Vol. 12, Supplement 5

A SPECIAL SUPPLEMENT TO BIOPROCESS INTERNATIONAL Innovations in Cell Culture

IN ASSOCIATION WITH New Flexsafe Bag Family. New PE Film. New Benchmark.

Our new Flexsafe bags ensure an excellent and reproducible growth behavior ONE FILM FOR ALL with the most sensitive production cell lines. The optimization of the resin formulation, the complete control of our raw materials, the extrusion process and the bag assembly guarantee a consistent lot-to-lot cell growth performance. www.sartorius-stedim.com/flexsafe USP – DSP – F+F VOLUME 12 SUPPLEMENT 5 SEPTEMBER 2014

From the Editor ...... 2 Robust and Convenient Single-Use Processing: The Superior Strength and Flexibility of Flexsafe Bags...... 38 SECTION 1: UPSTREAM TECHNOLOGIES Elisabeth Vachette, Christel Fenge, Jean-Marc Cappia, Automated Miniature Bioreactor Technology: Optimizing Early Lucie Delaunay, Gerhard Greller, and Magali Barbaroux Process Development in Fermentation and Bioprocessing . . . . . 3 Enhanced Assurance of Supply for Single-Use Bags: Mwai Ngibuini Based on Material Science, Quality By Design, Single-Use, Stirred-Tank Bioreactors: Efficient Tools for Process and Partnership with Suppliers ...... 43 Development and Characterization ...... 7 Jean-Marc Cappia, Elisabeth Vachette, Carole Langlois, Andre Grebe, Christel Fenge, and Jean-François Chaubard Magali Barbaroux, and Heiko Hackel Design of Experiments with Small-Scale Bioreactor Systems for Development and Qualification of a Scalable, Disposable Efficient Bioprocess Development and Optimization...... 10 Bioreactor for GMP-Compliant Cell Culture ...... 47 Andree Ellert and Conny Vikström Anne Weber, Davy De Wilde, Sébastien Chaussin, Thorsten Adams, Susanne Gerighausen, Gerhard Greller, and Christel Superior Scalability of Single-Use Bioreactors ...... 14 Fenge Davy De Wilde, Thomas Dreher, Christian Zahnow, Ute Husemann, Gerhard Greller, Thorsten Adams, and Christel Fenge Verification of New Flexsafe STR Single-Use Bioreactor Bags: Using a CHO Fed-Batch Monoclonal Antibody Production Process Integrated Optical Single-Use Sensors: Moving Toward a True at 1,000-L Scale...... 53 Single-Use Factory for Biologics and Vaccine Production...... 20 Regina Reglin, Sebastian Ruhl, Jörg Weyand, Davy De Wilde, Henry Weichert, Julia Lueders, Joerg Weyand, Ute Husemann, Gergard Greller, and Christel Fenge Thorsten Adams, and Mario Becker Pressure Decay Method for Post-Installation Single-Use Diatomaceous Earth Filtration: Innovative Single-Use Concepts Bioreactor Bag Testing ...... 