Welcome to fokve | Welcome My Forum

[Hoca]

The BioPharma sector is under constant pressure to bring more novel therapies to market, more quickly and affordably. Attaining this goal requires improved operational control and continuous real-time quality assurance. These processes are the potential benefits of process analytical technology (PAT), i.e., the analysis and control of manufacturing through the timely measurement of critical quality and performance attributes of material and processed.

While significant advances in PAT have been accomplished with regard to the sector’s ability to monitor key processes and attributes, it is broadly acknowledged that more needs to be done with regard to utilizing those data for subsequent process controls.1 BioPharma’s future success rests on the use of advanced PAT, as well as platforms that facilitate the utility and application of PAT data.

Application of PAT widely accepted​

The application of PAT for biologics development and manufacturing is now widely accepted, thanks to its technological, economic, and regulatory advantages. Typically, chromatographic, spectroscopic, and/or mass spectrometric sensors are integrated into upstream and downstream unit operations in addition to a photosensitive sensor or redox electrode, to enable process monitoring and control.

A recent study evaluated a range of PATs in terms of their technological attributes, such as process understanding, control, and high-throughput capabilities, and business attributes such as simplicity of implementation, lead time, and cost reduction.2 On-line liquid chromatography, in-line Raman, and gas analysis techniques were identified as high value PAT, especially for upstream processing.

However, in reality, the choice of PAT depends on a company’s analytical needs. For example, we chose to invest in Raman-based PAT that enables real-time, high-frequency monitoring of cell culture processes and supports various control strategies, along with a biocapacitance-based PAT platform that monitors biomass and supports the automation of cell culture processes.

Hang Zhou, PhD, senior vice president and head of bioprocess research and development , WuXi Biologics. [WuXi Biologics]

Raman spectroscopy has become a first-choice PAT for monitoring and controlling upstream bioprocesses because it facilitates advanced process control and enables consistent process quality.3 Recent applications of Raman, from basic cell and molecular biology to downstream process monitoring, illustrate its value throughout a biologic product’s lifecycle.3 An all-in-one Raman-based PAT platform centred around Raman analyzers can monitor multiple parameters with high accuracy and at a high frequency.

A Raman-based PAT platform is used at our sites in process development labs, pilot plants, and GMP manufacturing sites. We have applied this technology in upstream culture modes with traditional fed-batch and advanced intensified and perfusion culture as well as in downstream purification processes.

Raman is non-invasive, so it has minimal impact on the culture system, and it has a wide application range across a variety of product types (e.g., monoclonal antibodies, bispecific antibodies, fusion proteins) and at a range of scales (e.g., 3 to 2000L bioreactors). Its measurement capabilities are extensive, including:

• Viable cell density (VCD), average diameter, cell viability;
• Quality attributes (e.g., purity, charge variants, N-glycan);
• Titer, pH, pCO2;
• Glucose, lactate, amino acids, osmolality;
• Na+, K+, NH4+, Fe3+;
• Other custom tests.

While Raman can be used to monitor several of these parameters simultaneously, biocapacitance is used for monitoring cell-related parameters such as VCD with higher accuracy. In our study, we employed a biocapacitance-based PAT platform that provides real-time monitoring of biomass and supports automation of cell culture processes. In alignment with general trends towards new modalities such as virus-based and cell-based therapies, biocapacitance provides new opportunities for process development and control.

Data application ​

The collection of large amounts of data generated by bioprocesses in an accurate and reliable way and their interpretation for implementation are key challenges. Therefore, a comprehensive PAT roadmap should address data analysis, aggregation, visualization, and the smart utility of those data for greater process understanding.4

For example, the signals collected by single- or multi-channel Raman analyzers can generate new data every five minutes, but those signals are merely raw laser intensities, which are incomprehensible until they are converted to usable data. As illustrated in Figure 1, this interpretation requires chemometric methods that are integrated into the PAT platform. Raman-based PAT must have robust technology support to enable advanced model predictive outcomes. Similarly, any data collected through a biocapacitance-based PAT for monitoring biomass and cell culture processes will be of no value in the absence of a system that analyzes and interprets the source data accurately and quickly.

It is no coincidence that advances in PAT technologies have taken place alongside the evolution of big data, the Internet of Things (IoT), and machine learning. In order to obtain sufficient insights from real-time PATs, these technologies must be linked to the IoT for rapid analysis and interpretation. A central PAT and IoT platform can empower human–machine and machine–machine interactions. PAT acquires data at appropriate process points, but cyber-physical systems are needed to facilitate automated analysis, visualization, and systematic storage of data.

