Emerging Strategies for Cancer Immunoprevention

Abstract

The crucial role of the immune system in the formation and progression of tumors has been widely accepted. On one hand, the surveillance role of the immune system plays an important role in endogenous tumor prevention, but on the other hand, in some special circumstances such as in chronic inflammation, the immune system can actually contribute to the formation and progression of tumors. In recent years, there has been an explosion of novel targeted immunotherapies for advanced cancers. In the present manuscript, we explore known and potential various types of cancer prevention strategies and focus on non-vaccine based- cancer preventive strategies targeting the immune system at the early stages of tumorigenesis.

INTRODUCTION

Since cancer is a collection of multiple complex individual diseases that happen in different sites of the body and progress in different ways, a large amount of information is required to understand how to prevent each cancer type. Existing interventions for cancer prevention focus on the use of chemopreventive agents or vaccines designed and tested for prevention of specific cancer types. Most of the existing cancer vaccines aim to manipulate the patient’s own immune system to destroy existing cancer cells and therefore are called “therapeutic vaccines” [1]. Cancer vaccines that target the immune system to prevent the development of cancer are called “prophylactic vaccines” and could potentially be used for cancer immunoprevention. However, there are many other strategies to prevent cancer besides from vaccines. In the present manuscript, we will review existing and potential cancer prevention strategies and focus on non-vaccine- approaches for the prevention of cancer targeting the immune-system.

Types of cancer interventions

There are two global interventions to target cancer: preventive and therapeutic strategies (Table 1). Cancer prevention includes all approaches that aim to reduce the risk and incidence of cancer. There are two types of cancer preventive strategies: Primary and secondary. Primary cancer prevention focuses on health counseling, education and environmental control to decrease exposure to risk factors [2]. More than 30% of cancer deaths could be prevented by avoiding key risk factors, including tobacco use, an unhealthy diet, alcohol abuse, sexually transmitted HPV infections, etc. Tobacco is the single most important risk factor for cancer causing 22% of global cancer deaths. Secondary cancer prevention encompasses interventions leading to the diagnosis and control of cancer and precancerous lesions at the earliest stage possible, including screening and active interventions. These active prevention strategies can be further divided in two types: chemoprevention and immunoprevention.

Table 1.

Differences between chemoprevention and immunoprevention

While immunoprevention focuses on directly or indirectly targeting the immune system, chemopreventive strategies utilize drugs or natural compounds to prevent cancer development [3]. Less than a dozen chemopreventive agents have been FDA approved: tamoxifen and raloxifene for breast cancer prevention [4, 5], celecoxib for colorectal polyps prevention in FAP [6, 7], BCG and valrubicin for bladder cancer prevention [8, 9] and several topical agents for prevention of skin cancer in patients with actinic keratosis: fluorouracil, diclofenac, masoprocol and ingenol mebutate [10].

Differences between immunoprevention and immunotherapy

The term “immunoprevention” refers to the manipulation of the immune system for disease prophylaxis, while “immunotherapy” is reserved for the treatment of existing disease by targeting the immune system. “Immunoprevention” or “immunoprophylaxis” represent one of the most successful types of interventions ever developed. Preventive vaccines have practically eliminated infectious diseases such as polio. In contrast, immunotherapies against chronic infectious diseases have not shown enough efficacy due to the mechanisms developed by infectious pathogens to avoid recognition and elimination by the immune system. Until very recently, the same was believed about cancer immunotherapies, with most of strategies failing due to the tolerance and escape of detection by the immune system [11]. However, in the past decade, huge advances have been made in the immunotherapy field, most of them based on a better understanding of the molecular events leading to tumor tolerance [12].

Immunoprevention of tumors caused by infectious agents

Two different types of tumors can be targeted by immunoprevention, tumors caused by infectious agents or the ones caused by non-infectious agents. Effective vaccines are available for viruses associated with cancer development. The vaccine against the hepatitis B virus (HBV) initially developed for hepatitis prevention has also been shown to reduce the incidence and mortality of liver carcinoma [13, 14]. In 2006, the FDA approved 2 vaccines developed primarily for the prevention of cervical cancer associated with HPV infections: Gardasil (Merck) and Cervarix (GlaxoSmithKline). Gardasil is a quadrivalent vaccine that protects against 4 HPV types: HPVs 6, 11, 16, and 18. Cervarix is a bivalent vaccine against HPVs 16 and 18. These vaccines can reduce the incidence of cervical cancer by 70% [15].

