It is imperative that new drugs demonstrate adequate pharmacokinetic properties, allowing an optimal safety margin and convenient dosing regimens in clinical practice, which then lead to better patient compliance. Such pharmacokinetic properties include suitable peak (maximum) plasma drug concentration (Cmax), area under the plasma concentration-time curve (AUC) and a suitable half-life t(½). The Cmax and t(½) following oral drug administration are functions of the oral clearance (CL/F) and apparent volume of distribution during the terminal phase by the oral route (V(z)/F), each of which may be predicted and combined to estimate Cmax and t(½). Allometric scaling is a widely used methodology in the pharmaceutical industry to predict human pharmacokinetic parameters such as clearance and volume of distribution. In our previous published work, we have evaluated the use of allometry for prediction of CL/F and AUC. In this paper we describe the evaluation of different allometric scaling approaches for the prediction of Cmax, V(z)/F and t(½) after oral drug administration in man.
Twenty-nine compounds developed at Janssen Research and Development (a division of Janssen Pharmaceutica NV), covering a wide range of physicochemical and pharmacokinetic properties, were selected. The Cmax following oral dosing of a compound was predicted using (i) simple allometry alone; (ii) simple allometry along with correction factors such as plasma protein binding (PPB), maximum life-span potential or brain weight (reverse rule of exponents, unbound Cmax approach); and (iii) an indirect approach using allometrically predicted CL/F and V(z)/F and absorption rate constant (k(a)). The k(a) was estimated from (i) in vivo pharmacokinetic experiments in preclinical species; and (ii) predicted effective permeability in man (P(eff)), using a Caco-2 permeability assay. The V(z)/F was predicted using allometric scaling with or without PPB correction. The t(½). was estimated from the allometrically predicted parameters CL/F and V(z)/F. Predictions were deemed adequate when errors were within a 2-fold range.
Cmax and t(½). could be predicted within a 2-fold error range for 59% and 66% of the tested compounds, respectively, using allometrically predicted CL/F and V(z)/F. The best predictions for Cmax were obtained when k(a) values were calculated from the Caco-2 permeability assay. The V(z)/F was predicted within a 2-fold error range for 72% of compounds when PPB correction was applied as the correction factor for scaling.
We conclude that (i) Cmax and t(½) are best predicted by indirect scaling approaches (using allometrically predicted CL/F and V(z)/F and accounting for k(a) derived from permeability assay); and (ii) the PPB is an important correction factor for the prediction of V(z)/F by using allometric scaling. Furthermore, additional work is warranted to understand the mechanisms governing the processes underlying determination of Cmax so that the empirical approaches can be fine-tuned further.