Development of a Physiologically Based Pharmacokinetic Model for Sinogliatin, a First-in-Class Glucokinase Activator, by Integrating Allometric Scaling, In Vitro to In Vivo Exploration and Steady-State Concentration–Mean Residence Time Methods: Mechanistic Understanding of its Pharmacokinetics

摘要

Aim The objective of this study was to develop a physiologically based pharmacokinetic (PBPK) model for sinogliatin (HMS-5552,?dorzagliatin) by integrating allometric scaling (AS), in vitro to in vivo exploration (IVIVE), and steady-state concentration–mean residence time ( C ~(ss)-MRT) methods and to provide mechanistic insight into its pharmacokinetic properties in humans. Methods Human major pharmacokinetic parameters were analyzed using AS, IVIVE, and C ~(ss)-MRT methods with available preclinical in vitro and in vivo data to understand sinogliatin drug metabolism and pharmacokinetic (DMPK) characteristics and underlying mechanisms. On this basis, an initial mechanistic PBPK model of sinogliatin was developed. The initial PBPK model was verified using observed data from a single ascending dose (SAD) study and further optimized with various strategies. The final model was validated by simulating sinogliatin pharmacokinetics under a fed condition. The validated model was applied to support a clinical drug–drug interaction (DDI) study design and to evaluate the effects of intrinsic (hepatic cirrhosis, genetic) factors on drug exposure. Results The two-species scaling method using rat and dog data (TS-~(rat,dog)) was the best AS method in predicting human systemic clearance in the central compartment (CL). The IVIVE method confirmed that sinogliatin was predominantly metabolized by cytochrome P450 (CYP) 3A4. The C ~(ss)-MRT method suggested dog pharmacokinetic profiles were more similar to human pharmacokinetic profiles. The estimated CL using the AS and IVIVE approaches was within 1.5-fold of that observed. The C ~(ss)-MRT method in dogs also provided acceptable prediction of human pharmacokinetic characteristics. For the PBPK approach, the 90% confidence intervals (CIs) of the simulated maximum concentration ( C ~(max)), CL, and area under the plasma concentration–time curve (AUC) of sinogliatin were within those observed and the 90% CI of simulated time to C ~(max)( t ~(max)) was closed to that observed for a dose range of 5–50 mg in the SAD study. The final PBPK model was validated by simulating sinogliatin pharmacokinetics with food. The 90% CIs of the simulated C ~(max), CL, and AUC values for sinogliatin were within those observed and the 90% CI of the simulated t ~(max)was partially within that observed for the dose range of 25–200?mg in the multiple ascending dose (MAD) study. This PBPK model selected a final clinical DDI study design with itraconazole from four potential designs and also evaluated the effects of intrinsic (hepatic cirrhosis, genetic) factors on drug exposure. Conclusions Sinogliatin pharmacokinetic properties were mechanistically understood by integrating all four methods and a mechanistic PBPK model was successfully developed and validated using clinical data. This PBPK model was applied to support the development of sinogliatin.