Abstracts

Rationale: Pediatric epilepsies caused by focal cortical dysplasia (FCD) can be singled out among medically refractory epilepsies (MREs) as an exceptionally severe type of epilepsy. In this study, we investigated the potential clinical utility of electric source imaging (ESI) on intracranial EEG (iEEG) as a new approach for surgical planning in patients with focal MRE undergoing intracranial long-term monitoring (LTM). The goal of the study is to perform ESI of ictal and interictal epileptiform discharges using iEEG and assess its predictive value in terms of surgical outcome in children with MRE and FCD. To this purpose, we also estimated the predictive value of the clinically-defined seizure onset zone (clinical-SOZ) and irritative zone (clinical-IZ) and compared it with the ESI-SOZ and ESI-IZ. Methods: We retrospectively analyzed long term monitoring iEEG data from 25 children with MRE confirmed to have FCD after resection. For each patient, ictal events recorded during the LTM period were analyzed as well as a 10-12 min segment of interictal data, recorded with subdural and/or depth electrodes. Ictal rhythmic activity at seizure onset and interictal spikes were marked by two reviewers who were blinded to each other's results (Fig. 1). We determined the location of each iEEG contact on the patient's cortex by co-registering presurgical MRI with post-implantation CT. For each patient, we defined the ESI-SOZ and ESI-IZ by localizing all the ictal and interictal events using the Equivalent Current Dipole (ECD) method. We also defined the clinical-SOZ as all the iEEG contacts active at seizure onset and the clinical-IZ as the most interictally active iEEG contacts. The resected volume were determined by co-registering the patient’s presurgical and postsurgical MRIs. For ESI, we calculated the distance of each SOZ or IZ dipole from the closest resection border (Dres). Finally, for each patient, we estimated the percentage of resection (proportion of dipoles with Dres <15 mm) of both ESI-SOZ and ESI-IZ and tested its predictive value in terms of surgical outcome, dichotomized into good (Engel 1) and poor. The optimal threshold on the percentage of resection to predict outcome was assessed using the receiver operating characteristic curve (ROC). Similarly, we assessed the predictive value of the clinical-SOZ and IZ. Results: We localized interictal spikes in all 25 patients (average: 11 spikes/min). Ictal data from 157 seizures of 18 patients were also included (average of 9 seizures/patient). ROC analysis showed that the resection of the ESI-IZ was more predictive of good outcome (Engel 1) than clinical-IZ (Fig. 2): (i) for ESI-IZ, we estimated a positive predictive value (PPV) of 100%, negative predictive value (NPV) of 63%, and informedness (IM) of 60% (Fisher’s test: p=0.003); (ii) for clinical-IZ, we had a PPV of 78%, NPV of 86%, and IM of 53% (Fisher’s test: p=0.007). Similarly, for the SOZ, the resection of the ESI solution was more predictive than the clinical-SOZ: (i) for ESI-SOZ, we estimated a PPV of 90%, NPV of 75%, and IM of 68% (Fisher’s test: p=0.013); (ii) For clinical-SOZ, we had a PPV of 77%, NPV of 80%, and IM of 48% (Fisher’s test: p=0.047). Conclusions: Our study showed that the ESI definition of both IZ and SOZ using iEEG has the potential to guide the surgical resection in children with MRE and FCD. Our findings indicate that the resection of the IZ and SOZ defined by ESI potentially increases the likelihood of postsurgical good outcome compared to the resection of clinically-defined IZ and SOZ. Funding: No funding