Simon Hawdon - Principal Scientist, CPI Biologics, Darlington UK
Sean Ruane - Senior Scientist – Process Automation, CPI Biologics, Darlington UK
Daniel Myatt - Senior Scientist – Analytical Characterisation, CPI Biologics, Darlington UK
The global biopharmaceutical market is estimated to grow to $3.9B by 2024 - a CAGR of 8.59%,7 with the majority of the growth in monoclonal antibodies. This growth will be through larger numbers and combinations of smaller quantities of biologics, which will put pressure on manufacturing supply chains and increase the demand for flexible manufacturing technology.
Efficient and agile production of high-quality biopharmaceuticals is therefore of the highest priority to the biopharmaceutical manufacturing industry. The use of intensified or continuous manufacturing systems for biotherapeutics has been shown to address these challenges.5,11,12 The advantage of integrating multiple unit operations into a continuous process was successfully demonstrated by a previous Innovate UK consortium project and has also been described in scientific literature.5 Here we describe an ongoing project at CPI which builds on the proof-of-concept system developed in previous projects and adds the next layer of configurability, automated control, process intelligence and next generation PAT (Process Analytical Technology) to allow:
- Critical control automation and predictive performance to show real-time continuity of product quality attributes and to predict consumable lifetime and changeover to avoid quality deviations (predictive maintenance). Next generation Advanced Process Control and Machine Learning techniques will be developed to create a vendor and scale-independent intelligent control solution.
- Optimization of unit operation interfaces by incorporation of 3D-printed microfluidics to minimize pooling volumes and therefore residence time distribution. These interfaces are a critical point for hygienic process sampling for at-line quality analytics.
- Synchronizing the process data generated by the system into models assessing manufacturing cost, investment required, environmental impact, production capacity and production facility requirements to enable production optimization.
- Use process models to predict how the integrated production unit could form a distributed manufacturing network for low dose biotherapeutics (e.g. pandemic vaccines).
Healthcare Need and Impact
The importance of manufacturing agility in production and supply of therapeutic biologics has been highlighted during the COVID-19 outbreak. However, efficient response to unpredictable demand and differing clinical project priorities in the pharmaceutical sector is a well-known manufacturing requirement.1 The goal during development of a biologic process or product is to demonstrate a well-understood, consistent, productive process before later-stage clinical trial processes begin. Integrated manufacturing, with improved process control and real-time data analytics, generates more robust, stable processes in a shorter time. This is referred to as speed-to-clinic, removing process development from the critical path to clinical results. Decreasing process development time has been shown to save project time, cost and drug development costs of $10sM.2 Many potential drugs remain undeveloped because of the large investment needed and the scale of risk involved. The cost of manufacturing clinical materials can range from 20-50% of the spend on clinical trials.3 Integrated manufacturing has the potential to decrease not only outsourced costs (such as for contract development and manufacturing organizations (CDMOs)), but also the capital investment required, when manufacturing is performed in-house.4 Upon commercialization of a therapeutic, the intensified manufacturing process reduces capital and operating costs for supply of a drug product portfolio (modelled by Sanofi;5) the critical advantages of integrated manufacturing seen by the pharmaceutical industry for commercial production are improved control and security of supply - reduced batch failures, real-time release, on-time delivery and improved facility utilization with lower capital outlay and lower operating costs with increased sustainability.6 Throughout the development and commercialization of drug products, the use of integrated manufacturing is seen as a considerable benefit in enabling lower manufacturing cost and adaptable supply chains. Integrated manufacturing can potentially provide simpler process implementations with lower investment risk. Thus, enabling greater freedom in drug development and the clinical experience around the trials process.
Biopharma 4.0
The Biopharma industry is currently undergoing radical change. This change has been brought about by new technologies such as advances in process analytical technologies, artificial intelligence, genetic modification, and robotics. This strategic shift has been referred to as a new industrial revolution.10 In response to this the Biophorum Operations Group (BPOG) commissioned a Technology Roadmap which looks at the future of the industry over the next ten years. The principles and recommendations it suggested comprise what has been termed Biopharma 4.0.
