Comprehensive analysis of immune escape-related prognostic signature in high-grade glioma

Comprehensive analysis of immune escape-related prognostic signature in high-grade glioma | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Comprehensive analysis of immune escape-related prognostic signature in high-grade glioma Chongkang Ren, Jinyi Cai, Yanhua Liu, Zeshang Guo, Xinyu Hong This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6330200/v1 This work is licensed under a CC BY 4.0 License Status: Under Review Version 1 posted 12 You are reading this latest preprint version Abstract Objective High-grade glioma (HGG) is one of the most lethal malignancies. Immune escape is considered to be a reason for the failure of immunotherapy for HGG patients, and there is currently no immune escape-related prognostic model for glioma. Therefore, we explored the relationship between immune escape-related genes and the prognosis of patients with HGG. Methods This study combined 101 machine learning algorithms to determine the best immune escape-related prognostic model. Subsequently, the TCGA and CGGA cohorts were used to verify the effectiveness of the model. Subsequently, molecular docking, Mendelian randomization and other comprehensive analyses were performed on the model genes. Finally, the biology function of the signature gene was further verified via CCK-8, and colony formation. Results Our differential expression analysis found that 41 immune escape-related genes were significantly related to the prognosis of HGG patients. We further selected 18 key genes through various machine learning methods to construct an immune escape-related prognosis model. This model can effectively distinguish between high-risk and low-risk groups, and shows good prediction results in both CGGA and TCGA data sets. Subsequently, the risk score was found to be an independent prognostic factor, and the nomogram including clinical characteristics was constructed. Immunotherapy response prediction results show that patients in the low-risk group respond better to immunotherapy and have longer survival. Furthermore, TAB1 knockdown reduced the ability of human glioma cells to proliferate and clone. Conclusion This study constructed a prognostic model related to immune escape through multiple machine learning methods and verified its clinical application value in HGG patients. It provides a theoretical basis for the exploration of immune escape treatment targets. High-grade glioma immune escape machine learning immune microenvironment immunotherapy Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 Figure 9 Figure 10 Figure 11 Figure 12 Introduction High-grade glioma (HGG, WHO grade III and IV) are common primary malignant brain tumors, particularly the higher-grade subtypes, such as glioblastoma multiforme (GBM) and anaplastic astrocytoma, both of which exhibit extremely high malignancy [ 1 ]. The heterogeneity of gliomas leads to significant differences in clinical manifestations, molecular characteristics, and treatment responses among patients. Although histological classification provides a foundation for clinical treatment, patients with similar molecular features may exhibit vastly different prognoses [ 2 ]. Among them, GBM is the most lethal subtype of gliomas, with a particularly poor prognosis and a five-year survival rate of 5.5% [ 3 ]. Despite the fact that treatment for HGG currently relies primarily on surgical resection, radiation therapy, and chemotherapy, the therapeutic outcomes remain limited, and most patients experience recurrence after surgery [ 4 ]. Recent studies have shown that immunotherapy offers new therapeutic avenues for various malignant tumors, particularly with immune checkpoint inhibitors (ICIs) demonstrating significant efficacy in lung cancer, melanoma, and other tumors [ 5 – 7 ]. However, despite the improved efficacy of ICIs in other types of cancer, immunotherapy responses in HGG remain suboptimal [ 8 ]. This phenomenon is closely related to the immune escape mechanisms present within the tumor microenvironment [ 9 ]. Therefore, there is an urgent need to develop novel therapeutic strategies and evaluation models to enhance the immunotherapy outcomes in HGG patients. Immune escape refers to the ability of tumor cells to evade host immune surveillance through various mechanisms, thereby promoting tumor growth, progression, and metastasis [ 10 ]. In recent years, immune escape mechanisms have become a focal point in cancer research. Tumor cells regulate immune responses through immune checkpoint molecules (e.g., PD-1/PD-L1, CTLA-4) or alter the function of immune cells in the tumor microenvironment, enabling them to evade immune surveillance [ 11 , 12 ]. In gliomas, immune escape mechanisms are particularly complex [ 13 ]. The unique immunosuppressive microenvironment of GBM, which includes a high density of regulatory T cells (Tregs), M2 macrophages, and cancer-associated fibroblasts (CAFs), provides an ideal setting for immune escape [ 13 ]. Additionally, glioma cells further suppress immune system function by secreting immunosuppressive factors, such as TGF-β and IL-10, and expressing immune checkpoint molecules, such as PD-L1, thereby exacerbating immune escape [ 13 – 15 ]. Although there has been some progress in the study of immune escape mechanisms in glioblastoma, no novel prognostic models targeting immune escape-related genes have been reported to date. Therefore, identifying immune escape-related genes with prognostic significance in HGG and exploring their interactions with the immune microenvironment has become a key research focus in this field. This study aims to construct a prognostic model based on the combination of 101 machine learning algorithms to identify immune escape-related genes, and further explore the role of these genes in the immune microenvironment of HGG and their relationship with patient responses to immunotherapy. By constructing and validating this model, we aim not only to provide more accurate prognostic assessments for HGG patients but also to offer new perspectives for the development of personalized immunotherapy strategies. Furthermore, we will further validate the efficacy of key genes in the model using techniques such as single-cell analysis, molecular docking, and the Mendelian randomization (MR). Through these comprehensive analyses, we aim to enhance the optimization of HGG immunotherapy, identify novel immune escape therapeutic targets, and provide more precise strategies for clinical treatment. Materials and methods Data collection and processing The data for this study were sourced from several public databases, including the Chinese Glioma Genome Atlas (CGGA, http://www.cgga.org.cn/ ) and The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/ ), to obtain gene expression and clinical data from high-grade glioma patients. Additionally, normal brain transcriptomic data were downloaded from the Adult Genotype Tissue Expression (GTEx, https://xenabrowser.net/ ). During data collection, samples with complete clinical and gene expression data were selected, while groups with insufficient sample sizes or significant missing data were excluded, ensuring data integrity and accuracy. A total of 182 immune escape genes were downloaded from previous studies (Table S1 ) [ 16 ]. In data preprocessing, all gene expression data were first standardized to eliminate measurement scale differences between samples, ensuring comparability. Additionally, to correct for batch effects between datasets, the ComBat algorithm was applied [ 17 ]. The flowchart of this study was shown in Figure S1 Identification and differential analysis of immune escape-related genes Differential expression analysis used the R package “limma” to evaluate the differences in expression of immune escape-related genes between the tumor group and the normal group. The Wilcoxon test was used to calculate the log 2 (Fold Change) and the false discovery rate (FDR) for the differentially expressed genes, followed by multiple hypothesis testing correction. A threshold for differential expression analysis was set at |log 2 FC| > 0.585 and FDR < 0.05, which identified significantly differentially expressed genes for subsequent analysis. Construction and validation of immune escape-related machine learning models We obtained 722 HGG samples and 20 normal control samples from the CGGA database. First, univariate Cox regression analysis was performed to identify differentially expressed immune escape genes significantly associated with the prognosis of high-grade glioma patients. Using p value < 0.05 as the significance criterion, the hazard ratio (HR) and its confidence interval of each gene were displayed to further evaluate its impact on patient survival. To construct a risk score model based on immune escape related genes, the machine learning algorithms used in this study included CoxBoost, stepwise Cox, survival support vector machine (survival-SVM), elastic network (Enet), least absolute shrinkage and selection operator (Lasso), partial least squares regression for Cox (plsRCox), Ridge, random survival forest (RSF), generalized boosted regression modeling (GBM), and supervised principal components (SuperPC) respectively [ 18 ]. These algorithms can handle high-dimensional data and select features through appropriate algorithm combinations. Next, we performed multivariate Cox regression using the “survival” R package to establish an immune escape-related prognostic model based on the following formula: Immune Escape Index (IEI) = Σ(Ci * Ei), where Ci represents the coefficient calculated for each signature gene in the multivariate Cox regression analysis, and Ei denotes the expression value of each signature gene. This risk score allows further prognostic stratification of patients. For model validation, the CGGA dataset was used as the training set and the TCGA dataset as the validation set. The model's performance was assessed using the concordance index (C-index), a commonly used metric in survival analysis that measures the correlation between the predicted risk scores and actual survival times. Additionally, the model's predictive ability was further validated by plotting the receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC). Risk stratification and survival analysis Based on the immune escape-related index calculated using the machine learning model, patients were classified into high-IEI and low-IEI groups. Kaplan-Meier survival curves were used to assess the survival differences between the risk groups, and the statistical significance of the survival differences between the two groups was determined using the Log-rank test. The "pheatmap" package was used to visualize the distribution of risk values between the two groups. The "timeROC" package was used to generate ROC curves to assess the predictive accuracy for 1-year, 3-year, and 5-year overall survival (OS) in HGG. The predictive accuracy was determined based on the AUC values. Construction and validation of a nomogram in HGG patients We conducted both univariate and multivariate Cox regression analyses to further assess whether the immune escape-related risk score could serve as an independent prognostic indicator. Furthermore, we developed a nomogram for HGG patients, which visualizes the combined impact of various clinical features and the risk score on prognosis. The predictive accuracy of the nomogram was evaluated using a calibration curve, which illustrates the relationship between the predicted survival probabilities and the actual observed values. To validate the clinical applicability of the model, we also employed the C-index, ROC curve, and decision curve analysis (DCA) to further evaluate the model's clinical net benefit and predictive accuracy. Immune escape-related gene set enrichment analysis This study employed the GeneMANIA tool ( http://genemania.org/ ) to analyze the association network of 18 immune escape-related model genes, integrating and visualizing their relationships. Subsequently, to explore the biological roles of the immune escape genes in HGG, we performed the Gene Set Enrichment Analysis (GSEA) to analyze the significantly enriched pathways in both the high-IEI and low-IEI groups. The GSEA enabled the identification of potential biological pathways related to immune escape, providing key insights for subsequent mechanistic studies. Analysis of tumor microenvironment We calculated differences in tumor microenvironment scores between different subgroups using the ESTIMATE algorithm, including the Stromal Score, Immune Score, and ESTIMATE Score. This analysis assessed the composition of the HGG tumor microenvironment to better understand tumor behavior and design targeted therapeutic strategies. Subsequently, immune microenvironment analysis was performed using the CIBERSORT algorithm based on 7 different immune databases (QUANTISEQ, XCELL, EPIC, TIMER, MCPCOUNTER, CIBERSORT, and CIBERSORT-ABS) to evaluate immune cell infiltration in HGG. The spearman correlation analysis was then used to further investigate the relationship between immune escape genes and immune cell infiltration, identifying immune cell types significantly correlated with immune escape genes. Additionally, the relationship between each immune escape-related model gene and the immune microenvironment was visualized to help elucidate the potential roles of immune escape genes in the tumor immune microenvironment. Assessment and prediction of immunotherapy We evaluated the differential expression of immune checkpoint-related molecules across different subgroups. The varying expression of immune checkpoints in this prognostic risk model may offer new therapeutic insights and predict the clinical effectiveness of corresponding ICIs. Additionally, to explore the differences in tumor mutational burden (TMB, mutations per million base pairs) between the two risk subgroups, we calculated the TMB for each HGG patient. Subsequently, the immune treatment cohort was downloaded from the Tumor Immune Dysfunction and Exclusion ( TIDE, http://tide.dfci.harvard.edu/ ). The TIDE scoring was used to assess the potential impact of immune escape genes on the effectiveness of immune therapy and evaluate the sensitivity differences to immunotherapy between the high-IEI and low-IEI groups. The response of this prognostic signature to immunotherapy was then predicted based on the IMvigor210 cohort. The IMvigor210 cohort was used to predict patients' responses to immunotherapy [ 19 ]. Under the Creative Commons 3.0 license, complete expression and clinical data were downloaded from http://research-pub.Gene.com/IMvigor210CoreBiologies . The raw data were then normalized, and the counts were converted into TPM values. Single-cell analysis The single-cell sequencing data of GBM, comprising 6,148 cells from 73 regions of 14 patients, were downloaded from the CGGA database and related previous literature [ 20 – 22 ]. Based on this, the distribution of immune escape-related model genes was explored. The R package "Seurat" was used to process the single-cell expression matrix, followed by further processing using the FindVariableFeatures, ScaleData function, principal component analysis (PCA), and FindNeighbors [ 23 ]. Finally, the UMAP was employed for visualization. Molecular docking and drug sensitivity The molecular docking analysis was performed using the CB-Dock2 tool ( https://cadd.labshare.cn/cb-dock2/index.php ) to investigate the molecular interactions between immune escape-related model genes and the chemotherapy drug Temozolomide. Molecular dynamics simulations were employed to validate the stability of the molecular docking complex. Additionally, drug sensitivity analysis was performed using the "oncoPredict" package to predict drug responses in the high-IEI and low-IEI groups, with boxplots used to visualize the sensitivity differences between the groups. MR analysis Two-sample MR was conducted using the "TwoSampleMR" package to assess the causal relationship between immune escape-related model genes and glioma risk. Significant single nucleotide polymorphisms (SNPs) from the Integrative Epidemiology Unit (IEU, https://gwas.mrcieu.ac.uk ) and the FinnGen ( https://www.finngen.fi/en ) datasets were selected as instrumental variables to ensure their relevance and independence. Sensitivity analysis was performed using leave-one-out analysis and MR-PRESSO to assess the robustness and pleiotropy of the instrumental variables [ 24 ]. In the pleiotropy test, a p-value greater than 0.05 indicates the absence of pleiotropic effects, meeting the criteria for further analysis. The SNPs with a p-value less than 5e-08 in the outcome data were selected as valid instrumental variables for MR analysis. Causal inference methods included IVW (Inverse Variance Weighted), MR-Egger, Weighted median, Simple mode, and Weighted