Pathology

PTLD results from the uncontrolled proliferation of lymphocytes in the immunosuppressed transplant recipient. It encompasses a remarkable diversity of pathologic conditions, challenging the development of a standard classification of disease. The most commonly used pathologic classification scheme is the World Health Organization (WHO) categorization, outlined with diagnostic components in Table 1 . Pathologic evaluation requires assessment of tissue architecture and cytologic features; excisional biopsies are preferred and tissue samples should be submitted fresh rather than in formalin. EBV-associated disease is best determined by EBV-encoded RNA in situ hybridization (ISH) testing of diagnostic tissues. Cytogenetic studies assist with determination of clonality (immunoglobulin heavy [IgH] gene rearrangement; T-cell receptor gene rearrangement) and identify disease-associated chromosomal abnormalities.

There are 4 categories of PTLD in the WHO categorization. Early lesions refer to lymphoid proliferations with preservation of normal tissue architecture and occur more frequently in previously EBV-naive SOT recipients. Polymorphic PTLD is the most commonly diagnosed PTLD among pediatric transplant recipients, often following primary EBV infection. Morphology can be variable and include areas that seem more monomorphic, suggesting a continuum between polymorphic and monomorphic disease. Monomorphic disease is classified according to the B-cell or T/natural killer (NK)-cell neoplasms described in the immunocompetent host. Markers of oncogenes and tumor suppressor genes (eg, C-myc, N-ras, and p53) may be used to facilitate diagnosis in more complex cases. Recurrent chromosomal abnormalities have been reported in some series of B-cell PTLD and may portend a worse prognosis. Reed-Sternberg-like cells may be seen in early, polymorphic, and some monomorphic PTLD. The diagnosis of true classic Hodgkin lymphoma (cHL)-type PTLD relies on morphologic and immunophenotypic features of the diagnostic sample. The expression pattern of EBV proteins (EBV-latency pattern) may aid in the diagnosis of cHL.

In contrast to the typical T/NK-cell disease spectrum among immunocompetent children, the most common T/NK-cell PTLD described is peripheral T-cell lymphoma (PTCL), followed by hepatosplenic T-cell lymphoma. T/NK-cell disease tends to occur later and is rare (approximately 15%), although it may be increasing. The disease is more common in certain geographic regions possibly because of the greater prevalence of human T-lymphotropic virus 1 (HTLV-1) infection in these areas. It more commonly presents late, is more likely to present at extranodal sites, and is more often EBV negative (two-thirds of cases). T/NK-cell pathology seems to be rare in the pediatric SOT population. The available literature suggests that T/NK-cell PTLD is clinically aggressive and associated with a poor prognosis, although the natural history of the disease in children remains to be defined.

Immunobiology of EBV

Most cases of PTLD following pediatric SOT are associated with EBV, a latent human lymphotropic virus discovered in 1964 by Epstein, Barr, and Achong. More than 90% of the world’s population is infected with EBV and infected individuals remain lifelong carriers of the virus. The life cycle of EBV is outlined in Fig. 1 . Primary infection occurs through the oropharynx, where the virus infects resting B cells that become proliferating lymphoblasts with expression of EBV nuclear antigen 2 (EBNA2), which activates the promoters necessary to produce all latent proteins (growth program or latency III) ( Table 2 ). The virus gains access to the memory B-cell compartment following the steps required for a normal B cell to move along the pathway from naive to memory cell. After migration of the infected cells into the follicle, EBNA2 is turned off, together with the latent genes that drive proliferation. Infected cells acquire a germinal center phenotype and start the default program (latency II), which is accompanied by the expression of latent membrane protein 1 (LMP1) and LMP2, which provide the necessary signals, T-cell help and B-cell receptor, respectively, that physiologically rescue latently infected germinal center cells into memory cells. These infected B cells that have the surface phenotype of long-lived memory B cells and do not express any of the known latent proteins (latency O) or only EBNA1 (latency I), mature and circulate in the peripheral blood. Release of the virus into the saliva occurs when circulating memory B cells transit the nasopharyngeal lymphoid tissue and differentiate into plasma cells.

EBV life cycle.

Control of virus spread and of unrestrained infected B-cell proliferation is guaranteed by the development of a specific immune response that occurs in healthy immunocompetent individuals and consists of T cells specifically recognizing immunodominant latent EBV proteins presented in the context of major histocompatibility complex (MHC) molecules. NK, CD8+, and CD4+ T cells generally control and contain the proliferation of EBV-infected B cells during primary infection. CD8+ cells, particularly those directed against lytic EBV antigens, expand dramatically during acute infection, and then contract at the resolution of the primary EBV infection. During persistent EBV infection, lytic and latent EBV antigen-specific T cells are maintained at a frequency of 1% to 5% of peripheral blood T cells to immune survey and eliminate reactivating/proliferating infected B cells when they express the growth program. CD4+ T cells directed against lytic and latent EBV antigens are also implicated in controlling EBV responses.

