January 5, 2012

IL28B genetic polymorphism testing in the era of direct acting antivirals therapy for chronic hepatitis C: ten years too late?

Liver International

Special Issue: Proceedings of the 5th Paris Hepatitis Conference. International Conference of the Management of Patients with Viral Hepatitis: Special Edition Hepatitis C

Volume 32, Issue Supplement s1, pages 74–78, February 2012

Review Article

Donald M. Jensen1,*, Stanislas Pol2

Article first published online: 29 DEC 2011

DOI: 10.1111/j.1478-3231.2011.02712.x

© 2012 John Wiley & Sons A/S

Abstract

An association between variations at the IL28B gene locus and HCV clearance (spontaneous recovery or sustained virological response under pegylated interferon (PEG-IFN) and ribavirin (RBV) has been extensively described. In genotype 1-infected patients, the new direct antiviral agents (DAA) including the two approved protease inhibitors boceprevir and telaprevir, in association with the PEG-IFN/RBV combination is the new standard of care making it necessary to redefine the interest of the IL28B genotype in the decision to treat and how to treat genotype 1-infected patients. In treatment-naïve patients, IL28B status can certainly identify those with a high probability of achieving SVR with response guided therapy and probably in whom the duration of treatment can be markedly reduced. In experienced patients, the impact of IL28B genotypes is limited and cancelled by early viral kinetics. However, the decision to initiate or withhold therapy remains a clinical one. In summary, although it was a major milestone in the treatment of patients with PEG-IFN/RBV,IL28B polymorphism testing entered the clinical arena almost 10 years too late.

Well established on treatment and baseline predictors of sustained virological response (SVR) to pegylated interferon and ribavirin (PEG-IFN/RBV) in patients with chronic hepatitis C virus (HCV) genotype 1 infection include rapid virological response (RVR; undetectable HCV RNA at week 4), low baseline viral load (<600 000 IU/ml), non-Black race, and absence of severe fibrosis or insulin resistance [1]. For some years, low serum levels of the 10 kDa interferon gamma-induced protein (IP-10) have also been associated with a better PEG-IFN/RBV response [2, 3].

Genetic variations have long been sought to explain the differences in host antiviral responses, and it is now well established that host genetics plays a role in the response to IFN-based therapy in HCV infection. Genome Wide Association Studies (GWAS) have described single nucleotide polymorphisms (SNPs) on chromosome 19 in the region of the IL28B gene that were highly predictive of spontaneous resolution of acute hepatitis C infection [4], response to PEG-IFN α/RBV therapy in the general population [5, 6, 7, 8, 9] as well as in human immunodeficiency virus (HIV) co-infected individuals [10] and liver transplant recipients in whom both donor and recipient IL28B haplotypes contribute to the probability of treatment response [11].

Although an association between outcomes of treatment response in HCV infection and variations in the IL28B gene locus were observed, a causal variant responsible for these effects remains unknown: the specific immunological mechanisms involved in HCV clearance associated with IL28B genotype remains elusive and a functional link between IL28B genotype and liver cytokine expression has not been established (12). It is however known that type III interferons, including IFN-λ3, are encoded by IL28B (13, 14). These type III interferons activate the Jak-STAT pathway which leads to induction of several IFN stimulated genes (ISGs) which are responsible for their antiviral activity (15). Thus, it is involved in the T-cell adaptive immune response [12]and IL28B has been associated with increased CD8+ cytotoxic T cell responses [16]. Interestingly, it has been demonstrated that in non-responders, some interferon-stimulated genes were upregulated before treatment [17, 18].

In addition, minor alleles of IL28B polymorphisms (i.e. rs8099917 G and rs12979860 T) have been associated with reduced IL28B expression in peripheral blood mononuclear cells [19]. Thus, IL28B genotypes may play a role in viral containment, and recent results suggest that IL28B polymorphisms associated with poor HCV clearance may actually be protective against hepatic necroinflammation and fibrosis progression, particularly in patients with HCV genotypes other than one [20]. The aim of this controversy is to re-define at the time of triple therapy the interest of the IL28B genotype in the decision to treat and how to treat genotype 1-infected patients.

