Summary: Purpose: To study the value of [18F]2-deoxyglucose (FDG)-positron emission tomography when surface ictal EEG is nonlocalizing.
Methods: FDG-PET scans were performed in 46 patients with complex partial seizures (CPS) not localized by ictal SUT-face-sphenoidal video-EEG (VEEG) telemetry. Interictal PET was performed with continuous EEG monitoring, and images were analyzed with a standard template. Forty patients subsequently had subdural and 6 had depth electrodes (invasive EEG, IEEG); 22 had bilateral implants. A focus was detected in 40, and 35 had temporal lobectomy based on IEEG localization.
Results: There was a close association between IEEG and PET localization (p < 0.01): 26 patients had relative unilateral temporal FDG-PET hypometabolism, all had congruent IEEG, and 18 of 23 were seizure-free after temporal lobectomy. Five patients had unilateral frontotemporal hypometabolism (3 of 5 were seizure-free), 1 had frontal hypometabolism, and 14 had no lateralized PET abnormality (4 of 7 were seizure-free). Patients who became seizure-free had significantly higher lateral temporal asymmetry index (AI). PET showed 315% relative temporal hypometabolism (AI) in 12 of 22 patients with non-lateralized surface ictal VEEG and was capable of distinguishing between frontal and temporal foci in 16 of 24 patients with lateralized, but not localized, surface ictal video-EEG.
Conclusions: FDG-PET provides valuable data in patients with unlocalized surface ictal EEG and can reduce the number of patients who require IEEG studies. Quantitation is necessary for optimal PET interpretation.