Transesophageal Echocardiography

Intraoperative transesophageal echocardiography (TEE) is commonly applied in pediatric heart surgery, including orthotopic heart transplantation. Intraoperative assessment allows the evaluation of graft function immediately after implantation and reperfusion, assessment of valve regurgitation, and examination of anastomotic sites.

The most common surgical technique used in pediatric heart transplantation is direct connection of recipient left atrial cuff (including the pulmonary veins) to the donor left atrium and anastomoses at the superior and inferior vena cava, the “bicaval” technique ( Figure 1 ). Obvious suture lines may not be present with the bicaval technique, and the right atrial size appears smaller. The “biatrial” technique with direct anastomosis of donor to recipient right atrial tissue is still performed at some centers, particularly in small children or in the setting of atypical venous anatomy. The biatrial technique results in the appearance of prominent suture lines along the atrial wall, which should not be confused with thrombus. Stenoses at the supravalvar pulmonary and supravalvar aortic suture lines ( Figure 2 ) are rare. Stenosis of the superior caval anastomosis is likewise uncommon on the whole, although there is increased risk in pediatric patients, especially those with prior cavopulmonary anastomoses or other surgical manipulation of the cava. Patients with congenital heart disease constitute nearly 40% of pediatric heart transplant recipients, and these patients are at increased risk for venous and arterial complications. It is therefore critical that the echocardiographer performing TEE attempt to exclude obstruction at these anastomoses.

This image contrasts the biatrial and the bicaval techniques. In the biatrial connection ( left ), the recipient right atrium is excised roughly at the crista terminalis laterally, the donor superior cava is ligated, and an anastomosis is created to an oblique incision extending from the donor right atrial appendage to the donor inferior cava. It results in prominent suture lines along the atrial wall and atrial septum. In the bicaval technique ( right ), the superior and inferior cava are connected separately and obvious suture lines may not be present. Either technique results in a prominent left atrium.

(Used with permission of Mayo Foundation for Medical Education and Research. All rights reserved.)

Parasternal long-axis 2D echocardiographic image in a 10-year-old male patient who underwent orthotopic heart transplantation demonstrating mild supravalvar aortic stenosis at the aortic anastomosis. Note the degree of narrowing in the mid ascending aorta (AAo) ( arrow ), where the surgical anastomosis was made. Color Doppler images from a parasternal long-axis view might not demonstrate the flow disturbance, because of unfavorable intercept angle, so continuous-wave Doppler interrogation of the LV outflow tract from an apical five-chamber view is recommended. AoV , Aortic valve.

Assessment of Right Heart Failure

One of the early causes of graft loss is acute right heart failure. Right heart failure may result from prolonged ischemic time, poor cardiac protection, elevation of pulmonary vascular resistance, or acute graft rejection. TEE offers an important opportunity to evaluate ventricular systolic function, chamber dimensions, valve regurgitation, and estimation of right ventricular and pulmonary artery pressures. Right ventricular fractional area change, tricuspid annular plane systolic excursion by M-mode, tissue Doppler assessment of the maximal systolic tricuspid annular velocity, and myocardial performance index are potentially helpful right heart function parameters in children and have been applied in the operative and perioperative assessment of the right heart in adults. Identification of significant right heart failure, tricuspid regurgitation or other evidence of cardiac dysfunction on TEE may prompt the clinician to consider intensified right heart inotropic support, pulmonary vasodilator therapy, or the need for mechanical circulatory support such as extracorporeal membrane oxygenation. Following return from the operating room, transthoracic echocardiography is a useful tool to help evaluate perioperative complications, including right heart failure, rejection or cardiac tamponade.

Transesophageal Echocardiography

Intraoperative transesophageal echocardiography (TEE) is commonly applied in pediatric heart surgery, including orthotopic heart transplantation. Intraoperative assessment allows the evaluation of graft function immediately after implantation and reperfusion, assessment of valve regurgitation, and examination of anastomotic sites.

