BackgroundGastric cancer(GC)is the fourth leading cause of cancer death,with about 769,000 deaths worldwide every year.GC is highly heterogeneous and can be divided into different subtypes according to histological characteristics,genotype and molecular phenotype.In 2014,the investigators of Cancer Genome Atlas(TCGA)divided GC into four molecular subtypes,including Epstein–Barr virus-positive,Microsatellite instability(MSI),Genomic stability(GS)and Chromosomal instability(CIN).Epstein-Barr virus-associated gastric cancer(EBVaGC)accounts for about 8.7%of GC.The prognosis of EBVaGC is better than that of Epstein-Barr virus-negative gastric cancer(EBVnGC),and the five-year overall survival(OS)rate of EBVaGC can reach 70%.Based on histology types,EBVaGC can be classified as lymphoepithelioma-like carcinoma(LELC)or non-lymphoepithelioma-like carcinoma(NLELC).The tumor microenvironment of EBVaGC exhibits a high level of Tumor infiltrating lymphocytes(TILs),particularly CD8~+T cells and mature dendritic cells.Because programmed cell death ligand 1(PD-L1)is substantially expressed,EBVaGC is more susceptible to immunotherapy than EBVnGC.ObjectivesThis study aims to summarize the clinicopathological characteristics and identify the prognostic factors of EBVaGC,and comprehensively compare the clinicopathological features and prognostic differences of different tissue subtypes of EBVaGC(LELC/NLELC).Through the analysis of EBVaGC transcriptome sequencing data,we revealed the immune microenvironment and preliminarily explored the prognostic biomarkers of EBVaGC.MethodsPart Ⅰ:In our center,we collected the clinical and pathological data of EBVaGC patients.Survival analysis was utilized after following up the patient mortality,we then screen the prognostic factors of EBVaGC through Cox proportional hazard model.The data of LELC and NLELC were statistically analyzed and the differences between the two groups were compared.Part Ⅱ:List the search formula according to the precise search terms,initially search the literature,identify the results based on the inclusion and exclusion criteria,gather the data of the included literature,and perform the clinical pathological characteristics and prognosis meta-analysis of LELC and NLELC.Part Ⅲ:Download the transcriptome sequencing data and clinical data of GC from the public database.Immune infiltration analysis was used to reveal the unique immune microenvironment of EBVaGC.After analyzing the differentially expressed genes(DEGs)between EBVaGC and EBVnGC,we then performed functional enrichment analysis on the basis of DEGs and identified the EBVaGC-related prognostic biomarkers.ResultsPart Ⅰ:A total of 122 EBVaGC patients were enrolled in our study.The male-to-female ratio is 4.8:1,the age range is 26-80,the prevalence of Helicobacter pylori infection is 10.7%,the Lauren type is mostly diffuse type(63.1%),the tumor is primarily located in the proximal stomach and body of the stomach(73.7%),the HER2 IHC score is mostly 0-1(86.8%),100%of the patients are Mismatch repair-proficient(pMMR),and the positive rate of PD-L1 is 63.2%.The 1-,3-,and 5-year OS rate of EBVaGC were 95.8%,78.8%and76.2%,and the 1-,3-,and 5-year disease-free survival(DFS)rates were 93.3%,80.8%and74.1%,respectively.Histological type is an independent factor of prognosis.The 5-year OS rate of LELC and NLELC is 95%and 63.7%respectively,and the 5-year DFS rate is 97.9%and 58.6%respectively.LELC is the unique histological subtype of EBVaGC.Compared with NLELC,the tumor stage(AJCC 8th)and T stage of LELC are earlier,and the proportion of diffuse type(Lauren type)and the positive rate of PD-L1 are higher.Part Ⅱ:A total of 643 EBVaGC patients were included in the meta-analysis of 8 articles,including 257 LELC and 386 NLELC.There was no difference between LELC and NLELC in gender and age composition ratio,both of which are more likely to occur in middle-aged men,but the tumor stage,T stage,and N stage of LELC are earlier.There was no difference between LELC and NLELC patients in HP infection rate,tumor location and size,Lauren classification or tumor differentiation,but LELC has less perineural invasion,a higher PD-L1 positive rate,and a higher 1-,3-,and 5-year OS rate than NLELC.Part Ⅲ:The immune infiltration analysis showed the immune scores were higher in EBVaGC,the proportions of CD8~+T cells,activated CD4~+memory T cell,M1 macrophages,etc.were higher,while the proportions of memory B cells,resting CD4~+memory T cells,M0 macrophages,etc.were lower.The expressions of immune checkpoint genes(CTLA4,LAG3,PD1,PD-L1,TIGIT,PD-L2,HAVCR2)and most human leucocyte antigen(HLA)family genes were higher in EBVaGC.Compared with EBVnGC,EBVaGC has 2912 down-regulated DEGs and 672 up-regulated DEGs.GO and KEGG analysis showed that EBVaGC-related DEGs mainly participated in immune-related biological processes and pathways.Gene Set Enrichment Analysis(GSEA)also showed that EBVaGC was related to immune-related biological processes.Through weighted gene co-expression network analysis(WGCNA)and protein-protein interaction(PPI)analysis,a total of 10 key node genes were screened,including CD4,STAT1,FCGR3A,IL10,C1QA,CXCL9,CXCL10,CXCR6,CD274,CCL18.Through survival analysis,CXCL9 and CXCR6 were identified as EBVaGC-related prognostic biomarkers.The expressions of CXCL9 and CXCR6 were positively correlated with the content of CD8~+T cells,M1 macrophages and other immune cells.CXCL9 and CXCR6 were found to be positively related with the expression of PD-L1,CTLA-4 and PD-1.CXCL9 and CXCR6 were highly expressed in human EBV positive gastric cancer cell lines.ConclusionPart Ⅰ:EBVaGC has unique clinicopathological characteristics.EBVaGC patients with histological type LELC have a better prognosis than NLELC.Part Ⅱ:LELC is a special histological subtype of EBVaGC,with earlier stage,less perineural invasion,higher positive rate of PD-L1,and better prognosis than NLELC.Part Ⅲ:EBVaGC has a unique tumor immune microenvironment with a high proportion of TILs,especially CD8~+T cells,activated CD4~+memory T cells,M1 macrophages,etc;CXCL9 and CXCR6 are EBVaGC-related prognostic biomarkers. |