Pregnant women infected by Toxoplasma gondii usually have no symptoms and are identifiable only by serologic testing. Many European countries routinely offer prenatal testing so that infected women can receive antibiotics in order to lessen the risk of mother-to-child transmission. If the fetus is already infected, it is believed that prenatal treatment may limit impairment in childhood. After three decades of prenatal screening, however, the efficacy of prenatal treatment remains uncertain. The authors conducted a systematic review of 26 observational cohort studies in which universal maternal screening for congenital toxoplasmosis was used to identify women who seroconverted during pregnancy. Women who had prenatal (fetal) diagnosis were excluded. Meta-analysis of data from individual patients was carried out to detect any relationship between the type and timing of prenatal treatment and the risk of transmission to the fetus.
There were 1721 infected mothers with 506 infected children. The risk of fetal infection increased with gestational age at the time of maternal seroconversion, from 15% (95% CI: 13–17) at 13 weeks, to 44% (95% CI: 40–47) at 26 weeks, and 71% (95% CI: 66–76) at 36 weeks. This represented a 12% increase in the transmission rate per week of gestation at seroconversion. However, the severity of the fetal infection decreased with increasing gestational age at seroconversion. The risk of intracranial lesions declined markedly with seroconversion at advancing gestational ages, while the decrease in risk of ocular lesions was less significant.
Analysis of the effect of prenatal treatment was based on 1438 infected mothers from 18 prenatal screening cohorts who were treated during pregnancy. Infection was confirmed in 398 of their children. There was evidence that fetuses whose mothers were treated within 3 weeks of seroconversion had half the incidence of infection (adjusted odds ratio, 0.48; 95% CI, 0.28–0.80, P = 0.05) than those whose mothers began treatment more than 8 weeks after seroconversion. Neither the type of initial treatment nor whether there was treatment of any kind appeared to influence the clinical manifestations in infected fetuses; the 550 infected live-born infants who were identified by prenatal or neonatal screening had similar clinical manifestations whether they were treated by spiramycin alone, by a combination of pyrimethamine and sulfonamide, or had no treatment. The adjusted odds ratio of any clinical manifestations was not significantly influenced by either of the two treatment regimens, and the gestational age at seroconversion did not influence the outcome of prenatal treatment. No significant treatment effect emerged from intention-to-treat analysis.
These findings provide weak evidence for an association between early prenatal treatment of the mother and a reduced risk of congenital toxoplasmosis. One possible reason is that early treatment does confer protection; another is selective treatment of mothers at high risk of fetal infection who were diagnosed at a relatively late stage. Only a large randomized, controlled clinical trial can provide valid evidence that prenatal treatment of maternal toxoplasmosis benefits the fetus.
Université Bordeaux, Bordeaux, France