Skip to main page content
U.S. flag

An official website of the United States government

Dot gov

The .gov means it’s official.
Federal government websites often end in .gov or .mil. Before sharing sensitive information, make sure you’re on a federal government site.

Https

The site is secure.
The https:// ensures that you are connecting to the official website and that any information you provide is encrypted and transmitted securely.

Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2020 Nov 13:7:581354.
doi: 10.3389/fmolb.2020.581354. eCollection 2020.

A Prognostic Model of 15 Immune-Related Gene Pairs Associated With Tumor Mutation Burden for Hepatocellular Carcinoma

Junyu Huo  1 , Liqun Wu  1 , Yunjin Zang  1
Affiliations

A Prognostic Model of 15 Immune-Related Gene Pairs Associated With Tumor Mutation Burden for Hepatocellular Carcinoma

Junyu Huo et al. Front Mol Biosci. .

Abstract

Introduction: Tumor mutation burden (TMB) is an emerging biomarker for immunotherapy of hepatocellular carcinoma (HCC), but its value for clinical application has not been fully revealed.

Materials and methods: We used the Wilcox test to identify the differentially expressed immune-related genes (DEIRGs) in groups with high and low TMB as well as screened the immune gene pairs related to prognosis using univariate Cox regression analysis. A LASSO Cox regression prognostic model was developed by combining The Cancer Genome Atlas Liver Hepatocellular Carcinoma (TCGA-LIHC) with the International Cancer Genome Consortium (ICGC) LIRI-JP cohort, and internal (TCGA, ICGC) and external (GSE14520) validation analyses were conducted on the predictive value of the model. We also explored the relationship between the prognostic model and tumor microenvironment via the ESTIMATE algorithm and performed clinical correlation analysis by the chi-square test, revealing its underlying molecular mechanism with the help of Gene Set Enrichment Analysis (GSEA).

Results: The prognostic model consisting of 15 immune gene pairs showed high predictive value for short- and long-term survival of HCC in three independent cohorts. Based on univariate multivariate Cox regression analysis, the prognostic model could be used to independently predict the prognosis in each independent cohort. The immune score, stromal score, and estimated score values were lower in the high-risk group than in the low-risk group. As shown by the chi-square test, the prognostic model exhibited an obvious correlation with the tumor stage [American Joint Committee on Cancer tumor-node-metastasis (AJCC-TNM) (p < 0.001), Barcelona Clinic Liver Cancer (BCLC) (p = 0.003)], histopathological grade (p = 0.033), vascular invasion (p = 0.009), maximum tumor diameter (p = 0.013), and background of liver cirrhosis (p < 0.001). GSEA revealed that the high-risk group had an enrichment of many oncology features, including the cell cycle, mismatch repair, DNA replication, RNA degradation, etc.

Conclusion: Our research developed and validated a reliable prognostic model associated with TMB for HCC, which may help to further enrich the therapeutic targets of HCC.

Keywords: hepatocellular carcinoma; immune; prognostic; signature; tumor mutation burden.