Immunoprevention of tumors caused by non- infectious agents

There are two kinds of approaches to non-infectious tumor immunoprevention: one that targets specific antigens (vaccines) or another that targets non-specific immune components. The drugs developed for the latter approach will be called “immune-modulators” in the present manuscript. At this point, immunopreventive approaches for tumors of non-infectious cause are at the preclinical level of development. Multiple transgenic mouse models harboring activated oncogenes have been produced in the past decade and they can be used to test prevention strategies by targeting the immune system [16].

The Immunosurveillance theory and implications for Immunoprevention

In the 1950s, Thomas and Burnet developed the theory of tumor immunosurveillance in which they proposed that the immune system protects against developing tumors by attacking abnormal cells before they progress to invasive malignancy. However, when this hypothesis was tested in CBA/H nude mice, the most congenitally immunodeficient mice available in 1970’s, mice were found to develop spontaneous tumors and methylcholanthrene (MCA)-induced sarcomas in a similar rate to wild type mice [17]. The findings brought skepticism about the immunosurveillance theory. However, two issues in this experiment were found decades later. First, nude mice are not entirely immunodeficient as they still maintain NK cells and low numbers of T cells. Second, the CBA/H strain of mice expresses a highly active isoform of the enzyme that metabolizes MCA to its carcinogenic form. Subsequent studies using neutralizing antibodies to IFN-g in tumor bearing mice or mice lacking the IFN-g receptor showed accelerated tumor growth compared with control mice [18, 19]. Studies using perforin-deficient mice have shown increased susceptibility to both chemically induced and spontaneous models [20]. Mice lacking the recombinase activating gene (RAG)-2 cannot somatically rearrange lymphocyte antigen receptors and cannot produce peripheral αβ T cells, B cells, NKT cells, or γδ T. Since RAG-2 is selectively expressed in the lymphoid system, RAG-2−/− mice provided an appropriate model to exclusively study the effects of host lymphocyte deficiency on tumor development. Following subcutaneous injection of a chemical carcinogen RAG-2−/−mice developed sarcomas at the injection site faster and with greater frequency than strain-matched controls [21, 22]. Data about the role of immunosurveillance in cancer prevention is not only restricted to preclinical mouse models [23] but also exists in humans. As an example of this, approximately a four-fold increase in the incidence of de novo malignant melanoma after organ transplantation has been reported [24]. http://www.nature.com/ni/journal/v3/n11/full/ni1102-991.html - B57 In spite of that, tumors do develop in the presence of a functional immune system. The protection provided by the immune system may not be sufficient and immune-preventive strategies aiming to promote antitumor immune defenses may bring a proportional decrease in tumor incidence.

Theory of Immunoediting and implications for Immunoprevention

The immunoediting theory later replaced the immunosurveillance theory to reflect the increasingly apparent dual role of immunity in not only preventing but also in sculpting the tumoral process. Dunn et al [25] proposed that the process of immunoediting was comprised by three phases and named them “three Es of cancer immunoediting”: elimination, equilibrium and escape. The first phase, elimination, represents the original concept of immunosurveillance and when successful, the whole process of immunoediting is completed. In the second phase, equilibrium, the host immune system and the tumor cells that survived the elimination step enter into a state in which the immune system imposes selective pressure on the tumor cells that is enough to contain, but not completely remove the tumor, resulting in a tumor composed of many genetically unstable and mutating tumoral cells. In the third phase, escape, tumor cells selected by the equilibrium phase now grow and become detectable clinically. Some of the escape mechanisms include: elaboration of immunosuppressive cytokines (TGF-β), growth factors (VEGF), recruitment of regulatory T cells, and changes in tumor cells that affect their recognition by immune effector cells (loss of MHC, development of IFN-γ insensitivity, expression of anti-apoptotic signals, etc) [26]. The accumulation of metabolic enzymes, indoleamine 2,3-dioxygenase (IDO) and arginase, that suppress T cell proliferation and activation as well as the high expression of tolerance-inducing molecules (PD-1 and CTLA-4) are additional mechanisms that contribute to tumor escape [27]. Further understanding of the mechanisms responsible for tumor equilibrium and escape will be invaluable to generate better strategies for cancer immune-prevention.

Advantages of Immunoprevention without vaccines

A key issue in developing effective cancer preventive vaccines is the large variability of targetable tumor antigens (even within the same patient tumor). Even in the event of a very common antigen selected as the target, results have been less than optimal [28]. Immunopreventive strategies that target the immune system directly would have to deal with less variability and therefore, potentially could result in more effective outcomes as evidenced by the recently FDA-approved non vaccine- cancer immunotherapies [12]

Risk profile of immune-preventive agents

Given the fact that preventive strategies are directed to healthy patients who, even in the case of pertaining to high-risk groups, may actually never develop cancer, safety is a major concern. As immune-preventive agents continue to being developed, major attention should be focused on dissecting their risk profile. Since some pro-tumoral targets may also have a role in host defense against pathogens, strategies that aim to suppress pro-tumoral targets may actually result in an increased risk for infections [29]. On the other hand, strategies that focus on inducing anti-tumoral targets may actually result in the development of autoimmune conditions [30].

