Abstracts

Rationale: Children exposed in utero to valproate (VPA) have an increased risk of birth defects. We sought to determine whether some of this risk could be attributable to genetically based maternal epilepsy syndromes, such as idiopathic generalized epilepsy, for which VPA is an effective treatment. Methods: The methods of data collection and identification of congenital malformations utilized by the North American Antiepileptic Drug Pregnancy Registry have been described previously (Neurology 2005;64:961-965). Two neurologists (EB, BD), blinded to pregnancy outcome, independently reviewed questionnaire responses and, when available, medical records of women taking VPA monotherapy and enrolled between 9/97 and 11/2006. Each mother was classified as having: non-epilepsy (NE, e.g., migraine or bipolar illness), idiopathic generalized epilepsy (IGE), partial epilepsy (PE), or non-classifiable epilepsy (NCE). Malformation risks among children of women in each diagnostic group were compared using the Fisher test. Results: 173 VPA-exposed pregnancies were identified, of which 19 (11.0%, including one set of dizygotic twins, counted as a single pregnancy) were associated with a major malformation. Maternal diagnoses were: NE 14 (8.1%), IGE 90 (52.0%), PE 21 (12.1%), and NCE 48 (27.7%). About half (89/173, 51.4%) furnished medical records. Proportions of children with malformations in each group are shown in Table 1. These rates did not differ significantly from each other. Limiting the analysis to IGE and PE mothers diagnosed using medical records yielded malformation rates of 9/60 (15%) for IGE and 1/8 (12.5%) for PE; these were also not significantly different (p=1.00). Conclusions: We found no significant differences in malformation rates among VPA-exposed children of mothers with different epilepsy syndromes. Since our study was not powered to evaluate this question, however, many more cases would be required to exclude such an effect.