The most common surgical technique used in pediatric heart transplantation is direct connection of recipient left atrial cuff (including the pulmonary veins) to the donor left atrium and anastomoses at the superior and inferior vena cava, the “bicaval” technique ( Figure 1 ). Obvious suture lines may not be present with the bicaval technique, and the right atrial size appears smaller. The “biatrial” technique with direct anastomosis of donor to recipient right atrial tissue is still performed at some centers, particularly in small children or in the setting of atypical venous anatomy. The biatrial technique results in the appearance of prominent suture lines along the atrial wall, which should not be confused with thrombus. Stenoses at the supravalvar pulmonary and supravalvar aortic suture lines ( Figure 2 ) are rare. Stenosis of the superior caval anastomosis is likewise uncommon on the whole, although there is increased risk in pediatric patients, especially those with prior cavopulmonary anastomoses or other surgical manipulation of the cava. Patients with congenital heart disease constitute nearly 40% of pediatric heart transplant recipients, and these patients are at increased risk for venous and arterial complications. It is therefore critical that the echocardiographer performing TEE attempt to exclude obstruction at these anastomoses.

This image contrasts the biatrial and the bicaval techniques. In the biatrial connection ( left ), the recipient right atrium is excised roughly at the crista terminalis laterally, the donor superior cava is ligated, and an anastomosis is created to an oblique incision extending from the donor right atrial appendage to the donor inferior cava. It results in prominent suture lines along the atrial wall and atrial septum. In the bicaval technique ( right ), the superior and inferior cava are connected separately and obvious suture lines may not be present. Either technique results in a prominent left atrium.

(Used with permission of Mayo Foundation for Medical Education and Research. All rights reserved.)

Parasternal long-axis 2D echocardiographic image in a 10-year-old male patient who underwent orthotopic heart transplantation demonstrating mild supravalvar aortic stenosis at the aortic anastomosis. Note the degree of narrowing in the mid ascending aorta (AAo) ( arrow ), where the surgical anastomosis was made. Color Doppler images from a parasternal long-axis view might not demonstrate the flow disturbance, because of unfavorable intercept angle, so continuous-wave Doppler interrogation of the LV outflow tract from an apical five-chamber view is recommended. AoV , Aortic valve.

Assessment of Right Heart Failure

One of the early causes of graft loss is acute right heart failure. Right heart failure may result from prolonged ischemic time, poor cardiac protection, elevation of pulmonary vascular resistance, or acute graft rejection. TEE offers an important opportunity to evaluate ventricular systolic function, chamber dimensions, valve regurgitation, and estimation of right ventricular and pulmonary artery pressures. Right ventricular fractional area change, tricuspid annular plane systolic excursion by M-mode, tissue Doppler assessment of the maximal systolic tricuspid annular velocity, and myocardial performance index are potentially helpful right heart function parameters in children and have been applied in the operative and perioperative assessment of the right heart in adults. Identification of significant right heart failure, tricuspid regurgitation or other evidence of cardiac dysfunction on TEE may prompt the clinician to consider intensified right heart inotropic support, pulmonary vasodilator therapy, or the need for mechanical circulatory support such as extracorporeal membrane oxygenation. Following return from the operating room, transthoracic echocardiography is a useful tool to help evaluate perioperative complications, including right heart failure, rejection or cardiac tamponade.

Conventional Echocardiography

Although conventional 2D echocardiography is the most commonly performed imaging study in the follow-up management of children with heart transplantation, its utility for detecting presymptomatic rejection is debated. Transthoracic echocardiography has advantages in that it is portable, is cost-effective, and does not require use of anesthesia or other personnel. Changes such as increased wall thickness, increased echogenicity of the ventricular myocardium, and presence of new valvular insufficiency or effusion have been associated with allograft rejection ( Figure 3 ). However, these changes may not be reliably present even in moderate cellular rejection.

Baseline parasternal long-axis images in a patient after heart transplantation with normal left ventricular posterior wall thickness ( top ) and in the same patient at presentation 3 months later with allograft rejection ( middle and bottom ). Note the increased left ventricular posterior wall thickness ( white arrow ) and small rim of pericardial effusion ( asterisk ). Middle image shows pulsed-wave Doppler of the mitral inflow with increased E/A ratio. LA , Left atrium; RV , right ventricle.

