Cerebrospinal fluid tau biomarkers in the prediction and concordance of neurofibrillary tangle and amyloid pathology: Biomarkers (non‐neuroimaging): Longitudinal and prognostic biomarker studies. (7th December 2020)

• Record Type: Journal Article
• Authors: Ashton, Nicholas J; Benedet, Andréa Lessa; Pascoal, Tharick A; Kvartsberg, Hlin; Karikari, Thomas K; Schöll, Michael; Stoops, Erik; Francois, Cindy; Vanmechelen, Eugeen; Mathotaarachchi, Sulantha; Savard, Melissa; Therriault, Joseph; Chamoun, Mira; Zetterberg, Henrik; Rosa‐Neto, Pedro; Blennow, Kaj
• Is Part Of: Alzheimer's & dementia. Volume 16(2020)Supplement 5
• Page Start: n/a
• Page End: n/a
• Publication Date: 2020-12-07
• Subjects: Alzheimer's disease -- Periodicals; Alzheimer Disease -- Periodicals; Dementia -- Periodicals; Démence; Maladie d'Alzheimer; Périodique électronique (Descripteur de forme); Ressource Internet (Descripteur de forme); 616.83
• DOI: 10.1002/alz.041849
• Languages: English
• ISSNs: 1552-5260

Abstract

Abstract: Background: Cerebrospinal fluid (CSF) measures of tau (p‐tau181 and t‐tau) are central to the research diagnostic criteria for Alzheimer's disease (AD). Nevertheless, alternative tau phosphorylation sites or tau fragments merit further evaluation to better index neurofibrillary tangle (NFT) and amyloid‐β (Aβ) pathology. Method: We evaluated 168 participants from the TRIAD cohort (young controls, cognitively normal controls (CN), mild cognitive impairment, frontotemporal dementia and AD patients). LUMIPULSE® G1200 (Fujirebio) was used to measure CSF p‐tau181 and t‐tau. CSF p‐tau231 was quantified using a prototype ELISA assay (ADx NeuroSciences). CSF p‐tau217 and tau368 was measured using an in‐house Simoa assay (University of Gothenburg). For PET ([ 18 F]MK6240 and [ 18 F]AZD4694), standard uptake value ratios (SUVR) were determined using the cerebellar and cerebellar cortex as reference tissue, respectively. Finally, ROC curves were generated to investigate the predictive power of the CSF biomarkers. Result: All CSF p‐tau biomarkers equally associated with [ 18 F]MK6240 (r>0.51), while CSF p‐tau231 correlated more to [ 18 F]AZD4694 (r=0.815) than other tau biomarkers. ROC curves revealed p‐tau biomarkers accurately detected Braak I‐II (AUC>0.905) and III‐IV positivity (AUC>0.940), with p‐tau217 significantly out performing p‐tau231 and p‐tau181 at both stages. Tau368/t‐tau ratio was the superior biomarker in the prediction of Braak V‐VI positivity. P‐tau217 and p‐tau231 demonstrated high accuracy in predicting Aβ positivity (ROC>0.950). This was significantly better than p‐tau181 (p‐tau217, P <0.01; p‐tau231, P <0.001). Prediction of Aβ positivity was also high at the pre‐symptomatic stage for p‐tau231 (ROC>0.9) and again was significantly superior to p‐tau181 ( P <0.01). All p‐tau biomarkers demonstrated full concordance (CSF ‐ /PET ‐ ) with [ 18 F]MK6240 and [ 18 F]AZD4694 for young controls and FTD subjects. P‐tau217 and p‐tau231 demonstrated high concordance with [ 18 F]MK6240 and discordant cases were restricted to the CSF + /PET ‐ quadrant (Figure 1). In the same manner, CSF p‐tau217 and p‐tau231 demonstrated the highest concordance with [ 18 F]AZD4694, although subjects were observed in both discordant quadrants (Figure 2). Conclusion: All CSF p‐tau biomarkers relate to NFT pathology to the same degree, however, tau368/t‐tau seemingly characterises NFT pathology at an advanced stage. CSF p‐tau217 and p‐tau231 are better predictors of Aβ pathology, with p‐tau231 a stronger predictor of Aβ pathology at the pre‐symptomatic stage.