How To Optimize Upstream Bioprocessing

Biopharma PAT Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 2 Content Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 1 3 2 4 5 6 7 8 PAT Building Blocks 5 PAT for Biopharma 8 Culture & Fermentation Process Types 11 Monitoring Methods 14 Critical Process Parameters 17 Critical Quality Attributes & Key Performance Indicators 23 Conclusions 31 References 33 3 4 5 6 1 2 3 Intelligent Arc Sensors for pH & DO In-Situ Measurement 18 Dissolved Oxygen User’s Experiences 19 Real-time Monitoring of DCO2 During Cell Culture for mAb Production 20 In-Situ Cell Density for Batch & Perfusion mAb Production 26 Validation of Density Measurement for Different Cell Types 27 The Benefits of Monitoring Cell Density During Various Applications 29 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Focus Spots 4 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Abstract This white paper introduces the concepts of Process Analytical Technology (PAT) in the context of Upstream Biopharma processes, focusing on the significance of Quality Attributes, Critical Process Parameters, and Key Performance Indicators at the Bioreactor. The PAT initiative is recommended by the US FDA as a strategy to minimize the risks associated with pharmaceutical product manufacturing. At Hamilton Process Analytics, we believe that monitoring and controlling several Critical Process Parameters and Key Performance Indicators greatly improves the reliability, efficiency, and productivity of bioproduction processes. In this white paper, we present real-world applications and case studies, showcasing the practical implications of implementing PAT monitoring the ubiquitous dissolved oxygen and pH, in addition to the often overlooked yet nonetheless essential critical process parameter dissolved carbon dioxide. We further demonstrate the advantages of monitoring key performance indicators for cell density, including both total cell density and the insightful viable cell density, providing examples for different cell types and fermentation process. Overall, this white paper serves as a comprehensive guide to understanding the critical elements of Biopharma PAT and its role in enhancing the efficiency and quality of bioreactor manufacturing processes. Keywords Biopharma Process Analytical Technology (PAT) Quality Attributes Critical Process Parameters Key Performance Indicators Bioreactor Pharmaceutical cGMP Automated Control Monitoring Methods Intelligent Arc Sensors Risk Minimization Biopharmaceutical Product Manufacturing Process Efficiency Process Optimization Upstream Processes In-situ measurement Real-time monitoring PAT initiative Bioprocesses Control systems Sensor technology Cell growth Mammalian Cell Cultures CHO cells In-line Sensors Dissolved Oxygen (DO) Dissolved CO2 (DCO2) Fed-batch Processes Perfusion Processes pH Total Cell Density Viable Cell Density Scale-up and Scale-down Quality-by-Design (QbD) Regulatory Framework Quality Assurance 5 PAT Building Blocks 1 6 PAT Building Blocks Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 The Process Analytical Technology (PAT) Initiative originates from the 2004 guidance published by the U.S. Food & Drugs Administration (FDA)[1]. It is a part of the broader initiative “Pharmaceutical cGMP for the 21st century – A risk based approach”[2]. The focus of this initiative is to minimize risks to public health associated with pharmaceutical product manufacturing. PAT established a regulatory framework intended to facilitate the voluntary development and implementation of innovation in pharmaceutical development, manufacturing, and quality assurance. It focuses on enhancing the understanding and control of the manufacturing process to achieve Quality-by-Design (QbD): quality should be built into a product with an understanding of the product itself and the process by which it is developed and manufactured along with a knowledge of the risks involved in the manufacturing process and how best to mitigate those risks. PAT promotes a process which starts with the identification of each product’s specific Critical Quality Attributes (CQAs), then proceeds with monitoring as often as possible the related Critical Process Parameters (CPPs) and of the Key Performance Indicators (KPIs), in order to automatically control them within pre-defined limits. The relationship between CQAs, CPP and KPI are described in their definitions from PAT literature: [3], [4] • Critical Quality Attribute (CQA): a physical, chemical, biological, or microbiological property or characteristic that should be within an appropriate limit, range, or distribution to ensure the desired product quality. • Critical Process Parameter (CPP): a process parameter whose variability has an impact on a critical quality attribute and, therefore, should be monitored or controlled to ensure the process obtains the desired quality. • Key Performance Indicator (KPI): a metric for the status of each production step. KPIs are related to CQAs and therefore influenced, as well, by the CPPs. As the CPPs remain within the pre-defined limits, the KPIs should indicate that each production step proceeds accordingly resulting, in the end, in a product having its CQAs within the appropriate limits, too. CQAs are still difficult to measure directly in production. Along the upstream and downstream portions of the manufacturing process it is most common to monitor the CPPs and KPIs related to the quality attributes. PAT recommends various tools for this purpose: • Multivariate tools for process design, data acquisition and analysis • Process analyzers (e.g. in-line sensors or automated at-line devices) • Process control tools (e.g. statistical process control softwares) • Continuous improvement and knowledge management tools 7 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 An appropriate combination of these tools may be applicable to a single-unit operation like a bioreactor, or to an entire manufacturing process portion like upstream or downstream. Process analyzers are a typical example of tools to measure process data. Their output is used for different scopes like univariate mechanistic modeling, process characterization or multivariate analysis such as the “golden-batch” prediction. This white paper focuses on process analyzers for the bioreactor. An overview of the most important performance indicators and process parameters will be provided, together with examples of the proper in-situ sensors and equipment used to monitor them. Figure 1 highlights an example of in-situ process sensors. Figure 1 Example of a bioreactor with in-situ process sensors for a microbial fermentation. The chart represents the real-time monitoring of CPPs such as the Dissolved Oxygen as % saturation (signal in black). 