Abstracts

Rationale: Incomplete resection of the epileptogenic zone, defined by subdural electroencephalography (EEG) and postoperative magnetic resonance imaging (MRI), is the most important predictor of poor outcome following resective surgery for intractable epilepsy. However, up to 25% of patients with incomplete resection become seizure-free. We sought to determine predictors of seizure freedom after incomplete resection in children. We also analyzed the contribution of MRI and EEG data to incompleteness as a predictor of outcome. Methods: Patients were identified from an epilepsy surgical database at the Miami Children’s Hospital Brain Institute. We included all patients <18 years of age with incomplete resection and 2 years of follow-up. Patients undergoing hemispherectomy, corpus callosotomy, or multiple subpial transections were excluded. Fifteen pre and perioperative variables were compared in seizure-free and non-seizure-free patients. We analyzed for effect of complete EEG resection on seizure outcome in patients with incompletely-resected MRI lesions and vice versa. To further control for the contribution of MRI and EEG data to incompleteness, we compared the outcome of lesional patients, categorized by reason for incompleteness (i.e. MRI only, EEG only, or both), to a group of lesional patients with complete resection. Results: Eighty-five patients met inclusion criteria with 34 (40%) becoming seizure-free. Patients with full scale IQ>80 (p=0.02), contiguous MRI lesions (p=0.006), and temporal resection (p=0.04) were more likely to remain seizure-free. Amongst 45 patients with incomplete lesional resection, those with complete EEG resection were more likely to become seizure-free (Table). Sixty-two patients had incomplete EEG resection and, when a lesion was present, completeness of lesional resection did not predict seizure freedom (Table). For the secondary analysis, the control group included 48 lesional patients with complete resection. When compared to lesional patients with incomplete resection, extraoperative subdural EEG was less commonly performed (p=0.006). Patients incomplete by both MRI and EEG or EEG alone were significantly more likely to continue seizing 2 years post-op (Table). Incomplete resection of the abnormal MRI alone was not predictive of seizure persistence. Conclusions: Our findings indicate that the classic epilepsy surgery paradigm mandating complete anatomic (MRI) and functional (EEG) resection of the epileptogenic region to achieve seizure freedom may be inaccurate. Contrary to prevailing beliefs, over a third of patients with incomplete resection become seizure-free. In lesional patients, seizure-freedom may be achieved by complete functional excision alone, a finding of significant importance in presurgical counseling.