Honing Discovery of Plasma Protein Biomarkers for Profiling Early Parkinson’s Disease Progression
David Brazel1, Ludmila Voloboueva1, Matt Burrill1, Katrina Paumier1, Ferhan Qureshi1, Evan Boyle1
1Octave Bio
Objective:
To identify plasma biomarkers associated with early Parkinson’s disease (PD) progression
Background:
Biofluid biomarkers are desirable for assessing progression of and risk for PD when motor impairment is within range of neurologically healthy controls (HCs). Here, we developed protein quantitation standardization for sensitive detection of biomarkers, before motor deficits can be accurately scored.
Design/Methods:
We harnessed Olink Explore 1536 data from the Parkinson’s Progression Markers Initiative (PPMI) cohort (http://www.ppmi-info.org/) to discover biomarkers linked to PD pathophysiology. We stratified plasma protein markers with k-means clustering and used cluster content, age, sex, and plate ID among HCs to standardize protein quantitations for all markers in healthy controls, prodromal cases, and sporadic PD cases under 2 years of disease duration. The average standardized quantitation per subject of prodromal and PD cases were separately tested against those of HCs. Candidate prodromal markers were queried against UK Biobank (UKB) early PD data.
Results:
7 plasma proteins passed multiple-test correction for prodromal cases versus HCs. The 6 markers also measured in UKB-PPP shared the same direction of effect (one-sided Binomial test p = 0.02). 4 of the 7 were nominally significant in all three comparisons, and one protein, ITGAV, exhibited Bonferroni significant depletion in all comparisons. Standardizing quantitations enhanced discrimination of prodromals from HCs (AUC for ITGAV increased from 0.719 to 0.754, paired t test p = 0.01 for AUCs of 7 significant markers).
Conclusions:
Adjusting for age, sex, and total protein levels during standardization of log-scale protein quantitations using trends from HCs is a straightforward approach to improve biomarker association and account for demographic factors correlated with disease status when evaluating marker performance. Accurate discrimination of HCs from early PD cases and monitoring molecular disease pathways is attainable with 5-10 plasma proteins.
Disclaimer: Abstracts were not reviewed by Neurology® and do not reflect the views of Neurology® editors or staff.