Shotgun vs. Targeted immune-preventive approaches

Broad anti-inflammatory agents such as celecoxib and aspirin have proven beneficious as preventive agents in high-risk populations but even these drugs have several toxicities, including cardiovascular toxicities and increased of bleeding risk, when used chronically [31]. It would be expected that better molecular characterization of the specific immune events surrounding pre-neoplastic lesions may theoretically result in high efficacy with few side effects through the use of targeted immune-preventive strategies.

Targets for tumor Immunoprevention

PRO-TUMORAL

Interleukin-6

The complex of IL-6 and IL-6R does not lead to signaling but instead associates with the protein gp130 inducing its dimerization and starting intracellular signaling via the JAK/STAT and the Ras-Raf-MAPK signaling pathways [32]. IL-6 is the only cytokine that in vivo uses both classical membrane-bound receptor signaling and trans-signaling through its soluble receptor [33]. Using mice in which the portions of gp130 have been genetically modified in models of inflammatory diseases led to the conclusion that the IL-6-gp130-STAT3 axis was fundamentally required for the orchestration of the inflammatory process [34].

IL-6 and its downstream targets are involved in the regulation of cell proliferation, survival, and metabolism, explaining the close association between IL-6 signaling and tumorigenesis [34]. IL-6 regulates normal and tumor stem cell self-renewal through Notch-3 activation and upregulation of the hypoxia response protein carbonic anhydrase IX (CA-IX), which creates a permissive environment for breast cancer stem cells survival in hypoxic conditions [35]. Gao et al have shown that activated EGFR induces epithelial expression of IL-6, which induces STAT3 activation in lung carcinoma [36]. Lesina et al have demonstrated pancreatic expression of oncogenic Kras(G12D) causes an IL-6 transsignaling-dependent activation of Stat3/Socs3 required to promote pancreatic tumorigenesis [37]. In a model of colitis-associated cancer (CAC), IL-6 produced by hematopoietic cells enhances adenoma growth [38] and its ablation reduces the number and size of colonic adenomas [39]. Toclizumab, a humanized anti–IL-6 receptor monoclonal antibody, globally blocks IL-6 activities since it competes for both the membrane-bound and soluble types of IL-6 receptors [40]. This antibody was approved by the FDA in 2010 for the treatment of patients with rheumatoid arthritis that have failed at least one anti-TNF therapy. When targeting IL-6 with antobiodies. the impact of blocking classical membrane-bound signaling and IL-6 trans-signaling will have to be considered. A recent Phase I clinical trial of a chimeric IL-6 monoclonal antibody, siltuximab, has reported good tolerance in cancer patients and phase II therapeutic trials are currently ongoing [41]. Clinical preventive strategies targeting IL-6 would be highly promising, particularly in inflammation-associated cancers. A proof of principle for anti-IL-6 immune prevention of pancreatic cancer in preclinical model with few long term side effects has been recently reported [42].

Interleukin-8

Interleukin-8 (IL-8), also known as CXCL8, is a proinflammatory CXC-chemokine that binds to CXCR1 and CXCR2, two cell-surface G protein–coupled receptors [43, 44]. The receptors CXCR1 and CXCR2 have also been found over-expressed on cancer cells and multiple reports have described an autocrine IL-8 signaling loop associated with cell proliferation, invasion and angiogenesis as mechanisms to induce tumorigenesis in various models [45–47]. Humanized monoclonal antibodies against IL-8 have been used in murine models and have shown to decrease bladder cancer and melanoma growth in xenografts models [45, 46]. A liposomal encapsulated small interfering RNA that supressed IL-8 expression in a model of ovarian cancer xenografts resulted in decreased microvessel density and growth retardation [48]. Targeting IL-8 expression for cancer prevention would fail to account for the signaling effect of other CXC-chemokines. In this regard, several CXCR1/2 receptors [49–55] inhibitors are now being developed and they may also have a potential for immunoprevention.

