Abstracts

Rationale: Simple partial seizures frequently escape scalp-EEG detection. Ictal EEG patterns in such seizures can be better detected on intracranial EEG (ICEEG). The location, orientation, extent, and degree of synchrony of the generating source, are factors determining the appearance of epileptic activities on scalp. We aimed to compare the extent of the intracranial ictal onset zone in patients with scalp-EEG undetectable seizures to patients with scalp-EEG positive seizures arising from identical sublobar regions. No previous systematic studies addressing this question in patients with epilepsy arising outside the temporal lobe were found in the published literature. Methods: We retrospectively reviewed the records of 426 patients who underwent ICEEG for epilepsy surgery from 1997-2008. Of these, 69 patients had seizures or auras with no ictal EEG pattern on their preoperative scalp video-EEG recordings. Patients who had seizures with identical semiology both during their scalp-EEG and subsequent ICEEG evaluation were included. Cases were age-matched to control patients, who had scalp-EEG positive seizures preoperatively, arising from an identical sublobar region on ICEEG, according to the Talairach system. The extent of the intracranial ictal onset zone was estimated by the number of ICEEG electrodes involved at ictal onset. Results: A total of 18 cases were identified. There were no differences in demographic data, epilepsy classification, MRI findings, pathology results, or surgical outcomes between cases and controls. Average number of implanted contacts was 104 (36-184) in cases and 115 (48-184) in controls (p=0.368). A total of 289 seizures were analyzed (156 in cases and 133 in controls). Scalp undetectable seizures were associated with auras in 13 patients and simple motor seizures with or without aura in 5 patients. Intracranial ictal EEG onset was located on the lateral convexity in 9 patients, mesial aspect in 6 patients, and basal surface in 3 patients. Mean number of contacts involved at the ictal onset was smaller in cases (2.88 /-1.83) compared to controls (4.50 /-2.66) (p<0.001). This difference was also notable when examining each of the 3 surfaces of the brain separately: lateral convexity (2.89 /-1.99 vs 4.58 /-3.54) (p=0.020), mesial aspect (3.46 /-1.73 vs 4.44 /-2.12) (p=0.031) and basal surface (2.32 /-1.14 vs 4.53 /-0.96) (p<0.001). On the whole, the estimated size of the ictal onset region was 2-3cm2 in cases and 4-5cm2 in controls.