Clinical Biomarkers and Prediction Models for Poststroke Epilepsy

In an era of time-dependent reperfusion and recanalization therapy for stroke leading to improved survival, there is a growing population at risk of poststroke epilepsy (PSE). Accumulating evidence suggests a multidirectional interaction among stroke, PSE, and dementia in stroke survivors. There is no evidence to justify prophylactic antiseizure medication (ASM) to reduce these morbidities. Although several predictive molecular biomarkers and scoring models have been proposed, they remain inadequately validated for stratifying risk and indicating who will benefit from prophylactic ASM. Studies leveraging advances in genetics, metabolomics, electrophysiology, imaging, and artificial intelligence (AI) may help to discover noninvasive molecular biomarkers and easy-to-score models. These discoveries should improve our understanding of epileptogenesis in PSE and identify new pharmacologic targets. Besides, accurately identifying high-risk patients and timely initiating prophylactic ASM therapy has the potential to disrupt the feed-forward multidirectional interaction among stroke, PSE, and dementia.