Update on Cardiac Transplantation Pathology

Heart transplantation remains the most effective therapy for end-stage heart disease of coronary and noncoronary etiology, with continued improvement in survival during the years. The most common indications for cardiac transplantation in the adult have not changed in the last 3 decades; 85% of cases are roughly equally divided between coronary heart disease and nonischemic cardiomyopathies.1 In the pediatric age group, congenital heart disease is the leading diagnosis for recipients younger than 1 year old. Cardiomyopathy and congenital heart disease are the two most common indications for transplantation in children.2 

The 10-year survival rate after cardiac transplantation currently approaches 50% and more in high-volume centers.1,3 The success of heart transplantation, for the most part, has been achieved through better understanding of the immunology of transplant rejection and the application of strategies for the recognition, treatment, and prevention of rejection. In the early years of cardiac transplantation, failure resulted from a high incidence of acute cellular rejection that limited graft survival. The signs and symptoms of acute cellular rejection are often vague and there are no serologic markers of cardiac allograft rejection. The treatment of rejection, in turn, was often complicated by infection, malignancy, and drug toxicities that result from the difficulty in titrating immunosuppression to the desired end point according to the severity of rejection. The introduction of percutaneous transvenous endomyocardial biopsy by Caves et al4 in 1973 provided an objective means of diagnosing rejection and allowed for careful monitoring and prompt treatment of cardiac allograft rejection.

Endomyocardial biopsy (EMB) remains the gold standard for rejection surveillance in the heart transplant patient.5 It has a high sensitivity and specificity for the diagnosis of acute cellular rejection.6,7 There are currently no cardiac imaging modalities or serum markers that can replace the performance of surveillance biopsies in the posttransplantation care and management of these patients.8 

Ideally, an initial biopsy of the donor heart should be obtained in the operating room at the time of transplantation. This biopsy can be valuable because it provides a means to assess the status of the donor myocardium for hypertrophy, ischemia, or the presence of any pathologic process such as myocarditis. The frequency of posttransplant surveillance biopsies varies highly between different institutions. Typically, surveillance biopsies are performed once weekly for the first month, every 2 weeks for the second month, and every 6 to 8 weeks between the third and 12th months. After the first year, the frequency can be decreased to quarterly, biannually, or annually. In some centers, protocol biopsies are not done after 2 or more years unless there is a clinical suspicion of rejection. If rejection is diagnosed, the patient is treated and undergoes repeat biopsy after 1 to 2 weeks.

The procurement of the tissue is made with a bioptome introduced from either the jugular or femoral vein to sample the right ventricular septal wall. Bioptomes are available in different sizes; therefore, the size of the pieces of tissue retrieved will differ slightly. The common sizes used are 7 F (French) and 9 F in adults and 3 F, 5 F, and 7 F in pediatric-age patients (Figure 1, A through D).9 

To prevent introducing artifacts in EMB, the tissue should not be allowed to sit on filter paper, gauze, or any other surface that is impregnated with saline or other solutions that are not iso-osmotic, for a prolonged period of time. The tissue should be fixed immediately in the desired fixative, the most commonly used being 10% phosphate-buffered formalin that has been allowed to reach room temperature (25°C). Cold fixative enhances contraction band artifact. To avoid crushing artifacts, the tissue should not be handled with forceps or divided with a scalpel. The cardiac catheterization suite personnel should not triage the tissue based on gross appearance. All the pieces obtained should be submitted because they may have valuable information when examined histologically. Pieces that look white, suggesting that they are made up of thick endocardium, or pieces that look like blood clot may harbor a piece of myocardium in their core. Tissue is not routinely fixed in glutaraldehyde for electron microscopy of allograft biopsies.

In the 1990 International Society for Heart and Lung Transplantation Working Formulation of Cardiac Allograft Pathology (ISHLT-WF1990), 4 to 6 pieces of tissue, depending on the size of the bioptome used, were required for light microscopic evaluation.10 Because acute cellular rejection is not uniformly distributed in the heart, it is important to take multiple samples during the biopsy procedure. It has been shown that if 3 biopsy pieces taken show no rejection, there is a 5% and 0% chance of missing a mild and moderate-to-severe rejection, respectively. However, if 4 pieces are examined, the false-negative rate of mild rejection is further reduced to 2%.11 Other investigators have suggested that the extent of infiltration is also important. Where mild rejection is the most severe grade observed in 3 or 4 fragments, the probability of missing moderate or severe rejection is 25.4% and 28.2%, respectively.12 The 2004 revised working formulation (ISHLT-WF2004), however, currently recommends an absolute minimum of 3 biopsy pieces for evaluation, each of which must contain at least 50% myocardium and exclude a previous biopsy site or scar.13 Studies of sensitivity to detect rejection with only 3 biopsy pieces using the current grading system have yet to be performed. Specimens that do not meet these criteria should be diagnosed as “inadequate biopsy.” If rejection is noted in a biopsy of fewer than 3 evaluable pieces, the rejection grade may be indicated in a diagnosis comment with the emphasis that a higher grade of rejection cannot be ruled out.

In addition to the demographic data of the patient, the gross description should include the number of tissue pieces, an aggregate measurement with the average size, and color. Careful gross examination provides, in most instances, important information regarding the presence of myocardium, thickened endocardium, adipose tissue, blood clot, or chordae tendineae (Figure 1, E and F).14 It is good practice to state the number of pieces submitted in the requisition form to be verified on gross examination and always correlated with the number of pieces present in the paraffin block and in the hematoxylin-eosin–stained slides.

The current working formulation suggests a minimum of 3 step levels for microscopic examination.13 No special stains are routinely required for evaluation. Unstained slides can be cut and saved for immunohistochemical staining if needed.

One or more pieces of tissue can be snap-frozen for immunofluorescence or other additional study (such as in situ nucleic acid hybridization, in situ polymerase chain reaction, and gene expression profiling) depending on the needs of a given patient and any research protocol used by the institution.

Following a careful freezing protocol is important in order to achieve the best preservation of morphology possible. The ISHLT-WF1990 suggests freezing 1 biopsy piece in OCT freezing compound (Miles Inc, Diagnostics Division, Elkhart, Ind). There is no specific recommendation in the ISHLT-WF2004 regarding either the manner of freezing or the number of biopsy pieces to be frozen. In our institution, 4 biopsy pieces are routinely obtained and all pieces are frozen. The tissue is quickly and gently blotted to remove any excess moisture before embedding them on a chuck containing partially frozen OCT. After proper orientation, the specimen is fully covered with OCT and submerged in liquid nitrogen until frozen. Three step levels are cut for hematoxylin-eosin staining. This technique yields excellent frozen sections that are comparable to those obtained from paraffin sections. Additional slides can be obtained for the application of immunoperoxidase and immunofluorescence studies. The tissue is then kept frozen and stored at −80°C for future study.

