Lung Cancer and Its Microenvironment

The association between inflammation and cancer was identified
in the 19th century. Initially, it was believed that leukocyte
infiltrates in tumors represented an attempt by the host to
eradicate malignant cells. It was later demonstrated that several
chronic inflammatory conditions, such as inflammatory bowel
disease, Helicobacter pylori infection, hepatitis B or C infection,
and prostatitis, predisposed people to cancer of the colon,
stomach, liver, and prostate, respectively. In addition, malignant
tissues that contain inflammatory cells such as macrophages
in breast cancer or neutrophils and mast cells in lung
cancer are associated with an unfavorable outcome.
Lung cancer risk is clearly enhanced by cigarette smoking,
and chronic inflammation associated with chronic obstructive
pulmonary disease probably enhances this risk, although an
increase is difficult to demonstrate. Some lung cancers that
occur in association with scars could have no relationship with
smoking habits. In patients with idiopathic fibrosis, lung cancer
incidence is much higher than in the general population. 1
Oncogene activation is often associated with inflammatory
response. For example, scar-associated cancers seem to more
often have KRAS (codon 12) mutations. Further, a transgenic
mouse model of lung cancer generated by KRAS activation
showed a robust inflammatory response compared to the wildtype
mice. 2,3 Lastly, in several mouse models, this inflammatory
response has been demonstrated to be not only associated
with but also required for tumor initiation or growth. 4,5
The tumor microenvironment is composed of structural
(extracellular matrix [ECM]), soluble (growth factors, chemokines,
cytokines, proteases, and hormones, among others),
and cellular components (tumor cells, fibroblasts, inflammatory
cells, vascular and lymphatic endothelial cells, and vascular
smooth muscle cells and pericytic cells, among others).
Characterization of the inflammatory cells within tumors has
revealed both the adaptive and innate arms of the immune
response. For example, dendritic cells (DCs) present tumor
antigens to T lymphocytes (CD4 , CD8 , and natural killer
[NK]), promoting an antitumor cytotoxic T-cell response.
This response is negated by a population of immature myeloid
cells called myeloid-derived suppressor cells that promotes
the development of FOXP3 CD4 T cells or Tregs, which
suppress the antitumor cytotoxic T-cell response and induce
polarized differentiation of monocytes into tumor-associated
macrophages (TAMs or M2 macrophages). TAMs, vascular
endothelial cells, and fibroblasts within the tumor stroma secrete
a number of growth factors and chemokines that promote
tumorigenesis. Thus, conflicting immunologic forces fight for
supremacy in the tumor microenvironment. This chapter will
deal with recent research on the tumorigenic and antitumorigenic
effects of the immune system on the lung. The latter was
discussed fully in a recent review. 6
ANGIOGENESIS
Angiogenesis is the growth of the new blood vessels, necessary
for cancerous tumors to keep growing and spreading (see
Chapter 8). Many proteins and other smaller molecules have
been identified as angiogenic , particularly vascular endothelial
growth factor (VEGF), basic fibroblast growth factor (bFGF),
and CXC chemokines, among others. The binding of these
molecules to their appropriate receptor activates a series of
relay proteins that transmits a signal into the nucleus of the endothelial
cells. The nuclear signal ultimately prompts a group
of genes to make products needed for new endothelial cell
growth. First, the activated endothelial cells produce matrix
metalloproteinases (MMPs), a class of degradative enzymes
that break down the extracellular matrix, thus permitting the
migration of endothelial cells that had been tethered to the
matrix. As they migrate into the surrounding tissues, activated
endothelial cells begin to divide. Soon they organize into hollow
tubes that evolve gradually into a mature network of blood
vessels. Cancer cells originating in a primary tumor can spread
to another organ and form metastases that can remain dormant
for years. The induction of this vasculature in primary
tumor or in metastases, termed angiogenic switch , can occur at
various stages of tumor progression, depending on the tumor
FIBROBLASTS
Normal stroma contains fibroblasts in association with
physio logical extracellular matrix. Reactive stroma is associated
with an increased number of fibroblasts, enhanced
capillary density, and type I collagen and fibrin deposition.
