therapy in patients with a history of pegIFN and RBV therapy.4 The study enrolled relapsers (undetectable HCV RNA at the end-of-treatment but no SVR), and partial-responders (week 12 reduction in HCV RNA > 2 log10 IU/mL but persistently detectable HCV RNA on-treatment). Prior null responders, defined by a < 2 log10 IU/mL reduction in serum HCV RNA at week 12 of the past treatment course, were excluded. The poor response IL28B genotypes were more common than in the SPRINT-2-naive population (C/C 24%, C/T 61%, and T/T 15%). The rates of SVR in each treatment arm according to IL28B genotype are presented in Table 1. BOC/PR48 was associated with an increase in SVR across all IL28B genotypes. In a multivariate logistic regression model using stepwise selection, IL28B genotype was not independently associated with treatment outcome. However, as in the SPRINT-2 study, IL28B was associated with on-treatment virological decline. 82% of C/C patients had undetectable plasma HCVRNA levels at week 8 and were eligible for short-duration therapy (36 weeks), compared to 51% of non-CC patients. The clinical utility of IL28B genotyping for these treatment-experienced patients is therefore not to predict treatment outcome, but rather to predict the likelihood of short-duration therapy. IL28B genotype was strongly associated with virological response during the lead-in phase (odds ratio [OR] 4.5, 1.5-13.7, P = 0.007).
The REALIZE study was a phase 3 study of the effectiveness of TVR-based triple therapy for treating pegIFN treatment-experienced patients, including prior relapsers, partial responders, and null-responders.34 The study compared three treatment arms: T12PR48, 4 weeks of lead-in PR plus T12PR44, and PR48 control. There was no response-guided protocol. The IL28B genotype distribution was C/C 18%; C/T 61%, and T/T 21%, with this nonresponder cohort being enriched for the poor response IL28B genotypes. SVR rates were numerically higher in C/C patients in the overall cohort, although this difference did not reach statistical significance (Table 4). Prior treatment response was the strongest predictor of treatment outcome, and no difference in SVR rate was seen according to IL28B genotype when patients were considered on the basis of treatment history (Table 4). No data on the relationship between IL28B genotype and on-treatment viral decline was presented. The REALIZE study did not have a response-guided arm and therefore could not evaluate whether shorter duration TVR-based therapy might be possible according to IL28B genotype in IFN-experienced patients.
Table 4. Data From REALIZE.41 The Association Between IL28B Genotype and Treatment Response Is Attenuated and No Longer Clinically Relevant Once Patients Were Stratified According to Prior IFN Response.
SVR n (%)
Prior Null Responders
Pooled TVR Arms N = 422
PR Control N = 105
Pooled TVR Arms N = 209
PR Control N = 52
Pooled TVR Arms N = 79
PR Control N = 20
Pooled TVR Arms N = 134
SVR rates for each treatment arm according to IL28B genotype (rs12979860).
160/266 (60% )
Comment on IL28B Genotyping in Patients With a Well-Characterized IFN Treatment History.
Once on-treatment responses are defined (e.g., quantitative HCV RNA reduction, week 4, week 8 viral clearance), the predictive power of IL28B for SVR is much less strong—measurement of viral decline has captured the information predicted by IL28B genotype. In fact, intensive monitoring of on-treatment viral decline is more accurate for predicting outcome, as it will also capture the influence of other important predictors of IFN treatment response such as liver fibrosis stage, ethnicity, and the metabolic syndrome.35 The clinical utility of IL28B genotyping for predicting response to DAA therapy is therefore limited in well-characterized patients who have previously failed pegIFN-based therapy, as the IFN responsiveness of these patients has been defined by their treatment history. It may remain more relevant in the real world, where patients often have poorly documented treatment histories, particularly where adherence may have been limited, or the primary treatment regimen involved standard IFN or pegIFN monotherapy. The most accurate measure of IFN responsiveness in this case would be a “real-time” assessment using a lead-in phase of pegIFN and RBV.
Genotype 2/3 HCV.
IL28B genotype is less relevant to pegIFN and RBV treatment of genotype 2/3 HCV, which are much more IFN-responsive. As yet there is no clear clinical utility for testing IL28B genotype prior to standard-of-care therapy. It may be relevant to non-RVR patients, as a trigger for 24 versus 48 weeks of treatment, but this requires prospective investigation.36 Newer DAAs, including the nucleoside NS5B inhibitors, NS5A inhibitors, and the cyclophilin inhibitors appear to have cross-genotypic actions. In the phase 2 ELECTRON study 40 patients with chronic genotype 3 HCV infection were randomized to 12 weeks of treatment with the NS5B pyrimidine nucleotide analog GS-7977 400 mg daily plus RBV, with or without the addition of variable duration pegIFN therapy (4, 8, or 12 weeks).37 All of the patients in the study achieved SVR12, including 10/10 in the IFN-free treatment arm. Such high response rates are seen in larger studies, IL28B genotype will not be clinically relevant.
How Will IL28B Genotyping Be Incorporated Into the BOC/TVR Treatment Paradigms in Treatment-Naive Patients?
