Abstracts

Rationale: Nonlesional neocortical resections result in freedom from seizures in approximately half of patients. However, noninvasive studies and the design of invasive electrode surveys vary among centers. Here we describe our approach to planning invasive electrode studies for nonlesional neocortical epilepsy. Methods: We reviewed intracranial electrode studies for nonlesional neocortical cases from 2005-2010. The impact of noninvasive studies on invasive electrode plans was assessed, and common survey strategies were identified. Results: The routine set of noninvasive studies included a detailed history and neurological examination, neuropsychological testing, anatomical magnetic resonance imaging (MRI), continuous audiovisual-electroencephalography monitoring (CAV-EEG), positron emission tomography (PET), interictal single positron emission computed tomography (iiSPECT), 7T magnetic resonance spectroscopy (MRS), ictal SPECT (iSPECT), and the Wada test. Functional MRI (fMRI) and spike-triggered fMRI were obtained in many patients. Factors that typically guided the design of intracranial electrode studies were seizure semiology, ictal CAV-EEG, neuropsychology, iSPECT, and language lateralization. PET and iiSPECT results less commonly affected electrode implantation, while MRS and spike-triggered fMRI were treated as investigational studies with minimal impact on decision-making. Intracranial electrode surveys were designed to cover the neocortex broadly, with denser sampling guided by noninvasive studies. Noninvasive studies sometimes lateralized seizure onset, but bilateral studies were common, especially with frontal lobe semiology and electrographic findings. Craniotomies were positioned to cover the cortical regions hypothesized to be the origin of seizures. All craniotomies included subdural strip electrodes extending radially from the perimeter of the craniotomy in a starburst pattern. When frontal lobe seizures were suspected, interhemispheric placement of bilateral medial strip electrodes from the nondominant side was included. Subdural grid electrodes were included when a certain region was highly suspicious or when stimulation mapping was planned. Electrode arrays were typically supplemented with one to three depth electrodes per side. Depth electrodes were placed in the bilateral hippocampi or frontal lobes, or were targeted to an iSPECT focus. Depending on results from the intracranial survey, a second localized intracranial study was often undertaken with subdural grid electrodes overlying the seizure origin and areas requiring stimulation mapping for resection planning. Conclusions: Details of current approaches for designing intracranial electrode studies are not widely known, and there is little evidence for choosing between approaches. The findings presented here illustrate the current approach at Yale. Quantitative comparison among approaches likely will be necessary to determine the optimal approach to localizing neocortical epileptic foci.