Unique regional patterns of amyloid burden predict progression to prodromal and clinical stages of Alzheimer’s disease

Background Regional differences in amyloid‐beta burden were examined between healthy controls (HC), mild cognitive impairment (MCI) and Alzheimer’s disease (AD) and its prognostic value was compared to other biomarkers. Method ADNI data (http://adni.loni.usc.edu/) of four age‐matched amyloid‐beta‐positive groups were included: cognitively stables (CN‐CN, n=38), CN to MCI or AD progressors (CN‐MCI/AD, n=38), MCI stables (MCI‐MCI, n=104), and MCI to AD progressors (MCI‐AD, n=104). Baseline 18F‐florbetapir scans were pre‐processed including normalization and intensity‐standardization (reference: cerebellum). To assess regional differences in amyloid‐beta burden, voxel‐wise group‐wise comparisons (FWE‐corrected) were performed between CN‐CN vs. CN‐MCI/AD and MCI‐MCI vs. MCI‐AD, respectively. Mean amyloid‐beta burden of resulting significant regions was subsequently used in logistic regression analyses together with global amyloid, genetic risk, and CSF biomarkers to determine the best predictive biomarker for progression. Result CN to MCI/AD progression was predicted by amyloid deposition in the precuneus, lingual and angular gyrus, pallidum, caudate, middle and superior temporal gyrus. MCI to AD progression was predicted by amyloid deposition in the anterior cingulate gyrus, medial frontal cortex, precuneus, middle and superior temporal gyrus, posterior and anterior insula. Conclusion Regional amyloid‐beta burden is a more sensitive predictor of progression to MCI or AD than global amyloid‐beta burden, pointing towards its usefulness as prognostic biomarker.