Towards prediction of in vivo exposure from a permeability perspective using a 96-well Caco-2 assay

In this study we systematically validated a reproducible 96-well Caco-2 assay via an extended test set of 93 marketed drugs consisting of diverse transport mechanisms and quantified by LC/MS/MS. We sought to investigate the predictive utility of the widely-accepted Caco-2 permeability assay while dealing with the impact of solubility and recovery on challenging discovery compounds. We observed that paracellular compounds could be flagged by absorptive permeability (logPapp(A-B) <-5.5 cm/s) and physicochemical property space (clogP <1). An examination of 8000 Novartis discovery compounds revealed 14% were rejected for testing based on low aqueous solubility, while 13% were subject to low recovery (<30%). Compound loss in the assay was examined using a comparison between cell monolayer and artificial membrane (PAMPA), while bovine serum albumin (0.5% BSA in both donor and acceptor compartments) was utilized to improve recovery in the assay. The caveat to using Vitamin E TPGS to reduce solubility drop-offs was evaluated. The second focus of this study was to investigate the advantages and limitations of the current 96-well Caco-2 screening assay for predicting in vivo exposure from the permeability perspective in the drug discovery stage. Caco-2 measurements for compounds with high aqueous solubility and low in vitro metabolic clearance were compared to 88 in vivo rat bioavailability studies. Despite the challenges posed by discovery compounds with sub-optimal physicochemical properties, Caco-2 data successfully projected low exposure. This platform set the stage for mechanistically evaluating subsets of compounds towards improving in vitro-in vivo correlations.