RESUMO
The purpose of this study is to develop glipizide push-pull osmotic pump (PPOP) tablets by using a formulation design expert system and an artificial neural network (ANN). Firstly, the expert system for the formulation design of osmotic pump of poor water-soluble drug was employed to design the formulation of glipizide PPOP, taking the dissolution test results of Glucotrol XL as the goal. Then glipizide PPOP was prepared according to the designed formulations and the in vitro dissolution was carried out. And in vivo evaluation was carried out between the samples which were similar to Glucotrol XL and the Glucotrol XL in Beagle dogs. The range of the factors of formulation and procedure, which could influence the drug release, was optimized using artificial neural network. Finally, the design space was found. It was found that the target formulation which was similar to Glucotrol XL in dissolution test could be obtained in a short period by using the expert system. The samples which were similar to Glucotrol XL were bio-equivalent to the Glucotrol XL in Beagle dogs. The design space of the key parameter coating weight gain was 9.5%-12.0%. It could be concluded that a well controlled product of glipizide PPOP was developed since the dissolution test standard of our product was more strict than that of Glucotrol XL.