A Review of Immunologic Aspects
With the continued failures of both early diagnosis and treatment options for pancreatic cancer, it is now time to comprehensively evaluate the role of the immune system on the development and progression of pancreatic cancer. It is important to develop strategies that harness the molecules and cells of the immune system to treat this disease. This review will focus primarily on the role of immune cells in the development and progression of pancreatic ductal adenocarcinoma and to evaluate what is known about the interaction of immune cells with the tumor microenvironment and their role in tumor growth and metastasis. We will conclude with a brief discussion of therapy for pancreatic cancer and the potential role for immunotherapy. We hypothesize that the role of the immune system in tumor development and progression is tissue specific. Our hope is that better understanding of this process will lead to better treatments for this devastating disease.
- pancreatic cancer
- immune response
Statistics from Altmetric.com
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.
Pancreatic cancer has a fatality rate of 95%. Early detection is rare, and most patients present with locally advanced or metastatic disease with no real hope of an effective treatment. With the current treatment strategies, the median life expectancy is 6 to 10 and 3 to 6 months for patients presenting with locally advanced disease or metastatic disease, respectively.
For the purposes of this review, we will be discussing the immunology of pancreatic ductal adenocarcinoma (PDAC), the major type of pancreatic cancer. We hypothesize that the role of the immune system in tumor development and progression is tissue specific. Our hope is that better understanding of the immunological aspects of PDAC will lead to better treatments for this devastating disease.
Basic Immunology Review
The Innate Immune Response
Several cell types of the innate immune system can recognize “danger,” that is, pathogens, tumors, and damaged tissues. These include neutrophils, macrophages, dendritic cell (DCs), mast cells, and natural killer (NK) cells. Natural killer cells can also be involved in recognizing cells infected with intracellular pathogens that downregulate major histocompatibility complex (MHC) antigens and express viral antigens or altered self-antigens. Activated cells of the innate immune system can release molecules such as cytokines (interferons [IFNs], interleukins [ILs], colony-stimulating factors [CSF]) and chemokines (CC). These molecules can lead to cell migration, local and systemic inflammation, and ultimately alert the adaptive immune system.
The Adaptive Immune Response
Although the specificity of the innate immune response is limited to toll-like receptors and several other conserved molecules that NK cells recognize, the cells of the adaptive immune response have enormous diversity and can recognize tens of millions of antigenic determinants or epitopes. A DC that has taken up an antigen matures as it leaves the site of a wound or infection and completes its journey to the regional lymph node. Dendritic cells can degrade large antigens and present peptides or lipids in human leukocyte antigen (HLA) molecules of either the class I or class II type or in CD1 molecules. In the first instance, CD8+ cytotoxic T cells (CTL or Tc) are activated, and in the second, CD4+ T helper (TH) cells are activated. Mature Tc cells can kill infected cells directly. The activated TH cells can interact with naive B cells that also recognize the corresponding specific epitope(s) on the native molecule and provide costimulation for further differentiation. Both Tc and TH cells make cytokines that interact with B cells so that they eventually produce antibodies that are specific for epitopes on that antigen. TH cells are also responsible for controlling affinity maturation and isotype switching so that the antibodies produced are highly effective at eliminating a pathogen. This antibody can neutralize, opsonize, and/or kill infected cells and pathogens.
Immune response against tumor cells can involve both innate and adaptive immune responses. An effective antitumor immune response involves recognition of tumor-associated antigens (TAAs) by the immune system and the generation of T- or B-cell responses that will kill the tumor cells but not damage life-sustaining normal tissue. Tc cells can kill the tumor cells directly. Paradoxically, both Tc and TH cells produce cytokines that can inhibit or enhance the growth of the tumor. They also help B cells differentiate into memory cells as well as plasma cells that make antibodies against the tumor. These antibodies can kill or opsonize the tumor cells, stimulate or inhibit their growth, or actually block Tc cells from killing the tumor cells. As in other tumor models, the immune system in patients with PDAC seems to have several roles. One role is to prevent tumor development by recognizing and removing abnormal cells arising from normal pancreatic cells. In this situation, the immune system exerts an “antitumor response.” But, the immune system can also provide a “protumor response,” whereby components of the immune response can stimulate the growth of tumor cells directly or indirectly by dampening the antitumor immune response.
THE ROLE OF THE IMMUNE SYSTEM—“ANTITUMOR VERSUS PROTUMOR RESPONSE”
Pancreatic ductal adenocarcinoma is an exocrine tumor that develops from the epithelial cells that line pancreatic ducts. However, it is a complex environment composed of many cell types including immune cells, pancreatic stellate cells (PSCs, fibroblasts), vascular endothelial cells, endocrine cells, and nerve cells. These cells can interact with tumor cells to disrupt the normal tissue architecture to form the dense stroma and the dynamic environment found in PDAC. Although it is currently well accepted that the immune response is determined by an invading pathogen or “danger” signal, Matzinger and Kamala1have recently provided a new perspective on how immune responses are determined. The authors suggest that the tissue rather than the pathogen itself determines the type of immune response. The idea that the tissues control the effector phase of the immune response has arisen from a better understanding of immunologic phenomena of immune-privilege sites, oral tolerance, and oral vaccination. This concept is slowly gaining support. The basic understanding of tissue-specific factors that control immune function may be critical in fully understanding the immune response not only for invading pathogens but also for tumor development and progression. Exploration of this new hypothesis may shed light on the highly variable immune responses observed when tumors arise from different organs. However, unless we truly understand the epidemiology (ie, genetic predisposition, chronic inflammation, viral infection) of a cancer, we will never fully understand the role and complexity of the immune response to that tumor. Nevertheless, we are making progress toward defining the role of the immune system in regulating the growth of malignant cells as recently reviewed by Schreiber et al.2
Lymphocytes are considered the main effector cells for the antitumor immune response. The lymphocytic cell populations are predominantly found in the stroma surrounding the tumor mass with few or no lymphocytes in the actual tumor mass.3,4This surrounding stroma has a large population of CD4+ T lymphocytes and macrophages with a small population of B lymphocytes and plasma cells.3In patients with PDAC, no correlation was found between the numbers of tumor-infiltrating lymphocytes (TILs) and the number of circulating lymphocytes. However, PDAC patients tend to have decreased numbers of circulating lymphocytes compared with healthy individuals and individuals with chronic pancreatitis.4
The role of CD4+ T cells in PDAC immunity is poorly understood, but depending on the cytokine environment, CD4+ T cells can differentiate into TH1, TH2, TH17, or Treg cells. TH1 cells produce IL-2 and IFN-γ and induce B cells to make opsonizing antibodies. TH2 cells produce IL-4, IL-5, and IL-6 and induce B cells to make neutralizing antibodies. In cancer, as a general trend, the main immune response is mediated through TH1 cells. Currently, therapy for PDAC focuses on cellular immunity and “direct tumor cell killing,” but humoral immunity could be just as important. Hence, understanding this complex balance between TH1 and TH2 responses in PDAC is crucial in developing better therapies for this disease. There are a few reports of the CD4+ T-cell responses in PDAC. Tassi et al.5compared CD4+ T-cell responses in patients with PDAC to those of healthy donors and found that the former had impaired anti–carcinoembryonic antigen (CEA) specific but not anti–viral-specific CD4+ T-cell immunity. Interestingly, in healthy donors, CEA-specific CD4+ T-cell immunity was significantly higher and produced mainly granulocyte-macrophage CSF (GM-CSF) and IFN-γ, whereas CD4+ T cells from patients with PDAC produced IL-5. However, there was no difference in the antiviral CD4+ T-cell response between the two. This study suggests that, in PDAC CD4+ T-cell immunity is skewed toward a TH2-type immune response and that this is locally mediated at the tumor site.5On the other hand, some studies support a more systemic TH2-like cytokine expression profile after CD4+ T-cell activation.6A possible explanation for this discrepancy may be best explained by the stage of disease of the patients in the studies. In the former study, the patients were at an earlier stage of disease either stage I or II, but in the latter study, most patients where at later stages of disease either stage III or IV. We assume that, in tumor progression, the immune response is initially capable of eliminating tumor cells and is therefore most likely a TH1-skewed response because this response is activated by intracellular “danger” signals (altered self-proteins produced by tumor cells) and leads to cell-mediated immunity (IFN-γ and activation of macrophages) to eliminate tumor cells. However, it is also possible that an early TH2 response occurs, an adaptive immune response more effective at removing extracellular “danger signals” that lead to IL-4 production and neutralizing antibodies. Hence, the 2 responses could be competing with or potentially enhancing each other, ultimately leading to the development of Treg cells that dampen the response as a protective measure to prevent autoimmunity. Therefore, understanding the type and function of immune cells in PDAC as well as the time line of the immune response will facilitate the development of immunotherapeutic strategies to use at different stages of disease. For example, if, at an early stage of disease in PDAC, both TH1 and TH2 responses are active, there might be a more effective antitumor response. However, at later stages of disease, if TH2 responses are more beneficial, the best strategy might be to shift the balance toward a TH2 response. The role of the immune system in the development and progression of pancreatic cancer is a powerful and dynamic tool that we must understand and apply strategically to promote antitumor responses at specific stages of disease.
