Population-Based Modeling to Predict Human PK/PD of TAK-500, an Anti-CCR2 Antibody-Drug Conjugate for First-in-Human Study in Cancer Patients.
TAK-500 is a novel immune cell-directed antibody-drug conjugate (iADC) composed of TAK-202, an anti-CCR2 monoclonal antibody, conjugated to the STING agonist dazostinag (TAK-676), and is designed to stimulate antitumor immunity by reprogramming CCR2-positive monocytes. To support clinical translation, we developed a population pharmacokinetic (PK)-receptor occupancy (RO) model that leveraged both preclinical and clinical data from the parental antibody. In cynomolgus monkeys, TAK-500 displayed nonlinear, target-mediated disposition with dose-dependent CCR2 RO and monocyte redistribution. A 2-compartment PK model confirmed that TAK-500 total antibody and TAK-202 exhibited comparable PK profiles in monkeys, suggesting translational concordance. In humans, TAK-202 demonstrated nonlinear PK with greater-than-dose-proportional exposure, which was not captured by allometric scaling from monkeys but provided a critical basis for model development. We established a population PK-RO model of TAK-202 in humans and integrated it with cynomolgus TAK-500 data to project human PK and RO across doses of 50-500 μg/kg. Simulations (n = 1000) predicted dose‑dependent decreases in clearance, from 40 to 31 mL/h for total antibody and from 148 to 124 mL/h for conjugated payload (median values in a 70‑kg adult), respectively. Corresponding terminal half-lives were estimated at 114-118 h for total antibody and 30-60 h for conjugated payload. These findings demonstrate the feasibility of predicting TAK-500 PK and RO in humans by integrating preclinical and parental antibody clinical data, providing a quantitative framework for first-in-human dose selection of immune cell-directed ADCs. Trial Registration: Clinical trials: NCT04420884, NCT04879849.
Authors
Saravanakumar Saravanakumar, Deng Deng, Dong Dong, Haridas Haridas, Xia Xia, Appleman Appleman, Abu-Yousif Abu-Yousif, Piatkov Piatkov, Wang Wang
View on Pubmed