Association Between Polymorphisms of CTLA-4 + 49A > G Locus and Susceptibility of Upper Gastrointestinal Cancer: A Systematic Review and Meta-Analysis

Objective: To explore the correlation between CTLA-4 + 49A > G polymorphism and risks of upper gastrointestinal cancer. Methods: Searched Pubmed, Embase, Cochran Library, Chinese Biomedical Literature Database, Chinese CNKI, VIP and Wanfang database, retrieve the deadline to Feb 2021. Document acquired independently by the two collectors according to inclusion and exclusion criteria. System analysis was performed by Revman 5.3 software. Results: A total of seven case-control studies were included in the study, including 3045 patients and 3545 controls. The results of the systematic analysis showed that all studies on upper gastrointestinal tumors were combined: random effects model, dominant genetic model GG + GA vs AA: the difference was not statistically signicant [OR=1.15; 95%CI: 0.74-1.81]; recessive genetic model GG vs GA + AA [OR=0.99; 95%CI: 0.74-1.31], codominant model GG vs AA [OR=1.14; 95%CI: 0.67-1.95], GA vs AA [OR=1.20; 95%CI: 0.80-1.79], allele model G vs A [OR=1.06; 95%CI: 0.82-1.38]. The above results show CTLA-4 + 49A > G locus. There was no correlation between gene polymorphism and susceptibility to upper gastrointestinal tumors including gastric cancer and esophageal cancer. After the ve studies on gastric cancer were combined, each genetic model showed that the CTLA-4 + 49. Conclusion: This study showed that there was no correlation between single nucleotide polymorphisms in CTLA-4 + 49A > G locus and upper gastrointestinal tumors including gastric cancer and esophageal cancer. GA + AA (OR = 1.07, 95% CI: 0.78–1.43), GG: AA (OR = 1.18, 95% CI: 0.64–2.18), GA: AA (OR = 1.00, 95% CI: 0.70–1.44), and G: A (OR = 1.08, 95% CI: 0.82–1.42), suggesting that the correlation between the risk of upper gastrointestinal tumor development and the polymorphism of the CTLA-4 + 49 locus was not statistically signicant.


Introduction
Gastrointestinal tumor is a common human disease, which seriously endangers people's health [1] . Stomach cancer is the 4th most common cancer in the world [2] , esophageal cancer is the 3rd common cause of cancer death. With the development of society, its incidence rate is also rising and increasing year by year [3][4][5] .
Cytotoxic T lymphocyte antigen 4 (CTLA-4) is an important inhibitory regulator of T cell proliferation and activation [6] , it induces Fas-independent apoptosis of activated T cells. The cytotoxic T lymphocyte-associated antigen 4 gene is located on chromosome 2q33, and its expression product is the leukocyte surface antigen CD152, which competes with CD28 for binding ligand B7-1/B7-2, thereby negatively regulating cell proliferation and activation and decreasing the occurrence of anti-tumor immune responses in humans, thus increasing tumor susceptibility [7] . It has been found that CTLA-4 single nucleotide polymorphisms (SNPs) can affect the expression of CTLA-4 gene from different expression links and are closely associated with genetic susceptibility to tumors such as gastric and breast cancers [8] . Zheng [9] suggested that CTLA-4 + 49 G/A polymorphism was associated with the development of solid tumors (including colorectal, gastric, esophageal, skin, thymoma, nasopharyngeal, and cervical squamous carcinomas). Zhang [10] showed that the polymorphisms were not associated with gastric and colorectal cancers. Liu [11] also reported no statistically signi cant association between CTLA-4 + 49G/A polymorphisms and the risk of digestive system cancers.
In this study, the collected studies were systematically analyzed by including case-control studies on the association between upper GI tract tumor susceptibility risk and CTLA-4 + 49 gene polymorphism, and then understanding the association between the mechanistic process of upper GI tract tumorigenesis and single nucleotide polymorphisms at the CTLA-4 + 49A > G locus

Materials And Methods
Literature Retrieval Strategy Search Pubmed, Cochran Library, Embase, China Biomedical Network, China Knowledge Network, Wanfang and Wipu databases, search the period from library constructioned to Feb 2021, with "CTLA-4", "Single Nucleotide Polymorphism" or "single nucleotide polymorphism", "gastriccancer oresophagealcancer or upper gastrointestinal cancer" or "stomach cancer, esophageal cancer, or upper gastrointestinal tract tumor" as search terms. At the same time, carried out necessary manual retrieval as a supplement.