In contrast to the direct anti/protumor activity of TH1 and TH2 CD4+ T cells, TH17 and Treg cells can regulate all T-cell responses. The differentiation of CD4+ T cells into either TH17 or Treg cells seems to involve a precarious balance between the transforming growth factor β (TGF-β)–driven expression of forkhead box P3 (FoxP3) expression (which drives the development of Treg cells) and the production of TGF-β/IL-6 that favor the development of TH17 cells and inhibits the development of Treg cells. TH17 cells secrete IL-17, a proinflammatory cytokine that mediates several effects on several different cell types.7,8In contrast, Treg cells inhibit the proliferation of T cells and dampen antitumor immunity.
Research on the TH17 CD4+ T-cell lineage in PDAC is limited, but recent studies have shown that, if the cytokine balance of the tumor environment is tipped in favor of the development of the TH17 cell lineage by inducing IL-6 or depleting the Treg cells, an antitumor effect is achieved.9,10The role of TH17 cells in cancer is currently under investigation. There is no definitive answer as to whether TH17 cell enhances or inhibits tumor growth. However, it has been suggested that the role of TH17 cells may change depending on the cause, type, location, and stage of the tumor.11If this were correct, it would further support the hypothesis that the role of the immune system in tumor development and progression is tissue specific and that an individual immune profile of each PDAC patient should guide therapy.
The role of Treg cells in PDAC is better understood. Both circulating Treg cells and PDAC tissue-specific Treg cells are significantly increased in patients with pancreatic cancer compared with healthy controls. The presence of Treg cells in the tumor tissue correlates with the stage and progression of disease.12,13Treg cells are known to induce tolerance against TAAs and suppress the antitumor activity of T cells.14,15In vivo studies suggest that a decrease or depletion of Treg cells in PDAC results in inhibited tumor growth and promotion of tumor-specific immune responses16,17and that the increase in Treg cells in PDAC is dependent on tumor-derived TGF-β.18,19
Pancreatic ductal adenocarcinoma has been characterized by the presence of few tumor-specific CD8+ T cells, B cells, and tumor-reactive antibody-producing plasma cells in the tumor mass. Antigens on PDAC cells can elicit either cellular or humoral immune responses and, in some instances, both. Tumor antigens that are recognized by the immune system and induce a response become tumor immunogens, and these are important for immune responses. However, not all tumor antigens are immunogenic. A few of these PDAC cell antigens are being explored as targets in clinical trials. Immunogens in PDAC include α-enolase,20coactosin-like protein,21mesothelin,22mucin 1 (MUC1),23–25mutant kirsten rat sarcoma virus oncogene,26–28cadherin 3 (CDH3)/P-cadherin,29CEA,30human epidermal growth factor receptor 2,31prostate stem cell antigen,32,33and mutant p53.34,35The peripheral blood from PDAC patients contain a high frequency of functional tumor-reactive T cells that can ultimately lead to tumor antigen–specific T-cell responses.36The bone marrow also contains tumor cell–reactive memory T cells.36Moreover, when evaluated together, the presence of both CD4+ and CD8+ T cells in malignant PDAC tissues correlates with a better prognosis than the presence of either alone.37Although this suggests that there is an antitumor response, unfortunately it is not enough. A possible explanation for the failure of the antitumor response may be provided by recent identification of the antibody-independent functions of B cells. Thus, effector and regulatory B cells may regulate T-cell immune responses by promoting the production of effector and memory CD4+ T cells as well as the proliferation and survival of Treg cells.38
Natural killer cells are a subset of cytotoxic lymphocytes that only recently received attention for their role in tumor development. Natural killer cells do not express unique antigen-specific receptors, but they play an important role in innate immunity and antitumor immunity.39They can induce target cell killing as a result of the complex integration of inhibitory and activating signals.40Natural killer cells can produce IFN-α, tumor necrosis factor α (TNF-α), GM-CSF, and IL-3.41Initially, in PDAC research, subsets of NK cells were not distinguished, but now they are divided into 2 phenotypically and functionally distinct types of cells. The majority expresses low densities of CD56 (CD56lo), secrete low levels of cytokines and exert potent effector cell cytotoxicity. In contrast, the minority group expresses high levels of CD56 (CD56hi) and IL-2 receptor α chains (CD25), secrete high levels of cytokines and are poorly cytotoxic.40Natural killer cells in PDAC have been reported to mediate tumor cell lysis,42and high levels of NK cells lead to a better prognosis.43However, even in the early stage of disease, NK cell activity is impaired and worsens with advancing disease.44,45Interestingly, CD56hi NK cells exhibited potent reactivity on several pancreatic cancer cell lines in addition to autologous tumor cells46that were identified from a pancreatic cancer patient undergoing immunotherapy with ipilimumab (a therapeutic antibody against CTLA-4, a T-cell coinhibitory molecule).46Although anti–CTLA-4 antibodies block the activity of CTLA-4 and sustained immune responses in T cells, this patient had an antitumor response with potent NK cell activity. This supports the possibility that the activation of NK cells as well as CD4/CD8+ T cells can lead to the killing of tumor cells. Several groups have evaluated the effect of modulating NK cell activity as well as CD4+ and CD8+ T-cell activity by administering IL-2 to patients, to try and promote antitumor responses.43,46–49
Polymorphonuclear leukocytes or neutrophils are often a neglected cell type in the tumor microenvironment, but a better understanding of their impact in tumor development is beginning to emerge. Neutrophils are the most abundant type of leukocyte found in the blood and are not usually found in normal tissues. In response to the production of IL-8 and C5a during acute inflammation, neutrophils can produce reactive oxygen species (ROS), serine proteases, and metalloproteases to kill invading pathogens. The activity of neutrophils is thought to follow a linear progression and, when recruited into the tumor microenvironment, can induce both protumor and antitumor responses.50Although the active states of neutrophils are not clearly defined, it has been proposed that moderate neutrophil activity in the tumor microenvironment can promote tumor growth and invasion due to the production of ROS and proteases. In contrast, robust neutrophil activity can be toxic to tumor cells and promote an antitumor response.50
Only a few studies have evaluated the potential role of neutrophils in PDAC. In 2 separate studies, it was found that an elevated neutrophil/lymphocyte ratio is a predictor of decreased patient survival.51,52In in vitro studies, it was found that activated neutrophils promote the adhesion of PDAC cells to microvascular endothelium53possibly promoting tumor migration and extravasation. Furthermore, in vivo studies have found that tumor-infiltrating neutrophils produce matrix metalloprotease type 9 (MMP-9), a potent vascular endothelial growth factor (VEGF)–independent angiogenic factor that mediates the initial angiogenic switch in PDAC.54,55Much more remains to be learned with regard to the role of neutrophils in PDAC. However, one might predict that by targeting tumor-associated neutrophils or their production of ROS and proteases, tumor invasion and growth might be inhibited.
Mast cells are typically studied in the context of type I hypersensitivity and autoimmunity. However, in a recent review by Khazaie et al.,56the role of mast cells as positive and negative regulators of the immune response in tumor development and progression is discussed. Mast cells typically surround blood vessels and nerves and are activated by inflammation, cross-linking of IgE, or complement proteins. After activation, mast cells can release several mediators including histamine, serine proteases, platelet-activating factor, and, importantly, VEGF.57In addition, mast cells can also produce cytokines typical of TH1 cells (GM-CSF, IFN-γ, and TNF-α) and TH2 cells (IL-4 and IL-13).56Thus, mast cells can play a pivotal role in both innate and adaptive immunity as well as modifying the tumor microenvironment by producing proinflammatory and angiogenic factors.
