J. Fan, X. Zhang, R. Lionberger; US Food and Drug Administration, Silver Spring, MD

BACKGROUND: The aim of this study was to investigate the impact of altered in vitro dissolution profile and pH dependent solubility on warfarin in vivo PK performance by mechanistic physiologically-based pharmacokinetic modeling and further to confirm whether it is necessary to develop a discriminating dissolution methodology at different pHs.
METHODS: Simulations were performed using the Simcyp Simulator and Gastroplus. The warfarin PBPK model was built based on physicochemical properties, such as solubility vs. pH dissolution profiles, formulation properties, such as particle size, or the in vitro dissolution profiles in various pH conditions (1.2, 4.5, and 6.8), and clinical observed plasma concentration-time data after oral administration of warfarin sodium in healthy subjects. The altered in vitro dissolution profiles due to the RH (room humidity) and thermal stress and various IPA levels and hypothetical various pH dependent solubility profiles were used in the simulation to evaluate their impact on warfarin in vivo PK performance.
RESULTS: Changes in dissolution profile due to the RH and thermal stress, low pH condition and IPA levels are less likely to result in the change in Cmax and AUC, provided dose loss did not occur. Slower dissolution in pH 6.8 may affect BE slightly. Slower dissolution rate and lower solubility in acidic pH caused a slightly delayed Tmax, which might not be of clinical importance.
CONCLUSION: Current models of absorption and dissolution indicate that warfarin exposure is robust against variation in the dissolution performance at lower pH as long as complete release occurs at higher pH. Dissolution in high pH (6.8) is more critical in determining PK profiles compared to dissolution in lower pH (1.2 and 4.5).