In the immunocompetent host, T cell–mediated responses to the immunogenic proteins prevent outgrowth of EBV-infected B cells. In contrast, when T cell–mediated responses are impaired, as in immunocompromised individuals, including hematopoietic stem cell transplant (HSCT) recipients and SOT recipients, uncontrolled proliferation of EBV-infected B cells will lead to the development of EBV-associated lymphoproliferative diseases (PTLDs).

Immunobiology of EBV

Most cases of PTLD following pediatric SOT are associated with EBV, a latent human lymphotropic virus discovered in 1964 by Epstein, Barr, and Achong. More than 90% of the world’s population is infected with EBV and infected individuals remain lifelong carriers of the virus. The life cycle of EBV is outlined in Fig. 1 . Primary infection occurs through the oropharynx, where the virus infects resting B cells that become proliferating lymphoblasts with expression of EBV nuclear antigen 2 (EBNA2), which activates the promoters necessary to produce all latent proteins (growth program or latency III) ( Table 2 ). The virus gains access to the memory B-cell compartment following the steps required for a normal B cell to move along the pathway from naive to memory cell. After migration of the infected cells into the follicle, EBNA2 is turned off, together with the latent genes that drive proliferation. Infected cells acquire a germinal center phenotype and start the default program (latency II), which is accompanied by the expression of latent membrane protein 1 (LMP1) and LMP2, which provide the necessary signals, T-cell help and B-cell receptor, respectively, that physiologically rescue latently infected germinal center cells into memory cells. These infected B cells that have the surface phenotype of long-lived memory B cells and do not express any of the known latent proteins (latency O) or only EBNA1 (latency I), mature and circulate in the peripheral blood. Release of the virus into the saliva occurs when circulating memory B cells transit the nasopharyngeal lymphoid tissue and differentiate into plasma cells.

EBV life cycle.

Control of virus spread and of unrestrained infected B-cell proliferation is guaranteed by the development of a specific immune response that occurs in healthy immunocompetent individuals and consists of T cells specifically recognizing immunodominant latent EBV proteins presented in the context of major histocompatibility complex (MHC) molecules. NK, CD8+, and CD4+ T cells generally control and contain the proliferation of EBV-infected B cells during primary infection. CD8+ cells, particularly those directed against lytic EBV antigens, expand dramatically during acute infection, and then contract at the resolution of the primary EBV infection. During persistent EBV infection, lytic and latent EBV antigen-specific T cells are maintained at a frequency of 1% to 5% of peripheral blood T cells to immune survey and eliminate reactivating/proliferating infected B cells when they express the growth program. CD4+ T cells directed against lytic and latent EBV antigens are also implicated in controlling EBV responses.

In the immunocompetent host, T cell–mediated responses to the immunogenic proteins prevent outgrowth of EBV-infected B cells. In contrast, when T cell–mediated responses are impaired, as in immunocompromised individuals, including hematopoietic stem cell transplant (HSCT) recipients and SOT recipients, uncontrolled proliferation of EBV-infected B cells will lead to the development of EBV-associated lymphoproliferative diseases (PTLDs).

Epidemiology of PTLD

It is difficult to precisely define the incidence and potential changes in the biologic characteristics of PTLD observed over time. First, there is a lack of large registry data with mandatory reporting of PTLD, established over the long periods of time required to appreciate changes in epidemiology of disease. Second, because of the heterogeneity of disease, a consensus for PTLD diagnosis has been lacking. Nonetheless, the incidence of PTLD seems to be related to several risk factors. The factor that has consistently been demonstrated to convey highest risk for PTLD is EBV seronegativity at time of transplant. Similarly, younger age at transplant is a strong risk factor for PTLD, but likely reflects the proportion of recipients who are EBV naive. The risk is highest following primary EBV infection, which may occur via transmission from the allograft or the natural route via salivary secretion.