The roadmap highlights the challenges to the industry posed by market growth, the range of new product classes, and cost pressures. It proposes addressing these challenges through meeting the goals of increasing quality through process robustness, reducing waste due to sub-quality production, decreasing cost of manufacture and capital expenditure, increasing flexibility by improving the response to variations in demand and product changeover time, and improving facility build time.
It identifies the key enabling technologies to address these challenges as process intensification, continuous processing, automated facilities, in-line monitoring, real-time release, and knowledge management. Implementation of continuous processing significantly reduces the size and cost of a facility required to produce a given amount of product. This translates into reduced fixed costs and therefore reduced cost of goods sold.
Two of the key aspirations of the roadmap are the development of portable systems, and small-scale systems for personalized medicines. Another key recommendation from the roadmap is that the industry evolve collaborations through joint projects.
Integrating Continuous Technologies For Rapid Delivery of Cost-Effective Biotherapeutics to Patients
Our current project brings together and further develops the work performed in two previous projects. The UK Integrated Continuous Bioprocessing project (which was completed 2020) built an integrated continuous downstream manufacturing platform for monoclonal antibodies. During the execution of this project it became apparent that a more practical approach was a semi-continuous, fully automated system rather than truly continuous. This has become more widely known as a hybrid approach.
The Digital Integrated and Intelligent Continuous Biomanufacturing project (completed 2021) developed the system further by integrating predictive models and maintenance on selected unit operations, demonstrating the impact of an intelligent digital operation able to predict and mitigate failure during process.
The current project enhances the existing digital control wrapper on the semi-continuous system to enable flexible PAT integration with adaptive Advanced Process Control (APC) with predictive/prescriptive maintenance to offer “plug and play” process development and integrated manufacturing. The system will incorporate 3D printed microfluidic devices to control and balance flow, reducing the need for traditional holding tanks. The system also represents an end-to-end process by including a continuous upstream module (comprising a single use bioreactor linked to hollow fiber perfusion equipment).
Using novel, spectroscopic, in-line PAT allows the collection of process data in real time. This in turn can be used to inform predictive models which will allow advanced, intelligent control over Critical Quality Attributes (CQAs) by altering Critical Process Parameters (CPPs). This brings the possibility of real-time control over critical quality attributes which in turn is a step along the road to real-time drug substance release.
To ensure the safety and efficacy of the final product the current product release process includes extensive end product quality testing which can take up to six weeks. This presents a roadblock in the production process. However, if the process can be shown to have been controlled such that all CQAs have been maintained within approved specifications for the entire process time, this would reduce the need for final product testing and allow more rapid release of the product to the patient, while reducing the risk and value held within drug product inventory.
This strategy can lead to improvements in:
Product quality: The product is produced consistently throughout the process. The quality is designed in rather than tested in.
Process Economics: Reduced waste. Product is available for release more rapidly. More efficient use of materials, reduced human resource requirements, and lower physical footprint.
Process Knowledge: The implementation of in-line PAT gives the ability to better understand the relationship between CQAs and CPPs. Through this better understanding of the operating and design space control models can be constructed to enable process failures to be predicted, identified and solved.
Because of their complex nature, the development and operation of continuous manufacturing systems will require cross-functional teams such as systems biologists, data scientists and automation engineers. Improved, intelligent system automation will, however, simplify the operation of systems. In order to gain the deeper understanding of the relationship between critical process parameters and critical quality attributes further innovation in the fi elds of sensor technology and PAT will be necessary. The effect of this is that substantial initial investment may be required in the development phase. However, this investment will yield longer term benefits in efficiency and productivity.
Stepwise Approach to Real-Time Release
Our vision is of a continuous production platform which supports real-time release of bio-therapeutics. It is our mission to develop a system which enables progress towards this. We have therefore taken a stepwise strategy, as shown in Figure 1, which correspond to systems with increasing levels of automation and complexity.
The first step is to bring together the modular unit operations, such as chromatographic separations or filtration stages which comprise the process.
Onto this basis we add a basic level of process control. This traditionally relies on control around established set points such as balance weight or pH.
