An investigation into the biotransformation and enzymology of a methyl-imidazole containing cyclin dependent kinase inhibitor

Abstract

The cyclin dependent kinase inhibitor AZD5438 was in development by AstraZeneca for treatment of solid tumours where it showed a pattern of emerging toxicity loosely correlated with exposure during clinical trials in patients. Metabolites identified from clinical and preclinical studies recognised the possibility of reactive metabolite formation from AZD5438, possibly from an identified Phase II sulfate metabolite which could react via a carbocation intermediate. This research project used an array of in vitro techniques to investigate the reactive metabolite formation potential of AZD5438. The sulfate metabolite of interest and its precursor, a hydroxyl metabolite, were detected in clinical plasma samples. Both metabolites were also produced by in vitro incubations with human hepatocytes and active uptake of [14C]- AZD5438 into hepatocytes was observed. Optimisation of an in vitro system to selectively produce the sulfate metabolite of interest was unsuccessful. In vitro assays with human liver microsomes and human hepatocytes indicated that CYP2C9 and CYP3A4/5 were responsible for the production of the covalently bound metabolite of AZD5438. Binding values showed a negative trend as the complexity of the metabolic system increased, resulting in binding of 124.95, 57.10, 29.26, 5.14 pmol equivalents per mg protein for incubation of [14C]-AZD5438 (10 µM) to human liver microsomes, human liver microsomes with uridine diphosphate glucuronic acid (UDPGA), human liver S9 fraction and fresh human hepatocytes, respectively. The sulfate metabolite of interest was therefore, unlikely to be the source of reactive metabolite formation. In vitro toxicity assays with an endothelial cell line (THLE-CYP) showed parent AZD5438 could cause cell death, although whether further toxicity due to reactive metabolite formation occurred was not determined. Extrapolation of the microsomal covalent binding data using literature calculations could not conclusively evaluate risk of drug induced liver injury. However, the same calculation using values from the hepatocyte incubations predicted a lower comparative risk.