Assessment of Biventricular Function in Clinically Well Pediatric Heart Transplantation Patients by Three-dimensional Speckle Tracking Echocardiography

Background Studies on pediatric heart transplantation (HTx) are uniquely challenging because pediatric HTx center volumes are generally low. And, the biventricular function plays an important role in the prognosis of pediatric HTx. The primary aim of our study was to evaluate biventricular function of pediatric HTx by three-dimensional speckle tracking echocardiography(3D-STE). Methods We enrolled 30 clinically well pediatric HTx patients and 30 sex- and age- matched healthy controls. All subjects underwent comprehensive echocardiographic examinations. Left ventricular (LV) global longitudinal strain (GLS), global circumferential strain (GCS), LV and right ventricular (RV) ejection fraction (EF) and RV longitudinal strain (RVLS) of free wall and septum were acquired by 3D-STE. And the correlations between strains and clinical data were explored. Results Compared with controls, LV GLS was decreased in pediatric HTx patients (P<0.05), while LV GCS and LVEF showed no difference. RVEF, RVLS (free wall) and RVLS (septum) in HTx group were diminished (P<0.05), but RVEF was still in normal range. Cold ischemic time was correlated inversely with LV GLS (β=-0.401, P<0.05). The mean pulmonary artery pressure (β=0.447, P<0.05) and postoperative tricuspid regurgitation pressure (β=0.607, P<0.05) were associated with RVLS (free wall). Conclusion Biventricular longitudinal systolic function rather than global systolic function was impaired after HTx. 3D STE may be able to evaluate the ventricular function better. Prolonged ischemic time leads to impaired LV longitudinal systolic function in pediatric HTx patients. It’s interesting that in HTx patients, it shows compensatory enhancement due to increased pulmonary vascular resistance.


Introduction
Orthotopic heart transplantation (HTx) is a well-established and effective therapeutic option for children with end-stage heart failure. [1][2][3][4][5] The assessment of ventricular function has been hotspot since the rst HTx in 1967. Until now, many efforts have been made to in the aspect of characteristics of the transplanted heart. [6][7][8][9][10][11] Studies on adult HTx patients have showed that the myocardial function of HTx patients was different from normal people and patients undergoing other cardiac surgery. [5][6][7][8][9][11][12][13][14][15] These studies found that the myocardial function decreased in both clinically stable and unstable HTx patients, and that left ventricular (LV) dysfunction after HTx was multifactorial and could be an indicator of graft rejection or coronary artery vasculopathy (CAV). [6][7][8]12 In their recent study, Ingvarsson and his peers have reported the reference ranges for adults after HTx with no acute transplant associated cardiac complications or CAV. 6 But, studies on pediatric HTx are uniquely challenging because pediatric HTx center volumes are generally low. 2,4,14 And, most of the existing studies were done by two-dimensional speckle tracking echocardiography (2D STE). As we all know, 2D STE algorithms only track speckles in 2D planes, while speckles move in 3-dimensions, so only a portion of the real motion can be detected. This limitation is particularly relevant in heart transplant recipients, in whom the typically greater size of the mediastinal cavity when compared with the donor heart, and the loss of support provided by the pericardial sac, results in a marked translational motion of the transplanted heart inside the chest during the cardiac cycle. 12 Fortunately, three-dimensional speckle tracking echocardiography (3D STE), as a novel technique, overcomes the limitation and has been suggested to be a more accurate tool for global and segmental assessment of LV function.
So, in this study, we tried to report 3D biventricular volume and strain in clinical well pediatric HTx population. And, we sought to describe possible impact of clinical variables on biventricular myocardial mechanics.

Method
The study was conducted in Union Hospital of Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. The study was approved by the ethics committee of the university and the entire studied population was enrolled after informed consent was obtained.

Study Population
Pediatric patients who underwent HTx in Union Hospital of Tongji Medical College, Huazhong University of Science and Technology after January 1, 2015 were enrolled during their routine follow up. The patients included were ≤ 18 years old when they underwent HTx. And, at the time of examination all patients were in sinus rhythm or a regular paced rhythm (10 patients were in regular paced rhythm). Exclusion criteria were as follows: (1) suspicion of graft rejection based on standard clinical ndings (changes in LV function, ECG, arrhythmia, serologic examinations, valve regurgitation); (2) evidence of coronary artery disease (de ned as the presence of a stenosis > 50% in any major artery or distal pruning of any secondary branch on the basis of angiographic or coronary CTA); (3) signi cant valvular regurgitation or stenosis (more than mild); (4) within 6 months after transplantation.
The control group consisted of age and sex matched healthy children. All children included were in sinus rhythm or a regular paced rhythm (4 were in regular paced rhythm).
Demographic and clinical data of the subjects were collected in detail and listed in Table 1. Mean pulmonary artery pressure pre-HTx was assessed by percutaneous cardiac catheterization. It's hard for pediatric patients to take percutaneous cardiac catheterization in their routine follow up, as a result, we failed to acquire mean pulmonary artery pressure post-HTx.

