Volume 247, Issue 5 p. 589-605
Invited Review
Free Access

The innate immune architecture of lung tumors and its implication in disease progression

Simon Milette

Simon Milette

Department of Medicine, Division of Experimental Medicine, McGill University, Montreal, Canada

Rosalind and Morris Goodman Cancer Research Centre, McGill University, Montreal, Canada

Search for more papers by this author
Pierre O Fiset

Pierre O Fiset

Department of Pathology, Faculty of Medicine, McGill University, Montreal, Canada

Search for more papers by this author
Logan A Walsh

Corresponding Author

Logan A Walsh

Rosalind and Morris Goodman Cancer Research Centre, McGill University, Montreal, Canada

Department of Human Genetics, Faculty of Medicine, McGill University, Montreal, Canada

Correspondence to: Daniela Quail and Logan Walsh, Rosalind and Morris Goodman Cancer Research Centre, 1160 Pine Ave. West, Montreal, QC H3A 1A3, Canada. E-mail: [email protected], [email protected]; Or Jonathan Spicer, Montreal General Hospital, 1650 Cedar Ave. Montreal, QC H3G 1A4, Canada. E-mail: [email protected]

Search for more papers by this author
Jonathan D Spicer

Corresponding Author

Jonathan D Spicer

Department of Medicine, Division of Experimental Medicine, McGill University, Montreal, Canada

Rosalind and Morris Goodman Cancer Research Centre, McGill University, Montreal, Canada

Department of Surgery, McGill University Health Center, Montreal, Canada

Correspondence to: Daniela Quail and Logan Walsh, Rosalind and Morris Goodman Cancer Research Centre, 1160 Pine Ave. West, Montreal, QC H3A 1A3, Canada. E-mail: [email protected], [email protected]; Or Jonathan Spicer, Montreal General Hospital, 1650 Cedar Ave. Montreal, QC H3G 1A4, Canada. E-mail: [email protected]

Search for more papers by this author
Daniela F Quail

Corresponding Author

Daniela F Quail

Department of Medicine, Division of Experimental Medicine, McGill University, Montreal, Canada

Rosalind and Morris Goodman Cancer Research Centre, McGill University, Montreal, Canada

Department of Physiology, Faculty of Medicine, McGill University, Montreal, Canada

Correspondence to: Daniela Quail and Logan Walsh, Rosalind and Morris Goodman Cancer Research Centre, 1160 Pine Ave. West, Montreal, QC H3A 1A3, Canada. E-mail: [email protected], [email protected]; Or Jonathan Spicer, Montreal General Hospital, 1650 Cedar Ave. Montreal, QC H3G 1A4, Canada. E-mail: [email protected]

Search for more papers by this author
First published: 25 January 2019
Citations: 31
Conflict of interest statement: JDS has received consulting honoraria from Bristol-Myers-Squib and AstraZeneca. He is also a Principal Investigator for clinical trials from Bristol-Myers-Squibb, Merck and AstraZeneca. POF has participated as a consultant on advisory boards for AstraZeneca Canada, Roche Canada, Merck Canada and Pfizer Canada. The other authors declare no conflict of interest.

Abstract

Lung malignancies are the leading cause of cancer-related mortality. By virtue of its unique physiological function, the lung microenvironment is highly dynamic and constantly subjected to mechanical, chemical and pathogenic stimuli. Thus, the airways rely on highly organized innate defense mechanisms to rapidly protect against pathogens and maintain pulmonary homeostasis. However, in the context of lung malignancy, these defenses often provide collateral inflammatory insults that can foster tumor progression. This review summarizes the interactions between cancer cells, recruited immune cells and tissue-resident cell subpopulations, such as airway epithelial cells and alveolar macrophages, during homeostasis and disease. Furthermore, we examine the role of the lung immune landscape in response to current therapeutic interventions for cancer. Given the prevalence of lung malignancies, we propose that consideration of lung physiology as a whole is necessary to understand and treat these lethal diseases. Copyright © 2019 Pathological Society of Great Britain and Ireland. Published by John Wiley & Sons, Ltd.

Introduction

Lung malignancies, including primary tumors and metastases, are the most frequently diagnosed and leading cause of cancer-related deaths 1. As with many cancer types, molecularly targeted therapies for lung malignancies have led to minimal improvements in patient survival due to the emergence of acquired resistance. However, recent progress with immune checkpoint inhibitors (ICIs) have highlighted the clinical potential of harnessing the tumor immune microenvironment (TIME) to combat disease. Interestingly, ICIs have been more effective in lung cancer compared with many other malignancies 2, suggesting that the unique immune network of the lung can be reprogrammed to foster therapeutic response. A better understanding of the pulmonary immune landscape may therefore provide clinically relevant insight into the therapeutic options for a vast number of patients with lung cancers.

The lung microenvironment is composed of a unique array of cell types, including airway epithelial cells (AECs), interstitial structural cells, endothelial cells and various inflammatory cells, such as macrophages, neutrophils, and lymphocytes 3-5. The lungs are also host to commensal microorganisms and exposed to exogenous inorganic and organic particulates during respiration 6. Under normal physiologic conditions, the lung is thus accustomed to frequent antigenic insults that must be overcome to maintain tissue homeostasis. This is particularly true for the innate immune compartment, which is constantly on high-alert for pathogen exposure. However, in cancer, the lung TIME becomes maladaptive as tissue homeostasis is chronically disrupted. This can either prevent or promote tumor progression, and the balance between these polarizing effects determines disease outcomes and therapeutic efficacy.

Several excellent reviews have broadly covered the contexture of the lung tumor microenvironment, including both immune and nonimmune compartments 7-9. Here we focus on the multifaceted functions of the lung TIME during cancer progression, with emphasis on components of the innate immune system that have more recently emerged in the context of lung malignancy. As the first responder to pathological stimuli, the innate immune system regulates the kinetics of early anti-tumor responses, which significantly impact the evolution of malignant disease in later stages. We focus on primary lung tumors including both small cell lung cancer (SCLC) and non-SCLC (NSCLC; ∼85% of all primary lung cancer cases), as well as lung metastases from various anatomical sites (e.g. breast and colorectal cancer). We first introduce some of the key cellular players that are critical to normal lung physiology and innate responses to inflammatory cues, and then discuss how these cell types may influence different aspects of neoplastic progression in the lung. Our goal is to shed light on how lung physiology as a multi-cellular, coordinated system can regulate pulmonary malignancy, and to discuss new potential targets in the TIME that could be harnessed to improve patient survival.