IL28B and the pegylated interferon ribavirin combination

In clinical trials investigating responses to PEG-IFN/RBV (not including DAA), the association of IL28B genotype with SVR seemed to be similar to that of the viral genotype. In CC homozygous patients of Caucasian origin (around 30% of patients) without severe fibrosis, the probability of achieving SVR is estimated at 86%, vs. 36% and 43% for TT and CT respectively. An RVR was obtained in 30% of CC patients vs. 5% of CT or TT patients but in cases of RVR, the SVR rates were greater than 90% irrespective of the IL28B genotype. This suggests that RVR is a stronger predictor of SVR than IL28B CC genotype. However, there are no data yet assessing the efficacy of 24 weeks of therapy in CC patients with RVR. In the absence of RVR, SVR was achieved in 60% of CC patients. It is noteworthy that the odds ratios (OR) that describe the strength of association between IL28B genotype and virological response to PEG-IFN/RBV therapy vary according to the study taken into consideration. Tanaka et al. reported an OR for the association of the rs12980275 and rs8099917 polymorphisms with ‘null virological response’ of 17.7 and 27.1 with highly significant P values of 2.84 × 10−27 and 2.68 × 10−32 respectively. [9] Ge et al. determined a combined SVR OR of 3.1 for rs12979860[6] and Suppiah et al. found a combined SVR OR of 1.98 for rs809917. [8] It is challenging to ascertain the predictive value of a particular IL28B allele in the first two studies cited; however, Suppiah and colleagues were careful to note that ‘according to a model of dominant inheritance, the rs8099917 G allele predicts non-response with 57% sensitivity and 63% specificity’. They went on to report a negative predictive value (NPV; indicating correct prediction of treatment failure) of only 64%. Polymorphisms in the region of IL28B gene have been associated with PEG-IFN/RBV treatment response mainly in genotype 1 HCV infections. Recently, in genotype four patients with chronic hepatitis C, a better treatment response rate was associated with IL28B gene SNP rs12979860. The response rates were 81.8%, 46.5% and 29.4% for genotype CC, CT and TT respectively [21].

IL28B and direct acting antiviral agents (DAA)

The IL28B genotype status is the most important pretreatment predictor to PEG-IFNα/RBV therapy and it is increasingly evident that after the development of chronic infection, the response to triple therapy is also partially governed by polymorphisms at the IL28B locus [22]. IL28B status and outcome to therapy has been evaluated in several phase 3 clinical trials of telaprevir and boceprevir when used in conjunction with PEG-IFNα/RBV. SPRINT-2 and RESPOND-2 are randomized, double-blind, placebo-controlled studies of boceprevir, PEG-IFN-α/RBV. SPRINT-2 registered treatment-naïve genotype 1 hepatitis C patients [23] whereas the RESPOND-2 trial enrolled genotype 1 patients who had failed prior therapy with PEG-IFNα/RBV (relapsers and partial responders) [24]. Poordad et al. presented data pointing to IL28B status as the most important pretreatment predictor of response in patients receiving triple therapy [25]. In both trials, the IL28B CC genotype was significantly predictive of a favourable response to therapy as defined by SVR in all arms. In addition, IL28B CC genotype can also be used as a predictor of eligibility for shortened therapy as it was significantly associated with undetectable HCV RNA after 8 weeks of therapy in both SPRINT-2 (89% for CC vs. 52% for CT and TT combined, respectively) and RESPOND-2 (89% and 82% ,respectively for the CC genotype vs. 52% and 51% for the CT and TT genotypes). In the SPRINT-2 study, IL28B genotype was an independent predictive factor for SVR (OR = 4.5; P < 0.001). SVR rates in the combined boceprevir-based treatment arms were 81% in CC patients, 68% in CT patients and 57% in TT patients. Since responses with PEG-IFN/RBV therapy alone were also high in CC patients, the difference in SVR rate for boceprevir treated patients compared with PR48 was smaller in CC vs. CT or TT patients (3% vs. 40% or 30%, respectively). However, in the RESPOND-2 study, IL28B polymorphisms did not have a significant effect on SVR which rates were 78% in CC patients, 67% in CT patients and 66% in TT patients treated with boceprevir-based therapy. This was consistently higher than those obtained with PEG-IFN/RBV alone: 46%, 17% and 50% in CC, CT or TT patients respectively.

