Study Pits Alzheimer's Drugs Against Each Other in FDA Side-Effect Reports

A new analysis of FDA adverse-event data finds lecanemab and donanemab produce different patterns of brain-swelling and bleeding signals, the first study to compare the two anti-amyloid drugs head to head rather than separately.

Researchers in China have published the first study to directly compare adverse-event reports for lecanemab and donanemab, two antibody drugs given for early Alzheimer's disease, using the FDA Adverse Event Reporting System (FAERS). The study, posted June 22, 2026 in Frontiers in Pharmacology, examined 3,640 reports naming either drug as the primary suspect (2,602 for lecanemab, 1,038 for donanemab) and ran four separate statistical methods to flag adverse events reported more often than expected for each drug. Disorders of the nervous system were the most frequently reported category for both drugs and the only category that met all four statistical thresholds for a signal in both.

At the level of individual adverse events, the reports clustered around amyloid-related imaging abnormalities, a known class of brain-scan findings involving swelling or fluid (called ARIA-E) and microbleeds (called ARIA-H), along with headache and infusion-related reactions. When the researchers compared signal strength between the two drugs, lecanemab showed a stronger disproportionality signal for ARIA-H, the microbleed pattern, while donanemab showed a stronger signal for ARIA-E, the swelling and fluid pattern. Both patterns are already listed as risks on the drugs' FDA labels; what is new here is a direct statistical comparison of how often each shows up in spontaneous reports tied to one drug versus the other.

Prior FAERS studies had looked at lecanemab and donanemab on their own, or compared each separately against older drugs like aducanumab, but researchers said no earlier work had set the two current anti-amyloid antibodies directly against each other in the same disproportionality analysis. The authors frame the finding as a hypothesis-generating signal for further pharmacovigilance monitoring, not as evidence that one drug carries greater real-world risk than the other for any individual patient.

The authors call for continued post-marketing surveillance and further studies to confirm whether the sex-stratified and event-type differences they observed hold up as more reports accumulate.