Analysis of the burst effect for small molecular weight solutes release from poly(vinyl alcohol) hydrogels

Controlled release of small molecular weight solutes from crosslinked poly(vinyl alcohol) (PVA) hydrogel matrices was investigated. Drugs were loaded at a known concentration during the hydrogel crosslinking procedure. The release profile was characterized by an initial rapid drug release, referred to as “burst effect,” followed by a sustained near-zero-order release. The burst effect was studied under different conditions, such as drug loading concentration, hydrogel crosslinking ratio, drug properties, and polymer aqueous solution concentration; influences of these factors on the burst effect were analyzed. A new parameter, degree of burst (DB) was defined to characterize the relative magnitude of the burst effect compared to the subsequent sustained release. It was found through the experimental studies that no simple design parameter could reduce the burst effect without decreasing the sustained release rate.; Two different techniques, surface extraction and surface preferential crosslinking, were investigated to minimize the burst effect. Both the methods involved changing the system design parameters on a partial basis.; Drug redistribution experiments suggested that the drug migration theory put forth in previous studies to account for the burst effect was not likely applicable in the systems investigated in this project.; A mathematical model based on a diffusion-convection model was studied to simulate the hydrogel swelling and drug release process. The solvent diffusion front velocity, which has been assumed constant in prior studies, was described by an exponential decay equation for the early stage for better simulation of the initial rapid solvent uptake. Since surface desorption of drug molecules is an important factor responsible for the burst effect, a surface desorption boundary condition was applied in the model to replace the perfect sink boundary that has been used in previous studies. According to the results of the drug redistribution experiments, an initial non-uniform drug concentration distribution, with the drug concentrated in the center, was applied and the sustained release stage that followed the burst effect was predicted more accurately. Effects of modeling parameters on the burst effect and the entire release profile were also investigated; results suggested that polymer characteristic relaxation time and the surface desorption constant had significant influences on the burst effect.