Abstract Introduction Conventional risk prediction algorithms are inadequate in precisely defining risk of adverse cardiovascular outcomes. This residual risk may be better defined by assessing the activation of specific pathophysiological pathways involved in atherosclerotic disease progression. We developed a 4-protein biomarker-based risk score (BRS) for cardiovascular event prediction in the general population. Methods Levels of high-sensitivity C-reactive protein (hsCRP), high-sensitivity cardiac troponin (hs-cTn), soluble urokinase plasminogen activator receptor (suPAR), and brain natriuretic peptides (BNP or NT-proBNP) were measured at baseline in 4,586 individuals without known CAD enrolled in the Multi-Ethnic Study of Atherosclerosis (MESA) and prospectively followed for the incidence of incidence of hard cardiovascular disease (CVHD) defined as myocardial infarction (MI), resuscitated cardiac arrest, coronary heart disease death, stroke, and stroke death. Sex-specific binary thresholds for each biomarker were derived in a discovery cohort (50% of MESA participants) using the absolute value of the log-rank statistic and replicated in the remaining validation cohort. Each biomarker value above the cut-off was scored 1, and the BRS varied between 0 and 4. Fine-gray sub-distribution or Cox proportional hazard models adjusted for traditional cardiovascular risk factors. Risk discrimination analyses were conducted in comparison to the 10-year Atherosclerotic Cardiovascular Disease (ASCVD) risk assessment using the Predicting Risk of Cardiovascular Disease Events (PREVENT) equations. Results The validation cohort included 2,293 patients, with mean ± SD age 62.9 ± 10.3, 49% women and 39% white. After 12.6±4.4 years follow-up, 268 (11.6%) developed CVHD. Each biomarker was predictive of cardiovascular events, independent of traditional risk factors and other biomarkers (Data not shown). In the validation cohort, a 1-unit increase in BRS was independently associated with CVHD events (adjusted HR 1.57, 95% CI 1.44, 1.72, Figure A). The BRS improved prediction beyond the PREVENT risk score with a delta Harrell’s c-statistic of (0.02 95% CI 0.01,0.04) and net reclassification improvement of 0.207. The BRS improved prediction, particularly among intermediate and high PREVENT risk subgroups, p for interaction=0.47(Figure B). Conclusions A BRS based on 4 circulating proteins is an independent predictor and reclassifies the risk of adverse cardiovascular events in MESA beyond traditional risk assessment tools. Whether optimizing treatment based on the BRS will improve outcomes needs to be studied.
Inojosa et al. (Sat,) studied this question.