Statistical experimental design based studies on placebo and mitoxantrone-loaded albumin microspheres

A new technique for the preparation of hydrophilic, nonaggregating placebo albumin microspheres was developed. The influence of albumin concentration, poloxamer concentration, pH of albumin solution, and dispersion energy on mean microsphere size was investigated using a central composite design. The microspheres exhibited mean particle sizes in the range from 15 to 69 mu m. Mitoxantrone-loaded microspheres were prepared by dissolving the drug in an albumin solution prior to dispersion in the coherent phase. Effects and interactions of four relevant process factors (type of dehydrating agent, amount of glutaraldehyde, amount of mitoxantrone, and stabilisation time) on drug incorporation and particle diameter were determined according to a factorial design. Microspheres with mitoxantrone content from 6.1 to 11.5% (w/w), drug entrapment efficiency from 58 to 91%, and mean particle sizes from 25 to 36 mu m were obtained. Variation of stabilisation time showed no influence on the response variables. Nonaggregating particles with a smooth surface and high drug entrapment efficiency were obtained using a mitoxantrone/albumin ratio of 1/10 (w/w). The drug content was increased by the application of butanol instead of acetone as the dehydrating solvent while simultaneously increasing the amount of glutaraldehyde. The drug release following first order kinetics was delayed by increasing the crosslinking degree. (C) 1998 Elsevier Science B.V. All rights reserved.