Blood Reviews (2008) 22, 261–281
www.elsevier.com
REVIEW
Immunodeficiency-associated lymphomas Huy Tran a, Jamie Nourse b, Sara Hall a, Michael Green c, Lyn Griffiths c, Maher K. Gandhi a,b,* a
Department of Haematology, Princess Alexandra Hospital, Ipswich Road, Brisbane, Queensland, 4102, Australia b Clinical Immunohaematology Lab, Queensland Institute of Medical Research, Floor I, CBCRC Building, 300 Herston Rd, Brisbane, 4006, Queensland, Australia c Genomics Research Centre, Gold Coast Campus, Griffith University, PMB 50 Gold Coast Mail Centre, Queensland, 9726, Australia
KEYWORDS
Summary This article covers lymphoproliferative disorders in patients with primary or acquired immunodeficiencies. Primary immunodeficiences include Ataxia Telangiectasia and X-linked disorders such as Wiskott-Aldrich syndrome. Acquired immunodeficiencies predominantly occur in the setting of infection with the Human Immunodeficiency Virus or arise following immunosuppressive therapy administered after organ transplantation. The rising incidence of HIV throughout the world and the dramatic increase in transplant surgery since the 1990’s suggest that these lymphomas will remain an important health problem. Evidence for lymphoma developing as a result of treatment with methotrexate or Tumour Necrosis Factor Antagonists for autoimmune entities will also be reviewed. The lymphoproliferations that occur with immunodeficiency are extremely heterogenous. In part this reflects the diversity of the causal immune defect. The most striking clinical characteristic is the high frequency of extranodal disease. Frequently, these lymphomas are driven by viruses such as Epstein-Barr virus (EBV), although the lack of EBV in a proportion indicates that alternate pathways must also be involved in the pathogenesis. Lastly, discussion will centre on mechanisms utilized by lymphomas in the immunodeficient as these may have applications to lymphomas in the ‘‘immunocompetent’’, by serving as a paradigm for the altered immunoregulatory environment present in many lymphoma sub-types. c 2008 Elsevier Ltd. All rights reserved.
Immunodeficiency; Lymphoma; Ataxia Telangiectasia; Combined variable immunodeficiency disorder; Post-transplantation lymphoproliferative disorder; Epstein-barr virus; Human herpes virus 8; Methotrexate; Infliximab; Human immunodeficiency virus; Primary central nervous system lymphoma; Primary effusion lymphoma; Hodgkin’s lymphoma; Diffuse large B cell lymphoma; Rituximab
* Corresponding author. Address: Clinical Immunohaematology Lab, Queensland Institute of Medical Research, Floor I, CBCRC Building, 300 Herston Rd, Brisbane, 4006, Queensland, Australia. Tel.: +61 7 3845 3792; fax: +61 7 3845 3510. E-mail address:
[email protected] (M.K. Gandhi).
0268-960X/$ - see front matter c 2008 Elsevier Ltd. All rights reserved. doi:10.1016/j.blre.2008.03.009
262
Introduction There are a multitude of factors known to be associated with cancer aetiology, with escape from host immunity only one facet. Transformation of a cell to one with malignant potential in an immunodeficient host may allow clonogenic expansion and eventually clinical cancer.1 The importance of defective immunosurveillance in malignancy is most marked in cells with strong antigenic potential including those that have undergone viral induction. Consistent with this is the observation that the predominant cancer sub-type in immunodeficient subjects is lymphoma (a neoplasm of the immune system), that frequently these lymphomas are driven by viruses such as Epstein-Barr virus (EBV) and that restoration of immunity can result in tumour regression.2 Yet the lack of EBV in a proportion indicates that alternate pathways must also be involved in lymphomagenesis. Indeed the wide clinical spectrum of lymphoproliferative disorders (LPD) encountered in immunodeficiency along with its associations with germ-line and acquired genetic defects, with oncogenic viruses, and with autoimmunity suggests an interplay in which genetic aberrations interact with viral oncogenes, impaired immunosurveillance and chronic antigen stimulation (Fig. 1). Increasing evidence suggests that the immune escape mechanisms utilized by lymphomas in the immunodeficient may have relevance to lymphomas in the overtly ‘‘immunocompetent’’.3–5 Thus the study of immu-
H. Tran et al. nodeficiency-associated lymphoproliferations may serve as a paradigm for the altered immunoregulatory environment present in many lymphoma subtypes.
