Closing the Gap: Counting and Sizing of Particles Across Submicron Range by Flow Cytometry in Therapeutic Protein Products

Abstract

Quantification and size distribution characterization of subvisible particles in parenteral biopharmaceutics, present as both proteinaceous and nonproteinaceous particles in the size range from 0.1 to 100 μm, are important for biopharmaceutical industry due to their potential safety and efficacy implications. Although a number of analytical techniques are available to count and size subvisible particles, characterization of particles ≤2 μm remains a significant challenge due to technical limitations of existing particle counting instruments. In this article, we demonstrate the ability of an optimized flow cytometry system to detect and quantify size distribution of subvisible particles without additional labeling that includes the critical submicron range in biopharmaceutical formulations. In addition, these qualitative and quantitative determinations are performed in a high-throughput manner using sample volumes that allow statistically significant evaluations. This approach can be used not only to ascertain the quality of therapeutic protein products but also to evaluate numerous conditions during the screening of drug candidates and their prospective formulations.

Introduction

An important product quality attribute required by many pharmacopoeias to be tested for parenteral biotherapeutics is the presence of subvisible particles. A sample passes the compendia subvisible particle test if it contains less than 6000 particles/container at ≥10 μm and less than 600 particles/container at ≥25 μm for low-volume drug products (≤100 mL of fill volume) using a compendia specified method that is mainly a light obscuration (LO) method.¹ The concerns with ≥10-μm and ≥25-μm particles have originated from the possible occlusion of small blood vessels by parenteral administration of extrinsic particles (unexpected contaminant particles, e.g., paint chips, clothing fibers) and intrinsic particles (undesirable particles from degradation of formulation components, or from manufacturing and packaging processes, e.g., silicone oil droplets, rubber from stoppers) in these size ranges. However, in biopharmaceutics, the inherent particles (expected particles from degradation of active pharmaceutical ingredient or excipients) are mainly proteinaceous aggregates originating from colloidal and structural instability of the protein molecules growing across a size continuum from soluble aggregates (<0.1 μm) to subvisible and visible size range.2, 3 These particles pose primary concerns because of their potential immunogenicity consequences that could negatively impact safety and efficacy of the product.1, 4, 5 Although the link between the presence of protein aggregates and potential immunogenicity has not been unambiguously established,6, 7 the uncertainty associated with the potential immunogenicity concerns from proteinaceous subvisible particles highlights the need to accurately count and size proteinaceous particles across the entire subvisible particle size continuum, with the emphasis on submicron range particles, as increasing number of biotherapeutic drug candidates continue to enter clinical development.8, 9

The strengths and limitations of analytical methods commonly used to characterize subvisible particles in biopharmaceutics are available in a number of reviews.1, 4, 10, 11, 12, 13, 14 The LO method, originally developed for foreign particulate analysis based on reduction in transmitted light intensity when a particle traversing the light beam scatters light, is the preferred method for subvisible particle testing listed in the pharmacopeias. LO instruments offer rapid determination of number of particulates and the size distribution of the particulates with a substantial commercial biotherapeutic product testing history.¹⁰ However, in part due to the possibility of LO method being less reliable in smaller than the 10-μm size range, depending on the type of particles and optical properties, the presence of proteinaceous particulates in biopharmaceutical drug product development has been initially poorly addressed, if not underestimated. Visually convincing demonstration of the abundance of proteinaceous particles in therapeutic protein formulations with distributions generally centered below 5 μm and extending below 1 μm by a quantitative Nile Red fluorescence microscopy method¹⁵ led to the recognition of the respective analytical gap as a major challenge in the process of the development of safe and efficacious therapeutic protein products.⁴ These concerns were only partially addressed by a concurrently introduced dynamic imaging analysis (DIA) capable of counting, sizing, and capturing images of subvisible particles as the sample passes through an optical cell positioned in the center of a field of view with a size limit of 0.7 μm.¹⁶ The DIA systems, allowing accurate differentiation of silicone oil droplets and air bubbles from protein aggregates for particles ≥4 μm based on differences in particle morphology,¹⁶ have become an important alternative method for protein-based subvisible particle analysis.16, 17, 18 LO and DIA systems are not suitable for samples that are high in turbidity, and they also have difficulties measuring particles that are transparent, or have refractive index (RI) close to that of the medium.10, 17, 19 LO instruments, such as the widely used HIAC, have a relatively low particle count threshold (18,000 particles/mL),²⁰ and for samples with high particle concentrations, such as the stressed samples, dilution maybe required to reduce coincidence for count and size accuracy. However, any dilutions may result in changes in subvisible particle counts and size distributions²¹ and the effects of sample dilution on subvisible analysis results should be evaluated when developing methods.

During the following decade, despite an emergence of a number of new approaches such as light scattering–based video imaging (NanoSight) and resonant mass measurement (RMM, Archimedes), characterization of protein aggregates in the submicron range remained difficult due to low throughput combined with limited size ranges, and most importantly, due to insufficient volumes (in nanoliter range) that could be effectively scanned compared to microliter range volumes required to obtain statistically significant readouts of particle numbers typically encountered in protein formulations.²²

Flow cytometry (FC), developed to identify subpopulations of cells by the scattered light or fluorescence emission, has been demonstrated to be capable of detecting protein aggregates and silicone oil droplets in submicron size ranges.23, 24, 25, 26 In addition, to probe submicron range particles as low as 200 nm,²⁶ FC, compatible with plate-based high-throughput platform that requires small sample volumes of <200 μL with counting and sorting speeds of thousands of particles per second, is an ideal method to obtain statistically significant protein aggregate analysis results that could support candidate selection and optimal formulation condition identification processes, where a large amount of samples from multiple combinations of pH, ionic strength, and excipients need to be screened for each candidate when the sample volumes are often too limited to use LO or DIA systems. One challenge in use of FC as a particle analysis method is that there are no standard methods or reference materials to convert the scattering or fluorescence signals obtained from FC into absolute measures of particle sizes, and the quantitative recovery of predicted standard particle counts in the submicron range has not been demonstrated to date. Moreover, in FC systems employing sheath fluid, sample dilutions as the samples enter the flow cell and breakage of unstable aggregates may happen.²¹ The reported particle analysis results are often presented only as a total number of particles, while using qualitative visual dot plots for pattern comparisons.

In this work, we demonstrate the use of FC as a high-throughput particle analysis method to characterize submicron and subvisible protein aggregate populations without the use of the sheath fluid and fluorescence labeling encompassing the size ranges from 0.16 μm to 10 μm in a both qualitative and quantitative manner. In addition, to obtain total particle counts, we used both fluorescence-labeled particle standards and nonlabeled particle standards to create particle size gates to obtain estimated size distribution information from submicron to subvisible size ranges. As a part of method assessments, we evaluated the particle counting linearity and reproducibility, demonstrated quantitative recovery of predicted standard particle counts at 0.5 μm and 1 μm, as well as compared the FC particle analysis results of protein samples with the results from a DIA and an RMM system. In effect, a reliable and practically applicable method for detecting the presence and estimating size distributions of submicron and subvisible particles in therapeutic protein products and their prospective formulations is obtained.