Abstract
BACKGROUND: There is no standardized validated experimental model used to predict human drug response, and the choice of model is not based on systematic evidence. Therefore, we decided to systematically investigate which models are currently used by selecting studies from literature that use prescribed inflammatory bowel disease medication as a positive control. METHODS: A search of PubMed was performed using terms describing experimental colitis models and the drugs used in corresponding clinical practice followed by the application of an animal filter. Each article was read and scored using a predesigned form describing the characteristics of the study (17 items), a quality assessment (10 items) completed by a meta-analysis. Results : One hundred ninety-four unique articles were included that met the selection criteria. A large heterogeneity was found regarding the characteristics of the animals used, induction methods, treatment protocol, and reporting quality. If categorized by colitis model only a small number of studies used a chronic model (10/194). Almost all use acute chemical models that investigate a response to epithelial damage, rather than chronic colitis. Fifty-six percent used a TNBS model and 20% used a dextran sodium sulfate model. In these models, an ameliorating effect of 5-ASA and corticosteroids was demonstrated and also a difference in outcome when male or female animals are used. CONCLUSIONS: This scope describes a huge heterogeneity in study designs for preclinical drug efficacy. In addition, more than three-quarters of the studies used an acute model irrelevant for testing new treatment options for inflammatory bowel disease.
