Abstracts

Rationale: On histopathological examination, the resected hippocampii from patients undergoing epilepsy surgery demonstrate a variety of findings. Intraoperative hippocampal electrocorticography (IH-ECoG) frequently captures electrographic seizures. The relationship between hippocampal pathology and presence of seizures on IH-ECoG is not well understood. Methods: Consecutive patients that underwent anterior temporal lobe resection along with IH-ECoG were included in the study. IH-ECoG was recorded for at least 10 min from the ventricular surface of the hippocampus after removal of the temporal neocortex, using a 1x4 subdural strip. The electrophysiologic, radiographic, and histopathologic characteristics were collected for each patient. Correlation between the electrographic seizures (recorded intraoperatively during IH-ECoG and/or during long-term extraoperative intracranial EEG in patients undergoing two-stage surgery) and preoperative MRI as well as hippocampal histopathology was analyzed. Results: A total of 17 patients were included in the study over a period of 16 months. Histopathological analysis of resected tissue revealed 3 patients with mixed oligoastrocytoma; 2 patients each with DNET, oligodendroglioma, hippocampal sclerosis, mild gliosis; and 1 patient each with infiltrating astrocytoma, ulygyria, cortical dysplasia, microheterotopia, epidermoid tumor, and meningioma. On IH-ECoG recording, 8 of 17 patients (47%) had electrographic seizures. Of these 8 patients, 6 (75%) had evidence of either neoplastic involvement or a developmental defect within the hippocampus proper (infiltrating tumors in 4 patients, cortical dysplasia and microheterotopia in 1 each). Of the 9 patients without seizures on IH-ECoG, only 4 (44%) had hippocampal neoplasm and two others had hippocampal sclerosis (22%). Similarly, hippocampal abnormality was detected on MRI in 9 patients. Five of these 9 patients (56%) had seizures on IH-ECoG. Of the remaining 8 patients without hippocampal abnormality on MRI, IH-ECoG recorded seizures in only 3 (38%). In 10 patients who had two-stage surgery, extraoperative long-term intracranial EEG recordings captured hippocampal seizures (alone or with neocortical seizures) in 5 patients (50%). Of these 5 patients, 4 (80%) had seizures during IH-ECoG as well. Of the other 5 patients without hippocampal seizures on extraoperative EEG, only 2 (40%) had seizures intraoperatively during IH-ECoG. Conclusions: IH-ECoG frequently captures hippocampal seizures following neocortical temporal resection. Intraoperative hippocampal seizures are more likely to occur in patients with hippocampal pathology (neoplasia, heterotopia, or dysplasia) compared to patients with either no pathology or with hippocampal sclerosis/mild gliosis. Similarly, an abnormal MRI signal involving the hippocampus (excluding hippocampal sclerosis) appears to be predictive for presence of seizures during IH-ECoG. In addition, hippocampal seizures are more likely to be seen intraoperatively when the epileptogenic zone (defined by extraopeartive intracranial recording) involves the hippocampus.