Abstracts

Rationale: Intracranial EEG (iEEG) recording is considered to be the gold standard for seizure localization in many patients with epilepsies that are intractable and complex. Due to inherent morbidities associated with invasive neurosurgery and intracranially indwelling foreign substance, the extent of iEEG has to be judiciously determined and limited to reduce the risk of complications. Even though iEEG implantation is guided by thorough presurgical evaluations, including scalp EEG and imaging studies, there are still situations in which iEEG is not able to localize the seizure-onset zone. Repeat surgery is sometimes required to revise the placement of the intracranial electrodes. Given the increased risk of repeat brain surgery, it is important to evaluate the diagnostic yield of repeat surgeries for revising intracranially implanted electrodes. Methods: We retrospectively reviewed our epilepsy surgery database between 1997 and 2010. We identified patients who had revision of intracranial electrode placements during one episode of inpatient evaluation, because the initial implantation did not adequately localize seizure onset. Data collected included seizure semiology, interictal and ictal scalp EEG discharges, MRI head, and initial iEEG findings. Fisher's exact test was used to evaluate for statistical significance of each factor. Results: Twenty patients consecutively fit our inclusion criteria. Two patients did not have another seizure after the revision of the electrode placements. Revised implantation and recording localized seizure onset in 10 of the remaining 18 patients (55.6%). Factors that were statistically significant in localizing a seizure focus after implantation revision are: 1) lateralizing seizure semiology (71.4% in lateralizing vs. 0 % in non-lateralizing; p = 0.023); 2) focal scalp interictal discharges (100% in focal vs. 38.5% in non-focal; p = 0.036); and 3) initial iEEG showing ictal onset at the edge of the electrode grid (100% in ictal onset at the edge of the electrode grid vs. 20% in indeterminate ictal onset; p = 0.001). Focal scalp ictal EEG onset could be a prognostic factor, because it nearly reached statistical significance (80% in focal vs. 25% in non-focal; p = 0.054). MRI head findings (normal vs. abnormal) did not show an association in this study. There was no permanent complication associated with revised implantation but one patient experienced transient right foot apraxia. 6/10 patients who underwent resective surgery after localizing revised implantation were seizure- free during a limited period of follow-up (1 month to 4 years). Conclusions: Our study shows that the diagnostic yield of repeat surgeries for revising intracranially implanted electrodes is about 55%. However, the yield is improved by a focal ictal onset at the edge of the original implanted grid, lateralizing seizure semiology, or focal scalp interictal discharges.