Three-Dimensional Echocardiographic Evaluation of Longitudinal and Non-Longitudinal Components of Right Ventricular Contraction Results from the World Alliance of Societies of Echocardiography Study

Study Design and Population

The WASE Study rationale was described in detail elsewhere ^{7, 8}. A comprehensive 2D and 3D TTE examination was performed in all study subjects using commercially available ultrasound imaging systems (GE, Philips, and Siemens). Image acquisition followed a standard protocol established by ASE/EACVI guidelines and data analyzed by the two WASE core laboratories (University of Chicago and MedStar Health Research Institute) ⁹.

Image Acquisition and Analysis

Wide-angle 3DE full-volume RV datasets were acquired with the patient in the left decubitus position over 4-6 cardiac cycles during suspended respiration from the RV-focused view, while carefully avoiding stitching and drop-out artifacts. Data was digitally stored and analyzed offline at the 3DE Core Lab (University of Chicago) using dedicated vendor-independent software (Image Arena; “4D RV-Function”, TOMTEC, Unterschleissheim, Germany), previously validated against cardiac magnetic resonance (CMR) reference ^{9, 10}. 3DE datasets were deemed adequate for analysis if all walls were visible throughout the cardiac cycle. A minimum frame rate of 15 volumes per second was required for the software to perform the analysis.

Auto-segmentation technology with the help of specific manually identified landmarks allowed the 3D RV dataset to be displayed in both the long and short-axis cut-planes with an initial endocardial border contour suggestion. End-diastole was automatically identified as the time point in which the RV cavity was the largest, while end-systole was identified as the time-point when the cavity was the smallest. Afterwards, the user could manually adjust the proposed end-diastolic and end-systolic endocardial border contours to optimize dynamic tracking throughout the cardiac cycle. Once this step was completed, the program generated and rendered a 3D RV surface model.

Decomposition of the RV motion

Figure 1. illustrates the workflow for the decomposition of RV systolic function into its different contractile components.

• Download figure
• Open in new tab

Fig. 1.

Right Ventricular Contraction Decomposition Workflow

Statistical Analysis

All data are presented as the mean ± standard deviation (SD). Group differences were evaluated using ANOVA and unpaired two-tailed student’s t-tests. Statistical significance was defined as p<0.05. Normal ranges for each parameter were defined as upper and lower limits using 2.5th and 97.5th percentile values from the relevant group. This is in accordance with the definition of “normal” as falling within 95% of the normal population, with the remaining 5% being distributed equally between the two tails of the distribution, irrespective of whether it is Gaussian. Pearson or Spearman correlation tests were applied, as appropriate. In addition, multivariable linear regression analyses were performed to identify independent associations between the decomposed RV metrics with demographic, anthropometric, basic hemodynamic, and LV functional parameters. In order to add race, we dichotomized our cohort accordingly, analyzing White (yes/no), Black (yes/no) and Asian (yes/no) races as separate variables in each model to avoid collinearity.

1. Sex-related differences

Table 3 shows the 3D RV morphological and functional parameters of the total study population and separated by sex. RVEDVi was significantly higher in men than in women (p <0.001). Global RVEF and absolute REF, AEF, LEF values were higher in women than in men (p <0.001). However, the relative contribution of each of the individual components to the global RVEF was similar in both sexes. While GLS and GAS were significantly higher in women than in men, GCS showed no differences between sexes.

2. Age-related differences

Table 4 shows the age-related differences in the components of 3D RV systolic function in both men and women. There were no significant differences in RVEDVi between age groups. Both sexes exhibited a decline in longitudinal shortening (i.e., LEF, LEF/RVEF, and GLS, p < .001) with increasing age. In women, this reduction in LEF/RVEF led to a lower global RVEF (p < 0.001), which was incompletely compensated by an increase in REF/RVEF.

3. Race-related differences

BSA and RVEDVi were significantly smaller in Asian compared to black and white subjects (Table 5). Both black men and women showed lower RVEF, and GAS values compared to white and Asian subjects of the same sex (p <0.001). Black men showed a distinctive RV mechanical pattern, consistent of significantly higher REF/RVEF and lower LEF/RVEF values, when compared to Asian and white men. White men showed higher LEF/RVEF than both black and Asian men (p <0.001). Of note, these differences in RV contraction patterns across races were not significant in women.

