Identification of a Metabolism-Related Risk Signature Associated With Clinical Prognosis in Glioblastoma Using Integrated Bioinformatic Analysis

Lactate dehydrogenase A regulates tumor-macrophage symbiosis to promote glioblastoma progression

Abundant macrophage infiltration and altered tumor metabolism are two key hallmarks of glioblastoma. By screening a cluster of metabolic small-molecule compounds, we show that inhibiting glioblastoma cell glycolysis impairs macrophage migration and lactate dehydrogenase inhibitor stiripentol emerges as the top hit. Combined profiling and functional studies demonstrate that lactate dehydrogenase A (LDHA)-directed extracellular signal-regulated kinase (ERK) pathway activates yes-associated protein 1 (YAP1)/ signal transducer and activator of transcription 3 (STAT3) transcriptional co-activators in glioblastoma cells to upregulate C-C motif chemokine ligand 2 (CCL2) and CCL7, which recruit macrophages into the tumor microenvironment. Reciprocally, infiltrating macrophages produce LDHA-containing extracellular vesicles to promote glioblastoma cell glycolysis, proliferation, and survival. Genetic and pharmacological inhibition of LDHA-mediated tumor-macrophage symbiosis markedly suppresses tumor progression and macrophage infiltration in glioblastoma mouse models. Analysis of tumor and plasma samples of glioblastoma patients confirms that LDHA and its downstream signals are potential biomarkers correlating positively with macrophage density. Thus, LDHA-mediated tumor-macrophage symbiosis provides therapeutic targets for glioblastoma.

Boosting the therapy of glutamine-addiction glioblastoma by combining glutamine metabolism therapy with photo-enhanced chemodynamic therapy

The complete treatment of high grade invasive glioblastoma (GBM) remains to be a great challenge, and it is of great importance to develop innovative therapeutic approaches. Herein, we found that GBM derived from U87 MG cells is a glutamine-addiction tumor, and jointly using glutamine-starvation therapy and photo-enhanced chemodynamic therapy (CDT) can significantly boost its therapy. We rationally fabricated tumor cell membrane coated Cu2-xSe nanoparticles (CS NPs) and an inhibitor of glutamine metabolism (Purpurin) for combined therapy, because glutamine rather than glucose plays a crucial role in the proliferation and growth of GBM cells, and serves as a precursor for the synthesis of glutathione (GSH). The resultant CS-P@CM NPs can be specifically delivered to the tumor site to inhibit glutamine metabolism in tumor cells, suppress tumor intracellular GSH, and increase H2O2 content, which benefit the CDT catalyzed by CS NPs. The cascade reaction can be further enhanced by irradiation with the second near-infrared (NIR-II) light at the maximum concentration of H2O2, which can be monitored by photoacoustic imaging. The NIR-II light irradiation can generate a large amount of reactive oxygen species (ROS) within a short time to kill tumor cells and enhance the CDT efficacy. This is the first work on the treatment of orthotopic malignant GBM through combined glutamine metabolism therapy and photo-enhanced CDT, and provides insights into the treatment of other solid tumors by modulating the metabolism of tumor cells.

Identification of DNA Damage Repair Gene Signature as a Novel Prognostic Marker in Glioblastoma Multiforme

BACKGROUND

The efficacy of treatment of glioblastoma multiforme (GBM) is limited. The effect of DNA damage repair is an important factor.

METHODS

Expression data were downloaded from The Cancer Genome Atlas (training dataset) and the Gene Expression Omnibus (validation dataset) databases. Univariate Cox regression analysis and the least absolute shrinkage and selection operator were used to construct a DNA damage response (DDR) gene signature. Receiver operating characteristic curve analysis and Kaplan-Meier curve analysis were used to estimate the prognostic value of the risk signature. Moreover, consensus clustering analysis was used to investigate the potential subtypes of GBM according to DDR expression.

RESULTS

We constructed a 3-DDR-related gene signature through the survival analysis. The Kaplan-Meier curve analysis suggested that patients in the low-risk group have significantly better survival outcomes compared with the high-risk group in the training and external validation datasets. The results from the receiver operating characteristic curve analysis indicated that the risk model has high prognostic value in the training and external validation datasets. Moreover, 3 stable molecular subtypes were identified and validated in the Gene Expression Omnibus and The Cancer Genome Atlas databases according to the expression of the DNA repair gene. The microenvironment and immunity of GBM were further investigated and showed that cluster 2 had higher immunity and a higher immune score compared with clusters 1 and 3.

