Quantitative assessment of right ventricular volumes using 3D echocardiography in right heart - involving conditions: validation with cardiac magnetic resonance

Background right ventricular (RV) volumes and function are strong prognostic markers in right heart diseases. Although cardiac magnetic resonance (CMR) remains the gold standard for such quantification, its limited availability and cost restrict its use for serial follow-up. However, 2D echocardiography, while widely available, is inadequate for volumetric assessment due to the complex geometry of the RV. In this context, 3D echocardiography (3DE), is arising to assess RV volumes and function, even thouugh in pathological settings, especially with marked RV dilatation, its role remains uncertain. Purpose To evaluate the reliability of 3DE-based RV volume measurements compared to CMR gold-standard, across a spectrum of right heart-affecting conditions. Methods 41 subjects with conditions affecting RV, were prospectively enrolled in our analysis, including repaired tetralogy of Fallot and congenital pulmonary stenosis, intracardiac shunts, pulmonary embolism, pulmonary hypertension, tricuspid valve disease, carcinoid syndrome, nonischemic dilated cardiomyopathy and athlete’s heart. 3DE and CMR were performed on the same day to minimize the variability related to loading conditions, especially preload. RV end-diastolic volume (RVEDV), end-systolic volume (RVESV), ejection fraction (RVEF) and stroke volume (RVSV) were measured using 3DE, by two expert operators with offline semiautomated software, and then compared to CMR values. Correlation, agreement and systematic bias regarding 3DE and CMR measurements were assessed using Pearson’s r, intraclass correlation coefficients (ICC) and Bland–Altman analysis, respectively. Reproducibility of 3DE measurements was evaluated through intra- and inter-observer ICCs. Results 3DE showed good correlation and agreement with CMR for RVEDV and RVESV (Pearson’s r 0.86 [p<0.001] and 0.85 [p<0.001], respectively; ICC 0.74 [0.152–0.901] and 0.81 [0.602–0.902], respectively) and moderate for RVSV and RVEF (Pearson’s r 0.79 [p<0.001] and 0.71 [p<0.001], respectively; ICC 0.61 [0.039–0.836] and 0.66 [0.419–0.806]). Mean bias and percent mean bias were respectively for RVEDV, RVESV, RVSV, and RVEF -31.38 [-41.12 to -21.64] and -15.33% [-20.09 to -10.57], -10.79 [-17.15 to -4.423] and -10.39% [-16.53 to -4.262], -20.6 [-26.93 to -14.27] and -20.41% [-26.69 to -14.14], -2.37 [-4.09 to -0.65] and -4.78% [-8.24 to -1.32] (figure and table). Intra- and inter-observer reproducibility for RVEDV and RVESV was excellent (ICC > 0.92, for both measurements in each comparison), with good reproducibility for RVSV (ICC 0.72 and 0.83, respectively). Conclusion 3DE correlates well with CMR for measuring RV size and EF, though it slightly underestimates volumes by 10-20%. It’s reproducible, more accessible, and cheaper, making it a good option for monitoring patients with RV-related diseases.Bias and agreement analysis Correlation and bias analysis