mode [ 25 ]. Among these methods, if all SNPs included in this analysis meet the assumptions of valid instruments, the IVW method will accurately estimate the causal effect between exposure and outcome, and thus will be used as the primary method. Cell culture and transfection Two human glioma cell lines, U251 and SW1088, were obtained from Procell Life Science & Technology Co., Ltd. (Wuhan, China). U251 and SW1088 were cultured in Dulbecco’s modified Eagle’s Medium (DMEM) supplied with 10% fetal bovine serum (FBS) and 1% penicillin–streptomycin. The cells were cultured at 37°C and 5% CO 2 . TAB1 inhibitors and its negative controls (NC) were produced by Sangon Biotech (Shanghai, China). Total RNA samples of the two cell lines were then extracted using TRIzol reagent and reverse transcribed using the SYBR Green Master Mix kit. CCK-8 assay CCK-8 kit was used to detect the proliferation ability of human glioma cells after TAB1 expression was downregulated. The cells were seeded into 96-well plates at a concentration of 5×10 3 cells per well. At 0, 24, 48, and 72 hours after attachment, 10 µL CCK-8 solution was added to each well and incubated in the incubator for 2 hours. Finally, the absorbance of each group at 450 nm was measured using an enzyme labeling instrument. Colony formation assay The differences in colony-forming ability of human glioma cells in different transfection groups were detected. U251 and SW1088 cells were seeded in each well of a 6-well plate. After 14 days of culture, the cell colonies were fixed with 4% paraformaldehyde and then stained with 0.1% crystal violet. Statistical analysis In this study, all statistical analyses were conducted using the R software (version 4.4.1). The Wilcoxon rank sum test, log-rank test, and the one-way ANOVA were used to compare statistical differences. The cox regression analysis and the Kaplan-Meier curves were used to assess prognostic value. Correlation analysis was performed using the Pearson and the Spearman correlation methods. All statistical analyses were conducted with two-sided tests, and a significance level of p-value < 0.05 was applied unless otherwise specified. Results Identification and survival analysis of differentially expressed genes related to immune escape Gene expression data from HGG patients in the CGGA and TCGA databases, along with normal samples from the GTEx, were standardized, and batch effects were corrected using the ComBat algorithm to ensure data comparability and accuracy. Differential expression analysis of immune escape-related genes between tumor and normal groups in the CGGA cohort was performed using the "limma" package, identifying 41 significantly differentially expressed genes, including 20 upregulated and 21 downregulated genes. The heatmap (Fig. 1 A) illustrates the expression patterns of these genes in the samples, while the volcano plot (Fig. 1 B) further shows the distribution of upregulated and downregulated genes. The univariate Cox regression analysis revealed that the 41 differentially expressed immune escape-related genes were significantly associated with OS in HGG patients (p < 0.05), with HR and 95% confidence intervals shown in Fig. 1 C. These genes may serve as potential prognostic biomarkers and provide a foundation for subsequent model development. Construction and validation of prognostic signature based on machine learning A total of 101 machine learning model combinations were employed to perform feature selection on the 41 prognosis-related genes (Fig. 2 A). The best model trained on the CGGA cohort identified 18 key genes for constructing the immune escape-related prognostic model. The optimal model algorithm combination was CoxBoost and Enet (alpha = 0.2). The multivariate Cox regression analysis was used to calculate the IMI for each patient. Afterward, HGG patients were classified into high-IEI and low-IEI groups based on the median IMI value. In the CGGA cohort, the distribution of risk scores and survival status are shown in Fig. 2 B. The survival rate of high-IEI patients was significantly lower than that of low-IEI patients (p < 0.001, Fig. 2 D). In the TCGA cohort, the model similarly stratified patients by the risk score (Fig. 2 C), with the survival rate of high-IEI patients significantly lower than that of low-IEI patients (p < 0.001, Fig. 2 E). The ROC curve showed that the signature achieved AUC values of 0.710, 0.777, and 0.813 for 1-year, 3-year, and 5-year survival in the CGGA cohort (Fig. 2 F), and 0.800, 0.850, and 0.808 in the TCGA cohort (Fig. 2 G), indicating good predictive performance. Construction of a nomogram of HGG patients The univariate and multivariate Cox regression analyses were conducted to assess whether the risk score is an independent prognostic factor (Fig. 3 A and 3 B). The results showed that, after adjusting for clinical features such as age, gender, and WHO grade, the risk score remained an independent prognostic factor for HGG patients (HR = 2.14, 95% CI: 1.69–2.73, p < 0.001). Furthermore, the prognostic ability of the risk score was superior to that of other clinical features, with the highest AUC value on the ROC curve (Fig. 3 C). The C-index analysis further supported this conclusion (Fig. 3 D). Based on the results of the multivariate Cox regression analysis, a nomogram was constructed, incorporating the risk score and key clinical features, to predict the 1-year, 3-year, and 5-year survival probabilities (Fig. 3 E). The calibration curve showed that the predicted survival rates from the nomogram closely matched the actual observed values, indicating good predictive accuracy of the model (Fig. 3 F). The DCA revealed that the nomogram model provided higher clinical net benefits across different threshold probabilities (Fig. 3 G). Gene set enrichment analysis The interaction network of the 18 immune escape-related model genes was visualized using the Circos plot (Fig. 4 A). The results revealed complex interrelationships between different genes. The gene association network generated by the GeneMANIA revealed the complex interactions of the 18 key genes within inflammation and immune signaling pathways (Fig. 4 B). The results indicated that genes such as TNFRSF1A, TNFAIP3, and TRAF2 exhibited high connectivity in the network, suggesting their potential role as core regulatory factors in the TNF/NF-kappaB signaling pathway. The GSEA was conducted to gain deeper insights into the biological functions of the immune escape-related genes. The results showed that functional and pathway enrichment analyses were performed on the gene expression profiles of the high-IEI and low-IEI groups. In the GO biological process enrichment analysis, the high-IEI group was significantly enriched in immune-related biological processes, including adaptive immune response and lymphocyte-mediated immunity (Fig. 4 C), suggesting more active roles in inflammation and immune regulation. In contrast, the low-IEI group was primarily enriched in neuro-related biological processes, such as postsynaptic membrane activity and cation channel complexes (Fig. 4 D), indicating its characteristics are more aligned with homeostatic regulation and neural function. In the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, genes in the high-IEI group were significantly enriched in pathways associated with tumor progression and immune signaling, including cell cycle and cytokine-cytokine receptor interactions (Fig. 4 E), reflecting their potential higher proliferative capacity and inflammatory characteristics. In contrast, the low-IEI group was predominantly enriched in calcium signaling pathway and neuroactive ligand-receptor interaction (Fig. 4 F), suggesting that its biological characteristics are more aligned with normal tissue function. Overall, the high-IEI group exhibited prominent features of immune activation and tumor proliferation, while the low-IEI group was more associated with neural signaling and tissue homeostasis, providing important insights into the biological significance of the risk stratification. Analysis of immune cell infiltration Immune cell infiltration was analyzed using the CIBERSORT algorithm and 7 immune databases. The bubble plot shows a significant correlation between the expression of certain immune cell types and the prognostic model genes (Fig. 5 A). The network plot illustrates significant interactions between immune cells and prognostic genes, suggesting that these genes may influence the remodeling of the immune microenvironment (Fig. 5 B). The tumor microenvironment characteristics were assessed using the ESTIMATE algorithm, revealing that the stromal score, immune score, and ESTIMATE score of the high-IEI group were significantly higher than those of the low-IEI group (p < 0.05, Fig. 6 A), indicating a more complex tumor microenvironment in the high-IEI group. The comparison of immune subtype proportions revealed that the high-IEI group was more likely to exhibit an immunosuppressive subtype (Fig. 6 B). Further analysis was conducted on the relative proportions of 21 tumor-infiltrating immune cells in each HGG sample. The results showed that the enrichment of tumor-infiltrating immune cells varied across different risk groups (Fig. 6 C). Tumor mutation burden and prediction of immunotherapy response This study examined the relationship between the risk score and TMB, finding a significant positive correlation (r = 0.5, p < 0.001, Fig. 7 A). The TMB of the high-IEI group was significantly higher than that of the low-IEI group (Fig. 7 B). The Kaplan-Meier survival analysis showed that patients with high TMB levels had lower survival rates (Fig. 7 C). Additionally, we investigated the differences in the immune checkpoint expression. In the high-IEI group, the expression of nearly all key immune checkpoint-related molecules was upregulated (Fig. 6 D). The TIDE scores were used to assess the sensitivity of HGG patients in different risk groups to immune therapy, revealing that the high-IEI group had significantly higher TIDE scores than the low-IEI group (Fig. 7 D), indicating a higher likelihood of immune evasion in the high-IEI group. In the IMvigor210 cohort, further stratification of patients based on immune therapy response type revealed distinct risk scores in the progressive disease, stable disease, and partial response groups (Fig. 7 E). Compared to the stable disease and complete response groups, the progressive disease group had a higher proportion of high-risk scores (Fig. 7 F). The results showed that patients in the low-risk group had a higher response rate to immune therapy and significantly better survival rates compared to the high-risk group (Fig. 7 G). The ROC curve analysis further validated the model's efficacy in predicting immune therapy responses (Fig. 7 H). Single-cell RNA sequencing data analysis Using the marker genes for each cell subtype from previous literature [ 26 ], the UMAP plot dimensionality reduction was applied to visualize the distribution of different cell types (Fig. 8 A, Figure S2 ). The expression patterns of 18 model genes across different cell types were then revealed (Fig. 8 B). The results showed that these 18 genes exhibited diverse cell type-specific expression patterns (Fig. 8 C). Most genes, including TNFRSF1A, TNFRSF1B, IFNGR2, and IRF1, were highly expressed in T cells and myeloid cells, suggesting their potential role in immune regulation and inflammatory responses. Meanwhile, certain genes, such as HDAC1, NDUFA13, and KMT2A, showed significant expression in neurons, astrocytes, and oligodendrocytes, suggesting their multifunctional regulatory roles in the tumor microenvironment. Furthermore, the expression of AHSA1, EIF3H, ERP44, UBE2N, and VDAC2 was more widespread, showing significant expression across multiple cell types. These findings provide a foundation for further research into the functions of immune escape-related genes in different cell types. Drug sensitivity and molecular docking analysis Using the "oncoPredict" package, we predicted the chemotherapy drug sensitivity of HGG patients in different risk groups, identifying 121 drugs with differential sensitivity (Table S2 ). The high-risk group was more sensitive to drugs such as Carmustine, Nelarabine, and MIRA-1. In contrast, the low-risk group showed higher sensitivity to drugs like Dabrafenib, LCL161, and CZC24832 (Fig. 9 ). These findings offer potential drug options for personalized treatment. Moreover, the molecular docking analysis further explored the interactions between 18 key model genes and Temozolomide. The results revealed that 7 gene proteins exhibited strong binding affinities with Temozolomide (Table S3 ), with binding sites and interacting residues shown in Fig. 10 . These results provide a theoretical foundation for developing new therapeutic strategies. Mendelian randomization analysis to verify the causal relationships To validate the causal relationship between the model genes and glioma risk, the MR analysis was performed. Our analysis confirmed that all SNPs are robust instrumental variables. The results indicate that the expression levels of TNFAIP3, TNFRSF1B, and ERP44 are significantly causally associated with glioma (IVW method: p < 0.05, Figs. 11 A, C, E). Moreover, the trends in the odds ratios derived from these methods were consistent, with an increase in exposure factor levels corresponding to a reduced risk of disease occurrence. Sensitivity analysis revealed no significant heterogeneity or pleiotropy, confirming the robustness of the results (Figs. 11 B, D, F). Functional verification of gene TAB1 in vivo experiments First, we explored the hub genes of the model genes. The top three genes according to the scores were TNFRSF1A, TNFAIP3, and TAB1 (Table S3 ). Among them, only the gene TAB1 was highly expressed in HGG patients and was validated in the Human Protein Atlas (HPA, https://www.proteinatlas.org/ ) (Figs. 12 A and 12 B). However, the biological role of gene TAB1 in HGG has not been explored. Therefore, we selected gene TAB1 for further experimental validation. We used qRT-PCR to detect the expression level of TAB1 in two human glioma cell lines (U251, SW1088) treated with si-TAB1 to silence TAB1. Apparently, si-TAB1-1 and si-TAB1-2 significantly reduced the expression of TAB1 in those cell lines (Fig. 12 C). To evaluate the effect of TAB1 knockdown on cell proliferation, the results of CCK-8 assay showed that cell proliferation was significantly reduced after TAB1 knockdown compared with the control group (Fig. 12 D). In addition, colony formation assay results showed that TAB1 knockdown reduced clone survival (Fig. 12 E). Therefore, these results indicate that TAB1 gene downregulation can effectively inhibit the growth and proliferation ability of U251 and SW1088 cells. Discussion The microenvironment of HGG is typically considered to be highly immunosuppressive and heterogeneous, characteristics that significantly contribute to tumor progression and deterioration [ 27 ]. However, due to the high heterogeneity of HGG and the complexity of the immune microenvironment, developing effective models to predict patient prognosis remains a challenge [ 28 ]. The survival time of patients with HGG varies significantly, depending on tumor grade, the microenvironment, and individual patient differences, with median survival ranging from 14 to 18 months [ 29 ]. Establishing prognostic models helps predict patient survival and treatment responses, thereby providing more precise evidence for clinical decision-making. In recent years, with growing attention to the importance of immune escape in the tumor microenvironment, unique molecular characteristics related to immune escape have gradually been recognized [ 29 – 31 ]. The mechanisms of immune escape in glioma patients have increasingly become a focal point of research [ 13 , 32 ]. This study is the first to incorporate immune escape genes into a prognostic signature for HGG, which expands the research ideas in this field and suggests that immune escape plays a certain role in the biological behavior of glioma. In this study, the prognostic model constructed using 101 machine learning combinations demonstrated good predictive performance in risk stratification of high-IEI glioma patients. The model, based on the 18 key immune escape genes, utilized a combination of CoxBoost and Enet algorithms for feature selection and successfully distinguished high-IEI and low-IEI patient groups in the validation set. HGG patients in the high-IEI group had significantly lower survival rates compared to the low-IEI group, indicating the critical role of immune escape genes in tumor progression and further validating their clinical value as potential prognostic biomarkers. Additionally, the ROC curve analysis and the construction of nomogram further confirmed the stability and predictive accuracy of the model. Moreover, the