Associations of plasma pT205/T205 with Alzheimer’s disease pathology and cognitive function: a comparison with pT217/T217 and pT181/T181
Shorena Janelidze, Nicolas R. Barthélemy, Erik Stomrud, Randall J. Bateman, Oskar Hansson- Psychiatry and Mental health
- Cellular and Molecular Neuroscience
- Geriatrics and Gerontology
- Neurology (clinical)
- Developmental Neuroscience
- Health Policy
- Epidemiology
Abstract
Background
Alzheimer’s disease (AD) related amyloid (Aβ) and tau pathologies both lead to changes in the cerebrospinal fluid (CSF) levels of phosphorylated tau (p‐tau). However, the extent of these changes appears to vary across different p‐tau variants. Recent evidence has indicated that CSF measures of tau phosphorylation occupancy at T205, pT205/T205, correlated strongly with tau deposition detected by PET and cognitive impairment but were less associated with brain levels of Aβ on PET. Here, we investigated associations of plasma pT205/T205 with tau‐PET and Aβ‐PET as well as cognitive performance in comparisons with two other established plasma p‐tau biomarkers, pT217/T217, and pT181/T181.
Method
This cross‐sectional study included cognitively unimpaired individuals (N = 231) and patients with mild cognitive impairment (N = 106), AD dementia (N = 66) and non‐AD neurodegenerative diseases (N = 18) from the Swedish BioFINDER‐2 study. pT205/T205, pT217/T217, and pT181/T181 were measured in plasma (all participants) and CSF (N = 118) using immunoprecipitation‐coupled mass spectrometry. Tau‐PET, Aβ‐PET and cognitive testing were available in 410, 320 and 406 participants, respectively. Aβ status (A) was defined using either CSF Aβ42/Aβ40 or Aβ‐PET.
Result
We found strong correlations between CSF and plasma levels for all 3 biomarkers (pT205/T205, Pearson r = 0.76; pT217/T217, r = 0.85; pT181/T181, r = 0.67; p<0.001, N = 118) (Figure 1). In A+ individuals (N = 195), associations with tau‐PET (Figure 2) were similar for plasma pT205/T205 (partial Spearman ρ = 0.57, p<0.001) and pT181/T181 (ρ = 0.51, p<0.001) and higher for pT217/T217 (ρ = 0.78, p<0.001). Associations with Aβ‐PET (N = 113) were moderate for plasma pT205/T205 (ρ = 0.38, p<0.001) and pT181/T181 (ρ = 0.36, p<0.001) and stronger for plasma pT217/T217 (ρ = 0.69, p<0.001) (Figure 2). We found similar associations of cognitive performance (based on MMSE, N = 185) with plasma pT205/T205 and pT217/T217 (standardized coefficient β, ‐0.42 and ‐0.46, respectively; p<0.001) whereas associations with pT181/T181 were weaker (β = ‐0.30, p<0.001) (Figure 3). In multivariable models, pT205/T205 (β = ‐0.22, p = 0.015) and pT217/T217 (β = ‐0.39, p<0.001) but not pT181/T181 were independently associated with MMSE. Associations of plasma biomarkers with Aβ‐PET and MMSE were similar in the whole samples but somewhat lower for tau‐PET (Figures 2‐3). A larger dataset will be presented at AAIC.
Conclusion
Plasma pT205/T205 is a promising novel biomarker that might prove particularly useful when combined with pT217/T217 for monitoring cognitive decline in AD.