Evaluation of MCAM expression in correlation with clinicopathological parameters of gastric cancer.

Introduction
Gastric cancer is a malignancy most frequently arising from glandular cells. Recent years have brought a continuous decrease in the incidence of gastric cancer. Annually, it is diagnosed in approximately one million of people, of whom around 783.000 die. Moreover, gastric cancer is the fifth most commonly detected malignancy worldwide and the third most frequently causing death. Mortality rate due to gastric cancers accounts for approximately 8.2% of all deaths due to malignant cancers worldwide1,2.
MCAM (Melanoma Cell Adhesion Molecule) is an adhesion protein belonging to the immunoglobulin superfamily, which was identified as a melanoma-specific protein. It is composed of the external part, transmembrane domain and a cytoplasmic tail3,4. Its major function is to contribute to the process of cell adhesion. MCAM is relatively poorly effective compared to other adhesion proteins, e.g. cadherins. It is involved both in physiological and pathological processes5.
High expression of MCAM protein is observed particularly in the epithelial cells of the developing nervous system, trachea, kidneys and uterus. It is also closely related to the immune system as it is responsible for the maintenance of the thymic structure. Moreover, it is involved in the mediation of transmembrane migration of lymphocytes. The expression of MCAM has been identified on a small population of T and B cells in the peripheral blood of healthy people. The protein is believed to take part in the recruitment of activated T cells to the inflammation site4,6–8.
The MCAM protein is intensively engaged in the process of angiogenesis. Its high level is observed during embryonic development of the vascular system. A drop in MCAM expression results in weaker development of blood vessels in the young body. Angiogenesis is also an important process in the neoplastic growth. A tumor can develop and metastasize to other organs through its own network of vessels. The presence of MCAM has been revealed not only in melanoma cells but also in its vascular epithelium. Then, it has been proved that MCAM takes an active part in the process of angiogenesis in different neoplastic tumors. The role of MCAM in this process results from the fact that this protein is a component of VEGF (vascular epithelial growth factor) considered to be a co-receptor with VEGFR2 (receptor for vascular epithelial growth factor)4,9–13.
Research has revealed that the level of MCAM is significantly elevated in malignancies and metastases of numerous cancers (affecting the prostate, the breast, the ovary, the lungs, etc.), which is associated with the promotion of growth and metastasizing of tumors. It has been noted that the mechanisms responsible for increased invasiveness of many neoplastic cells involve the ability of MCAM to modulate the expression of proteins taking part in apoptosis, proliferation or angiogenesis. An increase in the level of MCAM is associated with the elevated level of VEGF and VEGFR2, i.e. factors indispensable for angiogenesis. The protein has been shown to contribute to lymphatic metastasizing4,14–20.
The aim of the study was to assess the expression of MCAM in gastric cancer using the immunohistochemical method and to analyze the correlation of MCAM expression with chosen clinical-pathological parameters such as age and sex of patients, tumor diameter, location, histological type, malignancy grade, infiltration depth (pT), lymph node involvement (pN), presence of distant metastases (pM), infiltration of blood and lymphatic vessels, Lauren classification type, desmoplasia grade, H. pylori infection grade and patients’ survival.

Results
Positive expression of MCAM was found in neoplastic cells in 42.5% (37/87) of gastric cancer patients. In cancer stroma was noted in 33.33% (29/87) of study patients (Fig. 1.).

The statistical analysis showed no correlation of MCAM expression in neoplastic cells with sex, tumor diameter and location, histological type and grade, infiltration depth (pT), lymph node involvement (pN), the presence of distant metastases (pM), infiltration of lymphatic and blood vessels in the Lauren classification.
Positive expression of MCAM was observed to be more frequent in patients with low desmoplasia (p = 0.019). A statistically significant correlation was found between MCAM expression and Helicobacter pylori grade (p = 0.037) (Table 1).

No statistically significant correlation was noted between MCAM expression in neoplastic cells and patients’ survival (p = 0.544) (Fig. 2.)

MCAM expression in cancer stroma was not found to correlate with clinical-histopathological parameters and patients’ survival. Data are not shown.

