|| Checking for direct PDF access through Ovid
Little is known about the underlying effects controlling in vitro-in vivo correlations (IVIVCs) for biodegradable controlled release microspheres. Most reports of IVIVCs that exist are empirical in nature, typically based on a mathematical relationship between in vitro and in vivo drug release, with the latter often estimated by deconvolution of pharmacokinetic data. In order to improve the ability of in vitro release tests to predict microsphere behavior in vivo and develop more meaningful IVIVCs, the in vivo release mechanisms need to be characterized. Here, two poly(lactic-co-glycolic acid) (PLGA) microsphere formulations encapsulating the model steroid triamcinolone acetonide (Tr-A) were implanted subcutaneously in rats by using a validated cage model, allowing for free fluid and cellular exchange and microsphere retrieval during release. Release kinetics, as well as mechanistic indicators of release such as hydrolysis and mass loss, was measured by direct analysis of the recovered microspheres. Release of Tr-A from both formulations was greatly accelerated in vivo compared to in vitro using agitated phosphate buffered saline + 0.02% Tween 80 pH 7.4, including rate of PLGA hydrolysis, mass loss and water uptake. Both microsphere formulations exhibited erosion-controlled release in vitro, indicated by similar polymer mass loss kinetics, but only one of the formulations (low molecular weight, free acid terminated) exhibited the same mechanism in vivo. The in vivo release of Tr-A from microspheres made of a higher molecular weight, ester end-capped PLGA displayed an osmotically induced/pore diffusion mechanism based on confocal micrographs of percolating pores in the polymer, not previously observed in vitro. This research indicates the need to fully understand the in vivo environment and how it causes drug release from biodegradable microspheres. This understanding can then be applied to develop in vitro release tests which better mimic this environment and cause drug release by the relevant mechanistic processes, ultimately leading to the development of mechanism based IVIVCs.