Abstract
Despite guideline-based therapy, patients with coronary artery disease (CAD) are at widely variable risk for cardiovascular events. This variability demands a more individualized risk assessment. Herein, we evaluate the prognostic value of 6 biomarkers: high-sensitivity C-reactive protein, heat shock protein-70, fibrin degradation products, soluble urokinase plasminogen activator receptor, high-sensitivity troponin I, and B-type natriuretic peptide. We then develop a multi-biomarker-based cardiovascular event prediction model for patients with stable CAD. In total, 3,115 subjects with stable CAD who underwent cardiac catheterization at Emory (mean age 62.8 years, 17% Black, 35% female, 57% obstructive CAD, 31% diabetes mellitus) were randomized into a training cohort to identify biomarker cutoff values and a validation cohort for prediction assessment. Main outcomes included (1) all-cause death and (2) a composite of cardiovascular death and nonfatal myocardial infarction (MI) within 5 years. Elevation of each biomarker level was associated with higher event rates in the training cohort. A biomarker risk score was created using optimal cutoffs, ranging from 0 to 6 for each biomarker exceeding its cutoff. In the validation cohort, each unit increase in the biomarker risk score was independently associated with all-cause death (hazard ratio 1.62, 95% confidence interval [CI] 1.45 to 1.80) and cardiovascular death/MI (hazard ratio 1.52, 95% CI 1.35 to 1.71). A biomarker risk prediction model for cardiovascular death/MI improved the c-statistic (Δ 6.4%, 95% CI 3.9 to 8.8) and net reclassification index by 31.1% (95% CI 24 to 37), compared with clinical risk factors alone. Integrating multiple biomarkers with clinical variables refines cardiovascular risk assessment in patients with CAD.
Patients with established coronary artery disease (CAD) are at high risk for recurrent cardiovascular events. Despite standardization of secondary prevention, a heterogeneous risk of recurrent events exists that demands a more individualized approach to risk assessment and management.1 Traditional risk factors and measures such as left ventricular ejection fraction (LVEF) are commonly used to assess risk but fail to reliably estimate the residual risk of recurrent events.2,3 The transition from stable disease to acute plaque disruption is complex and involves activation of multiple pathophysiologic pathways including inflammation, immune activation, cellular stress, coagulation, myocardial stretch, and ischemia.4–6 Circulating biomarkers representing dysregulation of these pathways may help identify the patients who are at greatest risk and thus aid individualization of preventive treatment. High-sensitivity C-reactive protein (hs-CRP) representing inflammation,7 soluble urokinase plasminogen activator receptor (suPAR) also representing immune activation,4,8 heat shock proteins (HSPs) representing cellular stress,9 fibrin degradation products (FDPs) representing coagulation,5,10 high-sensitivity troponins (hs-TnIs) representing myocardial stress/injury,11 and natriuretic peptides as a marker of myocardial stretch12–15 have shown prognostic significance in patients at risk of or with established CAD. Herein we (1) evaluate the prognostic value of these pathophysiology-specific biomarkers, and (2) develop and validate a biomarker-based model for prediction of adverse events in patients with stable CAD.
Methods
Participants were recruited from the Emory Cardiovascular Biobank, a prospective registry of subjects who underwent cardiac catheterization for the diagnosis of suspected CAD at 3 Emory Healthcare sites between 2004 and 2010. Details of the study population have been published previously.5 Briefly, subjects between ages 20 and 90 years who underwent cardiac catheterization for the diagnosis of suspected or confirmed CAD were interviewed to collect demographic information, medical history, risk factor prevalence, medication use, and behavioral habits as previously described.5 Subjects with acute coronary syndrome, congenital heart disease, cardiac transplantation, severe valvular heart disease, severe anemia, a recent blood transfusion, myocarditis, active inflammatory diseases, cancer, limited follow-up time (<30 days), and missing biomarker data were excluded. The study was approved by the Emory University Institutional Review Board. All subjects provided written informed consent.
Gender and race were self-reported. Anginal symptom data were self-reported using the Seattle Angina Questionnaire. The presence of hypertension, hyperlipidemia, diabetes mellitus (DM), peripheral vascular disease, and atrial fibrillation was determined by physician diagnosis and/or treatment prescribed in the medical chart. Smoking was classified as nonsmoker or current/past smoker. Blood pressure, weight, and height were measured on the enrollment date. Body mass index (BMI) was defined as weight in kilograms divided by height in meters. Serum create-nine measurements at enrollment before cardiac catheterization were obtained using data from routine follow-up clinic visits or hospitalizations within the Emory Healthcare system. The estimated glomerular filtration rate (eGFR) was computed using the chronic kidney disease (CKD) Epidemiology Collaboration equation.16 CKD was defined as eGFR <60 ml/min/1.73 m2. Echocardiographic LVEF was abstracted after reviewing medical records. Medications and medical history were obtained by self-reported history and followed by review of relevant medical records. The following medications were included in this analysis: aspirin, β blockers, angiotensin-converting enzyme inhibitor/angiotensin II receptor blocker (ACEi/ARB), and statins.
Luminal narrowing of coronary arteries was quantified using a modified American Heart Association/American College of Cardiology classification of the coronary arteries and using the Gensini score.17,18 Angiography outcomes data were extracted from reports published after catheterization.
Fasting arterial blood samples for serum and plasma were drawn before catheterization and stored at −80°C (mean, 4.9 years). Details of the biomarker assays have been previously described.4,5 Briefly, serum hs-CRP and FDP measurements were determined using a sandwich immunoassay (FirstMark Inc., San Diego, California). Serum HSP-70 was measured with a sandwich enzyme-linked immunosorbent assay (R&D Systems, Minneapolis, Minnesota) and optimized by FirstMark. Plasma suPAR levels were measured using commercially available kits (suPARnostic kit, Virogates, Copenhagen, Denmark). Plasma hs-TnI was measured using the Abbott ARCHITECT analyzer (Abbott Laboratories, North Chicago, Illinois). Plasma B-type natriuretic peptide (BNP) was also measured on the Abbott ARCHITECT BNP chemiluminescent microparticle immunoassay (Abbott Laboratories, reported in pg/ml). The limit of detection of hs-TnI is 1.2 pg/ml and an interassay coefficient of variation of <10% at 4.7 pg/ml. Minimum detectable hs-CRP, FDP, HSP-70, suPAR, and BNP were 0.1 mg/L, 0.06 μg/ml, 0.313 ng/ml, 0.1 ng/ml, and 10 pg/ml, respectively.
Primary end points of all-cause death and the composite end point of cardiovascular death and nonfatal myocardial infarction (MI) within 5 years of follow-up were adjudicated (Supplementary Materials). Major adverse cardiovascular events were defined as a composite of cardiovascular death, hospitalization for heart failure, nonfatal MI, or nonfatal stroke within 5 years.
The entire analysis data set (n = 3,115) was randomly split equally into training (n = 1,557) and validation (n = 1,558) sets with similar baseline characteristics. Using the training set, Cox regression model and Fine-Gray model were used to assess the association between all-cause death and cardiovascular death/MI (treating noncardiovascular death as competing risk event) and the 6 biomarkers (simultaneously included in the model as continuous variables). Covariates had age, gender, race (Black vs non-Black), BMI, history of smoking, LVEF, hypertension, DM, hyperlipidemia, history of MI, history of heart failure, Gensini score (categorical variable based on quartiles), eGFR, use of aspirin, ACEi/ARBs, β blockers, and statins. The optimal cut point was identified for each biomarker with gender-specific cut points for suPAR and hs-cTnI. In the validation data set, each biomarker was dichotomized into high and low according to the optimal cut point. An aggregated biomarker risk score (BRS) was created for each patient to indicate the number of elevated biomarkers, ranging from 0 to 6. Subsequently, a Cox model was applied to evaluate the prediction of the BRS. Interactions between the continuous BRS and patient age, gender, DM, CKD, history of heart failure, and use of statins were examined.