58 for Clarification of High-Density Mammalian Cell Cultures ....25 Magnus Stering, Martin Dahlberg, Thorsten Adams, Tjebbe van der Meer, Benjamin Minow, Bertille Lagrange, Davy De Wilde, Christel Fenge Franziska Krumbein, and François Rolin Total Solutions Support the Growth of a Dynamic Industry: An Designing Full Solutions for the Future: Interview with Reinhard Vogt and Stefan Schlack...... 62 An Interview with Christel Fenge...... 29 Brian Caine and S. Anne Montgomery Brian Caine and S. Anne Montgomery LAST WORD SECTION 2: DEVELOPMENT OF Disposables for Biomanufacturing: A User’s Perspective ...... 67 THE NEW FLEXSAFE BAG FAMILY Berthold Boedeker Consistently Superior Cell Growth Achieved with New Polyethylene Film Formulation...... 34 Christel Fenge, Elke Jurkiewicz, Ute Husemann, Thorsten Adams, Lucy Delaunay, Gerhard Greller, and Magali Barbaroux EDITORIAL OFFICES SALES AND ADMINISTRATIVE OFFICES PO Box 70, Dexter, OR 97431 One Research Drive, Suite 400A Production and Creative Manager Westborough, MA 01581 Genevieve McCarthy 1-212-520-2752 Editor in Chief (sales inquiries, media kits, advertising) [email protected] S. Anne Montgomery [email protected] Publisher Director of Audience (article and supplement queries, editorial policies) Brian Caine 1-508-614-1443 Development and Manufacturing [email protected] Nora Pastenkos 1-212-520-2733 Senior Technical Editor [email protected] Cheryl Scott Eastern Regional Sales Manager [email protected] Christopher Johnson 1-508-614-1273 Marketing and Digital Content Strategist (press releases, art submissions, design) [email protected] Leah Rosin 1-508-614-1167 [email protected] Managing Editor Western Regional Sales Manager Maribel Rios [email protected] Mike Kelly 1- 646-957-8974 List Rental Amy Miller (article queries, special projects) [email protected] 1-508-614-1251 [email protected] Reprints Rhonda Brown Editorial Assistant Alison Center European Sales Manager 1-866-879-9144 x194 [email protected] Joanna Taylor 44-(0)-20-7551-9392 [email protected] [email protected] Find BPI citations online in the Chemical Sales and Marketing Coordinator For subscription inquiries, call 1-847-763-4930, Abstracts Database (www.cas.org). Kim Rafferty 1-508-614-1226 toll free 1-877-232-2399, or email bioprocess@ [email protected] halldata.com.

