Background: Inflammation has been implicated in the pathogenesis of coronary heart disease, but the relevance and independence of individual inflammatory proteins is uncertain.
Objective: To examine the relationships between a spectrum of inflammatory proteins and myocardial infarction (MI).
Methods and results: A panel of 92 inflammatory proteins was assessed using an OLINK multiplex immunoassay among 432 MI cases (diagnosed < 66 years) and 323 controls. Logistic regression was used to estimate associations between individual proteins and MI, after adjustment for established cardiovascular risk factors and medication use, and stepwise regression to identify proteins with independent effects. Machine learning techniques (Boruta analysis and LASSO regression) and bioinformatic resources were used to examine the concordance of results with those obtained by conventional methods and explore the underlying biological processes to inform the validity of the associations. Among the 92 proteins studied, 62 (67%) had plasma concentrations above the lower limit of detection in at least 50% of samples. Of these, 15 individual proteins were significantly associated with MI after covariate adjustment and correction for multiple testing. Five of these 15 proteins (CDCP1, CD6, IL1-8R1, IL-6, and CXCL1) were independently associated with MI, with up to three-fold higher risks of MI per doubling in plasma concentrations. Findings were further validated using machine learning techniques and biologically focused analyses.
Conclusions: This study, demonstrating independent relationships between five inflammatory proteins and MI, provides important novel insights into the inflammatory hypothesis of MI and the potential utility of proteomic analyses in precision medicine.
The PROCARDIS study conducted a hypothesis-free proteomic study using a panel of 92 inflammatory proteins in cases with early onset myocardial infarction (MI) and healthy controls and identified 15 proteins that were significantly associated with MI, including five proteins that independently contributed to risk of MI. The study used state-of-the-art analytical methods including conventional statistical analysis and machine learning approaches to characterize the proteomic associations with MI. It also integrated bioinformatic and genomic data to consider the biological relevance of the proteins independently associated with MI. The findings provide novel insights into the ‘inflammatory basis’ of MI and provide support for prioritizing a wider array of inflammatory proteins for further study than have been previously considered in order to discover if therapeutic modification could be used for treatment and prevention of MI.
© The Author(s) 2023. Published by Oxford University Press on behalf of the European Society of Cardiology.