Cation exchange chromatography on a monodisperse 3 µm particle enables extensive analytical similarity assessment of biosimilars.

Biosimilarity assessment requires extensive characterization and comparability exercises to investigate product quality attributes of an originator product and its potential biosimilar(s) and to highlight any differences between them. Performing a thorough comparison allows a shortened approval path, which also eliminates lengthy and expensive clinical trials, ensuring comparable product quality and efficacy but at lower drug prices. The wide variety of analytical methods available for biosimilar assessment ranges from biological to analytical assays, each providing orthogonal information to fully characterize biosimilar candidates. Intact native mass spectrometry (MS) has been shown to be an excellent tool for detection and monitoring of important quality attributes such as N-glycosylation, deamidation, sequence truncation and higher order structures. When combined with efficient upfront separation methods, simplification of the proteoform heterogeneity and associated complexity prior to MS analysis can be achieved. Native mass spectrometry can provide robust and accurate results within short analysis times and requires minimal sample preparation. In this study we report the use of a monodisperse strong cation exchange chromatography phase hyphenated with Orbitrap mass spectrometry (SCX-MS) to compare the best-selling biopharmaceutical product Humira® with 7 commercially approved biosimilar products. SCX-MS analysis allowed for the identification of previously described as well as so far unreported proteoforms and their relative quantitation across all samples, revealing differences in N-glycosylation and lysine truncation, as well as unique features for some products such as sialylation and N-terminal clipping. SCX-MS analysis, powered by a highly efficient separation column, enabled deep and efficient analytical comparison of biosimilar products.

Size-based characterization of adalimumab and TNF-α interactions using flow induced dispersion analysis: assessment of avidity-stabilized multiple bound species.

The understanding and characterization of protein interactions is crucial for elucidation of complicated biomolecular processes as well as for the development of new biopharmaceutical therapies. Often, protein interactions involve multiple binding, avidity, oligomerization, and are dependent on the local environment. Current analytical methodologies are unable to provide a detailed mechanistic characterization considering all these parameters, since they often rely on surface immobilization, cannot measure under biorelevant conditions, or do not feature a structurally-related readout for indicating formation of multiple bound species. In this work, we report the use of flow induced dispersion analysis (FIDA) for in-solution characterization of complex protein interactions under in vivo like conditions. FIDA is an immobilization-free ligand binding methodology employing Taylor dispersion analysis for measuring the hydrodynamic radius (size) of biomolecular complexes. Here, the FIDA technology is utilized for a size-based characterization of the interaction between TNF-α and adalimumab. We report concentration-dependent complex sizes, binding affinities (Kd), kinetics, and higher order stoichiometries, thus providing essential information on the TNF-α-adalimumab binding mechanism. Furthermore, it is shown that the avidity stabilized complexes involving formation of multiple non-covalent bonds are formed on a longer timescale than the primary complexes formed in a simple 1 to 1 binding event.

Quality by Design-Based Assessment for Analytical Similarity of Adalimumab Biosimilar HLX03 to Humira®.

Quality by design (QbD) is an efficient but challenging approach for the development of biosimilar due to the complex relationship among process, quality, and efficacy. Here, the analytical similarity of adalimumab biosimilar HLX03 to Humira® was successfully established following a QbD quality study. Quality target product profile (QTPP) of HLX03 was first generated according to the public available information and initial characterization of 3 batches of Humira®. The critical quality attributes (CQAs) were then identified through risk assessment according to impact of each quality attribute on efficacy and safety. The anticipated range for each CQA was derived from similarity acceptance range and/or the corresponding regulatory guidelines. Finally, a panel of advanced and orthogonal physicochemical and functional tests and comparison of 6 batches of HLX03 and 10 batches of the reference standard demonstrated high similarity of HLX03 to Humira®, except for slightly lower percentage of high mannosylated glycans (%HM) in HLX03 which had no effect on FcγRIII binding and antibody-dependent cell-mediated cytotoxicity (ADCC) activity in human peripheral blood mononuclear cell (PBMC). All above demonstrated the feasibility and efficiency of QbD-based similarity assessment of a biosimilar monoclonal antibody (mAb).

