Exploiting natural anti-tumor immunity for metastatic renal cell carcinoma

Abstract

Clinical observations of spontaneous disease regression in some renal cell carcinoma (RCC) patients implicate a role for tumor immunity in controlling this disease. Puzzling, however, are findings that high levels of tumor infiltrating lymphocytes (TIL) are common to RCC. Despite expression of activation markers by TILs, functional impairment of innate and adaptive immune cells has been consistently demonstrated contributing to the failure of the immune system to control RCC. Immunotherapy can overcome the immunosuppressive effects of the tumor and provide an opportunity for long-term disease free survival. Unfortunately, complete response rates remain sub-optimal indicating the effectiveness of immunotherapy remains limited by tumor-specific factors and/or cell types that inhibit antitumor immune responses. Here we discuss immunotherapies and the function of multiple immune system components to achieve an effective response. Understanding these complex interactions is essential to rationally develop novel therapies capable of renewing the immune system's ability to respond to these tumors.

Keywords:

• renal cell carcinoma
• apoptosis
• immunotherapy
• T cell
• NK cell
• MDSC

Introduction

Renal cell carcinoma (RCC) is the third most common urogenital cancer in the world, and affects ∼60,000 people in the US each year.¹ Clinical management of RCC is variable, and depends on tumor stage at the time of diagnosis. Standard treatment for localized RCC tumors with no evidence of metastases or a single metastasis is surgery to remove the tumor-bearing kidney (nephrectomy) and metastatic lesion, and 5-year survival rates for these patients are high (˜90%).² Yet, ∼30% of RCC patients have multiple metastases at diagnosis, and a third of patients who undergo nephrectomy will develop metastases. Metastatic RCC (mRCC) is incurable, with a median survival time of only 7–11 months and a 5-year survival rate of less than 10%; poor survival is mainly attributed to resistance to chemotherapy and radiotherapy in these patients.³

Patients with locally mRCC can benefit from surgical resection of the tumor-bearing kidney and metastatic lesions. Rare patients with mRCC (∼0.1%) can experience spontaneous remission following nephrectomy supporting the notion that RCC could be an immunologically controlled malignancy.⁴ However, patients with multiple, systemic metastases generally do not benefit from nephrectomy and require augmented immune intervention. Two immunotherapies have shown efficacy for the treatment of mRCC: interleukin (IL)-2 and interferon (IFN)-α⁵. The exact mechanisms by which these therapies work are not completely understood, but it is thought that both stimulate an immune response to the tumor.^6-9 IFN-α is also thought to have direct anti-proliferative effects on tumor cells and anti-angiogenic effects on the tumor microenvironment.¹⁰ Though these agents have demonstrated some efficacy, the effects have been limited. IFN-α has only increased the median survival of mRCC patients by 7 months, with a response rate of less than 20%.¹¹ IL-2 has demonstrated slightly greater survival benefits (reaching 13–17 months); however, the high dose needed to achieve these results is often intolerable to the patient.¹² Small molecule inhibitors have demonstrated efficacy in extending survival. The multi-kinase inhibitor, Sunitnib, was able to prolong survival by 6 months compared to IFN-α, but this benefit was only afforded to 31% of patients.^13,14 Sorafenib, another multi-kinase inhibitor approved for the treatment of RCC, can extend progression-free survival for 10% of patients an additional 3 months over Sunitnib.¹⁵ While these therapies have shown “significant” clinical increases in response rate and progression-free survival compared to placebos and controls, it is clear that improved therapies are needed for mRCC.

The Goal of Cancer Immunotherapy

Immunotherapy for cancer is designed to mobilize an antitumor immune response sufficient to selectively kill malignant cells and promote long-term tumor-free survival.¹⁶ Clinical trials with tumor vaccines utilizing synthetic peptides derived from tumor antigens have had limited success.^16,17 One limitation of peptide-based tumor vaccines is major histocompatibility complex (MHC) restriction, and therefore diminished utility in diverse populations.¹⁸ Suboptimal efficacy could also result from tumor heterogeneity, emergence of tumor cells that lose the targeted antigen, and immunological modulation that limit a productive antitumor immune response. Autologous whole tumor cells include all antigens that a person's immune system could potentially respond to, representing a more exhaustive repertoire of tumor antigens. However, generation of autologous whole tumor cell vaccines requires tumor tissue collection with careful and costly ex vivo processing.¹⁹ To circumvent these technical/financial constraints, but at the same time harness the most relevant tumor antigens for an individual patient, protocols based on the premise of generating an in situ whole tumor cell-based vaccine may prove most beneficial. The success of such an approach requires the coordinated function of multiple immune system components to achieve an effective response, as well as reducing the immunosuppressive activity of other cells and molecules augmented by the presence of the tumor.

