Influence of human immune cells on cancer: studies at the University of Colorado

Abstract

There will be over half a million cancer-related deaths in the United States in 2012, with lung cancer being the leader followed by prostate in men and breast in women. There is estimated to be more than one and a half million new cases of cancer in 2012, making the development of effective therapies a high priority. As tumor immunologists, we are interested in the development of immunotherapies because the immune response offers exquisite specificity and the potential to target tumor cells without harming normal cells. In this review, we highlight the current advances in the field of immunotherapy and the current work being completed by laboratories at University of Colorado School of Medicine in multiple malignancies, including breast cancer, lung cancer, melanoma, thyroid cancer, and glioblastoma. This work focuses on augmenting the anti-tumor response of CD8 T cells in the blood, lymph nodes, and tumors of patients, determining biomarkers for patients who are more likely to respond to immunotherapy, and identifying additional anti-tumor and immunosuppressive cells that influence the overall response to tumors. These collaborative efforts will identify mechanisms to improve immune function, which may elucidate therapeutic targets for clinical trials to improve patient health and survival.

Introduction

Cancer is a significant public health problem; a total of 1,638,910 new cancer cases and 577,190 deaths from cancer are projected to occur in the United States in 2012 [1]. One in four deaths in the United States is due to cancer; it is the second leading cause of death after heart disease. Although data between 1998 and 2008 indicate that cancer death rates have declined by more than 1 % per year in men and women of almost every racial and ethnic group [1], there is profound interest in developing successful therapies to augment this decline.

The immune system and inflammation contribute to tumor elimination and progression. In fact, Hanahan and Weinberg highlight both “evading immune destruction” and “tumor-promoting inflammation” in their 2011 Hallmarks of Cancer [2]. However, traditional therapeutic drug development focused solely on the tumor itself, not on the other components of the tumor microenvironment. Although it was unknown when they were developed, many common treatments alter the immune system and its microenvironment. More recently, the field of immunotherapy has broadened to target immune infiltrates and evasion mechanisms. The ultimate goal of these therapies is to shift the balance from tumor escape to tumor destruction.

Cytotoxic T cells (CTL) have long been the focus of tumor immunotherapy. The first immunotherapies generated a non-specific increase in T-cell proliferation with the stimulatory cytokines IL-2 [3] or IFN-α [4]. Later, specific antibody therapies were developed that targeted the activating molecules expressed on T cells (41BB, OX40). Recently, an antibody against cytotoxic lymphocyte activation gene 4, CTLA-4, was approved by the FDA for metastatic melanoma [5]. Patients with different cancers in a phase I clinical trial of a programmed cell death-1, PD-1, antibody also showed evidence of improved anti-tumor activity [6]. These non-specific immunotherapies have shown response rates in clinical trials ranging from 5 to 20 % (reviewed in [7, 8]).

A strong correlation between tumor-infiltrating lymphocytes (TILs) and patient survival is reported in many cancers (Table 1). Other targeted immunotherapies for the stimulation of CD8 TILs include vaccines for identified tumor antigens and various adjuvants [9, 10] or the adoptive transfer of dendritic cells differentiated in vitro in the presence of tumor antigens [11]. Increased clinical response and improved production methods for the expansion of TILs from patients are leading to the use of adoptive transfer of TILs as a cancer treatment. Several immunotherapies that activate CD8 TIL responses against the tumor have demonstrated clinical benefit or even improved survival (Table 2).

Table 1.

