Continuous Bioprocessing: PAT Strategies in Support of Process Monitoring and Control to Enable Rapid Product Release
Posted on April 11,2017
Continuous bioprocessing offers potential to enhance productivity and product quality uniformity while simultaneously decreasing facility footprint and associated operational overhead. Advances in technology and increasing commercial pressures are leading to an increased interest in continuous processing across the biopharmaceutical sector. A number of companies are exploring and advancing continuous bioprocessing and this presents a range of opportunity and challenges, including the use of Process Analytical Technology (PAT) for process characterization, process control, and process robustness, in support of a Rapid Product Release (RPR) strategy.
The predominant production method for biopharmaceutical drug substance manufacturing has been based on batch processing. However, as the industry matures, the drivers for improved process economics and product quality uniformity1 have led to the evaluation of the business impact2 and demonstration3 of continuous processing. Several benefits are expected as a result of shifting from a batch to continuous manufacturing processing platform, these include improvements in productivity and cycle time along with reduced capital and equipment costs4.
The realization of these improvements through the implementation of continuous processing is dependent largely on the appropriate use of PAT. The gap between currently available technologies and its critical role has become almost synonymous in reviews describing the implementation of continuous manufacturing1, 3, 4, 5, 6.
The use of PAT is fundamental in ensuring a continuous process can be controlled to reach and maintain “steady state”. In addition to the element of control, the use of PAT is acknowledged as an enabler to the implementation of real-time release testing (RTRt)7, 8. By definition, RTRt is the ability to evaluate and ensure the quality of in-process and/or final product based on process data, which typically include a valid combination of measured material attributes and process control9. Here, the term Rapid Product Release (RPR) is considered a step forward in the application of RTRt and practical through near-term solutions that can reduce the release testing of a batch to only a few days.
How can PAT strategies for monitoring and control enable RPR in continuous bioprocessing?
The Elements of a Rapid Product Release Strategy and the Role of PAT
The foundational elements of a RPR strategy (see Figure 1) are based on the concepts described in the International Conference on Harmonization (ICH) Guidelines Q8, Q9, Q10, and Q119, 10, 11, 12. These elements are independent of the manufacturing mode, batch or continuous, and consist of Process Characterization, Control Strategy, and Process Robustness. Each element builds upon the prior until a holistic understanding of the process and the correlations to product quality are achieved and, in turn, can be used to enable an alternative analytical strategy.
The analytical strategy (see Figure 1) leverages the application of in-process and parametric control, advancing methods, and validating out of testing to support a case for RPR. While the approach can be similar for batch or continuous bioprocessing, there is likely more benefit to applying rapid product release to continuous processing. The benefits of RPR for continuous processing include, but are not limited to, handing an increase number of batches, dealing with fully integrated zero pool systems, and manufacturing high volumes at risk.
Process Characterization is achieved through a Quality by Design (QbD) process development approach. By varying experimental conditions within the design space during process development, the impact on the Critical Quality Attributes (CQA) and the Quality Target Product Profile (QTPP) becomes well understood. The role of PAT during process characterization includes both analysis and control of the incoming raw material variation and process variation to ensure continuous real time quality assurance7.
The role of process characterization in a RPR strategy is to establish an appropriate control strategy based on the process understanding achieved during development. Advancements supporting process characterization include high throughput screening methods, Multivariate Data Analysis (MVDA), and the integration of multiple unit operations in mechanistic models8.
The main challenge to process characterization during development is the short development cycle as products are advanced to market. Here, the application of miniaturized/high throughput systems have aided in the application of Design of Experiment (DOE) studies. Miniaturization though, can be at odds with the use of PAT tools as sample volumes and interfaces diminish. The application of scale-independent PAT tools can be a successful approach in mitigating the risks associated with process scale-up and control. The use of MVDA in process development can be instrumental in identifying and resolving variations such as scale dependencies, medium feed rates, operational errors or instrument failure13.
Another challenge for process characterization is the fact that a process is typically developed using a limited number of raw material lots. The ability of a process to deal with raw material variation is a challenge to characterize during process development. In some cases, processes under development may only be manufactured with a small number of vendor lots making raw material characterization an important element in process characterization. The development and application of attribute based control strategies will help in the ability to adjust the process in response to raw material variation particularly at commercial scale.
Process characterization ultimately supports the risk assessment process and classification of critical processing parameters (CPPs) and critical quality attributes (CQAs). PAT forms part of the solution for monitoring and controlling these CPPs and CQAs to ensure that the QTPP is met.
