Umeå University's logo

The biopharmaceutical industry is transitioning towards more efficient and cost-effective production methods, driven by the need for more affordable treatments. Process intensifications techniques, such as perfusion are the means by which the biopharmaceutical industry is trying to lower costs, enhance productivity and product quality, and reduce facility footprints. The dynamic nature of perfusion processes presents considerable challenges for real-time monitoring and control, requiring the advancement of process analytical technologies (PAT) with at-line, in-line, or on-line capabilities. Raman spectroscopy has emerged as a pivotal technology, providing real-time, noninvasive measurements of multiple analytes simultaneously, contingent upon the availability of sufficient data for calibration modeling. This study outlines the implementation of an automated data generation workflow for Raman calibration modeling within a high-throughput perfusion miniature bioreactor system, specifically the Ambr 250 HT Perfusion. Additionally, we demonstrate the effectiveness of Raman calibration models in monitoring various cell culture parameters within perfusion cultivations, spanning multiple cell lines and monoclonal antibody products. Finally, we present the feasibility of a Raman-based bleed-rate control system and how it compares to the conventional cell counter-based approach.

Biopharmaceuticals have transformed modern medicine over recent decades due to their efficacy in treating various severe diseases such as cancers, autoimmune disorders, and genetic conditions. While representing a wide class of different treatments, biopharmaceuticals are commonly produced through genetically engineered living organisms. In practice, the therapeutics of interest are typically manufactured through cell culture in bioreactor systems. However, this process is complex as it relies on fragile biological systems that need to be monitored and tightly controlled. To achieve such monitoring and control, process analytical technology (PAT) is needed. Spectroscopic methods, such as Raman and bio-capacitance spectroscopy, have been presented as potential PAT candidates due to their real-time, non-invasive measurement capabilities of various critical cell culture parameters (e.g., glucose and cell concentration). However, the use of spectroscopic sensors as PAT tools greatly depends on robust multivariate calibration models. These models are required to translate spectral data into actual process parameter values.

This thesis addresses fundamental challenges in calibration modeling for PAT implementation in biopharmaceutical manufacturing. Specifically, the use of Raman and bio-capacitance spectroscopy as PAT tools in upstream cell culture is investigated. We explore how biological variation impacts the transferability and robustness of Raman-based monitoring models in Paper I. In Paper II, we extend beyond monitoring by demonstrating how the combination of classical chemometric calibration models and simplified mechanistic models can yield accurate forecasts during cell culture, effectively developing a predictive decision support system. Paper III explores calibration data generation by presenting an automated workflow using a miniature-scale high-throughput bioreactor system. Its usefulness is further demonstrated by developing and deploying calibration models for monitoring and control in perfusion culture. Finally, Paper IV explores a novel validation framework for calibration models that tests the specificity and robustness of developed models.

Recombinant adeno-associated virus (rAAV) vector production is a complex process in which the robust cultivation of human embryonic kidney cells (HEK293) plays a critical role in generating high-quality viral vectors. Tracking the viable cell concentration (VCC) during upstream production is essential for process monitoring and for implementing actions that ensure optimal process management. The advent of inline capacitance probes has introduced a crucial process analytical technology (PAT) tool for real-time VCC measurement. Here, we present the development and application of a method for real-time monitoring of VCC in HEK293-based rAAV vector production. In a first step, BioPAT Viamass probes were used to record capacitance data of individual 10 L rAAV-8 batches within a frequency range of 50 kHz–20 MHz. Based on the capacitance data, a linear single-frequency model and an orthogonal partial least square (OPLS) multifrequency model for VCC prediction were developed. Subsequently, these models were deployed inline, and predictions were exposed into BioPAT MFCS bioprocess control software, enabling real-time VCC monitoring in subsequent rAAV-8 production batches. In addition, the continuous VCC signal was used as input for an exponential cell growth model that was deployed inline to provide accurate real-time forecasting of the transfection time point. To the best of our knowledge, this is the first example of inline deployment of VCC and Time-Till-Transfection predictive models to the bioprocess control system for real-time monitoring and forecasting of these parameters in HEK-cell-based transient rAAV vector production.

Raman spectroscopy is widely used in monitoring and controlling cell cultivations for biopharmaceutical drug manufacturing. However, its implementation for culture monitoring in the cell line development stage has received little attention. Therefore, the impact of clonal differences, such as productivity and growth, on the prediction accuracy and transferability of Raman calibration models is not yet well described. Raman OPLS models were developed for predicting titer, glucose and lactate using eleven CHO clones from a single cell line. These clones exhibited diverse productivity and growth rates. The calibration models were evaluated for clone-related biases using clone-wise linear regression analysis on cross validated predictions. The results revealed that clonal differences did not affect the prediction of glucose and lactate, but titer models showed a significant clone-related bias, which remained even after applying variable selection methods. The bias was associated with clonal productivity and lead to increased prediction errors when titer models were transferred to cultivations with productivity levels outside the range of their training data. The findings demonstrate the feasibility of Raman-based monitoring of glucose and lactate in cell line development with high accuracy. However, accurate titer prediction requires careful consideration of clonal characteristics during model development.