Application of simple mathematical expressions to relate half-lives of drugs in mice to those in humans.

The use of rat half-life data to predict human half-lives for drugs and other xenobiotics has previously been shown to be of value. Since transgenic and knockout mice are increasingly used in early stages of preclinical drug development, we wondered whether the estimation of half-life values in mice might be used in the same way that has been recommended for rats. A dataset of mouse and human half-life values was assembled for 88 drugs. Three different regression models were then applied to the prediction of human half-life values from mouse values. The results showed that none of the models was superior to any other, and that all models, including the simplest linear regression model, could predict a human half-life value from a mouse half-life value such that there would be an 80% chance that the predicted value would be within three-fold (i.e. between 0.33 times and 3 times) of the actual human value.