Abstracts

We set out to describe the natural history of first, second and third seizure recurrence in a large epilepsy surgery cohort and to determine if timing of the initial seizure recurrence predicted future recurrences. We also also studied whether temporal lobectomy patients had different recurrence patterns than patients with extratemporal resections.
All patients who had resection for refractory epilepsy at Graduate Hospital or Thomas Jefferson University Hospital in Philadelphia between April 1986 through January 2001 were included in this study. Latency from surgery to the first, second and third days on which postoperative seizures occurred were registered. Where appropriate, Chi square tests were used for bivariate comparisons. Kaplan-Meier curves were generated, and groups were compared using the log-rank test
A total of 561 patients (298 male vs. 263 female) underwent resective surgery during the study period; 491 had a temporal lobectomy and 70 had an extratemporal resection. Two hundred eighty nine patients (51.5%) had at least one post-operative seizure. The median time to first recurrence was 0.3 years . Of patients who had a first recurrence, 69.9% had a second recurrence. Of patients who had a second recurrence, 89.6% had more seizures. We split the cohort into five groups depending on the time of their first post-operative seizure (group 1; [lt] 1 month post operatively to group 5; [gt]5 years post-op). Latency from surgery to the first recurrence predicted the likelihood of future recurrences. The proportion experiencing a second seizure was 92.4%, 75.5%, 58.5%, 36.4% and 26.6% in groups 1 [ndash] 5. . Early seizure recurrence significantly decreased the chance of being seizure free at last contact (30-40%) and increased the chance of being intractable (50%). Although latency to subsequent seizure decreased with each successive seizure, it did not appear to influence the outcome as of last contact. There was no difference in recurrence patterns between temporal lobectomy patients and those with extratemporal resections.
These results underscore the importance of the timing of postoperative seizure recurrence with regard to long-term prognosis. The lack of difference in recurrence latency between temporal and extratemporal patients is surprising given the higher rates of long-term remission after temporal lobectomy, and highlights the need for using multiple methods of outcome assessment. Prompt consideration of further surgery may be wise in patients who recur soon after surgery. A less aggressive approach seems in order for patients whose first recurrence is two years or more after surgery especially as some of these recurrences may occur in the context of planned drug withdrawal.