Development and validation of a prognostic and immunotherapeutically relevant model in hepatocellular carcinoma

Identify novel inflammation-related prognostic signature in pancreatic cancer patients

Pancreatic cancer (PC) is a malignant tumor of the digestive system with a poor prognosis. PC patients with pancreatitis have a worse prognosis. But nobody reported the relationship between inflammation and prognosis in PC. Based on this, we are going to explore inflammation-related prognostic signature to predict patients' survival and potential therapeutic target. We screened gene expression profile and corresponding clinical information of patients from The Cancer Genome Atlas (TCGA) database. Gene set enrichment analysis (GSEA) was performed to identify differentially expressed genes (DEGs) between tumor and normal tissues with P value < .05. Univariate and multivariate Cox regression analysis was applied to identify possible prognostic inflammation genes and establish an inflammation-related risk score system, which was validated by Kaplan-Meier and Receiver operating characteristic (ROC) curves. Finally, we used the TISIDB database to predict targeted drugs for up-regulated gene hepatocyte growth factor receptor (MET) and used AUTODOCK software for molecular docking. We built a prognostic model consisted of 3 inflammation-related genes (tumor necrosis factor receptor associated factor 1/TFAR1, tyrosine kinase 2/TYK2, MET). According to the median value of those genes' risk score, PC patients were ranked into high- (88) and low-risk (89) groups. Then, the results of the Kaplan-Meier curves and the area under the curve (AUC) of the ROC curves showed this model had a good predictive power (P < .001, AUC = 0.806). The result of human protein atlas (HPA) database showed the expression of TRAF1 and TYK2 were low in pancreatic cancer, the expression of MET was high. TISIDB database founded brigatinib could target to MET. And AUTODOCK showed brigatinib had a nice docking with MET. Taken together, our study suggested that inflammation-associated prognostic signature might be used as novel biomarkers for predicting prognosis in PC patients and potential therapeutic target of the disease.

A necroptosis-related gene signature to predict prognosis and immune features in hepatocellular carcinoma

BACKGROUND AND AIM

Necroptosis plays an important role in hepatocellular carcinoma (HCC) development, recurrence, and immunotherapy tolerance. We aimed to build a new prognostic necroptosis-related gene signature that could be used for survival and immunotherapy prediction in HCC patients.

METHODS

We found that necroptosis was associated with HCC progression and survival outcomes and was involved in the immune infiltration of HCC. Multiple bioinformatics methods including WGCNA, LASSO-Cox regression, stepwise Cox regression, and Random Forest and Boruta model analysis, were used to establish a prognostic profile related to necroptosis. The necroptosis-related gene signature was validated in ICGC and GSE14520 datasets.

RESULTS

This five-gene signature showed excellent predictive performance and was an independent risk factor for patients' overall survival outcome in the three cohorts. Moreover, this signature was an exact predictor using fewer genes than previous gene signatures. Finally, qRT-PCR and immunohistochemical staining investigations were performed in previously collected fresh frozen tumor tissues from HCC patients and their paracancerous normal tissues, and the results were consistent with the bioinformatics results. We found that LGALS3 not only affected the proliferation and migration ability of HepG2 cells but also affected necroptosis and the expression of inflammatory cytokines.

CONCLUSION

In summary, we established and validated an individualized prognostic profile related to necroptosis to forecast the therapeutic response to immune therapy, which might offer a potential non-apoptotic therapeutic target for HCC patients.

Clinical significance and immunogenomic landscape analysis of glycolysis-associated prognostic model to guide clinical therapy in hepatocellular carcinoma

Background

Hepatocellular carcinoma (HCC) is a common malignant tumor with a poor prognosis and high mortality rate worldwide. Glucose metabolism disorder is one of the most important characteristics of HCC. However, as the primary risk factors for the prognosis of HCC patients are unclear, the survival prognosis and therapy response of patients cannot be accurately predicted.

Methods

First, gene sets of 29 cancer hallmarks were collected from public databases. The z-score of various cancer hallmarks were quantitively analyzed by a single-sample gene set enrichment analysis (ssGSEA) of HCC patients. Next, a glycolysis-related gene signature (GRS) was constructed using a series of bioinformatics methods, which were used to predict the survival prognosis of HCC patients and the immunotherapy benefits. The prediction accuracy of the GRS was validated in different HCC cohorts and clinical subgroups. Additionally, a decision tree and nomogram were also established based on the GRS and other clinical variables. Finally, the genomic alterations and tumor immune microenvironment of the HCC patients were examined.

