Mélanie Wilbaux1, Zhengqi Tan1, Joseph Grossman2, Gerald Fetterly2, Dhan Chand2, Amit Roy1
1Pumas-AI, inc., 2Agenus
Background: Optimizing dosing regimens for therapeutic antibodies through population pharmacokinetic modeling is critical for maximizing clinical benefit while maintaining practical administration schedules. Balstilimab, an IgG4 anti-PD-1 antibody, is being developed in combination with botensilimab, a novel multifunctional Fc-enhanced IgG1 anti-CTLA-4 antibody designed to optimize Fc?R co-engagement and enhance anti-tumor immunity against ‘cold’ tumors [1]. This combination has demonstrated clinical activity across multiple refractory solid tumors, including advanced microsatellite-stable colorectal cancer [2]. While the current reference regimen for balstilimab is 3 mg/kg administered every 2 weeks (Q2W), multiple dosing strategies have been evaluated, including both weight-based (Q2W and Q3W) and fixed-dose regimens, necessitating comprehensive pharmacokinetic analysis to optimize the dosing approach. Objectives: To develop a population pharmacokinetic model for balstilimab, alone or combined with botensilimab, in advanced solid tumors to characterize drug exposure and support the selection of optimal fixed and extended-interval dosing regimens. Methods: A population pharmacokinetic analysis was conducted using 6,339 balstilimab concentration measurements from 721 patients across five clinical studies (2 Phase I, 3 Phase II) in advanced solid tumors. Patients received balstilimab intravenously as monotherapy or with botensilimab, with doses ranging from 1-10 mg/kg Q2W or 6-10 mg/kg Q3W for weight-based dosing, and 240 mg Q2W or 300-450 mg Q3W for fixed dosing. Model development followed a stepwise approach: base model development, covariate model building including pre-specified parameter relationships, and backward elimination for the final model. Model selection criteria included Bayesian Information Criterion (BIC), condition number, parameter precision, shrinkage values, and diagnostic plots including prediction-corrected visual predictive checks. Model-based simulations were performed to evaluate fixed versus weight-based dosing and alternative dosing schedules. Analyses were performed using Pumas v2.5.1 on the JuliaHub platform [3]. Results: Balstilimab pharmacokinetics was best described by a linear 2-compartment model with time-varying clearance (CL), characterized by a sigmoid-Emax function [4]. Inter-individual variability was included on CL, central and peripheral volumes of distribution (Vc and Vp) and maximal change in CL (Emax), with correlation between CL, Vc and Vp. The base model incorporated body weight effects on both CL and Vc. After covariate evaluation, only albumin and age significantly influenced CL, with dose-normalized Cavg 13.7% lower in patients with albumin <3.5 g/dL and minimal age effect (geometric mean ratio for Cavg of 1.06 in patients =65 vs <65 years). Notable non-significant covariates included botensilimab co-administration, tumor type, anti-drug antibodies, sex, ECOG status, eGFR, and hepatic impairment. Model-based simulations demonstrated bioequivalence between fixed and weight-based dosing, with 240 mg Q2W showing 14% higher geometric mean exposure compared to 3 mg/kg Q2W (within 0.80-1.25 equivalence range). Extended interval dosing at 450 mg Q3W provided comparable exposure to 240 mg Q2W, with Cavg and Ctrough within 20%. While all evaluated regimens showed higher Cmax than the reference 3 mg/kg Q2W, values remained below those observed at 10 mg/kg Q2W, suggesting acceptable safety profiles. Conclusions: The population pharmacokinetic model characterized balstilimab disposition with time-varying clearance, consistent with observations for other monoclonal antibodies. The minimal impact of intrinsic and extrinsic factors, including botensilimab co-administration, supports a simplified dosing approach without need for adjustments. Model-based simulations demonstrated comparable exposure between weight-based and fixed dosing regimens, supporting conversion to more convenient fixed dosing of 240 mg Q2W or an alternative extended-interval regimen of 450 mg Q3W, according to the FDA guidance on pharmacokinetic-based criteria for supporting alternative dosing regimens of PD-1/PD-L1 blocking antibodies [5].
[1] Chand D. et al., Botensilimab, an Fc-Enhanced Anti-CTLA-4 Antibody, Is Effective against Tumors Poorly Responsive to Conventional Immunotherapy. Cancer Discov. 2024 Dec 2;14(12):2407-2429. [2] Bullock A. et al., Botensilimab plus Balstilimab in Relapsed/Refractory Microsatellite Stable Metastatic Colorectal Cancer: a Phase 1 Trial. Nat Med. 2024 Sep;30(9):2558-2567. [3] Rackauckas C. et al., Accelerated Predictive Healthcare Analytics with Pumas, a High Performance Pharmaceutical Modeling and Simulation Platform. BioRxiv. 2020;11.28.402297. [4] Bajaj G. et al., Model-Based Population Pharmacokinetic Analysis of Nivolumab in Patients With Solid Tumors: Model-Based Population Pharmacokinetic Analysis of Nivolumab. CPT Pharmacometrics Syst Pharmacol. 2017 Jan;6(1):58–66. [5] U.S. Food and Drug Administration, Pharmacokinetic-Based Criteria for Supporting Alternative Dosing Regimens of Programmed Cell Death Receptor-1 (PD-1) or Programmed Cell Death-Ligand 1 (PD-L1) Blocking Antibodies for Treatment of Patients with Cancer: Guidance for Industry. 2022 Dec.
Reference: PAGE 33 (2025) Abstr 11749 [www.page-meeting.org/?abstract=11749]
Poster: Drug/Disease Modelling - Oncology