CD8 tumor-infiltrating lymphocytes correlate with patient survival

Cancer type	CD8 TIL counts	Patient survival (months)	Correlation	No. of patients	References
Colorectal	100–400 cells per mm2	10–35	Lower cell density correlates with tumor recurrence	415	[113]
Metastatic melanoma	35 % of total T cells in tumor	23–130	Decrease in survival when CD8 TILs are around tumor versus throughout tumor	147	[114]
Glioblastoma	1–10 cells per mm2	30–40	Increase in CD8 TILs correlates with increase in malignancy	93	[115]
Lung	0.4–45 cells per mm2 Median = 10.8 cells	240 (20–40 % survival)	Higher survival in patients with CD8 TILs >median	99	[116]
Ovarian	0–89.5 cells per field Median = 7.60 cells	25.5–54.9	Higher survival correlates with higher intraepithelial CD8 TILs	117	[117]
Breast	90–557 cells per field	240 (50–70 % survival)	Higher survival in patients with CD8 TILs and correlation with grade	1334	[118]

Table 2.

Immunotherapy in the treatment for human cancer: a summary of recent successful clinical trials

Cancer type	Trial phase	Compound name	Treatment description	No. of patients	Outcome	References
Numerous	I	MKC1106-PP-PRAME and PSMA antigens	Intralymph node injection of DNA & peptide vaccine	26	7 patients with stable disease at 6 months	[119]
Numerous	I	MDX-1106-I.V. MDX-1106	I.V. MDX-1106	39	1 CR and 2 PR at 3 months	[6]
Non-small cell lung cancer	IIB	L-BLP25-MUC1 antigen	I.V. cyclohexamide and multiple S.C. vaccine injections	171	24 % (L-BLP25 + BSC) v. 12 % (BSC) survival at 3 years	[120]
Non-small cell lung cancer	I	Chemotherapy/CEA peptide-pulsed DC/cytokine-induced killer cells	Chemo followed by multiple I.V. injections of cells	28	Median progression-free survival 6.9 months v. 5.2 months (chemo only)	[121]
Glioblastoma multiforme	I/II	DC vaccine with autologous tumor lystates	Multiple S.C. injections near lymph node	17	37.5 % (vaccine) v. 3.2 % (historical ctrl) survival at 3 years	[122]
Melanoma	II	Lymphodepletion, autologous TIL, and high-dose IL-2	Infusion of TIL expanded in culture	93	36 % survival at 3 years, 29 % survival at 5 years	[123]
Melanoma	III	Ipilimumab-anti-CTLA- 4 ± gp100 peptide vaccine	I.V. ipilimumab ± S.C. gp100 in IFA	676	Median overall survival 10.1 months v. 6.4 months (gp100 alone)	[5]
Melanoma	III	gp100 peptide and high-dose IL-2	I.V. injection of IL-2 and S.C. injection of peptide in IFA	185	17.8 months v. 11.1 months overall survival (IL-2 only)	[10]
Renal cell carcinoma	I/II	GM-CSF, IL-2, and IFN-γ	S.C. injection of cytokines	60	Median progression-free survival 6.0 months	[124]
Prostate cancer	III	Sipuleucel-T-GM-CSF/PAP antigen fusion protein	I.V. injection of activated PBMC	512	31.7 v. 23 % (placebo) survival at 3 years	[11]

PRAME preferentially expressed antigen in melanoma, PSMA prostate-specific membrane antigen, MUC1 Mucin 1, IV intravenous, SC subcutaneous, V versus, BSC best supportive care, DC dendritic cell, TIL tumor-infiltrating lymphocytes, GM-CSF granulocyte-macrophage colony-stimulating factor, IL-2 interleukin 2, IFN-γ interferon gamma, IFA incomplete Freund’s adjuvant, PAP prostatic acid phosphatase, PBMC peripheral blood mononuclear cells, CR complete response, PR partial response

Although these promising new therapies significantly improve overall survival of cancer patients and demonstrate that manipulation of the immune system can lead to successful cancer treatments, they rarely establish durable cures for cancer in most patients. One hypothesis for these inconsistent responses is that increased regulatory mechanisms in cancer patients overcome active immunity and may also need to be targeted for successful treatment of cancer. In many patients, regulatory T cells (T_reg) create an immunosuppressive environment within the tumor and throughout the periphery. Human T_reg express the forkhead box transcription factor (FOXP3) [12] and suppress autoreactive T cells. T_reg are detected in TILs of many human solid tumors [13–16], and a decreased ratio between CD8 TILs and T_reg in the microenvironment of the tumor correlates with poor survival [17, 18]. T_reg suppress the function of tumor-specific T cells by secretion of the immunosuppressive cytokines TGF-β and IL-10 and engaging suppressive cell surface molecules (i.e., CTLA-4-CD80/86) [19].

