Single-Cell RNA-Sequencing of 6 Million Tumor and Immune Cells in 365 Patients with MGUS and SMM Demonstrates Complex Progressive Changes in Immune Cell Functionality with a Distinct Footprint on Tumor Cell Biology and Disease Control

INTRODUCTION

Despite the absence of symptoms, patients with Monoclonal Gammopathy of Undetermined Significance (MGUS) and Smoldering Multiple Myeloma (SMM), show significant changes in bone marrow (BM) immune cell composition, which may translate into impaired immune cell functionality. Here, we describe how immune dysregulation evolves with disease progression to Multiple Myeloma (MM), identify networks of dysregulation spanning tumor and immune cells, and nominate immune biomarkers of progression that may help to improve patient risk stratification and inform therapeutic strategies.

METHODS

We performed single-cell RNA-seq coupled with single-cell B cell receptor and T cell receptor (TCR) sequencing (10X Genomics) on ~6 million tumor and immune cells from 533 BM and peripheral blood (PB) samples of 365 patients with MGUS, SMM, and MM, as well as healthy donors. For 241 individuals, both tumor and immune cells were sequenced for integrative analyses of tumor and immune biology. Moreover, we performed bulk TCR-seq (Adaptive Biotechnologies) on PB-derived DNA from 100 patients with MM in the CoMMpass cohort.

RESULTS

We observed a significant decrease in the abundance of cytokine-expressing myeloid cells and a significant increase in the abundance of GZMB+ CD8+ effector memory T cells (TEMs) in patients with MM compared to MGUS. To assess if these changes were driven by altered expression of particular gene programs, we performed SignatureAnalyzer (which is based on Bayesian NMF) on lymphoid and myeloid cells together and extracted cross cell-type signatures whose activity we compared between disease stages. In line with our prior observation, a gene program specific to CD8+ TEMs, marked by granzyme-encoding genes, CCL5, and CD8A/CD8B, appeared to increase significantly with disease progression, as did an interferon (IFN) stimulation program that was active across cell types. A gene expression program marked by THBS1, PLAUR, TIMP1, and HIF1A, which was most active in cytokine-expressing myeloid cells, decreased significantly with progression to MM, in line with the observed decrease in the proportion of these cells. The IFN stimulation program spanned both tumor and immune cells in a cross-cell type network of dysregulation, and showed significantly higher activity in patients with SMM who progressed to MM. The observed increase of GZMB+ CD8+ TEMs was underlain by increased clonal expansion coupled with a bias towards terminal differentiation. A gradual decrease in TCR repertoire diversity was observed with progression to MM, which was independent of chronological age and was associated with a higher risk of progression. In the CoMMpass cohort, patients with overt MM and higher T cell clonality showed significantly shorter overall survival, independent of age, suggesting that T cell clonality may also be associated with suboptimal response to therapy.

Leveraging the large number of immune cells in our cohort, we could systematically assess for changes in immune cell functionality this time using SignatureAnalyzer within each cell type separately. We derived a comprehensive compendium of immune functionality programs within each cell type and extracted cross-cell type networks. We identified a network of distinct, but co-regulated signatures that spanned lymphoid and myeloid cells and decreased in activity with disease progression to MM. This network included a novel signature of T cell functionality that decreased with progression and was associated with a transcriptional footprint of immune control on tumor cells as well as a lower tumor cell proliferative rate, suggestive of more effective disease control. The immune control signature derived from tumor cells of patients with higher activity of the T cell functionality signature was then shown to decrease with progression in both our cohort as well as an external validation cohort.

CONCLUSION

We generated a comprehensive compendium of immune alterations with disease progression from MGUS to MM and revealed complex cross-cell type networks of dysregulation with a distinct footprint on tumor cell biology and disease control. This study may have implications for the risk stratification of patients with MM and its precursor conditions, provides a rationale for administering immunotherapy early in patients with better preserved immune function, and may unveil new targets for the development of immunotherapeutics.