Heart transplantation (HT) is an established life-saving treatment option for patients with severe end-stage heart failure. Although many improvements over the past few years in organ donation, organ preservation, and antirejection therapy have resulted in improved survival, acute allograft rejection (AAR) and cardiac allograft vasculopathy (CAV) remain significant causes of morbidity and mortality. Historically, surveillance for AAR and CAV has been based on invasive procedures, which carry inherent risks and high costs. Different noninvasive imaging modalities have been used in the evaluation of cardiac allograft structure and function and in the detection of AAR and CAV after HT. Echocardiography is particularly useful for the assessment of HT recipients since it is easily performable and it is not associated with the risks of the invasive procedures.
The features of transplanted heart
The donor cardiectomy involves a partial atrial resection. With the biatrial surgical approach (Shumway technique) an anastomosis is made between the residual recipient atrial tissue and the donor atria. Therefore, transplanted hearts have a characteristic atrial morphological appearance with a increased size in both atria, primarily caused by an increase in their long-axis dimension. The site of anastomosis between the residual recipient atrial tissue and the donor atria is visualisable as a ridge at mid-atrial level. The Shumway technique has been reported to contribute to abnormal patterns of LV filling and to predispose to atrial thrombus formation, due to blood stasis in the dilated atrial cavities with sub-optimal hemodynamics. Over the past decade the bicaval operative technique has become predominant. In this approach the entire recipient right atrium and much of the left atrium are excised, and the anastomosis is made between the small right atrial cuffs on the superior and inferior cavae and the left atrial cuff remaining around the four pulmonary veins to the donor atria. This surgical approach is associated with better preservation of atrial morphology without the marked increase in longitudinal dimension and lower risk for development of atrial thrombus. Similar to patients who have undergone other forms of cardiac surgery, HT patients often have abnormalities of interventricular septal motion with an abnormal movement of the left ventricle toward the right ventricle during systole and a decrease in septal thickening. This abnormal septal motion and thickening had no effect on overall ejection fraction. The normalisation of left ventricular (LV) systolic function after successful HT is responsible for early improvement of symptoms and has a strong impact on prognosis. LV chamber dimension and systolic function are well maintained 10–15 years after surgery. LV systolic dysfunction late after HT is often due to the effects of CAV and is associated to a poor prognosis. After HT, there is an evident increase in LV wall thickness and mass. LV hypertrophy is due to several causes (peritransplantation injury, repetitive rejections, arterial hypertension, immunosuppressive therapy, chronic tachycardia and denervation) and its progression occurs mainly in relation to cyclosporine levels and blood pressure levels. Early changes of diastolic filling after HT are related to the status of the donor heart, the acute effects of transplantation, and the cardiovascular abnormalities in the recipient. The elevated filling pressures and a restrictive filling pattern often seen early after HT cause Doppler filling abnormalities on echocardiography. Isovolumetric relaxation time (IVRT), early filling deceleration time, and early filling velocities gradually return to a normal pattern by 4 weeks after surgery. Despite this, abnormal diastolic filling can be identified in some long-term transplant recipients. The E/Ea ratio, has been reported as a validated parameter for noninvasive estimation of LV filling pressure in the transplanted heart. Valvular regurgitation is common immediately post HT but structural abnormalities of the aortic and mitral valves are not observed frequently after HT. In contrast to the left-sided valves, tricuspid valve regurgitation is very common after HT and its etiology is multifactorial. Significant tricuspid regurgitation early after surgery is likely related to elevated pulmonary arterial pressures and vascular resistance in the recipient as well as to atrial structure and function in patients who had undergone biatrial surgical approach. In the majority of the cases, the elevated pulmonary pressures and resistance normalize over the first month with a resolution of significant tricuspid regurgitation. Tricuspid regurgitation may be also due to alterations to the tricuspid annulus (i.e. related to the right atrial anastomosis), or the effect of multiple rejection episodes, or the injury to the chordal apparatus caused by the repeated endomyocardial biopsy. The persistence of tricuspid regurgitation has been related to the development of symptomatic right heart failure, reduced functional status, and increased mortality. The assessment of RV systolic function is important in HT recipients. Although the majority of the cases of early RV dilation improve progressively over the first weeks after HT, RV failure is a recognised cause of in-hospital death and some survivors have residual RV dilation. Pericardial effusions are not uncommon in the first year following transplantation. They can be large and close echocardiographic monitoring is required in order to avoid heart tamponade. The cause of pericardial effusions in these patients is unclear. A mismatch between recipient and donor hearts, the presence of AAR, and the effect of some immunosuppressive drugs could be involved. Pericardial effusion is not associated, however, with any adverse clinical outcomes and progressively disappears. If a large pericardial effusion accumulates slowly it may be of a little hemodynamic impact.
