João Pedro Ferreira1, Abhinav Sharma2, Cyrus Mehta3,4, George Bakris5, Patrick Rossignol6, William B White7, Faiez Zannad6. 1. Université de Lorraine, Centre D'Investigation Clinique- Plurithématique Inserm CIC-P 1433, Inserm U1116, CHRU Nancy Hopitaux de Brabois, F-CRIN INI-CRCT (Cardiovascular and Renal Clinical Trialists), Nancy, France. j.ferreira@chru-nancy.fr. 2. Division of Cardiology, McGill University Health Centre, Montreal, QC, Canada. 3. Cytel Corporation, Cambridge, MA, USA. 4. Harvard T.H. Chan School of Public Health, Boston, MA, USA. 5. Department of Medicine, American Heart Association Comprehensive Hypertension Center, University of Chicago, Chicago, USA. 6. Université de Lorraine, Centre D'Investigation Clinique- Plurithématique Inserm CIC-P 1433, Inserm U1116, CHRU Nancy Hopitaux de Brabois, F-CRIN INI-CRCT (Cardiovascular and Renal Clinical Trialists), Nancy, France. 7. Calhoun Cardiology Center, University of Connecticut School of Medicine, Farmington, Connecticut, USA.
Abstract
BACKGROUND:Patients with diabetes who had a recent myocardial infarction (MI) are at high risk of cardiovascular events. Therefore, risk assessment is important for treatment and shared decisions. We used data from EXAMINE trial to investigate whether a multi-proteomic approach would provide specific proteomic signatures and also improve the prognostic capacity for determining the risk of cardiovascular death, MI, stroke, heart failure [HF], all-cause death, and combinations of these outcomes. METHODS:93 circulating proteins (92 from the Olink® CVDII plus troponin) were assessed in 5131 patients. Cox, competing risks, and reclassification measures were applied. RESULTS: The clinical model showed good discrimination and calibration for all outcomes. On top of the clinical model that included age, sex, smoking, diabetes duration, history of MI (prior to the index MI of inclusion), history of HF hospitalization, history of stroke, atrial fibrillation, hypertension, systolic blood pressure, statin therapy, estimated glomerular filtration rate, and study treatment (alogliptin or placebo), troponin and BNP added prognostic information to the composite of cardiovascular death, MI, or stroke (∆C-index + 5%) and cardiovascular death alone (∆C-index + 7%). Troponin, BNP, and TRAILR2 added prognostic information on all-cause death and the composite of cardiovascular death or HF hospitalization. HF hospitalization alone was improved by adding BNP and Gal-9. For MI, troponin, FGF23, and AMBP added prognostic value; whereas for stroke, only troponin added prognostic value (multi-proteomics improved C-index > 3% [p < 0.001] for all the studied outcomes). The addition of the final biomarker selection to the clinical model improved event reclassification (cNRI from + 23% to + 64%). Specifically, the addition of the biomarkers allowed a better classification of patients at low risk (as having "true" low risk) and patients and high risk (as having "true" high risk). These results were consistent for all the studied outcomes with even more marked differences in the fatal events. CONCLUSIONS: The addition of multi-proteomic biomarkers to a clinical model in this population with diabetes and a recent MI allowed a better risk prediction and event reclassification, potentially helping for better risk assessment and targeted treatment decisions. T2D type 2 diabetes, MI myocardial infarction, CV cardiovascular, HFH heart failure hospitalization, Δ delta, cNRI continuous net reclassification index, BNP brain natriuretic peptide, TRAILR2 trail receptor 2 (or death receptor 5), Gal-9 galectin-9, FGF23 fibroblast growth factor 23.
RCT Entities:
BACKGROUND:Patients with diabetes who had a recent myocardial infarction (MI) are at high risk of cardiovascular events. Therefore, risk assessment is important for treatment and shared decisions. We used data from EXAMINE trial to investigate whether a multi-proteomic approach would provide specific proteomic signatures and also improve the prognostic capacity for determining the risk of cardiovascular death, MI, stroke, heart failure [HF], all-cause death, and combinations of these outcomes. METHODS: 93 circulating proteins (92 from the Olink® CVDII plus troponin) were assessed in 5131 patients. Cox, competing risks, and reclassification measures were applied. RESULTS: The clinical model showed good discrimination and calibration for all outcomes. On top of the clinical model that included age, sex, smoking, diabetes duration, history of MI (prior to the index MI of inclusion), history of HF hospitalization, history of stroke, atrial fibrillation, hypertension, systolic blood pressure, statin therapy, estimated glomerular filtration rate, and study treatment (alogliptin or placebo), troponin and BNP added prognostic information to the composite of cardiovascular death, MI, or stroke (∆C-index + 5%) and cardiovascular death alone (∆C-index + 7%). Troponin, BNP, and TRAILR2 added prognostic information on all-cause death and the composite of cardiovascular death or HF hospitalization. HF hospitalization alone was improved by adding BNP and Gal-9. For MI, troponin, FGF23, and AMBP added prognostic value; whereas for stroke, only troponin added prognostic value (multi-proteomics improved C-index > 3% [p < 0.001] for all the studied outcomes). The addition of the final biomarker selection to the clinical model improved event reclassification (cNRI from + 23% to + 64%). Specifically, the addition of the biomarkers allowed a better classification of patients at low risk (as having "true" low risk) and patients and high risk (as having "true" high risk). These results were consistent for all the studied outcomes with even more marked differences in the fatal events. CONCLUSIONS: The addition of multi-proteomic biomarkers to a clinical model in this population with diabetes and a recent MI allowed a better risk prediction and event reclassification, potentially helping for better risk assessment and targeted treatment decisions. T2D type 2 diabetes, MI myocardial infarction, CV cardiovascular, HFH heart failure hospitalization, Δ delta, cNRI continuous net reclassification index, BNP brain natriuretic peptide, TRAILR2 trail receptor 2 (or death receptor 5), Gal-9 galectin-9, FGF23 fibroblast growth factor 23.
Authors: Sean D Pokorney; Sana M Al-Khatib; Jie-Lena Sun; Phillip Schulte; Christopher M O'Connor; John R Teerlink; Paul W Armstrong; Justin A Ezekowitz; Randall C Starling; Adriaan A Voors; Eric J Velazquez; Adrian F Hernandez; Robert J Mentz Journal: Eur J Heart Fail Date: 2017-12-20 Impact factor: 15.534
Authors: Emil Wolsk; Brian Claggett; Marc A Pfeffer; Rafael Diaz; Kenneth Dickstein; Hertzel C Gerstein; Francesca C Lawson; Eldrin F Lewis; Aldo P Maggioni; John J V McMurray; Jeffrey L Probstfield; Matthew C Riddle; Scott D Solomon; Jean-Claude Tardif; Lars Køber Journal: J Am Heart Assoc Date: 2017-05-29 Impact factor: 5.501
Authors: Michal Mihalovic; Petr Mikulenka; Hana Línková; Marek Neuberg; Ivana Štětkářová; Tomáš Peisker; David Lauer; Petr Tousek Journal: J Clin Med Date: 2022-05-02 Impact factor: 4.241
Authors: João Pedro Ferreira; Patrick Rossignol; George Bakris; Cyrus Mehta; William B White; Faiez Zannad Journal: Cardiovasc Diabetol Date: 2021-09-14 Impact factor: 9.951