The unique innate immune machinery of the lungs

The lung microenvironment is at the interface between the environment and the host, as it extracts oxygen from the atmosphere and delivers it to the circulation. Through this process, the lung is constantly exposed to inflammatory stimuli, such as environmental particulate matter, tobacco and other forms of recreational smoke or inhaled products, viruses, fungi and bacteria. As such, the pulmonary immune machinery has a remarkable ability to rapidly defend against potentially harmful environmental exposures. Continuous exposure of the respiratory tract to airborne antigens, extreme fluctuations in pressure and variable pH, requires complex and highly coordinated defense mechanisms that rely heavily on the innate immune system 10. These innate processes involve both cellular and humoral mechanisms of immunity, which work rapidly to maintain tissue homeostasis. In the following section, we will briefly introduce some of the key cellular immunological barriers within the lung and how they work together to maintain tissue homeostasis. We will later discuss how these players are relevant to lung tumorigenesis and progression.

AECs mediate innate immunity

AECs act as a first line of defense against antigens that have escaped mucociliary clearance. Single cell-RNA sequencing studies have shown that AECs can be subdivided into multiple subtypes 3, 5. The best characterized subtypes include squamous type I and cuboidal type II cells, which are responsible for gas exchange with the endothelium and production of pulmonary surfactant, respectively 11. Being endowed with an arsenal of pattern recognition receptors (PRRs) such as Toll-like receptors (TLRs), Nod-like receptors and retinoic acid-inducible gene (RIG)-I-like receptors, AECs promptly sense pathogen-associated molecules and initiate NF-κB-dependent inflammatory cascades through the release of antimicrobial peptides, cytokines and chemokines 11. Of particular importance are microbial ssRNA- and CpGDNA-sensing TLR-3, TLR-7, TLR-9, RIG-I and melanoma differentiation-associated protein-5, which have been implicated in the rapid release of interferon (IFN)-β, granulocyte-macrophage colony-stimulating factor (GM-CSF), interleukin (IL)-6 and IL-8, a potent neutrophil-recruiting chemokine 12. This robust cytokine response is critical to subsequently stimulate adaptive immunity, and neutralize the harmful effects of foreign pathogens.

However, owing to the hypersensitivity of AEC PRRs, these receptors can sometimes become diverted to mediate autoimmunity. For example, in mouse models of allergic asthma, activation of TLR-4 by dust mites on AECs is associated with local production of IL-25, IL-33, GM-CSF and thymic stromal lymphopoietin (TSLP) 13. These cytokines work in concert to stimulate activation and pulmonary infiltration of dendritic cells (DCs), lymphocytes, neutrophils and eosinophils 13. In addition, studies of pulmonary alveolar proteinosis have shown that GM-CSF is specifically necessary for lung homeostasis, such that GM-CSF autoantibodies or genetic deletion causes abnormal lung development and surfactant accumulation, despite normal peripheral hematopoiesis 14-16. Together these studies highlight a unique ability of AECs to undergo ‘innate immune mimicry’, and orchestrate inflammation and autoimmunity. Therefore, the contribution of AECs in the pulmonary microenvironment is essential to understanding both normal and pathologic lung function.

Myeloid cell diversity in the lung parenchyma

Given that the lung is exposed to the external environment, it hosts a plethora of myeloid cells with strong innate immune functions. One of the most abundant myeloid cell types in the lung is the macrophage. Lung macrophages encompass at least two populations, including alveolar macrophages (AMΦ) and interstitial macrophages (IMΦ), which are found in separate anatomical sites. AMΦs are the predominant macrophage population of the two, comprising 90–95% of resident immune cells, and live on the luminal side of the alveoli in close proximity to AECs 17, 18. Interestingly, AMΦs do not express integrin αM (CD11b; a canonical myeloid-lineage marker). Like most tissue-resident macrophage populations 19, AMΦs arise during embryonic development from fetal liver precursors in response to GM-CSF, and replenish in adults through self-renewal rather than peripheral recruitment of monocytes 20, 21. As professional phagocytes, the main homeostatic function of AMΦs is the regulation of surfactant levels by type II AECs, as well as clearance of cellular debris and inhaled foreign particles. In contrast, IMΦs reside between the alveoli and capillaries. Unlike AMΦs, they express CD11b and are thought to be maintained from monocytic progenitors that are recruited from the periphery 22. Relative to AMΦs, less is known about the physiological role of IMΦs. However, studies suggest that they share many properties with antigen-presenting cells such as DCs, including overexpression of the class II major histocompatibility complex (MHC-II) and integrin αX (CD11c) 23, 24, and play an important role in maintaining tissue homeostasis in response to harmless inhaled antigens that would otherwise provoke airway allergy 25. It is important to note that many studies do not discriminate between AMΦs and IMΦs, and the ontogeny of specific tissue-resident macrophage subsets is still somewhat controversial. Therefore, we will refer to this mixed population simply as ‘macrophages’ herein, unless specified otherwise, which are predominantly AMΦs.

In all tissues, macrophages display remarkable plasticity and can adopt both pro- and anti-inflammatory phenotypes, allowing them to switch between homeostatic and tissue repair functions. The activation status of macrophages depends on the type of soluble factor present in the local microenvironment. Classic examples include cytokines induced by bacterial lipopolysaccarindes (LPS; e.g. IFN-γ), which skew macrophages toward a pro-inflammatory phenotype (i.e. classically activated, anti-tumorigenic), or IL-4/10/13, which promote an immunosuppressive phenotype (i.e. alternatively activated, pro-tumorigenic) 26, 27. The duration of these exposures influences phenotype switching. For instance, during acute bacterial infections, macrophages respond to LPS (a TLR-4 agonist) by phagocytosing bacteria and secreting tumor necrosis factor (TNF)-α, IL-1β and IL-6 28, 29. Macrophage-derived IL-1β can in turn regulate neutrophil-mobilizing IL-8 release from AECs 30. Thus, acute infections enable macrophages to adopt a pro-inflammatory phenotype, and limit bacterial outgrowth by amplifying innate immune responses. Supporting this notion, during influenza infection, macrophage depletion is associated with synchronous, secondary pneumoccocal infections, which can be prevented by GM-CSF treatment 31. In contrast, chronic airway inflammation (for example, in response to infection with parainfluenza virus) causes reprogramming of macrophages toward an IL-13-producing immunosuppressive phenotype 32. This is a maladaptive reaction that results from persistent disruption of tissue homeostasis that cannot be resolved. Transcriptional profiling of macrophages has similarly revealed that chronic exposure to tobacco smoke can promote an immunoregulatory gene program associated with an anti-inflammatory alternative activation state 33. Whether similar immunosuppressive effects occur in lung tumors 3 and how macrophage subsets can be targeted therapeutically in cancer are active areas of investigation.

