Screening Of Dual Genes Combined Prognostic Markers Based On Big Data Of Liver Cancer And Analysis Of Prognostic Efficacy Of INTS8

Objective Hepatocellular carcinoma(HCC)is the sixth most common tumor in the world and the third most common cause of cancer-related death.Currently,the treatment of HCC is mainly guided by clinical staging,especially Barcelona staging(BCLC),which has been widely accepted and applied.HCC is mainly treated by surgical resection,local ablation,chemotherapy,liver transplantation,transcatheter arterial chemoembolization(TACE)or intra-arterial infusion chemotherapy.However,the 5-year survival rate of the patients after hepatectomy was only about 60%.Postoperative recurrence is a major threat to long-term survival of patients with HCC.Recurrence of HCC is a multi-factor,multi-gene process,and its complex molecular mechanism which has not been fully elucidated.In order to better understand and reveal the new mechanism of liver cancer recurrence,find new targets for liver cancer treatment,and further improve the prognosis of liver cancer patients,we urgently need to find effective biomarkers for diagnosis and prognosis.Pubmed was used to search for genes related to the prognosis of liver cancer,and TCGA cohort data were used for preliminary screening to find the genes related to the prognosis in cancer tissues.The obtained genes were randomly combined into gene combinations,and the gene combinations with better prognosis were further screened through the TCGA cohort data.Combined with the two screeningresults,the genes with better prognosis were identified.The UNIHI website was used to predict the protein interactions of the better-performing genes.Methods 1.In the search for the genome with better performance,the genes related to cancer were first found on pubmed by searching the keywords of HCC,OS and DFS.The TCGA data were downloaded and the cancer gene expression levels were evaluated.Kaplan-meier was used to analyze the single gene and genome to evaluate and find the genes with better prognosis of liver cancer.The UNIHI database was then used to predict the protein interactions of the better performing genes.2.Immunohistochemistry: 90 cases of liver cancer tissues and matched para-cancer immunohistochemistry staining.The operation period was from February 2006 to May 2007.The follow-up period was from June 2012,and the follow-up period was 5 ~ 6 years.The donor age ranged from 27 to 84 years old,with an average age of 52.31 years.All cases were pathologically confirmed as hepatocellular carcinoma without any preoperative treatment.Double-blind evaluation by two experienced pathologists: the samples were scored according to the different staining strength of the samples: negative-0;Light yellow,l points;Yellow,2 points;Brown,3 cents.The positive staining rate of tumor cells was used to distinguish the samples: the positive rate could be expressed as less than 5%,with a score of 0;Positive rate between 5% and 25%,1 point;Positive rate between 25% and 50%,2 points;Positive rate: 50-75%,3 points,75-100%,4 points.Then,the score of item A was calculated according to the degree of staining,and the score of item B was calculated according to the number of positive cells.Finally,the IRS of the section = the score of item A × the score of item B.For this experiment,IRS is as follows: 10 positive cells in the field of view under 400-fold microscope were randomly selected for careful observation,200 cells were counted in each field of view,and then B score was obtained according to the proportion of positive cells.IRS score > 4 was divided into high expression group(H);IRS score ? 4 was divided into low expression group(L).Results 1.Bioinformatics data analysis showed that:(1)Kaplan-meier methods showed that 17 genes in the TCGA cohort were related to OS and/or DFS.Among those genes,15 genes were related to OS and 11 genes were related to DFS.(2)Kaplan-meier survival showed that the gene combinations with better OS prognosis were CXCL8-STAT4,CDT1-CXCL8,CXCL8-MCM7,CXCL8-LCAT,CXCL8-MMP9,MCM7-SLC1A5,LCAT-STAT4,HIF1A-STAT4,HIF1A-PAMR1,AND SLC1A5-STAT4.The gene combinations with better prognosis in DFS were FOXM1-STAT4,CDT1-STAT4,MCM7-STAT4,FOXM1-LCAT,LCAT-STAT4 ? LCAT-RUNX2,FOXM1-RUNX2,MCM7-RUNX2,SLC1A5-STAT4,SLC1A5-LCAT.There are 18 gene combinations.The gene combinations slc1a5-stat4 and lcat-stat4 performed well in OS and DFS.(3)protein interactions in UNIHI database showed that:CDT1,SLC1A5,MCM7,CXCL8 and INTS8 were related,and it was found that UBC was the key protein for the interaction of CDT1,SLC1A5,MCM7 and INTS8.HIST3H3 is the key protein that interacts with CXCL8 and MCM7.STAT4 and LCAT were not associated with other 5 proteins.2.Immunohistochemical assay: the expression level of INTS8 in cancer tissues was significantly lower than that in adjacent tissues by immunohistochemical microarray analysis.IRS: HCC(4.378 2.790)was significantly lower than that in adjacent tissues(5.878 2.597),P<0.000.OS of liver cancer patients with low expression of INTS8 in cancer tissues was superior to that with high expression of INTS8(P=0.044).However,there wasno significant difference in DFS between the high expression group and the low expression group(P = 0.931).The positive staining was mainly found in cytoplasm with only a few nuclei.Conclusion(1)based on the TCGA cohort,17 genes are related to the prognosis of HCC patients.(2)in the single-gene screening and gene combination screening of CDT1,LCAT,SLC1A5,STAT4,MCM7 and CXCL8,the prognosis of OS and DFS of patients was outstanding.(3)high expression of INTS8 was associated with low 3-year survival rate in patients with liver cancer.High expression of INTS8 was associated with more disease progression in liver cancer patients in 2 years.Low expression of INTS8 may be associated with poor prognosis of OS.