Abstracts

Rationale: High Frequency Oscillations (HFOs) are seen during seizures and interictally, during slow wave sleep and seem to play a role in seizure genesis. This study investigates whether interictal and ictal HFOs occur in the same brain regions and how they relate to each other, to the preictal period and to the seizure onset zone (SOZ). Methods: Epochs of interictal slow wave sleep (5-10 minutes), preictal iEEG (10 seconds) and seizure onset (5 seconds) were studied for spikes, ripples (80-250 Hz) and fast ripples (FR: 250-500 Hz) and their co-occurrence on all artefact-free channels. The three different states (interictal, preictal and ictal) were compared for percentage of time occupied by the different events. We studied which event type in which state was the most specific for the SOZ channels, by comparing ratios between events in the SOZ channels to events in channels outside the SOZ. We studied whether the events occurred in the same channels interictally, preictally and ictally by comparing the ranking distance (RKD) of channels between states. The RKD was determined by ranking channels from those with the highest to those with the lowest percentage of time occupied by specific events, and then comparing ranks between two states. More than 0.22 was considered high (z-test). Also, bar graphs from individual patients were evaluated for the occupation of channels by events during ictal and interictal periods. Results: 25 patients were included. HFOs and spikes were more frequent in the ictal period than in the interictal period (interictal to ictal: ripples 2.6% to 6.7%, FR 0.2% to 0.8%, spikes 1.1% to 4.4% over all channels). HFOs already increased in the preictal period compared to the interictal night (interictal to preictal to ictal: ripples outside spikes 2.3% to 3.7% to 4.6%, FR outside spikes 0.11% to 0.25% to 0.51%), in contrast to spikes (1.1% to 0.9% to 4.4%). The ictal ripples had the highest ratio of events in the SOZ compared to events outside the SOZ (0.63). The RKD of channels between interictal, preictal and ictal was higher than 0.22 for spikes, but not for ripples and FRs. Ictal increase of all occurred mostly in the channels with interictal HFOs, but the increase in FRs was most confined to the same channels, while the increase in spikes was relatively more widespread. Conclusions: In contrast to spikes, HFOs increase during the preictal period. Interictal and ictal HFOs occur in the same channels, while ictal spikes are seen more extensively than interictal spikes. HFOs appear to remain confined to the epileptogenic area interictally and ictally, while spikes might reflect the spread of epileptic activity and thus represent reactive more than causal phenomena.