Results

Among the 29 cancer hallmarks, glycolysis was the most predominant risk factor for a poor prognosis in HCC. We subsequently constructed a novel GRS comprising 12 glycolysis-related genes. The high-GRS patients had a poorer survival prognosis than the low-GRS patients. The GRS exhibited a powerful ability to predict survival prognosis in different HCC cohorts and clinical feature subgroups. Additionally, the decision tree and nomogram aided in the risk stratification and prognosis evaluations of HCC patients. Further, we found that a high GRS was characterized by a severe tumor stage, pathological grade, and other clinical features. There were significant differences in the genomic alterations, immune cells, and immune checkpoints between the low- and high-GRS patients, especially in relation to the tumor protein p53 mutation and immunosuppressive cells. Notably, we also found that the GRS could be used to identify HCC patients who are more sensitive to chemotherapy and immunotherapy.

Conclusions

In summary, the GRS may be a useful tool for predicting the prognosis and guiding the clinical therapy of HCC patients.

Construction and Validation of an Immune Cell Signature Score to Evaluate Prognosis and Therapeutic Efficacy in Hepatocellular Carcinoma

Background: Hepatocellular carcinoma (HCC) is the most common primary liver malignancy with high morbidity and mortality worldwide. Tumor immune microenvironment (TIME) plays a pivotal role in the outcome and treatment of HCC. However, the effect of immune cell signatures (ICSs) representing the characteristics of TIME on the prognosis and therapeutic benefit of HCC patients remains to be further studied.

Materials and methods: In total, the gene expression profiles of 1,447 HCC patients from several databases, i.e., The Cancer Genome Atlas (TCGA), International Cancer Genome Consortium, and Gene Expression Omnibus, were obtained and applied. Based on a comprehensive collection of marker genes, 182 ICSs were evaluated by single sample gene set enrichment analysis. Then, by performing univariate and multivariate Cox analysis and random forest modeling, four significant signatures were selected to fit an immune cell signature score (ICSscore).

Results: In this study, an ICSscore-based prognostic model was constructed to stratify HCC patients into high-risk and low-risk groups in the TCGA-LIHC cohort, which was successfully validated in two independent cohorts. Moreover, the ICSscore values were found to positively correlate with the current American Joint Committee on Cancer staging system, indicating that ICSscore could act as a comparable biomarker for HCC risk stratification. In addition, when setting the four ICSs and ICSscores as features, the classifiers can significantly distinguish treatment-responding and non-responding samples in HCC. Also, in melanoma and breast cancer, the unified ICSscore could verify samples with therapeutic benefits.

Conclusion: Overall, we simplified the tedious ICS to develop the ICSscore, which can be applied successfully for prognostic stratification and therapeutic evaluation in HCC. This study provides an insight into the therapeutic predictive efficacy of prognostic ICS, and a novel ICSscore was constructed to allow future expanded application.

Development and Validation of a TNF Family-Based Signature for Predicting Prognosis, Tumor Immune Characteristics, and Immunotherapy Response in Colorectal Cancer Patients

In this study, a comprehensive analysis of TNF family members in colorectal cancer (CRC) was conducted and a TNF family-based signature (TFS) was generated to predict prognosis and immunotherapy response. Using the expression data of 516 CRC patients from The Cancer Genome Atlas (TCGA) database, TNF family members were screened to construct a TFS by using the univariate Cox proportional hazards regression and the least absolute shrinkage and selection operator- (LASSO-) Cox proportional hazards regression method. The TFS was then validated in a meta-Gene Expression Omnibus (GEO) cohort (n = 1162) from the GEO database. Additionally, the tumor immune characteristics and predicted responses to immune checkpoint blockade in TFS-based risk subgroups were analyzed. Eight genes (TNFRSF11A, TNFRSF10C, TNFRSF10B, TNFSF11, TNFRSF25, TNFRSF19, LTBR, and NGFR) were used to construct the TFS. Compared to the high-risk patients, the low-risk patients had better overall survival, which was verified by the GEO data. In addition, a high TFS risk score was associated with high infiltration of regulatory T cells (Tregs), nonactivated macrophages (M0), natural killer cells, immune escape phenotypes, poor immunotherapy response, and tumorigenic and metastasis-related pathways. Conversely, a low TFS risk score was related to high infiltration of resting CD4 memory T cells and resting dendritic cells, few immune escape phenotypes, and high sensitivity to immunotherapy. Thus, the eight gene-based TFS is a promising index to predict the prognosis, immune characteristics, and immunotherapy response in CRC, and our results also provide new understanding of the role of the TNF family members in the prognosis and treatment of CRC.

