Abstracts

Simultaneous cortical grid and intracerebral depth electrodes allow localization of seizure onset in patients with temporal lobe epilepsy undergoing intracranial evaluation, but the relationship to magnetic resonance imaging (MRI) findings is not established. We reviewed all temporal lobectomy epilepsy patients who underwent simultaneous cortical grid and intracerebral depth electrode monitoring at the University of Maryland Medical Center from 1987 - 2004. Patients had intracranial monitoring for further localization of seizure onset or for language mapping. We reviewed the electrophysiologic findings of seizure onset and correlated the findings with magnetic resonance imaging (MRI) and clinical outcomes. We analyzed 63 temporal lobectomy patients with simultaneous grid and depth electrodes. Adequate MRI and seizure data was available for 47 patients. Intracranial monitoring revealed mesial onsets in 32 patients (68%) and lateral onsets in 9 patients (19%). 6 patients had evidence of mesial and lateral onset (13%). 24 temporal lobectomy patients (52%) had evidence of mesial temporal lobe abnormalities on MRI and 23 (49%) had normal mesial temporal structures on MRI. The presence of mesial abnormalities on MRI was predictive of mesial onset in 20 of 24 patients (83%). In patients with normal mesial structures on MRI, only 12 of 23 (52%) had mesial onsets on EEG, 8 had lateral onsets (17%), and 3 (14%) had mesial and lateral onsets. In patients with mesial temporal abnormalities on MRI, 23 of 24 (96%) were seizure free (Engel Class I), and patients without mesial temporal abnormalities on MRI were less likely, 10 of 23 patients (43%), to be seizure free. The patients with mesial temporal onsets were more likely to be seizure free (81%) and patients with either lateral or bilateral onsets did more poorly, with only 47% with Class I outcomes. The 20 patients with both mesial EEG onset and mesial abnormalities on MRI had the best outcomes with 100% seizure freedom. Mesial seizure onsets were found in 68% of temporal lobectomy patients with combined grid and depth intracranial monitoring and were predictive of better outcomes than with lateral or mesial and lateral seizure onsets. In addition, the findings of mesial temporal abnormalities on preoperative magnetic resonance imaging were also associated with a better prognosis, even in patients without mesial onset on EEG. Patients with both mesial onsets on EEG and mesial abnormalities on MRI had the best prognosis with 100% seizure freedom. The technique of combining subdural grid and depth electrode recording is complementary to MRI findings and adds valuable information in predicting outcomes for patients undergoing temporal resection for refractory epilepsy.