Findings from The Neurodevelopmental Effects of Antiepileptic Drugs (NEAD) Study

Yellow2 MONEAD Logo


The predecessor to the Maternal Outcomes and Neurodevelopmental Effects of Antiepileptic Drugs study (MONEAD) was the NEAD study. The NEAD study group was led by Dr. Kimford J. Meador who is also the co-Principal Investigator of the MONEAD study.

The NEAD study was a prospective observational study that enrolled pregnant women with epilepsy receiving monotherapy of carbamazepine, lamotrigine, phenytoin, or valproate over a period of several years (1999-2004). The goal of the investigation was to determine the differential effects of in utero exposure to antiepileptic drugs (AEDs) on children. The data collected from this study led to the publication of over 60 research articles, book chapters, reviews, and abstracts (NEAD Publications). The findings of the NEAD study provided critical insight that caused the Food and Drug Administration (FDA) to amend the warning label for valproate to include the risk of cognitive impairment in children exposed to the drug during pregnancy.

Here we offer an overview of a few key findings of the NEAD study.

Exposure to valproate in utero negatively affects child IQ

Children who were exposed to valproate during pregnancy had lower IQ than children exposed to carbamazepine, lamotrigine, or phenytoin. This effect was analyzed in children at three years of age and at six years of age (Meador et al, 2009 and Meador et al, 2013). IQ was lower in valproate-exposed children at age three years, and this adverse effect persisted through age six years.

The NEAD study team analyzed child cognitive outcomes in 309 children whose mothers were taking antiepileptic drug monotherapy during pregnancy (carbamazepine, lamotrigine, phenytoin, or valproate). Several factors are known to contribute to IQ scores including maternal IQ, maternal age, gestational age at birth, maternal preconception use of folate, socioeconomic status and tobacco or alcohol use. After controlling for these variables, analysis revealed that there is a dose-dependent effect of valproate on child IQ at three years of age. Children exposed to valproate in utero had IQ scores that were six to nine points lower than children exposed to the other antiepileptic medications. This effect was predominantly due to exposure to higher doses of valproate, but a safe dose remains uncertain. Average IQ was indistinguishable between children exposed to carbamazepine, lamotrigine, or phenytoin. The NEAD findings at age three led to a new FDA warning about the cognitive risks of fetal valproate exposure.

Subsequent analysis of 224 children at six years of age again revealed that children exposed to valproate during pregnancy still had lower IQ than children exposed to other medications. IQ scores improved with age across all drug groups, but the children with fetal valproate exposure remained seven to ten IQ points lower than children exposed to the other antiepileptic medications. An additional finding of this study was that children exposed to periconceptional folate had significantly higher IQ than those who were unexposed, corroborating findings from other research studies.

Effect of AED exposure on other cognitive outcomes

In addition to evaluating the impact of fetal AED exposure on IQ, the NEAD study group also specifically investigated the effect of fetal exposure on verbal and non-verbal cognitive abilities (Meador et al, 2011 and Meador et al, 2013). Here again, children whose mothers received monotherapy with carbamazepine, lamotrigine, phenytoin, or valproate during pregnancy were included in this investigation.

Cognitive outcomes at three years of age were calculated from the Differential Ability Scales, Preschool Language Scale, Peabody Picture Vocabulary Test, and the Developmental Test of Visual-Motor Integration.

The data revealed that children exposed to valproate in utero had significantly lower verbal and non-verbal index scores than any other drug group. Verbal abilities were significantly more affected than non-verbal abilities for each of the four antiepileptic medication groups. A dose-dependent effect on both verbal and non-verbal abilities was seen in children exposed to valproate. The carbamazepine group only showed a dose-dependent effect on verbal abilities. Dose did not contribute to the effect seen in either the lamotrigine or phenytoin groups.

Among the findings of this study, it is particularly interesting that the verbal abilities appear to be more negatively impacted than non-verbal abilities. The investigators propose that impairment of cerebral lateralization may be the likely underpinning of this selective effect. Cerebral lateralization most broadly refers to the developmental process through which certain cognitive functions become preferentially housed in one of the two cerebral hemispheres. Lateralization is more important for the development of some cognitive functions than for others. Of relevance in this case is the fact that development of language abilities involves a greater contribution from lateralization compared to the development of non-verbal cognitive function, providing a basis for an explanation for the observed differential effect. It is also possible that other proposed neurodevelopmental effects of AEDs, such as apoptosis or alteration of neuronal physiology, may contribute to this effect but it is difficult to explain why these effects would more specifically target language abilities.

