Abstracts

Rationale: Cognitive outcomes after temporal lobectomy for treatment of intractable epilepsy are well characterized in the literature. However, despite being the second most common type of partial epilepsy, outcomes after frontal resection are not well known. This study examined changes in verbal fluency in patients who underwent surgery for frontal lobe epilepsy (FLE). Methods: A retrospective chart review of patients who underwent epilepsy surgery at the Cleveland Clinic from 1991 through 2008 was performed. Inclusion criteria were: aged 18 or older, diagnosis of frontal lobe epilepsy, completed pre and postoperative neuropsychological testing that included phonemic verbal fluency, and available postoperative MRI. Patients with prior epilepsy surgeries were excluded. Resection site and extent was characterized using imaging software (MRICro) to identify the involvement of the following regions: superior medial frontal (SMF), inferior medial frontal (IMF), orbitofrontal (OF), and lateral frontal (LF) lobe (with or without Brodmann area 44). Patients were categorized into two groups based on postoperative change in their verbal fluency scores using standardized regression based change scores (i.e., decline n=10 and no decline n=27). A series of logistic regression analyses were then conducted to identify predictors of decline in verbal fluency. Results: 37 patients were identified who met inclusion criteria (19 right FLE; 18 left FLE), 41% were female. Average age was 30.24 10.58 years. Preoperative seizure frequency was 54.4 seizures per month, and 68% were seizure-free at the time of postoperative neuropsychological testing. There were no significant differences between the two study groups in age, education, age at seizure onset, duration of epilepsy, Full Scale IQ, or preoperative verbal fluency performance. Patients without verbal fluency declines were more likely to be seizure free after surgery, and there was a trend for them to have a longer test-retest interval as compared to the decline group. Preoperative imaging showed MRI abnormalities in 62% and PET abnormalities in 89%. The area of resection included: SMF (95%), IMF (66%), OF (50%), LF with Brodmann 44 (18%), and LF without Brodmann 44 (63%). Logistic regression using relaxed input criteria (p = .1 to add to model, p = .15 to remove from model) indicated both lesion location (p = .051) and preop FAS score ( p = .112) are predictive of verbal fluency decline (R2 = .155; p = .052). These two variables accurately predicted decline in 71% of patients. As noted, patients who declined on verbal fluency were also less likely to be seizure free after surgery. Conclusions: Twenty seven percent of patients who underwent FLE surgery showed a decline in verbal fluency. This decline did not appear to be due to general intellectual decline or post-operative aphasia. Preoperative predictors of decline include a high pre-surgical verbal fluency score and a left lateral frontal lobe resection (with sparing of Broca s area). Cognitive deficits can be noted postoperatively even if patients have resections of the non-eloquent cortex.