21.04.10 08:00:11 0 00:30 0 01:00 0 01:30 0 02:00 0 02:30 0 03:00 0 03:30 0 04:00 0 04:30 0 05:00 Induction 50% sat 0 0.0 0.00 0.00 Stirrer U/min pO %2 Luft l/min Abgas C O %2 1000 100.0 10.00 10.00 8 PAT for Biopharma 2 9 PAT for Biopharma Manufacturing of biopharmaceutical products is a complex process[5]. The complexity is due to the heterogeneity typical for the bioprocesses. In the upstream process the heterogeneity arises from the cells who are living subjects. According to minimal variations of the environment, they can produce higher or lower yield of product with a quality within the pre-defined limits. Or, in other words, with a molecular conformation bioactive enough to deliver the expected healing effect on patients. The upstream heterogeneity transfers to the subsequent product purification steps in downstream processes, too. Figure 2 Biopharmaceutical manufacturing process. This map shows the commonly monitored real-time Critical Control Parameters pH, ORP, DO, Conductivity and the Key Performance Indicator Cell Density. Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Buffer Preparation Centrifuge Protein A Chromatography IEX Chromatography TFF Filtration (Or Diafiltration) TFF Filtration (Or Diafiltration) Liquid Waste Air Filter Bioreactor Harvest Tank Liquid Waste Media Prep Tank Air Water Nutrients CIP Cleaning Compressor Fill & Finish Virus Inactivation 10 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Existing scientific literature[6] already describes how the application of PAT to these complex processes could enable significant improvement in upstream through the use of performance indicators (e.g. viable cell density). In downstream the application of PAT results in higher quality and purity of the final product (e.g. protein, vaccines, etc.). It is universally acknowledged that to properly apply PAT, it is essential to move from the manual sampling and laboratory measurement procedures to automated control[7]. As even minimal variations of process parameters have a major influence on the final product, controlling them in real-time minimizes the risk of lower yield and purity. Real-time monitoring is possible due to sensors which can withstand the Cleaning-In-Place (CIP) and Sterilization-In-Place (SIP) procedures required to minimize the risk of contamination. This is already common for the fundamental CPPs: e.g. pH, dissolved oxygen (DO) and conductivity (Figure 2). Further advances in CIP and SIP compliant sensors have been developed to directly measure KPIs such as Total Cell Density (TCD) and Viable Cell Density (VCD). These will be described in the following paragraphs. 11 Culture & Fermentation Process Types 3 12 Culture & Fermentation Process Types Bioreactors are vessels used for cultivating mammalian cells (such as CHO), microbial cells (such as E. Coli) and yeast or small plant cells (such as moss). These cells work like small factories to produce the desired compounds. Culture – Mammalian MAbs and therapeutic proteins are produced mainly by means of mammalian cells, especially if the needed therapeutic agent must be compliant with human biology. The main cell lines are: CHO, BHK, and NSO-GS. Such cells typically exhibit a growth rate of doubling their numbers every 24 hours. This is relatively slow, therefore monitoring & control strategies should benefit from longer working times. Nonetheless, mammalian cells have a less robust outer membrane compared to microbial cells, thus they are more fragile against changing process conditions: they have to be constantly controlled. Huge cost comes from batch loss due to equipment not working properly to maintain the desired process conditions[9]. Equipment failures – particularly at the commercial scale – are extremely costly, resulting in lost batches, a repeat of the bioprocess studies to satisfy regulators, and other such problems[10]. Due to these reasons, the PAT push for real-time monitoring and automated control means significant improvements for bioprocesses using mammalian cells. Culture – Microbial Several recombinant proteins and vaccines are produced with microbes such as bacteria[11] and yeasts[12], applications of these cells are widespread due to their robustness and ease of cultivation. Compared to mammalian cells, microbials typically have shorter fermentation times as well as higher chemical and physical protein stability. Bacterial cultures can double within 20 to 30 minutes, which is why CPPs such as pH, DO, cell density and feed rates need to be measured as frequently as possible. Once again real-time control becomes necessary to achieve true QbD. Fermentation Process Type – Batch Batch fermentation processes are often considered the first processes adopted by the biopharmaceutical industry. Microorganisms are added to culture media in the bioreactor, which has been pre-filled with nutrients like glucose, glutamine, other amino acids and minerals. The media remains the same during the entire process and is not supplemented, refilled or exchanged at any time. Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Through a fermentation process they transform nutrients like glucose and amino acids into high-value biopharmaceutical products as vaccines, monoclonal antibodies (mAbs), and other therapeutic proteins. Such variability in the processes makes the application of PAT at the bioreactor unique according to the culture used and the fermentation process type required[8]. 