Methods- Mechanistic PBPK models for ertugliflozin, dapagliflozin and canagliflozin were developed using clinical intravenous and ADME data and in vitro CYP reaction phenotyping results with commercially available software (Simcyp V15, Certara). The modeling assumptions and parameters were verified using clinical data in healthy volunteers. The Simcyp rifampin PBPK compound file, which only incorporates CYP3A4/5 induction, was used without further modification in the simulated DDI studies.
Results- A 29% reduction in AUC of ertugliflozin with rifampin coadministration was predicted using PBPK modeling while a 39% reduction was observed in the ertugliflozin-rifampin clinical study. Since the rifampin PBPK compound file only incorporated CYP3A4/3A5 induction, these results suggest that the induction of the minor CYP3A4/3A5 pathways (~11% of overall clearance) by rifampin contributes to a majority of the observed DDI while induction of the major UGT1A9 pathway (~70% of overall clearance) by rifampin is relatively minor. Similar results were obtained by PBPK modeling of the DDI between rifampin and dapagliflozin and rifampin and canagliflozin.
Conclusion- PBPK modeling suggests that induction of the minor CYP pathway is a key contributor to the observed DDI between rifampin and SGLT2 inhibitors, which are predominantly metabolized by UGT enzymes.