prognostic model built using machine learning revealed the significant impact of these key genes on tumor progression and patient survival. In particular, genes such as TNFRSF1A, TNFAIP3, and TRAF2, through modulation of the immune microenvironment, influencing immune cell infiltration and the expression of immune checkpoint molecules, ultimately promote tumor immune escape [ 33 – 35 ]. We then explored the hub genes of the model genes. It was found that the hub gene transforming growth factor beta-activated kinase 1-binding protein 1 (TAB1) with a high score was highly expressed in HGG patients. TAB1 is a key regulator of TAK1, which promotes its activation by interacting with TAK1, thereby regulating the MAPK and NF-κB signaling pathways [ 36 ]. TAB1 has important pathophysiological significance in various diseases. Studies have shown that TAB1 mediates GFAT1-dependent p38 MAPK signaling activation through S438 phosphorylation, promotes cell autophagy and survival under nutritional stress, and is associated with poor prognosis in patients with lung adenocarcinoma [ 37 ]. Furthermore, TAB1 promotes TAK1 activation by forming the TAB1-TAK1-TAB2 complex with TAB2 and TAK1, thereby driving pro-inflammatory signaling in microglia during ischemic injury [ 38 ]. However, the biological role of TAB1 in HGG has been poorly studied. Our study reported for the first time that TAB1 is a tumor promoter factor in glioma, and knockout of TAB1 significantly inhibited the proliferation and cloning of glioma cells. Further results indicated that the significant differences in immune-related pathways between high-IEI and low-IEI groups may be key factors in determining patient prognosis. Furthermore, the tumor microenvironment of high-IEI patients exhibited more complex immunosuppressive characteristics, which correlated with their poorer survival rates, suggesting that these genes could serve as important targets for future immunotherapy. The immune escape mechanisms in gliomas are unique. Compared to solid tumors such as pancreatic cancer and non-small cell lung cancer, the immune microenvironment in gliomas is more significantly immunosuppressive, particularly in high-IEI patients, where immune checkpoint molecules are notably upregulated [ 9 , 39 , 40 ]. These molecules, through their synergistic interaction with immunosuppressive pathways, impair anti-tumor immune responses, thereby enhancing the tumor's immune escape ability [ 41 ]. By comparing the immune escape mechanisms across different tumor types, this study provides valuable insights into the unique immune characteristics of gliomas and offers guidance for future targeted research. Moreover, the prognostic signature based on the immune escape related genes can be used for precise risk stratification of HGG patients, providing a reliable reference for the development of personalized treatment plans. Due to their pronounced immunosuppressive characteristics, high-IEI patients may be more suitable for combination therapy with ICIs to overcome their immune escape mechanisms, whereas low-IEI patients may exhibit poor responses to monotherapy and require additional treatment strategies. Drug sensitivity analysis further revealed significant differences in chemotherapy responsiveness between high-IEI and low-IEI patients, providing potential evidence for personalized drug selection. Furthermore, this study reveals the crucial role of immune escape-related genes in regulating immune cell infiltration and reshaping the tumor microenvironment. In the high-IEI group, immunosuppressive cells, such as regulatory T cells, significantly increase in the tumor microenvironment, while the proportion of effector immune cells, such as Th1 and Th2 cells, remains relatively low. The changes in the proportion of these immune cells reflect an enhanced immunosuppressive state in the tumor microenvironment and a weakened anti-tumor immune response[ 42 , 43 ]. Targeting the immunosuppressive components of the tumor microenvironment may be an effective strategy to enhance the efficacy of immunotherapy. Subsequently, this study explores the relationship between risk scores and TMB, revealing that TMB is significantly higher in the high-IEI group compared to the low-IEI group. However, despite the higher TMB in the high-IEI group, these patients still have lower survival rates, suggesting that their glioma cells may evade immune system clearance through other immune escape mechanisms. Additionally, drug sensitivity analysis and molecular docking in this study identified potential therapeutic targets and drug combination strategies. Molecular docking analysis shows that temozolomide strongly binds to several key model gene proteins, offering potential pathways to improve its therapeutic efficacy. Based on these findings, future research may focus on developing novel immunotherapy combination strategies targeting the immunosuppressive features of HGG patients, such as combining ICIs with specific targeted therapies, which may effectively improve patient prognosis. This study has achieved some good results based on the integrated application of multiple machine learning algorithms, but there are still some limitations. First, the study mainly relies on transcriptome data provided by public databases and lacks verification of real clinical samples. More clinical samples need to be included in the future to improve the clinical applicability of the model. Secondly, this study only focuses on the predictive prognostic role of immune escape-related genes at the transcriptome level. It does not deeply explore its regulatory mechanisms at multiple molecular levels such as epigenetics and proteomics. In addition, although we have identified multiple key genes and preliminarily explored the biological role of TAB1 in HGG, the specific molecular mechanisms of these genes still need more in-depth experimental verification. For example, although this study found that TAB1 is closely related to tumor cell proliferation. However, its specific role in the immune microenvironment of glioma has not yet been elucidated, and it needs to be confirmed by more in-depth in vivo animal studies in the future. The prognostic model of immune escape-related genes proposed in this study performed well in predicting the risk stratification of HGG patients. However, it still needs in-depth optimization at multiple levels to promote its real application in clinical practice. In summary, this study established a prognostic model for HGG patients based on immune escape-related genes, suggesting that immune escape may play an important role in the biological behavior of glioma. This model showed good predictive ability in risk stratification, providing a potential reference for personalized treatment decisions, especially for the precise application of immunotherapy. However, these findings still need to be further verified by follow-up studies. In the future, the specific mechanism of glioma immune escape should be further explored through multi-omics data integration and in vivo and in vitro experiments, in order to provide more reliable theoretical support for the formulation of personalized treatment plans and improve the clinical efficacy of HGG patients. Conclusion This study constructed the prognostic signature based on immune escape-related genes, revealing the prognostic value of immune escape-related genes in HGG patients. This signature provides novel ideas for personalized treatment decisions, especially in the precise application of immunotherapy. The study provides direction for developing personalized treatment strategies targeting immune escape mechanisms. Declarations Acknowledgment The authors are grateful to all patients who provided samples to the public databases. Conflict of Interest The authors declare that they have no competing interests. Author Contributions XYH and ZSG conceived and planned this study. CKR, JYC, and YHL helped interpret the results. CKR took the lead in writing the manuscript. All authors provided vital feedback to shape the final version of the manuscript. Funding The present study was supported by the Jilin Provincial Medical and Health Talent Project (grant no. JLSWSRCZX2023-24). Data availability statement All data analyzed in the present study are publicly available in The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/), Chinese Glioma Genome Atlas (CGGA, http://www.cgga.org.cn/), Adult Genotype Tissue Expression (GTEx, https://xenabrowser.net/), Integrative Epidemiology Unit (IEU, https://gwas.mrcieu.ac.uk), the FinnGen (https://www.finngen.fi/en) and the Human Protein Atlas (HPA, https://www.proteinatlas.org/). The raw data and code for this study are reviewed via this link (https://www.jianguoyun.com/p/DSULx4AQ38nsChjHj9EEIAA). Informed consent not applicable. Ethics approval and consent to participate Our research was analyzed using publicly available data and did not involve direct human participation. Therefore, approval from the Medical Ethics Committee was not required. All analytical procedures strictly adhered to relevant policies regarding data usage and distribution. Clinical Trial Number not applicable. Consent to Publish not applicable. References Weller, M., et al., EANO guidelines on the diagnosis and treatment of diffuse gliomas of adulthood. Nat Rev Clin Oncol, 2021. 18 (3): p. 170-186. Hu, X., et al., Multigene signature for predicting prognosis of patients with 1p19q co-deletion diffuse glioma. Neuro Oncol, 2017. 19 (6): p. 786-795. Ostrom, Q.T., et al., CBTRUS Statistical Report: Primary brain and other central nervous system tumors diagnosed in the United States in 2010-2014. Neuro Oncol, 2017. 19 (suppl_5): p. v1-v88. Ren, J., et al., The Importance of M1-and M2-Polarized Macrophages in Glioma and as Potential Treatment Targets. Brain Sci, 2023. 13 (9). Mamdani, H., et al., Immunotherapy in Lung Cancer: Current Landscape and Future Directions. Front Immunol, 2022. 13 : p. 823618. Klobuch, S., et al., Tumour-infiltrating lymphocyte therapy for patients with advanced-stage melanoma. Nat Rev Clin Oncol, 2024. 21 (3): p. 173-184. Verma, S., et al., Immunotherapy and Radiation Therapy Sequencing in Breast Cancer: A Systematic Review. Int J Radiat Oncol Biol Phys, 2024. 118 (5): p. 1422-1434. Persson, M.L., et al., The intrinsic and microenvironmental features of diffuse midline glioma: Implications for the development of effective immunotherapeutic treatment strategies. Neuro Oncol, 2022. 24 (9): p. 1408-1422. Guo, X. and G. Wang, Advances in research on immune escape mechanism of glioma. CNS Neurosci Ther, 2023. 29 (7): p. 1709-1720. Yuan, Z., et al., Extracellular matrix remodeling in tumor progression and immune escape: from mechanisms to treatments. Mol Cancer, 2023. 22 (1): p. 48. McGranahan, N., et al., Allele-Specific HLA Loss and Immune Escape in Lung Cancer Evolution. Cell, 2017. 171 (6): p. 1259-1271.e11. Yamamoto, K., et al., Autophagy promotes immune evasion of pancreatic cancer by degrading MHC-I. Nature, 2020. 581 (7806): p. 100-105. Li, J., et al., Tumor-Associated Macrophage-Derived Exosomal LINC01232 Induces the Immune Escape in Glioma by Decreasing Surface MHC-I Expression. Adv Sci (Weinh), 2023. 10 (17): p. e2207067. Iwata, R., et al., ICOSLG-mediated regulatory T-cell expansion and IL-10 production promote progression of glioblastoma. Neuro Oncol, 2020. 22 (3): p. 333-344. Ding, X.C., et al., The relationship between expression of PD-L1 and HIF-1α in glioma cells under hypoxia. J Hematol Oncol, 2021. 14 (1): p. 92. Lawson, K.A., et al., Functional genomic landscape of cancer-intrinsic evasion of killing by T cells. Nature, 2020. 586 (7827): p. 120-126. Wang, Q., et al., Distinct molecular subtypes of systemic sclerosis and gene signature with diagnostic capability. Front Immunol, 2023. 14 : p. 1257802. Zhou, Q., et al., Machine learning-based identification of a cell death-related signature associated with prognosis and immune infiltration in glioma. J Cell Mol Med, 2024. 28 (11): p. e18463. Wang, S., et al., Immune cell infiltration-based signature for prognosis and immunogenomic analysis in breast cancer. Brief Bioinform, 2021. 22 (2): p. 2020-2031. Li, G., et al., An MRI radiomics approach to predict survival and tumour-infiltrating macrophages in gliomas. Brain, 2022. 145 (3): p. 1151-1161. Zhao, Z., et al., Chinese Glioma Genome Atlas (CGGA): A Comprehensive Resource with Functional Genomic Data from Chinese Glioma Patients. Genomics Proteomics Bioinformatics, 2021. 19 (1): p. 1-12. Yu, K., et al., Surveying brain tumor heterogeneity by single-cell RNA-sequencing of multi-sector biopsies. Natl Sci Rev, 2020. 7 (8): p. 1306-1318. Li, P.H., et al., Recent developments in application of single-cell RNA sequencing in the tumour immune microenvironment and cancer therapy. Mil Med Res, 2022. 9 (1): p. 52. Zhou, T., et al., Inflammatory markers and the risk of idiopathic sudden sensorineural hearing loss: A Mendelian randomization study. Front Neurol, 2023. 14 : p. 1111255. Chen, L., et al., Genetic Insights into Obesity and Brain: Combine Mendelian Randomization Study and Gene Expression Analysis. Brain Sci, 2023. 13 (6). Jiang, Y., et al., Single-cell RNA sequencing highlights intratumor heterogeneity and intercellular network featured in adamantinomatous craniopharyngioma. Sci Adv, 2023. 9 (15): p. eadc8933. Quail, D.F. and J.A. Joyce, The Microenvironmental Landscape of Brain Tumors. Cancer Cell, 2017. 31 (3): p. 326-341. Hu, L.S., et al., Integrated molecular and multiparametric MRI mapping of high-grade glioma identifies regional biologic signatures. Nat Commun, 2023. 14 (1): p. 6066. Regmi, M., et al., From glioma gloom to immune bloom: unveiling novel immunotherapeutic paradigms-a review. J Exp Clin Cancer Res, 2024. 43 (1): p. 47. Zhang, I.Y., et al., RAGE ablation attenuates glioma progression and enhances tumor immune responses by suppressing galectin-3 expression. Neuro Oncol, 2023. 25 (5): p. 886-898. Sharma, P., et al., Hyperglycosylation of prosaposin in tumor dendritic cells drives immune escape. Science, 2024. 383 (6679): p. 190-200. Zhu, W., et al., OMA1 competitively binds to HSPA9 to promote mitophagy and activate the cGAS-STING pathway to mediate GBM immune escape. J Immunother Cancer, 2024. 12 (4). Zhang, P., et al., K63-linked ubiquitination of DYRK1A by TRAF2 alleviates Sprouty 2-mediated degradation of EGFR. Cell Death Dis, 2021. 12 (6): p. 608. Zou, Y., et al., The Single-Cell Landscape of Intratumoral Heterogeneity and The Immunosuppressive Microenvironment in Liver and Brain Metastases of Breast Cancer. Adv Sci (Weinh), 2023. 10 (5): p. e2203699. Takeuchi, Y., et al., Highly immunogenic cancer cells require activation of the WNT pathway for immunological escape. Sci Immunol, 2021. 6 (65): p. eabc6424. Tien, T.Y., et al., Hsa-miR-134-5p predicts cardiovascular risk in circulating mononuclear cells and improves angiogenic action of senescent endothelial progenitor cells. J Cell Mol Med, 2024. 28 (13): p. e18523. Wei, S., et al., GFAT1-linked TAB1 glutamylation sustains p38 MAPK activation and promotes lung cancer cell survival under glucose starvation. Cell Discov, 2022. 8 (1): p. 77. Xia, Q., et al., TRIM45 causes neuronal damage by aggravating microglia-mediated neuroinflammation upon cerebral ischemia and reperfusion injury. Exp Mol Med, 2022. 54 (2): p. 180-193. Kloesch, B., et al., A GATA6-centred gene regulatory network involving HNFs and ΔNp63 controls plasticity and immune escape in pancreatic cancer. Gut, 2022. 71 (4): p. 766-777. Wang, D.R., X.L. Wu, and Y.L. Sun, Therapeutic targets and biomarkers of tumor immunotherapy: response versus non-response. Signal Transduct Target Ther, 2022. 7 (1): p. 331. Kim, J.M. and D.S. Chen, Immune escape to PD-L1/PD-1 blockade: seven steps to success (or failure). Ann Oncol, 2016. 27 (8): p. 1492-504. Liu, X., et al., FOXP3(+) regulatory T cell perturbation mediated by the IFNγ-STAT1-IFITM3 feedback loop is essential for anti-tumor immunity. Nat Commun, 2024. 15 (1): p. 122. Wu, Y., et al., Myeloid-derived suppressor cells: an emerging target for anticancer immunotherapy. Mol Cancer, 2022. 