As in any other solid organ, cardiac rejection can result from humoral and cellular rejection. These are, in turn, subclassified into hyperacute, acute, and chronic rejection on the basis of mechanism and duration of the process.

Hyperacute rejection is graft injury triggered by preformed antibodies and occurs rapidly after implantation of the graft, usually within minutes to hours. In the older literature, hyperacute rejection has also been referred to as humoral rejection, vascular rejection, and antibody-mediated rejection. This type of rejection is extremely rare in the current practice of allograft cardiac transplantation. The morphologic findings are well described in experimental discordant xenografts15 with similar findings in autopsy cases of cardiac allograft recipients.16 Predisposing factors that may play a role are preformed antibodies to epitopes of the ABO and HLA systems and vascular endothelial cells,17 previous pregnancies, multiple surgeries with the use of blood products and, especially, previous cardiac or other organ transplants. The pathogenesis of hyperacute rejection is believed to be an antibody-mediated activation of the complement cascade, producing severe damage to the endothelial cells, as well as platelet activation followed by the clotting cascade and thrombosis. Although the widely accepted concept is injury to the capillary network of the graft, some investigators have suggested that endothelial damage occurs primarily in cardiac venules, resulting in venular thrombosis.18 On gross examination, the heart is swollen and it is dusky on external inspection. The ventricles are dilated with scattered hemorrhages, mostly in the subendocardium. Histopathologic changes include swelling of the endothelial cells, vascular thrombosis, extravasation of red blood cells, prominent interstitial edema, and subsequent polymorphonuclear inflammatory infiltrates followed by tissue necrosis. These changes initially occur focally but rapidly spread through the organ. Immunohistochemical studies may show deposits of immunoglobulin (Ig) M, IgG, and complement in the vessel walls as well as fibrin deposits.

Morphologically, acute cellular rejection consists of a mononuclear inflammatory infiltrate that is predominantly a T-cell–mediated response directed against the cardiac allograft. In severe cases, there is also participation of granulocytes in the rejection process. Characterization of the phenotype of lymphocytes in cardiac biopsy tissue has shown no good correlation between the extent and composition (CD4:CD8 ratio) of T lymphocytes infiltrating the graft and the histologic grading of rejection.19,20 However, other studies report a good correlation between the mean number of CD8⁺ T cells and the severity of rejection grade.21 The discrepancy in these studies may be related to the fact that the immune response to the allograft is a continuous process in flux that is usually dissected in small “time-lapsed” views for pathologic study. Some support to this notion is provided by the observation that if subsets of T lymphocytes are further classified on the basis of the presence of naive cells (CD45RA) and memory or activated cells (CD45RO), naive cells of the CD4 phenotype are more abundant in biopsy tissue during mild rejection. A shift toward activated CD8 phenotype is seen in moderate rejection.22 An increase in the number of antigen presenting cells (ie, macrophages and dendritic cells) is also observed as a function of the severity of rejection.23–26 B-cell infiltrates are rarely present in mild rejection. However, a substantial increase in activated B lymphocytes and natural killer cells are seen in moderate rejection, suggesting their important role as promoters and effectors of cellular rejection.26 

Historically, several methods to assess the histologic grade of rejection have been used by different transplant centers and will not be reviewed here. In 1990, the ISHLT published a standardized international grading system for the purpose of effectively communicating outcomes in multicenter drug trials and among institutions using different treatment regimens. The grades proposed in the ISHLT-WF1990 were mainly based on the amount of inflammatory infiltrate and the presence of myocyte damage.10 The absence of cellular rejection was called grade 0 (Figure 2, A and B). Because rejection is a patchy process, the severity of inflammation may differ from one fragment to the next. Rejection is generally graded on the worst area of involvement. The pattern of inflammatory infiltration was reflected in the subdivisions A and B in grades 1 and 3. In mild rejection, it played a minor role and does not imply that a diffuse pattern (1B) is worse than a focal infiltrate (1A) (Figure 2, C through H).27 It must also be noted that the grading of rejection was designed to assess rejection in endomyocardial biopsies and not the whole grafts.

As this grading scheme was widely adopted after its publication, variability in the interpretation of histologic grading among pathologists became evident and resulted in a lack of consensus with regard to the treatment of specific grades of cellular rejection. In 2001, the Banff Allograft Pathology Group invited pathologists, cardiologists, and cardiac surgeons to discuss their experiences after more than 10 years of using the ISHLT-WF1990. These discussions pointed out some of the more difficult issues for clinical practice and for use of the pathology information as end points in clinical trials.5 In 2004, under the direction of the ISHLT, a working group composed of an international, multidisciplinary team of subspecialists in cardiac transplantation met to review the ISHLT-WF1990 definitions of cellular and antibody-mediated rejection, identify areas of difficulty in interpreting transplant biopsies, and revise the grading system. There was strong consensus that any changes in the formulation should reflect current pathologic practice and should not affect the grading of historic samples. The issue then was not one of changing the 1990 ISHLT grading scales, but one of more clearly defining how pathologists and cardiologists should interpret the grading system.

A major controversy in the ISHLT-WF1990 is the diagnosis and clinical significance of grade 2 rejection (Figure 2, I and J).28,29 It is a grade that has been used in many transplant centers as a discrete defining point in therapeutic decisions. The misdiagnoses of grade 2 lesions by pathologists and the clinical data indicating that grade 2 rejections resolve without treatment in the majority of cases prompted the working group to now include grade 2 rejection with the revised mild rejection category. The old grade 3A (Figure 2, K through N) has been reclassified as grade 2R in the new working formulation (Table). Disagreement in the diagnosis between grade 3B (diffuse, borderline, severe acute rejection) and grade 4 (severe acute rejection) (Figure 2, O through R) also occurred previously as both of these can show the same severity of diffuse destructive infiltrates. The difference rests mainly on finding additional neutrophilic infiltrates and demonstrating edema and hemorrhage in the biopsy. It seemed more logical then that grades 3B and 4 were placed together in the severe category of the revised grading system because these minor discrepancies do not affect clinical therapeutic decisions.

The different histologic grades in the revised ISHLT-WF2004 classification are indicated by a suffix, “R” (Table). Absence of inflammation is reported as no rejection. A perivascular or interstitial infiltrate of mononuclear cells without architectural distortion is considered mild rejection. A focus of inflammation with myocyte damage, previously termed grade 2 in the ISHLT-WF1990 classification, has been incorporated in the mild rejection category. Moderate, intermediate-grade rejection consists of 2 or more foci of mononuclear cell infiltrates associated with myocyte damage. Eosinophils may be present in moderate rejection. Severe, high-grade rejection is a diffuse process with multiple areas of myocyte damage and often a polymorphous inflammatory infiltrate that may be accompanied by edema and hemorrhages. A comparison of the 1990 working formulation and the revised grading system is presented in the Table.

