# Plasma Protein Panel for Assessing the Risk of Alzheimer’s Disease by MRM-MS Analysis: The Study of Two Independent Clinical Cohorts

## TL;DR

Researchers identified a blood protein panel that can help detect Alzheimer's disease and track cognitive decline using a reliable and scalable method.

## Contribution

A universal plasma protein panel was developed for Alzheimer's risk assessment with strong classification performance across independent cohorts.

## Key findings

- A panel of 13 proteins showed strong classification performance in distinguishing Alzheimer's patients from controls (ROC-AUC = 0.90).
- The same panel effectively separated stable and progressing mild cognitive impairment cases (ROC-AUC = 0.81).
- The MRM method demonstrated high potential for validating candidate biomarkers in independent clinical cohorts.

## Abstract

Early recognition of a risk of Alzheimer’s disease (AD) remains a global challenge, and blood proteomic markers are of particular interest for wide-scale diagnostic use. Quantitative multiple reaction monitoring (MRM) approach demonstrates good reproducibility in the characteristic changes in the levels of reported candidate biomarkers (CBs) in different cohorts in AD. Following up on our previous study, we performed a joint analysis of 331 blood plasma samples from two different clinical cohorts of participants, comprising a total of 95 samples from patients with AD, 136 samples from patients with mild cognitive impairment (MCI), and 100 samples from controls. The obtained results confirm the significance of 37 CBs. A logistic regression-based algorithm was used to build protein classifiers, and a total of 21 important proteins were selected, 13 of which (ORM1, APOA4, LBP, HP, FN1, BCHE, APOE, PZP, A1BG, TF, SERPINA7, TTR, and F12) formed a universal panel that demonstrated strong classification performance in distinguishing AD patients from controls (ROC-AUC = 0.90) and in separating stable and progressing patients with MCI (ROC-AUC = 0.81). Overall, the analysis confirms the high potential of the MRM method for validating CBs in independent cohorts.