1. Published OnlineFirst March 5, 2012; doi: 10.1158/1535-7163.MCT-11-0915

T_H17 and Interleukin-17

IL-17A, a pro-inflammatory cytokine produced mainly by CD4+T cells known as T_H17 (T Helper 17) cells, plays a key role in chronic inflammation-related human diseases and autoimmunity [56, 57]. In addition to T_H17 cells, IL-17A can also be produced by NKT cells [58], gamma-delta T cells [59] and CD8+T cells [60]. IL-17A binds the heteromeric IL-17R complex consisting of at IL17RA and IL17RC [61]. TGF-β and IL-6 are the most important cytokines required for T_H17 differentiation [62], while IL-23 is crucial for its phenotype maintenance [63]. The differentiation of T_H17 cells also requires the expression of retinoic acid receptor-related orphan nuclear receptor gamma (RORγt), which is a STAT-3 dependent transcription factor [64]. T_H17 cells are increased in the tumor microenvironment during tumor development [65, 66] and several manuscripts have reported the role of IL-17 in promoting tumorigenesis [66–72]. The mechanisms for the pro-tumorigenic effects include the induction of angiogenic and anti-apoptotic factors, migration and invasion and recruitment of myeloid derived suppresors cells. Conversely, T_H17 cells have been also reported to mediate anti-tumor effects by promoting CTLs activities, MHC antigen expression, and production of IFNγ. The anti-tumor effects reported were mostly described in the setting of established cancer [73] and some of them relied on IFNγ, rather than on IL-17A secretion. The FDA has approved the use of ustekinumab, an anti-IL-23 antibody that would decrease T_H17 responses, for use in moderate-to-severe psoriatic arthritis [74]. Ixekizumab and secukinumab, IL-17A–specific antibodies, and brodalumab, monoclonal antibody directed against IL-17RA, have shown safety and efficacy in early phase trials for autoimmune and chronic inflammatory conditions [75–77] and are currently being tested in phase III trials for autoimmune diseases. RORγt small molecule inhibitors are currently being developed and under pre-clinical testing. In any case, the potential idea of targeting IL-17 producing cells as a strategy for cancer immunoprevention should consider their impacts in autoimmunity as well as the potential pro-tumoral effects [78–80]. Potential preventive strategies targeting IL-17 would be of particular interest for inflammation-induced cancer.

Interleukin-1

The IL-1 family encompasses three proteins: IL-1α, IL-1β and the IL-1 receptor antagonist (IL-1ra) [81, 82]. There are two IL-1 receptors, the biologically active IL-1 receptor type I (IL-1RI), agonistic receptor ubiquitously expressed with preferential binding to IL-1a, and the IL-1 receptor type II (IL-1RII or decoy IL-1R), antagonist receptor expressed on B cells, neutrophils, and monocytes with preferential binding to IL-1 β. [83, 84]. The mechanisms responsible for the IL-1- tumor growth promotion are: induction of pro-metastatic genes (ex: matrix metalloproteinases) and stimulation of adjacent cells to produce angiogenic proteins and growth factors (VEGF, IL-8, IL-6, TNFα, and TGFβ) [81, 83, 85]. Cancer cells are able to produce IL-1 or can induce its production from cells within the tumor microenvironment [86]. IL-1 has autocrine behavior by inducing proliferation and invasion of the tumor cells and is also a paracrine effector on the stromal cells. From a clinical perspective, elevated IL-1 levels within the serum have been associated with more aggressive tumors [87].

The IL-1 receptor antagonist (IL-1ra) is an inhibitor of IL-1 and has been shown to decrease tumor growth, angiogenesis, and metastases in murine xenograft models [84]. IL-1 receptor antagonist (IL-1ra) is an approved treatment for patients with rheumatoid arthritis, has shown some promising results in indolent multiple myeloma treatment [88] and is now being tested on other cancers. IL-1ra could represent effective novel immune-preventive agents but further studies are needed [82]

STAT3

STATs (signal transducers and activators of transcription) are a family of transcription factors discovered in the early 1990’s that regulate cell growth and differentiation through phosphorylation by receptor-associated kinases [89]. STAT3 was found to mediate the acute phase response observed with inflammation and stress that is often mediated by interleukin-6 (IL-6) and other proinflammatory cytokines [90]. Activated STAT3 is phosphorylated on tyrosine and forms a dimer through phosphotyrosine/src homology 2 (SH2) domain interaction. The dimer enters the nucleus via interaction with importins and binds genes with critical role in cellular proliferation, survival, pluripotency, invasion, and angiogenesis, explaining its association with cancer development. Fibroblasts engineered to express an activated Stat3 formed tumors in nude mice, thus establishing Stat3 as an oncogene [91]. Increased STAT3 activity is present in more than 70% of solid tumors [92, 93], and the inhibition of STAT3 can mediate tumor regression [94]. Tyrosine kinase oncoproteins can target constitutive STAT3 phosphorylation [95], and the ablation of STAT3 results in impairment of malignant transformation by tyrosine kinase oncoproteins [96–98]. Stat3 signaling within the tumor microenvironment induces a procarcinogenic cytokine, IL-23, while inhibiting a central anticarcinogenic cytokine, IL-12, thereby shifting the balance of tumor immunity toward carcinogenesis [99].

