The statistical mixture design for the optimization of the controlled release matrix systems

Phuriwat Leesawat, Purdue University

Abstract

This research was designed to develop oral controlled release matrix systems using a statistical mixture design. Mixtures of three cellulosic polymers were used to control the drug release. These polymers were hydroxypropyl methylcelluose, cellulose acetate phthalate, and ethylcellulose. The augmented simplex centroid design was employed to evaluated the effects of the polymer mixtures on the dependent responses. The contour plots of each response were drawn based on the equation given by the statistical fitted models. With the optimization of more than one criterion, a combined contour plot was made so that the optimum formulation to satisfy the overall goal was obtained. The formulation containing 33.33% of each polymer gave an optimal result. This final product met not only the requirements placed on it, but also the practical mass production criteria of process and product reproducibility. Moreover, it showed a good stability at ambient condition. Artificial neural networks were utilized for modeling and predicting the characteristics of the granule properties, the tablet properties, and the dissolution profiles. The artificial neural networks satisfactorily predicted all the granulation properties, the tablet properties, and the dissolution profiles in both the training and the testing step. The artificial neural network methodology showed to be an excellent modeling tool, in comparison to the traditional response surface methodology both for data fitting and prediction abilities. The granulation method was found to produce similar tablet properties and dissolution profiles, but had a significant effect on the granule properties. The drug loading had a high correlation to the intrinsic dissolution rate of the drug. The effects of the intrinsic dissolution rate were differences in a swelling matrix, an erosion matrix, and a porous matrix. The different model drugs had significant effect on the granule properties, tablet properties and dissolution profiles. The release of the given drugs from the matrix systems was influenced by the aqueous solubility of the drugs. The aqueous solubility is not the only factor governing the release of drug through the mixture systems. The aqueous pH of the model drugs was also a critical factor, which affected the release mechanisms of the mixture matrix systems.

Degree

Ph.D.

Advisors

Peck, Purdue University.

Subject Area

Pharmaceutical sciences

COinS