Prediction of pH dependent absorption using in vitro, in silico, and in vivo rat models: Early liability assessment during lead optimization.

Abstract Weakly basic compounds which have pHdependent solubility are liable to exhibit pHdependent absorption. In some cases, a subtle change in gastric pHcan significantly modulate the plasma concentration of the drug and can lead to sub-therapeutic exposure of the drug. Evaluating the risk of pHdependent absorption and potential drug–drug interactionwith pHmodulators are important aspects of drug discovery and development. In order to assess the risk around the extent of decrease in the systemic exposure of drugs co-administered with pHmodulators in the clinic, a pHeffect study is carried out, typically in higher species, mostly dog. The major limitation of a higher species pHeffect study is the resource and material requirement to assess this risk. Hence, these studies are mostly restricted to promising or advanced leads. In our current work, we have used in vitro aqueous solubility, in silicosimulations using GastroPlus™ and an in vivo rat pHeffect model to provide a qualitative assessment of the pHdependent absorption liability. Here, we evaluate ketoconazoleand atazanavirwith different pHdependent solubility profiles and based on in vitro , in silicoand in vivo results, a different extent of gastric pHeffect on absorption is predicted. The prediction is in alignment with higher species and human pHeffect study results. This in vitro , in silicoand in vivo (IVISIV) correlation is then extended to assess pHabsorption mitigation strategy. The IVISIV predicts pHdependent absorption for BMS-582949 whereas its solubility enhancing prodrug, BMS-751324 is predicted to mitigate this liability. Overall, the material requirement for this assessment is substantially low which makes this approach more practical to screen multiple compoundsduring leadoptimization.