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

Thomas K Karikari, Tharick A Pascoal, N. J Ashton, S. Janelidze, A. Lessa Benedet, J. Lantero Rodriguez, M. Chamoun, M. Savard, M. Su Kang, J. Therriault, M. Schöll, G. Massarweh, JP. Soucy, K. Höglund, G. Brinkmalm, N. Mattsson, S. Palmqvist, S. Gauthier, E. Stomrud, H. Zetterberg, O. Hansson, P. Rosa-Neto, K. Blennow

Lancet Neurol 2020; 19: 422–33 

Selezionato dal Lettore: Giuseppe Piga - Università degli studi di Cagliari

Motivation: Amyloid β and tau pathology are the pathological features of Alzheimer's disease (AD). At the moment, blood-based biomarkers are not reliable, accessible, cost-effective and highly specific for the detection in-vivo of AD pathology. CSF and PET biomarkers of amyloid β and tau are highly accurate for detecting AD pathology, but their invasiveness, high cost and poor availability limits their use in clinical practice; on the other hand, mass spectrometry-based assays for plasma amyloid β, which reflects brain amyloidosis, have proved to be not specific to AD. 

Previous studies using immunoassays targeting distinct tau species reported promising results for blood tau phosphorilated at threonine 181 (p-tau181) as a biomarker of AD. However, these assays lack analytical sensitivity, and it is unclear whether they are able to detect Alzheimer-specific tau pathology. The aim of this study was to assess whether blood p-tau181could be used as a biomarker of AD and for prediction of cognitive decline and hippocampal atrophy.   

Methods: The authors developed and validated an ultrasensitive blood immunoassay measuring p-tau181 on a novel N-terminal form of tau, specific to AD, using four clinic-based prospective cohorts: the discovery cohort (n=37), two indipendent validation cohorts, from the TRIAD and BioFINDER-2 studies (respectively n=226 and n=763), and the primary care cohort (n=105), with a total of 1131 individuals. The discovery cohort recruited patients with AD and aged-matched controls; the indipendent validation cohorts included older people and patients with AD, mild cognitive impairement, dementia, fronto temporal dementia, and young adults (20-30y); finally, the primary care cohort included controls without a diagnosis of a neurological condition and patients with clinical diagnosis, which however had not yet undergone biomarkers checks. Partecipants underwent blood p-tau181 biomarker, CSF and PET biomarkers of amyloid β and tau, and were assessed according to sex, APOE ε 4 genotype and education. Brain atrophy was measured by grey matter density on T1 weighted MRI images. 

Data and statistical analysis: Only individual with complete data were included in each specific analysis. Each variable, and its association with the outcomes, is deeply examinated. Despite the complexity of the study, data are reported clearly, in a fully comprehensive manner. The performance of plasma p-tau181 to accurately identify AD diagnosis and increased amyloid and tau PET was compared with the other plasma and CSF biomarker, using AUC analyses. Likewise, it was compared with two of the most well known risk factors for AD (age, APOE ε 4 genotype, or both).

Clinical and scientific impact: This view confirms the specificity of blood p-tau181 to AD, as previously shown in CSF, detects a strong correlation between blood and CSF concentrantions of p-tau181, indicating that it specifically measures brain-derived p-tau181, and suggests potential biological links between tau production and amyloid β plaques. 

Furthermore, blood p-tau181 differentiated AD from several other neurodegenerative disease, distinguished the phases of the disease in AD continuum, predicted amyloid and tau PET, prognosticated cognitive decline and hyppocampal atrophy over a period of 1 year. 

The ultrasensitive blood p-tau181 immunoassay described in this original study could represents the first simple, cost-saving, time-saving, practical and scalable test for the screening, diagnosis and prediction of the progression of Alzheimer's disease.