Using the entire cohort, we attempted to estimate 5-year risk of cardiovascular death/MI. All the clinical variables listed in Table 1 and the 6 biomarkers were considered in the development of a prediction model using Cox regression. Final models were selected based on the smallest Akaike information criterion values. Interactions between gender and race and biomarkers were considered. Internal validation of the final model for the proposed risk score was performed using 100 bootstrap samples. For each of the 100 bootstrap samples, we obtained the parameter estimates for the final model and then estimated the risk of cardiovascular death/MI using the observed data set.
Table 1.
Baseline demographics, risk factors, events, and biomarker profiles
| Baseline Characteristics | All Patients (n=3115) | Training Set (n=1557) | Validation Set (n=1558) |
|---|---|---|---|
|
| |||
| Age (years) | 62.8 (11.6) | 62.7 (11.6) | 63.0 (11.5) |
| Male | 2013 (64.6%) | 995 (63.9%) | 1018 (65.3%) |
| Black Race | 513 (16.5%) | 242 (15.5%) | 271 (17.4%) |
| BMI, kg/m2 | 29.7 (6.0) | 29.8 (6.1) | 29.6 (6.0) |
| Current or Former Smoker | 2040 (65.5%) | 1003 (64.4%) | 1037 (66.6%) |
| Obstructive CAD* | 1770 (56.8%) | 894 (57.4%) | 876 (56.2%) |
| Non-obstructive CAD* | 431 (13.8%) | 221 (14.2%) | 210 (13.5%) |
| Gensini Score | 8 [0–34] | 8 [0–32] | 8 [0–35] |
| Angiogram outcomes | |||
| Medical management | 1584 (59.8%) | 789 (59.5%) | 795 (60.1%) |
| Stent | 445 (16.8%) | 226 (17.0%) | 219 (16.6%) |
| Angioplasty | 404 (15.3%) | 198 (14.9%) | 206 (15.6%) |
| Surgery | 195 (7.4%) | 105 (7.9%) | 90 (6.8%) |
| Other | 21 (0.8%) | 8 (0.6%) | 13 (1.0%) |
| Seattle angina questionnaire score | |||
| None over the past 4 weeks | 1295 (46.5%) | 648 (46.9%) | 647 (46.1%) |
| <1 time weekly | 352 (12.6%) | 182 (13.2%) | 170 (12.1%) |
| 1–2 times per week | 312 (11.2%) | 164 (11.9%) | 148 (10.5%) |
| ≥ 3 times per week | 344 (12.3%) | 179 (13.0%) | 165 (11.8%) |
| 1–3 times per day | 318 (11.4%) | 137 (9.9%) | 181 (12.9%) |
| ≥ 4 times per day | 165 (5.9%) | 72 (5.2%) | 93 (6.6%) |
| History of MI | 702 (22.6%) | 340 (21.9%) | 362 (23.3%) |
| History of heart failure | 932 (29.9%) | 451 (29.0%) | 481 (30.9%) |
| History of PCI | 1339 (43.0%) | 661 (42.5%) | 678 (43.5%) |
| History of CABG | 789 (25.3%) | 394 (25.3%) | 395 (25.4%) |
| History of PVD | 471 (15.1%) | 232 (14.9%) | 239 (15.3%) |
| History of atrial fibrillation | 213 (6.8%) | 107 (6.9%) | 106 (6.8%) |
| Hyperlipidemia | 2178 (70.1%) | 1074 (69.3%) | 1104 (70.9%) |
| Hypertension | 2136 (72.5%) | 1119 (72.2%) | 1124 (72.4%) |
| Diabetes | 919 (31.2%) | 494 (31.9%) | 479 (30.8%) |
| LV Ejection Fraction, % | 54.4 (11.4) | 54.6 (11.7) | 54.2 (11.2) |
| eGFR, mL/min/1.73 m2 | 73.5 (22.0) | 72.1 (22.2) | 74.0 (22.2) |
| Total Cholesterol, mg/dL | 164 [139–194] | 163 [139–194] | 165 [140–195] |
| LDL-C, mg/dL | 91 [71–117] | 91 [71–117] | 92 [71–117] |
| HDL-C, mg/dL | 40 [33–48] | 40 [33–48] | 40 [34–49] |
| Triglycerides, mg/dL | 127 [85–189] | 129 [84–193] | 125 [87–184] |
| Medications | |||
| ACEi/ARB Use | 1804 (57.9%) | 908 (58.3%) | 896 (57.5%) |
| Beta-blocker Use | 1940 (62.3%) | 977 (62.8%) | 963 (61.8%) |
| Statin Use | 2217 (71.2%) | 1114 (71.6%) | 1103 (70.8%) |
| Aspirin Use | 2334 (74.9%) | 1181 (75.9%) | 1153 (74.0%) |
| Plavix Use | 1421 (45.6%) | 716 (46.0%) | 705 (45.3%) |
| Events at 5 years | |||
| All-cause Death | 440 (14.1%) | 219 (14.1%) | 221 (14.2%) |
| Cardiovascular Death | 263 (8.4%) | 124 (8.0%) | 139 (8.9%) |
| Myocardial Infarction | 104 (3.3%) | 42 (2.7%) | 62 (4.0%) |
| MACE | 672 (21.6%) | 321 (20.6%) | 351 (22.5%) |
| Biomarkers | |||
| Hs-CRP (mg/L) | 2.7 [1.2–6.6] | 2.7 [1.1–6.3] | 2.7 [1.2–6.8] |
| HSP-70 (>0)† | 19.9% [0–8120] | 19.8% [0–6743] | 20.0% [0–8120] |
| FDP (ug/mL) | 0.54 [0.36–0.82] | 0.53 [0.36–0.82] | 0.54 [0.36–0.81] |
| suPAR (pg/mL) | 3016 [2363–3973] | 3033 [2383–4044] | 2989 [2351–3889] |
| Hs-cTnI (ng/mL) | 4.8 [2.7–10.9] | 5.0 [2.8–11.4] | 4.6 [2.6–10.2] |
| BNP (mg/dL) | 62.5 [24.9–144.0] | 66.4 [26.4–147.2] | 58.4 [23.4–138.0] |
Mean (standard deviation), median [interquartile range], and frequency count (percentage) are reported.
Obstructive CAD = ≥50% stenosis in at least one major epicardial vessel. Non-obstructive CAD = <10% stenosis
Percentage of non-zeros and range for HSP-70 are reported.
CABG = coronary artery bypass graft; CAD = coronary artery disease; eGFR = estimated glomerular filtration rate; FDP = fibrin degradation products; hs-CRP = high-sensitivity C-reactive protein; hs-cTnI = high-sensitivity troponin I; HSP-70 = heat shock protein 70; MACE = major adverse cardiovascular events; MI = myocardial infarction; PCI = percutaneous coronary intervention; PVD = peripheral vascular disease; suPAR = soluble urokinase plasminogen activator receptor.
Discrimination testing was performed using the c-statistic, continuous net reclassification improvement (NRI), and integrated discrimination improvement metrics. p <0.05 were considered statistically significant. Statistical analyses were performed with SAS (version 9.3; SAS Institute, Cary, North Carolina) and R 3.4.1 (The R Foundation for Statistical Computing, Vienna, Austria). Details of analyses appear in the Supplemental Materials.
