Abstracts

Rationale: Neuroimaging plays a fundamental role in the evaluation of intractable epilepsy patients. While MRI represents the mainstay modality, interictal FDG-PET imaging has an established role in epilepsy imaging but may be underutilized. Specifically, quantification of brain metabolism is not routinely performed during clinical evaluation. This study seeks to assess the role of FDG-PET neuroimaging in the presurgical evaluation of patients with medically refractory epilepsy, in a cohort of patients at the Hospital of the University of Pennsylvania from 2003 through 2011. The goals of our study were (1) to determine if quantification of brain FDG-PET in patients with medically refractory epilepsy will detect additional epileptogenic lesions not identified on the qualitative interpretation and (2) to compare qualitative and quantitative FDG-PET assessment with post-surgical seizure freedom. Methods: 93 consecutive intractable epilepsy patients who underwent surgical resection between January 2003 and December 2011 were included. All had MRI studies, 77 also had brain FDG-PET studies. PET imaging studies were analyzed using Neurostat/3D-SSP. Qualitative and quantitative FDG-PET imaging findings were compared with MRI imaging findings as well as neurological data (EEG, seizure semiology, WADA testing), pathologic data, and postsurgical seizure freedom (Engel classification). Results: Of all 93 patients who underwent MRI imaging, mesial temporal sclerosis was the most commonly reported diagnosis, comprising 44% of cases. A normal study was reported in 25% of cases. Of all 77 patients who underwent both MRI and FDG-PET imaging, a definite hypometabolic focus suggestive of an epileptogenic focus was reported on qualitative interpretation in 48% of cases. A possible hypometabolic focus was reported is 16% of cases. A seizure focus could not be localized in 30% of cases. Preliminary quantitative analysis representative of the available PET studies (n=27/77) revealed all studies with a reported definite hypometabolic focus (n=13/27) were correctly identified using a quantitative approach when compared to surgical pathology. Furthermore, cases initially identified as containing a possible hypometabolic focus (n=6/27) were incorrectly localized by qualitative interpretation in approximately 33% of cases (2/6), but correctly localized using a quantitative approach. Cases initially reported as "non-localized" (n=6/27) were correctly reported using a quantitative approach in 50% of studies (3/6). Analysis of postsurgical outcome data (by Engel classification) is ongoing. Conclusions: Our preliminary results suggest that quantitative analysis of FDG-PET brain imaging using Neurostat/3D-SSP in patients with medically refractory epilepsy may be more sensitive than qualitative interpretation in up to 50% of studies where a definite hypoemetabolic lesion is not identified.