There are few studies involving the role of mast cells in PDAC. In a study by Esposito et al.,58mast cells were associated with lymph node metastasis as well as increased tumor microvessel density, suggesting that their presence promotes an angiogenic phenotype. However, this study did not find a correlation between mast cell number and patient survival.58In a subsequent study by Strouch et al.,59mast cell infiltration was significantly increased in pancreatic cancer compared with normal controls and correlated with higher-grade tumors, as well as decreased recurrence-free and disease-specific survival. In contrast to the previous study, Strouch et al.59did not find a correlation between the number of mast cells and lymph node status. The discrepancies between these 2 studies may again be explained by the grade of tumor evaluated or the heterogeneity between pancreatic tumors. Hence, in the former study, higher-grade metastatic tumors were studied, whereas in the later study, all tumors were grade 3 or less. Strouch et al.59also evaluated the in vitro mechanism by which mast cells can contribute to the poor prognosis in patients with PDAC. They found that, in the absence of direct tumor cell contact, mast cells mediated tumor cell migration, proliferation, and invasion via MMPs. These studies provide evidence that mast cells are emerging as promoters of angiogenesis and tumor progression in patients with PDAC.
Gabrilovich and Nagaraj60have recently reviewed myeloid-derived suppressor cells (MDSCs) and their role as regulatory cells in the immune system. Myeloid-derived suppressor cells consist of myeloid progenitor cells and immature myeloid cells with immunosuppressive activity in cancer and other diseases.60In cancer, MDSCs are characterized by the expression of CD33 and the lack of expression of markers for mature myeloid or lymphoid cells.61Increased numbers of MDSCs have been associated with high levels of GM-CSF62or VEGF63in the circulation. However, these MDSCs do not differentiate in a normal way.64Once activated, MDSCs can serve as immunosuppressive cells by upregulating arginase 1, and nitric oxide synthase and increasing nitric oxide production from M-MDSCs and ROS production by G-MDSCs.60,65–67Myeloid-derived suppressor cells can also inhibit the function of T cells in several ways that are not yet entirely clear. However, there are reports that they can downregulate T-cell–mediated antigen-specific responses,68downregulate TCRs/CD3-ζ chains,69and promote the development of Treg cells.70,71Other mechanisms of MDSCs immunosuppression include secretion of TGF-β,72up-regulation of cyclooxygenase 2 (COX-2) and prostaglandin E2 73as well as negatively regulate NK cells by inhibiting effector functions.72These issues are discussed in recent reviews.60,66,74
Several in vivo studies of pancreatic cancer have shown increased numbers of MDSCs in the tumor microenvironment.75–77In a study of spontaneous pancreatic carcinoma, it was shown that not only are the MDSCs increased in frequency but they also have arginase activity and suppress T-cell responses.76Moreover, by evaluating the suppressive mechanisms from tumor inception throughout tumor development, results suggest that the suppressive mechanism exists in early premalignant lesions and increases during tumor progression.76In another study of mouse pancreatic cancer, the number of MDSCs inversely correlated with CD8+ T cells infiltrates and MDSCs were present in both the primary and the metastatic lesions and not merely correlated with chronic inflammation.77
It is easy to appreciate that therapeutic strategies designed to either inhibit MDSCs and their products or possibly promote their differentiation should be considered to treat tumor development and progression.
Most cells in the immune system have both protumor and antitumor activity. Macrophages can induce T-cell recruitment and activation at the tumor site, as well as promote tumor cell growth, angiogenesis, and immunosuppression. Tumor-associated macrophages (TAMs) are derived from blood monocytes in response to tumor-derived signals such as macrophage-CSF (M-CSF), chemokine ligand 2, VEGF, and angiopoietin 2.78–83Tumor-associated macrophages are functionally divided into 2 subtypes: M1 and M2. Tumor-associated macrophages M1 subtype are activated in response to IFN-γ or microbial products and are characterized by the production of high IL-12, IL-23, toxic intermediates, and proinflammatory cytokines including TNF-α. The M2 subset is induced by IL-4, IL-10, IL-13, glucocorticoids, and immunoglobulin (Ig) complexes. They produce TGF-β and IL-10 and promote adaptive TH2 immunity, angiogenesis, tissue remodeling, and repair.83
In PDAC, the role of macrophages is beginning to be explored. Macrophages are significantly more numerous in PDAC than in normal pancreatic tissue, and their accumulation does not correlate with chronic pancreatitis-like features in the surrounding tissue.58The TAM M2 subtype has been associated with a poor prognosis.84In an in vivo mouse model when large numbers of human monocytes were coengrafted with human tumor cells, tumor growth was enhanced. However, when a low ratio of human monocytes was cografted with human tumor cells, inhibition of tumor growth was observed.85This group has shown that repeated contact of monocytes with tumor cells leads to decreased production of cytotoxic molecules (TNF-α, reactive oxygen intermediates, and IL-12) and increased production of immunosuppressive cytokine IL-10.85,86This suggests that there may be a maximum ratio of monocytes to tumor cells and a threshold of the molecules they produce that when exceeded no longer has antitumor effects. In an in vitro study by Baran et al.,87the production of TNF-α by TAMs lead to more pancreatic tumor cells as well as macrophage motility, ultimately inducing phenotypic tumor cell changes characteristic of a epithelial-to-mesenchymal transition. These studies support the hypothesis that the increase in number of TAM and their products such as TNF-α in PDAC may overcome a certain threshold and switch from an antitumor to a protumor response, but further studies are needed to better understand the significance of the number and type of TAMs that play a role in PDAC.
Tumor immunology, as applied to pancreatic cancer, is in its infancy, but several studies support the notion that immune cells are actively engaged in eliminating tumor cells and generating antitumor memory cells but that the response is either not robust enough to control tumor growth or, potentially, too robust and causes damage that triggers immunosuppression and subsequent tumor growth. In a review by Sica et al.,83the role of M1/M2 macrophages in tumor development was hypothesized. The authors hypothesized that early in the course of tumor development, macrophages with an M1 phenotype (high IL-12, IL-23, toxic intermediates, and TNF-α production) dominate, and the production of proinflammatory cytokines and toxic intermediates supports tumor formation. Once the tumor is established, macrophages with the M2 phenotype (TGF-β, IL-10 production) dominate, thereby impairing the antitumor TH1 response and promoting tumor growth. This supports the hypothesis of an early active TH1 response (production of antitumor T cells, NK cells, antibodies, and cytokines) that becomes less effective as the disease progresses. This is accompanied by increased numbers of Treg cells and TH2 cytokine production. However, the time line of the immune response in PDAC is questionable. Clark et al.77studied the immune response in pancreatic cancer from early disease to invasive cancer in a murine model of spontaneous PDAC. They found that, early on, there were few effector T cells and that most infiltrating cells were macrophages, MDSCs, and Treg cells. Their findings suggest that, from the inception of PDAC, the immune system is suppressed and is never able to mount a robust antitumor response.77,88This might represent normal immune tolerance to self-tissue because most antigens on tumor cells are not recognized as foreign. Moreover, because of the vast heterogeneity seen in PDAC, both an early active TH1 response and suppression may occur. The balance between the 2 effects could be dependent on the etiology of the disease as well as the immune system of the patient in question.
