Abstract
Background/Aim: Glioma-associated oncogene 1 (GLI1) is an important transcription factor in the hedgehog signalling pathway and tumour formation. We evaluated the clinical significance of GLI1 expression as a prognostic factor in patients with locally advanced gastric cancer (GC). Patients and Methods: GLI1 expression levels were measured by quantitative real-time polymerase chain reaction analysis of cancerous and adjacent normal mucosa specimens obtained from 142 patients with Stage II/III GC administered adjuvant chemotherapy with S-1 after curative resection. The associations of GLI1 expression with clinicopathological features and survival were evaluated. Results: Clinicopathological features and GLI1 expression showed no association. Overall survival was significantly poorer in the high compared to the low GLI1 expression group (p=0.04). Multivariate analysis revealed that GLI1 expression was a significant independent prognostic factor [p=0.019, hazard ratio (HR)=1.94, 95% confidence interval (CI)=1.70-3.38]. Conclusion: GLI1 expression may be a useful prognostic marker in patients with locally advanced GC.
Gastric cancer (GC) is considered as a disease of global importance. With approximately 1 million new cases reported annually, it is the fifth most common malignancy worldwide. In 2018, 784,000 people died worldwide from GC, making it the third leading cause of cancer-related deaths (1). The standard treatment for locally advanced GC in Asian countries is curative resection followed by adjuvant chemotherapy. Based on the findings of the adjuvant chemotherapy trial with S-1 for gastric cancer, adjuvant chemotherapy with S-1 is currently considered as a standard treatment protocol for preventing cancer recurrence after curative resection (2). Additionally, the findings of the OPAS-1 trial reinforced the importance of S-1 adjuvant chemotherapy for treating Stage II GC (3). Recent findings from the JACCRO-07 trial have indicated the effectiveness of co-administration of docetaxel and S-1 in patients with Stage III GC (4). However, treatment outcomes in patients with locally advanced GC remain inadequate. Therefore, biomarker-based treatments may be an effective strategy for improving the outcomes of patients with locally advanced GC.
Glioma-associated oncogene 1 (GLI1) is one of the GLI proteins, which are important zinc finger transcriptional activators of the hedgehog (Hh) signalling pathway; GLI1 plays a critical role in embryonic development, differentiation, proliferation, and maintenance (5). Aberrant activation of this pathway is commonly observed in various cancers (6). Gli1, which is encoded by GLI1 mRNA, has been associated with epithelial-mesenchymal transition (EMT) and the Hh/Gli1 pathway, which has been implicated in the regulation of cancer stem cells (CSCs) (7-9). CSCs have been reported to be associated with chemoresistance (10, 11). Previous clinical studies have reported an association between GLI1 overexpression and poor prognosis in patients with different cancer types (12-18). Based on the importance of GLI1 in cancer development, we focused on studies that have reported the role of GLI1 expression in survival and chemoresistance in patients with cancers. In this study, we assessed the clinical relevance of GLI1 expression in cancer tissues and hypothesized that GLI1 expression can serve as an effective prognostic marker in patients with locally advanced GC administered S-1 adjuvant chemotherapy after curative surgery.
Patients and Methods
Patients and samples. Surgical specimens of cancerous tissues and adjacent normal mucosa from 142 patients with pathological Stage II/III GC who did not undergo preoperative treatment were collected. The patients underwent curative gastrectomy at the Department of Gastrointestinal Surgery, Kanagawa Cancer Center; Department of Surgery, Yokohama City University; and Gastroenterological Center, Yokohama City University Medical Center between January 2002 and December 2012. Approval of this study was received from the Ethics Committees of Kanagawa Cancer Center Hospital (approval: epidemiological study-29), Yokohama City University and Yokohama City University Medical Center (approval: 18-7A-4) and before study commencement. The participants were provided informed consent. The procedures followed were in accordance with the ethical standards of the Helsinki Declaration of 1975, revised in 1983. Each specimen was immediately embedded in optimum cutting temperature compound (Sakura FineTechnical Co. Ltd., Tokyo, Japan) and stored at deep freezer. The specimens were stained with haematoxylin and eosin for histopathological evaluation. Specimens containing >80% cancerous cells were used to prepare RNA extracts.
RNA extraction and complementary DNA (cDNA) synthesis. RNA extraction from GC tissue and adjacent normal mucosa specimens were performed using TRIzol Reagent (Gibco, Grand Island, NY, USA). cDNA was synthesized from 200ng of total RNA using an iScript cDNA Synthesis Kit (Bio-Rad Laboratories, Hercules, CA, USA). The cDNAs were diluted with pure water and stored at - 80°C.