Results
A total of 3,115 patients who had complete biomarker data (hs-CRP, HSP-70, FDP, suPAR, hs-cTnI, and BNP) were included in the analysis, after excluding patients with acute MI (hs-cTnI > 12 ng/ml) at enrollment and those with limited study follow-up time (<30 days). Their mean age was 62.8 years (±11.6 years), 17% were Black, 35% were female, 63% with obstructive CAD (>50% stenosis), and 31% with DM (Table 1). Overall, during a median follow-up of 5 years (interquartile range 3.3 to 5 years), there were 440 (14.1%) all-cause deaths, 263 (8.4%) cardiovascular deaths, 104 (3.3%) MIs, and 672 (21.6%) major adverse cardiovascular events. Spearman correlation coefficients between individual biomarkers were weakly significant, Supplementary Table 1.
Supplementary Figure 1 shows the Kaplan–Meier curves for all-cause death and for cardiovascular death/MI in subjects with significant differences in survival between those with high compared with low levels of individual biomarkers stratified by the median value in the training set. Table 2 lists the estimated hazard ratio (HR) and subdistribution HR (sHR) corresponding to a 10th to 90th percentile increase for each biomarker after adjustment for previously mentioned covariates. The results indicate that each biomarker was significantly and independently (while other biomarkers were present in the model) associated with increased risk of all–cause death in the adjusted Cox model. In the Fine-Gray models for the analysis of cardiovascular death/MI, FDP (sHR 2.25, 95% CI 1.29 to 3.95), suPAR (sHR 3.05, 95% CI 1.48 to 6.31), and BNP (HR 2.43, 95% CI 1.20 to 4.93) remained significant.
Table 2.
Predictive ability of individual biomarkers in the training set
| Outcome | Biomarker (per increase from 10th to 90th percentile) | HR/sHR (95% confidence limits) | P-value |
|---|---|---|---|
|
| |||
| Death | Hs-CRP | 1.64 (1.06,2.52) | 0.025 |
| HSP-70 | 1.30 (0.97,1.74) | 0.078 | |
| FDP | 2.23 (1.40,3.56) | 0.0008 | |
| SuPAR | 4.83 (2.76,8.45) | <.0001 | |
| Hs-cTnI | 1.97 (1.18,3.30) | 0.010 | |
| BNP | 1.98 (1.19,3.30) | 0.009 | |
| CV Death/MI | Hs-CRP | 1.27 (0.72,2.23) | 0.404 |
| HSP-70 | 1.35 (0.92,1.98) | 0.123 | |
| FDP | 2.25 (1.29,3.95) | 0.005 | |
| SuPAR | 3.05 (1.48,6.31) | 0.003 | |
| Hs-cTnI | 1.72 (0.88,3.36) | 0.114 | |
| BNP | 2.43 (1.20,4.93) | 0.014 | |
Model adjusted for age, gender, race (black), BMI, history of smoking, hypertension, diabetes, hyperlipidemia, history of MI, history of heart failure, Gensini score, eGFR, use of ACEi/ARBs, beta-blocker use and statins with all 6 biomarkers simultaneously in the model.
BNP = B-type natriuretic peptide; CV = cardiovascular; FDP = fibrin degradation products; HR = hazard ratio; Hs-CRP = high-sensitivity C-reactive protein; hs-cTnI = high-sensitivity troponin I; HSP-70 = heat shock protein 70; MI = myocardial infarction; sHR = subdistribution hazard ratio; suPAR = soluble urokinase plasminogen activator receptor.
The biomarker risk prediction model included the 6 biomarkers (as continuous variables) and clinical variables (LVEF, Gensini score, and history of hypertension, hyperlipidemia, CKD, and heart failure) that reached statistical significance in Cox regression models. Table 3 lists the coefficient estimates of this final model and the corresponding HR estimates that were used to generate the predicted risk of cardiovascular death/MI for each patient. Internal validation of the final model was performed with 100 bootstrap samples. Adequate goodness-of-fit for the final risk prediction model (p = 0.46) was observed with the Hosmer-Lemeshow test. To assess whether the predicted risk aligns with the observed risk (model calibration), we split the data into deciles based on the estimated risk and showed that the model performed well as compared with clinical variables alone (Figure 1).
Table 3.
Biomarker risk prediction model for cardiovascular death or MI in 5 years
| Coefficient Estimate | Hazard Ratio (95% Confidence Interval) | P-value | |
|---|---|---|---|
|
| |||
| Ln(hs-CRP) | 0.07 | 1.08 (1.01, 1.15) | 0.020 |
| HSP-70 >11.3 ng/mL | 0.43 | 1.54 (1.20, 1.96) | 0.001 |
| Ln(FDP) | 0.06 | 1.06 (1.00, 1.13) | 0.044 |
| Ln(suPAR) | 0.57 | 1.78 (1.45, 2.17) | <.0001 |
| Ln(hs-cTnI) | 0.08 | 1.08 (1.02, 1.14) | 0.014 |
| Ln(BNP) | 0.16 | 1.17 (1.10, 1.26) | <.0001 |
| History of heart failure | 0.31 | 1.36 (1.04, 1.78) | 0.024 |
| LVEF (per 10% less) | −0.12 | 0.89 (0.81, 0.98) | 0.015 |
| CKD | 0.27 | 1.30 (1.01, 1.68) | 0.039 |
| Hypertension | 0.22 | 1.24 (0.93, 1.67) | 0.150 |
| Hyperlipidemia | 0.26 | 1.30 (0.99, 1.69) | 0.067 |
All the covariates listed in Table 1 and the six continuous biomarkers (natural log (Ln) transformed except HSP-70) were considered in the development of a prediction model using Cox regression. Stepwise selection based on AIC was used to create the final set of predictor variables. Baseline survival probability is 0.99, which is the calculated likelihood of survival by setting all the covariate values as 0.
Figure 1.
Risk prediction utilizing biomarkers in the entire cohort. Cohort split into deciles based on estimated risk. Gray bar shows observed risk, blue bar shows risk predicted by clinical variables alone, and pink bar shows risk predicted by biomarker risk prediction model utilizing both biomarkers as continuous variables and clinical variables. CV = cardiovascular.
Based on the maximum likelihood approach, the optimal cut points derived from the training set of 1,557 subjects for hs-CRP, HSP-70, FDP, suPAR, hs-TnI, and BNP were 3 mg/L, 11.3 ng/ml, 0.82 μg/ml, 2,858 mg/100 ml (men) and 3,908 mg/100 ml (women), 5.87 ng/ml (men) and 6.66 ng/ml (women), and 261 mg/100 ml, respectively.
In the validation set of 1,558 subjects, an aggregated BRS was created for each patient to indicate the number of elevated biomarkers based on the optimal cut points; 18%, 28%, 24%, 16%, 10%, 4%, and 1% of the patients had a BRS of 0, 1, 2, 3, 4, 5, and 6, respectively, Table 4. Subjects with higher BRS were more likely to be older and Black, with a greater prevalence of hypertension, DM, and hyperlipidemia, and a history of heart failure, MI, and revascularization, higher Gensini score, and lower eGFR and LVEF, and higher utilization of prescribing medications (including clopidogrel, β blockers, and ACEi/ARB).
Table 4.
