Combining 3D cell-based data with in silico prediction of in vivo tissue concentration to improve prediction of DILI

Although exposure within the plasma is an important determinant of drug-induced adverse effects, it is the tissue concentrations which ultimately determine the organ-specific toxicity. In this research, Cyprotex used PBPK computational modelling to extrapolate plasma Cmax concentrations to liver tissue Cmax for a set of 29 compounds (20 were known to cause DILI in vivo and 9 where negative for DILI). These compounds were then assessed in two in vitro 3D cell-based models for hepatotoxicity and the data normalised using the predicted liver tissue Cmax.

Hepatotoxicity (or drug-induced liver injury) remains the primary cause of compound failure in drug development. While many standard 2D assays can provide adequate understanding of key endpoints like cytotoxicity, glutathione depletion, and lysosomotropism, not all models are amenable to long-term dosing or exhibit important organotypic features such as canaliculation or albumin production. By contrast, certain 3D liver models cultured either from HepaRG cells or cryopreserved primary human hepatocytes display uniform size, shape, albumin production, functional bile canaliculi and CYP activity over extended dosing regimens (beyond 21 days in culture).

In the study, the 3D cell based models were exposed to the selected 29 compounds over a 14 day period before being analysed by confocal HCS. ATP content was also assessed. The data illustrate that the HepaRG spheroid model provided better sensitivity over the primary human liver microtissue model. Furthermore, by normalising the data for liver Cmax, the HepaRG model was able to predict 70% of the DILI-positive compounds.

This research was presented at Eurotox 2016. Download the poster.

This website is intended to assist industry participants, customers and employees to understand Cyprotex's global operations and ambitions.