In chickens that are cancer-prone because they have been
infected with Rous sarcoma virus, wounding leads to invasive
carcinoma, demonstrating that reactive stroma provides
oncogenic signals that facilitate tumorigenesis. 8 Fibroblasts
are associated with cancer cells (tumor-associated fibroblasts
[TAFs], carcinoma-associated fibroblasts [CAFs]) at all stages
of cancer progression. The growth factors, chemokines, and
extracellular matrix—these fibroblasts produce facilitate angiogenic
recruitment of endothelial cells and pericytes. They
are phenotypically and functionally distinct from fibroblasts
that are not in the tumor microenvironment. The modified
phenotype they acquire is similar to that of fibroblasts
associated with wound healing. Smooth muscle differentiation
(myofibroblasts) is prominent in stromal cells of malignant
breast tissue but rarely seen in normal breast tissue. 9
The signals that mediate the transition of normal fibroblasts
into TAF or CAF are not fully understood, but transfor ming
growth factor- (TGF- ), platelet-derived growth factor
(PDGF), and fibroblast growth factor 2 (FGF2) are the main
mediators to induce fibroblast activation. TGF- induces the
acquisition of activated phenotype of fibroblasts in culture 10
and has been shown to be correlated with desmoplastic reaction
and poor prognosis in breast cancer. 11 PDGF induces
the proliferation of fibroblasts and has been shown to be associated
with cancer progression in breast cancer. 12 FGF2 also
stimulates proliferation of fibroblasts and is also recognized
for its potential to induce angiogenesis. 13
The fact that fibroblasts contribute to tumor initiation,
growth, and metastasis have been demonstrated by both in
vivo and in vitro study. 14 Whereas normal fibroblasts are required
to maintain epithelial homeostasis, CAFs probably
initiate and promote tumorigenic alterations in epithelial
cells. Fibroblasts cultured from malignant tumors have stimulatory
effects on breast tumor cell lines, whereas fibroblasts
cultured from normal tissue are inhibitory. 15 If CAFs are coinoculated
with prostate, breast, or bladder tumor cell lines into
nude mice, tumor latency is shortened and tumor growth increased,
16 whereas normal fibroblasts do not have this effect.
Increased cell proliferation and angiogenesis also result. Lastly,
fibroblasts could promote metastasis by secreting growth factors
that create a niche that promotes the growth of cancer
cells at distant sites. 17
CAFs could also modulate the immune response. CAFs
isolated from primary non–small cell lung cancers (NSCLCs)
were able to enhance or suppress tumor-associated T-cell
function. 18
The epithelial-to-mesenchymal transition (EMT) might
be an additional source of fibroblast-like cells (with an altered
genome). In EMT, epithelial cells lose cell–cell contacts and acquire
mesenchymal properties. Cancer cells undergoing EMT
develop invasive and migratory abilities and express EMT
markers (E-Cadherin, Vimentin) that have been shown to be
markers of tumor progression. 19–21 This phenotype has also
been shown to be associated with resistance to certain therapies
such as epidermal growth factor receptor (EGFR) tyrosine
kinase inhibitor (TKI) in NSCLC. 21,22
MACROPHAGES
Most solid tumors are abundantly populated with TAMs. These
cells can compromise clinical outcome. Clinical stu dies have
shown a correlation between TAM density and poor prognosis
for several types of cancer, an association that is particularly
strong for breast, prostate, ovarian, and cervical cancers. 23 For
NSCLC, TAM density correlated significantly and negatively
with overall survival or relapse-free survival in two of three
published studies. 24–26
Macrophages are recruited to tumors by a wide variety
of growth factors—granulocyte colony-stimulating factor
(G-CSF), granulocyte-monocyte colony-stimulating factor
(GM-CSF), macrophage-stimulating protein (MSP), VEGF,
TGF- —and chemokines, which include CC chemokines,
monocyte chemoattractant protein family, macrophage inflammatory
protein-1 (MIP-1), and macrophage migration
inhibitory factor (MIF). 27
Tumor-derived molecules probably influence TAM phenotype.
