Predicting outcome following reoperation for medically intractable epilepsy

The purpose of this study was to determine predictors of probable outcome following reoperation for medically intractable partial epilepsy. We reviewed outcome at least 1 year after reoperation in 21 patients with intractable seizures, for whom an earlier operation had failed. We examined age of onset of epilepsy, duration of seizures, gender, details of the history and clinical examination, pre-operative magnetic resonance (MRI) findings, electroencephalographic (EEG) studies obtained before and after the failed surgery, presence or absence of lateralizing neuro-psychological deficits, sites of operation and pathology of resected tissue to identify the factors associated with outcome. We found two factors that were significantly related to outcome: (1) no individual with a history of central nervous system (CNS) infection which predated the onset of epilepsy had a seizure-free outcome after reoperation (P = 0.04). (2) Reoperations that extended previous resections, based on new ictal EEG recordings that were concordant with both EEG ictal onsets and MRI findings obtained before the first, failed surgery resulted in a seizure-free outcome or >95% reduction in seizures for 100% (7/7) of such patients. This compares to 29% (4/14) of the remaining individuals without such concordance who had a similar outcome (P = 0.009). Site of operation (temporal or extratemporal) did not, in and of itself, predict outcome. A portion of patients who fail surgery for intractable partial seizures will achieve significant improvement following reoperation. Furthermore, we may be able to identify those individuals most likely to have an excellent result from a second operation.