Factors associated with failed focal neocortical epilepsy surgery

Abstract

Background: Seizure outcomes after focal neocortical epilepsy (FNE) surgery are less favorable than after temporal lobectomy, and the reasons for surgical failure are incompletely understood. Few groups have performed an in-depth examination of seizure recurrences to identify possible reasons for failure.

Objective: To elucidate factors contributing to FNE surgery failures.

Methods: We reviewed resections for drug-resistant FNE performed at our institution between 1998 and 2011. We performed a quantitative analysis of seizure outcome predictors and a detailed qualitative review of failed surgical cases.

Results: Of 138 resections in 125 FNE patients, 91 (66%) resulted in freedom from disabling seizures (Engel I outcome). Mean ± SEM patient age was 20.0 ± 1.2 years; mean follow-up was 3.8 years (range, 1-17 years); and 57% of patients were male. Less favorable (Engel II-IV) seizure outcome was predicted by higher preoperative seizure frequency (odds ratio = 0.85; 95% confidence interval, 0.78-0.93), a history of generalized tonic-clonic seizures (odds ratio = 0.42; 95% confidence interval, 0.18-0.97), and normal magnetic resonance imaging (odds ratio = 0.30; 95% confidence interval, 0.09-1.02). Among 36 surgical failures examined, 26 (72%) were related to extent of resection, with residual epileptic focus at the resection margins, whereas 10 (28%) involved location of resection, with an additional epileptogenic zone distant from the resection. Of 16 patients who received reoperation after seizure recurrence, 10 (63%) achieved seizure freedom.

Conclusion: Insufficient extent of resection is the most common reason for recurrent seizures after FNE surgery, although some patients harbor a remote epileptic focus. Many patients with incomplete seizure control are candidates for reoperation.

FIGURE 1

Predictors of Engel I seizure outcome from multivariate analysis. Shown are the results of multivariate logistic regression analysis of potential predictors of favorable (Engel I) seizure outcome, with data represented as odds ratio 6 95% confidence interval. Among all variables examined (see Methods and Tables 1-5), only those with a value of P < .20 on univariate analysis were included in the multivariate model. A higher preoperative seizure frequency, history of generalized tonic-clonic seizures (GTCS), and normal magnetic resonance imaging (MRI) were associated with a less favorable (Engel II-IV) seizure outcome. Seizure frequency was entered into the multi-variate model as a continuous variable, whereas history of GTCS and MRI findings are dichotomized categorical variables.

FIGURE 2

Flow chart summarizing factors associated with seizure recurrence. Among 47 cases with less favorable (Engel II-IV) seizure outcome, 36 had sufficient postoperative diagnostic workup available for further qualitative analysis. Of these 36 resections, 26 cases of seizure recurrence (72%) were associated primarily with extent of resection, with evidence suggesting residual epileptogenic tissue adjacent to the resection cavity, and 10 cases (28%) were more closely associated with location of resection, with evidence of an additional epileptic focus distant from the resection. Within these 2 categories, other commonly observed factors potentially related to seizure recurrence are also listed, with >1 factor noted in some cases. ECoG, electrocorticography; EEG, electroencephalography; MRI, magnetic resonance imaging; MTS, mesial temporal sclerosis.