Development And Validation Of A Anoikis-Related Prognostic Signature In Lung Adenocarcinoma Based On Bioinformatics

Objective:Lung cancer is the second most common malignant tumor after breast cancer worldwide,and has a high mortality rate.Among them,lung adenocarcinoma(LUAD)is the most common subtype.Surgery and chemotherapy are the traditional treatment methods.In recent years,with the further research on the molecular mechanism of the occurrence and development of lung cancer,targeted therapy and immunotherapy have been increasingly used in clinical practice.However,the overall survival rate of lung adenocarcinoma is still very low,and further research on the treatment monitoring and prognosis evaluation of LUAD is crucial for clinicians and patients.The aim of this study is to investigate the anoikis-related genes(ARGs)in LUAD and construct a prognostic risk model by bioinformatics methods,in order to identify the disease and predict the prognosis early,and to study the relationship between the prognostic model and the tumor microenvironment,so as to provide a new strategy for the treatment oflung adenocarcinoma.Methods:The transcriptome data and related clinical data of LUAD were downloaded from The Cancer Genome Atlas(TCGA).The expression matrix and platform files of dataset GSE72094 in GEO database were downloaded.R language was used to perform differential analysis of the transcriptome data of TCGA,and the obtained differentially expressed genes(DEGs)were used as candidate genes.Atotal of 801 ARGs were downloaded from Gene Cards database,and the genes with Relevance score > 1 were retained,finally 363 anoikis related genes were obtained.The expression data from TCGA and GEO databases were merged,and the expression data of ARGs in LUAD were extracted from them to obtain anoikis-related differentially expressed genes(AR-DEGs).Then univariate COX regression analysis was performed to screen out related genes according to p value,and cluster analysis was performed on the expression data of prognosis-related ARGs.LUAD samples were divided into 2subtypes,and pathway enrichment analysis was performed.Next,the Least Absolute Shrinkage and Selection Operator(LASSO)regression analysis was performed on the prognostic factors,and the statistically significant ARGs were selected to construct a linear prediction model related to survival.The samples were randomly divided into training set and validation set,and the risk scores of each sample were calculated.Taking the median as the cut-off value,the samples were divided into high and low risk groups according to the risk score.The 1-,3-,and5-year Overall Survival(OS)of the model in the training set was analyzed by using R language,and the survival curves of the high-and low-risk groups were drawn to verify the predictive ability of the prognostic model in the validation set.Multivariate COX Analysis was used to analyze the independent prognostic factors of lung adenocarcinoma,and a nomogram was constructed.The calibration curve and Decision Curve Analysis(DCA)were performed to evaluate the ability of the nomogram to predict the prognosis of patients.The CIBERSORT algorithm was used to analyze the infiltration of 22 tumor immune cells in thehigh andlow risk groups,and the immune microenvironment was scored,and the tumor purity of the samples in the high and low risk groups was calculated.The Genomics of Drug Sensitivity in Cancer(GDSC)database was used to analyze the differences in drug sensitivity between high and low risk groups.Results:Basedon the expression of anoikis genes,LUAD samples could be divided into two types,Aand B,and the survival of the two groups was significantly different(p<0.001).A linear risk score model consisting of the expression levels of PBK,KL,SLC2A1 and risk coefficient was constructed by lasso regression analysis.KL was a low-risk gene,and PBK and SLC2A1 were high-risk genes.This model could predict the survival rate of LUAD patients.The risk of group B is higher than that of group A,and the proportion of high-risk LUAD patients who died was higher than that of low-risk LUAD patients,suggesting that high-risk is associated with worse prognosis.The results of GSVA pathway enrichment analysis showed that metabolism-related pathways such as drug metabolism and histidine metabolism were more active in group A,and pathways related to cell proliferation such as oocyte meiosis,DNAdamage repair and cell cycle were more active in group B.The prognostic risk model not only can distinguish between high and low risk LUAD patients,but also has the ability to predict partial chemotherapy and targeted drug sensitivity in lung adenocarcinoma,which has potential value in guiding the selection ofanti-tumor drugs inlung adenocarcinoma.Conclusion:This study revealed a potential relationship between ARGs and LUAD outcome.The prognostic predictors identified in this study can be used as potential biomarkers for clinical application.