Supervised and unsupervised machine learning approaches for monitoring subvisible particles within an aluminum-salt adjuvanted vaccine formulation.

Suspensions of protein antigens adsorbed to aluminum-salt adjuvants are used in many vaccines and require mixing during vial filling operations to prevent sedimentation. However, the mixing of vaccine formulations may generate undesirable particles that are difficult to detect against the background of suspended adjuvant particles. We simulated the mixing of a suspension containing a protein antigen adsorbed to an aluminum-salt adjuvant using a recirculating peristaltic pump and used flow imaging microscopy to record images of particles within the pumped suspensions. Supervised convolutional neural networks (CNNs) were used to analyze the images and create "fingerprints" of particle morphology distributions, allowing detection of new particles generated during pumping. These results were compared to those obtained from an unsupervised machine learning algorithm relying on variational autoencoders (VAEs) that were also used to detect new particles generated during pumping. Analyses of images conducted by applying both supervised CNNs and VAEs found that rates of generation of new particles were higher in aluminum-salt adjuvant suspensions containing protein antigen than placebo suspensions containing only adjuvant. Finally, front-face fluorescence measurements of the vaccine suspensions indicated changes in solvent exposure of tryptophan residues in the protein that occurred concomitantly with new particle generation during pumping.

Machine learning approaches to root cause analysis, characterization, and monitoring of subvisible particles in monoclonal antibody formulations.

Processing stresses on therapeutic proteins may cause formation of subvisible particles. Different stress mechanisms generate particle populations with characteristic morphological "fingerprints," and machine learning techniques like convolutional neural networks (CNNs) allow classification of microscopy images of these particles according to known stresses at their root cause. Using CNNs to classify novel particle types not included during network training may lead to inaccurate classification, however, using CNNs to monitor the presence of particulate matter not explicitly used in training could serve as a useful process analytical technology. We used CNNs to classify and identify the root cause of particles generated by subjecting three monoclonal antibodies (mAbs) to various common manufacturing stresses. We probed the generality of particles generated by stressing different mAbs in different formulations and showed that CNN analyses were sensitive not only to the applied stress, but also the buffer conditions and the particular mAb that generated particle populations. Thus, models trained on images of particles created with one mAb and buffer system may not provide accurate root cause analysis when applied to particles generated by other mAb and buffer systems. A lever-rule analysis of CNN-derived fingerprints was used to characterize the composition of mixtures of particle types. Finally, we monitored the temporal evolution of CNN-derived fingerprints when novel populations of particles, which were not included during training, were generated by pumping mAb solutions through a peristaltic pump.

Testing Precision Limits of Neural Network-Based Quality Control Metrics in High-Throughput Digital Microscopy.

OBJECTIVE: Digital microscopy is used to monitor particulates such as protein aggregates within biopharmaceutical products. The images that result encode a wealth of information that is underutilized in pharmaceutical process monitoring. For example, images of particles in protein drug products typically are analyzed only to obtain particle counts and size distributions, even though the images also reflect particle characteristics such as shape and refractive index. Multiple groups have demonstrated that convolutional neural networks (CNNs) can extract information from images of protein aggregates allowing assignment of the likely stress at the "root-cause" of aggregation. A practical limitation of previous CNN-based approaches is that the potential aggregation-inducing stresses must be known a priori, disallowing identification of particles produced by unknown stresses. METHODS: We demonstrate an expanded CNN analysis of flow imaging microscopy (FIM) images incorporating judiciously chosen particle standards within a recently proposed "fingerprinting algorithm" (Biotechnol. & Bioeng. (2020) 117:3322) that allows detection of particles formed by unknown root-causes. We focus on ethylene tetrafluoroethylene (ETFE) microparticles as standard surrogates for protein aggregates. We quantify the sensitivity of the new algorithm to experimental parameters such as microscope focus and solution refractive index changes, and explore how FIM sample noise affects statistical testing procedures. RESULTS & CONCLUSIONS: Applied to real-world microscopy images of protein aggregates, the algorithm reproducibly detects complex, distinguishing "textural features" of particles that are not easily described by standard morphological measurements. This offers promise for quality control applications and for detecting shifts in protein aggregate populations due to stresses resulting from unknown process upsets.

Machine learning and statistical analyses for extracting and characterizing "fingerprints" of antibody aggregation at container interfaces from flow microscopy images.

