A Novel Approach to Personalized Prediction of Progression of Age-Related Macular Degeneration

More than 90% of patients with advanced AMD have severe vision loss. Predicting AMD progression from an early or intermediate stage is crucial, since prompt intervention after a choroidal neovascularization (CNV) event and geographic atrophy (GA) monitoring can greatly improve visual outcomes. Patients at higher risk of progression should have more frequent follow-up visits, since progression often occurs before any visual changes are noticed by the patient. Previous work has determined the risk factors for AMD progression based on drusen features in fundus photos, Optical Coherence Tomography (OCT) and from genetic factors. However, current models are limited by their ability to make predictions over short intervals, which limits their utility in guiding screening intervals.

In this study we will recruit patients with early and intermediate AMD in at least one eye who are at risk of converting to wet AMD or GA expansion. We will perform a randomized trial where we will randomly assign them to a control or test group (personalized follow-up care starting at 3 months based on their predicted risk from algorithm results), and collect baseline genetic, demographic, imaging, and clinical data and first follow-up data at the 3 month and 6 month follow-up time points. Outcomes will be measured to determine if an algorithm predicting early follow-up for high-risk patients (3 month) is advantageous over the standard 6 month follow-up time point.