Predictive In Vitro Vitreous and Serum Models and Methods to Assess Thiol-Related Quality Attributes in Protein Therapeutics.

In vitro models that mimic the in vivo environment can greatly facilitate and support criticality assessment of product quality attributes for therapeutic drugs to ensure product quality. An in vitro model is established to study and predict the impact of thiol-related attributes on safety or efficacy of intraocular antibody products. This model simulates the physiological redox environment of rabbit vitreous and maintains a steady-state redox potential using reduced and oxidized forms of glutathione. A similar in vitro model that mimics the thiol redox conditions of human blood has been previously established and has become a predictive tool to study intravenous (IV) therapeutic proteins. We utilized both vitreous and serum models to study the potential impact of antibody variants (trisulfides and free-thiols) on product qualities of different antibodies. The studies demonstrate that both models are effective tools to monitor changes of thiol-related attributes under physiological conditions, providing insights on these thiol-related attributes and allowing for more informed assessment of biological relevance and criticality of the attributes. Furthermore, we propose that the approach using an in vitro study for the product quality attribute assessment can be used to predict in vivo effects for future molecules during the development of biopharmaceuticals, reducing the need for live subject studies.