Articles
Blood phosphorylated tau 181 as a biomarker for Alzheimer's disease: a diagnostic performance and prediction modelling study using data from four prospective cohorts

https://doi.org/10.1016/S1474-4422(20)30071-5Get rights and content

Summary

Background

CSF and PET biomarkers of amyloid β and tau accurately detect Alzheimer's disease pathology, but the invasiveness, high cost, and poor availability of these detection methods restrict their widespread use as clinical diagnostic tools. CSF tau phosphorylated at threonine 181 (p-tau181) is a highly specific biomarker for Alzheimer's disease pathology. We aimed to assess whether blood p-tau181 could be used as a biomarker for Alzheimer's disease and for prediction of cognitive decline and hippocampal atrophy.

Methods

We developed and validated an ultrasensitive blood immunoassay for p-tau181. Assay performance was evaluated in four clinic-based prospective cohorts. The discovery cohort comprised patients with Alzheimer's disease and age-matched controls. Two validation cohorts (TRIAD and BioFINDER-2) included cognitively unimpaired older adults (mean age 63–69 years), participants with mild cognitive impairment (MCI), Alzheimer's disease, and frontotemporal dementia. In addition, TRIAD included healthy young adults (mean age 23 years) and BioFINDER-2 included patients with other neurodegenerative disorders. The primary care cohort, which recruited participants in Montreal, Canada, comprised control participants from the community without a diagnosis of a neurological condition and patients referred from primary care physicians of the Canadian National Health Service for specialist care. Concentrations of plasma p-tau181 were compared with established CSF and PET biomarkers and longitudinal measurements using Spearman correlation, area under the curve (AUC), and linear regression analyses.

Findings

We studied 37 individuals in the discovery cohort, 226 in the first validation cohort (TRIAD), 763 in the second validation cohort (BioFINDER-2), and 105 in the primary care cohort (n=1131 individuals). In all cohorts, plasma p-tau181 showed gradual increases along the Alzheimer's disease continuum, from the lowest concentrations in amyloid β-negative young adults and cognitively unimpaired older adults, through higher concentrations in the amyloid β-positive cognitively unimpaired older adults and MCI groups, to the highest concentrations in the amyloid β-positive MCI and Alzheimer's disease groups (p<0·001, Alzheimer's disease vs all other groups). Plasma p-tau181 distinguished Alzheimer's disease dementia from amyloid β-negative young adults (AUC=99·40%) and cognitively unimpaired older adults (AUC=90·21–98·24% across cohorts), as well as other neurodegenerative disorders, including frontotemporal dementia (AUC=82·76–100% across cohorts), vascular dementia (AUC=92·13%), progressive supranuclear palsy or corticobasal syndrome (AUC=88·47%), and Parkinson's disease or multiple systems atrophy (AUC=81·90%). Plasma p-tau181 was associated with PET-measured cerebral tau (AUC=83·08–93·11% across cohorts) and amyloid β (AUC=76·14–88·09% across cohorts) pathologies, and 1-year cognitive decline (p=0·0015) and hippocampal atrophy (p=0·015). In the primary care cohort, plasma p-tau181 discriminated Alzheimer's disease from young adults (AUC=100%) and cognitively unimpaired older adults (AUC=84·44%), but not from MCI (AUC=55·00%).

Interpretation

Blood p-tau181 can predict tau and amyloid β pathologies, differentiate Alzheimer's disease from other neurodegenerative disorders, and identify Alzheimer's disease across the clinical continuum. Blood p-tau181 could be used as a simple, accessible, and scalable test for screening and diagnosis of Alzheimer's disease.

Funding

Alzheimer Drug Discovery Foundation, European Research Council, Swedish Research Council, Swedish Alzheimer Foundation, Swedish Dementia Foundation, Alzheimer Society Research Program.

