The purpose of this study was to determine corneal permeability and uptake in rabbit, porcine, and bovine corneas for twenty-five drugs using an N-in-1 (cassette) approach and relate these parameters to drug physicochemical properties and tissue thickness through quantitative structure permeability relationships (QSPRs). A twenty-five-drug cassette containing β-blockers, NSAIDs, and corticosteroids in solution at a micro-dose was exposed to the epithelial side of rabbit, porcine, or bovine corneas mounted in a diffusion chamber, and the corneal drug permeability and tissue uptake were monitored using an LC-MS/MS method. Data obtained were used to construct and evaluate over 46,000 quantitative structure–permeability (QSPR) models using multiple linear regression, and the best-fit models were cross-validated by Y-randomization. Drug permeability was generally higher in rabbit cornea and comparable between bovine and porcine corneas. Permeability differences between species could be explained in part by differences in corneal thickness. Corneal uptake between species correlated with a slope close to 1, indicating generally similar drug uptake per unit weight of tissue. A high correlation was observed between bovine, porcine, and rabbit corneas for permeability and between bovine and porcine corneas for uptake (R² ≥ 0.94). MLR models indicated that drug characteristics such as lipophilicity (LogD), heteroatom ratio (HR), nitrogen ratio (NR), hydrogen bond acceptors (HBA), rotatable bonds (RB), index of refraction (IR), and tissue thickness (TT) are of great influence on drug permeability and uptake. When data for all species along with thickness as a parameter was used in MLR, the best fit equation for permeability was Log (% transport/cm²·s) = 0.441 LogD − 8.29 IR + 8.357 NR − 0.279 HBA − 3.833 TT + 10.432 (R² = 0.826), and the best-fit equation for uptake was Log (%/g) = 0.387 LogD + 4.442 HR + 0.105 RB − 0.303 HBA − 2.235 TT + 1.422 (R² = 0.750). Thus, it is feasible to explain corneal drug delivery in three species using a single equation.

Developing bioequivalent (BE) generic products of complex dosage forms like intravitreal implants (IVIs) of corticosteroids such as dexamethasone prepared using hot-melt extrusion (HME), based on biodegradable poly (lactide-co-glycolide) (PLGA) polymers, can be challenging. A better understanding of the relationship between the physicochemical and physicomechanical properties of IVIs and their effect on drug release and ocular bioavailability is crucial to develop novel BE approaches. It is possible that the key physicochemical and physicomechanical properties of IVIs such as drug properties, implant surface roughness, mechanical strength and toughness, and implant erosion could vary for different compositions, resulting in changes in drug release. Therefore, this study investigated the hypothesis that biodegradable ophthalmic dexamethasone-loaded implants with 20% drug and 80% PLGA polymer(s) prepared using single-pass hot-melt extrusion (HME) differ in physicochemical and/or physicomechanical properties and drug release depending on their PLGA polymer composition. Acid end-capped PLGA was mixed with an ester end-capped PLGA to make three formulations: HME-1, HME-2, and HME-3, containing 100%, 80%, and 60% w/w of the acid end-capped PLGA. Further, this study compared the drug release between independent batches of each composition. In vitro release tests (IVRTs) indicated that HME-1 implants can be readily distinguished by their release profiles from HME-2 and HME-3, with the release being similar for HME-2 and HME-3. In the early stages, drug release generally correlated well with polymer composition and implant properties, with the release increasing with PLGA acid content (for day-1 release, R² = 0.80) and/or elevated surface roughness (for day-1 and day-14 release, R² ≥ 0.82). Further, implant mechanical strength and toughness correlated inversely with PLGA acid content and day-1 drug release. Drug release from independent batches was similar for each composition. The findings of this project could be helpful for developing generic PLGA polymer-based ocular implant products.