Identification Of A Robust Prognostic Signature Based On Tumor Immune Microenvironment Of Lung Adenocarcinoma For Predicting Immunotherapy Response

Background:Immunotherapy has remarkably improved patients’ outcomes in lung adenocarcinoma(LUAD).Clinical study has shown that the first-line treatment of PD-1 inhibitor in patients with PD-L1 expression>50%can significantly improve the 5-year overall survival rate to more than 30%,which is much higher than traditional chemotherapy.However,the overall response rate to immunotherapy is less than 50%,few efficient biomarkers have been identified to distinguish patients who would obtain immunotherapeutic benefits.Studies have shown that the tumor immune microenvironment(TIME)is closely related to the efficacy of immunotherapy in cancer patients,which is a hot spot in the field of cancer research.We tried to construct a prognostic model based on the TIME of LUAD to predict the prognosis and immunotherapy efficacy of LUAD patients.Objective:We aimed to screen the tumor immune microenvironment-related genes of LUAD patients through bioinformatic methods,and construct a LUAD tumor immune microenvironment prognostic signature(LATPS)to evaluate the prognosis of LUAD patients and predict the efficacy of immunotherapy.Methods:We collected gene expression and corresponding clinical information from TCGA and GEO databases.Then,we utilized CIBERSORT and ESTIMATE algorithms to describe the immune infiltrative landscape of LUAD patients and used the unsupervised clustering method to categorize LUAD patients into different subgroups.Differential expression analysis was applied to screen out transcriptional variations between LUAD subgroups and identify TME related prognostic genes.A LUAD’s tumor immune microenvironment prognostic signature(LATPS)was then constructed using the least absolute shrinkage and selection operator and Cox regression analyses.Subsequently,we explored the correlation between the LATPS and tumor-infiltrating immune cells(TIICs),immune-related pathways,tumor immune dysfunction and exclusion(TIDE),and its predictive value for predicting immunotherapy response in immunotherapy cohorts.Results:A total of 1088 LUAD samples were analyzed in this study;149 TME related genes were screened out by differential expression analysis.Then,12 TME related prognostic genes were identified by Univariate Cox regression and Lasso analyses.The LATPS was further constructed based on the 4 hub genes by Multivariate Cox regression analysis.LUAD patients were stratified into LATPS-high and LATPSlow subgroups according to the median LATPS score.The LATPS-low subgroup had a better overall survival(OS),which was verified in the internal and external LUAD cohorts.Meanwhile,the LATPS-low subgroup tended to be a hot immune phenotype,which was characterized by an elevated abundance of immune cell infiltration and increased activity of immune-related pathways.Additionally,TIDE was markedly decreased in the LATPS-low subgroup,indicating an enhanced opportunity to benefit from immunotherapy.Survival analysis in the immunotherapy cohorts revealed that the LATPS-low subgroup had better immunotherapy efficacy.Conclusion:The LATPS established based on the tumor immune microenvironment of LUAD is a promising biomarker that can effectively predict the prognosis of LUAD patients.The LATPS score can more easily and effectively predict the efficacy of immunotherapy compared with TIDE.