Weekly Journal Scan: incremental utility of circulating biomarkers for cardiovascular risk prediction

Key Points• The present study is an individual-level data analysis from multinational population-based cohorts aiming to assess the comparative predictive value of circulating cardiovascular biomarkers for incident atherosclerotic cardiovascular disease (ASCVD) events in the general population, alone and in addition to established clinical risk factors, and to evaluate their differential prognostic value according to age.• Inclusion criteria for eligible cohorts were (i) enrolment of apparently healthy individuals (i.e.no history of major atherothrombotic cardiovascular events); (ii) at least one measurement of high-sensitivity troponin I (hs-TnI), high-sensitivity troponin T (hs-TnT), B-type natriuretic peptide (BNP), N-terminal pro-BNP (NT-proBNP), or high-sensitivity C-reactive protein; and (iii) follow-up duration of at least 2 years.• The primary outcome was incident ASCVD, defined by the first coronary heart disease event, ischaemic stroke event, coronary revascularization, coronary heart disease death, ischaemic stroke death, or unclassifiable death.The secondary outcomes were all-cause death, heart failure (HF), ischaemic stroke, and myocardial infarction.Sub-distribution hazard ratios (HRs) for the association of biomarkers with outcomes were assessed after adjustment for conventional CV risk factors (i.e.age, total cholesterol, HDL cholesterol, smoking status, diabetes, systolic blood pressure, and self-reported use of antihypertensive drugs).The incremental predictive value of biomarkers was estimated using the C-statistic and re-classification analyses.• A total of 164 054 individuals (median age, 53 years; 52% female) from 28 general population-based cohorts across 12 countries and 4 continents were included in the analyses.The median 10-year ASCVD risk Systematic Coronary Risk Evaluation 2 (SCORE2) was 4% [interquartile range (IQR), 2%-9%], and the median 10-year risk using the Pooled Cohort Risk Equation was 5% (IQR, 1%-13%).A total of 17 211 ASCVD events occurred during a median follow-up of 12 years (IQR, 6-18 years).After adjusting for sex, cohort, and conventional CV risk factors, biomarker concentrations were associated with incident ASCVD events {sub-distribution HR per 1 SD change: 1.13 [95% confidence interval (CI), 1.11-1.16]for hs-TnI; 1.18 [95% CI, 1.12-1.23]for hs-TnT; 1.14 [95% CI, 1.08-1.22]for BNP; 1.21 [95% CI, 1.18-1.24]for NT-proBNP; and 1.14 [95% CI, 1.12-1.16]for high-sensitivity C-reactive protein}.The associations of biomarkers with all-cause death and HF were larger than those with ASCVD.The highest sub-distribution HRs were observed for HF with NT-proBNP (HR, 1.62; 95% CI, 1.56-1.68),BNP (HR, 1.59; 95% CI, 1.43-1.77),and hs-TnT (HR, 1.44; 95% CI, 1.38-1.51).• The addition of each single biomarker to a model including conventional CV risk factors improved the C-statistic for ASCVD at 1, 5, and 10 years.The incremental value of biomarkers was higher when hs-TnI, NT-proBNP, and high-sensitivity C-reactive protein were combined in a single model.This combination was associated with a C-statistic improvement from 0.81 (95% CI, 0.80-0.82) to 0.82 (95% CI, 0.81-0.83)for 10-year incident ASCVD events in individuals aged <65 years, and from 0.63 (95% CI, 0.59-0.66) to 0.66 (95% CI, 0.62-0.68) in older people.Improvements in risk prediction were larger for HF and all-cause death than those for ASCVD.