Tumor Immune Cell Infiltration Patterns and Predictable Clinical Benefit with Nivolumab in Combination with Ipilimumab for 3P Implementation

Abstract Background The combination immunotherapy (CIT), administering both nivolumab and ipilimumab, has demonstrated the durable efficacy in patients with skin cutaneous melanoma (SKCM), kidney renal cell carcinoma (KIRC), and lung adenocarcinoma (LUAD). However, the survival benefits of the CIT vary among responders according to tumor type and thus, the biomarker development for CIT is necessary for predictive, preventive, and personalized (3P) medicine. Purpose To understand the mechanisms underlying the differential clinical efficacy and to explore the predictive biomarkers of the CIT, we analyzed transcriptome-based immune landscapes in SKCM, KIRC, and LUAD. Methods We obtained bulk tumor RNA-Seq data of LUAD (n=517), KIRC (n=506), and SKCM (n=472) from the Cancer Genome Atlas (TCGA) consortium and examined gene signature-based tumor-infiltrating immune cell profiles, as well as the correlations among expression levels of immune checkpoints and individual tumor mutational burden (TMB). Results Immunoprofiling revealed three subgroups of hot, intermediate, and cold clusters according to immune cell infiltration patterns for three tumor types examined. Among the relationship between immune checkpoints, CTLA-4 and PD-1 levels from LUAD and KIRC tumors were predominantly upregulated in immune-hot subgroups and exhibited strong concordance with each other (Spearman’s r=0.75 for LUAD; r=0.75 for KIRC). SKCM tumors were distinguished from LUAD and KIRC by manifesting relatively weaker correlations between PD-1 and CTLA-4 expression (r=0.58). Further analyses in the LUAD cohort presented that expression levels of immune checkpoints were dependent upon individual patient TMB, while overall tumor-infiltrating patterns of immune cells were poorly correlated with the mutational burden except the CD56dim NK cell subset. Conclusion Our data suggest that the gene signature-based profiling of tumor-infiltrating immune cells guides us to a better understanding of an immune landscape of the tumor immune microenvironment (TIME), and to predict the clinically demonstrated efficacy of the CIT in each cancer type. Therefore, CIT implemented through a more comprehensive characterization of immune features in individual patient’s tumors may enhance the clinical benefit of the predictive, preventive, personalized medicine (3PM) in cancer patients.