21 (1): p. 184. Additional Declarations No competing interests reported. Supplementary Files legendsinsupplementarymaterial.docx SupplementaryTableS1.xlsx SupplementaryTableS2.xlsx SupplementaryTableS3.xlsx SupplementaryTableS4.xlsx FigureS1.tif FigureS2.tif Cite Share Download PDF Status: Under Review Version 1 posted Editorial decision: Revision requested 23 May, 2025 Reviews received at journal 19 May, 2025 Reviewers agreed at journal 18 May, 2025 Reviews received at journal 18 May, 2025 Reviewers agreed at journal 17 May, 2025 Reviews received at journal 16 May, 2025 Reviewers agreed at journal 16 May, 2025 Reviewers agreed at journal 08 May, 2025 Reviewers invited by journal 07 May, 2025 Editor assigned by journal 28 Apr, 2025 Submission checks completed at journal 22 Apr, 2025 First submitted to journal 22 Apr, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6330200","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":454694588,"identity":"e775a2d4-a431-430b-b004-26484041cda3","order_by":0,"name":"Chongkang Ren","email":"","orcid":"","institution":"The First Bethune Hospital of Jilin University","correspondingAuthor":false,"prefix":"","firstName":"Chongkang","middleName":"","lastName":"Ren","suffix":""},{"id":454694589,"identity":"7eff3694-ed53-4932-8cc3-11c5f672a85f","order_by":1,"name":"Jinyi Cai","email":"","orcid":"","institution":"Jining Medical University","correspondingAuthor":false,"prefix":"","firstName":"Jinyi","middleName":"","lastName":"Cai","suffix":""},{"id":454694590,"identity":"0a056e08-431c-45d3-9122-2dcdf8ba5037","order_by":2,"name":"Yanhua Liu","email":"","orcid":"","institution":"The First Bethune Hospital of Jilin University","correspondingAuthor":false,"prefix":"","firstName":"Yanhua","middleName":"","lastName":"Liu","suffix":""},{"id":454694591,"identity":"16132a85-f00e-4f47-9c7d-494be24902ac","order_by":3,"name":"Zeshang Guo","email":"","orcid":"","institution":"The First Bethune Hospital of Jilin University","correspondingAuthor":false,"prefix":"","firstName":"Zeshang","middleName":"","lastName":"Guo","suffix":""},{"id":454694592,"identity":"1f69cea0-558d-428b-8d3c-49e021eec186","order_by":4,"name":"Xinyu Hong","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAw0lEQVRIiWNgGAWjYBACAwhlw8wHonhI0JLGzEaqlsMMxGsxZ+89/Jq37Tw7m0QC44O3bQzy5oS0WPacS7Oc2XabGaiF2XBuG4PhzgZCDruRY2bwEaKFTZq3jSHB4AAxWhLbzoG0sP8mVovxg49tB8C2MBOn5cwZM8YZ55KZ2XgeNkvOOSdhuIGgluM9xp95yuyS+dmTD354U2YjT9AWIGCTYGRjSGZgYGwAciQIqwcC5g8MfxjsiFI6CkbBKBgFIxMAABKwOfHbhHXOAAAAAElFTkSuQmCC","orcid":"","institution":"The First Bethune Hospital of Jilin University","correspondingAuthor":true,"prefix":"","firstName":"Xinyu","middleName":"","lastName":"Hong","suffix":""}],"badges":[],"createdAt":"2025-03-28 18:08:04","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6330200/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6330200/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":82607320,"identity":"86a9d326-825b-4311-9af4-7280ca992b66","added_by":"auto","created_at":"2025-05-13 10:14:20","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":10313688,"visible":true,"origin":"","legend":"\u003cp\u003eDifferential gene expression and survival analysis related to immune escape. (A,B) The heat map and volcano plot illustrate the expression of immune escape-related genes in tumor and normal samples. Red indicates high expression, blue indicates low expression, and black dots indicate insignificant genes. Data are based on Wilcoxon rank sum test. (C) The forest plot showing the HRs and confidence intervals of significantly differentially expressed immune escape genes in survival analysis, calculated based on the Cox regression model.\u003c/p\u003e","description":"","filename":"Figure1.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/7262f2dcfc1fbb5bd3868b02.png"},{"id":82606308,"identity":"3e811df4-8c61-49af-a574-ad118f644a3f","added_by":"auto","created_at":"2025-05-13 10:06:20","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":43499020,"visible":true,"origin":"","legend":"\u003cp\u003eConstruction of immune escape-related prognostic model for HGG patients. (A) Heat map showing the performance index (C-index) of 101 machine learning model combinations in immune escape gene prognosis prediction based on the CGGA and TCGA cohorts. Darker colors represent better performance. (B) The best model trained on the CGGA dataset for patient risk stratification, with heatmap showing differences between high-risk and low-risk groups. (C) Patient risk stratification of the best model validated on the TCGA dataset, and the heat map shows the difference between high-risk and low-risk groups. (D,E) Kaplan-Meier survival curves show the survival difference between high-risk and low-risk groups on the CGGA and TCGA datasets, demonstrating the predictive effect of the model. (F,G) ROC curves show the predictive performance of the model at 1, 3, and 5 years on the CGGA and TCGA datasets.\u003c/p\u003e","description":"","filename":"Figure2.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/d26b63c44de486dd7db457ce.png"},{"id":82607857,"identity":"308b9203-2de1-4975-8918-1e206cfdfebe","added_by":"auto","created_at":"2025-05-13 10:22:20","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":16280240,"visible":true,"origin":"","legend":"\u003cp\u003eConstruction and evaluation of an immune escape-related nomogram for survival prediction in HGG patients. (A,B) Forest plots show the results of univariate and multivariate Cox regression analysis of the risk score and clinical characteristics. The HR and its confidence interval for each covariate are shown. (C) ROC curve of the risk score and clinical characteristics compared their prognostic abilities. (D) C-index analysis compared the contribution of different clinical characteristics to survival prediction. (E) The nomogram predicts 1-year, 3-year, and 5-year overall survival probabilities, combining the risk score and clinical characteristics. (F) Calibration curve were used to evaluate the accuracy of the nomogram in predicting prognosis. (G) Decision curve analysis was used to evaluate the clinical application value of the model under different thresholds.\u003c/p\u003e","description":"","filename":"Figure3.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/474c8fe1d62105ad46d0f37d.png"},{"id":82606309,"identity":"d0088fa3-db09-46d5-9311-0aabdcad8fde","added_by":"auto","created_at":"2025-05-13 10:06:20","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":36901364,"visible":true,"origin":"","legend":"\u003cp\u003eFunction and pathway enrichment analysis of immune escape-related genes. (A) Circle plot showing correlations among immune escape genes in the model. Red represents a positive correlation and green represents a negative correlation. (B) Network analysis based on the GeneMANIA, showing the functional interactions and pathway connections of immune escape genes. (C,D) Gene set enrichment analysis demonstrating significantly enriched biological processes in high- and low-risk groups. (E,F) KEGG analysis showing differences among different risk groups.\u003c/p\u003e","description":"","filename":"Figure4.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/25750bfe41366a0339334655.png"},{"id":82606279,"identity":"bf4e904c-205c-4b79-9c4e-fd20c8ad0d52","added_by":"auto","created_at":"2025-05-13 10:06:18","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":36820687,"visible":true,"origin":"","legend":"\u003cp\u003eCorrelation between immune cell infiltration and immune escape-related genes. (A) Bubble plot showing the results of correlation analysis based on immune cell infiltration in 7 immune databases by Spearman correlation analysis. (B) The network diagram shows the significant interactions between immune cells and 18 model genes. The color and thickness of the connecting lines indicate significance and correlation strength.\u003c/p\u003e","description":"","filename":"Figure5.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/3bfb9d16f30d3e10aef734b2.png"},{"id":82607318,"identity":"fc96b09d-d870-4f7f-a323-f2e7ae7bb60e","added_by":"auto","created_at":"2025-05-13 10:14:20","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":6532645,"visible":true,"origin":"","legend":"\u003cp\u003eAnalysis of tumor microenvironment score. (A) Violin plot showing stromal score, immune score, and ESTIMATE score for different risk groups, with higher scores representing higher infiltration or stromal content. (B) Comparison of immune subtype proportions showing significant differences between high and low risk groups, statistically analyzed using the chi-square test. (C) Boxplot showing differences in immune function between high- and low-risk groups. (D) Differences in immune checkpoint expression among different risk groups are shown by the boxplot and evaluated using the Wilcoxon rank sum test.\u003c/p\u003e","description":"","filename":"Figure6.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/ef4966bdd1bc8225e4c24137.png"},{"id":82606310,"identity":"018a3a9c-ed00-48b2-a4d6-5bf81108ae3a","added_by":"auto","created_at":"2025-05-13 10:06:20","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":12491545,"visible":true,"origin":"","legend":"\u003cp\u003eTumor mutation burden (TMB) and immunotherapy response. (A) Scatter plot shows the correlation between the risk score and TMB, and the fitted line shows the trend relationship, based on the Spearman correlation analysis. (B) Boxplot comparing TMB distribution between high and low risk groups, with significant difference assessed using the Wilcoxon rank sum test. (C) Kaplan-Meier survival curve demonstrating survival difference among different groups. (D) Violin plot showing difference in TIDE scores across different risk subgroups. (E) Boxplot based on the IMvigor210 cohort shows the differences in risk score between different treatment response categories (CR, PR, PD, SD). (F) Stacked bar chart showing proportional differences in treatment response types among patients in different risk groups. (G) Kaplan-Meier survival curve shows the survival difference between high-risk and low-risk groups in the IMvigor210 cohort. (H) ROC curve is used to evaluate the performance of the model in predicting immunotherapy response.\u003c/p\u003e","description":"","filename":"Figure7.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/4ae9f42e002c0955922cd1b7.png"},{"id":82607321,"identity":"2d5a33ef-216f-4fe8-b619-cc96f12e657d","added_by":"auto","created_at":"2025-05-13 10:14:20","extension":"png","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":22220858,"visible":true,"origin":"","legend":"\u003cp\u003eSingle-cell RNA sequencing analysis. (A) The UMAP plot provides annotations and color coding of different cell types in tumors and adjacent normal tissues, with data from the CGGA database. (B) Average expression levels of model genes in different cell types. (C) Expression of each model gene, highlighting its differential expression in specific cell types.\u003c/p\u003e","description":"","filename":"Figure8.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/38040ac17f68b17d651bcb37.png"},{"id":82607333,"identity":"4c7a8825-2744-426c-88b5-89824ff8ee1a","added_by":"auto","created_at":"2025-05-13 10:14:21","extension":"png","order_by":9,"title":"Figure 9","display":"","copyAsset":false,"role":"figure","size":7754315,"visible":true,"origin":"","legend":"\u003cp\u003eDrug sensitivity analysis. (A-T) The oncoPredict package is used to perform drug sensitivity analysis in patients with HGG. The top 20 drugs with the highest drug sensitivity between high-risk and low-risk groups are shown, evaluated using the Wilcoxon rank sum test.\u003c/p\u003e","description":"","filename":"Figure9.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/74b19664acf1875b3f622779.png"},{"id":82606318,"identity":"da413596-dd99-4431-b29e-f77a86f59a6c","added_by":"auto","created_at":"2025-05-13 10:06:20","extension":"png","order_by":10,"title":"Figure 10","display":"","copyAsset":false,"role":"figure","size":34954819,"visible":true,"origin":"","legend":"\u003cp\u003eMolecular docking analysis. (A-G) Molecular docking results of 7 immune escape-related proteins and the chemotherapy drug temozolomide are shown. Important hydrogen bonding and interacting residues are annotated, showing potential sites for drug binding to proteins.\u003c/p\u003e","description":"","filename":"Figure10.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/52b701444dc5a0e93d34a760.png"},{"id":82606293,"identity":"1f8e1906-72e2-4bdc-a9d8-91ae0ff7c1f1","added_by":"auto","created_at":"2025-05-13 10:06:20","extension":"png","order_by":11,"title":"Figure 11","display":"","copyAsset":false,"role":"figure","size":9152127,"visible":true,"origin":"","legend":"\u003cp\u003eMendelian randomization analysis. (A, C, E) The analysis results of the causal relationship between model genes TNFAIP3, TNFRSF1B and ERP44 and glioma are shown. The figure includes 5 Mendelian randomization methods (Inverse Variance Weighted, MR-Egger, Weighted median, Simple mode and Weighted mode). (B, D, F) Leave-one-out sensitivity analysis shows the effect on causality after removing each SNP, ensuring the robustness of the results. The red line indicates a significant overall effect and black lines indicate the effect of each SNP and its confidence interval.\u003c/p\u003e","description":"","filename":"Figure11.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/7397d2d3223aae8adeab6f1b.png"},{"id":82606278,"identity":"12923d0a-b927-454a-8de7-adc6f8ce59f5","added_by":"auto","created_at":"2025-05-13 10:06:18","extension":"png","order_by":12,"title":"Figure 12","display":"","copyAsset":false,"role":"figure","size":22865079,"visible":true,"origin":"","legend":"\u003cp\u003eBiological role of TAB1 in glioma. (A) Sectional image of differential expression of the TAB1 gene from the HPA database. (B) Statistical column stacked plot of characteristic protein expression of gene TAB1 from the HPA database. (C) Validation of the efficiency of TAB1 knock-down. (D) CCK-8 assay. (E, F) Colony formation assay.\u003c/p\u003e","description":"","filename":"Figure12.png","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/80b75bb7f7b6e9191e44e0e2.png"},{"id":82606268,"identity":"8e6c885a-3f54-4ffa-a755-9660e91b574b","added_by":"auto","created_at":"2025-05-13 10:06:16","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1080364,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/75043230-11a9-4490-a433-37a89242f37f.pdf"},{"id":82606290,"identity":"4d39d86f-3030-458c-b3c0-88502123c94f","added_by":"auto","created_at":"2025-05-13 10:06:19","extension":"docx","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":14954,"visible":true,"origin":"","legend":"","description":"","filename":"legendsinsupplementarymaterial.docx","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/6d24d0e66a4d380c760ab3f2.docx"},{"id":82606271,"identity":"1b2a0dfd-4f7a-4187-9fa0-84854459f9c0","added_by":"auto","created_at":"2025-05-13 10:06:16","extension":"xlsx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":12351,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS1.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/cc23140c9f7281bf5c840bda.xlsx"},{"id":82606295,"identity":"4a951de4-d5a8-4637-96e3-b9b5cdb0a1ec","added_by":"auto","created_at":"2025-05-13 10:06:20","extension":"xlsx","order_by":3,"title":"","display":"","copyAsset":false,"role":"supplement","size":13470,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS2.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/fd78562906aad2b902c9bf29.xlsx"},{"id":82606291,"identity":"c1460a2d-b749-4573-8a09-05253923f273","added_by":"auto","created_at":"2025-05-13 10:06:20","extension":"xlsx","order_by":4,"title":"","display":"","copyAsset":false,"role":"supplement","size":9888,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS3.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/647a060efe8b6d304c46c1cd.xlsx"},{"id":82606284,"identity":"a9fae0a0-f122-4390-9f7b-e9989c4e0f19","added_by":"auto","created_at":"2025-05-13 10:06:18","extension":"xlsx","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":9892,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryTableS4.