Tumoricidal Agents

The first step in the generation of an antitumor immune response is the production of a tumor antigen-laden debris field. A number of treatments can induce tumor cell death within the kidney, including radiation, chemotherapy, radiofrequency ablation, and thermal ablation.^20-22 These techniques can potently induce tumor cell death; however, none discriminate between tumor cells and healthy cells. As a consequence of these treatments, cell death occurs via an apoptotic or necrotic mechanism. It is becoming more appreciated that the means by which a cell dies can determine the degree of immunogenicity or tolerogenicity of the dead cells.^23,24 An alternative to these cytotoxic approaches is the use of reagents that specifically target the genes or proteins expressed by tumor cells or the surrounding tumor microenvironment. For example, drugs that target tyrosine kinases (TKIs; sorafenib, sunitinib, pazopanib, and axitinib) are designed to inhibit vascular endothelia growth factor (VEGF) and angiogenesis.^25,26 TKIs can work well in slowing the growth of kidney tumors, but their use can be associated with significant side-effects and tumor cell growth resumes when treatment stops. Selective tumor cell death can also result from use of TNF-related apoptosis-inducing ligand (TRAIL).²⁷ Considerable effort has been devoted to development of recombinant TRAIL protein, small molecules to augment TRAIL-induced killing, agonistic antibodies specific for the death-inducing TRAIL receptors, and TRAIL-based gene therapies for cancer therapy.^28-36 Direct induction of tumor cell apoptosis is the typical means by which a TRAIL receptor agonist is thought to function in vivo, but recent data suggest that endothelial cells within a tumor also selectively express TRAIL receptor.³⁷ Disruption of tumor vasculature by TRAIL can also lead to tumor size reduction – even for tumors that lack TRAIL receptor. Regardless of the means by which tumor cell death occurs, generation of a debris field of tumor antigens can serve as a powerful stimulus for an antitumor immune response.

Dendritic Cells

Following tumor cell death, the next component important for the generation of an antitumor immune response is dendritic cells (DCs). DCs sample their environment and capture, process and present antigens; functions for which they are known as professional antigen-presenting cells (APC).³⁸ The ability of DCs to process and present antigens acquired from internalized apoptotic cells is well recognized,^39-41 and immunity or tolerance can be induced depending on the context of antigen uptake by DCs.⁴² Phagocytosis of apoptotic cells by immature DCs can be subimmunogenic and even tolerogenic under normal conditions or within the immunosuppressive tumor microenvironment.⁴³ Thus, a successful immunotherapeutic approach for treating cancer should consider the possible necessity for an adjuvant to activate DCs and elicit a strong antitumor immune response.

DCs present antigen as peptide fragments on MHC molecule complexes to T cells to induce adaptive immunity. Typically, DCs in peripheral tissues and circulation capture antigen and migrate into draining lymph nodes where they present antigen to naïve T cells. Alternatively, antigens can reach the lymph node-resident DCs through the lymphatic drainage system. In addition to capturing and processing antigens, DCs must also become activated to upregulate costimulatory molecules. DCs express a plethora of “sensing” molecules including pattern recognition receptors (PRRs) that detect pathogen-associated molecular patterns (PAMPs). Activation of DCs, specifically plasmacytoid DC (pDC) and CD8α DC, via toll-like receptors (TLRs) leads to upregulation of costimulatory molecules CD40, CD80 and CD86. Though both CD8α DC and pDC express MHC II at basal levels, MHC II expression increases with DC activation.

Following TLR-mediated activation, both pDC and CD8α DC are capable of making type I IFN, albeit to varying degrees, and CD8α DC can also produce IL-12 and IL-15, all of which are essential for optimal antitumor immunity.^44-50 Type I IFN is important for activation of the adaptive immune response, and additional mechanisms of DC action.^51,52 Data suggest that for DCs, in particular CD8α DC, type I IFN is important for optimal functionality, such as uptake, processing and retention of antigens.⁵³ The efficiency of these functions by CD8α DC is essential for the cross-presentation and priming of CD8⁺ T cells. With costimulatory molecules expressed and a peptide-loaded MHC I complex, these cross-presenting DC are capable of activating naïve CD8⁺ T cells into antitumor effector cells.

