(424a) Risk Analysis-Based Fault Diagnosis in mRNA Biotherapeutics Manufacturing: A Real-Time Benchmark Approach

In the current landscape of the continuous biotherapeutics manufacturing literature, fault detection and diagnosis methods are evaluated in highly simplified simulation studies or on proprietary industrial datasets [1–3], with the former missing key aspects of the real industrial systems and latter hindering result reproducibility. Within the context of biotherapeutics manufacturing, especially for mRNA production [4,5], the published simulation models that are reasonably realistic in describing the underlying mechanistic phenomena do not include models of the faults affecting components, sensors, and actuators.

This study presents a systematic framework for applying risk analysis methodologies to biotherapeutics manufacturing, and provides a real-time benchmark for the design and evaluation of risk-based fault diagnosis systems. A simulated mRNA biotherapeutics manufacturing process is described that emulates a continuous biomanufacturing system that has industrial components and devices. Communication protocols mimic those of a Supervisory Control and Data Acquisition system used in the pharmaceutical industry. A general problem formulation for risk-based fault diagnosis in biotherapeutics manufacturing systems is followed by its application to the mathematical model of the mRNA benchmark problem. We provide a calibrated analytical model encompassing various fault scenarios, informed by observations from the real mRNA biotherapeutics manufacturing system. Lastly, we compare datasets generated from experiments incorporating the aforementioned faults, thereby contributing to the design of enhanced fault detection diagnosis strategies in biotherapeutics manufacturing systems.

This research was supported by the U.S. Food and Drug Administration under the FDA BAA-22-00123 program, Award Number 75F40122C00200.

References:

[1] Cenk Ündey, Sinem Ertunç, Thomas Mistretta, and Bryan Looze. Applied advanced process analytics in biopharmaceutical manufacturing: Challenges and prospects in real-time monitoring and control. Journal of Process Control, 20(9), 1009-1018, 2010.

[2] Carina L. Gargalo, Simoneta Caño de las Heras, Mark Nicholas Jones, Isuru Udugama, Seyed Soheil Mansouri, Ulrich Krühne, and Krist V. Gernaey. Towards the Development of Digital Twins for the Bio-manufacturing Industry. In Digital Twins: Tools and Concepts for Smart Biomanufacturing, edited by Christoph Herwig, Ralf Pörtner, and Johannes Möller, Springer Nature, Switzerland, pages 1–34, 2021.

[3] Vincent Brunner, Manuel Siegl, Dominik Geier, and Thomas Becker. Challenges in the development of soft sensors for bioprocesses: A critical review. Frontiers in Bioengineering and Biotechnology, 9, 722202, 2021.

[4] Heribert Helgers, Alina Hengelbrock, Axel Schmidt, and Jochen Strube. Digital twins for continuous mRNA production. Processes, 9(11), 1967, 2021.

[5] Damien van de Berg, Zoltán Kis, Carl Fredrik Behmer, Karnyart Samnuan, Anna K. Blakney, Cleo Kontoravdi, Robin Shattock, and Nilay Shah. Quality by design modelling to support rapid RNA vaccine production against emerging infectious diseases. npj Vaccines, 6, 65, 2021.