Circadian and ultradian patterns of epileptiform discharges differ by seizure-onset location during long-term ambulatory intracranial monitoring.

Abstract

Previous studies reporting circadian patterns of epileptiform activity and seizures are limited by (1) short-term recording in an epilepsy monitoring unit (EMU) with altered antiepileptic drugs (AEDs) and sleep, or (2) subjective seizure diary reports. We studied circadian patterns using long-term ambulatory intracranial recordings captured by the NeuroPace RNS System. Retrospective study of RNS System trial participants with stable detection parameters over a continuous 84-day period. We analyzed all detections and long device-detected epileptiform events (long episodes) and defined a subset of subjects in whom long episodes represented electrographic seizures (LE-SZ). Spectrum resampling determined the dominant frequency periodicity and cosinor analysis identified significant circadian peaks in detected activity. Chi-square analysis was used to compare subjects grouped by region of seizure onset. In the 134 subjects, detections showed a strongly circadian and uniform pattern irrespective of region of onset that peaked during normal sleep hours. In contrast, long episodes and LE-SZ patterns varied by region. Neocortical regions had a monophasic, nocturnally dominant rhythm, whereas limbic regions showed a more complex pattern and diurnal peak. Rhythms in some individual limbic subjects were best fit by a dual oscillator (circadian + ultradian) model. Epileptiform activity has a strong 24h periodicity with peak nocturnal occurrence. Limbic and neocortical epilepsy show divergent circadian influences. These findings confirm that circadian patterns of epileptiform activity vary by seizure-onset zone, with implications for treatment and safety, including SUDEP.