Objective: To assess the ability of the cirrhosis risk score (CRS) to predict liver fibrosis progression in HIV/hepatitis C virus (HCV)-coinfected patients.
Design: Retrospective follow-up study.
Methods: Based on a minimum follow-up time of 10 years with HCV infection, 190 HIV/HCV-coinfected patients were classified according to their METAVIR score: (1) 25 nonprogressor patients who did not develop fibrosis (F0) and (2) 165 progressor patients who developed fibrosis (F ≥ 1). Seven polymorphisms of CRS signature and IL28B genotype were performed using the GoldenGate assay. The CRS signature was calculated by naive Bayes formula as previously described.
Results: Nonprogressors had CRS values significantly lower than progressors (0.61 versus 0.67; P = 0.043). Among the progressors, we observed similar CRS values through all the fibrosis stages (F1/F2/F3/F4). The percentage of patients with CRS > 0.70 (high risk of developing fibrosis) was higher in progressors than in nonprogressors; but the percentages with values between 0.50 and 0.70 (intermediate risk) and <0.50 (low risk) were quite similar for each of the fibrosis stages (P = 0.047). The area under the receiver-operating characteristic curve of CRS for discriminating nonprogressor versus progressor was 0.625 (P = 0.043). When clinical variables were considered (age at HCV infection, intravenous drug use, gender, IL28B, and HCV genotype), the area under the receiver-operating characteristic curve of CRS improved up to 0.739 (P < 0.001).
Conclusions: CRS itself seems not to be a good marker for identifying HIV/HCV-coinfected patients who are at high risk of developing liver fibrosis. However, CRS score coupled with clinical factors might help to distinguish between nonprogressors and progressors patients.