Abstracts

Rationale: High-frequency oscillations (HFOs) have been proposed as an independent marker of seizure-onset zone (SOZ) in patients with intractable localization related epilepsy such as temporal lobe epilepsy (TLE). HFOs are termed ripples (80-250 Hz) and fast ripples (250-500 Hz), and are recorded using conventional subdural grids or strips, and depth electrodes. The occurrence of HFOs in the hippocampus is associated with epileptogenesis in experimental TLE models. The present study examines the relationship of anatomical location and electrophysiological parameters of HFOs, interictal spikes, and SOZ defined by conventional phase II video-EEG monitoring in a small cohort of patients with TLE. Clinical seizure outcomes were also examined. Methods: Retrospective review of intracerebral EEGs of 5 patients with intractable TLE were obtained using macroelectrodes. The SOZ was determined by monitoring clinical seizures; and after patients were restarted on AEDs, a two hour sample of interictal EEG was obtained on the day prior to resective surgery. There were 2 additional patients that were unresected due to SOZ overlap with eloquent cortex. The spatial distribution, amplitude, power spectra and spiking rates of HFOs were correlated with similar parameters for interictal spikes, and the anatomical location of SOZ. The resection of the HFOs, interictal spikes, and SOZ were also correlated with clinical seizure outcomes using Engel’s classification. Resections were performed based on SOZ and HFOs were not considered. Results: HFOs were found in all 7 patients in the anatomical location of SOZ. HFOs, interictal spikes and SOZ co-localized in 6/7 patients. However, HFOs and SOZ co-localized in electrodes without interictal spikes in 4/7 patients. HFOs were found outside of SOZ in 6/7 patients, but the rate, duration and power spectra of HFOs were higher in the SOZ than outside. HFOs were found associated to primary electrode source of interictal spike foci in 6/7 patients, but were also recorded in electrodes without interictal spikes in 5/7 patients. Nevertheless, most of the HFOs recorded were associated with SOZ or interictal spikes. All 5 TLE patients with complete resection of SOZ were at Engel’s class IA outcome at follow-up (~4.7 mo). HFOs were only completely resected in 3/5 patients. Conclusions: The interictal HFOs obtained in 2 hr intracranial recording are highly predictive of the location of the seizure onset zone in our limited series of TLE patients. Our experience has similarities, but some differences to the two published series (Staba et al., 2004 - UCLA; Urrestarazu et al., 2007 - MNI). For example, the presence of HFOs outside the SOZ occurred in 6/7, which is higher than in the two published series. Differences might be due to the clinical definitions used to narrow SOZ to a minimum number of electrodes as HFOs locations outside SOZ were often adjacent.