Kinetics of the Tau PET Tracer ^sup 18^F-AV-1451 (T807) in Subjects with Normal Cognitive Function, Mild Cognitive Impairment, and Alzheimer Disease

We report kinetic modeling results of dynamic acquisition data from 0 to 100 min after injection with the tau PET tracer ^sup 18^F-AV-1451 in 19 subjects. Methods: Subjects were clinically diagnosed as 4 young cognitively normal, 5 old cognitively normal, 5 mild cognitive impairment, and 5 Alzheimer disease (AD). Kinetic modeling was performed using Logan graphical analysis with the cerebellum crus as a reference region. Voxelwise binding potential (BP) and SUV ratio (SUV R^sub 80-100^) images were compared. Results: In AD subjects, slower and spatially nonuniform clearance from cortical regions was observed as compared with the controls, which led to focal uptake and elevated retention in the imaging data from 80 to 100 min after injection. BP from the dynamic data from 0 to 100 min correlated strongly (R^sup 2^ > 0.86) with corresponding regional SUV R^sub 80-100^ - 1 values. In the putamen, the observed kinetics (positive SUV R^sub 1-5^ - 1 at the tracer delivery stage and plateauing time-SUVR curves for all diagnostic categories) may suggest either additional off-target binding or a second binding site with different kinetics. Conclusion: The kinetics of the ^sup 18^F-AV-1451 tracer in cortical areas, as examined in this small group of subjects, differed by diagnostic stage. A delayed 80- to 100-min scan provided a reasonable substitute for a dynamic 0- to 100-min acquisition for cortical regions although other windows (e.g., 75-105 min) may be useful to evaluate.