PubMed Disclaimer

Figures

FIGURE 1
FIGURE 1
The workflow chart of this study.
FIGURE 2
FIGURE 2
The relationship among tumor mutation burden, immune infiltration, and prognosis of HCC. (A) The waterfall plot mutation profiles in HCC samples from the combined TCGA and ICGC datasets. (B) The Kaplan–Meier survival curve regarding the TMB and overall survival. (C) The bar plot of 22 specific immune fractions represented by various colors in each sample are shown in the bar plot. (D) Correlation matrix of all 22 immune cell proportions. (E) The violin plot of different infiltration levels of immune cells between high- and low-TMB patients.
FIGURE 3
FIGURE 3
Identification of differentially expressed immune-related genes (DEIRGs) between high- and low-TMB patients. (A) Heatmap of DEIRGs between high- and low-TMB patients. (B) Protein–protein interaction plot of DEIRGs between high- and low-TMB patients (green represents upregulated genes in the low-TMB group, and red represents upregulated genes in the high-TMB group). (C,D) GO enrichment analysis of DEIRGs. (E,F) KEGG enrichment analysis of TRMGs.
FIGURE 4
FIGURE 4
The building process of the prognostic signature. (A) Forrest plot of prognostic DEIRGs identified by univariate Cox and Kaplan–Meier survival analyses. (B–D) The establishment of the prognostic model based on LASSO penalized Cox analysis.
FIGURE 5
FIGURE 5
Construction of the prognostic model in the training cohort. (A,B) Kaplan–Meier survival analysis and time-dependent ROC analysis for predicting the overall survival of patients in the training cohort using the risk score. (C–E) Heatmap of the 15 gene pairs, the distribution of the risk score, and the survival status of patients. (F) Kaplan–Meier survival analysis for predicting the overall survival of patients in the low-TMB group using the risk score. (G) Kaplan–Meier survival analysis of predicting the overall survival of patients in the high-TMB group using the risk score.
FIGURE 6
FIGURE 6
Internal validation of the prognostic model in the TCGA and ICGC cohorts. (A,B) Kaplan–Meier survival analysis and time-dependent ROC analysis of predicting the overall survival of patients in the TCGA cohort using the risk score. (C,D) Univariate and multivariate regression analyses of the relation between the risk score and clinicopathological characteristics regarding overall survival in the TCGA cohort (green represents univariate analysis, and red represents multivariate analysis). (E,F) Kaplan–Meier survival analysis and time-dependent ROC analysis for predicting the overall survival of patients in the ICGC cohort using the risk score. (G,H) Univariate and multivariate regression analyses of the relation between the risk score and clinicopathological characteristics regarding the overall survival in the ICGC cohort (green represents univariate analysis, and red represents multivariate analysis).
FIGURE 7
FIGURE 7
External validation of the prognostic model in the GSE14520 cohort. (A,B) Kaplan–Meier survival analysis and time-dependent ROC analysis for predicting the overall survival of patients in the GSE14520 cohort using the risk score. (C,D) Univariate and multivariate regression analyses of the relation between the risk score and clinicopathological characteristics regarding the overall survival in the GSE14520 cohort (green represents univariate analysis, and red represents multivariate analysis). (E,F) Kaplan–Meier survival analysis and time-dependent ROC analysis for predicting the recurrence-free survival of patients in the GSE14520 cohort using the risk score. (G,H) Univariate and multivariate regression analyses of the relation between the risk score and clinicopathological characteristics regarding the recurrence-free survival in the GSE14520 cohort (green represents univariate analysis, and red represents multivariate analysis).
FIGURE 8
FIGURE 8
Validation of the universal applicability of the prognostic model in the whole cohort. (A,B) Kaplan–Meier survival analysis and time-dependent ROC analysis for predicting the overall survival of patients in the whole cohort using the risk score. (C–E) Subgroup Kaplan–Meier survival analysis according to different clinical features.
FIGURE 9
FIGURE 9
The relationship between the prognostic model and tumor microenvironment. (A–C) Kaplan–Meier survival analysis of patients in the whole cohort using the immune score, stromal score, and ESTIMATE score, respectively. (D–F) Comparison of the immune score, stromal score, and ESTIMATE score in the low- and high-risk groups, respectively. (G–I) Pearson correlation analysis between the risk score and immune score, stromal score, and ESTIMATE score, respectively.
FIGURE 10
FIGURE 10
Gene set enrichment analysis between different risk groups in the (A) TCGA cohort, (B) ICGC cohort, and (C) GSE14520 cohort.
See this image and copyright information in PMC

Similar articles

See all similar articles

Cited by

See all "Cited by" articles

References

    1. Alborelli I., Leonards K., Rothschild S. I., Leuenberger L. P., Savic Prince S., Mertz K. D., et al. (2020). Tumor mutational burden assessed by targeted NGS predicts clinical benefit from immune checkpoint inhibitors in non-small cell lung cancer. J. Pathol. 250 19–29. 10.1002/path.5344 - DOI - PMC - PubMed
    1. Antonarakis E. S. (2019). A new molecular taxonomy to predict immune checkpoint inhibitor sensitivity in prostate cancer. Oncologist 24 430–432. 10.1634/theoncologist.2018-0819 - DOI - PMC - PubMed
    1. Cai H., Zhang Y., Zhang H., Cui C., Li C., Lu S. (2020). Prognostic role of tumor mutation burden in hepatocellular carcinoma after radical hepatectomy. J. Surg. Oncol. 121 1007–1014. 10.1002/jso.25859 - DOI - PubMed
    1. Calderaro J., Meunier L., Nguyen C. T., Boubaya M., Caruso S., Luciani A., et al. (2019). ESM1 as a marker of macrotrabecular-massive hepatocellular carcinoma. Clin. Cancer Res. 25 5859–5865. 10.1158/1078-0432.ccr-19-0859 - DOI - PubMed
    1. Carmo R., Aroucha D., Vasconcelos L., Pereira L., Moura P., Cavalcanti M. (2016). Genetic variation in PTX 3 and plasma levels associated with hepatocellular carcinoma in patients with HCV. J. Viral Hepat. 23 116–122. 10.1111/jvh.12472 - DOI - PubMed