Amyloid time: Quantifying the onset of abnormal biomarkers and cognitive impairment along the Alzheimer's disease continuum.

Background: Quantifying the temporal progression of key pathophysiological events in Alzheimer's disease is crucial for staging disease, dementia risk assessment, and informing clinical trials. This work extends a novel algorithm (sampled iterative local linear approximation; SILLA) developed for modeling PET trajectories to: 1) generate a unified CSF and PET amyloid accumulation timeline, and 2) characterize the temporal progression of biomarker and cognitive changes in AD. Method: 941 participants from the Wisconsin Registry of Alzheimer's Prevention, Wisconsin Alzheimer's Disease Research Center, and Statins in Healthy, At‐Risk Adults: Impact on Amyloid and Regional Perfusion studies underwent [C‐11]PiB PET and/or lumbar puncture; and a subset underwent [F‐18]MK‐6240 PET and anatomical MRI using standardized protocols (Table 1). Amyloid burden was quantified from cortical PiB DVR and CSF Aβ42/40. Amyloid Time, quantified as years from CSF Aβ42/40 positivity, was estimated by simultaneously modeling Aβ42/40 and PiB DVR using an extension of the SILLA method in a subset of 248 participants with both measurements. CSF biomarkers measured using the exploratory Roche NeuroToolKit assays, imaging biomarkers, and cognitive impairment (global CDR, CDR SOB, and diagnosis) were plotted as a function of Amyloid Time. Time between key AD events was estimated using LOWESS curves and positivity thresholds for each AD biomarker and CDR scores. Result: Figure 1 depicts biomarker thresholds. Amyloid Time suggested PET Aβ+, CSF T+, PET T+, CDR=0.5, and CDR=1 occurred 4.0, 12.0, 13.8, 17.7, and 22.0 years following Aβ42/40 positivity (Figure 2, Table 2). Heterogeneity was observed in tau levels relative to Amyloid Time for individual cases. Those with worsening CDR generally had high amyloid and at least one T+ biomarker. Conclusion: Amyloid Time based on combined CSF and PET amyloid biomarkers and the SILLA algorithm provides a novel way to quantify the duration of amyloid exposure from the earliest pathophysiologic CSF changes to dementia onset in AD within individuals. This enables characterization of the time between key AD events, which could inform prognosis, identify intervention timing for novel therapies, and enable exploration of mechanisms related to resistance and/or resilience. [ABSTRACT FROM AUTHOR]