Clinical value of plasma pTau181 to predict Alzheimer's disease pathology in a large real-world cohort of a memory clinic

Abstract

Background: The identification of patients with an elevated risk of developing Alzheimer's disease (AD) dementia and eligible for the disease-modifying treatments (DMTs) in the earliest stages is one of the greatest challenges in the clinical practice. Plasma biomarkers has the potential to predict these issues, but further research is still needed to translate them to clinical practice. Here we evaluated the clinical applicability of plasma pTau181 as a predictive marker of AD pathology in a large real-world cohort of a memory clinic.

Methods: Three independent cohorts (modelling [n = 991, 59.7% female], testing [n = 642, 56.2% female] and validation [n = 441, 55.1% female]) of real-world patients with subjective cognitive decline (SCD), mild cognitive impairment (MCI), AD dementia, and other dementias were included. Paired cerebrospinal fluid (CSF) and plasma samples were used to measure AT(N) CSF biomarkers and plasma pTau181.

Findings: CSF and plasma pTau181 showed correlation in all phenotypes except in SCD and other dementias. Age significantly influenced the biomarker's performance. The general Aβ(+) vs Aβ(-) ROC curve showed an AUC = 0.77 [0.74-0.80], whereas the specific ROC curve of MCI due to AD vs non-AD MCI showed an AUC = 0.89 [0.85-0.93]. A cut-off value of 1.30 pg/ml of plasma pTau181 exhibited a sensitivity of 93.57% [88.72-96.52], specificity of 72.38% [62.51-79.01], VPP of 77.85% [70.61-83.54], and 8.30% false negatives in the subjects with MCI of the testing cohort. The HR of cox regression showed that patients with MCI up to this cut-off value exhibited a HR = 1.84 [1.05-3.22] higher risk to convert to AD dementia than patients with MCI below the cut-off value.

Interpretation: Plasma pTau181 has the potential to be used in the memory clinics as a screening biomarker of AD pathology in subjects with MCI, presenting a valuable prognostic utility in predicting the MCI conversion to AD dementia. In the context of a real-world population, a confirmatory test employing gold-standard procedures is still advisable.

Funding: This study has been mainly funded by Ace Alzheimer Center Barcelona, Instituto de Salud Carlos III (ISCIII), Biomedical Research Networking Centre in Neurodegenerative Diseases (CIBERNED), Spanish Ministry of Science and Innovation, Fundación ADEY, Fundación Echevarne and Grífols S.A.

Keywords: Alzheimer's disease; Mild cognitive impairment; Plasma biomarkers; Prodromal AD; Real-world cohort; pTau181.

Fig. 1

Scatter plots show the correlation between z-transformed CSF and plasma pTau181 levels in both modelling and testing cohort (n = 1628).

Fig. 2

Comparison of plasma pTau181 levels among a) the different phenotypes of the whole study and b) MCI sub-groups. Non-parametric Kruskal–Wallis test. Statistic values correspond to adjusted p Value. SNAP, suspected non amyloidosis profile. #Prodromal AD is referred to those subjects with MCI due to AD.

Fig. 3

a) Contribution of different clinical variables in plasma pTau181 performance. Naïve plasma pTau181 (“all” group) was set as reference. b) Comparison of plasma pTau181 levels between Aβ(+/−). Patients with MCI stratified by age. Statistic values correspond to p Value.

Fig. 4

ROC curves developed from the modelling cohort to stablish the optimal cut off value for plasma pTau181.

Fig. 5

Plasma pTau181 results of conversion in the population with MCI. a) Differences of plasma pTau181 levels between converters and non-converters subjects with MCI. b) Rates of conversion of the subjects with MCI clustered by plasma pTau181 cut off (1.30 pg/ml). c) Proportions of subjects with MCIa and MCIb in all the MCI subgroups. d) Survival curves from non-adjusted cox regression analysis of MCI conversion to AD dementia grouped by plasma pTau181 cut off value (1.30 pg/ml). Follow up data from the validation cohort belong to GR@CE/DESCO cohort. “Accumulated survival” refers to the time until conversion to dementia. ∗MCIa, subjects with MCI and plasma pTau181 levels above the cut off value; MCIb, subjects with MCI and plasma pTau181 levels below the cut off value. SNAP, suspected non amyloidosis profile. ^#Prodromal AD is referred to those subjects with MCI due to AD.