Abstracts

Rationale: Epileptic auras are simple partial seizures associated with a localized ictal discharge close to the seizure onset zone. This retrospective study looked at the outcome of temporal lobectomy in medically refractory patients based upon the characterization of the aura. Methods: Patients were selected from the epilepsy surgery database of the South Texas Comprehensive Epilepsy Center at San Antonio, Texas. Patients with lesional temporal lobe epilepsy (including hippocampal sclerosis) were included who had at least one year postoperative follow-up. The characterization of the aura was compared between patients with right (RTLE) or left (LTLE) temporal lobe resections and between Engel Class 1 and 2-4 outcomes at one-year follow-up. The auras were classified as visual, auditory/aphasic, gustatory/ olfactory, somatosensory, viscerosensory (epigastric, headache), psychic (déjà vu, jamais vu, anxiety/fear), vertigo/dizziness, autonomic (cold chills, sensation of heat, sweaty palms, need to urinate, palpitations) and others. Results: Thirty-five patients had right TLE and 38 patients had left TLE. Auras were characterized in 25 (71%) of the RTLE patients and 32 (84%) of the LTLE patients. Seventeen (23%) of all patients had multiple auras. 4 (0.5%) patients could not describe and 16 (22%) did not have auras. The most frequent type of aura was viscerosensory, which was identified in 23 (32%) patients (14 RTLE vs 9 LTLE). Lateralization of the auras was noted with gustatory/olfactory (3 RTLE vs 0 LTLE), somatosensory (4 RTLE vs 0 LTLE) and visual auras (4 RTLE vs 0 LTLE). Class 1 outcomes were associated predominantly with viscerosensory ( 61%), somatosensory (75%) and visual (75%), psychic (60%) auras, whereas vertigo/nausea (83%), olfactory/gustatory (67%) and (56%) autonomic auras were associated with Class 2-4 outcomes. Conclusions: Epileptic auras can help to localize the epileptogenic zone. Furthermore they may help to predict the postoperative outcome after TLE. Some auras appeared to be associated with a right hemispheric predominance. Whether this association is due to inter-hemispheric differences of circuitry and propagation is unclear, but it may reflect a preoperative selection bias.