From Journal of Viral Hepatitis
Does IL28B Genotyping Still have a Role in the Era of Direct-acting Antiviral Therapy for Chronic Hepatitis C Infection?
J. A. Holmes; P. V. Desmond; A. J. Thompson
Posted: 09/28/2012; J Viral Hepat. 2012;19(10):677-684. © 2012 Blackwell Publishing
Abstract
IL28B genotype has been shown to be the strongest pretreatment predictor of sustained virological response (SVR) in patients with genotype 1 chronic hepatitis C infection (CHC) treated with pegylated interferon (peg-IFN) and ribavirin (RBV). Patients carrying the good response genotype have a two- to threefold higher chance of SVR than those with a poor response genotype, manifest as dramatically improved early viral kinetics. However, the treatment paradigm for CHC is changing with the introduction of potent direct-acting antivirals (DAAs). IL28B genotype remains relevant to both telaprevir and boceprevir treatment regimens, although the strength of association with virological response is attenuated. The association between IL28B genotype and outcomes of treatment regimens that involve peg-IFN plus combination DAA therapy, or IFN-free regimens, is currently being evaluated. IL28B genotype may remain relevant to individualizing the choice of treatment regimen in the future.
Introduction
Chronic infection with the hepatitis C virus (HCV) is a global epidemic affecting 130 to 170 million individuals, who are at risk of progressive liver fibrosis, decompensation and hepatocellular carcinoma. These complications have been shown to be reduced by viral eradication. Pegylated-interferon-alpha (peg-IFN) combined with ribavirin (RBV) has been the standard-of-care therapy for chronic hepatitis C (CHC) since 2003. However, the rate of response is suboptimal for genotype 1 HCV (HCV-1), with only 40–50% of patients being cured by dual therapy. Furthermore, the ability to identify which individuals would respond to treatment was limited. In the last 3 years, the field has made two major advances. In late 2009, genome-wide association studies (GWAS) identified genetic variants in the region of the IL28B gene that are strongly associated with the outcome of peg-IFN and RBV (PR) therapy.[1–3] In the IDEAL study pharmacogenetics cohort, carriage of the favourable IL28B genotype was associated with sustained virological response (SVR) rates of 70–80% in Caucasian patients treated with 48 weeks of PR, compared to 30–40% in patients carrying one of the unfavourable genotypes. Difference in population frequency of the favourable IL28B genotype explained much of the recognized ethnic disparity in PR response rates. IL28B genotyping immediately proved useful for pretreatment counselling for HCV-1 patients. However, PR is no longer standard-of-care for HCV-1 in many parts of the world. In 2011, the first direct-acting antivirals (DAAs) for HCV were approved for the treatment of HCV-1, increasing overall SVR rates almost twofold. 2011 also saw proof-of-concept that HCV could be cured using IFN-free regimens. DAA therapy clearly attenuates the association between IL28B genotype and HCV treatment response, but emerging data suggest that IL28B genotype will remain relevant to emerging treatment paradigms, including IFN-free therapy. In this article, we will discuss the clinical utility of IL28B genotyping for pretreatment counselling in the rapidly evolving era of DAA therapies for HCV.
