2014 Volume 78 Issue 8 Pages 1967-1973
Background: Many studies have investigated tricuspid annular plane systolic excursion (TAPSE) as a longitudinal right ventricular (RV) contraction. The aim of this study was to clarify the mechanism of RV systolic function compared with longitudinal and radial RV contractions in healthy children.
Methods and Results: A total of 815 consecutive healthy children and adolescents were enrolled. We measured TAPSE on M-mode echocardiography as a longitudinal RV contraction. RV wall displacement (RVWD) toward the center of the left ventricle (LV) was measured in the short-axis view on M-mode echocardiography. RV stroke volume (RVSV) was obtained using pulse Doppler echocardiography as an indicator of RV global systolic function. RVSV and TAPSE had a positive but non-linear correlation with a definite inflection point around 15 mm of TAPSE. Subjects were stratified into 2 groups according to TAPSE (≤15 mm or >15 mm). In subjects with TAPSE ≤15 mm, RVWD and TAPSE were identified as independent predictors of RVSV. In contrast, in subjects with TAPSE >15 mm, TAPSE were identified as an independent predictor of RVSV.
Conclusions: Normal RV contraction pattern shifts from radial to longitudinal directions at approximately 15 mm of TAPSE. RVSV is primarily generated by longitudinal contraction, but in neonates, RVSV is supported not only by longitudinal contraction but also by radial contraction. (Circ J 2014; 78: 1967–1973)
Although many studies involving right ventricular (RV) function have been reported, the complex geometry of the RV precludes the understanding of its function.1 Many studies have shown that tricuspid annular plane systolic excursion (TAPSE) – represented as a longitudinal RV contraction – is important for evaluating RV systolic function.2 The role of radial RV contraction, however, has not been well documented. Pettersen et al reported that a longitudinal RV contraction was predominant under normal conditions, but that circumferential RV contraction was predominant in response to increased afterload.3 The RV contraction pattern is therefore expected to change in response to the changing RV hemodynamics, particularly in the neonatal period because RV hemodynamics undergoes great change during this period.
The aim of this study was to elucidate the mechanism underlying RV systolic function compared with that of longitudinal and radial RV contraction in healthy neonates and children.
We examined 815 consecutive children and adolescents without heart disease, ranging from newborns to 22.2 years of age (mean age, 4.4±4.0 years). All subjects who were referred for heart murmur, chest pain, or those with a history of Kawasaki disease underwent complete physical examination, electrocardiogram, and routine echocardiography. Exclusion criteria in this study were as follows: (1) irregular heart rhythm; (2) intraventricular conduction disturbance such as complete or incomplete right bundle branch block; (3) coronary arterial involvement in patients with a history of Kawasaki disease; (4) pulmonary stenosis with peak-flow velocity >2.0 m/s; (5) moderate-severe pulmonary or tricuspid regurgitation; or (6) history of open-heart surgery. Patients with a history of Kawasaki disease who were within 6 months after onset of disease were also excluded from this study even if they did not have any cardiac involvement because we were concerned about prolonged myocarditis due to Kawasaki disease.4
All patients were diagnosed as anatomically normal with normal RV and left ventricular (LV) function after a complete examination. In patients who complained of chest pain, the causes of chest pain were confirmed to be non-cardiac origin such as intercostal neuralgia on complete medical examination. All patients or their parents provided informed consent for this study. This study complied with all institutional guidelines relating to patient confidentiality and research ethics, including institutional review board approval.
General Echocardiography and TAPSEWe used EUB-6000 (Hitachi Medical, Tokyo, Japan) for echocardiography with a 7–3-MHz or 4–2-MHz phased array sector probe. All subjects underwent echocardiography in the supine position. If subjects were uncooperative, sedation was provided as oral 10% triclofos sodium syrup. After obtaining routine echocardiographic data, we measured TAPSE as a longitudinal RV contraction.5
We used M-mode scanning to measure tricuspid lateral annulus through several cycles (Figure 1A). TAPSE was measured as the distance between the peak and the bottom of the M-mode tracing curve, and at least 4–5 consecutive beats were averaged. To investigate the influence of LV longitudinal motion to TAPSE, mitral annular systolic excursion (MAPSE) was measured.
