SEARCH

SEARCH BY CITATION

Keywords:

  • EBV;
  • posttransplant malignancies;
  • PTLD;
  • rapamycin;
  • skin cancer

Abstract

  1. Top of page
  2. Abstract
  3. Background
  4. Immune Surveillance and the Impact of Immunosuppression
  5. Immune Surveillance Revisited
  6. Other Considerations Relevant to Immunosuppressive Agents
  7. Tumor Cell Growth Factors and Soluble Mediators
  8. Conclusion
  9. Acknowledgments
  10. References

The increased risk for the development of malignancies in transplant recipients is generally attributed to the debilitated immune system that results from chronic exposure to potent immunosuppressive drugs required to prevent graft rejection. While impaired immunity is clearly a key determinant, there is strong evidence that a constellation of other factors contribute to the pathogenesis of posttransplant cancers. In this article we discuss the underlying molecular and immunologic mechanisms that contribute to the development of de novo malignancies in transplant recipients, with particular focus on the two leading posttransplant neoplasia, skin cancer and Epstein–Barr virus (EBV)-associated posttransplant lymphoproliferative disorder (PTLD).


Background

  1. Top of page
  2. Abstract
  3. Background
  4. Immune Surveillance and the Impact of Immunosuppression
  5. Immune Surveillance Revisited
  6. Other Considerations Relevant to Immunosuppressive Agents
  7. Tumor Cell Growth Factors and Soluble Mediators
  8. Conclusion
  9. Acknowledgments
  10. References

The pathogenesis of cancer is a multifactorial process involving genetic, immune, environmental and in some cases viral components. Alterations in the regulation of cell growth, resistance to apoptosis, defects in DNA repair, mutations in tumor suppressor genes and oncogenes and sustained angiogenesis can all contribute to tumor development. In transplant recipients immunosuppressive drugs severely disrupt immune function but can also have direct effects at the site of tumor formation. Together, these changes create a setting in which cell transformation is enhanced and tumor cells can gain advantage. Malignancies in transplant recipients can result from transfer of neoplastic cells from organ donors, de novo cancers or recurrence of primary tumors. In this article we focus on the most common de novo posttransplant malignancies, skin cancer and Epstein–Barr Virus (EBV)-associated B-cell lymphoma, and discuss the immunologic and molecular mechanisms that promote development of these cancers in transplant recipients.

Immune Surveillance and the Impact of Immunosuppression

  1. Top of page
  2. Abstract
  3. Background
  4. Immune Surveillance and the Impact of Immunosuppression
  5. Immune Surveillance Revisited
  6. Other Considerations Relevant to Immunosuppressive Agents
  7. Tumor Cell Growth Factors and Soluble Mediators
  8. Conclusion
  9. Acknowledgments
  10. References

The tenet that the immune system is responsible for ongoing detection and elimination of newly arising transformed cells in the host was first proposed by Burnet (1) and Thomas (2). As the complexity of host-tumor interactions has been unraveled in the intervening decades, the validity of this model has been debated. Nevertheless, several lines of evidence support the role of T lymphocytes, natural killer (NK) cells and cytokines in protecting the host from spontaneous or chemically induced tumors in experimental animal models (3). In transplant recipients the concept of cancer immune surveillance is dramatically demonstrated. Graft recipients that are immunologically naïve for EBV are at greater risk of developing EBV-associated PTLD than patients who are seropositive for the virus at transplant. Further, in those patients that do develop EBV+ B-cell lymphomas, reduction or withdrawal of immunosuppression can result in tumor regression in a significant number of patients. Similarly, tapering or withdrawal of immunosuppression in transplant patients can result in regression of Kaposi's sarcoma (KS) associated with human herpes virus (HHV)-8 (4), and can diminish or abrogate the occurrences of skin carcinomas associated with human papilloma virus (HPV) infections, but a causal link is not entirely clear (5). In the transplant scenario, those tumors linked to oncogenic viruses are potentially more immunogenic through presentation of viral peptides than tumors arising from chemical or environmental carcinogenesis, where tumors may not as readily be recognized as nonself. Because of this, growth of virus-associated tumors may be more amenable to reversal by the host immune system when immunosuppression is reduced. The specific lymphoid populations that can mediate tumor regression in transplant recipients following reduction of immunosuppression are not established but T lymphocytes and NK cells are likely to participate. We (group of O.M.M.) have shown with the use of MHC/peptide tetramers that immunosuppressed transplant recipients can develop and maintain a high proportion of EBV-specific CD8+ T cells poised to eliminate virally infected cells when the restraints of immunotherapy are removed (6). Nevertheless, restoration of host immunity following withdrawal of immunosuppression does not uniformly result in tumor regression, indicating other factors are at play (Figure 1). Indeed, once EBV+ B-cell lymphomas have converted from the polyclonal form to the monoclonal form, their aggressive properties and large growth advantage may override the ability of immune surveillance to control the tumor. Finally, allograft recipients are at increased risk for a few specific malignancies but are not broadly predisposed to all cancers as the immune surveillance model might predict.

