Abstracts

Rationale: Resective surgery for temporal lobe epilepsy carries a failure rate ranging from 20-40%. The cause of failure is complex and multifactorial, and some patients are amenable to reoperation. To assess factors that play a role in failed surgery, the decision to offer a repeat surgery, and the success rate of reoperation, we undertook a systematic review of the literature. Our aim was to assess not only the results of the published evidence, but also its methodology, shortcomings, and areas that require addressing in order to understand this problem better. Methods: We systematically and exhaustively searched the literature from 1990 to 2007, using a broad array of MESH terms and text words to explore the Medline database, including “failed surgery”, “re-operation”, “second surgery”, and “repeat surgery”, and combined these with our broad target entity “epilepsy” or “seizure”. Because this yielded only nine articles, we then searched ISI-web of science for articles citing these references. We also searched manually the references of the key articles, book chapters, and asked experts. A total of nine papers were found, of which eight dealt with outcomes. For each paper we abstracted data on demographics, clinical features, surgery, etiology, and surgical results.Results: Eight articles included 186 patients, ranging from 5 to 44 (median 22) per article. The proportion of patients undergoing reoperation ranged from 4% to 15%. The minimum interval between 1st and 2nd surgery was 3.5 years, but varied widely. The most common etiology was mesial temporal sclerosis (median 52%). Only two papers (25%) gave information on the initial surgery. Only one paper (12.5%) provided data on all relevant figures (number initially operated, number failed, number eligible for reoperation, and number reoperated). No paper gave information on the use of intracranial EEG for reoperation. Six (75%) studies attributed surgical failure to the presence of residual MTS, and one to dual pathology. Seizure freedom was obtained in nearly 50% of reoperated patients, with minimal permanent neurological complications, especially in patients with residual MTS. Conclusions: There is a profound deficiency of sound evidence to inform clinical decision making regarding re-operation after failed surgery. Reports are few, involve few patients and do not contain the standardized information. With this limited information, the best results were obtained with residual MTS. Arguably; residual MTS will become infrequent with better localization and surgical techniques, including intraoperative neuronavigation and intraoperative MRI. However, much remains to be done to understand the determinants of failure, reoperation and outcome following the latter. At the very least, it is important to examine all the failed cases, and decrease the current selection bias in reports (only 4-15% are included). This gives a false sense of security with regards to outcomes of reoperation.