Abstracts

Rationale: Patients (pts) with intractable nonlesional focal epilepsies (NLFE) constitute a most challenging group undergoing presurgical evaluation, and are often denied epilepsy surgery. Yet surgery can be effective in well-selected pts with no visible MRI lesion. Literature on pathology of intractable NLFE is limited. These pts suffer from a heterogeneous group of epileptic pathologies that are not well understood. We sought to characterize the underlying histopathology and correlate this with seizure outcome in a large series of strictly-defined MRI-negative pts undergoing epilepsy surgery. Methods: All pts with NLFE who had surgery over a 10-year period (2002-2011) were identified. High-res 1.5 or 3T preop MRI was performed using a uniform epilepsy protocol. We included pts who had a preop MRI between 6-65 years of age, and further restricted our study to pts with strictly-defined normal MRI: Preop MRIs were visually analyzed by dedicated neuroradiologists and re-reviewed by neuroradiology and neurosurgery during pt management conference guided by results of noninvasive testing. Pts with MRIs that were initially interpreted as normal were excluded, if a probable/questionable lesion was identified during focused re-review of the MRI. Pathological specimens were carefully re-reviewed by a single neuropathologist and characterized as focal cortical dysplasia (FCD) type I (Palmini) or balloon-cell type, hippocampal sclerosis (HS), gliosis, normal tissue and other. Resections were classified as frontal, parieto-occipital, neocortical temporal (T), standard T with removal of mesial T structures and multilobar. Results: 94 consecutive pts (mean age 30.4yrs, range 7-64yrs; 50% females) fulfilled the above criteria. 34 underwent hippocampal-sparing partial T resections, 32 standard T lobectomies, 20 frontal, 6 parieto-occipital and 2 bilobar resections. The majority had type I FCD (n=36; 38.3%). Balloon-cell FCD was identified in 3; HS in 10 (3 with FCD; dual pathology); gliosis in 34 and others in 4. Lastly 7 had no identifiable pathological changes. Follow-up data for >6 months postop were available in 87. 43 (or 46% of the entire cohort) attained Engel class 1 outcome. Among seizure-free pts 19 (44.2%) harbored type I FCD, and 2 balloon-cell FCD, 7 (16.2%) had HS, 2 other pathologies, 11 (25.5%) gliosis and 2 had no abnormalities. Conclusions: We present results of surgical pathology in a large single-center series of pts with strictly-defined NLFE undergoing epilepsy surgery. The most common pathological correlate not visualized by MRI was FCD, identified in 39 pts of whom only 3 had balloon-cell FCD. Pts with identifiable pathology appeared to have better chance for seizure-free outcome compared to pts without specific epileptic pathology (gliosis or normal). At the same time paucity of pathological changes in the resected specimen did not preclude favorable seizure outcome. These results highlight the heterogeneity of epileptic pathologies that cannot be visualized with available MRI technology, and the limitations of microscopic histopathological examination of resected human epileptic tissue.