Abstracts

Rationale: Insular epilepsy can mimic temporal/frontal lobe epilepsy and vice versa. Insular coverage during invasive EEG monitoring is often prompted by classical clinical semiology, opercular seizures, frontal lobe semiology with EEG localization in the temporal lobe, nonlesional MRI and insular involvement on functional imaging. Mapping the insula carries some risks. We aimed to determine insular ictal onset and spread, as well as safety and yield of insular intracranial EEG. Methods: We reviewed adult patients who underwent intracranial EEG monitoring for focal epilepsy between 2009 -2016. Demographic data, scalp EEG findings, structural and functional imaging, pre-implantation seizure onset zone (SOZ) localization hypothesis, results from intracranial EEG, surgical complications and outcomes were obtained. Results: One hundred seventy nine patients underwent intracranial monitoring, of whom 24 had insular implantation. One to three depth electrodes were placed in the insula to map the anterior, middle and posterior insula, using stereotaxic and navigation-assisted techniques. No procedural complications were reported. Six patients had typical insular symptoms. The pre-implantation hypothesis was frontal, orbitofrontal, opercular, insular or temporal lobe. Scalp ictal EEG showed electrodecremental respose in two patients, and no change in two patients, and the rest demonstrated a frontal, temporal or frontotemporal onset. Ten patients had prominent interictal discharges in the insula. Primary insular ictal onset was seen in nine (37.5%) patients. Low voltage fast activity was the predominant ictal onset pattern, except for two cases where 2-3 Hz repetitive spiking was seen and the onset was simultaneous with the orbitofrontal region. Secondary spread to insula was seen in twelve patients (50%) and no involvement of insula was seen in three patients. The mean duration to insular spread was 49 seconds (range 2-373 seconds). Several EEG spread patterns were seen. Insular seizures spread to mesial temporal and then orbitofrontal regions and vice versa. Mesial temporal seizures commonly spread to insula and gave rise to clinical symptoms. Frontal seizures spread to insula and then to temporal region. Nine patients were recommended for insular resection, five of whom underwent resection. Three patients were Engel class I and two were Engel class II at mean follow-up of 40 months. Conclusions: Primary (37.5 %) or secondary (50%) ictal involvement of the insular cortex was common in these patients. Based on semiology alone, only half the patients with suspected insular epilepsy actually had insular onset seizures. Low voltage fast activity is the most common ictal rhythm, except with simultaneously involvement of the orbitofrontal region when it is low frequency repetitive spiking. The epileptic network triad involving Insula, mesial temporal, and orbitofrontal structures is confirmed and should be sought in these patients. Insular depth electrode placement is a safe and effective adjunct to routine invasive EEG investigations. Funding: no funding support