Discussion
For many years the role of adhesion proteins in the development and metastasizing of numerous cancers has been emphasized. The proteins perform major functions both in physiological and pathological processes. Since they are responsible for cell adhesion, loss of their expression may be associated with migration and metastatic potential. Moreover, they are capable of interacting with the epidermal growth factor receptor (EGFR), due to which they may be involved in the development, growth and differentiation of neoplastic tumors. Their function that facilitates leukocyte adhesion, rolling and permeation to tissues may indicate their role in peritumoral inflammation21.
The role of all adhesion proteins in gastric cancer has not been explicitly described. Among a large number of these molecules some contribute to tumor development, others inhibit it. E-cadherin is the most frequently studied adhesion protein. It has been proved that a decrease in the level of E-cadherin is associated with the induction of epithelial-mesenchymal transformation (EMT) and with acquiring the ability to infiltrate and metastasize by neoplastic cells22. Likewise, MCAM takes part in the process of cell adhesion. It is responsible for intercellular junctions and connections of cells with the extracellular space. MCAM is involved in the process of adhesion in trophoblastic cells, which equips them with differentiating potential. This is essential for normal fetus growth23.
The results of our study showed positive expression of MCAM in gastric cancer cells in 42.5% of patients. This corresponds with the findings reported by Liu et al.24, who reported positive MCAM expression in neoplastic cells in 59/144 (41.0%) patients with gastric cancer. Moreover, our study revealed positive expression of MCAM in cancer stroma in 33.33% of patients. Thus, it seems that MCAM is more frequently found in neoplastic cells than in stromal cells. CAFs occur in tumor micro-environment and are responsible for tumor growth. They produce many growth factors that allow for extracellular matrix remodeling, growth of cancer cells and angiogenesis24. MCAM in turn is a marker of mesenchymal matrix cells that are CAF precursors4. Considering the above role of these cells as well as the previously described function of MCAM in the neoplastic process we decided to assess MCAM expression in stroma.
Our statistical analysis of the relationships between MCAM expression in the stroma of gastric cancer and chosen clinical-histopathological parameters and overall survival rate failed to reveal any significant correlations. Zheng et al.25 investigated the correlation of MCAM expression in gastric cancer stroma with clinical-histopathological parameters and assessed the expression of MCAM on cell lines derived from pancreatic cancer fibroblasts. They found statistically significant correlations between reduced MCAM expression in fibroblasts and higher tumor stage, higher histological grade and more frequent residual tumor status. Moreover, they showed higher overall survival rate in patients with positive MCAM expression in fibroblasts. Investigations conducted on cell lines revealed that the increase in MCAM expression was associated with the inhibition of migration and invasion of these cells. These results indicate that a drop in MCAM expression is connected with higher tumor stage.
We found no statistically significant correlations of the expression of MCAM protein in gastric cancer cells with age and sex of patients, tumor diameter and location, histological type, grade (G), infiltration depth (pT), lymph node involvement (pN), presence of distant metastases (pM), infiltration to lymphatic and blood vessels, lymph node involvement, the presence of distant metastases or type in Lauren classification.
In turn, a study performed by Liu et al.24revealed a correlation between increased MCAM expression and the presence of gastric cancer metastases to lymph nodes. Similar correlations with clinical-histopathological parameters have been observed in other neoplasms. Wang et al.26 studying tissues with urinary bladder cancer, Zabouo et al.27 tissues and cell cultures with breast cancer, and Wu et al.28 tissues and cell lines with prostate cancer observed a correlation between high expression of MCAM and metastasizing. Moreover, the increased expression of MCAM is associated with the activation of epithelial-mesenchymal transformation in breast tumors29 or angiogenesis in melanoma30. Like in other neoplasms, this shows that high expression of MCAM may be associated with poorer prognosis in gastric cancer and accelerate the disease progression.
