Prediction of human PK remains a central goal of the drug discovery process. Given the importance of distribution to efficacy and duration of drug action, much effort has gone into predicting distribution. Among the various models used to predict drug distribution some have been purely theoretical, some have been experimental, or a hybrid of both. Most models center around three main approaches: (1) QSAR-type models, (2) in vitro dialysis or (3) allometric scaling.
As an example of the first approach, Lombardo et al. described a method for the prediction of volume of distribution in humans based on two experimentally determined physicochemical parameters, ElogD74  and the fraction of compound ionized at pH 7.4 (derived from pKa), and on the fraction of free drug in plasma ( fu) determined from protein binding data . The fraction unbound in tissues ( fut), was determined via a regression analysis from 64 compounds using the parameters described, and was then used to predict Vd via the Oie-Tozer equation :
Vd = Vp(1 + R)+fu Vp( Ve / Vp — Re) + Vnfu/fut, (2.20)
where the parameters Vp, VE, and R e are taken to be the plasma and extracellular fluid volumes and the ratio of extravascular to intravascular proteins, respectively, with corresponding values in human of 0.0436 and 0.151 L/kg body weight for Vp and VE, respectively and approximately 1.4 for R e. Accuracy of this method was determined using a test set of 14 compounds, and it was demonstrated that human Vd values could be predicted to within about two-fold of the actual value.
It has been proposed that distribution of unbound drug is similar across species and that species differences in Vd can be explained by differences in plasma protein binding, giving rise to estimation of Vd by allometric models. Such approaches are typified by the report of Obach et al. who estimated Vd in humans from Vd determined in dog corrected for the differences in plasma protein binding in man and dog :
Was this article helpful?