Under pressure: right ventricular and right atrial strain predict in-hospital death in cancer patients with pulmonary embolism

Background Cardio-oncology is a rapidly evolving field. Malignancy and oncologic therapies (chemotherapy, surgery), by determining a procoagulant state, are predisposing factors for venous thromboembolism and pulmonary embolism (PE). Moreover, patients with PE and cancer carry greater short- and long-term mortality than non-cancer patients with PE. Therefore, a thorough risk stratification is of utmost importance. Echocardiography plays a crucial role in evaluating the hemodynamic consequences and the risk of PE. Our purpose was to analyse cancer patients with PE hospitalised in a tertiary center, while focusing on echocardiographic parameters and their short-term prognostic impact. Methods The study retrospectively included all patients with active cancer hospitalised with PE in our cardiology department between 2016-2025. We collected data regarding the malignancy (site, stage, treatment, time from diagnosis) and the PE (risk category, anticoagulant therapy, NT pro BNP), and also a range of echocardiographic parameters with focus on the right atrium (RA) (size, atrial reservoir strain [RASres]) and the right ventricle (RV) (size, TAPSE, S wave velocity, RV global longitudinal strain [RV-GLS], RV free wall longitudinal strain [RVFW-LS]). Patients were followed for in-hospital death. Results 202 cancer patients with PE were included. Mean age in the study group was 67±12 years and 48% were men. Colo-rectal cancer (13.4%), followed by pulmonary (12.4%) and breast cancer (11.4%) had the greatest prevalence. In-hospital death was 6.4% (13 patients). Patients who died were older (74±12 vs. 67±12 years, p=0.05) and had a greater prevalence of high-risk PE (38% vs. 7%, p<0.001). Among patients who died, the RV was larger (47±5 vs. 35±6, p<0.001) and more dysfunctional (16±4 vs. 19±4, p=0.007 for TAPSE, 10±3 vs. 13±3, p=0.02 for S wave). Both RV strain components were significantly more impaired in patients who died: -10±2% vs. -15±5%, p<0.001 for RV-GLS, -8±3% vs. -17±6%, p<0.001 for RVFW-LS. Pulmonary artey systolic pressure (PASP) was higher in patients reaching the endpoint: 48±16 vs. 33±15, p=0.006, while RASres was lower: 14±4% vs. 24±11%, p<0.001. The following parameters predicted death in univariable analysis: age, systolic blood pressure (BP) on admission, RV diameter, TAPSE, S wave, RV-GLS, RVFW-LS, RASres. In ROC analysis, the best predictors of in-hospital death were RVFW-LS (AUC=0.90, p<0.001) and RV-GLS (AUC=0.86, p<0.001). Both RV strain parameters, as well as RASres remained independent predictors of in-hospital death after adjusting for age, PASP and systolic BP on admission: HR=1.52 [95% CI, 1.15–2.01], p=0.003 for RV-GLS, HR=1.46 [95% CI, 1.16–1.86], p=0.002 for RVFW-LS, HR=0.82 [95% CI, 0.69–0.97], p=0.02 for RASres. Conclusions RV-GLS, RVFW-LS and RASres were independent predictors of in-hospital death in cancer patients with PE.