Analytical performance of the VERIS MDx system HCV assay for detecting and quantifying HCV RNA

Abstract

BackgroundThe diagnosis of HCV relies on the detection of viral RNA. ObjectiveTo evaluate the performance of the VERIS/MDx System HCV Assay, a new automated system for quantifying HCV RNA, and to compare with the COBAS® Ampliprep/COBAS® Taqman™ (CAPCTM) HCV Test version 2.0. Study designThe limit of detection was determined by Probit analysis with the 3rd International WHO HCV standard and precision by assaying in duplicate control samples with HCV RNA concentrations of 7.9; 5.0; 3.4; 1.6 and 0logIU/ml over 20 days. Analytical specificity was assessed by assaying 180 samples from negative anti-HCV and HCV RNA blood donors and linearity with replicates of serial dilutions of a clinical plasma (6.4–0.6logIU/ml). We compared the VERIS MDx HCV and CAPCTM HCV assays by testing 209 samples. ResultsThe limit of detection was 6.1IU/ml [CI 95%: 5.0–8.3] and the precision, given by the standard deviation, was ≤0.11logIU/ml. Specificity was 100%. The linearity ranged from 1.5 to 6.4logIU/ml. Passing-Bablok regression analysis gave: VERIS logIU/ml=−0.33+[1.04× CAPCTM] logIU/ml, with biases for the 25th, 50th, 75th percentiles of 0.18, −0.10 and −0.06logIU/ml. The two assays were well correlated (ρ=0.92, p<0.001) and Bland-Altman analysis gave biases of 0.12, logIU/ml for genotype 1, −0.19 for genotype 2, −0.26 for genotype 3, and −0.77 for genotype 4. ConclusionThe VERIS MDx HCV assay performed well. But, we observed an under-quantification of the genotype 4 samples.