Process characterization is the basis for defining a process control strategy. A fully integrated continuous process will be reliant on rapid analytical methods to ensure both the CPPs identified during process characterization remain in control and measure in real-time information about product quality attributes4. It is possible to break down the control strategy into six control blocks. These blocks consist of direct monitoring and control of product quality attributes or functionally linked parameters, direct monitoring and control of process performance attributes or functionally linked parameters, control of raw materials, and facility and equipment controls. Examples of currently available PAT solutions within each control block are listed in Table 1.
The key to using PAT in support of RPR is that it enables a dynamic control strategy for producing consistent product even when process, facility, or materials vary21. With limited experience, on the part of both industry and regulatory agencies, control strategies for continuous processing will continue to evolve. As part of that growth, control strategies supporting the implementation of use of RPR should be considered from the outset. The Office of Pharmaceutical Quality (OPQ) has initiated a regulatory science and research program on continuous manufacturing with focus areas in integrated process modelling, critical material attributes, and advanced process control strategies. This program is expected to provide support implementation of continuous manufacturing by providing the regulatory policies reflecting state of the art manufacturing science.22
Early integration of PAT during development for process characterization, monitoring, and control can help to promote and overcome the challenge of bolt-on applications post commercialization of a new manufacturing process.
Complementary to the process control strategy is the demonstration of process capability, i.e., the ability to deliver the QTPP batch after batch. Process robustness is the product of a well characterized process with appropriate control strategy. The effectiveness of the control strategy can be monitored using classical statistical process control, multivariate statistical process monitoring, and process capabilities indices while maintaining the QTPP.
The batching scheme for continuous processing may be based on volume, duration, or another measure of maturity. The use of process capability indices and process monitoring techniques will need to be adapted for a continuous process.
Due to variations in starting material, as well as other variations in critical and non-critical process parameters, a robust process is less susceptible to drifts or transient upsets that can occur during the long runs of continuous processing. Real-time data can be used to effectively monitor a continuous process to identify and highlight the occurrence of unwanted trends or process upsets and to enable optimal control and rapid troubleshooting23.
The current testing paradigm for most mAb drug substance manufacturing revolves around the paradigm of “Make it, Test it, Release it”. This paradigm ultimately ensures the product leaving the manufacturing facility has met the required endpoints for release and achieves a quality consistent with the filed limits. The current state is fundamentally reliant on the fixed control of input materials and in-process controls in the delivery of the target product profile. This strategy has been successful in the past but has significant costs, technical complexity, time implications, and associated risk. An added complexity when considering this strategy for continuous manufacturing is the strategy for defining a “batch”. During continuous mode operation a “batch” could be based on time, volume, or the change out of a raw material component. In contrast, during batch processing, a common schedule for a production bioreactor is to operate for twelve days, allowing two days for harvest and equipment turnaround, thus, producing one terminal harvest batch every 14 days. Assuming a daily3 harvest and batching, a continuous process would result in a 14:1 increase in the number of release tests.
In the example above, simply building out the end product release testing capabilities through new equipment purchases, laboratory efficiencies, and/or staffing could support daily batching in continuous manufacturing mode on a standard release schedule. Furthermore, through testing optimization and analytical performance improvements, a state of “on-demand” test results is likely possible for many end product release parameters. While there are advantages to “on-demand” testing results, the paradigm of “Make it, Test it, Release it” persists. To obtain the full benefit and efficiencies of continuous processing, a new paradigm is required that allows real-time or rapid monitoring of quality attributes in the process.
Taking a tactical approach, identification of the release tests that can be removed during process validation is a good starting point. The remaining attributes are considered for either reassignment as an in-process control or measurement with an advancing method.
Building on the foundation of a well characterized, well controlled, and robust process, several of the release tests associated with a mAb can be considered for removal at the time of process validation and/or after a significant number of commercial batches post process validation. These include: Residual Host Cell Protein, Residual Host Cell DNA, Residual Protein A, and other process related impurities.
In-Process and Parametric Control
Through the use of rapid in-line and at-line analytics, many of the remaining release parameters can be measured during processing. This strategy requires more than just rapid analytics and benefits from the complementary use of parametric monitoring of online data throughout the remainder of the process. This strategy builds further on the foundation of a holistic process control strategy and the demonstration of process robustness, especially important for a continuous manufacturing process, where steady state control of process conditions is essential.
This scenario includes the use of rapid at-line technologies including real-time quantitative polymerase chain reaction (RT-qPCR) for mycoplasma, rapid LAL for endotoxin and performance monitoring in-situ sensors including cell mass, inline UV, and fingerprinting spectroscopy. Additional technology development is needed in the areas of raw material conformance, viral adventitious agents, etc.
