Abstract
Background: Currently, there is no effective therapy for advanced gastric cancer. In this study, we investigated whether protein expression of CXCL12 and/or its receptor CXCR4 is associated with clinicopathological features and/or survival of gastric cancer. Materials and Methods: Primary tumor specimens from patients (n=137) with pathologically-confirmed gastric cancer, collected between 2001 and 2009, were analyzed by immunohistochemistry using anti-CXCL12 and anti-CXCR4 antibodies. Results: Expression of CXCL12 was directly associated with tumor differentiation (p=0.0143) but inversely associated with depth of invasion (p=0.0255), lymphatic invasion (p=0.0173), venous invasion (p=0.0022) and stage (p=0.049). Expression of CXCR4 was associated with depth of invasion (p=0.005) and stage (p=0.028). Increased CXCR4 expression, but not CXCL12 expression, was associated with 5-year cancer-specific survival (p=0.0079). Conclusion: CXCL12 was not associated with survival. Positive CXCR4 expression in gastric carcinoma was significantly associated with poor survival and, therefore, may be a potential biomarker for predicting poor survival.
Gastric cancer is a common disease and remains a major health concern in the Far East as well as in other countries worldwide (1). Prognosis of early-stage gastric cancer is excellent and various minimally-invasive procedures have been exploited to dissect and cure these lesions (2). However, patients with advanced gastric cancer have a poor prognosis and although various treatments have been developed, currently there is no effective therapy available. Therefore, biomarkers that are more precise are required to predict prognosis or aid in the pathological diagnosis.
At present, the properties of chemokines, including trafficking of leukocytes during immune and inflammatory reactions and organ selective migration of cancer cells during metastasis, have been widely recognized. Chemokines are a superfamily of small peptide molecules involved in a number of physiological processes. In humans, this superfamily includes more than 46 ligands and 18 chemokine receptors (3). Under normal physiological conditions, chemokines play a role in both pro-inflammatory and non-inflammatory cell homing. Chemokines are 8-10 kDa cytokines that are classified into four groups (CXC, CC, C and CX3C). Chemokine receptors belong to the G-protein-coupled receptor superfamily. Chemokines/chemokine receptors regulate a variety of immune responses to infection, inflammation and tissue repair. Besides controlling the trafficking of immune cells, chemokines also regulate the migration of seven different cell types during embryogenesis. Recently, the chemokine CXCL12, also known as stromal cell-derived factor-1 (SDF-1), and its receptor CXCR4 were shown to have prominent roles in the initiation and progression of primary and metastatic breast, pancreatic, colorectal, melanoma, neuroblastoma, esophageal, prostate, kidney and lung cancers (4-12).
The aims of this retrospective study were to examine the relationship between CXCL12 and CXCR4 expression and their immunostaining patterns, as well as to explore the role of these proteins, along with clinicopathological factors, play in gastric cancer-specific survival using surgical specimens from gastric cancer patients.
Materials and Methods
Patient characteristics and tissue samples. Our study included 137 patients who were diagnosed with gastric cancer in the Department of Surgery I, Dokkyo Medical University Hospital, between 2001 and 2009. The cohort comprised of 102 males and 35 females with a mean age of 68.9 years (range, 42-93 years) who had all undergone surgical resection. Patients who had distant metastasis or peritoneal dissemination were excluded from the study. Surgical diagnosis of residual tumor in these patients was R0 and R1.
Clinical follow-up data, including overall survival, was available for all patients. Median follow-up time of the surviving patients at the time of analysis was 79 months. The pathological tumor stage was classified according to the TNM Classification of Malignant Tumors, 7th Edition. Clinicopathological findings were assessed according to criteria established by the Japanese General Rules for Gastric Cancer Study (13, 14). Written informed consent for pathology analysis of surgical samples was obtained from each patient in accordance with our institution's guidelines prior to enrollment.
Surgical specimens were fixed in 10% formalin, embedded in paraffin and cut into 5-μm thick sections that were then stained with hematoxylin and eosin, and evaluated microscopically by two pathologists.
