1. Browse Program
2. Posters/Exhibits/Break
3. Posters/Exhibits/Break
4. Presentation

Presentation Details

Mechanistic population PK/PD modeling of factor XI-targeting monoclonal antibodies REGN7508CAT and REGN9933A2 in healthy volunteers: A foundation for dose selection and clinical translation Oleg Milberg1, Hisham Abdallah1, Rachel Kudgus Lokken2, Robert Dingman1, Karoline A.Meagher1, Michael E.Burczynski1, Ethan Marin1, Aaron Kithcart1, David E.Gutstein1. 1Regeneron Pharmaceuticals, Inc., Tarrytown, NY, USA.2Allucent, Cary, NC, USA

Abstract

Background: Factor XI (FXI) inhibition is a promising antithrombotic strategy with potentially lower bleeding risk than standard anticoagulants. REGN7508^CAT and REGN9933^A2 are high-affinity, human monoclonal antibodies (mAbs) binding distinct FXI epitopes (catalytic and A2 domains, respectively) to block downstream intrinsic pathway activation. We developed quantitative pharmacokinetic (PK)/pharmacodynamic (PD) models in healthy volunteers (HVs) to characterize antibody disposition, FXI engagement, and activated partial thromboplastin time (aPTT) response, and to identify clinical subcutaneous (SC) regimens that maximize PD effect throughout dosing for both mAbs. Objectives: (1) Build mechanistic target-mediated drug disposition (TMDD) models for total and functional drug and aPTT for REGN7508^CAT and REGN9933^A2 in HVs; (2) quantify covariate effects on PK/PD; (3) run model-based simulations to identify SC regimens expected to maintain maximal PD effect in ≥95% of individuals under variable physiologic and pathologic conditions. Methods: Data were integrated from two first-in-human, single-ascending-dose studies in HVs (6:2 active:placebo). REGN7508^CAT was administered intravenously (IV, 5–250 mg; six cohorts) or SC (125–600 mg; four cohorts; NCT05603195); REGN9933^A2 was administered IV (3–300 mg; five cohorts) or SC (100 and 300 mg; NCT05102136). PK/PD modeling was conducted in Monolix 2024R1 using a bivalent TMDD framework accounting for free antibody, single-arm-bound functional complex, and fully-bound complex. Functional drug (free + single-arm-bound antibody) directly drove aPTT via a direct-response model. Allometric body-weight scaling was applied to PK; covariates were screened with the COnditional Sampling for Stepwise Approach based on Correlation (COSSAC) procedure. Simulations were performed in Simulx to compare biweekly (Q2W) and monthly (Q4W) SC regimens for REGN7508^CAT and evaluate Q2W for REGN9933^A2 under baseline settings and stress-test conditions (increased antibody clearance and reduced bioavailability), representing diverse patient populations and disease states. Results: Both TMDD models adequately captured nonlinear PK and aPTT dynamics following IV and SC dosing (Figure 1). REGN7508^CAT and REGN9933^A2 achieved dose-dependent aPTT prolongation (at sustained PD plateaus, with medians of ~3.2× for REGN7508^CAT and ~2.5× for REGN9933^A2 relative to baseline aPTT), consistent with target saturation and epitope binding. Model-predicted 90% intervals indicated that ≥95% of individuals would achieve maximal or near-maximal aPTT prolongation at ≥250–300 mg SC Q2W. Simulated dose–response relationships showed a durable PD effect and minimal rightward shift when clearance doubled, supporting regimen robustness under stress (Figure 2). Estimated antibody half-lives were ~23 days (REGN7508^CAT) and ~26 days (REGN9933^A2); free FXI half-life was ~50–80 hours. Age emerged as a positive covariate on EC50 for REGN7508^CAT, suggesting reduced sensitivity to catalytic domain inhibition with increasing age, with negligible impact on predicted dose response. Conclusions: Mechanistic TMDD modeling for REGN7508^CAT and REGN9933^A2 quantitatively linked exposure to aPTT response in HVs and supported 300 mg SC Q2W as the more robust regimen to maintain maximal PD effect in ≥95% of individuals. Robustness persisted under simulated increases in clearance and decreases in bioavailability. These HV models provide a mechanistic basis for dose selection across multiple indications, and translation to patient populations with variable physiology or disease-related covariates, including patients in post-operative venous thromboembolism and other future Phase 2/3 programs.

No part of this publication may be reproduced, distributed, or transmitted in any form or by any means, including photocopying, recording, or other electronic or mechanical methods, without the prior written permission of the author.