Baseline characteristics stratified by number of elevated biomarkers (biomarker risk score) in the validation set
| Baseline characteristics | Number of Elevated Biomarkers* |
|||||||
|---|---|---|---|---|---|---|---|---|
| 0 (n=276) | 1 (n=438) | 2 (n=373) | 3 (n=248) | 4 (n=152) | 5 (n=55) | 6 (n=16) | P-value | |
|
| ||||||||
| Demographics and Risk Factors | ||||||||
| Age (years) | 60.0 (11.3) | 62.0 (11.2) | 63.0 (11.4) | 64.2 (11.6) | 65.3 (11.7) | 70.2 (9.8) | 71.5 (8.9) | <0.001 |
| Male | 178 (64.5%) | 271 (61.9%) | 250 (67.0%) | 170 (68.6%) | 105 (69.1%) | 33 (60.0%) | 11 (68.6%) | 0.46 |
| Black race | 47 (17.0%) | 66 (15.1%) | 61 (16.4%) | 50 (20.2%) | 33 (21.7%) | 8 (14.6%) | 6 (37.5%) | 0.12 |
| BMI, kg/m2 | 28.5 (5.0) | 29.8 (5.5) | 30.1 (6.2) | 30.0 (6.5) | 29.7 (6.8) | 28.0 (6.4) | 28.6 (4.6) | <0.001 |
| Current or former smoking | 166 (60.1%) | 300 (68.5%) | 251 (67.3%) | 180 (72.6%) | 92 (60.5%) | 34 (61.8%) | 14 (87.5%) | 0.01 |
| Obstructive CAD* | 130 (47.1%) | 238 (54.3%) | 213 (57.1%) | 153 (69.9%) | 98 (64.5%) | 34 (61.8%) | 10 (62.5%) | <0.001 |
| Non-obstructive CAD* | 60 (25.5%) | 67 (17.8%) | 41 (13.9%) | 31 (15.1%) | 8 (6.2%) | 3 (6.4%) | 0 (0%) | <0.001 |
| Gensini score | 5 [0–21] | 7 [0–26] | 10 [0–48] | 10 [0–34] | 10 [0–38] | 20 [0–52] | 9 [4–82] | 0.17 |
| Angiogram outcomes | 0.40 | |||||||
| Medical management | 159 (67.1%) | 242 (64.2%) | 176 (56.2%) | 109 (53.2%) | 78 (57.8%) | 25 (54.3%) | 6 (60.0%) | |
| Stent | 32 (13.5%) | 57 (15.1%) | 54 (17.3%) | 42 (20.5%) | 24 (17.8%) | 8 (17.4%) | 2 (20.0%) | |
| Angioplasty | 31 (13.1%) | 49 (13.0%) | 60 (19.2%) | 33 (16.1%) | 21 (15.6%) | 10 (21.7%) | 2 (20.0%) | |
| Surgery | 12 (5.1%) | 24 (6.4%) | 19 (6.1%) | 20 (9.8%) | 12 (8.9%) | 3 (6.5%) | 0 (0%) | |
| Other | 3 (1.3%) | 5 (1.3%) | 4 (1.3%) | 1 (0.5%) | 0 (0%) | 0 (0%) | 0 (0%) | |
| Seattle angina scoring | ||||||||
| None over the past 4 weeks | 219 (45.2%) | 343 (43.8%)† | 303 (45.5%) | 211 (47.6%) | 133 (49.8%) | 64 (61.0%)† | 22 (61.1%) | 0.019 |
| <1 time weekly | 56 (11.5%) | 115 (14.7%) | 77 (11.6%) | 58 (13.1%) | 34 (12.7%) | 9 (8.6%) | 3 (8.3%) | |
| 1–2 times per week | 60 (12.4%) | 78 (9.9%) | 86 (12.9%) | 47 (10.6%) | 26 (9.7%) | 10 (9.5%) | 5 (13.9%) | |
| ≥ 3 times per week | 72 (14.8%) | 99 (12.6%) | 81 (12.2%) | 52 (11.7%) | 29 (10.9%) | 9 (8.6%) | 2 (5.6%) | |
| 1–3 times per day | 57 (11.8%) | 98 (12.5%) | 80 (12.0%) | 46 (10.4%) | 27 (10.1%) | 8 (7.6%) | 2 (5.6%) | |
| ≥ 4 times per day | 21 (4.3%) | 51 (6.5%) | 39 (5.9%) | 29 (6.5%) | 18 (6.7%) | 5 (4.8%) | 2 (5.6%) | |
| History of MI | 43 (15.6%) | 84 (19.2%) | 104 (28.0%) | 62 (25.0%) | 48 (31.6%) | 12 (22.2%) | 9 (56.3%) | <0.001 |
| History of heart failure | 39 (14.1%) | 98 (22.3%) | 116 (31.0%) | 102 (41.1%) | 82 (53.9%) | 33 (60.0%) | 10 (62.5%) | <0.001 |
| History of CABG | 45 (16.3%) | 89 (20.3%) | 108 (29.0%) | 72 (29.0%) | 53 (34.9%) | 18 (32.7%) | 10 (62.5%) | <0.001 |
| History of PCI | 105 (38.0%) | 175 (40.0%) | 171 (45.8%) | 119 (48.0%) | 73 (48.0%) | 25 (45.5%) | 10 (62.5%) | 0.06 |
| History of PVD | 26 (9.5%) | 50 (11.4%) | 60 (16.1%) | 45 (18.1%) | 33 (21.7%) | 15 (27.3%) | 10 (62.5%) | <0.001 |
| History of atrial fibrillation | 14 (5.1%) | 28 (6.4%) | 26 (7.0%) | 15 (6.0%) | 16 (10.5%) | 5 (9.1%) | 2 (12.5%) | 0.40 |
| Diabetes | 51 (18.5%) | 99 (22.6%) | 139 (37.4%) | 89 (35.9%) | 64 (42.1%) | 26 (47.3%) | 11 (68.8%) | <0.001 |
| Hypertension | 169 (61.5%) | 300 (68.8%) | 284 (76.3%) | 194 (78.5%) | 120 (79.0%) | 44 (80.0%) | 13 (81.3%) | <0.001 |
| Hyperlipidemia | 177 (64.4%) | 309 (70.6%) | 284 (76.2%) | 173 (69.8%) | 107 (70.4%) | 41 (74.6%) | 13 (81.3%) | 0.06 |
| LVEF, % | 57.5 (8.3) | 56.3 (9.4) | 54.6 (10.1) | 52.2 (12.3) | 48.0 (14.1) | 48.3 (13.4) | 41.2 (19.2) | <0.001 |
| eGFR, mL/min/1.73 m2 | 82.2 (18.0) | 78.5 (18.8) | 75.8 (19.9) | 68.4 (22.1) | 62.7 (26.3) | 49.2 (26.8) | 44.1 (18.8) | <0.001 |
| Total cholesterol, mg/dL | 167 [142–193] | 168 [142–203] | 162 [140–190] | 164 [139–192] | 168 [140–195] | 150 [126–165] | 141 [125–153] | 0.001 |
| LDL-C, mg/dL | 95 [72–116] | 94 [72–121] | 89 [70–113] | 97 [74–118] | 90 [71–122] | 75 [60–103] | 72 [61–97] | 0.016 |
| HDL-C, mg/dL | 42 [35–53] | 41 [34–51] | 40 [34–47] | 39 [32–45] | 40 [33–50] | 39 [32–46] | 40 [32–46] | 0.003 |
| Triglycerides, mg/dL | 117 [78–173] | 129 [89–185] | 138 [92–190] | 125 [88–191] | 120 [86–182] | 129 [75–168] | 106 [74–132] | 0.06 |
| Medications | ||||||||
| Statin use | 193 (69.9%) | 300 (68.5%) | 269 (71.1%) | 180 (72.6%) | 109 (71.7%) | 40 (72.7%) | 12 (75.0%) | 0.90 |
| Aspirin use | 190 (68.8%) | 317 (72.4%) | 275 (73.7%) | 193 (77.8%) | 120 (79.0%) | 44 (80.0%) | 14 (87.5%) | 0.10 |
| Clopidogrel use | 93 (33.7%) | 186 (42.5%) | 181 (48.5%) | 126 (50.8%) | 78 (51.3%) | 31 (56.4%) | 10 (62.5%) | <0.001 |
| Beta-blocker use | 146 (52.9%) | 255 (58.2%) | 226 (60.6%) | 166 (66.9%) | 117 (77.0%) | 41 (74.6%) | 12 (75.0%) | <0.001 |
| ACEi/ARB use | 131 (47.5%) | 245 (56.0%) | 229 (61.4%) | 155 (62.5%) | 92 (60.5%) | 31 (56.4%) | 13 (81.3%) | 0.002 |
| Events at 5 years | ||||||||
| All-cause Death | 8 (2.9%) | 29 (6.6%) | 44 (11.8%) | 48 (19.4%) | 52 (34.2%) | 29 (52.7%) | 11 (68.8%) | <0.001 |
| Cardiovascular Death | 4 (1.5%) | 12 (2.7%) | 30 (8.0%) | 29 (11.7%) | 33 (21.7%) | 22 (40.0%) | 9 (56.3%) | <0.001 |
| Myocardial Infarction | 3 (1.1%) | 12 (2.7%) | 16 (4.3%) | 13 (5.2%) | 10 (6.6%) | 4 (7.3%) | 4 (25.0%) | <0.001 |
| MACE | 34 (12.3%) | 67 (15.3%) | 92 (24.7%) | 64 (25.8%) | 57 (37.5%) | 25 (45.5%) | 12 (75.0%) | <0.001 |
Mean (standard deviation), median [interquartile range], and frequency count (percentage) are reported.