Because of their potent immunostimulatory capacity, DC-based immunotherapy for RCC has been tested in a number of clinical trials.^54,55 The DCs used in these clinical protocols are typically generated from autologous peripheral blood monocytes loaded with RCC-derived mRNA or dead RCC cells and infused back into the patient in the form of a “vaccine.” Such an approach has been tolerated well by the recipients, with few adverse effects that lead to toxicity or decreased quality-of-life.⁵⁶ Cellular immune responses, evidenced by the increased activity of tumor-specific CD8 T cells and NK cells, have been reported in many patients.^57,58 Unfortunately, objective response rates have been low in mRCC patients receiving DC therapy.⁵⁹ It is becoming evident that use of DC-based immunotherapy for RCC may require additional components to enhance the stimulatory capacity of DCs. For example, the combination of DCs with other immunostimulatory cytokines (e.g., IL-2), TLR agonists, agents that reverse tumor-induced immune suppression (e.g.,, denileukin diftitox to deplete regulatory T cells,⁶⁰ or drugs to alter myeloid-derived suppressor cells number/recruitment^61-65), and/or chemotherapeutics to reduce tumor burden may stimulate a more robust antitumor immune response.⁶⁶

CD8⁺ T Cells

Eliciting a therapeutic antitumor immune response in cancer has been a long-standing goal of immunology. While chemotherapeutics and radiation can confer some initial benefit, the ultimate goal of cancer therapy is to induce antitumor immunity sufficient to clear detectable cancerous lesions and small undetectable foci and inhibit recurrence. However, activation of a robust antitumor CD8⁺ T cell response is a multifaceted process. Tumors adopt various mechanisms to escape and/or suppress the antitumor immune response, and therapies must not only activate tumor-specific CD8⁺ T cells but also overcome immunosuppression exerted by the tumor.

Activation and subsequent development of effector and memory CD8⁺ T cells⁶⁷ requires that T cells must come in contact with their cognate peptide bound to MHC I on an APC and be costimulated by the APCs and cytokines they produce. The activation of antitumor CD8⁺ T cells results from cross-presentation of tumor cell-derived antigens by APCs. Concomitantly, as CD8⁺ T cells are binding their cognate peptide, they also engage costimulatory molecules on the APC.⁶⁸ Activation typically occurs via ligation of CD28 on the T cell to CD80/CD86 on the APC, leading to increased cytokine gene expression, promotion of T cell survival and maintenance of T cell responsiveness.⁶⁸ Lack of costimulation results in T cells becoming hyporesponsive (anergic).⁶⁹ Following TCR/CD8 activation and costimulation, T cells must be exposed to cytokines, such as type I IFN and IL-12 for optimal activation, proliferation/survival and subsequent effector function.^67,69,70 Moreover, additional stimuli generated in response to ligation of a number of other costimulatory receptors found on T cells, such as CD27, CD40, CD120b (TNFR2), CD134 (OX-40), and CD137 (4–1BB), are needed to achieve maximal T cell responses.⁷¹ Together, the presentation of tumor-derived peptides in the context of MHC I on an APC, combined with costimulation and cytokines provided via activated APC are all necessary for the induction of a robust antitumor CD8⁺ T cell effector response. Several studies have examined the therapeutic benefit of administering agonists specific for key costimulatory receptors on CD8 T cells in preclinical models of RCC. In particular, agonists against CD40, CD134, and CD137 have proven effective in stimulating CD8⁺ T cell-mediated antitumor responses alone or in combination with immunostimulatory cytokines or chemotherapeutics.^9,72-74

Once CD8⁺ T cells have been primed and activated via antigen-loaded DC and cytokine stimulation, massive clonal expansion of the tumor antigen-specific T cells will follow. These CD8⁺ T cells acquire “effector” function and homing capabilities to traffic to the tumor bed where they will recognize their cognate antigen (presumably being expressed by the tumor) and subsequently carryout destruction of the tumor.^67,69,70,75 These effector functions include the ability to directly kill tumor cells via TRAIL⁷⁶ or FasL mechanisms,⁷⁷ or to secrete cytolytic molecules such as IFN-γ, TNF-α, perforin and granzymeB,⁷⁷ all of which can play a role in direct tumor cell death.^78,79 IFN-γ and TNF-α can also have cytolytic effects on the tumor stroma and vasculature,⁸⁰ as well as activating effects on other immune cells that can offer help in the antitumor response. Once the tumor has been eradicated, the majority of the expanded CD8⁺ T cells will undergo apoptosis, leaving behind a small but effective memory population.⁶⁷

