Comparison of different non-invasive estimates of right ventriculo-arterial coupling as outcome predictors in dilated cardiomyopathy

Abstract Right ventriculo-arterial coupling (RVAC) has emerged as an outcome predictor in pulmonary hypertension, but its role in left-sided HF remains to be clarified. Different non-invasive estimates for RVAC have been proposed, but a direct comparison of their prognostic ability is lacking. We sought to evaluate RVAC non-invasively using five different methods in a cohort of patients with dilated cardiomyopathy (DCM) and to assess their prognostic role. We enrolled 121 consecutive patients with non-ischemic DCM in sinus rhythm, who underwent comprehensive echocardiography, including speckle-tracking and 3D analysis. They were prospectively followed for 19±11 months for a composite endpoint of death, nonfatal cardiac arrest and hospitalization. RVAC was measured using five non-invasive estimates: (1) as the tricuspid annular plane systolic excursion (TAPSE)/pulmonary artery systolic pressure (PASP) ratio; (2) as the right ventricular (RV) global longitudinal strain (RV-GLS)/PASP ratio; (3) as the longitudinal strain of the RV free wall (LS-RVFW)/PASP ratio; (4) as the 3D RV ejection fraction (RVEF)/PASP ratio; and (5) as the 3D RV stroke volume (SV)/3D RV end-systolic volume (ESV) ratio. Mean age in the study cohort was 59±14 years and the majority (74%) were men. 55 patients (46%) reached the endpoint: there were 27 deaths, 5 nonfatal cardiac arrests and 23 readmissions for heart failure. Patients with adverse outcome had lower TAPSE/PASP (0.42±0.23 vs. 0.62±0.33, p<0.001), lower RVEF/PASP (0.96±0.49 vs. 1.47±0.67, p<0.001), lower RV SV/ESV (0.57±0.20 vs. 0.93±0.29, p<0.001) and higher RV-GLS/PASP (-0.25±0.20 vs. -0.38±0.23, p=0.002) and LS-RVFW/PASP (-0.28±0.30 vs. -0.51±0.33, p<0.001), reflecting greater ventriculo-vascular decoupling. In univariate Cox regression, all three RVAC indices were outcome predictors: HR=0.09 [95% CI, 0.02-0.37], p=0.001 for TAPSE/PASP, HR=13.67 [95% CI, 2.77-67.48], p=0.001 for RV-GLS/PASP, HR=8.72 [95% CI, 3.24-23.45], p<0.001 for LS-RVFW/PASP, HR=0.27 [95% CI, 0.14-0.50], p<0.001 for RVEF/PASP, HR=0.004 [95% CI, 0.001-0.019], p<0.001 for RV SV/ESV. In ROC analysis, the SV/ESV ratio had the highest AUC (AUC=0.88, p<0.001), a value lower than 0.74 having 86% sensitivity and 77% specificity for event prediction. After adjustment for age, NYHA class, left ventricular ejection fraction, and maximal left atrial volume, all coupling indices remained independent predictors of outcome: HR=1.90 [95% CI, 1.01-3.57], p=0.047 for TAPSE/PASP, HR=2.15 [95% CI, 1.17-3.96], p=0.01 for RV-GLS/PASP, HR=2.48 [95% CI, 1.34-4.60], p=0.004 for LS-RVFW/PASP, HR=3.19 [95% CI, 1.19-8.52], p=0.02 for RVEF/PASP, HR=7.50 [95% CI, 3.47-16.22], p<0.001 for RV SV/ESV. In DCM, RVAC is significantly more impaired in patients with major adverse events. Non-invasive RVAC was an independent predictor of adverse outcome irrespective of the formula used for estimation, but 3D RV SV/ESV ratio had the greatest prognostic power.