DEVELOPMENT OF IN VITRO MODELS FOR PREDICTING CUTANEOUS DRUG REACTIONS

Abstract

Severe cutaneous drug reactions can be life-threatening, but the existing in vitro models cannot adequately predict the drug reactions due to the complex underlying mechanism. Therefore, this thesis aims to develop in vitro systems that can predict the potential of drugs to induce cutaneous reactions. Firstly, an in vitro bioactivation model (HepaRG) and a functional skin sensitizing assay (U937 cells) was developed. Since the phenotypes of U937 cells were susceptible to different types of media, a compartmentalized liver immune coculture array was then designed to separate these media while allowing the exchange of metabolites. This platform can distinguish three skin sensitizing drugs from one non-skin sensitizing drugs. Lastly, a multi-culture array(MCA) was established by adding dermal cells (HaCaT and human dermal fibroblasts) with a quantitative prediction model using Boolean logic gates system, to classify large panels of drugs, with a sensitivity, specificity and accuracy of 71.43%, 57.14% and 81.81% respectively.