Therapeutic proteins are exposed to numerous stresses during their manufacture, shipping, storage and administration to patients, causing them to aggregate and form particles through a variety of different mechanisms. These varied mechanisms generate particle populations with characteristic morphologies, creating "fingerprints" that are reflected in images recorded using flow imaging microscopy. Particle population fingerprints in test samples can be extracted and compared against those of particles produced under baseline conditions using an algorithm that combines machine learning tools such as convolutional neural networks with statistical tools such as nonparametric density estimation and Rosenblatt transform-based goodness-of-fit hypothesis testing. This analysis provides a quantitative method with user-specified type 1 error rates to determine whether the mechanisms that produce particles in test samples differ from particle formation mechanisms operative under baseline conditions. As a demonstration, this algorithm was used to compare particles within intravenous immunoglobulin formulations that were exposed to freeze-thawing and shaking stresses within a variety of different containers. This analysis revealed that seemingly subtle differences in containers (e.g., glass vials from different manufacturers) generated distinguishable particle populations after the stresses were applied. This algorithm can be used to assess the impact of process and formulation changes on aggregation-related product instabilities.

Challenges in Predicting Protein-Protein Interactions from Measurements of Molecular Diffusivity.

Dynamic light scattering can be used to measure the diffusivity of a protein within a formulation. The dependence of molecular diffusivity on protein concentration (traditionally expressed in terms of the interaction parameter kD) is often used to infer whether protein-protein interactions are repulsive or attractive, resulting in solutions that are colloidally stable or unstable, respectively. However, a number of factors unrelated to intermolecular forces can also impact protein diffusion, complicating this interpretation. Here, we investigate the influence of multicomponent diffusion in a ternary protein-salt-water system on protein diffusion and kD in the context of Nernst-Planck theory. This analysis demonstrates that large changes in protein diffusivity with protein concentration can result even for hard-sphere systems in the absence of protein-protein interactions. In addition, we show that dynamic light scattering measurements of diffusivity made at low ionic strength cannot be reliably used to detect protein conformational changes. We recommend comparing experimentally determined kD values to theoretically predicted excluded-volume contributions, which will allow a more accurate assessment of protein-protein interactions.

Molecular trajectories provide signatures of protein clustering and crowding at the oil/water interface.

Using high throughput single-molecule total internal reflection fluorescence microscopy (TIRFM), we have acquired molecular trajectories of bovine serum albumin (BSA) and hen egg white lysozyme during protein layer formation at the silicone oil-water interface. These trajectories were analyzed to determine the distribution of molecular diffusion coefficients, and for signatures of molecular crowding/caging, including subdiffusive motion and temporal anticorrelation of the instantaneous velocity vector. The evolution of these properties with aging time of the interface was compared with dynamic interfacial tension measurements. For both lysozyme and BSA, we observed an overall slowing of protein objects, the onset of both subdiffusive and anticorrelated motion (associated with crowding), and a decrease in the interfacial tension with aging time. For lysozyme, all of these phenomena occurred virtually simultaneously, consistent with a homogeneous model of layer formation that involves gradual crowding of weakly interacting proteins. For BSA, however, the slowing occurred first, followed by the signatures of crowding/caging, followed by a decrease in interfacial tension, consistent with a heterogeneous model of layer formation involving the formation of protein clusters. The application of microrheological methods to single molecule trajectories described here provides an unprecedented level of mechanistic interpretation of interfacial events that occurred over a wide range of interfacial protein coverage.

Investigation of the immunogenicity of different types of aggregates of a murine monoclonal antibody in mice.

PURPOSE: The potential contribution of protein aggregates to the unwanted immunogenicity of protein pharmaceuticals is a major concern. In the present study a murine monoclonal antibody was utilized to study the immunogenicity of different types of aggregates in mice. Samples containing defined types of aggregates were prepared by processes such as stirring, agitation, exposure to ultraviolet (UV) light and exposure to elevated temperatures. METHODS: Aggregates were analyzed by size-exclusion chromatography, light obscuration, turbidimetry, infrared (IR) spectroscopy and UV spectroscopy. Samples were separated into fractions based on aggregate size by asymmetrical flow field-flow fractionation or by centrifugation. Samples containing different types and sizes of aggregates were subsequently administered to C57BL/6 J and BALB/c mice, and serum was analyzed for the presence of anti-IgG1, anti-IgG2a, anti-IgG2b and anti-IgG3 antibodies. In addition, the pharmacokinetic profile of the murine antibody was investigated. RESULTS: In this study, samples containing high numbers of different types of aggregates were administered in order to challenge the in vivo system. The magnitude of immune response depends on the nature of the aggregates. The most immunogenic aggregates were of relatively large and insoluble nature, with perturbed, non-native structures. CONCLUSION: This study shows that not all protein drug aggregates are equally immunogenic.