IL28B Genotype and Response to Dual Therapy With peg-Interferon-A and Ribavirin
The first-generation DAAs are not yet universally available, and it is pertinent to briefly review the literature describing the association between IL28B polymorphism and response to peg-IFN and RBV therapy. IL28B genotype is the most important pretreatment predictor of response to peg-IFN and RBV therapy for HCV-1, where patients who carry the good response genotype (e.g. C/C at rs12979860) have reported SVR rates ≥70%, a two- to threefold increase over patients who carry one of the poor response genotypes (e.g. C/T, T/T at rs12979860).[4,5] The frequency of the good response IL28B genotype varies between individuals of different ethnic backgrounds, being >80% in certain Asian populations, 35–55% in Caucasians and <20% in patients of African ancestry. This variation explains much of the disparity in treatment response rates observed according to ethnicity. It is also relevant to discussion of the health economics of novel DAA therapies (see below). The mechanism explaining the association between IL28B genotype and IFN responsiveness is not known, but it manifests as profound differences in on-treatment viral kinetics. The good response IL28B genotype is associated with increased phase-1 and phase-2 viral kinetics,[5,6] resulting in increased rates of rapid virological response (RVR) and complete early virological response (cEVR) on-treatment (Table 1). In fact, most RVR patients carry the good response IL28B genotype.[5] It is important to note that all patients who achieve an RVR do well, and a high rate of SVR is observed even in the minority of patients who carry a poor response IL28B genotype. It is also important to note that the good response IL28B genotype is not sufficient to indicate short-duration PR therapy using response-guided protocols for HCV-1. Recent data have shown that short-duration therapy (24 weeks) is only sufficient for patients who carry the good response IL28B genotype and have a low baseline viral load (<400–800 000 IU/ml) and achieve an RVR.[7] Finally, the importance of IL28B genotype as a pretreatment predictor of IFN responsiveness should not discount other predictors of response, especially liver fibrosis stage. In the IDEAL study cohort, SVR rates in patients with advanced liver fibrosis were considerably lower (METAVIR F3-4, SVR in C/C = 41% vs C/T = 22% vs T/T = 11%).5]
IL28B Genotyping in the Setting of Telaprevir and Boceprevir Therapy
In the phase-3 registration studies of treatment-naïve HCV-1 patients, the addition of the NS3 protease inhibitors telaprevir (TVR) or boceprevir (BOC) to a PR backbone increased overall SVR rates from 40–45% to 68–79%.[8, 9] In the setting of such high SVR rates, the association between IL28B genotype and treatment response is attenuated (Table 2). However, it remains clinically relevant.
A retrospective analysis of the relationship between IL28B polymorphism (rs12979860) and treatment outcomes from the SPRINT-2 study of BOC-PR therapy for treatment-naïve patients has recently been published.[10] The analysis included 62% (653/1048) of the SPRINT-2 cohort who consented to genetic testing. Very high SVR rates were observed in all patients with the C/C genotype; in fact, SVR rates were equivalent in the BOC treatment arms and the PR control arm (Table 2). However, the C/C patients were much more likely to be eligible for short-duration therapy than the non-C/C patients (89% vs 52% non-C/C patients achieved undetectable HCV RNA levels at week 8). It was also notable that 97% of treatment-naïve C/C patients achieved ≥1 log10 IU/mL reduction in HCV RNA at the end of the 4 week lead-in phase of PR treatment.[10,11] A ≥1 log10 reduction in HCV RNA identifies good IFN responders, with SVR rates >80% and low risk of selection of resistance-associated variants. BOC therapy was associated with much more benefit in non-C/C patients, with rates of SVR of 55–71% compared to 27–28% with PR therapy alone in SPRINT-2 (Table 2). The lead-in phase of PR was more useful for stratifying IFN responsiveness in the non-C/C naïve population, in which 27% C/T and 46% T/T patients were identified to have poor IFN responsiveness, predicting for low SVR rate and higher risk of selecting resistant variants. Patients with poor IFN responsiveness might be considered for deferral of DAA exposure to minimize risk of selection of resistant variants, awaiting future quadruple therapy or combination DAA regimens. This should be an individualized decision, according to the urgency for antiviral therapy. IL28B genotyping can therefore identify patients in whom the lead-in has greatest clinical utility for stratifying IFN responsiveness.