Real-time quantitative reverse transcriptase polymerase chain reaction (qRT-PCR). The oligonucleotide primers used for GLI1 were: sense primer 5’-GAGCACGAGGGCTGCAGTAA-3’; antisense primer 5’-TCGCAGCGAGCTAGGATCTGTA-3’. β-actin was used as the internal control. The oligonucleotide primers used for β-actin were following: sense primer 5’-AGTTGCGTTACA CCCTTTCTTGAC-3’; antisense primer 5’-GCTCGCTCCAACC GACTGC-3’. iQSYBR Green Supermix (Bio-Rad Laboratories) was used for real-time qRT-PCR experiments. The reactions were conducted in a total volume of 15 μl containing 200 ng of cDNA, 400 nM of each primer, 7.5 μl of iQ SYBR-Green Supermix containing 400 μM each of dCTP, dATP, dTTP, and dGTP, and 50 μU/μl of iTag DNA polymerase. The following conditions were used: 3 min at 95°C, 40 cycles of denaturation (10 s for GLI1 and 15 s for β-actin), annealing (10 s at 63°C for GLI1 and 15 s at 60°C for β-actin), and extension at 72°C (20 s for GLI1 and 30 s for β-actin), followed by 10 min at 72°C. The melting curves were analysed to distinguish specific from nonspecific products and primer dimmers. A standard curve was generated for each run and the readings for three points corresponding to the human control cDNA (Clontech Laboratories, Inc., Mountain View, CA, USA) were recorded to evaluate specific mRNA expression in the samples. The concentration of each sample was calculated based on the point of intersection with the standard curve.
Statistical analysis. The GLI1 mRNA expression levels in cancerous tissue and adjacent normal mucosa specimens were compared using the Wilcoxon test. The associations between GLI1 expression and the values of explanatory variables were analysed using the χ2 test. The association between GLI1 expression and overall survival (OS) was assessed using the Kaplan-Meier method and analysed using the log-rank test. A Cox proportional-hazards model was used to perform uni- and multivariate analyses to identify prognostic factors. Statistical analyses were performed using EZR (Saitama Medical Center, Jichi Medical University, Saitama, Japan), which is a graphical user interface for R (The R Foundation for Statistical Computing, Vienna, Austria). Two-tailed p-values were calculated; p-values <0.05 were considered as statistically significant.
Results
GLI1 mRNA expression in cancerous tissue and normal adjacent mucosa specimens. There were no significant differences between GLI1 mRNA expression levels in the cancerous tissue and normal adjacent mucosa by the Wilcoxon test (p=0.438; Figure 1).
Association between GLI1 mRNA expression and clinicopathological features in GC tissues. The study specimens were divided into high GLI1 expression group (n=62) and low GLI1 expression group (n=80) based on the levels of GLI1 mRNA expression. Subsequently, the association between each group and clinicopathological features were evaluated; we found no association between these parameters (Table I).
Association between GLI1 mRNA expression and outcomes in patients with GC. The 5-year OS in the high GLI1 expression group was significantly poorer than that in the low GLI1 expression group (57.1% vs. 73.3%, respectively; p=0.04; Figure 2). In Stage II patients, there was no significant difference between the 5-year OS in the patients of the high and low GLI1 expression groups (89.2% vs. 85.4%, respectively; p=0.53; Figure 3). In Stage III patients, the 5-year OS was significantly poorer in the patients of the high GLI1 expression group than in the patients of the low GLI1 expression group (42.5% vs. 68.9%, respectively; p=0.006; Figure 4). In contrast, there was no significant difference between the 5-year OS in patients who underwent curative surgery without adjuvant chemotherapy with S-1 in the high and low GLI1 expression groups (54.8% vs. 64.8%, respectively; p=0.50; Figure 5).
Uni- and multi-variate analyses of GLI1 mRNA expression, clinicopathological factors, and outcomes. Univariate Cox regression analyses revealed that high GLI1 expression and cancer stage were significant prognostic factors. Multivariate Cox regression analyses revealed that high GLI1 expression (p=0.019) and pathological Stage III (p=0.002) were independent prognostic factors (Table II).
Discussion
To evaluate the clinical significance of GLI1 mRNA expression in patients with locally advanced GC who underwent adjuvant chemotherapy with S-1 after curative surgery, we measured GLI1 mRNA expression in cancerous tissue and adjacent normal mucosa specimens from patients and analysed the association of GLI1 mRNA expression with clinicopathological factors and treatment outcomes.
