Predicting Clinical GI Safety Outcomes Using an In Vitro Human Intestinal Epithelial Model

The gastrointestinal (GI) epithelium is the body’s first line of defense to luminal contents such as food, microbiota, pharmaceuticals, and toxins. A small pool of proliferative stem and progenitor cells are responsible for maintaining and repairing the entire GI epithelium. Oncology chemotherapeutics target these rapidly dividing cells, contributing to adverse event risks, with 50-80% of patients developing diarrhea. Historically, these compounds have the lowest overall success rate for regulatory approval. A preclinical model for rapid screening of a candidate compound’s GI toxicity profile would provide a better safety profile prior to entering clinical trials. Although animal models, immortalized cell lines, and other complex in vitro systems exist for GI tox testing, RepliGut Planar, a simpler approach utilizing a monolayer of primary human epithelial cells, can more accurately predict GI toxicity.

Utilizing human colonic monolayers isolated from multiple donors, we investigated how chemotherapeutics affect the epithelium during proliferation and differentiation. Using an industry-backed test set of compounds, which included known positive and negative compounds for GI toxicity, each compound was tested to provide a dose-response curve. For proliferative cell testing, compounds were given to actively dividing cells and read-outs included resistance measurements indicative of barrier formation potential, total cell counts indicative of cell viability, and total EdU-positive cell counts indicative of monolayer proliferative capacity. For differentiated cell testing, compounds were given to non-dividing, cells and read-outs included barrier resistance measurements. Overall, the toxic concentration in the epithelial cell model were well-aligned with clinically relevant concentrations that produce toxicity (as measured by Cmax).