Log in to MyACC
Targeted Proteomics Improves CV Risk Prediction in Secondary Prevention
Quick Takes
- This study compares a proteomics-based risk model to a traditional risk factor model in predicting recurrent atherosclerotic events in two observational cohorts.
- The proteomics-based model outperformed the clinical risk factors model in both cohorts.
- The proteins associated with the outcomes differed when patients were stratified according to CRP levels, with interleukin-6 being prominent in the high CRP group, while neutrophil-signaling proteins were more prominent in the low CRP group.
Study Questions:
Does targeted plasma proteomics improve cardiovascular (CV) risk prediction in secondary prevention patients? Are different pathways contributing to CV risk in high and low C-reactive protein (CRP) patients?
Methods:
Targeted plasma proteomics was performed using the Olink Cardiovascular II, III, and Cardiometabolic panels, in two secondary prevention cohorts: the SMART (Second Manifestations of ARTerial disease) cohort (n = 870), which was the derivation cohort; and the Athero-Express cohort (n = 700), which was the validation cohort. The SMART cohort is a prospective single-center cohort in which patients aged <80 years with myocardial infarction, stroke, or transient ischemic attack (TIA) with a high risk of atherosclerotic events were included. The Athero-Express cohort included patients who had a stroke or TIA following endarterectomy. The primary outcome was first recurrent atherosclerotic CV event (myocardial infarction, ischemic stroke, or CV death). Predictive models with the top 50 proteins were derived in the SMART cohort using machine learning approaches, then validated in the Athero-Express cohort, and finally compared to a clinical risk score model including traditional risk factors in addition to CRP levels. Pathway analysis was performed for the subgroup of patients with high CRP (>2 mg/L) and low CRP (≤2 mg/L) levels to understand differences in inflammatory pathways between these two subsets.
Results:
In the derivation cohort, 263 (30.2%) participants experienced a recurrent event during a median follow-up of 8.0 (4.6-12.2) years. In the validation cohort, 130 (18.6%) participants experienced a recurrent event during a median follow-up of 3.0 (2.2–3.1) years. The final proteomic analysis included 267 unique proteins. The protein model outperformed the clinical model in both the derivation cohort (area under the curve [AUC], 0.810 vs. 0.750; p > 0.001) and validation cohort (AUC, 0.801 vs. 0.765; p < 0.001). The top 10 proteins in terms of relative importance were NT-proBNP, KIM1, MMP7, GDF-15, HAOX1, TGFBI, ENG, BNP, ADM, and UPAR. Interleukin-6 was a top protein in the high CRP subset, but not in the low CRP group. Conversely, neutrophil-signaling-related proteins (AMBP, NID1, VASN, TF) were associated with recurrent events in the low CRP patients.
Conclusions:
A proteomics-based risk prediction model outperformed traditional clinical risk factors in predicting recurrent atherosclerotic events. The top proteins in the model differed between patients with high and low CRP levels, suggesting that differing pathways underlie events in these subgroups of patients.
Perspective:
Understanding determinants of residual atherosclerotic CV disease risk and how to reduce represents the next frontier in the management of atherosclerotic diseases. The predictive ability of inflammatory biomarkers such as CRP pointed to the central role of chronic inflammation in CV disease. Inflammation, however, is a simple term for very complex processes that cannot be captured or represented by a single biomarker. The promise of proteomics lies in a more comprehensive assessment of the various pathways associated with atherosclerosis. This study exemplifies the potential benefits of proteomics for both improving risk prediction and our understanding of the underlying pathways. Translating findings into advances in clinical management faces many barriers extending from the limitations of measurement technique such as the poor correlation between proteomic measurements and enzyme-linked immunosorbent assay (ELISA), to logistics and cost of implementation in a clinical setting. The most important next step is determining whether the additional information provided by proteomic measurements can lead to a change in clinical management, whether diagnostic or therapeutic. Until then, proteomics remains an invaluable research tool for pathway and biomarker discovery.
Clinical Topics: Arrhythmias and Clinical EP, Cardiovascular Care Team, Invasive Cardiovascular Angiography and Intervention, Prevention, Vascular Medicine, Interventions and Vascular Medicine
Keywords: Atherosclerosis, Biomarkers, Brain Ischemia, Cardiometabolic Risk Factors, C-Reactive Protein, Endarterectomy, Heart Disease Risk Factors, Inflammation, Interleukin-6, Ischemic Attack, Transient, Ischemic Stroke, Machine Learning, Myocardial Infarction, Neutrophils, Proteomics, Secondary Prevention, Stroke, Vascular Diseases
< Back to Listings
