Comprehensive evaluation of the analytical toolbox for commercial cysteine-linked antibody-drug conjugates

The aim of this study was to evaluate the applicability of a broad range of analytical strategies for the detailed characterization and assessment of critical quality attributes (CQAs) of five cysteine-linked antibody-drug conjugates (ADCs), representative of the diversity of ADCs currently on the market. ADC integrity was evaluated using electrophoretic techniques, including non-reducing and reducing sodium dodecyl sulfate - polyacrylamide gel electrophoresis (SDS-PAGE) and capillary electrophoresis-sodium dodecyl sulfate (CE-SDS). These methods provided consistent and complementary information on structural integrity but lacked sufficient resolution to directly determine the average drug-to-antibody ratio (DAR_av) and payload distribution. DAR_av and payload distribution were therefore evaluated using three orthogonal approaches: hydrophobic interaction chromatography with UV detection (HIC‑UV) and native size‑exclusion chromatography coupled to mass spectrometry (nSEC‑MS), both at the intact level, as well as reverse‑phase liquid chromatography coupled to UV and MS (RPLC‑UV‑MS) at the middle‑up level. While HIC-UV is considered a reference technique, its performance was limited for ADCs containing hydrophilic payloads. Conversely, nSEC-MS was challenging for highly hydrophobic payloads and required the addition of organic modifiers to reduce adsorption. Middle-up RPLC-UV-MS also showed limitations, as sample preparation could introduce artefacts, UV detection did not always clearly resolve DAR species, and MS data processing was time-consuming with a risk of in-source fragmentation. Overall, each method exhibited specific strengths and weaknesses, and method selection should be guided by the chemical properties of the ADC. Charge variants of ADCs and their corresponding mAbs were characterized using imaged capillary isoelectric focusing (icIEF) and cation-exchange chromatography (CEX). ADCs exhibited a higher level of charge heterogeneity than their parent antibodies, likely due to disulfide bond reduction, linker-payload induced conformational changes, succinimide hydrolysis, and intrinsic differences between mAbs and ADCs and starting Materials. CEX provided higher resolving power than icIEF, probably because additional interactions beyond electrostatics contribute to separation. Moreover, payload-linker chemistry influenced the relative abundance of basic variants, leading to method-dependent differences between CEX and icIEF profiles. In conclusion, this work demonstrates that a wide range of electrophoretic, chromatographic, and mass spectrometric techniques can be applied for the detailed characterization of cysteine-linked ADCs, provided that analytical strategies are carefully selected according to payload-linker chemistry.

Antibody drug conjugates; Cation exchange chromatography; Drug-to-antibody ratio; Hydrophobic interaction chromatography; Imaged capillary isoelectric focusing; Native size exclusion chromatography – mass spectrometry.