Abstracts

Rationale: MEG is increasingly used in presurgical evaluation of patients with refractory partial epilepsy (RPE) although its clinical value compared to other localization techniques is still under debate. This multicenter prospective blinded study assessed the clinical added value of MEG in presurgical evaluation of patients with RPE. Methods: Between September 2007 and April 2010, 70 consecutive patients (42 males, mean age: 31,5 years, age range: 3-63 years) were prospectively included. Fourty-seven patients were followed at ULB-H pital Erasme (Brussels) and 23 patients at Ghent University Hospital. All patients had RPE and were not formally excluded from surgery after conventional non-invasive presurgical evaluation (CNIPE). All patients underwent whole-head MEG recording (Elekta Neuromag, Elekta Oy) during one hour (eyes-closed rest, lying position). MEG data were visually inspected for interictal epileptiform discharges (IED). Corresponding equivalent current dipoles (g/%>80%) were fitted in patients spherical head model and coregistered on their MRI. Results of CNIPE were first discussed blinded to MEG results in respective multidisciplinary epilepsy surgery meetings to determine the presumed localization of the epileptogenic zone and the therapeutic attitude (A: focal resective surgery, B: invasive EEG monitoring (iEEGm), C: rejected except if new decisive information from MEG). MEG results were then multidisciplinary discussed. The way MEG influenced the therapeutic attitude was assessed. Results: Based on CNIPE, twenty-five patients had extra-temporal lobe epilepsy, 36 had temporal lobe epilepsy and 9 had unclear localization. The therapeutic attitude was A in 24 patients, B in 32 patients and C in 14 patients. MEG was unreadable in 3 patients, normal in 14 patients and showed IED in 53 patients (76%). MEG results did not change the therapeutic attitude in 56 patients (80%) but confirmed it in 97% of the cases when abnormal. MEG changed the therapeutic attitude in 14 patients (20%). MEG-related changes consisted in a reorientation from C to B in 3 patients and from C to A in 1 patient, in modifications of iEEGm electrode implantation plan in 5 patients and in additional neuroimaging investigations in 5 patients. MEG-related changes involved 36% of patients with extra-temporal lobe epilepsy, 44% of patients with unclear localization and 3% of patients with temporal lobe epilepsy. Conclusions: Novel information brought by MEG changed the therapeutic attitude in 20% of patients with RPE who are potential candidates for surgery. MEG-related changes mainly involved patients with extra-temporal lobe epilepsy and unclear localization. MEG confirmed the management option in 97% of patients with no MEG-related change and abnormal MEG. We conclude that MEG is a clinically relevant non-invasive localization technique for presurgical evaluation of patients with RPE.