Abstract:Background: Converging evidence suggests that markers of Alzheimer's disease (AD) pathology in cognitively unimpaired older individuals are associated with high risk of cognitive decline and progression to functional impairment. The Anti-Amyloid Treatment in Asymptomatic Alzheimer's disease (A4) and Longitudinal Evaluation of Amyloid and Neurodegeneration Risk (LEARN) Studies enrolled a large cohort of cognitively normal older individuals across a range of baseline amyloid PET levels. Recent advances in AD blood-based biomarkers further enable the comparison of baseline markers in the prediction of longitudinal clinical outcomes. Objectives: We sought to evaluate whether biomarker indicators of higher levels of AD pathology at baseline predicted greater cognitive and functional decline, and to compare the relative predictive power of amyloid PET imaging, tau PET imaging, and a plasma P-tau217 assay. Design: All participants underwent baseline amyloid PET scan, plasma P-tau217; longitudinal cognitive testing with the Primary Alzheimer Cognitive Composite (PACC) every 6 months; and annual functional assessments with the clinical dementia rating (CDR), cognitive functional index (CFI), and activities of daily living (ADL) scales. Baseline tau PET scans were obtained in a subset of participants. Participants with elevated amyloid (Aβ+) on screening PET who met inclusion/exclusion criteria were randomized to receive placebo or solanezumab in a double-blind phase of the A4 Study over 240+ weeks. Participants who did not have elevated amyloid (Aβ-) but were otherwise eligible for the A4 Study were referred to the companion observational LEARN Study with the same outcome assessments over 240+ weeks. Setting: The A4 and LEARN Studies were conducted at 67 clinical trial sites in the United States, Canada, Japan and Australia. Participants: Older participants (ages 65-85) who were cognitively unimpaired at baseline (CDR-GS=0, MMSE 25-30 with educational adjustment, and Logical Memory scores within the normal range LMIIa 6-18) were eligible to continue in screening. Aβ+ participants were randomized to either placebo (n=583) or solanezumab (n=564) in the A4 Study. A subset of Aβ+ underwent tau PET imaging in A4 (n=350). Aβ- were enrolled into the LEARN Study (n=553). Measurements: Baseline 18-F Florbetapir amyloid PET, 18-F Flortaucipir tau PET in a subset and plasma P-tau217 with an electrochemiluminescence (ECL) immunoassay were evaluated as predictors of cognitive (PACC), and functional (CDR, CFI and ADL) change. Models were evaluated to explore the impact of baseline tertiles of amyloid PET and tertiles of plasma P-tau217 on cognitive and functional outcomes in the A4 Study compared to LEARN. Multivariable models were used to evaluate the unique and common variance explained in longitudinal outcomes based on baseline predictors, including effects for age, gender, education, race/ethnic group, APOEε4 carrier status, baseline PACC performance and treatment assignment in A4 participants (solanezumab vs placebo). Results: Higher baseline amyloid PET CL and P-tau217 levels were associated with faster rates of PACC decline, and increased likelihood of progression to functional impairment (CDR 0.5 or higher on two consecutive measurements), both across LEARN Aβ- and A4 Aβ+ (solanezumab and placebo arms). In analyses considering all baseline predictor variables, P-tau217 was the strongest predictor of PACC decline. Among participants in the highest tertiles of amyloid PET or P-tau217, >50% progressed to CDR 0.5 or greater. In the tau PET substudy, neocortical tau was the strongest predictor of PACC decline, but plasma P-tau217 contributed additional independent predictive variance in commonality variance models. Conclusions: In a large cohort of cognitively unimpaired individuals enrolled in a Phase 3 clinical trial and companion observational study, these findings confirm that higher baseline levels of amyloid and tau markers are associated with increased rates of cognitive decline and progression to functional impairment. Interestingly, plasma P-tau217 was the best predictor of decline in the overall sample, superior to baseline amyloid PET. Neocortical tau was the strongest predictor of cognitive decline in the subgroup with tau PET, suggesting that tau deposition is most closely linked to clinical decline. These findings indicate that biomarkers of AD pathology are useful to predict decline in an older asymptomatic population and may prove valuable in the selection of individuals for disease-modifying treatments.

Predicting Amyloid Positivity in Cognitively Unimpaired Older Adults: A Machine Learning Approach Using A4 Data

CAPS: a simple clinical tool for β-amyloid positivity prediction in clinical Alzheimer syndrome

PREDICTING AMYLOID BURDEN TO ACCELERATE RECRUITMENT OF SECONDARY PREVENTION CLINICAL TRIALS

Predicting Brain Amyloid Using Multivariate Morphometry Statistics, Sparse Coding, and Correntropy: Validation in 1,101 Individuals From the ADNI and OASIS Databases

A Data-Driven Cognitive Composite Sensitive to Amyloid-β for Preclinical Alzheimer's Disease

Integrated Algorithm Combining Plasma Biomarkers and Cognitive Assessments Accurately Predicts Brain Β-Amyloid Pathology

Prediction of Amyloid β-Positivity with both MRI Parameters and Cognitive Function Using Machine Learning

Development and assessment of algorithms for predicting brain amyloid positivity in a population without dementia

Predicting amyloid‐PET and clinical conversion in apolipoprotein E ε3/ε3 non‐demented individuals with multidimensional factors

Predicting Alzheimer's Disease with Interpretable Machine Learning.

Amyloid and Tau Prediction of Cognitive and Functional Decline in Unimpaired Older Individuals: Longitudinal Data from the A4 and LEARN Studies

Comparison of deep learning architectures for predicting amyloid positivity in Alzheimer's disease, mild cognitive impairment, and healthy aging, from T1-weighted brain structural MRI

Prediction of Amyloid Positivity in Patients with Subcortical Vascular Cognitive Impairment

Using the Artificial Neural Network to Discriminate Between Normal Controls with Different APOE E4 Genotypes and Probable AD Cases in PIB-PET Studies

Predicting progression and cognitive decline in amyloid-positive patients with Alzheimer’s disease

Machine learning prediction of future amyloid beta positivity in amyloid-negative individuals

Detecting amyloid positivity in early Alzheimer's disease using combinations of plasma Aβ42/Aβ40 and p-tau

Predicting Brain Amyloid Status Using the National Institute of Health Toolbox (NIHTB) for Assessment of Neurological and Behavioral Function

The A4 study: β-amyloid and cognition in 4432 cognitively unimpaired adults

Plasma Amyloid‐beta42/40 Ratio As Biomarker of Cerebral Amyloidosis in Cognitively Unimpaired APOE‐e4 Homozygotes, Heterozygotes and Non‐carriers

Application of Machine Learning to Predict the Risk of Alzheimer's Disease: An Accurate and Practical Solution for Early Diagnostics