Exposure to IL-4 and IL-10 in tumors may induce
TAMs to develop into M2 macrophages, which are characterized
by poor antigen-presenting capacity and production
of factors that suppress T-cell proliferation and activity and
induce angiogenesis, whereas M1 macrophages are efficient
immune cells. 28
Macrophages induce angiogenesis. Correlations between
number of macrophages and microvessel count have been
observed for many tumor types including lung cancer. 24,25,29
Macrophage infiltration into tumor is not homogeneous: studies
using hypoxic markers have shown that TAMs accumulate
in hypoxic and necrotic areas. Hypoxia induces synthesis of
macrophage chemoattractants such as VEGF by upregulating
hypoxia-inducible factor (HIF), which recruits and immobilizes
macrophages in such areas. 30 Here, these cells synthesize
angiogenic regulators, which results in formation of new blood
vessels. These regulators are angiogenic factors (VEGF, PDGF,
IL-8) and angiogenesis-modulating enzymes (MMPs and cyclooxygenase-
2 [COX-2]). In vitro studies, based on coculture experiences,
showed that exposure of macrophage to tumor cells
increases synthesis of angiogenic factors. 25,31,32 In transgenic
mice susceptible to mammary cancer (PyMT mice), malignant
transition was demonstrated to be regulated by infiltrated
macrophages in primary mammary tumors. Inhibition of macrophage
recruitment into the tumors delayed the angiogenic
switch and malignant progression, while genetic restoration of
the macrophage population rescued angiogenesis. 33
Macrophages have been shown to stimulate proliferation
of tumor cells. In K- ras LA1 mice, which develop lung adenocarcinoma
through somatic activation of a K- ras allele, intraepithelial
and airspace macrophage infiltration is observed
beginning at the earliest stage of neoplasia and increasing with
malignant progression. 3 In this model, inhibition of malignant
progression by a targeted treatment directed against the mTOR
pathway was accompanied by macrophage loss. Conditioned
media from primary cultures of macrophages stimulated the
proliferation of lung tumor cells, which was consistent with
previous reports demonstrating a stimulatory effect of alveolar
macrophages on the proliferation of normal and distal airway
epithelial cells in other animal models. 34–36 In a transgenic
mouse model of preneoplastic progression in the mammary
gland, conditional depletion of macrophages inhibited epithelial
cell proliferation and lateral budding. 37
Macrophages are involved in invasion and metastasis. In
the PyMT mouse model of breast cancer progression, leukocytic
infiltrates were present in areas of basement membrane
breakdown 38 suggesting their involvement in tumor invasion;
macrophage depletion resulted in reduced formation of lung
metastases. 4 In coculture experiments, interaction between
macrophages and tumor cells facilitated invasion of tumor cells
into a collagen matrix. 39 In a chemotaxis-based in vivo invasion
assay, a paracrine loop involving macrophages and tumor cells
was essential for motility and invasion of tumor cells in mammary
tumor. 40 Colony-stimulating factor-1 (CSF-1) secreted
by carcinoma cells leads to the activation of macrophages to
secrete EGFR ligands, leading to stimulation of carcinoma cell
movement. 40
NEUTROPHILS
Neutrophil infiltration has been described in NSCLC, and
particularly in the bronchioloalveolar subtype. 41,42 Recent
studies support the fact that this could be induced by Ras activation,
one of the most common oncogenic events in pulmonary
ade nocarcinoma. 5 Neutrophils could be recruited to
tumors by CXC chemokines with an N-terminal Glu-Leu-Arg
(ELR) motif. These chemokines are also autocrine growth factors
for certain types of cancer cells. 43–45 Mutations in the
proto- oncogene KRAS occur in 10% to 30% of lung adenocarcinomas,
46 and expression of mutant KRAS in the alveolar
epithelium leads to the development of lung adenocarcinoma
in mice. 45–50 In addition to its role in the transformation of
alveolar epithelial cells, the presence of KRAS mutations is a
predictor of shorter survival in NSCLC patients 51 and of resistance
to therapy. 52 Sparmann et al.5 demonstrated that the
CXC chemokine CXCL8 (interleukin-8) is a transcriptional
target of Ras signaling and is required for the initiation of
