Five Blood Biomarker Pathways That Predict Who Dies From Heart Failure

Heart failure affects a growing, increasingly older, and multimorbid population worldwide. Despite guideline-directed therapy, mortality after decompensation remains high. Established biomarkers—natriuretic peptides and cardiac troponins—are essential but capture only a fraction of mortality risk, particularly in biologically heterogeneous older adults where chronic inflammation, vascular stiffness, sarcopenia, impaired renal clearance, and endocrine remodeling alter baseline concentrations and clinical meaning.

This narrative review, published in Frontiers in Cardiovascular Medicine, synthesizes evidence from a PubMed search spanning database inception through June 2026 and proposes a mechanistic, pathway-based framework for HF mortality biomarkers. Rather than seeking the single strongest univariate predictor, the authors organize novel blood-based markers into five interconnected domains, each reflecting a distinct pathophysiological mechanism relevant to cardiac aging.

The first domain—fibro-inflammatory remodeling and myocardial stress—includes soluble ST2 (sST2), galectin-3, GDF-15, and heart-type fatty acid-binding protein (H-FABP). These markers reflect myocardial strain, macrophage-driven fibrosis, cardiomyocyte injury, and the senescence-associated secretory phenotype, with GDF-15 having older-adult-enriched evidence. The second domain covers congestion, endothelial dysfunction, and vascular aging, represented by carbohydrate antigen 125 (CA125), bioactive adrenomedullin (bio-ADM), mid-regional pro-adrenomedullin (MR-proADM), endothelin-1, and adhesion molecules. CA125 has older-adult-specific evidence and may detect fluid redistribution before overt edema appears. The third domain—cardiorenal and multiorgan reserve—includes NGAL, albumin, the creatinine-to-albumin ratio, the prognostic nutritional index, the De Ritis ratio, and the FT3/FT4 ratio, emphasizing that HF death often reflects systemic vulnerability rather than isolated ventricular failure. The fourth domain captures metabolic stress and inflammation through FGF21, phenylalanine, kynurenine, IL-6, hsCRP, the neutrophil-to-lymphocyte ratio, IL-8, and HMGB1—markers of mitochondrial dysfunction, dysregulated amino acid catabolism, and chronic inflammaging driving cachexia and frailty trajectories. The fifth domain encompasses proteomic and metabolomic phenotyping, where plasma proteomic clustering has been shown to identify HF subgroups with distinct outcomes that clinical clustering alone cannot detect, and metabolomic panels improve long-term mortality risk calibration in older cohorts.

The authors stress that aging fundamentally complicates biomarker interpretation and classify evidence as older-adult-specific, older-adult-enriched, or extrapolated from broader HF populations. Most biomarker evidence falls into the extrapolated category, highlighting a major evidence gap.

The proposed framework is explicitly conceptual and hypothesis-generating. Barriers to clinical translation include assay heterogeneity, insufficient age-specific validation, uncertain incremental utility beyond standard markers, and limited bedside interpretability. The authors conclude that prospective biomarker-guided studies in older, multimorbid HF populations are required before pathway-based panels can routinely inform treatment intensity, device decisions, or palliative care planning.