Changes in ventricular systolic function detectable by 2D transthoracic echocardiography are often late findings seen only in progressive or severe rejection. However, because early changes in shortening fraction or ejection fraction can be correlated to cellular rejection, 2D transthoracic echocardiography remains routine in the early postoperative phase and at intervals during later follow-up.

Spectral Doppler and DTI

As 2D imaging has been shown to be limited in the assessment of rejection, there has been increasing interest in the use of spectral Doppler to detect more subtle changes associated with myocardial edema and rejection. Myocardial inflammation affects ventricular relaxation and tissue characteristics, which become evident before overt systolic dysfunction. Therefore, a number of investigators are actively evaluating various measures of diastolic function as indicators of graft rejection. Not surprisingly, studies assessing mitral inflow velocities by Doppler, including early diastolic (E) peak velocity, late diastolic (A), and their ratio (E/A) ( Figure 3 ), as well as isovolumic relaxation time and E-wave pressure half-time, have all shown some promise.

The data on these various measures have produced conflicting results, especially in pediatric patients, leaving an unclear picture for clinicians. Because of the inconsistency of single echocardiographic parameters to demonstrate rejection, echocardiographic scoring systems have been developed that combine various measures of systolic function, wall thickness, diastolic function, and Doppler velocities. One of these scoring systems has reported sensitivity of 88% and specificity of 83% for rejection, although these results have not been reproduced in other publications. The same author group updated their scoring system using intrapatient changes in Doppler parameters and reduced the false-positive rate from 72% to 10% and enhanced the specificity from 90% to 99%.

DTI allows better characterization of myocardial mechanics and has been a focus of investigation in attempts to use noninvasive imaging to monitor for acute cellular rejection in both adults and children. Studies of DTI of septal and posterior basal segments of the left ventricle assessing early mitral excursion (Ea or E′) and systolic radial velocity (Sm) have shown changes associated with graft rejection. The utility of the myocardial performance index, derived as a composite ratio of isovolumetric contraction time and isovolumetric relaxation time to ejection time, has also been studied in the evaluation of rejection. There was a 98% mean increase in myocardial performance index during rejection compared with baseline, which returned to baseline after treatment, suggesting sensitivity as a marker of early diastolic and systolic performance.

Others have shown that attempts to use DTI measures in children following heart transplantation met with mixed results. The variability of such studies is not surprising, as there are many limitations to DTI in children, particularly in pediatric heart transplantation cohorts. Specifically, these groups represent a wide age range of patients with an even more diverse range of graft size and donor age. Abnormal interventricular septal motion complicates DTI interpretation and is common following heart transplantation because of factors such as elevation of right ventricular pressure or conduction abnormalities. The normal ranges of values of diastolic function in children are poorly defined and inconveniently broad. Moreover, DTI measures are generally decreased even in well-functioning cardiac grafts, so the application of normal ranges to these patients may be misleading. Finally, collection of these measures is site and angle dependent, requiring a practiced team to obtain useful data, as even small changes in the acquisition angle will cause changes in DTI velocity.

To circumvent the issues associated with these wide ranges of normal values and the baseline abnormal velocities seen in many posttransplantation patients, authors have focused on changes in DTI measures as a more predictive metric to evaluate acute rejection. Initial reports in adult heart transplant recipients demonstrated that a reduction in either E′ or Sm of >10% is associated with increased risk for rejection in adults following heart transplantation, with sensitivity of 90% and 86% and specificity of 94% and 96%, respectively. Recently, Lunze et al. reported a retrospective review of echocardiographic studies performed within 24 hours of EMB and compared with prior echocardiographic examinations in the same patients at a time of proven nonrejection, showing that a decline of <15% in S′ velocity and a decrease of <5% in A′ velocity predicted nonrejection with accuracy of >99%, without misclassifying any rejection episodes. This study is important in that it may better represent the role of noninvasive imaging in the pediatric population in which rejection is currently rare outside of the first year to be one of screening for normalcy and not for rejection. It is notable that this was a retrospective and single-center review, with a single reviewer providing the majority of interpretation of results. It has yet to be replicated in a prospective cohort, and small changes in velocity needed to have a “positive” result (indicating rejection) could occur with minor errors in data acquisition.

