Background
The impact of pulmonary hypertension (PHT) on right ventricular and left ventricular (LV) function in children with PHT is unknown, and echocardiographic data combining conventional and Doppler tissue imaging (DTI) on PHT in children are sparse.
Methods
Forty-one children (18 male; mean age, 7.9 ± 5.6 years) with PHT and structurally normal hearts (27 with idiopathic PHT, 14 with associated PHT) and 44 age-matched healthy controls were assessed using conventional echocardiography and DTI.
Results
Children with PHT had enlarged tricuspid valve diameters, right atrial areas, pulmonary artery dimensions, and LV eccentricity indices. In addition, pulmonary acceleration time and tricuspid annular plane systolic excursion were significantly reduced in patients compared with controls. DTI revealed that children with PHT had significantly lower systolic (S) and early diastolic (E) velocities at the tricuspid and septal levels. Despite preserved LV ejection fractions, left lateral free wall systolic velocities were significantly reduced in patients with PHT. Significantly reduced LV rapid filling velocities (E) suggested an underloaded left ventricle or LV diastolic dysfunction in children with PHT compared with controls. Pulmonary acceleration time and tricuspid annular plane systolic excursion correlated best with DTI systolic tricuspid and septal velocities.
Conclusions
Despite not being evident on conventional two-dimensional echocardiography, LV systolic performance appears to be impaired in children with PHT. Quantitative DTI assessment of ventricular function and ventricular-ventricular interactions in this setting might provide further insights into the mechanisms leading to end-stage PHT and may guide clinicians to optimize antifailure treatment.
Despite advanced specific therapies, pulmonary hypertension (PHT) remains a chronic progressive disease with significant mortality. Patients present with deteriorating exercise tolerance, syncope, or symptoms related to worsening right-heart failure. Eventually, it is right ventricular (RV) performance that most likely determines exercise tolerance and prognosis. Clinical symptoms are related to RV failure, which is often a leading cause of death in this population. However, because of the complex morphology and myocardial architecture as well as dependence on loading conditions, objective echocardiographic assessment of RV function on the basis of conventional volumetric variables such as ejection fraction (EF) remains a challenge. Cardiac magnetic resonance (CMR) imaging, with its wide field of view and no anatomic plane restriction, could overcome some of these limitations. However, CMR imaging requires general anesthesia or heavy sedation in young children, thus limiting its usefulness in this setting. Echocardiographic measures of longitudinal contractility, using tricuspid annular plane systolic excursion (TAPSE) on M-mode echocardiography, or Doppler tissue imaging (DTI) measures of annular and septal basal myocardial segments, represent alternatives and may be especially useful in assessing RV systolic function. In contrast to the left ventricle, which ejects a significant proportion of its stroke volume as a result of torsional shape changes, RV stroke volume grossly depends on longitudinal shortening. Anatomic studies have demonstrated that the deeper RV muscle fibers are predominantly arranged in a longitudinal fashion from the tricuspid valve annulus to the apex. Thus, longitudinal tricuspid annular motion and velocity closely reflect the RV free wall function and may precede circumferential RV dysfunction. Parameters of longitudinal RV function can be assessed quantitatively on DTI. Furthermore, RV dilatation, hypertrophy, and dysfunction affect left ventricular (LV) diastolic and systolic function. Bowing of the interventricular septum toward the left ventricle can distort LV geometry and negatively affect LV function due to myocardial fibers shared between the ventricles and limited pericardial space. The LV eccentricity index (LVEI) measures the departure of the left ventricle from a circular shape in a short-axis view. Beyond affecting ventricular function, altered LV geometry may also violate geometric assumptions underlying the calculation of LV EF, for example by using the Teichholz equation. Tissue Doppler echocardiography has been demonstrated to detect LV systolic and diastolic dysfunction in various settings, and it has been argued that longitudinal ventricular function may be more sensitive in detecting early myocardial damage compared with conventional measures such as EF.
In this study, we aimed to evaluate LV and RV longitudinal function, assess ventricular-ventricular interaction, and test the hypothesis that because of RV pressure overload, LV function may also be impaired due to underfilling and altered geometry in children with PHT. We therefore set out to investigate ventricular function using both conventional echocardiographic techniques and DTI in this pediatric cohort.