Dendritic cells are responsible for the recognition of “danger,” that is, pathogens as well as damaged tissue, activation of immunity, and preservation of tolerance to self-antigens.89They constitute the critical link between the innate and adaptive immune systems because they can traffic from damaged/invaded tissue sites to regional lymph nodes and present antigen to T cells. Once this occurs, the adaptive immune system is activated. Dendritic cells play a critical role in initiating the immune response against developing tumors. However, tumor progression and the influence of the tumor microenvironment can inhibit DC recruitment, differentiation, maturation, and survival.90In a recent review by Ma et al.,90the mechanisms by which tumor cells regulate DCs are thoroughly discussed. In brief, several factors are involved in the DC-tumor cell cross-talk including GM-CSF, VEGF, TGF-β, IL-10, and ROS. Tumor cells can produce or express various metabolites or proteins that can prevent DCs from engulfing, recruiting, differentiating, migrating, activating, and cross-presenting antigens, thereby inhibiting a tumor-specific T-cell response.90Furthermore, tumor cell death may either establish tumor-induced tolerance or enhance immune responses by exposing “cell death-associated patterns” that can ultimately induce a variety of innate immune responses.90
In PDAC, DCs are rare but when present, are located on the outside margin of the tumor.91A study evaluating the influence of circulating myeloid DCs, circulating lymphoid DCs, and DCs within the tumor on patient survival, it was found that a high percentage of circulating myeloid DCs or high numbers of DCs in the tumor prolonged survival.92Other studies found that blood myeloid DCs in PDAC were only “partially mature” and the change in their expression of surface markers led to an impairment of their immunostimulatory function.93This change was also observed in patients with chronic pancreatitis, suggesting that systemic inflammatory factors may play a role in this change.93,94In addition, it seems that preservation of mature blood DCs correlates with disease control and prolonged survival.93,94Several therapeutic strategies involve the vaccination of enhanced DC number and function in combination with other immune modulators and/or chemotherapy.
The future of immunotherapy will be dependent on elucidating the roles of immune cell subtypes and their capacity to function or dysfunction at various stages during the development of pancreatic tumors.
IMMUNE INTERACTION WITH MICROENVIRONMENT
Pancreatic ductal adenocarcinoma is a hypoxic environment with a dense stroma, which can comprise up to 90% of the tumor volume.95–97It is now understood that this dense stroma is derived from overgrowth of the extracellular matrix (ECM) and can “protect” and enhance tumor development by forming a barrier against both chemotherapeutic drugs and the immune system.98The production of fibroblast growth factor type 2 (FGF-2), epidermal growth factor (EGF) as well as the EGF receptor (EGFR),99TGF-α,100TGF-β1, insulin-like growth factor 1 (IGF-1), platelet-derived growth factor, and VEGF by tumor cells, immune cells, and other stromal cells contribute to stromal production as well as tumor cell survival and growth.99,101–105
Pancreatic stellate cells are myofibroblast-like cells and considered to be the main pancreatic cancer-associated stromal fibroblasts. These cells are recognized to be the key players in the development of desmoplasia as recently reviewed by Duner et al.106When PSCs are activated by stress, cytokines, or growth factors, they become ECM protein-producing fibroblasts. In addition to producing ECM protein, stellate cells secrete periostin,107as well as produce MMP-2/gelatinase A, MMP-9/gelatinase B,108and MMP-12.109Periostin is a protein that enhances the fibrogenic activity of PSCs while promoting endothelial cell growth and motility.107,109MMP-2/9 can break down components of the basement membrane and help promote angiogenesis, ultimately leading to local invasion and disease progression.108,110,111On the other hand, MMP-12 can induce the production of an antiangiogenic molecule, endostatin. Endostatin is a cleavage product of type XVIII collagen and can inhibit the proliferation of endothelial cells and ultimately, angiogenesis.112Pancreatic stellate cells are the main producers of VEGF in the tumor microenvironment.107In a study of tumor cell–stellate cell interactions in PDAC by Erkan et al.,109it was found that whereas tumor cells induce secretion of VEGF by PSCs, PSCs increase endostatin production of tumor cells. Although this balance between proangiogenic and antiangiogenic effects probably involves a variety of factors in addition to VEGF and endostatin, it is useful in understanding the stimulatory and inhibitory forces at play in the PDAC microenvironment (Fig. 1A). Pancreatic stellate cells provide a promising target for pancreatic cancer therapy. A recent study has shown that the specific up-regulation of hedgehog receptor smoothened (SMO) gene expression activates the sonic hedgehog signaling pathway in pancreatic cancer–associated stromal fibroblasts but not in normal pancreatic tissue.113This finding led to the development of SMO antagonists to target stromal fibroblasts in the tumor microenvironment. This is a challenging approach that relies heavily on interstitial fluid pressure for drug delivery. However, with the improved penetration of chemotherapeutic drugs into the tumor mass along with antiangiogenic agents that help stabilize local vasculature, SMO antagonists may find clinical utility for the treatment of PDAC.
Endothelial cells line all blood vessels, and without a constant blood supply, tumors cannot enlarge beyond 1 to 2 mm and cannot grow at distal sites.103Hence, angiogenesis involves the growth of these cells. The factors that can promote angiogenesis are VEGF, FGF-2, TGF-β1, and platelet-derived growth factor.102–105The most potent of which is VEGF type A, a soluble growth factor commonly known as VEGF. Soluble VEGF binds to the VEGF tyrosine kinase receptors type 1, 2, 3 (VEGFR-1, 2, 3). VEGFR-2 is restricted to endothelial cells. VEGF/VEGFR-2 complexes on endothelial cells can result in several downstream events that promote angiogenesis.103Neuropilin 1 (NRP-1) and neuropilin-2 (NRP-2) are coreceptors for VEGFRs on endothelial cells.114Recently, it was reported that NRP-1 and NRP-2 are expressed on tumor cells,115and their expression correlates with a more malignant phenotype.116,117In vivo, decreased expression of NRP-2 in PDAC slowed tumor growth. The inhibition of tumor growth was attributed to indirect effects on angiogenesis as opposed to antiproliferative effects on tumor cells,111making NRP-2 a potential therapeutic target as recently reviewed by Muders118and Dallas et al.119This study, as well as others, supports the idea that the normal balance between proangiogenic and antiangiogenic effects can be lost in PDAC. Although PDAC is not considered a “vascular” tumor, it has areas of enhanced endothelial cell proliferation with significant correlations between blood vessel density and disease progression, suggesting that antiangiogenic targets might be attractive candidates for antitumor therapy. Moreover, although antiangiogenic strategies that target VEGF alone have not yet shown efficacy for the treatment of pancreatic cancer, some antiangiogenic molecules have been shown to reduce the immunosuppression associated with cancer.120In a review by Tartour et al.,120the link between angiogenesis and immunity is discussed. In brief, antiangiogenic molecules can decrease immunosuppressive cells (MDSC, Treg cells), immunosuppressive cytokines (IL-10 and TGF-β), as well as inhibitory molecules (PD-1).120This suggests that antiangiogenic therapy may be most beneficial when used in conjunction with immunotherapy.
Endocrine cells play a role in the development and progression of pancreatic cancer has not been well established. In vitro studies have demonstrated that insulin can enhance the growth of pancreatic tumor cells.121However, in patients with PDAC, the association of hyperglycemia and hyperinsulinemia is questionable.122Although the link between endocrine disturbances such as diabetes and pancreatic cancer is still under debate,122it is hypothesized that increased proliferation and function of beta cells as a result of systemic insulin tolerance is involved in the progression of pancreatic cancer.123Patients with type 2 diabetes become unresponsive to insulin and the pancreas compensates by producing more insulin. Insulin can act as a tumor growth factor when tumor cells overexpress both insulin receptor substrates, 1 and 2, as is the case for pancreatic cancer.124,125Insulin can also act as a growth factor when circulating IGF-1 is low,126and the IGF-1 receptor is available to bind to insulin. Moreover, IGF-1R activation also leads to the activation of the mammalian target of rapamycin (mTOR)–mediated phosphatidylinositol 3-kinase (PI3K)/Akt pathway, providing antiapoptotic signals to the cell.127Thus, the inhibitions of IGF-1/IGF-1R activity as well as mTOR are potential therapeutic targets for pancreatic cancer.127In addition, under hypoxic conditions, insulin can also stimulate the expression of hypoxia-inducible factor 1α (HIF-1α) in pancreatic cancer cells,128,129thereby promoting angiogenesis and further tumor development. Moreover, hyperglycemia can stimulate the expression of VEGF by human vascular smooth muscle cells.130This demonstrates a potential role for hyperglycemia in endothelial cell dysfunction.130Patients with type 2 diabetes can have impaired immune cell functions, particularly regarding neutrophils and cytokines, leading to an immunosuppressive state. Thus, endocrine cell dysfunction and its relationship to the development and progression of PDAC may be more closely related than is currently appreciated (Fig. 1B).
IMMUNE FAILURE AND TUMOR ESCAPE IN PDAC
Tumor cell escape involves a complex network of dynamic interactions among cells of the tumor, the immune system, and the stroma. Although we assume that the immune response in PDAC has the potential to eliminate tumor cells, during cancerogenesis, it is possible that the immune system is activated but fails to eliminate the tumor. For example:
The tumor antigens are not recognized as foreign or dangerous.
