Right ventricular stroke work index in adult patients with pulmonary arterial hypertension: - a comparison between echocardiography and right heart catheterization

Abstract

Funding Acknowledgements Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Swedish Society of Pulmonary Hypertension Background In adult patients with pulmonary arterial hypertension (PAH), right ventricular (RV) failure may worsen rapidly, constituting a negative prognostic factor. In this population, non-invasive assessment of RV function is challenging. RV stroke work index (RVSWI) reflects right ventricular function and has been proposed to predict outcome in PAH. However, RVSWI assessed by echocardiography (ECHO) has not been thoroughly compared to measures with right heart catheterization (RHC) in adults. The aim of the present study was to therefore evaluate RVSWI derived by echocardiography (RVSWIECHO) vs. RHC (RVSWIRHC). Methods Fifty-four consecutive treatment naïve adult patients with PAH, were retrospectively analysed. All patients performed echocardiography and RHC with a median time of 1 day [IQR 0-1 days]. RVSWIRHC was calculated as: (mean pulmonary arterial pressure (mPAP) – mean right atrial pressure (mRAP)) x stroke volume index (SVI)RHC. Four methods for RVSWIECHO were evaluated: RVSWIECHO-1 = Tricuspid regurgitant maximum pressure gradient (TRmaxPG) x SVIECHO, RVSWIECHO-2=(TRmaxPG-mRAPECHO) x SVIECHO, RVSWIECHO-3 = TR mean gradient (TRmPG) x SVIECHO and RVSWIECHO-4=(TRmPG–mRAPECHO) x SVIECHO. Vena cava inferior diameter and its collapsibility were used for estimation of mRAPECHO.. Pearson’s correlation coefficients were used and data was expressed as mean ± standard deviation. Results Mean RVSWIRHC was 1132 ± 352 mmHg*mL*m-2. Mean RVSWIECHO-1-4 was: 1904 ± 568, 1732 ± 531, 1090 ± 366 and 918 ± 336 mmHg*mL*m-2, respectively. There was no significant difference between RVSWIRHC and RVSWIECHO-3 in mean values, although they exhibited the lowest correlation, but moderate (r = 0.66). The strongest correlation was demonstrated for RVSWIECHO-2 (r = 0.78, p < 0.001), followed by a moderate correlation for RVSWIECHO-1 and RVSWIECHO-4 (r = 0.75 and r = 0.69, p < 0.001). The absolute (relative) bias for RVSWIECHO-1 was -772 ± 385 (-50 ± 20%) mmHg*mL*m-2, RVSWIECHO-2 -600 ± 339 (-41 ± 20%) mmHg*mL*m-2, RVSWIECHO-3 42 ± 286 (5 ± 25%) mmHg*mL*m-2 and for RVSWIECHO-4 214 ± 273 (23 ± 27%) mmHg*mL*m-2. Conclusion RVSWIECHO-3, using the mean tricuspid gradient and SVIECHO, showed no significant difference to RVSWIRHC, albeit a moderate correlation between the methods. RVSWIECHO-1, RVSWIECHO- 2 and RVSWIECHO-4 exhibited moderate to strong correlations to RVSWIRHC, but poor concordance between absolute values. The clinical utility of RVSWIECHO and RVSWIRHC in assessing RV function in relation to PAH prognosis, risk stratification and treatment response remains to be evaluated in a larger clinical context.