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/2744481e528cda7bee1fa3f1.xlsx"},{"id":82606274,"identity":"b8b7606b-f6f1-4282-92d5-7dee2fe4c128","added_by":"auto","created_at":"2025-05-13 10:06:17","extension":"tif","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":12328778,"visible":true,"origin":"","legend":"","description":"","filename":"FigureS1.tif","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/17c76f54318a5e70bb13db5c.tif"},{"id":82606364,"identity":"2c9f9c14-c651-4689-9d51-09eb1519af3c","added_by":"auto","created_at":"2025-05-13 10:06:26","extension":"tif","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":144344036,"visible":true,"origin":"","legend":"","description":"","filename":"FigureS2.tif","url":"https://assets-eu.researchsquare.com/files/rs-6330200/v1/3bfe400c539f61ea891a88a2.tif"}],"financialInterests":"No competing interests reported.","formattedTitle":"Comprehensive analysis of immune escape-related prognostic signature in high-grade glioma","fulltext":[{"header":"Introduction","content":"\u003cp\u003eHigh-grade glioma (HGG, WHO grade III and IV) are common primary malignant brain tumors, particularly the higher-grade subtypes, such as glioblastoma multiforme (GBM) and anaplastic astrocytoma, both of which exhibit extremely high malignancy [\u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e1\u003c/span\u003e]. The heterogeneity of gliomas leads to significant differences in clinical manifestations, molecular characteristics, and treatment responses among patients. Although histological classification provides a foundation for clinical treatment, patients with similar molecular features may exhibit vastly different prognoses [\u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2\u003c/span\u003e]. Among them, GBM is the most lethal subtype of gliomas, with a particularly poor prognosis and a five-year survival rate of 5.5% [\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e3\u003c/span\u003e]. Despite the fact that treatment for HGG currently relies primarily on surgical resection, radiation therapy, and chemotherapy, the therapeutic outcomes remain limited, and most patients experience recurrence after surgery [\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e4\u003c/span\u003e]. Recent studies have shown that immunotherapy offers new therapeutic avenues for various malignant tumors, particularly with immune checkpoint inhibitors (ICIs) demonstrating significant efficacy in lung cancer, melanoma, and other tumors [\u003cspan additionalcitationids=\"CR6\" citationid=\"CR5\" class=\"CitationRef\"\u003e5\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e7\u003c/span\u003e]. However, despite the improved efficacy of ICIs in other types of cancer, immunotherapy responses in HGG remain suboptimal [\u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e8\u003c/span\u003e]. This phenomenon is closely related to the immune escape mechanisms present within the tumor microenvironment [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e]. Therefore, there is an urgent need to develop novel therapeutic strategies and evaluation models to enhance the immunotherapy outcomes in HGG patients.\u003c/p\u003e \u003cp\u003eImmune escape refers to the ability of tumor cells to evade host immune surveillance through various mechanisms, thereby promoting tumor growth, progression, and metastasis [\u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e10\u003c/span\u003e]. In recent years, immune escape mechanisms have become a focal point in cancer research. Tumor cells regulate immune responses through immune checkpoint molecules (e.g., PD-1/PD-L1, CTLA-4) or alter the function of immune cells in the tumor microenvironment, enabling them to evade immune surveillance [\u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e11\u003c/span\u003e, \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e12\u003c/span\u003e]. In gliomas, immune escape mechanisms are particularly complex [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. The unique immunosuppressive microenvironment of GBM, which includes a high density of regulatory T cells (Tregs), M2 macrophages, and cancer-associated fibroblasts (CAFs), provides an ideal setting for immune escape [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e]. Additionally, glioma cells further suppress immune system function by secreting immunosuppressive factors, such as TGF-β and IL-10, and expressing immune checkpoint molecules, such as PD-L1, thereby exacerbating immune escape [\u003cspan additionalcitationids=\"CR14\" citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e15\u003c/span\u003e]. Although there has been some progress in the study of immune escape mechanisms in glioblastoma, no novel prognostic models targeting immune escape-related genes have been reported to date. Therefore, identifying immune escape-related genes with prognostic significance in HGG and exploring their interactions with the immune microenvironment has become a key research focus in this field.\u003c/p\u003e \u003cp\u003eThis study aims to construct a prognostic model based on the combination of 101 machine learning algorithms to identify immune escape-related genes, and further explore the role of these genes in the immune microenvironment of HGG and their relationship with patient responses to immunotherapy. By constructing and validating this model, we aim not only to provide more accurate prognostic assessments for HGG patients but also to offer new perspectives for the development of personalized immunotherapy strategies. Furthermore, we will further validate the efficacy of key genes in the model using techniques such as single-cell analysis, molecular docking, and the Mendelian randomization (MR). Through these comprehensive analyses, we aim to enhance the optimization of HGG immunotherapy, identify novel immune escape therapeutic targets, and provide more precise strategies for clinical treatment.\u003c/p\u003e"},{"header":"Materials and methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eData collection and processing\u003c/h2\u003e \u003cp\u003eThe data for this study were sourced from several public databases, including the Chinese Glioma Genome Atlas (CGGA, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://www.cgga.org.cn/\u003c/span\u003e\u003cspan address=\"http://www.cgga.org.cn/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and The Cancer Genome Atlas (TCGA, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://portal.gdc.cancer.gov/\u003c/span\u003e\u003cspan address=\"https://portal.gdc.cancer.gov/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e), to obtain gene expression and clinical data from high-grade glioma patients. Additionally, normal brain transcriptomic data were downloaded from the Adult Genotype Tissue Expression (GTEx, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://xenabrowser.net/\u003c/span\u003e\u003cspan address=\"https://xenabrowser.net/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). During data collection, samples with complete clinical and gene expression data were selected, while groups with insufficient sample sizes or significant missing data were excluded, ensuring data integrity and accuracy. A total of 182 immune escape genes were downloaded from previous studies (Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e) [\u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e16\u003c/span\u003e]. In data preprocessing, all gene expression data were first standardized to eliminate measurement scale differences between samples, ensuring comparability. Additionally, to correct for batch effects between datasets, the ComBat algorithm was applied [\u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e17\u003c/span\u003e]. The flowchart of this study was shown in Figure \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eIdentification and differential analysis of immune escape-related genes\u003c/h3\u003e\n\u003cp\u003eDifferential expression analysis used the R package \u0026ldquo;limma\u0026rdquo; to evaluate the differences in expression of immune escape-related genes between the tumor group and the normal group. The Wilcoxon test was used to calculate the log\u003csub\u003e2\u003c/sub\u003e(Fold Change) and the false discovery rate (FDR) for the differentially expressed genes, followed by multiple hypothesis testing correction. A threshold for differential expression analysis was set at |log\u003csub\u003e2\u003c/sub\u003eFC| \u0026gt; 0.585 and FDR\u0026thinsp;\u0026lt;\u0026thinsp;0.05, which identified significantly differentially expressed genes for subsequent analysis.\u003c/p\u003e\n\u003ch3\u003eConstruction and validation of immune escape-related machine learning models\u003c/h3\u003e\n\u003cp\u003eWe obtained 722 HGG samples and 20 normal control samples from the CGGA database. First, univariate Cox regression analysis was performed to identify differentially expressed immune escape genes significantly associated with the prognosis of high-grade glioma patients. Using p value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 as the significance criterion, the hazard ratio (HR) and its confidence interval of each gene were displayed to further evaluate its impact on patient survival. To construct a risk score model based on immune escape related genes, the machine learning algorithms used in this study included CoxBoost, stepwise Cox, survival support vector machine (survival-SVM), elastic network (Enet), least absolute shrinkage and selection operator (Lasso), partial least squares regression for Cox (plsRCox), Ridge, random survival forest (RSF), generalized boosted regression modeling (GBM), and supervised principal components (SuperPC) respectively [\u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e18\u003c/span\u003e]. These algorithms can handle high-dimensional data and select features through appropriate algorithm combinations. Next, we performed multivariate Cox regression using the \u0026ldquo;survival\u0026rdquo; R package to establish an immune escape-related prognostic model based on the following formula: Immune Escape Index (IEI) = Σ(Ci * Ei), where Ci represents the coefficient calculated for each signature gene in the multivariate Cox regression analysis, and Ei denotes the expression value of each signature gene. This risk score allows further prognostic stratification of patients. For model validation, the CGGA dataset was used as the training set and the TCGA dataset as the validation set. The model's performance was assessed using the concordance index (C-index), a commonly used metric in survival analysis that measures the correlation between the predicted risk scores and actual survival times. Additionally, the model's predictive ability was further validated by plotting the receiver operating characteristic (ROC) curve and calculating the area under the curve (AUC).\u003c/p\u003e\n\u003ch3\u003eRisk stratification and survival analysis\u003c/h3\u003e\n\u003cp\u003eBased on the immune escape-related index calculated using the machine learning model, patients were classified into high-IEI and low-IEI groups. Kaplan-Meier survival curves were used to assess the survival differences between the risk groups, and the statistical significance of the survival differences between the two groups was determined using the Log-rank test. The \"pheatmap\" package was used to visualize the distribution of risk values between the two groups. The \"timeROC\" package was used to generate ROC curves to assess the predictive accuracy for 1-year, 3-year, and 5-year overall survival (OS) in HGG. The predictive accuracy was determined based on the AUC values.\u003c/p\u003e\n\u003ch3\u003eConstruction and validation of a nomogram in HGG patients\u003c/h3\u003e\n\u003cp\u003eWe conducted both univariate and multivariate Cox regression analyses to further assess whether the immune escape-related risk score could serve as an independent prognostic indicator. Furthermore, we developed a nomogram for HGG patients, which visualizes the combined impact of various clinical features and the risk score on prognosis. The predictive accuracy of the nomogram was evaluated using a calibration curve, which illustrates the relationship between the predicted survival probabilities and the actual observed values. To validate the clinical applicability of the model, we also employed the C-index, ROC curve, and decision curve analysis (DCA) to further evaluate the model's clinical net benefit and predictive accuracy.\u003c/p\u003e \u003cdiv id=\"Sec8\" class=\"Section2\"\u003e \u003ch2\u003eImmune escape-related gene set enrichment analysis\u003c/h2\u003e \u003cp\u003eThis study employed the GeneMANIA tool (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://genemania.org/\u003c/span\u003e\u003cspan address=\"http://genemania.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to analyze the association network of 18 immune escape-related model genes, integrating and visualizing their relationships. Subsequently, to explore the biological roles of the immune escape genes in HGG, we performed the Gene Set Enrichment Analysis (GSEA) to analyze the significantly enriched pathways in both the high-IEI and low-IEI groups. The GSEA enabled the identification of potential biological pathways related to immune escape, providing key insights for subsequent mechanistic studies.\u003c/p\u003e \u003c/div\u003e\n\u003ch3\u003eAnalysis of tumor microenvironment\u003c/h3\u003e\n\u003cp\u003eWe calculated differences in tumor microenvironment scores between different subgroups using the ESTIMATE algorithm, including the Stromal Score, Immune Score, and ESTIMATE Score. This analysis assessed the composition of the HGG tumor microenvironment to better understand tumor behavior and design targeted therapeutic strategies. Subsequently, immune microenvironment analysis was performed using the CIBERSORT algorithm based on 7 different immune databases (QUANTISEQ, XCELL, EPIC, TIMER, MCPCOUNTER, CIBERSORT, and CIBERSORT-ABS) to evaluate immune cell infiltration in HGG. The spearman correlation analysis was then used to further investigate the relationship between immune escape genes and immune cell infiltration, identifying immune cell types significantly correlated with immune escape genes. Additionally, the relationship between each immune escape-related model gene and the immune microenvironment was visualized to help elucidate the potential roles of immune escape genes in the tumor immune microenvironment.\u003c/p\u003e\n\u003ch3\u003eAssessment and prediction of immunotherapy\u003c/h3\u003e\n\u003cp\u003eWe evaluated the differential expression of immune checkpoint-related molecules across different subgroups. The varying expression of immune checkpoints in this prognostic risk model may offer new therapeutic insights and predict the clinical effectiveness of corresponding ICIs. Additionally, to explore the differences in tumor mutational burden (TMB, mutations per million base pairs) between the two risk subgroups, we calculated the TMB for each HGG patient. Subsequently, the immune treatment cohort was downloaded from the Tumor Immune Dysfunction and Exclusion ( TIDE, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://tide.dfci.harvard.edu/\u003c/span\u003e\u003cspan address=\"http://tide.dfci.harvard.edu/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e). The TIDE scoring was used to assess the potential impact of immune escape genes on the effectiveness of immune therapy and evaluate the sensitivity differences to immunotherapy between the high-IEI and low-IEI groups. The response of this prognostic signature to immunotherapy was then predicted based on the IMvigor210 cohort. The IMvigor210 cohort was used to predict patients' responses to immunotherapy [\u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e19\u003c/span\u003e]. Under the Creative Commons 3.0 license, complete expression and clinical data were downloaded from \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://research-pub.Gene.com/IMvigor210CoreBiologies\u003c/span\u003e\u003cspan address=\"http://research-pub.Gene.com/IMvigor210CoreBiologies\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. The raw data were then normalized, and the counts were converted into TPM values.\u003c/p\u003e \u003cdiv id=\"Sec11\" class=\"Section2\"\u003e \u003ch2\u003eSingle-cell analysis\u003c/h2\u003e \u003cp\u003eThe single-cell sequencing data of GBM, comprising 6,148 cells from 73 regions of 14 patients, were downloaded from the CGGA database and related previous literature [\u003cspan additionalcitationids=\"CR21\" citationid=\"CR20\" class=\"CitationRef\"\u003e20\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e22\u003c/span\u003e]. Based on this, the distribution of immune escape-related model genes was explored. The R package \"Seurat\" was used to process the single-cell expression matrix, followed by further processing using the FindVariableFeatures, ScaleData function, principal component analysis (PCA), and FindNeighbors [\u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e23\u003c/span\u003e]. Finally, the UMAP was employed for visualization.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec12\" class=\"Section2\"\u003e \u003ch2\u003eMolecular docking and drug sensitivity\u003c/h2\u003e \u003cp\u003eThe molecular docking analysis was performed using the CB-Dock2 tool (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://cadd.labshare.cn/cb-dock2/index.php\u003c/span\u003e\u003cspan address=\"https://cadd.labshare.cn/cb-dock2/index.php\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) to investigate the molecular interactions between immune escape-related model genes and the chemotherapy drug Temozolomide. Molecular dynamics simulations were employed to validate the stability of the molecular docking complex. Additionally, drug sensitivity analysis was performed using the \"oncoPredict\" package to predict drug responses in the high-IEI and low-IEI groups, with boxplots used to visualize the sensitivity differences between the groups.