Glycosylation affects clinical efficacy and safety; therefore, is a critical quality attribute of therapeutic monoclonal antibodies. Glycans are often labile and complex in patterns, giving rise to macro- and micro-heterogeneity. Recombinant production, diverse geographical locations, associated transportation and storage conditions further compound the problem. Two-way studies comparing glycoprofile of the originator and its given biosimilar are aplenty. However, the extent of analytical variation and similarity in glycoprofile across all approved versions of a drug is hardly explored. Using UHPLC and mass spectrometry, we compared the glycoprofiles of eight rituximab drug samples licensed for sale in India. While the types of glycans were found identical, the abundance of some glycans varied significantly within the tested population. The quality range of glycosylation parameters of the tested sample population differed significantly from the previously established values for US/EU licensed rituximab. As the mean abundance of the 90% of identified glycans falls within ±3SD, the extent of mutual variations amongst tested lots is less significant compared to the extreme deviation from previously established QR limits. Thus, we propose this approach as an orthogonal method to capture glycan variations in licensed versions of mAbs for quality surveillance and in cases where originator samples' are limiting. SIGNIFICANCE: As fluctuation in glycosylation may be of clinical significance, we identify that a one-to-one comparison with originator alone is insufficient in sensing the extent of variations in glycosylation parameters in licensed biosimilars of a given therapeutic mAb. Here we propose that future biosimilarity analysis may include an orthogonal approach of generating an additional combined QR range representing variations across the originator and its biosimilars. The glycosylation profiles of eight rituximab drug samples of different make obtained from the point of sale in India were found identical amongst the tested rituximab versions. However, the QR limits corresponding to important glycosylation parameters differed significantly across all tested samples from the previously established QR limits of US- and EU-licensed rituximab in statistical terms. Such an approach may be useful in defining the true range of glycan variations in licensed versions of therapeutic mAbs.

Foreign