Logistic regression modelling found that IL28B genotype was independently associated with the outcome of BOC-based therapy, OR = 2.6 (1.3–5.1) for C/C vs T/T and OR = 2.1 (1.2–3.7) for C/C vs C/T.[10] The other pretreatment predictors of SVR were low baseline HCV RNA level (< = 800, 000 IU/mL), the absence of cirrhosis, HCV-1 sub-type (1b vs 1a) and race (non-black vs black). IL28B genotype was the strongest predictor of good IFN response (≥1 log10 decline at week 4), OR = 26.5 (7.6–92.6) for C/C vs T/T and 16.4 (5.0–55.6) for C/C vs C/T.[10] When week 4 response was included in the SVR model, IL28B genotype was no longer independently associated with SVR. This is consistent with the observation from the PR literature that IL28B genotype informs IFN responsiveness, but once IFN responsiveness is defined in real-time by a trial of treatment, IL28B genotype is no longer clinically informative.[12]
IL28B genotyping has also been shown to be informative in the context of TVR therapy for treatment-naïve patients. Retrospective analysis of the association between IL28B polymorphism and treatment outcome was performed in a subset of the ADVANCE study population (n = 454/1088 [42%]).[13] The strength of association between IL28B genotype and treatment outcome was attenuated in the TVR treatment arms compared to the PR control arm. Rates of SVR were higher in the TVR treatment arms compared to the PR control arm among both C/C and non-C/C patients (it should be noted that the response rate among C/C patients in the PR control arm was lower than that observed in SPRINT-2, Table 2). SVR rates increased relatively more in non-C/C patients than C/C patients (Table 2). Similar to SPRINT-2, C/C patients were more likely to be eligible for short-duration therapy by achieving an extended rapid virological response (eRVR), defined as HCV RNA undetectable at weeks 4 and 12) with 12 weeks of TVR (C/C 78% vs C/T 57% vs T/T 45%).[13] SVR rates were very high in all patients who achieved an eRVR regardless of IL28B genotype (97% in CC patients and 88% in CT/TT patients). In patients who did not attain an eRVR, patients carrying the CC IL28B genotype had higher SVR rates (67% vs 38% for CT/TT patients). No data have yet been presented linking IL28B polymorphism to risk of resistance during TVR therapy, but given the key role for IFN responsiveness in controlling the emergence of resistance-associated variants (RAVs), this can be assumed.
In patients who have previously failed PR therapy, IL28B genotype is less informative for the outcome of TVR or BOC-based therapy. This was evaluated in retrospective analyses of the RESPOND-2 study of BOC therapy for prior relapsers and partial responders to PR therapy,[10] and the REALIZE study of TVR therapy for prior relapsers, partial responders and null responders (Table 2).[14] There was no significant association between IL28B genotype and treatment response noted in either study. This reflects again that IL28B genotype is no longer informative for SVR once IFN responsiveness is known, here defined by the prior course of PR therapy. IL28B genotyping may have a clinical role in real world patients where the prior treatment history is not well defined.
Can IL28B Genotype Identify HCV-1 Patients for Whom Dual Therapy With peg-IFN and RBV Should Remain First-line Therapy in the DAA Era?
The introduction of TVR and BOC represents a significant advance for the treatment of HCV-1. However, both drugs are very expensive. Modelling studies suggest that it may be possible to develop treatment algorithms in which PR is more cost-effective first-line therapy for a subgroup of patients. Patients who carry the C/C IL28B genotype achieve high rates of SVR, and a number of groups have suggested that in this population, PR is more cost-effective than universal triple therapy.[15–17] These patients also avoid the added morbidity of triple therapy, although this must be weighed against the need for an extra 24 weeks of total treatment time (most C/C patients will require 48 weeks of PR, but would be eligible for 24 weeks of triple therapy). There are a number of caveats to these modelling studies. They may be strongly influenced by underlying assumptions, as well as the set price for drug, and therefore may not be universally applicable to all patients and regions. One particular issue pertains to efficacy assumptions for patients with advanced liver fibrosis. In the IDEAL study, the rate of SVR was only 41% in C/C patients with F3-4 fibrosis,[5] and although there are little data concerning outcomes of triple therapy in this sub-population, it seems reasonable that triple therapy will be more effective, and cost-effective, for these patients.
