For more than 20 years, the risk of neural tube defects (NTDs) associated with first-trimester exposure to valproic acid has been well known: The estimated risk is 2%, about 10− to 20-fold higher than the baseline risk. With more widespread use of valproic acid, partly due to increasing use of the drug for psychiatric conditions, more data and larger controlled studies on its teratogenic effects have accumulated over the past 3–4 years, revealing an association with major malformations that previously had been reported anecdotally. The main anomalies that have been identified are cardiac and limb malformations.
The results of these studies reflect what we found in a metaanalysis of data from 13 cohort studies in the medical literature, published between 1982 and 2005. The studies in the metaanalysis, published in April in a Canadian journal, compared rates of major malformations among women who reported taking valproic acid during the first trimester with rates among pregnant women who were taking other antiepileptic drugs (AEDs) and among women who were not taking any such drugs.
Nearly 1,000 pregnant women were exposed to valproic acid in the 13 studies. The risk of major malformations, including NTDs, associated with exposure to valproic acid was twofold greater than the risk with exposure to other AEDs. The risk was 4.4-fold greater than in the healthy controls, representing a highly significant increase in risk among valproic acid-exposed pregnancies.
We could not include three studies comparing the neurobehavioral risks of in utero exposure to valproic acid and other AEDs in the metaanalysis, because of their different designs and the variety of cognitive tests used. Still, all three reported an association between valproic acid and developmental delays and cognitive deficits. The most prominent effect was on verbal IQ. More studies on the neurodevelopmental effects of in utero exposure that control for maternal education and other confounding factors need to be conducted to further examine these associations.
On the positive side, in 3 of the 13 studies that also looked at the dose-dependent effects of valproic acid, the threshold dose to cause malformations was about 1,000 mg/day, which has been reported over the past few years. This is true for all malformations associated with valproic acid, including NTDs. In one study, first-trimester valproic acid plasma levels in women were higher among those who had a child with a malformation; in another, a daily dose of 1,000 mg was associated with a significantly increased risk for major malformations, especially NTDs; and in the third, mothers who had a child with spina bifida were on a mean dose of 1,640 mg/day vs. a mean of 941 mg/day among those whose children had no malformations. The same studies indicated that at less than 600 mg/day, there was no increased risk.
These relatively new findings of major malformations other than NTDs and the potentially increased risk of cognitive effects of valproic acid are important for women and physicians to consider when women are planning a pregnancy. Sometimes, women who have been on a drug for epilepsy for many years may no longer need it. If a switch to another drug is not possible, patients need to be monitored closely for malformations, as has been the practice for NTDs, although there is no way to monitor for potential cognitive effects.
Another approach for women who are dependent on valproic acid is to make every effort to keep the daily dose at 600 mg or below or, if that is not feasible, under 1,000 mg/day. Patients should be monitored with ultrasound, fetal echocardiography, and maternal and amniotic α-fetoprotein testing.
Evidence that reducing the dose can effectively reduce malformations came in a report last year from the Australian pregnancy registry for women on AEDs, which found the risk of fetal malformations was 13 times higher among women taking more than 1,100 mg of valproic acid per day as monotherapy, compared with women not taking an antiepileptic drug. Although the fetal malformation rate among those on lower doses was greater than the 2%–3% risk in the general population, the difference was not significant.
If possible, a different medication for controlling seizures should be considered. Carbamazepine (Tegretol) is considered by many neurologists and obstetricians to be the AED of choice in pregnancy, because the cumulative data to date do not reveal any risks of major malformations, except for spina bifida at about 1%, which is half the rate associated with valproic acid.
However, there are far fewer data on the reproductive risks of newer antiepileptics such as lamotrigine and gabapentin. Of the newer drugs, lamotrigine seems to be the most promising in terms of adverse fetal outcomes, but the number of pregnancies with data is much smaller than is available with valproic acid and carbamazepine.