# Robustness Evaluation of a Legacy N-Glycan Profiling Method for a Therapeutic Antibody Under ICH Q14 Lifecycle Principles

## TL;DR

This study evaluates the reliability of an old method for analyzing sugar structures on a therapeutic antibody, considering changes in enzyme sources, extraction tools, and reagent stability.

## Contribution

The study provides a robustness evaluation of a legacy N-glycan profiling method under ICH Q14 lifecycle principles.

## Key findings

- Commercial PNGase F enzymes showed significant performance differences.
- Some SPE cartridges had poor recovery and high variability for sialylated glycans.
- Labeling reagent degraded rapidly within a few days.

## Abstract

Background: This study assesses the robustness of a legacy N-glycan profiling method for the therapeutic antibody MAB1 with different Peptide-N-glycosidase F (PNGase F) enzyme sources, solid phase extraction (SPE) cartridges, and reagent stability, aligning with ICH Q14 lifecycle management principles. Glycosylation profiling is critical for therapeutic antibodies as it influences both function and pharmacokinetics. Method: The legacy N-glycan profiling method, 2-aminobenzoic acid (2-AA) labeling combined with normal-phase HPLC, was re-evaluated to confirm consistent analytical performance in the context of evolving regulatory expectations. The evaluation focused on three key factors: PNGase F enzyme sources, solid-phase extraction (SPE) cartridges, and reagent stability. Results: Commercial PNGase F enzymes showed various performances, with some sources yielding significant differences. Several SPE cartridges were also tested, with certain formats displaying poor recovery and high variability, particularly for sialylated glycans. In addition, reagent stability studies revealed rapid degradation of the labeling reagent within a few days. Conclusions: These results underscore the importance of risk control, continual improvement, and lifecycle management to ensure reliable glycosylation analysis and the sustained robustness of legacy methods.