A Multi-Marker Approach for Risk Stratification in Patients With Chronic Heart Failure: Insights From the HF-ACTION Trial

Introduction: It is unclear whether multi-marker approaches improve risk stratification among patients with heart failure (HF). We evaluated the prognostic utility of a multi-marker approach with biomarkers of HF status; Growth-differentiation factor-15 (GDF-15), galectin-3 (Gal-3), ST2, NT-proBNP, C- reactive protein (CRP), and high sensitivity troponin (hs TnT) in the setting of established clinical outcomes markers.Methods: The NHLBI Heart Failure: A Controlled Trial Investigating Outcomes of Exercise Training (HF-ACTION) trial randomized 2331 patients with HFrEF to exercise training or usual care. Median follow-up was 30 months. Cox proportional hazard models assessed the prognostic utility of adding the biomarkers to the validated HF-ACTION risk model for the following outcomes: all-cause death; all-cause death or rehospitalization; and cardiovascular (CV) death or HF hospitalization. Hazard ratios (HR; per doubling of biomarker), C-statistics, and integrated discrimination improvement (IDI) with 95% confidence interval (CI) were calculated.Results: Among patients with baseline biomarker measurements (n=912), the median age was 59.2 years (50.6-68.4), 28.9% were female, and 50.5% had an ischemic etiology of HF. When the 6 biomarkers were added to the clinical model, NT-proBNP had the strongest association with outcomes (figure). In addition, the panel of biomarkers improved discrimination for all-cause death (c-statistic from 0.71 to 0.77; IDI 0.08 95% CI 0.06, 0.11), all-cause mortality or rehospitalization (c-statistic from 0.61 to 0.64; IDI 0.05, 95% CI 0.04, 0.07), and CV death or HF hospitalization (c-statistic from 0.71 to 0.76; IDI 0.06, 95% CI 0.04, 0.08).Conclusion: In well characterized patients with HFrEF, a multi-marker approach improved risk stratification over established clinical outcome predictors. Routine use of a multi-marker approach for risk stratification in clinical practice warrants further evaluation.Author Disclosures: A. Sharma: Other Research Support; Modest; Takeda, Roche diagnostic. S.R. Stevens: None. A. Coles: None. M. Fiuzat: Other Research Support; Significant; Roche diagnostics. K.F. Adams: Other Research Support; Significant; Roche diagnostics. D.J. Whellan: None. M.P. Donahue: None. D.W. Kitzman: None. I.L. Pi na: None. F. Zannad: Honoraria; Significant; Roche Diagnostic. W.E. Kraus: None. C.M. O’Connor: Research Grant; Significant; Roche diagnostics. G. Felker: Research Grant; Significant; Amgen, Roche Diagnostics. Other Research Support; Significant; Amgen. Consultant/Advisory Board; Significant; Amgen.

Duke Scholars

Author William Erle Kraus Medicine, Cardiology

Author Mona Fiuzat Medicine, Clinical Pharmacology

Author Gary Michael Felker Medicine, Cardiology