4. Association with 3D LV functional parameters and independent predictors of the contractile patterns

Correlations between LEF, REF, AEF and basic demographic, anthropometric, hemodynamic and LV functional parameters are summarized in Supplementary Table 1. Notably, three-dimensional echo-derived LV EF correlated weakly with 3D RV EF (r=0.161, p<0.001), LEF (r=0.112, p=0.001), and with AEF (r=0.126, p<0.001). 3D LV GLS also correlated weakly with 3D RV EF (r= -0.158, p<0.001), 3D RV GLS (r=0.166, p<0.001) and LEF (r= -0.238, p<0.001).

Results of multivariable linear regression analyses are shown in Supplementary Tables 2,3 and 4. Beyond age, sex, and LVEF, white and black races were independent predictors of LEF, whereas sex, LVEF, and white and black races were independent predictors of AEF.

Rationale for the evaluation of non-longitudinal components of RV contraction

The RV myo-architecture and contraction are extremely complex. While the septum is characterized by oblique longitudinal and spiral fibers, the RV free wall contains predominantly transversal myofibers ¹³. Due to this unique spatial anatomical disposition of myofibers, RV contraction is the result of the combination of three distinctive contractile patterns i.e., shortening along the longitudinal axis, radial contraction of the free wall (“bellows effect”), and anteroposterior shortening as a result of left ventricular (LV) contraction.

However, all 2DE techniques used for the evaluation of RV contraction (i.e. TAPSE and S wave tissue Doppler velocity) provide information on regional longitudinal, rather than global RV function. Although 2DE RV FWS has the advantage of being a more sensitive indicator of subclinical impairment of RV contraction than conventional 2DE measurements ^{2, 14, 15}, it has been reported that in situations where RV function is not predominantly affected by reduced longitudinal contraction or where global RV function remains preserved despite decreased longitudinal motion, the sole use of 2DE RV FWS might be insufficient to characterize RV systolic function ^{16, 17}.

Whenever examiners are familiar with 3DE imaging and image quality is acceptable, it should be used the evaluation of RV function. This is supported by previous reports showing that 3DE assessment of RV volumes, EF and GLS are feasible and reproducible ¹⁸, and provides additional prognostic information in patients with diverse cardiac conditions ^{19, 20}. Moreover, constant advances and development of matrix-array transducers with higher spatial and temporal resolutions has made possible the development of dedicated software for 3DE deconstruction of RV systolic function in all its motion components. The software used in our study has previously shown excellent intra- and inter-observer reproducibility as well as a robust correlation with currently available echocardiographic software and a modest correlation with CMR ^{11, 21}. These findings and the existing evidence about the clinical and prognostic value of the measurements performed in our study, support the feasibility and potential role of this approach in enhancing our understanding of the mechanisms of RV adaptation or maladaptation to pathologic conditions.

Evidence about the clinical utility of the different components of RV systolic function

The prognostic value of the parameters included in our study has been described by Kitano et al. who showed that 3D RV GCS, GLS and GAS were significantly associated with cardiac death, ventricular tachyarrhythmia, or heart failure hospitalization at 20 months of follow-up in a cohort of patient with diverse cardiac diseases ²¹. Moreover, these findings persisted after adjusting for factors like age, renal function, LVEF and average mitral E/e’. Kaplan-Meier analysis showed that median values of 3D RV GCS, GLS and GAS stratified these patients by survival rates ²¹. Tolvaj et al have described the association between reduced RV GCS and significantly increased risk of mortality, even in scenarios where LV GLS is preserved ²².

Interestingly, 3D RV GAS was shown to be a better prognostic marker than 3DE RV GLS or GCS, most likely due to the fact that the information provided by GAS corresponds to an multiplanar evaluation of RV contraction and is not limited to either the longitudinal or circumferential directions as are GLS and GCS ²¹.

It is key to understand that the RV adaptive mechanisms vary according to the underlying pathophysiological process. These different mechanical responses can be identified only through the separate assessment of the multiple components of RV contraction. This approach has allowed to describe distinctive RV mechanical adaptations in different populations ^{21, 23–25}, including heart transplant recipients ²⁶, congenital cardiomyopathies ²⁷, and elite athletes ²⁸.

