Abstract
The intercellular cell-adhesion molecule-1 (ICAM-1) is a transmembrane molecule and a distinguished member of the Immunoglobulin superfamily of proteins that participates in many important processes, including leukocyte endothelial transmigration, cell signaling, cell-cell interaction, cell polarity and tissue stability. ICAM-1and its soluble part are highly expressed in inflammatory conditions, chronic diseases and a number of malignancies. In the present article we present the implications of ICAM-1 in the progression and prognosis of one of the major global killers of our era: lung cancer.
- Intercellular cell-adhesion molecule
- immunoglobulin
- cell signaling
- cell-cell interaction
- cell polarity
- prognosis
- lung cancer
- review
Lung cancer, despite all therapeutic efforts, remains the leading cause of cancer-related death worldwide, with approximately 160,340 deaths in the United States in 2012 (1). Its high mortality is largely attributed to early local expansion and metastasis, as most patients present with locally advanced or metastatic (up to 75%) disease at the time of diagnosis (2, 3), while in a large percentage of the remaining patients, early stage disease rapidly progresses to metastatic. Since early-diagnosis provides the only chance for a potentially curative treatment, several studies have focused on the identification of serum biomarkers who could be used as diagnostic, staging, prognostic or predictive markers in lung cancer. Adhesion molecules have been the subject of many of these studies.
Adhesion between cells and their microenvironment, cell proliferation and cell migration are fundamental for the embryonic development and the establishment and maintenance of cell polarity and tissue architecture (4, 5). These processes are promoted by a variety of cell adhesion molecules in normal tissue. In malignancy, the aberrant expression of cell adhesion molecules results in impaired adhesive properties and cell-cell communication (6, 7). This provides a substratum for neoplastic cells to circumvent the control of differentiation, acquire a malignant phenotype, become motile and invasive and, finally, achieve colonization of distant organs (8, 9). Changes in expression of cell adhesion molecules belonging to the integrin, cadherin and immunoglobulin superfamilies (4) have been observed in several cancer types (10) and associated with tumor progression (6).
The intercellular cell adhesion molecule-1 (ICAM-1) is a member of the immunoglobulin (Ig) superfamily of proteins expressed in several cell types. ICAM-1 plays a key role in inflammatory conditions, nervous system development, immune responses through antigen recognition and lymphocyte circulation and activation (11, 12). Particularly, ICAM-1 is involved in signal transduction across cell membranes, leukocyte-leukocyte, leukocyte-endothelial and leukocyte-epithelial cell interactions, transendothelial migration and adhesion-dependent respiratory bursts (13-15). ICAM-1 shows altered expression in many benign and malignant diseases. The evaluation of the expression of ICAM-1 and its soluble form in lung cancer has gathered significant interest in recent years, although there still remain many questions unanswered. This narrative review focuses on the implications of ICAM-1 in lung cancer evolution and prognosis. A literature review was conducted through the Pubmed and Medline databases, using the lung cancer, intercellular adhesion molecule-1 or ICAM-1, and prognosis as key words. An attempt was made to gather all relevant articles and appropriate references extracted from these articles to compose this review and, ultimately, evaluate the role of ICAM-1 in lung cancer.
