Estimating prognosis: biomarkers

Abstract

Risk stratification informs the management of heart failure (HF). A range of clinical variables are associated with outcome in HF and have been incorporated into risk calculators which perform adequately at a population level, especially for those with more severe or recently decompensated HF. Their performance in predicting individual patient outcomes among the ambulant community-based populations with chronic HF is less assured. Risk score performance is enhanced by incorporation of selected circulating biomarkers. The cardiac natriuretic peptides, particularly B-type natriuretic peptide (BNP) and N-terminal proBNP (NT-proBNP), are independently predictive of mortality in both acute and chronic HF. Risk at a given level of BNP is similar regardless of left ventricular ejection fraction. The prognostic performance of NT-proBNP and BNP enriches therapeutic trials for clinical events. Trials of marker-guided therapy in HF consistently indicate that lowering plasma NT-proBNP (preferably to <1000 pg/mL) is associated with improved outcomes. The cardiac natriuretic peptides may be combined with other markers including mid-regional pro-adrenomedullin, ST2, growth/differentiation factor 15, and cardiac troponin to refine risk stratification. There is a need for an improved, independently validated, risk score calculator derived from a large well-annotated HF population incorporating both key clinical predictors and one or more circulating biomarkers.