Global cancer statistics 2020 showed that there were 1,089,103 new cases of gastric cancer (GC) and 768,793 deaths, leading to the 5th and 4th incidence and mortality rates, respectively (1). Recently, although some progress has been made in surgery, chemotherapy, radiotherapy, targeted therapy, anti-angiogenic therapy, and immunotherapy (2), the survival of patients with advanced GC remains dismal (3). Therefore, it is necessary to actively explore its pathogenesis, effective prognostic markers, and elements of poor therapeutic outcome.

Nowadays, although the pathogenesis of GC has not been fully elucidated, various studies have shown that helicobacter pylori (HP) infection is considered to be an essential contributor to the development of GC (4, 5). HP infection could stimulate an inflammatory response in the body, which in turn induces oncogenic mutations and malignant cell transformation, ultimately leading to intestinal epithelial metaplasia, cellular dysplasia, and gastric carcinogenesis (6, 7). In addition, approximately 10% of patients with GC manifest an association with Epstein-Barr virus (EBV) infection (8, 9). This virus is involved in the malignant cells in 80% of patients with lymphoid stromal GC and can also promote inflammatory changes in the gastric mucosa, thereby promoting the development of GC (10, 11). It is evident that chronic inflammation plays a significant part in the progression of GC (12). It could facilitate tumorigenesis and progression through the release of inflammatory mediators that suppresses the immune response, promote mitogenesis, and chemotaxis of cancer cells (13, 14). Inhibition of the associated inflammatory response may be a crucial part of the strategy to tackle GC. However, it is well known that GC subtypes are molecularly distinct, different subtypes of patients with GC have the different characteristics and therapeutic strategies (15, 16). This indicates that the specific inflammatory status of GC subgroups will help the patients to find the suitable treatment. In recent years, there are many new molecular classification methods for GC classification (17, 18). However, the association of inflammation-related genes and prognosis of GC is remained unclear.

Simultaneously, the inflammatory response was closely correlated with the immune status (19). The persistence of inflammation of the tumor in most cases can lead to a depression of the body’s innate immunity, thereby reinforcing tumor-mediated immunity and leading to tumor development, local infiltration, vascular regeneration, and distant metastasis (19–21). Previous studies have also suggested that the potent inflammatory effect is connected to the regulation of the tumor immune microenvironment (22). Thus, understanding the relationship between inflammation-related genes and the prognosis of GC and the interaction of tumor microenvironment (TME) and its relevance to inflammatory response is an attractive area to explore.

Herein, we constructed a prognostic model based on inflammation-related genes of patients with GC. Besides, we analyzed the associations between inflammation-related genes and the tumor immune microenvironment to explore the potential strategies to overcome immunosuppressive tumor microenvironment, and improving the efficiency of immunotherapy. The analysis showed that the prognostic inflammation-related genes may be a potential immunotherapy target and prognostic biomarker for GC.

Inflammatory etiology attributed to various cancers, such as GC (30). HP infection is regarded as the main contributor to GC and it could cause chronic inflammation of the stomach and leads to atrophic gastritis, intestinal metaplasia, heterogeneous hyperplasia, and GC (5). Inflammation-related genes are considered to be potential prognostic biomarkers for patients with GC.

In our study, we obtained 29 differentially expressed and prognostic inflammation-related genes. Then, 11 inflammation-related gene signatures were constructed by the LASSO algorithm analysis. Among these 11 inflammation-related genes (APOA1, CYP19A1, F5, HBB, IGFBP1, MATN3, MTTP, PON1, PVT1, RNASE3, and SERPINE1), APOA1, HBB, and MTTP were downregulated, the rest of the genes expressed highly in GC tissues. Moreover, we found that the high-risk group was significantly correlated with shorter OS.

