Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2025 Jul 31;16(1):7042.
doi: 10.1038/s41467-025-62218-7.

Biomarker panels for improved risk prediction and enhanced biological insights in patients with atrial fibrillation

Pascal B Meyre  1   2 , Stefanie Aeschbacher  3   4 , Steffen Blum  3   4 , Tobias Reichlin  5 , Moa Haller  6   7 , Nicolas Rodondi  6   7 , Andreas S Müller  8 , Alain Bernheim  8 , Jürg Hans Beer  9 , Giorgio Moschovitis  10   11 , André Ziegler  12 , Bianca Wahrenberger  3   4 , Elia Rigamonti  10   11 , Giulio Conte  11 , Philipp Krisai  3   4 , Leo H Bonati  13 , Stefan Osswald  3   4 , Michael Kühne  3   4 , David Conen  14   15   16
Affiliations

Biomarker panels for improved risk prediction and enhanced biological insights in patients with atrial fibrillation

Pascal B Meyre et al. Nat Commun. .

Abstract

Atrial fibrillation (AF) increases the risk of adverse cardiovascular events, yet the underlying biological mechanisms remain unclear. We evaluate a panel of 12 circulating biomarkers representing diverse pathophysiological pathways in 3817 AF patients to assess their association with adverse cardiovascular outcomes. We identify 5 biomarkers including D-dimer, growth differentiation factor 15 (GDF-15), interleukin-6 (IL-6), N-terminal pro-B-type natriuretic peptide (NT-proBNP), and high-sensitivity troponin T (hsTropT) that independently predict cardiovascular death, stroke, myocardial infarction, and systemic embolism, significantly enhancing predictive accuracy. Additionally, GDF-15, insulin-like growth factor-binding protein-7 (IGFBP-7), NT-proBNP, and hsTropT predict heart failure hospitalization, while GDF-15 and IL-6 are associated with major bleeding events. A biomarker model improves predictive accuracy for stroke and major bleeding compared to established clinical risk scores. Machine learning models incorporating these biomarkers demonstrate consistent improvements in risk stratification across most outcomes. In this work, we show that integrating biomarkers related to myocardial injury, inflammation, oxidative stress, and coagulation into both conventional and machine learning-based models refine prognosis and guide clinical decision-making in AF patients.

PubMed Disclaimer

Conflict of interest statement

Competing interests: P.B.M. received funding from the Swiss National Science Foundation outside the submitted work. S.A. received funding from the Swiss Heart Foundation and speaker fees from Roche Diagnostics outside of the submitted work. S.B. received funding from the Swiss National Science Foundation, the Mach-Gaensslen Foundation, and the Bangerter-Rhyner Foundation outside the submitted work. T.R. reports research grants from the Swiss National Science Foundation, the Swiss Heart Foundation, and the sitem insel support fund, all for work outside the submitted study. Speaker/consulting honoraria or travel support from Abbott/SJM, AstraZeneca, Brahms, Bayer, Biosense Webster, Biotronik, Boston Scientific, Daiichi Sankyo, Medtronic, Pfizer BMS, and Roche, all for work outside the submitted study. Support for his institution’s fellowship program from Abbott/SJM, Biosense Webster, Biotronik, Boston Scientific, and Medtronic for work outside the submitted study. A.M. reports fellowship and training support from Biotronik, Boston Scientific, Medtronic, Abbott/St. Jude Medical, and Biosense Webster; speaker honoraria from Biosense Webster, Medtronic, Abbott/St. Jude Medical, AstraZeneca, Daiichi Sankyo, Biotronik, MicroPort, Novartis, and consultant honoraria for Biosense Webster, Medtronic, Abbott/St. Jude Medcal and Biotronik. G.M. has received consultant fees for taking part in advisory boards from Novartis, Boehringer Ingelheim, Bayer, AstraZeneca, and Daiichi Sankyo, all outside of the current work. A.Z. is an employee of Roche Diagnostics, a commercial provider of diagnostic tests. M.K. reports personal fees from Bayer, personal fees from Böhringer Ingelheim, personal fees from Pfizer BMS, personal fees from Daiichi Sankyo, personal fees from Medtronic, personal fees from Biotronik, personal fees from Boston Scientific, personal fees from Johnson&Johnson, grants from Bayer, grants from Pfizer, grants from Boston Scientific, grants from BMS, grants from Biotronik. Grants from the Swiss National Science Foundation, the Swiss Heart Foundation, the Foundation for Cardiovascular Research Basel, and the University of Basel. D.C. has received consultant fees from Roche Diagnostics and Trimedics, outside of the current work. The remaining authors have nothing to disclose.