Biomarkers were dichotomized as high or low based on optimal cut points and an aggregate biomarker risk score was created to indicate the number of elevated biomarkers, ranging from 0 to 6. Optimal cut points for hs-CRP, HSP-70, FDP, suPAR, and hs-cTnI were 3 mg/L, 11.3 ng/mL, 0.82 ug/mL, 2858 mg/dL (men) and 3908 mg/dL (women), 5.87 ng/mL (men) and 6 ng/mL (women), 261 mg/dL, identified using the training set.
Obstructive CAD = ≥50% stenosis in at least one major epicardial vessel. Non-obstructive CAD= <10% stenosis.
ACEi = angiotensin-converting enzyme inhibitor; ARB = angiotensin II receptor blocker; CABG = coronary artery bypass graft surgery; CAD = coronary artery disease; eGFR = estimated glomerular filtration rate; FDP = fibrin degradation products; HDL-C = high density lipoprotein cholesterol; hs-CRP = high-sensitivity C-reactive protein; hs-cTnI = high-sensitivity troponin I; HSP-70 = heat shock protein 70; LDL-C = low density lipoprotein cholesterol; LVEF = left ventricular ejection fraction; MACE = major adverse cardiovascular events; MI = myocardial infarction; PCI = percutaneous coronary intervention; PVD = peripheral vascular disease; suPAR = soluble urokinase plasminogen activator receptor.
The Kaplan–Meier survival curves, cumulative incidence functions, and 5-year event rates stratified by BRS in the validation cohort are shown in Figure 2. As the BRS increased, the event risk increased substantially (<10% with a BRS of 0 or 1 and nearly ≥50% with a BRS of 5 or 6). After adjusting for age, gender, race, smoking history, BMI, LVEF, hypertension, DM, hyperlipidemia, eGFR, history of MI and heart failure, Gensini score, and medications, the HRs for those with 1, 2, 3, 4 and 5 of 6 elevated biomarkers compared with those with BRS 0 are listed in Table 5. Compared with those with a BRS of 0, the HR of all-cause death was 12.6 (95% CI 5.6 to 28.3), and the sHR of cardiovascular death/MI was 11.6 (95% CI 4.6 to 29.3) for those with a BRS of 5 or 6. Also, each 1 unit increase in the BRS was associated with a HR of 1.62 (95% CI 1.45 to 1.80) in all-cause death and a sHR of 1.52 (95% CI 1.35 to 1.71) in cardiovascular death/MI (Table 5). Sensitivity analyses investigating the association between the BRS and patient outcomes for subgroups determined by age, gender, DM, CKD, heart failure, and use of statins are illustrated in Supplementary Figure 2. There were no significant interactions between the BRS and the covariates tested (p >0.05).
Figure 2.
(A) Kaplan–Meier curves for all-cause death. (B) Cumulative incidence functions for CV death or MI with the competing risk of non-CV death. (C) Five-year Event Rates of All-cause Death and CV Death or MI stratified by the BRS. All results are derived from the validation set. The results of log-rank test show significant different survivals in different BRS groups. For illustration purposes, the event rates were calculated using the baseline risk set without accounting for censoring. CV = cardiovascular.
Table 5.
Association between the aggregate biomarker risk score (BRS) and adverse outcomes in the validation set
| Outcome | Effect | HR/sHR (95% confidence limits) | P-value |
|---|---|---|---|
|
| |||
| Death | BRS=1 vs 0 | 2.05 (0.93, 4.49) | 0.074 |
| BRS=2 vs 0 | 3.33 (1.56, 7.12) | 0.002 | |
| BRS=3 vs 0 | 4.90 (2.28, 10.50) | <0.0001 | |
| BRS=4 vs. 0 | 8.84 (4.09, 19.10) | <0.0001 | |
| BRS≥5 vs 0 | 12.60 (5.57, 28.30) | <0.0001 | |
| 1 unit increase in BRS | 1.62 (1.45, 1.80) | <0.001 | |
| CV Death/MI | BRS=1 vs 0 | 1.99 (0.80, 4.94) | 0.14 |
| BRS=2 vs 0 | 3.70 (1.57, 8.71) | 0.003 | |
| BRS=3 vs 0 | 4.28 (1.77, 10.40) | 0.001 | |
| BRS=4 vs. 0 | 6.24 (2.59, 15.10) | <0.0001 | |
| BRS≥5 vs 0 | 11.60 (4.57, 29.30) | <0.0001 | |
| 1 unit increase in BRS | 1.52 (1.35, 1.71) | <0.0001 | |
Estimated hazard ratios (HR) and subdistribution hazard ratios (sHR) are presented, adjusted for age, gender, race (black), BMI, history of smoking, hypertension, diabetes, hyperlipidemia, history of MI, history of heart failure, Gensini score, eGFR, use of ACEi/ARBs, beta-blocker use, and statins.
