Post-transplant lymphoproliferative disorders (Review for Seminars in Nephrology)

Introduction

Post-transplant lymphoproliferative disorders (PTLDs) are a heterogenous set of unregulated lymphoid cell proliferations after organ or tissue transplant (1, 2). While many cancers occur at higher frequency in the setting of reduced immune surveillance because of the non-specific immune suppression by the immunosuppressant medications used to prevent acute rejections (3), PTLDs are among the most common (4, 5) and most feared. Some of these unregulated lymphoid proliferations can evolve into full-fledged cancers such as frank lymphomas, but even other subtypes carry a significant risk of morbidity and mortality.

Epidemiology:

PTLDs are relatively rare, more so after kidney transplantation. Their frequency is best expressed as an incidence density, with time under immunosuppression as a denominator. Renal allografts that were lost early and had their immunosuppression stopped early would otherwise be inappropriately be counted in the denominator if raw proportions are used. From the Organ Procurement and Transplant Network database in the USA, the incidence density for PTLD after kidney transplant was reported as 1.58 per 1000 person-years (6), while from the United Kingdom, the incidence density of the PTLD sub-type non-Hodgkin lymphoma was 2.6 per 1000 person-years (7). The ANZDATA registry reported a 25-year cumulative PTLD incidence of 3.3% in adults and 3.6% in children, adjusted for competing risks such as death (8). In children after kidney transplant, PTLD incidence density jumped in the 1990s-2000s (9), but more recent eras are associated with a reduction in incidence across both children and adults (8, 10). In a very recent prospective multicentre study, where 944 children were enrolled and 872 received some type of solid organ transplant (including higher risk intestinal transplant), only 34 PTLDs had occurred in a group followed till at least 1 year (11).

The risk factors for PTLDs that are consistently reported can be divided into 2 groups: non-modifiable and modifiable. Among the non-modifiable risk factors are: a) the extremes of recipient age – the youngest and oldest are at higher risk. Since cancers are much less common in children, their standardized incidence ratios are magnified 20–40 fold (8, 12); b) recipient Epstein-Barr virus (EBV) seronegativity (13, 14) (see Pathogenesis section for a detailed discussion of EBV relationship to PTLD); c) recipient race/ethnicity – white recipients are at higher risk than Black, Hispanic or Asian recipients, even after multivariable adjustments for recipient EBV serostatus (13); d) type of organ transplant – intestinal, lung and heart transplant recipients are at comparatively higher risk than liver, kidney or pancreas transplant recipients (15).

Among the modifiable risk factors are: a) the overall intensity of immunosuppression – a greater number of agents or higher doses/levels increase the risk (16–18). Though many studies report on individual immunosuppressive agents, the only agents that consistently showed a higher risk were the monoclonal antibody OKT3 (19) or the fusion protein belatacept (20). For the latter, PTLD rates were 10-fold higher in EBV-seronegative recipients (5% rate vs 0.5% in EBV-seropositive recipients), and CNS localization of PTLDs were seen in 9/15 (60%) of patients who received belatacept, a very high proportion compared to 10–12% in other series (9, 18, 21, 22). Most cases occurred within the first 18 months. Other agents such as anti-thymocyte globulin induction, de novo sirolimus use or tacrolimus-mycophenolate combinations are inconsistently reported, with conflicting data on risk, suggesting that totality of immunosuppression plays the major role. Other inconsistently reported risk factors are recipient HLA alleles or the use of anti-viral prophylaxis (23, 24).

Pathogenesis:

Majority of PTLD cases are associated with EBV (25, 26). EBV infection is almost ubiquitous, with more than 90% of adults showing EBV seropositivity (27). EBV has several differential gene expression patterns, defining the distinct latency programs. During type III latency, which occurs shortly after infection and is also known as proliferation program, full range of EBV latent genes is expressed including six nuclear antigens (EBNA-1, −2, −3A, −3B, −3C, and -LP) and three latent membrane proteins (LMP-1, −2A, and −2B) as well as several untranslated RNAs (28, 29). In immunocompetent individual, the proliferation of EBV-infected B-cells is controlled by vigorous cytotoxic T-lymphocyte response which rapidly clears the vast majority of the infected B cells (30, 31). A small subpopulation of EBV infected B-cells persists for lifetime of the host by downregulating viral antigen expression and escaping immune surveillance (28).