Figures

Fig. 1
Fig. 1. Associations between selected biomarkers and adverse cardiovascular outcomes from combined Cox models.
This figure shows standardized hazard ratios and 95% CIs of associations between backward-selected biomarkers and different adverse cardiac outcomes, derived from combined multivariable Cox models. A shows effect estimates of selected biomarkers for composite outcome of cardiovascular death, nonfatal ischemic stroke, nonfatal systemic embolism, or nonfatal myocardial infarction. B shows effect estimates of selected biomarkers for heart failure hospitalization. C shows effect estimates of selected biomarkers for major bleeding. D shows effect estimates of selected biomarkers for all strokes. E shows effect estimates of selected biomarkers for ischemic stroke. F shows effect estimates of selected biomarkers for myocardial infarction. G shows effect estimates of selected biomarkers for cardiovascular death. H shows effect estimates of selected biomarkers for all-cause death. I shows effect estimates of selected biomarkers for any bleeding. J shows effect estimates of selected biomarkers for clinically relevant non-major (NM) bleeding. All outcomes were assessed in N = 3817 AF patients. Dots and whiskers represent hazard ratios and 95% CIs. ALAT Alanine aminotransferase, ANG-2 Angiopoetin-2, GDF-15 growth differentiation factor‑15, hsTropT high-sensitivity troponin T, IGFBP-7 Insulin-like growth factor-binding protein-7, IL-6 Interleukin-6, NT-proBNP N-terminal pro-B-type natriuretic peptide, OPN Osteopontin.
Fig. 2
Fig. 2. Relative importance of predictors from combined Cox models.
This figure shows the relative importance of each clinical variable and backward-selected biomarkers for different adverse cardiac outcomes, derived from combined multivariable Cox models. A shows relative importance of variables for the association with the composite outcome of cardiovascular death, nonfatal ischemic stroke, nonfatal systemic embolism, or nonfatal myocardial infarction. B shows relative importance of variables for the association with heart failure hospitalization. C shows relative importance of variables for the association with major bleeding. D shows relative importance of variables for the association with all strokes. E shows relative importance of variables for the association with ischemic stroke. F shows relative importance of variables for the association with myocardial infarction. G shows relative importance of variables for the association with cardiovascular death. H shows relative importance of variables for the association with all-cause death. I shows the relative importance of variables for the association with any bleeding. J shows relative importance of variables for the association with clinically relevant non-major (NM) bleeding. All outcomes were assessed in N = 3817 AF patients. Dots represent the partial χ2 – degree of freedom values. Source data are provided as a Source Data file. ALAT Alanine aminotransferase, ANG-2 Angiopoetin-2, BMI body mass index, CAD coronary artery disease, CKD chronic kidney disease, eGFR estimated glomerular filtration rate, GDF-15 growth differentiation factor‑15, hsTropT high-sensitivity troponin T, IGFBP-7 Insulin-like growth factor-binding protein-7, IL-6 Interleukin-6, NT-proBNP N-terminal pro-B-type natriuretic peptide, OPN Osteopontin, SBP Systolic blood pressure.
Fig. 3
Fig. 3. Discriminatory performance of base Cox models with biomarkers vs. clinical scores for stroke and major bleeding.
This figure shows ROC curves of base Cox models with and without biomarkers and clinical risk scores for all strokes, ischemic stroke, and major bleeding. All outcomes were assessed in N = 3817 AF patients. ABC age, biomarkers, clinical history stroke risk score, CHA2DS2-VASc Congestive heart failure, Hypertension, Age (2 points if age >75 y), Diabetes, Stroke, Vascular disease, Sex category, HAS-BLED Hypertension, Abnormal renal/liver function, Stroke, Bleeding history or predisposition, Labile international normalized ratio, Elderly (>65 years), Drugs/alcohol concomitantly.
Fig. 4
Fig. 4. Predictive performance of Cox and machine learning models for outcomes with and without biomarkers.
The figure shows the AUC with 95% CI for Cox and machine learning models, comparing the performance of base and base + biomarker models for different adverse cardiac outcomes. The combined Cox models include age, sex, body mass index, current smoker, systolic blood pressure, history of diabetes, prior stroke or TIA, history of heart failure, chronic kidney disease, coronary artery disease, and backward-selected biomarkers. The machine learning models include all variables listed in the Supplementary Table 16 and all biomarkers. All outcomes were assessed in N = 3817 AF patients. Dots represent AUC values and whiskers indicate 95% CIs. AUC area under the curve.
See this image and copyright information in PMC

Similar articles

See all similar articles

References

    1. Conen, D. et al. Risk of death and cardiovascular events in initially healthy women with new-onset atrial fibrillation. JAMA305, 2080–2087 (2011). - PMC - PubMed
    1. Soliman, E. Z. et al. Atrial fibrillation and the risk of myocardial infarction. JAMA Intern. Med.174, 107–114 (2014). - PMC - PubMed
    1. Wolf, P. A., Abbott, R. D. & Kannel, W. B. Atrial fibrillation as an independent risk factor for stroke: the Framingham Study. Stroke22, 983–988 (1991). - PubMed
    1. Wang, T. J. et al. Temporal relations of atrial fibrillation and congestive heart failure and their joint influence on mortality: the Framingham Heart Study. Circulation107, 2920–2925 (2003). - PubMed
    1. Meyre, P. B. et al. Bleeding and ischaemic events after first bleed in anticoagulated atrial fibrillation patients: risk and timing. Eur. Heart J.43, 4899–4908 (2022). - PubMed

MeSH terms