PII-178 - PROSPECTIVE APPLICATION OF PHYSIOLOGICALLY BASED PHARMACOKINETIC (PBPK) MODELING TO INFORM THE DESIGN OF A CLINICAL DRUG-DRUG INTERACTION (DDI) STUDY: CASE STUDY OF MEZIGDOMIDE

J. Burnett¹, C. Sychterz², Y. Zhu², F. Shakeel², K. Dingley², W. Chen², P. Rajadhyaksha³, A. Shahane², X. Wang², L. Gaohua², M. Lamba⁴, A. Gaudy²; ¹Bristol Myers Squibb, 556 Morris Ave, United States, ²Bristol Myers Squibb, ³University of Pittsburgh, ⁴Bristol Myers Squibb, Summit, NJ, United States.

Background: PBPK modeling approaches are increasingly used for efficient drug development. In the current application, a PBPK model was used to inform the design of a clinical DDI study of Mezigdomide (MEZI), a novel, highly potent Cereblon-E3 ligase modulating drug demonstrating promising clinical activity. In vitro studies demonstrated that MEZI is metabolized primarily by CYP3A4/5 and may cause inhibition of P-gp and BCRP in the gut.

Methods: PBPK modeling was used to predict the magnitude of potential DDI of MEZI as a victim of itraconazole (ITZ) and rifampin (RIF). Based on the projected magnitude of exposure changes, doses of MEZI were selected in a clinical DDI study (NCT05389722) in adult healthy subjects that were quantifiable, safe, and clinically relevant. Part 1 of the DDI study evaluated PK of MEZI alone (1 mg, n = 18) and in combination (on day 10, n = 15) with CYP3A4 inducer RIF (600 mg, days 1-13). Part 2 of the study evaluated the PK of MEZI alone (0.6 mg, n = 16) and in combination (0.2 mg on day 6, n = 16) with CYP3A inhibitor ITZ (200 mg, days 1-12). Part 3 evaluated the PK of digoxin (DIG) and rosuvastatin (ROS) (administered as a cocktail at 0.25 and 10 mg, respectively) alone (n = 15) or in combination with MEZI (1 mg, n = 13). Plasma concentrations were determined by a validated LC-MS assay.

Results: The table shows ratio of geometric means from PBPK model predictions and observed results of Cmax and AUC(INF) for Part 1 and 2. Coadministration of MEZI with DIG and ROS demonstrated no significant changes in exposure as predicted, with 90% confidence intervals of geometric mean ratio generally within limits of bioequivalence interval.

Conclusion: Prospective PBPK modeling was successfully used to predict DDI outcomes and inform appropriate dose selection for MEZI DDI study in healthy subjects. Observed results from the DDI study were well aligned with a priori predictions of the PBPK model and confirm its validity for predicting MEZI exposure changes with moderate CYP inhibitors and inducers.