Construction And Validation Of An Immunogenic Cell Death-related LncRNA Prognostic Signature For Hepatocellular Carcinoma

Objective:Hepatocellular carcinoma(HCC),as the most common type of liver cancer,has strong heterogeneity,poor prognosis,and lack of early biomarkers.Immunotherapy can improve the prognosis in many tumors,but most of the immune checkpoint blocking therapy is not effective in liver cancer.Immunogenic cell death(ICD)has been proven to enhance the effectiveness of immunotherapy in many studies.The purpose of this study is to establish a prognostic model combining immunogenic cell death and long-chain non-coding RNA(lncRNA)to improve the accuracy of prognosis prediction of HCC patients and screen patients suitable for immunotherapy.Methods:The transcriptome data and clinical data of HCC were downloaded and sorted from TCGA database,34 ICD-related genes extracted by extensive searching from published researches.Spearman correlation analysis was used to identify ICD-related lncRNA,differential analysis of ICD-related lncRNAs yielded 841 differentially expressed genes,and single-factor Cox regression analysis was used to determine 56 prognostication-related lncRNA for subsequent analysis.HCC patients subtypes were identified by ConsensusClusterPlus,the correlation between clustering subtypes and prognosis and immune cell infiltration in HCC patients was analyzed.Five optimal prognosis-related lncRNAs generated by LASSO stepwise regression analysis were used to build the prognosis prediction model.The accuracy of the model was evaluated by ROC curve,Kaplan-Meier survival curve and nomogram.Univariate regression and multivariate regression were used to verify whether the risk score was an independent prognostic factor.Risk score calculated by prognostic model,we grouped HCC patients to high risk group and low risk group according to the median risk score.Differential analyzes on the clinical characteristics of high and low risk groups were conducted.CIBERSORT,ssGSEA and other algorithms were used to evaluated the immune cell infiltration of patients,differential expression of immune checkpoint molecules in two groups was analyzed.Lastly,the drug sensitivity of high and low risk groups was analyzed.Results:56 ICD-related lncRNAs related to prognosis were obtained by univariate analysis,and two significantly different subtypes were obtained by cluster analysis,which were different in prognosis and immune infiltration.Based on LASSO regression analysis,a prognostic model composed of five ICD-related lncRNAs(AC243654.3,LINC01060,LINC01747,MKLN1-AS,ZNF529-AS1)was established.The model shows excellent accuracy in predicting the survival probability of patients at 1,3 and 5 years through nomogram.It could be used as an independent prognostic factor to predict the prognosis of HCC patients individually.Based on the median risk score,patients could be divided into high-risk group and low risk group.HCC patients in high-risk group tended to have a worse prognosis and were closely related to later stage and higher pathological grade.In addition,the content of macrophages and regulatory T cells were higher in the high-risk group,the expression of immune checkpoint molecules was also significantly higher in the high risk group,which showing a cold immune characteristics.Showed by drug sensitivity analysis,high risk group was more sensitive to sorafenib.Conclusion:The prognostic signature constructed in this study contains five lncRNAs associated with ICD,it shows a high accuracy in predicting the prognosis of HCC patients.This model is expected to providing a reference for HCC patients to choose personalized treatment.