Our findings suggested that these inflammation-related genes may serve as prognostic markers and therapeutic targets for GC, and there were numerous studies supporting our results. Ma et al. showed that ApoA1 was significantly related with the clinical outcome of patients with GC (31). Chong et al. discovered that human APOA1 expression levels in mice with larger tumors were significantly lower than those in mice with smaller tumors. Moreover, APOA1 expression levels in the plasma of mice with high tumor burden was lower than that of mice with low tumor burden, indicating that the decrease APOA1 expression correlated closely with tumor progression (32). Sadeghi-Amiri et al. identified that CYP1A1 expression increased significantly in GC tissues compared with their normal tissue cohort (33). Wang et al. showed that CYP1A1 was a possible new molecular target for GC therapy (34). Yang et al. found that CYP19A1 was a prognostic biomarker of GC (35). Ma et al. showed that high CYP1A1 expression may be a favorable factor for patients with GC (36). Zhang et al. found that CYP1A1 was expressed in a comparatively early stage of gastric carcinogenesis and played its parts in the whole process of sequential carcinogenesis (37). Liu et al. found that F5 was markedly increased in GC tumor tissue and possibly a promising prognostic biomarker (38). Sato et al. showed that an increased expression of IGFBP1 was more likely to lead to hematogenous metastasis and exhibited a worse survival, IGFBP1 might be a possible new predictive factor and candidate for molecular targeted therapy in GC (39). Luo et al. demonstrated that there was an increased expression and release of IGFBP-1 in HP-infected GC cells, and IGFBP-1 could inhibit the migration of GC cells, which possibly had a protective effect in HP-induced GC (40). Wu et al. confirmed that MATN3 was highly overexpressed in patients with GC and MATN3 might serve as an independently predictable prognostic factor for the poor prognosis in patients with GC. (41). Krzystek-Korpacka et al. displayed that the activity of PON1 was reduced in gastroesophageal cancer, which was paralleled with the level of inflammation and cancer-associated anemia. The reduction of PON1 appeared to present with lymph node metastasis, whereas PON1 failed to serve as an independent indicator for clinical application (42). Ding et al. found that PVT1 expression was higher in GC tissues compared with adjacent non-cancerous tissues and it was correlated with lymph node metastasis in GC. Moreover, PVT1 exhibited promising therapeutic targets for the treatment of GC and for enhancing paclitaxel sensitivity (43). Kong et al. revealed that an increased expression of PVT1 was distinctly related to the depth of tumor invasion and late TNM stage and it could act as an independent predictive factor for OS (44). Zhang et al. discovered that LncRNA PVT1 was overexpressed in cisplatin resistant GC patient tissues and cisplatin resistant GC cells, and its overexpression facilitated the development of multidrug resistance (45). Yuan et al. suggested that the expression of PVT1 was markedly increased in both GC tissues and cell lines versus normal controls, and there was a notable association between its upregulation and the depth of invasion, advanced TNM, and lymph node metastasis in GC (46). Xu et al. revealed that LncRNA PVT1 was notably increased in GC tissues and the high PVT1 expression was closely associated with the poor prognosis in patients with GC (47). Chen et al. found that circPVT1 was highly increased in GC tissues and it could competitively bind miR-125 to promote the GC cell proliferation (48). Zhao et al. suggested that SERPINE1 was considered associated with carcinogenicity and adverse prognosis in GC (49). Yang et al. revealed that GC patients with the high expression of SERPINE1 were associated with poorer OS and DFS (50). Yang et al. identified that GC patients with higher SERPINE1 expression had shorter OS and it could promote the proliferation, invasion, and migration of GC cells, indicating that SERPINE1 might serve as a new therapeutic target for GC (51). Liao et al. revealed that SERPINE1 and SPARC were expressed exclusively higher in GC tissue and they were correlated with an unfavorable outcome (52). These results were consistent with our study in general. There are several genes that have been less studied in GC at present, but possibly could be candidates for further study in the future for new insight into the prognosis and treatment of GC. In addition, our study was analyzed based on the mRNA expression level of differentially expressed and prognostic inflammation-related genes, which had some limitations. The expression levels of proteins which could be regulated by various factors, and the mRNA expression pattern is far away from the real protein expression in the clinical sample (53). Therefore, a multi-center clinical study with a larger sample size needs to be implemented to validate the expression of the hub genes at the mRNA and protein levels, respectively.