Intercellular Cell Adhesion Molecule-1 (ICAM-1)
Intercellular adhesion molecules (ICAMs) are structurally-related members of the immunoglobulin (Ig) superfamily. Five molecules have already been identified (ICAM-1 to 5) and classified together as they contain two or more of an immunoglobulin-like domain of 90 to 100 amino acids. These immunoglobulin-like domains are expressed on the extracellular portion of the protein (16). ICAM-1 (CD54) is a type I transmembrane protein with a molecular weight of 74-114 kDa, depending on its level of glycosylation. The extracellular portion of ICAM-1 consists of 5 immunoglobulin-like domains. The endoomain has a cytoplasmic tail containing a tyrosine with signaling properties that appears to interact with the cytoskeleton protein, α-actinin, the linker molecule ezrin or cortactin (17). A molecule with a suggested crucial role in ICAM-1 signaling cascades is the hydrolase of guanosine triphosphate (GTPase) Rho member of the Ras family of G-proteins, which interferes with the ligation of ICAM-1 on the plasma membrane during the re-arrangement of the actin cytoskeleton (18). Antibody-crosslinked ICAM-1 molecules lacking cytoplasmic tails were not capable of activating Rho proteins (19). ICAM-1 is expressed constitutively at low levels on vascular endothelial cells and on some lymphocytes and monocytes, macrophages, B- and T-lymphocytes, fibroblasts and epithelial cells (20). Furthermore, ICAM-1 is an accessory molecule stabilizing the T-cell receptor-mediated binding between antigen-presenting cells (APC) and T-lymphocytes. ICAM-1 is a specific ligand for leukocyte-function associated antigen-1 (LFA-1), interleukin-1 (IL-1), tumor necrosis factor-1 (TNF-1) and interferon-γ (IFN-γ) (21). Today, there is enough evidence that ICAM-1 modulates and participates in many key processes such as the maintenance of cell-cell interactions and signal transduction, cytoskeleton rearrangement, leukocyte transmigration during inflammation, T-cell stimulation and neovascularization (7, 12, 22-24). Despite ICAM-1's crucial role in immune responses, it has also been associated with tumor progression and prognosis in various types of cancer, including gastrointestinal cancer, breast cancer and melanoma (25-27). ICAM-1 expression in cancer cells most likely reflects the interaction between the tumor and the immune system. ICAM-1 shows altered expression in many benign and malignant diseases, potentially playing a role in tumorigenesis and enhancing the metastatic ability of malignant tumors (28, 29).
Soluble ICAM-1 (sICAM-1)
Soluble ICAM-1 (sICAM-1) is a molecule with 5 extracellular domains, but no transmembrane and cytoplasmic domains have been detected in the serum. In vitro studies have demonstrated that sICAM-1 stimulates proinflammatory cascades (30). Thus, increased levels of sICAM-1 have been found in patients with atherosclerosis and cardiovascular disease, autoimmune disease and in patients with airway inflammation disorders such as asthma and chronic obstructive pulmonary disease (COPD), in many cases reflecting the severity of the disease (31-33). sICAM-1 is the product of the proteolytic cleavage of cell surface ICAM-1, as a part of a negative feedback loop (34). Tumor necrosis factor-1 (TNF-1), interleukin-1 (IL-1) and matrix metalloproteinases (MMPs) act as proteases which are responsible for the shedding of sICAM-1 from the cell surface through phosphathidyl inositol 3-kinase (PI3K), mitogen-activated protein kinase (MAPK), nuclear factor kappa light-chain-enhancer of activated B cells (NF-KB), extracellular regulated kinase 1/2 (ERK-1/2) and non-receptor tyrosine protein (Src) kinase signaling pathways (35). In malignant disorders, sICAM-1 has the ability to bind to circulating cytotoxic lymphocytes, allowing tumors to escape immune recognition (36). Additionally, sICAM-1 has a substantial role in angiogenesis and stimulates tumor cell growth (37). Increased sICAM-1 levels have been detected in patients with solid and hematological tumors. In most cases, increased sICAM-1 serum concentrations have been correlated with advanced disease stage and dismal prognosis (38-40).