Acute allograft rejection
AAR is characterised histologically by inflammatory cell infiltrates, interstitial edema and myocite necrosis which translate into structural and functional abnormalities. Because AAR is frequent in the first months after HT and is initially asymptomatic, regular rejection surveillance is mandatory. Endomyocardial biopsies (EMBs) represent the gold standard for AAR detection. The management strategy depends on the histologic type and grade of rejection (using the International Society of Heart and Lung Transplantation revised grading system) and the presence or absence of hemodynamic compromise (decreased allograft systolic function and/or hemodynamic instability). With current advances in immunosuppression, the majority of patients who develop AAR have no significant changes in left ventricular (LV) ejection fraction. However, monitoring cardiac allograft systolic function is important in suspected or proven AAR because more aggressive immunosuppression can lead to improvement in LV function in patients with depressed function.
Cardiac allograft vasculopathy
CAV is characterized by diffuse intimal hyperplasia that is likely the result of cumulative endothelial injuries. The early diagnosis of CAV is challenging because typical clinical symptoms of ischemia are lacking, given cardiac denervation and the fact that coronary angiography can underestimate the severity of disease. Early recognition is important because rapid CAV progression, as defined by intravascular ultrasound (IVUS), has been shown to be a powerful predictor of all-cause mortality and myocardial infarction. Although IVUS is considered the most sensitive tool to detect CAV, coronary angiography is still the standard in many transplant centers. Given the invasive nature and high cost of both coronary angiography and IVUS, a noninvasive imaging modality is needed to identify patients at risk for CAV and several imaging modalities have been used for the detection of CAV.
Echocardiography and acute allograft rejection
Although EMB is considered the gold standard for the diagnosis of AAR, different noninvasive imaging techniques have been used to detect AAR. Echocardiography is a well-established, widely used noninvasive diagnostic technique to assess global contractile function and regional wall motion abnormalities in a variety of cardiac disorders, but in AAR LV systolic function is rarely affected. Furthermore, the conventional measures of cardiac function have shown their dependence on hemodynamic conditions and failure to detect subtle alterations in LV systolic function. More sensitive echocardiographic markers of cardiac dysfunction are required for the noninvasive detection of early signs of rejection in the transplant setting. Myocardial edema may manifest by subtle changes in relative wall thickness, accounting in part for the conflicting results on the usefulness of these echocardiography parameters to detect AAR. Serial increases in wall thickness and LV mass in cardiac transplant patients have been associated with the occurrence of rejection episodes, however, this is an insensitive marker, particularly for mild rejection. The presence of a pericardial effusion in cardiac transplant recipients is associated with a higher incidence of AAR. However, pericardial effusion occurs frequently in the post-operative period with low sensitivity and specificity to detect ACR. The presence of a pericardial effusion has been associated with acute rejection, but the sensitivity and specificity for diagnosis of rejection were only 49% and 74%, respectively. Early investigations demonstrated altered LV diastolic mechanics in the absence of systolic dysfunction in patients with AAR. Multiple studies have evaluated LV Doppler inflow indexes to detect AAR, including early diastolic (E) peak velocity, late diastolic (A) peak velocity, E/A ratio, PHT, and IVRT with not consistent results. Standard Doppler appears as a method with an excellent specificity but insufficient sensitivity. Limitations in predicting AAR are related to the influence of several parameters on these indexes including donor age, heart rate(variable in the setting of cardiac denervation), loading conditions, influence of recipient atrial contraction. In addition to transmitral Doppler indexes, several studies have examined changes in pulmonary vein flow indexes, LV diastolic flow propagation, and the myocardial performance index to detect rejection with conflicting results. Because of the relatively load independency of myocardial velocities demonstrated also in HT recipients, Tissue Doppler Imaging (TDI) has been proposed for the evaluation of LV diastolic properties due to AAR. Using pulsed wave TDI obtained from the basal posterior wall, some investigators have demonstrated that a reduction in peak systolic velocity (Sm) and peak early diastolic velocity (Em) is helpful in detecting cardiac rejection. Based on the high negative and positive predictive values for rejection of changes in diastolic velocity and serial TDI screening could guide the effective use of EMBs. Discordant results have been seen using pulsed wave TDI obtained from the mitral valve annulus: some studies have found significant decreases in systolic velocities with rejection, whereas other investigators have not. These discrepancies with regard to the clinical usefulness of these parameters may be secondary to differences in methodology (i.e., using pulsed wave versus color TDI, and using tissue Doppler Em from different myocardial walls or mitral valve annulus locations). Finally, also the E/Ea ratio, a validated parameter for estimation of LV filling pressure in the transplanted heart, has been evaluated to detect AAR, with conflicting results. Quantitative echocardiographic techniques of regional and global myocardial function, that include strain and strain rate imaging, have been used to identify subclinical LV dysfunction in a small number of cardiac transplant recipients with promising results.
