Multiple myeloma (MM) is characterized by uncontrolled proliferation of clonal plasma cells, leading to myeloma-defining events (1, 2). Despite advancements in treatment, a substantial number of patients with MM experience relapse and develop resistance to conventional therapies, rendering the disease largely incurable (3, 4). With a deepening understanding of disease biology, innovative therapeutic approaches continue to emerge.

Bispecific antibodies (BsAbs) therapy is a novel approach that has shown potential in early phase trials for the treatment of relapsed or refractory MM (RRMM) (5, 6). BsAbs bind a target on both tumor cells and effector T-cells, which results in T-cell activation and thereby tumor cell apoptosis (7, 8). There are several BsAb formats, most of BsAbs used in MM target B-cell maturation antigen (BCMA), whereas others target non-BCMA antigens, including G protein-coupled receptor, class C group 5 member D (GPRC5D), Fc receptor-like protein 5 (FcRH5), and CD38 (9, 10).

In several ongoing trials, however, severe adverse events (AEs) have been observed with BsAbs, such as cytopenias, infections, cytokine release syndrome (CRS) and neurotoxicity (11, 12). To enhance clinical understanding of these therapies, we summarized pivotal data, including benefits and risks between BCMA-targeted and non-BCMA-targeted BsAbs in this study.

Patients with RRMM often encounter challenges associated with multiple lines of treatment and poor clinical outcomes, highlighting the imperative to investigate novel and effective therapeutic alternatives (29). BsAb recognize two epitopes or antigens and are among the most promising immunotherapeutic drugs for RRMM today (30, 31). In this first, large-scale systematic review and meta-analysis, we quantified the reported efficacy and safety of BsAbs in RRMM.

In this analysis, the pooled ORR for the entire cohort was 61%. Monotherapy trials in this population that led to the U.S. Food and Drug Administration (FDA) approval, including teclistamab, elranatamab and talquetamab, achieved 40% to 74% ORR and 5.1 month to NR mPFS (18, 24, 28, 32, 33). Importantly, these trials were at different stages of maturity or recruited different groups of patients, and therefore had variable response rates. In the long-term follow-up from the MajesTEC-1 study, the ORRs were 63% and 40% for cohort A (previous treatment with a BCMA-targeted therapy was not allowed) and cohort C (patients enrolled had prior exposure to anti-BCMA treatment), respectively (20, 28). Similarly, in the pivotal cohorts of the MonumenTAL-1 study, ORRs were consistent across subgroups (74% and 73%), but in the prior T-cell redirection cohort, the ORR was as low as 63% (24). Future recommendations may need to take into account these variables.

Our findings highlight that non-BCMA-targeted BsAbs treatment is associated with a significantly improved ORR for patients with RRMM. Notably, except for one study, patients enrolled in other non-BCMA-targeted BsAbs studies previously received anti-BCMA treatment, accounting for 20% to 33.8% (17, 19, 22, 27). In a phase 1 study of RRMM treated with 1-year, fixed-duration cevostamab, all four patients who were refractory to anti-BCMA treatment achieved a response (19). The non-BCMA-targeted BsAbs in this analysis included GPRC5D, FcRH5 and CD38. BCMA, or CD269, is ubiquitously present on the surface of plasma cells including MM cells (34, 35). Whereas GPRC5D, FcRH5 are preferentially expressed on multiple myeloma cells (36). Patients with RRMM who have received prior anti-BCMA therapies may also benefit from non-BCMA-targeted BsAbs therapy.

Among BCMA-targeted BsAbs, Moreau showed that the median PFS and duration of response (DOR) for teclistamab were 11.3 and 18.4 months, respectively (27); Lesokhin reported that fifteen-month DOR and PFS rates for elranatamab were 71.5% and 50.9%, respectively (17); Wong revealed that the median DOR in patients with IV Alnuctamab was 146.1 weeks (15). These studies suggested that BCMA-targeted BsAbs exhibited deep and durable responses in patients with RRMM. When it comes to non-BCMA-targeted BsAbs, talquetamab displayed varied PFS across different patient populations and dosage groups, with median PFS observed at 7.5, 11.9, and 5.1 months in the 0.4mg/kg QW, 0.8mg/kg Q2W, and prior T-cell redirection cohorts, respectively (23). It is noteworthy that the current data on PFS and DOR is limited, necessitating longer-term follow-ups to provide extensive information.

