Clinical And Prognostic Implications Of An Immune-related Model(IRM) Based On TP53 Status In Lung Adenocarcinoma

Objective:The tumor suppressor gene TP53 is the five most conspicuous mutations in lung adenocarcinoma and has been certified that it is associated with immune response and inflammatory diseases.This study aims to construct an immune-related model that can be used for clinical prognosis analysis based on the mutation status of TP53.Method:We obtained the clinical baseline data and transcriptome from TCGA(The Cancer Genome Atlas)database.These patients included 236 patients with TP53 mutant lung adenocarcinoma and 260 patients with TP53 wild type lung adenocarcinoma.Differentially expressed genes(DEGs)between TP53 mutant and TP53 wild-type patients were sifted out,using "DESeq2" R software package.We further screened immune related differentially expressed genes(IRDEGs)through Imm Port database.GO analysis(Gene Ontology)and KEGG analysis(The Kyoto Encyclopedia of Genes and Genomes)were performed on DEGs and IRDEGs.We used LASSO Cox model to establish a six gene immune related model(IRM)to predict the survival rate of lung adenocarcinoma patients in the TCGA cohort based on the status of TP53.Then,we use GEO(Gene Expression Omnibus)dataset and the Nanjing cohort to verify the predictive ability of this model.A stratified survival analysis of this model were performed in clinically relevant subgroups.The differences in the immune microenvironment of the two groups of patients were verified by immunohistochemistry and q RT-PCR.The independent prognostic ability of the IRM index was tested by univariate and multivariate Cox analyses.Using the R package "rms",a concise nomogram for predicting theoverall survival(OS)of lung adenocarcinoma was established.Result:Compared with TP53 wild-type group,1829 genes were differentially expressed in TP53 mutant group.We found that 1230 genes were down-regulated while 599 genes were up-regulated.A total of 75 IRDEGs were further screened.According to LASSO cox regression analysis,among these 75 IRDEGs,six genes(including CRHR2,BPIFB2,INHA,SSTR5,SCGB3A1 and BPIFB1)with non-zero regression coefficients were found to have the greatest prognostic value,and IRM was constructed based on this.The best cut-off point of the IRM index is-0.959 524 3.Kaplan-Meier analysis suggest that high index patients had worse overall survival.The timedependent ROC curve analysis of the IRM index in the TCGA cohort showed that the prognosis of OS was good(3-year AUC = 0.717,5-year AUC = 0.730,7-year AUC =0.731).These results have been verified in the GEO dataset and the Nanjing cohort.In different clinical subgroups,the IRM index is still an effective indicator for predicting OS.These subgroups include younger(≤65 years)(P = 0.00036)or older(> 65 years)(P <0.0001),male(P <0.0001)or female(P <0.0001),early tumor(TNM stage I)(P =0.0026)or advanced patient(TNM stage II,III,IV)(P <0.001).In the IRM high index group,there were more memory B cells and regulatory T cells around the tumor tissue(P <0.001,P <0.001),while in the IRM low index group,more neutrophils and resting memory CD4 + T cells(P <0.001,P = 0.007)infiltrated around the tumor tissue.The IRM index was significantly correlated with the expression of immune checkpoints such as TIM3,TIGIT,PDCD1 and CD274(P <0.05).The expression of PDCD1(P = 0.005)and CD274(P <0.001)in the high index group was markedly more than that in the low index group.q RT-PCR indicate that patients with high index expressed more PDCD1(P = 0.0038)and CD274(P = 0.004)in the Nanjing cohort.The results of immunohistochemistry proved that the IRM index is obviously positively correlated with IHC scores of these two immune checkpoint proteins.Univariate and multivariate Cox analysis showed lymphatic invasion(HR = 3.201,95%CI: 2.930～4.058),IRM index(HR = 5.395,95%CI: 4.059～6.921),age(HR =1.124,95%CI: 1.002～1.303)and TNM staging(HR = 1.420,95%CI: 1.117～1.806)are independent risk factors.Conclusion:There are differences in the immune microenvironment between TP53 mutant and wild-type lung adenocarcinoma patients.The IRM model constructed based on the immune-related differential genes between them can effectively predict the prognosis of lung adenocarcinoma.IRM index is an independent prognostic factor for the overall survival of lung adenocarcinoma patients.The nomogram constructed based on the IRM index provides clinicians with a quantitative method to predict the prognosis of patients with lung adenocarcinoma.