Abstracts

Rationale: Paroxysmal high frequency oscillations (HFOs) in the interictal intracranial EEG has been the subject of recent attention as markers of epileptogenic cortex. HFOs may occur simultaneously, independently, or at times with consistent propagation across subdural recording sites. We sought to determine if the propagation patterns of interictal HFOs, when observed, predict the propagation of seizures. Methods: We reviewed all adult intracranial EEG studies performed at the Medical College of Wisconsin’s Comprehensive Epilepsy Program between January 2007 and May 2008. EEG sampling rate was 500 Hz in all but two patients where it was 1000 Hz. Intracranial EEG records were reviewed at a sensitivity of >20 uV/mm with the low frequency filter set at 53 Hz, and high frequency cut-off at 120 Hz. Localization of HFOs and their propagation patterns, as well as ictal onset and propagation was performed by visual analysis. Results: A total of 26 intracranial EEG studies were screened. Nine patients with mesial temporal sclerosis and four patients with no ictal data to allow definition of ictal propagation patterns were excluded. We identified 13 patients with neocortical seizure onsets. Ictal onset zones were identified in the lateral temporal (3), frontal (3), temporo-parietal (2), temporo-occipital (1), and fronto-temporal (4) neocortices. Multichannel HFOs were noted in all 13 patients. Frequency of the HFOs encountered ranged from 60-120 Hz (mean 93.5 Hz, SD16.9Hz). In 7 of 13 patients (53%) a consistent HFO propagation pattern was present, with propagation across lobar boundaries or within a lobe. Propagation time was always less than 200 ms. In 6 of 7 patients (85%) with a consistent HFO propagation pattern, ictal onset zone was predicted by the subdural electrodes with the leading HFOs, and ictal spread by electrodes showing HFO propagation. In one of these patients, only one of two different ictal patterns that were observed was predicted by HFO propagation. In these 6 patients, HFO and ictal propagation pattern was interlobar in 4 cases, and intra- plus inter-lobar in 2 cases. Frequencies of the leading HFOs were identical to those of the propagated HFOs in 4 of these patients, but in two patients the leading HFOs were lower in frequency compared to the propagated activity. Amplitudes of the HFOs showed no consistent relationship between areas of seizure onset and propagation. Conclusions: Our observations suggest that seizure spread patterns may be predicted by propagation patterns of interictal high frequency oscillations. Although our observations are limited to frequencies < 120 Hz, they appear to argue against the notion that higher frequency HFOs are more localizing of ictal onset zones. Instead, they support the view that the location of leading HFOs may be more predictive of ictal onset zones than HFO amplitude or frequency.