Early studies have shown that AMΦs in particular are poor antigen presenting cells (APCs). Despite being equipped with antigen presenting machinery 34, 35, AMΦs lack expression of several co-stimulatory molecules such as CD28 and CD86 36, thereby instructing T cells to remain quiet in the presence of specific antigens. AMΦs can also indirectly restrain adaptive immunity by blunting pro-inflammatory signaling cascades in AECs 37. Combined, these innate mechanisms allow AMΦs to safely maintain immune tolerance to innocuous, commensal antigens. To balance these functions during infection, efficient antigen presentation is performed by lung DCs. However, in cancer, the immune-stimulating functions of these professional APCs become compromised through education by the tumor 38, 39, thus contributing to cancer immune evasion.

During infection, neutrophils are one of the first circulating immune cell types to be recruited to the lungs in response to inflammation, due to their ability to rapidly respond to bacterial exposure. They exit from the peripheral circulation at the lung capillaries through the process of margination, which requires direct interactions with the endothelium 40. As granulocytes, the main innate defense functions of neutrophils are phagocytosis, degranulation, and release of neutrophil extracellular traps (NETs) 40. Neutrophil recruitment is frequently controlled by IL-8, GM-CSF and macrophage inflammatory protein 2; their egress from the bone marrow and retention in the lungs is a dynamic process that requires the cooperative expression of chemokine receptors CXCR2 and CXCR4 41, 42. Some of the most abundant proteolytic enzymes contained within neutrophil granules are neutrophil elastase (NE), myeloperoxidase (MPO) and cathepsin G. Upon degranulation, these enzymes stimulate mucus production 43, 44, and cleave cytokine pro-peptides to amplify immune responses 45. Similar to macrophages, neutrophils are also endowed with phenotypic plasticity which allows them to participate in both antigen elimination (pro-inflammatory phenotype) and immune resolution (immunosuppressive phenotype). Although traditionally neutrophils have been viewed as short-lived, monolithic cytotoxic immune cells, it is becoming more widely appreciated that neutrophils have highly complex and adaptive biology.

As such, neutrophil plasticity allows them to rapidly respond to both sterile and nonsterile challenges, and many studies are now highlighting the diversity of neutrophil subsets in cancer and other pathologic conditions. For instance, during pulmonary infections, activated neutrophils can release NETs decorated with proteolytic peptides to trap pathogens 46, 47; however, excessive NET formation can also lead to acute lung injury and amplify inflammation in response to damaged tissue 48, 49. In neutrophil-dominated chronic inflammatory conditions such as cystic fibrosis (CF) or chronic obstructive pulmonary disease (COPD), neutrophil proteases contribute to emphysema 50, T cell exhaustion through the release of arginase 1 (Arg1) 51, and extracellular matrix (ECM) remodeling 52. Together these studies highlight the diverse functions of neutrophils in tissue homeostasis and disease, somewhat similar to the plasticity exemplified by other myeloid cell types such as macrophages. Whether neutrophil diversity is the result of distinct differentiation lineages, or simply differences in activation states, is yet to be determined.

Lymphoid cells at the intersection of innate and adaptive immunity

Innate lymphoid cells (ILCs), a more recently characterized lymphoid cell subset with innate functional properties, sit at the intersection between innate and adaptive immunity. In contrast to T cells, ILCs lack antigen-specific receptors and play a protective role in antimicrobial responses, particularly at mucosal interfaces 53. In the human lung, approximately 60% of ILCs are group 3 (ILC3), which include natural cytotoxicity receptor-positive (NCR)+ cells and NCR lymphoid tissue inducer cells 54. Although less frequent in number, ILC2 cells (representing 30% of lung ILCs) are critical for maintaining lung homeostasis and respiratory tissue repair 54. In response to AEC-derived IL-25, IL-33 or TSLP, resident ILC2 can rapidly secrete important quantities of amphiregulin and IL-13, which promote airway fibrogenesis and goblet cell mucus secretion, respectively 53, 55. During acute lung inflammation, ILCs are also implicated in macrophage activation, eosinophil mobilization and T helper 2 cell priming 56, 57. Although they remain poorly understood in lung physiology, these studies implicate ILCs and their vast repertoire of pleiotropic cytokines as central immune modulators in the respiratory mucosa.

Similar to ILCs, γδ T cells are a rare subset of T cells that display both innate and adaptive immune functions 58. These cells are highly abundant at mucosal and epithelial surfaces where innate defenses are critical, such as the reproductive, digestive and respiratory tracts, as well as the skin 59-62. Based on their function, γδ T cells can be subdivided into two distinct populations: pro-inflammatory effector γδ T cells and immunosuppressive regulatory γδ T cells. During microbial exposure in the lungs, pro-inflammatory γδ T cells secrete a repertoire of cytokines that are critical during bacterial infection, such as IFN-γ, TNF-α, IL-1β, and IL-17A, which facilitate recruitment of innate mononuclear cells (e.g. DCs and peripherally derived macrophages) and promote microbial clearance 63-65. During allergic responses in the airways, γδ T cells promote IL-5-mediated eosinophilia and IgE class switching via secretion of IL-4 66. In the context of cancer, high transforming growth factor (TGF)-β and IL-15 concentrations in the local microenvironment induce the polarization of immunotolerant γδ T cells (such as Foxp3+ regulatory γδ T cells and γδ T17 cells 67-69), which have been shown to suppress effector T cell responses, block DC maturation, recruit immunosuppressive myeloid cells and promote angiogenesis 70, 71.

Finally, natural killer T (NKT) cells are also important lymphocytes that can have innate immune functionality; they are abundant in lung tissue and recognize antigens in the context of CD1d molecules. In lung, NKT1 and NKT2 subsets are found predominantly in the vasculature (where NKT1 are more abundant), while NKT17 cells are predominantly located within the lung parenchyma 72. In response to inflammation (e.g. during pneumonia infection or allergic asthma), NKT1 cells are activated by CD1d-expressing AMΦs and DCs, and extravasate into the lungs in response to neutrophil-secreted cytokines 73-75. This primes NKT1 and NKT17 subsets to expand and produce pro-inflammatory cytokines such IFN-γ and IL-17, respectively. These reactions help myeloid cells amplify innate immune reactions, to promote pathogen clearance. However, in the case of allergic asthma, this also leads to eosinophilia 74. Whether these cells play a role in lung tumor expansion remains an open question.

Coordinated response of the TIME during key phases of lung tumor progression

An increasing body of evidence suggests that both the pro- and anti-inflammatory activities of pulmonary host-defense cells are intrinsically linked to lung tumorigenesis. Cooperation between cancer cells, AECs, macrophages, and other peripherally-recruited innate immune cells can determine the fate of lung tumors at different stages of both metastatic (Figure 1) and primary (Figure 2) disease, including preneoplastic, early, and late lesions.

