Abstract 996P
Background
Hepatocellular carcinoma (HCC) is characterized by poor prognosis and a high recurrence rate. Here we conduct a genomic data driven classification of immune microenvironment subtypes in HCC. Additionally, we demonstrate the prognostic value and suggest the potential for therapeutic targeting based on immune subtype.
Methods
Whole exome RNAseq data from the cancer genome atlas HCC project (TCGA-LIHC) was utilized (n = 366). Abundance of immune cells was imputed using CIBERSORT, and unsupervised hierarchical clustering was used to visualize the data. 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 PD-1/PD-L1 co-expression.
Results
Of four clusters, two were characterized by increased macrophages, with distinct M0Hi and M2Hi subtypes. M2Hi and M0Hi were independently prognostic for OS on multivariable analysis (table). Kaplan-Meier estimates demonstrated that M0Hi and M2Hi clusters had decreased OS and post-treatment survival after sorafenib, and angiogenesis hallmark genes were enriched in the M0Hi group (table). CXCL6 and POSTN were overexpressed in both the M0Hi and the PD-1Hi / PD-L1Hi groups. A score consisting of CXCL6 and POSTN expression and absolute M0 macrophage content was discriminatory for OS (Favorable = Ref, Intermediate HR 1.59, 95% CI = 1.02 – 2.5, p = <0.001, Unfavorable HR 2.08, 95% CI 1.37 – 3.23, p = 0.04). Table: 996P
Characteristics and prognostic indicators of immune clusters
Cluster 1 (CD8Hi) | Cluster 2 (CD4Hi) | Cluster 3 (M2Hi) | Cluster 4 (M0Hi) | |
Macrophages (M0, M1, M2)* | 0.84 | 0.99 | 1.28 | 1.36 |
M0* | 0.29 | 0.31 | 0.16 | 0.77 |
M2* | 0.38 | 0.43 | 0.95 | 0.41 |
Multivariate HR (95% CI, p-value) | REF | 0.89 (0.34-2.3, p=0.8) | 2.04 (1.04-3.99, p=0.04) | 2.62 (1.18-5.84, p=0.02) |
Overall Survival (In months, range) | 80.7 (55.4 - N/A)p = 0.003 | 45.6 (27.5 - 60.9) | ||
Post-Sorafenib Survival (In months, range) | 29.9 (3.47 - N/A)p = 0.04 | 2.6 (1.4 - N/A) | ||
Angiogenesis Genes Enrichment Score (p-value) | REF | NS | NS | 0.66 (p = 0.008) |
*Absolute immune score, derived from cibersort.Abbreviations: HR: hazard ratio; NS: not significant; N/A: not available.
Conclusions
Distinct immune clusters with macrophage predominance characterize an aggressive HCC phenotype, defined molecularly by angiogenic gene enrichment and clinically by poor prognosis and sorafenib response. The novel immuno-genomic signature described above may aid in stratification of unresectable patients to receive checkpoint inhibitor and anti-angiogenic therapy combinations rather than sorafenib.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
The authors.
Funding
Has not received any funding.
Disclosure
I. El Naqa: Advisory/Consultancy: Endecxtra LLC; Advisory/Consultancy: Resero AI; Speaker Bureau/Expert testimony: Sun Nuclear. All other authors have declared no conflicts of interest.