Real-time interpretation of data is now an achievable goal. The PAT & IoT platform illustrated in Figure 1, which includes machine learning, modeling, and data management, enables the real-time monitoring of more parameters with higher accuracy. Combining real-time data with extensive experience in bioprocesses, which provides a better interpretation of real-time high-frequency parameters, results in an even deeper understanding of bioprocesses and the ability to make decisions that achieve optimal outcomes.

Figure 1: Raman-based PAT must be supported by a robust technology platform to enable advanced model predictive outcomes. [WiXi Biologics]

Improving bioprocess monitoring and control ​

In contrast to conventional off-line testing in which samples were typically analyzed remotely in testing labs and result generation could take several hours or days, in-line analysis implements the analytical tool in the bioprocess to provide data in real-time as the process is occurring. This enables integration with bioreactors, purification columns, analytical instruments, and software for seamless monitoring of analytes.

For example, we recently published a paper explaining how in-line Raman-based PAT can be applied in monitoring product quality attributes (PQA) in intensified perfusion culture.5 Real-time Raman spectra were collected from a 3L benchtop intensified perfusion cell culture system, and the data were correlated with historical PQA data from purified samples. Modeling results showed that the Raman spectrum could reflect the PQA changes as a function of elapsed time, could capture the PQA profiling trends, and the models were maintained irrespective of process changes like pH strategy, perfusion rate, or additional feeding.5

Biocapacitance-based PAT can also contribute to bioprocess monitoring. For example, timely detection of cell apoptosis onset allows opportunities for preventive controls that ensure high productivity and consistent product quality. We have been able to perform in-line monitoring of cell apoptosis, with earlier detection of apoptosis being achieved with biocapacitance-based PAT than with a conventional trypan blue methods. Although these observations are still considered a proof-of-concept stage, they appear to be supported by other scientists who have also reported using capacitance spectroscopy to capture apoptosis‐related cellular changes and quantify the percentage of dying cells.6

Improving bioprocess control can only be achieved with comprehensive, real-time insights into the state and stage of bioprocesses inside a bioreactor. Comprehension of bioprocesses combined with IoT platforms can grant a deeper understanding of the parameters monitored by PAT. For example, Raman spectroscopy has emerged as a robust technique to control mammalian cell culture process in real-time.

In a recent study performed by our process development team, a state-of-the-art perfusion cell culture and Raman integrated system achieved auto-control of VCD, with profound effects on in-process product quality control.7 The results helped to deepen our understanding of perfusion cell culture process control, broaden the application for Raman spectroscopy in continuous manufacturing, and offer a powerful control strategy that might benefit the shaping of product quality.

Optimizing models for accurate and precise monitoring​

Advanced PAT permits better live monitoring of product purity, cell population growth and health, heterogeneity, and functionality. Added to this, with appropriate calibration models, it is possible to perform simultaneous monitoring of different parameters to support robust biomanufacturing. A common challenge with this approach is that chemometric models for spectrometers tend to be process specific. Therefore, generic models are useful, especially for CRDMOs that need to handle a variety of projects within very tight timelines and that rigorously protect the IP rights of their clients.

For example, a recent publication by our process development team outlined a novel approach for developing generic metabolic Raman calibration models for in‐line glucose and lactate concentrations analysis.7 This is just one example of many generic models that we have established with data from solution titration studies or cell culture runs using our internal cell lines. The results demonstrated a high prediction accuracy and established a generic metabolic calibration model that could be extended to other metabolites.

Improving bioprocess control can only be achieved with comprehensive, real-time insights into the state and stage of bioprocesses inside a bioreactor. Comprehension of bioprocesses combined with IoT platforms can grant a deeper understanding of the parameters monitored by PAT. [Ultramansk/Getty Images]

Generic models such as this can be applied in a “plug and play” approach with acceptable accuracy but, in reality, they are just a starting point. When a researcher encounters a new cell line, product, culture medium or culture mode, they typically start with a generic model to provide acceptable predictions for new processes. However, we would then expect to develop a customized model to better fit specific bioprocesses, in parallel with development of the upstream process, and while continuously optimizing the model. Project-specific models can be continuously optimized as more data become available, or as we improve the modelling platform.

Finally, we would expect to calibrate, evaluate, and validate the model to maintain its performance throughout the life cycle of the corresponding bioprocess, through process development, technology transfer, process scale-up, and manufacturing.

