# Development and Validation of a Noninvasive Model of Mixed Venous Oxygen Saturation in Heart Failure

**Authors:** Adam K. McDiarmid, Bradley S. Chambers, David A. Broadbent, Roshan Patel, Gareth Matthews, Oscar Gonzalez-Fernandez, Sven Plein, Pankaj Garg, Peter P. Swoboda

PMC · DOI: 10.1016/j.jacadv.2025.102484 · JACC: Advances · 2026-01-28

## TL;DR

Researchers developed a noninvasive way to estimate mixed venous oxygen saturation in heart failure patients using MRI, which could predict future health risks.

## Contribution

A noninvasive CMR model (iSvO2) was developed and validated as an independent predictor of adverse outcomes in heart failure.

## Key findings

- iSvO2 strongly correlates with invasive SvO2 measurements (R = 0.82).
- iSvO2 is an independent predictor of mortality and heart failure hospitalization.
- The model remains significant after adjusting for multiple clinical variables.

## Abstract

Mixed venous oxygen saturation (SvO2), measured with right heart catheterization, is a crucial prognostic tool in patients with heart failure. The prognostic significance of SvO2 estimated noninvasively using cardiovascular magnetic resonance (CMR) from the T2 of intracardiac blood pools remains unknown.

The objective of the study was to develop a CMR model of mixed venous saturation (imaging-derived SvO2 [iSvO2]), and establish if it is associated with future adverse events in heart failure.

The iSvO2 was modeled in the discovery cohort (N = 30), who underwent CMR T2 mapping and invasive right heart catheterization, by linear regression. The validation cohort of 628 patients with recently diagnosed heart failure underwent clinical assessment, CMR, and follow-up (median 3 years [IQR: 1.5-4.8]) for a primary endpoint of all-cause mortality or heart failure hospitalization.

Significant positive correlation was found between the ratio of right ventricular blood pool T2/left ventricular blood pool T2 and invasive mixed venous oxygen saturation (R = 0.82; 95% CI: 0.66-0.91; P < 0.001), giving the equation: iSvO2 = 95·(RV−T2BP/LV−T2BP). In the validation cohort, there was a strong association between iSvO2 and the primary endpoint (HR: 0.66 for 10% change in iSvO2; 95% CI: 0.54-0.81; P < 0.001), which remained significant after adjusting for age, sex, left ventricular ejection fraction, right ventricular ejection fraction, N-terminal pro–B-type natriuretic peptide, NYHA functional class, and diabetes.

The CMR iSvO2, measured from simple T2 maps of left and right ventricular blood pool, allows accurate estimation of the invasive SvO2. In a real-world heart failure registry, iSvO2 is an independent predictor of mortality and heart failure hospitalization.

## Linked entities

- **Diseases:** heart failure (MONDO:0005252)

## Full-text entities

- **Diseases:** cardiogenic shock (MESH:D012770), stroke (MESH:D020521), pulmonary edema (MESH:D011654), hypertension (MESH:D006973), anemia (MESH:D000740), pump failure (MESH:D051437), angina (MESH:D000787), myocarditis (MESH:D009205), pulmonary hypertension (MESH:D006976), depression (MESH:D003866), diabetes (MESH:D003920), atrial fibrillation (MESH:D001281), hypoxemia (MESH:D000860), myocardial edema (MESH:D004487), Death (MESH:D003643), CMR (MESH:D002318), complication (MESH:D008107), HF (MESH:D006333), coronary stenosis (MESH:D023921), myocardial infarction (MESH:D009203), chronic kidney disease (MESH:D051436), congenital heart disease (MESH:D006330), coronary artery disease (MESH:D003324), cardiomyopathies (MESH:D009202), arrhythmia (MESH:D001145), hypertrophic cardiomyopathy (MESH:D002312), amyloidosis (MESH:D000686), pulmonary disease (MESH:D008171), cardiac disease (MESH:D006331), anxiety (MESH:D001007)
- **Chemicals:** SvO2 (-), Oxygen (MESH:D010100)
- **Species:** Homo sapiens (human, species) [taxon 9606]
- **Mutations:** A1C

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC12869880/full.md

## Figures

5 figures with captions in the complete paper: https://tomesphere.com/paper/PMC12869880/full.md

## References

28 references — full list in the complete paper: https://tomesphere.com/paper/PMC12869880/full.md

---
Source: https://tomesphere.com/paper/PMC12869880