Although an enormous effort has been put forth to create a standard method for grading rejection that is easily reproducible, there were some controversial points that have been identified by both pathologists and clinicians in using the ISHLT-WF1990 and these warranted further clarification in the revised grading scheme.5 Some of these controversies are discussed in the following sections.

A major source of discordance in histologic grading is the criteria used for the interpretation of “myocyte damage” in light microscopy, which is a required feature in higher grades of rejection.5 The morphologic spectrum of myocyte damage is wide and has subtle changes that can be difficult to ascertain. Various forms of myocyte injury described by experienced cardiac pathologists include vacuolization, perinuclear halo, ruffling of the cytoplasmic membrane, irregular myocyte border, splitting or branching of myocytes, and myocyte encroachment with partial disruption of the myocytes.6,30 Hypereosinophilia and nuclear pyknosis would indicate myocyte necrosis. Ultrastructural studies have shown that actual myocyte necrosis is rare, and reversible myocyte injury and myocyte regeneration occur even in moderate-to-severe acute cellular rejection.31–33 In the revised working formulation, myocyte damage is described as “clearing of the sarcoplasm and nuclei with nuclear enlargement and occasionally prominent nucleoli.” 13 Architectural distortion, myocyte encroachment with irregular myocyte borders, and myocyte dropout also frequently indicate myocyte damage in cellular rejection.

One of the criticisms in relying on EMB to monitor rejection is the low interobserver agreement in the diagnosis of grade 2 rejection.34,35 A corollary to this is the controversy of whether or not grade 2 rejection exists. Recognition of a grade 2 lesion is indeed problematic because of the obvious implications for therapy. Earlier on, most centers treated moderate rejection (grade 2 or higher) with adjustment in the immunosuppressive regimen. It is believed that a major source of confusion in grade 2 rejection is the difficulty in distinguishing the histologic features of this grade from Quilty lesions. Quilty lesions, named after the first patient in whom they were observed at Stanford University, are also known as endocardial lymphocytic infiltrates, which we think is a better term (Figure 3, A through D).36,37 These are collections of predominantly T lymphocytes with admixed B cells, occasional macrophages, and plasma cells seen in the endocardium of transplanted hearts that vary in size from 0.007 to 1.89 mm².38 (The detailed pathology of endocardial lymphocytic infiltrates is discussed under “Redefinition of the Quilty Effect.”) Small capillaries, sometimes with prominent endothelial cells, and dense endocardial collagen (Figure 3, B) are seen within the infiltrate and are diagnostically useful clues. Quilty infiltrates can extend deep into the subjacent myocardium and the lesion is designated type B in the ISHLT-WF1990 (Figure 3, C and D). Quilty B lesions can be big and may be associated with architectural distortion that does not represent acute rejection. One may imagine how a tangential section through the deeper (myocardial) end of a Quilty B lesion may show inflammatory infiltrates with myocyte encroachment that can easily be mistaken for moderate rejection if only a few levels of section are examined. However, if additional sections are made, one can usually ascertain the continuity of such a lesion from the myocardium to the endocardium. This type of artifact has prompted some observers to question whether or not grade 2 cellular rejection even exists.39 Our personal experience shows that sectioning through the entire tissue block and examining alternatively stained slides almost always resolves the question (Figure 4).

Another solution offered to this problem is to stain the biopsy section with antibodies to RANTES (regulated on activation, normal T cell expressed and secreted). This is helpful in differentiating a focus of cellular rejection from Quilty B lesions because the RANTES-positive cells are more abundant in acute rejection.40 

In the ISHLT-WF1990, the inflammatory infiltrates are called “aggressive” but are not further defined. Pathologists have difficulty in determining what is meant by “large aggressive lymphocytes.” This descriptive term is therefore deleted in the current grading system. Immunostains for phenotyping inflammatory cells are not routinely performed for diagnostic or prognostic purposes.

The following sections show morphologic findings that may be confusing for the novice pathologist in the differential diagnosis of rejection. Some of these features do not represent rejection but need to be clearly recognized. Furthermore, the ISHLT-WF2004 requires that these features be recorded in the report.

The presence or absence of ischemic damage should always be documented. The ISHLT-WF1990 makes a distinction during allograft monitoring between ischemia commonly seen in the biopsy up to 3 weeks posttransplant representing perioperative injury (ischemia A) and late ischemia that occurs after 3 or more months (ischemia B). In the revised grading system, ischemia is divided into early (up to 6 weeks) and late ischemic injury (Figure 5). Late ischemic injury may explain cardiac allograft dysfunction secondary to severe allograft atherosclerosis.

Perioperative ischemia is seen in a majority of transplanted hearts and is strongly associated with prolonged total ischemic time.41 Other causes of ischemic injury include events that affect the donor such as catecholamine discharge, pressor therapy given during acute care, severe donor trauma, reimplantation damage, or early postoperative damage. In the early stage, it consists of subendocardial foci of myocytes showing coagulation necrosis (with or without contraction bands) and macrophages with variable amounts of polymorphonuclear leukocytes. These areas are usually sharply demarcated with necrotic myocytes occurring in small groups and highlighted by staining with Masson's trichrome. Some lesions can lack an acute inflammatory reaction (Figure 5, A and B). Ischemic foci may persist for several weeks because of a depressed inflammatory response in these immunosuppressed patients. In the healing phase, these ischemic foci usually show pigment-laden macrophages with a few lymphocytes, a somewhat loose connective tissue stroma, and scant granulation tissue (Figure 5, C). Once they mature, ischemic lesions are indistinguishable from scars produced by previous endomyocardial biopsies (Figure 5, D and E).

Ischemic injury should be differentiated from cellular rejection. The extent of myocyte necrosis is usually out of proportion to the inflammatory infiltrate in ischemic injury, with the infiltrates consisting mostly of neutrophils and macrophages. In cellular rejection, the infiltrates are predominantly lymphocytic. A more difficult distinction to make is between the healing phase of ischemic injury and the resolving phase of moderate rejection in the early posttransplant period. This is usually resolved with clinical correlation and proper communication with the cardiologists.