image

Figure 1. Immune and tumor-derived factors that contribute to the development of EBV+ B-cell lymphomas in transplant recipients. The EBV-encoded latent cycle gene LMP1 induces production of cellular IL-10 that acts as an autocrine growth factor for EBV+ lymphoblastoid B cells and can inhibit antiviral T cells. IL-6 also acts as an autocrine growth factor and can be produced by non-B cells, including monocytes, in PTLD lesions. LMP1 and other latent cycle EBV genes induce survival and antiapoptotic proteins. Immunosuppressive drugs, including cyclosporine (CS) and Rapamycin impair T-cell immunity. However, unlike CS, Rapamycin also can directly inhibit production of IL-10 and can interfere with cell cycle progression and proliferation of EBV+ B cells.

Download figure to PowerPoint

Immune Surveillance Revisited

  1. Top of page
  2. Abstract
  3. Background
  4. Immune Surveillance and the Impact of Immunosuppression
  5. Immune Surveillance Revisited
  6. Other Considerations Relevant to Immunosuppressive Agents
  7. Tumor Cell Growth Factors and Soluble Mediators
  8. Conclusion
  9. Acknowledgments
  10. References

Clearly there exists a complex and dynamic interplay between tumor-intrinsic and host-derived factors that determine tumor development and progression. Along these lines, the classic tumor immune surveillance model has been updated and refined by Schreiber and colleagues (3) to incorporate the concept of ‘immunoediting’, which suggests that the immune system can influence and shape tumor growth through three processes that may occur independently or in sequence. First, as in the classical immune surveillance model, the immune system can recognize and eliminate tumor cells. Second, there exists a state of ‘equilibrium’ such that the immune system holds existing cancer cells in check. Third, during ‘escape’, tumor variants with low or absent immunogenicity, or with the ability to suppress immune responses, can develop into clinical malignancy. A classic example of the latter, demonstrated in the 1970s in mice (7), are UV radiation-induced skin cancers that are highly antigenic, but nevertheless escape rejection because of an acquired UV tumor-specific tolerance by the host (7) involving transferable ‘suppressor’ T cells, now dubbed ‘regulatory’ T cells (8). These skin cancers are readily rejected when transplanted into naïve syngeneic mice. Here, chronic UV irradiation and the tumor apparently ‘edit’ the immune system.

