Heart failure & cardiomyopathiesResearch article

A new protein panel for the diagnosis of HFpEF: combining machine-learning and liquid-chromatography mass-spectrometry proteomics

ESC Heart FailPublished

  • LC-MS plasma proteomics with machine learning
  • technically validated using the SomaScan 7K assay

Authors' conclusion

We identified a protein combination with higher specificity than NT-proBNP for diagnosing HFpEF. Integrating these biomarkers into clinical practice could improve HFpEF diagnosis and management, although external validation is required before clinical implementation.

Study context

Population
Patients with symptomatic HFpEF (LVEF ≥50%) from BIOSTAT-CHF and control subjects without heart failure from the PREVEND studyHFpEF patients (BIOSTAT-CHF) 249 · control subjects without heart failure (PREVEND) 99
Participants
249 vs 99
Comparison
Combined biomarker model compared to NT-proBNP for diagnosing HFpEF
  • HFpEFn = 249
  • Non-HF controlsn = 99

Main finding

As reported in the abstract

Baseline characteristics

female

OutcomeHFpEFNon-HF controls
median age77 years75 years
median LVEF59%Not reported in the abstract
median plasma NT-proBNP1056 pg/mL77 pg/mL

Further findings

Diagnostic performance of combined biomarker model vs NT-proBNP

Net Reclassification Index 1.57 (1.26–1.83)

specificity

AUC NT-proBNP 0.96 · Combined biomarker model 0.99

sensitivity Combined biomarker model 97%

OutcomeEstimateP value
Diagnostic performance of combined biomarker model vs NT-proBNP · AUC improved from 0.96 to 0.99(95% CI −0.00072 to 0.065)P = 0.036

Protein profiles

differentially abundant proteins
211
Principal component analysis showed a clear distinction between the protein profiles of patients with HFpEF versus controls
Two strongly downregulated proteins (TMP4 and YWHAZ) identified as biomarker candidates

Original publication

A new protein panel for the diagnosis of HFpEF: combining machine-learning and liquid-chromatography mass-spectrometry proteomics. ESC Heart Fail. Published 9 July 2026. doi:10.1093/eschf/xvag189

Summarised from the published abstract; every number is checked against it