Natural Killer Cells

Natural killer (NK) cells are innate immune cells that have the capacity to kill cancerous cells while sparing normal cells. In humans, mature NK cells make up approximately 10% of the nucleated cells in the peripheral blood, and are identified as CD3⁻ lymphocytes that express CD56 with or without CD16 (also called FcγRIIIA). Most (90–95%) peripheral NK (pNK) cells are CD56^dim and express high levels of CD16; the remaining fraction is CD56^bright CD16^dim/neg 81. It was originally thought that NK cells in the blood and tissues were the same; however, recent work has revealed the existence of unique sub-populations of NK cells within uterus,⁸² thymus,⁸³ skin,⁸⁴ and liver.⁸⁵ In these tissues, CD56^bright NK cells are more abundant than CD56^dim counterparts, a finding that contrasts the peripheral blood. The significance of these differences is that CD56^dim NK cells have potent cytotoxic activity, while CD56^bright versions have limited cytolytic responses and are more efficient producers of cytokines and chemokines.

NK cells express a number of inhibitory and activating receptors on the cell surface.^86,87 Activating receptors include natural cytotoxicity receptors (NCRs: NKp46, NKp30 and NKp44), C-type lectin receptors (CD94/NKG2C, NKG2D, NKG2E/H and NKG2F) and killer cell immunoglobulin-like receptors (KIRs: KIR-2DS and KIR-3DS). Inhibitory receptors comprise additional members of the C-type lectin (CD94/NKG2A/B) and KIR (KIR-2DL and KIR-3DL) receptor families. Unsolicited killing of host cells by NK cells is controlled by inhibitory receptor recognition of MHC I.⁸⁸ NK cells are activated by and destroy host cells that lack or have abnormal MHC I expression. The critical role of NK cells in tumor immunosurveillance has fueled a long-standing interest in therapeutic strategies designed to stimulate antitumor responses by exploiting the natural activity of NK cells.

NK cells are the predominant lymphocyte populations in the decidua during embryo implantation. These decidua NK (dNK) cells are CD56^brightCD16^neg and demonstrate little-to-no cytotoxicity capacity. Interestingly, dNK cells have an increased ability to secrete cytokines and angiogenic molecules,^82,89-91 particularly PGF and VEGF, angiopoietins, TGFβ and, in mice, vasoactive proteins such as IFN-γ, angiotensin II receptors and atrial natriuretic peptide.^92,93 Accumulation of dNK cells correlates with angiogenesis, lymphangiogenesis, endometrial edema and vasodilation leading to increased uterine artery blood flow.⁹⁴ Furthermore, CD56^bright CD16^neg NK cells are critical for lessening cardiac damage by increasing angiogenesis, vascular remodeling and cardiac function⁹⁵ by inducing endothelial cell proliferation.⁹⁶ Mediators of this phenotypic and functional differentiation in endometrium are not well known, but one inducer of pNK to dNK cell phenotypic differentiation is TGFβ.⁹⁷ TGFβ is highly expressed in decidua in situ⁹⁸ and mediates conversion of pNK cells from healthy donors to dNK cells in vitro.⁹⁷ Thus, CD56^brightCD16NK^neg cells could have significant, yet under-appreciated, roles in promoting angiogenesis in a number of pathological situations, including tumor growth and metastasis.

A distinct feature of RCC is high levels of tumor infiltrating lymphocytes (TIL), and NK cells are consistently identified.⁹⁹ Poor survival is indicated when NK cells make up less than 20% of the total TIL population for advanced RCC tumors.^100-103 Other studies suggested that insufficient activation of RCC tumor NK cells contributed to therapy failure for patients treated with combinations of IL-2, IFNα, and histamine.^101,104,105 Even when RCC tumors are heavily infiltrated with NK cells, these tumor-infiltrating NK cells (TiNK) were unable to lyse target cell lines.¹⁰⁶ Phenotypic analysis has demonstrated that RCC TiNK cells are different compared to matched peripheral blood samples for several receptors including the cytotoxic marker CD16.^103,107 Specifically, cytotoxic CD56⁺CD16⁺ NK cells, resembling pNK cells, were less frequent in RCC tumors. For some RCC tumor NK cells, these changes are transient and reversed by ex vivo culture in medium containing IL-2,^106,108 suggesting tumor-specific alterations.¹⁰⁸