Recently, elegant studies by Koebel et al. (9) have demonstrated in a mouse model of methylcholanthrene (MCA)-induced sarcomas that occult tumor cells can persist in immunocompetent hosts but that outgrowth of sarcomas can result when components of the adaptive immune system, including CD4+ T cells, CD8+ T cells, IFN-γ and IL-12, are eliminated. Further, those tumor cells that do escape equilibrium with the immune system appear to be ‘edited’ and are less immunogenic since they grow when transferred to either immunodeficient Rag−/– mice or to wild type recipients. Conversely, ‘unedited’ tumor cells that are maintained in equilibrium in wild type hosts are immunogenic because they can grow in Rag−/− mice but are rapidly rejected when transferred into immunocompetent mice. These findings may help explain donor transfer of occult cancer that is maintained in equilibrium by the donor but becomes clinically apparent in an immunosuppressed graft recipient. Escape may also be relevant to PTLD since EBV has been shown to actively subvert host immunity through a variety of evasive maneuvers (10). In particular, recent studies indicate that the latent membrane protein 1 (LMP1) of EBV, expressed in PTLD-associated lymphomas, can actively block apoptotic signals delivered through the Fas/Fas ligand and TRAIL/death receptor pathways (11). Thus, the current paradigm that it is simply immunosuppression that prevents host effector pathways from eliminating tumor cells may have to be revised to include the concept that the virus plays an active role in this process as well.

Other Considerations Relevant to Immunosuppressive Agents

  1. Top of page
  2. Abstract
  3. Background
  4. Immune Surveillance and the Impact of Immunosuppression
  5. Immune Surveillance Revisited
  6. Other Considerations Relevant to Immunosuppressive Agents
  7. Tumor Cell Growth Factors and Soluble Mediators
  8. Conclusion
  9. Acknowledgments
  10. References

The cumulative amount of immunosuppressive drugs, and in some cases, the use of particular immunosuppressive drugs such as OKT3 (antibody directed at CD3 on T cells) have been associated with the development of skin cancer (12,13) and PTLD (14). This observation may be linked to suppressive effects of these medications on the immune system, but direct effects of these medications at the site of tumor development may also play a role, most notably, effects raising mutagenesis in cells and speeding up tumor growth.

Besides impairing the ability of the host immune system to eliminate tumor cells, immunosuppressive drugs also dampen antiviral immunity. This can increase the likelihood of infection by oncogenic viruses such as EBV, HCV, HBV, HPV or HHV-8 and predispose the patient to malignancy. Recent studies demonstrate that certain types of beta-HPVs can hamper UV-induced apoptotic responses (15) or render the host cell ‘immortal’ (16) and thus enhance skin carcinogenesis. At the same time, immunosuppressants may have direct effects on viral infection as was shown in early studies demonstrating the frequency of outgrowth of immortalized EBV-infected B lymphoblasts is increased in the presence of cyclosporine A (CS).

The strong increase in skin carcinomas in immunosuppressed transplant recipients was assumed to be a logical consequence of a compromised immune system that would otherwise eliminate the antigenic skin cancers. However, it was subsequently shown that Azathioprine (Aza) and CS could lower the repair of DNA lesions induced by solar UV radiation (17,18). CS also appears to hamper apoptotic responses of overly damaged cells, which may further enhance UV mutagenesis and carcinogenesis. Finally, CS enhances angiogenesis and tumor growth (19). Recent studies indicate that Aza results in incorporation of 6-thioguanine pseudobases in the DNA, which photosensitize the DNA and thus enhance genotoxicity and mutagenicity from solar UV exposure (20). Furthermore, the 6-thioguanine is a substrate of mismatch repair (MMR), which slows down DNA replication and cell growth. Thus, a loss of MMR bestows a selective growth advantage on cells—at the cost of genetic instability-–which would explain the loss of MMR in acute myeloid leukemia (21) and sebaceous carcinoma (22) in organ transplant recipients receiving Aza. The nonimmunologic, local effects of CS and Aza on the skin's defenses against UV radiation may result in an increase of microscopic clones of cells overexpressing a mutant form of the p53 tumor suppressor protein, as indicated by a recent study in renal transplant recipients by one of our groups (23). These ‘p53 patches’ are putative precursors of actinic keratoses and squamous cell carcinomas (SCC), and were found to not be immunogenic in mice (23,24), in contrast to the ultimate skin tumors. Hence, there is clear evidence that the conventional immunosuppressants Aza and CS increase the risk of skin carcinomas through direct adverse effects on skin cells (depicted in Figure 2). The immunosuppressant Rapamycin operates through an entirely different mechanism, and blocks the outgrowth of tumors (19). The initial clinical results indeed show that the skin cancer rate is considerably lower after a conversion to this novel immunosuppressant (25).

image

Figure 2. Flow diagram of the process of UV-induced skin carcinoma, and the interference from the immunosuppressive drugs, azathioprine (Aza), cyclosporine (CS) and Rapamycin (Rapa).

