Abstracts

RATIONALE: Faster frequency initial ictal discharges (IIDs) in scalp EEG correlate both with greater degrees of hippocampal pathology and with hippocampal seizure onsets in temporal lobe epilepsy (TLE) (Vossler et al. Ann Neurol 1998;43:756-62; Ebersole and Pacia, Epilepsia 1996;37:386-99). The goal of this study was to determine whether the site of seizure onset within the temporal lobe correlates with hippocampal pathology. Objective: after this presentation participants should be able to discuss how hippocampal pathology may influence site of seizure onset in the temporal lobe.
METHODS: All consecutive patients with no hippocampal atrophy (HA) on MRI who had scalp EEG-video evidence of TLE (non-lesional TLE) were prospectively studied with intracranial electrodes to determine spatial extent of the IID in a protocol beginning in 1996. Twenty-six patients with either no HA, or slight HA ([lt] 3 S.D. below normal hippocampal absolute volumes and asymmetry index), were included. Twenty-three had longitudinal hippocampal depth electrodes and orbitofronal and temporal subdural strip electrodes. Three had only strip electrodes for technical reasons. Three other patients with significant HA who required depth and strip electrodes for clinical purposes served as a comparison group. Electrode placement was verified using intraoperative X-ray or postoperative MRI. Hippocampal pathology was assessed by measurement of HA and by histopathologic grade of hippocampal sclerosis (HS).
RESULTS: A total of 29 patients were investigated. Twenty-three patients with depth electrodes had no, or mild, HA. The ictal EEG onsets in those 23 subjects showed: simultaneous onsets in the hippocampus/amygdala, parahippocampal gyrus (PHG) and lateral neocortex in seven cases, IIDs in hippocampus and PHG in six patients, IIDs outside the hippocampus in eight individuals (three in PHG, three in lateral neocortex, and two in both PHG and lateral neocortex), a bilateral diffuse neocortical IID in one subject, and an IID confined to the hippocampus in only one case. The three patients with no depth electrodes and no HA all had IIDs involving the PHG and temporal neocortex simultaneously. The three patients with moderate-marked HA all had IIDs confined to the hippocampus. Moderate-marked HA significantly correlated ([chi]-square=18.65, p=0.001, n=26 depth electrode subjects) with IIDs restricted to the hippocampus. Fifteen patients had the degree of HS graded. High grade (3-5) HS also significantly ([chi]-square=6.24, p=0.012) correlated with an IID limited to the hippocampus.
CONCLUSIONS: Absence of HA and lower grade HS both correlate with initial ictal electrographic discharges involving the hippocampus and medial (with or without lateral) temporal cortex or only the medial and/or lateral temporal neocortex. By contrast, substantial HA and high grade HS both correlate with IIDs confined to the hippocampus/amygdala.