CONCLUSIONS

The DNA damage repair-related gene signature was an independent and powerful prognostic biomarker in GBM. Knowledge of the GBM subtypes could have important implications in the subclassification of GBM.

Identification and Validation of a Metabolism-Related Prognostic Signature Associated with M2 Macrophage Infiltration in Gastric Cancer

High levels of M2 macrophage infiltration invariably contribute to poor cancer prognosis and can be manipulated by metabolic reprogramming in the tumor microenvironment. However, the metabolism-related genes (MRGs) affecting M2 macrophage infiltration and their clinical implications are not fully understood. In this study, we identified 173 MRGs associated with M2 macrophage infiltration in cases of gastric cancer (GC) using the TCGA and GEO databases. Twelve MRGs were eventually adopted as the prognostic signature to develop a risk model. In the high-risk group, the patients showed poorer survival outcomes than patients in the low-risk group. Additionally, the patients in the high-risk group were less sensitive to certain drugs, such as 5-Fluorouracil, Oxaliplatin, and Cisplatin. Risk scores were positively correlated with the infiltration of multiple immune cells, including CD8+ T cells and M2 macrophages. Furthermore, a difference was observed in the expression and distribution between the 12 signature genes in the tumor microenvironment through single-cell sequencing analysis. In vitro experiments proved that the M2 polarization of macrophages was suppressed by Sorcin-knockdown GC cells, thereby hindering the proliferation and migration of GC cells. These findings provide a valuable prognostic signature for evaluating clinical outcomes and corresponding treatment options and identifying potential targets for GC treatment.

Development and validation of a two glycolysis-related LncRNAs prognostic signature for glioma and in vitro analyses

BACKGROUND

Mounting evidence suggests that there is a complex regulatory relationship between long non-coding RNAs (lncRNAs) and the glycolytic process during glioma development. This study aimed to investigate the prognostic role of glycolysis-related lncRNAs in glioma and their impact on the tumor microenvironment.

METHODS

This study utilized glioma transcriptome data from public databases to construct, evaluate, and validate a prognostic signature based on differentially expressed (DE)-glycolysis-associated lncRNAs through consensus clustering, DE-lncRNA analysis, Cox regression analysis, and receiver operating characteristic (ROC) curves. The clusterProfiler package was applied to reveal the potential functions of the risk score-related differentially expressed genes (DEGs). ESTIMATE and Gene Set Enrichment Analysis (GSEA) were utilized to evaluate the relationship between prognostic signature and the immune landscape of gliomas. Furthermore, the sensitivity of patients to immune checkpoint inhibitor (ICI) treatment based on the prognostic feature was predicted with the assistance of the Tumor Immune Dysfunction and Exclusion (TIDE) algorithm. Finally, qRT-PCR was used to verify the difference in the expression of the lncRNAs in glioma cells and normal cell.

RESULTS

By consensus clustering based on glycolytic gene expression profiles, glioma patients were divided into two clusters with significantly different overall survival (OS), from which 2 DE-lncRNAs, AL390755.1 and FLJ16779, were obtained. Subsequently, Cox regression analysis demonstrated that all of these lncRNAs were associated with OS in glioma patients and constructed a prognostic signature with a robust prognostic predictive efficacy. Functional enrichment analysis revealed that DEGs associated with risk scores were involved in immune responses, neurons, neurotransmitters, synapses and other terms. Immune landscape analysis suggested an extreme enrichment of immune cells in the high-risk group. Moreover, patients in the low-risk group were likely to benefit more from ICI treatment. qRT-PCR results showed that the expression of AL390755.1 and FLJ16779 was significantly different in glioma and normal cells.

CONCLUSION

We constructed a novel prognostic signature for glioma patients based on glycolysis-related lncRNAs. Besides, this project had provided a theoretical basis for the exploration of new ICI therapeutic targets for glioma patients.