The ADVANCE study, a randomized, double-blind, placebo-controlled phase III trial examined the safety and efficacy of PEG-IFN-α, RBV, and telaprevir in treatment-naïve patients with genotype 1 hepatitis C virus [26]. A stepwise decrease in overall SVR rates occurred with increasing stage of fibrosis (81% in F0-F1, 75% in F2 and 62% in F3-F4) [27]. Only Caucasian patients were analysed for IL28B status (evaluated in 42% (454/1088) of the patient population) and demonstrated an increased SVR across all IL28B genotypes, but patients with the favourable CC genotype still had the best outcome as determined by SVR in all arms of the study taken into consideration. In addition, patients with the CC genotype were most likely to attain RVR and to have shortened durations of therapy [26]. In summary, although the results of the multivariate analysis have not yet been presented, the IL28B genotype had some impact on treatment outcome with 90% of CC, 71% of CT and 73% of TT patients achieving a SVR. Although the SVR rate was higher with triple therapy vs. PEG-IFN/RBV alone whatever the IL28B genotype, the difference between treatment arms was more striking in CT and TT patients than in CC patients (SVR difference in T12/PR arm vs. PR48: 47% for CT and TT genotypes combined vs. 26% for CC).

The REALIZE study has demonstrated that treatment with telaprevir in combination with PEG-IFN/RBV significantly increases SVR rates in prior relapsers and non-responders, including prior partial and null responders, compared with PEG-IFN/RBV-only control. The subanalysis of IL28B genotypes (SNP rs12979860) did not show differences in SVR rates across the genotypes and according to prior response [28]. Although there was a trend for increased eRVR (undetectable HCV RNA at week 4 and week 12 of therapy) and decreased relapse rates in prior partial or null responder patients with the CC vs. CT or TT genotypes, it is difficult to draw firm conclusions owing to the small number of patients in some groups. Thus, IL28B genotype may have limited value as a prognostic marker in patients with prior PEG-IFN/RBV treatment failure who are being evaluated for retreatment with a boceprevir- or telaprevir-based regimen.

Finally, as more potent DAA combinations are developed, some of which will not require an IFN backbone, it will be important to note how valuable these pretreatment predictors will be. There is also increasing evidence that more potent DAA regimens reduce the importance of IL28B genotype status as a determinant of the likelihood of response. In patients enrolled in PILLAR, a phase IIb study of TMC435, a second generation NS3-4A protease inhibitor in combination with PEG-IFNα/IFN for treatment-naïve genotype 1 patients, high rates of viral response were achieved in all patients regardless of IL28B genotype or pretreatment IP-10 levels.

IL28B and clinical decision-making

So how does IL28B genotype knowledge really affect clinical decisions? This is the fundamental question. Does it impact decision-making? Although it is comforting to know that IL28B status relates to antiviral response, will knowledge of IL28B status impact the doctor-patient decision to initiate treatment? Or to chose one treatment over another?

Pro

For treatment-naïve patients, IL28B status certainly can identify those who would have a high likelihood of SVR with response guided therapy. This is valuable not only for the patient (information on the chance of recovery and of reduction of treatment duration which affects safety and tolerability) but also for health authorities (cost of therapies). Other than for counselling, knowledge of IL28B genotype will be useful if this information is used to avoid treatment; an additional predictor for a decision that more rightfully should be based upon disease severity characteristics.