tumor-associated inflammation and neovascularization in xenograft
models. In this model, neutralization of CXCL8 in
RasV12-expressing subcutaneous tumors attenuates neoplastic
growth; CXCL8 inhibition does not affect tumor cell proliferation
but leads to an increase in tumor cell death and an impairment
of tumor vascularization coincident with an impairment
of stromal infiltration of neutrophils. In the heterotopic and
orthotopic Lewis lung cancer models, tumor growth is associated
with enhanced neovascularization, neutrophil inflammation,
and expression of CXC chemokines. Neutralization of
CXC chemokine receptor decreases tumor size and increases
tumor necrosis. 53
In K ras LA1 mice, a mouse model in which lung adenocarcinoma
develops through somatic activation of a KRAS allele
carrying an activating mutation in codon 12 (G12D), 50 neutrophils,
and vascular endothelial cells infiltration increased
during malignant progression, and the murine functional homologues
of human CXCR2 chemokines (KC, MIP-2) and
their receptor CXCR2 are highly expressed. 54 CXCR2 inhibition
blocks the expansion of early alveolar neoplastic lesions,
but this antitumor effect does not occur outside the presence
of the tumor microenvironment.
In humans, adenocarcinomas with bronchoalveolar features
are also characterized by an intense inflammatory reaction,
predominantly consisting of alveolar neutrophils and
macrophages. Increased numbers of tumor-infiltrating neutrophils
are linked to poorer outcome in these patients. 41
The tumor environment drives local neutrophil recruitment
and activation via CXC chemokine release, but it also prolongs
alveolar neutrophil survival through the production of
soluble antiapoptotic factors GM-CSF and G-CSF. 55 The
mechanisms by which neutrophils influence the prognosis
of adenocarcinomas with bronchoalveolar features could
be multiple. It has been postulated that the persistence of
neutrophil alveolitis would result in persistent release of proinflammatory
mediators such as cytokines, proteases, and
reactive oxygen and nitrogen species that can damage DNA
and activate oncogenes. 56,57 Among these factors released by
neutrophils, hepatocyte growth factor (HGF) seems to be
particularly involved in the progression of these types of tumors,
especially through its mitogenic and scattering properties,
favoring c-Met–expressing tumor cell migration along
the alveolar basal membrane. 58 Lastly, neutrophils might be
involved in luminal tumor spread by promoting tumor cell
shedding, 59 which is described pathologically as the presence
of micropapillary clusters that are also involved in the mechanism
of aerogenous progression. 60
MAST CELLS
Several different studies showed a significant association between
mast cell density, angiogenesis, and poor prognosis in
NSCLC. 61–64 Using monoclonal antibodies for tryptase—a
specific marker for mast cells—and for endothelial cell surface
molecules, several studies quantified mast cell and microvessel
density in lung cancer tissue. Takanami et al.61 showed a
correlation between mast cell density and microvessel count
in a study of 180 patients with resected pulmonary adenocarcinoma.
Mast cell density was also associated with N classification
and was an independent factor for survival duration.
Production of angiogenic factors such as VEGF or other proinflammatory
cytokines by mast cells is probably involved in
this phenomenon. 64
DENDRITIC CELLS
Effective antitumor responses require antigen-presenting cells
(APCs), lymphocytes, and NK effectors. DCs are bone marrow–
derived leukocytes characterized by a high level of expression
of major histocompatibility complex (MHC) and
costimulatory molecules. They are the most effective APCs.
To initiate and maintain an effective antitumor response after
antigen uptake, DC should migrate to draining lymph nodes
and to prime T cells. This priming reaction is triggered by
an activation-driven maturation process of DC characterized
by upregulation of costimulatory molecules (CD40, CD80,
and CD86), a switch in the chemokine receptor repertoire,
and production of immunomodulatory cytokines (IL-12
and IFN-a) necessary for the generation of cytotoxic T lymphocytes.
65 However, immunosuppressive cytokines such as
IL-10, TGF-b, prostaglandin E2 (PGE2), and VEGF interfere
with DC maturation and migration, altering tumor response.