The activation of the response is either not rapid or robust enough to eradicate the tumor cells in early stages. In late stages, tumors grow too rapidly to be controlled by Tc cells.
Antigens, to which an immune response has been generated, signal cytokine overproduction that alters the immune response or actually suppresses/kills cells of the immune system.
The immune system can also provide a “protumor response,” whereby components of the immune response can stimulate the growth of tumor cells themselves or dampen (alter) the antitumor immune response.
The products of the immune system such as antibodies, activated T cells, and cytokines might also have many collateral effects on normal tissues causing immune dysregulation as well as tissue damage.
Clearly, there are a variety of other mechanisms that might also be involved in the ability of tumor cells to evade or circumvent the immune system either during the activation phase or during the effector phase of the immune response. This is not a new concept, and it has been extensively reviewed.131–133However, the likelihood for several more mechanisms that have yet to be elucidated is high. A general view of tumor escape in pancreatic cancer is depicted in Figure 2.
Pancreatic tumor cells can “escape” immune surveillance by several mechanisms. They can avoid apoptosis, immune detection, and the effector phase of the immune system and kill tumor-specific Tc cells. Moreover, pancreatic tumor cells can migrate to other tissues and promote immune suppression and dysregulation. This raises the question of whether “tumor escape” is a true failure of the immune system to recognize an altered normal cell and mount an antitumor response or whether the tumor cells silence and/or attack the immune system. Indeed, both might occur.
Tumor Cells Avoid Undergoing Apoptosis and Induce Apoptosis in Other Cells
Apoptosis is natural programmed nonnecrotic cell death. It plays a crucial role in maintaining homeostasis as well as immune-mediated cell killing. Pancreatic ductal adenocarcinoma cells have developed several mechanisms to avoid undergoing apoptosis and or induce apoptosis in immune cells (Tc cells) and surrounding normal epithelial cells. Either mechanism could promote tumor progression. Samm et al.134extensively reviewed the role of apoptosis in the pathology of pancreatic cancer demonstrating a correlation between disease occurrence with failures in apoptotic mechanisms. This section will review the key apoptotic evasion strategies used by PDAC cells.
Apoptosis can normally occur through “extrinsic” or “intrinsic” pathways. Death receptors (DRs), Fas (CD95, Apo-1), TNF-related apoptosis-inducing ligand receptor (TRAILR, Apo-2) and TNF receptor and their corresponding ligands FasL, TRAIL, and TNF-α mediate the extrinsic pathway. Proapoptotic and antiapoptotic molecules of the mitochondria mediate the intrinsic pathway. Molecules involved in apoptotic pathways are ideal targets for killing tumor cells particularly if they are overexpressed. However, in PDAC, tumor cells have developed mechanisms to downregulate apoptotic receptors and/or upregulate the apoptosis-inducing ligands as well as mutate regulatory apoptotic pathways.
The FAS system is composed of Fas ligand (FasL) that, when bound to Fas receptor (CD95), can induce apoptosis in cells that express functional Fas receptor.135Tumor escape can be achieved through down-regulation or loss of Fas, dysfunctional Fas signal transduction, or expression of functional FasL.136,137In PDAC, Fas is expressed on most established cell lines. However, most are resistant to Fas-ligand–mediated apoptosis.137–139This resistance can be attributed to Fas-associated phosphatase-1, an inhibitor of Fas signal transduction, that is overexpressed in Fas-resistant pancreatic cancer cell lines.137,140However, several pancreatic cancer cell lines and surgical specimens express functional FasL, allowing these tumor cells to induce apoptosis in activated Tc cells137,141,142as well as other FasR-expressing cell types.
TRAIL can induce apoptosis in susceptible cells through interaction with membrane receptors DR4 and DR5143and decoy receptors DcR1 and DcR2.144The TRAIL death receptor pathway is regulated by inhibitory proteins such as bcl-2–related proteins, bcl-2, bcl-xL, and fas-like IL-1 converting enzyme (FLICE)-like inhibitor protein and stimulated by Bax.145Pancreatic cancer cell lines are heterogeneous in their expression of TRAIL, its receptors DR4, DR5, DcR1, and DcR2, as well as regulatory proteins Bax, bcl-2, and bcl-xL.146Therefore, some pancreatic cancer cell lines are susceptible to TRAIL-mediated apoptosis, whereas others are completely resistant.146The cell lines that express TRAIL can induce apoptosis in TILs as well as other TRAIL-receptor expressing cell types. This provides TRAIL-expressing tumor cells with a way to escape the immune system. It also confers a growth advantage for an aggressive tumor cell by eliminating the less aggressive clones.146,147In addition, this promotes metastasis through the apoptosis of surrounding normal cells.148Initially, TRAIL-based therapy was postulated to be a good treatment option for PDAC considering that a high TRAIL expression correlated with an increased apoptotic index.149,150Unfortunately, primary human tumors are often resistant to TRAIL-induced apoptosis despite the expression of TRAIL receptors, DR4 and DR5, as well as the mediators for the pathway.151–155Several proteins that can promote resistance to TRAIL-mediated killing in PDAC have been identified, and these include histone deacetylase 2 (HDAC2),154STAT3,156CUX1,157cFLIP,158–160XIAP,161–165MCL1,156,166bcl-xL156,167,168or survivin,169and SKP2.155Because of the heterogeneity in tumors from different patients, immunotherapy targeted to TRAIL could be beneficial in some, but not all, patients with pancreatic cancer.
Receptor-binding cancer antigen 1 (RCAS1) was identified from cancer cells and can induce apoptosis in immune cells that express RCAS-1 receptor (RCAS1R).170Although evidence for the role of RCAS1 in PDAC is limited, a few studies have found trends in tumor cell positivity and up-regulation of RCAS1, correlating to histopathologic grade and poor patient prognosis, respectively.171–173This suggests that up-regulation of RCAS1 may play an important role in PDAC progression by evading the immune system. However, further investigation is warranted. In addition, increased serum levels of RCAS1 correlated with tumor stage and when compared to CA-19-9, RCAS1 showed greater fidelity as a diagnostic marker for pancreatic cancer. However, when used together, diagnostic efficiency was enhanced.172
Both proapoptotic and antiapoptotic molecules of the mitochondria mediate the intrinsic pathway of apoptosis. In PDAC, alterations of the Bcl-2 protein family regulated intramitochondrial signal transduction pathway have been reported.134Tumor cells promote their own survival, progression, and metastasis by manipulating both “intrinsic” and “extrinsic” apoptotic pathways.
Tumor Cells Avoid Immune Detection and the Effector Phase of the Immune System
Normal cells undergo many alterations in the progression to adult cancer cells. These alterations can be advantageous to the tumor cell and lead to down-regulation or up-regulation of various genes and their corresponding expression of molecules such as HLA, TAAs, or complement regulatory proteins (CRPs). In addition, alterations can lead to expression of abnormal genes and proteins that can provide specific targets for therapy.
Dysregulation of HLA
For the immune system to initiate a response against a tumor, DCs must transport tumor antigens to the regional lymph nodes. Tumor antigens are processed with the help of transporter for antigen presentation (TAP) and presented to T cells by HLA class I and class II molecules on the surface of DCs. For the tumor cells to be killed by the resulting Tc cells, they must express the specific tumor antigens in class I molecules. In PDAC, tumor cells downregulate or lose expression of HLA class I, its associated β2-microglobulin174and TAP.175Therefore, some tumor cells no longer present antigen to immune cells and avoid immune detection as well as killing by Tc cells. Although this does render them sensitive to NK cell-mediated killing, NK cells are far less effective in eliminating tumor cells.
Tumor cells can express HLA class II molecules de novo.175This suggests that tumor cells are promoting a TH2/humoral immune response, by influencing the type of HLA molecule expressed on their cell surface. Ultimately, this can have detrimental effects on tumor cell killing by preventing cellular immunity, that is, Tc cells, and promoting humoral immunity, that is, TH cells. Interestingly, however, it has been reported that HLA class I and TAP expression can be reinduced in PDAC cell lines in vitro by treatment with IFN-γ,175thus providing a possible means of altering the balance between cellular and humoral immunity by promoting TH1/cell–mediated immunity.
