Hadi El Radi

Background: Process analytical technology (PAT) enhances product quality by monitoring and controlling critical quality attributes (CQAs), offering better insights into process performance. Raman spectroscopy is a promising PAT tool to support the development of control systems for continuous and automated processes. In this study, Raman spectroscopy was used to monitor in real-time ultra- and diafiltration (UF/DF) of adeno-associated virus (AAV) and lentiviral vector (LV). Method: The incorporation of a Raman probe in a tangential flow filtration (TFF) system, when operated in open loop, enabled spectral acquisition during a five-fold concentration of clarified bulk followed by a buffer exchange to PBS with five diafiltration volumes. Results: Principal Component Analysis reduced the dimensionality of the training set, and a Partial Least Squares model was developed for each parameter in each phase. After validation, Raman monitoring platforms showed good performance in predicting CQA of both viral vectors. Batch-to-batch variability was identified as the principal source of spectral variation, and its impact on UF/DF operation was observed. Discussion: This study emphasises the potential of Raman spectroscopy as a PAT for monitoring and control strategies in the downstream processing of viral vectors. Conclusions: These monitoring platforms enhance understanding of TFF’s operation and enable timely decision making by providing real-time information on multiple parameters.

Process analytical technology (PAT) is a framework that encourages the implementation of analytical tools for real‐time monitoring of critical batch components during the production of pharmaceuticals. In mammalian cell cultures, Raman spectroscopy as a PAT tool has demonstrated its value for accurate in‐line monitoring of various process parameters as well as quality attributes. However, as process development cultures are transferred to manufacturing scales, the small‐scale models are often adjusted either with new calibration data or with transfer learning algorithms. This model adjustment step makes the overall process transfer to manufacturing less straightforward from an analytical standpoint, requiring additional resources and specific data treatment expertise to fine tune the models. In this work, we compare a specific spectra processing approach against existing approaches reported in the literature for processing Raman spectra in the context of scale‐up challenges. The proposed novelty is a preprocessing strategy which computes spectra normalization based on the 3100–3700 cm ⁻¹ band. Data generated at 3 L scale were used to monitor a 3000 L bioreactor without model recalibration or transfer learning with equivalent accuracy as a same scale 3 L bioreactor monitoring. While broader validation case would be warranted, these results demonstrate a direct scale‐up model implementation in large scale manufacturing under the studied conditions.