Multivariate modeling of encapsulation and release of an ionizable drug from polymer microspheres

In the formulation of polymer microspheres (MSs) loaded with verapamil hydrochloride (VRP), a low molecular weight ionizable drug, by W/O/W emulsification, the pH of the external aqueous phase proved to be a primary determinant of both IE and drug release behavior. Increasing the pH of the external aqueous phase enhanced IE (≈100% at pH 8.4). This was associated with a considerable increase in initial release rate at pH 1.2. Two multivariate methods, factorial analysis (FA) and artificial neural network (ANN), were used to investigate the impact of the combined effect of the external phase pH and other parameters (polymer concentration and initial drug load) on MS characteristics; IE, initial drug release, MS size and yield. FA indicated that the external aqueous phase pH affected all responses, with a particularly strong correlation with IE in addition to a combined synergistic effect with polymer concentration on MS size. ANN showed better internal and external predictive ability of responses compared to FA. The ANN model developed in the study can be successfully used for multivariate modeling of the encapsulation and release of VRP and similar drug salts from hydrophobic polymer MSs prepared by multiple emulsification in addition to other MS characteristics. © 2009 Wiley-Liss, Inc. and the American Pharmacists Association J Pharm Sci 98:4603-4615, 2009