Prediction of viremia for cases of hepatitis C virus (HCV) infection using a third-generation anti-HCV enzyme immunoassay test

Background/Aims: Positive results for anti-hepatitis C virus (HCV) testing reveal subjects are infected by HCV, with presence of HCV RNA indicating persistent infection. In this study, we attempted to evaluate the validity of the HCV viremia using a commercially available, third-generation anti-HCV test. Methodology: Sample rate/cut-off rate (S/CO) ratios for 1,907 anti-HCV-positive tests (S/CO >1, AxSYM HCV 3.0; Abbott, IL, USA), which had been performed during the last three years, were retrospectively analyzed. Cases with S/CO values between 1 and 100 were divided into 20 groups according to S/CO range (in increments of 10) and ALT (normal or elevated). Ten random cases were obtained for each of the 20 groups. If cases in any group numbered ≤10, all were recruited. Totally, 193 cases were enrolled for HCV RNA detection (COBAS Amplicor; Roche Diagnostics, NJ, USA). Results: The S/CO distribution was biphasic, with two S/CO peaks in the ranges 1-10 (10.7%) and 81-90 (24.2%). Regardless of the ALT level, all samples with S/COs ≤10 were negative for HGV RNA. Of the samples with S/CO values >10, the optimal cut-off was 40 with sensitivity and specificity for both of 81%. In conclusion, subjects with S/CO values ≤10 (10.7%) were more likely to be cases of past infection or of non-specific reaction. Most (90%, 108/120) of the subjects with S/COs >40 represent current or persistent infection. To predict viremia in subjects with S/COs between 10 and 40, 6.7% of all anti-HCV-positive subjects was invalid by a cross-sectional observation. Conclusions: Follow-up or further study is recommended. The third-generation EIA test plays a semiquantitative role for the prediction of viremia in HCV infection.