Incidence, temporal profile and neuroanatomic correlates of poststroke epilepsy

Abstract

Background and Purpose:The relationship between ischemic stroke site and occurrence of poststroke epilepsy (PSE) is incompletely understood. This study intended to evaluate incidence and temporal profiles of seizures and to correlate ischemic lesion sites with PSE using voxel-based lesion symptom mapping (VLSM). Methods:Patients with imaging-confirmed first-ever ischemic stroke without prior history of epilepsy were prospectively included. Demographic data, cardiovascular risk factors, and National Institute of Health Stroke Scale (NIHSS) scores were assessed. Data on seizures and modified Rankin scale scores were determined within a 90-day period after stroke onset. Ischemic lesion sites were correlated voxel wise with occurrence of PSE using nonparametric permutation test. Age-and sex-matched patients with first-ever ischemic strokes without PSE after 90 days served as controls for the VLSM analysis. Results:The stroke database contained 809 patients (mean age: 68.4 ± 14.2 years) with first-ever imaging-confirmed ischemic strokes without history of epilep. Incidence of PSE after 90-day follow-up was 2.8%. Five additional patients were admitted to the emergency department with a seizure after 90-day follow-up. Fifty percent of the seizures occurred in the acute phase after stroke. PSE patients had higher NIHSS scores and infarct volumes compared to controls without PSE (p < .05). PSE patients had infarcts predominantly involving the cerebral cortex. The hemisphere-specific VLSM analysis shows associations between PSE and damaged voxels in the left-hemispheric temporo-occipital transition zone. Conclusions:The data indicate that PSE occurs in a small proportion of patients with rather large ischemic strokes predominantly involving the cerebral cortex. Especially patients with ischemic lesions in the temporo-occipital cortex are vulnerable to develop PSE.