Methods
We assessed 41 children (18 male; mean age, 7.9 ± 5.6 years) with PHT and structurally normal hearts using echocardiographic techniques, including DTI. A structurally normal heart was defined as the absence of any intracardial or extracardial shunt lesion or relevant valve abnormality, with the exception of tricuspid regurgitation (TR) and/or at most moderate pulmonary regurgitation, commonly seen in PHT. In addition, 44 age-matched healthy controls (23 male; mean age, 7.7 ± 4.1 years; P vs patients = .82) were investigated. All control subjects recruited were healthy (eg, siblings of patients), had no medical histories of heart disease, and were not treated with any cardiac medications. A careful routine echocardiographic investigation was performed to exclude any structural abnormalities. Patient demographics are presented in Table 1 .
| Variable | Controls ( n = 44) | Patients with PHT ( n = 41) | P |
|---|---|---|---|
| Age (y) | 7.7 ± 4.1 | 7.9 ± 5.6 | .82 |
| Height (cm) | 126.1 ± 23.7 | 117.6 ± 34.2 | .20 |
| Weight (kg) | 31.1 ± 16.5 | 26 ± 17.5 | .14 |
| Heart rate (beats/min) | 97 ± 28 | 105 ± 23 | .43 |
Twenty-seven children had idiopathic PHT, whereas 14 children had structurally normal hearts and associated PHT (interstitial and chronic lung disease, n = 6; connective tissue or autoimmune disease, n = 5; late closure of a persistent arterial duct, n = 2; drug associated, n = 1).
The presence of significant PHT was confirmed on cardiac catheterization according to current guidelines (i.e., mean pulmonary artery pressure ≥ 25 mm Hg at rest) in 31 children. In the remaining 10 children, PHT was diagnosed on the basis of two-dimensional (2D) echocardiographic evaluation (showing RV dilatation and dysfunction), Doppler estimation of RV systolic pressures (TR velocity > 3 m/sec was used as a threshold above which a patient was considered to have PHT), and routine clinical assessment. All patients were on advanced PHT-specific therapies.
Conventional Echocardiography
All echocardiographic studies were performed in children using Vivid 7 machines (GE Healthcare, Milwaukee, WI), equipped with matrix-array 3.5-MHz and 7-MHz probes. On 2D echocardiography, we measured right and left atrial areas (apical four-chamber view) and the diameters of the tricuspid, mitral (apical four-chamber view), pulmonary (tilted parasternal long-axis view), aortic valve (parasternal long-axis view), and branch pulmonary arteries (high parasternal short-axis view). To quantitatively determine the degree of interventricular septal shift, we measured the LVEI from a parasternal short-axis view, using the ratio of the diameter of the LV cavity perpendicular to the interventricular septum and its diameter on a perpendicular view, as described previously. From M-mode parasternal long-axis views, RV and LV end-diastolic and end-systolic dimensions and interventricular septal as well as posterior LV free wall measures were obtained. EF was assessed using the Teichholz formula, as described in previous reports. An M-mode measurement from an apical four-chamber view of the tricuspid valve annulus was performed to assess TAPSE.
Pulmonary artery pressures and RV pressures were estimated according to the modified Bernoulli equation in patients in whom complete Doppler envelopes of the TR could be obtained. Complete Doppler envelopes could be obtained in 36 of 41 patients (88%). In all patients, right atrial pressures were estimated on the basis of 2D measurements of the diameter of the inferior vena cava and its collapsibility during inspiration, as suggested by Lang et al. We found, however, that adding the estimated right atrial pressures to RV systolic pressures (estimated from TR Doppler velocities) did not alter any correlations relevantly, while potentially introducing an additional source of error. We have thus not included the data in the statistical analysis. Pulmonary acceleration time (PAcT) was determined from the pulmonary artery pulsed Doppler trace signal.
DTI
DTI spectral analysis velocity measurements were performed at the tricuspid valve level of the RV free wall, the mitral level of the LV free wall, and the level of the interventricular septal crest, as illustrated in Figure 1 . Three cardiac cycles were recorded, and measures at the tricuspid valve level were taken in breath-hold in all cooperative children. Imaging was optimized to acquire a tissue Doppler frame rate of >160 frames/sec.
We aimed to compare DTI data with conventional echocardiographic parameters reflecting pulmonary artery pressures (PAcT, TR velocity, TR severity [0 = none, 1 = trivial, 2 = mild, 3 = moderate, 4 = severe]) and RV functional parameters such as TAPSE, semiquantitative RV function (1 = normal, 2 = mildly impaired, 3 = moderately impaired, 4 = severely impaired), and LVEI as a quantitative measure of the degree of interventricular deviation, caused by the interaction of ventricular pressure-loading conditions.