Primary immune deficiency and lymphoproliferative disorders The occurrence of lymphoproliferative disorders in patients with a primary immunodeficiency (PID) has been documented in the literature for nearly 40 years.6 These PIDs are a heterogeneous group of genetically determined disorders; hence the lymphoproliferative diseases (LPD) that arise are diverse and variable. The risk of developing LPD is related to the type of PID. Accurate quantification of this risk is difficult because PID is rare; hence the incidence of lymphoma is low, leading to a reliance on small case series for estimates. These range from 0.7% to 15%.7 However the PIDs more commonly associated with LPD can be classified as follows:8 T and B cell immunodeficiencies e.g. Severe combined immunodeficiency (SCID), and X-linked hyper-IgM syndrome (XHIGM); antibody deficiencies e.g. Common variable immunodeficiency (CVID); DNA repair defects e.g. Ataxiatelangectasia (A-T); Immune dysregulation e.g. X-linked lymphoproliferative syndrome (XLP); Autoimmunity e.g. Autoimmune lymphoproliferative disorder (ALPS); and other syndromes such
Figure 1 Interaction of factors contributing to lymphomagenesis in immunosuppressed patients. Darker shading indicates an increased risk of developing lymphoma resulting from the combination of factors. Abbreviations: A-T, ataxia-telangiectasia; BCL6, B cell lymphoma 6 gene; WAS, Wiskott-Aldrich syndrome; CVID, common variable immunodeficiency; SCID, severe combined immunodeficiency; HHV8, human herpes virus 8.
Summary of Primary Immune Disorders with an increased risk of Lymphoproliferative Disorders and Lymphoma.
Syndrome/Frequency
Clinical Features
Genetics/Pathogenesis
Immune Defects
LPD/Lymphomas
SCID 1 in 100,000 live births. High prevalence in Navajo & Apache Indians. X-linked (XLSCID) & AR (ADA-SCID) inheritance.
If untreated, death within 1 year due to severe, recurrent infections. Chronic diarrhea, ear infections, recurrent PJP & oral candidiasis.
X-linked SCID: c chain mutations cause defective interleukin signaling; ADASCID: defective adenosine deaminase (ADA), resulting in build-up of lymphotoxins.
Dependent on sub-type, but in general low or absent T cells & NK cells & nonfunctional B cells observed.
Common Variable Immunodeficiency 1 in 10–50,000 live births. Cases are generally sporadic, but 10% are familial.
Several different clinical phenotypes. Prone to recurrent bacterial infections, AID, LPD & granulomatous disease. Elevated risk of malignancy. Typically present aged 20–40 years.
In a proportion, homozygous defects in genes encoding ICOS, CD19, & BAFFR. Defects in TACI can be dominant or recessive. Loci at 4q, 5p, 16q implicated.
Low IgG & IgA; variable IgM. Reduced B cells.
X-linked Hyperimmunoglobulin M Syndrome I in 20,000,000 live male births.
Neutropenia, thrombocytopenia, anaemia, biliary tract/liver disease, recurrent PJP infection and diahorrea. Increased risk of AID.
Low/absent IgG & IgA, normal/increased IgM. Variable defect in effector T cell and macrophage function.
Ataxia Telangiectasia 1 in 40–100,000 live births. AR. Carrier rate 1%.
Cerebellar ataxia, chromosomal instability, oculocutaneous telangiectasia, thymic aplasia, radiosensitivity, recurrent sinopulmonary infections, lymphoid (80%) & other tumours.
X-linked Lymphoproliferative Syndrome-1. (XLPS-2 caused by mutation in XIAP shares many features). 400 documented cases from 100 US families. Autoimmune Lymphoproliferative Syndrome Rare. Frequency unknown. AD & AR forms of inheritance.
Clinical & immunological abnormalities triggered by EBV infection.
CD40 ligand mutation on Xq26–27.2, leading to defective B cell & dendritic cell signaling. CD40L required for isotype class-switching from IgM to IgG or IgA. AT Mutated gene on 11q22– 23, encodes for protein kinase with >40 known substrates. Disorder in cellcycle check-point & impaired DNA double-stranded break repair. Mutations in SH2D1A on Xq25, encodes for SAP which is involved in T–B cell interactions.