ICAM-1 Expression in Benign and Malignant Bronchial Epithelium
The bronchial epithelium over its large surface area is exposed to physical and infectious agents. As part of the defense system, leukocytes are recruited in order to neutralize any possible threat. However, airway hyper-responsiveness is the result of excessive numbers of interstitial leukocytes, as observed in COPD and asthma (41, 42). The accumulation of leukocytes within the interstitium depends on the recruitment of leukocytes from the blood and interstitial space and the survival of tissue leukocytes. The interaction between LFA-1 and ICAM-1 is a prerequisite for lymphocyte accumulation on lung epithelial cells (43). ICAM-1 is expressed mainly on the surface of activated endothelial cells in the pulmonary vasculature and in type II pneumocytes (44). Migration of leukocytes across the endothelial monolayers is both transcellular and paracellular. On the other hand, the movement of T-cells across the bronchial epithelial barrier requires a 3-step process involving adhesion of T-cells onto the basal surface of the epithelium, recognition of the epithelial junctions by T cells promoted by of LFA-1 on T-cells and ICAM-1 on epithelial cells, and T cell transmigration across the epithelial tight junctions (45). The functional consequences of cytokine-induced up-regulation of ICAM-1 expression by lung cancer cells have been investigated since the early 1990s. The presence of monocytes and T-cell lymphocytes within the neoplastic cells triggers the release of cytokines which induce the expression of ICAM-1 (46). ICAM-1 has been identified in most cases of all types of non-small cell lung cancer cultured lines (NSCLC) and, rarely, in small-cell lung cancer (SCLC) cell lines. Early experimental in vitro studies with squamous cell carcinomas showed that LFA-1-positive tumor-infiltrating lymphocytes actively express messenger ribonucleic acid (mRNA) transcripts for IFN-γ, suggesting that the ability of squamous carcinoma cells to express ICAM-1, usually with basal distribution, is associated with the recruitment and activation of tumor-infiltrating lymphocytes, which, in turn, might up-regulate ICAM-1 expression by tumor cells (46, 47). It was also found that increased surface expression of ICAM-1 on TNF treated NCI H460 and H211 cells is associated with increased adherence to LAK (Lymphokine-activated killer) cells and increased susceptibility to LAK-cell cytotoxicity and that these effects are related, at least partly, to ICAM-1/LFA-1 interaction. The expression of ICAM-1 by tumor cells may be involved in the regulation of their susceptibility to the cytotoxic effect of autologous and allogeneic defensive cells based on the fact that LFA-1 is expressed on mononuclear phagocytes, cytotoxic T-cells, natural killers (NK) and LAK (48).
Other studies, by Passlick et al. (49) and Jiang et al. (50) have reported that ICAM-1 expression in adenocarcinomas has been found selectively increased in alveolar epithelial cells. Interestingly, all surrounding non-malignant tissue compartments, including lung endothelial cells, pulmonary lymphocytes and interstitial fibroblasts, demonstrated ICAM-1 expression. On the contrary, SCLC lines showed no basal ICAM-1 expression. More recently, the potential role of ICAM-1 in lung cancer progression has been evaluated with the use of monoclonal antibodies against cell surface ICAM-1. Brooks et al. (51) examined ICAM-1-positive lung cancer cell lines established in SCID mice. UV3, an antibody against ICAM-1, was administered subcutaneously into the tumor. The antibody managed to slow the growth of all tumors, although it had no curative effect. The researchers observed simolar inhibitory results in cell lines from breast, prostate and pancreatic cancer. In SCLC cell lines (H24, H69, H82), the adherence of cancer cells to cultured vascular endothelium in stasis and flow depends on the expression of ICAM-1. After blocking endothelial ICAM-1, adhesion was reduced. It may be speculated that ICAM-1 is a molecule crucially involved in SCLC endothelial adhesion (52). The decrease in ICAM-1 expression on lung cancer cells may also be the result of TNF suppression through the inhibitory effect of natural killer-B cells (NK-B) to the ICAM-1 promoter. These observations were made after administration of thalidomide in lung cancer patients and immunohistochemical evaluation of tumor specimens (53). The important role of NF-kB in the activation of ICAM-1 was demonstrated in human lung cancer A549 cells. The receptor activator of NF-kB ligand (RANKL), known for its substantial role in osteoclastogenesis and tumor metastasis, directed the migration of A549 cells through interaction with MEK/ERK leading to the activation of NF-kB and resulting in ICAM-1 expression and cell migration (54).
Serum Soluble ICAM-1 in Lung Cancer Patients
In the last 20 years, a growing body of data concerning sICAM-1 serum levels in patients with lung cancer has been collected in an attempt to shed more light on the potential role of sICAM-1 in lung cancer growth, disease progression and prognosis (all available data are summarized in Table I). As already mentioned, in other types of malignancies, elevated serum sICAM-1 levels are associated with extensive tumor burden and poor outcome. Early studies have examined the possible differences in sICAM-1 levels between lung cancer patients, patients with benign lung diseases, healthy smokers and non-smokers (55-61). It has been discovered that the smoking habit per se is a main cause of sICAM-1 serum levels increase. Moreover, patients with COPD and asthma showed significantly increased sICAM-1 levels when compared with non-smokers. Patients with lung cancer also demonstrated greatly increased sICAM-1 concentrations when compared to all other studied groups of subjects. Interestingly, no difference in sICAM-1 serum levels and histological ICAM-1 expression between various histological tumor types was observed. The correlation between sICAM-1 levels and ICAM-1 tumor expression has also been evaluated.