Myeloid-derived suppressor cells (MDSCs) are also associated with immunosuppression in multiple cancers. MDSCs are potent inactivators of both CD4 and CD8 T cells. The mechanisms that MDSCs use to induce immunosuppression involve activity of amino acid-metabolizing enzymes (such as arginase I and nitric oxide synthase) and production of reactive oxygen species, which have debilitating effects on the T-cell receptor (TCR) [20].

Due to challenges in translating promising research discoveries into human cancer treatments, the use of immunotherapies in cancer patients has not yet reached its potential. However, recent successes described in this review and the current trend toward combinatorial therapies suggest that the immune system will ultimately play a pivotal role in future cancer treatments. This review highlights the progress and objectives of the clinical tumor immunology research being conducted at the University of Colorado School of Medicine (UCSOM) in the context of the rest of the field.

Activation of TILs from human solid tumors (Slansky laboratory)

Several studies demonstrate functional inertness of CD8 TILs in solid tumors [21–24]. One example is the CD8 TILs in prostate cancer [25, 26]. These CD8 TILs are antigen-experienced (CD45RO+) and do not proliferate when removed from the tumor microenvironment ([25], Bruno, unpublished). In addition, they are refractory to stimulation through either the TCR or downstream signaling pathways via PHA, or PMA and ionomycin ([25], Bruno, unpublished). One reason that the T cells may be tolerized is that they are desensitized due to constant exposure to self-antigen; removal from this stimulation does not fully restore function. However, these CD8 TILs do contain perforin granules and produce cytokines, which suggests some degree of prior stimulation ([25], Bruno, unpublished). The CD8 TILs may have been activated by matured antigen presenting cells (APCs) because of the inflammatory cues present in the prostate. Thus, intrinsic and extrinsic factors contribute to functional unresponsiveness or incomplete activation of the CD8 TILs.

In multiple studies, the presence of CD8 TILs correlates with improved patient survival in various cancer types (Table 1). In human breast cancer, it has been reported that 11–57 % of tumors are infiltrated with CD8 TILs, which display an activated phenotype [27, 28]. Unlike prostate CD8 TILs, the Slansky laboratory has shown that CD8 TILs in human breast cancer tumors are functional when removed from the tumor microenvironment (Fig. 1). The laboratory is studying breast cancer CD8 TILs to determine the best way to enhance their cytolytic activity against tumors. One approach is to identify antigens that trigger a robust clonal proliferation of these CD8 TILs. To test methods of enhancing CTL activity of breast cancer TILs, we developed an effective protocol for isolating these cells from tumor tissue (Fig. 2). Determining the anti-tumor function and molecular characterization of these CD8 TILS will aid in the development of targeted immunotherapy for these patients. We ultimately envision an immunotherapy that targets the appropriate breast cancer antigen in combination with a blockade of one or multiple immune checkpoints, such as PD-1.

Fig. 1.

Breast cancer CD8 TILs proliferate (Ki-67) and produce IFN-γ in response to TCR stimulation. CD8 T cells were isolated from total PBMC post-Ficoll separation (left). CD8 TILs were isolated according to the protocol in Fig. 2 (right). Cells were stimulated postselection for 5 days with αCD3αCD28. On day five, the cells were restimulated with PMA, ionomycin, and brefeldin-A for 6 h, stained intracellularly for Ki-67 and IFN-γ, and analyzed by flow cytometry

Fig. 2.

