INTRODUCTION: Down syndrome (DS) is associated with elevated risk for Alzheimers disease (AD) and lifelong accumulation of beta amyloid (Aβ). The current work addresses two major gaps in our understanding of Alzheimers disease (AD) in Downs Syndrome (DS): (1) regional amyloid aggregation and reliability in predicting onset or progression of disease in sporadic AD, and (2) the contributions of regional amyloid burden to AD progression in DS. We hypothesized that the spatial distribution of Aβ plaque burden predicts transition to dementia in individuals with DS. METHODS: We acquired 18F-Florbetapir Positron Emission Tomography (PET) scans on nineteen non-demented adult individuals with DS at baseline and monitored them clinically over a four year period, identifying five individuals who transitioned to dementia. We used machine learning classification to determine features on 18F-Florbetapir standardized uptake value ratio (SUVR) maps that were most predictive of transition. RESULTS: Of the nineteen participants, fourteen remained non-demented (age 50.4±4.3 yrs.;10M/4F) and five transitioned to dementia (age 52.1±5.7;2M/3F) over the study period. Compared with non-transitioned participants, those who transitioned showed increased amyloid binding in the middle and superior orbital-frontal and superior temporal lobes (t(14)=3.8;p<0.001 unc.) with corresponding Cohens d effect sizes ranged from 1.4-3.2. Increased amyloid burden in subjects who transitioned temporally closer to the baseline PET scan were found in the middle and superior temporal lobes (t(3)=10.2; p<0.001 unc.). The coefficient of determination (r squared) measures ranged from 0.87-0.98 corresponding to a strong relationship between regional amyloid burden and the transition time to clinical dementia. Using regional amyloid burden to predict future transition using 10-fold cross-validated logistic regression classifiers resulted in inferior parietal lobe (area under receiver operating characteristic curve (AUC) 0.96±0.04), inferior and superior temporal lobes (AUC=0.92±0.11; 0.93±0.10), and lateral orbital-frontal lobes (AUC=0.92±0.11) as the most predictive signals. DISCUSSION: The spatial distribution of Aβ burden in individuals with DS, and in particular 18F-Florbetapir uptake in the orbitofrontal and inferior parietal cortices, may serve as biomarkers, predicting clinical transition to dementia in DS.

[18F]-Florbetapir PET: Towards Predicting Dementia in Adults with Down Syndrome