M-mode measurements of longitudinal and radial RV contractions. (A) Longitudinal RV contraction was displayed as TAPSE and (B) radial RV contraction was displayed as radial RVWD. RV/LV diameter was calculated as RVD/LVD. LV, left ventricle; LVD, left ventricular end-diastolic diameter; MV, mitral valve; RV, right ventricle; RVAW, right ventricular anterior wall; RVD, right ventricular end-diastolic diameter; RVWD, right ventricular wall displacement toward the center of the LV; TAPSE, tricuspid annular plane systolic excursion; TV, tricuspid valve.
RVWD was measured in the same short-axis view as the LV at the papillary muscle level, using the M-mode. RVWD is the RV anterior wall (RVAW) displacement toward the center of LV and is in the orthogonal direction of TAPSE. We set the M-mode scan line on the center of the LV avoiding the RV outflow tract. We measured RVWD as the distance between the peak and the bottom of the M-mode tracing curve of RVAW, averaging at least 4–5 consecutive beats (Figure 1B). The ratio of radial to longitudinal RV contractions was calculated as RVWD/TAPSE.
RV/LV Diameter RatioThe ratio of RV/LV diameter was measured for evaluating RV size among children with various body sizes (Figure 1B).6,7
Standard RV Performance and Pulmonary Vascular ResistanceWe used RV stroke volume (RVSV) as a geometry-independent indicator of RV performance. RVSV was determined as the product of the cross-sectional area and the velocity time integral (VTI) over systole with the use of the following equation:
RVSV = 1/4 × π × VTI × D2
where D is the midsystolic diameter of the pulmonary valve.8
To evaluate pulmonary vascular resistance, pulmonary acceleration time (AcT) was measured from the pulmonary flow velocity profile obtained on pulsed Doppler. AcT was measured from the beginning of the ejection to the peak of the flow profile. AcT/ET was also calculated by dividing AcT by RV ejection time (ET; Figure 2).
Measurements of right ventricular (RV) stroke volume and pulmonary vascular resistance. AcT, acceleration time; D, midsystolic diameter of pulmonary valve; ET, ejection time; LPA, left pulmonary artery; MPA, main pulmonary artery; RPA, right pulmonary artery; RVOT, right ventricular outflow tract; VTI, velocity time integral.
In all subjects, body length and body weight were measured and body surface area (BSA) was calculated using the Haycock formula.
Statistical AnalysisStatistical analysis was done using Stat View 5.01 (SAS Institute, Cary, NC, USA). All data are expressed as mean ± SD. Linear regression was used to evaluate the correlation between RVSV and TAPSE or BSA. A change point regression analysis was used to identify the optimal splitting point of the linear regression line.9 Analysis of covariance was used to compare 2 regression slopes. To study the strongest predictors for RVSV, variables such as TAPSE, MAPSE, RV/LV ratio and PA Act/ET were examined using multiple regression analysis. P<0.05 was considered significant.
There was no gender difference of TAPSE (male, 19.1±4.2 mm; female, 19.0±4.6 mm). RVSV ranged from 4.4 to 136.6 ml (mean, 29.3±14.6 ml) and had a strong linear relationship with BSA (RVSV=39.8×BSA+2.6, r=0.89, P<0.0001; Figure 3A). In contrast, RVSV and TAPSE had a non-linear correlation with a definite inflection point at approximately 15 mm of TAPSE (arrow; Figure 3B). On change point regression analysis the optimal TAPSE splitting point was 14.5 mm for right- and left-side linear correlations. There was a linear relationship between RVSV and TAPSE ≤14.5 mm (RVSV=1.12×TAPSE–2.06, r=0.63, P<0.0001). And there was also a linear relationship between RVSV and TAPSE >14.5 mm (RVSV=3.08×TAPSE–30.51, r=0.73, P<0.0001). The slopes of these regression lines were significantly different (P<0.001). Subjects were stratified into 2 groups according to TAPSE (≤15.0 mm or >15.0 mm). Fifteen millimeters of TAPSE corresponded to approximately 4 months of age (Figure 4A). Multivariate analysis identified predominant predictors for RVSV in subjects with TAPSE ≤15.0 mm and subjects with TAPSE >15.0 mm, respectively (Table). In subjects with TAPSE ≤15.0 mm, RVWD and TAPSE were identified as independent predictors of RVSV. In contrast, the influence of RVWD on RVSV became smaller and TAPSE predominated in subjects with TAPSE >15.0 mm.