Most early ischemic injury is clinically silent, but if the injury is extensive, myocyte necrosis can compromise the function of the graft postoperatively. Another possible implication in hearts that had damage during the peritransplant period is the subsequent development of interstitial fibrosis.42 

The ISHLT-WF1990 recommends that the presence or absence of the Quilty effect should be recorded. In the revised grading system, distinction between infiltrates exclusively confined to the endocardium (Quilty A) (Figure 3, A) and those that extend into the underlying myocardium (Quilty B or invasive Quilty) (Figure 3, B through D) is no longer indicated. There appears to be no clinical significance in subtyping Quilty lesions into A and B.36 Both these lesions are now referred to as the Quilty effect. Several hypotheses have been proposed to explain the pathogenesis of these infiltrates and include the use of cyclosporine-based immunosuppression,43 idiosyncratic responses to cyclosporin A,37 reduced endocardial levels of cyclosporine A,44 and concomitant infection with Epstein-Barr virus.45 None of these have been proven conclusively. One striking observation is that the Quilty effect was not found in the hearts of patients who were also treated with cyclosporin A for other solid-organ transplantation including the liver and kidney.46 The Quilty lesion seems to be a phenomenon that occurs only in the endocardium of cardiac allografts. Clear and consistent associations of Quilty lesions with grade of cellular rejection, viral infection, subsequent development of vasculopathy, or survival have not been established. As alluded in the section “Does Grade 2 Lesion Exist?,” Quilty effect lesions are sometimes misinterpreted by inexperienced pathologists and the diagnosis of rejection is rendered. Serial sections are very useful to differentiate these two lesions, as shown in Figure 4, A through D. Furthermore, the histologic detail of these two lesions is rather distinct. The Quilty lesions usually have extracellular matrix (collagen) between the lymphocytes as these cells are infiltrating the endocardium (Figure 4, E). These lesions frequently show capillaries in the middle of the infiltrate. On the other hand, the rejection lesions that were previously called grade 2 are indeed foci of rejection in which the lymphocytes are attacking the graft and not infiltrating connective tissue. Thus, one does not find collagen bundles surrounding the lymphocytes (Figure 4, F).

A previous biopsy site is a common finding in transplant surveillance biopsies and can be seen in up to 69% of biopsies.47 This high frequency occurs because, for a given patient, the anatomy of the inflow tract to the right ventricle is constant. During the biopsy procedure using the transjugular approach, the ridges of the atrial or caval anastomotic sites, the right ventricular trabeculations, and the moderator band all contribute to guide the tip of the bioptome toward the same site in the interventricular septum. Figure 6, A through G, illustrates different stages of lesions related to previous biopsy site. Gross examination at autopsy may show a patch of thickened endocardium measuring 1 to 2 cm in diameter in the mid third of the right ventricular septum in patients who survived several months to years after the transplant. On light microscopy, the findings of this repetitive sampling of a small area of the septum will include several stages of healing. Recent biopsy sites will show thrombus and granulation tissue (Figure 6, A). Later, there is fibrosis with entrapped myocytes that often exhibit disarray and a variable amount of mononuclear cell infiltrate (Figure 6, E). Old biopsy sites present as endocardial scars (Figure 6, F and G).

Posttransplant lymphoproliferative disease has been reported to occur in 1.2% to 9% of cardiac transplant patients, more commonly within the first year of transplantation.48–50 Recent studies are lacking, and this diagnosis is indeed rare in large-volume centers, perhaps as a result of modern immunosuppression regimens. Identified risk factors for the development of lymphoid neoplasms in these patients are infection with Epstein-Barr virus and type of immunosuppressive regimen received, particularly OKT3.50–53 Histology of lymphoid neoplasia can range from polymorphic lymphoid hyperplasia to monomorphic malignant lymphomas.54 Posttransplant lymphoproliferative disease can be diagnosed in the transplant biopsy and should be distinguished from that of acute rejection because early diagnosis and reduction of immunosuppression may lead to regression.55 Molecular studies can be performed using allograft biopsy material to confirm the diagnosis, including DNA analysis for immunoglobulin gene rearrangement and detection of Epstein-Barr virus genome by in situ hybridization or polymerase chain reaction.48,56–59 The majority of posttransplant lymphoproliferative diseases seen today are malignant lymphomas of B-cell origin. Their clinical presentation, in decreasing order of frequency, involves lymph nodes, lung, gastrointestinal tract, liver, central nervous system, spleen, and the heart itself.60 T-cell lymphomas also occur and usually present in extranodal sites.61–63 Development of multiple myeloma after cardiac transplantation is rare.64,65 

Chronic immunosuppressive therapy to control rejection predisposes transplant patients to a large number of opportunistic infections. Bacterial infection is the most common type of infection, accounting for 47% of the cases. Viral infections are second in frequency (41%), with fungal and protozoal pathogens being responsible for the remaining 12%.66,67 Identification of infectious pathogens in cardiac biopsy is rare. The two most commonly reported opportunistic infections seen in EMB specimens are Toxoplasma and cytomegalovirus (Figure 7, A and B, respectively).68–70 When examining a biopsy, unusual inflammatory infiltrates such as the presence of granulocytes, plasma cells, and/or macrophages in a focus of inflammation without overt myocyte necrosis or dropout should alert the pathologist to consider a possible infectious process. One should also look for viral inclusions in the nuclei of endothelial cells, smooth muscle cells, or miscellaneous perivascular cells. Cytomegalic inclusions within cardiac myocytes are extremely rare. Both infections can also be associated with a paucity of inflammatory infiltrates and can therefore be easily overlooked. Figure 7, C and D, show examples of fungal infections.

Development of interstitial fibrosis in the transplanted heart has been associated with cyclosporine therapy, total ischemic time, rejection episodes, and donor cause of death.42,71–75 Other investigators, however, did not find a significant association between increase in myocardial collagen and prolonged ischemic time or cyclosporine immunosuppression.76 The perception of the amount of fibrosis in endomyocardial biopsies may be influenced by the size of the bioptome used; larger pieces of biopsy fragments appear to have lesser quantitated area of fibrosis.77 Perimyocytic fibrosis is seen most often in areas adjacent to previous biopsy sites. A causal relationship between interstitial fibrosis and diastolic dysfunction in the cardiac allograft is still uncertain.78 Furthermore, fibrosis may in fact be a feature already present in a donor heart. Low availability of hearts has led cardiologists and surgeons to the practice of accepting hearts from older donors, which, in some instances, show interstitial and/or replacement fibrosis despite having “no history” of coronary artery disease.

