Natural killer (NK) cells (expressing the surface markers CD16 and CD56, but not CD3) are lymphocytes that play an important role in the rejection of tumors without previous sensitization and without restriction by the major histocompatibility complex (MHC) [7, 8]. NK cells eradicate tumors through multiple killing pathways, including direct tumor cell killing. They also secrete cytokines and chemokines like interleukin (IL) IL-10, tumor necrosis factor (TNF)-α, and the principal NK-derived cytokine interferon (IFN)-γ, which can coordinate the innate and adaptive immune responses to tumor cells and may lead to apoptosis of the attacked cells. A large cohort study showed that an increase in NK cells in tumor tissue is a strong independent prognostic factor for the survival of lung cancer patients [9]. Natural killer T (NKT) cells (CD16+, CD56+, CD3+) are a subset of NK cells that are found in the peripheral blood, tumor tissue and pleural effusions of lung cancer patients in decreased numbers and with reduced functions [10, 11]. It has been shown that NKT cells in cancer patients produce a decreased amount of IFN-γ and are therefore less effective than NKT cells in healthy controls [12, 13].

Accumulation of mast cells is common in angiogenesis-dependent conditions, like cancer, as mast cells are major producers of proangiogenic molecules as vascular endothelial growth factor (VEGF), IL-8, transforming growth factor (TGF)-β [14]. Mast cells also play a central role in the control of innate and adaptive immunity by interacting with B and T cells (in particular regulatory T cells) and dendritic cells. The density of mast cells in non-small cell lung cancer (NSCLC) tumors is correlated with microvessel density [15] and mast cells/histamine has a direct growth promoting effect on NSCLC cell lines in vitro [16]. Tumor-infiltrating mast cells can directly influence proliferation and invasion of tumors, by histamine, IL-8 and VEGF while the production of TNF-α and heparin can suppress tumor growth [16, 17]. It has been shown that in NSCLC mast cell counts were enhanced as tumor stage increased while another study did not found this correlation [14, 18]. The controversy of mast cells in other cancer types seems to be related to the type, microenvironment and stage of cancer and their role may depend on the tumor environment [19, 20].

Neutrophils play a major role in cancer biology. They make up a significant portion of the infiltrating immune cells in the tumor and the absolute neutrophil count and the neutrophil to lymphocyte ratio in blood are independent prognostic factors for survival of NSCLC [21–23]. Neutrophils are attracted to the tumor under the influence of specific chemokines, cytokines and cell adhesion molecules. Tumor-associated neutrophils (TAN) have polarized functions and can be divided into the N1 and N2 phenotype in a context-dependent manner [24]. The N1 phenotype inhibits tumor growth by potentiating T cell responses while the N2 phenotype promotes tumor growth [25]. The antitumor activities of N1 neutrophils include expression of immune activating cytokines (TNF-α, IL-12, GM-CSF, and VEGF), T cell attracting chemokines (CCL3, CXCL9, CXCL10), lower expression of arginase, and a better capacity of killing tumor cells in vitro. N2 neutrophils support tumor growth by producing angiogenic factors and matrix-degrading enzymes, support the acquisition of a metastatic phenotype, and suppress the anti-tumor immune response by inducible nitric oxide (NO) synthase and arginase expression [26]. Neutrophils also influence adaptive immunity by interacting with T cells [27], B cells [28], and DC [29]. In resectable NSCLC patients, intratumoral neutrophils were elevated in 50 % of the patients and this was associated with a high cumulative incidence of relapse [30].

B cells may affect the prognosis of patients with lung cancer , as patients with stage I NSCLC contain more intratumoral germinal centers with B lymphocytes than patients with stages II to IV [31]. These tertiary (T-BALT) structures provide some evidence of an adaptive immune response that could limit tumor progression in some patients. For instance, the production of antibodies by B cells can activate tumor cell killing by NK cells and other inflammatory cells [32]. Auto-antibodies against tumor antigens are commonly found in patients with lung cancer [33–35] and can inhibit micrometastasis [36]. The role of B cells seems depending on the context.

