Two-dimensional speckle tracking echocardiography assessed right ventricular function and exercise capacity in pre-capillary pulmonary hypertension.

Resting two-dimensional speckle tracking echocardiography (2D-STE) identified right ventricular (RV) systolic function were reported to predict exercise capacity in pulmonary hypertension (PH) patients, but little attention had been payed to 2D-STE detected RV diastolic function. Therefore, we aim to elucidate and compare the relations between 2D-STE identified RV diastolic/systolic functions and peak oxygen consumption (PVO) determined by cardiopulmonary exercise testing (CPET) in pre-capillary PH. 2D-STE was performed in 66 pre-capillary PH patients and 28 healthy controls. Linear correlation and multivariate regression analyses were performed to evaluate and compare the relations between RV 2D-STE parameters and PVO. Receiver operating characteristic curves were used to compare the predictive value of 2D-STE parameters in predicting the cut-off-PVO < 11 ml/min/kg. There were significant differences of all the 2D-STE parameters between PH patients and healthy controls. In patients, RV-peak global longitudinal strain (GLS, r = - 0.498, P < 0.001), RV- peak systolic strain rate (GSRs, r = - 0.537, P < 0.001) and RV- peak early diastolic strain rate (GSRe, r = 0.527, P < 0.001) significantly correlated with PVO, but no significant correlation was observed between RV- peak late diastolic strain rate (GSRa, r = 0.208, P = 0.093) and PVO. The first multivariate regression analysis of clinical data without echocardiographic parameters identified WHO functional class, NT-proBNP and BMI as independent predictors of PVO (Model-1, adjusted r = 0.421, P < 0.001); Then we added conventional echocardiographic parameters and 2D-STE parameters to the clinical data, identified S,(Model-2,adjusted r = 0.502, P < 0.001), RV-GLS (Model-3, adjusted r = 0.491, P < 0.001), RV-GSRe (Model-4, adjusted r = 0.500, P < 0.001) and RV-GSRs (Model-5, adjusted r = 0.519, P < 0.001) as independent predictors of PVO, respectively. The predictive power was increased, and Model-5 including RV-GSRs showed the highest predictive capability. ROC curves found RV-GSRs expressed the strongest predictive value (AUC = 0.88, P < 0.001), and RV-GSRs > - 0.65/s had a 88.2% sensibility and 82.2% specificity to predict PVO < 11 ml/min/kg. 2D-STE assessed RV function improves the prediction of exercise capacity represented by PVO in pre-capillary PH.