ST indicates comparable distribution of CD16+ and CD16− NK cells across EMM samples. A and B, Estimated spatial cell type abundance of CD16+ and CD16− NK cells, PCs, and T cells in EMM_19 (left) and EMM_23 (right) samples by cell2location. The color scale represents the estimated relative abundance of each cell type in the corresponding spatial spot: Yellow shows higher abundance of the cell type.

Median cytotoxicity and exhaustion scores of CD8+ T cells

Summary of clinical characteristics of EMM patients in FC and scRNA cohort

Infection remains a leading cause of morbidity in multiple myeloma. Preventing infections is paramount and immune profiling could reflect the cumulative effect of host, tumor and treatment-related immunosuppression. However, current understanding of immune dysfunction and its association with infection is limited. To address this gap in knowledge and identify immune biomarkers of increased infection risk, we performed immune profiling using next-generation flow cytometry in bone marrow and peripheral blood samples from 1,786 patients at various disease stages and treatment scenarios. Patients developing infection had significantly lower percentages of CD27+ B cells and CD27- NK cells, as well as increased CD27-/CD27+ T-cell ratio in bone marrow. These immune risk factors were validated in three independent datasets. An immune score was developed to stratify patients with ≤1 vs ≥2 of the aforementioned risk factors, which was associated with higher infection incidence (35% vs 60%, P <.001). The immune score (odds ratio: 2.31, P <.001), disease stage and CD38, BCMA or GPRC5D targeted therapy were independently associated with infection incidence. All cell types detectable in bone marrow and peripheral blood were significantly correlated, suggesting that immune biomarkers of increased infection risk could be monitored using minimally-invasive methods that are available in routine laboratories.

Myeloid subclusters measure by FCM and scRNAseq

Increased proliferation in EMM cells

Comparison of scRNA-seq and FCM cohorts

Extramedullary multiple myeloma (EMM) is a high-risk feature of multiple myeloma associated with increased resistance to treatments, including modern immunotherapies, and shorter survival. The composition and functional state of immune cells within the EMM tumor microenvironment (TME) remain poorly understood. Using single-cell RNA sequencing, flow cytometry, and spatial transcriptomics, we revealed significant differences in the EMM TME compared with multiple myeloma bone marrow (BM). T and NK cells were verified as the most abundant immune subsets in the EMM TME. Compared with the BM counterparts, we found these tumors to have a significantly reduced effector-to-tumor cell ratio, a significantly lower number of CD4+ T cells, and an increased proportion of regulatory CD16- NK cells. We observed a high proportion of exhausted, tumor-reactive CD8+ T cells in roughly half of EMM tumors. Furthermore, we identified elevated expression of immune checkpoints, such as PD-1 on CD8+ T cells and KLRC1 (NKG2A) on CD16- NK cells.Significance: This study characterizes the TME in EMM lesions and paired BM from patients with multiple myeloma, revealing a high proportion of less cytotoxic CD16- NK cells in EMM tumors and suggesting that direct cell-cell interactions may underlie the CD8+ T-cell exhaustion observed in a subset of these tumors.

Summary of clinical characteristics of RRMM patients in FC and scRNA cohort

Figure S1 showing correlation between cell types measured by scRNAseq and FCM

Infections remain a key challenge during treatment of multiple myeloma (MM) patients with anti-BCMA and -GPRC5D bispecific antibodies (bsAbs). However, the underlying mechanism behind different rates and severity of infections induced by the two bsAbs remains poorly understood. Single-cell RNA-sequencing performed in bone marrow aspirates of 11 MM patients and 8 healthy donors revealed BCMA expression on mature B cells and, surprisingly, in small pre-B cells within B-cell precursors. By contrast, GPRC5D expression was restricted to normal and malignant plasma cells (PCs). Next-generation flow cytometry immune profiling showed that anti-BCMA bsAbs severely depleted bone marrow (BM) mature B cells (4.9%→0%; p<0.001) and normal PCs (0.17% → <0.0002%; p<0.001) during treatment of 62 relapsed MM patients. This was observed in early and late time points of therapy. Additional flow cytometry (N=31) and single-cell RNA-sequencing studies (N=8) demonstrated that, in contrast to anti-GPRC5D, anti-BCMA bsAbs also depleted immature and small pre-B cells. The MIcγ1 mouse model was used as a negative control of BCMA expression in all stages of the B-cell lineage, which confirmed no depletion of any B-cell subset after anti-BCMA treatment. In conclusion, we show that while GPRC5D bsAbs selectively target PCs, anti-BCMA bsAbs target both PCs and B cells from the small pre-B stage onwards. Our study provides mechanistic insight into the increased infection risk with anti-BCMA therapy and lays a foundation for individualized bsAb strategies in MM. Moreover, dual targeting of B cells and PCs may have therapeutic potential in other B cell malignancies or autoimmune diseases.

