Abstracts

Large series reporting seizure outcome following temporal resection in patients with bitemporal EEG abnormalities are few (Hirsch et al., Ann Neurol 1991;30:347-356; Sirven et al., Ann Neurol 1997;42:873-878; So et al., Ann Neurol 1989;25:432-439). Many centers are reluctant to offer surgery to medically refractory temporal lobe epilepsy patients who have bilateral abnormalities, in particular those with bilateral seizures. A cumulative database (1990-2001) of temporal lobectomies performed on adults at the Cleveland Clinic Foundation was retrospectively reviewed for those with neurophysiological evidence of bilateral epileptiform activity (seizures or interictal spikes). Data abstracted included demographic variables, duration of epilepsy, epilepsy risk factors, seizure frequency and type, neuroimaging findings, and results of extended EEG evaluations, including invasive recordings. Primary outcome variable was postoperative seizure recurrence. Data was screened with a univariate analysis using [chi]² and t-test; variables thus identified were further analyzed using multivariate logistic regression. 100 patients were identified-23 with bilateral seizures, and 77 with unilateral seizures and bilateral spikes. Seizure-free rates were 95% at 2 years, 73% at 5 years, 53% at 8 years, and 40% at 10 years. There were no differences in seizure outcome between those with bilateral seizures (15/23) versus those with only spikes bilaterally (46/77, p[gt]0.50 by [chi]²). Results suggested lesser seizure-free rates associated with a history of generalized seizures (p=0.006), more than 20 seizures a month before surgery (p=0.038), a positive family history of epilepsy (p=0.015), or absence of history of febrile seizures (p=0.014). Results suggesting higher seizure recurrence in those with bilateral abnormalities on MRI and PET did not achieve significance. Patients who required invasive recording to clarify laterality of disease burden prior to lobectomy had similar seizure-free rates (18/32) to those not invasively evaluated (41/68, p[gt]0.50). Regression modeling broadly confirmed these results. Good seizure-freedom rates can be achieved in carefully selected patients with pharmacoresistant epilepsy and bilateral epileptic EEG features, including those with bilateral seizures. Other characteristics of this cohort, in particular the strategies utilized in selecting those with bilateral seizures who were offered surgery, will be discussed.