Participant characteristics
The study enrolled a total of 35 pediatric patients, and 32 healthy children. After exclusion (2 within 6 months after transplantation, 1 with moderate tricuspid regurgitation, 2 with poor image quality), 30 HTx patients were nally included. The median elapsed time since HTx was 11.3 (7.5-22.2) months. Meanwhile, 30 healthy children were nally included as control group (2 was excluded for poor image quality). According to the clinical records, 23 HTx patients underwent percutaneous cardiac catheterization before transplantation, and the preoperative mean pulmonary artery pressure was 30.3 ± 12.9 mmHg. Baseline characteristics of HTx group and control group were shown in Table 1. The heart rate and blood pressure of HTx group were signi cantly higher (P < 0.05), but blood pressure was in normal range.
Conventional TTE parameters 2D echocardiographic, color Doppler and tissue Doppler ndings of the left and right heart are presented in Table 2 and Table 3, respectively. TASPE, s' (Free wall) and RVFAC were decreased in HTx group (P < 0.05), but RVFAC was still in normal range. 13 HTx patients had mild tricuspid regurgitation at the time of examination. The mean tricuspid regurgitation velocity was 2.4 ± 0.4 m/s, and the mean tricuspid regurgitation pressure was 23.5 ± 6.8 mmHg, as estimated from the tricuspid regurgitation velocity (using the modi ed Bernoulli equation).

Lv Strain Derived From 3d Ste Analysis
Data of the LV, derived from 3D STE, was showed in Table 4. In HTx group, GLS was decreased (P < 0.05), while GCS and LVEF remained (P > 0.05). All the clinical variables listed in Table 1 and tricuspid regurgitation pressure at the time of examination were entered into a simple linear correlation with GLS and GCS. Only LVEF pre-HT, donor sex and cold ischemic time showed signi cance or trends toward signi cance with GLS, while no variables showed signi cance or trends toward signi cance with GCS. They were entered into multivariate linear regression analyses. Only cold ischemic time was an independent predictor of GLS (β=-0.401, P < 0.05). The result was showed in Table 5.

Rv Strain Derived From 3d Ste Analysis
Data of the RV, derived from 3D STE, was showed in Table 6. All the clinical variables listed in Table 1 and tricuspid regurgitation pressure at the time of examination were entered into a simple linear correlation with RVLS (free wall). Only mean pulmonary artery pressure pre-HTx, previous cytomegalovirus infection and tricuspid regurgitation pressure post-HTx showed signi cance or trends toward signi cance with RVLS (Free Wall). The result of multivariate linear regression analyses of RVLS (Free Wall) was showed in Table 7. Mean pulmonary artery pressure (β = 0.447, P < 0.05) and postoperative tricuspid regurgitation pressure (β = 0.607, P < 0.05) were independently associated with RVLS (free wall).

Intraobserver And Interobserver Reproducibility
The ICCs and Bland-Altman analyses for the intra-and inter-observer reproducibility of the strain parameters derived from 3D STE were showed in Table 8. All ICCs were consistent with good to excellent reproducibility and the range of the difference could be tolerated.

Discussion
We conducted this study to report 3D biventricular volume and strain and validate whether 3D STE can show more details in biventricular function in pediatric HTx population with stable clinical status. And we also tried to describe possible impact of demographic and clinical data on biventricular myocardial mechanics. To our knowledge, this is the rst comprehensive evaluation of biventricular function in clinically well pediatric HTx patients by 3D STE. Though some investigators are actively trying to use parameters assessing diastolic function, but the results are con icted, especially in pediatric patients. [16][17][18][19][20][21][22][23] In this study, we pay more attention to the systolic function.

Left Ventricular
Previous studies in both clinically well pediatric and adult HTx patients have showed a mean LVEF in the normal range, 6, 7, 12, 24-26 but some were statistically signi cantly lower than reference. 6,12 Our pediatric HTx patients showed a mean LVEF in the normal range, no statistically signi cantly compared to control group, which agrees with some of the previous studies. 7,[24][25][26] HTx patients in our study maintained normal global LV systolic function.
We saw a reduction in GLS in our study, which agrees with the previous pediatric and adult study. 6,7,24,25,27 This re ect that the LV longitudinal systolic function was impaired even in clinical well HTx patients.
GLS is mainly produced by subendocardial myocardium, and subendocardial myocardium is sensitive to ischemia, myocardial alignment disorder, and brosis. 28,29 So, ischemic time during surgery, changes in post-operative load status, myocardial reperfusion injury, and some postoperative complications may affect its function. In multivariate linear regression, we found that cold ischemic time is an independent impactor on longitudinal systolic function, though the correlation was weak. This result indicated a promising trend that the longer the cold ischemic time, the worse the LV longitudinal systolic function.
Studies on GCS were less than GLS. Some of the previous studies showed a reduction in GCS. 7,26 But, in our study, GCS remained normal, which agrees with Ingvarsson's study in 2017. 6 This re ect that the LV circumferential systolic function in HTx patients was remained. The circumferential systolic function remained, while longitudinal systolic function decreased. That might possibly indicate a compensatory mechanism, and this may also explain why LVEF remained normal. We did not nd any signi cant indicators related to GCS. This may due to that circumferential systolic function was less in uenced by the patients' clinical status pre-HTx and the habitats of donor heart.
We also found that the LV wall thickness and LVM increased in HTx patients. Wall thickness has been discussed in detail in previous adult HTx studies. Researchers explored that the LV morphology is characterized by an increase in LVM and wall thickness during the rst month after HTx. After three months, the wall thickening mostly improved. 30,34 A secondary increase in LVM and wall thickness may occur as a consequence of many factors such as repetitive rejection episodes, chronic tachycardia, and systemic hypertension, usually induced by immunosuppressive agents. 30,[34][35][36] Some patients in our study suffered from systemic hypertension and renal dysfunction after the surgery, though treated, the effect on the ventricular wall may not disappear completely.

