BACKGROUND Elranatamab (ELRA) is a heterodimeric humanized full-length bispecific antibody comprised of one B-cell Maturation Antigen (BCMA) binding arm and one cluster of differentiation 3 (CD3) binding arm paired through hinge mutation technology. Results from the MagnetisMM-3 (NCT04649359) study indicated deep and durable responses in patients with relapsed or refractory multiple myeloma (RRMM). The current analysis was conducted a) to characterize the relationship between free (ie, unbound) and total (ie, unbound and bound to soluble BCMA - sBCMA) ELRA exposure and objective response rate (ORR), b) to evaluate the effect of potential covariates in the ELRA exposure-response (E-R) relationship with ORR, and c) to characterize the relationship between ELRA exposure and duration of response (DOR). METHODS Data from ELRA monotherapy from the MagnetisMM-1 (first-in-human Ph1, NCT03269136), MagnetisMM-2 (Japan Ph1, NCT04798586), MagnetisMM-3 (Ph2 with pivotal Cohort A in BCMA-naïve patients and Cohort B with BCMA-exposed patients), and MagnetisMM-9 (Ph1/2, NCT05014412) studies were pooled for analyses. The dataset included a wide dose range from 0.1 µg/kg IV to 1000 µg/kg / 76 mg SC weekly (QW) dosing. Free and total exposure metrics (C ave,Day28, ie, average ELRA concentration on Day 28) and (C ave,Event, ie, average ELRA concentration up to the time of first response or progression/end of treatment) were obtained using the individual post-hoc pharmacokinetic (PK) parameters from the population PK model 1. The E-R relationship for the efficacy endpoint ORR by Investigator and the effects of various baseline demographic and disease risk factors (eg, age, sex, baseline sBCMA level, cytogenetics, extramedullary disease [EMD], and number of prior lines of therapy) were characterized using binomial logistic regression method. A Cox proportional hazard model was used to quantify the effects of C ave,Day28 and C ave,Event and the reduction of the dosing intensity from QW to once every two weeks (Q2W) on DOR. The relationship between tertiles and DOR was also explored using Kaplan-Meier (KM) analysis. The analyses were performed with R software. RESULTS A significant E-R relationship was identified between both total and free ELRA C ave,Day28 and C ave,Event and ORR. Higher ELRA exposure was associated with higher ORR. For the total ELRA PK analysis, lower baseline sBCMA level was associated with higher ORR. For both total and free ELRA E-R relationships, number of prior lines of therapy ≤ 5 and absence of baseline EMD were associated with higher ORR. None of the other covariates were significant predictors of ORR. The models developed using either C ave,Day28 or C ave,Event identified same set of significant covariates. Figure 1 shows the results of free ELRA C ave,Day28 E-R analysis. The Cox proportional hazard analysis and KM plots of DOR by tertiles of exposure revealed no significant association between ELRA exposure, Q2W switch, and DOR. CONCLUSION The efficacy endpoint ORR demonstrated E-R relationships with total and free ELRA exposures. Patients with higher ELRA exposure, lower number of prior lines of therapy, , were more likely to achieve objective responses. Lower sBCMA was also a significant predictor of ORR in the total PK analysis. The findings supported the 76 mg QW ELRA, which is the highest tested dose in the MagnetisMM-1 study. This regimen is projected to provide meaningful clinical benefit in the majority of RRMM patients, including those with relatively high baseline sBCMA levels. The identified covariates are known prognostic factors for RRMM, yet the ORR achieved in these subgroups is clinically meaningful. There was no significant association between total or free ELRA exposures, Q2W switch and DOR, which further supports the maintenance of responses with Q2W switch for the RRMM population.