FIM following primary EBV infection. Lymphoma not well documented. T cell LPD has developed in a proportion of patients as a result of retroviral gene therapy for X-linked SCID. LPD in the GI tract & lungs. HHV8 implicated in pathogenesis of granulomatous/lymphocytic interstitial lung disease. Increased risk of (typically) B cell, extranodal, Ig secreting NHL. LPD in 65%, predominantly peripheral & abdominal adenpathy. Nodes lack germinal centres. Lymphoma not well documented but LPD can be fatal. Bimodal distribution: age 1–5: T cell acute lymphoblastic leukaemia/ lymphoma; young adults: T prolymphocytic leukaemia.
3 groupings, characterized by the defect in the FASmediated apoptosis pathway.
Increased circulating CD4-/ CD8-T cells due to defective lymphocyte apoptosis. Normal B cells & Igs.
Present in infancy with autoimmune cytopenias, AID, adenopathy & splenomegaly.
IgA, IgE, IgG2 & IgG4 decreased. Increased IgM monomers. Progressive decrease in T cells. Normal B cell numbers.
Variably decreased Igs. B cells normal or reduced. Defective lysis & polarization of EBVspecific T and NKT cells.
Immunodeficiency-associated lymphomas
Table 1
FIM, aplastic anaemia & hepatitis. Survivors acquire hypogammaglobulinaemia & extranodal B cell NHL, particularly of the terminal ileum. 50· increased risk of HL; 10– 15· increased risk of NHL.
263
(continued on next page)
7% develop NHL, with CNS involvement common. Lymphomatoid Granulomatosis also seen. Progressive decrease in T cells, normal B cells. Decreased IgM & antibody to polysaccharides. IgA & IgE maybe raised. PJP: Pneumocystis Jirovecii Pneumonia. FIM: Fulminant Infectious Mononucleosis. LPD: Lymphoproliferative Disease. NHL: Non-Hodgkin’s Lymphoma. AID: Autoimmune Disease. ICOS: Inducible Co-Stimulator Protein. BAFFR: B Cell Activating Factor Receptor. TACI: Transmembrane Activator and Calcium Modulator and Cyclophilin Ligand Interactor. SAP: (Signaling Lymphocyte Activation Molecule: SLAM) Associated Protein. Igs: Immunoglobulins. HL: Hodgkin’s Lymphoma. CNS: Central Nervous System.
LPD/Lymphomas Immune Defects
Thrombocytopenia with small platelets, eczema, AID, IgA nephropathy, bacterial & viral infections.
Mutations in WAS Protein encoded on Xp11.22–23, resulting in abnormal cytoskeletal architecture of haemopoietic cells.
Clinical Features
Wiskott-Aldrich Syndrome X-linked. 4 in 1,000,000 live male births.
Genetics/Pathogenesis
H. Tran et al.
Syndrome/Frequency
Table 1 (continued)
264
as Wiskott-Aldrich syndrome (WAS). Features of these conditions are outlined in Table 1. What these heterogenous disorders have in common are a primary deficiency in immunity and a predisposition to lymphoma. Furthermore, treatment with allogeneic stem cell transplantation (alloSCT) leads to reduction in susceptibility to LPD, emphasizing the causal nature of PID in the development of lymphoma.9 At an experimental level, engraftment of Epstein-Barr virus (EBV) seropositive peripheral blood mononuclear cells in SCID mice causes delayed but lethal LPDs.10 Yet it is overly simplistic to ascribe lymphomagenesis solely to defective immunosurveillance. For example, sporadic mutations in the FAS gene are associated with extranodal lymphomas in the absence of immune abnormalities.11 In A-T the underlying defect is a deficiency in the surveillance, initiation, and integration of cellular responses to DNA double stranded breaks/damage (DSBs).12 DSBs are created during normal lymphoid development and this renders lymphoid tissues particularly susceptible to oncogenic transformation, as elegantly demonstrated in ATM knock-out mice.13 The above-mentioned PIDs do share some common features. They usually present in childhood (except for CVID) and affect more males (many are X-linked inherited) than females. Associated lymphomas generally have a predilection for extranodal sites (central nervous system and gastro-intestinal tract); are rapidly progressive if untreated and are usually B cell in origin. PID associated lymphomas are frequently associated with EBV,7,14 consistent with the notion that immunodeficiency is permissive for EBV-driven LPD. The lymphoproliferation maybe polyclonal, and even if monoclonal may not necessarily progress to lymphoma in all cases.15 Histological classification of lymphomas is according to the World Health Organization (WHO) and resembles those occurring in patients without predisposing immune defects. Most common is diffuse large B cell lymphoma (DLBCL), although specific congenital immunodefiencies have associations with particular lymphoproliferations. For example, WAS is characterized by lymphomatoid granulomatosis (an EBV-driven T-cell rich B cell LPD), and in A-T there is a pronounced predisposition to developing T-cell leukaemias and lymphomas. Patients generally present with infectious complications and it is rare that lymphoma is the cardinal sign of an underlying immunodeficiency.8 Of interest is a condition primarily seen in XLP and SCID patients, termed Fulminant Infectious Mononucleosis (FIM), characterized by an abnormal immune
