Objectives: Residual pulmonary hypertension challenges the right ventricular function and worsens the prognosis in heart transplant recipients. The complex geometry of the right ventricle complicates estimation of its function with conventional transthoracic echocardiography. We evaluated right ventricular function in heart transplant recipients with the use of 3-dimensional echocardiography in relation to systolic pulmonary artery pressure.
Materials and Methods: We performed 32 studies in 26 heart transplant patients, with 6 patients having 2 studies at different time points with different pressures and thus included. Right atrial volume, tricuspid annular plane systolic excursion, peak systolic annular velocity, fractional area change, and 2-dimensional speckle tracking longitudinal strain were obtained by 2-dimensional and tissue Doppler imaging. Three-dimensional right ventricular volumes, ejection fraction, and 3-dimensional right ventricular strain were obtained from the 3-dimensional data set by echocardiographers. Systolic pulmonary artery pressure was obtained during right heart catheterization.
Results: Overall mean systolic pulmonary artery pressure was 26±7 mm Hg (range, 14-44 mmHg). Three-dimensional end-diastolic (r = 0.75; P < .001) and end-systolic volumes (r = 0.55; P = .001)correlated well with systolic pulmonary artery pressure. Right ventricular ejection fraction and right atrium volume also significantly correlated with systolic pulmonary artery pressure (r = 0.49 and P = .01 for both). However, right ventricular 2- and 3-dimensional strain, tricuspid annular plane systolic excursion, and tricuspid annular velocity did not.
Conclusions: The effects of pulmonary hemodynamic burden on right ventricular function are better estimated by a 3-dimensional volume evaluation than with 3-dimensional longitudinal strain and other 2-dimensional and tissue Doppler measurements. These results suggest that the peculiar anatomy of the right ventricle necessitates 3-dimensional volume quantification in heart transplant recipients in relation to residual pulmonary hypertension.
Key words : Echocardiography, Right ventricle
Residual pulmonary vascular resistance in heart transplant recipients threatens the right ventricular (RV) function, and RV failure is a serious complication early after heart transplant.1-3 In addition, during the first year after transplant, failure of adaptation of the donor heart to pulmonary artery pressure (PAP) may lead to allograft failure.3,4 The complex geometry and interplay between longitudinal and circumferential deformations of the RV complicate the estimation of RV function in relation to residual pulmonary hypertension. Therefore, assessments of 1- or 2-dimensional longitudinal parameters incompletely reflect RV function. Three-dimensional (3D) RV volume and ejection fraction (EF) assessments by echocardiography have been proposed as a reliable method for global RV function assessment.5 The novel 3D RV quantification tools also offer the possibility to quantify RV deformation on a 3D data set.6 Accordingly, we aimed to (1) evaluate 3D RV volumes and 3D RV strain in relation to systolic pulmonary artery pressure (sPAP) in heart transplant recipients and (2) compare those with conventional 2D and Doppler parameters to test the hypothesis that RV volume assessment is a more accurate reflector of the adaptation of RV to residual pulmonary hemodynamic burden in heart transplant recipients.
Materials and Methods
Study population and protocol
All patients underwent a comprehensive transthoracic echocardiography evaluation, including 2D assessment, conventional Doppler, tissue Doppler studies, and 3D assessment of the RV function at the same day of cardiac catheterization and endomyocardial biopsy procedures. The study protocol was approved by the Institutional Ethics Committee and conformed to the ethical guidelines of the 1975 Helsinki Declaration. Written informed consent was obtained from patients.