Details are in the caption following the image
The premetastatic niche and early metastatic colonization in the lungs. Schematic depiction of the various components of early metastasis, including the premetastatic niche, circulating tumor cells, colonization, and dormancy. Most metastases form adjacent to the lung capillaries as they exit from the vasculature. (A) Integrin α6β1-bearing tumor-derived exosomes accumulate in the premetastatic lung, where they promote the recruitment of pro-inflammatory myeloid cells and deposition of fibronectin 76. Exosomal snRNA activate AECs, which in turn secrete neutrophil-recruiting S100 proteins 77. (B) C5a activates AMΦs, which in turn secrete TGF-β and participate in ECM remodeling 78. (C) Inflammatory stimuli such as tobacco smoke or microbial infection can induce the formation of NETs in the lung, which cleave laminin 79. Laminin cleavage generates epitopes that activate integrin signaling in dormant cancer cells to stimulate their awakening. (D) Circulating tumor cells in the vasculature are captured by NETs, which can be induced by inflammation or soluble G-CSF 80, 81.
Details are in the caption following the image
Cell–cell interactions in the TIME of primary lung tumors. Schematic representation of the interactions between cancer cells and various pulmonary cell types that regulate tumor progression, as well as the soluble factors involved. Most primary lung tumors develop from bronchial or alveolar epithelial cells, and are therefore often found adjacent to the airspaces. (A) Malignant transformation of AECs can be triggered by activated myeloid-cell derived ROS 82. (B) Cancer cells expanding in the lung parenchyma (alveolus or bronchial epithelium) can be attacked by the immune system and undergo apoptosis through several surveillance mechanisms. Pro-inflammatory neutrophils and AMΦ can secrete ROS and TNF-α. Neutrophils can also act as antigen presenters to activate tumor-infiltrating lymphocytes 83, 84. (C) Vγ9Vδ2 T cells can promote DC maturation and activate macrophages via TNF-α and IFN-γ secretion 85-87. Cancer cell death is also induced by NK cell- and CTL-derived GrzB 88. The cytotoxicity of these cells can be further enhanced by the activity of ILC2s and ILC3s 89, 90. (D) The formation of TLSs promotes the generation of anti-tumor plasma cells. (E) Cancer cells can evade immune attack through the activity of tolerogenic lymphoid cells including Tregs, γδ T17 and Foxp3+ γδ T cells 91, 92. These cells shut down cell-mediated responses through the expression of immune checkpoints, including PD-L1 and CTLA-4. Macrophages and neutrophils can also be reprogrammed to acquire an immunosuppressive phenotype (e.g. in response to hypoxia). This promotes tumor expansion through the release of immunosuppressive and pro-angiogenic factors such as IL-10, TGF-β, Arg1 and VEGF. The tolerogenic activity of these cells can be further enhanced under hypoxic conditions 93-95.

Innate immune defenses of the lungs contribute to premetastatic niche formation

During lung metastasis, primary tumors reprogram secondary organ sites to be permissive to metastatic seeding prior to the arrival of cancer cells, via long-range communication between tumor cells and lung-resident cells through the blood. Over the past decade, a myriad of studies have shown that this communication is mediated by various bone marrow-derived progenitor cell populations 96, 97, cytokines and soluble factors 98-101, extracellular vesicles 76, 77, 102, among other factors. These are collectively thought to contribute to the pattern of metastatic organotropism, akin to the ‘seed and soil’ hypothesis as first proposed by Stephen Paget in the 1800s 103.

One of the best characterized cell types that colonizes the premetastatic niche to facilitate cancer dissemination to the lung is the neutrophil. In both the MMTV-PyMT 104 and KEP (K14-cre;Cdh1f/f;Trp53f/f) 91 mouse models of breast cancer, neutrophils (which canonically express CD11b and lymphocyte antigen 6G (Ly6G) on the cell surface) accumulate in the lung prior to the arrival of tumor cells, and are further increased once metastases form. In these studies, neutrophil accumulation promotes metastatic progression, such that targeting neutrophils using anti-Ly6G antibodies improves metastatic outcomes. Several studies have proposed that the mechanism of neutrophil expansion and recruitment to lung is driven by G-CSF that is produced by the primary tumor, which signals systemically to the bone marrow 100, 105. In experimental metastasis models (i.e. absence of a primary tumor), neutrophil homing to the premetastatic lung can also be achieved through chronic inflammation, e.g. from obesity or smoking tobacco 79, 106. Release and redirection from bone marrow is mediated in part by a switch in CXCR2 and CXCR4 signaling axes, which favor homing to the lung tissue 41, 107. Once neutrophils arrive in the lung, they have multifaceted functions that can mirror their behavior during infection and auto-immunity, including NETosis 79, 80, 108, 109, immunosuppression 110-113, or production of leukotrienes 104, which can stabilize single cell metastases 114 and ultimately support a pro-tumorigenic niche 115. Of note, granulocytic myeloid-derived suppressor cells (MDSCs; frequently defined by their expression of CD11b and Gr1) were named for their ability to suppress T cell responses; it is debated whether or not these cells represent a distinct lineage, or whether they are neutrophils with a specific immunosuppressive function.

Additional studies have shown that tissue-resident innate immune cells, such as AMΦs, also contribute to the premetastatic niche prior to the arrival of tumor cells. For example, in mouse models of breast cancer, AMΦs proliferate in the premetastatic lung in response to complement C5a signaling 78 (Figure 1A). C5a-activated AMΦs then impair anti-tumor T cell activity via suppression of IFN-γ production by Th1 cells, inhibition of DC maturation and downregulation of MHC in tumor cells. AMΦ and other innate immune cells can also prepare the lungs for metastatic seeding through means of immune stimulation. In mouse models of hepatocellular carcinoma, AMΦ-derived leukotriene B4 has been shown to increase metastatic permissiveness of the lungs, such that inhibition of the leukotriene-generating enzyme, 5-lipoxygenase (Alox5), is sufficient to prevent pulmonary colonization 116 (Figure 1A). Neutrophils also contribute to Alox5 production in breast cancer models, which facilitates the production of leukotrienes 104, suggesting that these mechanisms may be more universally applicable to lung metastases from multiple primary sources.

Although it is well-documented that primary tumors release cytokines systemically 117, recent studies have reported more creative modes of communication between tumors and the premetastatic niche. Of particular interest are exosomes, which are released by tumor cells into the circulation, carrying informative cargo that can reprogram the secondary tissue. Efforts to characterize the cargo of cancer exosomes have greatly improved our understanding of the molecular processes underlying the remote stimulation of lung resident cells. For example, lung carcinoma- and melanoma-derived exosomes deliver small nuclear (sn)RNA to type II AECs that acted as ligands for TLR-3 77, a PRR that normally recognizes dsRNA from viral pathogens (Figure 1B). In response to TLR-3 stimulation, these AECs actively secrete chemokine C-X-C motif ligand (CXCL)1, CXCL2, S100A8 and S100A9, leading to bone marrow-derived neutrophil recruitment and fibronectin deposition. Similarly, in human NSCLC tissue sections, high expression of TLR-3 in adjacent tissues of lung tumors correlated with massive neutrophil infiltration, high S100 immunohistochemistry scores and poor overall survival 77. Other studies have shown that exosomes direct metastasis organotropism through the expression of integrins 76; in lung, α6β4 and α6β1 integrins are associated with enhanced site-specific metastasis, such that inhibition can reverse this effect. Taken together, each of these studies highlight a key role for innate immunity in the formation of a tumor-supportive niche.