Immunodeficiency-associated lymphomas response to primary EBV infection. FIM is a lifethreatening condition marked by fever, rash, generalized lymphadenopathy, hepatosplenomegaly, and cytopenias. Characteristic findings include the uncontrolled systemic expansion of polymorphous B lymphocytes, involving lymphoid and non-lymphoid organs (such as the terminal ileum). EBV-specific cellular immunity is defective.16,17 Haemophagocytois is frequent, with extensive marrow infiltration by lymphoid cells and cellular necrosis resulting in severe pancytopenia. Opportunistic infections and liver failure, often associated with acute hemorrhage are the major causes of death.18 In XLP, those surviving frequently acquire hypogammoglobulinaemia (involving lymph nodes necrosis although peripheral B cell numbers maybe maintained) or malignant B cell lymphoma. EBV may also trigger aplastic anaemia and/or hepatitis. Pre-emptive administration of the anti-CD20 antibody rituximab may reduce the devastating sequelae of primary EBV infection in these patients.19 There is limited data on treatment and prognosis in PID associated lymphomas given the rarity of this disease and the wide variety of predisposing immunodeficencies. In the absence of randomized clinical trials, it is recommended that treatment be tailored according to histological subtype and prognosis. Replacement immunoglobulin is indicated for hypogammaglobulinaemia, as is early and aggressive treatment for infections with special consideration for Pneumocystis Jirovecii Pneumonia antibiotic prophylaxis. There are currently two clinical trials underway examining the efficacy of anti-oxidants in the treatment of A-T. AlloSCT has been shown to restore immune function and prevent lymphoma in ATM-deficient mice.20 Clinically, alloSCT has been used successfully for lymphomas arising in XLP, ALPS, SCID, WAS and XHIGM.21,22 Gene therapy has been successfully used for SCID. Although cases of T-cell LPD driven by integration of the retroviral vector have been reported,23 these results should be seen in context, and it is still gene therapy that has the greatest potential for long-term cure of PID.
Post-transplant lymphoproliferative disorders Post-transplant lymphoproliferative disorder (PTLD) represent the group of lymphoid disorders that arise following solid-organ transplantation (SOT) and stem cell transplantation.24–27 Relative to the immunocompetent, the risk of lymphoma after transplantation is increased by the order of
265 30 times.28 The incidence of PTLD varies, depending on the intensity of immunosuppression, recipient age, the organ transplanted, the number of previous allografts, and in liver transplant recipients by the presence of Hepatitis C cirrhosis or primary biliary cirrhosis.29 These risk factors are reflected in the highest incidence of PTLD occurring after lung and small bowel transplants (up to 30%,30 in contrast to 1–5% for renal, cardiac and liver transplants. After SCT, the cumulative incidence of PTLD at 10 years is 1%.27 As immunosuppression protocols become less intensive, so the relative frequency of PTLD after SOT may diminish. However since the 1990’s there has been a dramatic rise in transplant surgery and this coupled with improvements in long-term SOT patient survival, will likely result in an absolute rise in the incidence of PTLD. Approximately 80% of PTLD biopsies are positive for EBV within the tumour cells. Reflecting the critical role of reduced cellular immuno-surveillance against Epstein-Barr virus (EBV) in the pathogenesis, other risk factors are the EBV-serostatus of the donor (D) and recipient (R) (for SOT: D+/R ; and for SCT D /R+ are at greatest risk respectively); cytomegaloviral disease31 and the use of T cell depleting agents. Furthermore, use of the anti-CD52 antibody alemtuzumab which depletes both B and T cells is associated with a lower incidence of PTLD than when T cells are depleted in isolation.32 In SOT recipients, the incidence is highest in the first year post-transplant (the period of most intense immunosuppression), and a reduced risk remains thereafter. Early post-transplantation EBV-seroconversion due to primary EBV infection places the recipient at a major risk of EBV-positive PTLD. Although at least 90% of the population are EBV-seropositive by the age of 40, children are frequently seronegative, and in the majority of PTLD seen in the paediatric setting primary EBV infection has occurred in the 6 months prior to presentation.33,34 In vitro, TOR inhibitors such as sirolimus reduce the proliferation of EBV transformed lymphoblastoid cell lines (LCL),35 and unlike calcineurin inhibitors such as cyclosporine and tacrolimus due not protect LCL from apoptosis.36 The role for sirolimus as a immunosuppressive agent following SOT is still evolving, and currently there is insufficient data for a definitive statement as to whether patients are less likely to develop PTLD. Although exceptions have been reported, in general following SOT, EBV-positive PTLD arises from host lymphocytes. In EBV-seronegative recipients, recipient cells are infected by the virus present in the donor organ; whereas in seropositive recipients the PTLD tissue frequently but not invariably