13 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 After an initial lag phase, the number of microorganisms increase sharply during the growth phase. Then, after a stationary phase of suspended population, the culture population drops off in a death phase. The cause for the population drop-off can be tied to the depletion of nutrient media and the accumulation of toxic substances. The PAT strategies are limited just to the Critical Process Parameters which can be modified in real-time and therefore can be controlled: e.g. pH, DO and temperature. Fermentation Process Type – Fed-Batch Fed-batch has been the dominant bioprocessing method for decades[8]. The fed-batch process differs from the traditional batch process by adding nutrients in stages to maximize cell growth. The bioreactor is filled with a base amount of media to support initial cell growth. Feed media is added when needed to replace nutrients depleted by the increasing cell population. The cells and their product(s) remain in the bioreactor until the end of the run. With this setup it is possible to automatically regulate the addition of feed media according to nutrient levels or viable cell density. Fermentation Process Type – Perfusion The term “continuous bioprocessing” generally refers to perfusion technologies. The bioreactor runs at a fixed volume and fixed cell concentration for 30–90 days or longer depending on cell line. During this time the feed media is constantly refreshed and the secondary toxic metabolites eliminated while cells are simultaneously harvested for further processing. Perfusion technology is one of the newest methods for cell culture processes. Despite the benefits of perfusions, regulatory issues are still a hurdle for its implementation: problems with the “batch” definition in a continuous process make release procedures more complex. Therefore, even more than for the other processes‘ types, perfusion is highly dependent on QbD and PAT in order to work properly and be accepted by regulatory authorities. 14 Monitoring Methods 4 15 Monitoring Methods According to the PAT guidance[1] and the scientific literature[8], the monitoring of the critical process parameters and quality attributes can be performed following different methods, like those represented in Figure 3. Although the in-line (or in-situ) and on-line are the methods of choice for real-time monitoring, at-line and off-line are still options for those CPPs, KPIs and even CQAs which cannot be accurately measured in the bioreactor. The following paragraphs detail such differences. Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 On-line In-line / In-situ At-line Off-line Off-line The sample is taken out of the bioreactor in sterile conditions and analyzed in the lab after physical pretreatments (e.g. filtration and dilution). The preparation and handling require clear Standard Operating Procedures (SOPs) as well as skilled personnel. If problems occur during these stages, the accuracy of the results will decrease. Together with the complexity involved in manual handling, the major disadvantage of off-line measurement is the time delay, which results in lower measurement frequency. Due to these issues off-line measurements should not be considered true PAT unless there are no other measurement possibilities (e.g. HPLC for product titer or mass spectroscopy for product quality). In these examples, automated controls are not a possibility. Off-line laboratory measurements are commonly used to monitor and validate the accuracy of the in-line/on-line process analyzers. However, factors such as temperature changes and de-gasing can negatively influence the accuracy of these reference measurements. Figure 3 Different methods of process monitoring according to the PAT guidance (2004) & later scientific literature 16 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 At-line In at-line measurement, the sample is removed and analyzed in close proximity to the production process, either manually or by using automated sampling devices. Similar to off-line measurement, sterile conditions must be maintained for accurate results. At-line measurement is most common for parameters which cannot be measured accurately in-situ or on-line. Advantages of at-line measurement include shortened time delay (relative to off-line), and the possibility for automated control; however the final results might be too slow to effectively monitor cultures with fast growth rates such as microbial cultures, according to PAT principles. On-line In on-line measurement, the sample is diverted from the manufacturing process with a by-pass stream and may be returned to the bioreactor. The sample is automatically measured in the by-pass by process sensors. The advantages of this method lie in the simple sterilization and the straightforward access to the sample in stationary conditions. The implementation of such a solution requires a specifically designed or modified bioreactor. The added complexity of set-up makes this method less common than in-line monitoring, yet it is one of the two methods with which constant monitoring and, therefore, control are possible in real-time. In-line or In-situ In in-line or in-situ, the measurement occurs directly in the bioreactor with a process sensor. The generated measurements are sent in real-time to PLC/SCADA systems for automated control. Process parameters such as pH, ORP (redox potential), dissolved oxygen, dissolved CO2 (DCO2), temperature, and conductivity are all common in-situ measurements. In-line and on-line sensors are the optimal choice for application of PAT principles. They are required to accurately measure without manual intervention over the entire process run, which can last several weeks or even months. Therefore, the operation and maintenance of the sensor should not be underestimated to guarantee reliable, accurate measurement. Preventative measures such as calibration and cleaning should be implemented at specified intervals to avoid drift or loss of signal. The sensors need to be compatible with repeated CIP and SIP cleanings. Extended time at temperatures of 120 to 130°C should not affect the sensor‘s performance. 17 Critical Process Parameters 5 Focus 1 Intelligent Arc Sensors for pH & DO In-Situ Measurement Efficient, reliable, compact design, and precise process control – these are the factors that GEA Diessel GmbH requires for monitoring the fermentation plant. They were found in Hamilton Arc sensors[A]. The measurement of the pH and dissolved oxygen takes place in the pre-fermenter as well as in the main fermenter, in this process. 