TOPLINE

In a large cohort study, prenatal exposure to antiseizure medication was associated with poor fetal growth. The risk was greatest with polytherapy, which was linked to a lower birth weight centile and higher risks for small for gestational age (SGA), severe SGA, and low birth weight compared with monotherapy.

METHODOLOGY

• Researchers conducted a prospective, observational, longitudinal cohort study using data from the International Registry of Antiepileptic Drugs and Pregnancy between June 1999 and November 2023.
• They included 15,893 singleton pregnancies in women aged 14-55 years with epilepsy who were exposed to antiseizure medications from conception to delivery; 12,911 were exposed to monotherapy and 2982 to polytherapy.
• The primary outcome was birth weight centile; secondary outcomes included SGA (birth weight below the 10th centile for gestational age), severe SGA (birth weight below the third centile), and low birth weight (under 2500 g).
• Researchers assessed whether prenatal exposure to antiseizure medication was associated with poor fetal growth, and examined how that risk varied between monotherapy and polytherapy, across individual medications and combinations, and by dose. 
• Follow-up data were acquired after each trimester and at delivery to track medication exposure and birth outcomes.

TAKEAWAY

• Exposure to polytherapy was associated with lower birth weight centile (adjusted unstandardized model coefficient [b], -2.74; 95% CI, -4.22 to -1.26) and increased risk for SGA (adjusted odds ratio [OR] 1.48; 95% CI, 1.29-1.70), severe SGA (OR 1.49; 95% CI, 1.22-1.83), and low birth weight (OR 1.50; 95% CI, 1.15-1.95) compared with monotherapy.
• Birth weight centile decreased with increasing number of concomitant antiseizure medications (adjusted b, -14.18; 95% CI -24.07 to -4.29, for four medications relative to monotherapy).
• Among monotherapies, compared with lamotrigine, topiramate showed the greatest reduction in birth weight centile (adjusted b, -11.93; 95% CI -16.72 to -7.15), followed by phenobarbital (b, -8.08; 95% CI -11.89 to -4.26), oxcarbazepine (b, -5.10; 95% CI -8.24 to -1.96), carbamazepine (b, -3.15; 95% CI -5.01 to -1.28), valproic acid (b, -2.54; 95% CI -4.50 to -0.58), and levetiracetam (b, -2.51; 95% CI -4.38 to -0.63).
• Carbamazepine combined with levetiracetam was associated with a lower birth weight centile (adjusted b, -6.27; 95% CI -12.03 to -0.50) compared with lamotrigine monotherapy, while lamotrigine combined with either levetiracetam or valproic acid showed no significant difference from lamotrigine monotherapy.
• Oxcarbazepine showed a dose-response pattern, with higher doses linked to a greater risk for SGA.

IN PRACTICE

"Consideration of these risks and their association with the type of antiseizure medication exposure should, similar to major congenital malformations and neurodevelopmental disorders, be part of routine preconception counselling and clinical decision making. For women with epilepsy, any change in antiseizure medication therapy aimed at protecting fetal health should generally be done preconception and take into consideration not only maternal and fetal risks related to antiseizure medication exposure, but also those associated with potential loss or worsening of seizure control," the authors of the study concluded.

SOURCE

The study was led by Piero Perucca, MD, Department of Medicine (Austin Health), University of Melbourne, Melbourne, Australia. It was published online  in September 2026 in The Lancet Neurology.

LIMITATIONS

The study could not fully account for all potential confounders affecting fetal growth, including maternal weight or nutritional status. Generalizability was limited due to underrepresentation of some ethnic groups.  The study also did not include cohorts exposed to antiseizure medications for indications other than epilepsy.

DISCLOSURES

The study was supported by Brain Australia, Neurological Foundation of New Zealand, Norman Beischer Medical Research Foundation, and Weary Dunlop Medical Research Foundation. Multiple authors reported receiving speaker honoraria, consultancy fees, conference travel support, or research funding from various pharmaceutical companies including Eisai, UCB Pharma, Novartis, Biogen, Merck, and others. Several authors hold editorial positions for medical journals. Additional author disclosures are reported in the original article.

This article was created using several editorial tools, including AI, as part of the process. Human editors reviewed this content before publication.