We revealed statistically significant correlations between MCAM expression and desmoplasia grade. We observed positive MCAM expression in 51.79% of gastric cancer patients with low desmoplasia grade and in 25.81% of those with high desmoplasia grade (p = 0.019). Desmoplasia, which is a process leading to the proliferation of connective tissue components of the stroma, has a major significance for the growth of tumor, which by building its own network of blood vessels can more easily develop and migrate. It is believed that low desmoplasia tumors grow faster and are more aggressive, and have thus poor prognosis31. There are few publications on the characteristics of the stroma in gastric cancer. Kemi et al.32 were the first to demonstrate the maturity of the stroma as an independent prognostic factor in gastric adenocarcinoma, using a two-level system to categorize tumors into mature and immature stroma groups. A high percentage of stroma was found to be an independent prognostic factor in both intestinal and diffuse histological subtypes of gastric adenocarcinoma. Pun et al.33 also assessed stromal maturity at both the invasive front and in the main tumor mass and used the most immature stromal type for their analysis, indicating its association with tumor progression and recurrence-free survival of intestinal-type gastric adenocarcinoma. In contrast, Chiaravalli et al.34 evaluated desmoplasia in diffuse gastric cancer and showed that tumors with embedded desmoplasia demonstrated an increased rate of early submucosal tumors, lower rates of lymph node metastasis and distant metastasis, and significantly more favorable survival. The use of desmoplastic response assessment in routine clinical practice is being considered, but differences in criteria between studies indicate the need for standardized this assessment.
MCAM plays a role in the formation of new vessels, which is an important desmoplasia component. In our study, we focused not on the maturity of the stroma but divided cancers with poor or prominent demoplasia. An increase in MCAM expression in gastric cancers with low desmoplasia may suggest its involvement in the growth of cancer which will acquire metastatic potential. There are only few publications describing the correlations between MCAM and desmoplasia, and thus further research is needed to elucidate the role of this protein in the pathogenesis of this phenomenon.
We also revealed a correlation between MCAM expression and the grade of coexisting H. pylori infection (p = 0.037). Positive expression was found in 37.88% of cancers without the accompanying infection, in 53.85% with moderate and 83.33% with high grade infection. No literature reports on this type of correlation are available. It is, however, known that the infection is accompanied by inflammation being a response of the body against the pathogen. The relationship of MCAM expression with higher H. pylori infection may indicate the role of this protein in the formation of inflammatory reaction. Some literature reports have described the presence of MCAM expression on the surface of inflammatory cells. Elshal et al.35 conducted research assessing the expression of MCAM on the surface of lymphocytes using flow cytometry. Their findings suggest that MCAM is present on approximately 2% of T cells and 1% of B cells. Moreover, they found out that the presence of this protein on the surface of these inflammatory cells supports their ability to adhere to endothelium, which seems to confirm the role of MCAM in the development of the inflammatory condition.
The analysis of the relationship between MCAM expression and overall survival rate in patients with gastric cancer showed no statistical significance. Liu et al.24 in their study conducted on 144 cases of gastric cancer found that patients with positive MCAM expression had a shorter overall survival time (p < 0.001), showing that the increased MCAM expression in gastric cancer patients is connected with poorer prognosis. Likewise, other researchers who examined MCAM expression in other neoplasms noted a correlation between higher expression of this protein and a shorter overall survival rate. Yuan-Ke et al.29 revealed this correlation in breast cancer, Zhang et al.36 in non-small cell lung cancer, Oka et al.37 in lung cancer, Jiang et al.38 in hepatic cell cancer, Aldovini et al.18 in ovarian cancer. However, the results reported by Bai et al.39 indicate that patients suffering from clear-cell renal-cell carcinoma with higher MCAM expression showed longer survival time as compared to those with low expression of the protein.
Our study contributes to our understanding of the involvement of the MCAM protein in gastric cancer, but it has several limitations. Firstly, we were unable to consider the molecular classification of gastric cancers due to a lack of data on EBV status, microsatellite instability (MSI) and the mutation status of other genes. These molecular types are associated with disease progression and help to select the most effective therapy. Secondly, as only one observer assessed the desmoplastic response, we were unable to assess inter-observer variability. The assessment of the desmoplastic response is not well defined, so we used our own method in this study.
Since a small number of literature reports deal with the role of MCAM in gastric cancer it is difficult to explicitly determine the involvement of this adhesion molecule in the development of this neoplasm. The assessment of MCAM does not seem to be a useful marker of gastric cancer stage or its prognostic factor. However, this protein may contribute to the process of desmoplastic stroma formation and be involved in the mechanism of inflammatory reaction.