Grothey et al. analyzed sICAM-1 serum levels in 51 NSCLC patients and 40 healthy individuals (smokers and non-smokers) and performed immunohistochemical analysis of ICAM-1 expression in tumor cells in 20 NSCLC patients. They observed that patients with tumors demonstrating ICAM-1 positivity between 30%-60% had higher sICAM-1 levels compared to those who showed ICAM-1 cell positivity of 0%-1% (55). Disease stage was found to be a factor with significant impact on sICAM-1 baseline levels in the majority of studies. Serum sICAM-1 levels were related to tumor stage in several studies involving NSCLC and SCLC patients, with a remarkable difference between metastatic and localized disease (56, 58-61). However, conflicting results have been published demonstrating no difference between patients with early and advanced disease (58, 62, 64). This controversy can be partially explained by the fact that samples were small and different types of ELISA kits were used.
Recent studies have evaluated the predictive and prognostic role of sICAM-1 during chemotherapy (61, 65). The main issues were, firstly, to investigate whether baseline sICAM-1 levels could be of independent prognostic value despite the responses achieved from chemotherapy; and, secondly, whether the decrease in serum sICAM-1 levels after administration of chemotherapy could be used as a surrogate marker of objective response. In 2008, Guney and colleagues (61) measured sICAM-1 serum levels in 57 NSCLC patients before and after combination chemotherapy containing cisplatin and compared them with those of 24 healthy controls. They found that serum sICAM-1 levels decreased after chemotherapy (1724,07 ng/mL before chemotherapy, 1514,17 ng/mL after chemotherapy; p=0.05), independently from chemotherapy response. In the study published by Young et al. in 2012 (65), plasma samples from 95 patients with NSCLC (n=48) and SCLC (n=47) chosen from a pool of 1,446 patients who participated in two large multi-center, randomized, phase III trials were analyzed. The first trial enrolled 724 patients with SCLC. They were treated with up to 6 cycles of carboplatin and etoposide chemotherapy every 3 weeks and randomly assigned to receive either oral thalidomide or matching placebo for up to 2 years. The second trial involved a total of 722 patients with stage IIIB or IV NSCLC who were treated with up to four cycles of gemcitabine and carboplatin chemotherapy every 3 weeks and randomly assigned to receive either oral thalidomide or placebo for 2 years. In the biomarker study of 95 patients, serum sICAM-1 was measured before and after 3 cycles of chemotherapy. Sub-group analysis identified a significant relationship between progression-free survival (PFS) and baseline sICAM-1 in SCLC for pooled treatment arms, but not in NSCLC, with a hazard ratio (HR) for progression with high baseline sICAM-1 in SCLC of 2.20 (95% confidence interval, CI. 1.16-4.18; p=0.016). In both NSCLC and SCLC patients, serum sICAM-1 was reduced after chemotherapy. However, reduced sICAM-1 levels could not predict response to chemotherapy (42, 46).
sICAM-1 levels in various studies in lung cancer.
In another study by Quian et al. published in 2011 (66), including 124 NSCLC patients who received 4 cycles of platimun-based chemotherapy, the atuhors evaluated serum levels of sICAM-1 at baseline and after every two courses of chemotherapy. Serum sICAM-1 levels were found significantly decreased compared to baseline levels (p=0.012). The median reduction was 21% in responders and 9% in non-responders (p=0.097). For an observed sICAM-1 post-treatment reduction of 11.5%, an objective response to chemotherapy could be predicted with a sensitivity of 62% and specificity of 82% (area under the curve, AUC=0.73. 95% CI=0.59-0.87). Patients with low baseline sICAM-1 levels survived longer than those who had higher baseline concentrations. Survival analysis also demonstrated an independent prognostic role for serum sICAM-1 (HR=1.12, 95% CI=1.01-1.21) (66). In the E4599 phase III study (Bevacizumab approval), 878 patients were randomized to receive paclitaxel and carboplatin with our without bevacizumab. Chemotherapy was repeated every 21 days up to a total of six cycles. Serum from 113/878 patients was collected both before treatment and after two cycles of chemotherapy. Baseline sICAM-1 showed significant associations with response and survival in both groups. Patients with low baseline sICAM-1 levels had a higher response rate (32% vs. 4%, p=0.02), better overall survival (p=0.00005) and 1-year survival (65% vs. 25%) than those with high sICAM levels, respectively. sICAM-1 levels were also predictive for response. Patients with low sICAM-1 levels (260 ng/ml) demonstrated a response rate of 32%. On the contrary, in the high-sICAM-1 group the response rate was 14% (p=0.02). However, there was no significant decrease in the sICAM-1 levels after 2 cycles of chemotherapy (67).