Common AEs of BsAbs therapy included CRS, infections and neutropenia. Cytopenias were mainly high-grade, which may lead to an increased risk of serious opportunistic infections, while CRS events were almost limited to low-grade. Compared with subcutaneous (SC) administration, intravenous (IV) administration was related to a higher incidence of CRS. In the phase 1 study of alnuctamab (ALNUC), any grade of CRS was reported in 89.5% of patients treated with IV ALNUC and 63% of patients treated with SC ALNUC (16, 21). Low-grade CRS was generally be treated with antipyretics, analgesics and corticosteroids (17, 24, 27). In addition, tocilizumab was highly effective and widely used for treating CRS and ICANS (17, 27, 37, 38). Van de Donk and colleagues found that the use of tocilizumab before teclistamab treatment appeared to reduce the incidence of CRS without new safety events or an impact on the response to teclistamab (39). ICANS, associated with a cytokine storm that allows high concentrations of cytokines to transit into the cerebrospinal fluid, usually occurred concurrently with or following CRS (40, 41). In this analysis, ICANS ranged from 2% to 11% at any grade and 0% to 2% at grade ≥ 3. However, Costa and colleagues reported that one patient treated with alnuctamab died in the study in the setting of CRS, with a potential infection as a contributing factor (21). In response to the serious concern of infections associated with BsAbs, a consensus recommendation from a panel of 13 global experts focused on infection monitoring, prophylaxis and treatment for patients with MM (42). Our analysis showed that any grade and grade ≥ 3 infections occurred in 48% (95%CI: 37%-59%) and 22% (95%CI: 19%-24%) of patients treated with BsAbs. The common infections were Covid-19, pneumonia and upper respiratory tract infection. Moreau and colleagues reported 12 deaths from Covid-19, and Lesokhin and colleagues reported 3 deaths from septic shock (18, 28). The infection risk factors in MM patients treated with BsAbs vary, such as dysfunction of the adaptive immune response, neutropenia, and the use of immunosuppressive agents (43–45). Future BsAb trials should take the incorporation of various prophylactic measures into consideration to prevent serious or even fatal infections. In addition, there are some similarities in the guidelines of BsAbs and CAR T-cell therapies, and a part of infection-related lessons can be drawn from CAR T-cell therapy (42, 46).

This study also presented significantly lower risks of neutropenia and lymphopenia among patients treated with non-BMCA-targeted BsAbs compared to those treated with BMCA-targeted BsAbs. Farah and colleagues thought that this was the result of the nuclear factor κB and c-Jun N-terminal kinase activation, which was caused by BCMA overexpression, and played a role in T-cell proliferation and cytokine release to increase the production of neutrophils (47, 48). The inhibition of BCMA expression led to a decrease in lymphocyte proliferation and neutrophil production. Moreover, a significant difference in the incidence of fatigue and pyrexia was observed between the two groups, which may be related to non-specific factors, such as disease characteristics and patient factors. Even without inclusion, skin-related events were observed in GPRC5D-targeted BsAbs therapy, with the most common events being exfoliation, pruritus, and dry skin. Despite their frequency, these events were primarily low-grade and responsive to both oral and topical glucocorticoid treatment (33).

While this study adhered to stringent selection and exclusion criteria, several limitations warrant consideration. Firstly, all enrolled studies were single-arm trials. Secondly, different targets (GPRC5D, FcRH5 and CD38) were classified as non-BCMA targets, and lastly, the drugs used varied among the studies. All of the above may cause bias. Although this study did not fulfill the above features completely, overall, the bias risk of study quality was acceptable.

In conclusion, BsAbs emerge as a promising therapeutic class for RRMM. This analysis indicated that opting for non-BCMA-targeted BsAbs therapy may present a more favorable treatment response and enhanced tolerability. On the contrary, BCMA-targeted BsAbs therapy appears to be associated with a heightened risk of ICANS. Our findings underscored the importance of carefully considering the choice of BsAbs therapy in clinical applications, with potential implications for optimizing patient outcomes and safety.

The original contributions presented in the study are included in the article/ Supplementary Material . Further inquiries can be directed to the corresponding author.

XW: Conceptualization, Data curation, Formal analysis, Investigation, Methodology, Writing – original draft. AZ: Conceptualization, Writing – review & editing. JZ: Writing – review & editing. TN: Funding acquisition, Writing – review & editing.