Next there is the need to monitor the performance of the system in operation. This is done by linking PAT analytics to traditional process control software and systems, and pulling together process and PAT measurements into a single, easily accessible stream.
Advanced Process Control allows the adjustment of the process in response to observed trends. The most commonly used form of APC is Model Predictive Control, in which statistical or mechanistic models are used to make real-time predictions of CQA (Critical Quality Attributes) based on live process information. The system then uses these models to determine which CPP (Critical Process Parameters) changes are required to maintain process specification, allowing greater process stability and consistent product quality.
The final goal of this stepwise approach is a system capable of real-time release of product. When the complete process can be demonstrated to be in control, and CQAs are monitored by PATs or accurately predicted by robust models, the need for end-product testing is reduced, allowing rapid deployment of product to patients.
Some of the milestones described above have been achieved by our earlier collaborative projects.
The system developed during the UK Integrated Continuous Bioprocessing project brought together unit operations, built an automation system and developed a method for monitoring the performance of the system. The project’s aims were to develop a flexible, integrated, automated downstream biologics manufacturing system with multiple unit operations and integrated analytics.
Three industrially relevant processes were converted from batch to continuous operation, and a proof-of-concept study was performed at clinical scale (48-hour continuous operation). For example, one of the processes comprised eleven unit operations, as shown in Figure 2.
The system we developed required the flexibility to perform different processes. Automation was developed to simplify this reconfiguration. The control strategy was required to integrate with existing control software (contained within the unit operation hardware) and the overarching control software.
To meet these requirements we defined logical envelopes around the unit operations, allowing each unit operation to be treated in the same way regardless of its type; simplifying the operation of the system and allowing us to hide the inner workings to simplify operator interactions.
Each unit operation was connected to the next by a surge tank. The control system monitored the status of each unit operation based on the mass in the surge tanks immediately before and after. The system fl ow was controlled by pausing and restarting the unit operation based on the status of the surge tanks. In this way each unit operation communicated directly with the control system but was independent of the other unit operations.
Analytical data was collected during the process operation from in-line (e.g. pH, conductivity, turbidity), at-line (titer and quality from SEC-HPLC and autosampler), and off -line (e.g. glycosylation from Mass Spectrometry).
In our second project - Digital Integrated and Intelligent Continuous Biomanufacturing - our aim was to build on the digital control from the first project by developing predictive and prescriptive models to identify failures in advance.
In the context of our stepwise approach, it advances unit operation monitoring towards advanced process control.
We identified three levels of control:
Predictive, which uses a model to predict an outcome based on process monitoring data;
Prescriptive, which predicts an outcome and suggests a corrective measure to the operator;
Advanced, which automatically instigates a process change based on a predicted outcome, and maintains optimised and stable performance.
The aim was to show predictive control over two individual unit operations (we selected Filtration and Protein A capture chromatography), reducing system scope to allow robust data collection. For each unit operation we used a design of experiments approach to investigate the experimental space and build large datasets which were then used by Applied Materials to create statistical models using their PharmaMV control software.
For filtration, a prescriptive maintenance model was developed that could predict remaining filter life to 92% accuracy and alert an operator when critical points were reached. In the case of Protein A chromatography a model was developed that could predicted the yield of the process step with a 95% accuracy.
Our current project – Integrating Continuous Technologies for Rapid Delivery of Cost-Effective Biotherapeutics to Patients – aims to develop a system capable of Advanced Process Control, including the use of novel PAT to make process changes based on live CQA measurements. This system will use 3D printed microfluidics to control and balance process flow, reducing control system complexity and greatly improving residence distribution times. Economic models will be created to calculate the process costs, including environmental and investment costs, which will enable future economically optimized process control. The scope also includes an integrated continuous upstream perfusion process. Our partners in this project are Pall, BiologIC, Biopharm Services and SCIEX.
This system will add an overarching advanced control layer, allowing the process to adapt to variations in the inputs and thereby tune itself and give robustness. To enable this advanced control we will be assessing in-line spectroscopic technology to provide real-time measurements of CQAs. Measurement data will be input into models which will provide control over critical process parameters.
