Ictal electrocorticographic findings related with surgical outcomes in nonlesional neocortical epilepsy

Abstract

Purpose: To characterize ictal electrocorticographic features related to surgical outcomes in nonlesional neocortical epilepsy (NE). Methods: We analyzed 187 ictal electrocorticograms (ECoG) obtained from 18 patients who had undergone presurgical evaluation and subsequent neocortical resections (frontal: seven, parietal: one, occipital: four, multilobar: six). None of them had any MRI-detectable lesions. Various ECoG data sets recorded from eight patients who achieved a favorable surgical outcome (either seizure free or more than 90% reduction of seizure frequencies) were compared with that from ten patients with unfavorable outcome (less than 90% reduction of seizure frequencies) (follow up duration: 47±11 months). Results: Reproducible ictal onset zone (IOZ) in recurrent seizures (P=0.013) and persistent ictal discharges in IOZ from the onset to the end of seizure (P=0.004) were found more frequently in the patients with good outcome. Ictal onset patterns consisting of low voltage fast or high amplitude beta spikes predicted a good surgical outcome while rhythmic sinusoidal activity or rhythmic spike/sharp waves of slow frequency were predictive of poor outcome (P=0.01). The ictal onset rhythm consisting of gamma or beta frequencies was more prevalent in the favorable group (P=0.015). Conclusions: The presence of stable ictal circuit suggested by the consistent earliest activation of specific electrodes in the repetitive seizures (reproducible IOZ) and the active participation of IOZ throughout the attack were valuable prognostic factors in addition to the morphology and frequency of ictal onset rhythm.

Introduction

The presence of a lesion detected by MRI or other neuroimaging studies has been known as the best prognostic factor for surgery of medically refractory neocortical epilepsy (NE) (Cascino et al., 1992, Van Ness, 1992, Berkovic et al., 1995, Lorenzo et al., 1995, Spencer, 1995, Zentner et al., 1996). Compared with lesional NE, nonlesional NE has been associated with much less favorable surgical outcomes, in which a seizure free rate after cortisectomy ranged from 0 to 20% despite a routine use of invasive EEG investigations (Zentner et al., 1996, Kutsy et al., 1999, Lee et al., 2000). Presumably, the lower yield of invasive EEG in NE might be related to a very rapid propagation of ictal discharges through extensive neural networks, a widely spread epileptogenic region, or improper placement of electrodes. Despite these problems associated with the EEG investigations in NE, a comprehensive analysis of electrophysiological features recorded from intracranial electrodes should be an essential exercise to identify certain characteristics strongly correlating with the surgical outcomes. Previous investigations (Zentner et al., 1996, Kutsy et al., 1999, Lee et al., 2000) of the intracranial ictal EEG features in NE have largely failed to identify specific features related with surgical outcomes aside from a contiguous ictal spreading pattern (Kutsy et al., 1999). However, prior studies had included both lesional and nonlesional NE, which might be inappropriate for various reasons; (1) significantly different surgical outcomes between lesional and nonlesional NE, (2) potential alterations of surrounding neuronal characteristics or circuitries by macroscopic lesions, (3) characteristic EEG patterns related to specific lesions, or (4) different concepts for the placement of intracranial electrodes between lesional and nonlesional cases. Therefore, we limited the analysis of ictal electrocorticographic (ECoG) features to a pure group of nonlesional NE.