Endomyocardial biopsy is commonly performed by a percutaneous technique using the right internal jugular or femoral vein or femoral artery with fluoroscopic guidance, 2-dimensional echocardiography, or both. Since the introduction of more flexible bioptomes, however, such as the Stanford-Caves Schultz and King’s bioptomes, the preferred site of access is now the right internal jugular vein, for access to the right ventricle. Biopsies should be taken from the interventricular septum, given that the right ventricular free wall is thin, and scraping too hard may cause perforation.

Due to its invasive nature, the test may provoke anxiety and discomfort for the patient and remains particularly challenging in the pediatric population, often requiring the use of general anesthesia. A major drawback to the endomyocardial biopsy is that it samples only a limited area of the endocardium. Inflammatory changes may be sporadic through the myocardium, or may predominantly affect the subendomyocardium; in these cases, the biopsy may miss the diagnosis. Thus, diagnosis of rejection also relies on the clinical presentation and echocardiographic findings, which may or may not be supported by histology [3, 4]. Furthermore, biopsy utilizes significant resources including physician time and is associated with substantial costs.

Although the procedure is considered safe with a complication rate well below 6% [5], there is a finite risk of injury. Such reported complications include transient right bundle branch block, tricuspid regurgitation, access site hematoma, transient arrhythmias and occult pulmonary embolism [5]. More rarely (<1%), right ventricular perforation has been reported [5]. Generally speaking, only those who undergo repeated biopsy are at risk of long-term complications, which may include severe tricuspid regurgitation and coronary artery to right ventricular fistula.

As the transplanted heart is denervated, symptoms resulting from graft rejection may remain silent and may not be recognized until late during the course of a rejection episode. Consequently, surveillance biopsies are traditionally performed at standard intervals from the time of transplantation. The recommended frequency for performing surveillance right ventricular biopsy varies by center. There has been a recent trend towards a reduction in the number of procedures being performed as improvements in immunosuppressive therapy and post-transplant management continue to show a decline in the number of rejection episodes. The development of alternative, non-invasive surveillance methods has further decreased the use of biopsy at some centers. A typical biopsy schedule consists of performing the procedure weekly during the first month, every 2 weeks for another month and monthly until 6 months and then every two or 3 months until the end of the first post-operative year, with yearly biopsies thereafter in higher-risk patients. This schedule is intended to reflect the general risk of allograft rejection which is highest in the first 6 months post-transplant. After the first year, any additional protocol biopsies are likely not to be of clinical significance given the very low rates of rejection observed in this period [6]. However, biopsies are performed anytime in cases of clinically suspected rejection. Repeat biopsies are performed 7–14 days after treatment of rejection in order to confirm resolution.

Histologically speaking, acute rejection is observed as an inflammatory response of the host to the transplanted organ. Though T-cell mediated mechanisms leading to acute cellular rejection (ACR) were initially described, there is now consensus that host antibody responses play an equally important role, and may result in antibody-mediated rejection (AMR). The diagnosis of AMR remains technically more challenging and a consensus on its definition and management has only recently evolved [7].

As rejection is a histological diagnosis, there are many cases where the patient may remain asymptomatic, especially with milder forms of rejection. In cases where there are clinical features, symptoms of rejection may include palpitations, tachycardia, arrhythmias, edema, dizziness or blackout spells, dyspnea, and a fever of 100 °F or greater.

Although now uncommon, the development of hyperacute rejection was the most feared complication prior to the advent of effective immunosuppressive therapy. Hyperacute rejection is mediated by preformed antibodies to the allograft in the recipient. It typically presents following surgical engraftment and restoration of native circulation as an almost immediate, aggressive and inevitably lethal immune attack on the organ. Hyperacute rejection is mediated by preformed antibodies to predominantly HLA antigens, although the phenomenon has also been observed in cases of ABO incompatibility [8]. It is characterized by thrombotic occlusions and hemorrhage of the graft vasculature that begins minutes to hours after the graft is placed. Antigen recognition activates the complement system, along with an influx of neutrophils. Endothelial cells and platelets are induced to shed lipid particles from their membrane that promote coagulation; the resulting inflammation prevents vascularization of the graft, which suffers irreversible damage from ischemia. While this is the most drastic consequence of preformed antibodies to the graft, the presence of donor-specific antibodies is also associated with adverse outcomes even after successful engraftment [9].

The development of the prospective cytotoxic crossmatch, and subsequently the virtual crossmatch (mentioned in Chap. 6) has been a major achievement in avoiding hyperacute rejection in solid organ transplantation [10]. Use of these strategies also helps identify patients who are at high risk of rejection in whom immunosuppression may need to be augmented after transplant. Such advances in perioperative management and improvements in immunosuppression in recent years have led to a general decline in the rates of allograft rejection, though it still remains a significant problem post-transplant.

Acute cellular rejection (ACR), the most common form of rejection in heart transplant, is characterized by a predominantly T-cell mediated response with infiltration of macrophages and lymphocytes, which in turn can lead to myocyte necrosis. Thus, histologically ACR is defined by an inflammatory infiltrate which is typically lymphocyte predominant with associated evidence for myocyte injury (see Table 12.1, Fig. 12.2). Most episodes of ACR occur within the first 6 months post-transplant.