Partial unfolding of a monoclonal antibody: role of a single domain in driving protein aggregation.

We have examined the effect of incubating a monoclonal antibody (mAb) in low (0-2.0 M) concentrations of guanidine hydrochloride (GdnHCl) on the protein's conformation and aggregation during isothermal incubation. In GdnHCl solutions at concentrations from 1.2 to 1.6 M, the mAb was partially unfolded. As demonstrated by fluorescence and circular dichroism spectroscopy, the partially unfolded state of the antibody had perturbed tertiary structure but retained native secondary structure. Furthermore, partial unfolding of the antibody was documented by analytical ultracentrifugation, dynamic light scattering, and limited proteolysis. Subsequent aggregation of the antibody was characterized using size-exclusion chromatography, analytical ultracentrifugation, and dynamic light scattering. Over the entire concentration range (0-2.0 M) of GdnHCl, protein-protein interactions were attractive, as quantified by negative osmotic second virial coefficients measured with static light scattering. However, during isothermal incubation at 37 °C, the aggregation of the antibody was detected only in solutions that induced partial unfolding. Differential scanning calorimetry studies showed that the antibody's CH2 domains were unfolded in antibody molecules that had been incubated in 1.2 M and higher concentrations of GdnHCl. These results suggest that unfolding of the CH2 domains leads to aggregation.

Stabilization of an Infectious Enveloped Virus by Spray-Drying and Lyophilization.

Enveloped viruses are attractive candidates for use as gene- and immunotherapeutic agents due to their efficacy at infecting host cells and delivering genetic information. They have also been used in vaccines as potent antigens to generate strong immune responses, often requiring fewer doses than other vaccine platforms as well as eliminating the need for adjuvants. However, virus instability in liquid formulations may limit their shelf life and require that these products be transported and stored under stringently controlled temperature conditions, contributing to high cost and limiting patient access. In this work, spray-drying and lyophilization were used to embed an infectious enveloped virus within dry, glassy polysaccharide matrices. No loss of viral titer was observed following either spray-drying (at multiple drying gas temperatures) or lyophilization. Furthermore, viruses embedded in the glassy formulations showed enhanced thermal stability, retaining infectivity after exposure to elevated temperatures as high as 85 °C for up to one hour, and for up to 10 weeks at temperatures as high as 30 °C. In comparison, viruses in liquid formulations lost infectivity within an hour at temperatures above 40 °C, or after incubation at 25 °C for longer periods of time.

Stain-Free Approach to Determine and Monitor Cell Heath Using Supervised and Unsupervised Image-Based Deep Learning.

Cell-based medicinal products (CBMPs) are a growing class of therapeutics that promise new treatments for complex and rare diseases. Given the inherent complexity of the whole human cells comprising CBMPs, there is a need for robust and fast analytical methods for characterization, process monitoring, and quality control (QC) testing during their manufacture. Existing techniques to evaluate and monitor cell quality typically constitute labor-intensive, expensive, and highly specific staining assays. In this work, we combine image-based deep learning with flow imaging microscopy (FIM) to predict cell health metrics using cellular morphology "fingerprints" extracted from images of unstained Jurkat cells (immortalized human T-lymphocyte cells). A supervised (i.e., algorithm trained with human-generated labels for images) fingerprinting algorithm, trained on images of unstained healthy and dead cells, provides a robust stain-free, non-invasive, and non-destructive method for determining cell viability. Results from the stain-free method are in good agreement with traditional stain-based cytometric viability measurements. Additionally, when trained with images of healthy cells, dead cells and cells undergoing chemically induced apoptosis, the supervised fingerprinting algorithm is able to distinguish between the three cell states, and the results are independent of specific treatments or signaling pathways. We then show that an unsupervised variational autoencoder (VAE) algorithm trained on the same images, but without human-generated labels, is able to distinguish between samples of healthy, dead and apoptotic cells along with cellular debris based on learned morphological features and without human input. With this, we demonstrate that VAEs are a powerful exploratory technique that can be used as a process monitoring analytical tool.

Effect of mechanical stresses on viral capsid disruption during droplet formation and drying.