266
H. Tran et al.
contains the recipient EBV isolate.37 In contrast, in allogeneic stem cell transplants (SCT) it is generally donor-derived lymphocytes that undergo transformation. Frequently, this is as a result of loss of endogenous EBV and acquisition of the donor virus isolate.38 PTLD is a heterogenous syndrome ranging from a benign self-limited form of lymphoproliferation to an aggressive, widely disseminated lymphoma. Patients with lympomatous-PTLD appear to have a more aggressive clinical course and poorer outcomes than lymphomas in immunocompetent hosts.39 The clinical presentation is highly variable (see Table 2), correlating to some extent with the type of transplant, as well as the specific morphologic subtype of PTLD. Early-onset PTLD commonly presents after an episode of graft rejection, occurs in younger patients, and typically consists of an infectious mononucleosis like syndrome with cervical and tonsillar enlargement or simply pyrexia of unknown origin. Like other immunodeficiency lymphomas, late-onset PTLD (occurring more than one year post-transplant) commonly presents with extranodal disease which may manifest as organ dysfunction often including the allograft. Identified prognosticators include lactate dehydrogenase, time from transplant to PTLD, presence of B symptoms, allograft involvement with tumour, performance status, response to rituximab monotherapy and EBV tumour status40–44 (and Hourigan personal communication). Based on these findings, a number of PTLD-specific prognostic indices have been proposed, but require validation in large-scale prospective clinical trials. The importance of interim positron emission tomography positivity remains to be elucidated.
Table 2
Approximately 90% are of B cell origin, and the majority of these are EBV-associated. T cell PTLD is generally EBV negative, and in common with non-EBV-associated B cell PTLD, tend to occur late (>1 year) after solid-organ transplantation (SOT). The malignant cells observed in PTLD may be reactive/hyperplastic, polymorphic (heterogenous) or monomorphic (homogenous) in appearance, and may arise from polyclonal, oligoclonal or monoclonal populations. The relationship between clonality and histology is complex, but in general early lesions are polyclonal, whereas polymorphic and monomorphic are monoclonal. Progressive transition from polyclonal lymphoproliferation to outgrowth of a malignant clone has been reported.34 The histological terms have been integrated under one system, which categorises lesions as ‘early’ (including reactive plasmacytic hyperplasia and infectious mononucleosis-like syndrome); polymorphic PTLD; monomorphic PTLD (individual B and T cell lymphomas are then classified according to the WHO lymphoma classification); and lastly miscellaneous lesions such as Hodgkin-like lymphoma. Recently, it has been suggested that the classification should include EBV status and clonality (Fig. 2).45 Classically, PTLD are believed to be the result of unchecked proliferation of polyclonal EBV transformed B cells in the absence of EBV-specific cellular surveillance. The expression of latent EBV oncogenes such as LMP1, EBNA2 and EBNA3A are strongly suggestive of a direct role of EBV in tumourogenesis. The presence of clonal EBV within monoclonal tumours places EBV infection as an early or potentially pre-malignant event. Subsequent genetic alterations involving oncogenes or
Unique features of PTLD as compared to Diffuse Large B Cell Lymphoma in the immunocompetent.
PTLD
DLBCL
80% EBV-tumour tissue positive 80% extranodal disease (may include allograft) 30% CNS disease at presentation (generally multicentric cerebral) 60% PCNSL Early-onset PTLD: B cell origin Late-onset PTLD: both B and T cell origin Poorly tolerant of chemotherapy Immunosuppressed prior to therapy Management includes prevention of organ rejection Initial therapy increasingly instituted by transplant physicians
10% EBV-tumour tissue positivea 30% extranodal disease 17% CNS recurrence if two risk factors present; 3% if one or no risk factorsb 173 Rarely presents as PCNSL Germinal centre and post-germinal centre B cell sub-types Chemotherapy generally well-tolerated Not immunosuppressed No transplant organ concerns Therapy supervised by haemato-oncologists
Abbreviations: CNS: central nervous system; PCNSL: primary CNS lymphoma. a Based on Korean and Japanese data,174,175 unknown in other populations. b Risk factors defined as P2 extranodal sites and raised LDH.