Small molecules inhibitors of STAT3 have been developed and used in the preclinical setting but they mostly target upstream receptor and non-receptor tyrosine kinases, resulting in low specificity of action. Sorafenib, a multikinase inhibitor, decreased STAT3 phosphorylation in hepatocellular carcinoma by inhibiting phosphoinositide 3-kinase (PI3K)/Akt and MEK/ERK pathways [100]. WP1–66, a JAK2 inhibitor, has also been used to inhibit stat3 activity in a model of renal cell carcinoma [101]. RNA interference approach has also been done to knock down stat3 and in preclinical models it supressed tumorigenesis [102]. Finally, a chemically modified double-stranded STAT3 oligonucleotide decoy that inhibited tumorigenesis in murine models has also been clinically tested in a phase 0 setting and proved to be safe and effective to selectively suppress STAT3 target genes in head and neck tumors [103]. Given the key role of STAT3 in promoting prostate, colon and pancreatic cancer [104–106], STAT3 inhibitors could be useful for cancer prevention in patients with high risk for these diseases.

PD-1/ PD-1L

Programmed cell death-1 (PD-1), a CD28 family receptor, plays a key role in tumor tolerance [107]. The interaction between PD-1/PD-L1 results in the inhibition of T lymphocyte effector functions, with subsequent down-regulation of antitumor immune responses [108]. PD-1 has two ligands, PD-L1 and PD-L2 with a different pattern of expression and regulation. PD-L1 is expressed in most hematopoietic cells and in multiple tumors including melanoma, pancreatic, breast and lung tumors [109–113] but PD-L2 expression is restricted to macrophages, mast cells and dendritic cells. High expression of tumoral PD-L1 has been associated with poor prognosis in multiple studies [109, 114]. Preclinical studies have suggested that PD-1/PD-L1 signaling favors tumor immune evasion and that blockade of the PD-1 pathway can restore antitumor immune responses. Tumor cells that express PD-1L have increased resistance to T cell- mediated lysis with increased tumorigenesis and invasiveness, when compared with tumor cells that do not express PD-L1 and the administration of anti-PD-L1 antibodies reversed these effects[115]. Also, PD-1 deficient mice presented inhibition of tumor growth, suggesting a strong antitumor immune response in the absence of PD-1/PD-L1 signaling [116]. Finally, in mice with established tumors, the administration of anti-PD-1 or anti-PD-L1 antibodies led to reduction in tumor size and increased survival [117].

In recent years, early phase studies reported clinical activity of PD-L1 [118] and PD-1 [119, 120] monoclonal antibodies in patients with advanced solid tumors. As larger therapeutic trials get conducted for efficacy validation, the utilization of these antibodies for tumor immunoprevention in high risk populations may become a potential effective strategy, particularly if PDL1/PD1 biomarkers are well defined in pre-neoplastic lesions and side effects are manageable. [121] Given the success of these agents in melanoma and non-small cell lung cancer, immunoprevention targeting PD1/PD1L might be of high relevance in patients with pre-neoplastic lesions associated with either disease.

CTLA-4

CTLA-4 is a cell surface protein constitutively expressed on T-cells, but up-regulated by the interaction of the T cell receptors (TCR) and CD28 [122]. A splice variant of CTLA-4 may also be secreted by T-cells in a soluble form that binds to T cell receptors. CTLA-4 interacts with CD80 (B7-1) and CD86 (B7-2) and competitively competes with CD28 for binding with these receptors, therefore acting as an inhibitor of T-cell proliferation [123]. The receptor CD28 is similar to CTLA4 and shares a homologous amino-acid sequence. CD28 binding causes co-stimulation and activation of T-cells, while CTLA-4 binding causes T-cell inhibition [122]. As CTLA-4 provides a regulatory signal that inhibits T-cell activation, it serves as an immune checkpoint inhibitor to prevent over activation of T cells [124]. Two antibodies have been tested against CTLA-4 clinically: Ipilimumab and Tremelimumab. Ipilimumab has become the standard of care for melanoma patients after a phase III trial showed survival benefit in patients with metastatic melanoma [125]. Tremelimumab failed to show a survival advantage in a phase III clinical trial in melanoma [126] and is currently being investigated in combination with other agents. The recent success in targeting CTLA-4, particularly in melanoma has led to increased interest in understanding the role that CTLA4 plays in immune surveillance and immunoprevention, particularly for patients with pre-neoplastic lesions.

