Long-acting injectable formulations based on poly(lactide-co-glycolide) (PLGA) have been commercialized for over 30 years in at least 20 FDA-approved products. These formulations offer several advantages, including reduced dosing frequency, improved patient compliance, and maintenance of therapeutic levels of drug. Despite extensive studies, the inherent complexity of the PLGA copolymer still poses significant challenges associated with the development of generic formulations having drug release profiles equivalent to those of the reference listed drugs. In addition, small changes to PLGA physicochemical properties or the drug product manufacturing process can have a major impact on the drug release profile of these long-acting formulations. This work seeks to better understand how variability in the physicochemical properties of similar PLGAs affects drug release from PLGA solid implants using Ozurdex (dexamethasone intravitreal implant) as the model system. Four 50:50, acid-terminated PLGAs of similar molecular weights were used to prepare four dexamethasone intravitreal implants structurally equivalent to Ozurdex. The PLGAs were extensively characterized by using a variety of analytical techniques prior to implant manufacture using a continuous, hot-melt extrusion process. In vitro release testing of the four structurally equivalent implants was performed in both normal saline and phosphate-buffered saline (PBS), yielding drastically different results between the two methods. In normal saline, no differences in the release profiles were observed. In PBS, the drug release profiles were sensitive to small changes in the residual monomer content, carboxylic acid end group content, and blockiness of the polymers. This finding further underscores the need for a physiologically relevant in vitro release testing method as part of a robust quality control strategy for PLGA-based solid implant formulations.
Keywords: Ozurdex; PLGA; acid number; blockiness; hot-melt extrusion; implant.