Constructing metabolism-protein interaction relationship to identify glioma prognosis using deep learning

Glioma is heterogeneous disease that requires classification into subtypes with similar clinical phenotypes, prognosis or treatment responses. Metabolic-protein interaction (MPI) can provide meaningful insights into cancer heterogeneity. Moreover, the potential of lipids and lactate for identifying prognostic subtypes of glioma remains relatively unexplored. Therefore, we proposed a method to construct an MPI relationship matrix (MPIRM) based on a triple-layer network (Tri-MPN) combined with mRNA expression, and processed the MPIRM by deep learning to identify glioma prognostic subtypes. These Subtypes with significant differences in prognosis were detected in glioma (p-value < 2e-16, 95% CI). These subtypes had a strong correlation in immune infiltration, mutational signatures and pathway signatures. This study demonstrated the effectiveness of node interaction from MPI networks in understanding the heterogeneity of glioma prognosis.

Assessing Metabolic Markers in Glioblastoma Using Machine Learning: A Systematic Review

Glioblastoma (GBM) is a common and deadly brain tumor with late diagnoses and poor prognoses. Machine learning (ML) is an emerging tool that can create highly accurate diagnostic and prognostic prediction models. This paper aimed to systematically search the literature on ML for GBM metabolism and assess recent advancements. A literature search was performed using predetermined search terms. Articles describing the use of an ML algorithm for GBM metabolism were included. Ten studies met the inclusion criteria for analysis: diagnostic (n = 3, 30%), prognostic (n = 6, 60%), or both (n = 1, 10%). Most studies analyzed data from multiple databases, while 50% (n = 5) included additional original samples. At least 2536 data samples were run through an ML algorithm. Twenty-seven ML algorithms were recorded with a mean of 2.8 algorithms per study. Algorithms were supervised (n = 24, 89%), unsupervised (n = 3, 11%), continuous (n = 19, 70%), or categorical (n = 8, 30%). The mean reported accuracy and AUC of ROC were 95.63% and 0.779, respectively. One hundred six metabolic markers were identified, but only EMP3 was reported in multiple studies. Many studies have identified potential biomarkers for GBM diagnosis and prognostication. These algorithms show promise; however, a consensus on even a handful of biomarkers has not yet been made.

Metabolic-related gene pairs signature analysis identifies ABCA1 expression levels on tumor-associated macrophages as a prognostic biomarker in primary IDHWT glioblastoma

Background

Although isocitrate dehydrogenase (IDH) mutation serves as a prognostic signature for routine clinical management of glioma, nearly 90% of glioblastomas (GBM) patients have a wild-type IDH genotype (IDH^WT) and lack reliable signatures to identify distinct entities.

Methods

To develop a robust prognostic signature for IDH^WT GBM patients, we retrospectively analyzed 4 public datasets of 377 primary frozen tumor tissue transcriptome profiling and clinical follow-up data. Samples were divided into a training dataset (204 samples) and a validation (173 samples) dataset. A prognostic signature consisting of 21 metabolism-related gene pairs (MRGPs) was developed based on the relative ranking of single-sample gene expression levels. GSEA and immune subtype analyses were performed to reveal differences in biological processes between MRGP risk groups. The single-cell RNA-seq dataset was used to examine the expression distribution of each MRG constituting the signature in tumor tissue subsets. Finally, the association of MRGs with tumor progression was biologically validated in orthotopic GBM models.

Results

The metabolic signature remained an independent prognostic factor (hazard ratio, 5.71 [3.542-9.218], P < 0.001) for stratifying patients into high- and low-risk levels in terms of overall survival across subgroups with MGMTp methylation statuses, expression subtypes, and chemo/ratio therapies. Immune-related biological processes were significantly different between MRGP risk groups. Compared with the low-risk group, the high-risk group was significantly enriched in humoral immune responses and phagocytosis processes, and had more monocyte infiltration and less activated DC, NK, and γδ T cell infiltration. scRNA-seq dataset analysis identified that the expression levels of 5 MRGs (ABCA1, HMOX1, MTHFD2, PIM1, and PTPRE) in TAMs increased with metabolic risk. With tumor progression, the expression level of ABCA1 in TAMs was positively correlated with the population of TAMs in tumor tissue. Downregulation of ABCA1 levels can promote TAM polarization towards an inflammatory phenotype and control tumor growth.

Conclusions

The metabolic signature is expected to be used in the individualized management of primary IDH^WT GBM patients.

