Purpose: To quantify the risks of intrauterine antiepileptic drug (AED) exposure in monotherapy and polytherapy.
Methods: Data from five prospective European studies totaling 1,379 children were pooled and reanalyzed. Data were available for 1,221 children exposed to AED during pregnancy and for 158 children of unexposed control pregnancies.
Results: Overall, when comparing a subgroup of 192 children exposed to AED with 158 children of matched nonepileptic controls, there was an increased risk of major congenital malformations (MCA) in children exposed to AED during gestation [relative risk (RR) 2.3; 95% confidence interval (CI): 1.2-4.7]. A significant increase in risk was found for children exposed to valproate (VPA) (RR 4.9; 95% CI: 1.6-15.0) or carbamazepine (CBZ) (RR 4.9; 95% CI: 1.3-18.0) in monotherapy. When comparing different AED regimens during all 1,221 pregnancies, risks of MCA were significantly increased for the combination of phenobarbital (PB) and ethosuximide (RR 9.8; 95% CI: 1.4-67.3) and the combination of phenytoin, PB, CBZ, and VPA (RR 11.0; 95% CI: 2.1-57.6). Offspring of mothers using > 1,000 mg VPA/day were at a significantly increased risk of MCA, especially neural tube defects, compared to offspring exposed < or =600 mg VPA/day (RR 6.8; 95% CI: 1.4-32.7). No difference in risk of MCA was found between the offspring exposed to 601-1,000 mg/day and < or =600 mg/day.
Conclusions: This reanalysis shows that VPA is consistently associated with an increased risk of MCA in babies born to mothers with epilepsy. Significant associations were also observed with CBZ. Larger prospective population-based studies are needed to evaluate the risks of many other less frequently prescribed treatment regimens, including newly marketed AEDs.