ESTABLISHMENT OF DOE-IVIVC MODEL FOR THE PHARMACOKINETIC PREDICTION OF RUXOLITINIB SUSTAINED-RELEASE DOSAGE FORMS

초록

This study aimed to demonstrate the utility of the in vitro-in vivo correlation (IVIVC) model combined with the design of experiment (DoE) to predict the in vivo pharmacokinetics of the sustained-release dosage form from its formulation composition. Ruxolitinib sustained-release formulations were designed by applying the DoE method and prepared by the wet granulation method. in vitro dissolution data were obtained using the paddle method. From in vitro dissolution profiles, the time of 80% dissolution (T80:) of each formulation was calculated, and the maximum dissolution rate (Vmax) was estimated by applying a modified MichaelisMenten equation. DoE model was developed for both dissolution parameters, i.e., T80 and Vmax. Both models exhibited good predictability (R2adj> 99.85%) and the average prediction error (%PE) was 4.81% for T80 and 1.70% for Vmax. To estimate in vivo dissolution of ruxolitinib SR formulations, in vivo pharmacokinetic profiles were obtained following oral administration of 3 different formulations to beagle dogs. The in vivo dissolution parameters were estimated by developing a population pharmacokinetic model. By correlating in vitro Vmax and the corresponding in vivo Vmax, a level A IVIVC was developed. Finally, the DoE-IVIVC model was developed by incorporating the mathematical equations of DoE and IVIVC models to predict in vivo pharmacokinetics from the formulation compositions. The average %PE was 7.90% and 9.43% for Cmax and AUC, respectively. External validation resulted in %PE of 14.27% for Cmax and 3.76% for AUC, demonstrating the acceptable predictability of the developed DoE-IVIVC model. The present DoE-IVIVC model could be used as a valuable tool for the development of SR formulations to achieve desired pharmacokinetic profiles.