TIL isolation protocol

B cells also infiltrate cancers. Tumor-infiltrating B cells (TIL-Bs) are present in about 25 % of breast cancers and comprise up to 40 % of the TIL population [29–31]. Detection of these cells correlates with positive survival rates in medullary breast cancer [30, 32, 33]; however, the function of these cells is unclear. There are several hypotheses offering explanation as to why these B cells are prevalent in tumors based on studies done in autoimmunity. Specifically, B cells are antigen-presenting cells and collaborate with T cells to generate immune responses with significant tissue damage. The B cells often enhance T-cell responses by producing antibodies, stimulatory cytokines, and chemokines and serve as a local antigen-presenting cell. Together, B cells and T cells often generate lymphoid structures at the site of the autoimmune reaction [34]. It has been hypothesized that B cells help generate a potent, long-term immune response against cancer. Although some B cells may amplify the immune response, regulatory B cells suppress the immune response via IL-10 and TGF-β-1 [35]. While these correlative data are interesting, there is not much known about the mechanism of TIL-Bs in cancer. Thus, we are also studying TIL-Bs to determine their function.

Our first analyses of TIL-Bs are in non-small cell lung cancer (NSCLC) patients. Like breast cancer, TIL-Bs correlate with positive survival in NSCLC patients [36, 37]. In collaboration with Dr. Jeffrey Kern, director of the Cancer Research Center at National Jewish Health, we obtain lung tumor and tumor-adjacent samples from primary NSCLC adenocarcinomas that undergo surgical resection at University of Colorado. We phenotype the memory and activation profile of TIL-Bs and dissect the function by isolating TIL-Bs from the tissue (Fig. 2) and monitoring proliferation and cytokine production. With a better understanding of the function of TIL-Bs, we hope to introduce them as a potential therapeutic target.

Immunotherapeutic biomarkers in the treatment for human melanoma (McCarter laboratory)

The recent successes in immunotherapy for cancer have largely been a result of research in melanoma patients (reviewed in [7]). Whether due to the antigens expressed by melanoma cells, the identification of these specific antigens, or the nature of the tumors themselves, melanoma uniquely generates detectable tumor antigen–specific T-cell responses [38]. In addition, melanoma is generally resistant to conventional chemotherapy and radiation, making it an ideal disease for studies of experimental immune therapies [39]. Several studies demonstrate that cells and molecules of the immune system correlate with survival and prognosis in melanoma patients; however, therapies designed to activate the immune system, such as IL-2 and IFN-α, rarely lead to long-term productive responses. To improve these treatments, there is a need for a better understanding of the immune response in treated patients and for biomarkers that predict which patients will respond.

A recent success in the treatment for melanoma is with the drug ipilimumab, an antibody therapy that targets CTLA-4. Although studies in knockout mice suggest that CTLA-4 controls effector T-cell function [40], CTLA-4 is constitutively expressed by T_reg, which makes them a target of anti-CTLA-4 therapy [41]. One study found no differences in the number of peripheral T_reg after anti-CTLA-4 treatment and no differences between patients that responded to the therapy and those that did not respond [42]. Others found that ipilimumab did not directly inhibit T_reg function in vitro and that post-treatment T_reg maintain suppressive function [43]. However, preclinical data suggest that blocking CTLA-4 on both effector and regulatory T cells is required for full benefit [44]. In stage IV melanoma patients, anti-CTLA-4 therapy increases the 1-year survival rate from 25 to 45 % and significantly extends overall survival [5, 45]. Further investigations into these mechanisms will identify why only some patients respond to this treatment.