Therapeutic use of NK cells in metastatic RCC was first attempted by infusion of lymphokine activated CD3⁻CD56⁺ cells (or LAKs) in combination with IL-2^109-111. Alternative approaches include adoptive transfer of unmodified or ex vivo activated allogeneic NK cells¹¹² and NK cell lines, such as NK-92.^113,114 Genetically modified NK cells engineered to express silencing RNA to inhibitory receptors or overexpress cytokines, activating receptors, or chimeric antigen receptor (CAR), are being studied for potential use in the clinic.

Tumor-Mediated Immune Suppression

High levels of tumor infiltrating lymphocytes (TIL) are common to RCC suggesting that immune mechanisms have a role in the natural course of disease.⁹⁹ Yet, aggressive tumors, like RCC, frequently adopt characteristics to evade immune surveillance. Immune escape and immune suppression are 2 of the main obstacles that need to be overcome when considering therapies to induce antitumor immune responses.¹¹⁵ Cancerous cells frequently downregulate MHC I to escape recognition by tumor-specific CD8⁺ T cells, but this still renders them susceptible to NK cell lysis.¹¹⁶ Tumors can also induce anergy of resident naïve T cells by presenting tumor-associated antigens in the absence of co-stimulation or by secreting suppressive molecules.¹¹⁷ Engagement of inhibitory receptors, such as CTLA-4 and PD-1, on T cells is an additional direct mechanism that tumor cells employ to suppress T cells.¹¹⁸ CTLA-4 and PD-1 are members of the CD28 family, but instead of activating the T cells, they induce anergy, and cells within the tumor milieu can express the ligands for these receptors.¹¹⁷ Development of reagents that specifically block the immunosuppressive nature of CTLA-4 and PD-1 has been one of the most significant advances in cancer immunotherapy, and blocking these “checkpoint inhibitors” has shown clinical efficacy in RCC.^119,120 Continued testing and development of these reagents will likely lead to even greater antitumor responses and improved patient survival. Tumor cells can also secrete immunosuppressive molecules such as IL-10 and TGF-β, which can inhibit proliferation and development of CTL. These molecules also likely recruit immunosuppressive cell populations into the tumor microenvironment, including the above-described dNK-like cells.

Myeloid Derived Suppressor Cells (MDSC)

Extramedullary hematopoiesis and neutrophilia were first described as characteristics of tumor progression in the early 1900s.¹²¹ These events were associated with atypical myeloid progenitor generation and differentiation resulting in abnormal myeloid cells that lacked conventional B cell, T cell, and macrophage lineage markers and were capable of decreasing CTL generation and function.¹²² In recent years these “tumor-induced” cells have been identified as myeloid derived suppressor cells (MDSC), which are present in high numbers and correlate with poor prognosis, and tumor evasion of host immunity.¹²³ MDSC are readily induced by inflammation and the presence of tumors.^124-127 Expansion and activation of MDSC can be induced by multiple factors made primarily by tumor/stromal cells and activated immune cells within the tumor microenvironment. These secreted factors can promote MDSC expansion by stimulating myelopoiesis, blocking differentiation of mature myeloid cells, and direct activation of MDSC. Granulocyte-macrophage colony-stimulating factor (GM-CSF), prostaglandin E2 (PGE₂),¹²⁸ and S100A^129-131 proteins are examples of molecules produced by tumor cells that induce the expansion and activation of MDSC both in the periphery (bone marrow, secondary lymphoid organs) and tumor. Molecules expressed by immune cells, such as activated T cells and myeloid cells, that can induce MDSC include IFN-γ¹²⁷, IL-6¹²⁴, IL-1β and cyclooxygenase (COX)-2/PGE₂.^128,132,133