Immunodeficiency-associated lymphomas
267
Figure 2 Relationship between EBV latency gene expression, histological subtype of PTLD and recovery of EBVspecific immunity. After transplant EBV-specific immunity is ablated but gradually returns. In PTLD a spectrum of EBV latent gene expression is observed, with a broad range of genes tending to be expressed in early onset PTLD, whereas in late onset cases expression is restricted.
tumour suppressor genes are thought to accumulate resulting in a survival advantage of an aggressive malignant clone.46 5’ non coding region mutations of BCL6 gene are frequent and are associated with an aggressive clinical outcome.25 Other factors are likely implicated. For example the predilection of PTLD to extranodal sites suggests that the tumour microenvironment is important. The heavy infiltration with CD4 T cells47 may be an indicator for the importance of immuno-regulatory factors, as seen in EBV-positive Hodgkins Lymphoma.4,5 Involvement of the transplanted organ which is itself subject to frequent sub-clinical rejection might induce a state of chronic antigen stimulation, as might co-infection with non-EBV pathogens. The incidence of non-EBV-positive PTLD remains higher than in non-immunocompromised subjects, and the possible role of other viruses in lymphomagenesis (such as Human Herpes Virus 6) remains unresolved.48 Although reduction of immunosuppression is recommended as initial therapy, as monotherapy this
generally leads to response in only a minority of patients, particularly those with early lesions or wildtype BCL-6.25,43 Nevertheless in all patients with PTLD the immunosuppression should be cautiously reduced as per current European and American guidelines (see Table 3).45,49 Indeed some centres report good results with complete cessation of immunosuppression during induction chemotherapy for renal-PTLD (D. Gill, personal communication). Chemotherapy remains the gold-standard therapy but is frequently poorly-tolerated and results remain worse than for equivalent lymphomas in nonimmunosuppressed patients.50 The anti-CD20 antibody rituximab is generally non-toxic and has activity as a single-agent but relapse is frequent.51,52 Sequential combination therapy with x4 cycles of weekly rituximab followed by x4 G-CSF supported CHOP-21 reports encouraging results, with response to rituximab predictive of survival .53 Importantly, this combination does not appear to impair EBV-specific cellular immunity,54 except in those patients receiving ongoing immunosuppression.55
268 Table 3
H. Tran et al. Recommended guidelines for the reduction of immunosuppression in PTLD.
Europe49 – Maintain only steroids OR – decrease by 50% the anti-calcineurin drugs and stop other immunosuppressive agents eg azathioprine/mycophenolate)
USA45 – Limited disease: 25% reduction in immunosuppression – Extensive disease: Critically ill: stop all except prednisolone 7.5–10mg/day. Not critically ill: decrease cyclosporine/tacrolimus by 50%, discontinue azathioprine/mycophenolate and maintain prednisolone 7.5–10mg/day
There are anecdotal reports of PTLD regression following therapy with the nucleoside analogues acyclovir and ganciclovir.56 Other authors have documented poor clinical outcomes with aciclovir,57 and the consensus is that antiviral therapy for treating PTLD is unlikely to be of benefit in the majority of patients. This is not surprising since nucleoside analogues require viral thymidine kinase (TK) and BGLF4 to be converted to its active cytotoxic form. This involves the action of cellular enzymes which incorporates the nucleoside triphosphate into DNA leading to the premature termination of the DNA. However, in its latent episomal form TK is not expressed. Transcriptional regulators such as arginine butyrate (a short-chain fatty-acid) have been shown to up-regulate TK58 and in combination with ganciclovir show promising responses in patients with EBV-positive lymphomas, with reasonable tolerability59 (and unpublished data M. Gandhi). By a similar mechanism, in vivo data using a SCID mouse model suggest that the addition of ganciclovir to either gemcitabine or doxorubicin may enhance the therapeutic efficacy of these drugs for EBV-driven lymphoproliferative disease in patients.60 The expression of EBV latent antigens in PTLD provides an ideal target for T cell-based immunotherapy. The importance of donor-derived CTL as prophylaxis against EBV-driven B cell expansions has been shown in allogeneic SCT patients transfused with EBV-specific CTLs.61 To date, more than 60 bone-marrow transplant patients have been infused with EBV-specific CTL lines as a prophylactic treatment, and none of these patients has shown any symptoms of PTLD. Interestingly, many of these adoptively-transferred EBV-specific T cells can be detected 2–3 years after the infusion. Although the idea of applying a similar rationale of adoptively transferring EBV-specific CTLs to resolve PTLD arising in solid organ recipients is attractive, there are fundamental differences between bone marrow and solid-organ transplantation that pose a major challenge. These include activating a CTL response in vitro in cells from patients receiving high levels of immunosuppressive
