Genomic Analysis Of The Prognostic Value Of Colony-Stimulating Factors(CSFs)and Colony-stimulating Factor Receptors(CSFRs)Across 24 Solid Cancer Types

Objective:Granulocyte-macrophage colony-stimulating factor(GM-CSF,also called CSF2)and granulocyte-colony stimulating factor(G-CSF,also called CSF3)are two common hematological growth factors in hematopoietic system,which can stimulate the growth,proliferation and differentiation of bone marrow stem cells into mature granulocytes by binding to specific receptors.In clinical oncology,recombinant G-CSF or GM-CSF is widely used to correct chemotherapy and radiotherapy-induced neutropenia.However,as widely expressed growth factors in a variety of tumor tissues,the expression levels and prognostic value of CSFs and CSFRs in different cancer types are still controversial.Studies have shown that CSFs can be up-regulated in multiple cancer types through immunodependent and non-immunodependent pathways and promote the growth,invasion and migration of tumor cells,such as lung cancer,breast cancer,melanoma,prostate cancer,head and neck cancer,colorectal cancer,bladder cancer,etc.In addition,CSFs have also been found to exert anti-tumor effects in various cancer types by regulating tumor microenvironment,such as head and neck cancer,prostate cancer,melanoma,colorectal cancer,etc.Therefore,this study aims to explore the prognostic value of CSFs and CSFRs in various solid cancer types at the genomic level,and to explore their potential molecular mechanisms of action.Methods:In this study,we used the public dataset of cBioPortal database(http://www.cBioportalportal.org/),a visualized online sub-database of The Cancer Genome Atlas(TCGA).The expression of "CSF2","CSF3”,"CSF2RA","CSF2R",and "CSF3R" and the clinicopathological characteristics in 24 solid cancer types,including age,sex,ethnicity,stage,which may affect the survival outcome were obtained.X-tile software was used to select the optimal cut-off value of gene expression level in CSFs(CSF2,CSF3)single-gene group,CSFRs(CSF2RA,CSF2RB,CSF3R)single-gene group and CSFs+CSFRs(CSF2+CSF2RA,CSF2+CSF2RB,CSF3+CSF3R)combination group.We divided the single-gene CSFs group into two subgroups with high expression and low expression,and the CSFs+CSFRS combination group into three subgroups with low expression,medium expression and high expression.Survival analysis was performed using Kaplan Meier survival curve.Then we analyzed the differentially expressed genes(DEGs)of high and low expressed subgroup of CSFs and CSFRs using "edgeR" package of R software to sceen out the DEGs that were up-regulated or down-regulated genes in each subgroup.Intersections of up-and down-regulated differential genes for cancer types with consistent prognostic trends were filtered out using Venny 2.10 software,an online Venn diagram mapping tool.Genome ontology(GO)enrichment analysis and Kyoto Encyclopedia of Genes and Genomes(KEGG)pathway enrichment analysis were performed for the selected common DEGs using Metascape(http://metascape.org/gp/index.html),a gene function annotation analysis tool.Data processing and analysis:SPSS 20.0 statistical software was used for data analysis.Survival differential curve was analyzed using Kaplan-Meier method.Crosstab chi-square test analysis were employed to examine the association between the expression of CSFs and CSFRs and clinical characteristic parameters.Univariate and multivariate analyses were performed using the Cox proportional hazards model.P<0.05 was considered statistically significant.Results1.Patient characteristicsA total of 8,565 patients with 24 cancer types were extracted from the cBioPortal database,including 3,062 and 5503 patients with high and low CSF2 expression,and 3463 and 5102 patients with high and low CSF3 expression.Crosstab analysis showed that the expression level of CSF2 was closely correlated with age and stage(P<0.05),and there was no significant correlation with sex and racial type.While the expression of CSF3 was significantly correlated with gender,age,ethnic type and AJCC TNM stage(P<0.05).2.Survival analysisThrough Kaplan-Meier analysis,we found that the expression level of CSFs and CSFRs had a significant difference in overall survival among 24 solid cancer types.Our results showed that the survival time of patients in the subgroup with high CSF2 expression was significantly longer than the subgroup with low CSF2 expression in 8 types of cancer,including breast invasive carcinoma(BRCA),lung adenocarcinoma(LUAD),and bladder urothelial carcinoma(BLCA).However,patients with high CSF2 expression subgroup were associated with poorer prognosis in 6 types of cancer,including head and neck squamous cell carcinoma(HNSC),esophageal carcinoma(ESCA),kidney renal clear cell carcinoma(KIRC),pancreatic adenocarcinoma(PAAD).For CSF3,patients with high expression subgroup had bad prognosis in 10 cancer types,including colon adenocarcinoma(COAD),BLCA and CESC,while patients with high expression subgroup had good prognosis in 3 cancer types,including BRCA,brain lower grade glioma(LGG)and prostate adenocarcinoma(PRAD)Subsequently,we further analyzed the effect of the CSFs and CSFRs combination group expression level on the prognosis of patients with different cancer types.In the CSF2+CSF2RA combination group,we found that the high expression subgroup had a better prognosis in 5 cancer types including BRCA,uterine corpus endometrial carcinoma(UCEC),skin cutaneous melanoma(SKCM),and a poorer prognosis in 5 caner types,including COAD,ESCA and LUAD.In the CSF2+CSF2RB combination group,the high expression subgroup had a favorable prognosis in 6 cancer types,such as BRCA,LUAD,SKCM,etc and had an unfavorable prognosis in 4 cancer types,such as ESCA,lung squamous cell carcinoma(LUSC),kidney clear cell carcinoma(KIRC).In the CSF3+CSF3R combination group,we found that the prognosis of the high expression subgroup was better in BRCA and LUAD,and was worse in bladder urothelial carcinoma(BLCA),LUSC,HNSC and other 6 cancer types.3.Enrichment analysisThe differential genes with up-regulated and down-regulated expression were separated using "edgeR" package,and the corresponding venn plots were generated for cancers with similar histological or anatomical characteristics and same prognostic trends to get the intersection of common DEGs.The results showed that up-regulated and down-regulated of DEGs were mainly enriched in immune and inflammatory response related pathways in the CSF2 and CSF2RA combined group.In the CSF2 and CSF2RB combination group,it was mainly enriched in the immune response and nervous system related pathways.In the CSF3 and CSF3R combination group,the main enrichment pathways were found to be related to inflammation and signal transduction pathways.ConclusionIn this study,through genomic analysis of the expression levels of CSFs and CSFRs in 24 cancer types,we found that CSFs and CSFRs were independent prognostic factors,and the prognostic value of CSFs and CSFRs in different cancers were cancer-type dependent.Therefore,personalized treatment based on CSFs and CSFRs enrichment pathways should be considered for cancer patients.