HAI Book 2025 - Flipbook - Page 516
Yakoub, Yara
Head-to-head comparison of longitudinal plasma assays in relation to
longitudinal amyloid pathology in Alzheimer9s disease
Yara Yakoub1,2, Ting Qiu1,2, Sylvia Villeneuve1,3,4, Alexa Pichet Binette5,6,7, Alzheimer9s Disease
Neuroimaging Initiative (ADNI) ADNI Foundation for the National Institutes of Health (FNIH)
Biomarkers Consortium8
1
Douglas Mental Health University Institute, Montréal, QC, CA
Integrated Program in Neuroscience, Faculty of medicine, McGill University, Montréal, QC, CA
3
Montreal Neurological Institute, McGill University, Montréal, QC, CA
4
Department of Psychiatry, Faculty of medicine, McGill University, Montréal, QC, CA
5
Clinical Memory Research Unit, Department of Clinical Sciences Malm., Lund University, Lund, SE
6
Department of Physiology and Pharmacology, Université de Montréal, Montréal, QC, CA
7
Centre de Recherche de l’Institut Universitaire de Gériatrie de Montréal, Montréal, QC, CA
8
Memory and Aging Center, Department of Neurology, Weill Institute for Neurosciences, University of California, San
Francisco, CA, US
2
Background: Blood-based biomarkers of AD, especially p-tau217, show high concordance with A´-PET load and
status. While many studies relied on cross-sectional data, assessing dynamic changes in these assays is
important for future clinical use and trial outcomes. We evaluated the longitudinal trajectories of multiple plasma
biomarkers and their associations with longitudinal changes on A´-PET.
Methods: We used data from the ADNI FNIH consortium (Table 1). It consists of 404 participants with on average
three plasma samples collected over 11 years, in which 14 assays were tested (4 p-tau217, 4 A´42/40, 2 p-tau181, 2
GFAP, 2 NfL), and with A´-PET available over the same period. We first assessed the changes over time for each
assay at the group level. Then, we calculated the individual rate of change of each plasma assay and A´-PET
Centiloids (slope) using linear mixed-effect models. Association between baseline plasma levels and plasma rate
of change with Centiloids accumulation were tested using linear regression models.
Results: Across the whole sample, all biomarkers, except for A´42/40 assays, showed increase over time (Fig.1A),
which was similar in the A´-positive and A´-negative groups assessed separately (Fig.1B). Almost all assays
showed A´-PET dependent changes (time × A´ status), with the greatest interactions seen with p-tau217 (Fig.1C).
In the whole group, both baseline plasma levels and plasma slopes were associated with Centiloids accumulation
across almost all assays (Fig.2A-B for statistics and C for scatter plots). In A´-positive participants, such
associations were no longer significant, except for %p-tau217. In A´-negative participants, most of the
associations with p-tau assays remained (Fig.2).
Conclusion: While all plasma assays levels, except for A´42/40, increased over time, at higher levels they were no
longer associated with insoluble amyloid accumulation. The associations between plasma markers and Centiloid
accumulation were also stronger in individuals who had not reached the positivity threshold.
HAI2025 - 516
