Allopurinol kinetics in humans as a means to assess liver function: comparison of different models.

To describe the mechanisms involved in allopurinol kinetics after intravenous injection in humans, a number of alternative computer-based biodynamic models were designed. Distribution processes were described with two-compartment as well as with three-compartment kinetics for both allopurinol and its metabolite oxipurinol. These two major physiological alternatives were combined with biochemical models assuming either competitive or tight-binding-complex inhibition kinetics. The four resulting basic models were evaluated (and successively improved) using sets of plasma allopurinol and oxipurinol concentration curves, measured after intravenous injection in healthy subjects and in patients with different degrees of liver function. A three-compartment model with tight-binding-complex inhibition was selected and used to analyze the 35 loading tests performed. One of the parameters estimated in this way, the fractional rate constant for transport of allopurinol from the central compartment to the metabolically active (liver) compartment (kA31), turned out to be a powerful discriminative parameter between a group of healthy subjects, a group of patients with slightly to moderately reduced overall liver function, and a group with severely reduced overall liver function [kA31(min-1) = 0.136 +/- 0.042 (mean +/- SD, n = 13), 0.072 +/- 0.024 (n = 13), and 0.025 +/- 0.015 (n = 8), respectively].