Download figure to PowerPoint

Finally, the chronic immune activation and inflammation associated with alloreactivity and viral infection, even in the face of immunosuppression, may contribute to a tumorigenic environment. Indeed, in recent years the concept that the inflammatory microenvironment within tumors may facilitate, rather than hinder, tumor progression has gained momentum. Immune mechanisms that have been implicated include tumor-infiltrating regulatory T cells that suppress antitumor immune responses, immature or tolerogenic dendritic cells (DC) that are poor stimulators of T cells, the provision of costimulatory survival signals to tumor cells by tumor infiltrating lymphocytes, expression of negative costimulatory pathways such as programed death 1 (PD-1) and PD ligand 1 that attenuate antitumor T-cell responses, and the secretion of immunosuppressive factors and tumor growth factors. Whether or not these tumor immunomodulatory mechanisms are operative in the context of the complex immune environment of immunosuppressed hosts bearing an allograft is an important question for future studies. So far, good evidence supports a role for the production of growth factors that modulate the tumor microenvironment and can directly augment tumor growth.

Tumor Cell Growth Factors and Soluble Mediators

  1. Top of page
  2. Abstract
  3. Background
  4. Immune Surveillance and the Impact of Immunosuppression
  5. Immune Surveillance Revisited
  6. Other Considerations Relevant to Immunosuppressive Agents
  7. Tumor Cell Growth Factors and Soluble Mediators
  8. Conclusion
  9. Acknowledgments
  10. References

The production of growth factors by tumor cells can promote tumor progression through autocrine and paracrine growth and survival pathways, suppression of host antitumor immunity or by enhancing tumor invasion and metastases.

  • (i) 
    IL-6. IL-6 is a multifunctional cytokine produced by DC, macrophages, fibroblasts, endothelial cells, T cells and B cells that can participate in an array of immune functions. Several lines of evidence suggest IL-6 may be involved in the pathogenesis of PTLD. IL-6 can act as an autocrine or paracrine growth factor in EBV-infected B cells (26) and expression of IL-6 by EBV+ B cells increases growth and tumorigenicity in vivo (27). Moreover, IL-6 transcripts are found in PTLD tissue and, interestingly, the primary source of IL-6 appears to be the adherent, non-B cells in PTLD specimens (28). Increased IL-6 levels have been measured in the serum of PTLD patients (28) although these results have been challenged in more recent studies (29). Immunosuppressive drugs may contribute to increased IL-6 levels in transplant recipients since CS has been shown to enhance IL-6 production in T cells and monocytes (30). OKT3 also induces production of IL-6 and this has been proposed as one mechanism by which the use of this T-cell-depleting agent is a risk factor for development of EBV-related B-cell lymphomas (31). A small multicenter phase I–II study (32) using anti-IL-6 monoclonal antibodies in 12 patients with PTLD was encouraging showing good safety and clinical efficacy in eight patients (5 complete responses and three partial responses).
  • (ii) 
    IL-10. IL-10 is an immunomodulatory cytokine produced by Th2 cells, B cells, DC and monocytes that can regulate the growth and differentiation of cytotoxic T cells (CTL), B cells and mast cells. IL-10 transcripts have been identified in primary basal cell carcinoma (BCC) and SCC lesions and a role for IL-10 in immune modulation has been proposed (33). Single nucleotide gene polymorphisms (SNP) in the promoter or coding region of cytokine genes have been identified that markedly influence the level of specific cytokine production. The human IL-10 gene contains several SNP and the polymorphism at position −1082, where the -1082GG genotype is associated with high IL-10 production while –1082 AA is associated with the low IL-10 production, has been studied extensively. Position –1082 was analyzed in 140 kidney recipients including 70 that developed either BCC or SCC. The genotype associated with low IL-10 production (GG negative) was less frequent in SCC patients but not BCC patients compared to unaffected controls. The high production genotype (GG) was more frequent in both SCC and BCC patients compared to unaffected control patients (34). Interestingly, the authors demonstrated that the levels of IL-10 produced by LPS-stimulated PBMC from patients correlated with levels predicted by IL-10 gene polymorphisms and that IL-10 levels tended to be higher in PBMC from SCC patients.

