Abstracts

Rationale: Determining the etiology of epilepsy, which affects ~1% of the population, is challenging but can have important implications for treatment, prognosis and recurrence risk counseling. Methods: We performed molecular testing for 1,600 individuals using next-generation sequencing and copy number analysis for 53 genes most often associated with major forms of primary and syndromic epilepsy. Results: A known or predicted pathogenic variant was observed in 261 individuals (16%), underscoring the role of genetic factors in disorders of neuronal excitability. The highest molecular diagnostic rate (21%) was obtained for a subset of genes associated with infantile-onset epilepsy, including infantile epileptic encephalopathy or spasms. 6% of positive findings were gene deletions or duplications, emphasizing that copy number analysis increases the diagnostic yield. 478 individuals (30%) had a variant of unknown significance (VUS), including 169 who only had one heterozygous VUS in an autosomal recessive gene, most likely unrelated to the phenotype or representing carrier status. The remainder had likely benign or no reportable variants. Our data showed a pronounced incidence of autism and/or intellectual disability not only in individuals with mutations in MECP2, STXBP1, SCN1A, and CDKL5, but also in those with variants in KCNQ2, SLC2A1, SCN2A, SPTAN1, and GABRG2. Although mutations in ion channel genes are a common cause of epilepsy, half of all pathogenic findings in our cohort were in non-ion channel genes. For example, 23% of individuals had mutations associated with disorders in the Rett/Angelman syndrome spectrum. We also found pathogenic mutations associated with rare forms of epilepsy, such as pyridoxine-dependent epilepsy and others, for which a molecular diagnosis can guide treatment. The vast majority of individuals with positive findings had definitive pathogenic variants in only one gene, suggesting that genetic forms of epilepsy are mostly monogenic. However, 35% of these individuals also had one or more VUS, which may contribute to the phenotype, and modifier effects due to variants in other genes not included in the assay may also exist. Conclusions: Our data provide new insight into the mutation spectrum and frequency of pathogenic variants for specific epilepsy subtypes, broaden the genotype-phenotype correlations for many genes studied in this cohort, and illustrate the utility of a multi-gene testing strategy.