A lipid metabolism-related risk signature for patients with gliomas constructed with TCGA and CGGA data

Lipid metabolism affects cell proliferation, differentiation, membrane homeostasis and drug resistance. An in-depth exploration of lipid metabolism in gliomas might provide a novel direction for gliomas treatment. A lipid metabolism-related risk signature was constructed in our study to assess the prognosis of patients with gliomas. Lipid metabolism-related genes were extracted. Differentially expressed genes (DEGs) were screened, and a risk signature was built. The ability of the risk signature to predict the outcomes of patients with gliomas was assessed using the log-rank test and Cox regression analysis. The relationships between immunological characteristics, drug sensitivity and the risk score were evaluated, and the risk-related mechanisms were also estimated. Twenty lipid metabolism-related DEGs associated with the patient prognosis were included in the risk signature. The survival rate of high-risk patients was worse than that of low-risk patients. The risk score independently predicted the outcomes of patients. Immunological parameters, drug sensitivity, immunotherapy benefits, and numerous molecular mechanisms were significantly associated with the risk score. A lipid metabolism-related risk signature might effectively assess the prognosis of patients with gliomas. The risk score might guide individualized treatment and further clinical decision-making for patients with gliomas.

Development and Validation of a Prognostic Classifier Based on Lipid Metabolism-Related Genes for Breast Cancer

Background

The changes of lipid metabolism have been implicated in the development of many tumors, but its role in breast invasive carcinoma (BRCA) remains to be fully established. Here, we attempted to ascertain the prognostic value of lipid metabolism-related genes in BRCA.

Methods

We obtained RNA expression data and clinical information for BRCA and normal samples from public databases and downloaded a lipid metabolism-related gene set. Ingenuity Pathway Analysis (IPA) was applied to identify the potential pathways and functions of Differentially Expressed Genes (DEGs) related to lipid metabolism. Subsequently, univariate and multivariate Cox regression analyses were utilized to construct the prognostic gene signature. Functional enrichment analysis of prognostic genes was achieved by the Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG). Kaplan-Meier analysis, Receiver Operating Characteristic (ROC) curves, clinical follow-up results were employed to assess the prognostic potency. Potential compounds targeting prognostic genes were screened by Connectivity Map (CMap) database and a prognostic gene-drug interaction network was constructed using Comparative Toxicogenomics Database (CTD). Furthermore, we separately validated the selected marker genes in BRCA samples and human breast cancer cell lines (MCF-7, MDA-MB-231).

Results

IPA and functional enrichment analysis demonstrated that the 162 lipid metabolism-related DEGs we obtained were involved in many lipid metabolism and BRCA pathological signatures. The prognostic classifier we constructed comprising SDC1 and SORBS1 can serve as an independent prognostic marker for BRCA. CMap filtered 37 potential compounds against prognostic genes, of which 16 compounds could target both two prognostic genes were identified by CTD. The functions of the two prognostic genes in breast cancer cells were verified by cell function experiments.

Conclusion

Within this study, we identified a novel prognostic classifier based on two lipid metabolism-related genes: SDC1 and SORBS1. This result highlighted a new perspective on the metabolic exploration of BRCA.

Metabolism-Associated DNA Methylation Signature Stratifies Lower-Grade Glioma Patients and Predicts Response to Immunotherapy

Metabolism and DNA methylation (DNAm) are closely linked. The value of the metabolism-DNAm interplay in stratifying glioma patients has not been explored. In the present study, we aimed to stratify lower-grade glioma (LGG) patients based on the DNAm associated with metabolic reprogramming. Four data sets of LGGs from three databases (TCGA/CGGA/GEO) were used in this study. By screening the Kendall’s correlation of DNAm with 87 metabolic processes from KEGG, we identified 391 CpGs with a strong correlation with metabolism. Based on these metabolism-associated CpGs, we performed consensus clustering and identified three distinct subgroups of LGGs. These three subgroups were characterized by distinct molecular features and clinical outcomes. We also constructed a subgroup-related, quantifiable CpG signature with strong prognostic power to stratify LGGs. It also serves as a potential biomarker to predict the response to immunotherapy. Overall, our findings provide new perspectives for the stratification of LGGs and for understanding the mechanisms driving malignancy.

