144: Population Pharmacokinetic Model for Tulisokibart in Healthy Participants

Statement of Purpose, Innovation or Hypothesis: To develop a population pharmacokinetic (PopPK) model for tulisokibart, a first-in-class monoclonal antibody directed against the membrane-bound and soluble tumor necrosis factor-like cytokine 1A (sTL1A) in healthy participants, and to evaluate the influence of covariates on tulisokibart PK variability. The final PopPK model serves as a simulation tool for the identification of the recommended phase 2 dose (P2RD) in patients with inflammatory bowel disease (IBD).

Description of Methods & Materials: The final integrated analysis dataset included 94 participants with 1,670 evaluable tulisokibart serum concentrations from two Phase 1 studies in healthy participants. Tulisokibart was administered either intravenously (IV) at 5-1,000 mg, 50- 500 mg every two weeks (Q2W) for a total of 3 doses, or subcutaneously (SC) at 240 mg. Free serum concentrations of tulisokibart and total plasma concentrations of sTL1A were quantified using a validated competitive ELISA method with a lower limit of quantification at 80 ng/mL for tuliokibart and 0.05 ng/mL for sTL1A. Mathematical modeling was performed using NONMEM. A stepwise PopPK model building approach was performed: 1) exploratory data analysis, 2) base structural model development and evaluation of covariate effects, and 3) model qualification. The final PopPK multivariate model was used to perform model-based simulations for the identification of the P2RD in IBD patients.

Data & Results: PK data indicated that tulisokibart exposures were dose proportional at doses ≥100 mg and non-dose proportional at lower doses ( < 100 mg). Hence, a 3-compartment model best described tulisokibart PK with 1) a depot compartment for the SC dosing with a first‑order rate constant (Ka) for absorption and a lag time (ALAG1) from depot compartment to 2) central compartment, where tulisokinard is cleared through a linear (CL) and nonlinear Michaelis-Menten (Km and Vmax) eliminations, and 3) a peripheral compartment where tulisokibart distributes via an intercompartment clearance (Q). The final PopPK model included estimates of interindividual variability (IIV) for CL, central volume (Vc), and Vmax while IIVs on peripheral volume (Vp), Q, ALAG1, Ka and Km were fixed to zero. No collinearity was observed between IIVs on CL, Vc, and Vmax. A combined additive plus proportional error model was used to describe the residual variability. Baseline body weight was found to be a statistically (but not clinically) significant covariate on CL and Vc, with typical CL and typical Vc predicted to increase with increasing body weight. Baseline albumin was a statistically significant covariate on CL, with typical CL predicted to decrease with increasing baseline albumin concentration. None of the remaining covariates tested (age, sex, race, sTL1A, renal function, and anti-tulisokibart antibody status) were identified as statistically significant. Prediction-corrected visual predictive checks showed that the median concentrations predicted by the final tulisokibart PK model tracked the median observed concentrations generally well across studies and dosing intervals.

Interpretation, Conclusion or Significance: The final PopPK model captured the observed tulisokibart concentration-time course concentrations in healthy participants reasonably well. Baseline body weight was found to impact tulisokibart CL and Vc, while baseline albumin concentration influenced tulisokibatd CL.

Disclosures:

Citations/References:

Additional Information/Authors: NA