An important lesson learned from the early clinical trials of anti-CTLA-4 treatment is that responses to drugs targeting the immune system are different from chemotherapy and radiation that directly and quickly kill tumor cells. Immunotherapies may result in an initial increase in tumor size, or they may take longer to generate a T-cell response required to decrease the tumor. Therefore, patients that initially fail to respond to these therapies using traditional Response Evaluation Criteria in Solid Tumors (RECIST) measurements may respond outside of the time limits imposed by the research study or may demonstrate an overall survival benefit in the absence of tumor shrinkage. Although not yet widely adopted, an important advance in immunotherapy was the development of new evaluation criteria, termed immune-related response criteria (irRC), which may prevent immune therapies from prematurely failing in clinical trials [46]. However, other response indicators, independent of tumor size, are actively under investigation and may be important for monitoring responses to immunotherapy. Several studies have found that changes in absolute lymphocyte and/or CD8 T-cell counts in the peripheral blood, particularly the rate at which these cell counts increase, correlate with positive responses to anti-CTLA-4 treatment [47]. A post-treatment increase in the expression of inducible costimulator (ICOS) on CD4 T cells correlates with overall survival [48]. Finally, one study found that CD4 memory T cells from patients responding to anti-CTLA-4 treatment develop resistance to suppression by T_reg, while cells from non-responding patients maintain sensitivity to T_reg inhibitory mechanisms [49].

Molecular biomarkers are currently in use for selecting targeted therapies in melanoma patients. For example, tumors expressing a BRAF mutation (40–60 % of all cutaneous melanoma) are more likely to respond to BRAF inhibitors [50, 51]. These types of mutational screens also help determine the mechanism of escape from drugs such as BRAF inhibitors, in which activation (or mutation) of other molecules in the MAPK pathway drives resistance [52]. In contrast, immune biomarkers are still exploratory, not standardized, and not currently directing patient care. Recent studies suggest that patients with increases in certain serum cytokines, such as IL-1α/β, TNF-α, IL-6, MIP-1α/β, are more likely to respond to IFN-α treatment [53]. Increased levels of serum C-reactive protein, fibronectin, and VEGF correlate with poor responses to IL-2 therapy [54, 55]. Finally, increased levels of intratumor T_reg, the T_reg-associated transcription factor FOXP3, and the T_reg-associated enzyme indoleamine 2,3-dioxygenase (IDO) correlate with better responses to anti-CTLA-4 treatment [56]. These results suggest that preexisting immune responses against the tumor may predict responses to anti-CTLA-4 therapy ([57]; reviewed in [58]). With an understanding of the factors that distinguish responders from non-responders, immune biomarkers may help guide patient care and predict treatment outcomes.

In the McCarter laboratory, we are investigating several potential biomarkers of responses to ipilimumab. Using peripheral blood from stage IV melanoma patients, the number of T cells, expression of ICOS and PD-1 on T cells, T_reg suppressive function, and plasma cytokine levels will be determined before and after treatment with 10 mg/kg ipilimumab. In agreement with previous studies [13, 59–61], the number of T_reg in the peripheral blood of untreated stage IV melanoma patients is increased relative to normal donors, and the number of T_reg in these patients correlates with the amount of TGF-β in the plasma (unpublished data). Important for this study and others, melanoma patients at UCSOM are also recruited to donate tumor tissue and serum samples to the Melanoma Tumor Bank, so that biomarker studies can be performed (Dr. William Robinson, Director). We are using live tumor cells isolated from tissue and human melanoma tumor lines from this bank for in vitro investigational studies to gain insight into identifying immunologic biomarkers that will help select patients that are most likely to benefit from specific immunotherapies.

