Primary transplantation for “lethal” congenital heart disease in the newborn

Perhaps the most controversial indication is for the newborn with unpalliated hypopolastic left heart syndrome. Survival rates in excess of 80% at one year may be achieved in experienced centers with primary transplantation for this condition.⁴ Similar results may be achieved with staged reconstruction, the so-called Norwood approach.⁵ “Head-to-head” comparisons of these therapies should ideally involve:

• Analysis of survival on the basis of intention to treat, with particular focus on pre-surgical mortality.
• Very long-term follow-up.
• Cost analyses.
• Quality of life measures.
• Impact of transplantation on the availability of organs for other infants with cardiac disease unsuitable for surgical palliation.

Most centers have moved away from transplantation, and towards staged reconstruction, for this group of neonates, partly in light of the high mortality prior to transplantation, which some have reported to approach 30%.⁶ In fact, only approximately 70 transplants are performed in infants each year in the United States of America. It has been estimated that only approximately 5% of infants with hypoplastic left heart syndrome or related disorders could be palliated with transplantation due to the scarcity of donor organs. A pragmatic approach would involve identification of patients with adverse risk factors for successful staged palliation, such as a miniscule aorta, or severe right ventricular dysfunction, and to reserve transplantation for this group. Of course, infants with certain anatomic variants, such as intact interatrial septum, represent poor candidates for either staged repair or primary transplantation.⁷

Recipient with prior malignancy

Survivors of childhood cancer previously treated with anthracyclines, with or without mediastinal radiation, represent a challenging group of patients. The period of time that a child should be free of disease prior to transplantation should be determined in collaboration with a pediatric oncologist, and will be dependent on the type of tumor, its stage, and the response of the disease. The risk of recurrent disease, or new onset malignancy, is poorly defined. In a recent review of 17 cases from 3 centers, only one case of recurrence was identified after transplantation, and no new malignancies were encountered.⁸ Much less experience exists with concurrent, or recently treated, malignancy,⁹ especially primary cardiac tumors. Our only experience in this regard is with one patient with primary angiosarcoma of the heart, who was treated with chemotherapy followed by transplantation. This case was associated with early recurrence.¹⁰

Pre-existing viral disease

Transplantation of children with chronic infection with either hepatitis B or C, or human immunodeficiency virus, remains controversial. Indeed, there is very little experience of transplantation in children in any of these areas, and information is generally extrapolated from data with adults, and from transplantation of other organs. Small trials of liver transplantation in patients with human immunodeficiency virus infection are underway. While most centers consider infection with this virus a contraindication to transplantation, decisions must ultimately be evidence-based rather than reflecting anecdotes, personal bias, or “gestalt”. Most importantly, we must predefine what we consider an acceptable outcome. If we believe, for example, that candidates should have an estimated one year survival of >70%, then we would not be justified in performing heart transplantation for infants and children receiving support with extracorporeal membrane oxygenation. For candidates with hepatitis B and C infection, it is important to evaluate the extent of any active hepatitis and cirrhosis. Several reports have suggested progression, or development, of chronic liver disease in many recipients having hepatitis B and C following transplantation.^11–14 The “natural” history of these infections in the population of children undergoing transplantation warrants study through multi-centric collaborations.

Transplantation for complex congenital heart disease

Children with complex congenital heart disease now represent approximately half of all referrals for consideration for transplantation. Other than the newborn with hypoplastic left heart syndrome, most have undergone multiple prior surgical palliative or corrective procedures.¹⁶ Frequently, there are stenoses, hypoplastic segments, or even discontinuities of the branch pulmonary arteries. Anomalies of systemic and pulmonary venous return, particularly in the context of isomerism or mirror imaged atrial arrangement, pose additional challenges for the surgeon. A series of reports have described specialized surgical techniques for dealing with these complex anomalies of the heart and great vessels, including abnormalities of atrial arrangement.^16–20 In experienced centers, survival has been accomplished comparable to patients with cardiomyopathy, though the donor ischemic times, post-operative stays in intensive care, and length of hospitalization are prolonged compared to patients transplanted for cardiomyopathy.^{16, 17} Failure to recognize the presence of significant aorto-pulmonary collateral circulation in patients with cyanotic heart disease may contribute to high output failure occurring soon after transplantation.²¹ This emphasizes the importance of comprehensive preoperative anatomic evaluation and planning in this difficult group of patients.

The patient with a “failed Fontan” procedure poses additional challenges related to premorbid states such as protein losing enteropathy, chronic liver disease, and pulmonary arteriovenous malformations. The latter are most common in those with a previous classical Glenn anastomosis, and when hepatic venous return is excluded from the cavo-pulmonary repair, especially in the setting of left isomerism. Resolution of small pulmonary arterio-venous malformations may occur after transplantation,²² though it would seem unlikely that large lesions will regress rapidly. In the latter setting, the patient may remain severely cyanotic after transplantation, potentially leading to primary failure of the transplanted heart from hypoxemia. It should also be recognized that increasing numbers of patients with complex palliated congenital heart disease are being referred for consideration of transplantation in adult life. It is important to note that we cannot readily extrapolate results from children to adults. In general, the results of transplantation for congenital heart disease in adults are poor compared to those in children, though the precise reasons for this are not clear.^{23, 24}