Immunohistochemistry (IHC). Immunohistochemical staining of CXCL12 and CXCR4 was performed as described previously (15). In brief, the 5-μm thick sections were de-paraffinized, rehydrated, placed in 0.01 mol/l citrate buffer (pH 6.0) and heated in a microwave oven for 20 min. Sections were then preincubated with 3% H2O2 in methanol for 30 min at room temperature to quench endogenous peroxidase activity. Sections were pretreated with 1% bovine serum albumin (BSA) in phosphate-buffered saline (PBS) and then incubated with an anti-CXCL12 antibody (R&M Systems Inc., Minneapolis, MN, USA; Catalog# MA350: dilution 1:50) and an anti-CXCR4 antibody (Epitomics Inc., Burlingame, CA, USA; Catalog# 3108-1; dilution 1:200) for 1 h at room temperature. Thereafter, sections were incubated with a biotinylated secondary antibody (Vectastain Elite ABC kit, Burlingame, CA, USA) for 30 min, washed with PBS and treated with peroxidase-conjugated streptavidin for 20 min. Finally, sections were visualized by incubating them with diaminobenzidine tetrahydrochloride (Liquid DAB + Substrate Chromogen System; Dako, Carpinteria, CA, USA). Sections were rinsed briefly in water, counterstained with hematoxylin and mounted.
Evaluation of immunohistochemical staining. Two independent investigators blinded to patients' clinicopathological data evaluated the immunohistochemical staining for CXCL12 and CXCR4. Positive expression was defined as the presence of CXCL12 and CXCR4 immunoreactivity in >30% of cancer cells (16).
Due to the heterogeneous content of proliferative tumor cells in sections, areas of highest proliferative activity were found by scanning tumor sections at low magnification. Tumor cell counts were then performed on these areas at a × cell magnification. CXCR4 immunoreactivity was detected in the plasma membrane, cytoplasm and nucleus. Some cancer cells showed clear CXCR4 immunoreactivity in the nucleus and a weak signal in the cytoplasm (nuclear type), whereas others showed no nuclear immunoreactivity but rather a diffuse signal in the cytoplasm and the plasma membrane (cytomembrane type). Every lesion was classified as a nuclear or cytomembrane type in accordance with its dominant immunostaining pattern. However, sections with no CXCR4 immunoreactivity were defined as negative (12).
Statistical analysis. The χ2 test was performed to determine correlations among the various parameters while Fischer's exact test was also used, as needed. The cumulative survival rate was assessed using the Kaplan-Meier method and analyzed using the log-rank test. Survival curves were estimated using the Kaplan-Meier product-limit method and the significance of difference between survival curves was determined using the log-rank test. A p-value of less than 0.05 was considered statistically significant. All statistical analyses were conducted using Stat View 5.0 (SAS Institute Inc., Cary, NC, USA).
Results
Patient demographics and disease characteristics. The study cohort consisted of 137 patients, out of which 102 were males and 35 were females, with a median age of 68.9 years. Based on histological typing, 88 cases (64.2%) were differentiated adenocarcinomas and 49 cases (35.8%) were undifferentiated adenocarcinomas. Thirty-six cases were stage I, 30 cases were stage II, 53 cases were stage III and 18 cases were stage IV (non-curative resection). During the follow-up period, 45 patients died due to relapse and metastasis while 17 patients died due to other causes.
Expression of CXCL12 and CXCR4 in primary tumor tissues. Among the 137 cases of primary gastric cancer, 43 cases (31.4%) stained positive for CXCL12 (Figure 1A, 1B). All positive CXCL12 immunoreactivity was of the cytomembrane type. In addition, 82 cases (59.8%) stained positive for CXCR4 (Figure 1C, 1D). CXCR4 immunoreactivity was of the cytomembrane type in 42 cases and of the nuclear type in 40 cases (Figure 2).