Comprehensive Analysis of the Tumor Immune Microenvironment Landscape in Glioblastoma Reveals Tumor Heterogeneity and Implications for Prognosis and Immunotherapy

Background

Glioblastoma (GBM) is a fatal brain tumor with no effective treatment. The specific GBM tumor immune microenvironment (TIME) may contribute to resistance to immunotherapy, a tumor therapy with great potential. Thus, an in-depth understanding of the characteristics of tumor-infiltrating immune cells is essential for exploring biomarkers in GBM pathogenesis and immunotherapy.

Methods

We estimated the relative abundances of 25 immune cell types in 796 GBM samples using single sample gene set enrichment analysis (ssGSEA). Unsupervised clustering was used to identify different GBM-associated TIME immune cell infiltration (GTMEI) patterns. The GTMEIscore system was constructed with principal component analysis (PCA) to determine the immune infiltration pattern of individual tumors.

Results

We revealed three distinct GTMEI patterns with different clinical outcomes and modulated biological pathways. We developed a scoring system (GTMEIscore) to determine the immune infiltration pattern of individual tumors. We comprehensively analyzed the genomic characteristics, molecular subtypes and clinicopathological features as well as proteomic, phosphoproteomic, acetylomic, lipidomic and metabolomic properties associated with the GTMEIscore and revealed many novel dysregulated pathways and precise targets in GBM. Moreover, the GTMEIscore accurately quantified the immune status of many other cancer types. Clinically, the GTMEIscore was found to have significant potential therapeutic value for chemotherapy/radiotherapy, immune checkpoint inhibitor (ICI) therapy and targeted therapy.

Conclusions

For the first time, we employed a multilevel and multiplatform strategy to construct a multidimensional molecular map of tumors with different immune infiltration patterns. These results may provide theoretical basises for identifying more effective predictive biomarkers and developing more effective drug combination strategies or novel immunotherapeutic agents for GBM.

m6A Regulatory Gene-Mediated Methylation Modification in Glioma Survival Prediction

The median survival of patients with gliomas is relatively short. To investigate the epigenetic mechanisms associated with poor survival, we analyzed publicly available datasets from patients with glioma. This analysis revealed 12 prognosis-related m6A regulatory genes that may be responsible for poor prognosis. These genes may be involved in genomic changes inherent to oxidative phosphorylation, adipogenesis, hedgehog signaling, and Myc signaling. We reconstructed a risk model with univariate and multivariate Cox analyses and identified older age and the m6A risk score as independent risk factors for predicting the prognosis of glioma patients, which is associated with glioma immune infiltration. In conclusion, m6A regulatory genes may serve as both reliable biomarkers and potential targets to increase the chance of survival of patients with glioma.

RPP25 as a Prognostic-Related Biomarker That Correlates With Tumor Metabolism in Glioblastoma

RPP25, a 25 kDa protein subunit of ribonuclease P (RNase P), is a protein-coding gene. Disorders associated with RPP25 include chromosome 15Q24 deletion syndrome and diffuse scleroderma, while systemic sclerosis can be complicated by malignancy. However, the functional role of RPP25 expression in glioblastoma multiforme (GBM) is unclear. In this study, comprehensive bioinformatics analysis was used to evaluate the impact of RPP25 on GBM occurrence and prognosis. Differential analysis of multiple databases showed that RPP25 was commonly highly expressed in multiple cancers but lowly expressed in GBM. Survival prognostic results showed that RPP25 was prognostically relevant in six tumors (CESC, GBM, LAML, LUAD, SKCM, and UVM), but high RPP25 expression was significantly associated with poor patient prognosis except for CESC. Analysis of RPP25 expression in GBM alone revealed that RPP25 was significantly downregulated in GBM compared with normal tissue. Receiver operating characteristic (ROC) combined with Kaplan-Meier (KM) analysis and Cox regression analysis showed that high RPP25 expression was a prognostic risk factor for GBM and had a predictive value for the 1-year, 2-year, and 3-year survival of GBM patients. In addition, the expression of RPP25 was correlated with the level of immune cell infiltration. The gene set enrichment analysis (GSEA) results showed that RPP25 was mainly associated with signalling pathways related to tumor progression and tumor metabolism.

