Evaluating Dextrose-Derived Interference in Low-Concentration Monoclonal Antibody Quantification for Clinical In-Use Studies
摘要
Clinical in-use study of monoclonal antibody (mAb) drug products is a critical component of product development, designed to assess the stability and compatibility of mAbs with clinical diluents and delivery materials. These studies often involve diluting mAbs into 5% dextrose (D5W), a common intravenous diluent that can pose analytical challenges. When D5W IV bags are subjected to terminal steam sterilization, the dextrose undergoes degradation into glucose degradation products (GDPs), including 5-hydroxymethylfurfural (5-HMF), which absorbs strongly at 284 nm and can interfere with UV absorbance-based protein quantification at 280 nm (A280). Although GDP formation in D5W IV bags is well documented, its impact on A280-based protein quantification, especially at low concentrations, has not been previously reported. In this study, we investigated the influence of GDPs on protein quantification in D5W-diluted samples, with a focus on mAb concentrations ≤ 1.0 mg/mL. We also evaluated the effects of temperature and storage duration on GDP accumulation. Our findings revealed that GDP levels varied across IV bags and increased with prolonged storage at elevated temperatures. Using a model mAb, we demonstrated that D5W interferes with A280-based protein measurements. While baseline correction using D5W from the same IV bag reduced interference, variability persisted when using D5W from different bags, even within the same lot. To address this challenge, we implemented a Protein-A-based affinity chromatography method that enabled accurate quantification of mAb concentrations in the range 0.01–1.0 mg/mL using a UV detector and 0.005–1.0 mg/mL using a fluorescence detector. These findings highlight that D5W can introduce significant discrepancies in A280-based protein measurements and demonstrate that Protein-A affinity chromatography provides a reliable alternative for quantifying mAbs in D5W-diluted samples by effectively mitigating GDP interference.
Graphical Abstract