Objective: We sought to develop a score to predict sustained virological response (SVR) in racially diverse HIV/hepatitis C virus (HCV)-coinfected and HCV-monoinfected pegylated interferon/ribavirin-treated patients.
Methods: We retrospectively evaluated 374 patients (259 monoinfected and 115 coinfected) treated at a single tertiary care center. The IL28B rs12979860 single nucleotide polymorphism genotyping was performed in 335 patients, and plasma CXCL10 levels were measured by enzyme-linked immunosorbent assay in 171 patients.
Results: Of the 374 patients, 64.9% were white, 17.2% were African American, 76.5% were HCV genotype 1 infected, and 49.3% had advanced fibrosis. Sustained virological response was achieved by 151 (40.4%) patients, 106 (40.9%) patients monoinfected, and 45 (39.1%) patients coinfected. Patients with IL28B C/C genotype were significantly more likely to achieve an SVR compared with non-C/C genotype patients, but only if they were infected with HCV genotypes 1/4 (59.1% vs 21.1%, P < 0.0001). No significant differences existed in IL28B predictive capacity between coinfected and monoinfected patients. Pretreatment CXCL10 levels were significantly higher in nonresponders, both monoinfected and coinfected, compared with SVR patients (P = 0.0018). Coinfected patients had higher CXCL10 levels compared with monoinfected patients (P = 0.03). The combination of IL28B genotype, pretreatment CXCL10 and HCV RNA levels, and HCV genotype had the best ability to predict treatment response in both patient groups (area under the receiver operating characteristic curve = 0.85). Among all patients, a cutoff score of -0.94 or more had a sensitivity of 0.93 and specificity of 0.59. In coinfected patients, a score of -0.55 or more had sensitivity of 0.81 and specificity of 0.80.
Conclusions: IL28B genotype, pretreatment CXCL10, and HCV RNA levels have very good capacity to predict pegylated interferon/ribavirin-treatment outcome in both HIV/HCV coinfected and HCV monoinfected patients.