Formulation and Delivery
Siddhi Hate, PhD
Senior Advisor
Eli Lilly and Company
Carmel, Indiana
To view all Prologues, visit our library here and click "Enroll".
Physiologically based biopharmaceutics model (PBBM) can greatly aid in formulation selection at early stage of drug development and plays a critical role in clinical study design during pivotal drug development phases. However, due to lack of in vivo relevant information, PBBM predictions can have uncertainties. Integrating data from in vitro experiments significantly enhances confidence in model predictions. This is particularly important when predicting in vivo performance of a salt of weakly basic drug due to several reasons including poor understanding of the rate of solid form conversion, relevant solubility that drives the dissolution, the effect of counterion on surface or bulk pH and precipitation kinetics during gastrointestinal transit. In this study, a systematic in vitro performance assessment of an investigational weakly basic drug was carried out to inform PBBM, aiding in form selection and predictions of the effects of food and proton-pump inhibitors (PPIs). Bottom-up PBBM was first built based on traditional equilibrium solubility and two-stage dissolution, however, it showed a poor fit to the observed preclinical and clinical data. To improve model prediction and robustness, in vitro experiments were carefully designed to capture the in vivo relevance. Accurate measurement of salt solubility through intrinsic dissolution, along with two-stage dissolution under PPI stomach conditions, elucidated observed preclinical exposures in dogs when co-dosed with PPI. Notably, minimal PPI effect was observed for the salt compared to a substantial effect observed for the free base. Insights into precipitation kinetics were gained from additional biorelevant dissolution experiments, including two-stage dissolution and artificial stomach duodenum (ASD) experiments. It was interesting to observe differences in precipitation kinetics based on the buffer capacity of the intestinal media used in in vitro experiments. Fast precipitation was observed in conventional two-stage dissolution, however, ASD experiment with low buffer capacity media showed no precipitation, likely due to decrease in intestinal pH as a result of the amount and pKa of the counterion. The impact of the acidity of salt counterion on the intestinal pH, as a function of buffer capacity was noteworthy, particularly as recent publications discussing composition of human intestinal fluid have reported low buffer capacity in duodenum in fasted state. PBBM refined with in vitro data successfully predicted the drug exposure observed in the clinic wherein data suggested no precipitation in vivo. Furthermore, based on the model validation with preclinical and clinical data, model was able to accurately predict food effect and effect of co-administration with PPI observed in the clinic, which provided confidence in extrapolating food/PPI effect for wide range of doses. The study underscores critical value of in vivo relevant considerations in in vitro experiments to improve accuracy of in silico model predictions.