The pharmaceutical industry is constantly developing new therapies and treatments, while the cost of the drug discovery process currently is estimated at two billion dollars, spent over a 12-15 year period. Adding to the cost associated with bringing a drug to market is the high attrition rate, with only 1 in every 10,000 compounds being approved by the Food and Drug Administration. Interest in reassessing existing research procedures for improved efficiency has recently been garnering attention. Specifically, pharmacology studies, which utilize in vivo studies to obtain pharmacokinetic (PK) and pharmacodynamic (PD) information during the preclinical stage of the drug discovery process, have been a focal point. By complimenting the in vivo studies with in vitro models, an increase in efficiency is able to be realized by a reduction in consumed materials. In this dissertation, a diffusion-based dynamic in vitro (DDIV) PK model, fabricated on a microfluidic polydimethyl siloxane (PDMS) platform, was used to characterize the loading and elimination of a PK profile. However, challenges traditionally associated with the microfluidic devices, such as the fragility of the membrane due to device flexibility, reusability, and lack of automation make long-term PK studies incredibly difficult to perform, as well as reproduce. DDIV models fabricated on a rigid three-dimensional (3D) printed platform are rugged, reusable, and amenable to automation when integrated with a disposable cell culture insert. The 3D printed DDIV PK/PD device was characterized using fluorescein (332.31 g/mol) and validated using the antibiotic levofloxacin (361.37 g/mol). The loading profiles were achieved by flowing concentrated analyte through the device channels while adding buffer to the membrane insert to create a concentration gradient across the porous membrane, thereby allowing diffusion from the channel into the insert. Parameters related to the loading portion of a PK curve, such as loading time, flow rate, volume of the insert, and initial concentration in the channel were characterized. The profiles obtained during the characterization of the initial concentrations (7.5, 15, 30 μM) in the channel yielded a prediction model for both the concentration along the loading profile and the maximum concentration (Cmax) at a given loading time. Elimination of analyte from the membrane insert was proven to undergo first order rate kinetics. The elimination profile, and the resulting elimination rate constant are used to obtain the half-life. Ultimately, a prediction model for the half-life will be crucial to the characterization of the DDIV model, however preliminary gradient studies highlighted the importance of a correction factor pertaining to the amount of analyte absorbed by the device. Upon complete characterization, the reusable 3D printed DDIV PK/PD millifluidic device will allow researchers to mimic in vivo dosing regimens on an in vitro platform, resulting in a useful tool to be used in tandem with animal models.
Read