Technique of implantation

Specialized techniques for implantation are required for complex congenital defects, as outlined above. Where anatomy is straightforward, two basic techniques may be utilized. For many years, most orthotopic transplantations were performed using biatrial anastomoses, based on the original technique of Lower and Shumway.²⁵ This technique avoids individual systemic and pulmonary venous anastomoses. The recipient is left with capacious atrial chambers, comprising donor and recipient components, which contract asynchronously. It has been suggested that atrial contribution to cardiac output may be superior with near to total cardiac transplantation.²⁶ A small cuff of left atrial tissue is left in place, incorporating all pulmonary veins, and the entire right atrium is removed. Bicaval anastomoses are then performed. This technique results in more normal anatomical result. It has been suggested that it improves sinus nodal function, invokes less tricuspid regurgitation, and improves exercise performance.²⁷ While we routinely utilize this technique, there is only preliminary data in children to show that outcomes are improved.²⁸

Pulmonary vascular resistance

A key issue in the evaluation of any candidate for transplantation is pulmonary vascular resistance. If this is excessive and fixed, the ischemic, thin walled right ventricle of the transplanted heart will fail acutely, and it will not be possible to wean the patient from cardiopulmonary bypass. The upper-limit of indexed pulmonary resistance that is acceptable remains a subject of debate. In adults, pulmonary vascular resistance >5 units, and transpulmonary gradients >15 mmHg, are associated with increased perioperative mortality, and are generally considered contraindications to orthotopic transplantation.²⁹ Children appear able to undergo successful transplantation with higher indexed resistances than their adult counterparts.³⁰ Using guidelines similar to those proposed by Addonizio,³¹ we consider children with indexed pulmonary vascular resistance <10 units to be acceptable candidates, though those in the range of 6 to 10 units are considered at increased risk of acute right heart failure. Selected patients with indexed pulmonary resistance >10 units are also accepted for transplantation if a significant drop in resistance to below 10 units can be achieved with vasodilator testing. In general, we evaluate hemodynamics with general anesthesia to allow for hyperventilation, administration of 100% oxygen, and administration of nitric oxide at up to 80 parts per million. We have found that the addition of intravenous vasodilators rarely produces further immediate fall in pulmonary vascular resistance in the catheterization laboratory. If the latter remains borderline, we admit the patient to hospital and treat for 1 to 2 weeks with intravenous dobutamine and milrinone, and then restudy. Impressive falls in pulmonary vascular resistance have been achieved using this strategy. Over the last 15 years, approximately one-fifth of our candidates have been transplanted with indexed pulmonary vascular resistance >6 units. Only one death due to right heart failure has occurred, and in that instance we ignored our own guidelines, in that a baseline resistance of 14.6 indexed units fell to only 12 with vasodilator therapy.

Mechanical circulatory support

Mechanical circulatory support has been utilized to bridge infants and children to cardiac transplantation for many years.³² Extracorporeal membrane oxygenation has been most widely used in infants and small children, since ventricular assist devices are generally not available for use in this population. In general, it is hard to support children more than two weeks using this technology, due to escalating problems with bleeding, sepsis, and secondary end organ dysfunction. Pre-transplant mortality is high, and the postoperative course is often difficult. The overall chance of achieving hospital discharge after listing for transplantation is around 50%.³² A recent report suggests that persistent renal failure while on extracorporeal membrane oxygenation support is a major adverse prognostic factor for survival in children being bridged to transplantation.³³ This should represent a relative, if not absolute, contraindication to proceeding with transplantation.

Older children can be successfully bridged to transplantation using ventricular assist devices originally designed for use in adults.^{34, 35} We have successfully bridged eight children to transplant using left heart or biventricular support with Thoratec (Thoratec Laboratories, Berkeley, CA) and Novacor (Novacor Division-World Heart Coorporation, Oakland, CA) assist devices. The Thoratec system has been successfully employed in children <20 kg. In Europe, more effort has been extended to produce pneumatically driven paracorporeal devices for children. The greatest experience is with the pediatric version of the “Berlin Heart” (Mediport Kardiotechnik, Berlin, Germany). A second, similar, system has also been developed in Germany (Medos HIA VAD, Stolberg, Germany). Both systems have been used successfully to bridge small children, including neonates, to transplantation.^{36, 37} The major stumbling blocks to further development of such devices for children are the small market, prolonged time to develop and test new systems, and the very high costs of such development.

Acute cellular rejection

Acute failure of the transplanted heart is the most common cause of death in the first 30 days after transplantation and occurs with greater frequency in infant recipients.² From 1 to 5 years after transplantation, acute cellular rejection and infection are the commonest causes of death. Beyond 5 years, chronic rejection becomes the dominant cause of loss of either the heart or the patient. Understanding the timing of, and risk factors for, acute rejection are vital for designing logical strategies for surveillance and prevention. It must also be noted that there is an important distinction between acute rejection and acute dysfunction of the heart. The two terms are not synonymous. Most acute rejection in the heart is not associated with overt dysfunction, though subtle subclinical abnormalities of diastolic function are common. When cardiac failure does occur in the first few years after transplantation, acute cellular rejection is the most likely diagnosis, though other pathologies must be entertained. These include humoral, antibody mediated, rejection,³⁸ acute presentation of chronic rejection due to post-transplantation coronary arterial disease, and even viral myocarditis. There then remains a mysterious group of patients in whom acute dysfunction, sometimes reversible with augmentation of immunosuppression, occurs without evidence of significant cellular, humoral, or chronic rejection.³⁹ The pathophysiology of this phenomenon remains an enigma, making it hard to develop logical strategies for treatment.

