The records of all pregnant Medicaid recipients in Michigan who were taking anti-epileptic agents during pregnancy were studied at the Center for Drug Evaluation and Research, Food and Drug Administration, Rockville, Maryland. Four cases of spina bifida were identified in infants born to 1,490 women who took anticonvulsants during pregnancy. Among 107 women taking carbamazepine three gave birth with spina bifida; one was taking valproic acid in addition and two were taking phenytoin, barbituates or primidone in combination with carbamazepine. Among 1,018 infants of mothers taking barbiturates alone only one had spina bifida. Among 444 women taking phenytoin and 50 taking primidone alone, none gave birth to an infant with spina bifida. When the data from approximately 20 published studies were pooled, nine cases of spina bifida occurred among a total of 984 in utero exposures to carbamazepine, and the relative risk was 13.7 times the expected rate (approximately one in 1,500 births). Five cases of spina bifida occurred in infants whose mothers took carbamazepine alone during pregnancy. While exposure to carbamazepine in utero carried a 1% risk of spina bifida the risk with valproic acid was close to 2% but the difference was not significant. The risks with other anticonvulsant drugs was only 0.14% or one in 748 cases. [1]

COMMENT. The incidence of meningomyelocele has declined steadily from 12 per 10,000 in 1970 to 6 to 8 per 10,000 in 1980 among babies born in the United States. The neural tube closes between 22 and 29 days after conception and exposure to anti-epileptic agents such as valproic acid and carbamazepine may interfere with closure. Meningomyelocele may be diagnosed in the fetus with 75 to 80% accuracy by measurements of serum alpha-fetoprotein levels, and ultrasonography can identify the extent and location of these defects. [2]

The outcome of a neural tube defect may be improved by cesarean section before labor begins [3]. Infants who had been exposed to labor were 2.2 times more likely to have severe paralysis than those delivered by cesarean section without labor. The mean functional motor level was at the second lumbar vertebra in fetuses exposed to labor and at the fourth lumbar vertebra in those delivered before labor began. Cesarean section before onset of labor may result in better subsequent motor function and the ability to walk with minimal mechanical assistance.