Mapping the immune checkpoint receptor-ligand landscape in newly diagnosed multiple myeloma may help prioritize patients for immune checkpoint inhibitor therapy

Abstract

Abstract Background: Recent clinical trials of immune checkpoint inhibitors (ICI) in multiple myeloma (MM) demonstrated the need to prioritize patients most likely to benefit from this type of therapy (Richard, et al., Nat. Cancer, 2024). Immune checkpoints in MM function through receptor-ligand (RL) interactions between CD8+ T cells and malignant plasma cells. However, there is a gap in our understanding of the immune checkpoint landscape in newly diagnosed MM (NDMM). To explore this, we mapped known immune checkpoint RL pairs in subjects enrolled in the CoMMpass Study (NCT01454297), a prospective, longitudinal observational trial. Methods Bone marrow aspirates from study participants were collected and separated into CD138 positive and CD138 negative fractions for comprehensive molecular characterization of the tumor (using bulk RNA sequencing and whole genome sequencing) and tumor immune microenvironment (using 3' single cell RNA sequencing (scRNA-seq)). 191 patients had paired bulk tumor and single cell immune expression profiles. To evaluate the co-expression of receptor-ligand pairs, we developed two approaches that combine gene expression values across single cell and bulk RNA-seq. The first approach uses joint percentiles to identify patients with high expression (top 75%) of both receptor and ligand (referred to as high-high). The second approach uses a rank-based interaction score to create a single value for each patient within a distribution defined for each receptor-ligand pair; patients with extreme values have high expression of both receptor and ligand. The two methods identified similar sets of patients with high co-expression of receptor and ligand. We then compared progression free survival (PFS) across receptor-ligand co-expression groups (high-high vs. others) and tested for other clinical associations. From a pre-defined list of 17 receptor-ligand pairs (Liu, et al., Exp. Hematol. Oncol., 2023), we focused on the following receptors (from CD8+ T cells) and ligands (from tumor plasma cells): PD-1 with PD-L1; LAG-3 with MHC-II; TIGIT with CD115; and TIM-3 with phosphatidylserine. Results Overall, 91 patients (91/191, 47.6%) showed high-high co-expression for at least 1 out of 17 receptor-ligand pairs, and 100 patients were high-high for no receptor-ligand pairs. Among patients with at least one high-high pair, there was a median of 2.0 high-high pairs (range: 1-11). Hierarchical clustering of high-high status and rank-based interaction score correlations showed that for a given receptor, patients often showed high-high co-expression across multiple related ligands. For TIM-3 (receptor, CD8+ T cells) and phosphatidylserine (ligand, tumor plasma cells), we identified 16 patients (16/191, 8.4%) with high-high co-expression and significantly worse PFS (median survival 485 vs. 1503 days, log-rank p-value 0.043). For PD-1-PD-L1, we identified 7 patients with high-high co-expression; for LAG-3-MHC-II, 12 patients (6.3%); and for TIGIT-CD115, 13 patients (6.8%). Conclusion ICI therapy remains an emerging area of study in MM. Here, we developed methods to explore immune checkpoint receptor-ligand co-expression from paired immune microenvironment and tumor cell populations. Our approach demonstrates the value of data integration across gene expression modalities, especially when standard sample collection methodologies segregate tumor and immune populations. Co-expression analysis may be one factor in prioritizing patients for immune checkpoint inhibitor therapy.