Abstract Body (Do not enter title and authors here): Background: Newly developed cardiac magnetic resonance (CMR) derived cardiac quantitative susceptibility mapping (QSM) is a non-invasive technique to quantify venous oxygen saturation (SVO2).

Objective: In this study, we test whether impaired non-invasive SVO2 (IΔO2) predicts prognosis independent of conventional imaging-based risk stratification.

Methods: The population comprised of prospectively enrolled patients undergoing CMR for known or suspected pulmonary hypertension. Dedicated QSM imaging was performed to calculate non-invasive SVO2. In accordance with prior validation study of health volunteers, impaired SVO2 (IΔO2) was defined as greater than 23.7%. Echo was performed as a secondary measure of chamber remodeling and right ventricular (RV) afterload. Clinical data including heart failure hospitalization and all-cause mortality were acquired blinded to imaging indices.

Results: In total, 227 patients underwent CMR (age 55.1 ± 15.2 years, 45.4% men); 41 patients (18.1%) had IΔO2. IΔO2 was associated with greater left ventricular (LV) and RV dysfunction on CMR (p<0.05 for all) which was paralleled by more advanced heart failure symptoms (p<0.001). IΔO2 was also closely linked to RV afterload; pulmonary pressure was significantly higher among patients with IΔO2 versus those without (29.6 ± 11.8 vs. 45.6± 22.8mmHg, p<0.001). During a median follow up of 1.2 years, there were 36 composite outcomes (mortality and heart failure admission). IΔO2 was independently associated with heart failure hospitalization and death even after controlling for LV and RV function (HR: 2.52; CI 1.20-5.30; p=0.02). Figure 1 provides Kaplan Meier survival curve demonstrating increased mortality and heart failure hospitalization among patients with IΔO2 (p=0.004).

Conclusion: CMR quantified IΔO2 improves prognostic stratification of pulmonary hypertension beyond conventional imaging indices.