Adipocytes are normal cellular components of the heart, mostly present in the epicardium. In addition, microscopic foci of adipose tissue are usually present in the subendocardium and, less frequently, within the myocardium. These foci can be seen in all chambers but are more commonly found in the right ventricular wall. In obese patients, older patients, and patients taking steroid hormones, fat infiltration is more common and can be grossly visible. Thus, the presence of adipose tissue per se is not pathologic. The goal of the right ventricular biopsy procedure is to obtain samples from the right side of the interventricular septum; however, on rare occasions, the bioptome may actually sample the right ventricular free wall. Therefore, when a focus of adipose tissue is found in an EMB, the pathologist should make an effort to determine if this is subendocardial or subepicardial adipose tissue. This can sometimes be easily determined by looking for the presence of mesothelial cell lining, indicating the epicardial surface (Figure 8, A). Because of the fibrinous and eventually fibrous pericarditis that usually develops after the transplant, it may be difficult to find mesothelial cells; in the latter case, the presence of nerves and ganglion cells or inflammation in the fat is suggestive of epicardial location (Figure 8, B). In time, the organized pericarditis usually forms a dense, fibrous, protective layer around the myocardium that prevents the development of tamponade if there is perforation. In one study, the presence of adipose tissue was reported to occur in 4.62% of transplant biopsies.79 There is also some tendency to see fat deposits in areas of previous biopsy site or foci of healing ischemic damage. Whether the use of steroids for the treatment of rejection increases the amount of adipose tissue in the subendocardium is not known.

Lymphocytes from Quilty lesions can be trapped in previous biopsy sites and then are crushed during subsequent biopsies (Figure 8, C). In other instances, lymphocyte clusters can be seen in postcapillary venules that become engorged with lymphocytes as these prepare to migrate into the interstitial space of the graft (Figure 8, D).

There have been reports of various forms of calcification in the heart after transplantation. In some patients, evidence of calcification has been shown histologically in biopsy tissue and radiographically in the native atria.80,81 In our experience, it is also uncommon to see dystrophic calcification of the ventricular myocardium in biopsies. Calcium deposition within mitochondria is known to occur during ischemia and catecholamine-induced myocardial injury. In the posttransplant patients, a relationship between calcification and cyclosporine therapy has been suggested.82 In some cases, several episodes of rejection requiring therapy, temporary uremia, and septicemia appear to be associated with the development of dystrophic calcification.80 On light microscopy, the dystrophic calcification of the mitochondria is easily recognized as dark blue granular material in the cytoplasm of myocytes, ranging from 1 to 2.5 μm in diameter (Figure 8, E and F). The granules may be seen in perinuclear location and in between the myofibers. When they are abundant, they follow the contour of the whole myocyte. Dystrophic calcification is usually found in the subendocardium, affecting single myocytes or small groups of myocytes.

When a small muscular artery is sampled by the bioptome, telescoping or intussusception occurs. Just before the jaws of the bioptome completely cut through the tissue, the small artery is stretched and then recoils into its own lumen as soon as it is severed. This can give the appearance of an occluded vessel or a small artery with vasculopathy. The birefringent internal elastic lamina within the lumen can be recognized easily on closer examination of small arteries (Figure 8, G).

Fragments of chordae tendineae are occasionally seen in the biopsy specimen and should be described in the report when present (Figure 8, H and I). Chordae to the tricuspid valve can arise from the septum and thus can be entrapped, torn, or biopsied during the procurement of tissue. Chordal rupture may or may not result in clinically significant tricuspid regurgitation.14,83–85 

Procedural artifacts are common and should be recognized in the interpretation of the endomyocardial biopsy.86  Contraction bands are a very common artifact seen in transplant and nontransplant heart biopsies. Several factors may influence the presence of contraction bands in the biopsy. It may be the result of trauma to the myocardium induced when the bioptome cuts the tissue. It may also be induced by poor osmolarity of the medium in which the biopsy is placed before and during fixation, as well as the cool temperature of the medium. We rarely observe contraction bands in frozen sections. Because of the high likelihood of finding contraction bands, they should never be the only criterion used to make a diagnosis of myocyte necrosis or ischemic damage in heart transplant biopsies. Pinching or forceps artifact represents mechanical distortion of the tissue induced by the bioptome itself during extraction. It can also be induced during processing of the tissue in the pathology laboratory. An effort should be made to handle biopsy tissue with care because this artifactual deformation may render the specimen uninterpretable. Foreign bodies introduced at the time of the transplant, such as gelatin foam, occasionally can be seen. At other times, actual sampling of fragments of indwelling catheters or the soft plastic cover of pacemaker leads may occur. Pseudohemorrhage occurs when red blood cells are embedded into the tissue by the pressure of the bioptome on the myocardium being sampled. This produces artifactual pools that mimic hemorrhage. They are usually not accompanied by inflammatory cells or pathologic changes in the myocytes, thus making the distinction between artifact and rejection fairly easy.

Transplants are capable of eliciting strong cellular and humoral immune responses. Antibody-mediated rejection (AMR) is an immunopathologic process in which injury to the graft is, in part, the result of activation of the complement system. This was first recognized in kidney transplantation as a distinct clinicopathologic entity characterized by acute allograft rejection associated with the production of antidonor reactive antibodies and poor prognosis.87 It is poorly responsive to conventional immunosuppression, which targets the cellular arm of the immune response. Old terminology such as vascular rejection, microvascular rejection, and humoral rejection should be avoided as it has only led to confusion in the literature. The preferred terminology in the ISHLT-WF2004 is AMR.

Risk factors for developing AMR include pregnancy, previous transplantation, blood transfusions, sensitization by OKT3 induction therapy, use of ventricular assist devices, presence of positive B-cell flow cytometry crossmatch, and elevated panel-reactive antibodies.13,88,89 The long-term outcome of AMR is not yet fully established in heart transplantation but it has been associated with the development of cardiac allograft vasculopathy (CAV) and with decreased survival.90,91 

A detailed pathologic classification of “humoral rejection” in biopsies was not well defined in the ISHLT-WF1990. Consequently, the true incidence of AMR is unknown and recognition of AMR as a real entity was not widely accepted for several reasons. There was no uniform set of diagnostic criteria provided to guide different transplant programs in the detection of this entity. The antibodies used in evaluation of immunofluorescence changed over time. Positive immunofluorescence with the markers suggested then (IgG, IgM, C3, C1q, and fibrinogen) did not always correlate with hemodynamic compromise or incidence of CAV, which resulted in decreased usefulness of this test.92 Lastly, it was believed that most AMR occurs early and the ISHLT-WF1990 recommends AMR monitoring by immunofluorescence on all biopsies up to 6 weeks posttransplant only. This is clearly incorrect, as it is now known that AMR can and most commonly does occur months and even years after transplantation.