HLA-related molecules, MHC class I chain–related genes A and B (MICA/MICB), are intestinal surface glycoproteins that can be upregulated in response to stress or by epithelial tumors.176MICA/MICB are ligands for the NKG2D-activating receptor found on NK and γ/δ T cells of the immune system. Natural killer cells can recognize cells that either downregulate MHC antigens or completely lose HLA class I molecules. In PDAC, cells express MIC and sera from PDAC patients contain elevated levels of soluble MIC that correlate with tumor stage and differentiation.177Moreover, soluble MIC in the sera of patients with PDAC can inhibit the cellular cytotoxicity of NK and γ/δ T cells, thereby inhibiting the ability of the innate immune response to eliminate PDAC cells.177
Down-Regulation of TAAs
Tumor cells can upregulate normally expressed molecules as well as express abnormal self-molecules. These phenomena are some of the main factors driving the development of targeted immunotherapy. Unfortunately, after tumor cells “escape” from immune surveillance, they often become “resistant” to TAA-specific induced immune effector cells. In addition, as the more aggressive tumor cells differentiate, the expression of TAAs can mutate or decrease to a point of complete loss of expression from the surface of the remaining tumor cells.
Expression of CRPs
Complement inhibitors CD46 (membrane cofactor protein), CD55 (complement decay accelerating factor), and CD59 (protectin) can protect tumor cells from lysis by activated complement.178–180Pancreatic ductal adenocarcinoma cell lines express high levels of these molecules on their surface (Pop, Vitetta et al., unpublished data). This suggests that tumor cells are able to regulate complement-dependent effector functions. Hence, antitumor antibodies made by the host or administered therapeutically (eg, anti–CA-19-9) would fail to kill the tumor cells by C′-mediated lysis.181
Tumor Cells Promote Immune Suppression and Immune Dysfunction
Interestingly, in tumor progression, tumor cells can also aberrantly express T-cell costimulatory molecules. These molecules are typically limited to cells of the immune system and are involved in lymphocyte signaling pathways. The overexpression of these molecules by tumor cells can lead to either amplification or dampening of local immunity with devastating consequences for normal body physiology. For example, tumor cells that express inhibitory costimulatory molecules can suppress or eliminate specific antitumor immunocytes, thereby allowing the tumor to progress. On the other hand, tumor cells that express activating costimulatory molecules can enhance the immune response such that the inflammatory milieu causes damage to the surrounding normal tissue and results in further progression of tumor growth.
The B7 superfamily is composed of costimulatory molecules expressed on antigen-presenting cells (APCs) that included B7-1 (CD80) and B7-2 (CD86) and their receptors CD28 and cytotoxic T-lymphocyte antigen 4 (CTLA-4), as well as the B7-homolog molecules B7-H1, B7-DC, B7-H2, B7-H3, and B7-H4. In the immune response, T cells and antigen-MHC complexes determine specificity whereas the costimulatory molecules of the B7 family determine the magnitude and type of immune response. Therefore, the B7 ligands can provide an activating or inhibitory signal depending on the receptor bound and the influence of the local environment.182
The B7/CD28 and B7/CTLA-4 systems are T-cell costimulatory pathways that act on APCs and T cells. The B7/CD28 interaction promotes T-cell and B-cell activation, TH1/TH2 differentiation, cell migration, and homeostasis of CD25+CD4+ Treg cells.183In contrast, B7/CTLA-4 interactions downregulate T-cell function and ongoing immune responses as well as help maintain peripheral tolerance.183By blocking costimulatory pathways, specific clones of activated T cells are turned off.183This pathway has been explored in developing strategies for immune intervention therapies (discussed in “The Role of the Immune System—“Antitumor Versus Protumor Responses”” section).
B7-H1 (programmed death-1 ligand, PD-L1; CD274) and B7-DC (PD-L2; CD273) are cell surface ligands for programmed death 1 (PD-1) receptor that is expressed on activated T cells, B cells, and monocytes.184–188The expression of B7-H1 is induced by IFN-γ on several cells types.189,190The expression of B7-DC is limited to DCs and activated macrophages and induced by IL-4 and IL-13.191Both ligands can induce PD-1 to negatively regulate both cellular and humoral immune responses.184,188,189,192However, the interaction of B7-DC with PD-1193also has stimulatory effects suggesting another receptor interaction or influence of other environmental factors.189,194B7-H1 inhibits antitumoral T-cell immunity by interacting with PD-1 on T cells resulting in tumor-specific T cell apoptosis or impaired cytotoxicity and cytokine production by activated T cells.195–198In addition, ligation of B7-H1 to T-cells can result in the preferential production of IL-10.199,200Interleukin 10 is an immunosuppressive cytokine that can inhibit TH1-type immune responses201,202by modulating APCs and DC function and promoting Treg cell responses.203,204In PDAC, the expression of both B7-H1 and IL-10 is upregulated compared with normal tissues. The expression of B7-H1 correlates with advanced tumor stage and poor prognosis and is inversely correlated with TILs.188,205This suggests that tumor cells are expressing B7-H1 to suppress the antitumor immune responses while promoting the production of the immunosuppressive cytokine IL-10. Furthermore, in an in vivo mouse model of pancreatic cancer, blocking the B7-H1, B7-DC,206or the PD-Ls/PD-1 pathways188with monoclonal antibodies can induce antitumor effects and promote infiltration of T cells into the tumor.188,206This is important because it identifies specific molecules in pathways that can be therapeutically targeted to restore the antitumor immune response. Moreover, the B7-DC blockade decreases IL-10 and FoxP3 levels, whereas B7-H1 blockade increases IFN-γ and FoxP3 in the tumor site.206Thus, further demonstrating the important role of each ligand not only as specific Tc cell inhibitors but also as general immunosuppressors involving Treg cells. As we better understand the mechanisms by which tumor cells inhibit antitumor response, we can design more effective therapies.
Although no studies to date report the expression of B7-H2 (CD275) on pancreatic cancer cells, it would not be surprising if it were expressed.
The role of B7-H3 (CD276) in antitumor immunity was recently reviewed by Loos et al.207and will be briefly discussed in this section. B7-H3 is expressed on several nonimmune cell types throughout the body. However, its expression can be induced on activated DCs, monocytes, T cells, and some tumor cell lines.208–211The role of B7-H3 in immune regulation is controversial because both stimulatory and inhibitory immune functions have been reported and possibly attributed to 2 distinct receptors.207–216However, only 1 receptor has been identified, it is the triggering receptor expressed on myeloid cells–like transcription 2 (TLT-2).217The interaction of B7-H3 with TLT-2 on T cells enhances T-cell activation, proliferation, cytokine production, and cytotoxicity.217In PDAC, B7-H3 tumor-related expression has been reported to be significantly higher than in noncancer tissue or normal pancreas.218Its expression correlated with lymph node metastasis and advanced pathologic stage.218In an in vivo mouse model of pancreatic cancer, B7-H3 blockade promoted CD8+ T-cell infiltration into the tumor and induced substantial antitumor effects that were synergistic with gemcitabine.218
B7-H4 is expressed predominantly on human epithelial cells of the female genital tract, kidney, lung, and pancreas with low/no expression on other cell types.219,220Although the receptor for B7-H4 is unknown, the expression of B7-H4 can be induced on monocytes, macrophages, and myeloid DCs by both IL-6 and IL-10 and downregulated by GM-CSF and IL-4.221–223Much remains to be elucidated concerning the role of B7-H4 in immune regulation. However, studies have shown that B7-H4 inhibits the proliferation of both CD4+ and CD8+ T cells as well as cytokine production by inducing cell cycle arrest.224–226B7-H4 is highly expressed on several human cancers,219,220and although data are limited, 1 study has investigated the expression of B7-H4 in PDAC. B7-H4 was expressed more often than p53, a potential marker for pancreatic cancer, and B7-H4–positive tumor cells were inversely correlated to tumor grade.227These findings suggest an early induction followed by loss of B7-H4 expression leading to a decrease in tumor-associated immunogenicity in higher-grade tumors.227The role of B7-H4 in normal pancreatic tissue has not been investigated. However, B7-H4 interactions in normal pancreas may block T-cell–mediated immunity whereas in PDAC, this protection may be lost because of B7-H4 expression.227
CD40 is a membrane glycoprotein member of the TNF receptor family. It is expressed on several cell types including B lymphocytes, DCs, and monocytes.228–230Its ligand, CD154 (CD40L), is expressed on the surface of T cells. The interaction of CD40+ B cells with CD154+ T cells induces B-cell proliferation, immunoglobulin production, somatic hypermutation of B-cell receptors, and immunoglobulin class switching.231–234The interaction of CD40+ DCs, and CD154+ T cells leads to up-regulation of costimulatory molecules (CD80, CD86) on APC cells to help T-cell activation, proliferation, and cytokine expression.235The normal expression and interaction of CD40 and CD154 by immune cells result in the proliferation of the immune response with the potential to ultimately affect antitumor immunity. In a recent study by Shoji et al.,236it was found that both CD40 and CD154 are expressed by PDAC cell lines and patient specimens, and although the study did not directly evaluate TILs, they found the frequency of CD154 expression on TILs to be low in their xenograft model. These findings suggest that PDAC cells can potentially use CD40 and CD154 expression as an autocrine mechanism to promote tumor cell proliferation as well as potentially alter CD154 expression on TILs. This alteration may be explained by the ligation of CD40 on PDAC cells inducing the secretion of several types of proinflammatory and anti-inflammatory cytokines (IL-6, IL-10, IL-12) as found by Shoji et al.236Moreover, studies on other malignant cell types support the secretion of cytokines after CD40 ligation.237–239The balance of protumor versus antitumor immune responses can tip in favor of either response or remain in equilibrium based on the expression of surface molecules by tumor cells, immune cells, and stromal cells, as well as the factors that they release. A good example of this was the finding that very high expression of CD154 in patient specimens correlated with a favorable prognosis.236On the one hand, PDAC cells promote their own growth with the expression of CD40-CD154 and immune cell suppression with secretion of IL-10. The ligation of CD40 on these tumor cells leads to the secretion of proinflammatory cytokines IL-6 and IL-12 that could ultimately result in an antitumor response. These seemingly contradictory findings within the same study best illustrate the complexity of the tumor-immune system interactions. Moreover, when pancreatic cancer patients were treated with an agonist CD40 antibody and gemcitabine chemotherapy, tumor regression was observed.240When Beatty et al.240evaluated this effect in a mouse model of PDAC, they found that CD40-activated macrophages but not T cells nor gemcitabine infiltrated tumors and mediated tumor regression and depletion of tumor stromal cells.
