Making Antibody Treatments More Effective in Early Alzheimer's Disease Using 3Tesla Magnetica Resonance

Alzheimer's disease causes progressive memory and cognitive decline, driven in part by the buildup of a protein called β-amyloid in the brain. New antibody therapies - lecanemab and donanemab - can remove amyloid and slow down the disease in its early stages. However, it is still unclear how long each patient should continue treatment or when it is safe to stop, because amyloid is cleared at different rates across individuals. Today, amyloid Positron Emission Tomography (PET) scans are used to measure whether amyloid has been removed from the brain, but these scans are expensive, not always available, and expose patients to radiation. Since repeated PET scans are not ideal, doctors need better ways to monitor treatment progress. This study will use advanced brain Magnetic Resonance Imaging (MRI) and blood tests to create personalized prediction models. These models will simulate how amyloid spreads or clears in each person's brain and help identify when treatment is still needed. With this approach, monitoring becomes safer, more efficient, and more affordable - helping ensure that each patient receives the right treatment for the right amount of time. This prospective monocenter study investigates the role of 3Tesla MRI-based predictive modeling in predicting treatment response to anti-amyloid monoclonal antibodies (lecanemab or donanemab administered as clinical practice) in 50 patients with early Alzheimer's disease (AD) at IRCCS Ospedale San Raffaele (Milan, Italy). Advanced MRI techniques, including high- resolution structural imaging for cortical thickness and volumetric atrophy, diffusion imaging for structural connectivity, and resting-state functional MRI for functional network analysis, will be acquired at baseline, 6, 12, and 18 months. These multimodal MRI measures will be integrated into computational approaches, such as the Aggregation Network Diffusion (AND) model, to simulate individual disease trajectories and predict the probability of achievin