In previous studies, GLI1 mRNA expression levels in cancerous tissues have been reported to be significantly higher than those in adjacent normal tissues in breast cancer (19), lung cancer (20), and endometrial carcinoma (21). In contrast, there has been no significant difference between GLI1 mRNA expression levels in cancerous tissues and adjacent normal tissues in ovarian cancer (22) and astrocytoma (23). We observed no significant difference in GLI1 mRNA expression levels between cancerous tissues and adjacent normal mucosa from patients with GC.
The association between GLI1 mRNA overexpression and multiple clinicopathological factors such as lymph node metastasis, the depth of tumour invasion, and TNM staging in GC has been reported previously (15, 24). Gli1 expression levels have been reported to be associated with the extent of primary tumour and lymph node metastasis in oesophageal cancer (25). Other studies have revealed no significant associations between GLI1 expression and clinicopathological factors such as histological type, lymph node metastasis, depth of tumour invasion, and TNM staging in GC (26) and breast cancer (27). Similarly, in this study, we found no association between GLI1 mRNA expression levels and the clinicopathological features evaluated.
OS in the high GLI1 expression group has been reported to be significantly poorer than that in the low GLI1 expression group in patients with oesophageal cancer (25), ductal breast carcinoma (12), and colorectal adenocarcinoma (28). Consistent with these findings, we observed that the OS in the high GLI1 mRNA expression group was significantly poorer than that in the low GLI1 mRNA expression group. Uni- and multi-variate Cox proportional-hazards regression analyses indicated that high expression of GLI1 mRNA served as an independent prognostic factor of 5-year OS in patients with locally advanced GC administered adjuvant chemotherapy with S-1 after curative surgery.
Although the association between high GLI1 expression and poor prognosis in patients with locally advanced GC administered adjuvant chemotherapy with S-1 after curative surgery has been reported, the underlying mechanism remains unclear. Previous reports have suggested that two major factors are responsible for this association: the effect of GLI1 on EMT and role of GLI1 expression in cancer stem cells.
EMT has been associated with invasion, metastasis, and chemoresistance in various types of cancer cells (29-31). Gli1 expression has been observed to induce the upregulation of the Snail family transcriptional repressor 1, which regulates EMT transition and the loss of E-Cadherin in EK3E cells in vitro (32). In addition, Gli1 has been reported to regulate the expression of several EMT-promoting factors including transforming growth factor beta, RAS, WNT, phosphoinositide 3 kinase/AKT, integrins, transmembrane 4 superfamily, and S100 calcium binding protein A4 (33). Moreover, elevation of the GLI1 signalling axis was observed to be a major genetic alteration associated with the regulation of EMT-related genes in 5-fluorouracil (5-FU)-resistant colorectal cancer cells (30). These findings suggest that GLI1 expression is associated with the regulation of multiple EMT-promoting factors in GC tissues, as well as with the induction of invasion, metastasis, and chemoresistance, which includes resistance to 5-FU.
The association between cancer stem cells and chemoresistance has been investigated previously (11, 34, 35). Various signalling pathways, including the Hh pathway that is significantly associated with GLI1 expression, are activated in cancer stem cells (35). Hh signalling inhibitors have been shown to inhibit the expression of cancer stemness markers CD44, Nanog, and C-Myc and increase the sensitivity toward anticancer drugs by inhibiting GLI1 expression in 5-FU-resistant colorectal cancer organoids (36). These findings indicate that GLI1 expression is responsible for regulating cancer stem cells and inducing chemoresistance in these cells.
Limitations
First, we only examined GLI1 mRNA expression in cancerous tissue and adjacent normal mucosa specimens. The expression of GLI1 protein and its association with GLI1 mRNA expression should be analysed by immunohistochemical studies using the same specimens. Second, heterogeneity in the GC specimens posed a challenge. The specimens used for mRNA extraction were 5-mm square stomach cancerous tissue specimens; although these samples consist of tissues collected from the maximum depth, they did not faithfully represent the entire tumour.
Conclusion
Based on our findings, GLI1 mRNA expression may serve as a useful prognostic marker in patients with locally advanced GC administered adjuvant chemotherapy with S-1 after curative surgery. Further studies involving immunohistochemical analyses of GC specimens are needed to confirm our results.
Acknowledgements
The Authors thank Kazue Yoshihara for her technical support.
Footnotes
Authors' Contributions
IH and TO designed the study, participated in the analyses, and drafted the manuscript.
All other Authors made substantial contributions to data collection, analysis, and interpretation, as well as to the editing and approval of the final manuscript.
Conflicts of Interest
The Authors declare that there are no conflicts of interest regarding this study.
- Received July 13, 2020.
- Revision received July 24, 2020.
- Accepted July 27, 2020.
- Copyright© 2020, International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved