Anti-BCMA immunotherapies deplete plasmacytoid dendritic cells and impair antiviral immunity in multiple myeloma

Abstract

Modern T-cell redirecting immunotherapy targeting B-cell maturation antigen (BCMA) and G protein–coupled receptor class C group 5 member D (GPRC5D) on malignant plasma cells (PCs) demonstrated remarkable efficacy in multiple myeloma (MM). Currently, approved therapies include bispecific antibodies (bsAbs) directed against BCMA—teclistamab (tec), elranatamab (elra), linvoseltamab, and against GPRC5D—talquetamab (tal),^1-4 as well as chimeric antigen receptor (CAR)-T cells targeting BCMA—ide-cel and cilta-cel.^{4, 5} Infections remain the most common and critical complication associated with their use, including severe and fatal cases. This increased risk is driven by a combination of factors, including T-cell exhaustion, neutropenia, lymphopenia, and hypogammaglobulinemia. The use of combination regimens may further increase the risk of infection.⁶ In this context, targeting BCMA was associated with significantly higher infection risk compared to GPRC5D due to on-target off-tumor killing of PCs and B cells.^7-9 In addition, viral infections represent a substantial proportion of overall infectious burden with their true incidence being probably underestimated by current standard of infection reporting.^10-12 Respiratory viral infections are frequent, dominantly low-grade complications, often recurring or persisting, that considerably impair the quality of life of MM patients and may serve as an initial trigger for more severe bacterial superinfections. Interestingly, breakthrough infections remain common despite the use of multiple anti-infective and prophylactic measures, including immunoglobulin replacement therapy,⁶ suggesting that additional unidentified mechanism may contribute to infection onset.

Among the primary immune components engaged in antiviral defense, dendritic cells (DCs) serve as essential co-activators and professional antigen-presenting cells (APCs). Although plasmacytoid DCs (pDCs) are less efficient APCs than myeloid/conventional dendritic cells (mDCs/cDCs), they play a central role in immune system activation against viruses through enormous Type I interferon (IFN) production, particularly IFN-α and IFN-β, following toll-like receptors (TLR7 and TLR9) engagement by viral nucleic acids.¹³ Notably, BCMA expression in pDCs was first reported as early as 2017.¹⁴ Here, we investigated whether anti-BCMA immunotherapy, including bsAbs and CAR-T cells, contributes to increased viral susceptibility by affecting the pDC compartment in MM patients.

Overall, 388 unique samples from 72 MM patients were assessed in this study in multiple cohorts (Table S1). Patient characteristics are summarized in Tables S2–S5. BCMA gene expression was analyzed using publicly available single-cell RNA sequencing (scRNA-seq) datasets from healthy donor (HD) bone marrow (BM) samples (the Broad Institute [N = 8] and Erasmus MC [N = 5]).^{9, 15} Furthermore, BCMA surface protein expression on pDCs and B-cell lineage including PCs was assessed using 11-color spectral cytometry panel in fresh BM samples of 15 newly diagnosed MM (NDMM) patients. The immunomonitoring part of the study included 373 fresh samples from 57 MM patients (tec: 114/15; elra: 53/10; tal: 100/19; and cilta-cel: 106/13) treated at University Hospital Ostrava and Hradec Kralove, Czech Republic. pDCs and B cells were analyzed by a high-dimensional spectral cytometry panel focused on lymphoid compartment analysis.¹⁶ mDCs were quantified using a 13-color spectral cytometry myeloid-focused panel. Peripheral blood (PB) samples were analyzed at D0, 7, 14, 21, 30, 60, 90, 180, 270, and 360. BM paired samples were assessed at baseline and during minimal residual disease (MRD) evaluation between D90 and D360. pDCs and B-cell proportions were reported as a percentage of lymphoid cells, mDCs as a percentage of leukocytes. The study was conducted in accordance with the principles of the Declaration of Helsinki, and all patients provided written informed consent before sample collection for correlative studies. Levels of IFN-α, IFN-γ, and TNF-α were assessed in culture supernatants using ELISA. Peripheral blood mononuclear cells (PBMCs) were isolated from 13 fresh PB MM patients treated with tal or tec (1–3 months on therapy) and cultured for 15 h unstimulated or stimulated with TLR9 agonist CpG-A oligonucleotide, mimicking virus stimulation. Comparisons between two independent groups were performed using the Mann–Whitney U test. Two-sided P-values < 0.05 were considered statistically significant. Kruskal–Wallis test was used to compare >2 variables in a single therapy group. Analyses were conducted in R version 4.2.3 (www.r-project.org) using the tidyverse and ggpubr packages. Additional methodology is provided in Supplementary Methods.