Identification of the mechanisms by which viruses lose activity during droplet formation and drying is of great importance to understanding the spread of infectious diseases by virus-containing respiratory droplets and to developing thermally stable spray dried live or inactivated viral vaccines. In this study, we exposed suspensions of baculovirus, an enveloped virus, to isolated mechanical stresses similar to those experienced during respiratory droplet formation and spray drying: fluid shear forces, osmotic pressure forces, and surface tension forces at interfaces. DNA released from mechanically stressed virions was measured by SYBR Gold staining to quantify viral capsid disruption. Theoretical estimates of the force exerted by fluid shear, osmotic pressures and interfacial tension forces during respiratory droplet formation and spray drying suggest that osmotic and interfacial stresses have greater potential to mechanically destabilize viral capsids than forces associated with shear stresses. Experimental results confirmed that rapid changes in osmotic pressure, such as those associated with drying of virus-containing droplets, caused significant viral capsid disruption, whereas the effect of fluid shear forces was negligible. Surface tension forces were sufficient to provoke DNA release from virions adsorbed at air-water interfaces, but the extent of this disruption was limited by the time required for virions to diffuse to interfaces. These results demonstrate the effect of isolated mechanical stresses on virus particles during droplet formation and drying.

Features in Backgrounds of Microscopy Images Introduce Biases in Machine Learning Analyses.

Subvisible particles may be encountered throughout the processing of therapeutic protein formulations. Flow imaging microscopy (FIM) and backgrounded membrane imaging (BMI) are techniques commonly used to record digital images of these particles, which may be analyzed to provide particle size distributions, concentrations, and identities. Although both techniques record digital images of particles within a sample, FIM analyzes particles suspended in flowing liquids, whereas BMI records images of dry particles after collection by filtration onto a membrane. This study compared the performance of convolutional neural networks (CNNs) in classifying images of subvisible particles recorded by both imaging techniques. Initially, CNNs trained on BMI images appeared to provide higher classification accuracies than those trained on FIM images. However, attribution analyses showed that classification predictions from CNNs trained on BMI images relied on features contributed by the membrane background, whereas predictions from CNNs trained on FIM features were based largely on features of the particles. Segmenting images to minimize the contributions from image backgrounds reduced the apparent accuracy of CNNs trained on BMI images but caused minimal reduction in the accuracy of CNNs trained on FIM images. Thus, the seemingly superior classification accuracy of CNNs trained on BMI images compared to FIM images was an artifact caused by subtle features in the backgrounds of BMI images. Our findings emphasize the importance of examining machine learning algorithms for image analysis with attribution methods to ensure the robustness of trained models and to mitigate potential influence of artifacts within training data sets.

Interfacial Adsorption Controls Particle Formation in Antibody Formulations Subjected to Extensional Flows and Hydrodynamic Shear.

During their manufacturing and delivery to patients, therapeutic proteins are commonly exposed to various interfaces and to hydrodynamic shear forces. Although adsorption of proteins to solid-liquid interfaces is known to foster formation of protein aggregates and particles, the impact of shear remains controversial, in part because of experimental challenges in separating the effects of shear from those caused by simultaneous exposure to interfaces. Extensional flows (occurring when solutions flow through sudden contractions) exert localized elongational forces that have been suspected to be damaging to proteins. In this work, we measured aggregation and particle formation in formulations of polyclonal and monoclonal antibodies subjected to extensional flow, high shear (105 s-1) and exposure to stainless-steel/water interfaces. Modification of the surface charge at the stainless steel/water interface changed protein adsorption characteristics without altering shear profiles, enabling shear and interfacial interactions to be separated. Even under conditions where antibodies were subjected to high hydrodynamic shear and extensional flow, production of subvisible particles could be inhibited by modifying the stainless-steel surface charge to minimize antibody adsorption. Digital images of particles recorded by flow imaging microscopy (FIM) and analyzed with machine learning algorithms were consistent with a particle formation mechanism by which antibodies adsorb and aggregate at the stainless-steel/water interface and subsequently form particles when shear displaces the interfacial aggregates, transporting them into the bulk solution. Topographical differences measured using atomic force microscopy (AFM) supported the proposed mechanism by showing reduced levels of protein adsorption on surface-charge-modified stainless-steel.

A Single Dose, Thermostable, Trivalent Human Papillomavirus Vaccine Formulated Using Atomic Layer Deposition.

