(T1030-08-54) Addressing Particulate Matter in Protein Therapeutics

Purpose: Protein therapeutics are the fastest growing class of drugs offering a versatile approach towards treating high-impact diseases. It is known that transportation, storage and presence of contaminants (e.g. silicone) can accelerate protein aggregation. For all proteins, the proper secondary and tertiary structure is critical to the targeted drug efficacy. When the proper folded structure is lost, the functionality is gone, and protein denaturation occurs. This irreversible aggregation of proteins is detrimental not only in reducing the concentration and biological availability of the active protein, but also in triggering immunogenic reactions in patients. In this work, we show how a combination of various orthogonal methods for in-depth physical and chemical characterization can be used to investigate the exposure of a parenteral product to stress. Specifically, we study inherent (aggregates), intrinsic (from the manufacturing process) and extrinsic particulates (contaminants) found in protein therapeutic products.

Methods: Stress study: A forced degradation study was designed using a IgG1-type monoclonal antibody (Infliximab) spiked with 0.5% silicone oil. To simulate mechanical stress during transportation, the vials were subject to shaking at 60 min^-1 for variable time periods (0-1020 min). Particulate characterization: Visible particles were classified using light microscopy. Chemical characterization of visible particles was performed using FT-IR and/or SEM-EDX depending on the nature of the particle. Further identification of aggregates of proteinaceous nature was performed using LC-MS/MS peptide mapping and screening of expression host cell line proteome data base for undesired proteins in the product.
Sub-visible particle counting, sizing and morphological classification of protein aggregates were performed using MFI, DLS and FlowCam Nano.

Results: The stressed and unstressed samples were characterized in regards of particulates using a combination of techniques.Light microscopy and chemical characterization were performed using spectroscopic techniques (FT-IR and/or SEM-EDX) depending on the nature of the particle. The visual assessment allows for general particle classification. Dynamic light scattering (DLS) enables particle size determination in the 0.3 µm to 10 µm range, the technique shows limited suitability for measuring stressed protein solutions as the result is an averaged fit over the entire particulate population in the sample and does not allow for differentiation of different particle types.Micro flow imaging (MFI) provides a valuable overview of the type of particulates present in the solution in the range of 1 µm-100 µm. Distinction between protein aggregates, silicon oil droplets and air bubbles is possible, however the technique shows limitations in the image quality in the lower size range.FlowCam Nano was used to characterize the particulates in the sub-micron range (0.3-15 µm). This technique is capable of performing high resolution imaging of protein aggregates that allow for particle classification similarly to MFI, but for a smaller size range. Comparison of FlowCam Nano and MFI images at different scales can be seen in Figure 1. Due to its large overlap in the nano size range, it can be considered as a complementary method to DLS. DLS, MFI and FlowCam Nano analyses of particle count demonstrated an increase in particle size and concentration with 0.5% silicone contamination (after subjection to mechanical stress). PepMap LC-MS/MS was used for further investigation of proteinaceous particles to determine the amino acid sequence of the aggregate. Comparison with the amino acid sequence of Infliximab confirmed that the aggregates were formed by the antibody (Figure 2). Furthermore, the proteome database of the expression host cells (CHO) was performed to detect potential protein derived impurities, which could lead to aggregate formations. As a result of this study, the workflow shown in (Figure 3) is suggested for the in-depth characterization and identification of particulates.

Conclusion: This study demonstrates importance of an orthogonal approach towards characterization of proteins and particles in protein therapeutics. Particles/contaminants identified in the study originated mainly from the single-use process equipment, PPE, and the production process. FT-IR and SEM-EDX allow fast and non-destructive chemical identification of unknown particulates. A good starting point for particulate characterization is the visual assessment and spectroscopic characterization (FT-IR, Raman, SEM-EDX) of large particles. For the subvisible particulate size range MFI, for the submicron size range DLS and the FlowCam Nano all confirmed that silicone induces protein aggregation. Although DLS showed limited suitability for measuring stressed protein solutions, MFI and FlowCam Nano can be used in parallel for particle characterization in the lower visible range (with high resolution images captured by the FlowCam Nano). Combined with an additional technique such as LC-MS-MS, identification can be product specific and enable a clear assignment to a root cause and classification as inherent, intrinsic or extrinsic particles for toxicological and risk assessment. Physical and chemical characterization of particulates are an integral part of the drug product quality control to ensure patient safety. Combined, this study highlights common issues and strategies applicable to deal with particulates from development until commercial manufacturing.

Comparison of FlowCam Nano and MFI images at different scales

Characterization of proteinaceous particles

Workflow for the in-depth characterization and identification of particulates