Innate Lymphoid cells

Innate lymphoid cells (ILCs) have been recently described as a group of innate immune cells that can regulate immunity, inflammation, and tissue repair in multiple anatomical compartments. ILCs can regulate commensal bacterial communities, contribute to resistance to bacterial pathogens, promote inflammation, and orchestrate tissue repair and wound healing [127]. RORgt+ ILCs, which express and require the nuclear hormone receptor retinoic acid (RA) receptor related orphan receptor (RORgt), are involved in lymphoid tissue genesis and can produce T helper 17 (T_H17)-associated cytokines IL-17 and IL-22. On the basis of functional characteristics, three subpopulations of RORgt-dependent ILCs can be distinguished: (a) lymphoid tissue–inducer (LTi) cells, (b) ILCs dedicated to IL-22 production, and (c) IL-17-producing ILCs. ILCs that produce both IL-17 and IL-22 have also been found that may be cells derived from either IL-17- or IL-22-producing cells [128]. RORgt ILCs have recently been the target of increasing attention due to their key role in multiple models of microbe-related tumorigenesis. In a murine model of bacteria induced- colon cancer, innate lymphoid cells have been shown to sustain tumorigenesis through production of interleukin-22 [129]. In a melanoma model, the chemokine CCL21 expressed by the tumor cells recruited not only myeloid-derived suppressor cells that can directly suppress antitumor immune responses but also recruited RORgt+ ILCs that mediated the enhancement of tumor growth as shown by the absence of effect in tumor growth in RORgt-deficient mice [130]. Further understanding of these cells will be required for the development of immune strategies targeting them.

Myeloid derived supressor cells

Myeloid derived suppressor cells (MDSC) are a heterogeneous population of early myeloid cells that are expanded in various disease states with the capability of suppressing T-cell responses [131–133]. In addition to their suppressive effect on adaptive immune responses, MDSCs have also been reported to regulate innate immune responses by modulating the cytokines secretion of macrophages [134].

Myeloid progenitor cells and immature myeloid cells (IMCs) constitute the major types of MDSCs. In healthy patients, IMCs that are generated in the bone marrow quickly differentiate into mature granulocytes, macrophages or dendritic cells (DCs). In contrast, during pathological states like cancer and infections, a block in the differentiation of IMCs into mature cells results in an increase of the immature populations. Another relevant event during these pathological states is the upregulation of arginase1 (ARG1) and inducible nitric oxide (iNOS) on IMCs, conferring them immune suppresive properties.

Murine MDSCs are characterized by the co-expression of two myeloid differentiation antigens: Gr1 and CD11b [135]. The normal murine bone marrow can contain up to 30% cells with this phenotype but the spleen contains only 2–4%. Human MDSCs have a different phenotype, being defined by CD11b+/CD14− expression [136]. The levels of immune cells with these phenotypes have been shown to correlate with prognosis and overall survival [137–139]. Multiple methods of inhibiting MDSCs are currently under investigation. These methods involve (1) induction of MDSC maturation to non-suppressive cells (all trans retinoic acid, vitamin D), (2) decrease MDSC levels (sunitinib, gemcitabine, 5-FU, CDDO-Me), or (c) functional inhibition of MDSC (PDE-5 inhibitors, cyclooxygenase 2 inhibitors) [140]. These approaches may enhance anti-tumor immunity through reduction in immune suppression and thus would be relevant in the immune-preventive scenario.

Tregs

Regulatory T cells were initially called suppressive CD4+ T cells as they were found to negatively regulate tumour immunity leading to tumour growth in mice [141]. The transcription factor Foxp3 is crucial for the development and functionality of T regulatory cells and the IL-2 receptor α-chain, CD25, serves as the phenotypic marker for CD4+ T regulatory cells. Therefore, the classic regulatory T cells are phenotyped as CD4+CD25+FOXP3+T cells (Tregs). Multiple possible suppressive mechanisms for regulatory T cells have been proposed: induction of B7-H4 expression by antigen presenting cells (APCs) which can induce T-cell cycle arrest, direct killing of T cells or APCs through perforin or granzyme pathways, CTLA4-mediated suppression of T cell activity and direct inhibition of T-cell activation by the secreted cytokines IL-10 and TGF-b [142]. The role of Tregs (CD4+CD25+Foxp3+) on cancer progression has been widely demonstrated in several preclinical models [143]. Tumors trigger an increase of T regulatory cells (CD4+CD25+Foxp3+) numbers in lymph nodes and in the tumor microenvironment, leading to immunosuppression. Several reports have linked the abundance of Tregs in tumors with worse prognosis [144, 145].

