Alzheimer's Imaging Consortium.

Area of Science:

• Neuroimaging
• Neuropathology
• Cognitive Impairment

Background:

• The clinical significance of amyloid-beta positron emission tomography (Aβ-PET) signal distribution remains unclear.
• Data-driven methods were employed to analyze Aβ-PET patterns in over 12,000 cognitively impaired individuals across four cohorts.

Purpose of the Study:

• To identify distinct Aβ-PET binding patterns using data-driven approaches.
• To investigate the clinicopathological associations of these identified Aβ-PET patterns in patients with cognitive impairment.

Main Methods:

• Analysis of multi-tracer template-space Aβ-PET SUVR images.
• Independent component analysis (ICA) and k-means clustering were used to group participants based on Aβ-PET binding topography.
• Validation of the model across three independent cohorts, assessing associations with clinical data, APOE-ε4 status, tau-PET, and neuropathology.

Main Results:

• Three distinct Aβ-PET binding clusters were identified: Aβ-negative, Aβ+ with posterior predominance (Aβ+ posterior), and Aβ+ with typical distribution (Aβ+ typical).
• The Aβ+ posterior group exhibited more severe clinical impairment, lower APOE-ε4 carrier rates, and higher posterior tau-PET levels compared to the Aβ+ typical group.
• Autopsy findings revealed more severe cerebral amyloid angiopathy (CAA) in the Aβ+ posterior cluster.

Conclusions:

• A reliable method for assigning patients into an Aβ+ PET binding group, particularly with posterior predominance, was established.
• Occipital Aβ-PET binding is a significant finding that should be considered in clinical assessments.
• Posterior Aβ+ PET binding may serve as a biomarker for CAA and more severe cognitive impairment.