Basic Science and Pathogenesis.

Area of Science:

• Neurology
• Radiology
• Gerontology

Background:

• Positron emission tomography (PET) is the gold standard for in vivo amyloid burden assessment, often quantified using standardized uptake value ratios (SUVR).
• Calculating SUVRs involves choices in regions of interest (ROI) and reference regions, which is critical in the oldest-old due to atrophy and MRI contraindications.
• This study validated SUVR methods against postmortem pathology in participants aged 90+.

Purpose of the Study:

• To compare the performance of four different SUVR calculation methods for amyloid quantification.
• To identify the optimal SUVR method for predicting postmortem amyloid pathology in the oldest-old population.
• To establish reliable in vivo amyloid assessment in a vulnerable elderly cohort.

Main Methods:

• Utilized florbetapir PET and postmortem neuropathological data from The 90+ Study.
• Employed an MRI-free pipeline for PET scan re-alignment.
• Calculated four SUVRs using combined posterior cingulate/precuneus (PC²) or cortical gray matter (CorticalGM) as ROIs, and white matter (WM) or cerebellar gray matter (CbllmGM) as reference regions.
• Assessed predictive performance using ROC analysis and Youden's index for optimal cutoffs.

Main Results:

• SUVRs using white matter (WM) as the reference region outperformed cerebellar gray matter (CbllmGM).
• The PC²+WM method demonstrated the highest area under the curve (AUC) for predicting diffuse plaques (0.84) and neuritic plaques (0.82).
• An optimal cutoff of 0.77 for PC²+WM yielded a sensitivity of 0.72 and specificity of 0.87 for diffuse plaque prediction.

Conclusions:

• The PC²+WM SUVR method is the most effective for predicting amyloid diffuse and neuritic plaques in the oldest-old individuals.
• This validated method can improve in vivo amyloid assessment in advanced aging.
• Findings support the use of PC²+WM for accurate amyloid quantification in clinical and research settings.