Detailed Phenotypic, Molecular and Functional Profiling of Myeloid Derived Suppressor Cells (mdscs) in the Tumor Immune Microenvironment (TIME) of Multiple Myeloma (MM)

Background:Deep understanding of the complexity and diversity of the TIME and its influence on response to therapy is needed to improve the ability to predict, monitor and guide immunotherapeutic responsiveness. Among cells composing the MM‐TIME, granulocytic‐MDSCs (G‐MDSCs) have a prominent role in promoting tumor growth and inducing immunosuppression; however, their phenotype in MM is not well‐established.Aims:To define the phenotype of G‐MDSCs based on the integration of phenotypic data, immunosuppressive potential, gene regulatory networks and impact on patients’ outcome of different granulocytic subsets present in the MM‐TIME.Methods:We used multidimensional flow cytometry (MFC) to evaluate the pre‐established phenotype of G‐MDSCs in bone marrow (BM) samples from healthy donors (HD) (n = 4) and MM patients (n = 5). We then used principal component analysis (PCA) to unbiasedly identify different granulocytic subsets in the MM‐TIME, and FACSorted them for in vitro experiments to analyze their immunosuppressive potential (n = 9) and for RNAseq to analyze the molecular profile of G‐MDSCs in MM (n = 5) vs HD (n = 5). The clinical significance of these subsets was determined comparing their numbers at diagnosis within the BM TIME of MM patients (n = 71) included in the GEM2012MENOS65 trial.Results:Human G‐MDSCs have been defined as a cell population displaying a CD11b‐CD14‐CD15+CD33+HLADR‐ phenotype, representing 1% of total BM nucleated cells in HD and nearly 25% in MM patients. However, we found that the percentage of cells with this phenotype was similar in the BM of HD and MM patients (median of 8% in both, P>.99). Using MFC and according to PCA, 3 granulocyte subsets were unbiasedly identified in the BM of HD and MM, based on homogeneous CD14‐CD15+CD33+HLADR‐ expression but differential reactivity against CD11b, CD13 and CD16: CD11b‐CD13lo/‐CD16‐, CD11b+CD13lo/‐CD16‐ and CD11b+CD13+CD16+. The percentage of all three subsets was similar (P>.5) between HD and MM. Afterwards, we used FACSorting to isolate (1) or deplete (2) individually each of the 3 subsets from MM‐BM and determine its immunosuppressive potential in 2 assays. In (1), we noted a significant decrease in T cell proliferation when these were stimulated with CD3/CD28 and in presence of the CD11b+CD13+CD16+ neutrophil subset (0.5‐fold, P = .01) but not of the CD11b‐CD13lo/‐CD16‐ and CD11b+CD13lo/‐CD16‐ subsets. In (2), we noted that the cytotoxic potential of T cells significantly increased in presence of the BCMAxCD3 bispecific antibody (P = .04) reaching its maximum with the depletion of the CD11b+CD13+CD16+ subset (4‐fold, P = .007). Additionally, RNAseq of the 3 subsets in HD and MM patients revealed that genes commonly associated with G‐MDSCs (e.g. PTGS2, TGFβ1) were specifically upregulated in the CD11b+CD13+CD16+ granulocytic subset. Finally, we analyzed the distribution of the 3 subsets in the BM of MM patients. Interestingly, we observed that patients with a high number of CD11b+CD13+CD16+ neutrophils at baseline had lower count of T cells and that patients with low ratio between CD11b+CD13+CD16+ and T cells had inferior 3‐year rates of progression‐free survival as compared to the remaining patients (100% vs 75%, P = .03).Summary/Conclusion:In this comprehensive analysis that integrated molecular, phenotypic, functional and clinical data, we determined a correlation between three well‐defined granulocytic subsets and their immunosuppressive potential, thereby providing for the first time, a set of optimal markers (CD11b/CD13/CD16) for monitoring G‐MDSCs in MM.