Construction Of A New Tumor Immunity-related Signature To Assess And Classify The Prognostic Risk Of Ovarian Cancer

ObjectiveThis study aimed to construct a new immune gene signature of ovarian cancer,and to explore the prognostic value of the new score and its correlation with the tumor microenvironment.MethodWe obtained antigen presentation and processing genes from the Imm Port database,and used the GSE26712 dataset to search for differentially expressed protein-coding genes in serious ovarian cancer(SOC).Immune score and DEGs score were constructed in SOC RNA-seq data from TCGA and ICGC databases,respectively.The gene signature construction method was as follows:1)Univariate Cox analysis was used to obtain protein-coding genes with prognostic value;2)Lasso regression to reduce the dimensionality of the data to screen candidate genes;3)and incorporated candidate genes into the multivariate Cox model.The risk score was the sum of the product of the candidate gene expression value and the Cox regression coefficient.Then,survival curves,multivariate Cox and nomogram were used to evaluate the prognostic value of gene signatures.Finally,the relationship between gene signatures and tumor immune infiltration and tumor microenvironment was compared.ResultImmune score and DEGs score had prognostic value in both training and external validation dataset(HR: 0.379 vs.0.450;0.333 vs.0.327).Multivariate Cox results showed that the area under the 30-month Time-ROC curve(AUC)in the training and validation datasets were 0.665 and 0.743,respectively;the calibration curves were basically consistent with their diagonals.Compared with the published 9-gene signature prediction model,the Immune sore,DEGs score and the composite variable had better discrimination(AUC: 0.657 VS.0.712,0.710 and 0.736).In addition,scores were linear correlated with biological scores such as CYT,IFN-γ score and TIDE,transcript levels of immune checkpoint genes(PD1,PD-L1 and CTLA4),and other immune molecules(PDCD1LG2,CD86,CD80,CD27,CD40,CD70,TNFRSF14,CD276,VTCN1,IDO1,TIGIT,ICOS,CD58,LAG3 and HAVCR2).This reflected the value of Immune score in predicting SOC immunotherapy response.To explore the relationship between risk values and SOC immune microenvironment and its immune infiltrating cells.The scores of B cell,Class-switched memory B cell,T cell CD8+,T cell CD8+ central memory,T cell gamma delta,Macrophage M1,Myeloid dendritic cell activated and Plasmacytoid dendritic cell in the Low-risk group were higher than High-risk group.And cancer associated fibroblast was highly expressed in the High-risk group.Conclusion1.Immune score can effectively reflect part of the internal immune heterogeneity of SOC,and can stratify the overall survival of SOC;2.DEGs score can also effectively stratify SOC prognosis,and combined with Immune score variable analysis can increase the ability to predict OS in SOC patients;3.Immune score is an effective biomarker for predicting response to immune checkpoint inhibitor therapy.