Biphasic regulation of lung tumorigenesis by innate immunity

Experimental and clinical evidence suggests that chronic inflammation promotes cancer initiation in various anatomical sites. While many mechanisms for these associations have been proposed, one notion is that activated myeloid cells can directly generate genetic alterations associated with tumor initiation through the release of free radicals. For example, in vitro experiments using PMA-stimulated neutrophils have shown that secretion of reactive oxygen species (ROS) can cause major genotoxic stress and induce malignant transformation of fibroblasts and epithelial cells 118 (Figure 2A). In a recent study, myeloid cell-derived H2O2 was implicated in the generation of K-Ras mutant epithelial lung tumors in the absence of any carcinogen challenge 82. Finally, production of nitric oxide (NO) from lung macrophages has also been identified as an initiating event in lung squamous cell carcinogenesis, and genetic ablation of the inducible NO synthase 2 (Nos2) gene was sufficient to prevent epithelial cell hyperproliferation and fibrosis 119, 120.

Paradoxically, although myeloid cell-derived free radicals might induce genotoxic stress conducive to malignant transformation, they also promote tumor cell cytotoxicity once cancer is initiated. For example, in murine cancer models of Lewis lung carcinoma (LLC), IFN-polarized pro-inflammatory neutrophils accumulate 7 days following implantation and display tumoricidal activities through the release of TNF-α, NO and H2O2 121 (Figure 2B). Neutrophil-derived H2O2 is also critical for mediating breast cancer cytotoxicity in the premetastatic lung 122. Interestingly, ROS is required for T cell activation 123-125; it is conceivable that this could in part explain its effects on tumor cell clearance. Together these studies highlight the plastic nature of innate immunity during cancer progression, and demonstrate that the anti- or pro-tumorigenic functions of certain cell types are highly context-dependent.

In contrast to their pro-tumorigenic role during early metastatic seeding, studies in early NSCLC have shown that neutrophils are pivotal to tumor cell clearance by stimulating adaptive immunity (Figure 2B). Subpopulations of tumor-infiltrating neutrophils that can stimulate T cells through antigen presentation were documented in early-stage human lung cancer 83, 84. These APC-like neutrophils differentiate from immature granulocytic progenitors in the presence of IFN-γ and GM-CSF, and trigger antigen-restricted effector T cell responses by providing co-stimulatory signaling via the OX40L, 4-1BBL, CD86 and CD54 receptors, and by cross-presenting antigens 84. They also promote anti-tumor T cell proliferation via IFN-γ secretion 83. Similarly, subpopulations of tumor-entrained neutrophils 122, ‘N1’ neutrophils 126, high-density neutrophils 127, or trogocytic neutrophils 128 have anti-tumor functions, suggesting that neutrophils are not simply obligatory immunosuppressive cells, but also exhibit highly diverse phenotypes that are likely stage-dependent. Supporting this notion, mass cytometry and single cell RNA-sequencing studies in NSCLC have shown that the composition of the immune compartment eventually switches to favor the infiltration of myeloid cells that suppress anti-tumor T cell immunity 4, providing evidence of acquired immune evasion as tumors progress.

By comparison, the role of neutrophils during later stages of primary lung cancer progression frequently appears to be pro-tumorigenic, as neutrophils represent the most abundant recruited cell type in more advanced NSCLC 129, Primary NSCLC tumors have been shown to systemically stimulate osteocalcin-expressing osteoblasts in the bone marrow to produce SiglecFhi neutrophils (a canonical eosinophil marker) that promote tumor progression remotely within the lung parenchyma 130. Further, NE promotes disease progression, such that deletion of the NE-encoding gene, Elane, in K-Ras-driven lung tumors prevents cancer-associated death 131. Of note, in the KP model of NSCLC (driven by alterations in K-Ras and p53), neutrophils significantly accumulate following orthotopic injection of syngeneic KP cell lines, but not in response to spontaneous KP tumors, which progress much slower 132. This suggests that, in addition to tumor stage, the speed of tumor progression may also influence neutrophil response. Collectively these studies highlight the stage-specific diverse nature of neutrophil functionality. Adding further complexity to this notion, studies have shown that neutrophil biology is dramatically affected by circadian rhythms 133, the microbiome 134, 135, sex 136, 137, among other factors.

Tertiary lymphoid structures confer protective immunity in human lung malignancies

In response to various pathological stimuli, including cancer, innate and adaptive immune cells that accumulate in the lungs can be organized into tertiary lymphoid structures (TLS; sometimes called induced bronchus-associated lymphoid tissues), which contain distinct T cell, B cell and DC zones 138. These ectopic lymphoid formations share many similarities with secondary lymphoid organs, require lymphotoxin-β-responsive stromal cells for their generation and orchestrate adaptive anti-tumor responses (Figure 2D). In lung cancer, TLS are often found at the invasive margins of the tumor regardless of disease stage, and are surrounded by lymphatic and PNAd+ blood vessels 139. In a retrospective analysis of 74 patients with NSCLC, high frequency of TLS correlated with improved overall and disease-free survival compared with low frequency of TLS 140, demonstrating their prognostic value and suggesting a potential role in anti-tumor immunity.

In addition to the adaptive immune contributions of TLS through canonical B and T cell responses, the unique innate immune properties of TLS involve several lymphoid cell types that have innate-like functions. For instance, NCR+ ILC3s have been identified as a critical component of human NSCLC TLS 89. Upon NKp44-dependent sensing of tumor cells, these cells release pro-inflammatory cytokines (e.g. IL-22, TNF-α, IL-8 and IL-2), and activate the endothelium. Although not directly cytolytic, activated NCR+ ILC3s cross-communicate with lung fibroblasts and help maintain anti-cancer lymphoid aggregates at tumor sites, such that a high frequency of intratumoral NCR+ ILC3 correlates with lower disease stage 89. Preclinical investigations have also revealed the potential relevance of ILC2 to the immunobiology of lung cancers. ILC2-deficient RAR-related orphan receptor (RORα)−/− mice have significantly increased tumor growth rates as well as elevated numbers of circulating tumor cells 90. In this context, ILC2 chemoattraction is dependent on tumor-derived IL-33, and their accumulation in cancerous lesions significantly improves granzyme B-mediated cytotoxic functions of CD8+ T cells (Figure 2C). Finally, ILC1s, which include conventional NK cells, can also be found in respiratory tissues. NK cells play a pivotal role against tumor and virus-infected cells through recognition of mismatched human leukocyte antigen (HLA) molecules. In Kras-driven lung cancers, the cytotoxic activity elicited by NK cells is indispensable to prevent tumor initiation, but is often compromised in later stages due to impaired glycolysis and inhibition by TGF-β 88.

