Cell therapies show promise for treating rare, severe, and prolonged diseases like cancers, inflammatory diseases, and chronic degenerative disorders. These innovative therapies involve the manipulation of cells and tissues in vitro using nutrient-rich stem cell cultures, making them vulnerable to microbe contamination. Cell therapy product (CTP) manufacturers need quick, effective ways to ensure that cells are free from contamination before being administered to patients. Delays due to contamination testing of CTPs can have life-threatening consequences for critically ill patients awaiting treatment. To provide microbial contamination detection early in the CTP manufacturing process, researchers from the Critical Analytics for Manufacturing Personalized-Medicine (CAMP) group at the Singapore-MIT Alliance for Research and Technology (SMART), in collaboration with MIT, A*STAR Skin Research Labs, and the National University of Singapore, developed a machine learning-aided, UV absorbance spectroscopy technique. The technique can be used at different stages in the CTP manufacturing process to monitor cell cultures continuously in real time, ensuring the safety of CTPs. SMART CAMP senior research engineer Shruthi Pandi Chelvam uses the UV absorbance spectrometer to measure the absorbance spectra of cell culture samples. Courtesy of SMART CAMP. “Specifically, our method supports automated cell culture sampling at designated intervals to check for contamination, which reduces manual tasks such as sample extraction, measurement, and analysis,” professor Rajeev Ram said. “This enables cell cultures to be monitored continuously and contamination to be detected at early stages.” The spectroscopy method can detect microbial contamination in less than 30 minutes, with minimal sample preparation and The researchers used a one-class support vector machine (SVM) to analyze the absorbance spectra of cell cultures and predict whether a sample was sterile or contaminated. They used mesenchymal stromal cell (MSC) cultures, which are widely used in cell therapy, as a demonstrator. To train the SVM model, they used the absorbance spectra of sterile cell culture samples measured by a commercial spectrometer. Using an anomaly detection approach, the team identified spectral differences in the region of interest and predicted contamination in test cell culture samples. They also identified the probable principles underlying the SVM model’s ability to predict contamination. In tests, the machine learning-aided method was able to detect contamination at the 21-hour timepoint when 10 colony-forming units of E. coli were spiked in an MSC culture. The researchers demonstrated the ability to detect down to 10 colony-forming units for 7 microorganisms using the new method. They demonstrated the robustness of the approach by testing it across different commercial donor MSCs. By measuring UV light absorbance of cell culture fluids and using machine learning to recognize light absorption patterns associated with microbial contamination, this testing method could reduce the time required for sterility testing, enabling cell therapies to get to patients faster. “Traditionally, cell therapy manufacturing is labor-intensive and subject to operator variability,” Ram said. “By introducing automation and machine learning, we hope to streamline cell therapy manufacturing and reduce the risk of contamination.” The new approach could offer significant advantages over traditional sterility tests, which can take up to 14 days, and advanced techniques like rapid microbiological methods, which entail complex processes and required highly skilled workers. It provides label-free detection of cell contamination, eliminating the need to stain cells for labeling. It also eliminates the invasive process of cell extraction. It delivers results in under one-half hour, providing an intuitive, rapid “yes/no” contamination assessment. The technique’s simple workflow requires no additional incubation period or growth enrichment mediums. Nor does it require specialized equipment, which keeps its cost low. “This rapid, label-free method is designed to be a preliminary step in the CTP manufacturing process as a form of continuous safety testing, which allows users to detect contamination early and implement timely corrective actions, including the use of RMMs only when possible contamination is detected,” researcher Shruthi Pandi Chelvam said. “This approach saves costs, optimizes resource allocation, and ultimately accelerates the overall manufacturing timeline.” In the future, the researchers plan to broaden the machine learning-assisted spectroscopy method to encompass a wider range of microbial contaminants, specifically those representing Current Good Manufacturing Practices (cGMP) environments and previously identified CTP contaminants. Additionally, the model’s robustness could be tested across more cell types, apart from MSCs. Beyond CTP manufacturing, this method could also be used by the food & beverage industry for microbial quality control testing to ensure food products meet safety standards. The research was published in Scientific Reports (www.doi.org/10.1038/s41598-024-83114-y).