18 Critical Process Parameters Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Physical and chemical critical process parameters are commonly monitored using in-line/on-line process sensors or at-line process analyzers. This chapter provides an overview of each parameter. Application examples are also presented in focus spots 1 and 2. pH The most classical example of PAT applied to bioprocesses is the maintaining of culture pH at a pre-defined set-point based on an in-situ electrochemical sensor signal. The signal is used to automatically regulate the addition of a base or acid (or the controls of CO2 for mammalian cells). The working range varies according to the applications. For example, mammalian cells vary between 6.8-7.4 pH, while others, like insect cells, are optimized around 6.3 pH. Tight control of this parameter is crucial. Drifting pH measurement often negatively influences the product’s yield in large scale manufacturing operations[7]. Keeping the pH in the correct working range has an impact both on cell viability as well as on the product’s quality. An example of the latter is how pH directly affects the therapeutic affect of mAbs: too low pH level negatively influences the protein’s glycosylation pattern resulting in a loss of their bioactivity[15]. Glycosylation is one of the most important CQAs for monoclonal antibody production. Cross flow Fermentation 750L 75L 150L Batching tanks 4-20 mA loop Arc Wi adapter Arc pH probe Arc DO probe Arc Cond. probe 150L 1 • The VisiFerm DO Arc optical dissolved oxygen sensor was chosen for its compact design and low maintenance costs. The optical sensing element is unaffected by pressure fluctuations and does not require polarization time at startup. • pH is measured with the EasyFerm Plus Arc sensor. Its pressurized reference and low drift after repeated sterilization cycles make it ideal for fermentation processes. • All Arc sensors include an integrated microtransmitter which transfers the measurement, CIP and SIP data, and sensor quality, as well as data regarding the operating life, via a wireless connection to the Hamilton ArcAir App for sensor management. 19 Focus 2 Dissolved Oxygen User’s Experiences The most common process pH sensors are electrochemical combination glass electrodes which are designed to withstand multiple autoclavations, CIP and SIP cycles. Alternative pH measurement solutions, such as optical sensors, exist as well; however they have limited measurement range, can only be sterilized by gamma radiation, and exhibit substantial drift to guarantee accurate measurement. Dissolved Oxygen Dissolved oxygen (DO or pO2) is another critical process parameter. Air or oxygen enriched air is supplied to the bioreactor to support cell demand. Oxygen is used for cellular respiration and cellular growth. While important, DO can be controlled over a broader range than pH without too significantly impacting cell growth rates or product quality. Typical DO operating ranges for aerobic cultures lie between 30 to 40% air saturation. DO levels below this range will affect cell viability, whereas excessive DO levels can oxidize the end-product. • Maintaining precision over multiple CIP/SIP cycles Dissolved oxygen concentration is directly related to cell growth and high protein yield. Roche Pharmaceuticals uses the optical VisiFerm DO sensor due to its robustness, enabling them to survive in their applications[B]. Each sensor is sterilized over 25 minutes at 121°C followed by a deionized water rinse. This procedure is repeated several times a week. Roche reports no degradation of the sensor and precise measurement over time. • Arc technology prevents downtime and allows reliable fermentation control UK based Albumedix focused their production on Saccharomyces to produce recombinant proteins[C]. An average fed-batch production run last approximately 5 days. During this time dissolved oxygen is reduced from 98% saturation to a designated control point using, as well, the optical VisiFerm DO sensor. The sensor outputs a 4-20 mA signal directly from the integrated microtransmitter into the biocontroller. Sensor status and actual measurement values can be easily checked through use of the Hamilton ArcAir app at any time. • VisiPro DO Ex for low dissolved oxygen content in ATEX zones The manufacturer of pharmaceutical agents (APIs) Medichem relies on an optical dissolved oxygen sensor[D]. The high dissolved oxygen sensitivity of sulfur compounds requires a sensor that measures reliably at ppb levels. In this application, the dissolved oxygen sensors were located in potentially explosive atmosphere thus ATEX certification was required. The optical VisiPro DO Ex Sensor provides quick response time and simplified maintenance compared to polarographic sensors. Its integrated microtransmitter outputs 4-20mA with HART protocol directly into the control system for simple integration. Time End of reagent addition End of reagent addition Deoxygenation Filtration 100 200 300 400 500 600 700 800 900 1000 0 2 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 20 Focus 3 Real-time Monitoring of DCO2 During Cell Culture for mAb Production Industrial bioproduction processes rely on innately variable living organisms to produce complex biomolecules, as such many parameters can affect the productivity and consistency of processes. Therefore, for producers to minimize negative effectors and maximize production, they must implement rigorous controls during production. Adhering to the FDA’s PAT initiative and implementing control of multiple parameters during production is one suggested method. At the bioreactor, in- / on-line monitoring of as many critical process parameters (CPPs) and key performance indicators (KPIs) as possible enables a better understanding of processes and immediate correction of deviations during production. TH OWL based in Germany applied this principle and monitored dissolved CO2 (DCO2) in addition to ubiquitous dissolved oxygen (DO2) during cell cultivation – to demonstrate the value of additional dissolved gas measurements during cell cultivation. In-line measurements continuously notify of process conditions in real-time, preventing knowledge gaps during production; this study measured in-line DO, DCO2 and pH. During this study, additional metabolic insights were achievable when gas measurements for both CO2 and O2 were also collected, as were process insights such as the effect of stirrer speed on parameter measurement and process performance. Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Two types of bioprocess DO sensors are commonly used: polarographic and optical. Polarographic sensors utilize an electrochemical cell (sometimes referred to as a Clark cell). This design was the first to market, and has limitations of high maintenance costs, extended startup time, and measurement errors due to fouling by CO2 and other gases. Optical DO sensors based on the newer quenched luminescent technology have begun to supplant polarographic technology and are now considered state-of-the-art for in-situ measurement. Dissolved Carbon Dioxide Dissolved CO2 (DCO2 or pCO2) is a parameter which is monitored due to its influence on pH values in mammalian cells and fatty acid synthesis. A higher dissolved carbon dioxide level can inhibit cell growth and reduce production of secondary metabolites. Carbon dioxide is especially critical in cell culture (mammalian) processes and must be kept within 5 - 10% air saturation[G]. Process DCO2 sensors based on the Severinghaus measurement principle have been available for many years, however this indirect method has received limited industrial uptake due to the significant maintenance efforts and costs required for accurate measurements. The Severinghaus principle underperforms as it combines the challenges of measuring pH and electrochemical DO in a single sensor[H,I]. Recent technological advances have enabled the development of a maintenance-free, solid-state optical CO2 sensor: CO2NTROL. An example of applications of optical CO2 sensor technology is shown in Focus box 3. 3 21 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Exhaust O2 & CO2 – Off-gas In some bioprocesses, like those using yeast for antibiotics production, culture viability is controlled by monitoring indirect indicators like Oxygen Uptake Rate (OUR) and Carbon Dioxide Evolution rate (CER). The ratio between O2 and CO2 entering the bioreactor and the exhaust/off-gas measured after air filter is used as a proxy of culture viability. Exhaust/off-gas can be measured with at-line mass spectrometers or with on-line process sensors based on galvanic measurement and infrared (IR) absorption. These measurements are mainly implemented for microbial culture, where the measurement is considered complicated and often not enough reliable. Nutrients & Metabolites Proper monitoring of nutrient (or substrates) concentration, as well as measurement of secondary metabolites, is important, especially for fed-batch and perfusion processes, as the feeding strategies can be controlled during the process. Glucose or glycerol are the main C-source (carbon source), while glutamine is the main N-source (nitrogen source), together with other amino acids in these bioprocesses. During the fermentation they are consumed, and secondary metabolites such as lactate, acetate and ammonium are produced. Suboptimal feeding strategies can produce excessive secondary metabolites which hinder cell viability and product yield. For example, the accumulation of lactate in mammalian cultures has long been recognized as an inhibitory factor for cell growth and recombinant protein production[13]. Therefore, control of nutrient feeding is of paramount importance (along with pH, DO and temperature), for process optimization. 3 0 0.2 60 0 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2 2.2 10 20 30 40 50 60 70 80 90 100 0 2 4 6 8 10 12 14 16 70 80 90 100 110 120 0.5 1 1.5 50 100 150 Cultivation time t/h Viable Cell Density [caells/mL] Dissolved O2 [% air saturation] Stirrer speed [RPM] cGlucose [mmcl/L] Dissolved CO2 [%] x106 cGlucos Viable Cell Density [cells/mL] e Dissolved CO2 [%] Dissolved O2 [% air saturation] Stirrer speed [RPM] [mmcl/L Figure 4 In-line measured values of DO and DCO2 along with off-line measuring of VCD and Glucose at different stirrer speeds. 22 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 In-line and on-line sensors are typically based on molecular spectroscopy technologies like Near-Infrared (NIR) and Raman. They are secondary measurement technologies, meaning measurement with off-line reference methods are required to calibrate them through use of statistical multi variate data analysis (MVDA). The measurement accuracy of these methods is strictly related to the specific bioprocess environment and to the quality of the off-line measurements used for calibration. There have been no published studies which show the use of the same global MVDA calibrations to predict different cell processes with an acceptable accuracy[7]. For all such reasons, they are considered too labor-intensive and too expensive for a successful implementation in production environments. The complexity of these measurements elucidate why at-line and/or off-line methods are still the most common option for nutrient and metabolite monitoring, despite not being optimal for PAT compliance. This equipment may utilize different technologies: HPLC, glucose oxidase, or other biochemical analysis to perform the measurement. These analyzers can be automated or semi-automated. They often require separate devices for sterile sampling and measurement cycles often require a relatively long measurement time (minutes)[14]. Yet, for the reason explained, they remain the only option available for measuring the culture’s substrate and secondary metabolites. Temperature, Pressure & ORP Temperature is a fundamental and well-controlled parameter in bioprocesses. Bioprocesses are typically monitored and controlled tightly between 0 and 60°C, including during sterilization cycles. Several devices and measuring principles are common to measure temperature in bioreactors such as thermistors, resistance and bimetallic thermometers[8]. Other physical and chemical CPPs, such as pressure and ORP can be controlled to optimize the cultures fermentation processes, as well. Pressure is an important control parameter because it affects not only the bioprocess but also safety. In general, it influences the saturation concentration of the gases dissolved in the liquid phase, like DO and DCO2. Most common measuring options are represented by piezoelectric-based and filled diaphragm transducers. Monitoring the oxidation reduction potential (ORP) provides information regarding the concentration of oxidizing or reducing molecules. ORP can provide important feedback for understanding the process: e.g. optimizing the yield from mammalian cells. Similar to pH sensors, the most common in-situ monitoring option is a combination reference/measurement electrode. 