Correlation of CXCL12 and CXCR4 expression with clinicopathological disease characteristics and cancer-specific survival. To explore the clinicopathological significance of CXCL12 and CXCR4 expression, we studied its relationship with several clinicopathological factors. Tables I and II show the correlation of CXCL12 and CXCR4 expression with each clinical feature. Positive expression of CXCL12 in cancer cells was associated with differentiation (p=0.0143), negative depth of invasion (p=0.0255), negative lymphatic invasion (p=0.0173), negative venous invasion (p=0.0022) and early stage (p=0.049). Positive expression of CXCR4 in cancer cells was associated with depth of invasion (p=0.005) and stage (p=0.028). However, there was no correlation between the immunoreactivity pattern and any clinicopathological parameter in either group (Table III). With regard to patient cancer-specific survival, gastric cancer with positive expression of CXCL12 was not associated with a lower 5-year probability of survival compared to those with negative expression of CXCL12 (p=0.9088). Gastric cancer with positive expression of CXCR4, however, was associated with a lower 5-year probability of survival than those with negative expression of CXCR4 (p=0.0079) (Figure 3). There was no correlation between the immunoreactivity pattern and cancer-specific survival (Figure 4).
Representative images of immunostaining of CXCL12 and CXCR4 proteins in gastric cancer tissues. Tumor cells with: (A) Positive and (B) negative expression of CXCL12; and (C) positive and (D) negative expression of CXCR4. Original magnification, ×100.
Immunohistochemistry of CXCR4 in gastric cancer. (A) Cytomembrane type: immunoreactivity was detected in the cytoplasm and plasma membrane of cancer cells. (B) Nuclear type: immunoreactivity was detected weakly in the cytoplasm of cancer cells and strongly in the nucleus. Original magnification, ×1,000.
Discussion
In 2001, Muller et al. published the first report stating that CXCR4 and CCR were highly expressed in human breast cancer tissues and cell lines as well as lymph nodes, lung, liver and bone marrow. These areas have been identified as common metastatic sites of primary breast cancer and highly express CXCL12 (CXCR4 ligand) and CCL21 (CCR7 ligand) (4).
We examined the expression of the chemokine receptor CXCR4 and its ligand CXCL12 in gastric cancer tissues using IHC and analyzed the relationship between their expression and clinicopathological features. CXCL12 was significantly expressed in gastric adenocarcinomas of the intestinal type rather than the diffuse type. In addition, CXCL12 positive expression was inversely associated with depth of invasion, lymphatic invasion and venous invasion. Increased expression occurred in early-stage gastric cancer, a trend that decreased with gastric tumor progression. Ishigami et al. (17) showed that CXCL12 positive expression in gastric cancer was associated with lymph node metastasis, depth of invasion, lymphatic invasion, tumor diameter, clinical stage and poor surgical outcome. Conversely, Ingold et al. (18) found that CXCL12 expression was negatively associated with distant metastasis and tumor grade, while other clinicopathological factors were not significantly associated.
Correlation of CXCL12 expression in tumor cells with clinicopathological patient characteristics.
CXCL12 is a ubiquitous α-chemokine expressed by a variety of cells and considered a constitutive chemokine (19). Shibata et al. demonstrated that overexpression of CXCL12 in the stomach caused a significant enhancement of gastric epithelial cell proliferation, induction of inflammation and tumor development, particularly in combination with Helicobacter felis infection (20). Moreover, Fakhari et al. reported that co-culture of H. pylori with a gastric epithelial cell line, AGS, produced an increase of CXCL12 in AGS cells (21). In our study, H. pylori infection was not examined. However, we surmise that carcinogenesis of gastric cancer can occur in a chronic inflammatory environment.