Several subsets of T cells have been identified, and they all develop in the thymus from common precursor T cells. Based on cytokine secretion and function, these cells are classified as CD4, CD8, Th17, regulatory T cells (Tregs), or TH17 cells, amongst others. CD4+ cells and CD8+ cells represent the strong effectors of the adaptive immune response against cancer [37]. There is controversy on the impact of T cells and their localization on the prognosis of lung cancer [38–43]. This may be caused by the presence of a special subset of T cells, the regulatory T cells (Tregs), and myeloid-derived suppressor cells which are both discussed below. Also tumor-derived factors can exhaust T lymphocytes or induce their apoptosis [44]. Recently it has been shown that cytotoxic T lymphocytes (CTL) within the tumor (the tumor-infiltrating lymphocytes [TIL]) are of beneficial prognostic influence in resected NSCLC patients in both adenocarcinoma [45] and squamous cell carcinoma [46].

Tregs, characterized by the expression of CD4, CD25, Foxp3, but absence of CD127, are T lymphocytes that are generated in the thymus (natural Treg) or induced in the periphery (induced Treg) when triggered by suboptimal antigen stimulation or stimulation with IL-35, TGF-β and IL-10 [47]. Tregs are further characterized by the expression of glucocorticoid-induced TNF-receptor-related-protein (GITR), lymphocyte activation gene-3 (LAG-3), and cytotoxic T-lymphocyte-associated antigen 4 (CTLA4). In cancer patients, Tregs confer growth and metastatic advantages by inhibiting anti-tumor immunity. They have this pro-tumoral effect by promoting tolerance via direct suppressive functions on activated T-cells or via the secretion of immunosuppressive cytokines such as IL-10 and TGF-β [48, 49]. Tregs are present in tumor tissue [50, 51] and increased in peripheral blood of NSCLC patients compared to healthy controls [52, 53]. This increase in Tregs was found to promote tumor growth and was correlated with lymph node metastasis [54, 55] and poor prognosis [50, 56]. Many factors can increase Tregs in NSCLC tumors, among them are thymic stromal lymphopoietin (TSLP) [57] and intratumoral cyclooxygenase-2 (COX-2) expression [58]. Tregs are considered the most powerful inhibitors of antitumor immunity [59].

Th17 cells are a subpopulation of CD4+ T helper cells that are characterized by the production of interleukin-17 (IL-17, also known as IL-17A). IL-17 plays an important role in the host defenses against bacterial and fungal infections by the activation, recruitment, and migration of neutrophils [60, 61]. In vitro experiments have shown that IL-1β, IL-6, and IL-23 promote Th17 generation and differentiation from naïve CD4+ T cells [62]. Among the other cytokines secreted by Th17 cells are IL-17 F, IL-21, IL-22, and TNF-α. The role of Th17 cells in cancer is poorly understood. Th17 cells accumulate in malignant pleural effusion from patients with lung cancer [62]. Also higher levels of IL-17A were detected in serum and in tumor lesions of lung adenocarcinoma patients, indicating a potential role of these cells in cancer [63]. It has been shown that Th17 cells encouraged tumor growth by inducing tumor vascularization or enhancing inflammation, but other studies revealed also opposite roles for Th17 cells. Recent data indicate that IL-17 may play a role in the metastasis of lung cancer by promoting lymph-angiogenesis and is therefore an independent prognostic factor in both overall and disease-free survival in NSCLC [64]. So, it is controversial whether Th17 cells in cancer are beneficial or antagonistic; this may be dependent on the tumor immunogenicity, the stage of disease, and the impact of inflammation and angiogenesis on tumor pathogenesis [65].

Myeloid-derived suppressor cells (MDSCs) are a heterogeneous population of immature myeloid cells and myeloid progenitor cells. MDSCs inhibit T cells activation [66, 67] in a nonspecific or antigen-specific manner, alter the peptide presenting ability of MHC class I molecules on tumor cells [68], influence B cells [69], block NK cell cytotoxicity [70–72], inhibit dendritic cell differentiation [73], and expand Tregs [74, 75] signifying their crucial contribution in constituting a tumor suppressive environment. Furthermore, there is compelling evidence that MDSC, by secreting MMP9 and TGF-β1, are also involved in angiogenesis, vasculogenesis, and metastatic spread [76]. MDSCs suppress the immune system by the production of reactive oxygen species (ROS), nitric oxide (NO), peroxynitrite and secretion of the cytokines IL-10 and TGF-β [77]. Upregulated arginase-I activity by MDSCs depletes the essential amino acid L-arginine, contributing to the induction of T cell tolerance by the down regulation of the CD3ζ chain expression of the T cell receptor [78–81]. However, the mechanisms that are used to suppress the immune responses are highly dependent on the context of the tumor microenvironment [82]. An increased subpopulation of MDSCs in the peripheral blood of NSCLC patients was detected that decreased in those patients that responded to chemotherapy and patient undergoing surgery [83].