The other potential reason to consider IL28B testing in treatment-naïve subjects would be to avoid protease inhibitor therapy entirely: one might also argue that an individual with a CC genotype could be spared the expense and adverse events associated with boceprevir or telaprevir therapy, and would achieve similarly excellent results with PEG-IFN/RBV alone, in those patients with RVR since SVR is not different between dual and triple therapy.

In future, IL28B genotype could be used to propose a shorter duration of triple therapy: a retrospective analysis of the French patients of the Prove2 study reports a 100% SVR rate in the CC patients treated by the short arm of 12 weeks of triple therapy with telaprevir, PEG-IFN/RBV compared to 40% in the CT and 20% in TT patients (work in progress).

Con

Although the IL28B genotype has a clear impact on SVR in naïve (and to a lesser extent treatment-experienced) patients treated with PEG-IFN/RBV as well as triple therapy, the on-treatment virologic responses provide equally compelling prognostic data. Given the advantage of triple therapy over dual therapy across all IL28B genotypes, the question is mainly a clinical one: treatment or no treatment. In particular, will patients benefit from treatment now or should they wait until further therapeutic improvements? Decisions should be based upon disease severity, cost, tolerability, potential for medication adherence, and not necessarily upon who will have the greatest SVR, although this is an important factor. If considered in this light, knowledge of IL28B status has somewhat less relevance. Would we not offer treatment to an IL28B TT prior partial responder with cirrhosis? Of course we would.

It is true that knowledge of a CC IL28B genotype might be used to consider dual therapy with PEG-IFN/RBV and the advantages of concomitant protease inhibitor therapy might be offset by the concern that stopping at 24 weeks might be associated with relapse in those subjects with any negative predictors. But the real advantage would be if PEG-IFN/RBV could also be shortened to 24 weeks. The high SVR results reported by Ge and colleagues in the IDEAL trial [2] in Caucasians with the CC genotype only included those who completed a 48 weeks course of therapy and at least 12 weeks of follow-up. The results in real life may not be comparable because the confounding effects of age, cirrhosis, obesity and insulin resistance need to be considered and may further diminish response to PEG-IFN/RBV. The ability to shorten PEG-IFN/RBV therapy to 24 weeks is restricted to rare patients with an on treatment RVR (‘super responders’, largely Caucasians) with low baseline viral loads (< 600 000 IU/ml) and thus is not equivalent to the response guided therapy using telaprevir or boceprevir in which baseline negative predictors (viral load, insulin resistance…) have less impact.

In summary, although it was a major milestone in the treatment of patients with PEG-IFN/RBV, IL28B polymorphism testing entered the clinical arena almost 10 years too late. Predictive medicine is still difficult and IL28B genotypes will probably help in the decision to treat or not to treat with PEG-IFN/RBV in countries with limited resources and in rare patients with intermediate fibrosis where there may be either a hesitation between using triple therapy or waiting for future combinations. Although the interest of testing for IL28B polymorphism in the era of DAA will probably be to reduce triple therapy from 24 to 12 weeks or from 48 to 24 weeks, depending on other baseline predictors including fibrosis, this must still be clearly established. As HCV therapy and the field moves away from IFN-based regimens, and as more and more potent antiviral agents are approved, IL28B will probably become a footnote in the history of hepatitis. What a pity!

Conflicts of interest

Dr Jensen has received honoraria from the following entities: Advisory Boards for scientific input towards clinical trial design and interpretation: Abbott, Boehringer-Ingelheim, BMS,Genentech, Merck, Pharmasset, Roche Global, Tibotec/J&J, Vertex. Research Funding for clinical trials and/or investigator-initiated studies: Abbott, Boehringer-Ingelheim, Genentech/Roche, Pharmasset. Dr. Pol declares the following: consulting and lecturing fees from Bristol-Myers Squibb, Boehringer Ingelheim, Vertex, Janssen/Tibotec, Novartis, Gilead, Roche, Schering-Plough/Merck, Abbott, Sanofi and GlaxoSmithKline, and grants from Bristol-Myers Squibb, Gilead, Roche and Merck/Schering Plough.

References

Source

No comments:

Post a Comment