During the early invasive stage, the tumor maintains its epithelial character. During the advanced invasive stage, the epithelial character is lost; therefore, positive expression of CXCL12 is downregulated during the advanced stages of gastric cancer. Lee et al. (22) reported that CXCL12 stimulated CXCR4 expression in gastric cancer through both autocrine and paracrine mechanisms. Thus, an autocrine mechanism may attenuate the secretion of CXCL12 helping to accelerate tumor invasion. However, in a previous report, CXCL12 in gastric cancer was not detected in the early stage but was upregulated in advanced invasive tumors. Although further studies are needed, CXCL12 positive expression did not correlate with patient survival in the current study. Therefore, CXCL12 may not be important as a predictive factor.
Correlation of CXCR4 expression in tumor cells with clinicopathological patient characteristics.
CXCR4 protein expression was associated with depth of invasion, stage and a lower 5-year probability of cancer-specific survival status after radical resection. Kwak et al. (23) detected the expression of CXCR4 mRNA in 10 gastric cancer cell lines and in 100% (43/43) of gastric cancer tissues using reverse transcription polymerase chain reaction (RT-PCR). This group also reported positive expression of CXCR4 protein in 36.5% of gastric cancer tissues using IHC. Arigami et al. (16) demonstrated that CXCR4 expression was related to gastric tumor metastasis to lymph nodes.
Jieer et al. (24) detected positive staining of CXCL12 and CXCR4 in gastric cancer cells and metastatic lymph nodes. Moreover, positive staining for CXCR4 was significantly related to worse disease-free survival and overall survival. These findings suggest that gastric tumor cells with a high expression of CXCR4 and potentially activated by CXCL12, appear more likely to metastasize to the peritoneum and lymphatics, thereby hastening the progression of gastric cancer. We demonstrated that gastric cancer cells had two distinct immunostaining patterns for CXCR4: either cytomembrane or nuclear types, although both types were not associated with clinicopathological features or outcomes, similar to that reported for cases of hepatocellular carcinoma (25), breast cancer (26, 27) and nasopharyngeal carcinoma (28). CXCR4 is a transmembrane protein and it was originally thought that its expression would be detectable at the plasma membrane and cytoplasm; however, IHC analysis detected a nuclear pattern of CXCR4 expression.
Kaplan-Meier plot of cancer-specific survival based on cytomembrane-versus nuclear-type expression of CXCR4 in gastric tumor specimens.
Relationship between CXCR4 immunostaining patterns and clinicopathological features in patients with CXCR4-positive gastric tumors.
Kaplan-Meier plots of cancer-specific survival based on (A) positive versus negative CXCL12 protein expression, and (B) positive versus negative CXCR4 protein expression in gastric tumor specimens.
Conversely, Spano et al. (29) showed that CXCR4-positive nuclear staining was associated with a significantly better outcome during early-stage non-small cell lung cancer (NSCLC). Yoshitake et al. (12) examined the localization of CXCR4 in colorectal cancer (CRC) cell lines and reported that CXCR4 immunoreactivity was detectable in both the cytomembrane and nuclear protein fraction, suggesting CXCR4 immunoreactivity in the nuclei of CRC cells was specific. Of note, CRC patients with nuclear-type CXCR4 expression showed more frequent lymph node metastasis, poor differentiation and worse outcomes than those expressing the cytomembrane type, suggesting the nuclear expression of CXCR4 might be associated with CRC progression. CXCR4 expression intensity was an important indicator that could be used to predict disease prognosis. Therefore, the CXCL12/CXCR4 axis has great potential as novel target molecules that could be used for gastric cancer therapy.
In summary, CXCL12 expression was not associated with survival. However, CXCR4 expression in gastric cancer was significantly associated with depth of invasion and poor survival. Thus, CXCR4 may be a potential biomarker for predicting poor survival of gastric cancer patients.
Acknowledgements
This study was supported by a grant (grant Number 25462065) from JSPS KAKENHI. The Authors would like to thank Ms. Noriko Suzuki and members of the Division of Clinical Science for their technical assistance.
- Received April 13, 2014.
- Revision received June 3, 2014.
- Accepted June 4, 2014.
- Copyright© 2014 International Institute of Anticancer Research (Dr. John G. Delinassios), All rights reserved