©2014 BioProcess International (USPS 0022-044, ISSN 1542-6319) is published eleven times a year by Informa Life Sciences Group at 52 Vanderbilt Ave., New York, NY 10017, Phone: 1-212-520-2777, fax 1-212-661-5052, www.bioprocessintl.com. Periodicals postage is paid in Westborough, MA and additional mailing offices. POSTMASTER: Send address changes to BioProcess International, PO Box 1170, Skokie, IL 60076. Canadian publication agreement No. 41067503. Canadian return address DPGM 7496 Bath Road Unit 2, Mississauga, ON L4T 1L2. Electronic subscriptions are available online. Articles are abstracted by the Chemical Abstracts database at www.cas.org.

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 1 FROM THE EDITOR his special issue has come about through a long Among products highlighted in these articles and interviews collaborative process with our contacts at Sartorius is the company’s new bag tester. Proper handling of single-use T Stedim Biotech in Göttingen, Germany. It is the third such materials is another common theme these days. Users want to sponsored issue that BPI has published over the years with the prevent leaks and tears as well as ensure bag compatibility with company and the first one that we have produced in-house. We the products that come into contact with them. The bag tester is could not be happier with the results. yet another element of the company’s goal to provide total The aim here is to highlight technologies and responses to solutions to its customers. Along with a focus on training, key industry challenges rather than focus over much on offering clients such a tool should increase confidence in Sartorius itself. But the company’s long-standing culture of adoption of single-use systems overall. developing its own and integrating acquired technologies and The topic of scalability is often the proverbial elephant processes — toward its goal of providing total biomanufacturing lounging in the back of the room at single-use conferences. How solutions — is worth underscoring. In the two interviews in this large can single-use bags go? The consensus in this issue is that issue you will read about a company that has a long history of 2,000 L may be perfectly adequate, now that titers are higher investing in and positioning its technologies and related and processes can be scaled to run in parallel. This is not the last equipment for future needs, even when the current industry has word on the topic, of course, but a number of authors herein not been quite ready for them yet. In this risk-averse era, such discuss reasons for that conclusion. long-term thinking can require no small amount of corporate One of those reasons is that with Sartorius Stedim Biotech’s courage. Through its own R&D activities and combinations of recent acquisition TAP Biosystems and its scalable line of ambr licensing, partnering, and acquisitions the company has crafted bioreactors, the company just doesn’t see the need for larger its current identity and continues on a well-defined path. systems. Customers can use TAP systems to effectively optimize As part of this issue’s preparation, BPI publisher Brian Caine their processes and titers. Also, many larger companies still have and I traveled to the Sartorius Stedim Biotech facility in Göttingen, stainless steel equipment to which they transfer processes for Germany. There we interviewed Christel Fenge (vice president of large-scale manufacturing — so questions remain in the marketing for fermentation technologies), Reinhard Vogt industry about what biomanufacturing companies want and (executive vice president of marketing, sales, and services and a truly need at later processing stages. member of the administrative board), and Stefan Schlack (senior Thus, authors in this issue are primarily examining bioreactor vice president of marketing and product management). Brian and design innovations, qualification and GMP development, use of I were grateful for the time they devoted to speaking with us — those bioreactors with the company’s line of Flexsafe bags, all during a very busy few days. That week the company was also testing and qualification of the new film, and the status of conducting a two-day conference at its Sartorius College. sensor development. The latter will be an essential component Academic and industry speakers presented analyses of PAT tools for any totally single-use manufacturing operation in the future, and how those can enhance process modeling and development when automation will be key. And although this issue focuses under the quality by design (QbD) initiative. That event on upstream development, we conclude with an assessment of highlighted Sartorius Stedim Biotech’s commitment to educating a single-use, filtration technology based on a diatomaceous- and training the coming generations of bioprocessors. earth cell removal approach — just to take a quick look at the Our general theme here is the development and continual interface between upstream and downstream. optimization of upstream technologies, so articles related to We thank the authors for contributing to this issue and for fermentation issues and to development and design of single- working closely with us to prepare their manuscripts for use bioreactors make up the first half of this supplement. This publication. We especially thank Sartorius Stedim Biotech’s continues BPI’s exploration of the present and future viability of corporate communications manager, Dominic Grone, who single-use technologies (highlighted recently in our April coordinated this project with us, circulated review copies to the supplement). The ultimate goal for many industry insiders is to authors, arranged for and helped us conduct the two interviews, put together a completely single-use process from production and generally helped us keep track of the many pieces and through formulation, fill, and finish. Certain assurances need to multiple sets of expectations. This is truly a Sartorius Stedim be brought to the industry at large regarding one-time use of Biotech product, and we enjoyed working with this company’s production and processing equipment. For one thing, what are team to provide you with this publication. these materials, really? This is a topic of increasing interest among our readers: how plastic materials and components are made and how companies are ensuring supply-chain consistency and robustness. Herein you’ll find Sartorius Stedim Biotech’s solution to that particular concern — forming the second half of this issue. The S. Anne Montgomery company is providing and qualifying its own film by a strong editor in chief cooperation with its partner, Südpack, and is taking control of associated risks in material suitability, reliability, and availability (sourcing of raw materials). Some articles in this issue explain the intricacies of that process.

SUPPLEMENT 2 BioProcess International 12(8)s SEPTEMBER 2014 B IOP ROCESS PROCESS DEVELOPMENT

Automated Mini Bioreactor Technology for Microbial and Mammalian Cell Culture Flexible Strategy to Optimize Early Process Development of Biologics and Vaccines by Mwai Ngibuini

