Machine learning aided UV absorbance spectroscopy for microbial contamination in cell therapy products

Area of Science:

• Biotechnology
• Analytical Chemistry
• Machine Learning

Background:

• Cell therapy product (CTP) manufacturing requires stringent microbial contamination monitoring.
• Current methods can be time-consuming, delaying product release and safety assessments.

Purpose of the Study:

• To demonstrate the feasibility of machine learning-aided UV absorbance spectroscopy for rapid, in-process microbial contamination detection in CTPs.
• To establish a label-free, rapid, and low-volume method for early contamination identification.

Main Methods:

• Utilized a one-class support vector machine (SVM) to analyze UV absorbance spectra of cell cultures.
• Developed a method for rapid output (<30 minutes) with minimal sample preparation and volume (<1 mL).
• Tested detection limits by spiking various microbial organisms into mesenchymal stromal cells supernatant from multiple donors.

Main Results:

• Successfully detected microbial contamination at low inoculums (10 Colony Forming Units [CFUs]).
• Achieved mean true positive and negative rates of 92.7% and 77.7%, respectively.
• Demonstrated contamination detection comparable to compendial USP <71> testing, with detection of E. coli at 21 hours.

Conclusions:

• Machine learning-aided UV spectroscopy is a feasible approach for in-process microbial contamination detection in CTP manufacturing.
• This technique offers rapid, label-free, and sensitive monitoring for enhanced CTP safety.
• Potential application for real-time, continuous culture monitoring across various CTP manufacturing stages.