Prediction of Liquid-Liquid Phase Separation at the Dissolving Drug Salt Particle Surface.

During the dissolution of drug salt particles, liquid-liquid phase separation (LLPS) of a free form can occur within the unstirred water layer (UWL) of the particles (UWL-LLPS). Theoretically, UWL-LLPS occurs when the free form concentration at the salt particle surface (C0) exceeds the intrinsic LLPS concentration (S0LLPS) of the free form. In the present study, we attempted to predict UWL-LLPS based on the intrinsic physicochemical properties of drugs. Cyproheptadine hydrochloride (CPH-HCl), diclofenac sodium (DCF-Na), papaverine hydrochloride (PAP-HCl), and propafenone hydrochloride (PRF-HCl) were selected as model drug salts. The pH0 and C0 values at pHs 4.0-9.5 (citric acid, phosphoric acid, and boric acid, buffer capacity = ca. 4 mM/ΔpH) were calculated using the pKa, solubility product (Ksp), and diffusion coefficient (D) of a drug. S0LLPS was measured using the pH-shift method. UWL-LLPS was predicted to occur when C0 ≥ S0LLPS. The prediction result was then compared with UWL-LLPS observed at each pH by polarized light microscopy (PLM). The pH-LLPS concentration (SpHLLPS) profile of each drug was also measured. UWL-LLPS was approximately correctly predicted for CPH-HCl, DCF-Na, and PRF-HCl. However, UWL-LLPS was not observable when C0 was close to S0LLPS. Furthermore, UWL-LLPS was not accurately predicted in the case of PAP-HCl. The pH-SpHLLPS profile of PAP did not follow the Henderson-Hasselbalch equation, probably because of the formation of cationic aggregates. In conclusion, UWL-LLPS was approximately predictable for drug salts using their intrinsic physicochemical properties (Ksp, pKa, D, and S0LLPS), except for PAP-HCl.

Physiologically Based Pharmacokinetic/Pharmacodynamic Modeling to Predict the Impact of CYP2C9 Genetic Polymorphisms, Co-Medication and Formulation on the Pharmacokinetics and Pharmacodynamics of Flurbiprofen.

Physiologically based pharmacokinetic/pharmacodynamic (PBPK/PD) models can serve as a powerful framework for predicting the influence as well as the interaction of formulation, genetic polymorphism and co-medication on the pharmacokinetics and pharmacodynamics of drug substances. In this study, flurbiprofen, a potent non-steroid anti-inflammatory drug, was chosen as a model drug. Flurbiprofen has absolute bioavailability of ~95% and linear pharmacokinetics in the dose range of 50-300 mg. Its absorption is considered variable and complex, often associated with double peak phenomena, and its pharmacokinetics are characterized by high inter-subject variability, mainly due to its metabolism by the polymorphic CYP2C9 (fmCYP2C9 ≥ 0.71). In this study, by leveraging in vitro, in silico and in vivo data, an integrated PBPK/PD model with mechanistic absorption was developed and evaluated against clinical data from PK, PD, drug-drug and gene-drug interaction studies. The PBPK model successfully predicted (within 2-fold) 36 out of 38 observed concentration-time profiles of flurbiprofen as well as the CYP2C9 genetic effects after administration of different intravenous and oral dosage forms over a dose range of 40-300 mg in both Caucasian and Chinese healthy volunteers. All model predictions for Cmax, AUCinf and CL/F were within two-fold of their respective mean or geometric mean values, while 90% of the predictions of Cmax, 81% of the predictions of AUCinf and 74% of the predictions of Cl/F were within 1.25 fold. In addition, the drug-drug and drug-gene interactions were predicted within 1.5-fold of the observed interaction ratios (AUC, Cmax ratios). The validated PBPK model was further expanded by linking it to an inhibitory Emax model describing the analgesic efficacy of flurbiprofen and applying it to explore the effect of formulation and genetic polymorphisms on the onset and duration of pain relief. This comprehensive PBPK/PD analysis, along with a detailed translational biopharmaceutic framework including appropriately designed biorelevant in vitro experiments and in vitro-in vivo extrapolation, provided mechanistic insight on the impact of formulation and genetic variations, two major determinants of the population variability, on the PK/PD of flurbiprofen. Clinically relevant specifications and potential dose adjustments were also proposed. Overall, the present work highlights the value of a translational PBPK/PD approach, tailored to target populations and genotypes, as an approach towards achieving personalized medicine.

PBPK Absorption Modeling: Establishing the In Vitro-In Vivo Link-Industry Perspective.

The establishment of an in vitro-in vivo correlation (IVIVC) is considered the gold standard to establish in vivo relevance of a dissolution method and to utilize dissolution data in the context of regulatory bioequivalence questions, including the development of dissolution specifications. However, several recent publications, including industry surveys and reviews from regulatory agencies, have indicated a low success rate for IVIVCs, especially for immediate-release formulations. In recent years, the use of physiologically based pharmacokinetics (PBPK) and absorption modeling, as a tool to facilitate formulation development, has been attracting increased attention. This manuscript provides an industry perspective on the current challenges with establishing IVIVCs and the potential PBPK and absorption modeling offer to increase their impact. Case studies across both immediate-release and extended-release formulations from five pharmaceutical companies are utilized to demonstrate how physiologically based IVIVC (PB-IVIVC) may facilitate drug product understanding and to inform bioequivalence assessment and clinically relevant specifications. Finally, PB-IVIVC best practices and a strategy for model development and application are proposed.