Initially, BCMA mRNA expression was evaluated across BM immune cell subsets including DCs and B-cell lineage using two publicly available scRNA-seq datasets: (1) Broad Institute (N = 8) HD BM and (2) Erasmus MC (N = 5) HD BM (Figures 1A,B and S5). Both datasets demonstrated detectable BCMA expression in pDCs at levels comparable to those observed in B cells, which was previously shown to be sufficient to trigger cell killing by anti-BCMA bsAbs.⁹ In contrast, BCMA expression in mDCs, as well as GPRC5D expression in both pDCs and mDCs, was negligible. Additionally, analysis of surface protein expression on pDCs and B-cell lineage, including precursor stages and PCs, was performed in 15 NDMM BM samples by spectral cytometry. The proportion of BCMA+ cells in the pDC pool was significantly higher compared to pro-B, immature B, and mature B cells (all P < 0.001) and similar to the pre-B-cell subset (Figure 1C). Analysis of median fluorescence intensity (MFI) showed no consistent statistical significance (Figure S6).

To investigate the effect of anti-BCMA bsAbs on DC subsets, PB and BM samples were analyzed using spectral cytometry. First, pDC levels were assessed longitudinally, together with B cells acting as reference, in a cohort of 257 PB samples from 44 MM patients treated with anti-BCMA and anti-GPRC5D bsAbs. pDCs represented a small fraction of lymphoid cells across all samples at D0 (median: 0.2%), in line with literature.¹⁷ Further analysis showed a rapid decline of pDCs immediately at D7 in the BCMA group (P = 0.011), followed by complete depletion at D14 (P < 0.001), which persisted till D360 (P < 0.001–0.008), in contrast to the GPRC5D group, where pDCs remained detectable throughout therapy. Comparable trends were observed when tec and elra were evaluated separately (Figure S7). Moreover, CD1c+ mDCs were analyzed in 133 PB samples from 43 patients, with no reduction detected at any time point between BCMA and GPRC5D groups (all P > 0.112). B-cell kinetics was nearly identical to pDCs (Figure 2A).

Next, to assess whether pDC depletion is a shared feature of anti-BCMA therapies including CAR-T cells, 100 PB samples from 13 MM patients treated with cilta-cel were analyzed. pDCs decreased at D7 compared to D0 (P = 0.039) and were absent at D14, D21, and D30. Recovery was observed from D60 (P = 0.009), becoming more evident at D90 compared to D30 (P = 0.001). B-cell kinetics followed the pDCs pattern (Figure 2B). The reappearance of pDCs and B cells is most probably due to short persistence of anti-BCMA CAR-T cells in the patient's body that allows for early regeneration of both cell subsets.^{4, 18} In contrast, analysis of 35 PB samples from the same cohort indicated that mDCs were not depleted at any timepoint.

To determine whether comparable findings were also present in the BM, aspirates from eight MM patients treated with tal or tec were analyzed at baseline and during MRD evaluation. As expected, pDCs were absent also in BM at the MRD timepoint in the BCMA group compared to GPRC5D (P = 0.037), while mDCs were detected in all samples. B cells were also depleted at MRD in the BCMA group (P = 0.012) (Figure 2C).

To further assess whether differences in pDC levels between BCMA and GPRC5D cohorts translate into altered antiviral capacity, a functional in vitro assay was used. Cytokine production was evaluated in fresh PBMCs isolated from 13 MM patients treated with tal (N = 6) and tec (N = 7). Specifically, IFN-α (pDCs; Type I IFN response), IFN-γ (lymphocytes; cellular antiviral immunity), and TNF-α (monocytes; innate inflammatory response) were quantified, with the latter two included as controls for broader immune activity. PBMCs were cultured unstimulated or stimulated with TLR9 agonist CpG-A oligonucleotide mimicking viral exposure. In controls, IFN-α was undetectable in both groups. Nevertheless, following CpG-A stimulation, robust IFN-α production was observed in GPRC5D samples, whereas production was lacking in BCMA samples (P = 0.016). Notably, levels of IFN-γ and TNF-α remained unchanged between groups and across stimulation conditions (Figure 2D).