Diagnosis of ACR is made by endomyocardial biopsy; the first standardized grading scale was proposed by Billingham [11] in 1990, which was later revised in 2004 to accommodate for the reporting of AMR [12]. The most recent ACR grading scale, which classifies rejection into mild (1R), moderate (2R) or severe (3R) grades has allowed standardization of reporting, although variability of interpretation and discordance between pathologists remains, particularly for higher grades of rejection [13]. The main benefit of the new grading scale allows improved guidance for appropriate therapy, in conjunction with clinical assessment. Generally speaking, mild grades of rejection (ISHLT Grade 1R) do not require augmentation of immunosuppressive therapy as the vast majority of these episodes resolve spontaneously, without increased risk of poor subsequent outcomes. However, higher grades (ISHLT ≥2R) invariably require aggressive supplemental immunosuppression (see Sect. 12.4.3).

While the role of antibodies in mediating acute myocardial injury has been appreciated since the early days of cardiac transplantation when sub-optimal immunosuppressive regimens and unidentified preformed circulating antibodies led to early post-operative graft failure from hyperacute rejection, only in recent years has there been official acknowledgement of the role of humoral (antibody) responses in causing allograft rejection in the later phases post-transplantation [16].

It is now known that AMR develops when recipient antibody is directed against donor-HLA antigens on the donor heart endothelium. The recipient antibody initiates fixation and activation of the complement cascade, resulting in donor tissue injury. This complement activation results in activation of the innate and adaptive immune responses. Complement and immunoglobulin are deposited within the allograft microvasculature, resulting in an inflammatory process characterized by endothelial cell activation, macrophage infiltration, cytokine upregulation, increased vascular permeability, and microvascular thrombosis [17]. This process ultimately manifests clinically as allograft dysfunction.

In 2005, the ISHLT revised the 1990 working formulation for the standardization of heart transplant rejection to officially recognize AMR as a distinct rejection entity alongside ACR. The new scale established immunohistologic criteria for the reporting of AMR [12]. It was defined by histopathological changes consisting of capillary endothelial changes, macrophage (in particular CD68-expressing) and neutrophil infiltration, interstitial edema, and linear accumulations of immunoglobulins and complement, especially complement component C4d (see Table 12.2, Fig. 12.3). Additional clinical and serological findings of donor-specific antibodies supported the diagnosis of AMR [18].

However, in subsequent years, the phenomenon of asymptomatic AMR associated with worse outcomes was raised [19–21], and the sensitivity and specificity of the immunohistologic features and C4d staining was questioned [22–27]. Furthermore, surveys revealed a variety of approaches to the biopsy specimen investigation and considerable discordance between pathologists in the diagnosis of AMR, with opinion growing that AMR should be classified by severity analogous to ACR [23, 28–30].

Thus, in 2013, following expert discussions and consensus of expert opinion [31], further revisions were made by the ISHLT to the diagnostic criteria for AMR, in an attempt to further standardize diagnosis, and acknowledge that AMR evolves along a worsening spectrum of pathologic changes similar to ACR [7]. The new system specifies that AMR is divided into 3 degrees of severity (see Table 12.2) and is diagnosed from combined histologic and immunopathologic review of the endomyocardial biopsy.

The histopathologic features of AMR include intravascular macrophage accumulation within distended capillaries/venules, and enlarged nuclei and expanded cytoplasmic projections within endothelial cells that may narrow or even occlude the vessel lumen. For more severe cases, there may be signs of hemorrhage, interstitial edema, myocyte degeneration and necrosis, mixed inflammatory infiltrates, and endothelial cell pyknosis/karyorrhexis. The immunopathologic component of AMR comprises of the application of a panel for various antibodies (including C4d, CD68 and anti-HLA-DR) using immunohistochemistry from paraffin sections or immunofluorescence from frozen graft sections. Based on the combination of these findings, an overall pAMR grade is assigned to the biopsy (Table 12.2).

Prior to the acknowledgement and standardization of diagnosis of AMR as a distinct entity, hemodynamic compromise in the absence of evidence of acute cellular rejection was termed “biopsy-negative rejection”. While the prevalence of this so-called “biopsy-negative rejection” has substantially decreased with clear criteria for AMR diagnosis, there continue to be incidences of patients who present with LVEF <45% but have no biopsy findings of ACR or AMR. Nevertheless, these are exceedingly rare [40]. Due to the inherent flaws with the endomyocardial biopsy, the existence of BNR is questioned in some circles. Cases of BNR tend to respond favorably to appropriate rejection therapy.

While endomyocardial biopsy-derived histology remains the gold standard for rejection diagnosis, the potential complications and disadvantages—in particular patient discomfort, sampling error and poor inter-pathologist concordance—are notable. Furthermore, the pathological finding of rejection is a relatively late phenomenon, with diagnosis only made once myocardial damage has already taken place. An ideal test would be non-invasive, utilize less economic resources (biopsy requires radiologists, anesthesiologists, cardiologists, pathologists, associated technical staff) and allow early detection for the onset of rejection before any significant myocardial necrosis has occurred. Many non-invasive modalities have been investigated for this purpose, with the aim of minimizing biopsies if possible.