Résumé
Objectives - Molecular detection of Toxoplasma gondii plays a crucial role in the prenatal and neonatal diagnosis of congenital toxoplasmosis (CT). Sensitivity of this diagnosis is partly related to the efficiency of parasite DNA extraction and amplification. DNA extraction methods with automated platforms have been developed. Therefore, it is essential to evaluate them in combination with adequate PCR amplification assays. Methods - In this multisite study, we investigated the suitability of two recent automated procedures for the isolation of Toxoplasma DNA from amniotic fluid (AF) (Magtration system 12GC, PSS and Freedom EVO VacS, Tecan), compared with three other automated procedures (MagNAPure Compact, Roche, BioRobot EZ1, Qiagen and modified NucliSens easyMAG, bioMérieux) and with the manual DNA extraction QIAamp DNA Mini kit (Qiagen). Two Toxoplasma PCR assays targeting the '529-bp' repeat DNA element were used, based upon dual hybridization (FRET) or hydrolysis (TaqMan) probes. A total of 1296 PCRs were performed including 972 Toxoplasma PCRs. Results - We showed variable efficacy (4.2%-100% positive results) among the DNA extraction procedures in isolating up to five T. gondii cells/mL in AF samples. Moreover, for a given DNA extraction method, variable results were obtained among the two Toxoplasma PCR assays for detecting up to five T. gondii cells/mL: when using TaqMan PCR, all the automated systems yielded more than 60% positive results. Nevertheless, when testing the DNA extracts in triplicate, four out of six extraction methods allowed a satisfactory detection of low amounts of T. gondii DNA (≥33% of positive results) independently of the PCR assay used. Conclusions - Despite the influence of the subsequent PCR method used, this study should help microbiologists in the choice of DNA extraction methods for the detection of T. gondii in amniotic fluid. The extraction method should be checked as adequate for the PCR assay used.