Infections are the second most common cause of death in MM after the disease itself and remain a major clinical concern associated with BCMA-targeted immunotherapy.¹⁹ To further improve outcomes of this effective treatment, a thorough understanding of factors driving infections is crucial. Here, we describe mechanism contributing to increased susceptibility to viral infections in patients receiving anti-BCMA T-cell–redirecting therapy. We demonstrated that pDCs, key mediators of antiviral immunity and major producers of IFN-α,²⁰ are depleted following both anti-BCMA bsAb and CAR-T-cell therapy (unlike anti-GPRC5D) in PB and BM. This is explained by BCMA expression detectable at both the mRNA and protein levels in pDCs, as first reported by Schuh et al. in 2017¹⁴; however, as pDCs are a rare population, their depletion has remained underrecognized. Our observation was independently confirmed by Rade et al., who also demonstrated pDC depletion following anti-BCMA CAR-T-cell therapies using different methodology (scRNA-seq). These findings also highlight the potential of BCMA-targeted therapy in blastic pDC neoplasm.²¹ Furthermore, our in vitro functional assays demonstrated a reduced capacity to produce IFN-α in MM patients receiving anti-BCMA therapy compared with anti-GPRC5D therapy, suggesting a potential reduction in antiviral immune responses. In this context, monitoring B-cell and pDC recovery could potentially serve as a useful biomarker of immune competence and infection risk, particularly given the recovery observed after treatment discontinuation.⁹ Moreover, the findings suggest that GPRC5D-targeted therapy may represent a preferable initial option for patients with increased risk or severe infections.

Overall, the data are consistent with clinical experience showing that especially patients receiving continuous anti-BCMA bsAb therapy have a high burden of predominantly low-grade respiratory viral infections, which significantly impair quality of life and may predispose them to more severe complications. Indeed, depletion of pDCs represents a missing piece of the puzzle explaining the higher rate and severity of infections in patients receiving BCMA-targeted immunotherapy, which may ultimately translate into individualized anti-infective strategies, including novel therapeutic approaches.

Ondrej Venglar: Conceptualization; data curation; formal analysis; writing—original draft; writing—review and editing; investigation; funding acquisition. Eva Radova: Data curation; writing—original draft; writing—review and editing. Maxima Warmuzova: Data curation; methodology; formal analysis; writing—review and editing. Daniel Bilek: Data curation; formal analysis; writing—review and editing. Marketa Kudelkova: Data curation; formal analysis; methodology; writing—review and editing. Katerina Jasinkova: Methodology; data curation; writing—review and editing; formal analysis. Ludmila Muronova: Data curation; writing—review and editing. Lucie Broskevicova: Data curation; writing—review and editing. Veronika Kapustova: Methodology; writing—review and editing. Jan Vrana: Methodology; writing—review and editing. Tereza Novakova: Methodology; writing—review and editing; data curation. Katerina Karasova: Methodology; writing—review and editing; data curation. Michaela Hornakova: Writing—review and editing; data curation; methodology. Tereza Popkova: Methodology; writing—review and editing. Jana Mihalyova: Methodology; writing—review and editing. Hana Plonkova: Methodology; writing—review and editing. Michal Simicek: Writing—review and editing; funding acquisition. Jakub Radocha: Writing—review and editing. Tereza Sevcikova: Writing—review and editing; funding acquisition. David Zihala: Writing—review and editing; funding acquisition. Juli Bago: Writing—review and editing; conceptualization; investigation. Roman Hajek: Writing—review and editing; funding acquisition. Tomas Jelinek: Writing—review and editing; conceptualization; writing—original draft; investigation; funding acquisition.

T.J.: research funding from Sanofi and Janssen; honoraria and consultancy/advisory role for Pfizer, Janssen, Sanofi, GlaxoSmithKline, and Bristol Myers Squibb.

The study was conducted in accordance with the principles of the Declaration of Helsinki and was approved by the institutional ethics committee under number 413/2012.

This work was supported by the OPJAK SALVAGE project (number CZ.02.01.01/00/22_008/0004644) under the Ministry of Education, Youth and Sports of the Czech Republic with co-financing from European Union and the State Budget; LERCO project number CZ.10.03.01/00/22_003/0000003 from the Operational Programme Just Transition (Ministry of the Environment of the Czech Republic); the Czech Health Research Council (NU23-03-00374, NW26-03-00323); institutional support (MH CZ-DRO-FNOs/2022, MH CZ-DRO-FNOs/2023, MH CZ-DRO-FNOs/2024, 1/OR/2025, MH CZ-DRO-FNOs/2026); the Ministry of Education, Youth and Sports of the Czech Republic through the e-INFRA CZ (ID: 90254). T.J. was supported by the European Hematology Association Physician Scientist Research Grant RG-202409-06772.

The data that support the findings of this study are available from the corresponding author upon reasonable request.