Impact of IL28B polymorphisms on response to peginterferon and ribavirin in HIV–hepatitis C virus-coinfected patients with prior nonresponse or relapse

Abstract

IL28B polymorphisms predict treatment response in chronic hepatitis C. However, no information exists in prior treatment failures. A total of 62 HIV/hepatitis C virus (HCV) patients who completed retreatment with peginterferon-α/ribavirin were examined, of whom 25 (40%) had been cured. Predictors of response [odds ratio, OR (95% confidence interval, CI)] were HCV genotypes 2/3 [16.1 (2.7–90.9)], prior relapse [9.6 (1.5–62.4)] and ribavirin plasma trough concentrations at week 4 [4.9 (1.3–18.4)]. IL28B-CC only predicted response in prior nonresponders carrying HCV genotypes 1/4 [25.1 (1.9–337)].

While awaiting for the arrival of new direct hepatitis C virus (HCV) antivirals, the accelerated course of liver fibrosis in HIV/HCV-coinfected individuals ^[1] makes treatment of chronic hepatitis C a priority ^[2], being patients who have failed interferon (IFN)α-based therapy in the past no exception. Besides the strong influence of HCV genotypes, the chances of a sustained virological response (SVR) after treatment rechallenge seem to mostly depend on patient's characteristics rather than viral factors, that is, extent of liver fibrosis or ribavirin (RBV) plasma exposure ^[3]. Single nucleotide polymorphisms (SNPs) nearby the IL28B gene are currently known to be strong predictors of response to first-line pegIFNα–RBV therapy in both HCV-monoinfected ^[4–6] and HCV/HIV-coinfected individuals ^[7,8]. At this time, the impact of IL28B variants on treatment rechallenge is unknown.

We have assessed the influence of IL28B rs12979860 SNPs in 62 HIV/HCV-coinfected patients who received a second course of therapy with pegIFNα-2a (180 μg/week) and RBV (1000–1200 mg/day) for 48 weeks, after having failed to suboptimal IFNα-based regimens in the past (i.e., conventional IFNα with/without RBV or pegIFNα and fixed 800 mg/day RBV dosing). Participants without a decline of more than 2 logs in serum HCV-RNA at week 12 or with serum HCV-RNA more than 10 IU/ml at week 24 were considered as virological failures and discontinued therapy ^[3]. Likewise, participants who showed HCV-RNA rebound after discontinuing treatment with undetectable viremia were considered as relapsers. Plasma HCV-RNA was measured using a real-time PCR assay (lower limit of detection of 10 IU/ml). HCV genotyping was performed using a commercial RT-PCR hybridization ^[9]. Plasma trough concentrations of RBV were measured at week 4 using high-performance liquid chromatography (HPLC) ^[10]. The IL28B rs12979860 SNP was examined in peripheral blood mononuclear cells using the 5′ nuclease assay with allele-specific TaqMan probes (ABI TaqMan allelic discrimination kit) and ABI7900HT Sequence Detection System (Applied Biosystems, Carlsbad, California, USA) ^[11].

In the study population, mean age was 43 years, most were men (82%) and were former injection drug users (IDUs; 97%); active alcohol abuse was rare (8%) and most patients were on antiretroviral therapy (94%), with undetectable plasma HIV-RNA (95%) and mean CD4 cell counts of 657 cells/μl. Most participants had serum HCV-RNA levels more than 500 000 IU/ml (73%) and were infected with HCV genotypes 1 or 4 (76%). More than a half of patients had advanced liver fibrosis (53%) and had failed to pegIFNα and low-dose RBV (58%). Nonresponse (63%) was the most frequent type of virological failure to first hepatitis C therapy, being HCV relapse recognized in only 21% of cases. In the remaining 16%, the prior course of therapy had been prematurely interrupted due to toxicity. Overall, 47% of patients had the IL28B rs12979860 CC genotype.

