Utility of Three Flow Imaging Microscopy Instruments for Image Analysis in Evaluating four Types of Subvisible Particle in Biopharmaceuticals

Application of one-class classification using deep learning technique improves the classification of subvisible particles

Capturing subvisible particles using flow imaging microscopy is useful for evaluating protein aggregates that may induce immunogenicity. Automated labeling is desirable to distinguish harmless components such as silicone oil (SO) from subvisible particles. The one-class classifier, which requires only target class data for model establishment, is suitable for machine learning and proposes a useful solution for distinguishing a subject with heterogeneous but stable distributions, such as SO. However, the effectiveness of the application of one-class classifiers to subvisible particles remains unclear. In this study, we investigated whether deep learning techniques can improve the performance on a variety of images. We prepared datasets using SO and two types of protein aggregates: immunoglobulin G-derived aggregates (AggIgG) and albumin-derived aggregates (AggAlb). The deep-learning technique improved the classification scores for both AggIgG and AggAlb. The classification scores for AggIgG were more satisfactory than those for AggAlb. Cluster analysis revealed that one-class classification using deep learning techniques achieved excellent effectiveness across almost all clusters in classifying AggIgG. Collectively, the deep learning technique remarkably improved the one-class classification of subvisible particles of AggIgG and AggAlb. Combined with deep learning, one-class classification can contribute to the evaluation of subvisible particles, particularly for AggIgG.

Representative training data sets are critical for accurate machine-learning classification of microscopy images of particles formed by lipase-catalyzed polysorbate hydrolysis

Polysorbate 20 (PS20) is commonly used as an excipient in therapeutic protein formulations. However, over the course of a therapeutic protein product's shelf life, minute amounts of co-purified host-cell lipases may cause slow hydrolysis of PS20, releasing fatty acids (FAs). These FAs may precipitate to form subvisible particles that can be detected and imaged by various techniques, e.g., flow imaging microscopy (FIM). Images of particles can then be classified using supervised convolutional neural networks (CNNs). However, CNNs should be trained on representative images of particles which, as we demonstrate in this work, may be challenging to obtain. Here, we tested several rapid techniques to create FA particles and examined whether CNNs trained on microscopy images of these rapidly formed particles could accurately classify images of particles that had been produced by kinetically slower lipase-catalyzed hydrolysis of PS20. CNNs trained on images of rapidly produced particles were less accurate in classifying images of FA particles that had been produced by enzymatic hydrolysis of PS20 than CNNs trained with images of particles generated by the same slow hydrolysis, highlighting the importance of using representative image data sets for training CNN classifiers.

Particle formation in response to different protein formulations and containers: Insights from machine learning analysis of particle images

Subvisible particle count is a biotherapeutics stability indicator widely used by pharmaceutical industries. A variety of stresses that biotherapeutics are exposed to during development can impact particle morphology. By classifying particle morphological differences, stresses that have been applied to monoclonal antibodies (mAbs) can be identified. This study aims to evaluate common biotherapeutic drug storage and shipment conditions that are known to impact protein aggregation. Two different studies were conducted to capture particle images using micro-flow imaging and to classify particles using a convolutional neural network. The first study evaluated particles produced in response to agitation, heat, and freeze-thaw stresses in one mAb formulated in five different formulations. The second study evaluated particles from two common drug containers, a high-density polyethylene bottle and a glass vial, in six mAbs exposed solely to agitation stress. An extension of this study was also conducted to evaluate the impact of sequential stress exposure compared to exposure to one stress alone, on particle morphology. Overall, the convolutional neural network was able to classify particles belonging to a particular formulation or container. These studies indicate that storage and shipping stresses can impact particle morphology according to formulation composition and mAb.