It was recently documented that the development of TLS is also dependent on activated γδ T cells and their production of IL-17 and IL-22 139, 141. Indeed, pulmonary γδ T cells can contribute to anti-tumor immune responses through several diverse mechanisms that cover both innate and adaptive functions. Vγ9Vδ2 T cells have been implicated in cancer cell apoptosis, either through secretion of perforin and granzymes 85, antibody-dependent cellular cytotoxicity 142, or TNF-related apoptosis-inducing ligand ligation 86 (Figure 2C). In addition, γδ T cells can also present cancer antigens via HLA-DR 143, promote DC maturation 87, help B cell antibody production 144 and enhance NK cell cytotoxicity 145. Although the presence of γδ T cell-derived IL-17 is important for pulmonary TLS maintenance, studies suggest it also plays a central role in the formation of a tumor-permissive state in the lungs. This cytokine was indeed shown to promote angiogenesis through induction of vascular endothelial growth factor (VEGF) secretion 92. In addition to immunosuppressive myeloid cell recruitment, γδ T cell-derived IL-17 can promote the expansion of an immunosuppressive subset of iNOS+ neutrophils, which facilitates dissemination of tumor cells to the lungs 91 (Figure 2E).

Taken together, these studies reveal an important role for TLS in lung cancer immunity, and suggest that TLS can be used as prognostic indicators for both incipient and advanced disease.

Hypoxia drives immune suppression in advanced lung tumors

Although the lung represents one of the most oxygenated organs in the body, advanced pulmonary tumors exhibit regions of hypoxia due to an inability of the vascular supply to meet the metabolic demands of rapidly dividing and highly invasive cancer cells 146, 147. This leads to regions of the tumor that are poorly perfused. The mechanisms of hypoxia signaling pathways are predominantly mediated by the hypoxia-inducible factors (HIFs)-1α and HIF-2α, which bind hypoxia-response elements on DNA to regulate the transcription of genes that promote angiogenesis to improve oxygen delivery, such as the VEGF gene, Vegfa 148.

Accumulating evidence suggest an intricate relationship between hypoxia, angiogenesis, and Gr1+ myeloid cells with immunosuppressive properties (often called MDSC, or immunosuppressive neutrophils). In mouse models of NSCLC, anti-Gr1 depletion was shown to re-oxygenate the TIME as a consequence of blood vessel normalization 93. HIF-1α also promotes the generation of an immunosuppressive phenotype in Gr1+ myeloid cells, via upregulation of Arg1 and iNOS 94 (Figure 2E). Genetic deletion of HIF-1α was sufficient to prevent the expansion of these immunosuppressive cells, restore CD8+ cytotoxic T cell surveillance and reduce tumor growth. In mouse models of LLC, HIF-1α was also shown to directly up-regulate the expression of the co-inhibitory receptor programmed death ligand-1 (PD-L1) in Gr1+ myeloid cells, which led to a progressive loss of T cell function 149. PD-L1 blockade under hypoxic conditions abrogated myeloid cell-dependent immune suppression and reduced their secretion of immunosuppressive cytokines (e.g. IL-6, IL-10 and TGF-β).

Similar to its effects on granulocytes, tumor hypoxia has also been shown to reprogram lung macrophages towards a pro-tumorigenic phenotype. Under hypoxic conditions, IL-6-activated lung tumor-infiltrating macrophages showed an immunosuppressive gene signature characterized by elevated expression of Arg1 and Il10 95 (Figure 2E). This was dependent on ERK activation within macrophages, whereby MEK inhibition reversed these effects in in vitro models. Hypoxic lung tumor-derived exosomes have also been implicated in lung macrophage reprogramming. For example, exosomal transfer of the miRNA, miR-103a, to tumor-associated macrophages promotes the expression of several anti-inflammatory genes, such as Il10, Vegfa and Ccl18, through signal transducer and activator of transcription 3 (STAT3) signaling 150. Studies suggest that similar HIF-driven processes are also responsible for Treg cell activation and inhibition of Th1 differentiation 151, however it is still unclear whether this process also takes place in the context of lung malignancies.

Collectively, these findings identify hypoxia as a potent diver of immune suppression in the lung and pave the way for the development of novel strategies targeting the HIF axis.

NETs foster pulmonary dissemination of metastatic tumor cells

NETs are pathogen-trapping DNA webs decorated with various proteolytic enzymes 152. Although NETs are critical for the control of microbial infections, they can exert pro-tumorigenic functions in the context of malignancy. Given that early studies implicated NETs in thrombosis and coagulation disorders, the first studies to link NETs with cancer were focused on tumor-associated hemorrhage. The presence of neutrophil-derived extracellular MPO and nicotinamide adenine dinucleotide phosphate (NADP)H were shown to be associated with vascular injury and leakiness in mouse models of LLC 153. Neutrophil depletion via anti-Gr1 antibody treatment protected against thrombocytopenia-induced tumor hemorrhage. In subsequent studies, neutrophils were found to be primed by tumor-supplied G-CSF to induce NETosis (Figure 1D), leading to the induction of a prothrombotic state in LLC-bearing mice 81. These studies were the first to provide evidence that neutrophil NETosis has effects beyond basic immunology, and may be relevant in cancer.

In mouse models of metastasis, several studies have since shown that NETs are directly involved in metastatic outgrowth of circulating cancer cells. NETs are released as early as 48 h following a bacterial infection, and can sequester circulating lung carcinoma cells within DNA webs to promote hepatic micrometastasis formation 80. This effect can be attenuated with DNAse to digest established NETs. Similar observations were reported for metastatic triple negative breast cancer, which frequently spreads to the lungs (Figure 1D). In mice, breast cancer cells induce NETosis in the absence of bacterial infection via G-CSF, NADPH, and peptidylarginine deiminase 4, which promotes tumor cell seeding within the lungs 108. Targeting NETs via treatment with DNase was sufficient to reduce pulmonary metastatic outgrowth.

