Identifying an Appropriate Drug for Comparative Safety Research

Antiepileptic Drugs in Pregnancy: Searching for a Reference Drug for Comparative Safety

AV Margulis, AS Oberg, S Hernandez-Diaz. Poster presented at the 33rd International Conference on Pharmacoepidemiology & Therapeutic Risk Management (ICPE); August 29, 2017. Montreal, Canada.


Around 0.5% of women in Europe and 2% of women in the US use antiepileptic drugs during pregnancy. It is important to know how safe these meds are.

To explore the comparative safety of these drugs, one would compare one antiepileptic drug to another. Currently, it is not clear which antiepileptic drug should serve as a reference in safety studies. The reference drug should have the same indication as other drugs that are of interest (epilepsy and mood disorders, in this case), be commonly used and have, ideally, a benign safety profile.

We conducted a literature review to identify an appropriate reference antiepileptic drug for comparative safety research in pregnancy. In this literature review, we identified 872 published papers that evaluated antiepileptic drugs, and preterm delivery or neonatal size (for example, birth weight).

From those, we identified 11 that provided information relevant to our question, and 3 that were similar enough to each other to enable us to conduct appropriate comparisons.

This figure shows results for preterm delivery.

The orange dots represent the study results; the black horizontal bars represent the 95% confidence interval. Bigger dots reflect larger numbers of patients in the study. Our literature review shows quite clearly that the use of carbamazepine in pregnancy increases the risk for preterm delivery in comparison to no use of antiepileptic drugs.

For valproic acid, results are less clear. Overall, they seem to indicate that there might be an increased risk for preterm delivery.

Results for lamotrigine are not crystal clear either, but overall, they seem to indicate that there might be less risk associated with the use of this drug.

The results for low birth weight are similar to results for preterm delivery: carbamazepine seems to be the most harmful of the three drugs, with less clear results for valproic acid and lamotrigine.

It has been known for decades that valproic acid carries an increased risk for birth defects and, more recently, for cognitive problems, for infants exposed in utero.

Taking all of this into consideration, we concluded that lamotrigine is an appropriate reference drug for use in future comparative safety research.

–We are applying these results already in our next pregnancy safety study.