drugs and the efficacy of adoptively-transferred CTLs in the face of high levels of immunosuppression in vivo. However, autologous EBV-specific CTL have successfully been used to prevent PTLD in high-risk patients with an elevated EBV viral load, and as therapy in patients with established PTLD.2,62 The combination of rituximab with reduced dose chemotherapy and subsequent autologous CTL has been used to treat PTLD in the paediatric renal transplant setting with durable complete remissions achieved in all five patients treated.63 Autologous CTL generation takes several months, and the EBV-transformed lymphoblastoid cell-lines required for their generation cannot be established in patients prior treated with the B cell depleting antibody rituximab. An alternate strategy, which overcomes this limitation is to establish banks of cryopreserved, partially HLA-matched EBV-specific CTL derived from healthy seropositive subjects,64 which offers a distinct logistic advantage of speed of access and ease of generation that may permit their wide-scale use in both PTLD and other malignancies, such as Hodgkin’s Lymphoma (HL).This strategy has proven clinical efficacy in peripheral and primary CNS PTLD65 as well as lymphoma in the setting of primary immunodeficiency.66 A phase II trial on relapsed/refractory patients demonstrated 64% response rates with no allo-reactivity despite HLA disparity.67 Homing of allogenic CTL was demonstrated at the tumour site68 (see Fig. 3). Measuring the EBV viral load in plasma and PBMC has been used to identify patients at risk of developing PTLD.69 Studies are hard to compare since study design as well as EBV detection methods and analysis are highly variable. Although both PBMC and plasma have value to quantitate viral load by real-time PCR, if PBMC is used then threshold levels are required to differentiate between the EBV-infected B cells observed in healthy subjects and EBV-infected neoplastic cells. Thus the specificity of the analysis is higher if plasma is used for analysis.70 Also, in patients treated with rituximab the disappearance of cellular viral load does not predict clinical response.71
Immunodeficiency-associated lymphomas
269
Figure 3 Allogeneic EBV-specific T cells home to sites of disease. Immunohistochemistry (IHC) and fluorescent in-situ hybridization (FISH) in left and right panels respectively demonstrating scattered CD8+ T cells and XY cells admixed within the B cell PTLD, taken at autopsy. The donor-CTL donor was male, and the recipient was female. The PTLD was of recipient origin. Reproduced with kind permission from an article originally published in the American Journal of Transplantation.68
Methotrexate-associated lymphoproliferative disorders The development of methotrexate (MTX) by Farber’s team in Boston was a pivotal moment in the modern era of chemotherapy. It remains a critical component of acute lymphoblastic leukaemia therapy, and high-dose MTX is the mainstay for the treatment of lymphomas within the nervous system. Ironically, low-dose MTX, used most commonly in the setting of treatment for RA, psoriasis, dermatomyositis and myasthenia, is implicated in lymphomagenesis. In their most recent classification, the World Health Organization included MTX associated lymphoproliferative disorders as a subcategory of lymphoma.72 The histology’s observed are variable and include DLBCL, HL, follicular lymphoma, lymphoplasmacytic lymphoma, mantle cell lymphoma and a picture analogous to polymorphous PTLD.73 Extranodal presentation appears common, but in contrast to AIDS patients high frequencies of neurological involvement are not seen.74 However, studies are conflicting and there is no definitive epidemiological evidence as to the extent to which MTX increases the risk of lymphoma in such patients, if at all. No excessive risk of lymphoma was observed in one retrospective study75 and in several longitudinal studies of patients receiving MTX, even after long-term follow-up.76,77 These studies included only several hundred patients, and may therefore have been insufficiently powered to detect an excessive risk of a rare event (lymphoma), and may be further confounded by reported increase in prevalence of lymphoma in patients taking cyclosporine A.78 A national French prospective study conducted over 3 years found a significant