Past, Current, and Future: Application of Image Analysis in Small Molecule Pharmaceutical Development

The often-perceived limitations of image analysis have for many years impeded the widespread application of such systems as first line characterisation tools. Image analysis has, however, undergone a notable resurgence in the pharmaceutical industry fuelled by developments system capabilities and the desire of scientists to characterize the morphological nature of their particles more adequately. The importance of particle shape as well as size is now widely acknowledged. With the increasing use of modelling and simulations, and ongoing developments though the integration of machine learning and artificial intelligence, the utility of image analysis is increasing significantly driven by the richness of the data obtained. Such datasets provide means to circumvent the requirement to rely on less informative descriptors and enable the move towards the use of whole distributions. Combining the improved particle size and shape measurement and description with advances in modelling and simulations is enabling improved means to elucidate the link between particle and bulk powder properties. In addition to improved capabilities to describe input materials, approaches to characterize single components within multicomponent systems are providing scientists means to understand how their material may change during manufacture thus providing a means to link the behaviour of final dosage forms with the particle properties at the point of action. The aim is to provide an overview of image analysis and update readers with innovations and capabilities to other methods in the small molecule arena. We will also describe the use of AI for the improved analysis using image analysis.

Combined 100 keV Cryo-Electron Microscopy and Image Analysis Methods to Characterize the Wider Adeno-Associated Viral Products

Adeno-associated viruses (AAVs) are effective vectors for gene therapy. However, AAV drug products are inevitably contaminated with empty particles (EP), which lack a genome, owing to limitations of the purification steps. EP contamination can reduce the transduction efficiency and induce immunogenicity. Therefore, it is important to remove EPs and to determine the ratio of full genome-containing AAV particles to empty particles (F/E ratio). However, most of the existing methods fail to reliably evaluate F/E ratios that are greater than 90 %. In this study, we developed two approaches based on the image analysis of cryo-electron micrographs to determine the F/E ratios of various AAV products. Using our developed convolutional neural network (CNN) and morphological analysis, we successfully calculated the F/E ratios of various AAV products and determined the slight differences in the F/E ratios of highly purified AAV products (purity > 95 %). In addition, the F/E ratios calculated by analyzing more than 1000 AAV particles had good correlations with theoretical F/E ratios. Furthermore, the CNN reliably determined the F/E ratio with a smaller number of AAV particles than morphological analysis. Therefore, combining 100 keV cryo-EM with the developed image analysis methods enables the assessment of a wide range of AAV products.

Stability of Protein Pharmaceuticals: Recent Advances

There have been significant advances in the formulation and stabilization of proteins in the liquid state over the past years since our previous review. Our mechanistic understanding of protein-excipient interactions has increased, allowing one to develop formulations in a more rational fashion. The field has moved towards more complex and challenging formulations, such as high concentration formulations to allow for subcutaneous administration and co-formulation. While much of the published work has focused on mAbs, the principles appear to apply to any therapeutic protein, although mAbs clearly have some distinctive features. In this review, we first discuss chemical degradation reactions. This is followed by a section on physical instability issues. Then, more specific topics are addressed: instability induced by interactions with interfaces, predictive methods for physical stability and interplay between chemical and physical instability. The final parts are devoted to discussions how all the above impacts (co-)formulation strategies, in particular for high protein concentration solutions.'

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.

Sub-Visible Particle Classification and Label Consistency Analysis for Flow-Imaging Microscopy Via Machine Learning Methods

Sub-visible particles can be a quality concern in pharmaceutical products, especially parenteral preparations. To quantify and characterize these particles, liquid samples may be passed through a flow-imaging microscopy instrument that also generates images of each detected particle. Machine learning techniques have increasingly been applied to this kind of data to detect changes in experimental conditions or classify specific types of particles, primarily focusing on silicone oil. That technique generally requires manual labeling of particle images by subject matter experts, a time-consuming and complex task. In this study, we created artificial datasets of silicone oil, protein particles, and glass particles that mimicked complex datasets of particles found in biopharmaceutical products. We used unsupervised learning techniques to effectively describe particle composition by sample. We then trained independent one-class classifiers to detect specific particle populations: silicone oil and glass particles. We also studied the consistency of the particle labels used to evaluate these models. Our results show that one-class classifiers are a reasonable choice for handling heterogeneous flow-imaging microscopy data and that unsupervised learning can aid in the labeling process. However, we found agreement among experts to be rather low, especially for smaller particles (< 8 µm for our Micro-Flow Imaging data). Given the fact that particle label confidence is not usually reported in the literature, we recommend more careful assessment of this topic in the future.