CD54 (intercellular adhesion molecule 1 [ICAM-1]) exists in both membrane-bound and soluble forms. The adhesion molecule CD54 is expressed on several different cell types241–243and can be secreted as soluble CD54 (sCD54) by mononuclear cells, endothelial cells, keratinocytes, hepatocytes, and some tumor cells.244The regulation of sCD54 is not well understood, but TNF-α, IFN-γ, and IL-1 can induce the expression of membrane-bound CD54, whereas glucocorticoids are the most well-known inhibitors.241–243CD54 binds to the β2 integrins lymphocyte function-associated antigen (CD11a/CD18) and macrophage 1 (CD11b/CD18) on leukocytes, as well as sialophorin (CD43) on leukocytes and platelets and soluble fibrinogen.242–246It functions predominantly as an adhesion molecule, but it can elicit a variety of effects including T-cell and NK cell activation and leukocyte migration.242–246CD54 is associated with disease states characterized by local or systemic inflammation,247,248and although CD54 is not tumor specific and is expressed on many normal cells in humans, it can play a crucial role in the tumor microenvironment. It has been hypothesized that CD54 dictates the metastatic potential and lethality of many types of cancer cells.249The overexpression of CD54 at the leading edge of tumor invasion has been correlated with a poor patient prognosis.250Although no published studies have evaluated the role of CD54 in pancreatic cancer, our unpublished data derived from pancreatic cell lines suggest that its expression is downregulated or lost on some PDAC cell lines.
CD70 (TNFSF7) ligand is a member of the TNF superfamily that interacts with CD27. The interaction of CD70 ligand with CD27 regulates long-term maintenance of T-cell immunity as well as B-cell activation and immunoglobulin synthesis.251–260CD70 expression is normally limited to antigen-activated T and B lymphocytes254,261,262and is found infrequently in a few other normal cell types.263,264However, aberrant expression of CD70 has been reported in several tumor types including pancreatic cancer cells.265–269Because CD70 expression has a limited normal distribution and aberrant cell surface expression in tumors, CD70 makes an attractive target for therapy.269In an in vivo model of human pancreatic cancer, mice were treated with an anti-CD70 drug conjugate (SGN-75) all 7 mice treated showed a delay in tumor growth, with 2 of 7 mice showing a complete and sustained regression.269
Loss of CD3-ζ Chain Expression
The T-cell receptor (TCR)/CD3-ζ chain is a crucial component in the T-cell signal transduction complex. Although it is important for the initial activation of Tc cells in the regional lymph node, and not in the effector function of Tc cells at the tumor site, it is important to note that specimens from PDAC patients have shown significant down-regulation or loss of TCRs/CD3-ζ chains on TILs.137The significance of this finding has yet to be determined, but it is proposed that environmental factors such as ROS and arginase produced by macrophages in tumor sites can decrease the expression of TCRs on effector T cells such that they can no longer recognize the tumor antigens expressed in HLA class I molecules. Thus, the effector T cells in the tumor microenvironment might not recognize their target cells and, hence, not kill them.
Production and Secretion of Immunosuppressive Factors
A more global mechanism of PDAC tumor escape is the production and secretion of immunosuppressive molecules. In addition to several preciously mentioned immunosuppressive molecules, PDAC cells can produce and secrete TGF-β, MUC1, MUC5AC, indoleamine 2,3-dioxygenase (IDO), galectin 1 (Gal-1), and ROS.
The role of TGF-β 270in blocking the activation of lymphocytes and monocytes has been covered in other sections of this review. However, its role in tumor cells will be addressed here. In PDAC, TGF-β has been shown to upregulate proteases such as MMP-2 and urokinase plasminogen activator,271to downregulate cell surface CD54,272and to stimulate the secretion of VEGF by tumor cells.271These effects can result in degradation of the ECM, thereby promoting tumor cell invasion and metastasis while providing an angiogenic stimulus to promote further development. Moreover, TGF-β2 has been shown to induce PDAC cells to express functional Foxp3, possibly allowing tumor cells to mimic the function of Treg cells.273This suggests another mechanism by which tumor cells can suppress antitumor responses, by functioning as suppressor cells themselves.
MUC1 is an epithelial cell membrane-bound glycoprotein that is approximately 80% carbohydrate.274,275It is associated with the progression of normal pancreatic ductal cells to infiltrating ductal carcinoma and has been shown to enhance the invasiveness of pancreatic cancer cells by inducing an epithelial to mesenchymal transition.276–282MUC1 is also an immunogen that elicits CD8+ T-cell responses23and induces the production of anti-MUC1 antibodies of both the IgM and IgG isotypes.24,25Increased serum levels of anti-MUC1 antibodies correlate with increased patient survival.25Tumor-derived mucin has been shown to profoundly affect the cytokine repertoire of monocyte-derived DCs, producing regulatory APCs (IL-10highIL-12low) that lead to a TH1 immune response.283Again, this supports the hypothesis that the tumor cells do elicit an antitumor immune response. However, it is either not robust enough or is quickly thwarted by other escape mechanisms used by tumor cells. To better elucidate the immunosuppressive effects of MUC1, Tinder et al.284compared pancreatic tumors that expressed MUC1 to those that lacked MUC1 expression in a mouse model of spontaneous PDAC. The tumors derived from MUC1+ mice expressed higher levels of COX-2 and IDO compared with tumors from MUC1− mice especially during early stages of development. In addition, MUC1+ mice had an increased proinflammatory milieu with elevated levels of Treg cells and myeloid suppressor cells within the tumor and draining lymph nodes.284In subsequent in vivo studies, Besmer et al.285showed that MUC1− mice have significantly slower tumor progression and rates of metastasis. Moreover, from their in vitro studies, it is suggested that MUC1 is necessary for MAPK activity and oncogenic signaling.285MUC1-mediated mechanisms can enhance the onset and progression of the disease, which, in turn, regulate the immune response.284It is possible that early in disease MUC1 expression is recognized by the immune system and initially promotes a robust TH1 antitumor immune response. However, over time, the progressive inflammation can evoke an immunosuppressive response established by either the efforts of the immune system to maintain balance or the attempts by the tumor cells to suppress the response.