Statistical Analysis
All values are expressed as mean ± SD. Comparisons between groups were made using nonparametric tests, Mann-Whitney U tests, or χ 2 tests as appropriate. Variables were tested for normality using the Kolmogorov-Smirnov test. Where there was evidence of significant deviation from the normal distribution, a log transformation was applied, and the parameters were retested for normality. For all analyses, a two-sided P value < .05 was considered statistically significant. MedCalc version 8.1 (MedCalc Software, Mariakerke, Belgium) and R version 2.12.1 (R Project for Statistical Computing, Vienna, Austria) were used for statistical analysis.
Results
Data on gender, age, weight, height, and heart rate for healthy controls and children with PHT are shown in Table 1 . Control subjects were taller and had higher body weights compared with patients with PHT, although this did not reach statistical significance ( P = .20 and P = .14, respectively). In addition, controls had higher resting heart rates compared with patients with PHT, but this difference was also not statistically significant.
Conventional Echocardiographic Measures of Ventricular Function and Dimensions
Planimetric measures of right and left atrial areas, pulmonary artery dimensions, tricuspid and mitral valve diameters, and diastolic right intraventricular diameter were significantly different between patients and controls, as shown in Table 2 . In addition, PAcT was shorter, and TAPSE was significantly decreased in patients. In the left ventricle, the LVEI was also found to differ significantly from normal healthy controls, as shown in Table 2 and illustrated in Figure 2 . The mean LV EF (calculated using the Teichholz formula) was higher in the children with PHT compared with normal children (66.3 ± 16.6% vs 58.7 ± 12.6%), but this did not reach statistical significance ( P = .10; Table 2 ).
| Variable | Controls ( n = 44) | Patients with PHT ( n = 41) | P |
|---|---|---|---|
| RA area | 9.6 ± 2.5 | 12.4 ± 6.2 | .02 |
| LA diameter | 9.2 ± 2.8 | 7.6 ± 3.7 | .044 |
| TV diameter | 2.15 ± 0.48 | 2.57 ± 0.75 | .005 |
| PV diameter | 1.85 ± 0.36 | 2.06 ± 0.65 | .10 |
| Right PA | 1.1 ± 0.3 | 1.3 ± 0.5 | .002 |
| Left PA | 1.1 ± 0.3 | 1.4 ± 0.5 | .002 |
| RVIDd (cm) | 1.83 ± 0.72 | 2.73 ± 1.29 | .0002 |
| RVIDs (cm) | 1.5 ± 0.52 | 3.6 ± 6.5 | .002 |
| TR velocity (m/sec) | 2 ± 0.4 | 4.5 ± 1 | <.0001 |
| PA acceleration (msec) | 119.7 ± 31.8 | 65.3 ± 22.7 | <.0001 |
| MV diameter | 2.10 ± 0.45 | 1.83 ± 0.59 | .03 |
| LVEId | 1 ± 0 | 1.6 ± 0.5 | <.0001 |
| LVEIs | 1.1 ± 0.1 | 2.1 ± 0.8 | <.0001 |
| TAPSE (cm) | 1.9 ± 0.2 | 1.4 ± 0.3 | <.0001 |
| EF (%) | 58.7 ± 12.6 | 66.3 ± 16.6 | .10 |
Tissue Doppler Measures of LV and RV Function
Children with PHT had lower systolic (S) and early diastolic (E) velocities at the tricuspid and septal levels ( Table 3 ). Despite preserved LV function on conventional M-mode echocardiography, we found significantly reduced tissue Doppler LV velocities (S, E, and A) at the lateral mitral valve annular level in children with PHT ( Table 3 , Figure 2 ).