23 Critical Quality Attributes & Key Performance Indicators 6 24 Critical Quality Attributes & Key Performance Indicators Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Monitoring CPPs makes it possible to maintain the related Critical Quality Attributes and Key Performance Indicators within the pre-defined limits. Collected process data are used in case of need for root cause analysis or for process characterization studies based on experimental design (e.g. for scale-up and scale-down). PAT will be best fulfilled when CQAs and KPIs can be directly measured as often as possible, as explained in the following paragraphs. Product Quality & Product Titer The main goal of bioreactor operation is to produce as much product as possible with the quality needed to make it functional for its therapeutic purpose. Quality and yield are important as the product often requires further purification in downstream processes that may cause additional modifications or losses. As previously mentioned, the most important biopharmaceutical products are: monoclonal antibodies, recombinant proteins, or other types of therapeutic proteins (like vaccines). Therefore, bioprocess CQAs are often considered attributes specific to the protein’s quality such as the glycosylation patterns or molecular-size distribution[4]. The most commonly used KPI at the bioreactor is the total protein titer and eventually specific titer for the protein type (e.g. IgG). The most promising results for in-situ measurement of product titer and quality have been obtained with spectroscopic technologies, which have the same limitations described for nutrient and metabolite measurements. Likewise, HPLC, mass spectrometers, NMR, fluorescence or super-resolution microscopy, capillary electrophoresis or biochemical analyzers installed at-line or off-line are often seen as more reliable solutions to quantify the mentioned CQAs and KPIs at the bioreactor[14]. Again, sterile sampling technologies and procedures make these quality attributes and process indicators limited with respect to PAT control guidelines. Total and Viable Cell Density Other Key Performance Indicators successfully used for in-situ control at the bioreactor are Total Cell Density (TCD) and Viable Cell Density (VCD). TCD indicates the total amount of cells in the bioreactor, while VCD is an indicator of the viable cells (alive and still productive). VCD is directly correlated with final product yield and is thus of high importance. Different off-line measurement technologies have been established over the years[8]. For example, total cell density, as well as viable cell density, can be measured via off-line automated cell counter systems. A major disadvantage is that these methods are based on time-consuming sampling procedures which reduce the possibility that the cell growth is monitored in a process-safe way compatible with PAT principles. For this reason several efforts have been put forth over the years to find technologies suited for accurate and reproducible real-time measurements. 25 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Some efforts for real-time cell density are based on the use of molecular spectroscopy, others on soft sensing techniques (e.g. algorithms based on the evolution of OUR and CER), both requiring MVDA to generate application specific calibrations which are labor-intensive to maintain. The most reliable measurements are obtained using near-infrared light to measure culture turbidity in microbial fermentations, or by using capacitance sensors to measure cell viability (especially for mammalian cell cultures). Turbidity and capacitance are currently the most common technologies used to measure TCD and VCD in real-time. These sensors can withstand autoclavations, CIP and SIP cycles to meet hygienic standards. Examples of the application of such technologies are provided in focus spot 4,5 and 6. 26 Focus 4 In-Situ Cell Density for Batch & Perfusion mAb Production The monitoring of KPIs such as Total Cell Density and Viable Cell Density are currently available as in-situ measurements with process sensors. Studying these indicators allows the real-time control of the nutrient feed rate based on the growth rate of cell cultures. The Cell Culture Research Team of the University of Bielefeld investigated the accuracy of Incyte and Dencytee (Hamilton VCD and TCD sensors), for both perfusion and batch production of mAb using CHO cells[E]. The test demonstrated the following benefits: • Accurate measurement of the cell growth enabling real-time control • Better insight about cell health based on the parallel measurement of TCD and VCD • Reliable and stable measurements for long-lasting continuous fermentations 4 The Incyte is based on capacitance principles. In an alternating electrical field, viable cells behave like small capacitors. The charge from these small capacitors is measured by the sensor and reported as permittivity (capacitance per area). The Dencytee is based on turbidity measurement of a suspension at NIR wavelengths. All particles and molecules that scatter the NIR light will be detected and can be correlated to the total cell density. Time / h Experiment 1 - Perfusion Process 1 20 40 60 80 100 120 140 0 2.30 3.30 4.30 5.30 6.30 1.30 0 50 100 150 200 250 300 350 10 20 30 40 50 60 0 Time / h Experiment 3 - Batch Process 1 1.80 2.30 2.80 3.30 4.30 3.80 1.30 0 20 40 60 80 1000 120 Performance comparison of cell density measurement for the batch set-up. Performance comparison of cell density measurement for the perfusion set-up. Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 27 Focus 5 Validation of Density Measurement for Different Cell Types Hamilton’s cell density sensors Dencytee (total cell density) and Incyte (viable cell density) have been validated for use with different cell types including both prokaryotic (Gram-positive and Gram-negative bacteria) and eukaryotic cells (mammalian, insect, fungi, yeasts and algae)[16] [F, L, M, N, O, P]. Shown below are examples of data collected using Dencytee and Incyte, either together or in combination, for different cell types, highlighting the compatibility of these measurement principles for different applications. CHO cells, Incyte and Dencytee[F] Human Cells, Incyte[L] Insect, Incyte[M] Yeast, Dencytee[N] 5 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Time / h A 6.75 7 7.25 7.5 6.5 25 50 75 100 0 0 65 130 195 260 pHonline pHoffline Dissolved oxygen Time / h C 65 110 165 220 0 0 65 130 195 260 measured VCD Incyte model Dencytee model Time / h D 65 110 165 220 0 0 65 130 195 260 Time / h B 6.75 7 7.25 7.5 6.5 25 50 75 100 0 0 50 100 150 200 0 4 12 18 6.00 0.00 0.10 3.62 7.00 50.00 2.55 102.69 8.00 100.00 5.00 201.76 24 pH pO2 VCD Time / h Time / h 50 100 150 200 250 300 350 400 450 500 550 600 0 0 5 10 15 20 25 30 35 40 45 50 50 100 150 200 250 300 350 400 450 0 DCW In-line automated DCW Off-line manual WCW In-line automated WCW Off-line manual OD In-line automated ODW Off-line manual Process time / h 2 4 6 8 10 12 0 0 24 48 72 96 Off-line Data polyclonal Off-line Data Wild-Type In-line Data polyclonal In-line Data Wild-Type 28 Cyanobacteria, Dencytee[O] Moss, Incyte[P] 5 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Time / day 0 10 20 30 40 50 60 70 80 90 100 110 120 130 140 150 160 170 180 0.5 1.0 1.5 2.0 2.5 3.0 3.5 0.0 0 20 40 60 80 100 120 140 160 180 200 220 240 260 280 300 320 340 360 2 4 6 8 10 0 7 14 21 28 35 0 Off-line cell density [g/L] LED Intensity Cell density correlation Cultivation time / h 29 Focus 6 The Benefits of Monitoring Cell Density During Various Applications Due to the necessity of producing large volumes of viable cells during bioprocesses in Biopharma, in-line viable cell density sensors, such as Incyte, offer the potential for more than just determining the percentage of viable cells in a process. The continuous, real-time data they collect enables users to understand the effects of process changes on performance (e.g., addition or limitation of nutrients on cell physiology) and to optimize conditions to prevent apoptosis and/or extend the production phase, in addition to determining timings for seed transfer during scaling for maximum productivity and efficiency with respect to cultivation recovery. The following examples demonstrate the breadth of application of these sensors and the versatility of the multivariate information available from a single continuous in-line measurement. Cells grown on Microcarriers[J] Continuous processing – perfusion processes using mammalian cells (CHO)[J],[K] Time / h This data point is an outlier that can be seen as compared to the Incyte measurement. Without Incyte, an incorrect process estimation would likely happen. 20 40 60 80 100 120 140 0 1.0E+06 2.0E+06 3.0E+06 4.0E+06 5.0E+06 6.0E+06 7.0E+06 0.0E+06 0.00 20.00 40.00 50.00 80.00 100.00 120.00 140.00 Viable Cell Count Incyte Measurement Cell Retention Bioreactor Feed Cell Bleed Media Harvest Waste 6 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Time / h 2 1 3 4 6 8 10 9 7 5 11 0 1 2 3 4 5 6 7 0 0 50 100 150 200 250 Glutamine [mmol/L] Alanine [mmol/L] Glucose [g/L] Incyte Measurement 30 Automated control of seed train transfer[J],[K] Improving the efficiency of processes during scale transfer[N],[17] Time Time / h 20 40 60 80 100 120 0 2.30 3.30 4.30 5.30 6.30 1.30 0 50 100 150 200 250 300 Off-Line Incyte Measurement Dencytee Measurement 6 Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 0 2 4 6 0 25 50 75 100 125 80 85 95 90 100 Time / days VCD [x106 cells/mL] Cell Line mAb4 Viability [%] 0 2 4 6 0 25 50 75 100 125 80 85 95 90 100 Time / days Cell Line mAb5 Viability [%] VCD [x106 cells/mL] 5L VCD 200L VCD 500L VCD 5L Viability 200L Viability 500L Viability 31 Conclusions 8 32 Conclusions Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 With the implementation of PAT in bioprocesses, Critical Quality Attributes, Critical Process Parameters and Key Performance Indicators must be monitored. The most valuable measurements are performed in-situ to allow for real-time control strategies. The data derived from these continuous measurements are also valuable for use in root cause analysis or for process characterization such as scale-up and scale-down studies. Table 1 summarizes the most important CPPs, CQAs and KPIs at the bioreactor, with an overview of the available measuring methods along with their typical accuracy and limitations. The crucial CPPs allowing for real-time control strategies are pH, DO, and temperature. Other parameter’s measurement such as DCO2, nutrients and metabolites are accurate and repeatable just through at-line or off-line analyzers, making them complicate for real-time control strategies. In regards to CQAs and KPIs, those related to the product quality and product titer are important, nonetheless they are likewise complicate to measure in-line with acceptable accuracy. In conclusion, the real-time monitoring of the mentioned crucial CPPs, together with the in-line measurement of TCD and VCD represents currently the best PAT option to control product yield and quality at the bioreactor. Table 1: Summary of the CPPs, CQAs and KPIs at the bioreactor PAT Method of Choice Alternative Methods In-line / In-situ Sensor On-line Analyzer At-line & Off-line Analyzer Critical Process Parameter pH DO DCO2 Temperature Exhaust O2/CO2 Nutrients e.g. Glucose, Glutamine Metabolites e.g. Lactate, Ammonium Critical Quality Attribute Product Quality e.g. Protein Glycosylation Key Performance Indicator Product Titer Total Cell Density Viable Cell Density The availability of monitoring methods according to the scientific literature is represented with the indication of the corresponding measurement accuracy and robustness: Accuracy, robustness and repeatability good enough to be commonly implemented for process control Accuracy, robustness and repeatability not commonly accepted for process control No true option available Not required 33 References 9 34 References [1] U.S. Department of Health and Human Services (2004): Guidance for Industry. PAT – A Framework for Innovative Pharmaceutical Development, Manufacturing, and Quality Assurance. Rockville. [2] FDA, Pharmaceutical cGMPs for the 21st century – A risk based approach; Final Report, September 2004 [3] M. Mitchell, Determining Criticality-Process Parameters and Quality Attributes, BioPharm International, Volume 26, Issue 12, 2013 [4] S. Haigney, QbD and PAT in Upstream and Downstream Processing, www.processdevelopmentforum.com, August 5, 2013 [5] Rakhi B. Shah et al., Application of PAT in Biotech Drug Substance Manufacturing, Biotechnology and Bioprocessing, Series Vol 33, Pages 3-4, 2012 [6] C. Undey, D. Low et al., PAT applied in Biopharmaceutical Process Development and Manufacturing, CRC Press, 2012 [7] A. V. Carvalhal, V. M. Saucedo, Process Analytical Technology