The biopharmaceutical need for high protein concentration formulations (i.e. > 25 mg/mL) that are necessary to attain acceptable efficacy within the volume limit for subcutaneous delivery has grown with increased development of therapeutics such as monoclonal antibody (MAb)-based products. In addition to typical stability challenges, high protein concentrations present unique obstacles during formulation, including Donnan effects and increases in viscosity due to reversible self-association. Furthermore, at relatively high protein concentrations, a nonlinear increase in protein activity coefficient may lead to a rate of aggregation of protein molecules that far exceeds what is expected based on behavior of the protein in more dilute solutions. Thus, an incremental increase in protein concentration (e.g., from 125 to 150 mg/mL) may cause a dramatic increase in aggregation rate.

The physical stability of a current therapeutic protein, recombinant human interleukin-1 receptor antagonist (rhIL-1ra), has been characterized at high protein concentrations. rhIL-1ra, was found to exist in a protein concentration dependent monomer-dimer equilibrium controlled by solution ionic strength. The equilibrium constants for the rhIL-1ra system were determined by sedimentation equilibrium experiments, and used to fit static light scattering and membrane osmometry data to an expanded second osmotic virial coefficient model accounting for separate monomer-monomer (B₂₂), monomer-dimer (B₂₃) and dimer-dimer ( B₃₃) interactions.

Accelerated aggregation experiments at 37°C were used to observe the reaction kinetics, and a population balance model was applied to a continuous mixed suspension, mixed product removal (MSMPR) reactor at steady-state to determine aggregate nucleation and growth rates. The rhIL-1ra aggregation results were consistent with hard sphere fluid thermodynamic activity coefficient predictions. These observations were explained in the context of a modified Lumry-Eyring model for aggregation that incorporates the proposed rhIL-1ra assembly mechanism combined with a hard-sphere model to account for volume-exclusion-based solution nonidealities found at high protein concentrations.