Article révisé par les pairs
Résumé : Background: Differences in human chorionic gonadotrophin (hCG) results provided by the commercial immunoassays reflect the heterogeneity of antibodies and the use of suboptimal standards. As a consequence, the principal forms of hCG and metabolites are differentially detected and the hCG tests are not suited for the same clinical applications. Conflicting results are available in the literature regarding which hCG variants are recognized by the Roche Elecsys hCG + β test. The aim of our study was to compare the hCG concentrations provided by the Siemens Immulite 2000 test and the Roche test as well as to assess the concordance between both assays. Methods: In this purpose, 152 samples obtained from women and 44 samples from men were analysed by both tests during the follow-up of pregnancy termination, gestational trophoblastic disease and malignancies. The intermediate precision of the Roche test was also investigated on a pool with a low hCG concentration. Results and conclusions: The hCG concentrations measured with the Roche test were slightly lower compared with the Siemens assay; mean biases of −34.2% and −8% were respectively obtained for hCG values ≤100 UI/L and higher than 100 UI/L. The overall agreement between both assays was 96.1% for women and 97.7% for men. By using an upper reference limit of 3.2 UI/L for women and 1.6 UI/L for men, the Roche test demonstrated a respective concordance of 98.7% and 100%. This test also yielded an excellent precision with a coefficient of variation of 2.8% at a mean hCG concentration of 7 UI/L.