Magnetoencephalography in epilepsy patients carrying a vagus nerve stimulator

Due to technical constraints, magnetoencephalography (MEG) is challenging in vagus nerve stimulation (VNS) patients. This study evaluates (1) the feasibility of MEG in VNS patients and (2) the added value of MEG in their presurgical evaluation (PE). Ten VNS patients were studied by MEG using the spatiotemporal signal space separation (tSSS) method. Equivalent current dipoles (ECD) were classified "clustered"/"scattered". It was evaluated whether MEG (1) confirmed localisation of the hypothesized epileptogenic zone (HEZ), (2) improved delineation of the HEZ, or (3) identified 1 out of multiple HEZs. Finally it was evaluated whether adding MEG to the PE improved patient management by changing or supporting the hypothesis. In 7/10 patients, tSSS allowed to obtain interpretable MEG data, with interictal epileptiform discharges in 6/7. ECD clustered within 1 lobe in 4/6; confirming the localisation of the HEZ in 2/4 and improving delineation of the HEZ in 2/4. When ECD clustered within 2 lobes (1/6) or were scattered (1/6), MEG could not identify 1 out of multiple HEZs. In 2 patients, MEG changed management to invasive video-EEG monitoring (IVEM) and resective surgery (RS). In 4 patients, MEG further supported the management; IVEM in 2/4 and unsuitability for RS in 2/4. So far IVEM, performed in 2, resulted into RS. This study demonstrates the feasibility of MEG in VNS patients. MEG changed management in 20% and further supported the proposed management in 40% illustrating the clinical value of MEG in the PE of VNS patients.