IL28B Genotyping in Other Clinical Scenarios
Telaprevir and boceprevir are approved for the treatment of HCV-1 mono-infection only at present. The standard-of-care treatment for chronic infection with other HCV genotypes remains PR dual therapy. IL28B genotype is associated with the outcome of PR treatment for HCV-4, with a similar effect size to that seen in HCV-1.[18–20] IL28B genotype is less relevant to the outcome of PR treatment for HCV-2 and HCV-3.[21–26] HCV-2/3 are more IFN sensitive, and the data are conflicting. IL28B genotype may be most relevant for HCV-2/3 patients who are slow responders to PR (e.g. non-RVR patients).[21] IL28B genotype is associated with the outcome of PR therapy for HCV-1 in the setting of HIV co-infection, where HIV does not seem to attenuate the association.[27] IL28B genotype has been associated with the outcome of PR treatment for HCV-1 after liver transplantation, where both donor and recipient IL28B genotype influence outcome.[28–31] IL28B genotype is strongly associated with the outcome of acute HCV infection, where patients with poor response IL28B genotypes, particularly if anicteric, are unlikely to spontaneously clear infection and should be considered for early antiviral therapy.[32,33]
IL28B Genotyping and Next-generation Therapies Involving peg-Interferon
Emerging data suggest that IL28B genotype will remain relevant to the outcomes of DAA therapy when used in combination with a PR backbone. This has been the case for the emerging protease inhibitors (PIs), NS5A inhibitors (NS5I) cyclophilin inhibitors and non-nucleos(t)ide inhibitors (NNI) of the NS5B polymerase.[34–39] In general, rates of on-treatment response as well as SVR have been higher in patients with the good response IL28B genotype, and the good response genotype has predicted for short-duration therapy using response-guided regimens. Very recent interim data from the ATOMIC study suggest that the nucleotide inhibitor (NI) sofosbuvir (GS-7977) may overcome the IL28B effect. 90% of patients treated for 12 weeks with sofosbuvir plus PR achieved an SVR12 (n = 52).[40] Quadruple therapy regimens are also being evaluated. These involve the combination of two DAAs with PR. A small cohort of prior null responders was treated using the combination of daclatasvir (NS5I) plus asunaprevir (PI) plus PR.[41] 9/10 patients carried a poor response IL28B genotype. 10/10 patients achieved an SVR12; and 9/10 had a SVR24. Larger phase-3 studies are required in larger cohorts to confirm these encouraging results. Therefore, it seems likely that it will be possible to overcome the effect of IL28B polymorphism with the combination of PR plus either a high potency, high genetic barrier to resistance drug (sofosbuvir), or the combination of 2 DAAs. Even so, it may be that IL28B genotype might be used to individualize treatment strategies, such that patients who carry the good response IL28B genotype might be eligible for shorter, simpler or cheaper regimens, and conversely poor responder patients might require longer therapy with multiple DAAs. Such individualized treatment algorithms are yet to be explored in detail.
