Recombinant proteins and monoclonal antibodies (mAbs) have revolutionized treatment regimes in many therapeutic areas. In addition, cell and gene therapies for personalized medicine will further contribute to improve the quality of life for people all around the world. What all these biopharmaceutical treatments have in common is that their production requires the use of living cells (typically mammalian, but also bacterial and yeast). After bioproduction, these biopharmaceutical products must be purified and concentrated before administered to patients. The production phase through cell culture or fermentation is called upstream processing (USP); the purification and concentration phase is called downstream processing (DSP).
The latter phase involves several unit operations, including buffer preparation/dilution, chromatography, ultrafiltration/diafiltration, viral inactivation, and other optional operations, such as centrifugation and protein refolding. In compliance with the U.S. Food and Drug Administration’s (FDA) Process Analytical Technologies (PAT) initiative,1 each unit operation must be developed so that critical process parameters (CPPs) impacting critical quality attributes (CQAs) are appropriately controlled over a broad process design space to ensure process robustness, reproducibility, and scalability, as well as product quality, safety, and efficacy.2 In other words, the CPPs are “switches” that can be controlled in real time to make sure that the CQAs stay within the predefined values, ensuring the product is safe and provides its intended therapeutic effect.
The most common CQAs for DSP relate to purity (protein purity and host-cell protein (HCP), DNA, and RNA contaminant levels), protein structure and stability (primary, secondary, and tertiary structures; posttranslational modifications (PTMs); and aggregate and product-related variant levels), sterility (microbial and endotoxin contaminant levels), and viral clearance.
The CPPs relating to these various CQAs include pH, conductivity (ionic strength), impurity analysis, temperature, flow rate/pressure, redox environment, dissolved oxygen (DO, for oxygen-sensitive proteins), and turbidity. Each of these CPPs correlates to two or more CQAs. The specific CPPs and their impacts on different CQAs will depend on the biologic drug substance, the production culture type, and the downstream process steps employed.
The more critical parameters that are controlled in real time, the higher the chance that the desired CQAs will be obtained. Real-time monitoring requires in-line and/or on-line monitoring technologies. Several such technologies are already available, such as the aforementioned pH, conductivity, protein fraction/concentration, and flow rate / pressure. For some CPPs, however, off-line analysis is still required, owing to the lack of suitable in-line solutions. Although both in-line/on-line and off-line technologies can be used for PAT, the more desirable “PAT tools” are the former, as they provide real-time measurements.
In-line CPP monitoring with PAT tools supports quality-by-design (QbD) approaches to process development — which aim to prevent system downtime and/or batch failures.
PATs are more commonly studied in upstream processing than for real-time monitoring of downstream purification operations due to the complexity and variability of the latter processes. At the same time, especially recently, progress has been made in overcoming some of the challenges regarding the use of PAT tools for real-time monitoring of downstream operations.
The recent COVID-19 pandemic offers an instructive example. The bottleneck for the production of the COVID-19 vaccines involved DSP rather than USP operations. The available DSP capacity could not keep pace with the demand to purify and concentrate huge amounts of antigens. This triggered greater attention on the part of all DSP technology groups of the need to enhance the productivity of their processes with in-line monitoring and control. It may be worthwhile to look in depth into each DSP operation unit to understand which CPPs best suits the tight control of CQAs. Below are some examples of approaches to improve DSP productivity.
Increasing the sustainability, efficiency, and cost-effectiveness of biopharmaceutical manufacturing remain primary goals for biologic drug manufacturers. In addition to improving process performance through in-line monitoring, identifying opportunities for process intensification is an active area for many companies and a focus for industry trade groups, such as the BioPhorum.
Each DSP operation unit requires different buffers to support its function: the biopharmaceuticals are always processed in liquid suspension state. As a result, huge volumes of buffers have to be available at the point of use. Buffer preparation thus dictates plant capacity and production schedules. Buffer preparation is traditionally labor intensive and involves the use of large quantities of reagents and large mixer tanks, which occupy a significant portion of the manufacturing footprint.
A recent approach leverages in-line dilution of buffer concentrates along with single-use fluid handling and mixing solutions. Concentrated mixtures of various buffer solutions are prepared in advance in fixed vessels and then diluted as needed for each process. With this approach, it is possible to reduce the floor space required for buffer storage by 75% and the cost per liter for buffer preparation by up to 12% and to make buffer management more time-efficient.
One of the keys to successful in-line buffer preparation is control of pH and conductivity. Reliable and precise real-time process monitoring is essential to maintain optimal conditions and ensure efficient, high-quality biopharmaceutical production, especially when buffers are prepared “just-in-time.”
The Cytiva Allegro™ Connect Buffer Management System is an example of an effective in-line buffer preparation skid.3 In this system, pH and conductivity are monitored with Hamilton SU OneFerm Arc 120 and Conducell 4USF Arc 120 in-line sensors, respectively.
