Characterization of the Tumor Immune Microenvironment Identifies M0 Macrophage-Enriched Cluster as a Poor Prognostic Factor in Hepatocellular Carcinoma

Abstract

Purpose: Hepatocellular carcinoma (HCC) is characterized by a poor prognosis and a high recurrence rate. The tumor immune microenvironment in HCC has been characterized as shifted toward immunosuppression. We conducted a genomic data-driven classification of immune microenvironment HCC subtypes. In addition, we demonstrated their prognostic value and suggested a potential therapeutic targeting strategy.

Methods: RNA sequencing data from The Cancer Genome Atlas-Liver Hepatocellular Carcinoma was used (n = 366). Abundance of immune cells was imputed using CIBERSORT and visualized using unsupervised hierarchic clustering. Overall survival (OS) was analyzed using Kaplan-Meier estimates and Cox regression. Differential expression and gene set enrichment analyses were conducted on immune clusters with poor OS and high programmed death-1 (PD-1)/programmed death-ligand 1 (PD-L1) coexpression. A scoring metric combining differentially expressed genes and immune cell content was created, and its prognostic value and immune checkpoint blockade response prediction was evaluated.

Results: Two clusters were characterized by macrophage enrichment, with distinct M0^Hi and M2^Hi subtypes. M2^Hi (P = .038) and M0^Hi (P = .018) were independently prognostic for OS on multivariable analysis. Kaplan-Meier estimates demonstrated that patients in M0^Hi and M2^Hi treated with sorafenib had decreased OS (P = .041), and angiogenesis hallmark genes were enriched in the M0^Hi group. CXCL6 and POSTN were overexpressed in both the M0^Hi and the PD-1^Hi/PD-L1^Hi groups. A score consisting of CXCL6 and POSTN expression and absolute M0 macrophage content was discriminatory for OS (intermediate: hazard ratio [HR], 1.59; P ≤ .001; unfavorable: HR, 2.08; P = .04).

Conclusion: Distinct immune cell clusters with macrophage predominance characterize an aggressive HCC phenotype, defined molecularly by angiogenic gene enrichment and clinically by poor prognosis and sorafenib response. This novel immunogenomic signature may aid in stratification of unresectable patients to receive checkpoint inhibitor and antiangiogenic therapy combinations.

FIG 1.

Hierarchic clustering based on immune cell subsets identifies macrophage-enriched clusters. (A) The results of unsupervised hierarchic clustering based on immune cell subsets comprising each sample in the Cancer Genome Atlas–Liver Hepatocellular Carcinoma (TCGA-LIHC) dataset. The columns correspond with immune cell subsets, and the rows represent patient sample IDs. Yellow shading represents higher relative fraction, whereas blue shading represents lower relative fraction. The clusters are defined as follows: cluster 1: macrophage^Mid and C4^Hi; cluster 2: macrophage^Mid and CD8^Hi; cluster 3: macrophage^Hi and M2^Hi; and cluster 4: macrophage^Hi and M0^Hi. (B) The mean absolute immune scores, derived from CIBERSORT, presented as a barplot grouped by clusters. Bars are colored according to the corresponding immune cell. (C) A correlation matrix of correlation coefficients for all possible immune cell subset pairings. Red shading represents positive correlation, whereas blue shading represents negative correlation. The size of each bubble and the intensity of the shading correspond to the magnitude of the correlation coefficient. Correlations that did not meet significance are marked with an x. T reg, regulatory T cells.

FIG 1.

Hierarchic clustering based on immune cell subsets identifies macrophage-enriched clusters. (A) The results of unsupervised hierarchic clustering based on immune cell subsets comprising each sample in the Cancer Genome Atlas–Liver Hepatocellular Carcinoma (TCGA-LIHC) dataset. The columns correspond with immune cell subsets, and the rows represent patient sample IDs. Yellow shading represents higher relative fraction, whereas blue shading represents lower relative fraction. The clusters are defined as follows: cluster 1: macrophage^Mid and C4^Hi; cluster 2: macrophage^Mid and CD8^Hi; cluster 3: macrophage^Hi and M2^Hi; and cluster 4: macrophage^Hi and M0^Hi. (B) The mean absolute immune scores, derived from CIBERSORT, presented as a barplot grouped by clusters. Bars are colored according to the corresponding immune cell. (C) A correlation matrix of correlation coefficients for all possible immune cell subset pairings. Red shading represents positive correlation, whereas blue shading represents negative correlation. The size of each bubble and the intensity of the shading correspond to the magnitude of the correlation coefficient. Correlations that did not meet significance are marked with an x. T reg, regulatory T cells.

FIG 2.

Macrophage enrichment is associated with decreased overall survival and response to adjuvant therapy, including sorafenib. (A) Multivariable analysis with hazard ratio (HR) represented as a forest plot adjusting for key covariates, including clinical and pathologic characteristics, risk factors, and immune cluster. The covariates are listed in the left column, whereas HRs, CIs, and P values are in the right column. (B) A multivariable analysis for the combination of alpha-fetoprotein (AFP) and cluster. The left column lists the various combinations, with A representing an AFP < 400 ng/mL and B representing an AFP > 400 ng/mL. The corresponding number represents the cluster. AIC, Aikake information criterion. * P = .01-.05; ** P = .001-.01; *** P = .0-.001.

FIG 3.

Macrophage enrichment is associated with decreased overall survival and response to adjuvant therapy, including sorafenib. (A) The Kaplan-Meier curve for overall survival for each of the clusters in The Cancer Genome Atlas–Liver Hepatocellular Carcinoma (TCGA-LIHC) cohort. Clusters are color coded with a legend at the top of the figure. A risk table is displayed at the bottom of the table. (B) The Kaplan-Meier curve for overall survival for combined clusters (1 and 2, 3 and 4). (C) The Kaplan-Meier curve for post-treatment survival for combined clusters receiving any adjuvant drug therapy. (D) The Kaplan-Meier curve for post-treatment survival for combined clusters receiving adjuvant sorafenib.

FIG 4.

Gene set enrichment analyses (GSEA) identifies enrichment of angiogenesis hallmark genes. Heatmap representing the enrichment scores of MSigDB cancer hallmarks. A red shade represents a gene set that is overenriched and a blue shade represents a gene set that is underenriched. The intensity of the shade corresponds to the magnitude of the enrichment score. Blanks in the heatmap represent gene sets that were not significantly overenriched or underenriched. Clustering was performed on both rows (hallmark gene sets, n = 3), and columns (clusters, n = 2). ES, enrichment score.

FIG 5.

Immunogenomic signature is prognostic for overall survival in both The Cancer Genome Atlas (TCGA) and FSE14520. (A) The Kaplan-Meier curve for overall survival for each of the score groupings in the TCGA–Liver Hepatocellular Carcinoma (LIHC) cohort. Clusters are color coded with a legend at the top of the figure. A risk table is displayed at the bottom of the table. (B) The Kaplan-Meier curve for overall survival for the score groupings in the GSE14520. (C) A box plot representation of T-cell dysfunction scores (y-axis) by score grouping (x-axis). Global Kruskal-Wallis P value is displayed along with Mann-Whitney pairwise comparisons. (D) A box plot representation of cancer-associated fibroblast signature (y-axis) by score grouping (x-axis). Kruskal-Wallis and Mann-Whitney P values are displayed.