Characterizing the flow behavior of protein and excipient solutions using Dripping-onto-Substrate extensional rheometry

Abstract

Injectable protein medications are lifesaving therapies for patients with cancer, COVID-19, and autoimmune diseases. However developing stable, concentrated protein therapies canbe challenging due to regulations requiring small volumes and viscosities. Accordingly,many therapies are administered intravenously at hospitals, requiring long, expensive stays.Developing ultra-concentrated (>150 mg/mL) medications that can be self-administered subcutaneously can reduce costs and improve flexibility for patients. However ultra-concentrated formulations often suffer from high viscosities and poor stability at rest and under flow. Further, protein medications can denature and lose efficacy when injected due to strong extensional “stretching” forces. These extensional flows can be more detrimental to protein structure and function than shear flows, although thorough studies of protein extensional rheology are limited due to volume constraints and instrumentation challenges. Further, pharmaceutical excipients – molecules like polymers or surfactants which are added to reduce shear viscosity and stabilize formulations – can produce complex flow effects in extension. The first goal of this thesis is to build and validate a novel instrument to measure extensional rheology and associated material properties of protein and protein-excipient solutions for the first time. To do so, a modified dripping-onto-substrate (DoS) extensional rheology device was designed and created. DoS is an extensional rheological technique that creates a semi-stable liquid bridge from a single drop. In time, the liquid bridge self-thins due to inertial, surface tension, viscous and elastic forces. The evolution of the liquid bridge radius is captured with high-speed imaging, which can then be fit to extract rheological parameters. The modified DoS instrument enables measurements of low-viscosity solutions in pure extensional flows in 10 μL or less, overcoming traditional limitations associated with measuring protein solutions in extension. This technique was validated using several test fluids with well-known literature values prior to measuring protein solutions. The second goal was to understand the flow behavior and potential synergistic or antagonistic effects between proteins and pharmaceutical excipients in extensional flows. Model protein ovalbumin (OVA) solutions with added FDA-approved excipients poloxamer 188 (P188), polysorbate 20 (PS20), or polysorbate 80 (PS80) were examined using the DoS technique. OVA is similar in size to insulin, which has previously demonstrated injection-dependant flow behavior. However promisingly, OVA-only solutions up to 300 mg/mL protein exhibited rapid thinning and breakup behavior characteristic of low viscosity fluids. Conversely, excipients typically added to prevent protein aggregation at restor in shear flow appeared to cause detrimental behavior in injection-like flows. P188, a poloxamer that is primarily composed of unimers in solution, demonstrated rapid thinning at low concentrations but transitioned to weakly elastic behavior at higher concentrations. P188 addition was required to observe elasticity in combined P188/OVA conditions since OVA alone did not demonstrate elasticity at any studied concentration. Compared to P188, PS20, and PS80 are smaller molecules and form micelles at the studied conditions. Although PS20 and PS80 are structurally similar, differences in surface activity result in observed flow differences at low concentrations for PS and PS/OVA solutions. However, at higher concentrations, PS20 and PS80 behavior becomes statistically identical due to crowded solution effects. The final goal was to examine the effect of substrate spreading behaviors on capillary-driven thinning, which is not well-explored, particularly for low-viscosity solutions. While capillary-driven thinning progresses in DoS experiments, fluid spreads on the substrate; these differences in spreading or other instrument parameters can lead to variations in the liquid bridge breakup times and rheological parameters. These discrepancies in behavior can be reconciled by correlating to the dimensionless Weber number as well as other fluid and instrument parameters such as aspect ratio or drop volume. Further, computing spreading distances and Weber numbers of spreading can elucidate the importance of Marangoni stresses in capillary thinning experiments for surface-active macromolecules. The results of this thesis demonstrate the first capillary thinning rheological measurements of protein excipient solutions and create a methodology for measuring future protein/excipient combinations. Identifying differences in flow behavior between formulations is important for pharmaceutical development to create more stable therapies in extensional flows. Additionally, an understanding of capillary thinning and spreading dependence can explain variation in DoS experiments and lead to more accurate comparisons of experiment results across samples and concentrations. While this thesis focused on protein-excipient solutions, many of the leanings on methodology are generally relevant to low-viscosity fluids, which can be useful for the broader rheological community.