Objectives

This study investigates the correlation between early-phase β-amyloid-PET perfusion deficits and neuropsychological test results in patients with Alzheimer's disease (AD).
It explores the potential of early-phase β-amyloid-PET as a biomarker for in vivo staging of disease severity.
It examines whether early-phase β-amyloid-PET can predict future cognitive decline.

Methodology

Regional perfusion deficits were evaluated in early-phase β-amyloid-PET scans of 82 patients with biologically defined AD.
Neuropsychological testing (MMSE and CERAD-Plus) was performed within 90 days of the PET scan.
Patients were classified into Braak-like stages (stage0, stageI−II+, stageI−IV+, stageI−VI+, and stageatypical+) based on regional perfusion deficits in predefined regions of interest (ROIs).
Multiple regression analysis, controlling for age, gender, and education, was used to assess the association between regional z-scores on perfusion-phase PET and clinical scores.
The analysis also included the prediction of annual cognitive decline in 23 patients with follow-up data.

Lower cognitive performance (MMSE and CERAD-Plus scores) was significantly associated with decreased global and regional perfusion in early-phase β-amyloid-PET (r2 values ranging from 0.09 to 0.41, p-values <0.05).
The strongest association was found in the left temporal lobe (r2=0.37, p<0.0001 for MMSE).
Temporal and parietal region z-scores predicted future annual decline in MMSE (r2=0.29, p=0.037 for temporal lobes) and CERAD-Plus sum scores (r2=0.39-0.41, p=0.043-0.048 for inferior parietal lobe).
Patients classified as stage0 and stageI−II+ had significantly better cognitive performance than those classified as stageI−IV+ and stageI−VI+ (p=0.014 for MMSE, p=0.044 for CERAD-Plus).

The study demonstrates a good correlation between early-phase β-amyloid-PET perfusion deficits and cognitive performance, supporting its potential as a biomarker for AD severity. However, the staging method based on Braak stages, while conceptually interesting, has limitations.
The classification of 19.5% of patients as 'stageatypical+' is a significant concern and needs further investigation. The authors acknowledge this, but a more detailed analysis of these cases (e.g., comparing them to typical AD progression patterns) would strengthen the study.
The study relies on a built-in normal cohort of FDG-PET for z-score calculation, which may not be directly comparable to early-phase β-amyloid-PET. Establishing a dedicated normal cohort for early-phase β-amyloid-PET would improve the accuracy of the z-score calculations.
The follow-up data is limited to 23 patients. While the results are promising, a larger cohort with longer follow-up is needed to confirm the predictive value of early-phase β-amyloid-PET for cognitive decline.
The study lacks a direct comparison with FDG-PET, which is a more established method for assessing neuronal injury. A head-to-head comparison would be valuable to determine the relative performance of early-phase β-amyloid-PET.