(50) Real-World Discontinuation Risk and Adverse Events with Anti-Amyloid Therapies in Patients with Alzheimer’s Disease

Background: : Amyloid-targeting therapies are the first disease-modifying agents and were recently approved for Alzheimer’s disease (AD) with mild cognitive impairment or mild dementia. Phase 3 trials (Clarity-AD and TRAILBLAZER-ALZ 2) showed slowed cognitive and functional decline, but higher rates of adverse events. Discontinuation due to adverse events occurred in 6.9% of lecanemab-treated patients and in 13.1% of those treated with donanemab. Common adverse events include amyloid-related imaging abnormalities (ARIA), such as ARIA-E (edema) and ARIA-H (hemorrhage), as well as headache, falls, and urinary tract infections. Patients carrying the ApoE ε4 homozygous genotype experienced ARIA-E more frequently than patients with heterozygotes or non-carriers. While clinical trial data highlight these adverse events, real-world evidence on discontinuation patterns, reasons for discontinuation, and differences by ApoE ε4 status remains limited.

Objective: : To describe the discontinuation risk and reasons among patients initiating anti-amyloid beta monoclonal antibody therapy for AD and to compare discontinuation and AE risk by APOE ε4 status.

Methods: : This retrospective cohort study will utilize electronic health records (EHR) from a large integrated delivery system in Texas between January 6, 2022, and July 31, 2025. Adults with AD initiating lecanemab or donanemab between January 6, 2023, and July 31, 2024 (index date = first infusion; follow-up through July 31, 2025, ensuring ≥12 months potential follow-up). A 12-month pre-index period will be used to report baseline characteristics. The primary outcome will be time to treatment discontinuation from the index date. Secondary outcomes will evaluate reasons for discontinuation, the risk of adverse events, and differences in discontinuation and adverse event risk by APOE ε4 status. Time to discontinuation will be analyzed using a Cox proportional hazards model.

Results: : Results In Progress

Conclusions: : Conclusions In Progress

Sponsorship:: Baylor Scott and White Health