Proteomics-informed physiologically-based pharmacokinetic modelling for aldehyde oxidase substrates in colorectal cancer liver metastasis

Aldehyde oxidase (AO) metabolises a broad range of compounds, including the oxidation of anticancer kinase inhibitors. Quantitative proteomics data reported in the literature for liver microsomes from colorectal liver metastases (CRLM) indicate decreased expression of various drug-metabolising enzymes and transporters compared to healthy liver, although information on AO remains limited. This study aimed to investigate, for the first time, potential alterations<br/>in AO abundance in CRLM patients and prospectively predict the pharmacokinetics of six AO/dual AO-CYP substrates in CRLM patients via physiologically-based pharmacokinetic (PBPK) models. Global proteomic analysis (using High N approach) was performed on human liver cytosols (HLC) and microsomes (HLM) from 16 cancerous and matched histologically normal (adjacent to tumour) liver tissues from CRLM patients. Commercial fractions from<br/>healthy donors were analysed as controls. After the application of physiological scaling factors (cytosolic/microsomal protein per gram of liver), a strong rank correlation (rs=0.86, n=24) was observed in AO abundance between donor-matching HLC and HLM samples from CRLM patients. Inter-individual variability was the highest within cancerous livers (CV=95%; n=9). Total hepatic AO abundance (nmol/g liver) was 4- and 6.9-fold lower in cancerous tissues compared to histologically normal and healthy, respectively. Hepatic CYP3A4 and AO<br/>abundance correlated well (rs= 0.76, n=22) across patient samples. Substrate-specific decrease in drug clearance was predicted for all 6 compounds in patients with CRLM with the proteomics-informed PBPK models. The implications of varying levels of tumour burden were explored in the developed PBPK models to enable prospective PK predictions for AO substrates in the CRLM cancer population.