Abstracts

Rationale: Around 10-17% of focal epilepsy patients have no interictal epileptiform discharges (IEDs) on scalp EEG [interictally-silent epilepsy (ISE)]. The role of MEG in the pre-surgical workup of ISE is unexplored. The purpose of our study is to fill this knowledge gap. In addition, we analyzed the IEDs noted during Stereo-EEG (SEEG) in ISE to correlate MEG findings.

Methods: After IRB approval, we used our prospectively maintained video-EEG monitoring (VEM) database from 01/01/2015 to 12/30/2020 to identify adults with ISE. Patients were classified as having ISE if they fulfilled the study inclusion and exclusion criteria. Inclusion criteria: underwent at least 72 hours of VEM, at least one typical seizure captured, lacked IEDs during the VEM, and underwent MEG at the Cleveland Clinic Epilepsy center. Exclusion criteria: IEDs captured on any pre-surgical EEG reports, including portable, ambulatory, and additional VEMs, along with outside records. Patients were dichotomized into MEG + or MEG – based on the presence or absence of MEG interictal epileptiform dipoles, respectively. MEG dipole clusters were classified into tight (single or multiple) or loose (single or multiple including scattered dipoles). In patients who underwent SEEG, IEDs were classified into spike (including repetitive spikes, polyspikes, slow with spikes) or low voltage fast activity (LVFA). Engel outcome scale was used for assessing seizure outcomes at last follow-up in patients with at least 6 months of post-surgery follow-up data.

Results: A total of 32 patients met the study criteria. Average length of VEM was 9.5 days (including all pre-surgical VEMs). A total of 12 (37.5%) patients with ISE were MEG +. Among MEG + patients, a total of 36 MEG clusters were identified. The most common location of interictal MEG dipoles was temporal lobe (14/36; 38.9%), including neocortical (n = 10), basal temporal (n = 3) and temporal pole (n = 1) followed by opercular region (8/36; 22.2%). Twenty-three patients underwent SEEG, including 7 of 12 (58.3%) MEG+ and 16 of 20 (80%) MEG- patients. On SEEG, the most common IEDs were spikes (100% of MEG+ and 93.75 % of MEG-patients). LVFA was more common in MEG- (43.8%) compared to MEG+ (14.2%) patients. Most common location of SEEG IEDs in MEG- ISE patients was anterior cingulate gyrus (7/16 patients; 43.8%). Of the 22 patients who underwent surgical intervention, 18 had at least 6 month of follow up data. MEG+ patients had 66.7% Engel I outcomes compared to 75% in MEG- patients at last follow-up.

Conclusions: MEG identifies interictal epileptiform dipoles in slightly more than a third of ISE patients, which are most commonly located in neocortical temporal and opercular region. MEG- ISE patients more often have low voltage fast activity IED on SEEG. IEDs in cingulate region may be missed by both EEG and MEG. Despite a negative MEG, good surgical outcomes can be achieved in ISE patients using SEEG.

Funding: Please list any funding that was received in support of this abstract.: None.