A Semi-Mechanistic Population Pharmacokinetic (PopPK) Model of Linvoseltamab With Considerations of Disease-Related Factors in Patients (Pts) with Relapsed/refractory (RR) Multiple Myeloma (MM)

Objectives: Linvoseltamab, a BCMA×CD3 bispecific antibody, induced deep and durable responses with generally manageable safety in pts with heavily pretreated RRMM in the LINKER-MM1 study (NCT03761108)1. A PopPK model was developed using PK data from RRMM pts who received linvoseltamab monotherapy in the dose escalation (Phase [Ph] 1) and dose expansion (Ph2) phases of LINKER-MM1. Objectives: 1) describe linvoseltamab PK, accounting for relevant intrinsic and extrinsic covariates of exposure; 2) estimate post hoc exposure metrics to allow subgroup comparisons and conduct further safety/response analyses; 3) simulate linvoseltamab exposure at the proposed registrational dose regimen.Methods: A PopPK analysis was performed using nonlinear mixed-effects modeling methodology in NONMEM. Potential intrinsic and extrinsic covariates were assessed using a stepwise covariate modeling procedure. Linvoseltamab concentrations from 281 pts (282 pts dosed, 6995 quantifiable serum samples) were available from LINKER-MM1. Pts received step-up doses (1–32 mg) in Weeks (Wks) 1–2 and full doses once weekly (QW) from Wks 2–3 to Wk 16 (Ph1) or Wk 14 (Ph2), followed by Q2W dosing. Ph2 evaluated two dose levels, 50 and 200 mg. Pts in the 200 mg cohort in Ph2 switched to Q4W dosing after ≥24 wks, depending on depth of response.Results: A two-compartment model with parallel linear and nonlinear, target-mediated (Michaelis Menten) elimination described linvoseltamab concentrations well. Linear clearance (CL) was consistent with proteolytic CL reported for antibodies, and was influenced by time-varying albumin and IgG concentrations in serum. Target-mediated CL showed a relationship with time-varying total free light chain (FLC) concentration in serum. The decline in target-mediated CL was consistent with a reduction in tumor burden by linvoseltamab. Estimated total CL at the end of QW dosing (Wk 14) and at steady state with Q4W dosing in the 5/25/200 mg cohort was reduced by 49.6% and 30%, respectively, compared with baseline CL (0.676 L/day). After the last 200 mg QW dose, predicted time for linvoseltamab concentrations to decline to 3% of the median Cmax was ~11 wks. Due to presence of non-linear and time dependent CL, dose proportionality (during step-up and full dose regimens) of linvoseltamab concentrations differed over the duration of treatment. Covariate analysis showed that albumin, IgG, total FLC, sex, and baseline body weight had statistically significant effects on linvoseltamab exposure, with IgG and albumin having the greatest impact.Conclusions: This evaluation indicates the complexity of the PK of linvoseltamab over a wide dose range (1–800 mg) in RRMM pts, that varies with response. This model described linvoseltamab PK in the context of pt response but did not capture effect of response on exposure yet. A future update to the model should encompass response as a function of exposure and changes to exposure as a function of response.Citations: [1] Jagannath S, et al. AACR 2024, Presentation CT001