Neurodevelopmental Teratogenicity of Antiseizure Medications: The Good, the Bad, and Unknown

Abstract

Maternal Outcomes and Neurodevelopmental Effects of Antiepileptic Drugs (MONEAD) Investigator Group. Neuropsychological Outcomes in 6-Year-Old Children of Women With Epilepsy: A Prospective Nonrandomized Clinical Trial

Meador KJ, Cohen MJ, Loring DW, Matthews AG, Brown C, Robalino CP, Carmack A, Birnbaum AK, Voinescu PE, Gerard EE, Kalayjian LA, Gedzelman ER, Hanna J, Cavitt J, Sam M, Hwang S, Pack AM, French JA, Tsai JJ, Taylor C, Pennell PB. JAMA Neurol. 2025 Jan 1;82(1):30-39. doi:10.1001/jamaneurol.2024.3982.

Importance: Antiseizure medications (ASMs) are potential teratogens commonly prescribed for multiple indications. ASM fetal exposure can impair neurodevelopment. Folate improves pregnancy outcomes, but higher doses may pose risks. Objectives: To compare the outcomes of 6-year-old children of women with epilepsy (WWE) vs those of healthy women (HW), and assess the association of outcomes to third-trimester ASM exposures. Design, Setting, and Participants: After informed consent, pregnant WWE and HW were enrolled from 2012 through 2016 in this prospective, multicenter, nonrandomized clinical trial. Children were assessed at 6 years of age (2019–2022). Participants were recruited from 20 US epilepsy centers. Study data were analyzed from August 2023 to August 2024. Exposures: Fetal ASM exposures. Main Outcomes and Measures: The a priori main neurodevelopmental outcome was the blindly assessed Verbal Index Score in 6-year-old children. The Verbal Index Score is calculated as the mean of the scores from the Word Definitions and Verbal Similarities subtests from the Differential Ability Scales, Expressive One-Word Picture Vocabulary Test, Phonological Processing, Comprehension of Instructions, and Sentence Repetition subtests from the Neuropsychological Assessment and Peabody Picture Vocabulary Test. The 2 primary analyses (1) compared children of WWE and HW using linear regression and (2) examined the outcomes of fetal exposure via ASM blood concentrations. Analyses were adjusted for multiple potential confounding factors. Other outcomes and folate exposure-related outcomes were assessed. Results: A total of 1123 pregnant women were screened, and 456 were enrolled (426 did not meet criteria, and 241 chose not to participate). A total of 298 children of WWE (mean [SD] age, 6.4 [4.2] years; 158 female [53.0%]; 140 male [47.0%]) vs 89 children of HW (mean [SD] age, 6.4 [4.2] years; 41 female [46.1%]; 48 male [53.9%]) did not differ on Verbal Index Score (parameter estimate, −0.6; 95% CI, −3.2 to 1.9; P = .64). Exposure-dependent outcomes differed across ASMs. Assessment of other ASMs was limited because 232 of 298 WWE (78%) were taking lamotrigine or levetiracetam alone or in combination. Folate supplementation during the first 12 weeks of pregnancy had positive associations with cognition and behavior with no signal for risks at higher folate doses. Conclusions and Relevance: Results of this prospective nonrandomized clinical trial suggest that verbal abilities in children of WWE vs HW did not differ. Exposure-dependent outcomes of ASMs highlight the importance of dosing high enough to protect the mother and fetus from seizures but low enough to protect the fetus. Folate supplementation early in pregnancy including higher doses was associated with improved cognitive and behavioral outcomes. Additional research is needed for ASMs with inadequate information on fetal exposure risks.

Commentary

The Maternal Outcomes and Neurodevelopmental Effects of Antiepileptic Drugs (MONEAD) study is a National Institutes of Health–National Institute of Neurological Disorders and Stroke (NIH–NINDS) funded, prospective, observational, multicenter investigation of pregnancy outcomes for mothers and their infants.

The MONEAD cohort comprises pregnant women with epilepsy (WWE) and healthy pregnant women (HW) who were enrolled at 20 USA epilepsy centers between 2012 and 2016. The study objectives were to compare the outcomes of 6-year-old children of WWE versus those of HW, and to assess the association of outcomes to third-trimester antiseizure medications (ASM) exposures. The focus on the third trimester period was driven by the fact that it has maximum brain growth and synaptogenesis during human pregnancy.

Although other large and similarly impactful pregnancy registries exist worldwide, the North American Anti Epileptic Drug (NEAD) registry (https://www.aedpregnancyregistry.org/) and its MONEAD study set a landmark in the life and care of WWE. They brought unprecedented information about fetal exposure to ASMs and their role in early childhood neurodevelopmental and behavioral outcomes.

