A diagnostic score for the prediction of spontaneous resolution of acute hepatitis C virus infection

Abstract

IL28B polymorphisms, jaundice, decline in HCV-RNA, IP-10, and gender have been proposed to be indicative of spontaneous clearance of acute hepatitis C virus infection. The aim of this study was to define a score enabling the discrimination of patients with spontaneous clearance of HCV from those with development of viral persistence and need for early antiviral treatment. 136 patients (74 male; 35 ± 15 years) were analyzed. From variables predictive of spontaneous clearance, calculated by univariate analysis, three scores were built. Analogous cut-offs were evaluated by computing area under the receiver operating characteristic curves. Candidate variables and cut-offs were: (I) presence of IL28B C/C (p=0.027), (II) age (p=0.031; cut-off: 35 years), (III) peak-bilirubin (p=0.018; cut-off: 6 mg/dl), (IV) HCV-RNA decline within 4 weeks (p<0.001;cut-off: >2.5 log), (V) serum IP-10 (p=0.003; cut-off: 546 pg/ml), (VI) presence of CD4(+) Th1 cells (p=0.024). Each variable was allocated to 0 or 1 point, an HCV-RNA decline of ≥ 1 log 10 but <2.5 log 10 to 1 point, a decline of ≥ 2.5 log 10 to 2 points. Three scores were evaluated (Score 1: I-IV; Score 2: I-V; Score 3: I-VI). A cut-off of ≥ 3 points out of 5 in Score 1 (AUROC: 0.82; DeLong 95% CI: 0.76-0.93) predicted spontaneous clearance with a sensitivity of 71% (95% CI: 0.53-0.86) and specificity of 87% (95% CI: 0.73-0.95). PPV and NPV were 79% and 82%. Corresponding findings for Score 2 including IP-10 (AUROC: 0.93; DeLong 95% CI: 0.86-0.93) at a cut-off of ≥ 4 were: sensitivity 81%, specificity 95% (PPV: 100%; NPV: 77%). A cut-off of ≥ 5 in Score 3 (AUROC: 0.98; DeLong 95% CI: 0.95-1.0) predicted spontaneous resolution with a sensitivity of 75% and specificity of 100% (PPV: 100%; NPV: 88%). The scores enable a reliable discrimination between AHC-patients with high potential for spontaneous clearance from candidates for early therapeutic intervention due to marginal chance of spontaneous resolution.