Prevalence of acute cellular rejection

The majority of recipients in childhood will experience at least one episode of moderate or severe acute cellular rejection. Data from the Pediatric Heart Transplant Study, a research study group of 23 North American centers undertaking transplantation in children, shows that approximately two-thirds of recipients are free from rejection by 1 month after transplantation, but this has fallen to under one-third by 1 year (Fig. 2). The peak hazard, or instantaneous risk, for rejection is around 1–2 months after transplantation. Late episodes of acute rejection, that is after more than 1 year after transplantation, have also recently been studied within the consortium.⁴⁰ Episodes of acute rejection occurring late after transplantation were diagnosed in one-quarter of recipients, with a probability of freedom from late rejection of 82% at 2 years, and 73% at 3 years after transplantation. In contrast to the first episodes of rejection occurring early after transplantation, the hazard for late rejection is more constant, with an ongoing risk for as long as the cohort has been followed. This ongoing risk for acute rejection correlates with the clinical observation that recipients do not tolerate discontinuation of immunosuppression, which at this time must be considered a life long therapy.

Figure 2. Probability of freedom from acute rejection among 1114 primary transplantations recorded in the files of the Pediatric Heart Transplant Study, from 1993 to 2000. The material is used with permission of the Pediatric Heart Transplant Study.

Risk factors for acute rejection

Several recent analyses from the Pediatric Heart Transplant Study sought the risk factors for early acute rejection, late acute rejection, and rejection with hemodynamic compromise.^40–42 Older age at transplantation was the strongest predictor of risk for first rejection, and risk for an increased number of episodes of rejection within the first 6 months after transplantation.⁴¹ These data are consistent with other observations that suggest that young infants may be less prone to acute rejection.⁴³ Risk factors for the initial episode of rejection when occurring more than one year after transplantation were greater than one episode of rejection in the first year after transplantation, black race of the recipient, and older age at transplantation.⁴⁰ Risk factors for rejection with severe hemodynamic compromise, a serious complication with high mortality, were older age at transplantation and black race of the recipient.⁴²

Recently, there has been significant interest in the investigation of genetic risk factors of the recipient for outcomes regarding both the transplanted organ and the wellbeing of the patient. Most attention has focused on genetic polymorphisms for cytokines, and other genes involved in immune responses. It is now well recognized that there is wide inter-individual variation in the level of production of various pro- and anti-inflammatory cytokines.⁴⁴ Part of the explanation for these inter-individual differences lies in the presence of functional genetic polymorphisms, most commonly in the regulatory region of various cytokine genes. Using sequence specific primer-polymerase chain reaction methodology, we have recently observed that “high producers” of the pro-inflammatory cytokine TNF-α, and “low producers” of the anti-inflammatory cytokine IL10, are more likely to experience recurrent cardiac rejection in the first year after transplantation.⁴⁵ We have also investigated polymorphisms of other genes that may influence outcomes. Of particular interest are genes that influence individual responses to various immunosuppressive drugs. The multi-drug resistance gene MDR1, encodes P-glycoprotein, a transport membrane glycoprotein that acts to expel many immunosuppressive drugs to the outside of the cell. We have recently noted that certain polymorphisms of this gene, at exons 21 and 26, correlate with increased chance of weaning children from corticosteroids within the first year after transplantation.⁴⁶ It is important to note that most African-American individuals carry the polymorphisms at these locuses that are associated with decreased immunosuppressive efficacy. It has long been known that the African-American population has poorer outcomes after transplantation, with worse profiles for rejection. These kinds of genetic studies hold promise that we will be able to predict the risk of an individual patient for acute rejection, and other outcomes, thus enabling us to tailor immunosuppressive therapy to the needs of the individual child.

Diagnosis of acute rejection

The optimal method for diagnosing acute rejection remains one of the most controversial issues in the management of children undergoing transplantation. In most centers, as with adults, endomyocardial biopsy remains the gold standard for the diagnosis of acute rejection. A large number of studies in adults have investigated the role of echocardiography in the diagnosis of acute allograft rejection.^{47, 48} The almost universal use of protocols based on endomyocardial biopsy for surveillance in adult practice is testament to the fact that no echocardiographic parameter, or combination of parameters, appears able reliably and reproducibly to diagnose allograft rejection. Because of the inconvenience, greater technical challenges, and possible increased morbidity of biopsy in smaller children, there has been much interest in evaluating the role of echocardiography in children undergoing transplantation.^49–52 A number of echocardiographic parameters have been investigated, including various measures of systolic and diastolic function, changes in left ventricular wall thickness and mass, and development of new mitral regurgitation or pericardial effusion. Standard M-Mode, cross-sectional, and Doppler studies,^49–51 automated techniques to detect borders, and tissue characterization have all been evaluated.^{52, 53} Unfortunately, most studies have involved too few patients, or have not used endomyocardial biopsy as a concurrent gold standard. Boucek and colleagues did compare an echocardiographic scoring system to the results of endomyocardial biopsy in a group of children.⁴⁹ Most of the biopsies were performed beyond the first few months after transplantation. It is difficult, therefore, to draw conclusions about the use of echocardiography for surveillance of rejection in the immediate period after transplantation, when patients are most at risk for the development of acute rejection. During this period, many other factors come into play that may effect cardiac function, dimensions, wall thicknesses, and mass. These include the known effects of cardiac hypoxia or ischemia and reperfusion on diastolic dysfunction, the effects of steroids on left ventricular wall thickness and mass, as well as the influence of mismatch between the size of the recipient and the transplanted heart on ventricular remodeling.^{54, 55} It is noteworthy that Santos-Ocampo and colleagues from St. Louis⁵⁶ found M-mode echocardiography to be an unreliable means of diagnosing rejection in the first month after transplantation in infants. The controversy over the role of echocardiography for the diagnosis of acute rejection is far from resolved.

