Construction Of Tumor Immune Microenvironment Evaluation Model Based On Functional Gene Sets

The Tumor Microenvironment(TME)is the environment in which tumor development occurs,consisting of tumor cells,immune cells infiltrating tumor tissue,fibroblasts,endothelial cells and other cell types,and non-cellular tissue components.In TME,components related to immunity construct tumor immune microenviroment(TIME).TIME is not only closely related to tumor genesis,progression,metastasis and prognosis,but also an important influencing factor in response to drug therapy.The development of evaluation models on TIME and its correlation with prognosis and treatment effects are important research topics in this field.Currently,most of the TIME evaluation models use single genes as important variables,which affects the evaluation accuracy;or overemphasize the characteristic genes of specific cancer type,which leads to poor generalization ability;moreover,the existing models are less concerned with the predictive function of tumor immunotherapy response.In view of this,this study constructs a TIME evaluation model that can be used to evaluate prognosis and immunotherapy response based on the immune function gene set,and validates it on a large cancer data set.Specifically including:(1)constructing IScore,an immune microenvironment scoring model based on immune-related functional gene sets.first,the immune functional gene sets significantly associated with prognosis were screened based on Gene set variation analysis(GSVA)scores and LASSO-Cox regression.Then,the Immune Score model(IScore model)was constructed using multifactorial Cox regression.Finally,the IScore model was evaluated in terms of clinical characteristics,immune infiltration characteristics and differentially expressed genes;and the IScore model was applied to pan-cancer data analysis.The results showed that the IScore model was an independent significant protective factor(HR=0.27)over other clinical features,the subgroup with high IScore score had good prognosis,and the upregulated genes were enriched in important immune-related pathways such as T-cell activation and its regulation,lymphocyte differentiation,and inter-lymphocyte adhesion and its regulation.In the application of 32 cancers,there was a significant difference in the survival curves of high IScore subgroups in 17 cancers(P<0.05).(2)Prediction of IT typing for immunotherapy based on IScore and Tumor Mutation Burden(TMB),a clinically certified prognostic marker for anti-PD-1 immunotherapy,was analyzed in this study to propose the relationship between four factors: IScore,TMB,immunotherapy efficacy and survival outcome,and proposed the IScore-TMB typing(IT typing).In IT typing,the IT-H subtype refers to the type in which both IScore and TMB are above the cut-off point,the IT-L subtype refers to the type in which both IScore and TMB are below the cut-off point,and the Other subtype refers to the other two combinations of IScore and TMB.In melanoma samples,there was a significant difference in survival curves among IT-H subtype,Other subtype,and IT-L subtype(P=0.000159),and survival rates decreased sequentially,with over 55% of patients in complete remission with immunotherapy coming from IT-H subtype.Validation in uroepithelial carcinoma yielded the same results,with 59% of patients responding to immunotherapy coming from the IT-H subtype and only3% from the IT-L subtype.Significant differences in survival curves were found for the three IT subtypes in 17 of the 32 cancer types(p<0.05).The IScore model is a statistically validated independent prognostic protective factor,and the combination of IScore model and TMB parameters for IT typing can predict the effect of immunotherapy,which has clinical utility and can also provide clues and basis for precision treatment of cancer.