The histologic features that allow for the identification of this type of rejection on endomyocardial biopsies as defined in the ISHLT-WF2004 and its companion article on AMR include: “capillary endothelial changes (swelling or denudation with congestion), macrophages in capillaries [Figure 9, A and B], neutrophils in capillaries, interstitial edema and/or hemorrhage and fibrin in vessels.” 93 If these features are observed in the biopsy and there is unexplained cardiac dysfunction, the revised working formulation proposed that immunofluorescence or immunohistochemistry, in the absence of frozen tissue, be performed. Immunopathologic evidence of AMR include13 

• “—Immunoglobulin (IgG, IgM and/or IgA) plus complement deposition (C3d, C4d and/or C1q) in capillaries by immunofluorescence on frozen sections; and/or

• —CD68 staining of macrophages within capillaries (CD31- or CD34-positive) by immunohistochemistry; and

• —C4d staining of capillaries by paraffin immunohistochemistry.”

Examples of the capillary pattern of complement deposition are shown in Figure 9, C through F.

It is also recommended that these patients undergo assessment for circulating antibodies to HLA class I or II as well as non-HLA donor antigens. An EMB with no histologic or immunopathologic evidence of AMR is graded 0 (AMR 0). If the immunofluorescence or immunohistochemical staining supports the histologic features of AMR, the biopsy is considered positive (AMR 1).

Although most AMRs are associated with absent, or at most mild, acute cellular rejection, mixed rejections have also been reported that carry a significant risk of mortality.94,95 Mixed rejections usually occur early in the course of transplantation and are also associated with allograft dysfunction.

The ISHLT-WF2004 recommends that if histologic features suggestive of AMR are not seen, no further testing (immunofluorescence or immunohistochemical) needs to be pursued. However, a recent report96 describes that the sensitivity of histologic criteria (ie, light microscopic features such as endothelial cell swelling, intravascular macrophages, edema, and hemorrhage) is too low to serve as screening parameters for AMR. The authors thus recommend the addition of immunostaining to screen for the presence of AMR.

Immunofluorescence methods for detection of AMR in tissues have evolved in the last decade. Some complement components, specifically C3d and C4d, are found to be more readily detected than antibodies and serve as very sensitive markers of rejection in endomyocardial biopsies for several reasons.97 Antibodies bind to antigens with different avidity and either dissociate at varying rates or are eliminated by shedding or internalization. In contrast, the process of complement activation yields split products of C4 and C3 that bind to the tissue where complement was activated. This increases the sensitivity of complement detection by prolonging their half-lives. Among the components of the complement system, C3 is present in the highest concentration, followed by C4; therefore, their split products are also deposited in tissues in the largest quantities (Figure 10).98 Furthermore, the amplification steps in the complement cascade results in the generation of more C3 split products.99 

Although complement is activated through antibody in the classic pathway, one must remember that complement can also be activated during procedures such as extracorporeal circulation during surgery,100,101 by ischemia/reperfusion injury,102 and by induction therapy before transplant with antithymocyte globulin.103 Thus, the mere presence of C4d and/or C3d in capillaries should not be equated with AMR.

In our experience, the use of C4d immunostaining alone is not a reliable tool. Instead, evaluation of endomyocardial biopsies for AMR should include staining for both C4d and C3d (Figure 9). A recent prospective study of heart transplant patients evaluated the usefulness of IgG, IgM, IgA, C1q, C4d, and C3d as markers for the diagnosis of AMR.104 In this study, the authors performed routine staining of all biopsies for these five markers. These authors' institution reported 3% incidence of AMR in 165 nonsensitized patients. Immunoglobulin G, IgM, IgA, and C1q did not prove to be useful in discriminating patients with AMR. Conversely, the usefulness of C4d and C3d was confirmed. Immunostaining for C4d alone can be misleading because about 10% of the patients showed either C4d or C3d deposits alone in capillaries without clinical evidence of dysfunction of the allograft. Within the study period of 3 years, some patients demonstrated persistent activation of complement with C4d deposition over time without the development of allograft dysfunction. Another important observation made was that AMR occurred many months to years after transplantation in most patients. This study showed that the diagnosis of AMR must be a correlative diagnosis in which pathologic and clinical criteria play a role.

Activation of the complement cascade detected by immunostains for C4d and/or C3d is not always accompanied by dysfunction of the graft. Some authors have referred to this apparent lack of graft injury despite evidence of complement activation as “accommodation” in animal models105 and in ABO-incompatible renal transplants.106 One possible explanation is that complement activation is interrupted by a protective mechanism in the host. This suggests that unless the complement cascade proceeds to the formation of the membrane attack complex, there is no expected injury to the allograft. This complex is needed to form a “pore” that leads to loss of integrity of the cell membrane. In humans, it is well known that there are regulators of complement activation that can prevent the completion of the complement cascade at different stages of activation.

Regulators of complement activation exert their effects at different points in the complement activation cascade, whether the activation occurs through the classic, alternate, or mannose binding lectin pathways. All these pathways converge at the point of generation of the enzymatic complexes known as the C3 convertases, which, in turn, proceed to activate the remaining complement components required for the formation of the membrane attack complex. There are two main types of proteins that can regulate the activation of complement. These can be divided into the membrane-bound and soluble types. In humans, the membrane-bound regulators are CD35 or complement receptor 1, CD46 or membrane cofactor protein, CD55 or decay-accelerating factor, CD59 or protectin, and C8-binding protein or homologous restriction factor.107,108 The soluble factors include the C1 inhibitor, C4 binding protein (C4bp), factor I, factor H, clusterin, and S protein (vitronectin). Their points of action are shown in Figure 10.

There is little information about the expression of these regulators of complement activation molecules in human heart transplantation. A recent abstracted study shows that decay-accelerating factor or CD55 is expressed locally in the myocardium in heart transplant patients. In this study, a group of patients with complement deposition in endomyocardial biopsies was examined. The biopsies were stained by immunofluorescence for C4d, C3d, and decay-accelerating factor (Figure 11, A through F). There were 2 subgroups identified on the basis of present or absent allograft dysfunction. All patients had biopsy-proven C4d (Figure 11, A and C) and C3d (Figure 11, B and D) deposits. Patients with good response to therapy and resolution of the AMR episode showed intense tissue expression of CD55 in the endothelium of the allograft (Figure 11, E and F). Patients with poor outcome had low or absent tissue expression of CD55. Thus, the local expression of decay-accelerating factor correlates with absence of allograft dysfunction in spite of C4d and C3d deposition in capillaries.109 In the same study, there was no evidence of detectable CD35, CD46, or CD59 in the biopsy tissue of this cohort of patients. At this juncture, there are no studies published that address the role of the soluble regulators of complement activation in human heart transplantation.

