Abstract

Background: Accurate localization of the probable Epileptogenic Zone (EZ) from presurgical studies is crucial for achieving good prognosis in epilepsy surgery. Objective: To evaluate the degree of concordance at a sublobar localization derived from non-invasive studies (Video-EEG, MRI, FDG-PET) and EZ estimated by StereoEEG, in forecasting seizure recurrence in a long-term cohort of patients with focal drugresistant epilepsy. Methods: We selected patients with a full pre-surgical evaluation and with post-surgical outcome at least one year after surgery. Multivariate Cox regression analysis for seizure freedom (Engel Ia) was performed. Results : 74 patients were included, 62.2% were in Engel class Ia with a mean followup of 2.8+2.4 years after surgery. In the multivariate analysis for Engel Ia vs >Ib , complete resection of the EZ found in Stereo-EEG (HR: 0.24, 95%CI: 0.09-0.63, p : 0.004) and full concordance between FDG-PET and Stereo-EEG (HR: 0.11, 95%CI: 0.02-0.65, p : 0.015) portended a more favourable outcome . Most of our results were maintained when analyzing subgroups of patients. Conclusions: The degree of concordance between non-invasive studies and StereoEEG may help to forecast the likelihood of cure before performing resective surgery, particularly using a sublobar classification and comparing the affected areas in the FDG-PET with EZ identified with Stereo-EEG.