A total of 25 (40%, by on-treatment analysis) attained SVR after completion of pegIFNα–RBV retreatment. Patients who achieved SVR had lower baseline serum HCV-RNA (5.8 vs. 6.2 log IU/ml, P = 0.06) and were less frequently infected with HCV genotype 1 or 4 (48% vs. 95%, P < 0.01) than failures. The likelihood of achieving SVR was significantly greater in prior relapsers than in nonresponders (85% vs. 31%; P < 0.001). Participants carrying IL28B CC more likely attained SVR than non-CC carriers (57% vs. 24%, respectively; P = 0.006). However, when the population was split out according to HCV genotype, the impact of IL28B on SVR was only seen in HCV genotypes 1 or 4 carriers (44% for CC vs. 14% for non-CC, P = 0.02), being not recognized in participants infected with HCV genotype 2 or 3 (82% SVR for CC vs. 100% for non-CC, P = 0.36). Patients who attained sustained HCV clearance had greater mean RBV plasma trough concentrations at week 4 of therapy than patients who failed therapy (2.41 vs. 1.75 μg/ml; P = 0.02). The best discriminatory RBV threshold was 2.0 μg/ml, which displayed a positive predictive value of 69% and a negative predictive value of 70% for SVR (P = 0.02).

Two models for the multivariate analysis were built considering or not RBV plasma trough concentrations at week 4 among the predictors of SVR. HCV genotype 2 or 3, relapse after prior IFNα-based therapy, and RBV plasma concentrations were all associated with SVR (Table 1). Interestingly, the impact of IL28B polymorphisms on SVR was only recognized in the subset of patients more difficult to treat, namely those infected with HCV genotypes 1 or 4 and with true nonresponse to a first course of therapy. In the multivariate analysis for this subpopulation, adjusting for sex, use of antiretroviral therapy, serum HCV-RNA levels, and liver fibrosis staging, the subset of patients carrying the CC genotype had a higher likelihood of response than CT/TT carriers [odds ratio (OR), 25.07 (95% confidence interval, CI 1.86–337), P = 0.01]. Moreover, in these patients, RBV plasma trough concentrations at week 4 did not predict SVR [OR, 2.79 (95% CI 0.58–13.03), P = 0.2].

Table 1: Predictors of sustained virological response in the study population (multivariate analysis).

The finding of a restricted influence of the favorable IL28B genotype in patients with history of true nonresponse instead of relapsers is in line with the recognition by others of a strong association between IL28B rs12979860 SNPs and early viral kinetics on therapy but not with prevention of viral rebound upon completion of treatment ^[12]. In contrast, RBV plasma trough concentrations predicted SVR to re-treatment of hepatitis C in the subset of patients with prior HCV relapse and/or HCV genotypes 2 or 3. In them, RBV trough concentrations more than 2 μg/ml were associated with SVR [OR, 11.67 (95% CI 0.92–147), P = 0.06] with almost statistical significance in the multivariate analysis.

In summary, re-treatment of chronic hepatitis C in HIV–HCV-coinfected patients must ensure optimal RBV exposure, especially in prior relapsers and/or in patients infected with HCV genotype 2 or 3. In contrast, in prior true nonresponders infected with HCV genotype 1 or 4, which is the most prevalent and difficult-to-treat population, optimization of RBV exposure seems to have little impact on SVR, while a favorable IL28B genotype plays a major role in the outcome of re-treatment.

Acknowledgements

The present work was supported by grants from Fundación Investigacion y Educacion en SIDA (IES), the European NEAT project, Red de Investigacion en SIDA (RIS, FIS-RD06/0006), Agencia Laín Entralgo, Instituto de Salud Carlos III (Río Hortega, ref. CM009) and Fundación para la Investigación y la Prevención del SIDA en España (FIPSE, 360799/09). V.S. and P.L. are recipients of intensification grants from Agencia Lain Entralgo, Comunidad Autonoma de Madrid. J.A.P. is recipient of an intensification grant from Fundación Progreso y Salud, Consejería de Salud de la Junta de Andalucía (AI-0021).

All authors declare no conflict of interest.