Objective: There are few modern data on the complications of surgery for epilepsy from the neurosurgeon's point of view. A survey of complications observed in a large current epilepsy surgery series is presented to facilitate the assessment of a risk:benefit ratio, which must be known when planning for epilepsy surgery and counseling patients.
Methods: A series of 429 consecutive patients operated on during 6.5 years in the newly established University of Bonn epilepsy surgery program was, in part, retrospectively, and, in larger part, prospectively analyzed for complications originating from 279 invasive diagnostic procedures and 429 therapeutic procedures. Neuropsychological and psychiatric complications as well as the rate of failure to control seizures are not addressed in this article.
Results: Two hundred and seventy-nine temporal operations, 59 frontal operations, 22 other extratemporal operations, 33 callosotomies, 3 multilobectomies, and 33 hemispherectomies were performed. Complications were grouped into general surgical and neurological complications. No mortality resulted from 708 invasive procedures. Two hundred and seventy-nine invasive diagnostic procedures (various combinations of strip, grid, and depth electrode insertions) resulted in 3.6% transient morbidity (2.9% surgical complications, 0.7% neurological complications) and 0.7% permanent morbidity (dysphasia). During 429 therapeutic procedures, 33 surgical complications were encountered. None of these resulted in permanent morbidity, except for the necessity for permanent shunt insertion in three patients. Wound infection was the most frequent surgical complication, but we were able to demonstrate a steady decrease during the 6.5-year observation period. The total rate of neurological complications in 429 therapeutic procedures was 5.4%, with 3.03% causing transient morbidity and 2.33% causing permanent morbidity.
Conclusion: Our data indicate that epilepsy surgery can be performed with an acceptable rate of resultant morbidity. The indications for epilepsy surgery, the learning curve determined, and the results from other series are discussed in the light of these figures.