(T-083) Application of quantitative systems pharmacology modeling in reverse translation

Osman Chaudhary, PhD – Research Scientist, Clinical Pharmacology, Pharmacometrics & Bioanalysis, Bristol Myers Squibb; Nidhi Sharda, PhD – Associate Director, Clinical Pharmacology, Pharmacometrics & Bioanalysis, Bristol Myers Squibb; Eun Mi Hur, PhD – Scientific Director, Discovery Immunology, TRC Immunology and CV, Bristol Myers Squibb; Jung Min Kim, PhD – Senior Principle Scientist, Nonclinical Research & Discovery Biotherapeutics, Bristol Myers Squibb; Murli Krishna, PhD – Senior Principle Scientist, Clinical Pharmacology, Pharmacometrics & Bioanalysis, Bristol Myers Squibb; Eileen Samy, PhD – Senior Principle Scientist, Discovery Immunology, TRC Immunology and CV, Bristol Myers Squibb; Devang Shah, PhD – Associate Director, Pharmaceutical Candidate Optimization, Bristol Myers Squibb; Radhika Goenka, PhD – Senior Principle Scientist, Discovery Immunology, TRC Immunology and CV, Bristol Myers Squibb; Michael Doyle, PhD – Scientific Director, Nonclinical Research & Discovery Biotherapeutics, Bristol Myers Squibb; Brian Schmidt, PhD – Executive Director, Clinical Pharmacology, Pharmacometrics & Bioanalysis, Bristol Myers Squibb; Panagiota Foteinou, PhD – Director, Clinical Pharmacology, Pharmacometrics & Bioanalysis, Bristol Myers Squibb

Objectives: Preclinical studies provide valuable insights into the efficacy and safety of potential drug candidates. However, the translation of these findings to the clinic is often challenging due to disease complexity and heterogeneity. Reverse translation helps bridge this gap by validating preclinical findings in clinical settings and refining preclinical studies based on clinical observations. Quantitative Systems Pharmacology (QSP) modeling has emerged as a powerful tool to facilitate this process by integrating quantitative approaches with biological knowledge to predict the behavior of drugs in the target population. By incorporating data from clinical trials, QSP models can be further used to gain insights into the underlying mechanisms of drug action and better understand how the interplay between drugs and their cellular targets influences pharmacologic activity. Here we present a case study by applying a clinically calibrated and validated QSP/IL-2 receptor agonist framework to explore in silico the impact of drug design parameters on downstream PD selectivity.

Methods: A multiscale human QSP model was developed linking IL-2 target engagement to induced signaling and downstream responses. At the molecular level, the model represents explicitly the physicochemical interactions with IL-2 and its cognate receptors: Rα, Rβγ and Rαβγ. At the cellular level, the active complexes (IL-2Rβγ, IL-2Rαβγ) trigger the phosphorylation of STAT5 which serves as surrogate for IL-2 signaling activation and mediates the proliferation of key cell types. In terms of clinical calibration and validation, the model was benchmarked against its capability to reproduce exposure and PD responses as seen clinically in Phase 1 BMS-986326 study. Specifically, a robust virtual population (Vpop) was generated based on our in-house algorithm using the two higher dose cohort data as the calibration dataset while holding out the lower dose data for validation. In terms of quality of fit, the Vpop was assessed based on its ability to reproduce data distributions as seen in the clinical study across multiple endpoints and dosing interventions. The clinically validated QSP/BMS-986326 model was then applied in the context of reverse translation by understanding key molecular drivers on the projected PD selectivity.

Results: A mechanistic QSP model for IL-2 mediated responses has been built and clinically calibrated to relevant data. The expression of Rα and Rβγ across different cell types was determined and further used in model development. By modeling both target expression variability and binding, the model was applied to understand better the combined effect of differential IL-2R expression and drug binding affinities on key biomarker outcomes.

Conclusion: A robust multiscale QSP reverse translation strategy has been successfully implemented to understand better the complex cross talk of IL-2 mediated responses on the desired vs. undesired pharmacology balance.

Citations: NA