Methods

Study group
The study involved a group of 87 patients with gastric cancer treated surgically in the Second Department of General and Gastroenterological Surgery, Medical University of Bialystok in the years 2005–2015. The material was subjected to histopathological analysis in the Department of General Pathomorphology, Medical University of Bialystok.
Each neoplastic tumor was cut parallel to the longest axis and at least one total 2–3 mm thick cross-section was collected for analysis. It was then divided into small blocks, 1–1.5 cm in diameter, to obtain in this way 4–8 segments from each tumor, comprising the tumor and the adjacent macroscopically unchanged tissues. For histopathological assessment local lymph nodes, resection margins and the greater omentum were also collected. The segments were fixed in 10% buffered formaldehyde solution for 24–48 h and next embedded in paraffin at a temp. of 56 °C.
The small paraffin blocks were cut into 4 μm thick sections. Next, they were stained with hematoxylin-eosin (HE). In routine histopathological assessment of the sections histological type, malignancy grade (G) and tumor stage (pTNM) were assessed. Distant metastases were detected by radiology and fine-needle biopsy.
The grade of Helicobacter pylori infection was assessed in the tissue adjacent to the tumor but macroscopically unchanged. The infection was classified according to the Sydney system40 and a 4-stage scale was used: 0 – lack, 1 – low, 2 – moderate, 3 – high.
A desmoplastic reaction was assesed at low magnification in the main tumor mass and scored semiquantitavely as poor (absent or less than 10% of the lining) and prominent (marked diffuse desmoplasia between tumor cells of any morphology). Tumor stroma in the submucosa or in the muscularis propria, as well as areas of stroma around microscopic abscesses were not taken into consideration. Evaluation was performed in HE and Trichrome stained specimens. Representative images are shown in Fig. 3.

Immunohistochemical analysis
Small paraffin blocks with tissues were cut on a microtome into 4 μm thick sections, which were deparaffinized in xylenes and hydrated in alcohols. Next, they were embedded in citrate buffer (pH = 6.0) and heated in a water bath for 20 min. at a temp. of 98.5oC to reveal the antigen and then for 20 min. at a room temperature. They were incubated with 3% hydrogen peroxide to block endogenous peroxidase and with 1% bovine serum to block the nonspecific bonds. In the next stage, the sections were incubated with rabbit polyclonal anti-MCAM antibody (clone HPA008848, Sigma Life Sciences, Sweden; dilution 1:200) for 60 min at a room temperature. Following the reaction conducted in the polymer technique (ImmPRESS Polymer Detection Kit, Vector Laboratories, Germany) the antigen-antibody complex was visualized by applying 3,3-diaminobenzidine chromogen (ImmPACT DAB, Vector Laboratories, Germany). Cell nuclei were stained with hematoxylin.

Assessment of MCAM expression
The immunohistochemical reaction for MCAM was observed both in the cytoplasm and cell membrane of neoplastic cells. Moreover, its expression was observed in the wall of blood vessels and in tumor stromal connective tissue cells (Fig. 1). The percentage of positive cells was calculated in 100 neoplastic cells and in stromal connective tissue cells in each preparation, at 400x magnification. The expression of MCAM in neoplastic cells was assessed semi-quantitatively: the reaction was considered negative when it was lacking or visible in < 10% of cells, and positive when observed in ≥ 10% of cells. In stromal connective tissue cells the reaction for MCAM was considered positive when present in ≥ 10% of cancer stroma cells and negative when seen in < 10% of cancer stroma cells. The expression of MCAM in the wall of blood vessels was not taken into consideration.

Statistical analysis
The statistical analysis used the Statistica 13.3, StatSoft program, Poland. The comparison between the expression of MCAM and clinical-histopathological parameters were calculated using the Mann-Whitney U test for two groups, and the Kruskal-Wallis test fo three and more groups. Additionally, for the Kruskal-Wallis test, the Dunn’s Multiple Comparison post hoc test was conducted. The likelihood of the total survival time was assessed by means of the Kaplan-Meier survival function estimator. The p < 0.05 was considered statistically significant. Missing data were deleted pairwise.