Although no broadly accepted single echocardiographic protocol currently exists to detect rejection, various similar echocardiographic scoring systems using multiple data points and intrapatient comparisons are in use at different centers. Unfortunately at the present time, these data have not been reproducible at multiple centers to allow dissemination of a defined methodology for noninvasive detection of rejection.

Advanced Functional Imaging

In a prospective investigation using 3D echocardiography, we demonstrated abnormal left ventricular (LV) mechanics and high prevalence of mechanical dyssynchrony in children after heart transplantation (mean systolic dyssynchrony index, 6.2 ± 4.3% in the transplanted left ventricle vs 2.2 ± 1.1% in normal). Both global and interventricular septal strain was lower in the transplantation group. The LV ejection fraction (LVEF) divergence was greater in transplantation patients and had a strong positive correlation with systolic dyssynchrony index and negative correlations with all measures of strain ( Figure 4 ). These findings indicate that abnormal LV mechanics possibly contribute to differences in LVEF measurements by 2D and 3D echocardiographic methods, and LVEF should therefore be calculated using 3D echocardiography in the transplantation population.

(A) Reconstructed LV model from a 3D echocardiographic data set in a pediatric patient after heart transplantation. (B) Sixteen-segment bull’s-eye maps of 3D strain derived from semiautomated endocardial and epicardial tracking of the left ventricle in the same patient showing abnormal regional strain in the interventricular septum ( yellow ). (C) Volume-time curves of the 16 LV segments with colored dots on each line representing the minimal systolic volume for each segment. Note that there is wide scatter in the timing of minimum segmental volume, indicating a dyssynchronous left ventricle.

Buddhe et al. studied 50 children with “normal” systolic function by traditional echocardiography 4 years after transplantation and demonstrated speckle-tracking-derived mitral early diastolic velocity–to–strain rate (E/SR[E]) to have a modest correlation with pulmonary capillary wedge pressure. Twenty-four percent of patients had global longitudinal strain (GLS) > −18%, while patients with coronary vasculopathy had significantly higher E/SR (E) (71.9 ± 28.4) compared with those with normal coronary arteries (45.2 ± 10.8). Sehgal et al. demonstrated the clinical utility of peak systolic strain for detecting acute allograft rejection in children. They reported significant decreases in peak systolic GLS (−11.7% vs −14.6%), circumferential strain (−14.4% vs −21.7%), and radial strain (18.3% vs 26.5%) during rejection. Mingo-Santos et al. reported similar utility of deformation imaging in detecting acute rejection and suggested that systolic strain measurements may reduce the burden of repeated biopsy. Others have confirmed usefulness of peak diastolic strain in predicting LV filling pressure and risk for rejection. Nawaytou et al. found baseline abnormalities in LV rotational mechanics in transplanted children without active rejection and proposed the slope of torsion-radial displacement loop as a marker of LV dysfunction. The European Association of Cardiovascular Imaging recommendations for the assessment and follow-up of patients after heart transplantation suggest GLS as a suitable parameter to diagnose subclinical allograft dysfunction and that GLS could be used in association with EMB to characterize an acute rejection or global dysfunction episode. The very high negative predictive value of speckle-tracking echocardiography to exclude rejection, and its potential to minimize the burden of frequent EMBs within the first year after heart transplantation, was highlighted in a recent editorial comment. However, others suggest that speckle-tracking echocardiography cannot be a replacement for biopsy, because of its inability to detect serial changes in patients with asymptomatic rejection.

