Abstract by Frederik Bock

Biotherapeutics are currently one of the fastest growing drug classes. Due to their limited oral bioavailability, biotherapeutics are generally administered parenterally via subcutaneous injection. Upon injection, the drug formulation is exposed to the subcutaneous environment, which includes the extracellular matrix and the interstitial fluid before reaching the circulation. The use of animal models has led to limited success in predicting absorption and bioavailability in humans, as a result biorelevant in vitro release-testing methods have received increased attention as an alternative. However, there are currently no regulatory-approved in vitro release-testing methods for injectables and the conventional in vitro release-testing methods used for injectable formulations may not mimic the subcutaneous environment accurately. Hydrogels have been used as a more biorelevant mimic for the extracellular matrix in combination with UV-Vis imaging for studying initial drug release and formulation behavior for injectables. The overall aim of this PhD project was to develop a biorelevant in vitro release-testing platform for injectable formulations using UV-Vis imaging in combination with hydrogels or hydrogel-like matrices.

A static UV-Vis imaging in vitro release-testing method was developed for injectable formulations consisting of a 3D-printed release cell and an agarose hydrogel emulating the extracellular matrix. It was found that the background absorbance of the matrix affected the dynamic range leading to deviations from Beer’s law, highlighting the importance of a transparent matrix in UV-Vis imaging. Novel MATLAB scripts provided advanced image analysis capabilities beyond the commercial software and were utilized throughout the project. Formulations were injected into a premade cavity, and cyclic AMP (cAMP) release was monitored at 280 nm, while formulation behavior was observed at 520 nm without interference from cAMP. The formulation geometry could be controlled using the static setup to visualize different diffusion patterns depending on the geometry. A limitation of the static setup was the inability of the applied hydrogel to sustain a convective flow leading to local drug buildup, i.e.  non-sink conditions.

For biorelevant hydrogel-based in vitro release-testing setups for s.c. injectables, the ability to sustain a convective flow is key in order to create sink conditions allowing drug release to continue and to emulate the hydrodynamic conditions in vivo. A porous matrix consisting of agarose beads commonly used in size-exclusion-, ion-exchange-, and affinity chromatography was identified as an alternative to the agarose gel utilized so far. A 3D-printed flow cell was used with the porous matrix facilitating the incorporation of convective flow and enabling injection directly into the matrix, mimicking in vivo conditions to a greater extent than the static setup. A disadvantage of the porous matrix was a high optical density (high matrix absorbance) caused by light scattering as compared to native agarose gels. However, refractive index matching using high molecular weight dextrans was to a large extent capable of mitigating the matrix absorbance effect. Using the UV-Vis imaging flow-through setup, the interplay between convective and diffusive contributions to dexamethasone drug transport was visualized at different flow rates and Péclet numbers. Ion-exchange agarose beads showed potential as mimics for endogenous polycharged extracellular matrix components and may be useful for characterizing or probing the potential of electrostatic interactions likely to occur at the injection site. The flow-through setup constitutes a significant breakthrough in terms of utilizing a flow-compatible matrix together with UV-Vis imaging for the first time.

To increase throughput, a so-called in vitro release cartridge (IVR cartridge) was explored comprising the porous matrix intended for size-exclusion- or affinity chromatography. The effect of incorporating biorelevant constituents into the system was investigated including medium composition and human serum albumin. Insulin was selected as a model compound for biotherapeutics and the insulin disappearance for solutions, suspension-based, and in situ precipitating insulins was determined. The presence of divalent cations in the biorelevant release medium (mimicking the ISF ionic composition) resulted in slower insulin disappearance for the suspension-based and in situ precipitating insulins as compared to the phosphate buffer. Correct rank-ordering regarding insulin disappearance for non-albumin binding insulins with both release media was achieved, and albumin incorporation on the agarose beads resulted in improved rank-ordering for albumin-binding insulins. The UV-Vis imaging flow-through setup visualized decreased insulin dissolution using the biorelevant release medium as well as the interaction of albumin-binding insulins with the albumin composite porous matrix. Plasma profiles obtained from the literature was used to establish an in vitro in vivo relation (IVIVR) resulting in correct rank-ordering for simulated in vivo release profiles for the non-albumin-binding insulins. Overall, it was shown that the porous matrix may hold potential as an extracellular matrix mimic for in vitro release-testing of other subcutaneous injectable formulations.

In conclusion, a biorelevant UV-Vis imaging in vitro release-testing platform for injectable formulations was developed utilizing a porous matrix emulating the subcutaneous tissue. The ability to sustain a flow constitutes a significant development resembling the interstitial fluid flow through the extracellular matrix while enabling the eluent to be collected and analyzed using complementary analytical methods. Furthermore, the biopredictive potential of the porous matrix was demonstrated by establishing an IVIVR for subcutaneously administered insulin products. The discoveries made in this project may allow for more biorelevant and biopredictive release testing for injectable formulations using UV-Vis imaging.