Epithelial-mesenchymal Transition-related Gene Prognostic Index And Phenotyping Clusters On Prognosis,tumor Microenvironment And Drug Efficacy Of Hepatocellular Carcinoma

Objective:Epithelial-mesenchymal transition(EMT)plays a role in the evolution of the highly tumour heterogeneous and immunosuppressive tumour microenvironment(TME)of primary hepatocellular carcinoma(HCC).The TME may or may not play a role in the process.Therefore,the development of an EMT-associated gene prognostic model and new subpopulations is needed to systematically assess their impact on HCC prognosis,TME and drug efficacy.Methods:1、HCC-related EMT gene modules were identified by WGCNA.The EMT-RGPI model was then constructed by combining Univariate Cox regression analysis with Least Absolute Shrinkage and Selection Operator(LASSO)analysis.The International Cancer Genome Consortium(ICGC)LIRI-JP Japanese HCC database was used as an external independent validation set.2、Clustered subgroup analysis of HCC cohorts based on liver cancer-specific EMT-associated HUB genes.Principal component analysis(PCA)was used to assess the gene expression patterns of different hepatocellular carcinoma subtypes.Kaplan-Meier analysis was used to compare the overall survival(OS)of subgroups of patients.Normograms were generated in combination with HCC clinical characteristics parameters and used to predict OS rates at 1,3 and 5 years in patients with hepatocellular carcinoma.The robustness of the nomogram model was evaluated based on the concordance index(C-index).3、Analysis of Differentially Expressed Genes(DEGs)in C1/C2 subgroups using KEGG and GO yielded up-regulated DEGs enriched in tumour-associated pathways,and 15 genes with the highest mutation rates could be identified based on the different enrichment in C1/C2 subgroups.4、The Ciber Sort tool was used to analyse the infiltration abundance of HCC immune cells and to analyse differences in the expression of immune checkpoint-related genes in C1 and C2 subpopulations.TME subtypes were scored according to tumour immune dysfunction and rejection(TIDE)and used to predict the therapeutic response of C1 and C2 subpopulations to immune checkpoint blocking agents.The pRophetic package was used to analyse the therapeutic response of C1 and C2 subgroups to commonly used targeted and chemotherapeutic agents for HCC.Results:1、After selecting the HCC-related gene modules,the EMT-RGPI model for 12 related genes was established by combining COX regression and LASSO analysis,and then its usefulness was verified using the ICGC LIRI-IP dataset,and it could be concluded that the EMT-RGPI model could accurately predict the prognosis of HCC patients.2、At the optimal level of clustering stability,PCA showed that HCC samples with different gene expression patterns were divided into two subgroups C1/C2,and combined with Kaplan-Meier analysis it could be concluded that overall survival was longer in the C1 subgroup.3、Analysis of Differentially Expressed Genes(DEGs)in the C1/C2 subgroups using KEGG and GO showed that up-regulated DEGs were enriched in tumour-associated pathways,and the 15 genes with the highest mutation rates could be identified according to the different enrichment in the C1/C2 subgroups.4、The C2 subpopulation was preferentially associated with unfavourable prognosis,higher stemness index(mRNAsi)values,upregulated immune checkpoint expression and significant immune cell infiltration.And features of EMT,glycolysis,Wnt β-catenin signalling,angiogenesis and TGF-β signalling were significantly enriched in the cluster C2 group.higher rates of TP53 and RB1 mutations were also in the C2 subgroup.Tumour immune dysfunction and rejection(TIDE)scores and TME subtypes showed that patients in subgroup C1 responded significantly to immune checkpoint inhibitors(ICI),while half-maximal inhibitory concentrations(IC50)showed that patients in subgroup C2 were more sensitive to anti-angiogenic or chemotherapeutic agents.Conclusion:1、The EMT-RGPI model can accurately predict the prognosis of HCC patients.2、Identification of C1 and C2 subgroups with significant tumour heterogeneity and significantly different immune checkpoint expression,TME and TIDE scores based on clustering analysis of hepatocellular carcinoma-specific EMT-associated HUB genes.3、Patients in subgroup C1 have a better immunotherapeutic effect on immune checkpoint inhibitors(ICI);patients in subgroup C2 are more sensitive to anti-angiogenic or chemotherapeutic drugs,which will help in the precise treatment of HCC patients.