Variation analysis of six HCV viral load assays using low viremic HCV samples in the range of the clinical decision points for HCV protease inhibitors.

In the range of clinical decision points for response-guided therapy of HCV, there is still insufficient data concerning the conformity of quantification results obtained by different assays and their correlation with the HPS/CTM v2 assay which was used for initial clinical studies. In a head-to-head comparison, assay accuracy and detection rates of six quantitative assays [artus HCV QS-RGQ, COBAS Ampliprep/COBAS TaqMan HCV v1/v2, High Pure System/COBAS TaqMan (HPS), RealTime HCV, and Versant HCV1.0] were assessed by measuring WHO and PEI standards at dilution steps near clinical decision points. Detection rates and mean differences between assays were evaluated by analyzing twenty clinical samples at 10, 100, and 1,000 IU/mL. Ten replicates from specimens with different HCV genotypes were used to analyze pan-genotypic intra-assay variation. At ≤ 25 IU/mL, RealTime demonstrated the highest detection rates. With 0.1 log difference when testing clinical samples, results obtained from the Versant and RealTime assays matched best with results from HPS. Mean difference analysis across all assay results revealed wide differences between 0.01 and 0.75 log IU/mL. RealTime showed the lowest intra-assay variation across genotypes 1-4 (25, 100, 1,000 IU/mL). There are substantial analytical differences between viral load assays clinicians should be aware of. These variations may have impact on clinical decisions for patients on HCV triple therapy and may argue for assay-specific decision points equivalent to reference values established in studies using HPS. A comparison of quantification is recommended prior to a switch of assays during ongoing therapy.