Plasma ratio of phospho‐ to non‐phospho‐tau at threonine 217 identifies brain amyloid burden as measured by quantitative amyloid PET in the PARIS sub‐study of IDEAS

Abstract Background Accumulation of amyloid beta plaques and tau tangles in the brain are sequential pathological hallmarks of Alzheimer’s disease. Specific measurement of tau phosphorylation sites in plasma samples can be used to investigate the presence of both amyloid plaque and tau tangle brain pathology. Method A tandem mass spectrometry assay that quantifies phosphorylation of tau at threonine residues 181 and 217 in plasma was developed. Both phosphorylated (phospho) and non‐phosphorylated (non‐phospho) peptides containing tau residues 181 and 217 as well as the ratio of phospho to non‐phospho at each of these phosphorylation sites can be calculated from a blood sample. We tested the concordance between these measures and brain amyloid burden in PARIS study participants, an IDEAS sub‐study in patients with mild cognitive impairment or dementia. K 2 EDTA plasma samples from PARIS participants (n=221) were analyzed for tau phosphorylation and Aβ42/40. Results The amount of phosphorylation at both 181 and 217 sited were significantly higher in patients who were amyloid PET positive, while PET negative patients showed very low amounts of phosphorylation at 217 (1.2 pg/mL vs. 4.9 pg/mL; p <2e ‐16 ). The ratio of phospho‐tau to non‐phospho‐tau 217 showed high correlation (Spearman’s rank correlation r=0.78; p <2e‐16) with quantitative amyloid PET, demonstrating that phosphorylation of tau on threonine‐217 can quantitatively identify the amount of amyloid burden. This ratio of phospho‐tau to non‐phospho‐tau also correlated better with amyloid status than the concentration of phospho‐tau alone, with area under the curve (AUC) for detecting brain amyloid burden improving from 0.74 (95%CI 0.67‐0.82) to 0.81 (95%CI 0.75‐0.88) for 181 and from 0.92 (95%CI 0.88‐0.96) to 0.95 (95%CI 0.92‐0.98) for 217. Combining the phospho‐tau‐217 with the Aβ42/40 ratios for the same sample further increased the AUC to 0.96 (95%CI 0.93‐0.99). Conclusion This novel mass spectrometry‐based quantitative assay for plasma phospho‐tau and non‐phospho‐tau species has excellent analytical and clinical validity performance characteristics. Combining the ratio of tau phosphorylation at 217 with plasma Aβ42/40 can help identify brain amyloid pathology with a performance that rivals that of CSF or PET tests.