Disclosures: The NEAD study is funded by the National Institutes of Health. Dr. Baker said he had no financial disclosures.
BANGKOK, THAILAND — Expressive and receptive language abilities are significantly poorer in 3-year-olds who were exposed to sodium valproate in utero than they are in children who were exposed to other individual antiepileptic drugs during gestation, based on a subanalysis of the Neurodevelopmental Effects of Antiepileptic Drugs study.
Valproate exposure was associated with a 10-point difference on both language measures compared with exposure to phenytoin, carbamazepine, or lamotrigine—a difference that is not only statistically significant, but clinically important as well, Gus A. Baker, Ph.D., said at the World Congress of Neurology.
The differences apparent in these 3-year-old subjects will likely expand as the groups grow older, said Dr. Baker, director of the division of neurosciences at the Walton Centre for Neurology and Neurosurgery in Liverpool, England. “We can expect the difference in the magnitude to get greater and not smaller with age,” he said. Already, Dr. Baker noted, valproate-exposed 3-year-olds in the U.K. portion of the study are lagging behind a group of matched controls. “Well over a third of those exposed to valproate have been referred for speech therapy, so we see that this 10-point difference has real meaning in terms of day-to-day practice.”
The prospective, observational Neurodevelopmental Effects of Antiepileptic Drugs (NEAD) study included 303 pregnant women who were taking sodium valproate, carbamazepine, lamotrigine, or phenytoin as monotherapy. Enrollment occurred during 1999-2004 in 25 epilepsy centers in the United States and the United Kingdom. Separate investigations in the United States and the United Kingdom were later combined. The primary outcome was cognitive performance of the children at 6 years of age.
Dr. Baker, a primary investigator in the U.K. study and coinvestigator in the overall study, presented the results of the drugs' effect on expressive and receptive language development among 234 children who were 3 years old at assessment. The children all underwent testing of verbal and nonverbal communication, including expressive and receptive language, visual motor construction, and nonverbal intellectual ability.
Their abilities in these areas were determined by calculating subscores on screening tests called the Differential Ability Scales (2nd ed.) and the Preschool Language Scale (4th Ed.). The scores were adjusted for factors known to affect child intellect.
“We saw that maternal IQ, antiepileptic drug (AED) dose, maternal age, gestational age, and preconceptional exposure to folate were significant factors predicting the scores, as we would expect,” he said. “But we also showed that overall, the scores for valproate-exposed children were significantly lower than all other drugs and the magnitude of the effect was greater for verbal than nonverbal language.”
Testing showed that the children exposed to valproate scored significantly lower on measures of expressive language (mean score of 91 vs. 102 for carbamazepine, 104 for lamotrigine, and 101 for phenytoin) and receptive language (mean score of 89 vs. 97 for carbamazepine, 101 for lamotrigine, and 101 for phenytoin). On visual motor construction and nonverbal intellectual ability, children exposed to valproate scored lower, but not significantly lower, than children exposed to the other drugs.
In terms of developmental milestones, this finding could bode ill for the valproate-exposed children, said Dr. Baker. “Without a cohesive and intact language system, a child's neurodevelopmental progress will be limited.”
Unlike the physical results of in utero valproate exposure, which can be surgically corrected to at least some degree, the cognitive effects cannot be erased, he pointed out. The best hope for such children is early identification and intervention. “If we identify them now, we have to think about an appropriate intervention now. If we leave it for later, the gains they might make will be limited.”
“For women for whom sodium valproate is the first choice because of the nature of their seizures, we should be thinking about reducing the dose to the least possible effective level,” he said.
“In an ideal world, we would have preconception counseling and would be thinking of an alternative drug several years before pregnancy.”