To improve antitumor immunity, tumor cells have
been transduced with genes encoding molecules able to attract
and to activate DC but with limited efficacy in curing
established tumors. To overcome tumor microenvironmentassociated
suppressive effect on the DC, a recent work used
a strategy that incorporates ex vivo–activated DC as the delivery
for chemokine expression. The authors transduced the
gene of the secondary lymphoid chemokine (CCL21, CCR7
receptor ligand) into DC ex vivo and delivered the genemodified
DC (DC-AdCCL21) in a mouse model of spontaneous
bronchoalveolar carcinoma. 66 A single intratracheal
administration led to a marked reduction in tumor burden
with extensive mononuclear cell infiltration of the tumors.
The reduction of tumor burden was accompanied by the
enhanced elaboration of type I cytokines (IL-12 and IFN-g
and GM-CSF) and antiangiogenic cytokines and a decrease
in immunosuppressive cytokines (IL-10, TGF-b, PGE2) in
the tumor microenvironment. 66 Continuous administration
of DC-AdCCL21 significantly prolonged survival of mice. 66
In another study, repeated treatments with a combination
of a microbial stimulus (a Toll-like receptor 9 ligand, CpG
oligonucleotide) and an antibody blocking the IL-10 receptor
reversed the functional paralysis of DC and reestablished
IL-12 production. 67 Lastly, a combination of local treatment
of CCL16 and cpG together with systemic administration of
antibody blocking the IL-10 receptor cured syngeneic tumors
in mice. 68
ADAPTATIVE IMMUNITY
Lung cancer cells themselves find a way to avoid activating
the adaptative immune system. Although they express tumor
antigens, the limited expression of MHC antigens, defective
antigen processing, and lack of costimulatory molecules make
them ineffective APC. 69 For example, the absence of expression
of costimulatory B7 molecules renders tumors invisi ble
to the immune system, whereas enhanced expression of inhibitory
B7 molecules protects them from effective T-cell
destruction. 70
Tumor-reactive T cells accumulate in the lung tumor
microenvironment but fail to respond because of suppressive
tumor cell–derived factors. These factors can reduce T-cell
survival. Lymphocytes exposed to lung tumor supernatant undergo
enhanced apoptosis with an impairment of nuclear factor
B activation due to reduced I B kinase (IKK) activity. 71
A high proportion of tumor-infiltrating lymphocytes
in the tumor microenvironment are regulatory T cells. The
CD4 CD25 T regulatory cells found in lung tumors have
been shown to selectively inhibit the host immune response
and contribute to the progression of lung cancer. They mediate
potent inhibition of autologous T-cell proliferation while they
fail to inhibit the proliferation of allogeneic T cells. 72
B cells also play a crucial role in the onset of chronic inflammation
associated with epithelial cancer development. 73 In
a recent study using a transgenic mouse model of skin carcinogenesis
where the gene of human papillomavirus 16 (HPV-16)
is expressed under control of the human keratin 14 promotor,
B cells were shown to be activated peripherally—with no need
to be recruited in neoplastic tissue. They were also shown to
initiate immunoglobulin deposition into neoplastic tissue, paralleling
the recruitment of inflammatory cells (mast cells and
granulocytes) and malignant progression. 73 Antibodies mediate
recruitment of innate immune cells via engagement of FcR
expressed on immune cells. Other studies have reported that
humoral immune responses potentiate in vivo growth and invasion
of injected murine and human tumor cell lines via recruitment
and activation of granulocytes and macrophages. 31 The
authors suggested that pharmacological intervention attenuating
B cell activation or blocking B cell–mediated recruitment
of innate immune cells may be effective in preven ting premalignant
epithelial progression.
CLINICAL IMPLICATIONS
Ongoing biochemical processes in the tumor microenvironment
create new targets for cancer therapy. One advantage of
therapies targeting the microenvironment is that these nontumor
cells are presumably genetically stable, whereas tumor
cells are genetically unstable and thus can accumulate adaptive
mutations and rapidly acquire drug resistance. Several
drugs directed against nontumor cells or their soluble mediators
have been developed and are now being evaluated in
clinical trials.
angiogenesis
and invasion of tumor cells, into both the surrounding
normal tissue and the blood and lymphatic systems. ECM is
also a rich source of sequestered heparin, binding progrowth
and proangiogenic factors, which are made available following
increased production of matrix-degrading enzymes. Clinical
trials were undertaken to determine if inhibitors of MMPs
(MMPI) improved overall survival in NSCLC or SCLC.
