Bhawana Rathore MD, Presenter: Nothing to Disclose

Vina Ravichandran BA, Abstract Co-Author: Nothing to Disclose

Pearce Korb MD, Abstract Co-Author: Nothing to Disclose

James Roland Galt PhD, Abstract Co-Author: Nothing to Disclose

Robert E. Gross, Abstract Co-Author: Nothing to Disclose

David M. Schuster MD, Abstract Co-Author: Nothing to Disclose

Bruce Jonathan Barron MD, Abstract Co-Author: Stockholder, Immunomedics Inc

Larry Olson, Abstract Co-Author: Nothing to Disclose

Jonathon Nye PhD, Abstract Co-Author: Consultant, Lantheus Medical Imaging, Inc Consultant, American College of Radiology

Hamilton Elizabeth Reavey MD, Abstract Co-Author: Nothing to Disclose

Approximately 30% of patients with epilepsy are refractory to medications and may require resective brain surgery. Identification of candidate regions of seizure onset is crucial to successfully guide resection or placement of surgically implanted electrodes for intracranial electroencephalography (iEEG). However many patients do not have visually identifiable lesions on brain MRI or PET making it a challenge for surgical planning. The primary aim of this proof of concept study is to determine if the most hypometabolic regions of the interictal brain PET using quantitative analysis software correlates with the seizure onset zone determined by subsequent iEEG. .

Eighteen interictal PET-CT scans of brain in patients who had also undergone iEEG were retrospectively reviewed. All patients originally had MR and qualitative PET interpreted as negative. The studies were then processed with quantitative analysis software (MimNeuro 5.6; Cleveland, Ohio) which compares PET images to a normal database. The 10 most hypometabolic foci were recorded on a scale of 1-10 with 1 being most hypometabolic. Foci which corresponded to white matter or cerebellum on co-registered CT were eliminated from analysis since these are rare locations for origin of seizures. Candidate foci based on rank order of hypometabolism were then compared to the actual location of the seizure onset zone as identified on iEEG.

Mean (±SD) age was 43.6 (±11.7); range 24-60 years. Ten patients were male and 8 female. For all 18 patients the seizure onset zone from the iEEG correlated with one of the hypometablic foci on the quantitative PET analysis. Mean (±SD) rank order of hypometabolism for the seizure onset zone was 1.7 (±1.1); range 1-5. In 10 out of 18 patients (56%) the most hypometabolic focus correlated with the proven seizure onset zone on iEEG.

Quantitative analysis of PET hypometabolism may be useful in identifying candidate seizure onset zones and guiding placement of intracranial electroencephalography electrodes even in patients with negative MR and unrevealing qualitative PET. Further analyses with a larger sample size and co-registration of PET and MR is warranted to determine diagnostic performance.

Quantitative analysis of PET hypometabolism may bring added value to pre-operative identification of seizure foci not detected on routine MR & qualitatively interpreted PET thereby improving outcomes.

Rathore, B, Ravichandran, V, Korb, P, Galt, J, Gross, R, Schuster, D, Barron, B, Olson, L, Nye, J, Reavey, H, Quantitative Analysis of FDG PET Hypometabolism in Pre-operative Identification of Seizure Foci Not Detected on Routine MR and Qualitative PET.  Radiological Society of North America 2013 Scientific Assembly and Annual Meeting, December 1 - December 6, 2013 ,Chicago IL. http://archive.rsna.org/2013/13016930.html