Abstracts

Rationale: The insular cortex is a complex structure enclosed in the depth of the Sylvian fissure. Covered by highly functional temporal, frontal and parietal operculae, it can generate a variety of ictal symptoms (e.g. viscerosensory, visceromotor, somatosensory, motor, speech and auditory) falsely leading to a diagnosis of temporal, frontal or parietal lobe seizures. MRI detection of an insular/periinsular lesion facilitates clinical recognition but some cases are MRI-negative. We wished to assess the role of other non-invasive diagnostic modalities in insular/periinsular epilepsy. Methods: Charts from all patients with insular/periinsular epilepsy seen by one of the authors (DKN) at our epilepsy clinic from August 2002-March 2010 were reviewed. Patients were included if (a) they had seizures related to an epileptogenic insular/periinsular lesion; or (b) had insular seizures/periinsular seizures documented by intrainsular depth electrodes; and (c) had undergone anyone of the following non-invasive tests: video-EEG monitoring, ictal single-photon emission computed tomography (iSPECT), interictal fluorodeoxyglucose-position emission tomography (PET) and magnetic resonance spectroscopy (MRS). Ictal EEG was classified as either localizing (ictal rhythm confined to T3/T4, C3/C4 or P3/P4), lateralizing (to the epileptogenic hemisphere), non-lateralizing (no ictal rhythm or diffuse ictal rhythm), false-localizing (ictal rhythm in the ipsilateral hemisphere other than T3/T4, C3/C4 or P3/P4), a false-lateralizing pattern (ictal rhythm in the contralateral hemisphere). The results of iSPECT and PET were classified as localizing (localize the epileptogenic lobe), lateralizing, non-lateralizing, false-localizing, or false-lateralizing. Results from MRS of the insula were classified as lateralizing, non-lateralizing or false-lateralizing. Results: Fifty-nine patients were identified as having insular/periinsular epilepsy (including 8 with intracerebral confirmation). Of these patients, 23 had undergone non-invasive diagnostic modalities: EEG (all patients), video-EEG (20/23 patients), MRI (all patients), iSPECT (15/23 patients), PET (12/23 patients) and MRS (4/23 patients). Ictal video-EEG correctly localized the insular/periinsular region of epileptogenicity in 10% and lateralized it an additional 50% of patients. Ictal SPECT correctly localized the insular/periinsular region of epileptogenicity in 27% and lateralized it in an additional 27%. PET correctly localized the insular/periinsular region of epileptogenicity in 25% and lateralized it in an additional 25%. MRS correctly lateralized the insular/periinsular region of epileptogenicity in 25% of patients. Of note, MRI was normal in one patient with depth-recorded insular/periiinsular seizures. Conclusions: Our results suggest none of the various presurgical evaluations studied here provide consistent localization and at best lateralizes to the correct hemisphere. Although some novel non-invasive techniques are not expected to provide much help (e.g. MEG-EEG and EEG-fMRI), others such as quantitative analysis of MRI are promising and need to be assessed.