Abstracts

A major complication rate of 2-6% for resective epilepsy surgery procedures, and 6-20% for intracranial electrode placements has been established based on several large series. These data have been reported principally from centers with either very experienced neurosurgeons, multiple surgeons with varying years of training and experience, and/or residency training programs, all factors that impact complication rates. Further, they do not take into account the [quot]learning curve[quot] of new surgeons. We reviewed the inital 5 year experience of a single neurosurgeon at a non-teaching center to determine complication rates in this setting.
We reviewed all epilepsy surgery procedures performed by a single surgeon (ANM) during the inital 5 years post training to identify surgery-related complcations. Both intracranial electrode placements and resective or ablative procedures were evaluated. Neurological deficits were considered minor if they resolved within 3 months, and major if they persisted [gt]3 months or were permanent. Analysis of complications by year since completion of neurosurgical training was performed.
During the 5 year period (1998-2002) 201 surgical procedures were performed, including 74 resections, 7 corpus callosotomies, 78 electrode implantations, and 42 electrode removals without resection. This included 46 temporal lobectomies, 3 amygdalohippocampectomies, 22 neocortical resections, 2 multilobar resections, and 2 hemispherectomies. There were 31 subdural grid implants, 37 depth electrode implants, and 10 combined subdural strip and depth electrode implants. 23 (11.1%) complications were noted. Eight (10%) new neurological deficits were observed in 7 (8.6%) patients, of which 4 resolved or dramatically improved within 3 months. These included hemiparesis (n=4), homonomous hemianopsia (n=2), and aphasia (n=2). All occured in patients undergoing temporal lobe resections, and all occured in the first 2.5 years of surgical practice, with no permanent deficits in the last 2.5 years. Infectious complications were noted in 9 patients, all of whom had subdural grids placed for periods of 2 or more weeks. These included meningitis (n=7), epidural empyema (n=1), and osteomyelitis (n=1). These complications were evenly distributed, with frequency correlating to the number of procedures done per year, not years of surgical experience. No major complications from depth electrode implantations were noted. There was one death due to status epilepticus 18 days after callosotomy, and 2 cases of deep venous thrombosis in the post-operative period.
Complication rates for new epilepsy surgeons may be slightly higher than those reported by more experienced surgeons, but not substantally different. Permanent neurological deficits following resective epilepsy surgery are more frequent in the first 2-3 years of surgical experience, and diminish in subsequent years of surgeon experience. In contrast, infectious complications are more common with subdural grids and do not appear to improve with increased surgical experience.