Proteomic predictors of mortality among people with Coronary Artery Disease

Background
Coronary artery disease (CAD) continues to pose a significant global health challenge, highlighting the crucial need to enhance prognostic accuracy and improve mortality outcomes for CAD management. Proteomics offers a nuanced perspective on the molecular intricacies underlying CAD progression. This study investigates proteomic profiles in people with CAD, aiming to identify protein markers associated with mortality outcomes.
Methods
Utilizing the UK Biobank (UKB) and the Mental Stress Ischemia Prognosis Study (MIPS), the study analyzed proteomic data from 2,768 CAD participants in the UKB and 91 CAD participants in the MIPS. Cox proportional hazards models, LASSO regression, and pathway enrichment analysis were employed to identify proteomic predictors of all-cause mortality. A proteomics score was derived using the LASSO-selected proteins.
Results
Out of 1,696 tested proteins, 341 were associated with all-cause mortality in the UKB. Seventeen proteins were selected by LASSO, with GDF15 and ANGPT2 validated in the MIPS. One standard deviation higher in the proteomics score was associated with all-cause mortality in both the UKB (HR 2.37, 95% CI 2.17 – 2.59, p 4.04 × 10-84) and the MIPS (HR 6.93, 95% CI 3.29 – 14.58, p 3.52 × 10-7). Mortality-associated proteins revealed enrichment in pathways related to inflammation and cellular adhesion.
Conclusion
This study expands our understanding of CAD prognosis by uncovering potential proteomic biomarkers. The development and validation of the proteomics score in two independent CAD cohorts in the UK and the US underscore the potential utility for enhancing prognostic precision in the context of CAD.