Introduction

More than 50 million people worldwide have dementia, and the cost of dementia care reached US$1 trillion in 2018.1 Amyloid β and tau pathology are the defining pathological features of Alzheimer's disease.2 In-vivo detection of these processes is central to disease diagnosis,3 its biological definition,4 and for selecting individuals for clinical trials.5 Although CSF and PET biomarkers of amyloid β and tau are highly accurate for detecting Alzheimer's disease pathology, the costs and low availability of the tools needed to detect these biomarkers hamper their feasibility for use in clinical diagnostic practice and for screening in clinical trials.6The accessibility and cost-effectiveness of blood-based biomarkers make them attractive for first-line clinical use and for facilitating clinical trial recruitment and monitoring.7 Blood neurofilament light chain, a marker of neuronal injury, is increased in Alzheimer's disease,8 but this biomarker has low specificity, because abnormal increases are also reported in several other neurodegenerative disorders, such as multiple system atrophy, corticobasal syndrome, and progressive supranuclear palsy.9

Research in context

Evidence before this study

We searched PubMed for all articles published from database inception to Jan 20, 2020, without language restrictions, using the keywords “tau”, “phosphorylated tau”, “CSF tau”, “CSF biomarker”, “Alzheimer's disease”, “plasma tau”, “amyloid”, “MRI”, “PET”, “cognitive decline”, and “hippocampal atrophy”. Previous attempts to develop a reliable blood assay for p-tau181 have been challenging due to the very low concentrations in blood samples. Furthermore, initial unsuccessful efforts focused on applying established mid-region CSF p-tau181 immunoassays directly to blood. Recent evidence has shown that tau in blood and CSF might be processed differently, with mainly N-terminal forms of tau present in measurable quantities in blood. A few studies, each targeting different tau species, have described blood p-tau181 immunoassays showing encouraging results in few patient cohorts. However, some of these assays lack the analytical sensitivity for examining cognitively unimpaired individuals, some of whom might be in the preclinical phase of Alzheimer's disease. Moreover, it is unclear if previously described blood p-tau181 assays detect either Alzheimer-specific tau pathology similar to CSF p-tau181 or tau pathology that is common to all neurodegenerative diseases, characterised by the presence of pathological tau.

Added value of this study

In this study, we present a blood-based immunoassay measuring p-tau181 on a novel N-terminal form of tau that is distinct from the mid-region forms targeted by the established CSF assays. This assay was validated to be specific for the p-tau181 site, does not capture non-phosphorylated tau species, and shows good diagnostic performance for Alzheimer's disease in both plasma and serum. The blood p-tau181 assay identified Alzheimer's disease at the very early stages of disease and demonstrated high diagnostic accuracy, with stepwise increases across the Alzheimer's disease continuum. Similar to mid-region CSF p-tau181, our blood p-tau181 assay appeared to be specific to Alzheimer's disease, differentiating it from other neurodegenerative diseases with high accuracy. Additionally, blood p-tau181 predicted cognitive decline and hippocampal atrophy over a period of 1 year, making it suitable as an Alzheimer's disease progression marker. Furthermore, plasma p-tau181 performed better than the most well-known Alzheimer's disease risk factors (age, APOE ε4 genotype, or both) and other plasma biomarkers (total-tau, amyloid β1–42, amyloid β1–42 to amyloid β1–40 ratio, and total-tau to amyloid β1–42 ratio) in predicting Alzheimer's disease diagnosis, increased tau PET, and increased amyloid β PET.

Implications of all the available evidence

The blood p-tau181 assay described in this study could represent the first simple, practical, and scalable test for the diagnosis of Alzheimer's disease. This technology has applications for diagnosis and recruitment for disease-modifying trials. The blood p-tau181 assay has the potential to be incorporated into clinical practice as a rapid screening test to identify or rule out Alzheimer's disease pathophysiology and to guide therapy and clinical management of patients with suspected neurodegenerative disorders. To facilitate widespread implementation of the blood p-tau181 assay, full clinical standardisation, including establishment of reference materials and methods to harmonise readouts across clinical laboratories, will be required.