MDSC that expand during a tumor challenge are not a defined population of cells; rather, they are a heterogeneous population of immature myeloid cells.¹³⁴ MDSC are poorly differentiated and lack differentiation markers expressed by other myeloid cells, such as mature DC and macrophages. Murine MDSC were historically characterized by expression of CD11b and Gr-1; however, it is now well appreciated that the Gr-1 antibodies used to identify MDSC bind to 2 epitopes from 2 different GPI-anchored cell surface glycoproteins, Ly6G and Ly6C.¹³⁵ Using differential antibodies to each Ly6G and Ly6C, 2 sub-populations of MDSC have been identified: CD11b⁺Ly6C^+/loLy6G^hi granulocytic MDSC, and CD11b⁺Ly6C^hiLy6G⁻ monocytic MDSC.^127,136,137 While other markers have been identified on MDSC, including F4/80, IL-4α receptor and CD115, these markers seem to be tumor model specific, and are not always constitutively expressed on MDSC. Though neutrophils and macrophages can express some of the same phenotypic markers as MDSC, MDSC specifically expand during a tumor challenge and demonstrate immunosuppressive capacities via a variety of mechanisms. Suppressive function is the main denominator that distinguishes MDSC from other myeloid cells, such as neutrophils and macrophage. MDSC have a multitude of mechanisms to inhibit the antitumor immune response, 2 of which are significant upregulation and activity of the Arg1 and iNOS enzymes.^138-141 These mechanisms induce non-specific T cell inhibition by catabolism of L-arginine from the surrounding environment into urea by Arg1 and nitric oxide (NO), by iNOS. Reactive oxygen species (ROS) produced by MDSC in response to cytokines have also been implemented in mediating immunosuppression. Recently, peroxynitrite has been shown to inhibit T cell activation by making them unresponsive to antigen stimulation.¹⁴² Many of these byproducts have direct effects on T cells by altering the CD3 and CD8 molecules on the surface of the T cell.

Though multiple MDSC subsets exist, they have all been linked to tumor outgrowth and poor prognosis in cancer patients. For these reasons, targeted therapy to deplete or modulate MDSC is currently the focus of many antitumor therapies. Over the last few years, therapeutic targeting of MDSC to reduce immune suppression for the purpose of mounting a robust antitumor immune response has become increasingly more common. Interestingly, many of the standard treatments for cancer may decrease or modulate the MDSC population in tumor-bearing hosts. Such chemotherapies as Gemcitabine and 5-Fluoruricil selectively deplete MDSC.^130,143 Small molecule inhibitors for the treatment of RCC, Sorafenib and Sunitnib, have the ability to decrease the suppressive capacity of MDSC in patients.^64,65 These agents are capable of alleviating immune suppression through altering the MDSC populations, but they possess off target effects and toxicities that can have negative effects on subsequent immune responses. Thus, immunotherapy strategies for overcoming tumor-associated MDSC immunosuppression that do not negatively affect global immune responses are of utmost importance.

Conclusions

Treatment options for RCC patients are poorly established and predominantly limited to surgical removal of either part or the entire kidney. While effective for individuals with localized early-stage disease, the majority of newly diagnosed patients will already have advanced (metastatic) disease at the time of diagnosis. The best opportunity for long-term survival is surgery in combination with therapies designed to enhance the body's natural immune responses. Unfortunately, the most advanced approaches are lacking efficacy; complete response rates remain <20%. One potential explanation for this shortfall is that therapeutic efficacy is limited by tumor-specific factors and/or cell types that inhibit natural and/or induced antitumor immune responses (Fig. 1). In addition, various risk factors (such as obesity, smoking, hypertension, renal disease, and viral hepatitis) can have profound effects on normal functioning of the immune system, which impacts tumor incidence/growth.^144-148 These points underscore efforts to achieve a more complete understanding of the mechanisms regulating RCC tumor immune suppression. Understanding these complex interactions has broad implications for successful application of immunotherapy for RCC.

Figure 1. Role for soluble RCC tumor products in reduced function of tumor infiltrating immune cells to promote tumor outgrowth and metastasis. Peripheral blood lymphocytes are recruited to the developing renal tumor (RCC) and their activity is impaired in the tumor environment when encountering soluble products. The effect is suppression and/or loss of cytotoxic capacity and conversion to a program that favors the production of angiogenic factors and cytokines that support tumor growth and promote dissemination (metastasis) to distant organ sites.

Disclosure of Potential Conflicts of Interest

No potential conflicts of interest were disclosed.

Authors' Contributions

All authors contributed to writing of the manuscript and approved the final version for submission. T.S.G. was responsible for organization of the effort.

Funding

This work was supported by grants from the National Institute of Child Health and Human Development (R15HD073868; D.S.T.) and National Cancer Institute (R15CA173657; D.S.T. and A.W.), (R01CA109446; T.S.G.). The funders had no role in preparation of the manuscript or decision to publish.