Abstracts

Rationale: Birth defects are major concerns of antiepileptic drug (AED) therapy in pregnancy. Mechanisms underlying their occurrence are poorly understood. Experimental and clinical evidence suggest that genetic factors play a role in AED-associated anatomical teratogenesis, but their identification remains elusive. Here, we investigated whether AED-induced de novo mutations are a potential mechanism for AED-associated birth defects. Methods: Whole exome sequencing (WES) was performed on child-parent trios recruited within a large international collaboration. Putative de novo mutations were first identified using an in-house bioinformatics pipeline, and subsequently confirmed using the Integrative Genomics Viewer (IGV). Fisher’s Exact test was used to compare the proportion of children with de novo mutations between children exposed to AEDs in utero (AED-exposed children) with birth defects and AED-exposed children without birth defects. The Wilcoxon rank sum test was used to test for differences in the number of de novo mutations per child between children exposed to valproate in utero (VPA-exposed children) and children either exposed to AEDs in utero other than valproate or not exposed to any AEDs in utero (VPA-unexposed children), irrespective of birth outcome. Results: A total of 61 child-parent trios (from 33 families) were included in the study, comprising 50 trios with AED-exposed children and 11 trios in which the child was not exposed to any AEDs in utero. 9/50 AED-exposed children had birth defects, including hypospadias (n=2), cryptorchidism (n=1), clubfoot (n=1), or multiple malformations (comprising cardiac, skeletal or neural tube defects; n=5). Initial WES analysis identified 78 putative de novo mutations; after further screening, 45/78 (58%) were considered to be high-confidence de novo mutations, of which eight have suggestive evidence of having arisen post-zygotically. The proportion of children harbouring de novo mutations did not differ significantly between AED-exposed children with birth defects (n=9) and AED-exposed children without birth defects (n=41) [mean/median (range): 0.55/0 (0-2) vs 0.63/0 (0-2); p=1.0]. It is noteworthy that the overall burden of de novo mutations in these two groups is comparable with the rates of ~1 exonic de novo mutation per control trio reported in larger trio exome sequencing studies. The number of de novo mutations per child did not differ significantly between VPA-exposed children (n=13) and VPA-unexposed children (n=48) (mean/median: 0.77/1 vs 0.73/1, p=0.42). Conclusions: This hypothesis-free, genome-wide analysis of AED-associated anatomical teratogenicity suggests that exposure to AEDs in utero does not increase the de novo mutation burden, and that this mechanism is not a major contributor to the occurrence of AED-associated birth defects. Funding: The study was supported by a National Health and Medical Research Council (NHMRC) Project Grant (1059858) and an NHMRC Program Grant (1091593).