In another publication, we described an approach to improving a Raman calibration model’s cross-scale prediction capabilities in cell cultures.8 The model was initially developed for just four 3L runs but was refined with data from a 50L scale-up. The integration of the 50L run data was found to be an effective strategy for improving prediction accuracy for a 1000L run. Additionally, it was shown that spectral variations stemmed from differences between the instruments, so two pairwise spectral transformation methods were added to the model to enhance it predictive accuracy in large-scale manufacturing.

Advancing continuous biomanufacturing ​

As in-line PAT tools are capable of real‐time analytical data generation, they enable continuous processing from one operation to the next. This involves connecting PAT systems to other automation systems, and consequently implementation of smart technology platforms. Two recent peer-reviewed publications have described our Ultra‐high Productivity integrated technology platform for continuous biomanufacturing (WuXiUP

).9,10 Through process intensification, the platform enables continuous manufacturing of almost any type of biologic, including monoclonal antibodies, fusion proteins, bispecific antibodies, and enzymes, and delivers bioprocesses with ultra‐high productivity.

Compared with continuous manufacturing in other industries, biologics production has more complex unit operations. PAT tools and automatic control systems link adjacent unit operations. Moreover, in order to seamlessly connect upstream and downstream processes, a DCS (Distributed Control System) collects real-time data from PAT, analyses the data, and provides instructions. The upstream output (like titer and volume) is dynamic, potentially resulting in fluctuating productivity and step yield.

To alleviate such fluctuation, an ingenious control solution is to trigger different capture chromatography methods based on real-time titer and load volume (weight) via a DSC communicating with the Raman platform and an automated liquid chromatography system. In that case, we can trade off load density and load time of capture step to achieve optimal daily productivity and capture resin usage.

Process monitoring and control support the maintenance of a state of control (ICH Q10) during production. For example, in continuous manufacturing, the impurities content of harvested cell culture may vary over time, but the quality of final drug substance must be consistent. On-line HPLC with an appropriate sampling frequency can monitor the monomer purity and divert potential non-conforming material from the product stream.

The Ultra‐high Productivity platform has demonstrated its capability in boosting the productivity of various biologics compared with traditional fed-batch culture (Figure 2). The capital expenditure and facility footprint are greatly reduced, and high flexibility of manufacturing is feasible. Compared with traditional perfusion culture which lasts 1–2 months, the platform can reduce media consumption and resin amount, while facilitating the manufacturing turnaround. Consequently, it can expedite product launch while lowering capital and operating costs.

Figure 2. Our integrated technology platform for continuous biomanufacturing (WuXiUP

) achieves ultra-high productivity for all kinds of pharmaceutical proteins with high cell densities and better product quality within similar production culture timelines. [Wuxi]

Conclusions​

In-line PAT can facilitate faster process development, meet specific demands, and ensure more robust manufacturing, while helping to secure product quality. However, while advanced PAT technologies are now widely available and commonly used, extensive expertise and technology platforms are required to analyze, interpret, and act on the data collected. In order to apply PAT strategies for fast process development and monitoring of large-scale processes, it is important to create business advantages with modelling and real-time data interpretation using proven effective technology platforms.

The foundation for future success in BioPharma lies in the use of advanced PAT, coupled with the implementation of smart technology platforms, machine learning algorithms for data analysis and interpretation, and the connection of all these systems for the deployment of automated feed-back control. This will not only enable improved bioprocess controls, robust biomanufacturing, and better continuous bioprocessing but it will enable us to do it more quickly and more affordably, allowing us to bring more novel therapies to market more reliably.



Hang Zhou, PhD, is senior vice president and head of bioprocess research and development (BRD) at WuXi Biologics.



References

1Rathore AS et al. Anal Bioanal Chem 2010;398:137-154.

2Gillespie C et al. Biotechnol Bioeng 2022;119:423-434.

3Esmonde KA et al. Anal Bioanal Chem 2022;414:969-991.

4Wasalathanthri DP et al. Biotechnol Bioeng 2020;117:3182-3198.

5Liu Z et al. Biochem Eng J 2021;173:108064.

6W S et al. Biotechnol Bioeng 2022;119: 857-867.

7Chen G et al. BioChem Eng J 2021;172:108063.

8Lang Z et al. AIChE J. 2024;e18608.

9Zhou H et al. Biotechnol Bioeng 2021;118: 3618-3623.

10Zheng X et al. Biotechnol Prog 2024 Jul 9:e3487



The post WuXi Biologics Stands on PAT appeared first on GEN - Genetic Engineering and Biotechnology News.