Inclusion and Exclusion Criteria
Inclusion criteria: (1) Studies of upper gastrointestinal tract tumors including esophageal and gastric cancers with susceptibility to single nucleotide polymorphisms at the CTLA-4 + 49A > G locus; (2) Case controlled study; (3) Data in the text must be complete.
Exclusion criteria: (1) No control group studies, for reviews, reports and other similar literature; (2) Duplicates or literature with incomplete information; (3) Studies with precancerous lesions such as esophagitis and atrophic gastritis as control groups.

Literature Quality Evaluation Criteria
Applying the Newcastle-Ottawa scale, NOS [12] criteria, all literature was evaluated for relevance by analyzing 3 aspects of casecontrol selection, comparability, and exposure information, and literature extraction was performed independently by ≥ 2 study participants. When the control genotype distribution deviated from HWE, studies with control genotype distribution deviating from HWE were included rst, and then sensitivity analyses were performed to compare the stability of the results or to test whether there were differences between groups [13][14][15][16] .

Statistical Analysis
The RevMan 5.3 software was used to perform the analysis of effect sizes, and in the heterogeneity test, the I 2 test was applied to the included studies, and when there was heterogeneity among the studies, P < 0.1 or I 2 > 50%, the effect sizes were combined by means of a random-effects model, and when the results were reversed, a xed-effects model was adopted. Processing of sensitivity analysis: meta-analysis was performed after excluding one literature at a time. Egger and Begg tests made with Stata 12.0 software were used to determine whether there was publication bias.

Results
Basic Characteristics of the Included Literature A total of 75 relevant papers were retrieved for this study. After de-duplication, reading the titles and abstracts, 40 duplicate and irrelevant literatures were excluded. After reading the full text, 8 papers were nally included, and the basic characteristics of the included papers are shown in Table 1, including 3 case-control studies of esophageal cancer [17][18][19] and 5 case-control studies of gastric cancer [18,[20][21][22][23] ; one [18] of the eight papers simultaneously analyzed the correlation between the risk of esophageal cancer, gastric cancer and CTLA-4 + 49A > G gene polymorphism.

Discussion
Gastrointestinal tumors are one of the major diseases that endanger human health, and their treatment and prevention are urgent issues to be addressed.
Nistico [24] reported an A-G dimorphism at position 49 of CTLA-4 exon 1 (rs231775), a polymorphism that causes amino acid exchange of threonine with alanine in the leading sequence of this protein. It was found that this polymorphism may also affect the ability of CTLA-4 to bind to B7-1/B7-2, which may affect T-cell activation [25] . The results suggest that the G allele at position + 49 but not the A allele attenuates the CTLA-4-driven downregulation of T-cell response [24] . However, there are con icting ndings regarding the association between the + 49A > G polymorphism and cancer risk. Therefore, the authors performed meta-analysis and systematic review of all eligible case-control studies to summarize and analyze data from published studies so as to assess the association between CTLA-4 + 49A-G polymorphism and risk of gastric and esophageal cancers In this study, the combined results across genetic models showed that there was no correlation between gene polymorphisms at the CTLA-4 + 49A > G locus and susceptibility to upper gastrointestinal tract tumors, including gastric and esophageal cancers. A subgroup analysis of studies on gastric cancer was performed, and there was also no signi cant correlation between the CTLA-4 + 49A > G locus gene polymorphism and gastric cancer in each genetic model. When the 2 papers [19,22] that did not t the H-w balance were excluded and combined again and compared with the total combined results, the results showed agreement with the total combined results and better stability of the pooled results.
Zheng [9] conducted a meta-analysis with only three studies on gastric cancer and only one on esophageal cancer and reported that CTLA-4 + 49A/G polymorphism was associated with the occurrence of solid tumors (including colorectal cancer, gastric cancer, esophageal cancer, etc.). Liu [11] conducted a meta-analysis with four studies on gastric cancer and only two on esophageal cancer and reported no statistically signi cant risk between CTLA-4 + 49 G/A polymorphism and digestive system cancers, and no signi cant association between CTLA-4 + 49 G/A polymorphism and gastric cancer was observed in the subgroup analysis. In this study, comparison of the results of the sensitivity analysis with the overall combined results showed that the pooled results may be more stable than the results of the above analysis.
This study has the limitation that the included literature is ethnically Asian, so a large number of well-designed epidemiological studies are still needed to validate the ndings.

Conclusion
This study showed that there was no correlation between single nucleotide polymorphisms in CTLA-4 + 49A > G locus and upper  Forest plot of association between CTLA-4 + 49 locus and susceptibility to gastric cancer (GG: GA + AA model)