Estimating the time course of biomarker changes in Alzheimer’s disease

Recent advancements in biomarkers have transformed Alzheimer's disease (AD) diagnosis from being purely symptom-based to include biological criteria. With new treatments targeting AD's core biology, understanding the timeline of biological changes is crucial as the disease progresses over decades. Longitudinal data from amyloid-beta (Aβ) PET and cognitive tests (MMSE and ADAS-cog) from the Alzheimer's Disease Neuroimaging Initiative (n=1,448) and BioFINDER (n=2,088) were used to stage patients against an estimated continuous disease timeline (predicted time since Aβ-PET positivity). The estimated timeline was validated by comparing correlations with unseen biomarkers and cognitive measures against alternative staging approaches. Trajectories for plasma, CSF, MRI, and PET biomarkers, measuring Aβ, tau, and neurodegeneration, were mapped along this AD continuum. The proposed staging approach was found to produce stronger correlations with unseen cognitive measures and biomarkers compared to alternative staging methods, including amyloid and tau PET clocks (all pairwise p<0.05). Findings related to biomarker trajectories were highly consistent across cohorts. The period from Aβ-PET positivity to end-stage AD dementia (MMSE = 0) was estimated at 20-25 years, with a presymptomatic phase of 7-11 years. CSF Aβ42/40 became abnormal about a year before Aβ-PET positivity, CSF p-tau231, p-tau217, and plasma p/np-tau217 1-3 years after, and tau-PET about 8 years after. Neurodegenerative biomarkers, such as hippocampal volume, became clearly abnormal in early dementia stages, 14-16 years after Aβ-PET positivity. The progression from initial biomarker abnormality to severe AD spans two decades. Disease progression modeling elucidates the evolution of AD biomarkers and cognition, highlighting the relative timing of biomarker abnormalities. These models can determine disease stages, aiding prognosis and evaluation for disease-modifying treatments.