Immune modulation in metastatic papillary thyroid cancer (French and Haugen laboratory)

Papillary thyroid cancer (PTC) commonly develops in patients with autoimmune thyroiditis. This association has led to much speculation about the role of inflammation in tumorigenesis and cancer progression [62]. Until recently, studies investigating the role of the immune system in PTC failed to distinguish between autoimmune thyroiditis and the tumor-directed immune response [63, 64]. Furthermore, these studies failed to fully characterize the types of leukocytes present in PTC. In a retrospective study, we assessed archived formalin-fixed paraffin-embedded primary thyroid tumors from PTC patients for evidence of lymphocytic infiltration. Patient samples were further distinguished based on the evidence of lymphocytic infiltration in normal thyroid tissue, a sign of autoimmune thyroiditis. While no significant difference was observed between patients without lymphocytic infiltration and those with concurrent autoimmune thyroiditis, patients with tumor-associated lymphocytes (TALs), from lymph nodes, displayed more severe disease [65]. T_reg were detected among TALs, and increased T_reg frequency correlated with a higher percentage of lymph node (LN) metastases [65]. These studies suggest that immune modulation in patients with PTC may play an important role in disease progression.

PTC is distinct from other types of cancer in that persistent metastatic disease is commonly confined to loco-regional LN. Analyses of central neck lymph node dissection (CNLD) specimens revealed that more than 60 % of patients with PTC develop LN metastases [66]. A total of 20–30 % of patients who have undergone standard therapies (i.e., primary thyroidectomy, CNLD, and radioactive iodine therapy) develop recurrent disease, most commonly in the locoregional LN [67, 68]. Metastatic LN in patients with PTC provide a unique model in which to study the interplay between the immune system and tumor. Our ongoing studies rely on postsurgical patient samples obtained by fine-needle biopsy from uninvolved (UILN) and tumor-involved lymph nodes (TILN). Live lymphocytes recovered from these samples are characterized using multicolor flow cytometry to assess cell surface activation markers, cytokine receptors, and intracellular expression of cytokines and transcription factors. These studies revealed that T_reg are enriched in TILN compared with UILN and their frequency is even further increased in TILN from patients with recurrent metastatic disease [69]. In line with these findings, CD8 T cells expressing PD-1, a marker of T-cell exhaustion, are enriched in TILN [69–71]. Furthermore, high frequencies of PD-1⁺ CD8 T cells in TILN are associated with extranodal invasion [69]. A majority of PD-1⁺ CD8 T cells in TILN from patients with PTC are CD45RA⁻, suggesting recent antigen exposure; however, these cells fail to downregulate CD27 and are not actively proliferating [69]. Of note, the frequencies of T_reg and PD-1⁺ CD8 T cells correlate directly in TILN [69]. Additional studies are necessary to determine the function of T_reg and PD-1⁺ CD8 T cells, respectively, in PTC.

Most patients with PTC have an excellent prognosis following standard therapies (97 % 5-year survival rate) [67]. However, patients with recurrent LN disease and distant metastatic lesions will benefit from innovative prognostic markers and adjuvant therapies. We are currently investigating the potential of T_reg and PD-1⁺ CD8 T-cell frequencies in TILN as prognostic markers for patients with recurrent lymph node disease. Such markers will aid in determining the best therapeutic approach for each patient (i.e., compartment LN excision vs. focused LN dissection vs. radioiodine therapy vs. monitoring). Furthermore, regional or systemic immune-based therapies that target T_reg and/or PD-1⁺ CD8 T cells may be viable options for patients with recurrent or advanced PTC that cannot be cured with surgery or standard therapies.

Identifying host factors that influence immune suppression in breast cancer (Borges laboratory)

Breast cancer is typically considered a disease of aging, as the dominant portion of cases is diagnosed in postmenopausal women. Similarly, the well-known risk factors for breast cancer are tightly tied to cumulative lifetime estrogen exposure, factors such as menarche, parity, late menopause, and exogenous hormone use, which only apply well to postmenopausal women. However, 25,000 cases of breast cancer are diagnosed each year in women under age 45 whose risk factors for the development of breast cancer are poorly understood. Diagnosis under age 40 is also associated with a significant increase in risk of metastasis and death for reasons that are not explained by known prognostic criteria [72–74].