he use of mammalian and Photo 1: ambr250 scale-down bioreactor system for parallel fermentation and cell culture microbial cells in the production of biologics and vaccines is well established, and Tthe majority of the top 10 drugs are now manufactured in this way. There is a significant and growing pipeline of new biologics (1), which in combination with increased pressure on cost reduction and generic competition from biosimilars (2), means that many biopharmaceutical companies are looking for ways to improve productivity in their development laboratories to ensure that upstream processes are efficient and robust. One of the main challenges for those companies is to minimise the high cost of goods that is typically associated with cell culture production processes to ensure the commercial viability of a therapy. For example, monoclonal antibody (MAb) therapies are required in large doses (6–12 g) to achieve clinical efficacy. That means they can cost tens of thousands of dollars per patient per year. Currently, the colorectal cancer treatments bevacizumab and cetuximab cost Manufacturing prophylactic produce affordable prophylactic $20,000–30,000 for an eight-week vaccines — which are a relatively vaccines for emerging markets, as well course (2), well over 60 times more small class in comparison with MAbs as for preventing seasonal influenza than comparable small-molecule in terms of global sales value — is also and responding to pandemic threats. therapies. challenging. Manufacturers need to Such cost and timeline issues are now SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 3 Figure 1: Optical density (OD) profiles of recombinant E. coli fermentation in the ambr250 bioreactor system, 15-L and 150-L fermentors developed by the application of complex DoE (design of experiment) Microbial data: ambr250 shows good OD methodologies. Such methodologies scalability data from 15-L through to 150-L scales have a very high experimental burden (4). All areas of the process must be fully understood, including media ambr2, 0.25 LOD components and all process parameters (AU) ambr1, 0.25 LOD (and the effects of changing them), B097, 15 LOD600

600nm B098, 15 LOD600 because such factors alter both titers B105,15 LOD600 OD and product quality attributes. B106, 15 LOD600 Because of the inherent problems B100, 150 LOD600 B104, 150 LOD600 associated with traditional B108, 150 LOD600 technologies, the resulting data can 0 12 24 36 48 60 72 include a number of errors that require Elapsed Fermentation Time (hours) constant experimental iterations. Such issues lead to costs that are prohibitive to the development of many Figure 2: CER profiles of recombinant E. coli fermentation in the ambr250 bioreactor system, 15-L and 150-L fermentors bioprocessing strategies. Well-characterized and accelerated Similar scalability results are seen with CER data process development and optimization from 15 L through to 150 L can be achieved only if the process can be sufficiently automated in a parallel miniaturised platform that is both easy to set up, scalable, and compatible B108, 150 L CER with disposable technologies. So there B104, 150 L CER is a need to move away from B100, 15 L CER traditional bioprocess equipment and B097, 15 L CER B098, 15 L CER into a platform that enables high- B105, 15 L CER throughput process development and B106, 15 L CER optimization. That type of technology ambr1, 0.25 L CER ambr2, 0.25 L CER should have three key characteristics: • miniaturization to enable faster experimental throughput at a low costs Carbon Evolution Rate (mMol/L/h) Rate Evolution Carbon 0 12 24 36 48 • automation for accurate, Elapsed Fermentation Time (hours) reproducible performance of a large number of individual operations driving biopharmaceutical companies quality, safety, and effectiveness. Part • parallel processing to allow to search for strategies to perform of those requirements is to provide evaluation of a wide experimental rapid process development for robust scientific evidence obtained space, resulting in process optimizing and scaling-up their during process development to support understanding. manufacturing processes. regulatory obligations. As a Here I present an automated mini Increasingly flat revenues and high consequence, there is a need for bioreactor technology that allows for development costs have made time-to- multifactorial statistically designed high-throughput process development market crucial for company research during process development and optimization for both microbial profitability. So there is a greater need and optimization, leading to a large fermentations and cell culture to release products with desired number of bioprocess experiments, processes. In addition, I demonstrate quality attributes as early as possible. which can be labour intensive, time the suitability of this technology by So for biopharmaceuticals, rapid and consuming, and expensive. analyzing data obtained from an efficient process development is Traditionally, technologies such as automated bioreactor and comparing essential for successful shake flasks and bench-top bioreactors data from laboratory- and pilot-scale commercialisation (3). or spinner flasks for vaccine production bioreactors. One challenge for biologics have been used for process manufacturers is the burgeoning development. However this approach is MATERIAL AND METHODS number of new regulatory manually intensive and prone to human Mini Bioreactor: The mini bioreactor requirements. Before a drug or vaccine error. It also has a high operating cost system chosen for scale-up comparison can be launched, its manufacturer and requires a large laboratory. is the ambr250 automated bioreactor must show that the product meets all A well understood biologics (from TAP Biosystems, a Sartorius regulatory requirements for high manufacturing process is best Company). This system has three

4 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 3: Growth profile of CHO clones cultured in ambr250 and 3-L bioreactors The clone cells were cultured in the Mammalian cell count data shows good scalability ambr250 system, which used a single- between ambr250 and bench-top bioreactors use microbial bioreactor with a dual 20-mm Rushton impeller. The same clones were cultured in 15-L bench-top fermentors and in 150-L cells/mL) 6 pilot-scale fermentors. 10

u 3 L All vessels were inoculated with 3 L 2–5% inoculum at an optical density at 3 L 3 L 600 nm (OD600) between 1 and 5. ambr250 The cell lines were cultured for ambr250 72 hours in chemically defined medium at 37 ˚C, pH 7.0 ± 0.03, 30% dissolved oxygen (DO), and an Sample ViCell VCC VCC ( Sample ViCell 0 2 4 6 8 10 12 14 16 impeller speed of gassed power per Offset (Days) volume between 5.33 = 102 and 1.43 = 104 W/m3. Samples were analyzed every 12 hours for cell density using a Figure 4: Titre profile of CHO clones cultured in stirred ambr250 and 3-L bioreactors SpectraMAX plus 384 spectrophotometer (from Molecular

Titer results also demonstrate good scalability Devices) to measure the OD600 value between the ambr250 and the benchtop bioreactors of cells. With the bench-top and pilot-plant fermentors, the carbon dioxide evolution rate (CER) was measured 3 L using a Prima PRO process mass 3 L spectrometer (from Thermo 3 L 3 L Scientific). With the ambr250 system, ambr 250 built-in off-gas analyzers measured ambr 250 the CER, with each bioreactor having its own dedicated off-gas analyzer. Scale-Up of Mammalian Cells: For

Sample Biacore Titer (mg/L) Titer Sample Biacore the ambr250 bioreactor system to be a viable early process development 0 2 4 6 8 10 12 14 16 model for mammalian cell culture, the Offset (Days) results from cultures grown in it and a in a bench-top bioreactor must be comparable. To show this, components: easy-connect single-use for either a 12 or 24 bioreactor recombinant CHO clones expressing a 250-mL bioreactors (available in both workstation. Both configurations therapeutic antibody were chosen as microbial and mammalian include an automated liquid handler the model organism. The clones were configurations), an automated for liquid transfer between bioreactors, cultured in the ambr250 system using workstation, and software (Photo 1). sample beds, or media bottles and can the single-use mammalian bioreactor The platform provides increased be used for automated sampling, and in 3-L bench-top bioreactors. The process volumes, pumped liquid inoculation or even media preparation. cell lines were cultured for 16 days in delivery for continuous feeds, and a proprietary, chemically defined automated individual bioreactor PROOF-OF-CONCEPT medium at 37 °C, pH 7.0 ± 0.3, 40% control for all parameters. These Scale-Up of Microbial Cells: To DO, and with an impeller tip speed of features allow a frequent feeding demonstrate that the ambr250 system 0.25 m/s. Cells were inoculated at 1 = regime and larger volumes to be is a viable process development and 106 viable cells/mL, and samples were sampled for performing a wide range scale-down model for microbial analyzed every 24 hours using a of analytical tests. Combined with the culture, the results from cultures ViCell cell viability analyzer (from parallel control of culture conditions grown in the mini bioreactor, a Beckman Coulter). Titres were and feeds, such capabilities provide a bench-top fermentor, and a pilot-scale assessed from day 2 and were then scale-down bioreactor model that fermentor must be comparable. To this measured every 24 hours using a supports quality by design (QbD). end, recombinant Escherichia coli Biacore 400 analyzer (from GE The ambr250 workstation is a class clones expressing a therapeutic protein Healthcare). II laminar flow hood that is designed were chosen as the model organism.

