The utility of 18F-fluorodeoxyglucose PET (FDG PET) in epilepsy surgery.

Purpose: To investigate the utility of 18F-fluorodeoxyglucose Positron Emission Tomography (FDG PET) in helping decision making for epilepsy surgery.

Methods: All patients with medically refractory focal epilepsy and MRI that was normal or discordant with clinical and EEG data underwent FDG PET. FDG PET scans were reported by two investigators blinded to clinical data using visual assessment aided by the semiquantitative assessment. All clinical, MRI and FDG PET data were reviewed in the multidisciplinary patient management conferences for the localization and further decisions, which were recorded in the electronic database. For this study, we reviewed the charts of all these patients to decide the usefulness of PET in further decision making. FDG PET was considered to be useful if led directly to surgery, helped in planning intracranial EEG or helped in excluding patients from further evaluation.

Results: 194 consecutive adult patients (median age, 32.5 years) underwent FDG PET; 158 had normal MRI, 12 had subtle MRI abnormalities and 24 had discordant non-invasive data. Final localization was temporal lobe epilepsy (TLE, n=64), frontal lobe epilepsy (FLE, n=66), temporal-plus epilepsy (n=26) and other extratemporal lobe epilepsies (ETE, n=38). PET scans were normal in 72 (37%) patients, showed unifocal hypometabolism in 98 (50.5%) and bilateral hypometabolism in 24 (12%) patients. The TLE group had a higher proportion of abnormal PET scans (67%) than FLE (52%) and ETE (61%). PET data were useful in 103 (53%) patients, more in TLE (63%) than FLE (38%) or ETE (50%). It led directly to surgery in 12 (6%) cases, helped in planning intracranial EEG in 67 (35%) patients and excluded 24 (12%) patients from further evaluation. Focal hypometabolism on FDG PET increased the odds of being selected for surgery or intracranial EEG by five fold [odds ratio, 5.1 (2.8-9.4); p<0.0001].

Conclusions: FDG PET scan can help decision making in 53% of presurgical patients with normal or discordant MRI. PET findings need to be evaluated in conjunction with other data.