Abstracts

Rationale: Epilepsy surgery is commonly done for patients with intractable epilepsy and has been showed to be superior to medical treatment in this population. We know that the presence of epileptic pathology outside of the expected surgical bed reduces the chances of postoperative seizure-freedom, but we don’t know whether such a negative prognostic effect is equally relevant in all types of epilepsy surgery. We report here the preliminary investigation of the predictive ability of 3 such measures on seizure recurrence: MRI brain abnormalities outside the surgical bed, bilateral interictal epileptiform activity and a non-localizable or generalized ictal EEG pattern.Methods: We retrospectively reviewed patients who underwent epilepsy surgery at the Cleveland clinic Epilepsy Center between 2010 and 2012. Data collected included general clinical characteristics, type of surgery, presence of MRI abnormalities outside the surgical bed, bilateral interictal epileptiform activity and a non-localizable or generalized ictal EEG pattern. A favorable outcome was defined as complete seizure-freedom.Results: The cohort included 147 patients of which 70 patients (48 %) were female. The cohort underwent 3 types of surgeries: Temporal Lobectomy (TL in 62%), Frontal Lobectomy (FL in 14%), Posterior Quadrant Surgery (PQ in 24%). MRI brain showed an abnormality in 125 patients (85%). Of those patients who had an abnormal MRI brain, 37 patients (25%) had MRI brain abnormalities outside the surgical bed. EEG data analysis revealed that 39 patients (27 %) had bilateral interictal epileptiform activity and 37 patients (25%) had a non-localizable or generalized ictal EEG pattern. Seizures recurred following surgery in 70 patients (48%). Patients with MRI abnormalities outside the surgical bed only had a non-significant tendency for seizure recurrence as a group (P = 0.1). However, when subgroup analysis was performed, patients who underwent FL and had MRI abnormalities outside the surgical bed were more likely to have seizure recurrence (Risk ratio: 2.90; P= 0.05) (Fig 1A), while patients who underwent TL had a smaller trend towards more seizure recurrence (Risk ratio: 1.62, P= 0.04) (Fig 1B) but not in patients who underwent PQ resections ( P=0.5) ( Fig 1C). Presence of bilateral interictal epileptiform activity or a non-localizable or generalized ictal EEG pattern had no effect on the likelihood of increased seizure recurrence.Conclusions: In this preliminary analysis, the presence of MRI brain abnormalities outside the surgical bed seems to have a differential effect on seizure recurrence following surgery, with a negative effect highest in the context of FL, followed by TL, and least relevant after PQ resections. Presence of bilateral interictal epileptiform activity or a non-localizable or generalized ictal EEG pattern had no effect on the likelihood of increased seizure recurrence after surgery, which may be attributed to either lack of effect or small sample size.