Abstracts

Rationale: Advances in genomic technologies, including targeted next generation sequencing and whole exome sequencing (WES), enables identification of pathogenic variants in 10 – 78 % of select patients with unexplained epilepsy. The clinical impact is significant and includes earlier diagnosis of disorders with specific treatment implications. In an ongoing study, we report the results of WES on 50 patients with early onset epilepsy of unknown cause.Methods: Between December 2014 - June 2015 WES was performed on 50 patients attending BC Children’s hospital with a history of early onset epilepsy (≤ 5 years) of unknown cause. Patients were classified as retrospective (epilepsy > 6 months) or prospective (epilepsy < 6 months). Detailed clinical data was abstracted in a REDCap database. WES was performed using the Ion AmpliSeq™ Exome Kit and Ion Proton™ System within 2 weeks of receiving samples. Reporting was restricted to the sequences of 557 genes previously implicated in epilepsy, with exon sequence average read depths >80X. Putative causative mutations were validated by Sanger sequencing and by parental testing (when possible). Diagnostic yield and time to diagnosis was obtained.Results: A definite diagnosis was made in 13 out of 50 patients (26%: 5/7 prospective; 8/43 retrospective). This included known pathogenic and novel variants in SCN1A, ATP1A2, ALG13, STXBP1, POLG, SCN1B, KCNQ2 x 3, PAFAH1B1, TUBB2B, GABRA1 and CACNA1H. A possible diagnosis was identified in 5 additional retrospective patients for which supporting evidence is pending (GABRB3, ARHGEF9, CHD2, KCNQ2, SCN3A). An additional retrospective patient was identified to have a diagnosis which did not fully explain her phenotype (2 variants in BTD recognized to cause partial biotinidase deficiency, with biochemical support of low biotinidase). Eight patients (16%: 2 prospective, 6 retrospective) had a diagnosis with possible treatment implications (ATP1A2, SCN1A, SCN1B, BTD, KCNQ2 x 3, POLG). The patient with compound heterozygous mutations in POLG was on valproic acid which was stopped due to study results. Her liver enzymes were increasing at time of diagnosis and normalized with valproic acid removal. The patient with SCN1A mutation (Dravet syndrome) was also heterozygous for a maternally inherited SCN5A variant of uncertain clinical significance. SCN5A is associated with epilepsy, cardiac arrhythmias and sudden death. Time from enrollment with genetic counselling to diagnosis with Sanger confirmation was 21 – 105 days (mean 48 days).Conclusions: The clinical utility of using WES in this cohort is supported by potential diagnoses in 19/50 patients (38%: 5/7 prospective; 14/43 retrospective) within a mean time of 48 days. Diagnoses in 8 (16%) patients had immediate treatment implications. WES also facilitated multi-locus variant identification that may further impact the phenotype (SCN5A). Genome-wide WES from patients/families with no genetic diagnosis will also be analyzed to implicate novel genes in epilepsy and likely increase the diagnostic yield.