Biomarkers

Area of Science:

• Neuroimaging
• Alzheimer's Disease Research
• Connectomics

Background:

• Preclinical Alzheimer's disease (AD) involves amyloid buildup without cognitive symptoms.
• Functional connectivity alterations are observed in preclinical AD, but their predictive value for pathology is unclear.

Purpose of the Study:

• To utilize connectome-based predictive modeling (CPM) to predict tau deposition using functional connectivity.
• To identify clinically meaningful subgroups within a preclinical AD cohort based on predictive models.

Main Methods:

• CPM was employed to predict baseline tau PET from resting-state functional connectomes in the A4 study cohort.
• Models were validated in the Alzheimer's Disease Neuroimaging Initiative (ADNI) cohort.
• Hierarchical clustering was performed on predictive network edges to identify subgroups, and their longitudinal cognitive trajectories were compared.

Main Results:

• Whole-brain functional connectivity robustly predicted regional tau PET, outperforming amyloid PET.
• Predictive models generalized to the ADNI cohort, particularly in participants with elevated tau.
• Clustering revealed two distinct subgroups, one exhibiting significantly faster cognitive decline.

Conclusions:

• The functional connectome modestly predicts regional tau PET across the AD spectrum.
• Predictive functional connectivity patterns correlate with cognitive trajectories, highlighting their role in early AD staging.
• Functional connectivity shows promise for enhancing early Alzheimer's disease staging when combined with other biomarkers.