Other noninvasive imaging modalities and acute allograft rejection
CMR provides excellent spatial resolution to accurately measure diastolic and systolic ventricular volumes and left and right ventricular systolic function. CMR may also be useful in the detection of AAR by its ability to quantitate changes in myocardial mass and to detect myocardial edema. CMR studies with the administration of contrast medium, gadolinium, have also been used to detect inflammatory changes in the myocardium.
Echocardiography and CAV
Advanced CAV affects myocardial properties at rest. Resting wall motion abnormalities as detected by 2-dimensional echocardiography are in general associated with low sensitivities (range 12% to 80%) but high specificity for the presence of CAV (range 69% to 100%). A few studies have examined resting TDI to detect CAV. After excluding AAR, a pulse wave TDI-derived radial peak systolic velocity (Sm value <10 cm/s) has been associated with a 97% likelihood for CAV, whereas Sm values >11 cm/s excluded accelerated CAV with 90% probability. Stress echocardiography using exercise, dipyridamole, and dobutamine has been evaluated to detect CAV. The blunted heart rate response to exercise because of cardiac denervation limits the sensitivity of exercise testing to detect CAV. In contrast, the use of pharmacologic stress testing (dipyridamole and dobutamine) to detect CAV has been shown to be more sensitive (range 50% to 100%). The specificity (range 41% to 95%) of dobutamine stress echocardiography is likely a reflection of the differences in defining the presence and severity of CAV by coronary angiography. Based on the overall results, annual evaluation with dobutamine stress echocardiography is an acceptable alternative to invasive means for detection of CAV. Preliminary studies in cardiac transplant recipients have demonstrated the validity of the flow coronary reserve as noninvasive markers of CAV.
Other noninvasive imaging modalities and CAV
Similar to stress echocardiography, several studies have examined myocardial perfusion imaging (MPI) using exercise, dipyridamole, and dobutamine to detect CAV. Although echocardiography has the advantage of lower cost, the use of MPI, especially with dobutamine, is a comparable alternative in cardiac transplant recipients given its prognostic value. MPI may be especially useful in patients with inadequate acoustic windows despite the use of a contrast echocardiography agent. The overall sensitivity and specificity range of MPI to detect CAV is broad (21% to 92% and 55% to 100%, respectively). This variability may be explained by differences in the timing of the examinations, the stressors and MPI agents used, and the variable criteria used to diagnosis CAV. MDCT coronary angiography has been shown to detect significant coronary disease with relatively high sensitivities (70% to 100%) and specificities (81% to 100%). MDCT offers the advantage of evaluating the coronary lumen as well as the wall. This may be potentially useful clinically to allow detection, grading, and follow-up of CAV, given that wall thickening and intimal hyperplasia are pathologic characteristics of CAV. Preliminary studies have demonstrated that changes in vessel wall can be accurately assessed using longitudinal and cross-sectional images of the coronary arteries to ensure adequate visualization of the concentric grey area surrounding the contrast-enhanced vessel lumen. Furthermore, up to 50% of the segments considered thickened by MDCT may be considered normal by coronary angiography, highlighting the potential for early CAV detection. There is, however, a paucity of studies correlating changes in vessel wall by MDCT to IVUS to assess its true sensitivity. Reducing heart rate enough to minimize motion artefact during MDCT could be a technical challenge in this subset of patients. As a result of cardiac denervation, the resting heart rate in cardiac transplant recipients is typically elevated, between 80 and 110 beats/min. With newer 64-slice scanner technology and multisector reconstruction algorithms, however, a heart rate of ∼80 beats/min may permit adequate visualization. Potential limitations of MDCT scanning in cardiac transplant recipients are related to the radiation and contrast media exposure associated with this technique, which is especially important given that chronic renal failure is frequent in cardiac transplant recipients. Delayed enhancement CMR may be useful indirectly for the detection of CAV, based on the ability to identify areas of delayed hyperenhancement that likely correspond to silent myocardial infarctions.