Abstracts

Rationale: It has been reported that an ictal discharge in temporal lobe epilepsy (TLE) does not appear in the scalp EEG until the electrocorticographic (ECoG) ictal discharge becomes both well-synchronized and present over a large surface area (>10 cm²) of the cortex (Tao et al., Epilepsia 2007). The goal of this study was to verify this observation in a group of patients who had intrahippocampal depth electrodes in addition to subdural strip and scalp electrodes. Methods: We reviewed the ictal ECoG-EEGs of 149 seizures in 19 consecutive TLE patients studied prospectively with concurrent scalp temporal, longitudinal hippocampal depth and lateral and inferior temporal lobe subdural strip electrodes. The degree of hippocampal atrophy (HA) on MRI and the grade hippocampal sclerosis (HS) in pathology studies were noted. The location and timing of the initial ictal discharge, later significant pattern, and subsequent significant pattern recorded in the ECoG from the hippocampus (HF), medial paleocortex (MPC), and lateral temporal neocortex (LNC) and in the EEG from the scalp temporal areas were examined. Results: Five patients had substantial HA or high grade HS, and 14 did not. In most subjects in both groups, an ictal pattern was detectable in the scalp temporal EEG with the same frequency and at the same time as when the ECoG discharge spread from its site of origin to become lobar (HF, MPC & LNC). In 5 subjects, however, the appearance on the scalp was delayed 5 or more seconds after the ECoG discharge had become lobar. Occasionally the scalp ictal pattern appeared when the ECoG discharge was only medial (HF & MPC, n=2) or only outside the HF (MPC & LNC, n=2). In 3 subjects, the scalp initial discharge was slower than the ECoG discharge. Conclusions: In patients with TLE, with or without HS, an ictal pattern is not detectable in the scalp EEG until (or after) the time the ECoG ictal discharge has spread from its site of onset to involve at least two of these areas: the HF, MPC and LNC.