Expected Centiloid Values for Patients with Positive Visual Reads of Amyloid PET Scans
Eric Siemers1, Maddelyn Hyland1, David Scott2, Todd Feaster1, Gopalan Sethuraman1
1Acumen Pharmaceuticals, Inc, 2Clario
Objective:
To determine the expected range of Centiloid (CL) values from amyloid PET scans assessed as positive by visual reads based on published literature and baseline clinical trial results.
Background:

While amyloid PET scans can be assessed by visual reads, some clinical trials now use a CL cut-off. The CL scale was developed to standardize comparison across PET tracers, with the intent to approximate a value of 0 for amyloid negativity and an approximate value of 100 for mild to moderate dementia.

Design/Methods:
The literature was reviewed to assess the range of CL values that might be expected for scans that were amyloid positive by visual read. For each PET ligand, a CL value can be calculated from the standard uptake value ratio (SUVr) using published equations.
Results:
Based on published results, for patients with autopsy findings showing an early amyloid stage, CL threshold values ranged from 7.4 to 12.2. Across 13 studies using different PET ligands, threshold values for amyloid positivity ranged from 12.4 CL to 40 CL. In the IDEAS study, visually positive PET scans (n=6332) had CL values ranging from approximately -40 CL to approximately 275 CL. In the phase 3 CLARITY study of lecanemab (n=1734), CL values at baseline ranged from -17 to 213.2 for visually positive PET scans. ALTITUDE-AD is an ongoing phase 2 study of sabirnetug that has completed enrollment of 542 participants with MCI or mild dementia due to AD. In this study, preliminary baseline CL values ranged from -20.37 to 220.32.
Conclusions:
The actual range of CL values for individuals with MCI or mild dementia with visually positive amyloid scans is likely wider than 0 to 100. The preliminary baseline CL values from the ALTITUDE-AD study are consistent with the published literature. Overreliance on CL cut-offs may bias against the detection of clinically relevant early amyloid deposition. 
10.1212/WNL.0000000000213131
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.