We used a commercially available cardiac ultrasonography machine (Vivid E9; General Electric Healthcare, Horten, Norway) equipped with M5Sc and 4V transducers. Data captured from 3 cardiac cycles were digitally stored, and off-line analyses were performed on a work station (EchoPAC BT 2.0.1, General Electric Healthcare, Horten, Norway) by a blinded researcher (HK) to invasive and clinical findings of the patients. We used an RV-focused apical 4-chamber view for a comprehensive evaluation of RV function, including right atrial volume, tricuspid annular plane systolic excursion (TAPSE), peak systolic annular velocity, 2D fractional area change (FAC), and 2D speckle tracking longitudinal strain. Fractional area change was calculated as follows: (RV area in diastole - RV area in systole)/RV area in diastole × 100. Two-dimensional speckle tracking longitudinal strain was obtained from 2D images, including both the septal and lateral walls of the RV. Images were acquired from 3 cycles triggered to QRS at a frame rate of 50 to 90 per second. Three-dimensional full volume data sets were obtained by a 3- to 4-cycle multibeat acquisition with 12-slice display during a breath-hold period to obtain a volume rate >25 and digitally stored. In addition to volumes, RV longitudinal strain was
also measured by post processing the 3D dataset (Figure 1).
Invasive hemodynamic measurements were obtained by interventionalists who were blinded to echocardiographic findings. A standard 6F triple Cournand catheter into the femoral vein and a 5F pigtail catheter into the femoral artery were introduced by the Seldinger technique and guided by using the pressure waveform and fluoroscopy. Pulmonary capillary wedge pressure (PCWP), mean right atrial pressure, sPAP and diastolic PAP, mean PAP, and cardiac output were measured. Cardiac output was measured by the Fick method. Pulmonary vascular resistance index was calculated as follows: 80 × (mean PAP– mean PCWP)/cardiac output.
Values are given as means ± standard deviation. Pearson correlation test was used to evaluate the correlation between the tested parameters. A P value of less than .05, based on a 2-sided test, was considered statistically significant. SPSS statistical software (version 15.0, SSPS Inc, Chicago, IL, USA) was used.
We performed 32 studies in 26 heart transplant patients, with 6 patients having 2 studies at different time points with different pressures and therefore included. Patient characteristics are presented in Table 1, and echocardiographic measurements are presented in Table 2. By invasive catheter measurements, peak sPAP was 26 ± 7 mmHg, mean PAP was 16±5 mmHg, PCWP was 12 ± 5 mm Hg, right atrial pressure was 5 ± 3 mm Hg, pulmonary vascular resistance was 0.57 ± 0.40 mm Hg•min/L, and cardiac output was 5.21 ± 1.7 L in the overall group. We observed that longitudinal function by TAPSE (r = 0.69; P < .001), FAC (r = 0.45; P = .01), 2D strain (r = 0.49; P = .01), and tricuspid annular velocity (r = 0.42; P = .02) correlated significantly with 3D RV EF, as expected. The strongest correlates of sPAP were 3D end-diastolic (r = 0.75; P < .001) and end-systolic volumes (r = 0.55; P < .001), whereas right atrial volume correlated (r = 0.49; P = .01) with sPAP moderately. Three-dimensional RVEF correlated with sPAP moderately as well (r = 0.49; P = .004). Longitudinal RV free wall strain by 3D and 2D speckle tracking, TAPSE, tricuspid annulus velocity, and FAC did not correlate with sPAP in posttransplant patients (P > .05 for all) (Figure 2). Right ventricular global longitudinal 3D strain correlated moderately (r = 0.36; P = .04) and 2D global strain (average of free wall and septum) correlated marginally (r = 0.36; P = .05) with sPAP.