Humoral rejection

The child who has been transfused at the time of previous surgery, or to treat severe anemia, may become sensitized to a wide variety of human leukocyte antigens. These patients are at risk for the development of severe humoral rejection after transplantation.³⁸ The management of the ambulatory candidate with an elevated panel reactive antibody screen remains controversial. Prospective cross matching of the candidate serum against donor lymphocytes is rarely feasible, since most donor hearts for children are procured from distant sites where recipient serum is not available. Attempts can be made to reduce the level of recipient anti-human leukocyte antigen antibodies. Strategies that have been used to achieve this include the use of intravenous pooled immunoglobulin,⁵⁷ plasmapharesis,⁵⁸ as well as agents that may inhibit production of antibodies by B cells.^{59, 60} Cyclophosphamide, azathioprine and mycophenolate mofetil have all been used in this regard.

In the critically ill child, a negative prospective cross match is unlikely to be achieved prior to the death of the patient. A decision must then be made as to whether the first compatible donor of suitable size and blood group will be used, irrespective of the potential result of cross-matching. These candidates can be managed with pretransplantation plasmapharesis, which can then be continued intraoperatively and postoperatively. This should minimize the risk of hyperacute rejection with immediate vascular thrombosis and graft failure. Ongoing problems related to humoral sensitization may occur, including severe acute rejection often associated with graft dysfunction. The endomyocardial biopsy will frequently show marked edema and endothelial activation, but with little cellular infiltrate. There are few experimental studies to guide the treatment of humoral rejection beyond plasmapharesis.⁶¹ Induction cytolytic therapy with anti-T cell antibodies may be used, since humoral responses may up-regulate cellular immune responses. Triple drug therapy is indicated, and cyclophosphamide, mycophenolate mofetil, or rapamycin are logical choices for adjunctive therapy in the immediate period after transplantation, since they may suppress production of antibodies with greater efficiency than azathioprine. Use of intravenous gamma globulin is considered once plasmapharesis is discontinued. Only one study has been reported in children using these types of strategies. Good intermediate term results were observed.⁶² It should be noted, however, that there is increasing evidence that antibodies play a role in the development of chronic rejection.⁶³ Thus, although an aggressive approach to the management of the sensitized patient may allow for short and medium term survival, the long terms consequences of transplanting sensitized children, especially with a positive cross match, remain unknown.

Chronic rejection or post-transplantation coronary arterial disease

Coronary arterial disease subsequent to transplantation is an accelerated vasculopathy that is the leading cause of death among late survivors of heart transplantation.² The pathology differs somewhat from that of ischemic heart disease in the normal adult population.⁶⁴ Typical allograft coronary arterial disease consists of myointimal proliferation that is concentric and involves the entire length of the vessel, including intramyocardidal branches. Eventually luminal occlusion occurs (Fig. 3a,b). There may also be a pronounced inflammatory component that is not typically seen in atherosclerotic disease. Progression of disease also tends to be more rapid subsequent to transplantation.

Figure 3. Chronic cardiac rejection. (a) Selective angiogram of left coronary artery obtained 3 years after transplantation in a 10 year-old child with diastolic dysfunction. Although post-transplantation coronary arterial disease is generally a diffuse process, severe focal lesions (arrows) may also be seen. (b) Histological section of an epicardial coronary artery with advanced vasculopathy. There is severe luminal narrowing secondary to intimal proliferation (*).

Prevalence

In adults, from one-third to half of recipients have angiographic or autopsy evidence of this condition by 5 years after transplantation. “Moderate-to-severe” disease occurred in one-sixth of patients registered in the multi-centric adult Cardiac Transplant Research database maintained at the University of Alabama, Birmingham. Using the same method of diagnosis, the incidence at five years after transplantation was 6% among recipients in the Pediatric Heart Transplant Study.⁶⁵ Overall, the incidence of any angiographic evidence of coronary disease was 3, 12, and 20% at 1, 3 and 5 years after transplantation of children in this dataset. Freedom from chronic rejection at Children's Hospital of Pittsburgh is approximately 90% at 5 years after transplantation (Fig. 4). Clearly, the incidence of this serious complication will be dependent on the method of survey. Most studies have relied on angiography for diagnosis. This will tend to underestimate mild disease that may be easily detected by intravascular ultrasound.

Figure 4. Probability of freedom from chronic rejection due to coronary arterial disease among children undergoing primary cardiac transplantation at Children's Hospital of Pittsburgh from 1982 to 2002.

Pathogenesis and risk factors

Both immune and non-immune mechanisms likely contribute to the development of coronary arterial disease, though immune mechanisms are probably of central importance in children. Even neonates, with no traditional risk factors, are now recognized to be at risk for the development of coronary arterial disease. The immunobiology of chronic rejection is discussed in detail elsewhere.⁶⁶ Identification of specific risk factors for development of disease in children has recently been investigated. Among 1032 recipients at 18 centers in the Pediatric Heart Transplant Study, older recipient age, older donor age, especially beyond 30 years, and greater number of episodes of rejection in the first year, were the main risk factors for development of coronary arterial disease.⁶⁵ Donor age beyond 30 years tripled the risk of disease at 5 years after transplantation.

