Fetal effects of anticonvulsant polytherapy exposure: Different risks from different drug combinations

Background: The fetal risks from prenatal exposure to anticonvulsant drugs have been considered greater when the treatment includes two or more drugs (polytherapy). Experience with the fetal effects of monotherapy has shown significant differences in the risk of malformation among anticonvulsant drugs.

Methods: 6,719 pregnant women have enrolled in the North American AED (antiepileptic drug) Pregnancy Registry since 1997. Information on AED use, demographic characteristics, vitamin use, medications and recreational drugs was collected in three telephone interviews. The diagnosis of malformations in infants up to 12 weeks of age was based on the mother's report and confirmed by medical records. The comparison group was pregnant friends and family members (not taking AED) of enrolled women. The relative risk (RR) and 95% confidence interval (CI) of major malformations for the most common polytherapies were determined.

Results: The most common polytherapy combinations included either carbamazepine (n=344) or lamotrigine (n=461). The risks of malformations were 1.7% for non-exposed control infants, in comparison to 15.8% for carbamazepine plus valproate exposed (RR 9.5; 95% CI 3.2-27.9) and 9.4% for lamotrigine plus valproate exposed infants (RR 5.7; 95% CI 1.8-17.9). In contrast, there was no significant increase for infants exposed to either carbamazepine (n= 306; RR=1.2; 95% CI 0.38-3.6) or lamotrigine (n=408; RR= 1.9; 95% CI 0.73-5.0) plus non-valproate anticonvulsant drugs.

Conclusions: Counseling women about the fetal risks from AED polytherapy must be based on the specific drugs included. Combinations which include valproate pose a higher risk than those without this drug.