Prognosis Implication of a Novel Metabolism-Related Gene Signature in Ewing Sarcoma

Ewing sarcoma (ES) is one of the most common bone cancers in adolescents and children. Growing evidence supports the view that metabolism pathways play critical roles in numerous cancers (He et al. (2020)). However, the correlation between metabolism-associated genes (MTGs) and Ewing sarcoma has not been investigated systematically. Here, based on the univariate Cox regression analysis, we get survival genes from differentially expressed genes (DEGs) from Gene Expression Omnibus (GEO) cohort. Multivariate Cox regression analysis and least absolute shrinkage and selection operator (LASSO) regression analysis were employed to establish the MTG signature. Comprehensive survival analyses including receiver operating characteristic (ROC) curves and Kaplan-Meier analysis were applied to estimate the independent prognostic value of the signature. The ICGC cohort served as the validation cohort. A nomogram was constructed based on the risk score of the MTG signature and other independent clinical variables. The CIBERSORT algorithm was applied to estimate immune infiltration. In addition, we explored the correlation between MTG signature and immune checkpoints. Collectively, this work presents a novel MTG signature for prognostic prediction of Ewing sarcoma. It also suggests six genes that are potential prognostic indicators and therapeutic targets for ES.

Prognostic Prediction Model for Glioblastoma: A Metabolic Gene Signature and Independent External Validation

Background: Glioblastoma (GBM) is the most common primary malignant intracranial tumor and closely related to metabolic alteration. However, few accepted prognostic models are currently available, especially models based on metabolic genes.

Methods: The transcriptome data were obtained for all of the patients diagnosed with GBM from the Gene Expression Omnibus (GEO) (training cohort, n=369) and The Cancer Genome Atlas (TCGA) (validation cohort, n=152) with the following variables: age at diagnosis, sex, follow-up and overall survival (OS). Metabolic genes according to Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways were contracted, and a Lasso regression model was constructed. Survival was assessed by univariate or multivariate Cox proportional hazards regression and Kaplan-Meier analysis, and an independent external validation was also conducted to examine the model.

Results: There were 341 metabolic genes showed significant differences between normal brain and GBM tissues in both the training and validation cohorts, among which 56 genes were dramatically correlated to the OS of patients. Lasso regression revealed that the metabolic prognostic model was composed of 18 genes, including COX10, COMT, and GPX2 with protective effects, as well as OCRL and RRM2 with unfavorable effects. Patients classified as high-risk by the risk score from this model had markedly shorter OS than low-risk patients (P<0.0001), and this significant result was also observed in independent external validation (P<0.001).

Conclusions: The prognosis of GBM was dramatically related to metabolic pathways, and our metabolic prognostic model had high accuracy and application value in predicting the OS of GBM patients.

Identification of Immune-Related lncRNA Signature to Predict Prognosis and Immunotherapeutic Efficiency in Bladder Cancer

Purpose

Identify immune-related lncRNA (IRL) signature related to the prognosis and immunotherapeutic efficiency for bladder cancer (BLCA) patients.

Methods

A total of 397 samples, which contained RNA-seq and clinical information from The Cancer Genome Atlas (TCGA) database, were used for the following study. Then the Lasso penalized Cox proportional hazards regression model was used to construct prognostic signature. According to the optimal cut-off value determined by time-dependent ROC curve, low and high-risk groups were set up. One immunotherapy microarray dataset as validation set was used to verify the ability of predicting immunotherapy efficacy. Furthermore, more evaluation between two risk groups related clinical factors were conducted. Finally, external validation of IRL-signature was conducted in Zhengzhou cohort.

Result

Four IRLs (HCP5, IPO5P1, LINC00942, and LINC01356) with significant prognostic value (P<0.05) were distinguished. This signature can accurately predict the overall survival of BLCA patients and was verified in the immunotherapy validation set. IRL-signatures can be used as independent prognostic risk factor in various clinical subgroups. According to the results of GSVA and MCP algorithm, we found that IRL-signature risk score is strikingly negative correlated with tumor microenvironment (TME) CD8+T cells and Cytotoxic lymphocytes infiltration, indicating that the better prognosis and immunotherapy might be caused partly by these. Then, the results from the TIDE analysis revealed that IRL could efficiently predict the response of immunotherapy in BLCA. External validation had similar results with TCGA-BLCA cohort.

Conclusions

The novel IRL-signature has a significant prognostic value for BLCA patients might facilitate predicting the efficacy of immunotherapy.