In general, it seems that risk of PTLD correlates with the intensity and cumulative exposure to immunosuppression. It is difficult to assess the effect of individual immunosuppressive agents on PTLD risk for several reasons. There is no reliable and reproducible method to measure intensity of immunosuppression. A patient is typically on more than 1 immunosuppressive agent. There is often a learning curve with new agents, such that the incidence of PTLD is often higher when an agent is first introduced to clinical care. To obtain sufficient numbers of patients to assess risk, large registries are required. Registry data are often limited by providing information only if an agent was used, not dosage or duration, or in combination with other agents. Despite the lack of conclusive data, it can be surmised that the more T cell–specific the immunosuppression used, the higher the incidence of PTLD. Therefore, the use of anti-T-cell antibodies is usually associated with the highest risk of PTLD. However, it seems that monoclonal antibodies convey a higher risk than polyclonal antibodies, and cytotoxic antibodies (ie, OKT3 and antithymocyte globulin) increase risk more than anti-interleukin 2 (anti-IL-2) antibodies. The use of calcineurin inhibitors seems to confer the next highest risk. Although early studies suggested the use of tacrolimus increases the risk of PTLD compared with cyclosporine, more recent studies suggest that if serum levels are monitored closely there is no difference in risk of PTLD. It seems that the risk of PTLD is low with the use of mycophenolate mofetil or mammalian target of rapamycin inhibitors such as sirolimus.

PTLD is an early event (ie, within the first 2 years after transplant) in most cases, likely because of more intense immunosuppression used for induction therapy or the exposure of the EBV-naive host to the virus, often from EBV-infected passenger lymphocytes in the allograft. However, in children who receive allografts from EBV-negative donors, primary exposure to EBV is via the natural route of salivary secretion and may occur during adolescence.

Most analyses suggest that the risk of PTLD is associated with the type of organ transplanted. Low-risk patients (eg, renal, heart, and liver recipients) are generally reported to have a risk of PTLD less than 5%, whereas high-risk patients (eg, those with lung, small bowel, and multiple organ grafts) have an incidence greater than 10%. In pediatrics the variability in PTLD incidence between various allografts is less. The incidence seems to be 5% to 10% for most pediatric recipients regardless of allograft, likely reflecting a stronger effect of EBV-naive status compared with type of allograft. The reasons for the observed differences in incidence of PTLD with various allografts are complex. There are recipient factors, such as age and EBV serostatus, but also allograft factors, such as the amount of immunosuppression required to prevent rejection and the differing risk of transmitting EBV-infected B cells (in associated lymphoid tissue) with intestine and lung transplants compared with liver, heart, and kidney transplants.

It seems that the incidence of PTLD in pediatric renal transplant recipients may have been increasing since the 1990s, whereas it may be decreasing in pediatric liver or intestinal transplant recipients. The reasons for these observations are not clear.

Clinical presentation

Every visit by an SOT survivor to a primary care provider represents an opportunity for surveillance for malignancy/PTLD. The clinical presentation of PTLD is variable, depending on the underlying pathologic condition, the type of transplant, and the time since transplant. It is important to maintain a high index of suspicion because the onset of PTLD can be insidious and nonspecific. Frequently, patients present with relatively benign findings (episodic and unexplained fever, weight loss, fatigue) before developing more significant symptomatology. Patients considered high risk and managed as such may be diagnosed with PTLD before becoming symptomatic. Rarely, SOT recipients may present with so-called fulminant PTLD. This term refers to rapidly progressive, disseminated disease resulting in multiorgan failure and is more commonly seen following HSCT than in the SOT context. Other EBV-associated diseases must be differentiated from PTLD, although the initial management is similar.

Early-onset PTLD occurring within 1 to 2 years of transplantation is more common among pediatric SOT recipients because of their risk for primary EBV infection. Early-onset disease is more likely to involve the allograft and present with declining allograft function, excepting heart allografts, in which direct involvement by PTLD is rare. The major differential diagnostic considerations for early onset PTLD include allograft rejection and infection. Later-onset disease is more likely to be EBV negative or include T/NK-cell disease and to involve extranodal sites (especially gastrointestinal [GI] sites) or present with dissemination.

Outside the allograft, common areas affected by PTLD include lymphoid tissues, GI tract, lung, and liver. Patients may present with constitutional symptoms, lymphadenopathy, and organ dysfunction. Disease classified pathologically as early lesions often presents with adenotonsillar involvement with associated sore throat and obstructive symptoms (new onset snoring or mouth breathing). Involvement of the GI tract may present with vomiting, diarrhea, bleeding, intussusception, or obstruction. Perforation may occur at presentation or immediately following initiation of therapy in the presence of transmural lesion necrosis. Chronic ulceration in intestinal transplant recipients should prompt a biopsy to rule out PTLD with samples from the ulcer edge and the intervening mucosa. New-onset anemia or hypoalbuminemia may indicate GI involvement. Lung disease may result in respiratory insufficiency or asymptomatic nodules. Liver disease may present as diffuse hepatitis or nodular lesions. PTLD of the central nervous system (CNS), isolated or as part of multiorgan disease, has been described but is rare. Patients may present with headache, seizures, or focal neurologic findings.

