Halo Labs proposes to develop âPartelligenceâ a particle ID technique that enables accurate and rapid identification of contaminating particles in biopharmaceutical formulations. Protein therapeutics currently represent between 15 and 30% of the overall pharmaceutical market. The primary concern for this class of therapeutics is that they can elicit an immune response from patients who develop anti- drug antibodies. The drugâs effect is therefore eliminated between 1 and 10 percent of patients who return to their original disease state. The presence of particulate matter in these therapeutics (e.g. shed glass from a syringe or a protein aggregate) can enhance this immune response and, due to the patient safety risk the FDA regulates the amount of particles that can be present. There are always some number of particles in each injected sample, and although their presence can be detected, they donât know what the particles actually are. A QC tool that can identify the particles would help manufactures trace them back to their source (e.g. a bad lot of syringes) and eliminate them. Partelligence aims to make particle identification routine in biopharma QC. The technology builds off our current instrument, Horizon, which was launched in mid-2017 and already sold to some of the worldâs largest pharmaceutical companies. The technique works by analyzing several combinatorial features including size, morphology, optical contrast, and intrinsic fluorescence, and in this proposal we will test which features are key to enable the most accurate and rapid particle recognition. To date, we have performed feasibility experiments validating our ability to identify a few commonly found particles in biopharma solutions. Given this, our goals in Phase I are to expand on these studies by building a comprehensive training set and by testing a number of different algorithms. We will first start with reference samples, and then move to real biopharmaceutical samples provided by our pharma collaborators. At the end of the study, we will do a feasibility analysis to determine if the throughput, specificity and reliability meets the needs of the industry.
Public Health Relevance Statement: Narrative We propose to evaluate a particle recognition technique to enable accurate and rapid identification of unwanted contaminating particles in biopharmaceutical formulations. Successful development of this analytical technique would improve bioprocess control by identifying dangers early on in development and throughout the manufacturing process, resulting in safer protein drugs, reduced recalls and shortened time to market.
Project Terms: Adverse effects; Algorithmic Analysis; Algorithms; Allergic Reaction; Alpha Particles; Anaphylaxis; Antibodies; Back; base; Biological Neural Networks; Biological Products; bioprocess; Cessation of life; Classification; Clinic; combinatorial; Dangerousness; Data; Development; Disease; Drug Industry; Effectiveness; Event; experience; experimental study; Failure; Fluorescence; fluorescence imaging; Forensic Medicine; forest; Formulation; Generations; Glass; Goals; Health; Image; Image Analysis; imaging modality; Imaging Techniques; Immune; Immune response; immunogenicity; improved; Industrialization; Industry; instrument; Investigation; Label; Learning; Letters; Machine Learning; manufacturing process; Membrane; microscopic imaging; Morphology; Optics; particle; Particulate; Particulate Matter; patient response; patient safety; Patients; Performance; Pharmaceutical Preparations; Pharmacologic Substance; Phase; predictive modeling; pressure; Process; protein aggregate; Proteins; Raman Spectrum Analysis; Regulation; Resistance development; Risk; Rubber; Sampling; Scientist; Shapes; small molecule; Source; Specificity; Spectroscopy, Fourier Transform Infrared; Syringes; System; Techniques; Technology; Testing; Therapeutic; therapeutic protein; Time; tool; Training; tr
