Abstracts

Rationale: Invasive ictal recordings with intracranial (IC) electrodes are often required to delineate the epileptogenic zone for neurosurgical treatment of medically refractory epilepsy. Although traditionally EEG frequencies have been limited to the range of 1.5-100Hz, recent studies have revealed that higher frequency activities could be determinant of epileptogenicity. Likewise, very low EEG frequencies, referred to as direct current (DC) potential shifts, have been associated with initial ictal EEG changes. This study examined these two extreme frequencies at the time of the seizure onset, explored the associations with epileptogenic zone and compared the results with traditional ictal EEG recordings. Methods: We examined 10 patients (ages 12-59) with refractory focal epilepsy who underwent invasive ictal recordings as part of their presurgical evaluation. A total of 17 seizures were reviewed under three different conditions: Traditional filtering for invasive recordings (1.6-100 Hz frequencies), wideband filtering up to 600 Hz (sampling rate 2000 Hz), and DC recording (high-pass filter at 0.8 hz). EEG data was acquired with higher sampling rate using “Multiview” analysis software provided from BESA by Nihon Kohden in 0.5-second epochs. DC potentials were defined as positive or negative shifts lasting at least 3 seconds that consistently occurred at the time of seizure onset in the same location. Initial ictal EEG onset time and electrode localization were determined for each seizure under these 3 different conditions, and findings were subsequently compared. Results: All patients had resective surgery of the presumed primary epileptogenic zone after invasive recordings. Each patient had 1-4 seizures. Results were summarized in the table below. Eight of 17 seizures showed no DC shifts, and seven of these seizures belonged to 5 patients. In every seizure that did have a DC-shift at the time of the ictal onset, the DC-shift preceded the onset determined by conventional EEG and HFO data. In follow-up of at least one year, eight out of 10 patients became seizure-free and their localizations by all three methods were congruent. The other two patients (total of 7 seizures), had no improvement in their seizures, and the presumed localization of the primary epileptogenic zone determined by DC-EEG recordings differed from that of the conventional ICEEG. In 7 of the seizures, HFO preceded onsets recorded with conventional ICEEG and were localizing, but the DC-EEG recordings consistently (when present) showed even earlier changes at the time of the seizure onset. Conclusions: Certainly DC-recordings are much less labor- and storage-intensive than wideband high-frequency recordings. Our study shows that DC potential shifts may often be recorded easily and may provide more useful information than conventional EEG settings and higher frequency settings. This finding may help to improve localization of the primary epileptogenic zone in refractory epilepsy.