Chromatography is the purification step where, for example, mAbs are captured and separated from other components using protein affinity or charge differences. For proteins, low-pressure liquid chromatography (LPLC) systems are typically used, operating at pressures less than 6 bar. It is used to desorb captured product from chromatography columns once impurities have been allowed to flow through the column. It is typically used for fine separation of high-value molecules (>$2/g), such as mAbs (as previously mentioned), separation of molecules of similar size, purification of fragile molecules and cells, and a range of other applications. Notably, it is an adaptable process that can be used for many types of molecules, including proteins, oligonucleotides, sugars, lipids, cells, viral vectors, enzymes, and nucleic acids.
To achieve the highest product recovery and purity, all CPPs, such as pH, conductivity, flow rate, protein titer, and temperature, must be carefully controlled. For processes to benefit from real-time monitoring, the sensors for monitoring these critical parameters must be very precise and reliable.
An example of a skid for LPLC, the Verdot FlexiProTM benchtop system is adaptable to a wide range of chromatography processes.4 The system also offers the capability to leverage in-line dilution or gradient preparation of aqueous buffers. Hamilton OneFerm and Conducell 4USF sensors are used to monitor buffer pH and conductivity and to control the flow rate and dilution conditions. Protein titer, flow rate, and temperature are also monitored.
Tangential-flow filtration (TFF) is used for ultrafiltration/diafiltration steps to remove impurities, exchange one buffer solution for another, and/or increase the concentration of the final product. In addition, by choosing the appropriate filter type, TFF can be used for particulate filtration, microfiltration, nanofiltration, and many other applications. Furthermore, because TFF enables continuous and gradual concentration with the product stream moving parallel rather than perpendicular to the filter, as is the case with traditional dead-end impact filters, the desired concentration can be achieved approximately 2.5 times faster.
The keys to implementing successful TFF processes that purify and concentrate the biopharmaceutical product within the expected timeframe while preserving product stability are realizing efficient functioning of the membranes and maintaining proper buffer conditions. Flow rates in production-scale TFF skids for monoclonal antibody purification, for example, can reach to 600 L/hour.
Monitoring of the CPPs needed to achieve these goals must be accurate and have a quick response time. These parameters include pH to maintain the desired buffer conditions, conductivity for monitoring the efficiency of buffer exchange and salt removal, transmembrane pressure to prevent membrane fouling and damage, and feed and permeate flow rates to ensure proper fluid dynamics.
Protein concentration and temperature are additional CPPs that can be valuable to monitor in real time. Protein concentration impacts product recovery, concentration, and quality. In-line UV sensors or spectrophotometers installed directly in the TFF skid’s permeate or retentate lines provide protein concentration values, as UV absorbance is proportional to protein concentration. It is important to note, however, that calibration curves established using a series of protein standards with known concentrations and compliant with each specific application must be used to accurately determine the protein concentration.
As is the case with temperature measurement during chromatography, in-line monitoring can provide information on whether the TFF process is proceeding within the optimal temperature range, and it is particularly important for temperature-sensitive molecules. However, it is not typically used for control purposes, because controlling the process temperature often involves approaches that do not rely on in-line sensor readings.
Examples of effective TFF skids incorporating PAT tools include the Donaldson (SolarisTM) Kronos and Tytan systems. Hamilton intelligent pH and conductivity arc sensors have been digitally integrated into the Solaris Process Control System to provide benefits over analog sensor outputs.5 For instance, the communication protocol is more robust and is not sensitive to electrical noise; the software directly tracks sensor health and diagnostic information in real time, helping to avoid lost batches due to probe fouling; and calibration is simplified and accelerated.
For downstream processing, each unit operation has its own set of CPPS and CQAs. In any scenario, the correct monitoring and control of CPPs via suitable process sensors is crucial for achieving the desired product quality and ensuring the success of biopharmaceutical manufacturing. Ultimately, for each biomolecule and bioprocess, a unique set of CPPs must be monitored with a unique combination of in-line — and currently off-line — analytical techniques/technologies.
Studies by the BioPhorum industry trade group confirm the real benefits to be gained from the use of in-line monitoring with respect to yield, quality, time savings, and reduced product/batch failures.6
At Hamilton, we pioneer sensor technology to enable biopharma users to solve their development and production challenges and improve their efficiency. Our goal is to provide accurate measurements that enable real-time control of the relevant CPPs. In our view, this must be done seamlessly, meaning that the process sensors should be easily implemented (e.g., digital integration) and not time-consuming to maintain (e.g., quick and easy to calibrate). When this is achieved, the advantages of real-time monitoring can be applied from R&D scale-up to pilot and production scale.
In the short term, it is likely that some improved in-line measurement solutions will be introduced for some process parameters, while others will remain challenging and continue to require off-line analysis (Table 1). Ideally, in the longer term, new technologies will be identified to make in-line monitoring of even these challenging parameters possible. Hamilton intends to remain at the forefront of those discoveries.
As an innovator company, Hamilton values partnerships with other innovators in the pharmaceutical industry — equipment vendors and drug developers and manufacturers. This includes every company with ideas about how to innovate and improve the efficiency of bioprocessing, both upstream and downstream. Our goal is to be the leading one-stop-shop provider of all relevant CPP monitoring solutions for both upstream and downstream processes.
References
.svg)