C-statistics were used to evaluate the prediction performance of the models for cardiovascular death/MI in 5 years with (1) clinical covariates, (2) clinical covariates and biomarkers as continuous variables, (3) clinical covariates and BRS, and (4) BRS alone. Addition of the biomarkers as continuous variables to the clinical variables improved the c-statistic by Δ 6.4% (95% CI 3.9 to 8.8) to 76.8%, compared with clinical variables alone, indicating good discrimination, Supplemental Figure 3, Table 6. Clinical variables with the BRS also had good discrimination with a c-statistic of 75.9% and an improvement of the c-statistic (Δ 5.4%, 95% CI 3.2 to 7.6) but were slightly inferior to the biomarkers used as continuous values. Importantly, the BRS alone, without any clinical or demographic information, outperformed the clinical variables with a c-statistic of 72.9%, as compared with 70.5% with clinical variables alone. Similarly, clinical variables plus the BRS were associated with significant improvement in risk reclassification metrics with an NRI of 25.7% (95% CI 19.9% to 30.6%), which improved further when biomarkers were used as continuous variables with an NRI improvement of 31.1% (95% CI 23.5% to 37.4%).
Table 6.
Discrimination statistics to evaluate the prediction performance of the biomarkers as continuous variables and BRS for cardiovascular death or MI
| C-statistic (%) | ∆C-statistic* (%) | Net reclassification improvement* (%) | Integrated discrimination improvement* (%) | |
|---|---|---|---|---|
|
| ||||
| (1) Clinical variables | 70.5 (66.8,74.1) | |||
| (2) Clinical variables + Biomarkers | 76.8 (74.4,79.2) | 6.4 (3.9,8.8) | 31.1 (23.5,37.4) | 6.5 (4.8,9.4) |
| (3) Clinical variables + BRS | 75.9 (72.8,78.9) | 5.4 (3.2,7.6) | 25.7 (19.9,30.6) | 5 (3.1,7.2) |
| (4) BRS alone | 72.9 (70.3,75.5) | 2.5 (−0.7,5.7) | 8.4 (−3.5,17) | 2.6 (−1.2,5.1) |
Compared to Model (1)
(2) Biomarkers as continuous variables.
Numbers in parentheses indicate 95% confidence limits.
Discussion
A number of pathophysiologic processes contribute to cardiovascular events and death. Current clinically utilized risk prediction methods focus on clinical variables that have been associated with high risk of adverse outcomes. Our risk prediction model that uses 6 clinical variables and 6 biomarkers provides a significant improvement on current clinical risk stratification methods, particularly in the low-risk and high-risk categories, where clinical covariates alone overestimate or underestimate risk, respectively.
The biomarkers used in the risk prediction model were each associated with increased risk of incident adverse events, and importantly their effects appeared to be additive. High levels of circulating hs-cTnI have been associated with prevalent CAD and with adverse events in patients with stable CAD or the general and older population.19–23 Within the Emory Cardiovascular Biobank cohort, higher hs-TnI levels are associated with the underlying burden of CAD, more rapid progression of CAD, and higher risk of allcause mortality and incident cardiovascular events.24 BNP and other natriuretic peptides are released from cardiac myocytes and their elevated levels are associated with higher event rates in the general population,25 in those with heart failure, and in survivors of acute MI.14 Importantly, in the Screening to Prevent Heart Failure (STOP-HF) trial, BNP-guided intervention reduced the rate of hospitalizations for heart failure.26
Similarly, hs-CRP levels independently predict incident cardiovascular disease (CVD) and adverse cardiovascular events.27,28 Importantly, hs-CRP levels are modifiable with statin and other therapies, and its lowering is independently correlated with the reduced rate of plaque progression and CVD events.29 HSPs are abundant intracellular proteins that aid in the cellular response to acute stress and are involved in protein folding and transport.30 HSP-70 is one of the more extensively studied HSPs.5,31 FDPs are measures of ongoing fibrin/fibrinogen degradation including d-dimer and fragments D and E and any additional intermediate products from fibrin degradation. Increased plasma FDP and d-dimer levels predict incident cardiovascular events.32 suPAR is an emerging marker of coronary and renal disease. The urokinase plasminogen activator (uPA) is a serine protease produced by smooth muscle cells, vascular endothelial cells, macrophages, monocytes, and fibroblasts, and when bound to its receptor (uPAR), leads to the generation of plasmin.33 uPAR is involved in several functions including migration, adhesion, fibrinolysis, and cell proliferation. Plasma suPAR reflects cellular shedding of suPAR, which is induced during inflammation; shedding appears to be free of circadian changes and is relatively stable during periods of acute stress. SuPAR levels independently predict incident CVD in healthy populations,34 and the severity of CAD and risk of incident events in patients with CAD.4 Additionally, suPAR levels are predictive of acute kidney injury,35 and incident CKD and renal function decrease.8
In addition to enhancement in prognostication, as shown in the BRS previously mentioned, several studies have examined whether intensification of medical therapy favorably impacts biomarkers and whether this improvement correlates with better outcomes. In the IMProved Reduction of Outcomes: Vytorin Efficacy International (IMPROVE-IT) trial, patients with higher levels of hsTn-T, NT proBNP, growth-differentiation Factor-15, and hs-CRP had a greater risk reduction in biomarker levels with therapy, suggesting that a biomarker-based strategy identifies a higher-risk group of patients with a correspondingly high absolute benefit from therapy intensification.36 Similarly, improvements in hs-CRP, FDPs, and HSP-70 levels in the Bypass angioplasty revascularization Investigation 2 Diabetes (BARI 2D) cohort after a course of intensive medical therapy was independently associated with improved outcomes.37
Incorporation of biomarkers for risk prediction, as we have demonstrated, surpasses predictive capabilities of clinical measures. The biomarkers selected represent critical pathways in the pathogenesis of atherosclerosis and heart failure, and their activation represents a vulnerable state for progression of disease.38 Without taking into account any demographic or clinical information, the BRS score was more predictive than the clinical risk factors of patients. Biomarker profiling identifies residual risk in patients with CAD and otherwise well-controlled clinical risk factors who continue to be at risk for secondary events. This may have important therapeutic implications, such as use of more aggressive treatment options and tailored behavioral modification counseling. Conversely, identification of patients at low risk may prevent unnecessary testing and therapy escalation in this population.
Our study has several strengths. We enrolled gender-diverse and racially-diverse subjects with CAD with a range of LV functions, reflecting a high-risk population typical of those who underwent cardiac catheterization. Each biomarker evaluation was performed by the same laboratory personnel, minimizing variability. We also replicated our findings derived from the training set in a validation set. The c-statistic, net reclassification index, and integrated discrimination improvements were calculated using survival models, which allows for better model discrimination and overall predictive ability. Limitations of our study include a one-time measurement of biomarkers that may not reflect fluctuations in their levels over time, a single healthcare system experience, and lack of data on medication adherence. Our results need to be further validated and should not be generalized to a population without known CAD. Future studies are needed to assess cost-effectiveness of using this BRS in routine clinical practice and to prospectively validate the use of BRS-guided model for therapeutic efficacy.
In conclusion, the model utilizing biomarkers representing inflammation, coagulation, cell stress, myocardial damage, and immune pathways and high-risk clinical variables independently predicts risk of incident MI and death and significantly improves risk reclassification beyond previous risk prediction models. Whether more intensive management of subjects with a high BRS would decrease the risk score and future risk remains to be studied.
A cardiovascular risk assessment model incorporating the BRS (a score based on blood concentrations of 6 pathophysiology-specific biomarkers) and clinical variables—including a history of heart failure, LVEF, hypertension, hyperlipidemia, CKD, and the Gensini score—improves prediction of incident death and MI beyond the use of clinical covariates alone in patients with stable CAD.
