Clinical implications of late gadolinium enhancement in patients with active cancer undergoing stress CMR

Background

Cardiovascular disease and cancer are the two leading causes of mortality worldwide. (1) Advances in oncologic therapies have improved cancer survival, resulting in a growing population of patients with active malignancy or cancer survivors who remain at increased risk of cardiovascular and non-cancer mortality. Cardiovascular magnetic resonance (CMR) with stress perfusion and late gadolinium enhancement (LGE) is well established for assessing myocardial ischaemia, ventricular function, and fibrosis. (2) However, its prognostic value in patients with active cancer remains poorly defined.

Purpose

To (i) characterize stress CMR findings in patients with active cancer compared with propensity-matched controls, and (ii) assess the prognostic value of LGE for all-cause mortality.

Methods

We included all consecutive active cancer patients undergoing stress CMR at a single EACVI-accredited centre (2008–2025). 325 patients with active cancer were identified and matched 1:1 with 325 patients without cancer using propensity-score matching based on age, sex, cardiovascular risk factors (obesity, hypertension, dyslipidaemia, diabetes mellitus, smoking), and known coronary artery disease. CMR parameters included left ventricular (LV) volumes, ejection fraction (LVEF), right ventricular function, inducible ischemia, and LGE. The primary outcome was all-cause mortality, analysed using Cox regression and Kaplan–Meier methods.

Results

Patients with active cancer showed a higher prevalence of adverse CMR findings (Figure 1). Inducible ischemia was more frequent in active cancer patients (19.7% vs. 11.4%, p=0.005), as were LV dysfunction (18.2% vs. 8.3%, p<0.001), LV dilatation (12.9% vs. 6.2%, p=0.005), and right ventricular dysfunction (7.1% vs. 0.9%, p<0.001). LGE prevalence was similar between groups (16.6% vs. 17.8%, p=0.755).

During follow-up, all-cause mortality differed significantly according to cancer status and LGE presence (Figure 2, p=0.036). Patients without cancer and without LGE had the best survival. In non-cancer patients, LGE was not associated with increased mortality (p=0.254). In contrast, active cancer was associated with higher mortality regardless of LGE status. Compared with non-cancer patients without LGE, cancer patients without LGE had a 70% higher mortality risk (HR 1.70, p=0.016), while cancer patients with LGE had the highest risk (HR 2.14, p=0.010).

Conclusion

In patients with active cancer, the prognostic significance of LGE appears context-dependent. While LGE is a recognized marker of adverse outcomes in general cardiology populations, in cancer patients it likely reflects a vulnerable myocardial substrate that may amplify the adverse impact of malignancy and its systemic consequences. Stress CMR may offer prognostic information beyond diagnosis alone. This study is the first to comprehensively characterize stress CMR findings in a propensity-matched cohort of patients with active cancer.Figure 1.CMR findingsFigure 2.Kaplan–Meier - all-cause mortality