Abstracts

Rationale: Surgical planning for medically refractory epilepsy primarily uses ictal onset location to guide resection. Interictal spikes may add important information to the intracranial EEG (iEEG) evaluation, but controversy exists regarding the relationship of interictal spikes to seizure onset location in iEEG recordings. To address this issue, we determined the spike density over long epoch iEEG and compared spike location to seizure onset.Methods: Twelve patients (ages 1-17), 11 eventually diagnosed by histology with cortical dysplasia, underwent surgical evaluations for medically intractable epilepsy between 2003 and 2006. Each patient had subdural grids and/or strips placed with contacts (ranging from 68 to 116) spanning 2 or more lobes. Seizures were marked by two reviewers for origin of clinical and sub-clinical seizures. The number of seizure onset contacts ranged from 0 to 37 (mean 12.2, median 5) for clinical and 0 to 33 (mean 11, median 6) for sub-clinical. A computerized spike detector located spikes over eight hours of randomly selected interictal iEEG. The relationship of spike density to seizure onset zone was compared. Seizure onset and spike density were considered to correlate if the contacts with highest spike density were within the seizure onset zone.Results: Clinical seizures were recorded in 9 patients and 2 had seizure onset within the region of highest spike density. Of the remaining 7 patients, 5 had seizure onset in regions adjacent to the highest spike density (within 1 or 2 contacts of the most active interictal spike focus), and 2 patients had seizure onset 3 or more contacts distant from the highest interictal spike density. Ten patients had sub-clinical seizures and 4 had seizure onset within the region of highest spike density. Of the remaining 6 patients, 3 had seizure onset within one contact of the highest spike density, and 3 patients had seizure onset 3, 4 and 6 contacts distant from the highest interictal spike density. Seven patients had both clinical and sub-clinical seizures. The clinical and subclinical seizure onset contacts were not identical for any patient and 2 patients had no contacts in common between sub-clinical and clinical onset.Conclusions: In children with histologically proven cortical dysplasia undergoing iEEG recordings, the region of most frequent interictal spikes are usually not within the seizure onset zone. Maximal spike density is typically adjacent to or several centimeters distant from the onset of clinical and sub-clinical seizures. Our data suggest that use of interictal spikes on iEEG is not sufficient for mapping the seizure onset zone. Because the region of maximal spike density frequently does not spawn seizures, the pathophysiology of seizure generation and spikes may differ. The work was supported by: CURE, NINDS, and the Woman's Fund of The Children's Hospital of Philadelphia.