Abstracts

Rationale: Resistance to antiepileptic treatment continues to be a major clinical problem for up to 40% patients with epilepsy. To date, efforts to identify genetic susceptibility to treatment-resistant epilepsy have focused on candidate genes and hypothesis-free genome-wide association analyses (GWAS) have not been performed. In this study, we present the results of a GWAS on treatment-resistant epilepsy of unknown etiology.Methods: Phenotyping of subjects was performed using an algorithm with positive predictive value of 89%, developed and validated by our group using electronic medical records. Samples were genotyped on HumanHap550 or Human610-Quad (Illumina). Imputation of non-observed genotypes was carried out using SHAPEIT and IMPUTE2. Population ancestry was determined by principal component (PC) analysis using Eigenstrat 3.0 and 10 PCs were generated and used as covariates in the association analysis with SNPtest.Results: 331 subjects (73 treatment-resistant and 258 treatment-responsive) identified by the algorithm with non-lesional epilepsy of unknown cause were selected for the analysis: 30% African Americans, 63.4% Europeans and 6.6% of other ancestries. 37.9 million genetic variants were imputed, however, only 8.2 million were considered for the analysis after removing poorly imputed variants and variants with a minor allele frequency < 3%. Three loci showed nominal significance (p=10e-7) with treatment-resistance: SPRED2 in 2p14, previously associated with immune disorders, SLC7A11, a Cystein/Glutamate Transporter in 4q31.1; and 13q14.2 harboring the serotonin transporter gene HTR2A, previously linked with the age of onset of epilepsy, and SUCLA2 encoding the mitochondrial succinyl-CoA ligase.Conclusions: We are currently conducting an ancestry-stratified analysis on this dataset that will help better elucidate the role of ancestry-specific genetic factors, however, these preliminary results show great promise in the identification of the genetic underpinnings of treatment-resistant epilepsy.