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec13\" class=\"Section2\"\u003e \u003ch2\u003eMR analysis\u003c/h2\u003e \u003cp\u003eTwo-sample MR was conducted using the \"TwoSampleMR\" package to assess the causal relationship between immune escape-related model genes and glioma risk. Significant single nucleotide polymorphisms (SNPs) from the Integrative Epidemiology Unit (IEU, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://gwas.mrcieu.ac.uk\u003c/span\u003e\u003cspan address=\"https://gwas.mrcieu.ac.uk\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) and the FinnGen (\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.finngen.fi/en\u003c/span\u003e\u003cspan address=\"https://www.finngen.fi/en\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) datasets were selected as instrumental variables to ensure their relevance and independence. Sensitivity analysis was performed using leave-one-out analysis and MR-PRESSO to assess the robustness and pleiotropy of the instrumental variables [\u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e24\u003c/span\u003e]. In the pleiotropy test, a p-value greater than 0.05 indicates the absence of pleiotropic effects, meeting the criteria for further analysis. The SNPs with a p-value less than 5e-08 in the outcome data were selected as valid instrumental variables for MR analysis. Causal inference methods included IVW (Inverse Variance Weighted), MR-Egger, Weighted median, Simple mode, and Weighted mode [\u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e25\u003c/span\u003e]. Among these methods, if all SNPs included in this analysis meet the assumptions of valid instruments, the IVW method will accurately estimate the causal effect between exposure and outcome, and thus will be used as the primary method.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec14\" class=\"Section2\"\u003e \u003ch2\u003eCell culture and transfection\u003c/h2\u003e \u003cp\u003eTwo human glioma cell lines, U251 and SW1088, were obtained from Procell Life Science \u0026amp; Technology Co., Ltd. (Wuhan, China). U251 and SW1088 were cultured in Dulbecco\u0026rsquo;s modified Eagle\u0026rsquo;s Medium (DMEM) supplied with 10% fetal bovine serum (FBS) and 1% penicillin\u0026ndash;streptomycin. The cells were cultured at 37\u0026deg;C and 5% CO\u003csup\u003e2\u003c/sup\u003e. TAB1 inhibitors and its negative controls (NC) were produced by Sangon Biotech (Shanghai, China). Total RNA samples of the two cell lines were then extracted using TRIzol reagent and reverse transcribed using the SYBR Green Master Mix kit.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec15\" class=\"Section2\"\u003e \u003ch2\u003eCCK-8 assay\u003c/h2\u003e \u003cp\u003eCCK-8 kit was used to detect the proliferation ability of human glioma cells after TAB1 expression was downregulated. The cells were seeded into 96-well plates at a concentration of 5\u0026times;10\u003csup\u003e3\u003c/sup\u003e cells per well. At 0, 24, 48, and 72 hours after attachment, 10 \u0026micro;L CCK-8 solution was added to each well and incubated in the incubator for 2 hours. Finally, the absorbance of each group at 450 nm was measured using an enzyme labeling instrument.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec16\" class=\"Section2\"\u003e \u003ch2\u003eColony formation assay\u003c/h2\u003e \u003cp\u003eThe differences in colony-forming ability of human glioma cells in different transfection groups were detected. U251 and SW1088 cells were seeded in each well of a 6-well plate. After 14 days of culture, the cell colonies were fixed with 4% paraformaldehyde and then stained with 0.1% crystal violet.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec17\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eIn this study, all statistical analyses were conducted using the R software (version 4.4.1). The Wilcoxon rank sum test, log-rank test, and the one-way ANOVA were used to compare statistical differences. The cox regression analysis and the Kaplan-Meier curves were used to assess prognostic value. Correlation analysis was performed using the Pearson and the Spearman correlation methods. All statistical analyses were conducted with two-sided tests, and a significance level of p-value\u0026thinsp;\u0026lt;\u0026thinsp;0.05 was applied unless otherwise specified.\u003c/p\u003e \u003c/div\u003e"},{"header":"Results","content":"\u003cdiv id=\"Sec19\" class=\"Section2\"\u003e \u003ch2\u003eIdentification and survival analysis of differentially expressed genes related to immune escape\u003c/h2\u003e \u003cp\u003eGene expression data from HGG patients in the CGGA and TCGA databases, along with normal samples from the GTEx, were standardized, and batch effects were corrected using the ComBat algorithm to ensure data comparability and accuracy. Differential expression analysis of immune escape-related genes between tumor and normal groups in the CGGA cohort was performed using the \"limma\" package, identifying 41 significantly differentially expressed genes, including 20 upregulated and 21 downregulated genes. The heatmap (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eA) illustrates the expression patterns of these genes in the samples, while the volcano plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eB) further shows the distribution of upregulated and downregulated genes. The univariate Cox regression analysis revealed that the 41 differentially expressed immune escape-related genes were significantly associated with OS in HGG patients (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05), with HR and 95% confidence intervals shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003eC. These genes may serve as potential prognostic biomarkers and provide a foundation for subsequent model development.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec20\" class=\"Section2\"\u003e \u003ch2\u003eConstruction and validation of prognostic signature based on machine learning\u003c/h2\u003e \u003cp\u003eA total of 101 machine learning model combinations were employed to perform feature selection on the 41 prognosis-related genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eA). The best model trained on the CGGA cohort identified 18 key genes for constructing the immune escape-related prognostic model. The optimal model algorithm combination was CoxBoost and Enet (alpha\u0026thinsp;=\u0026thinsp;0.2). The multivariate Cox regression analysis was used to calculate the IMI for each patient. Afterward, HGG patients were classified into high-IEI and low-IEI groups based on the median IMI value. In the CGGA cohort, the distribution of risk scores and survival status are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eB. The survival rate of high-IEI patients was significantly lower than that of low-IEI patients (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eD). In the TCGA cohort, the model similarly stratified patients by the risk score (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eC), with the survival rate of high-IEI patients significantly lower than that of low-IEI patients (p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eE). The ROC curve showed that the signature achieved AUC values of 0.710, 0.777, and 0.813 for 1-year, 3-year, and 5-year survival in the CGGA cohort (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eF), and 0.800, 0.850, and 0.808 in the TCGA cohort (Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003eG), indicating good predictive performance.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec21\" class=\"Section2\"\u003e \u003ch2\u003eConstruction of a nomogram of HGG patients\u003c/h2\u003e \u003cp\u003eThe univariate and multivariate Cox regression analyses were conducted to assess whether the risk score is an independent prognostic factor (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eA and \u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eB). The results showed that, after adjusting for clinical features such as age, gender, and WHO grade, the risk score remained an independent prognostic factor for HGG patients (HR\u0026thinsp;=\u0026thinsp;2.14, 95% CI: 1.69\u0026ndash;2.73, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001). Furthermore, the prognostic ability of the risk score was superior to that of other clinical features, with the highest AUC value on the ROC curve (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eC). The C-index analysis further supported this conclusion (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eD). Based on the results of the multivariate Cox regression analysis, a nomogram was constructed, incorporating the risk score and key clinical features, to predict the 1-year, 3-year, and 5-year survival probabilities (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eE). The calibration curve showed that the predicted survival rates from the nomogram closely matched the actual observed values, indicating good predictive accuracy of the model (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eF). The DCA revealed that the nomogram model provided higher clinical net benefits across different threshold probabilities (Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003eG).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec22\" class=\"Section2\"\u003e \u003ch2\u003eGene set enrichment analysis\u003c/h2\u003e \u003cp\u003eThe interaction network of the 18 immune escape-related model genes was visualized using the Circos plot (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eA). The results revealed complex interrelationships between different genes. The gene association network generated by the GeneMANIA revealed the complex interactions of the 18 key genes within inflammation and immune signaling pathways (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eB). The results indicated that genes such as TNFRSF1A, TNFAIP3, and TRAF2 exhibited high connectivity in the network, suggesting their potential role as core regulatory factors in the TNF/NF-kappaB signaling pathway. The GSEA was conducted to gain deeper insights into the biological functions of the immune escape-related genes. The results showed that functional and pathway enrichment analyses were performed on the gene expression profiles of the high-IEI and low-IEI groups. In the GO biological process enrichment analysis, the high-IEI group was significantly enriched in immune-related biological processes, including adaptive immune response and lymphocyte-mediated immunity (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eC), suggesting more active roles in inflammation and immune regulation. In contrast, the low-IEI group was primarily enriched in neuro-related biological processes, such as postsynaptic membrane activity and cation channel complexes (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eD), indicating its characteristics are more aligned with homeostatic regulation and neural function. In the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis, genes in the high-IEI group were significantly enriched in pathways associated with tumor progression and immune signaling, including cell cycle and cytokine-cytokine receptor interactions (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eE), reflecting their potential higher proliferative capacity and inflammatory characteristics. In contrast, the low-IEI group was predominantly enriched in calcium signaling pathway and neuroactive ligand-receptor interaction (Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003eF), suggesting that its biological characteristics are more aligned with normal tissue function. Overall, the high-IEI group exhibited prominent features of immune activation and tumor proliferation, while the low-IEI group was more associated with neural signaling and tissue homeostasis, providing important insights into the biological significance of the risk stratification.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec23\" class=\"Section3\"\u003e \u003ch2\u003eAnalysis of immune cell infiltration\u003c/h2\u003e \u003cp\u003eImmune cell infiltration was analyzed using the CIBERSORT algorithm and 7 immune databases. The bubble plot shows a significant correlation between the expression of certain immune cell types and the prognostic model genes (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eA). The network plot illustrates significant interactions between immune cells and prognostic genes, suggesting that these genes may influence the remodeling of the immune microenvironment (Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003eB). The tumor microenvironment characteristics were assessed using the ESTIMATE algorithm, revealing that the stromal score, immune score, and ESTIMATE score of the high-IEI group were significantly higher than those of the low-IEI group (p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eA), indicating a more complex tumor microenvironment in the high-IEI group. The comparison of immune subtype proportions revealed that the high-IEI group was more likely to exhibit an immunosuppressive subtype (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eB). Further analysis was conducted on the relative proportions of 21 tumor-infiltrating immune cells in each HGG sample. The results showed that the enrichment of tumor-infiltrating immune cells varied across different risk groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eC).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec24\" class=\"Section2\"\u003e \u003ch2\u003eTumor mutation burden and prediction of immunotherapy response\u003c/h2\u003e \u003cp\u003eThis study examined the relationship between the risk score and TMB, finding a significant positive correlation (r\u0026thinsp;=\u0026thinsp;0.5, p\u0026thinsp;\u0026lt;\u0026thinsp;0.001, Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eA). The TMB of the high-IEI group was significantly higher than that of the low-IEI group (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eB). The Kaplan-Meier survival analysis showed that patients with high TMB levels had lower survival rates (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eC). Additionally, we investigated the differences in the immune checkpoint expression. In the high-IEI group, the expression of nearly all key immune checkpoint-related molecules was upregulated (Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003eD). The TIDE scores were used to assess the sensitivity of HGG patients in different risk groups to immune therapy, revealing that the high-IEI group had significantly higher TIDE scores than the low-IEI group (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eD), indicating a higher likelihood of immune evasion in the high-IEI group. In the IMvigor210 cohort, further stratification of patients based on immune therapy response type revealed distinct risk scores in the progressive disease, stable disease, and partial response groups (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eE). Compared to the stable disease and complete response groups, the progressive disease group had a higher proportion of high-risk scores (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eF). The results showed that patients in the low-risk group had a higher response rate to immune therapy and significantly better survival rates compared to the high-risk group (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eG). The ROC curve analysis further validated the model's efficacy in predicting immune therapy responses (Fig.\u0026nbsp;\u003cspan refid=\"Fig7\" class=\"InternalRef\"\u003e7\u003c/span\u003eH).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cdiv id=\"Sec25\" class=\"Section3\"\u003e \u003ch2\u003eSingle-cell RNA sequencing data analysis\u003c/h2\u003e \u003cp\u003eUsing the marker genes for each cell subtype from previous literature [\u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e26\u003c/span\u003e], the UMAP plot dimensionality reduction was applied to visualize the distribution of different cell types (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eA, Figure \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). The expression patterns of 18 model genes across different cell types were then revealed (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eB). The results showed that these 18 genes exhibited diverse cell type-specific expression patterns (Fig.\u0026nbsp;\u003cspan refid=\"Fig8\" class=\"InternalRef\"\u003e8\u003c/span\u003eC). Most genes, including TNFRSF1A, TNFRSF1B, IFNGR2, and IRF1, were highly expressed in T cells and myeloid cells, suggesting their potential role in immune regulation and inflammatory responses. Meanwhile, certain genes, such as HDAC1, NDUFA13, and KMT2A, showed significant expression in neurons, astrocytes, and oligodendrocytes, suggesting their multifunctional regulatory roles in the tumor microenvironment. Furthermore, the expression of AHSA1, EIF3H, ERP44, UBE2N, and VDAC2 was more widespread, showing significant expression across multiple cell types. These findings provide a foundation for further research into the functions of immune escape-related genes in different cell types.