In the post-transplant setting, these latent EBV-positive B cells may give rise to EBV-positive PTLD, as a consequence of immunosuppression-induced deficiencies in EBV-specific T cells. Type III latency is found in majority of EBV-positive PTLDs (29, 32, 33), with full expression of EBV latent genes providing critical growth and survival signals for infected B-cells. Several of EBV-encoded proteins, including LMP1, have been shown to be essential for B-cell transformation (32). LMP1 is the major EBV oncogene, that mimics CD40 receptor signaling pathway (34) and prevents apoptosis by upregulating anti-apoptotic protein Bcl-2 by activating NF-kB pathway (35, 36). Interactions between LMP1 and adaptor molecules also lead to signal transduction through the MAPK p38, ERK, and JNK, and PI3K signaling pathways (37–40). LMP-1 can increased expression of regulatory cytokine IL-10 (41) leading to the attenuation of anti-viral adoptive immune and proliferation of EBV-positive B-cells (42). Expression of LMP1 in B cells in genetically engineered mice was sufficient to induce immune surveillance of the LMP1-positive cells by cells of the adaptive immune system, and weakening of this surveillance mechanism by depleting T cells resulted in the rapid development of B cell lymphomas in LMP1-positive mice model (43).

Recent studies have utilized advanced sequencing techniques to focus on EBV genome diversity and gain additional insights into pathogenesis of EBV-positive PTLD. Sequencing of EBV genomes from FFPE PTLD samples revealed that EBNA-1, EBNA-LP and EBNA3C genes had increased number of variations compared to other EBV genes (44). Peripheral blood sequencing from solid organ transplant recipients revealed that EBV genomes from PTLD patients had significantly more variations than EBV genome from transplant recipients without PTLD, including variations within EBNA3C and LMP2A epitopes targeted by T-cell response. Computational modeling suggested that the PTLD-associated variations could impact antigen presentation and subsequent T-cell response, altering host immunity to EBV and promoting escape from the immune response (45). Multicenter prospective study demonstrated that presence of G212S and S366T mutations in LMP1 is associated with a significantly increased risk of EBV-positive PTLD in pediatric solid organ transplant recipients, while absence of both mutations is associated with low probability of developing PTLD (11).

EBV-negative PTLD is less common than EBV-positive PTLD but accounts for up to 40% of PTLD cases, with incidence of EBV-negative PTLD increasing over the years. In contrast to EBV-positive PTLD, EBV-negative PTLD cases occur more commonly in adults and later after transplantation, with higher proportion of monomorphic PTLD cases (46, 47). Despite these differences, proportion of EBV-negative cases may respond to decreased immunosuppression, and are considered a subset of PTLD category (48).

Pathogenesis of EBV-negative PTLD is uncertain, with several proposed etiologies including another viral infection, undetectable EBV infection due to loss of EBV or technical reasons, prolonged immunosuppression, and transplant-related chronic antigenic stimulation (49). With the exception of rare human herpesvirus 8 (HHV8)-positive PTLDs (50, 51), association of PTLD development with viral infection other than EBV remains unproven. In a metagenomic shot gun sequencing analysis of 69 PTLD tissue samples, no viruses were detected in higher proportion of EBV-negative PTLD cases as compared to EBV-positive PTLDs (44). Further evidence that viral oncogenesis is not a key pathological pathway in EBV‐negative PTLD comes from the studies comparing gene expression profiles and genomic aberrations of EBV-positive and EBV-negative diffuse large B-cell lymphoma (DLBCL) PTLD to that of DLBCL arising in immunocompetent hosts. These studies have shown that EBV-negative DLBCL PTLD cases have molecular signatures that cluster with that of DLBCL in non-immunosuppressed patients and are distinct from EBV-positive DLBCL PTLD (32, 52–54).

