Abstract
Background
Neuroimaging and fluid biomarkers are being increasingly used for the diagnosis, prognosis, as well as recruitment and outcome measures in disease-specific therapeutic trials of Alzheimer's disease (AD).
Methods
We examined different neuroimaging markers in the context of predicting disease progression and cognitive trajectories, alone and in combination with fluid and genetic biomarkers (f.e. APOE or BDNF). We assessed the predictive power of cross-sectional and longitudinal approaches of neuroimaging measures of Aβ-pathology, brain volumetrics, and what predictive information can be derived from the combination of these markers. With the advent of tau imaging, we also tested the performance of neuroimaging and fluid biomarkers against the new proposed categorical classification that combines measures of Aβ, tau, and neuronal injury. Finally, we tested the performance of blood surrogate markers as predictors of pathology and disease progression.
Results
When assessed alone, baseline measures of Aβ burden were a much stronger predictor of disease progression (OR 11.3) than rates of Aβ accumulation established over 4.5 years (OR 5.1). Furthermore, the predictive value of Aβ imaging increases with increasing Aβ burden (from PPV 46% to PPV 84% in MCI). While BDNF polymorphism had no effect on Aβ accumulation, APOE4 carriers showed faster rates of Aβ deposition at the early stages of the accumulation process. Individuals with high Aβ burden carrying APOE4 or BDNF polymorphism declined faster than non-carriers. Cognitively unimpaired individuals who presented only with signs of neurodegeneration but no Aβ-pathology, did not show cognitive decline nor cortical grey matter atrophy over 6 years. The combination of markers of Aβ-pathology and markers of neuronal injury seem to be a stronger predictor of cognitive decline than the combination of markers of tau pathology and markers of neuronal injury. Blood markers, validated against similar cohorts (ADNI), showed high accuracy in predicting brain Aβ pathology and cognitive decline.
Conclusions
Markers of Aβ-pathology are highly predictive of cognitive decline and disease progression. Combination with neuroimaging markers of neurodegeneration and/or genetic markers refines the predictive power of markers of AD pathology. Improvement in blood biomarkers are allowing accurately prediction of Aβ-pathology, providing a platform for developing a population-based screens.