In the FDAs Guidance on Drug Product Sterilization24, Parametric Release is defined as a “release program where demonstrated control of the sterilization process enables a firm to use defined critical process controls, in lieu of the sterility test”. In application, the in-depth knowledge of the process along with a critical process parameter control strategy is used to replace direct attribute testing. This strategy is important in sterility testing because the method of statistical sampling “restricts the ability to capture microorganisms dispersed in a large volume”.
This challenge is extendable to continuous processing. In batch processing, the product quality profile is typically measured on a representative sample from a static process bulk pool. It is possible, however, that in continuous processing, the product is not static and/or pooled; in this instance the application of parametric release seems applicable.
The deployment of PAT in support of parametric release for continuous processing is likely to vary. For example, in a production perfusion bioreactor the PAT may be used for controlling a steady state condition across the duration of a culture. However, in simulated moving bed chromatography, PAT may be used to monitor the reproducibility of a dynamic operation. Demonstration of the control and reproducibility of the process using PAT can provide meaningful data and demonstrate consistency between sample measurements. A case for parametric prediction models can be supported through process monitoring and/or predictive modeling of product quality attributes.
The complexity of the process needs to be considered in development of prediction models for parametric release. Consider a production perfusion bioreactor. As a steady state process, both raw material changes and process excursions need to be considered in a prediction model. In the case of raw materials, full characterization and process development knowledge will be necessary to provide either the control strategy or correlations for determining the impact on the product quality profile. However, depending on the raw material batch sizing, the quantity of raw materials used during process development could be small and only expose the development dataset to a small number of batches. This reality would limit the exposure to variation and associated knowledge of the impact on process parameters and product quality.
These examples underline the importance of process understanding as the foundation for any rapid release strategy in support of continuous manufacturing.
Advancing methods are also an important part of a RPR strategy. There are opportunities to incorporate recent advancements in direct measurement such as microarray-based matrix-assisted laser desorption/ionization mass spectrometry (MALDI-MS)25 and automated sampling handling technologiesThe analytical tools needed to support near real-time measurement for rapid release in continuous bioprocessing can only be achieved by targeting the appropriate technology gaps which includes the commercialization of multi-attribute test methods. These systems offer the promise of rapid, single sample analysis with benefits extending to the analytical footprint and technical transfer. Release parameters measurable with a multi-attribute test method are listed in Table 2.
Continuous bioprocessing will become increasingly attractive as technology evolves and as more companies follow the lead of current innovators. Process Analytical Technology (PAT) will play a key role in helping development scientists and manufacturing teams to understand, control and optimize continuous bioprocesses. This capability will support the strategy for RPR of continuous processing but challenges remain, including the need for additional technology development. These challenges will gradually be addressed as the level of activity in this field increases and through close collaboration between industry, regulators, and technology partners.
1. A. L. Zydney, “Perspectives on integrated continous bioprocessing – opportunities and challenges,” Current Opinion in Chemical Engineering, vol. 10, pp. 8-13, 2015.
2. J. Walther, R. Godawat, C. Hwang, Y. Abe, A. Sinclair and K. Konstantinov, “The business impact of an integrated continuous biomanufacturing platform for recombinant protein production,” Journal of Biotechnology, vol. 213, pp. 3-12, 2015.
3. R. Godawat, K. Konstantinov, M. Rohani and V. Warikoo, “End-to-end integrated fully continuous production of recombinant monoclonal antibodies,” Journal of Biotechnology, vol. 213, pp. 13-19, 2015.
4. K. B. Konstantinov and C. L. Cooney, “White Paper on Continuous Bioprocessing May 20-21, 2014 Continuous Manufacturing Symposium,” Journal of Pharmaceuticals Sciences, vol. 104, no. 3, pp. 813-820, 2015.
5. V. Girard, N.-J. Hilboldb, C. K. Ng, L. Pegon, W. Chahim, F. Rousset and V. Monchois, “Large-scale monoclonal antibody purification by continuous chromatography, from process design to scale-up,” Journal of Biotechnology, no. 213, pp. 65-73, 2015.
6. M. Munk and E. Langer, “What is holding Industry back from Implementing Continuous Processing: Can Asia Adopt More Quickly?,” BioPharma Asia, pp. 22-27, November/December 2015.
7. FDA, “Guidance for Industry: PAT- A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance,” 2004.
8. A. S. Rathore, “QbD/PAT for bioprocessing: moving from theory to implementation,” Current Opionion in Chemical Engineering, vol. 6, pp. 1-8, November 2014.
