This study aimed to investigate the applicability of the Øie-Tozer model to predict human distribution volume (Vd) in the central compartment (V1 ), Vd at steady state (Vdss ), and Vd at beta phase (Vdβ ) based on animal Vd. Twenty compounds that have a human V1 /Vdss of 0.053-0.66 were selected from the literature. After intravenous administration of the compounds at 0.1 mg/kg to rats, dogs, and monkeys, plasma concentrations were determined, and pharmacokinetic parameters were obtained by one/two-compartmental analyses. The human V1 , Vdss , and Vdβ were predicted from animal Vd using the Øie-Tozer model, and the predictability was compared with that using proportionality and simple allometry. The Øie-Tozer model was the most reliable method for the overall prediction of Vd and applicable for accurately predicting human V1 , Vdss , and Vdβ (89%, 85%, and 68% of the compounds within a 3-fold error, respectively) when data of monkey for V1 and data of three animal species for Vdss and Vdβ were used. Additionally, the predicted human Vd with the two-compartment model was applicable for predicting pharmacokinetic profiles/parameters in humans after intravenous administration of 18 compounds [except for valproic acid (monophasic elimination profile) and chlorpromazine (deviation: Vdss < V1 )]. The prediction was more accurate than that using the predicted Vdss with the one-compartment model (e.g., underestimation of maximum plasma concentrations: 2 vs 8 compounds within a 3-fold error, respectively). In summary, the Øie-Tozer model was applicable for predicting human V1 , Vdss , and Vdβ , and their predicted Vd with the two-compartment model can lead to accurate pharmacokinetic prediction of compounds that show biphasic elimination.
Read full abstract