F09. Predictors of post-surgical seizure outcomes in refractory temporal lobe epilepsy

Abstract

Listen

Translate article

Introduction The effectiveness of epilepsy surgery depends on precise identification of epileptogenic zone and historically temporal lobe epilepsy has been described to have highest seizure freedom rates after surgery. Researches have previously analyzed multiple variables with distinct predictive values of postsurgical outcome. We attempted to identify prognostic variables in temporal lobe epilepsy using the diagnostic studies that are routinely performed during presurgical evaluation. Methods Retrospective analysis of electroencephalogram (EEG), Magnetic resonance imaging (MRI), Positron emission tomography (PET), Single-photon emission computed tomography (SPECT) and Magnetoencephalography (MEG) was performed on 93 adult patients with refractory temporal lobe epilepsy who underwent selective temporal lobectomy was performed. Outcomes was assessed on a clinical basis using modified Engel’s classification. Results At one year follow up, there were 73 patients (78.5%) with good surgical outcome (E1 & E2) and 10 patients (10%) with poor outcome (E3 & E4) and 10 patients were lost to follow up. EEG findings such as consistently unilateral interictal epileptiform discharges (IEDs) during NREM sleep (91%), consistent lateralized focal ictal onset (90%) and IEDs concordant with ictal onset on scalp and intracranial EEG were highly predictive of a good outcome. Mesial temporal sclerosis and other hippocampal abnormalities on MRI correlated with good outcome (82–100%). Concordant findings on ictal and interictal EEG MRI, PET and MEG also correlated with good outcome. The type of ictal onset pattern did not have significant difference. In patients with poor outcome (E3 & E4), there was lack of correlation between MRI/PET with interictal and ictal EEG. The presence of any lesion on the MRI did not seem to make any difference in outcome, especially in patients who skipped an invasive monitoring. In two-thirds of the patients with poor outcome, having concordant MRI findings and ictal onset did not change the surgical outcome. Subset analysis of non-lesional patients with hypometabolism on PET had a good outcome (33%). Conclusion Several factors have been identified with some postsurgical predictive value and majority of the studies have included patients with mesial temporal sclerosis (MTS), which is usually considered to be of excellent prognostic value. Contrary to this, the presence of MTS on MRI, did not improve the outcome in our study. This may indicate that the presence of MTS is an independent variable, and other diagnostic studies should be considered to make a reasonable prognostication and such information can be used to counsel the patients appropriately.