Aspherical, Nano-Structured Drug Delivery System with Tunable Release and Clearance for Pulmonary Applications

Addressing the challenge of efficient drug delivery to the lungs, a nano-structured, microparticulate carrier system with defined and customizable dimensions has been developed. Utilizing a template-assisted approach and capillary forces, particles were rapidly loaded and stabilized. The system employs a biocompatible alginate gel as a stabilizing matrix, facilitating the breakdown of the carrier in body fluids with the subsequent release of its nano-load, while also mitigating long-term accumulation in the lung. Different gel strengths and stabilizing steps were applied, allowing us to tune the release kinetics, as evaluated by a quantitative method based on a flow-imaging system. The micro-cylinders demonstrated superior aerodynamic properties in Next Generation Impactor (NGI) experiments, such as a smaller median aerodynamic diameter (MMAD), while yielding a higher fine particle fraction (FPF) than spherical particles similar in critical dimensions. They exhibited negligible toxicity to a differentiated macrophage cell line (dTHP-1) for up to 24 h of incubation. The kinetics of the cellular uptake by dTHP-1 cells was assessed via fluorescence microscopy, revealing an uptake-rate dependence on the aspect ratio (AR = l/d); cylinders with high AR were phagocytosed more slowly than shorter rods and comparable spherical particles. This indicates that this novel drug delivery system can modulate macrophage uptake and clearance by adjusting its geometric parameters while maintaining optimal aerodynamic properties and featuring a biodegradable stabilizing matrix.

Optimization of Flow Imaging Microscopy Setting Using Spherical Beads with Optical Properties Similar to Those of Biopharmaceuticals

Flow imaging microscopy (FIM) is widely used to characterize biopharmaceutical subvisible particles (SVPs). The segmentation threshold, which defines the boundary between the particle and the background based on pixel intensity, should be properly set for accurate SVP quantification. However, segmentation thresholds are often subjectively and empirically set, potentially leading to variations in measurements across instruments and operators. In the present study, we developed an objective method to optimize the FIM segmentation threshold using poly(methyl methacrylate) (PMMA) beads with a refractive index similar to that of biomolecules. Among several candidate particles that were evaluated, 2.5-µm PMMA beads were the most reliable in size and number, suggesting that the PMMA bead size analyzed by FIM could objectively be used to determine the segmentation threshold for SVP measurements. The PMMA bead concentrations measured by FIM were highly consistent with the indicative concentrations, whereas the PMMA bead size analyzed by FIM decreased with increasing segmentation threshold. The optimal segmentation threshold where the analyzed size was closest to the indicative size differed between an instrument with a black-and-white camera and that with a color camera. Inter-instrument differences in SVP concentrations in acid-stressed recombinant adeno-associated virus (AAV) and protein aggregates were successfully minimized by setting an optimized segmentation threshold specific to the instrument. These results reveal that PMMA beads can aid in determining a more appropriate segmentation threshold to evaluate biopharmaceutical SVPs using FIM.

Sensitivity and Uncertainty Analysis of Micro-Flow Imaging for Sub-Visible Particle Measurements Using Artificial Neural Network

PURPOSE

During biopharmaceutical drug manufacturing, storage, and distribution, proteins in both liquid and solid dosage forms go through various processes that could lead to protein aggregation. The extent of aggregation in the sub-micron range can be measured by analyzing a liquid or post-reconstituted powder sample using Micro-Flow Imaging (MFI) technique. MFI is widely used in biopharmaceutical industries due to its high sensitivity in detecting and analyzing particle size distribution. However, the MFI's sensitivity to various factors makes accurate measurement challenging. Therefore, in light of the inherent variability of the method, this work aims to explore the capabilities of an adopted coupled sensitivity analysis and machine learning algorithm to quantify the influencing factors on the formed sub-visible particles and method variability.

METHODS

The proposed algorithm consists of two interconnected components, namely a surrogate model with a neural network and a sensitivity analyzer. A machine learning tool based on artificial neural networks (ANN) is constructed with MFI data. The best fit with an optimized configuration is found. Sensitivity and uncertainty analysis is performed using this network as the surrogate model to understand the impacts of input parameters on MFI data.

RESULTS

Results reveal the most impactful reconstitution preparation factors and others that are masked by the instrument variabilities. It is shown that instrument inaccuracy is a function of size category, with higher variabilities associated with larger size ranges.

CONCLUSION

Utilizing this tool while assessing the sensitivity of outputs to various parameters, measurement variabilities for analytical characterization tests can be quantified.