Using non-animal methods to predict and avoid human adverse drug reactions

Many drugs cause undesired and unexpected human idiosyncratic human adverse reactions (IADRs). These are an important cause of serious human ill health, yet cannot be detected and predicted from nonclinical safety studies undertaken in animals.  IADRs are a major reason why drug development is so costly and inefficient and why many new drugs prove to be unsafe when used in large numbers of patients.  Research undertaken in the last few decades has shown that IADRs arise as a consequence of a combination of chemical insults caused by drugs to human tissues cells and patient-related susceptibility factors.  The patient-related factors are multi-factorial, involve both genetically determined and non-genetic aspects, and do not occur in non-susceptible humans or in experimental animals.  Hence human IADRs cannot be predicted from conventional toxicity studies, which are undertaken in non-susceptible animals.  One important chemical insult that can trigger IADRs is the formation of chemically reactive metabolites.  Others include potent cell cytotoxicity, impaired mitochondrial function and inhibition of bile salt excretion from hepatocytes by the bile salt export pump.  Recently, a multi-parametric and humanised in vitro assay panel has been described which enables the quantification of each of these processes.  When the data provided by the assay panel were integrated and adjusted to take account of in vivo human drug exposure,  highly specific and sensitive discrimination was achieved between drugs which caused severe IADRs and safe drugs (see: Chem Res Toxicol 2012, 25:1616; J Pharmacol Exp Ther 2015;352:281).  This highlights the need to use human based and mechanistically relevant in vitro non-animal methods to support selection of safe drugs during drug discovery, and to help underpin the human risk assessment process.