PII-171 - IDENTIFY BIOPREDICTIVE DISSOLUTION FOR PREDICTING IN VIVO PERFORMANCE OF A BCS II DRUG PRODUCT UNDER FED CONDITION

P. Du¹, L. Fang², L. Zhao², F. WU¹; ¹U.S. Food and Drug Administration, ²Division of Quantitative Methods and Modeling, Office of Research and Standards, Office of Generic Drugs, Center for Drug Evaluation and Research, U.S. Food and Drug Administration, Silver Spring, MD, USA.

Background: One of the challenges in physiologically based pharmacokinetic (PBPK) modeling is to identify biopredictive dissolution conditions and demonstrate the bio-discriminating capability/sensitivity of the PBPK model. The applicant developed a PBPK model to waive the in vivo fed bioequivalence (BE) study for generic Drug Tablets, which is a Biopharmaceutics Classification System (BCS) II drug containing amorphous solid dispersion (ASD) formulation. Initially, reviewers found that the developed PBPK model was not sensitive with dissolution input from dissolution quality control (QC) method. This research explored the methodology of incorporating dissolution data at different conditions into PBPK model and explored the possibility of using theoretical dissolution to evaluate the sensitivity of established PBPK model for detecting potential formulation differences.

Methods: A series of theoretical dissolution profiles for test products were generated by manually varying the dissolution rates based on the submitted dissolution profiles in acetate buffer pH 4.5 and in phosphate buffer pH 6.8. Gastroplus® was used to simulate in vivo PK of the varied dissolution profiles for test products using Z-factor approach under fed condition. The in vivo PK of reference products was also simulated by Gastroplus® using measured dissolution profiles. With the simulated PK generated, virtual bioequivalence was conducted and dissolution safe space was evaluated.

Results: Different Z-factors were calculated with dissolution profiles from different dissolution media. (Figure 1). It was found that Z-factors have positive relationship with dissolution. Parameter Sensitivity Analysis (PSA) from the developed PBPK modeling indicated that PK parameters were sensitive to Z-factor changes if Z-factor was low (e.g., Z-factor less than 1.00 x 10-3). Further, It was found that Z-factors have negative relationship with solubility.

Conclusion: In the current study, predicted in vivo PK parameters were affected by both solubilities and dissolution. For this case example with ASD formulation, it is a good practice to use measured solubilities and biopredictive dissolution as PBPK model inputs to establish dissolution safe space and support waiving in vivo fed BE study.