Macrophages are part of the innate immune system and play important roles in the first line of defense against foreign pathogens. They can be divided into M1 macrophages (classical activation) and M2 macrophages (alternative activation). M1 macrophages attract and activate cells of the adaptive immune system and have anti-tumor and tissue destructive activity, while the M2 phenotype has been linked to tumor-promoting activities by subversion of adaptive immunity, promoting tumor angiogenesis and supporting cancer cell survival, proliferation, invasion and tumor dissemination. Macrophages in tumors are usually referred to as tumor-associated macrophages (TAM) and their presence can be substantial (10–65 % of the tumor stroma). In the beginning, the TAM mainly consist of M1-like macrophages however, when the tumor starts to invade and vascularize, there is a skewing towards the M2 phenotype [84, 85]. This takes place especially at those regions in the tumor that are hypoxic [86]. It has been reported by several groups that there is an association between the number of tumor islet macrophages and NSCLC survival [87–91]. Moreover, when looking at the different phenotypes of TAM (M1 and M2), it is shown that high numbers of M1 macrophages infiltrating the tumor are correlated with improved survival. On the other hand, the presence of M2-like macrophages is associated with poor clinical outcome.

Dendritic cells (DC) are widely acknowledged as the central surveillance cell type and play an important role in the activation of lymphocyte subsets to control or eliminate human tumors. Upon encountering tumor cells or tumor-associated antigens, DC engulf this material and begin migrating via lymphatic vessels to regional lymphoid organs. The density of immature DC (Langerhans cell and interstitial DC) and mature DC, present in the tumor microenvironment, is highly predictive of disease-specific survival in early-stage NSCLC patients [92] and the presence of DC in resected NSCLC material is a good prognostic factor [9]. Interaction between the DC and tumor cells results in the release of anti-tumor cytokines [93, 94]. This suggests that DC within the tumor microenvironment of early-stage NSCLC are capable in initiating adaptive immune responses in situ [95–97]. In the peripheral blood and regional lymph nodes of lung cancer patients, the number and function of mature DC is dramatically reduced [98, 99], partly due to abnormal differentiation of myeloid cells (e.g. MDSC) [100]. Tumor cells, stromal cells like fibroblasts, and tumor-infiltrating immune cells and/or their secreted products, like VEGF, M-CSF, IL-6, IL-10, and TGF-β are also responsible for systemic and local DC defects [101–104]. Affected DC are impaired in their ability to phagocytose antigen and to stimulate T cells, leading to a defective induction of anti-tumor responses. NSCLC-derived DC produce high amounts of the immunosuppressive cytokines IL-10 and TGF-β [105]. It has been shown that the T cell co-inhibitory molecule B7-H3 and programmed death receptor-ligand-1 (PD-L1) are upregulated on tumor residing DC and these molecules conveys mainly suppressive signals by inhibiting cytokine production and T cell proliferation [106, 107].