Detection of coronary vasculopathy

The recipient will often not experience ischemic chest pain, since the heart is denervated. Heart failure, syncope or sudden death is often the initial presentation.⁶⁷ Standard treadmill and nuclear scans for the detection of coronary arterial disease are generally insensitive in this population. It has been suggested that dobutamine stress echocardiography is the most promising noninvasive technique. Only limited experience has been reported at this time in children.^{68, 69}Furthermore, many children do not agree that this is a non-invasive test! Nausea and palpitations are common, and intravenous access is required. The sensitivity, specificity, and predictive values for detection of disease in adults are fairly high, and superior to other non-invasive methods. In addition, serial dobutamine stress studies have proved useful in predicting acute cardiac events in adults.⁷⁰ Undoubtedly further experience in this field with children will be forthcoming in the future.

Angiography has traditionally been the gold standard for detection of coronary arterial disease in children and adults. It does, however, carry serious limitations, beyond the obvious facts that it is invasive and expensive as a screening tool. Angiography is insensitive, though severe epicardial disease that portends a poor prognosis is readily apparent. It has proven very difficult to predict prognosis for patients with mild to moderate angiographic disease, many of whom will do well for several years after diagnosis. Some pediatric centers advocate listing for retransplantation at the time of diagnosis of any angiographic abnormality because of the progressive nature of the disease, and the tendency for angiography to underestimate the degree of vessel involvement. We are more conservative, and would re-list only children with either abnormal hemodynamics, rapid progression, a positive stress echocardiogram, or decreased left ventricular systolic function. Since little data is currently available regarding natural history in children, it seems appropriate to follow patients closely with serial angiography and stress echocardiography when “mild” disease is identified.

Intravascular coronary ultrasound has the greatest sensitivity for detection of graft vasculopathy. Studies in adults have reported the frequent occurrence of abnormalities on intravascular ultrasound, despite angiographically normal appearances of the coronary arteries.⁷¹ Two studies from large programs involving children have also shown a discrepancy between findings from coronary angiography and intravascular ultrasound, abnormalities commonly being predicted ultrasonically when angiography is “normal”.^{72, 73} Interestingly, in the Loma Linda experience, maximal intimal thickness was greater in children transplanted in infancy and early childhood compared to those transplanted as neonates.⁷² This observation correlates with the recently reported survival advantage for neonatal recipients.⁷⁴ At the present time, however, intravascular ultrasound should be considered experimental in children, with the risks of the procedure not yet being adequately established in this population.

Prevention and treatment of chronic rejection

No effective treatment exists for established coronary arterial disease. Bypass grafting, and coronary angioplasty, have limited utility due to the diffuse nature of the disease. Similarly, stenting is of limited utility due to intimal proliferation through the stent, and due to progression of the disease in other areas. Rapamycin coated stents have shown great promise for preventing restenosis in ischemic heart disease.⁷⁵ We are currently using them in selected children after transplantation with focal lesions. Re-transplantation is limited due to the shortage of donors, especially for the older child who will be competing with adults for donor organs.

Several pharmacological approaches have been used for prevention of coronary arterial disease after transplantation. Anecdotal reports, small clinical trials in adults, and experimental models in animals suggest that a large number of agents may be potentially useful in prevention of graft vasculopathy and coronary arterial events. These include calcium antagonists, angiotensin converting enzyme inhibitors, anti-oxidants such as vitamin E, “statins”, aspirin, and antiproliferative agents such as mycophenolate mofetil and rapamycin. The latter agent is of particular interest, since it has been shown to prevent the development and progression of graft vasculopathy in a number of animal transplantation models, including primates.⁷⁶ Unfortunately, it will take many years before it is known whether this agent, or any other, can reduce the incidence and severity of chronic rejection in human recipients of transplanted hearts. Furthermore, a major side effect of this agent is hyperlipidemia, widely considered to be a risk factor for graft coronary arterial disease. Since adherence to complex medical regimens is a significant challenge for many patients and families, it is hard to justify routine use of all these agents in the absence of proof of efficacy. On the other hand, it must be recognized that large-scale multicentric trials of most of these therapies are unlikely to ever be performed, especially in children.

Infectious complications

The spectrum of infections after transplantation in children, and their prevention and treatment, has been the focus of a recent review.⁷⁷ An increased prevalence of all forms of infection is seen. Most infections are caused by pathogens that also cause infection in non-immunocompromised children. Common examples include respiratory viruses, Streptococcus pneumoniae, and Varicella Zoster virus. All infections that may be seen in non-immunocompromised patients can cause greater disease severity in the recipient of a transplanted heart, and, thus, strategies to achieve prevention and early treatment are imperative. Of particular note in this respect are infections due to cytomegalovirus and Epstein-Barr virus, which only rarely cause severe disease in the immunocompetent host. More rarely, opportunistic infections are seen such as that due to Pneumocystis carinii (Fig. 5). Although most infections are well tolerated, it should be noted that infection ranks comparable to early graft failure and rejection as the main causes of death after heart transplantation in children (Table 1).

Figure 5. Rapid evolution of infection with Pneumocystis carinii in a 15 year-old child presenting with cough. The chest radiographs are taken 48 hours apart.

Table 1. Causes of death after pediatric heart transplantation. Data from 1114 primary transplants in the Pediatric Heart Transplant Study 1993–2000 (total 248 deaths). (Used with permission.)