Recent evidence suggests that targeting NETs may not only prevent pulmonary colonization by metastatic cells, but also inhibit awakening of dormant cells that have already spread to the lungs 79. During sustained inflammation (for example, in response to LPS stimulation or tobacco smoke), NET-associated NE and matrix metalloproteinase 9 (MMP9) remodel the lung ECM, and generate laminin-111 neo-epitopes (Figure 1C). Integrin α3β1 expressed on dormant metastatic cells can then contact this newly-exposed laminin epitope and transmit proliferative signals that promote tumor cell awakening. Importantly, blockade of remodeled laminin or NET formation can prevent metastatic progression. Interestingly, exposure to tobacco smoke was sufficient to initiate this proteolytic cascade, supporting the notion that environmental triggers can have dramatic consequences on immunological responses of the host. Indeed, various other environmental factors such as air pollution 154, 155 or diet/cholesterol 46, 106, 156-158 have been shown to affect neutrophil biology and lung metastasis in various contexts. Whether these observations extend to human cancer, and whether additional environmental stimuli can initiate similar processes in lung tissue, are areas of active investigation.

Dysbiosis facilitates pulmonary tumor progression and alters response to ICI

Dysbiosis (unfavorable imbalance of microbial diversity) within the gut microbiome can remotely influence the lung TIME, and predict ICI efficacy. In a study on 140 NSCLC patients, it was shown that antibiotics significantly compromise the therapeutic efficacy of PD-1 inhibitors, and was associated with reduced overall survival 159. In addition, metagenomic analyses of fecal samples showed a proportional increase in relative abundance of Akkermansia muciniphila in responders compared with nonresponders. In a mouse model of LLC, fecal microbiota transplantation of stool samples from responders followed by oral supplementation with A. muciniphila restored the efficacy of PD-1 blockade. Functionally, A. muciniphila induced IL-12 secretion (a Th1-polarizing cytokine) by DCs, which increased the CD4+:Foxp3+ cell ratio and increased the recruitment of CCR9+CXCR3+CD4+ T cells at tumor sites. Interestingly, obesity is associated with improved response to ICI in LLC-bearing mice 160 and humans (particularly men) 161, and presence of gut A. muciniphila inversely correlates with onset of metabolic disorders during obesity in mice 162. Whether enhanced ICI efficacy in association with obesity is regulated by the gut microbiome is an open question.

Clinical studies have traditionally focused on the effects of the gut microbiota on intratumoral immune responses. Interestingly, the lung microbiome is also emerging as an important predictor of pulmonary malignancy outcomes. The lung is not a sterile organ and is predominantly populated by Firmicutes, Proteo-bacteria and Bacteroidetes in healthy individuals 163. However, the lung microbiome varies considerably as a function of pH, oxygen tension, ethnicity, sex, and geographic location. Therefore, the lung microbiome can be considered as an integral part of the pulmonary TIME. Indeed, imbalances in certain taxa have been linked to various lung diseases, including COPD, CF and asthma 164. Similarly, pulmonary dysbiosis has been observed in patients with lung cancer. For example, two independent studies have correlated increased frequency of Granulicatella species with lung cancer incidence in humans 165, 166 (Figure 2B). A recent study has also linked lung dysbiosis with lung cancer in mouse models, mediated by neutrophils and γδ T cells [167].

Taken together, investigation of the human microbiome and metagenomics is a rapidly expanding field of study in cancer, and future investigations will determine whether modulation of lung microbial balance is possible in lung cancer patients. Within the context of gut, ongoing clinical trials are investigating whether fecal transplants can be used to improve therapeutic efficacy of ICI in melanoma patients (NCT03341143, NCT02770326), which could potentially extend to clinical management of lung cancer patients as well.

Targeting the intersection between innate and adaptive immunity: ICIs in lung cancer

Innate immune cells must provide three signals to trigger adaptive immunity, including antigen presentation by MHC molecules, co-stimulation, and cytokine release. To restrain inflammatory signals, co-inhibitory checkpoint molecules act as immunoregulatory safeguards. This basic facet of immunology has spurred an entire field of cancer therapy focused on targeting co-inhibitory molecules in cancer, in an effort to boost adaptive immunity against the tumor by mimicking the innate immune system. Indeed, significant improvements have been made towards the clinical management of primary and metastatic lung malignancies with the introduction of PD-1- and PD-L1-targeting agents. Pembrolizumab, a humanized monoclonal antibody against PD-1, was shown to significantly extend the progression-free and overall survival of NSCLC patients, with fewer adverse events than platinum-based chemotherapy 168. Moreover, the addition of pembrolizumab to standard of care chemotherapy can markedly increase the response rate (from 18.9 to 47.6%), as well as the response duration 169.

In addition to TIME characteristics that predict ICI efficacy 170, tumor mutational burden (TMB) is one of the main factors that correlates with improved response to immunotherapy in different cancer types no., owing largely to its effect on neo-antigen expression 171-173. Neoantigens are recognized as nonself by the innate and adaptive immune systems, to trigger inflammation and tumor cell clearance. It can be hypothesized that the positive stimulatory effects of these tumor-associated antigen synergize with the blockade of inhibitory signaling, which represents the optimal conditions for immune priming. Incidentally, lung squamous cell carcinoma, adenocarcinoma and small cell carcinoma generally have very high mutational loads 174, which is not surprising given their strong ties to smoking tobacco. Tobacco smoke is indeed a potent genome destabilizer that can induce diverse DNA alterations, mainly C > A substitutions 175. This, combined with the lung's unique arsenal of immune defenses, may in part explain why ICIs have been particularly effective in lung cancer patients compared with other forms of cancer.

In some contexts, ∼20–40% of patients with advanced lung cancer benefit from PD-1 inhibitors 176. However, predicting which patients will benefit before treatment has proven challenging. This is an important consideration given the potential adverse side effects of immunotherapy. Compensatory up-regulation of T cell immunoglobulin mucin-3 (TIM-3) and lymphocyte-activation gene-3 (LAG-3) have been identified as mechanisms of resistance to PD-1 blockade 177, 178, providing a basis for combinatorial checkpoint inhibition. To this end, numerous TIM-3- or LAG-3-targeting agents, with or without combination with PD-1 inhibitors have entered early phase clinical trials (e.g. NCT03489343, NCT01968109). Clinical trials in NSCLC patients with a high TMB have indeed proven that a significant benefit can be expected from co-targeting different inhibitory molecules over chemotherapy alone 179. Combination of nivolumab (anti-PD-1) and ipilimumab (anti-CTLA-4) was shown to improve the 1-year overall survival from 13 to 43%, and the response rate from 25 to 68%, when compared with chemotherapy in patients with stage IV or recurrent NSCLC 179. Importantly, the efficacy of combination immunotherapy was not dependent on tumor PD-L1 expression status, which is consistent with a recently released meta-analysis of clinical trials with ICIs 180. The clinical data therefore indicate that PD-L1 expression alone is not sufficient for determining which patients will benefit from ICI. Despite the remarkable effects on long term survival, the treatment-associated adverse events leading to treatment discontinuation were significantly higher in the combination immunotherapy group (17.4%) compared with chemotherapy alone (8.9%) 179, which is a major concern. Cases of idiopathic pulmonary fibrosis (IPF) and other interstitial lung diseases in anti-PD1-treated lung cancer patients have been reported, and suggest that further investigation is critical to establish the long term safety of ICI in this population 181, 182. Inflammatory lung diseases such as IPF and pneumonitis are likely to remodel the immune architecture of the lung microenvironment and thereby influence the kinetics of immune-based treatments.