PTLD Classifications:

Since first histological description of PTLD more than 50 years ago (55), PTLD designation has come to encompass a spectrum of lymphoid or plasmacytic proliferations, ranging from usually EBV-positive polyclonal proliferations to EBV-positive or EBV-negative malignancies that have the same histological appearance as non-Hodgkin or Hodgkin lymphoma in immunocompetent patients. Due to this broad range of histological and immunophenotypic findings and varying degree of tissue destructiveness and clinical behavior, PTLD lesions require further classification. Several editions of the World Health Organization (WHO) classification of Tumours of Haematopoietic and Lymphoid Tissues have incorporated diagnostic criteria for different types of PTLD. Revised 4th edition of WHO classification recognized four major categories of PTLD: non-destructive PTLD, polymorphic PTLD, monomorphic PTLD, and Classic Hodgkin Lymphoma PTLD (48).

As the name suggests, non-destructive PTLDs preserve architecture of the involved tissue, commonly tonsils and lymph nodes. This group of PTLDs was previously designated as “early lesions” but this term was retired due to confusion with PTLD cases that occur shortly after transplantation. These lesions are polyclonal mixed lymphocytic and plasmacytic infiltrates showing a spectrum of B-cell maturation. They lack of histological features diagnostic for a malignant lymphoma. Non-destructive PTLD category is further divided into three histologic subtypes: infectious mononucleosis, plasmacytic hyperplasia, and florid follicular hyperplasia. Infectious mononucleosis PTLD has histological appearance associated with primary EBV infection in immunocompetent host, with interfollicular expansion composed of small lymphocytes, plasma cells and usually frequent larger transformed lymphoid cells known as immunoblasts. Plasmacytic hyperplasia is composed of polytypic plasma cells, small lymphocytes and occasional immunoblasts. Florid follicular hyperplasia is composed of reactive secondary follicles with prominent germinal centers (Figure 1). Most non-destructive PTLD cases are EBV-positive and form mass lesions, and these characteristics help to distinguish plasmacytic hyperplasia and florid follicular hyperplasia, which otherwise have non-specific histological findings, from reactive lymphoid or plasmacytic proliferations (48, 49).

Figure 1.

Non-destructive PTLD, florid follicular hyperplasia (revised 4^th WHO classification/ICC) / Florid follicular hyperplasia, EBV-positive, post-transplant (5th edition of WHO classification). Low-power view of tonsil with preserved architecture and florid follicular hyperplasia with multiple reactive secondary follicles ((A) H&E stain, original magnification 2x). Germinal center ((B) H&E stain, original magnification 20x) with increased EBV-positive cells ((C) In situ hybridization for EBV (EBER), original magnification 20x).

Polymorphic PTLDs are also heterogeneous mixtures of lymphocytic and plasmacytic components that show a full range of B-cell maturation, including small- to intermediate-sized lymphocytes, plasma cells and immunoblasts (Figure 2). However, unlike non-destructive PTLD, polymorphic PTLD cases are associated with effacement of underlying tissue architecture and usually show clonal IgH rearrangements, reflecting emergence of clonal B-cell population. The histological findings in polymorphic PTLD do not fulfill the criteria of any of the recognized types of lymphoma described in immunocompetent hosts but they may have necrosis and/or increased large lymphoid cells or Reed-Sternberg-type cells, making distinction from monomorphic PTLD difficult in some cases (48, 49).

Figure 2.

Polymorphic PTLD (revised 4^th WHO classification/ICC) / Polymorphic lymphoproliferative disorder, EBV-positive, post-transplant (5th edition of WHO classification). Effacement of architecture ((A)H&E stain, original magnification 20x) by heterogeneous lymphoid proliferation ((B) H&E stain, original magnification 50x) composed of mixture by CD20-positive B cells ((C), original magnification 20x) and CD3-positive T cells ((D), original magnification 20x) with scattered admixed CD30-positive immunoblasts ((E), original magnification 20x). In situ hybridization for EBV ((F), original magnification 20x) highlights frequent positive cells.