Relatively few major advances in the prevention and management of infection after transplantation have been realized in recent years. Two important exceptions are infections with cytomegalovirus and Epstein-Barr virus (see below). Cytomegalovirus is an important cause of morbidity after solid organ transplantation. Symptomatic disease in children is relatively unusual.⁷⁸ Some groups have suggested that cytomegalovirus infection is a risk factor for development of coronary arterial disease.⁷⁹ It is unknown if this is true in children, though it is of interest to note that, in a recent intravascular ultrasound study from Loma Linda, an increase in maximal intimal thickness and intimal index correlated with cytomegalovirus seropositivity.⁷² Advances in diagnosis, prevention, and treatment of cytomegalovirus have been realized in the last decade. These include widespread availability of ganciclovir as an intravenous and oral antiviral agent, the development of cytomegalovirus immune globulin preparations, and assays for rapid diagnosis of active infection from peripheral blood samples, such as pp65 antigenemia test and polymerase chain reaction detection of viral genome. The optimal strategies for preventing infection with cytomegalovirus remain to be determined in children.

Epstein-Barr virus infections and post-transplant lymphoproliferative disorders

The last few years have seen major advances in our understanding of the nature of Epstein-Barr infection in the immunocompromised host. The immunobiology, nomenclature, management, and possibilities for prevention of symptomatic Epstein-Barr virus disease and lymphoproliferative disorders occurring after transplantation have been the topic of several recent reviews.^80–85 Recent review of the Pediatric Heart Transplant Study database identified 56 cases among 1184 primary transplants (4.7%) at 19 North American centers.⁸⁶ Almost nine-tenths were driven by Epstein-Barr virus. The pathology of early onset disease was usually polymorphic, though frequently monoclonal, and almost invariably associated with primary infection with Epstein-Barr virus. Post-transplant lymphoproliferative disorders developing late, usually beyond 3 years, may be polymorphic, but are often monomorphic and lymphomatous.

The clinical manifestations of lymphoproliferative disorders occurring after transplantation are protean.⁸⁰ Fever and malaise are the most common presenting symptoms in children. In contrast to other transplanted solid organs, disease is rarely seen within the allograft. The majority of children who develop these disorders are diagnosed with lung nodules, mediastinal or abdominal adenopathy, or gastro-intestinal involvement. Abdominal pain, vomiting, and diarrhea are the major signs of gastro-intestinal involvement. Bleeding and perforation may also occur.

The recent development of quantitative Epstein-Barr virus–polymerase chain reaction on peripheral blood samples has proven to be a very useful simple technique for diagnosing and monitoring disorders driven by this virus.^{81, 82} The development of very high viral loads should alert the physician to look for sign or symptoms of disease, even in the asymptomatic patient. We routinely monitor seronegative patients at transplantation who are at risk by monthly polymerase chain reaction for the first 6–12 months after transplantation, and at the time of development of symptoms suggestive of the disease. This has enabled us to diagnose primary Epstein-Barr virus infection at a very early stage, that is before the development of symptoms. Studies are underway to determine if preemptive decrease in immunosuppression, and/or use of immunoglobulin preparations with high titers of anti-Epstein-Barr virus antibodies, can safely reduce the risk of lymphoproliferative disease when high viral loads are detected in an otherwise healthy child. We have also used weekly monitoring of polymerase chain reaction to assess response to treatment in patients with symptomatic infection by the Epstein-Barr virus or those with lymphoproliferative disease.^80–82 Response to therapy has been associated with a precipitous fall in viral load, even before radiographic improvement has occurred. Furthermore, preliminary experience has suggested that monitoring of the viral load may help predict time to development of rebound rejection after cessation or reduction of immunosuppression. When very high viral loads persist, we have not observed acute rejection in recipients of either transplanted hearts or livers having lymphoproliferative disorders. Fall of the viral load to <200 genome copies/10⁵ peripheral blood mononuclear cells has been associated with development of rebound rejection in several patients. This preliminary experience suggests that serial monitoring by quantitative polymerase chain reaction may help determine appropriate changes of immunosuppressive therapy and timing of endomyocardial biopsy in patients being treated for lymphoproliferative disorders occurring subsequent to transplantation.

The mainstay of treatment in children remains reduction or temporary discontinuation of immunosuppression. In our recent experience, this course of action is associated with “cure” in approximately four-fifths of cases with polymorphic histology. Rebound rejection is common, and must be sought by increased echocardiographic and endomyocardial biopsy surveillance. It has been rare for the disorders to recur, even if treatment with pulsed corticosteroid therapy for rebound rejection is required. The monomorphic forms are less likely to respond to a reduction in immunosuppression, but it has been our practice to attempt this as first line therapy in selected patients, since complete remissions may be observed with this therapy alone. Monomorphic disease may require treatment with conventional chemotherapeutic agents, as do overtly malignant lymphomas, such as Burkitt's lymphoma.