This is the first study where the novel 3D echocardiographic quantification tool has been used to assess RV function in relation to PAP in heart transplant recipients. The novel 3D RV quantification algorithm is easy to use, quick, and reproducible by an almost fully automated process.6 No geometric assumptions are incorporated, and few operator-defined markings are needed. In the present study, we found that RV end-diastolic volume was the most important correlate of sPAP. However, the correlation of sPAP with RV end-systolic volume was less strong, although statistically significant, as was the correlation of RV EF with sPAP compared with RV end-diastolic volume. This is quite understandable because, in contrast to many other parameters, including RV end-systolic volume, RV EF, TAPSE, and tissue velocities, end-diastolic volume does not tend to pseudonormalize with the tricuspid regurgitation that facilitates the RV unloading toward a much lower pressure compartment. Right ventricular end-diastolic volume is a direct reflection of pulmonary hemodynamic load. Ejection fraction, on the other hand, reflects the adaptation of the RV to pulmonary hemodynamic load. Of note, RV contractile function is a more important determinant of outcome than pulmonary hypertension itself.7 It is well known that the occurrence of a maladaptive ventricular remodeling and ventriculoarterial uncoupling but not pulmonary vascular resistance is an independent determinant of clinical outcome in patients with pulmonary arterial hypertension. This (mal)adaptation process may vary from patient to patient, meaning that despite reduction in PAP the RV EF may remain depressed.7 Therefore, RV volumes and EF need to be precisely quantified. On the other hand, as the RV dilates with increasing PAP, the RV longitudinal myofibers shift from a longitudinal to a more circumferential orientation.8-10 This in turn translates into a compensation of diminishing longitudinal function by radial function and necessitates the assessment of circumferential deformation along with longitudinal deformation.8 This is the reason why TAPSE, annular velocities, and longitudinal strain, despite the ease of their use, cannot reliably reflect the RV function in dilated ventricles. Instead, the assessment of RV volume and EF straightforward without any oversimplification and geometric assumption by new 3D RV quantification tools is preferable, as supported by the present findings. D’Andrea and associates11 have also found a good correlation between 3D RVEF by cardiac magnetic resonance imaging and 3D echocardiography (r = 0.89; P < .0001) in heart transplant patients.
Some limitations of TAPSE and tissue Doppler velocities are well recognized. These are angle dependent and affected by tethering and translation. In addition, these parameters oversimplify the complex mechanics of the RV function to the displacement of a single segment in 1 dimension. Speckle tracking strain may be more suitable as it quantifies the deformation of the myocardium. Therefore, it is a better indicator of the intrinsic contractile function.12 However, the assessment of only longitudinal strain, ignoring circumferential and shear strains, likely deviates from a realistic quantification of the RV function.9 Similarly, the FAC, which is based on area tracings, does not take into account many parts of the RV and oversimplifies the complex RV geometry. In a previous study, Clemmensen and associates13 also found that RVEF was a better determinant of functional capacity and mean PAP than annular velocities, TAPSE, and RV free wall strain in heart transplant patients.
Of note, several studies have already described reduced RV systolic function in both the immediate and the long-term posttransplant period.14-16 Our results are in accordance with those, indicating that the transplanted heart cannot be considered as a healthy heart due to several reasons. There are also wide variations in echocardiographic measurements from patient to patient that complicate the definitions of normal limits in heart transplant recipients.17 As a result, robust and reproducible quantification tools for serial follow-up are necessary.
This study has a limited patient population. The accuracy of the technique requires good image quality, and capture requires more than one beat, which necessitates breath holding by the patient. The echocardiographic and catheter measurements were not obtained simultaneously, but we did not include patients with unstable clinical condition. Therefore, the few hours between the 2 measurements are unlikely to represent different hemodynamic states. Right ventricular stroke index was not assessed invasively to make direct comparison with 3D RVEF. Nevertheless, other potential confounding factors such as allograft vasculopathy, coronary flow reserve, and ischemia-reperfusion time should be taken into account in larger series to better establish the relation between RV function parameters and PAP.
In conclusion, the effects of pulmonary hemodynamic burden on RV function were better estimated by 3D volume evaluation than by 3D longitudinal strain and other 2D and tissue Doppler measurements. These results suggest that the peculiar anatomy of the RV necessitates 3D volume quantification in heart transplant recipients threatened by residual pulmonary hypertension.
Volume : 15
Issue : 1
Pages : 231 - 235
DOI : 10.6002/ect.mesot2016.P104
From the Departments of 1Cardiology and 2Cardiovascular Surgery, Baskent University Medical School, Ankara, Turkey
Acknowledgements: The authors have no relations with industry and no conflicts of interest. All financial support for this study was provided by Baskent University.
Corresponding author: Leyla Elif Sade, University of Baskent, Department of Cardiology, 10 Sok No:45 Bahcelievler, 06490 Ankara, Turkey
Phone: +90 532 474 4998
Figure 1. Assessment of Right Ventricular Function
Table 1. Patient Characteristics
Table 2. Echocardiographic Findings
Figure 2. Assessment of Systolic Pulmonary Artery Pressure