BACKGROUND
•CLARITY-AD and TRAILBLAZER-ALZ2/ALZ6 showed a combined amyloid-related imaging abnormality (ARIA) rate of 21.5% for lecanemab and 29-36% for donanemab.1,2
•Infusion-related reactions (IRRs) were higher (25.8%) for lecanemab compared to donanemab (9%) in those trials.
•Relative safety of the two FDA-approved AATs, lecanemab and donanemab is not well characterized in real-world clinical practice.
OBJECTIVE
To compare real-world rates of ARIA, IRRs, and treatment discontinuation between patients treated with lecanemab and donanemab across three institutions
METHODSDesign. Prospective rolling enrollment cohort across three institutions — two academic (Univ. of Kansas; Univ. of Washington) and one private (Abington Neurological Associates) — over 3 years.
Patients. Early-stage AD patients (biomarker confirmed) initiating lecanemab or donanemab, followed from first infusion.
Outcomes. ARIA (ARIA-E, ARIA-H, combined) for full cohort+≥6 months; grade ≥2 IRRs; and treatment discontinuation.
Analysis. χ2 test and multivariable logistic regression via R (v4.5.2). Additional sensitivity analyses including diagnosis and APOE-e2-carriership as well as restricting cohort to those treated >6mo performed. Models:
1) Outcome ~ Drug + e4/e4 + site
2) IRR ~ Drug + e4/e4 + diagnosis
RESULTS/CONCLUSION
•Both AATs were well-tolerated with ARIA incidence at ≥6 months (17.3%) comparable to pivotal trials.
•Overall ARIA rates did not differ significantly between lecanemab and donanemab (p=0.77).
•Donanemab showed higher symptomatic ARIA (OR 2.93, p=0.044).
•Lecanemab discontinuation rate higher than donanemab; result impacted by duration of lecanemab use and withdrawal for progression to moderate stage.
•APOE ε4/ε4 carriers had the highest ARIA risk (ARIA-E OR 2.78; p=0.002).
•Donanemab caused markedly more grade ≥2 IRRs than lecanemab (24.6% vs 5.2%; adj. OR 6.3, p<0.001).
LIMITATIONS AND FUTURE INVESTIGATION
•Observational, non-randomized; groups differed at baseline (diagnosis, APOE, follow-up).
•Some risk covariates (i.e. baseline MRI) not included.
•IRRs were captured at a single site (n=203) only.
•Rolling enrollment → variable follow-up; crude rates underestimate cumulative risk.
•Further real-world studies required to understand long-term effects of ARIA events with AATs.