Cyclophosphamide is an alkylating agent that causes DNA crosslinking and is utilized as a chemotherapeutic agent for several types of tumors. High dose cyclophosphamide can induce effective anti-tumor activity. However, low dose cyclophosphamide can also induce improved anti-tumor immune responses as a result of Tregs depletion. This was demonstrated by the evidence that cyclophosphamide treatment caused an immune-mediated regression of a cyclophosphamide-resistant lymphoma [142]. Several preclinical studies have confirmed the effect of cyclophosphamide on Tregs depletion and clinical studies have also been conducted with similar results [146]. Studies have also shown that anti-CD25 antibodies induce tumor rejection in mice [147, 148] and humans. However, this strategy is not as effective in advanced tumors, highlighting the critical role that Tregs may have in the tumor immune preventive setting.

Macrophages

Macrophages are a major component of the inflammatory tumor stroma and macrophages that infiltrate tumors are called tumor-associated macrophages (TAMs) [149]. Macrophages are classified in M1 (classically activated) and M2 (alternatively activated) phenotypes in analogy with the Th1 and Th2 classification [150, 151]. Macrophages have functional plasticity and can change their functional status in response to changes in the microenvironment [152]. When macrophages are exposed to lipopolysaccharides and interferon (IFN)-ã, they polarize to an M1 phenotype with antitumor properties and when they are exposed to Th2 cytokines (IL-4, IL-13, IL-10), they polarize to an immune suppressive M2 phenotype that promotes tumor growth [150, 153]. TAMs are a major source of COX-2 and the selective COX-2 inhibitor celecoxib has been shown to change the TAM phenotype from M2 to M1, in parallel to a reduction in number of colon polyps in a murine model of colon cancer [154]. A large clinical trial has shown that celecoxib reduces the incidence of adenomas, but increased the risk of cardiovascular events [155]. Other selective COX2 inhibitors, with potentially less side effects, are currently being tested and could represent potential colorectal cancer chemopreventive strategies. In a different manner, a recent murine and human study has shown that CD40 agonists can cause macrophage activation with rapid stromal infiltration and tumoricidal effect in pancreatic adenocarcinoma [156]. This led us to speculate that similar agents could be used for prevention in patients with pancreatic pre-neoplastic lesions.

Mast cells

Tumor-infiltrating mast cells remodel the tumor microenvironment and promote tumor growth. Mast cell infiltration and activation in tumors is mainly mediated by tumor-derived stem cell factor (SCF) and its receptor c-Kit. [157]. Mast cells accumulate in the tumor microenvironment and exacerbate immunosuppression by inducing MDSC and Treg infiltration in an IL-17-dependent fashion [158]. The presence of mast cells in many precancerous lesions and tumors is associated with a poor prognosis, suggesting that mast cells may promote an immunosuppressive tumor microenvironment and impede the development of protective antitumor immunity [159–161].

Targeting mast cells can be achieved indirectly via targeting mast cells mediators with agents like anti-TNF monoclonal antibodies (infliximab) or directly through the use of mast cell stabilizing agents like sodium cromoglycate. Preclinical work with DSS-induced colitis with infliximab has decreased development of colorectal tumors in association with decreased mast cell numbers [162]. Stabilization of mast cells through sodium cromoglycate has also shown efficacy in preclinical studies [163] but the specificity of this compound has been questioned [164]. Interventions aimed at inhibiting mast cells function hold promise as potential effective cancer preventive strategy, particularly because of the low side effects of this agents.

ANTI-TUMORAL

IFN-α

Interferons alpha (IFNs-α) are pleiotropic cytokines from the type I IFN family, initially described for their antiviral activity achieved by directly inhibiting viral replication in infected cells [165]. IFN-α can induce anti-tumoral effects through induction/promotion of apoptosis and inhibition of cell growth [166]. Multiple studies have shown that IFN-α can induce the rapid differentiation of monocytes into activated dendritic cells (DCs) that are involved in effective anti-tumor T-cell immunity [167]. Moreover, IFN-α can induce polarization of THelper (TH) cells to TH1 cells and activation of cytotoxic T cells [168]. Finally, IFN-α can promote B cell differentiation, antibodies production and Ig class switching [169]. In sum, IFN-α has a broad effect on immune system activation and prolonged activation can lead to autoimmune reactions. IFN-a is used for the treatment of multiple malignancies including melanoma and renal cell carcinoma [170, 171] but its utilization for cancer prevention has not been explored yet.