Cardiac Magnetic Resonance Imaging

More recent research has focused on the use of complementary imaging techniques such as CMR to assess graft function and rejection. CMR has great appeal because of reliable whole-heart imaging throughout the cardiac cycle, excellent border definition allowing accurate measures of ventricular mass, volume, and LVEF, and tissue characterization superior to that obtained with echocardiography. Specifically, contrast-mediated imaging and analysis of T2 signal of heart tissue can reveal evidence of myocardial edema. Cellular rejection occurs because of the infiltration of lymphocytes and activation of cytokine pathways, leading to inflammation and edema in the muscle that can be quantified via T2 relaxation. Marie et al. suggested that a T2 relaxation time of >56 msec was sensitive (89%) and relatively specific (70%) for the detection of significant rejection (≥2R). However, Wisenberg et al. had previously demonstrated that T2 prolongation is common early after transplantation, even in the absence of rejection, and suggested that this represents myocardial edema secondary to the transplant process.

Miller et al. evaluated all patients receiving heart transplants at a single center with serial CMR on the same day as EMB. CMR included T1 mapping and T2 mapping, as well as evaluation for extracellular volume (ECV), late gadolinium enhancement (LGE), and absolute myocardial blood flow. Although improvements in function were noted with passage of time from transplantation, only circumferential strain was significantly different in patients with acute rejection (by EMB) compared with those patients without rejection. Unfortunately there was still significant overlap in circumferential strain between the two groups.

Others have shown high sensitivity and high negative predictive value for CMR in the diagnosis of acute cardiac allograft rejection. The benefits of quantitative T2 mapping and its potential use in characterizing rejections have been highlighted by some, whereas in other studies, multiparametric CMR was not able to accurately detect acute rejection during the early posttransplantation period. Despite its promise, these studies leave the true predictive value of T2-weighted imaging open to debate.

Other CMR markers (native T1 times and ECV) have shown correlation with the degree of fibrosis on EMB in children after heart transplantation. Coelho-Filho et al. focused on CMR markers of tissue remodeling—myocardial ECV and intracellular lifetime of water (τ _ic ), a measure of cardiomyocyte hypertrophy—in a study comparing 26 transplant recipients (mean age, 47 ± 7 years; median follow-up after transplantation, 6 months) with age-matched healthy volunteers. Transplant recipients had normal LVEFs (65.3 ± 11%) with higher LV mass relative to volunteers (114 ± 27 vs 85.8 ± 18 g). ECV and τ _ic were higher after transplantation (ECV, 0.39 ± 0.06 vs 0.28 ± 0.03; τ _ic , 0.12 ± 0.08 vs 0.08 ± 0.03), and ECV was associated with LV mass ( r = 0.74). In follow-up, transplant recipients with normal biopsies (International Society for Heart and Lung Transplantation grade 0R) in the first posttransplantation year exhibited lower ECV relative to patients with any rejection ≥2R (0.35 ± 0.02 for 0R vs 0.45 ± 0). Higher ECV but not LVEF was associated with reduced rejection-free survival. Further research on the impact of graft preservation and early immunosuppression on tissue-level remodeling of the allograft is necessary to delineate the clinical implications of these findings.

In summary, CMR has been studied as a potential means for noninvasive detection of rejection because it can identify areas of hyperemia, inflammation, and areas of scar and fibrosis using LGE. Single-center studies have reported using LGE patterns to identify areas of active inflammation, and others have examined signal intensity patterns in the early postcontrast phase to identify inflammation and necrosis. Large prospective studies are needed to better understand the true sensitivity and specificity of these techniques. Currently, for cellular rejection, CMR is limited primarily to the detection of edema, reduced LVEF, or other alterations of tissue character, whether by prolonged T2, altered T1, or LGE. Each of these signifies some degree of myocardial tissue damage. In the future, detection of rejection before these findings may be able to trigger intervention to minimize the irreversible damage to the allograft. Investigations are under way to detect immune activity in the allograft using CMR contrast agents specific to immune cells. At the time of this review there are no studies directly addressing the use of CMR to detect acute rejection in children. Such a study is uniquely challenging because pediatric heart transplantation center volumes are generally low, rejection rates are much lower than adults, and many children require anesthesia for adequate CMR.