Monomorphic PTLDs, which account for most cases of PTLD, are clonal lymphoid or plasmacytic proliferations that fulfill the diagnostic criteria for one of the non-Hodgkin lymphomas or plasma cell neoplasms recognized in non-immunosuppressed patients. The monomorphic PTLD cases may be of B-cell or T/NK-cell origin and are further categorized according to the lymphoma they resemble in immunocompetent hosts. The majority of cases of monomorphic PTLD are B-cell neoplasms, most commonly diffuse large B-cell lymphoma. Histologically, these are overtly malignant proliferations composed of large atypical lymphoid cells expressing B-cell markers (Figure 3) and usually exhibiting increased proliferation rate, apoptosis and necrosis. Post-transplant DLBCLs may be EBV-positive and EBV-negative. Gene-expression profile and immunohistochemical staining have been used to classify DLBCL cases in immunocompetent subjects into germinal center cell or non–germinal center cell origin (56, 57). In PTLD setting, EBV-positive DLBCL cases predominantly have non-germinal center cell immunophenotype, while significant subset of EBV-negative DLBCL cases is of germinal center cell origin (32, 58). It is important to note that indolent B-cell neoplasms arising in the post-transplant setting, with the exception of uncommon EBV-positive MALT lymphomas occurring preferentially in skin or subcutaneous tissue, are not considered PTLD.

Figure 3.

Monomorphic B-cell PTLD, diffuse large B-cell lymphoma (revised 4^th WHO classification/ICC) / Diffuse large B-cell lymphoma, EBV-positive, post-transplant (5th edition of WHO classification). Diffuse proliferation ((A) H&E stain, original magnification 20x) of large atypical cells with irregular nuclear contours and prominent nucleoli ((B) H&E stain, original magnification 50x) positive for B-cell marker CD20 ((C) original magnification 20x) and EBV ((D) in situ hybridization for EBV (EBER), original magnification 20x).

Monomorphic T-cell PTLDs are less frequent and are more commonly EBV-negative. Most of subtypes of T/NK-cell lymphomas have been reported in the post-transplant setting, with peripheral T-cell lymphoma, not otherwise specified, representing the most common subtype, followed by hepatosplenic T-cell lymphoma (48).

Classic Hodgkin lymphoma PTLD is the least common category of PTLD. These PTLDs meet the diagnostic criteria for classic Hodgkin Lymphoma, with relatively low numbers of malignant Reed-Sternberg cells in a non-neoplastic inflammatory background (Figure 4) and are overwhelmingly EBV-positive. Because Reed-Sternberg-like cells may be seen in other subtypes of PTLD, diagnosis of classic Hodgkin lymphoma PTLD requires presence of both morphological and immunophenotypic findings characteristic of classic Hodgkin lymphoma outside of post-transplant setting (48, 49).

Figure 4.

Classic Hodgkin Lymphoma PTLD (revised 4^th WHO classification/ICC) / Classic Hodgkin Lymphoma, EBV-positive, post-transplant (5th edition of WHO classification). Non-neoplastic inflammatory background composed of small lymphocytes and histiocytes ((A), H&E stain, original magnification 20x) with scattered admixed Reed-Strernberg cells ((B) H&E stain, original magnification 50x) positive for CD30 ((C), original magnification 20x) and EBV ((D) in situ hybridization for EBV (EBER), original magnification 20x).

Two new/updated classifying proposals of hematolymphoid neoplasms have been published in 2022: 5th edition of the WHO classification and International Consensus Classification (ICC) (59, 60). Updated WHO classification has integrated PTLD categories into overarching framework for immunodeficiency-associated lymphoproliferative disorders arising in the different settings of immune dysfunction (60) (Table 1). It has been previously observed that PTLD-like lesions arise in patients are immunosuppressed due to other iatrogenic or genetic causes. In prior classifications including revised 4^th edition of WHO classification (48), these lymphoproliferative disorders were organized according to the disease background in which they arose, with separate topics pertaining to post-transplantation, HIV infection, other iatrogenic immunodeficiencies and primary immunodeficiencies. Citing the evidence that morphological features and possibly the pathogenesis of many of these entities overlap, 5^th edition of WHO classification introduced the term immunodeficiency and dysregulation associated lymphoproliferative disorders (IDD- LPSDs) (60). Integrated classification of IDD-LPSDs was first proposed the Workshop on Immunodeficiency and Dysregulation in 2015 (61, 62). The new standardized multi-part nomenclature for reporting the diagnosis organizes information around common histologic and pathogenetic features IDD-LPSDs and includes: (1) histological diagnosis (hyperplasia, polymorphic LPD, lymphoma as for immunocompetent patients); (2) presence or not of virus such as EBV; (3) clinical setting/immunodeficiency background (post-transplant, HIV, iatrogenic/autoimmune) (60).