Several other therapies have been proposed in recent years. Two therapies are of particular interest, as they may effectively treat the condition without requiring global immunocompetence to be reestablished. In this way, rebound rejection may be avoided. The first is use of monoclonal antibodies directed against B cell antigens. More than nine-tenths of lymphoproliferative disorders are of recipient B cell origin after heart transplantation. In the United States of America, one such agent, rituximab (Rituxan, Genentech Inc./IDEC pharmaceuticals) is approved for treatment of certain types of adult B cell non-Hodgkin lymphomas. This is a human/mouse chimeric monoclonal antibody directed against the CD20 antigen carried on almost all B cells. We have found that most lesions express this antigen. Several reports have now suggested that this agent may be a useful therapy for B cell post-transplant lymphoproliferative disorders (Fig. 6).^87–89 A trial of this agent in children with refractory disease is in progress. A second novel approach is cellular immunotherapy, whereby the patient is given an infusion of cytotoxic T-lymphocytes directed against Epstein-Barr virus-specific antigens. This should result in control of the proliferation of the B cells infected with the Epstein-Barr virus, but without the risk of rejection. The use of Epstein-Barr virus-specific cytotoxic T cell therapy has already been developed, and applied successfully, for the management of post-transplant lymphoproliferative disorders in recipients of transplanted bone marrow.⁹⁰ Efforts to achieve this in recipients of transplanted solid organs are under way in several centers, and preliminary data suggests this should be feasible.⁹¹

Figure 6. Response of refractory post-transplant lymphoproliferative disorder to the anti-CD20 monoclonal antibody rituximab. The disease progressed despite withdrawal of immunosuppression (panels a and b) (arrows), but responded completely to a 4-week course of the antibody (panel c). Panel (d) shows a histological section of the lesion stained with an anti-CD20 antibody. There is dense staining of the membranes of the B lymphocytes for CD20, the target of the therapeutic antibody.

Reports of malignant neoplasia other than lymphoma are rare in transplanted children.⁹² Unfortunately, the risk of malignancy in immunosuppressed patients appears to persist through the lifetime of the patient. This is of particular concern for children destined to receive lifelong immunosuppression. Other than lymphoma, squamous cell carcinomas and other skin neoplasia are the commonest malignancies seen after heart transplantation.

Non-immune complications

Our focus on rejection, infection, and malignancy emphasizes the careful balance that must be achieved between over and under-immunosuppression. In addition to these complications, recipients experience a wide array of other complications that are unrelated to their immune status. Most reflect end-organ toxicities of immunosuppressive therapies.⁹³ Systemic hypertension is a common complication, especially in those maintained on cyclosporine-steroid based immunosupression. It generally responds to standard anti-hypertensive agents, but on occasion can be quite resistant to treatment. Hyperlipidemia is also common in children after heart transplantation.^{94, 95} Recent data show that children treated with cyclosporine experience more adverse lipid profiles compared to those treated with tacrolimus, and this effect occurs independent of the effects of corticosteroids.⁹⁶ Treatment with statins is generally quite effective, though rare cases of rhabdomyolysis indicate the need for careful monitoring during therapy. Dose-related and idiosyncratic neurological complications of current immunosuppressive regimens include seizures, tremors, paraesthesias, and even encephalopathies.⁹⁷ Common metabolic complications include hyperkalemia, hyperuricemia, hypomagnesemia and, most importantly, diabetes mellitus.⁹⁸ The relative diabetogenic effect of cyclosporine and tacrolimus is the subject of considerable controversy.⁹⁹ We have found that the development of diabetes is rare subsequent to transplantation in pre-adolescents.¹⁰⁰ Adolescent age, obesity, and the need for high doses of corticosteroids, have been major risk factors in our population. The diabetes reflects both high peripheral insulin resistance and pancreatic insufficiency. This complication is likely to become less frequent with newer immunosuppressive regimens that allow for lower doses of calcineurin inhibitors and corticosteroids.

One of the most worrisome end-organ toxicities is that of progressive renal dysfunction due to calcineurin inhibitor renal toxicity. In our own experience, up to one-twentieth of recipients have developed severe renal dysfunction or end-stage renal failure. We have observed no significant difference in renal function between cyclosporine and tacrolimus-based immunosuppressive regimens up to 7 years after transplantation.¹⁰¹ With increasing numbers of patients surviving into the second decade after transplantation, the number of patients developing end-stage renal failure will inevitably increase. We have also found that estimates of creatinine clearance based on serum creatinine levels significantly underestimate the extent of renal dysfunction when compared to nuclear medicine measurements of glomerular filtration rate.¹⁰¹ Indeed, approximately two-fifths of our patients during long-term follow-up have glomerular filtration rates >2 standard deviations below mean normal (Fig. 7). Increased use of adjunctive therapies without renal toxicity, for example rapamycin and mycophenolate mofetil, should allow for reduced dosing of tacrolimus and cyclosporine. This will hopefully reduce the extent of chronic nephropathy associated with these agents. Some groups have even attempted calcineurin free immunosuppressive regimens, though this has not yet been attempted in cardiac recipients. Use of multiple immunosuppressive agents with differing end-organ toxicities, each at low dosage, represents a logical approach to minimizing side effects while maintaining immunosuppressive efficacy. Such complex regimens are only suitable for the most motivated and compliant of patients. We struggle with compliance using the simplest of monotherapy regimens in our population of adolescents. Studies are also underway on a new generation of potent calcineurin inhibitors that appear to exhibit minimal renal toxicity.¹⁰² This represents a potential major advance in immunosuppressive therapy.

Figure 7. Renal function during long-term follow-up of children undergoing heart transplantation at Children's Hospital of Pittsburgh. Glomerular filtration rate (GFR), assessed using the Tc-99m DTPA technique, is shown for 69 patients with serum creatinine below 2 mg/dl. Although end-stage renal failure is rare, approximately two-fifths of recipients have filtration rates more than 2 standard deviations below the mean normal.

