Assessing Amyloid Pathology in Cognitively Normal Subjects Using 18F-Flutemetamol PET: Comparing Visual Reads and Quantitative Methods

Objective: Determine the optimal approach for assessing amyloid pathology in a cognitively normal elderly population. Methods: Dynamic [18F]flutemetamol PET scans acquired using a coffee-break protocol (0-30 and 90-110 min. scan) from 190 cognitively normal elderly (mean age 70.4 years, 60% female) were included. Parametric images were generated from standard uptake value ratio (SUVr) and non-displaceable binding potential (BPND) methods, with cerebellar grey matter as a reference region and were visually assessed by three trained readers. Inter-reader agreement was calculated using Kappa statistics and (semi-)quantitative values were obtained. Global cut-offs were calculated for both SUVr and BPND using a ROC analysis and the Youden Index. Visual assessment was related to (semi-)quantitative classifications. Results: Inter-reader agreement in visual assessment was moderate for SUVr (κ = 0.57) and good for BPND images (κ = 0.77). There was discordance between readers for 35 cases (18%) using SUVr and for 15 cases (8%) using BPND, with 9 overlapping cases. For the total cohort, the mean (±SD) SUVr and BPND values were 1.33 (± 0.21) and 0.16 (± 0.12), respectively. Most of the 35 cases (91%) where SUVr image assessment was discordant between readers, were classified as negative based on (semi-) quantitative measurements. Conclusion: The use of parametric BPND images for visual assessment of [18F]flutemetamol in a population with low amyloid burden improves inter-reader agreement. Implementing semi-quantification in addition to visual assessment of SUVr images can reduce false-positive classification in this population.