Supplementary Material
Acknowledgments
This work was supported partially by Abbott Laboratories, North Chicago, Illinois. Dr. Quyyumi is supported by National Institutes of Health, Bethesda, Maryland, grants 4R61HL138657–04, U54AG062334–01, 1P30DK111024–03S1, 15SFCRN23910003, 5P01HL086773–09, 1R01HL141205–01, 5P01HL101398–05, 1P20HL113451–04, 3RF1AG051633–01S2, and American Heart Association, Chicago, Illinois, grant 15SFCRN23910003. Drs. Dhindsa, Sandesara, Mehta, and Tahhan have been supported by the Abraham J. & Phyllis Katz Foundation, Atlanta, Georgia. Dr. Mehta is supported by American Heart Association grant 19POST34400057. Dr. Desai has been supported by T32 HL130025.
Footnotes
Declaration of Competing Interest
The authors have no conflicts of interest to declare.
Supplementary materials
Supplementary material associated with this article can be found in the online version at https://doi.org/10.1016/j.amjcard.2023.06.115.
References
- 1.Bhatt DL, Eagle KA, Ohman EM, Hirsch AT, Goto S, Mahoney EM, Wilson PW, Alberts MJ, D’Agostino R, Liau CS, Mas JL, Röther J, Smith SC Jr, Salette G, Contant CF, Massaro JM, Steg PG, REACH Registry Investigators. Comparative determinants of 4-year cardiovascular event rates in stable outpatients at risk of or with atherothrombosis. JAMA 2010;304:1350–1357. [DOI] [PubMed] [Google Scholar]
- 2.Sabatine MS, Morrow DA, de Lemos JA, Gibson CM, Murphy SA, Rifai N, McCabe C, Antman EM, Cannon CP, Braunwald E. Multimarker approach to risk stratification in non-ST elevation acute coronary syndromes: simultaneous assessment of troponin I, C-reactive protein, and B-type natriuretic peptide. Circulation 2002;105:1760–1763. [DOI] [PubMed] [Google Scholar]
- 3.Vittinghoff E, Shlipak MG, Varosy PD, Furberg CD, Ireland CC, Khan SS, Blumenthal R, Barrett-Connor E, Hulley S. Heart and Estrogen/progestin Replacement Study Research Group. Risk factors and secondary prevention in women with heart disease: the Heart and estrogen/progestin Replacement Study. Ann Intern Med 2003;138:81–89. [DOI] [PubMed] [Google Scholar]
- 4.Eapen DJ, Manocha P, Ghasemzadeh N, Patel RS, Al Kassem H, Hammadah M, Veledar E, Le NA, Pielak T, Thorball CW, Velegraki A, Kremastinos DT, Lerakis S, Sperling L, Quyyumi AA. Soluble urokinase plasminogen activator receptor level is an independent predictor of the presence and severity of coronary artery disease and of future adverse events. J Am Heart Assoc 2014;3:e001118. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Eapen DJ, Manocha P, Patel RS, Hammadah M, Veledar E, Wassel C, Nanjundappa RA, Sikora S, Malayter D, Wilson PW, Sperling L, Quyyumi AA, Epstein SE. Aggregate risk score based on markers of inflammation, cell stress, and coagulation is an independent predictor of adverse cardiovascular outcomes. J Am Coll Cardiol 2013;62:329–337. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Ghasemzedah N, Hayek SS, Ko YA, Eapen DJ, Patel RS, Manocha P, Al Kassem H, Khayata M, Veledar E, Kremastinos D, Thorball CW, Pielak T, Sikora S, Zafari AM, Lerakis S, Sperling L, Vaccarino V, Epstein SE, Quyyumi AA. Pathway-specific aggregate biomarker risk score is associated with burden of coronary artery disease and predicts near-term risk of myocardial infarction and death. Circ Cardiovasc Qual Outcomes 2017;10:e001493. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Sabatine MS, Morrow DA, Jablonski KA, Rice MM, Warnica JW, Domanski MJ, Hsia J, Gersh BJ, Rifai N, Ridker PM, Pfeffer MA, Braunwald E, Investigators PEACE. Prognostic significance of the Centers for Disease Control/American Heart Association high-sensitivity C-reactive protein cut points for cardiovascular and other outcomes in patients with stable coronary artery disease. Circulation 2007;115:1528–1536. [DOI] [PubMed] [Google Scholar]
- 8.Hayek SS, Sever S, Ko YA, Trachtman H, Awad M, Wadhwani S, Altintas MM, Wei C, Hotton AL, French AL, Sperling LS, Lerakis S, Quyyumi AA, Reiser J. Soluble urokinase receptor and chronic kidney disease. N Engl J Med 2015;373:1916–1925. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.Pockley AG, Frostegård J. Heat shock proteins in cardiovascular disease and the prognostic value of heat shock protein related measurements. Heart 2005;91:1124–1126. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Kannel WB, Wolf PA, Castelli WP, D’Agostino RB. Fibrinogen and risk of cardiovascular disease. The Framingham study. JAMA 1987;258:1183–1186. [PubMed] [Google Scholar]
- 11.Westermann D, Neumann JT, Sörensen NA, Blankenberg S. High-sensitivity assays for troponin in patients with cardiac disease. Nat Rev Cardiol 2017;14:472–483. [DOI] [PubMed] [Google Scholar]
- 12.Wang TJ, Larson MG, Levy D, Benjamin EJ, Leip EP, Omland T, Wolf PA, Vasan RS. Plasma natriuretic peptide levels and the risk of cardiovascular events and death. N Engl J Med 2004;350:655–663. [DOI] [PubMed] [Google Scholar]
- 13.European Association for the Study of the Liver (EASL), European Association for the Study of Diabetes (EASD), European Association for the Study of Obesity (EASO). EASL-EASD-EASO clinical practice guidelines for the management of non-alcoholic fatty liver disease. J Hepatol 2016;64:1388–1402. [DOI] [PubMed] [Google Scholar]
- 14.Berger R, Huelsman M, Strecker K, Bojic A, Moser P, Stanek B, Pacher R. B-type natriuretic peptide predicts sudden death in patients with chronic heart failure. Circulation 2002;105:2392–2397. [DOI] [PubMed] [Google Scholar]
- 15.Tapanainen JM, Lindgren KS, Mäkikallio TH, Vuolteenaho O, Leppäluoto J, Huikuri HV. Natriuretic peptides as predictors of non-sudden and sudden cardiac death after acute myocardial infarction in the beta-blocking era. J Am Coll Cardiol 2004;43:757–763. [DOI] [PubMed] [Google Scholar]
- 16.Levey AS, Stevens LA, Schmid CH, Zhang YL, Castro AF 3rd, Feldman HI, Kusek JW, Eggers P, Van Lente F, Greene T, Coresh J. CKD-EPI (Chronic Kidney Disease Epidemiology Collaboration). A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–612. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Austen WG, Edwards JE, Frye RL, Gensini GG, Gott VL, Griffith LS, McGoon DC, Murphy ML, Roe BB. A reporting system on patients evaluated for coronary artery disease. Report of the Ad Hoc Committee for Grading of Coronary Artery Disease, council on cardiovascular surgery, American Heart Association. Circulation 1975;51:5–40. [DOI] [PubMed] [Google Scholar]
- 18.Sinning C, Lillpopp L, Appelbaum S, Ojeda F, Zeller T, Schnabel R, Lubos E, Jagodzinski A, Keller T, Munzel T, Bickel C, Blankenberg S. Angiographic score assessment improves cardiovascular risk prediction: the clinical value of SYNTAX and Gensini application. Clin Res Cardiol 2013;102:495–503. [DOI] [PubMed] [Google Scholar]
