Abstracts

Rationale: We retrospectively reviewed EEGs, electrocorticography (ECoG) and MRIs in children status post resection to determine if follow up EEGs and or ECoG findings during resection could be a predictor of relapse. Methods: We identified patients from our epilepsy surgery database since 2011 who underwent resection of FCD for medically refractory epilepsy. The database and medical records were queried for age of seizure onset, type, MRI findings, EEG findings, and outcomes. Results: Twenty seven patients were identified (age 19 months to 20 years). Average age of seizure onset was 30 months, all with focal epilepsy except 2 with facilitated spasms. FCD was found on all patients' MRI, 3 with mesial temporal sclerosis and 1 with nodular heterotopia. Three initially had a "normal" MRI and with ancillary tests such as FDG PET and MEG, FCD was seen. All three had deep lesions. Out of 27 patients who had resection of FCD, 19 patients had follow up EEGs. Two patients had a normal EEG (1 relapsed and 1 remained seizure free). Seventeen patients had an abnormal EEG. Seven of 17 patients relapsed from 5 to 15 months post-resection (Engel III-IV). Of these 7, 6 had abnormal ECoG findings post resection (residual spike waves near the resection cavity) and 1 patient had grids with 1 electrographic seizure from a deep lesion (resection limited since it abuts the language area). Follow up EEGs showed focal spike waves +/- slowing and in 2, seizures noted in the same area of the FCD. Majority (6/7) showed residual dysplasia in MRI. Pathology were Type 1a (5), 1b (1), and 11b (1). Nine out of 17 were seizure free despite an abnormal follow up EEG. Post resection ECoG in 7 patients showed no spikes (2), significantly decreased spikes (4) and slowing (1). Two had grids placed, 1 had an ictal event with a depth electrode and 1 had language mapping only. Follow up EEGs showed slowing (5), or attenuation (1), slowing with rare spikes (2), initial sharp and slow wave then normalized (1). Post op MRI showed no residual FCD (8), and suspected FCD (1). Pathology were Type 1a (5), 1b (1), 2a (2) and 2b (1). One of the patients with an abnormal EEG was lost to follow-up. Conclusions: We looked at our patients with FCD who have undergone surgery to determine if follow up EEGs or ECoG findings can predict seizure recurrence. Majority of patients with relapse had residual spikes post resection and follow up EEG were abnormal in the previous resection site. Their MRI continued to show residual dysplasia. This shows that complete resection is crucial to attain seizure freedom. The majority of patients who remained seizure free had subtle follow up EEG abnormalities and also had relatively "quiet" ECoG post resection and correlated with no residual dysplasia by MRI. The study also showed that an initial ‘normal" MRI may be seen in FCD and additional tests are needed especially in deep lesions. This is our initial review and the study is limited by the number of subjects and the lack of time standardization for follow up EEGs.