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec26\" class=\"Section3\"\u003e \u003ch2\u003eDrug sensitivity and molecular docking analysis\u003c/h2\u003e \u003cp\u003eUsing the \"oncoPredict\" package, we predicted the chemotherapy drug sensitivity of HGG patients in different risk groups, identifying 121 drugs with differential sensitivity (Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). The high-risk group was more sensitive to drugs such as Carmustine, Nelarabine, and MIRA-1. In contrast, the low-risk group showed higher sensitivity to drugs like Dabrafenib, LCL161, and CZC24832 (Fig.\u0026nbsp;\u003cspan refid=\"Fig9\" class=\"InternalRef\"\u003e9\u003c/span\u003e). These findings offer potential drug options for personalized treatment. Moreover, the molecular docking analysis further explored the interactions between 18 key model genes and Temozolomide. The results revealed that 7 gene proteins exhibited strong binding affinities with Temozolomide (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e), with binding sites and interacting residues shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig10\" class=\"InternalRef\"\u003e10\u003c/span\u003e. These results provide a theoretical foundation for developing new therapeutic strategies.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec27\" class=\"Section3\"\u003e \u003ch2\u003eMendelian randomization analysis to verify the causal relationships\u003c/h2\u003e \u003cp\u003eTo validate the causal relationship between the model genes and glioma risk, the MR analysis was performed. Our analysis confirmed that all SNPs are robust instrumental variables. The results indicate that the expression levels of TNFAIP3, TNFRSF1B, and ERP44 are significantly causally associated with glioma (IVW method: p\u0026thinsp;\u0026lt;\u0026thinsp;0.05, Figs.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eA, C, E). Moreover, the trends in the odds ratios derived from these methods were consistent, with an increase in exposure factor levels corresponding to a reduced risk of disease occurrence. Sensitivity analysis revealed no significant heterogeneity or pleiotropy, confirming the robustness of the results (Figs.\u0026nbsp;\u003cspan refid=\"Fig11\" class=\"InternalRef\"\u003e11\u003c/span\u003eB, D, F).\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003c/div\u003e \u003cdiv id=\"Sec28\" class=\"Section2\"\u003e \u003ch2\u003eFunctional verification of gene TAB1 in vivo experiments\u003c/h2\u003e \u003cp\u003eFirst, we explored the hub genes of the model genes. The top three genes according to the scores were TNFRSF1A, TNFAIP3, and TAB1 (Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e). Among them, only the gene TAB1 was highly expressed in HGG patients and was validated in the Human Protein Atlas (HPA, \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://www.proteinatlas.org/\u003c/span\u003e\u003cspan address=\"https://www.proteinatlas.org/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e) (Figs.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003eA and \u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003eB). However, the biological role of gene TAB1 in HGG has not been explored. Therefore, we selected gene TAB1 for further experimental validation. We used qRT-PCR to detect the expression level of TAB1 in two human glioma cell lines (U251, SW1088) treated with si-TAB1 to silence TAB1. Apparently, si-TAB1-1 and si-TAB1-2 significantly reduced the expression of TAB1 in those cell lines (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003eC). To evaluate the effect of TAB1 knockdown on cell proliferation, the results of CCK-8 assay showed that cell proliferation was significantly reduced after TAB1 knockdown compared with the control group (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003eD). In addition, colony formation assay results showed that TAB1 knockdown reduced clone survival (Fig.\u0026nbsp;\u003cspan refid=\"Fig12\" class=\"InternalRef\"\u003e12\u003c/span\u003eE). Therefore, these results indicate that TAB1 gene downregulation can effectively inhibit the growth and proliferation ability of U251 and SW1088 cells.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e"},{"header":"Discussion","content":"\u003cp\u003eThe microenvironment of HGG is typically considered to be highly immunosuppressive and heterogeneous, characteristics that significantly contribute to tumor progression and deterioration [\u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e27\u003c/span\u003e]. However, due to the high heterogeneity of HGG and the complexity of the immune microenvironment, developing effective models to predict patient prognosis remains a challenge [\u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e28\u003c/span\u003e]. The survival time of patients with HGG varies significantly, depending on tumor grade, the microenvironment, and individual patient differences, with median survival ranging from 14 to 18 months [\u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e]. Establishing prognostic models helps predict patient survival and treatment responses, thereby providing more precise evidence for clinical decision-making. In recent years, with growing attention to the importance of immune escape in the tumor microenvironment, unique molecular characteristics related to immune escape have gradually been recognized [\u003cspan additionalcitationids=\"CR30\" citationid=\"CR29\" class=\"CitationRef\"\u003e29\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e31\u003c/span\u003e]. The mechanisms of immune escape in glioma patients have increasingly become a focal point of research [\u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e13\u003c/span\u003e, \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e32\u003c/span\u003e]. This study is the first to incorporate immune escape genes into a prognostic signature for HGG, which expands the research ideas in this field and suggests that immune escape plays a certain role in the biological behavior of glioma.\u003c/p\u003e \u003cp\u003eIn this study, the prognostic model constructed using 101 machine learning combinations demonstrated good predictive performance in risk stratification of high-IEI glioma patients. The model, based on the 18 key immune escape genes, utilized a combination of CoxBoost and Enet algorithms for feature selection and successfully distinguished high-IEI and low-IEI patient groups in the validation set. HGG patients in the high-IEI group had significantly lower survival rates compared to the low-IEI group, indicating the critical role of immune escape genes in tumor progression and further validating their clinical value as potential prognostic biomarkers. Additionally, the ROC curve analysis and the construction of nomogram further confirmed the stability and predictive accuracy of the model. Moreover, the prognostic model built using machine learning revealed the significant impact of these key genes on tumor progression and patient survival. In particular, genes such as TNFRSF1A, TNFAIP3, and TRAF2, through modulation of the immune microenvironment, influencing immune cell infiltration and the expression of immune checkpoint molecules, ultimately promote tumor immune escape [\u003cspan additionalcitationids=\"CR34\" citationid=\"CR33\" class=\"CitationRef\"\u003e33\u003c/span\u003e\u0026ndash;\u003cspan citationid=\"CR35\" class=\"CitationRef\"\u003e35\u003c/span\u003e]. We then explored the hub genes of the model genes. It was found that the hub gene transforming growth factor beta-activated kinase 1-binding protein 1 (TAB1) with a high score was highly expressed in HGG patients. TAB1 is a key regulator of TAK1, which promotes its activation by interacting with TAK1, thereby regulating the MAPK and NF-κB signaling pathways [\u003cspan citationid=\"CR36\" class=\"CitationRef\"\u003e36\u003c/span\u003e]. TAB1 has important pathophysiological significance in various diseases. Studies have shown that TAB1 mediates GFAT1-dependent p38 MAPK signaling activation through S438 phosphorylation, promotes cell autophagy and survival under nutritional stress, and is associated with poor prognosis in patients with lung adenocarcinoma [\u003cspan citationid=\"CR37\" class=\"CitationRef\"\u003e37\u003c/span\u003e]. Furthermore, TAB1 promotes TAK1 activation by forming the TAB1-TAK1-TAB2 complex with TAB2 and TAK1, thereby driving pro-inflammatory signaling in microglia during ischemic injury [\u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e38\u003c/span\u003e]. However, the biological role of TAB1 in HGG has been poorly studied. Our study reported for the first time that TAB1 is a tumor promoter factor in glioma, and knockout of TAB1 significantly inhibited the proliferation and cloning of glioma cells. Further results indicated that the significant differences in immune-related pathways between high-IEI and low-IEI groups may be key factors in determining patient prognosis. Furthermore, the tumor microenvironment of high-IEI patients exhibited more complex immunosuppressive characteristics, which correlated with their poorer survival rates, suggesting that these genes could serve as important targets for future immunotherapy.\u003c/p\u003e \u003cp\u003eThe immune escape mechanisms in gliomas are unique. Compared to solid tumors such as pancreatic cancer and non-small cell lung cancer, the immune microenvironment in gliomas is more significantly immunosuppressive, particularly in high-IEI patients, where immune checkpoint molecules are notably upregulated [\u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e9\u003c/span\u003e, \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e39\u003c/span\u003e, \u003cspan citationid=\"CR40\" class=\"CitationRef\"\u003e40\u003c/span\u003e]. These molecules, through their synergistic interaction with immunosuppressive pathways, impair anti-tumor immune responses, thereby enhancing the tumor's immune escape ability [\u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e41\u003c/span\u003e]. By comparing the immune escape mechanisms across different tumor types, this study provides valuable insights into the unique immune characteristics of gliomas and offers guidance for future targeted research. Moreover, the prognostic signature based on the immune escape related genes can be used for precise risk stratification of HGG patients, providing a reliable reference for the development of personalized treatment plans. Due to their pronounced immunosuppressive characteristics, high-IEI patients may be more suitable for combination therapy with ICIs to overcome their immune escape mechanisms, whereas low-IEI patients may exhibit poor responses to monotherapy and require additional treatment strategies. Drug sensitivity analysis further revealed significant differences in chemotherapy responsiveness between high-IEI and low-IEI patients, providing potential evidence for personalized drug selection.\u003c/p\u003e \u003cp\u003eFurthermore, this study reveals the crucial role of immune escape-related genes in regulating immune cell infiltration and reshaping the tumor microenvironment. In the high-IEI group, immunosuppressive cells, such as regulatory T cells, significantly increase in the tumor microenvironment, while the proportion of effector immune cells, such as Th1 and Th2 cells, remains relatively low. The changes in the proportion of these immune cells reflect an enhanced immunosuppressive state in the tumor microenvironment and a weakened anti-tumor immune response[\u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e42\u003c/span\u003e, \u003cspan citationid=\"CR43\" class=\"CitationRef\"\u003e43\u003c/span\u003e]. Targeting the immunosuppressive components of the tumor microenvironment may be an effective strategy to enhance the efficacy of immunotherapy. Subsequently, this study explores the relationship between risk scores and TMB, revealing that TMB is significantly higher in the high-IEI group compared to the low-IEI group. However, despite the higher TMB in the high-IEI group, these patients still have lower survival rates, suggesting that their glioma cells may evade immune system clearance through other immune escape mechanisms. Additionally, drug sensitivity analysis and molecular docking in this study identified potential therapeutic targets and drug combination strategies. Molecular docking analysis shows that temozolomide strongly binds to several key model gene proteins, offering potential pathways to improve its therapeutic efficacy. Based on these findings, future research may focus on developing novel immunotherapy combination strategies targeting the immunosuppressive features of HGG patients, such as combining ICIs with specific targeted therapies, which may effectively improve patient prognosis.\u003c/p\u003e \u003cp\u003eThis study has achieved some good results based on the integrated application of multiple machine learning algorithms, but there are still some limitations. First, the study mainly relies on transcriptome data provided by public databases and lacks verification of real clinical samples. More clinical samples need to be included in the future to improve the clinical applicability of the model. Secondly, this study only focuses on the predictive prognostic role of immune escape-related genes at the transcriptome level. It does not deeply explore its regulatory mechanisms at multiple molecular levels such as epigenetics and proteomics. In addition, although we have identified multiple key genes and preliminarily explored the biological role of TAB1 in HGG, the specific molecular mechanisms of these genes still need more in-depth experimental verification. For example, although this study found that TAB1 is closely related to tumor cell proliferation. However, its specific role in the immune microenvironment of glioma has not yet been elucidated, and it needs to be confirmed by more in-depth in vivo animal studies in the future. The prognostic model of immune escape-related genes proposed in this study performed well in predicting the risk stratification of HGG patients. However, it still needs in-depth optimization at multiple levels to promote its real application in clinical practice.\u003c/p\u003e \u003cp\u003eIn summary, this study established a prognostic model for HGG patients based on immune escape-related genes, suggesting that immune escape may play an important role in the biological behavior of glioma. This model showed good predictive ability in risk stratification, providing a potential reference for personalized treatment decisions, especially for the precise application of immunotherapy. However, these findings still need to be further verified by follow-up studies. In the future, the specific mechanism of glioma immune escape should be further explored through multi-omics data integration and in vivo and in vitro experiments, in order to provide more reliable theoretical support for the formulation of personalized treatment plans and improve the clinical efficacy of HGG patients.\u003c/p\u003e"},{"header":"Conclusion","content":"\u003cp\u003eThis study constructed the prognostic signature based on immune escape-related genes, revealing the prognostic value of immune escape-related genes in HGG patients. This signature provides novel ideas for personalized treatment decisions, especially in the precise application of immunotherapy. The study provides direction for developing personalized treatment strategies targeting immune escape mechanisms.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgment\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors are grateful to all patients who provided samples to the public databases.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConflict of Interest\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors declare that they have no competing interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor Contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eXYH and ZSG conceived and planned this study. CKR, JYC, and YHL helped interpret the results. CKR took the lead in writing the manuscript. All authors provided vital feedback to shape the final version of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe present study was supported by the Jilin Provincial Medical and Health Talent Project (grant no. JLSWSRCZX2023-24).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAll data analyzed in the present study are publicly available in The Cancer Genome Atlas (TCGA, https://portal.gdc.cancer.gov/), Chinese Glioma Genome Atlas (CGGA, http://www.cgga.org.cn/), Adult Genotype Tissue Expression (GTEx, https://xenabrowser.net/), Integrative Epidemiology Unit (IEU, https://gwas.mrcieu.ac.uk), the FinnGen (https://www.finngen.fi/en) and the Human Protein Atlas (HPA, https://www.proteinatlas.org/). The raw data and code for this study are reviewed via this link (https://www.jianguoyun.com/p/DSULx4AQ38nsChjHj9EEIAA).