Table 1.

Different PTLD classification systems

WHO classification, revised 4th edition	International Consensus Classification	WHO classification, 5th edition
Non-destructive PTLD  Subtypes:  ● Plasmacytic hyperplasia  ● Infectious mononucleosis  ● Florid follicular hyperplasia	Non-destructive PTLD  Subtypes:  ● Plasmacytic hyperplasia  ● Infectious mononucleosis  ● Florid follicular hyperplasia	Hyperplasia arising in immune deficiency/dysregulation  Subtypes:  ● Plasmacytic hyperplasia  ● Infectious mononucleosis Florid follicular hyperplasia
Polymorphic PTLD	Polymorphic PTLD	Polymorphic lymphoproliferative disorder arising in immune deficiency/dysregulation
Monomorphic B- and T-cell PTLD further categorized based on classification of B-cell and T-cell lymphomas in immunocompetent hosts	Monomorphic B- and T-cell PTLD further categorized based on classification of B-cell and T-cell lymphomas in immunocompetent hosts	Lymphomas arising in immune deficiency/dysregulation further categorized based on classification of B-cell and T-cell lymphomas in immunocompetent hosts
Classic Hodgkin Lymphoma, PTLD	Classic Hodgkin Lymphoma, PTLD	Lymphomas arising in immune deficiency/dysregulation further categorized as classic Hodgkin lymphoma

In contrast to the 5th edition of WHO classification changes, consensus ICC opinion acknowledged that EBV-positive B-cell lymphoproliferative disorders arising in different immunodeficiency backgrounds share some histological findings but concluded that further studies are needed. Thus, in ICC system PTLD category remains separate and unaltered from revised 4^th edition of WHO classification (Table) due to significant differences in clinical management (63).

Clinical features and diagnosis:

The clinical features for PTLD can be both diverse and non-specific, necessitating a high index of suspicion. Routine physical examination should include looking for any visible or palpable lymph node masses in the neck, axilla, groin or abdomen. Non-specific clinical features can include fever, night chills, weight loss or unexplained anemia. Some of the clinical features relate to the location of the lymphoid mass, such that gastrointestinal symptoms (diarrhea, vomiting) are seen if the PTLD is within the intestine. Similarly, central nervous system tumors may show symptoms and signs related to the location of the mass and the areas of the nervous system that are compressed.

With the increasing use of regular EBV nucleic acid testing in peripheral circulation, the diagnosis of PTLD can often be suspected even when asymptomatic. While in general, higher EBV DNA loads associate to higher PTLD risk (64), the specificity of individual prediction is low (65). The reasons for this relate to the varying sample sources (lymphocytes, whole blood or plasma), variations in the EBV PCR assay methodologies across laboratories (66) and the differing values and units (67), now attenuated but still quite variable after the implementation of an international EBV standard and increasing use of international units/ml (68). Both EBV peak load and rate of rise in EBV load exhibit areas-under-the-curve value in the 0.6–0.7 range, not high enough to be diagnostic in a given individual (69). For EBV-negative PTLDs, the EBV peripheral load is not useful.

Imaging, especially with radiation, should be reserved for those situations where suspicion of PTLD is high. Imaging can identify which lesions are amenable to biopsy. A tissue diagnosis is still the gold standard, though in some inaccessible locations such as the brain, a combination of clinical, laboratory and imaging features are used to make a presumptive PTLD diagnosis.

