Analytical Comparability Strategies for Originators and Biosimilars

Abstract

Biologics are pharmaceuticals produced by living systems. Biosimilars, the follow-on counterparts to biologics are unlike small-molecule generics in the fact that they are large and complex and may contain minor variations from the originator product. The presence of these variations, derived from post-translational modifications (PTMs), necessitates a more rigorous approval procedure than that which is implemented for small-molecule generics. Extensive physicochemical characterization must be undertaken to confirm that the biosimilar’s quality attributes closely match the originator in molecular and functional features. If high similarity is confirmed by these studies, the extent of preclinical and clinical trials may be reduced. Based on this, we performed extensive analytical comparability studies for filgrastim and several mAbs using various methods including several different liquid chromatography (LC) separations and mass spectrometry (MS) techniques. In the first study, we compared Neupogen® and its proposed biosimilar filgrastim in terms of structural (primary and higher order) and chemical variants (size, oxidation, deamidation) using tandem MS, intact MS, 2D NMR and LC separations. Both molecules showed identical primary structure, comparable higher order structure and low levels of each variant except in deamidation levels, where originator filgrastim showed higher levels than that of the biosimilar. In the second study, we combined the multi-attribute method (MAM) analysis with multiple orthogonal techniques to structurally compare Rituxan® and a proposed rituximab biosimilar while focus specifically on glycosylation and resulting biological activity. We hypothesized that different glycosylation distributions between the originator and biosimilar rituximab could result in different biological activities, including differences in binding affinity to the FcγIIIA receptor and differences in antibody dependent cell cytotoxicity (ADCC). We show that both mAbs had identical primary structures by tandem MS and similar higher order structures by ion mobility (IM) MS and hydrogen deuterium exchange (HX) MS. We observed similar levels of deamidation and oxidation for both products, but significant differences in the levels of specific glycoforms. In particular, the biosimilar not only had a higher level of afucosylated glycans but also showed a higher FcγIIIA binding affinity and higher ADCC potency, thus suggesting a possible difference in clinical efficacy. Finally, we identified initial structural differences/similarities and attempted to identify whether or not these differences could be amplified through the application of thermal stress through three originator-biosimilar pairs; rituximab, bevacizumab and trastuzumab. Initially, we detected highly similar secondary and tertiary structures and different levels of size and charge variants for each pair. After 4 weeks of incubation at 40 ºC, we measured differences in charge variant distributions and unfolding patterns. Taken together, our study shows the ability to establish comparability by creating a profile of initial differences for multiple mAb pairs and determining how those differences change when subject to thermal stress. In conclusion, our studies provide an exemplary analytical exercise that can be implemented in the development of future biosimilar products.