At age six years, reduced verbal to non-verbal abilities were only seen in the valproate group and to a lesser extent for lamotrigine. Interestingly, the proportion of right-handed children in the valproate and lamotrigine groups was significantly lower than the proportion seen in healthy children in the U.S. population. These findings are consistent with our hypothesis that fetal exposure to some antiepileptic medications may alter normal development of cerebral lateralization.

At six years of age, performance by the children was also evaluated for several other cognitive domains including the Children’s Memory Scale (CMS), the Behavior Rating Inventory of Executive Function (BRIEF), the Developmental Neuropsychological Assessment (NEPSY), the expressive one-word picture vocabulary test, and the Developmental Test of Visual Motor Integration (DTVMI). At this age high doses of valproate were negatively associated with verbal ability, non-verbal ability, memory, and executive function. Exposure to other AEDs evaluated did not negatively impact these cognitive measures at six years of age.

Differential effects of antiepileptic drugs on neonatal outcomes

Women with epilepsy have higher risks for several types of birth complications that are attributed to both seizure-related factors and fetal AED exposure during pregnancy.The NEAD study revealed that AEDs impose differential risks for various types of neonatal outcomes (Pennell et al, 2012).

Carbamazepine and valproate are associated with a higher risk of low birth weight

Analysis included the birth weights of 308 neonates whose mothers were taking a single AED (carbamazepine, lamotrigine, phenytoin, or valproate). After accounting for potentially confounding variables (e.g. tobacco use during pregnancy) this analysis revealed that carbamazepine and valproate are associated with a higher risk for a small for gestational age birth when compared to rates associated with exposure to either lamotrigine or phenytoin.

To determine if difference in birth weight was also seen at later ages, children’s weight was compared at two years of age. Indeed, children exposed to either carbamazepine or valproate still had significantly lower weight compared to those who were exposed to lamotrigine or phenytoin. These differences did disappear, however, by three years of age.

Carbamazepine, valproate and phenytoin are associated with greater risk of microcephaly

Microcephaly refers to decreased head size, typically assessed by head circumference measurement. Microcephaly is commonly associated with neurodevelopmental delay. In this analysis, head circumference measurement was collected for children at birth and several additional time points through three years of age.

Analysis revealed that at birth there were no significant differences between AED groups in head circumference. At one year of age, however, both carbamazepine and valproate were associated with smaller head circumference compared to lamotrigine and phenytoin.Head circumference is positively correlated with parental head circumference, so secondary analysis was performed to control for the potential genetic influence on head circumference in children. This analysis revealed that only carbamazepine was associated with a smaller head circumference compared to the lamotrigine group at birth and 12 months old.

In addition to birth weight and head circumference, this analysis also showed additional differences between drug groups with respect to APGAR scores but none of these findings correlated with cognitive performance at three years of age. Taken collectively, these results show that there are differential effects of certain AEDs on neonatal outcomes.

Effects of breastfeeding in children of women taking AEDs

Breastfeeding is known to have many beneficial effects, but it also provides a vehicle for transfer of medication from mother to child. Until the NEAD study, there had been no investigations of the potential risks of AED exposure via breastfeeding on child cognitive outcomes. As in other NEAD investigations, data were collected from cognitive measures of children whose mothers were taking a single AED (carbamazepine, lamotrigine, phenytoin, or valproate). Investigators compared cognitive outcomes in these children as a function of whether or not the children were breastfed or not. Additionally, comparisons were made across drug groups.

The findings of this study revealed that there were no differences in cognitive ability at three years of age between breastfed and non-breastfed children (Meador et al, 2010). Additionally, no significant differences were seen between AED drug groups. Taken together, these findings suggest that, for those AEDs that were studied, there was no adverse effect of exposure via breastfeeding on cognitive function in children already exposed in utero. It is important to remember, as the authors point out, that the findings of this study need to be replicated and that there are limitations of the study that influence interpretation of these findings. These limitations include relatively small sample size, loss of enrolled subjects to analysis, lack of randomization, lack of an unexposed control group during pregnancy, lack of details to fully quantify the amount of breastfeeding, absence of AED concentrations in breast milk or in children’s serum, and relatively young age of the children at time when cognitive performance was measured.

Conclusions

The NEAD study has improved our knowledge of how to care for women with epilepsy during pregnancy and in anticipation of pregnancy. However, many questions remain, which we hope to address in our extension of the NEAD study, which is called Maternal Outcomes and Neurodevelopmental Effects of Antiepileptic Drugs (MONEAD) study because it not only investigates outcomes in children of women with epilepsy, but also will study important issues related maternal outcomes in this population.

 

References