Characterization of the Immune Impact of Daratumumab By Mass Cytometry in Multiple Myeloma

Background With the advent of antibodies targeted against CD38 the ability to determine which patients will respond and the mechanism by which such response is mediated is imperative for therapeutic success. Immune alterations in multiple myeloma (MM) are known to involve multiple cell types and, therefore we used mass cytometry to study a diverse set of single cells simultaneously. In this work we have studied the immune sequelae of anti-CD38 antibody therapies in a set of multiple myeloma patients.

Methods A series of 12 cases of MM were treated with daratumumab (DARA) in a number of disease settings including post-transplant; all had low tumor burden (bone marrow MM plasma cell range 0.001-9.09%). Of the 12 cases assessed, 2 had a partial response (PR) and 3 were deceased within 1 year after start of DARA treatment. Cryopreserved peripheral blood mononuclear cells (PBMC) from 6 healthy donors and 12 myeloma patients were analyzed pre-DARA and at days 28 and 56 after start of therapy.

A mass cytometry panel comprising 35 cell surface and 3 intracellular antibodies including cell lineage markers and functional markers was developed. Cells were stained with cell surface antibodies, washed and stained with cisplatin for live-dead discrimination, fixed and permeabilized with FOXP3 staining buffers, and stained with intracellular antibodies. Cells were fixed in PFA containing Iridium intercalator for cell identification, washed and resuspended in ddH2O containing 10% EQ beads and acquired on a Helios mass cytometer.

Data were normalized using beads and were transformed using the inverse hyperbolic sine function with a cofactor of 5 and gated for live, intact, singlets for global analysis by manual gating (FCS Express 6) and clustering by viSNE for visualization. Differences in population abundance were identified in an unbiased manner by CITRUS using 13 lineage markers for clustering. Analysis was performed using the multiple testing permutation procedure (SAM), with an FDR of 1% and minimum population size of 1%.

Results Manual gating of immune populations revealed that myeloma cases had significant immune perturbations compared to healthy donor PBMCs including loss of CD8⁺ effector memory, CD4⁺ T cells, and B cells. Comparison of pre-treatment MM PBMCs to post-DARA therapy showed a significant loss of activated NK cells, B regulatory cells, plasmacytoid dendritic cells, and plasma cells. Interestingly, a significant loss of T regulatory cells (CD4⁺CD25^hiCD127^loFOXP3⁺) was not observed in this analysis.

CITRUS identified a total of 26 significant clusters that differentiated the 3 timepoints. Based on the evaluation of phenotypic markers within the clusters, CITRUS identified 4 major cell types that were changed including CD8⁺ T cells (2 groups, 1 up 1 down post-DARA), NK cells (2 groups, reduced post-DARA), myeloid/monocytes (4 groups, 2 up, 2 down post-DARA) and B cells (reduced post-DARA). CD8⁺ T cells reduced post-DARA were distinguished by the presence of CD38, elevated CD57 and lack of HLA-DR compared to CD8 T cells that increased at D28 but returned to baseline levels by D56. The two NK cell groups were differentiated by the presence of CD16, representing activated cells. The 4 myeloid groups altered were characterized by the phenotype (CD14⁺CD11b⁺CD11c⁺CD33⁺), and the addition of CD38 in the two groups reduced post-DARA, which may represent myeloid derived suppressor cells and myeloid dendritic cells.

Another CITRUS analysis was performed with identical settings to examine changes in median expression of markers. Two groups of 5 clusters were identified based on CD55 or CD59 intensity, markers associated with complement inhibition. The CD55 group (4 clusters, CD56⁺CD16⁺) was upregulated at days 28 and 56 compared to pre-DARA. The CD59 group (1 cluster, CD38^hiCD45^lo) was reduced overall at day 28 and 56, however the cases were bimodal in terms of the extent of reduction. As this population most likely represents myeloma PCs, CD59 may be a surrogate for the presence of circulating tumor cells.

Conclusions

Significant changes in immune status post-DARA were identified by the comprehensive analysis of immune population abundance and marker intensity using mass cytometry. Significant changes in CD8⁺ T cells, NK cells, B cells, monocytes and in CD55 and CD59 intensity were observed post-DARA.