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eInformed consent\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003enot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOur research was analyzed using publicly available data and did not involve direct human participation. Therefore, approval from the Medical Ethics Committee was not required. All analytical procedures strictly adhered to relevant policies regarding data usage and distribution.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eClinical Trial Number\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003enot applicable.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to Publish\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003enot applicable.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eWeller, M., et al., \u003cem\u003eEANO guidelines on the diagnosis and treatment of diffuse gliomas of adulthood.\u003c/em\u003e Nat Rev Clin Oncol, 2021. \u003cstrong\u003e18\u003c/strong\u003e(3): p. 170-186.\u003c/li\u003e\n\u003cli\u003eHu, X., et al., \u003cem\u003eMultigene signature for predicting prognosis of patients with 1p19q co-deletion diffuse glioma.\u003c/em\u003e Neuro Oncol, 2017. \u003cstrong\u003e19\u003c/strong\u003e(6): p. 786-795.\u003c/li\u003e\n\u003cli\u003eOstrom, Q.T., et al., \u003cem\u003eCBTRUS Statistical Report: Primary brain and other central nervous system tumors diagnosed in the United States in 2010-2014.\u003c/em\u003e Neuro Oncol, 2017. \u003cstrong\u003e19\u003c/strong\u003e(suppl_5): p. v1-v88.\u003c/li\u003e\n\u003cli\u003eRen, J., et al., \u003cem\u003eThe Importance of M1-and M2-Polarized Macrophages in Glioma and as Potential Treatment Targets.\u003c/em\u003e Brain Sci, 2023. \u003cstrong\u003e13\u003c/strong\u003e(9).\u003c/li\u003e\n\u003cli\u003eMamdani, H., et al., \u003cem\u003eImmunotherapy in Lung Cancer: Current Landscape and Future Directions.\u003c/em\u003e Front Immunol, 2022. \u003cstrong\u003e13\u003c/strong\u003e: p. 823618.\u003c/li\u003e\n\u003cli\u003eKlobuch, S., et al., \u003cem\u003eTumour-infiltrating lymphocyte therapy for patients with advanced-stage melanoma.\u003c/em\u003e Nat Rev Clin Oncol, 2024. \u003cstrong\u003e21\u003c/strong\u003e(3): p. 173-184.\u003c/li\u003e\n\u003cli\u003eVerma, S., et al., \u003cem\u003eImmunotherapy and Radiation Therapy Sequencing in Breast Cancer: A Systematic Review.\u003c/em\u003e Int J Radiat Oncol Biol Phys, 2024. \u003cstrong\u003e118\u003c/strong\u003e(5): p. 1422-1434.\u003c/li\u003e\n\u003cli\u003ePersson, M.L., et al., \u003cem\u003eThe intrinsic and microenvironmental features of diffuse midline glioma: Implications for the development of effective immunotherapeutic treatment strategies.\u003c/em\u003e Neuro Oncol, 2022. \u003cstrong\u003e24\u003c/strong\u003e(9): p. 1408-1422.\u003c/li\u003e\n\u003cli\u003eGuo, X. and G. Wang, \u003cem\u003eAdvances in research on immune escape mechanism of glioma.\u003c/em\u003e CNS Neurosci Ther, 2023. \u003cstrong\u003e29\u003c/strong\u003e(7): p. 1709-1720.\u003c/li\u003e\n\u003cli\u003eYuan, Z., et al., \u003cem\u003eExtracellular matrix remodeling in tumor progression and immune escape: from mechanisms to treatments.\u003c/em\u003e Mol Cancer, 2023. \u003cstrong\u003e22\u003c/strong\u003e(1): p. 48.\u003c/li\u003e\n\u003cli\u003eMcGranahan, N., et al., \u003cem\u003eAllele-Specific HLA Loss and Immune Escape in Lung Cancer Evolution.\u003c/em\u003e Cell, 2017. \u003cstrong\u003e171\u003c/strong\u003e(6): p. 1259-1271.e11.\u003c/li\u003e\n\u003cli\u003eYamamoto, K., et al., \u003cem\u003eAutophagy promotes immune evasion of pancreatic cancer by degrading MHC-I.\u003c/em\u003e Nature, 2020. \u003cstrong\u003e581\u003c/strong\u003e(7806): p. 100-105.\u003c/li\u003e\n\u003cli\u003eLi, J., et al., \u003cem\u003eTumor-Associated Macrophage-Derived Exosomal LINC01232 Induces the Immune Escape in Glioma by Decreasing Surface MHC-I Expression.\u003c/em\u003e Adv Sci (Weinh), 2023. \u003cstrong\u003e10\u003c/strong\u003e(17): p. e2207067.\u003c/li\u003e\n\u003cli\u003eIwata, R., et al., \u003cem\u003eICOSLG-mediated regulatory T-cell expansion and IL-10 production promote progression of glioblastoma.\u003c/em\u003e Neuro Oncol, 2020. \u003cstrong\u003e22\u003c/strong\u003e(3): p. 333-344.\u003c/li\u003e\n\u003cli\u003eDing, X.C., et al., \u003cem\u003eThe relationship between expression of PD-L1 and HIF-1\u0026alpha; in glioma cells under hypoxia.\u003c/em\u003e J Hematol Oncol, 2021. \u003cstrong\u003e14\u003c/strong\u003e(1): p. 92.\u003c/li\u003e\n\u003cli\u003eLawson, K.A., et al., \u003cem\u003eFunctional genomic landscape of cancer-intrinsic evasion of killing by T cells.\u003c/em\u003e Nature, 2020. \u003cstrong\u003e586\u003c/strong\u003e(7827): p. 120-126.\u003c/li\u003e\n\u003cli\u003eWang, Q., et al., \u003cem\u003eDistinct molecular subtypes of systemic sclerosis and gene signature with diagnostic capability.\u003c/em\u003e Front Immunol, 2023. \u003cstrong\u003e14\u003c/strong\u003e: p. 1257802.\u003c/li\u003e\n\u003cli\u003eZhou, Q., et al., \u003cem\u003eMachine learning-based identification of a cell death-related signature associated with prognosis and immune infiltration in glioma.\u003c/em\u003e J Cell Mol Med, 2024. \u003cstrong\u003e28\u003c/strong\u003e(11): p. e18463.\u003c/li\u003e\n\u003cli\u003eWang, S., et al., \u003cem\u003eImmune cell infiltration-based signature for prognosis and immunogenomic analysis in breast cancer.\u003c/em\u003e Brief Bioinform, 2021. \u003cstrong\u003e22\u003c/strong\u003e(2): p. 2020-2031.\u003c/li\u003e\n\u003cli\u003eLi, G., et al., \u003cem\u003eAn MRI radiomics approach to predict survival and tumour-infiltrating macrophages in gliomas.\u003c/em\u003e Brain, 2022. \u003cstrong\u003e145\u003c/strong\u003e(3): p. 1151-1161.\u003c/li\u003e\n\u003cli\u003eZhao, Z., et al., \u003cem\u003eChinese Glioma Genome Atlas (CGGA): A Comprehensive Resource with Functional Genomic Data from Chinese Glioma Patients.\u003c/em\u003e Genomics Proteomics Bioinformatics, 2021. \u003cstrong\u003e19\u003c/strong\u003e(1): p. 1-12.\u003c/li\u003e\n\u003cli\u003eYu, K., et al., \u003cem\u003eSurveying brain tumor heterogeneity by single-cell RNA-sequencing of multi-sector biopsies.\u003c/em\u003e Natl Sci Rev, 2020. \u003cstrong\u003e7\u003c/strong\u003e(8): p. 1306-1318.\u003c/li\u003e\n\u003cli\u003eLi, P.H., et al., \u003cem\u003eRecent developments in application of single-cell RNA sequencing in the tumour immune microenvironment and cancer therapy.\u003c/em\u003e Mil Med Res, 2022. \u003cstrong\u003e9\u003c/strong\u003e(1): p. 52.\u003c/li\u003e\n\u003cli\u003eZhou, T., et al., \u003cem\u003eInflammatory markers and the risk of idiopathic sudden sensorineural hearing loss: A Mendelian randomization study.\u003c/em\u003e Front Neurol, 2023. \u003cstrong\u003e14\u003c/strong\u003e: p. 1111255.\u003c/li\u003e\n\u003cli\u003eChen, L., et al., \u003cem\u003eGenetic Insights into Obesity and Brain: Combine Mendelian Randomization Study and Gene Expression Analysis.\u003c/em\u003e Brain Sci, 2023. \u003cstrong\u003e13\u003c/strong\u003e(6).\u003c/li\u003e\n\u003cli\u003eJiang, Y., et al., \u003cem\u003eSingle-cell RNA sequencing highlights intratumor heterogeneity and intercellular network featured in adamantinomatous craniopharyngioma.\u003c/em\u003e Sci Adv, 2023. \u003cstrong\u003e9\u003c/strong\u003e(15): p. eadc8933.\u003c/li\u003e\n\u003cli\u003eQuail, D.F. and J.A. Joyce, \u003cem\u003eThe Microenvironmental Landscape of Brain Tumors.\u003c/em\u003e Cancer Cell, 2017. \u003cstrong\u003e31\u003c/strong\u003e(3): p. 326-341.\u003c/li\u003e\n\u003cli\u003eHu, L.S., et al., \u003cem\u003eIntegrated molecular and multiparametric MRI mapping of high-grade glioma identifies regional biologic signatures.\u003c/em\u003e Nat Commun, 2023. \u003cstrong\u003e14\u003c/strong\u003e(1): p. 6066.\u003c/li\u003e\n\u003cli\u003eRegmi, M., et al., \u003cem\u003eFrom glioma gloom to immune bloom: unveiling novel immunotherapeutic paradigms-a review.\u003c/em\u003e J Exp Clin Cancer Res, 2024. \u003cstrong\u003e43\u003c/strong\u003e(1): p. 47.\u003c/li\u003e\n\u003cli\u003eZhang, I.Y., et al., \u003cem\u003eRAGE ablation attenuates glioma progression and enhances tumor immune responses by suppressing galectin-3 expression.\u003c/em\u003e Neuro Oncol, 2023. \u003cstrong\u003e25\u003c/strong\u003e(5): p. 886-898.\u003c/li\u003e\n\u003cli\u003eSharma, P., et al., \u003cem\u003eHyperglycosylation of prosaposin in tumor dendritic cells drives immune escape.\u003c/em\u003e Science, 2024. \u003cstrong\u003e383\u003c/strong\u003e(6679): p. 190-200.\u003c/li\u003e\n\u003cli\u003eZhu, W., et al., \u003cem\u003eOMA1 competitively binds to HSPA9 to promote mitophagy and activate the cGAS-STING pathway to mediate GBM immune escape.\u003c/em\u003e J Immunother Cancer, 2024. \u003cstrong\u003e12\u003c/strong\u003e(4).\u003c/li\u003e\n\u003cli\u003eZhang, P., et al., \u003cem\u003eK63-linked ubiquitination of DYRK1A by TRAF2 alleviates Sprouty 2-mediated degradation of EGFR.\u003c/em\u003e Cell Death Dis, 2021. \u003cstrong\u003e12\u003c/strong\u003e(6): p. 608.\u003c/li\u003e\n\u003cli\u003eZou, Y., et al., \u003cem\u003eThe Single-Cell Landscape of Intratumoral Heterogeneity and The Immunosuppressive Microenvironment in Liver and Brain Metastases of Breast Cancer.\u003c/em\u003e Adv Sci (Weinh), 2023. \u003cstrong\u003e10\u003c/strong\u003e(5): p. e2203699.\u003c/li\u003e\n\u003cli\u003eTakeuchi, Y., et al., \u003cem\u003eHighly immunogenic cancer cells require activation of the WNT pathway for immunological escape.\u003c/em\u003e Sci Immunol, 2021. \u003cstrong\u003e6\u003c/strong\u003e(65): p. eabc6424.\u003c/li\u003e\n\u003cli\u003eTien, T.Y., et al., \u003cem\u003eHsa-miR-134-5p predicts cardiovascular risk in circulating mononuclear cells and improves angiogenic action of senescent endothelial progenitor cells.\u003c/em\u003e J Cell Mol Med, 2024. \u003cstrong\u003e28\u003c/strong\u003e(13): p. e18523.\u003c/li\u003e\n\u003cli\u003eWei, S., et al., \u003cem\u003eGFAT1-linked TAB1 glutamylation sustains p38 MAPK activation and promotes lung cancer cell survival under glucose starvation.\u003c/em\u003e Cell Discov, 2022. \u003cstrong\u003e8\u003c/strong\u003e(1): p. 77.\u003c/li\u003e\n\u003cli\u003eXia, Q., et al., \u003cem\u003eTRIM45 causes neuronal damage by aggravating microglia-mediated neuroinflammation upon cerebral ischemia and reperfusion injury.\u003c/em\u003e Exp Mol Med, 2022. \u003cstrong\u003e54\u003c/strong\u003e(2): p. 180-193.\u003c/li\u003e\n\u003cli\u003eKloesch, B., et al., \u003cem\u003eA GATA6-centred gene regulatory network involving HNFs and \u0026Delta;Np63 controls plasticity and immune escape in pancreatic cancer.\u003c/em\u003e Gut, 2022. \u003cstrong\u003e71\u003c/strong\u003e(4): p. 766-777.\u003c/li\u003e\n\u003cli\u003eWang, D.R., X.L. Wu, and Y.L. Sun, \u003cem\u003eTherapeutic targets and biomarkers of tumor immunotherapy: response versus non-response.\u003c/em\u003e Signal Transduct Target Ther, 2022. \u003cstrong\u003e7\u003c/strong\u003e(1): p. 331.\u003c/li\u003e\n\u003cli\u003eKim, J.M. and D.S. Chen, \u003cem\u003eImmune escape to PD-L1/PD-1 blockade: seven steps to success (or failure).\u003c/em\u003e Ann Oncol, 2016. \u003cstrong\u003e27\u003c/strong\u003e(8): p. 1492-504.\u003c/li\u003e\n\u003cli\u003eLiu, X., et al., \u003cem\u003eFOXP3(+) regulatory T cell perturbation mediated by the IFN\u0026gamma;-STAT1-IFITM3 feedback loop is essential for anti-tumor immunity.\u003c/em\u003e Nat Commun, 2024. \u003cstrong\u003e15\u003c/strong\u003e(1): p. 122.\u003c/li\u003e\n\u003cli\u003eWu, Y., et al., \u003cem\u003eMyeloid-derived suppressor cells: an emerging target for anticancer immunotherapy.\u003c/em\u003e Mol Cancer, 2022. \u003cstrong\u003e21\u003c/strong\u003e(1): p. 184.\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"discover-oncology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"dion","sideBox":"Learn more about [Discover Oncology](https://www.springer.com/12672)","snPcode":"","submissionUrl":"","title":"Discover Oncology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"High-grade glioma, immune escape, machine learning, immune microenvironment, immunotherapy","lastPublishedDoi":"10.21203/rs.3.rs-6330200/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6330200/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003ch2\u003eObjective\u003c/h2\u003e \u003cp\u003eHigh-grade glioma (HGG) is one of the most lethal malignancies. Immune escape is considered to be a reason for the failure of immunotherapy for HGG patients, and there is currently no immune escape-related prognostic model for glioma. Therefore, we explored the relationship between immune escape-related genes and the prognosis of patients with HGG.\u003c/p\u003e\u003ch2\u003eMethods\u003c/h2\u003e \u003cp\u003eThis study combined 101 machine learning algorithms to determine the best immune escape-related prognostic model. Subsequently, the TCGA and CGGA cohorts were used to verify the effectiveness of the model. Subsequently, molecular docking, Mendelian randomization and other comprehensive analyses were performed on the model genes. Finally, the biology function of the signature gene was further verified via CCK-8, and colony formation.\u003c/p\u003e\u003ch2\u003eResults\u003c/h2\u003e \u003cp\u003eOur differential expression analysis found that 41 immune escape-related genes were significantly related to the prognosis of HGG patients. We further selected 18 key genes through various machine learning methods to construct an immune escape-related prognosis model. This model can effectively distinguish between high-risk and low-risk groups, and shows good prediction results in both CGGA and TCGA data sets. Subsequently, the risk score was found to be an independent prognostic factor, and the nomogram including clinical characteristics was constructed. Immunotherapy response prediction results show that patients in the low-risk group respond better to immunotherapy and have longer survival. Furthermore, TAB1 knockdown reduced the ability of human glioma cells to proliferate and clone.\u003c/p\u003e\u003ch2\u003eConclusion\u003c/h2\u003e \u003cp\u003eThis study constructed a prognostic model related to immune escape through multiple machine learning methods and verified its clinical application value in HGG patients. It provides a theoretical basis for the exploration of immune escape treatment targets.\u003c/p\u003e","manuscriptTitle":"Comprehensive analysis of immune escape-related prognostic signature in high-grade glioma","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-05-13 10:06:10","doi":"10.21203/rs.3.rs-6330200/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-05-23T07:57:38+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-19T14:59:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"67383683132069633559861744977827097065","date":"2025-05-18T06:13:29+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-18T04:54:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"222870736336510699465105143099339316033","date":"2025-05-18T03:42:35+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-05-16T07:29:49+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"103237678294694897786777601994948150549","date":"2025-05-16T06:26:34+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"33959889617549949916883983474855869890","date":"2025-05-08T08:41:15+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-05-07T16:29:31+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-04-28T14:26:50+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-04-22T16:39:39+00:00","index":"","fulltext":""},{"type":"submitted","content":"Discover Oncology","date":"2025-04-22T16:38:31+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"discover-oncology","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"dion","sideBox":"Learn more about [Discover Oncology](https://www.springer.com/12672)","snPcode":"","submissionUrl":"","title":"Discover Oncology","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Discover Series","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"f0808c9c-ad4c-45b3-b295-5739e50f4c02","owner":[],"postedDate":"May 13th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"under-review","subjectAreas":[],"tags":[],"updatedAt":"2025-09-02T08:53:59+00:00","versionOfRecord":[],"versionCreatedAt":"2025-05-13 10:06:10","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6330200","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6330200","identity":"rs-6330200","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}