Prevention and pre-emptive interventions for PTLD

Given the morbidity and mortality of PTLDs, prevention is highly desirable. Unfortunately, no EBV vaccine has reached the clinical realm. One vaccine to the EBV glycoprotein gp350 was tested in children with chronic kidney disease who were to receive a kidney transplant, but did not prevent either EBV replication or PTLD (70). Vaccines with other mechanisms are under development. The use of intravenous immunoglobulin or CMV-Ig, both of which would carry some EBV-directed antibodies, has conflicting results. Large retrospective data suggested benefit (71), whereas small prospective studies did not (72). Chemoprophylaxis with antiviral agents such as acyclovir, ganciclovir or their analogues is similarly controversial. A meta-analysis by AlDabbagh et al showed no benefit in PTLD prevention (23), but European registry data do show a benefit in reducing primary infection EBV DNAemia (73) and another group showed a benefit in preventing late PTLDs, those occurring after the first year post-transplant (24).

Rituximab use has become established in hematopoietic stem cell transplants as an effective pre-emptive measure at the time of EBV DNAemia. But unlike in kidney transplants, 90% of PTLDs are EBV positive in these populations, and the EBV DNAemia is more monophasic, occurring predominantly within the first year post-transplant. Two retrospective studies, one small (74) and one larger (75), have suggested that rituximab can prevent the progression to PTLD in adult kidney transplants recipients too. The use of EBV-directed cytotoxic T lymphocytes as a pre-emptive measure at the time of EBV DNAemia is attractive conceptually, but remains to be studied.

Treatment of PTLD

Treatment of PTLD typically involves a stepwise approach, balancing control of the PTLD with allograft preservation and side effects of therapy. Collaboration between specialists is essential for evaluating timing of PTLD diagnosis in relation to transplant, degree of necessary immunosuppression, allograft function, histological subtype, and overall health of the patient when considering available treatment options. The first line treatment is reduction of immunosuppression (RIS) to enhance alloreactive T-cell immunity. Response rates to RIS alone range widely from 43 to 63% in retrospective studies and are higher in patients with non-destructive and EBV positive PTLD (76, 77). Close monitoring is necessary due to the increased risk of organ rejection and graft loss with RIS. In rare circumstances when PTLD is localized, surgery may also play a role in treatment.

For patients who either do not respond to RIS, or are at high risk of developing rejection and are not candidates for RIS, rituximab monotherapy is often the next line of treatment for CD20 positive PTLD. Rituximab is an anti-CD20 monoclonal antibody which targets CD20, a transmembrane protein present on almost all B cells, and decreases B cell proliferation. Prospective trials of rituximab monotherapy evaluating 3 or 4 weekly doses reported overall response rates ORR of 44–79% with complete response of 25%–53% (78–81). Long-term survival for patients who achieve a CR to rituximab alone is excellent with disease-specific survival at 10 years of 88% (82).

The international prospective phase II PTLD-1 trial of adults with solid organ transplant PTLD introduced a risk stratified sequential treatment approach (83). Patients with CR after four doses of rituximab induction then received for additional doses every 21 days with estimated three-year progression-free survival (PFS) of 89% and overall survival (OS) of 91%. Patients with a partial response (PR) or progressive disease then received chemotherapy with R-CHOP (rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone). For this group of patients requiring chemotherapy, ORR was 88%, with 70% CR, and treatment-related mortality of 8%, highlighting the risk of chemotherapy toxicity in this patient population. Since PTLD arises from an underlying immunodeficient environment, treatment regimens developed for acquired immunodeficiency syndrome‐associated lymphomas, such as DA-EPOCH-R (dose adjusted etoposide, prednisone, vincristine, cyclophosphamide, doxorubicin, and rituximab) have also been trialed in advanced PTLD with comparable efficacy and tolerability to R-CHOP (84, 85). Overall, the use of risk stratified sequential treatment in patients with PTLD, especially polymorphic or diffuse large B-cell (DLBCL) sub-type, has led to good outcomes and decreased treatment-related mortality.

