Proteomic phenotyping with machine learning for cardiovascular outcomes in haemodialysis: insights from the AURORA trial

Salib M, Girerd S, Pinet F, März W, Scharnagl H, Massy ZA, Leroy C, Duarte K, Bresso E, Lacomblez C, Jardine AG, Schmieder R, Fellstrom B, Lopez-Andres N, Rossignol P, Zannad F, Girerd N (2026)

Publication Type: Journal article

Publication year: 2026

Journal

European Heart Journal – Digital Health Oxford University Press

Book Volume: 7

Journal Issue: 3

DOI: 10.1093/ehjdh/ztag044

Abstract

Abstract Aims Cardiovascular (CV) trials have yielded neutral results in haemodialysis. A better understanding of patient profiles is needed to personalize treatment strategies in order to improve CV outcomes in this setting. This study sought to identify biological phenotypes based on proteomic data using machine learning approaches in patients undergoing haemodialysis. Methods and results A clustering analysis using 253 plasma protein biomarkers was performed in 382 patients (machine learning derivation analysis) from the AURORA trial, which tested the effect of rosuvastatin on CV outcomes in patients on haemodialysis. A decision tree was subsequently constructed to predict cluster membership and assess its association with CV outcomes in another subset of the trial (n = 389 patients, validation analysis). Four phenotypes were identified, namely ‘cytokine storm signalling’, ‘toll-like receptors (TLRs) signalling’, ‘multiple pathways related to inflammation and fibrosis’ phenotypes, as well as a ‘reference phenotype’ which exhibited the least biological abnormalities. In multivariable analysis of the validation study, after adjusting for key prognostic factors, the TLRs phenotype was significantly associated with CV death, all-cause mortality, and MACE (HR = 1.65 [1.13–2.41], 1.43 [1.03–1.98], and 1.48 [1.04–2.10], respectively). Conclusion Using unsupervised machine learning on proteomic data, we identified four mechanistic biological phenotypes involving cytokine storm and TLRs signalling, inflammation and fibrosis. These biological phenotypes may contribute to CV prognosis and pave the way for personalized therapy in haemodialysis.

Authors with CRIS profile

Roland Schmieder Department of Medicine 4 – Nephrology and Hypertension

Involved external institutions

Université de Lorraine FR France (FR) Institut de Pasteur Lille FR France (FR) Universitätsklinikum Mannheim / University Medical Centre Mannheim (Universitätsmedizin Mannheim) DE Germany (DE) Medizinische Universität Graz AT Austria (AT) Hôpital Ambroise-Paré FR France (FR) University of Glasgow GB United Kingdom (GB) Uppsala University Hospital / Akademiska sjukhuset SE Sweden (SE) Hospital Universitario de Navarra ES Spain (ES)

How to cite

APA:

Salib, M., Girerd, S., Pinet, F., März, W., Scharnagl, H., Massy, Z.A.,... Girerd, N. (2026). Proteomic phenotyping with machine learning for cardiovascular outcomes in haemodialysis: insights from the AURORA trial. European Heart Journal – Digital Health, 7(3). https://doi.org/10.1093/ehjdh/ztag044

MLA:

Salib, Madonna, et al. "Proteomic phenotyping with machine learning for cardiovascular outcomes in haemodialysis: insights from the AURORA trial." European Heart Journal – Digital Health 7.3 (2026).

BibTeX: Download