Moreover, GSEA analyses suggested that the two different groups of patients with GC were primarily involved in extracellular matrix, which is the component of the TME. In recent years, immunotherapy for GC has achieved promising outcomes, but not all patients can benefit from immunotherapy due to tumor heterogeneity (2). Since the TME is closely related to immunotherapy, understanding the condition of immune cells in the TME and exploring promising therapeutic targets for TME remodeling to relieve the immunosuppressive TME would be beneficial in the treatment of GC. The TME is the partial biological environment in which the tumor develops. The TME is largely made up of immune cells, endothelial cells, fibroblasts, extracellular matrix, and the necessary growth factors, adhesion factors, and oxygen for the cell microenvironment (54). In the TME, tumor-associated immune cells can be divided into two main categories: anti-tumor immune cells and pro-tumor immune cells, and they have different roles in the different stages of tumor progression (55). Anti-tumor immune cells mainly contain CD8 + cytotoxic T cells, effector CD4 + T cells, natural killer cells, DCs, M1 macrophages, and N1 neutrophils. In addition, there are large number of tumor-promoting immune cells, such as regulatory T cells, N2-polarized neutrophils, and M2-polarized macrophages. According to the cell types in the TME, some studies have shown that if tumor-inhibiting cells are predominant in the TME, it could play a role in inhibiting tumor formation by killing tumor cells. On the contrary, if tumor-promoting cells occupy a superior position, the TME could play a pro-tumor activity in promoting the proliferation of tumor cells, thus playing a role in promoting tumors (56). Based on the characteristics of the TME, tumors can be classified into “hot tumor” and “cold tumor.” Tumors characterized as highly immunogenic usually described as “hot tumor” with an immunogenic microenvironment, such as inflated T cells, memory T cells, and cytokines (57). Contrary to “hot tumor,” the low-immunogenic tumors with a non-immunogenic microenvironment and the absence of the above-mentioned components are called “cold tumor.” Numerous reports have demonstrated that only “hot tumor” is able to respond well to immunotherapy, whereas “cold tumor” often undergo immune escape. GC is a typical inflammatory-related malignant tumor whose microenvironment contains a large immune cell. The status and function of tumor-infiltrating immune cells regulates the biological behavior of GC. Therefore, understanding the relationship between the TME and inflammation-related genes of GC is of great importance.

Our study demonstrated that the immune cell subpopulations of DCs, macrophages, mast cells, and neutrophils were upregulated in the high-risk groups. They are responsible for promoting tumor progression in the TME. As for the immune-related pathways, para-inflammation and CCR were markedly upregulated in the high-risk group, which could facilitate tumor development. To further explore the precision of the above results, the CIBERSORT approach was applied to gain further insight into the composition of immune and stromal cells in the TME of GC patients. The results demonstrated that the pro-tumor immune cells of M2 macrophages and resting memory CD4 + T cells were found to be higher in the high-risk group, perhaps this is the reason for their poor prognosis. Besides, the levels of CD8 + T cells and activated memory CD4 T + cells which could inhibit the tumor progression were significantly increased in GC with low-risk score and they have a better prognosis. The above results show that patients with GC in the high inflammatory risk group have the characteristics of “cold tumor” and have a poorer prognosis. Therefore, targeting inflammatory response-related genes and remodeling the TME to turn “cold tumor” into “hot tumor” may be a promising solution to improve the efficacy of immunotherapy for patients with GC.

However, there are still some limitations in our study. First, the prognostic model was constructed based on TCGA data, on which the original samples were probably based on single biopsies, further well-designed studies and a multi-center clinical study with a larger sample size need to be implemented to validate our results. Second, we mainly used the preclinical models to reveal that the inflammation-related genes maybe a prognostic biomarker and involve in the immunosuppressive microenvironment in patients with GC. The larger samples need to be implemented to validate them at the mRNA and protein levels, respectively. Third, we analyzed the correlation between risk score and the different immune cell subpopulations in the TME of patients with GC. Nevertheless, more experiments are needed to provide support for our findings. In future studies, the exact roles of inflammation-related genes in the microenvironment of GC worthy of thorough investigation.

We constructed a prognostic model based on inflammation-related genes associated with prognosis, and evaluated the proportion of immune cell subtypes in the tumor microenvironment. We found that the 11 inflammation-related genes could serve as prognostic markers and patients with GC in the high inflammatory risk group have the characteristics of “cold tumor” and have a poorer prognosis. These results provided novel ideas for improving the therapeutic efficacy of GC patients by overcoming immunosuppressive tumor microenvironment. However, there are still some limitations in our study. Further well-designed studies and a multi-center clinical study with a larger sample size need to be implemented to support for our findings.

All codes are available from the corresponding author on reasonable request and we could offer the version information as necessary and any restrictions on availability.

The original contributions presented in the study are included in the article/Supplementary Material, further inquiries can be directed to the corresponding author.

YW and CY: conceptualization. YW and XY: writing the original draft. LG, XX, and YW: methodology, software, and revision. YW, XY, LG, XX, and CY: validation. All authors approved the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.