During lyophilisation cycle design, primary drying parameters (chamber pressure and shelf temperature) are adjusted to maximize the sublimation rate and prevent cake collapse, by maintaining the product continuously below its critical temperatures. The objective of this study was to employ mixture design of experiments to generate empirical models capable of predicting glass transition of the maximally freeze concentrated solution (Tg') and collapse temperature (Tc) of amorphous protein (BSA and IgG1) formulations. Additionally, the models developed aid the design of high concentration protein formulations with maximised critical temperatures to obtain shorter and more cost-effective lyophilisation cycles. Formulations contain sucrose as cryo/lyo-protectant and arginine/arginine-HCl as multifunctional excipient (e.g. solubility enhancer, viscosity and aggregation suppressor). The impact of formulation components at varied ratios on critical temperatures was evaluated; the amorphous excipients decrease critical temperatures, on the contrary, the protein increases critical temperatures. The robustness of the empirical models generated with BSA formulations was verified with BSA and IgG1 formulations. The models showed greater accuracy in predicting Tg' than the Fox-Flory equation. For the first time, empirical models are reported to predict both critical temperatures. Finally, unconventional collapse events observed for formulations with and without arginine/arginine-HCl at different protein concentrations are also discussed.
Keywords: Collapse temperature; Critical temperatures; Freeze-drying; Glass transition temperature; Mixture design of experiment; Quality by design.
Copyright © 2019 Elsevier B.V. All rights reserved.