Polygenic Risk and Genetic Predisposition in Post-Traumatic Epilepsy: A Framework for Risk Estimation

BackgroundPost-traumatic epilepsy (PTE) can be a lifelong complication of traumatic brain injury (TBI). We hypothesize that PTE develops according to a Two-Hit Hypothesis, in which the first "hit" is a genetic predisposition and the second "hit" is the TBI. Using two independent datasets, we provide the first evidence that PTE is a polygenic disorder, and we propose a whole-exome sequencing (WES) based framework to estimate the risk of developing PTE. MethodsFrom a pool of thousands of screened veterans, we recruited a cohort of PTE subjects (n=28) and a control cohort of TBI subjects without PTE (n=22) and then performed WES. Approximately 375,000 variants identified in each subject were compared to approximately 15,000 verified epilepsy-associated variants from ClinVar. Fishers exact test was used to identify variants associated with PTE, which were subsequently classified as either PTE-prone or PTE-protective. Odds ratios (ORs) were calculated at both the variant and subject levels. Receiver operating characteristic (ROC) curve analysis was used to evaluate the predictive model. A second dataset was used to validate the model, and logistic calibration was performed to estimate the probability of developing PTE. ResultsThirty PTE-prone and 56 PTE-protective variants were identified, with corresponding large-effect ORs. ROC analysis demonstrated excellent discrimination (AUC=0.97). Polygenic variant patterns differed between cohorts, with a predominance of PTE-prone variants in affected individuals and PTE-protective variants in controls. ConclusionOur findings support a polygenic framework consistent with the Two-Hit Hypothesis. In addition, we developed a framework to estimate individual polygenic risk for PTE.