Dépistage de l'hépatite C: performances analytiques d'une trousse commerciale automatisée (CMIA-ARCHITECT® anti-VHC)

Abstract

In spite of improvement of the third-generation enzyme immunoassay (EIA) for screening HCV antibodies, some non-specific reactions persist. With commercialisation of a new chemiluminescence microparticle immunoassay (CMIA), we assessed the specificity of 2 assays providing by Abbott Diagnostics: CMIA-ARCHITECT anti-HCV and MEIA-AxSYM HCV 3.0 for qualitative detection of HCV antibodies in serum sample of patients collected in CHU of Caen. Anti-HCV results of 9753 serum samples tested by MEIA-AxSYM V.3 (2004), 6135 tested by CMIA-ARCHITECT1 (April to December 2005) and 5598 tested by CMIA-ARCHITECT2 (February to August 2006) were retrospectively analysed. Prevalences were calculated according to S/C ratio. The serum samples with an average S/C ratio from 1 to 2 for CMIA-ARCHITECT2 were confirmed with an immunoblot assay (Chiron RIBA HCV 3.0 SIA). The CMIA-ARCHITECT assays showed a strong discrimination between negative and positive samples. We observed a tiny distribution of negative results. The percentage of "low positive" was respectively 1.26% for the MEIA-AxSYM, 0.68% for the CMIA-ARCHITECT1 and 0.36% for the CMIA-ARCHITECT2. Thirty-three of 54 (61%) samples yielding S/C ratio between 1 and 2 in the initial screening analysis with the CMIA-ARCHITECT1 were tested negative with CMIA-ARCHITECT2. Among the 21 remaining, 62% of RIBA results were interpretable. CMIA-ARCHITECT assays improve the anti-HCV screening with a decrease of low-positive reactivity. However, low-positive results persist for which it is difficult to distinguish false-positive from low titer of antibodies. Supplemental assays such as immunoblot can be recommended in particularly context to more improve specificity and HCV-RNA detection should exclude a seroconversion.