Abstracts

Rationale: Focal cortical dysplasia (FCD) is a common cause of drug-resistant epilepsy (DRE) that is amenable to surgical treatment. To be successful, epilepsy surgery requires a comprehensive pre-operative evaluation to estimate the epileptogenic zone (EZ). To this purpose, non-invasive tests, such as high-density EEG (HD-EEG) and the magnetoencephalography (MEG), are usually performed to localize the sources of interictal epileptiform discharges and estimate the EZ to target during surgery. The possibility to tailor the surgical resection non-invasively via interictal magnetic or electric source imaging (MSI/ESI) would largely improve the pre-surgical workup of pediatric patients with epilepsy. The main goal of this study is to quantify and compare the ability of MSI and ESI to guide successful resection of the EZ in children with FCD. To pursue this goal, we analyzed interictal MEG and HD-EEG data in children undergone successful resective surgery (Engel 1) and compared the distances of MSI ESI and solutions to resection. Methods: We studied 17 children with DRE and FCD diagnosis based on histological exam who underwent successful surgery. We analyzed 12 minutes of interictal data from simultaneous MEG (306 sensors) and HD-EEG (72 channels) recordings (Fig 1). IEDs were identified on each modality by two independent reviewers and localized at the peak using an Equivalent Current Dipole (ECD). The patient’s presurgical MRI was used to build a realistic head model and a Boundary Element Model (BEM) consisting of skin, skull and brain tissue (three layers). We calculated the Euclidean distance of each ECD from the resected volume determined by co-registering the patient’s presurgical and postsurgical MRIs. Such distance was regarded as a measure of the localization error (Eloc) to the EZ and compared between ESI and MSI. For each dipole, we also calculated a measure of cluster density as the number of other adjacent dipoles within the radius of 15 mm. Dipoles with a cluster density ≥ 5 were classified as clustered or scattered otherwise. Results: All patients showed IED on both HD-EEG and MEG with a mean rate of 3 and 2.6 IEDs/min respectively (p=0.4, Wilcoxon sign-rank). Overall, MEG dipoles showed a lower Eloc than HD-EEG dipoles (median: 18 vs. 31 mm; p<0.001). Looking at each individual patient (Fig. 2), we observed that: (i) in 5 patients (29%), Eloc did not differ between MSI and ESI; (ii) in other 5 patients (29%), Eloc of ESI was lower than MSI (median localization improvement: 13-mm, p<0.05); and (iii) in 7 patients (41%), MSI was lower than MSI (median localization improvement: 19-mm difference). The proportion of clustered dipoles was higher in MEG than HD-EEG (68% vs. 54%; χ2: p<0.001). For MSI, the cluster density showed a moderate negative correlation with Eloc (p<0.001; R=-0.4); while this was not the case for ESI. Finally, Eloc of MSI, but not of ESI, improved significantly when considering clustered rather than scattered dipoles (MSI: 15 vs. 32 mm, p<0.001; MSI: 39 vs. 37 mm, p=0.5). Conclusions: MSI and ESI are valuable, complimentary tools for the presurgical estimation of the EZ in children with FCD. Our findings show that 306-channel MEG is able to localize the EZ more accurately than 72-channel HD-EEG though ESI was superior to MSI in about one-third of our patients. Our data indicate that the ability of MSI to estimate the EZ, but not of ESI, can be largely improved by targeting regions with dense dipoles clusters. Funding: No funding