Assessment of Antibody Self-Interaction by Bio-Layer-Interferometry as a Tool for Early Stage Formulation Development.

PURPOSE: To speed up the drug development process in the biopharmaceutical industry, high throughput methods are indispensable for assessing drug candidates and potential lead formulations, in particular during late stages of discovery and early phases of development. This study aimed to establish a bio-layer-interferometry based high throughput assay for assessing formulation dependent mAb self-interaction (SI-BLI) and to compare the results with kD values obtained by dynamic light scattering (DLS). METHODS: Self-interaction of proprietary and commercially available mAbs was analyzed by SI-BLI and dynamic light scattering (DLS). RESULTS: We found significant correlations of the SI-BLI results and kD-values obtained by DLS for both, different mAbs in one platform formulation and for mAbs formulated in several buffer compositions. In total, we assessed self-interaction propensity of different mAbs in 58 formulations and found significant Pearson correlation (p < 0.05) between kD and results of SI-BLI. CONCLUSIONS: The SI-BLI results correlate with kD and enable fast ranking of both different drug candidates and potential lead formulations. Thus, SI-BLI might decrease the risk to lose potent mAb candidates during transition from discovery to development, and help to accelerate the development of high concentration liquid formulations.

Development and Qualification of a Scale-Down Mammalian Cell Culture Model and Application in Design Space Development by Definitive Screening Design.

Scale-down models are indispensable and crucial tools for process understanding and continuous process improvement in product life-cycle management. In this study, a scale-down model representing commercial-scale cell culture process of adalimumab biosimilar HS016 was first developed based on constant power per volume (P/V) principle and then qualified by multivariate data analysis (MVDA) and equivalence test method. The trajectories of the bench-scale process lie in the middle of the control range of large-scale process, built by multivariate evolution model based on nutrients, metabolites, and process performance datasets. This indicates that the small-scale process performance is comparable with that of the full-scale process. The final product titer, integrated viable cell density (iVCD), viability, aggregates, acid peak content, total afucosylation level, and high mannose content recognized as key process attributes (KPAs) or critical quality attributes (CQAs) were equivalent across the scales upon comparison using equivalence test method. The qualified scale-down model was then used for process characterization using a definitive screening design (DSD) where five independent variables including pH, shifted temperature, inoculation seeding density, viable cell density (VCD) at first feeding, VCD at temperature shift were evaluated. Three quadratic polynomial models for final product titer, aggregation, and high mannose were then established using the DSD results. The design space was finally developed using a probability-based Monte Carlo simulation method and was verified with the operation setpoint and worst-case condition. The case study presented in this report shows a feasible roadmap for cell culture process characterization.

Assessment of the higher order structure of Humira®, Remicade®, Avastin®, Rituxan®, Herceptin®, and Enbrel® by 2D-NMR fingerprinting.

The advent of monoclonal antibody biosimilar products has stimulated the development of analytical methods that can better characterize an important quality attribute, namely the higher order structure (HOS). Here, we propose a simple approach based on heteronuclear 2D NMR techniques at natural abundance for generating spectral fingerprints of the HOS at high resolution. We show that the proposed method can assess the HOS of six therapeutic products, adalimumab (Humira®), bevacizumab (Avastin®), infliximab (Remicade®), rituximab (Rituxan®), trastuzumab (Herceptin®), and Etanercept (Enbrel®). After treatment with immobilized papain, the purified fragments (Fab and Fc) were analyzed by 2D proton-nitrogen and proton-carbon NMR correlations. All Fab and Fc fragments produced high-resolution 2D-NMR spectra from which assessment of their higher order structure can be performed in the context of comparability studies. In particular, the two different sequences of Fc fragments could be unambiguously distinguished. The results show that it is possible to obtain structurally dependent information at amino acid resolution of these important therapeutic agents.