Proteomic Signatures in Atrial Fibrillation: Deciphering the Risk of Heart Failure Hospitalization and Ejection Fraction Phenotypes

Introduction: The Interplay Between Atrial Fibrillation and Heart Failure

Atrial fibrillation (AF) and heart failure (HF) are often described as the twin epidemics of modern cardiovascular medicine. Their relationship is both bidirectional and synergistic; AF predisposes patients to HF through tachycardiomyopathy and loss of atrial kick, while HF promotes the structural and electrical remodeling of the atria that sustains AF. Despite advancements in anticoagulant and rate/rhythm control therapies, patients with AF remain at a significantly elevated risk for HF-related hospitalization.

One of the primary challenges in clinical practice is the early identification of patients at the highest risk for clinical decompensation. Traditional clinical risk scores often fail to capture the underlying molecular heterogeneity of these patients. Furthermore, the differentiation between HF with reduced ejection fraction (HFrEF) and HF with preserved ejection fraction (HFpEF) remains difficult in the context of AF, as both conditions share overlapping symptoms and diagnostic markers. Recently, multiplex screening of plasma proteins has emerged as a powerful tool to identify novel biomarkers and pathophysiological pathways, offering a window into the personalized risk profiles of patients with AF.

Highlights

This study provides several critical insights into the molecular landscape of patients with AF and their risk of subsequent heart failure events:

1. Multiplex proteomic screening identified 10 key biomarkers, including NT-proBNP, hs-cTnT, and FGF-23, that are most strongly and significantly associated with the risk of HF hospitalization after rigorous clinical adjustment.
2. Distinct pathophysiological signatures were identified for HF subtypes: HFrEF is more closely linked to cardiorenal dysfunction and systemic inflammation.
3. HFpEF is uniquely associated with proteins involved in adipose metabolism and tissue repair, suggesting a distinct metabolic etiology.
4. Beyond natriuretic peptides, markers like Spondin 1 and IGFBP-7 offer incremental prognostic value, potentially refining future risk stratification models.

Study Design and Methodology

The study utilized a case-cohort design within the landmark ARISTOTLE (Apixaban for Reduction in Stroke and Other Thromboembolic Events in Atrial Fibrillation) trial. This trial originally compared apixaban with warfarin for stroke prevention in patients with AF and at least one additional risk factor for stroke. For the current proteomic analysis, researchers evaluated 596 cases who experienced HF hospitalizations during the follow-up period and compared them with 4,029 randomly selected controls who did not experience HF hospitalization.

Plasma samples obtained at the time of randomization were analyzed using a combination of conventional immunoassays and proximity extension assay (PEA) panels. The PEA technology allowed for the simultaneous measurement of 268 biomarkers across various biological pathways. To identify the most relevant biomarkers, the researchers employed advanced statistical techniques, including random survival forest, Boruta feature selection, and Cox regression analyses. To account for the large number of comparisons, a stringent Bonferroni-Holm adjustment was applied (P≤0.00027) to ensure statistical robustness.

Key Findings: Biomarkers Associated with HF Hospitalization

Among the 268 biomarkers evaluated, the study identified a core group of proteins that were most strongly associated with an increased risk of HF hospitalization. These associations remained significant even after adjusting for baseline clinical characteristics, renal function, and established cardiac biomarkers (NT-proBNP and hs-cTnT).

The Top Prognostic Biomarkers

The most significant biomarkers identified included:
– NT-proBNP (N-terminal pro-B-type natriuretic peptide) and BNP (B-type natriuretic peptide): As expected, these markers of myocardial stretch remained the strongest predictors.
– hs-cTnT (High-sensitivity cardiac troponin T): A marker of ongoing myocardial injury.
– Fibroblast Growth Factor 23 (FGF-23): Involved in phosphate metabolism and linked to left ventricular hypertrophy and fibrosis.
– Spondin 1: An extracellular matrix protein associated with cardiac remodeling.
– Insulin-like Growth Factor Binding Protein 7 (IGFBP-7): A marker of cellular senescence and myocardial stiffness.
– Urokinase-type Plasminogen Activator Receptor (uPAR): A marker of inflammation and tissue remodeling.
– Osteopontin and Pentraxin-related protein 3 (PTX3): Markers of systemic inflammation and vascular stress.
– Transferrin Receptor Protein 1 (TfR1): Suggesting a link between iron metabolism and HF risk.

