Alzheimer's Imaging Consortium.

Area of Science:

• Neuroimaging
• Alzheimer's Disease Research
• Biomarker Development

Background:

• In vivo Braak staging stratifies Alzheimer's disease (AD) spectrum patients.
• Harmonizing tau PET tracer staging is crucial for consistent AD assessment.
• This study compares harmonization procedures for Braak staging using MK6240 and Flortaucipir tau PET tracers.

Purpose of the Study:

• To compare and test harmonization procedures for Braak staging using MK6240 and Flortaucipir tau PET tracers.
• To evaluate different cutoff methods and staging categories for optimizing concordance.
• To assess inter- and intra-tracer agreement in Alzheimer's disease staging.

Main Methods:

• Assessed 437 participants (245 cognitively unimpaired, 192 impaired) using head-to-head MK6240 and Flortaucipir scans.
• Computed Standardized Uptake Value Ratios (SUVRs) in Braak regions of interest (ROIs).
• Evaluated four cutoff methods (SD, GMM, Youden index) and two staging categories (7-class vs. 4-class) for Braak staging.

Main Results:

• Intra-tracer Braak staging concordance ranged from 63% to 94%, with Gaussian mixture modeling (GMM) cutoffs yielding the highest (94% for MK6240, 89% for Flortaucipipir).
• Inter-tracer agreement ranged from 56% to 76%, with the CU Elderly Aβ- cutoff showing highest concordance.
• A simplified 4-class Braak staging improved intra-tracer concordance and inter-tracer agreement (86.5%), with most discrepancies at stages II-IV.

Conclusions:

• Preliminary findings show discrepancies in Braak staging between MK6240 and Flortaucipir tracers.
• Adjustments in cutoffs and ROIs can partially mitigate inter- and intra-tracer divergences.
• Robust concordance is achievable with adjusted cutoffs and a 4-class staging system (0, I-II, III-IV, V-VI).