Marimastat, a nonselective MMPI, has also been tested in a
number of malignancies, including small cell lung cancer and
breast, gastric, and pancreatic cancers; the results were negative.
74–77 Musculoskeletal toxicity was a significant problem in
all studies. The failure of the broad-spectrum MMP inhibitors
(MMP-I) in the clinic has been explained by the fact that some
MMPs can also release antiangiogenic proteins. Prinomastat,
a more targeted MMPI with activity mainly against MMP2
and MMP9, was given versus placebo in patients with advanced
NSCLC in combination with gemcitabine–cisplatin
chemotherapy. This study was closed after an interim analysis
showed a lack of efficacy. 78 A parallel study of similar design
found no benefit when prinomastat was administered in addition
to pacli taxel and carboplatin in patients with advanced
NSCLC. 79 Another selective MMPI, BAY 12-9566, has been
evaluated in several disease settings, but after disappointing
results in studies of SCLC and pancreatic cancer, its development
has been suspended.
Fibroblasts might be a novel therapeutic target in cancer.
The cell-surface serine protease known as fibroblast activation
protein (FAP) is mostly expressed in wound healing and in
tumor stroma. A phase I dose escalation study with an antibody
directed to human FAP (sibrotuzumab) in patients
with colorectal cancer or NSCLC has shown that the antibody
bound specifically to the tumor sites. 80 Targeting CAFs as a
therapeutic strategy against cancer needs further study.
A plethora of antiangiogenic agents inhibiting either angiogenic
growth factors or their receptors have been developed
and tested in preclinical experiments. More recent data from
the clinical trials of the VEGF-specific antibody, bevacizumab
(Avastin), showed that in patients with metastatic colorectal
cancer, breast cancer, and NSCLC, there was a significant survival
benefit when combined with chemotherapy, 81,82 leading
to the Food and Drug Administration (FDA) approval of bevacizumab.
Treatment with thalidomide, another antiangiogenic
agent, was not associated with a significant improvement in survival
of SCLC patients. However, there was pronounced heterogeneity
in survival outcomes between groups of patients. 83
Some benefit was observed among patients with a performance
status (PS) of 1 or 2, showing that angiogenesis deserves further
study as a therapeutic target in this disease.
Epidemiological studies have demonstrated that people
taking nonsteroidal anti-inflammatory drugs (NSAIDs) have a
clear reduction in their risk of developing colorectal cancer, 84
and possibly other tumors. As a result, there were high expectations
for the next-generation NSAIDs, the selective COX-2
inhibitors, in the prevention and treatment of cancers associated
with chronic inflammation. Celecoxib had demonstrated
ability to reduce the incidence of colorectal cancer. 85 The addition
of rofecoxib did not improve overall survival compared
with first-line treatment with cisplatin plus gemcitabine in
patients with advanced NSCLC in a prospective, open-label,
randomized phase III trial. 86 Most of the clinical trials have
closed early because long-term high-dose COX-2 inhibitor elevates
the risk of cardiovascular events 87 ; alternative drugs will
need to be identified.

A growing body of evidence demonstrates that cancer cells
have accomplices. Quite early in tumor development, cancer
cells co-opt blood vessels and recruit leukocytes and fibroblasts,
reprogramming them to provide nourishment in the
form of peptides that support cell proliferation and metastasis.
Although their ability to dupe the host into becoming an ally
provides cancer cells with a selective advantage, it may also be
their Achilles heel. Initial efforts to elucidate the mechanisms
by which cancer cells interact with surrounding cells within the
tumor has revealed several potential therapeutic opportunities.
Future research will better define these bidirectional interactions
between tumor and host, and future clinical trials should
be designed to capitalize on this understanding.