With respect to IL-10 and EBV-associated PTLD, the viral homolog, vIL-10, is expressed during the lytic (productive) phase of EBV infection and may play a role in the transformation process. vIL-10 also exhibits immunosuppressive properties that can enhance establishment of viral infection. However, once B cells are transformed, vIL-10 is no longer expressed. Instead the virus persists in a latent state and the viral latent cycle gene LMP1 induces cellular IL-10. New findings indicate this is accomplished by hijacking signaling pathways including the p38 mitogen-activated protein (MAP) kinase and the PI3K/Akt/mTOR axis (35). Importantly, it has also been shown that EBV+ B-cell lymphomas produce and utilize cellular IL-10 in an autocrine growth pathway (36). The involvement of mTOR in the IL-10 pathway is consistent with findings that Rapamycin inhibits IL-10 production by EBV-infected B cells in vitro and markedly inhibits growth of EBV+ B-cell lymphomas in a xenogeneic SCID mouse model of PTLD (37). Both vIL-10 and cellular IL-10 are increased in the circulation of PTLD patients (38) and PTLD biopsy material expresses IL-10 mRNA. In addition to its role as an autocrine growth factor, IL-10 is likely to promote viral persistence and lymphomagenesis by interfering with antigen presentation and cytokine production by monocytes and by dampening the antiviral CTL response. IL-10 also prevents programed cell death of EBV+ B-cell lymphomas in SCID mice. Measurement of IL-10 in serum has been proposed as a useful means to monitor PTLD and exhibited similar utility as EBV load measurements for early diagnosis (29). In a study of 446 solid organ transplant patients, including 38 with late onset EBV-associated PTLD, there was an increased frequency in the −1082GG genotype (high producer) in controls compared to PTLD patients (39). However, IL-10 gene polymorphisms did not correlate with disease course or outcome.

