Clinical impact of amyloid PET in memory clinic: A 2-year longitudinal real-world study in Taiwan.

INTRODUCTION: Longitudinal evidence on amyloid positron emission tomography (PET) for Alzheimer's disease (AD) diagnosis in Taiwan and Asian memory clinics remains limited. This study evaluated the diagnostic and prognostic impact of florbetaben PET in a Taiwanese cohort. METHODS: This prospective cohort study enrolled 206 memory clinic participants who underwent florbetaben amyloid PET, volumetric magnetic resonance imaging, apolipoprotein E genotyping, and neuropsychological assessment, with 2-year longitudinal follow-up. RESULTS: Amyloid positivity was present in 48% of participants and drove diagnostic reclassification in 49.5% of the cohort. Receiver operating characteristic curve analysis identified an optimal Centiloid (CL) threshold of 21.60 (sensitivity 91.18%, specificity 99.04%, area under the curve [AUC] = 0.98). Amygdala, inferior parietal, and superior temporal atrophy were strong structural predictors of amyloid positivity (AUC = 0.79). DISCUSSION: These findings establish a locally validated amyloid threshold and provide actionable diagnostic and prognostic stratification in a real-world Asian memory clinic.