Pharmacokinetics
For a compound to be useful as a medicine, just being bioactive is not enough. It is increasingly recognised that knowing how it enters and is processed by the body is as critical as knowing its interaction with target sites. Which explains why pharmacokinetics now plays such a pivotal role in drug discovery, development and use.
Yet even today many promising drug candidates still fail to make it through human drug development owing to poor pharmacokinetic properties, such as inadequate absorption, poor stability, or excessive interaction with other drugs, while many clinical trials remain relatively uninformative, inefficient and expensive owing to poor design.
Predicting pharmacokinetics
The aim of this team is to improve prediction in all the steps between discovery and drug use, by moving from the current essentially descriptive approaches to more mechanistic modelling methodologies. Given its importance it is not surprising that this research is particularly well funded by industry, especially but not only through the Centre for Applied Pharmacokinetic Research, and also by Biotechnology and Biological Sciences Research Council (BBSRC), Department of Health and the EU. A lectureship in Mechanism-based Prediction has been funded by Pfizer.
The problem of predicting the likely pharmacokinetics of a compound in humans is being approached from several complementary angles, with some success. One is the development and validation of in-vitro tissue and cell-based systems to permit quantitative prediction of the individual processes in vivo. In this, the group continues to play a worldwide role in quantitative drug metabolism, and has highlighted the need to incorporate variability in expression of individual cytochrome P450 enzymes among human livers within the extrapolation process.
The group has characterised the kinetics of the major human enzyme CYP3A4, critical to prediction of drug-drug interactions, and was first to show that many properties of CYP3A4, one not peculiar to microsomal preparations, also occur in intact hepatocytes. This in-vitro approach is also applied to drug-efflux transporters, focusing on the intestine and the brain.
Whole body physiologically-based pharmacokinetic models
Another parallel approach has been the development of whole body physiologically-based pharmacokinetic models that form the natural vehicle for incorporating the individual processes. Here, the group has had success in scaling preclinical and in-vitro data to predict events in humans.
In addition, by incorporating biological variability and uncertainty into each of the components of the model, it has proved possible to make reasonably good predictions of the important likely variability in the pharmacokinetics of some drugs in humans. Such models also form the basis for predicting pharmacokinetics from physiochemical and structural properties, as the group has shown within a homologous series of small barbiturate molecules of increasingly lipophilicity.
Although this area is still in its infancy there is the tantalising promise of being able to make adequate predictions in silico.
Pharmacokinetic/pharmacodynamic modelling
As a third approach, supported by grants through an EU Biomed programme and from industry, the group continues to develop statistical modelling methodology for analysis of combined pharmacokinetic/pharmacodynamic data to improve clinical drug development and trial design, having developed a Bayesian method to estimate inter-occasional variability and devised methods to both handle categorical clinical response data and estimate pharmacokinetic parameters in patient populations.
Based on this combined approach there is every reason to believe that pharmacokinetic/pharmacodynamic modelling will become increasingly used as a predictive tool in drug selection and development.