MUC5AC, another glycoprotein from the mucin family, is overexpressed only by PDAC cells and not by normal pancreatic cells. A recent study revealed that, by knocking down MUC5AC expression of wild-type MUC5AC-positive pancreatic cell lines by small interfering RNA, there was a decrease in tumorigenicity and tumor development.286This suggests that MUC5AC expression may play an important role in the development of PDAC as well as provide a potential tumor specific target for therapy.
The up-regulation of enzymes or proteins crucial for immune cell function can be an important mechanism by which tumor cells control the tumor environment and prevent tumor specific immune responses. Indoleamine 2,3-dioxygenase is an IFN-γ–induced immune regulatory enzyme that catabolizes tryptophan.287Indoleamine 2,3-dioxygenase can create an immunosuppressive environment by deleting tryptophan, a crucial metabolite for T cells undergoing antigen-dependent activation, ultimately leading to T-cell arrest, anergy, or death.288–291In PDAC, up-regulation of IDO in tumor cells is associated with more Treg cells.289Recently, another immunoregulatory molecule secreted by pancreatic tumor cells and activated PSCs, Gal-1 was identified.292Gal-1 is a carbohydrate-binding protein that is thought to be a regulator of T-cell homeostasis, survival, and inflammation.293,294Up-regulation of Gal-1 was suggested to induce apoptosis in T cells, activate DCs, regulate immune cell trafficking, and promote proliferation and invasion of tumor cells.294,295Thus, PDAC cells can use basic nutrient metabolism and secretion of immunosuppressive factors to locally regulate the immune response.
Although ROS is mainly implicated to promote cell death and is a key component of immune defense against invading microbes,296–299there is increasing evidence to suggest that it also plays a role in cell survival and signaling.300–304We previously reviewed ROS as products of neutrophils as well as macrophages. However, PDAC cells have been reported to produce ROS as well.305In a study by Vaquero et al.,305when PDAC cells were stimulated by IGF-I or FGF-2, they produced ROS, which protected the cells from apoptosis.
It has been proposed that tumor cells require the presence of chemotactic molecules for growth, invasion, and metastasis. Ultimately, these “homing factors” may promote immune escape.
A strong expression of the chemokine receptor CXCR4 has been reported to correlate with advanced pancreatic cancer.306There is much more that remains to be uncovered about the role of CXCR4 and other chemotactic molecules in PDAC.
Unfortunately, tumor escape from normal immunity is not the main or only mechanism by which tumor cells survive. Genetic alterations and microenvironment factors that both promote tumor cell development and eliminate less robust tumor cells ultimately produce immortal cells with an infinite capacity for reproduction, if nutrients are available. In this regard, the precursors of cancer stem cells can evolve into new blood vessel progenitors,307and adult cancer cells, especially in hypoxic states, can induce and sustain new blood vessel formation that ensures nutrient supply for further tumor growth. Moreover, one of the most potent angiogenic factors, VEGF, is also an immunosuppressive molecule secreted by tumor cells.308Thus, as part of the extraordinary innate and acquired abilities of tumor cells to defeat the host, they develop a variety of “escape” mechanisms that can overcome almost any antitumor approach. However, the lessons learned from the various ways tumor cells continually adapt to ensure their survival should be used to develop rational multitargeted immunotherapies.
A ROLE FOR IMMUNOTHERAPY IN PDAC
The dismal prognosis of PDAC is due to late detection and limited treatment options. It is often diagnosed at an advanced stage of disease when the tumor is inoperable and frequently resistant to standard therapy. Clinical trials in patients with PDAC have focused on improving both surgery and radiation therapy as well as determining better drug-treatment combinations. Despite this, PDAC is almost uniformly fatal.
Several biological approaches have been studied for the treatment of pancreatic cancer, for example, gene therapy, signal transduction modulators, antiangiogenics, MMP inhibitors, oncolytic viral therapy, as well as immunotherapy. However, these have not improved patient survival as recently reviewed by Wong et al.309It is important to note that clinical trials with targeted biologics are used in all patients with disease regardless of whether they express the target, making conclusions difficult and potentially erroneous. Several proposals have been made to address this issue by selecting patients with the appropriate targets on their tumors or in the tumor microenvironment.310Therefore, there is a need to establish the genetic and proteomic profile of the tumor cells in each patient as well as understand the key molecules involved in multidrug resistance and use this information to successfully target the most effective agents to cells. This profile can be used to determine the treatment regimen as well as monitor responses to treatment. Ultimately, the hope is that, by using this profile to screen study patients for the expression of the target in question in clinical trials with targeted therapeutics, we may actually begin to see clinically meaningful responses.
The rationale for immunotherapy is to augment a patient’s natural immune response to their pancreatic cancer or introduce components of an immune system to slow disease progression. The consequences of impaired immune function are significant. Tumor cells are capable of functioning as immunocytes with the ability to secrete immunosuppressive cytokines, ultimately impairing the immune system’s function to recognize TAAs and destroy tumor cells.
Currently, immunotherapy for PDAC is only available in clinical trials. There are several ongoing clinical trials to evaluate single-agent as well as combined cytotoxic therapy and combinations of targeted therapies (including monoclonal antibodies).311However, considering how refractory to conventional agents this disease is, clinical trials may offer the best treatment option as well as teach us how to define specific combinatorial therapy guidelines (eg, dosing, timing, route of administration, adjuvant therapy). Unfortunately, the benefits of immunotherapy have yet to meet initial expectations. Clinical studies have yielded undesirable results such as the stimulation of incorrect immune responses, cytokine storm, tumor progression, and metastasis. From a mechanistic point of view, the failure of immunotherapy in treating pancreatic cancer is a consequence of the genomic instability inherent in cancer cells, which allow them to highjack immune defenses. Moreover, the potential existence of cancer stems cells may help explain tumor rescue and epithelial-to-mesenchymal transition, a phenomenon believed to play a role in tumor progression and ineffectiveness of the current therapy. Pancreatic ductal adenocarcinoma research should accelerate the understanding of the rescue mechanisms (tumorigenicity, invasiveness and resistance to therapy, angiogenesis, etc) that tumors use through the self-renewal “stem cell” subpopulations. Understanding these cell types can provide a new avenue for cancer-targeted therapeutics, as well as identify the patients with a high risk of unfavorable disease evolution or recurrence.
The future of immunotherapy stands on either identification of unique specific TAAs using proteomic analysis292or finding avenues to “teach” the tumor cell to express a stable specific antigen that can be further targeted in an immunotherapy setting. The best-case scenario will be a simultaneous immunostrategy, which will result in overcoming the tumor tolerance, stimulating the specific targeted antitumor response, and shutting off the suppressor arm of immunity at the same time. This scenario relies on the development of novel agents to target specific biological processes unique not only to the tumor cells but also to their microenvironment. Novel agents include growth factor inhibitors, antiangiogenic factors, MMP inhibitors, as well as intracellular mediator or pathway inhibitors.311,312Moreover, to enhance the efficacy of pancreatic cancer immunotherapy, we should also establish methods for earlier diagnosis with new and very specific molecular biomarkers.292This should facilitate the design of “early protective” measures such as vaccination (preventive immunotherapy).
The lessons learned from tumor escape mechanisms and failed clinical trials should represent the platform for the development of rational and successful combination therapy (eg, cytotoxic and targeted/immunotherapy along with surgery) for patients with PDAC. However, one of the major challenges of combining immunotherapy with conventional chemotherapy is timing. Administering chemotherapy before immunotherapy eliminates the immunocytes that are activated with the latter. However, the two may work synergistically with one serving as the necessary adjuvant for a robust antitumor immune response. For example, an antitumor response may be facilitated by the up-regulation of Fas, TAAs, and cell chemosensitization313or by cross-presentation of TAAs by DCs after tumor cytolysis. Ultimately, to improve old therapies and develop new ones, we must understand why our past attempts at treating PDAC have failed and apply that knowledge to our future approaches.
The authors thank Dr. Rolf Brekken, for his intellectual contribution in the discussion of pancreatic cancer, Dr. Michael McPhaul, for his editorial advice, and Dr. Helen Mayo, Reference and Liaison Librarian at The University of Texas Southwestern Medical Center Dallas, for her assistance with finding references for this article.