Common artifactual staining seen in immunofluorescence microscopy of transplant biopsy includes autofluorescent lipofuscin deposits (Figure 12, A), nonspecific binding to collagen in the interstitium (Figure 12, B), and to the internal elastic lamina of arteries (Figure 12, C). Necrotic myocytes likewise bind complement (Figure 12, D).

Currently, the most challenging problem in attaining a long-term successful outcome in cardiac transplantation is the development of CAV (Figure 13, A through E), also known as graft coronary artery disease, graft coronary vascular disease, transplant coronary artery disease, accelerated graft arteriosclerosis, and chronic rejection. This problem is not unique to the heart; it occurs in other solid organ grafts in a somewhat similar manner.110,111 Cardiac allograft vasculopathy develops in a majority of transplanted hearts at a variable rate, sometimes as early as 3 months after transplantation.112 According to the most recent ISHLT registries, only 47% of adults are free of CAV as detected by angiography at 9.5 years; in children, the incidence is much lower compared with adults, with 75% of patients free of CAV at 7 years posttransplant.113,114 Most patients cannot experience typical angina associated with myocardial infarction or ischemia because of denervation of the donor heart. Therefore, CAV commonly presents clinically as congestive heart failure, ventricular arrhythmias, and sudden death. Risk factors for developing early CAV (occurring within 3 years posttransplant) include donor hypertension, infection within 2 weeks posttransplant requiring intravenous antibiotics, and rejection during the first year.1,2,115 Risk factors associated with late CAV (occurring within 7 years posttransplant) include donor history of diabetes and intracranial hemorrhage as donor cause of death. Independent continuous risk factors for both early and late CAV are donor age, recipient age (inverse relationship), center volume, and recipient pretransplant body mass index.1,116 Female donors are associated with a lower risk.1 

The events leading to this type of vasculopathy are complex and interdependent. The mechanisms can be divided into immunologic and nonimmunologic. Endothelial cells express major histocompatibility complex class I and class II antigens, and thus appear to be primary targets of cell-mediated and humoral immune response.117–119 Activated T lymphocytes secrete cytokines (interleukins, interferons, and tumor necrosis factors), which promote proliferation of alloreactive T cells, activate monocytes and macrophages, and stimulate expression of adhesion molecules by endothelial cells.120 Macrophages are then recruited to the intima where they elaborate cytokines and growth factors, leading to smooth muscle cell proliferation and synthesis of extracellular matrix.121 The role of humoral immune response in CAV relates to the antibody production against HLA and endothelial cell antigens.122–125 The relationship between acute cellular rejection, histocompatibility mismatch, and development of CAV remains controversial.126–132 Endothelial cell dysfunction resulting from sustained inflammatory injury also predisposes to thrombosis, vasoconstriction, and vascular smooth muscle proliferation.133–135 

Some of the nonimmune factors that have been associated with the development and progression of CAV include myocardial ischemia,136,137 donor-transmitted coronary atherosclerosis,138,139 cytomegalovirus status,140–143 lipid profile,144,145 arteritis,146 deficient fibrinolysis,147,148 hormonal milieu,149 and immunosuppressive therapy.150–152 Excellent reviews of the pathobiology of vasculopathy have been written.110,153–155 

Allograft vasculopathy involves both epicardial and intramural coronary arteries. The whole length of the coronary vessels is usually affected. Formation of collateral vessels is lacking. In some patients, the pathology predominantly involves only the small intramyocardial branches.156 In these cases, early diagnosis is limited by inaccessibility of distal lesions to evaluation by coronary angiography or intravascular ultrasonography.

The endomyocardial biopsy has limited sensitivity in the recognition of vasculopathy because it samples only the smallest of intramyocardial arteries and arterioles, which often do not show histologic features of CAV.157 Proliferative intimal lesions are usually not prominent in the coronary microvasculature (vessels less than 100 μm in diameter) within the first few years posttransplantation when most of the surveillance biopsies are being performed on a regular basis.158 Moreover, investigators have suggested discordant progression of CAV because of differences in the structural and functional abnormalities between small intramyocardial and large epicardial arteries.158–160 Reported histologic changes seen in the small arteries in endomyocardial biopsies include concentric intimal thickening with or without foamy macrophages, subendothelial accumulation of lymphocytes (called by some, endothelialitis), and perivascular fibrosis.156,157 Evidence of myocardial ischemia, such as colliquative myocytolysis, frank coagulation necrosis, and healing ischemic lesions as well as interstitial, perivascular, and replacement fibrosis, can be seen in endomyocardial biopsies.161 Identification of myocardial injury should raise the suspicion of CAV as the cause of graft dysfunction. Absence of these findings, however, does not necessarily rule out the presence of CAV. One study of the predictive value of endomyocardial biopsies in a series of patients with CAV confirmed on autopsy reveals a sensitivity of only 21% for the detection of myocardial ischemic changes.162 

The classic feature of CAV is that of diffuse concentric narrowing with luminal stenosis (Figure 13, A and B). If atherosclerotic plaques were present in the donor heart prior to the transplantation, the morphology of the lesion is one of eccentric atheromatous plaques with superimposed intimal proliferation of transplant-related vasculopathy. Sometimes, long-term lesions of epicardial coronary arteries may eventually look like conventional atherosclerosis and be indistinguishable from CAV. Careful examination of the cut surfaces of ventricles often reveals intramural arteries (with a range in diameter from 0.2 to 0.5 mm) that are thickened with abundant perivascular fibrosis. In addition, focal areas of myocardial scarring may be evident.

The histology of allograft vasculopathy is slightly different in epicardial arteries compared with medium-sized or small arteries.146,163,164 Microscopically, allograft vasculopathy in large epicardial vessels shows concentric intimal proliferation composed of smooth muscle cells and less-differentiated spindled cells (myofibroblasts or “myointimal” cells) (Figure 13, B and C). There is accompanying abundant deposition of proteoglycans with different immunohistochemical staining pattern and distribution, compared with conventional atherosclerosis, and more similar to angioplasty-related restenotic lesions (Figure 13, C).165 Calcification and large pools of extracellular lipid are rare unless associated with atheromatous plaques that may develop in long-term survivors. Early lesions tend to be more cellular than those in the late stages, where the smooth muscle cells decrease in number and the intima becomes fibrotic. Mononuclear inflammatory cells are usually present in variable number, consisting mostly of T lymphocytes admixed with macrophages and foam cells. The internal elastic lamina is intact or only focally disrupted. The media is of normal thickness and shows little to no lipoprotein deposition. Medial fibrosis in the outer half is associated with lymphocyte-mediated injury of the vasa vasorum. An adventitial cuff of fibrous tissue with or without mononuclear inflammatory infiltrates is commonly observed (Figure 13, B and C). Atheromatous plaques, if present, are found in the proximal to middle segments of large epicardial arteries, produce an eccentric type of luminal stenosis, and histologically are indistinguishable from those of conventional atherosclerosis.