- 19.Everett BM, Brooks MM, Vlachos HEA, Chaitman BR, Frye RL, Bhatt DL, Group BARI 2D Study. Troponin and cardiac events in stable ischemic heart disease and diabetes. N Engl J Med 2015;373:610–620. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Neumann JT, Havulinna AS, Zeller T, Appelbaum S, Kunnas T, Nikkari S, Jousilahti P, Blankenberg S, Sydow K, Salomaa V. Comparison of three troponins as predictors of future cardiovascular events–prospective results from the FINRISK and BiomaCaRE studies. PLoS One 2014;9:e90063. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Sinning C, Keller T, Zeller T, Ojeda F, Schlüter M, Schnabel R, Lubos E, Bickel C, Lackner KJ, Diemert P, Munzel T, Blankenberg S, Wild PS. Gutenberg Health Study. Association of high-sensitivity assayed troponin I with cardiovascular phenotypes in the general population: the population-based Gutenberg health study. Clin Res Cardiol 2014;103:211–222. [DOI] [PubMed] [Google Scholar]
- 22.Blankenberg S, Salomaa V, Makarova N, Ojeda F, Wild P, Lackner KJ, Jørgensen T, Thorand B, Peters A, Nauck M, Petersmann A, Vartiainen E, Veronesi G, Brambilla P, Costanzo S, Iacoviello L, Linden G, Yarnell J, Patterson CC, Everett BM, Ridker PM, Kontto J, Schnabel RB, Koenig W, Kee F, Zeller T, Kuulasmaa K. BiomarCaRE Investigators. Troponin I and cardiovascular risk prediction in the general population: the BiomarCaRE consortium. Eur Heart J 2016;37:2428–2437. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Ndrepepa G, Braun S, Schulz S, Mehilli J, Schömig A, Kastrati A. High-sensitivity troponin T level and angiographic severity of coronary artery disease. Am J Cardiol 2011;108:639–643. [DOI] [PubMed] [Google Scholar]
- 24.Samman Tahhan A, Sandesara P, Hayek SS, Hammadah M, Alkhoder A, Kelli HM, Topel M, O’Neal WT, Ghasemzadeh N, Ko YA, Gafeer MM, Abdelhadi N, Choudhary F, Patel K, Beshiri A, Murtagh G, Kim J, Wilson P, Shaw L, Vaccarino V, Epstein SE, Sperling L, Quyyumi AA. High-sensitivity troponin I levels and coronary artery disease severity, progression, and long-term outcomes. J Am Heart Assoc 2018;7. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Whang W, Kubzansky LD, Kawachi I, Rexrode KM, Kroenke CH, Glynn RJ, Garan H, Albert CM. Depression and risk of sudden cardiac death and coronary heart disease in women: results from the Nurses’ Health Study. J Am Coll Cardiol 2009;53:950–958. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Ledwidge M, Gallagher J, Conlon C, Tallon E, O’Connell E, Dawkins I, Watson C, O’Hanlon R, Bermingham M, Patle A, Badabhagni MR, Murtagh G, Voon V, Tilson L, Barry M, McDonald L, Maurer B, McDonald K. Natriuretic peptide-based screening and collaborative care for heart failure: the STOP-HF randomized trial. JAMA 2013;310:66–74. [DOI] [PubMed] [Google Scholar]
- 27.Ridker PM, Glynn RJ. Hennekens CH. C-reactive protein adds to the predictive value of total and HDL cholesterol in determining risk of first myocardial infarction. Circulation 1998;97:2007–2011. [DOI] [PubMed] [Google Scholar]
- 28.Hemingway H, Philipson P, Chen R, Fitzpatrick NK, Damant J, Shipley M, Abrams KR, Moreno S, McAllister KS, Palmer S, Kaski JC, Timmis AD, Hingorani AD. Evaluating the quality of research into a single prognostic biomarker: a systematic review and meta-analysis of 83 studies of C-reactive protein in stable coronary artery disease. PLoS Med 2010;7:e1000286. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Ridker PM, Everett BM, Thuren T, MacFadyen JG, Chang WH, Ballantyne C, Fonseca F, Nicolau J, Koenig W, Anker SD, Kastelein JJP, Cornel JH, Pais P, Pella D, Genest J, Cifkova R, Lorenzatti A, Forster T, Kobalava Z, Vida-Simiti L, Flather M, Shimokawa H, Ogawa H, Dellborg M, Rossi PRF, Troquay RPT, Libby P, Glynn RJ, CANTOS Trial Group. Antiinflammatory therapy with canakinumab for atherosclerotic disease. N Engl J Med 2017;377:1119–1131. [DOI] [PubMed] [Google Scholar]
- 30.Xu Q Role of heat shock proteins in atherosclerosis. Arterioscler Thromb Vasc Biol 2002;22:1547–1559. [DOI] [PubMed] [Google Scholar]
- 31.Zhu J, Quyyumi AA, Wu H, Csako G, Rott D, Zalles-Ganley A, Ogunmakinwa J, Halcox J, Epstein SE. Increased serum levels of heat shock protein 70 are associated with low risk of coronary artery disease. Arterioscler Thromb Vasc Biol 2003;23:1055–1059. [DOI] [PubMed] [Google Scholar]
- 32.Sobel BE, Hardison RM, Genuth S, Brooks MM, McBane RD 3rd, Schneider DJ, Pratley RE, Huber K, Wolk R, Krishnaswami A, Frye RL, BARI 2D Investigators. Profibrinolytic, antithrombotic, and antiinflammatory effects of an insulin-sensitizing strategy in patients in the Bypass angioplasty revascularization Investigation 2 Diabetes (BARI 2D) trial. Circulation 2011;124:695–703. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Fuhrman B The urokinase system in the pathogenesis of atherosclerosis. Atherosclerosis 2012;222:8–14. [DOI] [PubMed] [Google Scholar]
- 34.Persson M, Engström G, Björkbacka H, Hedblad B. Soluble urokinase plasminogen activator receptor in plasma is associated with incidence of CVD. Results from the Malmo diet and cancer study. Atherosclerosis 2012;220:502–505. [DOI] [PubMed] [Google Scholar]
- 35.Hayek SS, Leaf DE, Samman Tahhan A, Raad M, Sharma S, Waikar SS, Sever S, Camacho A, Wang X, Dande RR, Ibrahim NE, Baron RM, Altintas MM, Wei C, Sheikh-Hamad D, Pan JS, Holliday MW Jr, Januzzi JL, Weisbord SD, Quyyumi AA, Reiser J. Soluble urokinase receptor and acute kidney injury. N Engl J Med 2020;382:416–426. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Cannon CP, Blazing MA, Giugliano RP, McCagg A, White JA, Theroux P, Darius H, Lewis BS, Ophuis TO, Jukema JW, De Ferrari GM, Ruzyllo W, De Lucca P, Im K, Bohula EA, Reist C, Wiviott SD, Tershakovec AM, Musliner TA, Braunwald E, Califf RM. IMPROVE-IT Investigators. Ezetimibe added to statin therapy after acute coronary syndromes. N Engl J Med 2015;372:2387–2397. [DOI] [PubMed] [Google Scholar]
- 37.Ghasemzadeh N, Brooks MM, Vlachos H, Hardison R, Sikora S, Sperling L, Quyyumi AA, Epstein SE. An aggregate biomarker risk score predicts high risk of near-term myocardial infarction and death: findings from BARI 2D (bypass angioplasty revascularization Investigation 2 Diabetes). J Am Heart Assoc 2017;6:e003587. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Mann DL. The emerging role of innate immunity in the heart and vascular system: for whom the cell tolls. Circ Res 2011;108:1133–1145. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.