9. ICH, “Guidance for Industry: Q8(R2) Pharmaceutical Development,” 2009.
10. ICH, “Guidance for Industry: Q9 Quality Risk Management,” 2006.
11. ICH, “Guidance for Industry: Q10 Pharmaceutical Quality System,” 2009.
12. ICH, “Guidance for Industry: Q11 Development and Manufacture of Drug Substances,” 2012.
13. S. M. Mercier, B. Diepenbroek, M. C. Dalm, R. H. Wijffels and M. Streefland, “Multivariate data analysis as a PAT tool for early bioprocess development data,” Journal of Biotechnology, vol. 167, no. 3, pp. 262-270, 2013.
14. S. M. McCarthy, T. E. Wheat, Y. Q. Yu and J. R. Mazzeo, “Two-Dimensional Liquid Chromatography for Quantification and MS Analysis of Monoclonal Antibodies in Complex Samples,” Waters Corporation, Milford, Ma, 2012.
15. D. D. Newbold, D. L. Millard, E. Y. Yu, P. T. Jeffers and J. W. Weber, “US20140087413 A1: Automatic aseptic sampling valve for sampling from enclosed containers,” Bend Research, Inc, Bend, OR, 2014.
16. N. R. Abu-Absi, B. M. Kenty, M. E. Cuellar, M. C. Borys, S. Sakhamuri, D. J. Strachan, M. C. Hausladen and Z. J. Li, “Real time monitoring of multiple parameters in mammalian cell culture bioreactors using an in-line Raman spectroscopy probe,” Biotechnol. Bioeng., vol. 108, no. 5, pp. 1215-1221, 2011.
17. A. S. Rathore, R. M. Kennedy, J. K. O’Donnell, I. Bemberis and O. Kaltenbrunner, “Qualification of a Chromatographic Column Why and How to Do it,” BioPharm International, pp. 30-40, March 2003.
18. J. Dowd, A. Jubb, K. E. Kwok and J. M. Piret, “Optimization and control of perfusion cultures using a viable cell probe and cell specific perfusion rates,” Cytotechnology, vol. 42, no. 1, pp. 35-45, 2003.
19. R. L. Green and C. D. Brown, “Raw-Material Authentication Using a Handheld Raman Spectrometer,” Pharmaceutical Technology, vol. 32, no. 3, 2008.
20. A. Walsh, “Cleaning Validation for the 21st Century: Overview of New ISPE Cleaning Guide,” Pharmaceutical Engineering, vol. 31, no. 6, pp. 1-7, November/December 2011.
21. A. S. Rathore, “Roadmap for implementation of quality by design (QbD) for biotechnology products,” Trends in Biotechnology, vol. 27, no. 9, pp. 546-553, 2009.
22. S. Lee, X. Yang, G. Tian and T. O’Connor, “Regulatory Approaches for Supporting Innovation in Pharmaceutical Manufacturing,” BioPharma Asia, pp. 6-11, January/February 2016.
23. C. Ündey, S. Ertunç, T. Mistretta and B. Looze, “Applied advanced process analytics in biopharmaceutical manufacturing: Challenges and prospects in real-time monitoring and control,” Journal of Process Control, vol. 20, no. 9, pp. 1009-1018, 2010.
24. FDA, “Guidance for Industry: Submission of Documentation in Applications for Parametric Release of Human and Veterinary Drug Products Terminally Sterilized by Moist Heat Processes,” 2010.
25. R. F. Steinhoff, D. J. Karst, F. Steinebach, M. R. Kopp, G. W. Schmidt, A. Stettler, J. Krismer, M. Soos, M. Pabst, A. Hierlemann, M. Morbidelli and R. Zenobi, “Microarray-based MALDI-TOF mass spectrometry enables monitoring of monoclonal antibody production in batch and perfusion cell cultures,” Methods, vol. 105, pp. 33-40, July 2016.
Jeffrey Doyle supports the delivery of innovative PAT at Pfizer’s mAb and Vaccine launch sites through his role as Manager of PAT Projects in the Process Analytical Sciences Group. He is active in industry technology road mapping and previously held positions in Technical Services and Validation. Jeff holds a B.S. in Chemical Engineering from Clarkson University in Potsdam NY, USA.
Paul Jeffers is Manager of PAT Projects in Process Analytical Sciences Group, Global Technology Services, Pfizer Global Supply. He supports the development and implementation of PAT technologies and approaches to bioprocess development and commercial manufacturing. Paul has a B.Sc. in Biotechnology from Dublin City University and a Ph.D. in Chemical Engineering from University College Dublin, Ireland.
Seamus O’Neill currently leads a team responsible for supporting PAT integration across a range of Pfizer manufacturing sites. His team’s project activity spans across API, Bio DS, Aseptic and Solid Oral dose manufacturing. Seamus is an Analytical Chemist and studied at Cork Institute of Technology in Cork Ireland.