Use of in vitro-in vivo correlation to predict the pharmacokinetics of several products containing a BCS class 1 drug in extended release matrices.

Purpose: To determine if an IVIVC model can predict PK profiles of varying formulations of a BCS Class 1 drug that is a salt of a weak base.

Method: An IVIVC model (Level A) was created by correlating deconvoluted in vivo absorption data obtained from oral administration of 50 mg, 100 mg, and 200 mg fast and slow extended release formulations with in vitro percent dissolved using residual regression analysis. The model was then used to predict the in vivo profile of five test products that varied in formulation characteristics.

Results: The model passed internal validation for predicted Cmax and AUC. For external validation, in vitro data of five different test formulations was utilized. The model passed external validation for two test formulations that were different but belonging to the same release mechanism as that of the reference formulation. Three formulations failed external validation because they belonged to either a mixed or different release mechanism. The model and results were further confirmed using GatstroPlus™ simulation software.

Conclusions: These observations indicate that an IVIVC model for a BCS class I drug may be applicable to varying formulations if the principle of the drug release is similar.