There are a range of spectroscopic technologies of interest to us. Index of refraction (IoR) has the flexibility of UV/Vis for determination of protein concentration but over a wider order linear range (up to 200g/L, and is therefore more process relevant. This wider range increases its in-line flexibility by reducing the need for sample dilution.
We will be investigating the use of mid-Infra-red (MIR) sensors. MIR has been shown to be useful in monitoring of bioreactor analytes (specifically glucose and lactate), but has also been used in the analysis of protein structure, glycosylation and aggregation. MIR technology mounted in-line brings the possibility of monitoring specific CQAs in real-time, a significant step in the direction of real-time release.
SCIEX, as a partner in the consortium, is developing rapid methods for mass spectrometry to analyse relevant CQAs. The aims of this work is to bring the analysis time down to around one hour, allowing near real-time measurements.
Data from in-line and at-line PATs will be aggregated into a single source. Advanced control software will make predictions based on this real-time data and make changes to set points in the distributed control system. This allows advanced predictive control to be implemented, while basic control is maintained by the robust distributed control layer.
An integrated, continuous manufacturing system for biotherapeutics brings with it the economic advantages of continuous manufacturing e.g. smaller footprint, more efficient use of resources. The development of real-time release will bring added advantages of improved quality; further economic advantages in reducing the time to batch release. It will also allow improved availability to the patient since the time to batch release will be minimized. Safety will be improved with increased process knowledge. Our project hopes to address some of these issues and bring together technology to take the industry a step closer to real-time release of biotherapeutics.
References
- Merkuryeva, G., Valberga, A. & Smirnov, A. Demand forecasting in pharmaceutical supply chains: A case study. Procedia Computer Science 149, 3--10 (2019)
- DiMasi, J. A., Smith, Z. & Getz, K. A. Assessing the Financial Benefits of Faster Development Times: The Case of Single-source Versus Multi-vendor Outsourced Biopharmaceutical Manufacturing. Clinical Therapeutics 40, 963--972 (2018)
- Applied Clinical Trials Online - Contract Development and Manufacturing Costs During Clinical Development of a New Drug (Aug 5, 2013)
- Pollard, D., Brower, M., Abe, Y., Lopes, A. G. & Sinclair, A. Standardized Economic Cost Modelling for Next-Generation MAb Production. Bioprocess International Sept, 2016
- Walther, J. et al. The business impact of an integrated continuous biomanufacturing platform for recombinant protein production. Journal of Biotechnology 213, 3--12 (2015)
- Konstantinov, K. B. & Cooney, C. L. White paper on continuous bioprocessing May 20-21, 2014 continuous manufacturing symposium. Journal of Pharmaceutical Sciences 104, 813--820 (2015)
- Mordor Intelligence LLP, Biopharmaceuticals Market - Growth, Trends, and Forecast (2020 - 2025)
- Dream, R; The Factory of the Future. The American Pharmaceutical Review Mar 24, 2022
- Biophorum Operations Group, Biomanufacturing Technology Roadmap: Executive Summary May 2017
- Joshua Hays, and Cornelia Thieme Biopharma 4.0: The Convergence Of Informational, Operational, And Processing Technologies. Bioprocess Online Aug 7, 2019
- Dirk Müller, Lukas Klein, Johannes Lemke, Markus Schulze, Thomas Kruse, Martin Saballus, Jens Matuszczyk, Markus Kampmann, Gerben Zijlstra. Process intensification in the biopharma industry: Improving efficiency of protein manufacturing processes from development to production scale using synergistic approaches. Chemical Engineering and Processing - Process Intensification Volume 171, January 2022, 108727
- Ou Yang, Maen Qadan, and Marianthi Ierapetritoua. Economic Analysis of Batch and Continuous Biopharmaceutical Antibody Production: A Review J Pharm Innov. 2019; 14: 1–19
- Ohnmar Khanal & Abraham M. Lenhoff. Developments and opportunities in continuous biopharmaceutical manufacturing. mAbs, Volume 13, 2021 - Issue 1
Subscribe to our e-Newsletters
Stay up to date with the latest news, articles, and events. Plus, get special
offers from American Pharmaceutical Review delivered to your inbox!
Sign up now!