| Variable | Controls ( n = 44) | Patients with PHT ( n = 41) | P |
|---|---|---|---|
| TV | |||
| S (m/sec) | 0.13 ± 0.02 | 0.11 ± 0.03 | <.0001 |
| E (m/sec) | 0.19 ± 0.15 | 0.12 ± 0.04 | <.0001 |
| A (m/sec) | 0.11 ± 0.02 | 0.11 ± 0.03 | .25 |
| IVS | |||
| S (m/sec) | 0.08 ± 0.01 | 0.06 ± 0.02 | <.0001 |
| E (m/sec) | 0.14 ± 0.02 | 0.07 ± 0.03 | <.0001 |
| A (m/sec) | 0.07 ± 0.01 | 0.06 ± 0.01 | .10 |
| MV | |||
| S (m/sec) | 0.09 ± 0.02 | 0.06 ± 0.02 | <.0001 |
| E (m/sec) | 0.18 ± 0.05 | 0.11 ± 0.04 | <.0001 |
| A (m/sec) | 0.07 ± 0.02 | 0.06 ± 0.02 | .0015 |
Correlations between Conventional Echocardiography and DTI
In the present study, higher values for systolic tissue Doppler (S) velocities at the tricuspid and septal levels were associated with longer PAcT (a marker of less severe pulmonary artery hypertension), as shown in Figure 3 . Systolic tissue Doppler (S) velocity at the tricuspid annular level was also associated with higher TAPSE values. At the mitral valve level, no relationship was found between S-wave velocity and PAcT. In contrast, shorter PAcT and higher TR peak Doppler velocity were related to lower early diastolic (E) tissue Doppler velocity, suggesting a higher degree of diastolic dysfunction in patients with more severe PHT ( r = 0.62, P = .0002, and r = 0.64, P = .0001 for E-wave velocity at the lateral mitral annular level and the interventricular septal position, respectively). In addition, systolic LVEI was correlated with early diastolic (E) tissue Doppler velocity at the septal and lateral mitral annular levels, indicating a relationship between LV shape and diastolic function ( r = −0.48, P = .007, and r = −0.62, P = .0002, for the lateral mitral annular level and the interventricular septal position, respectively).
Tissue Doppler velocities of systolic and diastolic function measured at the interventricular septum were significantly associated with semiquantitative RV function (S: r = −0.34, P = .03; E: r = −0.63, P < .0001; A: r = −0.39, P = .01), whereas no such association could be established at the tricuspid valve or mitral level.
Results
Data on gender, age, weight, height, and heart rate for healthy controls and children with PHT are shown in Table 1 . Control subjects were taller and had higher body weights compared with patients with PHT, although this did not reach statistical significance ( P = .20 and P = .14, respectively). In addition, controls had higher resting heart rates compared with patients with PHT, but this difference was also not statistically significant.
Conventional Echocardiographic Measures of Ventricular Function and Dimensions
Planimetric measures of right and left atrial areas, pulmonary artery dimensions, tricuspid and mitral valve diameters, and diastolic right intraventricular diameter were significantly different between patients and controls, as shown in Table 2 . In addition, PAcT was shorter, and TAPSE was significantly decreased in patients. In the left ventricle, the LVEI was also found to differ significantly from normal healthy controls, as shown in Table 2 and illustrated in Figure 2 . The mean LV EF (calculated using the Teichholz formula) was higher in the children with PHT compared with normal children (66.3 ± 16.6% vs 58.7 ± 12.6%), but this did not reach statistical significance ( P = .10; Table 2 ).
| Variable | Controls ( n = 44) | Patients with PHT ( n = 41) | P |
|---|---|---|---|
| RA area | 9.6 ± 2.5 | 12.4 ± 6.2 | .02 |
| LA diameter | 9.2 ± 2.8 | 7.6 ± 3.7 | .044 |
| TV diameter | 2.15 ± 0.48 | 2.57 ± 0.75 | .005 |
| PV diameter | 1.85 ± 0.36 | 2.06 ± 0.65 | .10 |
| Right PA | 1.1 ± 0.3 | 1.3 ± 0.5 | .002 |
| Left PA | 1.1 ± 0.3 | 1.4 ± 0.5 | .002 |
| RVIDd (cm) | 1.83 ± 0.72 | 2.73 ± 1.29 | .0002 |
| RVIDs (cm) | 1.5 ± 0.52 | 3.6 ± 6.5 | .002 |
| TR velocity (m/sec) | 2 ± 0.4 | 4.5 ± 1 | <.0001 |
| PA acceleration (msec) | 119.7 ± 31.8 | 65.3 ± 22.7 | <.0001 |
| MV diameter | 2.10 ± 0.45 | 1.83 ± 0.59 | .03 |
| LVEId | 1 ± 0 | 1.6 ± 0.5 | <.0001 |
| LVEIs | 1.1 ± 0.1 | 2.1 ± 0.8 | <.0001 |
| TAPSE (cm) | 1.9 ± 0.2 | 1.4 ± 0.3 | <.0001 |
| EF (%) | 58.7 ± 12.6 | 66.3 ± 16.6 | .10 |
Tissue Doppler Measures of LV and RV Function
Children with PHT had lower systolic (S) and early diastolic (E) velocities at the tricuspid and septal levels ( Table 3 ). Despite preserved LV function on conventional M-mode echocardiography, we found significantly reduced tissue Doppler LV velocities (S, E, and A) at the lateral mitral valve annular level in children with PHT ( Table 3 , Figure 2 ).