Advances and Applications in Recombinant Protein Cell Culture Processes, Biotechnology and Bioprocessing, Series Vol 33, Pages 93-126, 2012 [8] M. Pohlscheidt, M. Jenzsch et al.: Bioprocess and Fermentation Monitoring. In: Upstream Industrial Biotechnology: Equipment, Process Design, Sensing, Control and cGMP Operations, Volume 2, First Edition, Edited by Michael C. Flickinger, John Wiley & Sons., 2013 [9] E. S. Langer, R. A. Rader Continuous Bioprocessing and Perfusion: Wider Adoption Coming as Bioprocessing Matures, BioProcessing Journal, p. 50-55, Spring 2014 [10] W. S. Langer, Average Batch Failure Rate Worsens, Genetic Engineering & Biotechnology News, Vol. 36, No. 17, 01 October 2016 [11] G. L. Rosano, E. A. Ceccarelli, Recombinant protein expression in microbial systems, Editorial, Frontiers in Microbilogy, Volume 5, Article 341, 8 July 2014 [12] G. Melmer, G. Gellissen, G. Kunze, Recombinant Vaccine Production in Yeast, BioPharm International, Volume 2008, Issue 1, 02 January 2008 [13] M.M. St. Amand, P. G. Millili et al. Strategic Vision for Integrated Process Analytical Technology and Advanced Control in Biologics Manufacturing, Series Vol 33, Pages 9-28, 2012 [14] W. Whitford, C. Julien, Analytical Technology and PAT, Supplement Bioreactors Chapter 3, BioProcess International, p. 32-41, January 2007 [15] B. C. Mulukutla, A. Yongky, S. Grimm et al., Multiplicity of Steady States in Glycolysis and Shift of Metabolic State in Cultured Mammalian Cells, PLoS One; 10(3): e0121561., 25 March 2015 [16] Bernadett Kiss, Áron Németh, Application of a High Cell Density Capacitance Sensor to Different Microorganisms, Periodica Polytechnica Chemical Engineering, Volume 40, Issue 4, p. 290-297, 2016. (View of Application of a High Cell Density Capacitance Sensor to Different Microorganisms (bme.hu)) [17] Rittershaus, E. S. C. et al. N-1 Perfusion Platform Development Using a Capacitance Probe for Biomanufacturing. Bioengineering 9, (2022). https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9029935/ Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 35 All white papers, application notes and Hamilton documentation are available for download at www.hamiltoncompany.com [A] M. Culina, C. Brokamp, More functionality, lower costs, better usability with the Arc system, Ref. 695099, 2012 695099_01_Arc_in_GEA_EN.pdf (hamiltoncompany.com) [B] C. Miscenic, Oxygen Measurement in Fermentation with VisiFerm DO, Ref. 695098, 2013 695098_03_AppNote_Visiferm_EN.pdf (hamiltoncompany.com) [C] M. Williamson, VisiFerm DO in the production of recombinant proteins, Ref. 695170, 2017 695170_AppNote_VisiFerm-Albumedix_EN_LR.pdf (hamiltoncompany.com) [D] J. Garcia, M. Benito, Dissolved oxygen quantification in a product sensitive to it, Ref. 695218, 2016 695218_AppNote_DO-Quantification_EN_LR.pdf (hamiltoncompany.com) [E] H. Büntemeyer, A. Schmidt, Real-time Cell Density Measurement for PAT Applications, Ref. 695234, 2017a Real-time cell density measurement for PAT applications | Process Analytics (hamiltoncompany.com) [F] K. Kandra, P. Kroll, M. Brunner, P. Wechselberger, C. Herwig Online Monitoring of CHO Cell Culture, Ref. 695172, 2014 Online monitoring of CHO cell culture | Process Analytics (hamiltoncompany.com) [G] Real-time Monitoring of DCO2 in Addition to DO, Ref: 111006479/00 Real-time Monitoring of DCO2 in Addition to DO | Process Analytics (hamiltoncompany) [H] Should CO2 Be A Critical Process Parameter? Ref: 111003179/00 Should CO2 Be A Critical Process Parameter? | Process Analytics (hamiltoncompany.com) [I] Are CO2 Measurement Technologies Good Enough? Ref: 111003180/00 Are CO2 Measurement Technologies Good Enough? | Process Analytics (hamiltoncompany.com) [J] Cell Density Guide, 2018 Ref: L30019 Cell Density Guide | Process Analytics (hamiltoncompany.com) [K] Sensing the Future, 2020 Ref: 10105112 Sensing the Future | Process Analytics (hamiltoncompany.com) [L] Human Platelet Protein Production from Human Cell Culture using an Advanced Bioreactor, Ref: 111004661/00. Human Platelet Protein Production from Human Cell Culture using an Advanced Bioreactor | Process Analytics (hamiltoncompany.com) [M] Correlation between Capacitance Signal and Viable Cell Density of Insect Cells, Ref: 111004662/00. Correlation between Capacitance Signal and Viable Cell Density of Insect Cells | Process Analytics (hamiltoncompany.com) [N] Real-Time Total Cell Density Measurement of Yeast Fermentations, Ref: 695241/00. Real-Time Total Cell Density Measurement of Yeast Fermentations | Process Analytics (hamiltoncompany.com) [O] Continuous Cyanobacteria - Limnospora (Spirulina) Monitoring in the MELiSSA Loop using Dencytee, Ref: 111004663/00. Continuous Cyanobacteria | Process Analytics (hamiltoncompany.com) [P] Real-Time VCD Monitoring of Moss for Therapeutic Protein Production, Ref: 695247/00. Real-Time VCD Monitoring of Moss for Therapeutic Protein Production | Process Analytics (hamiltoncompany.com) Biopharma PAT: Quality Attributes, Critical Process Parameters & Key Performance Indicators at the Bioreactor 05. May 2024 Throughout this document, protected product names may be used without being specifically marked as such. Research use only. Not for use in diagnostics procedures. All rights reserved. All other trademarks are the sole property of their respective owners. © 2024 Hamilton Company. All rights reserved. All trademarks are owned and/or registered by Hamilton Company in the U.S. and/or other countries. Lit. No. 695237/02 – 05/2024 To find a representative in your area, please visit: www.hamiltoncompany.com/contacts Hamilton Europe, Asia & Africa Hamilton Bonaduz A.G. Via Crusch 8 CH-7402 Bonaduz, Switzerland Tel: +41-58-610-10-10 [email protected] Hamilton Americas & Pacific Rim Hamilton Company Inc. 4970 Energy Way Reno, Nevada 89502 USA Tel: +1-775-858-3000 Fax: +1-775-856-7259 Web: [email protected] www.hamiltoncompany.com USA: 800-648-5950 Europe: +41-58-610-10-10 Unlock the Secrets of Bioprocessing Excellence Download and Find Out More O2 Measurement Guide Measurement Challenges Optical Dissolved Oxygen Sensors White Paper Biopharma Downstream Critical Process Parameters Cell Density Applications eBook pH Measurement Guide Dissolved CO2 Series Part 2 Are current dissolved CO2 measurement technologies good enough? MEASUREMENT GUIDES WHITE PAPERS Dissolved CO2 Series Part 1 Should CO2 be a critical process parameter?