In the small epicardial and intramyocardial branches, allograft vasculopathy is also concentric but foam cells are not prominent. Endothelialitis is frequently observed in autopsy material (Figure 13, D and E). Occasionally, vasculitis with transmural inflammation by lymphocytes and plasma cells and disruption of internal elastic membrane is present in distal coronary arteries, usually associated with acute cellular rejection in the myocardium.164,166 A second pattern of vasculitis that is characterized by severe inflammation in the adventitia that extends to the medial layer with destruction of external elastic membrane and is correlated with myocardial rejection has also been recognized by some investigators.146 Fibrinoid necrosis of the media and thrombosis in small epicardial and intramural arteries can sometimes be seen.146,163,164 Recanalized vessels may represent healed vasculitis with thrombosis. It is not clear whether this necrotizing vasculitis is due to cellular or humoral rejection or a combination of both.

The myocardium oftentimes show bilateral, patchy microscopic acute and healing ischemic injury161 because it is believed that intramyocardial vessels are totally occluded first before the large epicardial arteries become critically stenosed. Chronic ischemic changes including myocytolysis and interstitial fibrosis are also frequent. Large myocardial infarcts are uncommon in the absence of thrombosis in the major epicardial vessels.164 The pathology of CAV in children is practically identical.166 

Complications of chronic immunosuppression include drug toxicities, development of malignancy, and increased risk of infection. In time, most patients also develop hypertension, hyperlipidemia, diabetes mellitus, and renal insufficiency. Other notable adverse effects of therapy include bone marrow suppression and gastrointestinal symptoms.

Despite the use of newer and less toxic immunosuppressive drugs that decrease the incidence of acute cellular rejection, immunosuppression is still a significant cause of morbidity and mortality in the first year posttransplant. The majority of patients will have at least one episode of rejection. The rates for freedom from rejection at 1 year ranged from 10% to 23% covering the era before and after the introduction of cyclosporine in the Stanford experience.167 Other centers have shown a significant decrease in the incidence of moderate and severe cellular rejection attributed to improved immunosuppressive therapy in the 1990s.168 Other major causes of death within the first year are early graft failure, multiorgan failure, and infection other than cytomegalovirus.1,169 Early graft failure in the absence of cellular rejection can be the result of right ventricular failure from pulmonary hypertension, acute humoral rejection, and ischemia related to donor atherosclerosis, prolonged ischemic time, or poor donor preservation.116 

Infections after cardiac transplantation often occur in the first few months after transplantation, with the highest risk of death at approximately 2 months after transplantation. Risk factors identified for fatal infections are very old and very young recipients, ventilator support at time of transplant, older (more than 50 years) donor heart, and prolonged donor ischemic time.67 The most common site of infection reported was the lung.66 Late infection is usually associated with recurrent high-grade rejection that requires augmentation of the immunosuppressive regimen.

Between 1 and 4 years, rejection, malignancy, and CAV account for most number of deaths. After 5 years, malignancy and CAV remain as the leading causes of death. A study involving 7290 patients who received transplants in multiple institutions between 1990 and 1999 reported malignancy (29%) as the leading cause of death, followed by CAV (23%) after the fourth year following transplantation.169 The ISHLT registry with an 8-year cumulative data shows a 26% incidence of malignancy.114 Transplant recipients do not appear to have increased risk of developing common cancers, including carcinomas of the lung, breast, prostate, and colon. However, an increased incidence of lymphoproliferative disorders, squamous cell carcinomas of the skin, sarcoma including Kaposi sarcoma, renal cell carcinoma, carcinomas of the vulva and perineum, and hepatobiliary tumors are observed.170 The etiopathogenesis of posttransplant malignancies can be multifactorial. Association between the development of lymphoproliferative diseases and cytolytic induction therapy has been reported in renal and cardiac transplant recipients.53,171 In another study, no increase in the incidence of malignant neoplasms was found between patients who received Thymoglobulin induction therapy and those who did not. However, it has been reported that patients who are treated with rabbit anti-thymocyte immunoglobulin develop malignant neoplasms earlier than those without induction therapy and have worse prognosis of their malignancies.172 Long-term use of azathioprine has also been implicated in the development of myelodysplastic syndrome and acute myelogenous leukemia.173 

Oncogenic viral infection may also play an important role in the development of malignancy in the immunocompromised hosts, including Epstein-Barr viruses, human papillomavirus, and hepatitis B and C virus. Recurrence of prior malignancy in a transplant patient may be because of defective immune surveillance. Transmission of cancer from donor to recipient is also possible. Transmitted malignancies in recipients of cardiac allografts include high-grade primary brain tumors, choriocarcinoma, lung adenocarcinoma, and melanoma.174 

Autopsy studies of sudden unexpected cardiac deaths in transplant recipients reveal arrhythmias, CAV with diffuse involvement of distal coronary arteries, and cellular or humoral rejection as the most frequent causes of death.175,176 

Endomyocardial biopsy is an invasive procedure that carries a low but finite risk of complications. There is associated patient discomfort with frequent biopsies. It is also a costly procedure and is resource-intensive. In addition, criticisms for EMB are the lack of concordance among pathologists in grading acute cardiac allograft rejection and failure to recognize AMR because of poorly defined histologic diagnostic criteria. However, no viable alternative exists to date.

Noninvasive monitoring of allograft rejection is an area of active research.177 The need for proteomic and genomic markers to predict cardiac transplant rejection, correlation with outcomes, and risk of graft failure has been well recognized.5 A recent study correlated gene expression profiling in peripheral blood mononuclear cells to the presence of acute cellular rejection in endomyocardial biopsies.178 Their results indicating a high negative predictive value for the test show a promising diagnostic role for molecular testing. Reproducibility of these results in large-scale clinical settings has to be further demonstrated.179 

In summary, the ultimate goal of any heart transplant team is a successful long-term outcome for the patients. This can be achieved with the pathologist working closely with the cardiac transplant team before and after the transplant. The recent revision of the ISHLT grading scheme should improve the interobserver reproducibility for cellular rejection and allow objective recognition of AMR. This, in turn, will result in more accurate diagnosis and better assessment of the effectiveness of therapy. Targeting proper therapy for host cellular or humoral response to the allograft may reduce the development and progression of CAV and other causes of graft dysfunction.

This work was supported by grants 5P01HL070295 and 5P01HL056091 from the National Institutes of Health.