  • (iii) 
    Vascular-endothelial growth factor (VEGF). Angiogenesis, the development of new blood vessels, is necessary for tumor growth and progression as the vasculature provides a mechanism to supply essential nutrients and oxygen and to remove waste products from the tumor. VEGF is a proangiogenic glycoprotein produced by most tumor cells and is regarded as a critical component in tumor invasion and metastases. However, numerous other receptor-ligand pairs are also involved in tumor angiogenesis (40). Thus, it is clear that the interplay between tumor and the surrounding stromal environment, including the vasculature, is complex and is likely to be impacted by immunosuppressive drugs. Along these lines, inflammatory cells within the tumor can participate in the switch to a proangiogenesis phenotype. For example, tumor-associated macrophages, neutrophils, NK cells and dendritic cells have been shown to produce VEGF and other proangiogenic factors such as TGF-β (41) and the production of these factors can be affected by immunosuppression. Experimental studies show that Cs promotes angiogenesis while Rapamycin seems to inhibit this process (19).
  • (iv) 
    Transforming growth factor (TGF-β). TGF-β is a complex cytokine regulating a variety of biologic functions in healthy individuals including embryonic development, cellular proliferation and differentiation, angiogenesis, and wound healing. Among its properties, TGF-β is antiproliferative, promotes production and deposition of extracellular matrix proteins, inhibits production of enzymes that degrade extracellular matrix, stimulates angiogenesis, and dampens immune surveillance by inhibiting the function of lymphocytes. A multitude of evidence indicates that dysregulation of TGF-β production or signaling promotes tumorigenesis (42). Along these lines, mutations in the TGF-β pathway can result in impaired cell cycle regulation resulting in uncontrolled proliferation while overproduction of TGF-β by tumor cells can promote angiogenesis, invasion and metastases. Suthanthiran and colleagues (43) demonstrated that CS induces marked morphologic changes that confer an invasive phenotype on a nontransformed human pulmonary adenocarcinoma cell line, A-549. It was further demonstrated that CS-stimulated secretion of TGF-β by the A-549 cells, that anti-TGF-β antibodies could prevent the CS-induced morphologic changes, and that recombinant TGF-β could recapitulate the phenotypic changes induced by CS. In vivo studies showed that CS enhanced invasion and increased metastases of murine renal cell adenocarcinoma (Renca) cells, murine Lewis lung carcinoma cells, and human bladder transitional carcinoma cells in an immunodeficient SCID-beige mouse model. Finally, anti- TGF-β antibodies could reverse the metastatic-promoting activity of CS in vivo. The induction of TGF-β by CS may be a general property of calcineurin inhibitors since more recent studies indicate that tacrolimus also induces TGF-β production and promotes tumor progression in murine models of renal cell lung metastases (44). Interestingly, the effect of the novel immunosuppressant Rapamycin appears to be the opposite of CS, that is, Rapamycin was found to suppress TGF-β and angiogenesis (19).

TGF-β can inhibit proliferation and differentiation of normal epithelial cells and appears to play a central role in maintaining epidermal homeostasis. The role of TGF-β in skin carcinogenesis has been extensively investigated in a variety of TGF-β transgenic and receptor/signaling pathway knockout animal models. The overarching paradigm from these studies suggests that the effect of TGF-β depends heavily on its cell expression pattern and at what stage of carcinogenesis TGF-β is expressed (45). During early stages of carcinogenesis, expression of TGF-β by suprabasal keratinocytes of the skin can inhibit tumor formation. However, at later stages of carcinogenesis TGF-β expression augments tumor progression by increasing the conversion of benign papillomas to malignant carcinoma.

Conclusion

  1. Top of page
  2. Abstract
  3. Background
  4. Immune Surveillance and the Impact of Immunosuppression
  5. Immune Surveillance Revisited
  6. Other Considerations Relevant to Immunosuppressive Agents
  7. Tumor Cell Growth Factors and Soluble Mediators
  8. Conclusion
  9. Acknowledgments
  10. References

The adverse side effects from the immunosuppressive drugs, such as an enhanced rate of cancer development, become more and more pronounced with the successful long-term retention of organ transplantations. The traditional notion that these side effects are an inevitable consequence of the immunosuppressive regimen per se is in need of revision. The conventional immunosuppressants, Aza and CS, exert direct adverse effects on cells targeted in carcinogenesis, e.g. lowering the DNA repair proficiency. A novel generation of immunosuppressants, most notably Rapamycin, differs in mode of action from the conventional drugs, and exerts antitumor effects while maintaining adequate immunosuppression for organ transplantation. Although the effects of these novel drugs on UV-induced skin carcinogenesis and PTLD are not fully studied, the first experimental results hold great promise for lowering substantially the risk of skin carcinomas and EBV-associated B-cell lymphomas in solid organ transplant recipients while maintaining adequate immune suppression to retain the graft.

References

  1. Top of page
  2. Abstract
  3. Background
  4. Immune Surveillance and the Impact of Immunosuppression
  5. Immune Surveillance Revisited
  6. Other Considerations Relevant to Immunosuppressive Agents
  7. Tumor Cell Growth Factors and Soluble Mediators
  8. Conclusion
  9. Acknowledgments
  10. References