Objective:

To provide an accessible and clinically viable plasma panel to assist physicians with the diagnosis of Alzheimer’s disease (AD).

Background:

The ATN framework has been widely utilized for the investigation of AD but has predominantly relied on neuroimaging or CSF measurements which are costly and invasive.  Plasma-based measurements of an ATN panel, however, would provide a more accessible and affordable analytical modality for investigation of AD.

Design/Methods:

Comprehensive analytical validation of assays on high-throughput clinical auto-analyzers were previously performed for measurements of amyloid-beta (Aβ) 1-42 and Aβ 1-40 (for determination of Aβ42/40), phosphorylated-tau181 (pTau181), and neurofilament light chain (NfL).  To investigate the clinical efficacy of each assay, 200 clinical specimens with Aβ status defined using PET imaging were measured.  Specimens were measured on the validated assays as well as predicate assays utilized in clinical studies.  Additionally, receiver operator characteristic (ROC) analysis was performed on assay results to assess the diagnostic ability of each assay.

Results:

Comparison of measurements to predicate assays produced a wide degree of correlations (R between 0.64-0.96) with mean biases ranging from -86.5 to 88.9%.  However, ROC analysis of clinical sample results from validated assays produced an area-under-the-curve (AUC) of 0.941 for Aβ42/40, 0.847 for pTau181, and 0.666 for NfL.  These results are consistent with the understood progression of AD along the ATN continuum.  Despite the relative low AUC, NfL results were found to increase as Mini Mental State Exam scores worsened from mild/negligible (21-30) to moderate scores (11-20, p < 0.0001) and from moderate to severe scores (0-10, p < 0.001).

Conclusions: