Category: Formulation and Quality
Purpose: Protein formulation development is a necessary step in protecting conformational, chemical, and colloidal stability of a molecule over the course of the drug product shelf-life and through dose preparation and administration. Large molecule drug products are formulated with excipients to maintain stability over the shelf-life of the product. Surfactants are added to the drug product to stabilize air-water interfaces known to induce protein aggregation. Specifically, IV bag preparation exposes the therapeutic protein to a new solution environment and concurrently dilutes the stabilizing excipient(s) formulated into the drug product. Furthermore, mixing in IV bags generates dynamic changes in the air-water interfacial area known to cause protein aggregation if not sufficiently protected.  Understanding the surfactant requirements for drug product end-to-end stability in early formulation development provides critical information required for a right-first-time approach to drug product formulation and robust clinical preparation. Coupling tensiometry with shaking studies, we set out to understand if interfacial properties of therapeutic proteins could predict formulation requirements for end-to-end stability.
Methods: Surface Pressure (π) - The surface tension of proteins was measured using a KRÜSS K100 tensiometer (Hamburg, GER). Proteins were diluted to 1 mg/mL in 0.9% saline and 5% dextrose. The surface pressure was calculated as the difference between the sample and the neat diluent surface tension. The change in surface pressure over time data were fit to a polynomial and the derivative of this curve at early time points was determined.
Bulk Properties (Data to be included in poster) - The bulk properties of the proteins were measured to compare the interfacial properties. Hydrophobicity was measured by hydrophobic interaction chromatography. The protein unfolding transition temperature (Tm) and aggregation onset temperature (Tagg) were measured by nanoDSF.
Shaking Studies and Subvisible Particle Analysis - Diluent shaking experiments were conducted to determine the minimum level of surfactant required to inhibit interfacial induced aggregation in saline and dextrose. Proteins were diluted in 0.9% saline and 5% dextrose with varying levels of polysorbate (PS20 and PS80 tested). The solutions were subjected to agitation in-vial and in-IV bag. Post overnight agitation samples were inspected for visible particles and subvisible particle size and concentration was measured using Halo Labs HORIZON (Toronto, CAN).
Results: The proteins selected represent three families (IgG1, IgG4, and a heavy chain heterodimer). The derivative of the surface pressure at time zero (dπ/dtt=0) is indicative of the molecule’s propensity to aggregate at the interface. A steeper slope at time=0 indicates a higher propensity for aggregation at the air-water interface. A wide range of interfacial activity is observed in saline and dextrose with four of the five molecules showing higher surface activity in saline (Figure 1).
The minimum surfactant concentration (PS20 and PS80) to suppress interfacial particle generation was determined for each molecule in saline and dextrose in two shaking geometries (vial and IV bag) for a total of 40 unique conditions. The surfactant concentrations were representative of clinical dosing, typically 1-2 orders of magnitude lower than surfactant concentrations in the drug product. All molecules displayed a minimum surfactant requirement where visible particle generation was eliminated and subvisible particle formation was significantly reduced compared to sub-optimal polysorbate concentrations (Figure 2). PS20 and PS80 provided protein stability at different concentrations in about half of the shaking studies, indicating that diluent shaking studies should be conducted in parallel with drug product in-vial shaking studies for formulation surfactant selection. In-vial diluent shaking required lower concentrations of surfactant compared to in-IV bag diluent shaking and did not correlate to the surface activity of the protein. The minimum level of surfactant required for stability in IV-bag shaking studies was linearly correlated to dπ/dtt=0 where higher surface-active compound concentrations were required in-IV bag to prevent visible and sub-visible particle formation (Figure 3).
Conclusion: The interfacial properties of a protein in a diluent are largely intrinsic. We have demonstrated that the surface pressure of a protein may predict the surfactant requirements in saline and dextrose. Importantly, the surface pressure method is fast with minimal material requirements thereby making it amenable to early formulation development. Here, the surfactant requirement prediction was tested under standard IV-preparation conditions with experimental output closely matching the fit prediction. By employing this prediction in early formulation development, large molecule drug product development may generate robust end-to-end stability across shelf-life, dose preparation, and administration.