Serum sICAM-1 levels in malignant pleural effusions have also been an area of interest. In a study of 79 adenocarcinomas of the lung, serum sICAM-1 was confirmed as an independent prognostic factor for PFS (68). sICAM-1 may increase in both inflammatory and malignant effusions. Secretion of sICAM-1 might be a sign of generalized inflammation within the pleural space (69).
Discussion
ICAM-1 expression in lung cancer tissue is particularly prominent in bronchial and alveolar epithelial cells. ICAM-1 expression is not affected by histological type. Tumor tissue ICAM-1 expression and serum sICAM-1 levels are significantly correlated, probably because tumor cells are the source of sICAM-1. This cannot be considered as evidence that the released sICASM-1 contributes to the elevated serum levels found in patients with lung cancer. On the other hand, sICAM-1 levels have been found increased in a number of benign and chronic diseases. Thus, sICAM-1 production could represent a process due to non-specific host defence mechanisms. This case is supported by the observation of elevated sICAM-1 levels in healthy smokers. Membrane-bound ICAM-1 levels stimulate T-cell receptor-mediated cellular immune response. Furthermore, the up-regulation of ICAM-1 expression on tumor cells is followed by the extracellular release of sICAM-1. Shedding of sICAM-1 in the extracellular space may be one of the mechanisms by which tumor cells escape cell-mediated cytotoxicity and lysis by the host cellular immune system. Also, sICAM-1 may prevent T-cell antigen-presenting cells from interacting or modifying the functional status of leukocytes. sICAM-1 may also be produced as a secreted splice variant of ICAM-1 without intracellular domains (70, 71). Circulating sICAM-1 can block lymphocyte attachment to endothelial cells, natural-killer cell toxicity and major histocompability complex restricted T cell-tumor interactions. sICAM-1 increases as part of the immune system process against inflammation or it may act as an immunomodulator.
ICAM-1 seems to play an active role in the metastatic process. We have already mentioned the ability of SCLC cell lines to adhere to vascular endothelium via ICAM-1 expression. Likewise, tumor cells may imitate leukocytes in their adhesion surface and use this pathway for vascular adhesion during haematogenous metastasis (72). Thus, through ICAM-1, two essential conditions of tumor growth are fulfilled – escape from immune surveillance and angiogenesis.
From a clinical point of view, baseline sICAM-1 levels seem to have a prognostic value and may be an independent predictor of the outcome in both NSCLC and SCLC. Also, sICAM-1 levels appear to be closely-dependent on the disease burden. The mechanisms resulting in sICAM-1 responses during disease progression and chemotherapy have not been elucidated. Monitoring sICAM-1 levels during treatment may be an adjunctive marker of response or progression. The increase of sICAM-1 expression during the progression of lung cancer is usually associated with poorer outcome compared to patients with dropping sICAM-1 levels. There might also be a good correlation between objective radiologic response and sICAM-1.The fashion in which serum sICAM-1 levels decrease after the induction of chemotherapy is unclear. However, these conclusions are not definitive. Racial differences among patient cohorts could constitute a source of inconsistency. So, are sICAM-1 levels useful and reliable for making clinical and treatment decisions? Our concluding remark is that sICAM-1 could be a potential surrogate and prognostic marker in lung cancer during first-line chemotherapy. However, taking into account that most of the presented studies enrolled small numbers of patients, the available data needs validation through prospective studies on a much broader scale.
Footnotes
-
Conflicts of Interest
The Authors report no declarations of interest.
- Received April 24, 2014.
- Revision received June 24, 2014.
- Accepted June 25, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved