Controlled and sustained drug delivery using polymeric matrices has been widely applied to the fields of medicine and agriculture with impressive results. The release rate of a particular compound from a polymer matrix can be tailored by altering the composition, porosity, and size of the carrier. Efforts have been made to model the effects of these parameters, particularly polymer composition, on drug release in order to expedite the development of new formulation. To date, many models are either regressive in nature, require obscure physical constants or involve cumbersome Monte-Carlo methods. Recent advances in computing power and numerical methods have made finite element analysis a feasible tool for examining controlled release in biodegradable polymers. Using finite element analysis, a system of non-linear differential equations can be solved which are based on the diffusion-reaction events governing controlled release from a biodegradable polymer matrix.

A finite element model was developed to predict drug release using the fundamental equations for Fickian diffusion and hydrolysis kinetics as applied to each of the basic components in the system: a) the polymer particle, b) the water reservoir, and c) the releasable drug. Further, explicit continuous equations were developed to describe the diffusivities of the components as they transition between the polymeric and aqueous phases of this system. In addition to being affected by the bulk phase, the polymer diffusivity is also a function of its molecular weight. Localized changes in the average polymer molecular weight were governed by the hydrolysis rate expression. In such a manner, the diffusion-reaction processes account for the degradation and erosion of the polymer matrix in this model. Therefore, simply given the component diffusivities, densities and molecular weights, along with the degradation rate expression, the release behavior of a given polymer can be predicted.

A direct comparison with data shows the diffusion-reaction model can effectively predict protein release from polyanhydride (PA) microparticles. Similarly, release from polylactic acid (PLA) has also been predicted and the resulting release profile compares with experimental data. In addition to predicting drug release profiles, the model also describes physical changes occurring in the polymer matrix due to degradation and erosion. Erosion (or the dissociation of the polymer matrix) can be monitored, providing a profile of carrier mass loss over the course of the matrix lifespan. Degradation (or the rate of polymer hydrolysis) is shown progressing not only in the original polymer matrix, but also in released oligomers. Such detailed predictions of polymer matrix behavior can help guide the creation of more effective carrier formulations without cumbersome and time consuming experimental release assays.

In summary, a physically relevant model describing release from eroding, biodegradable polymer matrices was solved using finite element analysis. The model employs common, experimental parameters without relying on regressions or Monte-Carlo methods. Finally, these model parameters can be varied to investigate the release behavior of potential polymer matrices.