| dc.description.abstract | Extending the release of therapeutic proteins is an area of study that has received a lot of attention of late due to the increasing use of antibodies in treating cancer, autoimmune diseases, viral infections, and even asthma. In many cases, long-term, local delivery systems are preferred over parenteral routes in order to reduce systemic exposure and avoid frequent injections. Hydrogels offer a promising potential for many drug delivery applications because they can be formed into a variety of shapes and sizes using biocompatible materials, while at the same time exhibiting a wide range of permeabilities. However, one of the challenges of developing an effective hydrogel delivery device is in controlling the diffusion of proteins within the hydrogel in order to provide the correct release rate and profile. The objective of this dissertation, therefore, was to investigate the permeability of therapeutic proteins in hydrogels to aid the development of long-term drug delivery vehicles. The work presented in this dissertation includes the study of protein partitioning in hydrogels, the development and modeling of a prototype release system, and the advancement of a hydrogel crosslinking method in an effort to expand the use of hydrogels for extended release applications. One particular area in which a long-term delivery system is needed is ocular drug delivery, where retinal diseases such as age related macular degeneration and diabetic macular edema require frequent injections into the vitreous in order to deliver drugs to the retina and prevent major vision loss. These diseases affect an estimated 16 million Americans and are the leading cause of blindness and vision loss in people over the age of 50. A hydrogel implant that could deliver sustained release to the retina would offer a number of advantages including convenience, safety, and financial benefits when compared to the current intravitreal dosing regimen. One of the difficulties in developing a hydrogel delivery device is achieving a significant protein concentration within the hydrogel. A method for loading proteins into monolithic hydrogels using the thermodynamic principles of aqueous two-phase extraction was investigated. By using a variety of polymers and partitioning salts, significant increases were seen in the partitioning of ovalbumin and IgG in PEG and dextran hydrogels. The results demonstrate the versatility of a method to overcome size exclusion of proteins, even as large as monoclonal antibodies, into many types of hydrogels, thus opening the door to new delivery strategies for therapeutic proteins using hydrogels. A successful hydrogel delivery device must be small enough to be implanted in the eye, yet carry a large enough drug load to achieve long-term release. Therefore, hollow mini cylinders were chosen as the structure for the intravitreal implants instead of a monolithic device. Achieving these goals requires a hydrogel with a very low diffusion coefficient (values below 10-10 cm2/s), an area of study that has received little attention in the literature yet has major implications in situations where long term, local delivery systems are needed to deliver therapeutic proteins such as ocular delivery. The swelling degrees required to achieve this target diffusion coefficient in a hydrogel were predicted using a free volume theory for protein diffusion in hydrogels, and were estimated to be in the range of 4-6 (g/g). In order to achieve such low swelling degrees, the cylinders were made from HA hydrogels that were highly crosslinked using a base-catalyzed Michael addition and the difunctional crosslinker divinyl sulfone (DVS). HA-DVS hydrogels with swelling degrees as low as 2.7 were achieved using HA concentrations that ranged from 15-30% (w/w) and HA:DVS ratios that ranged from 3:1–1:1. Using a custom mold, prototype cylinders were developed and proved capable of successfully releasing an antigen-binding fragment (Fab) for over 4 months in vitro with a maximum release rate of 4 micrograms a day. The results showed that a hollow cylinder made from hydrogels can achieve 3-6 months release of anti-vascular endothelial growth factors to the retina. In an effort to determine the target diffusion coefficients and cylinder dimensions required for ocular delivery the release of proteins from hollow cylinders was modeled using the physics software COMSOL. The largest size cylinders that were studied (1mm outer diameter) were able to load the largest amount of drug and provide the longest release. These cylinders were capable of delivering over 1 mg of drug while achieving a release rate of 2.5 micrograms a day for over 4.5 months. Moving to smaller cylinders makes the insertion process faster, safer, and less painful, yet it comes at the expense of smaller drug loading and shorter release times. The results indicated that the small dimensions required for an intravitreal implant leads to a narrow range of diffusion coefficients (1-3 x 10-11 cm2/s) that are capable of producing an effective delivery device, highlighting the importance of being able to tune the diffusion coefficient of hydrogels. Being able to control the network structure and the diffusion coefficient of a hydrogel was the main motivation for modifying HA with pentenoic anhydride and crosslinking it into hydrogels using the dithiol crosslinker dithiothreitol via a photoinitiated thiol-ene reaction. This crosslinking reaction is hypothesized to allow for greater control of the hydrogel network structure, while at the same time providing a way to quickly and efficiently tune the mechanical properties of the hydrogel. Evidence of a more uniform network included gelation at lower polymer concentrations and higher fracture strains (85%) when compared with hydrogels crosslinked via the chain polymerization of methacrylated HA. The hydrogel network can also be controlled simply by varying the ratio of thiols:ene, which can increase the crosslinking and reduce the swelling degree. Additionally, in most cases the photoinitiated reaction was completed after only 60 seconds of irradiation with UV light using initiator concentrations of only 0.1 mM, indicating a more efficient crosslinking reaction when compared to HA-DVS and methacrylated HA. The reaction can also occur under physiological conditions, a necessary requirement for the encapsulation of cells and proteins. | |
