Objective: Our purpose was to evaluate different methods of diagnosing congenital toxoplasmosis prenatally by amniocentesis and cordocentesis.
Study design: In a retrospective multicenter study, we investigated consecutive women who had seroconversion for Toxoplasma gondii during pregnancy and who underwent either amniocentesis or cordocentesis or both to obtain a prenatal diagnosis of fetal toxoplasmosis. Data were obtained from 122 patients recruited in 6 different European Toxoplasma reference centers. Infants born to these mothers were followed up until 1 year of age to confirm or exclude congenital toxoplasmosis. Sensitivity, specificity, positive predictive value, and negative predictive value were measured for the following parameters: (1) detection of the parasite in amniotic fluid by mouse inoculation, (2) detection of the parasite in amniotic fluid by in vitro cell culture, (3) detection of Toxoplasma deoxyribonucleic acid in amniotic fluid by a polymerase chain reaction assay, (4) detection of the parasite in fetal blood by mouse inoculation, (5) detection of specific immunoglobulin M antibodies in fetal blood, and (6) detection of specific immunoglobulin A antibodies in fetal blood.
Results: The polymerase chain reaction test performed on amniotic fluid had the highest level of sensitivity (81%) and also a high level of specificity (96%). The combination of the polymerase chain reaction test and mouse inoculation of amniotic fluid increased sensitivity to 91%. The sensitivity of immunoglobulins M and A in fetal blood was 47% and 38%, respectively. In congenitally infected fetuses a negative correlation was observed between positive serologic parameters and gestational age at the time of maternal infection and at prenatal diagnosis.
Conclusion: Congenital toxoplasmosis is best predicted by prenatal examination with the combination of T gondii polymerase chain reaction and mouse inoculation of amniotic fluid. The role of cordocentesis in the diagnosis of congenital toxoplasmosis is limited.