In a large cohort of patients with mildly and moderately reduced LVEF, Surkova et al. demonstrated that despite initially reduced LEF and AEF, REF increased, leading to maintain the global RV performance. Once the contribution of REF decreased, the global RV EF became severely reduced ²⁹. Notably, in patients with preserved RV EF (>45%), the AEF component has shown to be a significant and independent predictor of outcome after a median of 6.7 years follow-up ²⁹. These findings raise the question about the role of the evaluation of RV contraction beyond the longitudinal shortening in patients with LV dysfunction, considering that identifying the decrease in other motion components could be useful as early markers or RV dysfunction, follow-up parameters or indicators for treatment escalation. Accordingly, in a meta-analysis of more than 1,900 patients, Sayour et al. have elegantly demonstrated that 3D RVEF has a stronger association with adverse cardiopulmonary outcomes than conventional RV functional indices, namely TAPSE, FAC and RV FWS ³⁰.

Severe mitral regurgitation (MR) represents one of the most frequent indications for valve surgery, and its progression leads to adaptive (and maladaptive) changes in both left and right ventricles. Due to its broadly recognized prognostic implications, timely diagnosis and appropriate risk stratification represent a major clinical need ³¹. Accordingly, Tokodi et al. reported that in patients with severe primary MR and RV EF > 45%, there was a reduction in REF and a predominant contribution of LEF to the global RV contraction. However, in the immediate post-operative period, this contractile RV pattern inverted, and REF prevailed over LEF. Interestingly, after 6 months, the contribution of both components had equalized ²⁴.

These reports reinforce the notion that the impact of different hemodynamic conditions on RV systolic function might be overlooked, if the evaluation of RV contraction is only performed using conventional parameters of longitudinal deformation. The question whether these changes in RV contraction patterns might be useful as surrogates of early RV dysfunction and represent additional criteria for medical/surgical intervention, will require further prospective research in patients with specific pathology.

Contributions of the current study

Although Lakatos et al. have previously described the contribution of non-longitudinal components of RV contraction in 300 healthy volunteers from a single European and Asian institutions ¹², our study broadens this analysis to the larger, multicentric and multiethnic WASE cohort, yielding a more statistically sound assessment of the influence of sex, age, and race on these parameters. The LLN and ULN values for 3D RVEDVi and RV EF provided by our study are similar to those reported by the two largest studies of 3DE normative values of this chamber, including Maffessanti et al. ³² and Addetia et al. ³³.

However, current clinical practice lacks RV functional parameters that represent the non-longitudinal shortening of this chamber. The results of our study expands the evidence provided by previous reports, showing that REF and AEF have similar contributions to that of LEF to global RV contraction, and that these components should not be neglected when evaluating the RV function ¹². Importantly, we were able to show that the RV contraction pattern is independently associated with age, sex, and race highlighting that the established changes across the different subgroups are not just due to anthropometric (BSA) or basic hemodynamic (heart rate, blood pressure) differences. The magnitude of longitudinal and anteroposterior shortenings are coupled with LV function; however, radial shortening is not. This observation is in line with the findings of Surkova et al. who showed that these mechanical directions are deteriorating in parallel with LV systolic dysfunction ²⁹.

Our findings agree with those reported in the 2D RV systolic function analysis on the WASE population, which have shown that RV dimensions were larger in men and 2D RV functional parameters were larger in women ³⁴. Age-related changes in these 2D parameters were not uniform ³⁴ and previous studies using 3DE assessment of RV systolic function have also noted that age is weakly correlated with RVEF ³². Notably, we were able to identify an age-related reduction in the longitudinal components of RV contraction in both sexes and a characteristic increase in radial RV contraction in women with age. These observations might be related to an age-related increase in pulmonary pressures and consequent changes in RV myofiber architecture. A preserved RV EF does not preclude the presence of changes in the RV contraction patterns, and the role of REF as an early marker of elevated pulmonary pressures has been proposed ¹⁶. Muraru et al. have identified LV GLS and PASP as independent predictors of RV longitudinal performance, suggesting that the assessment of the RV adaptation to increased afterload should not be limited to the RV free wall ³⁵. Accordingly, our findings support the concept that the evaluation of RV systolic function in individuals > 65 years should not be restricted to its longitudinal motion.

The use of 3DE allowed to show the presence of higher longitudinal and anteroposterior contraction as well as lower radial contributions to global RVEF in Asian population ¹². Although noticeable differences in 3D RV volumes and EF across ethnic groups have been recently identified ³³, to the best of our knowledge, this is the first study describing distinctive RV contractile patterns in black and white men, which had been previously overlooked when only using 2DE parameters ³⁴. Our findings contribute to expand the understanding of race-related differences in echocardiographic measurements.