Research Scientist II Takeda Development Corporation Americas
Subasumstat, an anti-tumor agent currently in Phase I/2 clinical trial is primarily eliminated by metabolism based on preclinical and in vitro human liver S9 data (CYP3A4 fm~0.8, aldehyde oxidase fm~0.2). In vitro, subasumstat exhibited extensive, concentration-dependent RBC partitioning across species. This concentration-dependent RBC-partitioning precludes bottom-up clearance prediction by IVIVE with static well-stirred model. We successfully used dynamic model incorporating concentration-dependent RBC-partitioning and in vitro human liver S9 metabolism data for bottom-up prediction of subasumstat human PK from 3 to 120 mg dose. We present a successful case study applying modeling to predict clinical PK of a compound with high and variable RBC-partitioning by IVIVE using in vitro and preclinical data.
Learning Objectives:
Appreciate the limitations of static IVIVE bottom-up clearance prediction when complicated scenarios such as concentration-dependent blood binding are present
Appreciate the utility of dynamic modeling in describing complicated PK data when blood binding changes with changing drug concentration (potential for extrapolation of learnings to plasma protein binding)
Develop insights on how to use preclinical data to develop confidence in modeling a novel PK phenomenon with the intent of later using this strategy for clinical translation
