Positron Emission Tomography in Evaluation of Dementia

Context Deficits in cerebral glucose utilization have been identified in patients with cognitive dysfunction attributed to various disease processes, but their prognostic and diagnostic value remains to be defined. Objective To assess the sensitivity and specificity with which cerebral metabolic patterns at a single point in time forecast subsequent documentation of progressive dementia. Design, Setting, and Patients Positron emission tomography (PET) studies of [18F]fluorodeoxyglucose in 146 patients undergoing evaluation for dementia with at least 2 years' follow-up for disease progression at the University of California, Los Angeles, from 1991 to 2000, and PET studies in 138 patients undergoing evaluation for dementia at an international consortium of facilities, with histopathological diagnoses an average of 2.9 years later, conducted from 1984 to 2000. Main Outcome Measures Regional distribution of [18F]fluorodeoxyglucose in each patient, classified by criteria established a priori as positive or negative for presence of a progressive neurodegenerative disease in general and of Alzheimer disease (AD) specifically, compared with results of longitudinal or neuropathologic analyses. Results Progressive dementia was detected by PET with a sensitivity of 93% (191/206) and a specificity of 76% (59/78). Among patients with neuropathologically based diagnoses, PET identified patients with AD and patients with any neurodegenerative disease with a sensitivity of 94% and specificities of 73% and 78%, respectively. The negative likelihood ratio of experiencing a progressive vs nonprogressive course over the several years following a single negative brain PET scan was 0.10 (95% confidence interval, 0.06-0.16), and the initial pattern of cerebral metabolism was significantly associated with the subsequent course of progression overall (P<.001)