Abstracts

Rationale: Epilepsy is one of the most common neurological disorders, with a lifetime incidence of 3%. The epileptic encephalopathies (EE) are the most severe of all the epilepsies. Patients typically present with refractory epilepsy with multiple seizure types, cognitive arrest or regression, and have a poor prognosis. Several EE genes and EE-associated CNVs have been described, but a systematic mutation screening approach across multiple EE has not been undertaken and most cases have an unknown etiology. Methods: We screened 32 known and 40 candidate genes for disease-causing changes in a cohort of 518 patients with a diverse spectrum of EE. This was achieved by targeted resequencing of all exonic regions of these 72 genes, using molecular inversion probes for targeted capture followed by multiplexed massively parallel sequencing. This data was analyzed using two different pipelines. Firstly, standard variant calling tools were used to identify all DNA sequence changes in EE patients, and compared to large control data sets to identify rare variants. The presence of these rare mutations was then assessed in familial samples to determine pathogenicity. Secondly, using read depth data followed by normalization and a standard score analysis we were also able to detect copy number variations (CNVs) in the target genes. Putative CNVs will be validated using a custom oligonucleotide array with high-density coverage of all target genes. Results: To date, we have analyzed data from 365 individuals for 13 known and 19 candidate EE genes with 93% of targeted nucleotides sequenced at >25 fold coverage required for accurate variant calling. At this coverage threshold we detected 100% of known single nucleotide variants (n=622) and 86% of known CNVs (n=15). Segregation analysis of rare, possibly damaging variants in known EE genes detected in 317 EE probands revealed pathogenic changes in SCN1A ,GABRG2, CDKL5, GRIN2A and PCDH19 in five families and possibly pathogenic, inherited changes in an additional five families. The clinical manifestations of mutation carriers often differed from those previously described, broadening the clinical spectrum of phenotypes associated with known EE genes. We also observed intra-familial variability and incomplete penetrance. Rare variants within our candidate genes, comprising glutamate transporters, the NA+/K+ transporter ATPase-interacting proteins and synaptic vesicle exocytosis proteins is ongoing. Conclusions: In summary, we present a cost-effective, efficient method of screening multiple EE genes in large cohorts that will transform molecular testing in EE, enabling rapid early diagnosis, informing prognosis, treatment and genetic counseling.