In vivo evaluation of zidovudine (AZT)-loaded ethylcellulose microspheres after oral administration in beagle dogs.

The purpose of this study was to evaluate the in vivo performance of sustained-release zidovudine (AZT) microspheres after oral administration in Beagle dogs, and to establish an in vitro-in vivo correlation. Two AZT microsphere formulations as well as AZT powder were administered to four Beagle dogs. Plasma samples were analyzed by HPLC. The plasma concentration-time data was analyzed by both compartmental and noncompartmental pharmacokinetic analyses. Based on the calculated pharmacokinetic parameters, in vivo release profiles were simulated and compared with in vitro release profiles in three different release media. Significantly longer mean residence time (MRT) was observed after administration of the sustained-release microspheres compared with AZT powder. Significantly lower maximum (Cmax) concentration values and longer times to Cmax (tmax) values were also observed. Formulation I showed the longest MRT (4.4 h). AZT plasma concentration was maintained above the minimum effective concentration for approximately 10 h after administration of Formulation I. The relative bioavailability of the microsphere formulations with respect to AZT powder was not significantly different from 1. The in vitro release of the three formulations was slower in simulated gastric fluid compared with simulated intestinal fluid. The addition of enzymes and mucin to the release media significantly lowered the in vitro release rate of AZT from the microspheres formulations, but not from AZT powder. A good level of in vitro-in vivo correlation (Level A correlation) was achieved with a release medium that was composed of simulated gastric fluid with pepsin and mucin for 2 h followed by simulated intestinal fluid with pancreatin and mucin for 8 h. This in vitro model may be used to predict the in vivo release of AZT, in the further development of controlled-release AZT formulations.