[Thesis]. Manchester, UK: The University of Manchester; 2018.
Predictive pharmacokinetics now forms a critical part of the drug discovery process.
However, metabolic data has been demonstrated to under-predict in vivo clearance,
while no large scale analysis has been performed for hepatic uptake data. The primary
aim of this thesis was therefore to investigate the utility of various clearance parameters
generated in hepatocellular assays for the prediction of in vivo clearance.
Large scale literature analyses were performed for uptake data in both rat and human
hepatocytes. In the rat, it was highlighted that over-prediction was the predominant
issue for suspension and media loss hepatocyte assays. Conversely, monolayer and SCH
assays suffered from under-prediction. However, in human hepatocytes under-prediction
was observed in all assay formats. Use of empirical scaling factors improved predictions
in both species, and are recommended for future use.
The media loss assay, a method described by Soars et al, was further developed
in rat hepatocytes through inclusion of transporter and metabolic enzyme inhibitors.
Using a two-compartment model, individual clearance parameters (CLmet, CLactive and
CLpassive) were estimated, and were also used to estimate binding and partitioning
terms (Kp, Kpu and fucell). IVIVE of data produced from this assay resulted in a lower
bias than had been noted from literature data. However, it was hypothesised that additional
clearance parameters could be used in a mechanistic approach to further improve predictions.
SCH assays were performed to generate estimates of uptake rates, as well as efflux
rates from both the sinusoidal and canalicular membranes. Combining clearance terms
from both the media loss and SCH assays using the CLint,total term led to less bias
when predicting in vivo clearance than observed using uptake or metabolism data alone.
Additionally, the use of empirical scaling factors identified from the literature
analysis led to further reduction in prediction bias. Future work must now focus on
the application of this research to human hepatocytes.
It is concluded that the work presented in this thesis provides evidence for the usefulness
of both uptake and extended clearance terms, in conjunction with empirical scaling
methods, for the prediction of in vivo clearance. Adaptation of the media loss assay
allowed the estimation of several key pharmacokinetic parameters. Although some of
these are not always useful in a quantitative fashion, they remain essential properties
of a compound that must be considered when predicting behaviour within the body.