Initial assessment includes a full physical examination, screening blood tests, including a complete blood count with differential, chemistry panel to assess for tumor lysis syndrome, allograft function screening, and EBV viral-load studies. Variable disease presentation can make interpretation of imaging difficult. Ultrasound seems to be effective for initial imaging in patients with suspected abdominal or soft-tissue PTLD. Other imaging modalities should be pursued based on symptoms and initial screening results. The recommended diagnostic and staging procedures are similar to other non-Hodgkin lymphomas (NHLs) and are outlined in Fig. 2 . Presentations according to selected patient series are outlined in Table 3 .

Diagnostic and staging procedures for PTLD.

The role of [ ¹⁸ F]2-fluoro-2-deoxyglucose (FDG)-positron emission tomography (PET) in the diagnosis of equivocal lesions, staging, and response assessment of PTLD is currently being defined. Bakker and colleagues demonstrated additional extranodal sites on PET not appreciated on computed tomography (CT) in 50% of patients and concordance of PET response with outcome. However, false positives have been described when PET is used for disease monitoring in children and for the evaluation of lung lesions. An example of CT-PET imaging for the staging of PTLD is included in Fig. 3 .

A lung transplant recipient presented within 6 months of transplant with worsening pulmonary function tests and new onset of abdominal pain, weight loss, and diarrhea. Endoscopy was performed and biopsy specimens showed polymorphic EBV-associated PTLD. ( A ) Shown here on CT chest is 1 of 2 peripherally enhancing, centrally hypodense mass lesions posterior to and involving the membranous posterior wall of the trachea. ( B ) FDG-PET scanning showed 2 foci of marked uptake in the right paratracheal and pretracheal regions corresponding to the nodes seen on CT. Uptake in the liver, duodenal wall, scattered foci in the bowel, and lymph nodes in the left iliac chain inguinal areas were also noted and were consistent with disease.

Therapy

Although PTLD continues to be a significant cause of mortality in pediatric organ transplant recipients, there are data to suggest that the risk is decreasing. It is unclear if this reflects monitoring of at-risk patients and earlier diagnosis, prevention (primary or secondary), or improved therapy.

Primary prevention

As the knowledge of the risk factors for PTLD increases, attention has turned to primary prevention of disease. Limiting as much as possible the total amount of immunosuppression by adjusting type, combination, and doses of drugs after transplant is key. For example, Opelz and Dohler provided evidence in their analysis of the Collaborative Transplant Study data that the incidence of PTLD among cardiac transplant recipients decreased from 13.9% to 6.2% between 1985 to 1989 and 1995 to 2001, concomitant with reduced use of OKT3 in recent years. Avoiding EBV-positive donors in EBV-naive recipients to limit primary infection may prevent PTLD. However, such donors are in limited supply, with 1 survey study of donor EBV serologic status documenting only 27 of 459 (6%) donors to be EBV negative. An alternate approach is to immunize against EBV before immune suppression. Such vaccines are in development and early phase I/II trials are under way.

The role for passive immunization with the use of anticytomegalovirus intravenous immune globulin (anti-CMV-IVIG) or the use of antiviral therapy for prophylaxis remains unclear. Green and colleagues conducted a randomized, controlled trial of CMV-IVIG for the prevention of EBV disease and PTLD in EBV-seronegative SOT recipients. There was no difference from placebo at 2 years for either end point, although the study was underpowered and confounded by the clinical practice change of decreasing immunosuppression with the finding of rising EBV titres after the start of study recruitment. Another study by Humar and colleagues was unable to demonstrate a benefit of the addition of CMV-IVIG to gancyclovir in EBV load or incidence of PTLD. Opelz and colleagues conducted a large retrospective analysis of the effect of prophylactic anti-CMV-IVIG or antiviral therapy for renal allograft recipients on the incidence of PTLD. There was a statistically significant benefit to the use of CMV-IVIG in the first year after transplant, but not thereafter. This group was unable to detect a benefit to antiviral therapy as prophylaxis for PTLD, although other studies do suggest benefit. Biologically, the role of antivirals in PTLD prevention has been questioned as they do not suppress EBV-driven B-cell proliferation. There may be a benefit to the EBV-naive recipient with an EBV-positive donor, with a reduction in the number of infected B cells.