In pediatric patients with EBV positive PTLD, two phase-two trials have evaluated the use of low-dose chemotherapy including cyclophosphamide and prednisone for six cycles with or without rituximab (86, 87). Patients treated with cyclophosphamide, prednisone and rituximab had a 2-year EFS of 71%, including functioning original allograft, and OS of 83% with median follow-up of 4.8 years (87). Response-adapted sequential treatment strategy with rituximab +/− chemotherapy in pediatric patients with CD20+ PTLD has also been investigated (88). Patients received three weekly infusions of rituximab and if at least a PR was obtained they then received three additional rituximab doses. 64% of patients attained a CR with rituximab alone and 81% of responders remain alive and in continuous complete remission with a median follow up of 4.9 years. Patients with stable or progressive disease then received a moderate chemotherapy regimen mCOMP (vincristine, prednisone, cyclophosphamide and methotrexate). 66% of patients who received chemotherapy achieved a CR with two-year EFS 67% and OS 86%. Notably patients with liver or renal transplant had superior EFS to patients after thoracic organ transplantation 79% vs 44%.

Rare subtypes of PTLD including Hodgkin lymphoma, T-cell lymphoma, and central nervous system (CNS) lymphoma typically do not respond to RIS or rituximab monotherapy and should be treated with the standard of care chemoimmunotherapy for the specific histology. Patients with Hodgkin lymphoma PTLD treated with standard Hodgkin chemotherapy protocols have favorable outcomes, though not as good as for non-immunocompromised patients treated on similar protocols (89, 90). Chemotherapy dose modifications are common due to patient comorbidities. T-cell lymphoma PTLD, including peripheral T-cell lymphoma, hepatosplenic T-cell lymphoma, and anaplastic large cell lymphoma are often not associated with EBV, occur later than B-cell PTLD, and typically have a poor prognosis (91, 92). PTLD involving the CNS also has a poor prognosis. Treatment approaches include chemotherapy with high-dose methotrexate, higher dose intravenous and intrathecal rituximab, radiotherapy, and EBV cytotoxic T-lymphocytes (EBV-CTLs) (93, 94).

Adoptive T-cell therapy using EBV-CTLs is a newer treatment approach designed to address the underlying immunological defect. For EBV positive PTLD and other immune deficiency–related lymphoproliferations, the goal of treatment is to eradicate EBV infected B-cells and recover the EBV-specific T-cell immune response. EBV-CTLs for patients following solid organ transplant can be autologous, generated from the patient‟s own cells, or third party CTLs, generated from healthy HLA matched, EBV seropositive donors. The benefit of third party CTLs is that they are rapidly available for patients, while autologous cells take several weeks to generate. Products are chosen based on best HLA match and HLA restriction of viral activity. Both types of EBV-CTLs are generally well tolerated and generate good responses without causing organ rejection (95–97). Maintaining adequate immunosuppression throughout treatment is critical for preventing organ rejection. Concurrent use of steroids has not demonstrated a significant difference in response to EBV-CTLs, though most studies have limited the dose to <0.5 mg/kg prednisone or its equivalent in order to maintain CTL functionality (98). EBV-CTLs have been used as consolidation following chemotherapy and used prophylactically to treat EBV viremia (99). Several Phase II/III trials of EBV-CTLs in PTLD patients are currently underway to further evaluate the efficacy of this approach.

The use of other immunotherapies in PTLD treatment have been tried with caution due to the potential risks of T-cell activation and allograft rejection. Chimeric antigen receptor T-cell (CAR-T) therapy targeting CD19 is being increasingly used for treatment of lymphomas, but has not been studied systematically in PTLD. Questions remain about use of immunosuppressive therapy during CAR-T, its impact of efficacy of treatment, and risk of rejection (100). Most patients with PTLD receiving CAR-T in a real world case series had their immunosuppression discontinued prior to CAR-T infusion and a few remained on prednisone 5mg daily, resuming immunosuppression after a median of 2.2 months (101). Of the 20 patients included in the series, seven remain in remission after two years and three experienced kidney allograft rejection with two requiring hemodialysis and one patient dying due to PTLD progression. Similarly, checkpoint inhibitors are being incorporated into lymphoma treatment, however, the PD-1 and CTLA-4 pathways are important for allograft tolerance, raising concerns about risk of rejection with use of these therapies. In a single center retrospective review of patients post-SOT who received checkpoint inhibitors as part of their treatment for a range of cancers, primarily metastatic melanoma, 41% of patients (16/39) experienced organ rejection and 81% of those experienced graft loss (102). For patients with PTLD, at this point in time the risks of checkpoint inhibitors likely outweigh the benefits.

