Exercise right ventricular–pulmonary artery coupling reserve by 3D echocardiography predicts early HFpEF progression

Background Heart failure with preserved ejection fraction (HFpEF) often remains occult at rest, particularly in patients with unexplained exertional dyspnea. While left-sided indices dominate diagnostics, right ventricular–pulmonary artery (RV–PA) coupling reserve during exercise is poorly characterized. We investigated whether 3D echocardiographic RV–PA coupling identifies early HFpEF and predicts short-term progression. Purpose To determine whether exercise-derived 3D RV–PA coupling identifies early HFpEF and predicts short-term clinical progression in patients with unexplained exertional dyspnea. Methods We retrospectively analyzed 120 subjects (2023–2025): 80 with unexplained dyspnea (LVEF ≥50%) and 40 age-matched controls undergoing resting and semi-supine exercise 3D echocardiography. RV–PA coupling was defined as 3D-RVEF/PASP. A subset (n=30) underwent invasive validation against Ees/Ea. The primary endpoint was HF hospitalization or ≥1 NYHA class worsening at 12 months. Results Eighty dyspneic patients (NYHA II–III) and 40 controls were included. Mean age was 66±8 years; 58% were female; BMI was 29.4±4.1 kg/m². Hypertension and diabetes were present in 62% and 28%, respectively. Resting LVEF (61±5%), RVPAC (0.68±0.12 vs. 0.72±0.10; p=0.14), and PASP (31±6 vs. 29±5 mmHg; p=0.09) were similar between groups. During peak exercise, controls showed preserved coupling (+8±6%), whereas dyspneic patients demonstrated a significant decline (−18±11%; p<0.001); 42% exhibited exertional uncoupling (>20% reduction). Peak PASP was higher (52±9 vs. 41±7 mmHg; p<0.001) and 3D-RVEF augmentation was blunted (+3±5% vs. +11±6%; p<0.001). At 12 months, 22 dyspneic patients (28%) met the primary endpoint (7 HF hospitalizations, 15 NYHA worsening). Exercise RVPAC predicted events (AUC 0.91, 95% CI 0.84–0.96; sensitivity 86%, specificity 82%). RVPAC <0.51 conferred HR 4.2 (95% CI 2.1–8.3; p<0.001). Exercise RVPAC correlated strongly with invasive Ees/Ea (r=0.84; p<0.001) and outperformed E/e′, NT-proBNP, and H2FPEF score (p<0.01). Conclusion Exercise-derived 3D RV–PA coupling unmasks early right heart dysfunction and independently predicts short-term HFpEF progression, outperforming conventional markers and providing a noninvasive tool for risk stratification in unexplained dyspnea.