Pathophysiological Divergence: HFrEF vs. HFpEF

A major contribution of this research is the identification of different proteomic profiles between patients who developed HFrEF versus those who developed HFpEF. Among patients with prevalent HF at baseline, nine biomarkers remained significant after multiplicity adjustment, revealing clear biological distinctions.

HFrEF: Cardiorenal and Inflammatory Dominance

Patients with HFrEF exhibited significantly higher levels of:
– NT-proBNP and BNP
– hs-cTnT
– Renin and Angiotensin-converting enzyme 2 (ACE2): Reflecting heightened activation of the renin-angiotensin-aldosterone system (RAAS).
– Growth Differentiation Factor 15 (GDF-15) and Interleukin-6 (IL-6): Indicating a higher burden of systemic inflammation and cellular stress.

These findings reinforce the traditional view of HFrEF as a state of neurohormonal overactivation and progressive myocardial damage.

HFpEF: Adipose Metabolism and Tissue Repair

In contrast, patients with HFpEF showed higher levels of:
– Stem Cell Factor (SCF): Involved in hematopoiesis and tissue repair mechanisms.
– Leptin: A hormone secreted by adipose tissue, highlighting the metabolic and obesity-related drivers of HFpEF.

This divergence suggests that while HFrEF is driven by injury and neurohormonal failure, HFpEF may be more closely tied to metabolic dysfunction, systemic microvascular inflammation, and altered tissue repair pathways.

Expert Commentary: Mechanistic Insights and Clinical Implications

The discovery of these biomarker associations provides a more nuanced understanding of the AF-HF continuum. For clinicians, the prominence of FGF-23 and IGFBP-7 is particularly noteworthy. FGF-23 has been increasingly recognized not just as a marker of renal health, but as a direct mediator of cardiac hypertrophy. Its strong association with HF hospitalization in AF patients suggests that phosphate-mineral pathways may be a potential target for future therapeutic interventions.

Furthermore, the association of uPAR and Spondin 1 with HF events underscores the role of the extracellular matrix and chronic low-grade inflammation in the progression of AF toward clinical heart failure. The fact that these markers provide prognostic information beyond NT-proBNP suggests that they capture biological processes—such as fibrosis and vascular remodeling—that are not fully reflected by myocardial stretch alone.

From a diagnostic perspective, the distinction between HFrEF and HFpEF using SCF and Leptin is revolutionary. It supports the growing consensus that HFpEF is a systemic syndrome rather than just a localized cardiac disease. The higher levels of Leptin in HFpEF patients align with the clinical observation that obesity is a dominant risk factor for this HF subtype. This could eventually lead to a more personalized approach to AF management, where biomarker profiles guide the use of specific agents, such as SGLT2 inhibitors or GLP-1 receptor agonists, particularly in those with a “metabolic” proteomic signature.

Summary and Future Directions

In conclusion, this large-scale proteomic analysis of the ARISTOTLE trial cohort identifies a robust set of biomarkers that predict HF hospitalization in patients with AF. The study confirms the prognostic power of established markers like NT-proBNP and hs-cTnT while introducing novel candidates like FGF-23, Spondin 1, and uPAR into the clinical conversation. Perhaps most importantly, the research highlights the biological heterogeneity of heart failure subtypes, linking HFrEF to cardiorenal stress and HFpEF to adipose metabolism.

Future research should focus on validating these biomarker panels in diverse real-world populations and determining whether biomarker-guided therapy can reduce the incidence of HF hospitalization in AF patients. As we move toward a precision medicine model in cardiology, these proteomic signatures may serve as the foundation for earlier diagnosis and more effective, targeted treatments.

References

Pol T, Lindbäck J, Oldgren J, Alexander JH, Siegbahn A, Wallentin L, Hijazi Z. Plasma Biomarkers Associated With Heart Failure Hospitalization Among Patients With Atrial Fibrillation and Subtypes of Heart Failure. J Am Heart Assoc. 2026 Jan 20;15(2):e045970. doi: 10.1161/JAHA.125.045970. Epub 2026 Jan 14. PMID: 41532553.