Formulations of human papillomavirus (HPV) 16, 18, and 31 L1 capsomere protein antigens were spray dried to obtain glassy microspheres that were then coated by atomic layer deposition (ALD) with nanometer-thin protective layers of alumina. Spray-drying was used to formulate human papillomavirus (HPV) 16, 18, and 31 L1 capsomere protein antigens within glassy microspheres to which nanoscopic protective layers of alumina were applied using ALD. Suspensions of alumina-coated, capsomere-containing microparticles were administered in a single dose to mice. ALD-deposited alumina coatings provided thermostability and a delayed in vivo release of capsomere antigens, incorporating both a prime and a boost dose in one injection. Total serotype-specific antibody titers as well as neutralizing titers determined from pseudovirus infectivity assays were unaffected by incubation of the ALD-coated vaccines for at 4, 50, or 70 °C for three months prior to administration. In addition, even after incubation for three months at 70 °C, single doses of ALD-coated vaccines produced both higher total antibody responses and higher neutralizing responses than control immunizations that used two doses of conventional liquid formulations stored at 4 °C.

Machine Learning Analysis Provides Insight into Mechanisms of Protein Particle Formation Inside Containers During Mechanical Agitation.

Container choice can influence particle generation within protein formulations. Incompatibility between proteins and containers can manifest as increased particle concentrations, shifts in particle size distributions and changes in particle morphology distributions. In this study, flow imaging microscopy (FIM) combined with machine learning-based goodness-of-fit hypothesis testing algorithms were used in accelerated stability studies to investigate the impact of containers on particle formation. Containers in four major container categories subdivided into eleven container types were filled with monoclonal antibody formulations and agitated with and without headspace, producing subvisible particles. Digital images of the particles were recorded using flow imaging microscopy and analyzed with machine learning algorithms. Particle morphology distributions depended on container category and type, revealing differences that would not have been obvious by analysis of particle concentrations or container surface characteristics alone. Additionally, the algorithm was used to compare morphologies of particles generated in containers against those generated using isolated stresses at air-liquid and container-air-liquid interfaces. These comparisons showed that the morphology distributions of particles formed during agitation most closely resemble distributions that result from exposure of proteins to moving triple interface lines at points where container-air-liquid interfaces intersect. The approach described here can be used to identify dominant causes of particle generation due to protein-container interactions.

Atomic-Layer Deposition Processes Applied to Phage λ and a Phage-like Particle Platform Yield Thermostable, Single-Shot Vaccines.

Especially in developing countries, the impact of vaccines can be limited by logistical obstacles associated with multiple dose regimens, pathogen variants, and challenges imposed by requirements for maintaining vaccines at low temperatures during shipping and storage. Thus, there is a need for vaccines that can be flexibly modified to address evolving pathogen landscapes, are stable outside of narrow "cold-chain" temperatures and require administration of only single doses. Here we demonstrate in proof-of-concept studies a vaccine platform that addresses these impediments to more widespread use of vaccines. The platform relies on bacteriophage-derived phage-like-particles (PLPs) that utilize a "plug-and-play" antigen delivery system that allows for fast, easy alteration of antigens on the surface of the PLPs. Thermostability of PLP-based vaccines can be achieved by embedding the PLPs within glassy particles produced by spray drying, and nanoscopic aluminum oxide layers applied using atomic layer deposition (ALD) can serve to control release of antigen in vivo, yielding vaccine formulations that elicit strong immune responses after administration of single doses. Bacteriophage λ was stabilized by spray drying to form powders that were incubated at 37 °C for up to a year without loss of infectious activity. PLPs derived from bacteriophage λ were expressed and purified from E. coli cultures, and an in vitro conjugation strategy was used to decorate specific PLP surface sites with T4-lysozyme, a model vaccine antigen. The resulting T4-lysozyme:PLP complexes (Lys-PLPs) were embedded in glassy dry powders formed by spray drying and coated with nanometer-thick layers of alumina deposited by ALD in a fluidized bed reactor. Alumina-coated Lys-PLP vaccines were stable for a least a month at 50 °C, and single doses of the alumina-coated vaccines elicited immune responses that were indistinguishable from responses generated by conventional two-dose, prime-and-boost dosing regimens of alum-adjuvanted Lys-PLP vaccines.

Lyophilized Filovirus Glycoprotein Vaccines: Peroxides in a Vaccine Formulation with Polysorbate 80-Containing Adjuvant are Associated with Reduced Neutralizing Antibody Titers in Both Mice and Non-Human Primates.

Zaire ebolavirus, Sudan ebolavirus, and Marburg marburgvirus are the filoviruses most commonly associated with human disease. Previously, we administered a three-dose regimen of trivalent vaccines comprising glycoprotein antigens from each virus in mice and non-human primates (NHPs). The vaccines, which contained a polysorbate 80-stabilized squalane-in-water emulsion adjuvant and were lyophilized from a solution containing trehalose, produced high antibody levels against all three filovirus antigens. Subsequently, single-vial formulations containing a higher concentration of adjuvant were generated for testing in NHPs, but these vaccines elicited lower neutralizing antibody titers in NHPs than previously tested formulations. In order to explain these results, in the current work we measured the size of adjuvant emulsion droplets and the peroxide levels present in the vaccines after lyophilization and reconstitution and tested the effects of these variables on the immune response in mice. Increases in squalane droplet sizes were observed when the ratio of adjuvant to trehalose was increased beyond a critical value, but antibody and neutralizing antibody titers in mice were independent of the droplet size. Higher levels of peroxides in the vaccines correlated with higher concentrations of adjuvant in the formulations, and higher peroxide levels were associated with increased levels of oxidative damage to glycoprotein antigens. Neutralizing titers in mice were inversely correlated with peroxide levels in the vaccines, but peroxide levels could be reduced by adding free methionine, resulting in retention of high neutralizing antibody titers. Overall, the results suggest that oxidation of glycoprotein antigens by peroxides in the polysorbate 80-stabilized squalane-in-water emulsion adjuvant, but not lyophilization-induced increases in adjuvant emulsion droplet size may have been responsible for the decreased neutralizing titers seen in formulations containing higher amounts of adjuvant.

An Interlaboratory Comparison on the Characterization of a Sub-micrometer Polydisperse Particle Dispersion.

The measurement of polydisperse protein aggregates and particles in biotherapeutics remains a challenge, especially for particles with diameters of ≈ 1 µm and below (sub-micrometer). This paper describes an interlaboratory comparison with the goal of assessing the measurement variability for the characterization of a sub-micrometer polydisperse particle dispersion composed of five sub-populations of poly(methyl methacrylate) (PMMA) and silica beads. The study included 20 participating laboratories from industry, academia, and government, and a variety of state-of-the-art particle-counting instruments. The received datasets were organized by instrument class to enable comparison of intralaboratory and interlaboratory performance. The main findings included high variability between datasets from different laboratories, with coefficients of variation from 13 % to 189 %. Intralaboratory variability was, on average, 37 % of the interlaboratory variability for an instrument class and particle sub-population. Drop-offs at either end of the size range and poor agreement on maximum counts of particle sub-populations were noted. The mean distributions from an instrument class, however, showed the size-coverage range for that class. The study shows that a polydisperse sample can be used to assess performance capabilities of an instrument set-up (including hardware, software, and user settings) and provides guidance for the development of polydisperse reference materials.

Flow cytometry: a promising technique for the study of silicone oil-induced particulate formation in protein formulations.

Subvisible particles in formulations intended for parenteral administration are of concern in the biopharmaceutical industry. However, monitoring and control of subvisible particulates can be complicated by formulation components, such as the silicone oil used for the lubrication of prefilled syringes, and it is difficult to differentiate microdroplets of silicone oil from particles formed by aggregated protein. In this study, we demonstrate the ability of flow cytometry to resolve mixtures comprising subvisible bovine serum albumin (BSA) aggregate particles and silicone oil emulsion droplets with adsorbed BSA. Flow cytometry was also used to investigate the effects of silicone oil emulsions on the stability of BSA, lysozyme, abatacept, and trastuzumab formulations containing surfactant, sodium chloride, or sucrose. To aid in particle characterization, the fluorescence detection capabilities of flow cytometry were exploited by staining silicone oil with BODIPY 493/503 and model proteins with Alexa Fluor 647. Flow cytometric analyses revealed that silicone oil emulsions induced the loss of soluble protein via protein adsorption onto the silicone oil droplet surface. The addition of surfactant prevented protein from adsorbing onto the surface of silicone oil droplets. There was minimal formation of homogeneous protein aggregates due to exposure to silicone oil droplets, although oil droplets with surface-adsorbed trastuzumab exhibited flocculation. The results of this study demonstrate the utility of flow cytometry as an analytical tool for monitoring the effects of subvisible silicone oil droplets on the stability of protein formulations.