Use of In Vitro Systems To Model In Vivo Degradation of Therapeutic Monoclonal Antibodies

Major degradation pathways such as deamidation, isomerization, oxidation, and glycation may be accelerated after administration of antibody therapeutics to the patient. Tracking in vivo product degradation is important because certain post-translational modifications can inactivate the protein and reduce product efficacy. However, in vivo characterization of protein therapeutics is not routinely performed because of technical challenges and limited sample availability. In vitro models offer several potential advantages, including larger sample supplies, simpler and faster methods for sample preparation and analysis, and the potential to distinguish differences in product degradation from differences in product clearance. In this study, we compared the rates of in vivo product degradation using mAb1 recovered from clinical serum samples with the rates of in vitro product degradation using mAb1 recovered from spiked phosphate buffered saline (PBS) and spiked human serum samples to determine if results from the in vitro model systems could be used to predict the in vivo results. The antibody samples were characterized by peptide mapping or intact mass analysis to quantify multiple quality attributes simultaneously, including deamidation, isomerization, oxidation, N-terminal pyroglutamate formation, and glycation. It was clearly demonstrated that both the spiked PBS and spiked serum models were effective in predicting in vivo results for deamidation, isomerization, N-terminal pyroglutamate formation and glycation, whereas only the spiked serum model was effective in predicting in vivo results for oxidation.