Finally, a brief note about IFN-lambda. IFN-lambda (IL29) is a type-3 IFN and member of the same family of IFNs as IL28B. The type-3 IFN receptor has a limited distribution of expression, and peg-IFN-lambda is being developed as a less toxic alternative to peg-IFNα. In a small phase-II dose escalation study, investigating the efficacy and safety of peg-IFN-lambda plus RBV, compared to peg-IFNα plus RBV, virological responses were similar, and similar associations between IL28B genotype and treatment response were observed comparing the two IFN preparations.[42]
IL28B Genotyping and IFN-free Therapy
Interferon-free therapy for HCV-1 is likely to be approved in the next 3–5 years. Recent data suggest that the good response IL28B genotype may represent an 'easy-to-cure' characteristic for certain IFN-free regimens also. The INFORM-1 study was the first study to demonstrate that IFN-free therapy could have a potent antiviral effect. Patients were treated with a combination of mericitabine (NI) and the danoprevir (PI) for 14 days, before follow-on PR therapy to 48 weeks. Analysis of the on-treatment viral kinetics in 15 patients during the 2 weeks of oral therapy revealed a significant difference in phase-II viral kinetics according to IL28B genotype,[43] suggesting that IL28B genotype influences the rate of clearance of infected hepatocytes during IFN-free therapy. This might be consistent with the association between IL28B genotype and spontaneous clearance of HCV.[44] SOUND-C2 is the largest study of IFN-free therapy to date (n = 362).[45] SOUND-C2 evaluated the combination of BI 201335 (PI), BI 207127 (NNI) ± RBV (Fig. 1a). Interim results demonstrated a clear difference in SVR12 according to IL28B genotype in HCV-1a patients (Fig. 1b,c). The low SVR rates observed in HCV-1a non-C/C patients resulted from virological breakthrough in most patients, suggesting that IL28B genotype influenced the emergence of RAVs. Future studies of this regimen plan to enrol only HCV-1a patients who carry the good response IL28B genotype. This IL28B effect may be overcome as more potent agents and combination regimens emerge. Analysis of the ELECTRON study observed SVR4 in 22/25 (88%) treatment-naïve HCV-1 patients treated with sofosbuvir plus RBV for 12 weeks.[46] 22 of the patients were infected with HCV-1a, and only 11/25 carried the good response IL28B genotype. Response rates with 'optimized' DAA combinations may be higher still. A recent study of the combination of sofosbuvir plus the daclatasvir reported SVR4 of 100% in 44 HCV-1 naive patients, with or without ribavirin.[47] Such optimized regimens are likely to be very expensive, however, and may not be necessary for all patients. IL28B genotype may continue to be useful to identify those patients for whom such regimens are most appropriate as first-line therapy.
Interim results from the SOUND-C2 study 42. (a) Patients were randomized to one of five arms, involving variable durations of treatment with the protease inhibitor BI 201335 once daily, the NS5B non-nucleoside inhibitor BI 207127 BID or TID, plus or minus ribavirin. Interim analysis of SVR12 results was recently presented (with SVR4 results for the 40-week treatment arm). Overall results according to treatment arm: A – SVR12 = 59%, B – SVR12 = 61%, C – SVR4 = 56%, D – SVR12 = 68% and E – SVR 12 = 39% (treatment arm E was halted early because of concerns about RBV-free treatment); (b) SVR12 results in treatment arm D, according to HCV-1 subtype and IL28B genotype (rs12979860); (c) SVR results in treatment arms A–D, comparing non-C/C patients infected with HCV-1a (grey columns), with a composite group including C/C patients infected with HCV-1a and all HCV-1b patients (white columns).
Conclusion
IL28B genotype is the strongest pretreatment predictor of response to dual PR therapy for patients chronically infected with HCV-1. The approval of the first generation of NS3 protease inhibitors represented a significant advance for the field. The association between IL28B polymorphism and treatment outcome is attenuated in the setting of triple therapy, but IL28B genotyping continues to be useful for pretreatment counselling, with the good response IL28B genotype identifying the likelihood of an individual being eligible for short-duration therapy. IL28B genotype may also be relevant to strategies for maximizing cost-effectiveness. IL28B genotype is also associated with the response to IFN-free regimens, and C/C patients remain easier to cure, particularly in the setting of HCV-1a. Future regimens involving potent NI plus PR, quadruple therapy or combinations of best-in-class DAAs are likely to achieve very high SVR rates, and IL28B polymorphism will no longer predict treatment outcome. However, IL28B genotyping may remain useful if it can be used to individualize treatment strategies, identifying patients who can be successfully treated with shorter, simpler or cheaper regimens.
No comments:
Post a Comment