Development and validation of an in vitro human model of drug-induced vascular injury

Drug safety is a major cause of attrition in contemporary drug development. Many promising candidate drugs are terminated in the developmental process or withdrawn in the post-approval stage because of adverse drug reactions (ADRs). Among these, cardiovascular (CV) ADRs occur with a high incidence. Affecting the heart, blood vessels and blood components, CV ADRs can lead to hypertension, heart failure, cardiac arrhythmias and thrombosis. The prompt identification of CV toxicity would reduce both the delay and cost of drug development and increase patient safety. Drug-induced vascular injury (DIVI) in pre-clinical toxicology studies involves damage to endothelial (EC) and smooth muscle cells (SMC) in small and medium sized vessels, but little is known about the mechanisms of DIVI. Damage includes microhaemorrhage, endothelial junction disruption, SMC necrosis and inflammation. DIVI is often observed preclinically, driving a need to study this in human tissue. The aim of this thesis was to develop and characterise an in vitro human model that would allow the study of DIVI in human tissue during drug development. The first stage was to effectively reproduce DIVI in an animal model. To reduce the utilisation of live animals, a vessel explant was used. To determine whether DIVI was recapitulated ex vivo von Willebrand factor (VWF) release, one of the hallmarks of DIVI, was analysed using four drugs reported in the literature to produce DIVI. The same four drugs were then tested in human EC and SMC for the following features of DIVI: (1) VWF release; (2) EC monolayer disruption; (3) inflammation; (4) EC and SMC death. Both ex vivo in rats and in vitro in humans, the drugs did not exhibit common characteristics and each expressed DIVI features in different ways. The human in vitro model was also assessed in a co-culture of EC and SMC, and in the presence of a flow-mimetic that reproduces in vivo-derived hemodynamics. It was concluded that DIVI cannot be fully recapitulated in vitro in such a reductionist cellular model, suggesting that exposure to the complex microenvironment of the vasculature may be required for the lesions to develop, or that DIVI may, in fact, not be relevant to humans at all.Astra Zenec