From the above it is apparent that different types of tumor-infiltrating immune cells have different effects on tumor progression. Based on their functions, immune types can be divided into cells with a potentially positive impact on the antitumor response and cells with detrimental effects, but cell phenotypes can adapt on the changing environment (e.g. the macrophage M1/M2 phenotype polarization). The net effect of the interactions between all these various cell types and their secreted products within the environment of an established tumor participates in determining anti-tumor immunity, angiogenesis, metastasis, overall cancer cell survival and proliferation. These immune infiltrates are heterogeneous between tumor types, and are very diverse from patient to patient. The importance of these tumor-infiltrating immune cells for tumorigenesis, and their secreted chemokines and cytokines, was recently acknowledged by revisiting the hallmarks of cancer as described by Hanahan and Weinberg in 2000 and now include “tumor promoting inflammation” and “avoiding immune destruction” [108, 109]. Besides the type and density of the cells, also the location in the tumor (in the centre or core; or within the invasive margins of the tumor), in adjacent tertiary or secondary lymphoid structures or their presence in peripheral blood will shape the immune contexture. Histopathological analyses of the location, density and functional orientation of the different immune cell populations in large annotated collections of human lung tumors has identified a good association of effector T-cells (CD3 + CD8+), memory T cells (CD3 + CD45RO+) [38, 88, 110–112], and Th1 cells (IL-2 and IFN-gamma secreting CD3+ cells) [110] with longer disease-free survival and/or a better overall survival while Th17 [64, 110] and Tregs [56, 110, 113, 114] have a poor association with the prognosis [115, 116]. However, it is important to realize the complex spatiotemporal dynamics in the tumor-immune interactions in time due to perturbations at the gene and protein level of the immune cells and tumor cells within the microenvironment [116]. For example, when myeloid cells are attracted to the tumor, they are influenced by several signals able to shape the new cells as “needed” by the tumor. In an early phase of tumor development, the tumor-associated macrophages mainly consist of an M1-like phenotype and later in the tumorigenic process, when the tumor changes its local environment, there is a skewing towards the M2 phenotype [84, 85, 117]. This takes place especially at those regions in the tumor that are hypoxic (Fig. 2) [86, 118]. A subpopulation of TAMs gather in hypoxic sites in the tumor as a result of chemoattractants produced by tumor cells [119]. Exposure to hypoxia stimulates TAMs to acquire a pro-angiogenic M2 phenotype with high production of pro-angiogenic factors like VEGF and MMP-9 [120]. This preferential polarization is also a result of the absence of M1-orienting signals, such as IFN-γ or bacterial components in the tumor environment as well as the presence of M2 polarization factors. Thus the cancer immunoediting concept as described by Schreiber et al. [121], that comprises editing, equilibrium and escape, takes place continuously and this plasticity is a major source of heterogeneity between patients and a cause for therapy resistance [4].

Immunotherapy attempts to stimulate or restore the body’s natural ability of the immune system to fight cancer. There are various strategies to activate the immune system and these are classified earlier by Aerts et al. and here into the following categories: biological response modifiers, monoclonal antibodies, peptide or tumor cell vaccines, and cellular immunotherapy (Fig. 3 and Table 1). There is no consensus regarding which of the four categories is the optimal approach for thoracic malignancies, this will probably be highly dependent on the tumor characteristics of each individual patient.

Immunotherapy represents a new class of agents in the treatment of lung cancer. As demonstrated for sipuleucel-T in prostate cancer and ipilimumab in melanoma, treatment responses and improvement of overall survival can be seen in lung cancer patients. However, often the agents did not change initial disease progression. Even a transient worsening of disease manifested either by progression of known lesions or the appearance of new lesions can be seen, before disease stabilization or tumor regression. The commonly accepted treatment paradigm, however, suggests that treatments should initially decrease tumor volume, which can be measured using CT scan. Also, progression-free survival is increasingly used as an alternative end-point of studies. This seems to be unfortunate for immunotherapy, which may initiate an immune response that ultimately slows the tumor growth rate, resulting in longer survival, but not a decrease in tumor volume on CT scan or an increased progression-free survival [257]. Immune-related response criteria which adapt the standard response criteria and include the potential for delayed clinical response and initial increase of tumor mass after immunotherapy are currently being developed [207]. Therefore, clinicians at this moment may need to reconsider how to measure success of their immunotherapeutic approach [208].

For a long time, stimulating the patient’s immune system to attack tumors has been viewed as a rather meaningless intervention, an assumption that has radically changed since the recent clinical successes of cancer immunotherapy. However, despite the promising results achieved in some patients, the overall response rates are low. It has been shown that the immune contexture (i.e., the type, density and location of tumor-infiltrating immune cells) can predict the clinical outcome of patients affected by multiple types of cancer, but with consistent intra-patient variations [209]. It is therefore critical to identify the unique immunological profile of individual patients as a means to identify the best-suited immunotherapeutic approach for each patient [153]. This personalized way could greatly improve the efficacy of current immunotherapies. In addition to an individualized treatment plan, the future of immunotherapy in lung cancer patients includes multimodality treatment. Surgery, irradiation or chemotherapy vigorously reduce the tumor mass and the induced tumor cell death results in tumor antigen exposure. These therapeutic effects enhance the efficacy of immunotherapy and warrant combinatorial treatment approaches. In addition, multiple chemotherapeutical agents have been described to modulate the tumor microenvironment and therefore enhance immunotherapy [210, 211]. Gemcitabine and vinorelbine are known to be able to reduce circulating MDSCs [212, 213], increase the ratio of M1 to M2 macrophages [214] and stimulate APCs [215]. Therefore, the combination of immunotherapy with conventional treatments can elicit a synergistic treatment response and takes advantage of the full potential of cancer immunotherapy.