MONEAD prospectively evaluated offspring of WWE at different ages and, to date, data at ages 2-year-old, ¹ 3-year-old, ² 4.5-year-old, ³ and 6-year-old ⁴ from children born after intrauterine exposure to ASMs have been published. The information at age 6 is likely to show more robust information, as cognitive abilities at this age impact early education and might be predictive of adult outcomes. In the past, the NEAD study had previously highlighted the negative impact of valproate on creativity and executive function. ⁵ In contrast, it also demonstrated that newer ASMs did not negatively impact these measures at age 4.5-year-old. ³

The decision to become a parent is not always easy for WWE, particularly for those with drug-resistant epilepsy. There is a fear of exposing the fetus to the teratogenic risks of ASMs that looms for the duration of gestation. There is apprehension that experiencing a seizure during pregnancy can cause injury to the developing fetus. There is anxiety about possible complications during labor and delivery. There is fear of being incapable of safely caring for the baby after birth because of seizures. And there is hesitation about breastfeeding the newborn, knowing that ASMs make their way to the breast milk.

For healthcare providers, navigating safe pregnancy planning with their patients requires timely moves, including patient empowerment and education. Information from pregnancy registries allows informed decisions to be made early, much before our patients try to conceive. This is important in the transition to adult care for patients with pregnancy potential for childhood-onset epilepsy, and it is pivotal in discussions on first elected ASM for adult-onset seizures.

With information regarding the behavioral and neurodevelopmental teratogenicity of ASMs, a healthy-appearing baby at birth no longer brings the ultimate peace of mind in reassurance. Additional data arose that behavioral and cognitive teratogenicity related to ASMs was also a risk for their offspring. Therefore, despite the limited number of mother–child pairs, the MONEAD cohort observations that at least certain ASMs were associated with a healthy and normal neurodevelopment are of ultimate importance for pre-pregnancy counselling of WWE.

Verbal abilities, defined as the primary outcome measure, did not differ significantly between children of WWE and those of HW, after adjusting for maternal and child factors.

Analyses of ASM exposure by maternal serum ASM levels showed that higher Lamotrigine (LTG) levels in the third trimester were linked to slightly better verbal scores. It might be speculated that higher levels indicate proper dose up titration during pregnancy, and lower risk of seizure recurrence, which could also contribute to a healthier prenatal brain development.

ASM-related neurodevelopmental teratogenicity likely mirrors dose-related observations for their anatomical teratogenicity. It has been demonstrated that anatomical teratogenicity risk can be dose dependent for valproate and topiramate, and further increases with polytherapy.^6,7 In addition, maternal history of neurodevelopmental impairments can be associated with an increase in the odds of neurodevelopmental teratogenicity in the offspring, suggesting a genetic predisposition or an individual's susceptibility to the teratogenic effects of ASMs. ⁸

In a recent systematic review, ⁹ including data from 43 heterogeneous studies (population-based, prospective cohort, retrospective cohort, and cross-sectional observational), the highest risk of adverse neurodevelopmental outcomes was confirmed for valproate, in addition to growing evidence for topiramate as well. Offspring from WWE taking valproate during pregnancy have a 2- to 4-fold increased risk of autism spectrum disorder (ASD), 2- to 5-fold increased risk of intellectual disability (ID), and 1.5-fold increased risk of attention deficit hyperactivity disorder (ADHD), as well as impaired adaptive functioning and social skills. Topiramate has been associated with a 2-fold increased risk of ASD, a 3- to 4-fold increased risk of ID, and more than a 2-fold increased risk of ADHD. Outcomes following carbamazepine (CBZ) and phenobarbital exposures showed variable results, whereas exposure to LTG, phenytoin, oxcarbazepine, and levetiracetam seems to carry a low risk.

However, differences might be noticeable later in life, and these studies are still needed. Danish adolescents with intrauterine exposure to CBZ performed poorer in standardized scores at ninth grade (age 16-year-old) final examinations (Danish and mathematics) as compared to unexposed individuals, of a magnitude comparable to that found for valproate. ¹⁰ A subanalysis for CBZ polytherapy demonstrated an even more prominent decrease in academic performance within this cohort. Noting, however, that many ASMs are also used to treat conditions other than epilepsy, the clinical indication for treatment can be a confounding factor. Excluding children of women taking the ASM for other reasons than epilepsy, the association between prenatal CBZ exposure and academic performance was no longer statistically significant.

Studies on the effect of ASM exposure on offspring neurodevelopmental outcomes should control for maternal factors, including indication, account for multiple comparisons, and, being mostly exploratory or observational, replication is needed for all ASMs. Importantly, long-term outcome data are still missing for most ASMs (ie, newer ASMs), and data on children at a later age remain scarce. Lack of data does not equal safety, and this needs to always be clearly explained during pregnancy planning.

Ultimately, physicians caring for WWE in reproductive age have a big responsibility to counsel their patients. As reviewed by Raine, ¹¹ under the ethical principle of beneficence, the healthcare provider will act in the best interest of the patient while also exercising the duty of nonmaleficence (the ethical obligation to minimize harm) and, ideally, bringing a net benefit to the patient. Potential fetal effects may be considered as part of the overall balance of risks and benefits, but this is an ongoing discussion that needs to take place at each follow-up, for joint and safe informed decisions, in particular when seizure control is only attained with less favorable ASMs. These ongoing discussions depend on studies such as the MONEAD to report on updated information on neurodevelopmental and anatomical teratogenicity of ASMs.