(A) RVSV vs. BSA; (B) RVSV vs. TAPSE. There was a non-linear relationship between TAPSE and RVSV. The reflection point was noted at around 15 mm of TAPSE (solid arrow). BSA, body surface area; RVSV, right ventricular stroke volume; TAPSE, tricuspid annular plane systolic excursion.
Age-dependent change of (A) TAPSE, (B) RVWD and (C) RVWD/TAPSE. RVWD, right ventricular wall displacement toward the center of the left ventricle; TAPSE, tricuspid annular plane systolic excursion.
| Variables | TAPSE | |||
|---|---|---|---|---|
| ≤15 mm | >15 mm | |||
| β | P-value | β | P-value | |
| TAPSE (mm) | 0.414 | <0.005 | 0.496 | <0.0001 |
| MAPSE (mm) | 0.125 | 0.334 | 0.367 | <0.0001 |
| RVWD (mm) | 0.206 | <0.05 | 0.054 | 0.0746 |
| RV/LV | –0.081 | 0.405 | –0.009 | 0.767 |
| PA AcT/ET | 0.045 | 0.588 | 0.011 | 0.684 |
AcT, acceleration time; ET, ejection time; LV, left ventricle; MAPSE, mitral annular plane systolic excursion; PA, pulmonary artery; RV, right ventricle; RVSV, right ventricular stroke volume; RVWD; right ventricular wall displacement; TAPSE, tricuspid annular plane systolic excursion.
Figure 4 shows age-dependent change of TAPSE, RVWD and RVWD/TAPSE. TAPSE increased rapidly after birth until approximately 4 months of age, and after that increased slowly and linearly with age (Figure 4A). RVWD remained constant, between 0 and 8 mm, independent of aging (Figure 4B). RVWD/TAPSE was high at birth and showed a rapid decrease until approximately 4 months of age. After that, RVWD/TAPSE remained constant, between 0 and 0.5, independent of aging (Figure 4C).
Figure 5A shows the relationship between RVWD and TAPSE. RVWD ranged from 3.35 to 11.6 mm in subjects with TAPSE ≤15.0 mm (5.56±1.3 mm) and from 0.15 to 11.8 mm in subjects with TAPSE >15.0 mm (4.40±1.79 mm). There was a significant mean difference of RVWD between them (P<0.0001). The RVWD/TAPSE ratio was inversely related to TAPSE (Figure 5B). The predominance of radial RV contraction in small TAPSE shifts to predominance of longitudinal RV contraction with increasing TAPSE.
(A) RVWD vs. TAPSE; (B) RVWD/TAPSE vs. TAPSE. RVWD, right ventricular wall displacement toward the center of the left ventricle; TAPSE, tricuspid annular plane systolic excursion.
Interobserver variability was tested for TAPSE and RVWD on 10 random patients between the first observer (I.H.) and second observer (K.W.). The mean interobserver difference was 0.67±2.7 mm for TAPSE and 0.39±1.6 mm for RVWD. There was no significant difference in TAPSE and RVWD between 2 observers.