Right Ventricular
RV dysfunction was sought to be one of the most important cause of death in the early post-HTx period. 30,[37][38][39] However, evaluation of RV function has always been a challenge. In our study, we used both conventional and 3D RV functional parameters to evaluate the RV function of pediatric HTx patients.
We found that RVLS (free wall), RVLS (septum) and RVEF all decreased, but RVEF still in a normal range in pediatric HTx patients. This re ected that the longitudinal RV systolic function rather than the whole RV function impaired more in pediatric HTx patients. This is mostly consistent with previous studies. 9,[40][41][42] Since RVLS (eptum) is usually greatly affected by LV, we did not discuss it in detail and pay more attention to RVLS (Free wall). In multivariate linear regression, we found that in pediatric HTx patients, RV longitudinal systolic function was enhanced as mean pulmonary artery pressure pre-HTx and tricuspid regurgitation pressure increased. It's an interesting nding. Compared with normal, RV longitudinal systolic function was decreased, but it showed a compensatory enhancement as a response to increased pulmonary vascular resistance and RV afterload among pediatric HTx patients. Most of the patient underwent HTx had increased pulmonary vascular resistance and RV afterload. These changes persisted on even after surgery and may resulted in compensatory enhancement in RV longitudinal systolic function. In our study, the mediate follow up time was 1 year, whether the compensatory mechanism would exist in the future was unknown and needs further study. Previous studies had found that RV function was also signi cantly correlated with the ischemic time and requirement for mechanical circulatory support pre-or post-HTx. [41][42][43] However, we didn't nd signi cant correlation between RV function with ischemic time. And, only few patients required mechanical circulatory support pre-or post-HTx, so we didn't take this factor into consideration.
In our study, both the conventional parameters, including TAPSE and s' (RV free wall), and RVLS derived from 3D-STE showed that the RV longitudinal systolic function was impaired. While, RV FAC and 3D RVEF showed normal global systolic function. However, previous studies revealed that TAPSE and s' may be reduced due a distorted anatomy in the context of a normal overall RV function and EF in HTx, and TAPSE, s' (RV free wall) and RV FAC measured by 2D echocardiography were insu cient to assess RV function due to the complex anatomy of the RV. 6,30,44 3D echocardiography parameters, such as RVLS (Free wall) and RVEF, might be able to evaluate the RV function better.
Some of the ndings in our study were also found by 2D STE. However, 2D STE algorithms only track speckles in 2D planes, only a portion of the real motion of the heart can be detected. Fortunately, 3D STE, as a novel technique, overcomes the limitation and has been suggested to be a more accurate tool for global and segmental assessment of LV function. What's more, 2D-STE is time-consuming. We need more than 30 minutes to analyses a patient by 2D STE, while we only need 10-15 minutes by 3D STE, while. So, we recommend 3D STE as a more appropriate technique for the analysis of LV function in HTx patients.

Limitations
The study volumes are generally low. It is di cult for children to take EMB, so whether rejection exists was judged according standard clinical ndings in our study. In our study, the control group was included only according to the recipients.

Conclusion
Biventricular longitudinal systolic function rather than global systolic function was impaired after HTx. 3D STE may be able to evaluate the ventricular function better. Prolonged ischemic time leads to impaired LV longitudinal systolic function in pediatric HTx patients. It's interesting that the RV longitudinal systolic function was impaired compared with normal, but in HTx patients, it shows compensatory enhancement due to increased pulmonary vascular resistance and increased RV afterload.

Declarations
Ethics approval and consent to participate: It's a retrospective study. The study was conducted in Union Hospital of Tongji Medical College, Huazhong University of Science and Technology, Wuhan, China. The study was approved by the ethics committee of the university.
Consent for publication: The consent was obtained from the patients or their legal guardian.
Availability of data and materials: The datasets used and/or analysed during the current study are available from the corresponding author on reasonable request.