Perspective

A broad understanding of the lung TIME could help identify new targets to enhance therapeutic efficacy of both cytotoxic and immune-targeted treatments 183. In addition to the various cell types and immunoregulatory processes discussed here, many additional microenvironmental cell types in the lung contribute to disease progression, such as adaptive immune cells, endothelial cells, fibroblasts, among others. The functional influence of these cells on tumor biology cannot be discounted; rather, we propose that a comprehensive evaluation of all aspects of the lung microenvironment – including innate, adaptive, and non-immune components – is necessary to optimize therapeutic efficacy and survival outcomes across a broader spectrum of patients.

Among the emergence of ICIs in lung cancer therapy, targeting the innate immune system is also gaining traction in the clinical setting; therefore consideration of this arm of the immune system is critical to realize the full potential of all forms of cancer immunotherapy. For example, inhibition of NKG2A, an NK cell negative regulator, is currently being tested in phase II clinical trials in combination with PD-L1 inhibitors in lung cancer patients 184. Preliminary results from phase Ib/II trials using NKG2A inhibitors in head and neck cancer patients have shown a favorable response-to-toxicity profile with only minor and reversible adverse events 184. In addition, numerous macrophage-targeting agents (e.g. CSF-1R inhibitors), and neutrophil chemotaxis inhibitors (e.g. CXCL2 inhibitors) are now being explored for their anti-tumor potential in lung cancer patients (for additional examples of TIME-targeting agents that are currently being investigated in clinical trials, see Table 1). Whether these treatments will work best as monotherapy, or in combination with other immune-targeted therapies including ICI, has yet to be determined.

Table 1. Investigational agents targeting the lung TIME
Phase Condition(s) Intervention (Target) Trial ID
I

NSCLC

Melanoma

Renal cell carcinoma

APX005M (CD40)

Cabiralizumab (CSF-1R)

Nivolumab (PD-1)

NCT03502330
I

Lung cancer

Head and neck cancer

Pancreatic cancer

Ovarian Cancer

Renal Cell Carcinoma

Malignant Glioma

FPA008 (CSF-1R)

BMS-936558 (PD-1)

NCT02526017
I

NSCLC

Melanoma

Head and neck cancer

Hodgkin lymphoma

Urothelial cancer

Bladder cancer

L-NMMA (iNOS)

Pembrolizumab (PD-1)

NCT03236935
I

NSCLC

Triple negative breast cancer

Pancreatic ductal adenocarcinoma

Colorectal cancer (microsatellite stable)

Ovarian cancer

Renal cell carcinoma

NZV930 (CD73)

PDR001 (PD-1)

NIR178 (adenosine A2A-R)

NCT03549000
I-II Advanced solid tumors

Durvalumab (PD-L1)

Monalizumab (NKG2A)

Cetuximab

NCT02671435
I-II

NSCLC

Solid tumor

Hepatocellular carcinoma

Galunisertib (TGF-β)

Nivolumab (PD-1)

NCT02423343
I-II

Lung cancer

Metastatic cancer

Solid tumor

Colorectal cancer

Gastric cancer

Head and neck cancer

Renal cell carcinoma

Urothelial cancer

Bladder cancer

Mesothelioma

INCB001158 (Arg1)

Pembrolizumab (PD-1)

NCT02903914
I-II (Any)

BMS-986253 (IL-8)

Nivolumab (PD-1)

NCT03400332
II Resectable or early stage NSCLC

Durvalumab (PD-L1)

Durvalumab + Oleclumab (CD73)

Durvalumab + Monalizumab (NKG2A)

Durvalumab + Danvatirsen (STAT3)

NCT03794544
II

NSCLC

Castration resistant prostate cancer

Colorectal cancer (microsatellite stable)

Navarixin (CXCR2)

Pembrolizumab (PD-1)

NCT03473925
III NSCLC

Docetaxel

Ganetespib (HSP90/HIF-1α)

NCT01798485
III Lung cancer

Nivolumab (PD-1)

Epacadostat (IDO1)

with platinum doublet chemotherapy

NCT03348904
  • Listed are examples of ongoing phases I II and III clinical trials with lung cancer patients using inhibitors targeting different elements of the lung tumor microenvironment. Myeloid cell mobilization can be targeted with CSF-1R, IL-8 or CXCR2 inhibitors. Immunosuppression can be targeted with PD-1, PD-L1, iNOS, CD73, NKG2A, adenosine A2A-R, Arg1 or IDO1 inhibitors. Hypoxia can be targeted with HIF-1α inhibitors.
  • Lung conditions are highlighted in bold text.

Although the TIME holds promise as a therapeutic target, this approach is not without challenges. Given that many immune cells in the lung TIME can exert both tumoricidal and tolerogenic effects, the window of opportunity for administrating pro- or anti-inflammatory agents may be limited, and long-term adverse effects are a major concern for patients. Thus, identifying biological determinants that predict response to immune-based treatments is likely to improve the quality and efficacy of personalized care. For example, this could include evaluation of the frequency of specific immune cells within tumor biopsies (e.g. CD8+ T cells, activated neutrophils), tumor mutation load, or analysis of the gut microbiome. Moreover, histologic examination of the lung TIME will be greatly improved with the use of emerging multiplexed technologies such as imaging mass cytometry. These tools will help to identify regional microenvironmental patterns in patients most likely to respond to certain therapies, and/or limit the risk of inducing deleterious immune-related adverse events in patients with low probability of response. Thus, a holistic view of the TIME, which acknowledges both intra- and extra-pulmonary factors that influence the lung immune architecture (e.g. adipokines, gut microbes), will undoubtedly open new avenues for prevention and treatment of malignant lung diseases.

Acknowledgements

SM is supported by a Canderel Graduate Studentship Award from the Goodman Cancer Research Centre, and the Division of Experimental Medicine (McGill University). POF is supported by a Nesbitt-McMaster Award for Excellence in Medicine and Surgery. LAW is supported by the Rosalind Goodman Chair in Lung Cancer Research, Canadian Institutes of Health Research, and the Goodman Cancer Research Centre (McGill University). JDS is supported by the Cancer Research Society, the Fonds de Recherche Santé du Québec, the American Surgical Association, and the Research Institute of the McGill University Health Centre. DFQ is supported by Susan G. Komen, Canadian Institutes of Health Research, a Tier II Canada Research Chair in Tumor Microenvironment, and the Goodman Cancer Research Centre (McGill University).

    Author contributions

    All authors contributed to writing, editing, and reviewing the article.