UTILITY OF AMYLOID IMAGING USING PET IN A CLINICAL SETTING

There is growing interest in biomarker use for the diagnosis of Alzheimer's disease (AD). Amyloid plaques are a core pathologic feature of AD that can be detected in the brain using positron emission tomography (PET) scans. The goal of the study was to assess the clinical utility of Florbetapir PET by examining outcomes of patients enrolled in the Imaging Dementia-Evidence for Amyloid Scanning (IDEAS) study at a single center, the Johns Hopkins Memory and Alzheimer's Treatment Center (JHMATC) that has ∼3,000 encounters/year with memory disorder patients. 13 JHMATC physicians referred 51 patients to IDEAS between June 2016 and December 2017. Referring physicians completed a diagnostic pre-PET on all patients and a post-PET on the subset of patients in Phase I of IDEAS. Patients were evaluated with Florbetapir PET within 30 days of signing consent. Scans were read by a single, trained nuclear medicine physician. Mean patient age was 73.4 (SD=5.4); 24 (47.1%) were male. 61% (31/51) had a positive scan with the following outcomes based on pre-PET diagnoses: 40 with a clinical AD prodrome (37.5% positive Florbetapir scan); 2 with clinical FTD (50% positive); 9 with other diagnoses (44% positive). 31/51 took part in Phase I of IDEAS had a completed post-PET form: 12/13 with a positive scan had an AD diagnosis at follow up. None of 18 patients with a negative scan had an AD diagnosis at follow up. A multivariable logistic regression was estimated to predict positive or negative scan results with MMSE scores, age, and sex. The model was significant (p = 0.026) and explained 22% of the variance in scan results (Nagelkerke's R2 of 0.228). Lower MMSE scores were independently associated with a positive scan (p= 0.019). Negative Florbetapir scans were more likely to lead to change in diagnosis. Lower MMSE scores were predictive of a positive scan. In combination with clinical assessment, results of Florbetapir scans are helpful primarily to rule out an AD diagnosis in patients with uncertain etiology of memory impairment.