Prognosis

Imaging plays a critical role in diagnosis, staging, and response assessment of PTLD. 18F-fluorodeoxyglucose-positron emission tomography (FDG-PET) is the most sensitive and specific imaging modality for Hodgkin and non-Hodgkin lymphoma and is increasingly being used for patients with PTLD (103). PET scans may be helpful in cases of suspected PTLD. A retrospective study of 125 PET scans performed for differentiating PTLD from other diseases found a sensitivity of 90%, specificity of 89%, positive predictive value of 85% and negative predictive value of 93% (104). For staging, PET is more appropriate than a conventional computed tomography (CT) scan because PTLD often has extra-nodal involvement that is not always adequately detailed on CT (105). A meta-analysis of three studies evaluating PET response in PTLD showed that PET detected more lesions than CT in 15% of the scans and showed no metabolic activity in 32% of the lesions detected on CT, ultimately changing treatment decisions in 29% of patients (106). End of therapy (EOT) PET may also be useful for identifying patients at low risk of relapse. In a retrospective study of patients with CD20+ PTLD treated with rituximab or rituximab and chemotherapy, the positive predictive value of EOT PET for PTLD relapse was 38%, and the negative predictive value was 92%. Further studies are needed to better understand the prognostic utility of EOT PET.

A number of prognostic indices have been investigated in adults with PTLD. The International Prognostic Index (IPI), developed as a predictive model for aggressive non-Hodgkin Lymphoma, is based on the baseline parameters age, stage of disease, Eastern Cooperative Oncology Group (ECOG) performance status, involvement of extranodal sites, and LDH (107). However, several retrospective studies of patients with PTLD demonstrated that the IPI and clinical prognostic factors used for NHL in immunocompetent patients may not be as accurate for immunosuppressed patients with PTLD (108–111). Alternate prognostic indices for patients with PTLD have been developed, utilizing elements of the IPI and other PTLD specific factors such as monomorphic disease and graft involvement (Table 2). Notably, the ‘Renal-PTLD’ Index was specifically developed for renal transplant and includes elevated LDH and B-symptoms as negative prognostic markers (111). The applicability of these indices has changed over time with the evolution of PTLD treatment, especially the introduction of rituximab to upfront therapy. The PTLD-1 study which used a risk-stratified sequential treatment approach evaluated the use of these scoring systems for predicting outcome and found that the IPI, PTLD Prognostic Index, and Ghobrial Score were predictive for overall survival (112). Age and ECOG performance status had the highest prognostic impact in the different scoring systems. Other variables not included in the scores that were found to impact overall survival and disease progression were the type of transplant and the response to rituximab at interim staging. Further refinement and validation of these scores through a prospective trial may help guide treatment stratification.

Table 2:

Comparison of PTLD prognostic indices

Variable	IPI	PTLD Prognostic Index	Leblond Score	Ghobrial Score	Renal-PTLD Index
Age >60	X	X
Ann Arbor Stage III-IV	X
ECOG >=2	X	X	X	X
LDH > 1x norrmal	X	X			X
> 1 Extranodal site*	X
> 1 Disease site			X
Monomorphic disease				X
Graph involvement				X
B-symptoms					X
Score	0–5	0–3	0–2	0–3	0–2

^*Extranodal = Bone marrow, GI tract, liver, lung, CNS, skin, testes, Waldeyer’s ring

Prognostic risk factors in pediatric PTLD are poorly characterized with no published prognostic scores due to its rarity. Studies of pediatric patients with PTLD have identified multiple prognostic factors including age at transplantation, EBV status, deceased donor graft, HLA mismatch, intensity of immunosuppression, organ rejection, bone marrow or CNS involvement, and PTLD time of onset (113–117). Lack of response to initial PTLD therapy has also been shown to be a poor prognostic factor (117). All of these factors are important to take into account when considering the best treatment approach for pediatric patients with PTLD.

Concluding remarks:

While we have learned much about PTLD etiopathogenesis in the last 2 decades, many knowledge gaps still remain. While some pre-emptive therapies and treatment strategies are available have reasonable success are available, they do not eliminate the high morbidity and significant mortality after PTLD.