Breast cancer has multiple subtypes that display distinct prognosis and potential for treatment response. These subtypes are based on genomic data, but are clinically identifiable through the expression of the estrogen receptor, progesterone receptor, and HER2/neu oncoprotein on tumor cells [75]. The nomenclature of these subtypes currently are Luminal A (ER+, PR+, HER2−), Luminal B (ER+ and either PR− or PR+ and HER2+), HER2 (ER−, PR− and HER2+), and triple negative (ER−, PR− and HER2−). To date, immunotherapy in breast cancer has been successful through targeting of the HER2 pathway. The monoclonal antibody trastuzumab has been standard of care for advanced breast cancer since 1998 and early stage breast cancer since 2005. HER2-targeted vaccine treatment also demonstrates very promising results in clinical trials [76, 77]. In our Young Women’s Breast Cancer Translational Program, we are studying the interaction of young age at diagnosis with the different biologic breast cancer subtypes. We are identifying the differences in the immune response to breast cancer depending on patient age and outcome. We are currently focused on the characterization of MDSCs and their function in newly diagnosed young women’s breast cancer, as well as potential mechanisms of reversal of their function that would be relevant in the treatment for breast cancer.

While MDSCs have been well defined in murine models, a clinical definition in humans remains more elusive. As recently reviewed by Montero et al., there are discrepancies over the phenotype of these cells between cancer types, with one of the only commonalities being their ability to suppress T-cell function, though through different mechanisms. In early descriptions, MDSCs were defined as a heterogeneous population made up of immature myeloid cells [78–83]. There have also been studies that show more differentiated MDSCs, most commonly expressing CD11b and CD33. These mature MDSCs can be broken down into two groups based on the markers CD14 and CD15. Granulocytic MDSCs tend to be CD15+ CD14− [84–89], while monocytic MDSCs are CD15− CD14+ [90–94]. To date, cancer patients, including breast cancer patients, have increased numbers of MDSCs intratumorally and in the blood [95]. In breast cancer, the number of MDSCs correlates with disease stage and T-cell suppressive function as well as adverse patient outcomes [96].

Through the identification of MDSCs and their suppressive function in newly diagnosed early-stage young women’s breast cancer, we hope to define their potential as a prognostic marker and potential immunotherapeutic target. As part of a larger prospective trial of stage I–IV newly diagnosed breast cancer patients and age-matched normal donors, we have quantitated the presence of MDSCs and their function in T-cell suppression in the peripheral blood (unpublished data). We are currently testing several potential drugs that may disrupt MDSC suppression with the goal of identifying adjunct treatment to improve standard breast cancer therapies.

Neutrophils and immunosuppression in glioblastoma (Waziri laboratory)

Glioblastoma (GBM) patients have suppression of cellular immunity within the tumor microenvironment as well as the systemic circulation, secondary to functional defects in both APCs and T cells (reviewed in [97–99]). Despite these defects, tumor antigen–specific T cells can be found within the circulation of affected patients [100]. This observation implies that tumor-specific immunity is blunted by tumor-mediated suppression of T-cell function, a consideration that is likely relevant to the success of immunotherapy. Therefore, targeting tumor-associated immunosuppression in patients with GBM will be critical for the development of meaningful immunotherapeutic strategies.

In Dr. Waziri’s laboratory, we have recently demonstrated that myeloid cells from GBM patients suppress normal donor T-cell function within a mixed lymphocyte reaction [101]. Through expression analysis of a range of classical myeloid lineage surface markers, we identified an expanded population of CD33lo cells within peripheral blood mononuclear cells (PBMC) from GBM patients that were not present in normal donors, or to a lesser extent, patients with other intracranial tumors. Further examination of the CD33lo population confirmed that these cells express the characteristic neutrophil markers CD15 and CD66 and exhibit the histological appearance of mature neutrophils. This finding was intriguing, as neutrophils are not normally found within PBMC following Ficoll density centrifugation, but rather separate with the flow-through fraction. The shift to the PBMC fraction suggested decreased density when compared to normal neutrophils, a phenomenon that could be induced through activation and degranulation. We confirmed this hypothesis by stimulating normal donor neutrophils within whole blood with formyl-methionyl-leucyl-phenylalanine (FMLP). Analysis of the PBMC from FMLP-stimulated whole blood confirmed the presence of increased numbers of neutrophils.