IL28B genotyping cannot be recommended as a routine test, but it provides useful clinical information. In C/C patients treated with TVR/BOC, there is little increment in SVR compared to pegIFN and RBV standard of care. Rather, the aim of triple therapy will be to shorten treatment duration to 24-28 weeks (Fig. 1). Uptake will therefore depend on the willingness of providers to fund triple therapy in this group (see below), as well as clinician/patient preferences regarding the risk-benefit of 12-28 weeks of triple therapy versus the known morbidity of pegIFN/RBV over 48 weeks. The benefit of the lead-in phase of pegIFN and RBV therapy prior to BOC therapy in these IFN-responsive patients is not clear. The lead-in strategy does not have great clinical utility for C/C patients where it is being used to stratify patients according to a 1 log10 IU/mL reduction in HCV RNA at 4 weeks—this is almost universal in these C/C patients, raising the possibility that it could be dispensed with altogether. It may be more useful in the difficult-to-cure patients as above. Unfortunately, there is no prospective data that have investigated the efficacy of BOC-based therapy in C/C patients without a lead-in phase. In non-C/C patients, TVR/BOC-based therapy is associated with > 2-fold increases in SVR.5, 7 For these patients, the cost-effectiveness and risk-benefit of triple therapy will be more favorable (see below, Fig. 1). The lead-in strategy will be more useful for refining IFN responsiveness by identifying a 1 log response, stratifying for likelihood of SVR 5, and the risk of selection of resistant variants also. In addition, IL28B genotyping may be relevant to clinical decision-making where there is concern about the added toxicity of TVR/BOC, or where there is potential for problematic drug/drug interactions. In such a scenario, pegIFN and RBV may be preferred if the patient carries the good response IL28B genotype.
What Is the Most Cost-Effective Strategy for the Introduction of DAA-Based Therapy for Genotype 1 HCV?.
This is an important question. The wholesale acquisition cost (WAC) of a 48-week course of pegIFN and RBV in the US is approximately $US30,000. The current pricing schedule for BOC will be $US1,100 per week, translating to $US26,400 for 24 weeks, $US35,200 for 32 weeks, and $US48,400 for 44 weeks of treatment. TVR has a $US49,200 WAC price for 12 weeks. The cost for both BOC and TVR will be in addition to the cost of the pegIFN and RBV backbone, meaning that for most patients the costs of treatment will more than double. IL28B genotyping has been proposed as a strategy for prioritizing treatment access (Fig. 1).38 A recent analysis of strategies for using TVR in good response IL28B patients concluded that pegIFN and RBV response-guided therapy, with TVR therapy reserved for patients who relapsed, was more cost-effective than universal TVR treatment.38 Such analyses are particularly important for Asia, where the majority of the population carry the good response IL28B genotype, and the cost:benefit ratio for population-based DAA therapy will differ from the West.
Should the Field Develop Different DAA Treatment Strategies According to IL28B Genotype?
It may be possible to develop personalized DAA treatment strategies according to IL28B genotype. Looking forward, IL28B genotyping might be used to identify a group of patients for whom IFN should continue to be part of the treatment regimen (Fig. 1). This will be predicated on IFN-containing regimens offering one of the following benefits: higher SVR rates and/or lower rates of resistance, shorter treatment duration, or lower cost, than emerging IFN-free regimens.
The majority of naive IL28B good responder patients treated with BOC or TVR will be candidates for short-duration therapy of 24-28 weeks. More abbreviated durations may be possible. Viral kinetic modeling of TVR effectiveness has predicted that 10-12 weeks of therapy should be sufficient to clear HCV in 95% of fully compliant patients.39 Even shorter-duration therapy may be possible with more potent single/dual combination DAAs. Investigation of this approach is warranted and a number of short-duration studies are now under way (www.clinicaltrials.gov).
In patients with the poor response IL28B genotypes, IFN has a weaker antiviral effect and is therefore less useful for preventing the emergence of resistant variants in combination DAA regimens. Combination therapy with potent DAAs targeting different steps in the viral lifecycle will be required to maximize efficacy and prevent resistance. For the immediate future pegIFN and RBV backbone will remain necessary, but it is this group for whom future IFN-free regimens are most attractive (Fig. 1). Effective IFN-free regimens may be available sooner than expected, particularly for HCV-1B infection, within a 3-5 year timeframe.
IL28B polymorphism is strongly associated with treatment response to pegIFN and RBV in patients chronically infected with genotype 1 HCV. Although the mechanism remains unclear, this occurs as a result of improved viral kinetics in the setting of the good response variant. The addition of DAA therapy with TVR or BOC also improves on-treatment kinetics, increases SVR rates, and attenuates the association between IL28B genotype and outcomes. In treatment-naive patients the IL28B genotype continues to be significantly associated with SVR. In patients with the good response IL28B genotype the major benefit of DAA appears to be to allow short treatment duration (24-28 weeks) with response-guided therapy, while still achieving SVR rates > 80%. The cost-effectiveness of this approach remains unproven. In the future, even shorter treatment duration of ≤ 12 weeks should be explored.
Poor-responder IL28B patients have the most to gain from the addition of DAAs, with SVR rates increasing 2-fold with BOC/TVR to 55%-70% among the treatment-naive.5, 7 The cost-effectiveness of DAAs for these patients appears favorable. In treatment-experienced patients planned for BOC/TVR therapy, there is little utility for IL28B genotyping once patients have been stratified for prior treatment response.
In summary, DAAs attenuate the association between IL28B polymorphism and treatment outcome. We believe that IL28B polymorphisms will retain some clinical relevance while IFN continues to be used as the backbone for antiviral therapy. The ongoing relevance of IL28B polymorphism to treatment decisions for HCV patients will require the development of specific treatment algorithms for good versus poor responder IL28B patients, where IL28B genotype identifies patients for whom IFN offers the benefit of improved rate of SVR, shorter treatment duration, or lower consumer cost.