Abstracts

Rationale: Magnetoencephalography (MEG) has become a valuable tool in pre surgical evaluation of children with refractory epilepsy. Our goal is to study MEG findings in children whose pathology after epilepsy surgery showed cortical dysplasia (CD). We analyzed MEG characteristics according to the subtype of CD, and correlated MEG with two key tools for pre-surgical evaluation, video EEG (VEEG) and brain MRI. Methods: Of 178 MEG exams (2008-2013) in children (≤18 years) undergoing pre-surgical evaluation for epilepsy, study population included 33 who underwent resective surgery with histopathological diagnosis of CD and had at least 6 months of post-operative follow up. CD classification was re-ascertained systematically by the Neuropathologist (RP) using ILAE classification. Other clinical, VEEG, imaging data was extracted from medical charts. Results: Of 33 children (mean age = 9.45 years) who underwent MEG, 17 (52%) had ILAE type I, 13 (39%) had type II, and 3 (9%) had type III CD. 30 had epileptiform activity (Interictal and Ictal) on MEG- 14(47%) with type I, 13 (43%) with type II and 3(10%) with type III CD. However, tight MEG spike clusters (defined as dipoles across 2 gyri and 1 sulcus) were seen more frequently in types III (67%) and II (46%) compared to type I (35%) CD. Of 28 with MEG interictal spikes, 25 (89%) also had interictal spikes on long term VEEG with higher grades of CD being more likely to have spikes on both: 3/3 (100%) with type III, 11/12 (92%) type II, and 11/13 (85%) type I had both MEG and EEG spikes. In 5 patients with no interictal MEG spikes (4 with type I and 1 with type II), 3 had EEG spikes while 2 patients had no spikes on MEG or EEG. 3 of 28 (2 type I and 1 type II; 11%) patients had MEG unique spikes that were not seen on EEG. Brain MRI showed CD in 21 of 33 (64%) patients, with 10/17 (59%) of type I, 8/13 (61%) of type II, and 3/3 (100%) of type III. In non-lesional brain MRI patients (12/33; 36%), 11/12 (92%) showed MEG spikes (5/5 Type II, 6/7 Type I) assisting in the localization of epileptogenic zone preoperatively. Overall successful outcome (ILAE 1&2) was found in 47% patients with type I CD, 46% with type II and 66% patient with type III CD (mean follow up 17.3 months, range 6-48 months). Of 30 patients with epileptiform MEG abnormalities, complete or partial resection of the spike sources was achieved in 23 patients with 48% achieving successful outcome. Whether the patients were MRI-negative made little difference with 4/8 (50%) achieving successful outcome. Conclusions: MEG spikes are more frequent in Type 1 CD in children but these are likely to be scatter or loose clusters rather than tight clusters. Tight clusters are more common in type III and type II dysplasia. MEG supplemented EEG in interictal spike source localization, and showed unique spikes in 11% patients. In non-lesional brain MRI patients, MEG was a useful non-invasive tool for spike localization in addition to scalp video EEG.