**5. Gene expression-based prognostic scoring systems**

Data from literature revealed the important role of the molecular biological characteristics of gastric cancer in the prognosis of the patients and determination of a most suitable clinical therapy for these patients. There are some dysregulated gene expressions found to be associated with the prognosis, such as the overexpressions of HER2 [45] and p53 genes [46]. Also, the hypoxia inducible factor-1 alpha (HIF-1α) seems to be related to the early development of gastric tumor [47].

Takeno et al. [48] elaborated a gastric cancer regulatory network with CDKNIA as the node and examined the expression levels of seven genes in different stages of gastric cancer (iMMP7, SPARC, SOD2, INHBA, IGFBP7, NEK6, and LUM). Their results showed that these seven genes were activated as the disease progressed, suggesting the association of these genes with cancer development.

Wang et al. [49] proposed the hypothesis that molecular features are determining the tumor behavior and can be used to establish prognostic scoring system. Based on the Cancer Genome Atlas (TCGA) data and using different multivariate clustering techniques to identify the key genes for prognostic classification of these analyses, they created a 53-gene expression prognostic scoring system and successfully implemented it to predict overall survival (OS) in the TCGA and the GSE15459 data (Gastric Cancer Project 2008).

1+DEFA5+DES+DMN+GFRA3+MUC17+OR9G1+REEP3+TMED6+TTN) and a 9-gene group (DPT+EIF1AX+FAM26D+IFITM2+ LOC401498+OR2AE1+PRRG1+REEP3+RTKN2). The overall classification accuracy obtained on the three groups, early, advanced stomach cancer, and control, was 71.4%. Among the early-stage signature genes, there are signaling and immunerelated genes that may represent the early changes of tissue cells during carcinogenesis. A few genes were found to be in both the cancer grading and staging signatures (e.g., CPS1, DES, GFRA3, TMED6, and DPT), indicating some functional connection between cancer differentiation and progression. LANCL3, MFAP2, and PPA1 were genes highly correlated with different pathological stages, showing consistent upregulation or downregulation along with

There were found 62 genes with consistent differential expression in gastric cancer versus control tissues, related to extracellular processes such as CAMs, tight junction, cytokinecytokine receptor interaction, and ECM receptor interaction, the plasminogen activation cascade, as well as signaling pathways (Wnt and Integrin signaling) related to the control of cell

The study revealed that the differential expression patterns of 15 genes are highly specific to gastric cancer (e.g., GKN2, CLDN7, THY1, GIF, and PGA4), while most others are general to numerous cancer types, including a few members of the collagen gene family, the carcinoembryonic antigen-related cell adhesion molecule, matrix metalloproteinases, topoisomerase, and secreted phosphoprotein. Only three genes, CLDN7, CLDN1, and DPT, were significantly differentiated in all grades and stages of gastric neoplasia; the consistent expression of dysregulation across all the cancer subgroups may indicate their involvement in major biological pathways leading to cancer development and progression. Dermatopontin (DPT) represents an extracellular matrix protein that creates a link between the dermal fibroblast cell surface and its extracellular matrix,

previously found to be downregulated in both uterine leiomyomas and keloids [44].

Data from literature revealed the important role of the molecular biological characteristics of gastric cancer in the prognosis of the patients and determination of a most suitable clinical therapy for these patients. There are some dysregulated gene expressions found to be associated with the prognosis, such as the overexpressions of HER2 [45] and p53 genes [46]. Also, the hypoxia inducible factor-1 alpha (HIF-1α) seems to be related to the early development of

Takeno et al. [48] elaborated a gastric cancer regulatory network with CDKNIA as the node and examined the expression levels of seven genes in different stages of gastric cancer (iMMP7, SPARC, SOD2, INHBA, IGFBP7, NEK6, and LUM). Their results showed that these seven genes were activated as the disease progressed, suggesting the association of these genes with

Wang et al. [49] proposed the hypothesis that molecular features are determining the tumor behavior and can be used to establish prognostic scoring system. Based on the Cancer Genome Atlas (TCGA) data and using different multivariate clustering techniques to identify the key

**5. Gene expression-based prognostic scoring systems**

tumor progression.

114 Gastric Cancer

growth and proliferation.

gastric tumor [47].

cancer development.

These prognostic scores are able to distinguish between patients with good prognosis and bad prognosis, respectively. These genes include TNFAIP2 [50], FGFR4 [51, 52], CXCL10 [53], CEP55 [54], CXCL1 [55], LIMK1 [56], LAMC2 [57], APOE [58], INHBA [59], OSMR [60], APOC1 [61], KLF4 [62], MMP14 [63], ADH1C [64], COL6A3 [65], CCT2 [66], NOL8 [67], EPHB4 [68], and MCM2 [69]. The high expression of FGFR4 protein was previously reported to be associated with a poor prognosis in patients with advanced gastric tumors [51], while the FGFR4 Gly388Arg polymorphism proved to be a useful prognostic marker for early gastric cancer patients [52]. CEP55 functions in cell cycle regulation; knockdown of CEP55 led to diminished proliferation in gastric cancer cell lines acting on the PI3K/AKT signaling pathway and the expression of cyclin-related proteins, suggesting a potential role of CEP55 as a target used in gastric cancer treatment [54]. Some studies show that MCM2 expression levels are a useful tool for the diagnosis and prognosis of gastric tumors [69] and that SNPs in miRNAbinding sites may represent susceptibility markers for gastric cancer [50]. Chemokine (C-X-C motif) ligand (CXCL1) seems to play a major role in tumor metastasis; it has been previously reported that its expression is associated with hepatocellular carcinoma survival [55]. The study of Wang showed that CXCL1 is also involved in gastric cancer overall survival. ATPbinding cassette E1 (ABCE1) known to play a crucial role in the metastasis of lung cancers, and therefore, a potential therapeutic target in this setting [70], was also elevated in gastric tumors and predicted the prognosis of patients.