increase in HL but not NHL in RA patients treated with MTX.79 The issue of causality is confounded by the estimated 2-fold increased risk of lymphoma (particularly NHL) in patients with rheumatoid arthritis (RA) (see Fig. 4).80,81 MTX has two mechanisms of action. It is an antimetabolite that competitively inhibits dihydrofolate reductase (DHFR), thus preventing DNA, RNA and proteins synthesis, resulting in cytotoxicity during the S-phase of the cell-cycle. At the lower doses used in the management of RA, inhibition of DHFR is not thought to be the principle mechanism. Rather, many of the anti-inflammatory effects of MTX are mediated by adenosine, or the inhibition of T cell activation and suppression of intercellular adhesion molecule expression by T cells.82 Several genes within the folate pathway have been associated with the risk of lymphoma in general, including 5,10-methylenetetrahydrofolate reductase (MTHFR), methionine synthase, serine hydroxymethyltransferase, thymidylate synthase, and folylpolyglutamate synthase.83–88 However, these associations are heavily confounded by variances in the dietary intake of folate, methionine and multiple co-factors for their metabolism.88 To our knowledge, no data exists as to whether the variations in these genes are linked to MTXassociated lymphoma susceptibility. Intriguingly, approximately 50% of the lymphoproliferative disorders are EBV-associated.89 The frequency of EBV infection within the malignant cell varies between histology’s, with rates highest in HL (approximately 75% of cases). Since diagnostic laboratories tend to use only EBER staining, the pattern of EBV viral latency seen in these lymphomas remains to be ascertained.
270
H. Tran et al.
Figure 4 Effects of methotrexate potentially contributing to the development of lymphoma. The action of methotrexate appears to result in two broad categories that could contribute to the development of lymphoma. The combination of these factors may result in an increased risk of lymphoma.
The strongest evidence for causality, is that there are numerous case reports of at least partial regression of lymphoma upon withdrawal of MTX, estimated as approximately 60% of cases.89,90 The majority of these have been EBV positive.89 Although reporting bias almost certainly distorts the true picture, this is an observation that has been made repeatedly and consistently. The reported incidence of regression varies with the specific histology involved, with regression more typical with lymphoplasmacytic lymphomas than with HL and DLBCL.91 Although it is known in RA that circulating EBVpositive B cells are abnormally high,92 the mechanism of transformation employed by MTX remains unclear. MTX is able to activate lytic viral replication of LCL in vitro and it has been shown that RA and polymyositis patients treated with MTX have higher PBMC viral loads than similar patients treated with other immunosuppressive regimens.93 EBV load was also higher than that observed in Wegener granulomatosis patients treated with MTX (consistent with the observation that these patients are not at an increased risk of EBV-driven lymphoproliferative disorders). Given the known transforming capacity of EBV, the additional immunosuppressive effects of MTX may be sufficient to explain the modest increase in lymphoma cases in this setting. Although EBV-specific cellular immunity has been shown to be impaired in RA patients (using a regression assay),94 no data on the role of EBV-specific cellular immunity in the setting of MTX associated lymphomas is to our knowledge published.
Similarly, the value of EBV viral load monitoring remains to be elucidated. Although 50% of cases are EBER-ISH positive, there is no definitive laboratory test to distinguish lymphomas in which MTX is causally implicated, versus those in which MTX administration is more likely coincidental. Current recommendations are to discontinue MTX. Depending on the pace of disease, the lymphoma should either be observed for regression with time, or alternatively to chemotherapy commenced. For B cell lymphomas, the addition of the CD20-depleting monoclonal antibody rituximab at induction and/or as maintenance has the dual advantage of anti-lymphomatous activity and prolonged anti-autoimmune activity.73
Tumour necrosis factor antagonists Experimental evidence and clinical experience with Tumour Necrosis Factor-a inhibitors (antiTNF) demonstrate the central role for this cytokine in the pathogenesis of Rheumatoid Arthritis (RA), Crohns Disease (CD) and other autoimmune diseases. Three agents are currently licensed: infliximab (a chimeric monoclonal antibody), etanercept (a TNF-a receptor fused to the Fc region of IgG) and adalimumab (a recombinant monoclonal antibody). These agents were initially approved for moderate to severe RA in patients who had insufficient response to one or more diseasemodifying antirheumatic drugs (DMARDs), and subsequently for use as first-line treatment in
Immunodeficiency-associated lymphomas Table 4
271
Key Features of Hepatosplenic T cell lymphoma.
Epidemiology Rare (approximately 200 cases in world-wide literature). Estimated at
