Abstracts

Rationale: It has been reported that FDG-PET studies performed in children with continuous spike-waves during slow sleep (CSWS) show increase of metabolism at the site of the epileptic focus, associated with decrease of metabolism in distant and connected brain areas. However, evaluation of the PET findings in patients with MRI abnormalities are complicated. We performed serial FDG-PET studies in patients with CSWS and related conditions to assess the utility of subtracted PET analysis in patients with or without MRI abnormalities.Methods: We studied six patients with CSWS and related conditions. Two patients had MRI abnormalities and four had normal MRIs. Patients with normal MRIs consisted of one patient with Landau-Kleffner syndrome (LKS), one with oro-motor deficits and two others with CSWS. MRI abnormalities in two patients were right posthemorrhagic porencephaly and right middle cerebral artery infarction. Interictal FDG-PET were performed during the acute period with CSWS and recovery period without CSWS after treatments. PET images in the acute and recovery periods were co-registered and radio-activities were normalized to the mean activities of the whole brains. The image in the acute period was subtracted from the image in the recovery period to create an image of the difference between the two data sets. The standard deviation (SD) of the distribution of the subtraction pixel intensities was calculated, and the subtraction image thresholded to display only pixels with values greater than 2 SDs above zero. The images are co-registered to individual's 3D MRI. The localization of the subtracted PET findings was assessed and compared with EEG focus of epileptiform activities. Two patients had another sets of the scans one year later.Results: In patients with normal MRIs, subtracted PET showed hypermetabolism in bilateral precentral and opercular areas in a patient with oro-motor deficits and in left superior temporal gyrus in a patient with LKS. One patient with MRI abnormality had increased uptake in the left frontal lobe contralateral to the MRI lesion on the original PET during acute stage. However, Subtracted PET showed significant changes in the right parietal lobe that showed EEG focus. Another patient with MRI abnormality had increased uptake in the right lateral frontal areas surrounding the lesion on the first subtracted PET at the first scan. On the other hand, second subtracted PET showed abnormalities in the bilateral medial frontal areas. The EEG focus recorded during each period was consistent with the distribution of subtracted PET abnormality.Conclusions: In patients with MRI abnormalities, visual inspection of original PET images may cause misleading of functional abnormalities related to CSWS. Subtracted PET analysis is useful to evaluate the dynamic change of cerebral glucose metabolism in patients with CSWS, especially with MRI abnormalities.