← Back
NeuroTrials.ai
Neurology Clinical Trial Database

North American AED Pregnancy Registry

Use of Antiseizure Medications Early in Pregnancy and the Risk of Major Malformations in the Newborn

Share Share Copied! Compare

Year of Publication: 2025

Authors: Sonia Hernandez-Diaz, Moira Quinn, Susan Conant, ..., Lewis Ball Holmes

Journal: Neurology

Citation: Neurology. 2025;105:e213786.

Link: https://www.neurology.org/doi/full/10.12...000000000213786

PDF: https://www.neurology.org/doi/pdf/10.121...000000000213786

Clinical Question

In pregnant women taking antiseizure medications as monotherapy during the first trimester, what is the risk of major congenital malformations for each specific ASM compared to lamotrigine?

Bottom Line

Among 7,311 pregnant women on ASM monotherapy (1997-2023), valproate (9.2%), phenobarbital (6.0%), and topiramate (5.1%) carried highest malformation risks vs lamotrigine (2.1%). RR vs lamotrigine: valproate 4.35, phenobarbital 2.84, topiramate 2.41. Levetiracetam (2.0%), oxcarbazepine (1.5%), gabapentin (1.5%), zonisamide (1.3%) were not elevated. Lacosamide: 0/88 malformations. Topiramate specifically increased oral clefts (14/1000 vs 1/1000 reference).

Major Points

  • Valproate highest risk: 9.2% (31/337); RR 4.35 vs LTG (95% CI 2.83-6.69). Dose-dependent (higher >650 mg/day).
  • Phenobarbital: 6.0% (12/200); RR 2.84 vs LTG (1.54-5.23). Associated with cardiovascular defects and oral clefts.
  • Topiramate: 5.1% (26/510); RR 2.41 vs LTG (1.52-3.83). Specific signal for oral clefts: 14/1000 (vs 1/1000 reference). SGA in 19%.
  • Lamotrigine (reference): 2.1% (52/2,461). Levetiracetam: 2.0% (26/1,283); RR 0.96 vs LTG — no excess risk.
  • Oxcarbazepine 1.5%, gabapentin 1.5%, zonisamide 1.3% — all similar to LTG (RR 0.62-0.72).
  • Lacosamide: 0/88 in monotherapy (0%; 95% CI 0-5.2%). Reassuring but imprecise.
  • Carbamazepine 2.8%, phenytoin 2.8% — modest elevation vs LTG (RR ~1.34, CIs include 1.0).
  • Prospective registry with blinded dysmorphologist adjudication. 7,311 monotherapy exposures over 26 years.
  • Dose-response for valproate, topiramate, carbamazepine. No dose trend for lamotrigine or levetiracetam.
  • Landmark registry: largest prospective ASM pregnancy safety database. Supports LTG/LEV as safest options.

Design

Study Type: Prospective pregnancy registry (observational cohort)

Randomization:

Blinding: Malformation adjudication by blinded dysmorphologist.

Enrollment Period: February 1997 to January 2023

Follow-up Duration: Through 12 weeks postnatal

Centers: North American multi-site registry

Countries: United States, Canada

Sample Size: 7311

Analysis: Logistic regression for RRs. Reference: lamotrigine.

Inclusion Criteria

  • Pregnant women enrolled in NAAEDPR.
  • Taking ASM as monotherapy during first trimester.
  • Live-born, stillborn, or terminated pregnancy due to fetal malformation.
  • ASMs with ≥50 exposed participants included.

Exclusion Criteria

  • Spontaneous abortion.
  • Withdrawal from registry or lost to follow-up.
  • Minor anomalies (birthmarks, deformations, prematurity complications, genetic syndromes).

Baseline Characteristics

All Monotherapy (N=7,311):

  • Mean maternal age: ~30 years
  • Caucasian: >80%
  • Mean gestational age at enrollment: 16 weeks

Arms

FieldObservational registry
InterventionProspective enrollment of ASM-exposed pregnancies. Phone interviews at enrollment, 7 months, and 8-12 weeks postpartum. Medical record verification.
Duration26 years (1997-2023)

Outcomes

OutcomeTypeControlInterventionHR / OR / RRP-value
Major congenital malformations by ASM (monotherapy, first trimester)PrimaryLamotrigine: 52/2,461 (2.1%)Valproate 31/337 (9.2%); Topiramate 26/510 (5.1%); Phenobarbital 12/200 (6.0%); Levetiracetam 26/1,283 (2.0%); CBZ 32/1,132 (2.8%); PHT 12/423 (2.8%)All significant vs LTG
Oral clefts — TopiramateSecondary7/510 (1.37%) vs reference 0.11%
Neural tube defects — ValproateSecondary4/337 (1.19%) vs reference 0.12%
Cardiovascular anomalies — ValproateSecondary9/337 (2.67%) vs reference 0.33%
SGA — TopiramateSecondary19% vs 9.5% unexposed
Not reportedAdverseNo adverse event data extracted for this trial

Subgroup Analysis

Dose-response: valproate risk higher >650 mg/day; topiramate higher >200 mg/day; carbamazepine higher >700 mg/day. No dose trend for lamotrigine or levetiracetam.

Criticisms

  • Non-population-based volunteer sample — differential participation and retention.
  • Fetal loss malformations not captured — underestimates true risk.
  • Unexposed comparison group differs sociodemographically from ASM-exposed.
  • Limited power for rare specific malformations.
  • Lacosamide (n=88) and pregabalin (n=62) estimates very imprecise.
  • No adjustment for confounders significantly changed results (crude RRs used).
  • Cross-registry comparison difficult due to different malformation definitions.

Funding

Multiple pharmaceutical companies (Advanz, Janssen, Jazz, Marinus, SK Life Sciences, Supernus, UCB, Viatris, others).

Based on: North American AED Pregnancy Registry (Neurology, 2025)

Authors: Sonia Hernandez-Diaz, Moira Quinn, Susan Conant, ..., Lewis Ball Holmes

Citation: Neurology. 2025;105:e213786.

Content summarized and formatted by NeuroTrials.ai.