CD8+T cells

It is well known that CD8⁺ T cells are important for antitumoral responses. The expansion of tumor-specific CD8⁺ T cells can trigger effective antitumoral immunity [172], and this constitutes the basis for multiple vaccine trials. Recent reports indicate that CD8⁺ T cells may also be capable of protumoral activity depending on the antigen specificity [173]. There is large evidence about the value of high numbers of cells at the tumor site with CD8 memory T cells as a prognostic factor for overall survival (OS) in patients with solid malignancies [174]. Peptide based vaccines with adjuvants containing cytokines, chemokines or costimulatory molecules have been developed to amplify and direct the immune response but are beyond the scope of the present review.

CD4+T cells (TH1)

TH1 CD4+T cells are driven by IL-12 activation of Stat4 and T-bet transcription factors on naïve T cells [175, 176] that induces and up-regulates IFNγ [177]. Ectopic expression of T-bet in T_H2 or T_H17 cells results in their conversion into IFNγ-producing T_H1-like cells (Szabo). The central role of IFNγ- producing CD4+T cells in antitumor immunity has been known for a long time [178]. CD4⁺ T cells play an indispensable role in the generation of vaccine-related therapeutic primary immune responses for the generation of long-term immune memory [179]. In multiple human tumors, TH1 CD4+T cell infiltrates have been found associated with a better prognosis [180]. Specific effector immune cells proliferate within the tumor microenvironment but the contribution to anti-cancer immunity is unclear, given the immunosuppressive factors within the tumor. Vaccines targeting specific antigens should be planned in combination with agents that aim to decrease the immunosuppressive environment.

Cancers with high mutational loads have a high number of tumor-specific neoantigens. Recently, Linnemann and colleagues developed a screening platform to evaluate CD4⁺ T-cell responses to tumor neo-epitopes in melanoma patients and the results indicated that CD4⁺ T-cell reactivity toward patient-specific neoantigens is a common feature of melanoma and suggest that therapeutic and potentially preventive strategies to boost CD4⁺ T-cell responses may be effective in melanoma and other cancers with similar mutational loads [181].

NKT cells

NKT cells (NKTs) are a subset of T cells that recognize self- and microbial-derived glycolipids such as α-galactosylceramide (α-GalCer) presented by an HLA class I-like molecule, CD1d. NKTs are capable of producing key cytokines of both Th1 and Th2 T cells that provide signals for other immune cells to start innate and adaptive responses [182]. Low levels of circulating NKTs can predict poor clinical outcome in certain human tumors [183] and tumor infiltration by NKT is a predictor of favorable outcomes [184]. Treatment of mice with α-GalCer has been shown to suppress tumor metastasis [185] yet clinical trials have not shown efficacy [186]. Novel α-GalCer analogs have been designed to favor Th1-biased immunity, with greater preclinical anticancer potential and are currently being tested [187].

Dendritic Cells (DC)

Dendritic Cells play a critical role in linking innate and adaptive immunity. They have a key role in role in capturing and presenting antigens in the form of MHC II receptors to T cells stimulating them to develop into effector T cells. In this role, DCs play a significant role in the tumor microenvironment, surrounding tumors and providing signals to other immune cells as to stimulate or suppress T cell activation [188]. Tumor tolerant DCs are one reason that budding tumors can escape immune surveillance. Interleukin-10 secreted from tumors can inhibit the maturation of dendritic cells, and promote dendritic cells apoptosis, thereby promoting tumor tolerance [188]. Potent dendritic cells are required for a strong anti-tumor immune response after chemotherapy or radiation. There is a great degree of variability in the location and number of dendritic cells in tumors [189]. However, most evidence shows that although there are some levels of dendritic cells present in tumors, suppressed by a variety of immunosuppressing cytokines and other factors from the tumor microenvironment. Dendritic cells can also have prognostic value in treating patients. Studies have shown that tumors with dendritic cells expressing low levels of CD86 and greater levels of IL-10, generally are resistant to therapy and have a poor outcome [190]. All of these reasons make dendritic cell stimulation or prevention of dendritic cell tolerance, potential good strategies for tumor immune prevention.

CONCLUSIONS

The potential combination of agents that decrease or suppress the pro-tumoral events with agents that boost the anti-tumoral immunity applied to selected high risk populations may have a strong impact in decreasing cancer incidence (Figure 1). By gaining further understanding on the cancer microenvironment, in the future we might be able to not only identify appropriate high risk patients who may benefit from immunoprevention, but also determine the precise inflammatory mediators that place them at risk, so that individualized immunoprevention might become a reality.

Figure 1.