Evaluation of TD-NMR for the detection of protein aggregation in liquid biopharmaceuticals

Monitoring of protein aggregation is crucial to avoid the administration of an altered drug substance to a patient. The techniques that are currently used to detect the formation of protein aggregates, e.g., Size Exclusion - High Performance Liquid Chromatography SE–HPLC = SEC) and Flow Imaging Microscopy (FIM), require a certain sample preparation and the destructive use of the protein drug product to perform the measurement. Furthermore, multiple orthogonal or complementary techniques need to be used in order to characterize the sample completely, e.g., to characterize aggregates in every possible size range such as the nanometer (<100 nm), submicrometer (100-1000 nm), subvisible (1-100 µm), visible (>100 µm) range. In this work, the readouts obtained from established techniques for the analysis of protein aggregates were compared to the one obtained from a new approach, Time Domain Nuclear Magnetic Resonance (TD-NMR). The latter determines the transverse relaxation rate (R2) of protons in aqueous solution and allows a fast, non-destructive bulk measurement of a sample inside its primary package. To better understand the relation of R2 to the formation of protein aggregates, particularly during the early stage of their formation, five different techniques were used alongside TD-NMR: size-exclusion chromatography (SEC), Flow imaging microscopy (FIM), dynamic light scattering (DLS), Nanoparticles Tracking Analysis (NTA) and Digital Visual Inspection (DVI) with an Enhanced and Transportable Automatic Checking system (ETAC). Different formulations containing an IgG1 subclass monoclonal antibody (mAb) in different concentrations with or without a surfactant (PS20) in different concentrations, were analyzed to test TD-NMR’s sensitivity for variations in protein and surfactant content. Further, a forced degradation study consisting of freezing-thawing cycles was performed to relate the formation of early aggregation to changes in R2 detected by TD-NMR. The main part of the current work comprised a stability study that lasted 13 weeks and was performed to monitor degradation processes in mAb formulations and placebo stored at different temperatures (2-8 °C, 25 °C, 40 °C and 50 °C). Again, changes in R2 were compared to the read-out of the above-mentioned analytical techniques. TD-NMR showed high sensitivity for protein concentration change in solution. Further, in both, the forced degradation study and the stability study, it was demonstrated that when protein aggregates of any size were detected by one or more of the orthogonal techniques, TD-NMR never failed to give an indication of their presence by detecting a change in R2 compared to a reference measurement of unstressed mAb samples. However, not all techniques were able to detect the presence of aggregates. For instance, in the stability study, MFI, NTA and ETAC did not show a change in signal over time. Furthermore, it was demonstrated that TD-NMR is not sensitive to changes in surfactant concentration in a bio- pharmaceutically relevant concentration range (0-1 mg/mL) and can therefore not yet be used to study surfactant degradation. The simplicity of the measurement, its high sensitivity for the detection of even early stages of protein aggregation and the possibility of sample measurement in the primary package, make TD-NMR a promising candidate for aggregate quality control in the pharmaceutical industry.