Abstract
Background
Anti-amyloid antibodies slow cognitive and functional decline in individuals with early symptomatic Alzheimer’s disease, but few studies have examined changes associated with these treatments in plasma biomarkers linked to neurodegeneration. We analysed changes in a panel of 130 plasma proteins and their associations with conversion to amyloid PET negative status and cognitive changes in patients receiving lecanemab infusions as part of their clinical care.
Methods
This longitudinal cohort study included patients with early symptomatic Alzheimer’s disease and positive amyloid biomarkers who were identified as candidates for biweekly lecanemab infusions based on the US Food and Drug Administration label and consented to blood collection at the Washington University Memory Diagnostic Center (St Louis, MO, USA) between July 17, 2023, and Oct 10, 2025. Lecanemab was administered as a 10 mg/kg intravenous infusion every 2 weeks. Patients underwent clinical assessments and blood draws approximately every 6 months over up to 2 years. Research participants enrolled at the Knight Alzheimer Disease Research Center (Knight ADRC; St Louis, MO, USA) between May 3, 1996, and Feb 27, 2025, and who did not receive anti-amyloid antibodies, were included as control groups (an amyloid β-negative [Aβ–] group without cognitive impairment and an untreated group with cognitive impairment). The primary outcome was change in plasma concentrations of protein biomarkers associated with Alzheimer’s disease pathology with lecanemab treatment. Associations between plasma biomarker changes and number of lecanemab infusions were examined using regression and mixed-effects models, with adjustment for demographics.
Findings
The study included 197 patients in whom lecanemab infusions had been started (lecanemab group), as well as 173 patients in whom lecanemab had been ordered but not yet started, who were combined with 1139 research participants with cognitive impairment from the Knight ADRC cohort to form the untreated control group with cognitive impairment (n=1312). An additional 458 research participants formed the Aβ− control group without cognitive impairment. In the lecanemab group, the median patient age was 73 years (IQR 69–77); 104 (53%) were female and 93 (47%) were male; and most participants were White (187 [98%] of 190 with available data) and non-Hispanic (183 [97%] of 188 with available data). 130 biomarkers were classified into four distinct trajectories, with 34 showing significant changes (false discovery rate <0·05) in the lecanemab group. The trajectories were: normalising, with biomarkers shifting towards concentrations found in Aβ−individuals without cognitive impairment (eg, phosphorylated tau [p-tau]217, MAPT, and GFAP); overcorrecting, with biomarkers shifting towards and then beyond concentrations found in Aβ− individuals without cognitive impairment (eg, amyloid β peptides); opposite, with biomarkers shifting further from concentrations found in Aβ− individuals without cognitive impairment (eg, NPTXR, NRGN, and TREM2); and unchanged (eg, NEFL and APOE).
Interpretation
Lecanemab treatment was associated with dynamic, multidirectional changes across plasma biomarkers of Alzheimer’s disease pathology, with partial normalisation of amyloid and tau biomarkers alongside abnormal concentrations of inflammatory and neurodegeneration-related biomarkers (eg, TREM2 and NPTXR), which might reflect an immune response to amyloid plaques and effects of treatment on neuronal and synaptic processes distinct from amyloid clearance. These divergent trajectories suggest that different biomarkers could eventually serve different clinical purposes, such as monitoring treatment response or informing patient prognosis, although further validation is needed.
Authors and Affiliations
Rui Mu, MSᵃ˒ᵇ˒ᶜ, Anna Hofmann, MDᵈ˒ᵉ, Katherine Gong, MSᵃ˒ᵇ, Menghan Liu, MSᵃ˒ᵇ, Alex Roussak, MDᵃ˒ᵇ, Joseph Lowery, BSᵃ˒ᵇ, Yike Chenᵇ, Madeline Paczynski, PA-Cᵈ˒ᵉ, Zachary Posey, MAᵈ˒ᵉ, Yuchen Cheng, MSᵃ˒ᵇ˒ᶜ, Ying Xu, MSᵃ˒ᵇ, Yunchang Xie, MSᵃ˒ᵇ˒ᶠ, William J B Powell, MSᶜ˒ᵍ, Inez Oh, PhDᵍ, Aditi Gupta, PhDᵍ, Nicholas Sykora, MSᵃ˒ᵇ, John Budde, MSᵃ˒ᵇ, Steven Funk, PhDᵃ˒ᵇ, Tammie L S Benzinger, MDᶠ, David M Holtzman, MDᵈ˒ᵉ, John C Morris, MDᵈ˒ᵉ, B Joy Snider, MDᵈ˒ᵉ, Randall J Bateman, MDᵈ˒ᵉ, Suzanne E Schindler, MDᵈ˒ᵉ*, and Carlos Cruchaga, PhDᵃ˒ᵇ˒ᶜ*
ᵃ Department of Psychiatry, Washington University School of Medicine, St Louis, MO, United States of America
ᵇ NeuroGenomics and Informatics Center, Washington University School of Medicine, St Louis, MO, United States of America
ᶜ Division of Biology and Biomedical Sciences, Washington University School of Medicine, St Louis, MO, United States of America
ᵈ Department of Neurology, Washington University School of Medicine, St Louis, MO, United States of America
ᵉ Knight Alzheimer Disease Research Center, Washington University School of Medicine, St Louis, MO, United States of America
ᶠ Mallinckrodt Institute of Radiology, Washington University School of Medicine, St Louis, MO, United States of America
ᵍ Institute for Informatics, Data Science and Biostatistics, Washington University School of Medicine, St Louis, MO, United States of America
* Suzanne E Schindler and Carlos Cruchaga contributed equally.