Background: Anti-CD38 monoclonal antibodies (mAbs) have become a standard part of multiple myeloma (MM) therapy. Their activity is mostly mediated by NK cells through antibody-dependent cellular cytotoxicity (ADCC). Although the association of increased CD38+ NK cells with worse outcomes was demonstrated, the underlying mechanisms of resistance to anti-CD38 mAbs remain poorly understood. Aims: To identify immune effector cell subsets predictive of treatment efficacy and driving resistance to anti-CD38 mAbs. Methods: Bone marrow (BM) samples (N=111) of relapsed/refractory MM (RRMM) treated with anti-CD38-based regimens (CD38-IMID [N=64]; CD38-PI [N=33]; CD38-mono [N=14] were assessed at baseline (BASE; N=110) and progression (PD; N=23), using conventional EuroFlow 8-color MM panel, followed by high-dimensional spectral cytometryesigned for deep profiling of lymphoid subsets and their activation/exhaustion status. Subset proportions were expressed as percentage of lymphocytes, unless stated otherwise. Results: Patients were treated predominantly with daratumumab over isatuximab (92%, 82%, and 100% of CD38-IMID, CD38-PI, and CD38-mono regimens). Treatment groups differed by line of therapy (p <0.001), with CD38-IMID mainly used in line 2 (56%), CD38-PI in lines 2-3 (55%, 30%), and CD38-mono in line ≥4 (86%). Median progression-free survival (PFS) for the CD38-IMID group was 37 months, compared to CD38-PI (16 months; p = 0.012) and CD38-mono (3 months; p <0.001). First, RRMM BM (N=110) collected at BASE were assessed by the EuroFlow MM panel. Lymphocyte pool consisted of median 13.9% B (CD19+CD56-), 19.7% NK (CD19-CD56+), and 61.6% T cell (CD19-CD56-) lineage. 79.7% NK cells were CD38+, while only 29.2% T cells were CD38+. Elevated total NK cells were observed in the PD group (p=0.07). This difference was driven by increased CD38+ NK cell proportion (p <0.028). Lasso-penalized multivariable Cox regression adjusted for therapy group and all immune subsets identified higher percentages of CD38+ NK cells and mature B cells (CD19+CD38low/−CD81low) as significant predictors of PFS at BASE. Increased CD38+ NK cells were associated with worse PFS (HR 1.65, p=0.003), while higher levels of mature B cells correlated with improved PFS (HR 0.79, p=0.002). These findings remained significant also in uniform CD38-IMID group (p <0.007). To better characterize the immune landscape driving these associations, detailed subset distribution was explored using spectral cytometry at BASE (N=14). CD16-, CD16+CD57-, CD16+CD57+ cells represented median 1.68%, 4.26%, and 6.17% of lymphocytes. Comparing CD38+/- NK compartments, CD16+CD57− subset was significantly enriched in the CD38+ NK pool (p=0.006). Gating on 8 other key NK markers revealed that KLRB1+ (p=0.063) cells were also enriched in the CD38+ pool, whereas NKG2C+ cells were more frequent in the CD38- pool (p <0.001). To further investigate the impact of anti-CD38 mAbs on immune cells, paired BASE and PD samples were analyzed using conventional (N=23) and spectral (N=14) panels. Basic exploratory analysis using both methods showed that proportion of total T cells increased (p <0.017), while total B cells (p <0.012) and total NK cells (p=0.023) dropped at PD. Furthermore, both CD38+ NK and CD38+ T cells decreased (p <0.004). Finally, spectral cytometry was utilized in both BASE and PD to identify NK subsets most affected by anti-CD38 therapy. As expected, most reduced NK cells in lymphocytes were CD38+ (median drop, log2-fold change: 6.84%, 1.72; p=0.008), followed by CD16+CD57+ (5.85%; 1.31; p=0.023), KIR2DL+ (5.11%; 1.52; p=0.008), KLRB1+ (4.58%, 1.46; p=0.008), and CD16+CD57- cells (2.42%, 1.04; p=0.023). Conclusion: This study provides in-depth analysis of immune cells in anti-CD38 mAb-treated patients using conventional and spectral cytometry. Higher BASE proportion of CD38+ NK cells was associated with worse PFS, while higher levels of mature B cells predicted better outcomes. Post-therapy, CD38+ NK and CD38+ T cells, total NK cells, and total B cells decreased, while total T cells increased. Importantly, deeper profiling revealed CD16+CD57- and KLRB1+ cells among those enriched in the CD38+ NK pool at BASE. Post-treatment, CD16+CD57+, KIR2DL+, and KLRB1+ NK cells showed the greatest reduction. These results indicate a complex interplay of fratricide, resulting in preferential depletion of mature effector NK cell phenotypes by anti-CD38 therapy.