Infection burden with BCMA versus GPRC5D-targeted bispecific antibodies in multiple myeloma: A systematic review and meta-analysis.

Abstract

e19511 Background: The approved B-cell maturation antigen (BCMA) and G protein-coupled receptor class C group 5 member D (GPRC5D)-targeted bispecific antibodies (BsAbs) have demonstrated efficacy in relapsed/refractory multiple myeloma. However, infectious complications remain a major concern, and comparative class-level infection data remain incomplete. Methods: We searched PubMed/MEDLINE, EMBASE, and Scopus from inception through January 10, 2026, using predefined eligibility criteria to include clinical trials only of approved BsAbs (teclistamab, elranatamab, linvoseltamab, and talquetamab). Arm-level safety data were grouped by target class and regimen (monotherapy or combination). Monotherapy arms were pooled using random-effects meta-analysis of proportions. Endpoints were infections (any-grade, grade ≥3, opportunistic) and infection-related deaths. Heterogeneity was assessed using I² and τ². Indirect class comparisons used arm-level random-effects meta-regression on the logit scale. PROSPERO ID CRD420261288395. Results: Ten monotherapy arms were included (BCMA k=5, n=500; GPRC5D k=5, n=375). Any-grade infections were higher in BCMA monotherapy, showing 74.3% (95% CI 70.2–77.9; I²=0%) vs 62.1% (50.5–72.5; I²=57.7%) for GPRC5D. Grade ≥3 infections were 41.0% (32.3–50.4; I²=71.5%) vs 23.5% (14.1–36.6; I²=73.7%), respectively. Infection-related deaths were 8.6% (6.1–11.8; I²=22.9%) vs 2.0% (0.9–4.0; I²=0.0%), representing a 4-fold higher pooled estimate. Opportunistic infections were reported in 8 monotherapy arms (BCMA k=3; GPRC5D k=5); GPRC5D pooled incidence was 5.0% (3.2–7.9; I²=0%), whereas BCMA was summarized descriptively (9.1–11.1%). In arm-level meta-regression, GPRC5D was associated with lower odds of any-grade infections (β=−0.531, p=0.0038), grade ≥3 infections (β=−0.798, p=0.0248), opportunistic infections (β=−0.783, p=0.0092), and infection-related deaths (β=−1.526, p=0.00064). Exploratory analysis of combination-regimen arms (n=540) showed high infection rates, with any-grade infections of 84.5% (68.1–93.3; I²=89.5%), grade ≥3 infections of 41.8% (27.4–57.8; I²=87.4%), and infection-related deaths of 6.0% (3.3–10.5; I²=47.3%). Given the limited number of arms and substantial heterogeneity, these estimates should be interpreted as descriptive. Conclusions: Infections were frequent with severe and fatal events. Across pooled monotherapy estimates and arm-level meta-regression, BCMA vs GPRC5D-directed therapy showed higher infection burden. Separation was most consistent for infection-related deaths, while grade ≥3 infections showed greater heterogeneity. Despite heterogeneity, cross-trial limitations, and incomplete opportunistic infection reporting, these findings reinforce the need for careful monitoring and optimized supportive care and prophylaxis during treatment.