Kawut et al recently reported that RV volume and hypertrophy are associated with the risk of heart failure and cardiovascular death.10 Although evaluating RV function and structure is important, the complex geometry of RV precludes accurate analysis. Three-dimensional analysis using magnetic resonance imaging (MRI) is geometry independent and has been regarded as the gold standard for RV studies.10,11 Echocardiography is not as accurate as MRI, but gives us a real-time image for diagnosis with a high frame rate and is, furthermore, applicable at the bedside. Among many parameters for assessing RV function, such as TAPSE, RV ejection fraction (RVEF), and myocardial performance index, TAPSE proved to be the most reliable and reproducible index.12,13 Although some investigators have reported that TAPSE significantly correlates with RVEF and contributes to approximately 80% of RV output, others have reported that TAPSE was affected by LV function and also by pulmonary condition – resulting in a lower correlation with RVEF.14–18 We therefore hypothesized that RVSV was generated by longitudinal and radial RV contraction with changes of myocardial fiber structure and contraction pattern.
Despite the strong linear correlation of RVSV with BSA, RVSV did not linearly correlate with increasing TAPSE. A definite inflection point was noted at approximately 15 mm of TAPSE. A change in RV contraction pattern is expected to occur at approximately 15 mm of TAPSE. Pettersen et al reported in their study using echo strain that a longitudinal RV contraction was predominant under normal conditions, but circumferential RV contraction was predominant in response to hemodynamic change.3 Several reasons are considered to explain a change of contraction pattern.
First, the RV is located in the thoracic cavity and the tricuspid valve motion is limited in its space. The size of the thoracic cavity increases with growth and the range of tricuspid valve motion also increases corresponding to the growth of thoracic cavity. In general, the body size of the newborn baby is smallest at that point and rapidly increases until 1 or 2 years of age.19 Growth speed, however, changes slowly after that. RV longitudinal motion is therefore limited by thoracic size in the neonatal period. Accordingly, radial RV contraction is predominant for generating RVSV in the neonatal period. With growth of the thoracic cavity, RV contraction pattern shifts from radial to longitudinal direction. In adults, TAPSE has been reported to have a good correlation with RVEF because RVSV is primarily generated by longitudinal contraction. TAPSE, however, does not correlate with RVEF under serious conditions such as pulmonary hypertension and severe right heart failure because radial RV contraction predominates.20,21 The present study has shown that although RVSV is supported not only by longitudinal contraction but also by radial contraction in neonates, longitudinal RV contraction subsequently plays a major role in the generation of RV output. We therefore conclude that there exists an alternation of contraction pattern between TAPSE ≤15 mm and >15 mm.
Second, MAPSE, which represents longitudinal LV contraction, was associated with RVSV for TAPSE >15 mm.22 The RV–LV interaction is thus considered to affect RV function.23–25
RVSV as Geometry-Independent Indicator of RV PerformanceWe used RVSV as a geometry-independent indicator of RV performance in this study. The study confirmed that RVSV had a strong correlation with BSA. BSA is a more important determinant of the size of each of the cardiovascular structures than age, height, or weight alone.26 RVSV obtained with our method may not be equal to the absolute value of pulmonary stroke volume. To more accurately measure the pulmonary flow on Doppler echocardiography, the pulmonary outflow track should be scanned in 2 orthogonal planes, or other techniques should be used.27,28 The aim of this study was not to obtain the absolute value of pulmonary flow. RVSV obtained in this study fitted the linear relationship predicted by BSA and is considered to be sufficiently valid and accurate for data analysis.
Advantage of TAPSE and RVWD for RV Systolic FunctionTAPSE, which is tricuspid annular motion, is considered to represent a global RV contraction and is recommended in clinical use for evaluation of RV systolic function in the American Society of Echocardiography guidelines.29 Compared to other echocardiographic parameters such as myocardial performance index, fractional area change and so on, TAPSE has higher reproducibility.30,31 Furthermore, measurement of this is very simple and practical with any echo machine. In contrast, RVWD was devised by us, and its advantage has not been studied sufficiently. Further investigation is needed to clarify the clinical usefulness of RVWD.
Normal RV contraction pattern shifts from radial to longitudinal directions at approximately 15 mm of TAPSE, which corresponds to approximately 4 months of age. RVSV is primarily generated by longitudinal contraction, but, in neonates, RVSV is supported not only by longitudinal contraction but also by radial contraction.
No conflicts of interest to disclose.
