Quantitative structure-pharmacokinetic relationships for systemic drug distribution kinetics not confined to a congeneric series.

Many attempts have been made to describe quantitative structure-pharmacokinetic relationships within a congeneric series of drug molecules. The goal is to develop a predictive relationship that could predict in vivo results for other drugs within that series. These studies typically evaluate pharmacokinetic parameters that are reflective of both distribution and elimination processes. This work utilizes the results from 17 noncongeneric drugs reported in 18 pharmacokinetic studies. The objective was to determine if drug distribution parameters that were independent of elimination could be predicted from easily measured physicochemical parameters with a data base that included a wide variety of drugs that were not congeners of one another. Regression models utilizing a linear and a quadratic response surface were used to predict the various distribution parameters from physicochemical parameters, including molecular weight, intrinsic solubility, alcohol solubility, protein binding, and the distribution coefficient. Analogous to the extent of absorption, the extent of drug distribution can be predicted reasonably well by the probability that the drug will distribute into the peripheral system before being eliminated and by the volume of distribution at steady state. The duration of distribution, analogous to the rate of absorption, can be predicted by the mean transit time through the peripheral system the mean residence time of the drug in the peripheral system and the intrinsic mean residence time in the peripheral system. The ability to use statistical models to approximate drug distribution parameters without the constraints of working within a congeneric series provides some valuable opportunities.