Abstracts

Rationale: There is currently a large interest in improving epilepsy surgery outcomes and the appropriate utilization of the numerous testing available at a clinician's discretion. Several studies have demonstrated the effectiveness of functional imaging including PET and SPECT scans as well as neurophysiologic testing such as EEG and MEG source localization. However, many of these studies are of inferior research quality being retrospective and limited to a single institutional experience. Currently it is still unclear if combined EEG and MEG source modeling is beneficial and if specific subtypes of patients are more likely to benefit from this practice. Our study aims to demonstrate the feasibility and benefits of a cloud based multicenter outcome database to better define the appropriate role of EEG and MEG source localization. Methods: Northwestern University created a cloud based database that allows sharing of detailed demographic and clinical information of epilepsy patients including large files of MRIs, MEG and EEG studies. This was performed using a combination of the Redcap and Owncloud interfaces. We performed a retrospective review of patients throughout the Chicago area who had EEG/MEG source localization performed and subsequently had surgery. Patients who had refractory focal epilepsy with presurgical workup including MRI, long term EEG recording and MEG/EEG source localization and had post-surgical follow-up of at least 6 months were included in this study. Results: Currently, complete results from 2 Chicago area institutions are included with pending analysis of 2 additional sites. A total of 22 patients were included with 11 females and 11 males with average age of 24 (range 4-57). 10 of the patients were suspected to have a temporal lobe epilepsy with 12 having suspected extra-temporal lobe epilepsy. 6/22 underwent a phase II implantation with the remaining 16 having intraoperative electrocorticography. A total of 34 EEG dipoles and 29 MEG dipoles were seen in the 22 patients. 2 patients had MEG dipole solutions only and 4 had only EEG dipoles. 18/22 patients had Engel Class I outcome. No significant difference in seizure freedom rates were seen in patients based on the EEG or MEG dipoles. Conclusions: In this small dataset, we demonstrate the use of a cloud base database as an effective tool for multicenter studies. Application of such technology can be a valuable tool in enrolling patients on a national or even international level, particularly in cases where patient enrollment is challenging or limited. Further studies evaluating the appropriate utilization and role of specific ancillary testing in the evaluation of patients for epilepsy surgery are needed to advance the current practice. Interestingly, almost 20% of the patients had EEG only dipole models. This is notable as the current practice at most MEG centers is to not report and model EEG only spikes. While these results need to be validated with large patient numbers, this suggests that perhaps more centers should model the EEG data as well. Continued recruitment of patients and further analysis will be needed to make more informative statements about patient outcomes. Funding: Supported by an American Academy of Neurology/American Brain Foundation Clinical Research Training Fellowship