Antibody induction therapy

The use of monoclonal or polyclonal induction therapy, such as OKT3 or antithymocyte globulin preparations, remain highly controversial in both adults and children. Single and multicentric randomized trials in adults have failed to resolve the question, and no randomized multicentric studies have been performed in children. In adults, the main benefit has been to delay onset of rejection, but no enduring benefits have been established. The influence of induction therapy was investigated in a recent retrospective multicenter study of 465 children.¹⁰³ Patients receiving polyclonal rabbit antithymocyte serum had overall lower mortality and reduction in deaths due to rejection when compared to patients receiving no induction or OKT3. This study was not randomized, and it should be noted that the polyclonal antibody was used in some of the most experienced centers for transplantation of infants and children in the United States of America. This may have contributed to the improved survival. It is therefore hard to draw firm conclusions about the efficacy of antibody induction. The results do, however, stress the need for randomized studies of induction therapy in children, both for establishing efficacy and safety. There are no reported trials of induction therapy with the two new monoclonal antibodies directed against the α-chain of the interleukin-2 receptor (basiliximab and daclizumab) in children undergoing heart transplantation. In adult transplantation, these agents have proven to have excellent profiles for adverse events, and have been shown to reduce the frequency and severity of episodes of early acute rejection.^{104, 105} They are likely to be increasingly used in children in the next few years, and may help facilitate avoidance of steroids, or their early withdrawal.

Maintenance immunosuppression

All immunosuppressive regimes in children are based around the use of a calcineurin inhibitor, either cyclosporine or tacrolimus. Approximately four-fifths of recipients currently receive cyclosporine, with the remainder being on tacrolimus.² The proportion of patients receiving tacrolimus is gradually increasing, though it should be noted that the two agents have not been compared in randomized trials in children after transplantation of thoracic organs. Most centers also use a second or third agent.² Azathioprine and corticosteroids are the most commonly used adjunctive therapies, though there is increasing use of mycophenolate mofetil. There have been no randomized trials of mycophenoate in children, but there has been one large multicentric trial in adults after cardiac transplantation.¹⁰⁶

This was a randomized, blinded study from 28 centers in Australia, Europe, and North America comparing therapy consisting of cyclosporine, corticosteroids, and azathioprine to therapy with cyclosporine, corticosteroids, and mycophenolte mofetil. There was a very high withdrawal rate from the study, approximately two-fifths of each group. Survival and rejection were similar in the 2 randomized groups. In patients who actually received mycophenolate, there was a statistically significant reduction in mortality at 1 year, 6.2% as opposed to 11.4% for those receiving azathioprine. Survival at 3 years was 88% in those treated with mycophenolate, and 82% in the patients getting azathioprine. There was also a reduction in the requirement for treatment of rejection at 1 year, and a trend to fewer patients having greater than, or equal to, grade 3A rejection, or needing cytolytic agents in the group treated with mycophenolate. Increasing use of this agent can be anticipated for children in the coming years.

A number of centers have tried to minimize the use of corticosteroids, in an attempt to reduce corticosteroid morbidity.^107–109 Approximately three-quarters of children receive corticosteroids at the time of hospital discharge, but this has fallen to half by 3 years from transplantation.² There are no randomized trials of routine use of steroids versus their avoidance in children subsequent to heart transplantation. It is clear, however, that withdrawal is feasible in many, if not most, children in the medium term of 6 months to 2 years after transplantation.^{15, 110} Use of antibody induction therapy, tacrolimus as primary immunosuppressant, and newer adjunctive agents such as mycophenolate mofetil and sirolimus, may enhance the ability to achieve complete avoidance or early withdrawal of steroids.

Rapamycin, or sirolimus, is a newly approved agent with potent immunosuppressive properties.¹¹¹ The main concerns have been hyperlipidemia and marrow suppression. Need for discontinuation of this agent due to side effects has been rare in adults, but a significant proportion of patients will require statins to control hyperlipidemia. There is increasing evidence that this agent can be safely used with both cyclosporine and tacrolimus, despite the fact that it targets the same binding protein as tacrolimus. A plethora of studies in adults is under way to assess the role of rapamycin agents as adjunctive therapies with calcineurin inhibitors after all forms of solid organ transplantation. Some groups have studied use of high dose sirolimus with complete calcineurin avoidance after renal transplantation. Such trials are unlikely in cardiac transplantation, since the consequences of failure of this therapy could be catastrophic. There is likely, nonetheless, to be increasing interest in this agent as a calcineurin-sparing therapy for patients with serious complications of tacrolimus or cyclosporine therapy, such as significant renal dysfunction.

The benefits of tacrolimus, as opposed to regimes based on cyclosporine, remains one of the most controversial issues in transplantation in children. Most experience has been gained at the University of Pittsburgh, where tacrolimus has been used since 1989.^{15, 98, 100, 101, 112–116} This extensive experience in over 200 children suggests a slight delay in time to first rejection, less episodes of rejection in the first year, and less need for lympholytics beyond the early period after transplantation. Furthermore, there was lower requirement for long-term steroid therapy, and many patients were managed with tacrolimus monotherapy by the end of the first year.¹⁵ The cosmetic results were also much better, with no evidence of hirsuitism, coarsening of facial features, or gingival hyperplasia. Renal function has been comparable between children treated with tacrolimus and cyclosporine,¹⁰¹ while hypertension and hyperlipidemia tend to be less prevalent in those treated with tacrolimus. One very important caveat should be noted. The comparison with patients receiving cyclosporine was based on historical controls, and other factors may have contributed to improved profiles for rejection in the more recent cohort treated with tacrolimus. Unfortunately, there is little experience to draw on from adults. No pivotal trials comparing these two agents have been performed in adults undergoing cardiac transplantation.