One of the major constituents of neutrophilic granules is the immunosuppressive enzyme arginase I (ArgI) [101], which is released upon neutrophil degranulation [102] and induces T-cell dysfunction through a well-described mechanism [103, 104]. To extend upon our prior phenotypic data suggesting the presence of activated neutrophils within the circulation of GBM patients, we provided confirmatory functional evidence by documenting increased levels of ArgI within in vitro immunofunctional assays and serum samples harvested from GBM patients. Critically, by targeting ArgI within in vitro immunofunctional assays, GBM patient T-cell function could be restored to the level of normal donors. As a direct result of this study, we have initiated a clinical trial to test the efficacy of oral L-argi-nine supplementation for improving T-cell function in newly diagnosed GBM patients. Through this work, we hope to confirm the role of targeting neutrophil-mediated immunosuppression as an adjuvant strategy to improve immunotherapy in GBM patients.

We continue to explore additional aspects of the biologic interactions between tumor and neutrophils. In general, neutrophils are not believed to release their granules outside the confines of inflamed tissue or move out of the tissue to reenter circulation once activated. Therefore, the presence of activated neutrophils within the peripheral circulation of GBM patients is of unclear etiology. Limited prior data suggests that activated neutrophils may exist in the circulation of other tumor patients [86, 87] although the mechanism responsible for this phenomenon remains to be described. Our initial comparative studies focusing on expression patterns of circulating activated neutrophils and intratumoral neutrophils have identified altered levels of several adhesion molecules necessary for transmigration on the circulating activated population (Sippel et al., manuscript in preparation). Further exploration of this mechanism may identify new targets for preventing the formation of activated neutrophils in circulation and the resulting suppression of cellular immunity.

In addition to a potential immunosuppressive role, neutrophils may also exert other significant effects (either advantageous or detrimental) on tumor growth [105–107]. Neutrophils may promote tumor growth by initiating angiogenesis, aiding in invasion and metastasis, and enhancing proliferative activity. In contrast, they have been associated with anti-tumor effects through direct cytotoxicity as well as promotion of the inflammatory response. We and others have observed the presence of neutrophils within areas of active necrosis in GBM [101, 108] although the overarching effects of these cells remain indeterminate. Interestingly, neutrophils have been associated with increased necrosis, ischemic tissue, and edema in brain injury and stroke [109, 110]. As necrosis is a distinguishing feature in GBM [111] and edema significantly contributes to morbidity in these patients [112], it is possible that neutrophils may be intimately involved in these processes in GBM. Our ongoing work continues to focus upon the involvement of neutrophils in GBM, with the hope of identifying new therapeutic targets for this disease.

Conclusion

While important strides have been made to improve detection and treatment, thousands of patients succumb to cancer. The tumor’s ability to escape immune destruction remains a constant variable that may be central to the host’s inability to eradicate the transformed cells. To this end, improving the immune system’s response to eliminate these malignant cells has become the focus of many research laboratories. In this review, we have focused on the recent efforts of multiple laboratories at UCSOM and in the field to improve immune function and identify targetable immunosuppressive mechanisms in multiple malignancies including breast cancer, lung cancer, melanoma, thyroid cancer, and glioblastoma.

The ongoing, collaborative efforts of these laboratories focus on mechanisms to improve CD8 TIL responses, determine biomarkers for patients who are more likely to respond to immunotherapy, and identify cells within the tumor, surrounding lymph nodes and peripheral blood that are immunosuppressive. These efforts will identify mechanisms to improve immune function, which may elucidate therapeutic targets that could directly translate to the design of clinical trials and directly impact patient health and overall survival.