Computational pan-cancer characterization of model-based quantitative transcription regulations dysregulated in regional lymph node metastasis

Cancer is one of the major causes of mortality worldwide. Regional lymph node metastasis is an important mechanism during the spread of human cancers, in which transcription regulation plays an essential role. This study formulated a regression-model-based quantitative transcription regulation (mqTrans) between one mRNA gene and multiple transcription factors (TFs). Computational pan-cancer screening was carried out to detect the quantitative dysregulation of transcription regulation in the regional lymph node metastasis of 18 cancer types. Only a few metastasis-dysregulated mqTrans models were shared among the cancer types. The mRNA genes of the metastasis-dysregulated mqTrans models were not differentially expressed in regional lymph node metastasis. The experimental data suggested that mqTrans technology provided a complementary approach to the evaluation of transcription regulation mechanisms and may facilitate its quantitative investigation in other phenotypes.

Novel Hypoxia-Related Gene Signature for Risk Stratification and Prognosis in Hepatocellular Carcinoma

Hepatocellular carcinoma (HCC) is the most common form of liver cancer with limited therapeutic options and low survival rate. The hypoxic microenvironment plays a vital role in progression, metabolism, and prognosis of malignancies. Therefore, this study aims to develop and validate a hypoxia gene signature for risk stratification and prognosis prediction of HCC patients. The Cancer Genome Atlas (TCGA) and International Cancer Genome Consortium (ICGC) databases were used as a training cohort, and one Gene Expression Omnibus database (GSE14520) was served as an external validation cohort. Our results showed that eight hypoxia-related genes (HRGs) were identified by the least absolute shrinkage and selection operator analysis to develop the hypoxia gene signature and demarcated HCC patients into the high- and low-risk groups. In TCGA, ICGC, and GSE14520 datasets, patients in the high-risk group had worse overall survival outcomes than those in the low-risk group (all log-rank P < 0.001). Besides, the risk score derived from the hypoxia gene signature could serve as an independent prognostic factor for HCC patients in the three independent datasets. Finally, a nomogram including the gene signature and tumor-node-metastasis stage was constructed to serve clinical practice. In the present study, a novel hypoxia signature risk model could reflect individual risk classification and provide therapeutic targets for patients with HCC. The prognostic nomogram may help predict individualized survival.

The novel immune-related genes predict the prognosis of patients with hepatocellular carcinoma

Hepatocellular carcinoma (HCC) is one of the main causes of cancer deaths globally. Immunotherapy is becoming increasingly important in the cure of advanced HCC. Thus it is essential to identify biomarkers for treatment response and prognosis prediction. We searched publicly available databases and retrieved 465 samples of genes from The Cancer Genome Atlas (TCGA) database and 115 tumor samples from Gene Expression Omnibus (GEO). Meanwhile, we used the ImmPort database to determine the immune-related genes as well. Weighted gene correlation network analysis, Cox regression analysis and least absolute shrinkage and selection operator (LASSO) analysis were used to identify the key immune related genes (IRGs) which are closely related to prognosis. Gene set enrichment analysis (GSEA) was implemented to explore the difference of Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway between Immune high- and low-risk score groups. Finally, we made a prognostic nomogram including Immune-Risk score and other clinicopathologic factors. A total of 318 genes from prognosis related modules were identified through weighted gene co-expression network analysis (WGCNA). 46 genes were strongly linked to prognosis after univariate Cox analysis. We constructed a seven genes prognostic signature which showed powerful prediction ability in both training cohort and testing cohort. 16 significant KEGG pathways were identified between high- and low- risk score groups using GSEA analysis. This study identified and verified seven immune-related prognostic biomarkers for the patients with HCC, which have potential value for immune modulatory and therapeutic targets.