10.1016/S0271-7964(08)70442-7

Vollständiger Abstract

Worum geht es in dieser Arbeit?

OBJECTIVE: Epileptic arousals (EAs) are seizures characterized solely by arousal from sleep. EAs are unrecognized in current seizure classifications. Their subtle semiology and inconsistently detectable ictal activity in scalp electroencephalography (EEG) complicate differentiation from physiological arousals (PAs). This study characterizes EAs using simultaneous intracranial electroencephalography (iEEG) and scalp EEG with electrocardiography (ECG), exploring their neural and autonomic dynamics. METHODS: We retrospectively analyzed presurgical video-EEG monitoring data of 20 focal epilepsy patients (2010-2023). EAs were identified visually. Their seizure-onset zones (SOZs) and seizure-propagation zones (SPZs) were defined in Montreal Neurological Institute (MNI) space and compared to other seizure types. The detectability of ictal activity of EAs in scalp EEG was evaluated. Associations with SOZs/SPZs were assessed using logistic regression. Heart rate dynamics were compared between EAs and PAs evaluating distribution modality, variability, and receiver operating characteristic (ROC) analysis. RESULTS: Among 507 seizures, 133 (26.2%) were EAs. SOZs of EAs partially overlapped with other seizure types (Jaccard similarity index 0.47). Two-thirds (66.9%) of EAs were undetectable by scalp EEG. Detectability increased with broader maximal SPZs (odds ratio [OR] = 3.16; ROC-AUC [area under the curve] = 0.891). Heart rate increased during both EAs and PAs (p < 0.001), yet modulation was greater during EAs (p = 0.007, ROC-AUC = 0.995) and showed a multimodal distribution (p < 0.001). SIGNIFICANCE: EAs are common, yet underrecognized. They arise from partially distinct networks and may exhibit a specific autonomic signature, supporting their recognition as a separate seizure type. Reliance on scalp EEG alone risks missing patients with a high EA burden. Undetected EAs can perpetuate epileptic activity and sleep fragmentation.

Abstract: PubMed · Datensatz