In utero exposure to valproate associated with increased risk of impaired cognitive function at age 3 years compared with exposure to other AEDs, data indicate

Children exposed to valproate in utero tend to have a lower IQ at age 3 years than children who are exposed to other antiepileptic drugs (AEDs) do, according to an analysis of interim data from the Neurodevelopmental Effects of Antiepileptic Drugs (NEAD) study. The prospective cohort study included pregnant women who were taking carbamazepine, lamotrigine, phenytoin, or valproate monotherapy. The primary analysis of the study will compare neurodevelopmental outcomes for the children at age...

Children exposed to valproate in utero tend to have a lower IQ at age 3 years than children who are exposed to other antiepileptic drugs (AEDs) do, according to an analysis of interim data from the Neurodevelopmental Effects of Antiepileptic Drugs (NEAD) study.

The prospective cohort study included pregnant women who were taking carbamazepine, lamotrigine, phenytoin, or valproate monotherapy. The primary analysis of the study will compare neurodevelopmental outcomes for the children at age 6 years after in utero exposure to any of the 4 AEDs; the current analysis was a planned interim analysis of outcomes for the children at age 3 years. The researchers who evaluated the neurodevelopmental outcomes were blinded to the children's AED exposure status.

The investigators conducted an intent-to-treat analysis of 309 live births. Among these children, 258 completed testing at age 2 years, age 3 years, or both time points, including 73 children who were exposed to carbamazepine, 84 children exposed to lamotrigine, 48 children exposed to phenytoin, and 53 children exposed to valproate.

In an analysis adjusted for maternal IQ, maternal age at delivery, AED dose, gestational age at birth, and maternal use of folate before conception, the mean IQ scores in the 4 AED-exposed groups ranged from 92 with valproate to 101 with lamotrigine. The mean IQ score in the valproate group was 6 points lower than the mean score in the carbamazepine group (95% CI, 0.6-12.0; P=.04), 7 points lower than the mean score in the phenytoin group (95% CI, 0.2-14.0; P=.04), and 9 points lower than the mean score in the lamotrigine group (95% CI, 3.1-14.6; P=.009). There were no significant differences in IQ scores among the carbamazepine, phenytoin, and lamotrigine groups.

The association between valproate use and IQ score was dose-dependent (P=.005).

The children's IQ scores were significantly associated with maternal IQ scores among the children exposed to carbamazepine, lamotrigine, or phenytoin, but not among the children exposed to valproate, "suggesting that valproate disrupted this normally robust relationship," according to the study authors.

The investigators acknowledged that this was not a randomized trial, but they noted that the association between in utero exposure to valproate and worse cognitive outcome persisted in adjusted analyses, including propensity analyses that considered the probability of receiving a certain treatment given certain baseline characteristics. (Meador KJ, et al. N Engl J Med 2009;360;16:1597-1605.)

New guidelines, which were developed by the American Academy of Neurology (AAN) and the American Epilepsy Society and were presented in Seattle at the AAN's 61st Annual Meeting, suggest that women with epilepsy should avoid taking valproate during pregnancy because of the risk of fetal malformations and impaired cognitive function in offspring.

The guidelines, which were published online April 27 ahead of print in the journal Neurology by Harden CL, et al., also caution pregnant women to avoid phenytoin and phenobarbital as well as AED polytherapy to minimize the risk of adverse outcomes for their children.