Prediction of human pharmacokinetics (PK) is a critical part of candidate selection and the identification of compounds which have appropriate exposure levels in man. Frequently, human PK in early drug discovery is predicted using allometric scaling from a number of different species. Up to 27 animals are used per candidate, including rats, dogs and non-human primates.
By analysing data on the clearance of 74 compounds we have shown that human liver microsomes can be used to predict PK for compounds cleared by hepatic cytochrome P450 enzymes (as shown in Figure 2) as accurately as the non-human primate, and that the rat alone can be used for renally cleared compounds (data not shown). The human liver microsome data has been validated by Huntingdon Life Sciences in a study commissioned by the NC3Rs. This work has provided the basis for a predictive framework which allows compounds with the most desirable PK properties to be selected using in vitro methods alone or in vitro methods combined with a single species study in the rat.
Comparison of human liver microsomes and rat and non-human primate scaling methods for predicting human PK for compounds cleared by hepatic cytochrome P450 enzymes. Note the human liver microsome studies were conducted in two different laboratories for reproducibility purposes; two different methods were used for rat single species scaling (based on published data); and the non-human primate data is based on the hepatic blood flow method (using published data).
1 This method of calculation values both underprediction and overprediction in the same manner. A perfect prediction would exhibit a value of 1; 1 2-fold error (i.e 50% below or 100% above) would exhibit a value of 2.
Lave T., Chapman K., et al. (2009). Human clearance prediction: shifting the paradigm. Expert Opinion on Drug Metabolism & Toxicology 5(9): 1039-1048. doi:10.1517/17425250903099649
Beaumont K., Gardner I., Chapman K., et al. (2011).Towards an integrated human clearance prediction strategy that minimizes animal use. Journal of Pharmaceutical Sciences 100:1167–1783. doi:10.1002/jps.22635