Hematotoxicity and immune deficits with bispecific antibodies: A systematic review and meta-analysis in lymphoma and multiple myeloma

Abstract

Abstract Introduction: Bispecific antibodies (BsAbs) are a powerful therapeutic modality in B-cell malignancies but are associated with distinct toxicities. Despite growing recognition of the clinical importance of immunosuppression and infections in patients receiving BsAbs, comprehensive data on hematologic toxicities and immune deficits remain limited. This meta-analysis addresses this gap by systematically analyzing their incidence following BsAbs and examining associations with non-relapse mortality (NRM). Methods: Following registration to PROSPERO (CRD42025639456), we performed a systematic PubMed search through March 2025. We included studies in adult patients with Non-Hodgkin Lymphoma (NHL) or Multiple Myeloma (MM) treated with FDA-approved or investigational BsAbs with fully published phase 1/2 data. Eligible studies reported ≥15 patients at a consistent BsAb dose. Non-efficacious escalation cohorts were excluded when possible. For overlapping cohorts, the report with longer follow-up and equal or greater hematotoxicity detail was prioritized. Cytopenia incidence was extracted by maximum CTCAE grade, with grade 3/4 events classified as severe. Random effects models were used to calculate pooled incidence rates and to perform subgroup analyses. Results: Of 322 screened reports, 33 studies (28 clinical trials and 5 real-world studies) were included, encompassing 37 cohorts (22 NHL, 15 MM) and 3,557 patients (NHL: 2,172, MM: 1,385). Median follow-up was 14.5 months. Among NHL cohorts, the BsAb distribution was: 7 glofitamab, 6 mosunetuzumab, 5 epcoritamab, and 4 odronextamab. Among MM cohorts, 6 received teclistamab, 3 talquetamab, 1 teclistamab and talquetamab, 2 elranatamab, 2 linvoseltamab and 1 etentamig (ABBV-383). HematotoxicityAcross all cohorts, the pooled incidence of anemia was 38.7%, with significantly higher rates in MM (49.2%) compared to NHL (30.7%, p=0.01). Grade 3/4 anemia occurred in 19.1% overall (MM 31.2% vs. NHL 12.4%, p<0.001).The pooled incidence of neutropenia was 41.3%, with MM patients showing higher rates (51.8%) than NHL (34.2%, p=0.01). Grade 3/4 neutropenia occurred in 34.0% overall, including 45.4% in MM and 27.1% in NHL (p<0.001).Thrombocytopenia affected 28.3% overall (MM 35.0% vs. NHL 23.6%, p=0.10); grade 3/4 events were more frequent in MM (21.2% vs. NHL 10.7%, p=0.006). Immune DeficitsThe pooled incidence rate of lymphopenia was 24.9%, with higher rates in MM compared to NHL (32.9% vs. 14.1%, p=0.02). Similarly, grade 3/4 lymphopenia affected 21.8% overall (MM: 28.8%, NHL: 13.9%, p=0.07). Notably, immunoglobulin (IG) substitution was more frequent in MM than NHL (37.6% vs. 12.5%, p=0.001), particularly with BCMA-targeting BsAbs vs. GPRC5D-targeting agents (54.2% vs. 10.9 %, p<0.001). Metaregression ModellingMultivariable meta-regression analyses confirmed MM as an independent predictor of hematotoxicity across endpoints. After adjusting for disease entity, follow-up, age, prior treatment lines, and study setting, MM remained independently associated with higher rates of anemia (any-grade p=0.01, severe p=0.01), neutropenia (any-grade p=0.07, severe p=0.04), thrombocytopenia (any-grade p=0.03, severe p=0.07) and any-grade lymphopenia (p=0.03).In a meta-regression analysis restricted to MM studies, GPRC5D-directed BsAbs were linked to significantly lower rates of anemia, neutropenia, lymphopenia (both any grade and severe), any-grade thrombocytopenia and IG substitution compared to BCMA-directed agents. For severe thrombocytopenia, a trend was observed (p=0.08).In a meta-regression analysis restricted to NHL studies, combination regimes were significantly associated with increased rates of severe anemia, neutropenia, thrombocytopenia (both any grade and severe).The pooled NRM estimates were 6.5% for MM and 5.5% for NHL. In MM, neutropenia (any grade and grade 3/4), grade 3/4 thrombocytopenia and grade 3/4 anemia were significantly associated with increased NRM in univariate analyses. These associations were not observed in NHL. Conclusions: Cytopenias affect a substantial proportion of patients treated with BsAbs, particularly in MM and in NHL with combination regimes. These findings support the need for systematic hematologic monitoring, IG surveillance and tailored pre-emptive strategies to mitigate infection risk.This study represents the first and most comprehensive meta-analysis of hematotoxicity and immune deficits with BsAbs, establishing a benchmark across clinical settings.