Quantitative amyloid PET in the AMYPAD diagnostic and patient management study

Abstract

AbstractBackgroundAMYPAD‐DPMS is a European, multicenter, prospective, interventional, randomized controlled study. It aims to assess clinical utility and cost‐effectiveness of amyloid‐PET. It recruited individuals with Subjective Cognitive Decline plus (SCD+), and syndromic diagnosis of Mild Cognitive Impairment (MCI) and dementia. Here, we describe the quantitative results of the basal amyloid PET performed in these participants.MethodUntil end of recruitment on October 30th, 2020, 844 participants were included in AMYPAD‐DPMS from 8 clinical sites. Of those, 186 are from the Geneva site, 143 from Amsterdam, 130 from Toulouse, 101 from Barcelona, 94 from Cologne, 73 from Stockholm, 64 from London and 53 from Lausanne. Amyloid PET was performed with either 18F‐florbetaben or 18F‐flutemetamol and analysed with AmyPype, a PET‐only Centiloid‐calibrated pipeline running on a GE Advantage Workstation, to render comparable estimates of global amyloid load for both tracers. The standard Global Cortical Average region of interest was used as target region and the whole cerebellum as reference. Distributions as well as Mean and Standard Deviations (SD) of CL values by tracer, site and by diagnostic groups were computed.ResultFrom the 844 recruited participants, 245 (29%) were SCD, 341 (40%) MCI and 258 (31%) dementia. By January 19th, 2021, 661 scans had been analyzed with AmyPype and had passed quality control. Of them, 308 (47%) were scanned using 18F‐florbetaben and 353 (53%) with 18F‐flutemetamol. The global mean (SD) [range] of Centiloid (CL) values was 45.0 (49.0) [ ‐52.4, 221.9] CL. By clinical group, the average (SD) CL values were 25.5(39.1) for SCD+, 45.8(47.9) for MCI, and 66.3(51.1) for dementia patients. CL distributions by clinical group are shown in Figure 1, and per‐site and tracer in Figures 2‐3 for the MCI and Dementia groups.ConclusionQuantitative amyloid PET in AMYPAD‐DPMS shows a very well‐balanced tracer use. As expected, CL values increased with increasing disease severity. The consistent distribution of the majority of CL values within/across clinical groups suggest that the Centiloid transformation implemented in AmyPype allowed the direct comparison of estimates of amyloid burden across the two tracers used in this trial.