This longitudinal cohort study examined changes in 130 plasma protein biomarkers over up to 2 years in 197 patients with early symptomatic Alzheimer's disease receiving biweekly lecanemab infusions (10 mg/kg IV), compared to an untreated cognitively impaired control group (n=1,312) and an amyloid-negative cognitively unimpaired control group (n=458).
34 of 130 biomarkers showed significant changes (FDR <0.05), falling into four trajectories: normalising (e.g., p-tau217, MAPT, GFAP moving toward amyloid-negative levels), overcorrecting (e.g., amyloid-β peptides moving beyond amyloid-negative levels), opposite (e.g., NPTXR, NRGN, TREM2 moving further from amyloid-negative levels), and unchanged (e.g., NEFL, APOE).
The cohort was predominantly White (98%) and non-Hispanic (97%), limiting generalisability. The study is observational without randomisation, so causal inference is constrained. No placebo arm was included; the untreated control group was drawn from a separate research cohort.
Clinicians monitoring lecanemab-treated patients should not expect all plasma biomarkers to normalise — while p-tau217 and GFAP trend toward normal, inflammatory and synaptic markers like TREM2 and NPTXR worsen, suggesting different biomarkers may eventually serve distinct roles in tracking treatment response versus patient prognosis.
