Qualification and Implementation of a Robust Multi-Attribute Method Platform Across Multiple Laboratories for High-Resolution Process Characterization of Biopharmaceutical Products

Background/Objectives: Process characterization (PC) of biopharmaceutical products is intended to identify critical process parameters (CPPs) based on their impact on critical quality attributes (CQAs), as well as to define acceptable process parameter ranges to ensure consistent product quality and process performance. However, conventional analytical methods with indirect readouts, e.g., the sum of size/charge variants, often fall short in differentiating CQAs that have safety and efficacy relevance from product quality attributes (PQAs) that do not, thus limiting their utility in establishing a thorough process understanding. The Multi-Attribute Method (MAM) addresses these limitations by monitoring CQAs using the required resolution and specificity. Methods: A MAM platform was developed, and its robust performance and reproducibility were demonstrated across six laboratories. The MAM platform was applied for the drug substance (DS) PC of two monoclonal antibodies and the analysis of up to a hundred PQAs across 472 and 644 samples, respectively. The MAM results were compared to data obtained with conventional methods like hydrophilic interaction chromatography–fluorescence detection (HILIC-FLD) and cation-exchange chromatography with UV detection (CEX-UV). Results: The levels of PQAs, including succinimide, deamidation, glycosylation, and oxidation, were reproducible between six laboratories. Artefactual oxidation was limited by controlling the quality of TFA reagent. CPPs impacting the level of, e.g., oxidation, glycation, O-glycosylation, and N-glycan sialylation, were identified, leveraging a simultaneous, consistent, and fast MAM data analysis across all samples. Conclusions: MAM enables high-resolution PC to identify CPPs and inform control strategies for CQAs that are not resolved by conventional methods.