Existence of a flip-flop kinetic model for zidovudine (AZT) after oral administration

Using a three-way crossover experimental design, the effects of food and propantheline bromide on the pharmacokinetics of orally administered zidovudine (AZT; 3'-azido-3'-deoxythymidine) were assessed in six adult male Sprague-Dawley rats by the urinary excretion rate method. Accordingly, a single-10 mg/kg oral dose of AZT was administered as an aqueous solution to either fasting (F) rats, nonfasting (NF) rats or fasting rats pretreated with a 5 mg/kg oral dose of propantheline bromide (P; an inhibitor of GI motility). Quantitative urine collections were made at predetermined intervals for 24-32 hr after AZT administration, and each urine specimen was assayed for unmetabolized AZT by a sensitive and specific high-performance liquid chromatographic method. The mean extent of AZT absorption, as reflected by mean total urinary excretion values expressed as % of AZT dose, ranged from 72.9% to 75.2% and was unaffected by study condition. Kinetic interpretation of the terminal linear phase of the urinary excretion rate vs. time data yielded mean (+/- SD) half-lives of 4.35 +/- 2.1 hr for NF and 3.42 +/- 0.86 hr for P, which were significantly greater than the mean 1.58 +/- 0.63 hr half-life observed for F. The reported mean biological half-life of AZT after a 10 mg/kg intravenous dose to the same strain of rats is 0.76 +/- 0.35 hr (n = 6). After intravenous AZT administration, the terminal half-life reflects drug elimination. However, after oral AZT administration, the observed terminal half-lives reflect drug absorption rather than drug elimination (i.e., a flip-flop kinetic model is operable). There is some evidence to suggest that the disappearance of AZT from the blood of orally dosed AIDS patients may also be controlled by drug absorption.