Senior Scientist Multiple Myeloma Research Foundation Durham, North Carolina, United States
Introduction: Immune components of the bone marrow (BM) microenvironment in multiple myeloma (MM) have emerged as critical factors for disease progression and therapy responses. Identifying dysregulated immune populations associated with clinical outcomes has thus attracted significant attention. To this end, we profiled the BM microenvironment (BMME) of newly diagnosed MM (NDMM) patients to identify immune populations and signaling pathways associated with cytogenetic risk and disease progression.
Methods: We established a harmonized consortium to generate an Immune Atlas of MM aimed at informing disease etiology, risk stratification, and potential therapeutic strategies. Together, we analyzed 361 CD138neg BM samples from 263 NDMM subjects enrolled in the CoMMpass clinical trial via scRNA-seq resulting in more than 1.1 million single cells representing immune subpopulations spanning myeloid, lymphoid, and erythroid lineages. These subpopulations were tested for significant association with cytogenetic risk and disease progression. Outcome-associated populations were characterized by differential expression, pathway enrichment, and intercellular communication analyses. Lastly, we developed risk stratification models incorporating cytogenetic events and immune signatures and assessed their performance.
Results: Compositional analysis revealed multiple significant immune alterations associated with poor outcomes. This includes depletion of B cells, depletion of naïve T cells, enrichment of terminally differentiated, senescent CD8+ T cells, and general enrichment of cells of myeloid lineage. Empiric analyses suggest that NDMM patients with poor outcomes exhibited a shift toward immunosenescence and late-activated CD8+ T cells producing IFN-II. In contrast, MM patients with better outcomes displayed expanding naïve and central memory CD8+ T cell subsets. This could be in part due to observed differential signaling of BAFF by IFN-I-stimulated monocytes and APRIL by IFN-II-stimulated monocytes. Finally, we assessed if incorporating various cell abundance signatures could improve risk stratification over prevailing cytogenetic risk models using a bootstrap validation approach.Incorporating cell abundance signatures resulting in an increase in model AUC from 0.73 with only demographic and cytogenetic metrics, to 0.81 when adding the top 11 predictive cell clusters, to 0.96 using all immune clusters.
Conclusions: Profiling the BMME of NDMM patients revealed dysregulated T cells and myeloid subpopulations associated with patient risk and disease progression. Key signaling pathways, namely IFN, APRIL, and BAFF, appear to drive these outcome-associated populations. Integrating these findings with cytogenetics and clinical characteristics improved the prediction of PFS. Collectively, this work underscores the importance of capturing the BMME as a prognostic marker for MM and may allow for rational optimization of treatment regimens.