Another advance in biomarkers for Alzheimer's disease is the use of mass spectrometry-based assays for plasma amyloid β (ratio of amyloid β1–42 to amyloid β1–40), which reflects brain amyloidosis.10, 11 However, these assays have limitations, including substantial peripheral amyloid β expression,12 which results in less pronounced decreases in the amyloid β1–42 to amyloid β1–40 ratio in plasma compared with CSF, and a larger overlap of amyloid β concentrations between individuals who are amyloid β PET-positive and PET-negative.10 Furthermore, brain amyloidosis is present in 10–30% of individuals who are cognitively unimpaired.13 By contrast, CSF tau phosphorylated at threonine 181 (p-tau181) is a highly specific pathological marker of Alzheimer's disease that remains normal in other dementias.14 Thus, a blood test for p-tau181 would be a major advance for diagnostics. Some previous studies using immunoassays targeting distinct tau species reported promising results for blood p-tau181 as a biomarker for Alzheimer's disease.15, 16, 17, 18 However, some of these assays had insufficient analytical sensitivity for examining preclinical and cognitively unimpaired individuals, and it is unclear whether Alzheimer-specific tau pathology was detected.

In this study, we report the performance of an ultrasensitive immunoassay for blood p-tau181 that can be implemented for a practical assessment of in-vivo Alzheimer's disease pathophysiology. We aimed to evaluate whether blood p-tau181 can: (1) differentiate Alzheimer's disease dementia from no cognitive impairment, mild cognitive impairment (MCI) due to Alzheimer's disease, and other neurodegenerative diseases; (2) reflect abnormalities in tau or amyloid PET scans; and (3) predict future cognitive decline and hippocampal atrophy.

Section snippets

Study design and population

We developed a new analytical method for blood p-tau181 and validated its performance for assessment of Alzheimer's pathophysiology using four independent, prospective cohorts recruiting consecutive cases. The discovery cohort included patients with Alzheimer's disease, with a typical Alzheimer's disease core CSF biomarker profile (CSF amyloid β1–42 <530 ng/L, p-tau181 >60 ng/L, and total-tau >350 ng/L19), and age-matched controls, who were patients examined at the memory or neurology clinics

Results

We studied 37 individuals in the discovery cohort, 226 in the first validation cohort (TRIAD), 763 in the second validation cohort (BioFINDER-2), and 105 in the primary care cohort (total of 1131 individuals). Blood p-tau181 concentrations were not affected by sex. Demographic characteristics are presented in Table 1, Table 2, and in the appendix (pp 19, 20).

The blood p-tau181 assay (appendix p 10) showed high analytical performance (appendix p 21), with high precision within and between runs,

Discussion

Our high-performance blood p-tau181 assay enabled the identification of brain tau pathology, showing increased concentrations of blood p-tau181 in individuals with amyloid β pathology who were tau PET-negative. Moreover, plasma p-tau181 provided high diagnostic accuracy for Alzheimer's disease in four independent cohorts, discriminated amyloid β-positive cognitively unimpaired older adults and amyloid β-positive individuals with MCI from amyloid β-negative cognitively unimpaired older adults

References (32)

  • A Lleó et al.

    Cerebrospinal fluid biomarkers in trials for Alzheimer and Parkinson diseases

    Nat Rev Neurol

    (2015)
  • JL Molinuevo et al.

    Current state of Alzheimer's fluid biomarkers

    Acta Neuropathol

    (2018)
  • H Hampel et al.

    Blood-based biomarkers for Alzheimer disease: mapping the road to the clinic

    Nat Rev Neurol

    (2018)
  • N Mattsson et al.

    Association between longitudinal plasma neurofilament light and neurodegeneration in patients with Alzheimer disease

    JAMA Neurol

    (2019)
  • O Hansson et al.

    Blood-based NfL: a biomarker for differential diagnosis of parkinsonian disorder

    Neurology

    (2017)
  • A Nakamura et al.

    High performance plasma amyloid-β biomarkers for Alzheimer's disease

    Nature

    (2018)
  • Cited by (632)

    • Menopausal vasomotor symptoms and plasma Alzheimer disease biomarkers

      2024, American Journal of Obstetrics and Gynecology
    View all citing articles on Scopus
    *

    Co-first authors

    Contributed equally

    View full text