T. Kobayashi,1 M. Tanoshima,1 R. Tanoshima,1 J. Beyene,2 G. Koren,1 S. Ito1; 1The Hospital for Sick Children, Toronto, ON, Canada, 2McMaster University, Hamilton, ON, Canada

BACKGROUND: Valproic acid (VPA) has been suspected to adversely affect fetal development. The recent surge of published data prompted us to systematically address: 1) time profiles of emergence of the VPA teratogenicity signals over the last 30 years; and 2) expand risk estimates of specific congenital malformations (CMs) associated with VPA which have not been addressed to date.
METHODS: A systematic literature search was conducted on Medline, Embase classics plus Embase, and Cochrane Central Register of Controlled Trials between 1947 and May 2014. Cumulative and conventional meta-analyses were performed. Pooled relative risk (RR) and 95% confidence intervals of VPA monotherapy-associated risks of combined and specific major CMs, compared to other antiepileptic drugs (AEDs), were calculated.
RESULTS: We identified 59 cohort studies. Cumulative meta-analyses showed the increased risk of combined major CMs associated with VPA exposure in utero has been statistically significant since 1990. Signals of significant risks of specific major CMs (neural tube defects, congenital heart defects, cleft lip/palate, genitourinary anomalies and musculoskeletal anomalies) all emerged between 1992 and 2006. Conventional meta-analyses showed RR of VPA-associated major CMs were 2- to 7-fold in combined and each of the specific CMs, compared to patients with other AED exposure. Subgroup analyses suggested lamotrigine conferred the lowest risk of major CMs among the 5 common AEDs.
CONCLUSION: The significant risk signal of each of the major specific CMs associated with VPA has become apparent as early as 22 years ago. Our analysis confirms for the first time that VPA causes congenital heart defects, cleft lip/palate, genitourinary and musculoskeletal anomalies.