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 5 development and optimization of biomanufacturing processes. The implementation of the ambr250 bioreactor system will improve data quality and allow more complex statistically designed experiments in both mammalian and microbial bioprocess development. This could lead to significantly shorter development timelines and lower costs associated with early stage process development and, ultimately, may contribute to quicker technology transfer and a faster time to market of more affordable biologic drugs and vaccines.

Photo 2: The fully automated ambr250 system controls up to 24 = 250-mL bioreactor experiments REFERENCES 1 Biot J, et al. From Orthoclone to RESULTS plant) for both mammalian and Denosumab, the Fast-Growing Market of Monoclonal Antibodies. Med. Sci. 25(12) 2009: Scalability: Optical density and CER microbial cultures, achieving similar 1177–1182. profiles of the E. coli strains cultured results (cell density and titer). This is 2 Cornes P. The Economic Pressures for in the ambr250 bioreactor and 15-L consistent with published data on Biosimilar Drug Use in Cancer Medicine. and 150-L fermentors showed very CHO, demonstrating that automated Target Oncol. 7(supplement 1) 2012: 57–67. similar results (Figures 1 and 2), with mini bioreactors can mimic bench-top 3 Rahul B, et al. High-Throughput 5, 6 Process Development for Biopharmaceutical the OD600 data of the 15-L and 150-L bioreactors ( ) and shows that an vessels lying either side of the automated mini bioreactor is a Drug Substances. Trends Biotechnol. 29(3) 2011: 127–135. ambr250 data, and CER peaks at the comparable model for key parameters 4 Bareither R, et al. Automated same time points. These results such as cell growth and titre in Disposable Small-Scale Reactor for indicate that the ambr250 system bench-top bioreactors. High-Throughput Bioprocess Development: A provides the capability to be a good The ambr250 bioreactory system Proof of Concept Study. Biotechnol. Bioeng. scale-down model for process enables the automated operation of up 110(12) 2013: 3126–3138. development and optimization of to 24 mini bioreactors in parallel, so a 5 Hsu WT, et al. Advanced Microscale microbial cultures. full DoE run can be performed in one Bioreactor System: A Representative Scale-Down Model for Bench-Top Bioreactors. The consistent growth and titre experiment. This means the study of Cytotechnology 64(6) 2012: 667–678. profiles of the CHO clones cultured multiple process parameters is no 6 Lewis G, et al. Novel Automated in the ambr250 and the 3-L longer limited by the availability of Micro-Scale Bioreactor Technology: A bioreactor, showed good comparability bench-top bioreactors, operator time, Qualitative and Quantitative Mimic for Early Process Development. BioProcess J. 9(1) 2010: (Figures 3 and 4) with peak viability, and facility infrastructure. The  cell densities, and maximum titres at application of an ambr250 bioreactor 22–25. the same time points. The data from system could be used instead of the ambr250 system also showed shake-flask and conventional Mwai Ngibuini is manager at Sartorius better consistency than did the bench- bench-top bioreactors models for Stedim Biotech/TAP Biosystems, York Way, top bioreactors. This indicates that process development and optimization. Royston, Hertfordshire, SG8 5WY, UK; process development can be This would save considerable time and 44-1763-227200. reproducibly performed in an resources by reducing manual labour, automated ambr250 bioreactor system laboratory support facilities, and large with mammalian cell cultures. volumes of media. Because this platform is fully DISCUSSION automated, easy to set up, fully Here I outlined a parallel automated disposable, and requires smaller stirred mini bioreactor technology to culture volumes, scientists can allow for rapid process development program and perform high-throughput and optimization for both microbial experiments while achieving highly and mammalian cultures. Results accurate data. This significantly demonstrated that the ambr250 reduces the need to run repetitive bioreactor system can replicate larger experiments and makes this system a scales processes (laboratory- and pilot- cost-effective tool for process

6 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT B IOP ROCESS UPSTREAM/PRODUCTION

Single-Use, Stirred-Tank Bioreactors Efficient Tools for Process Development and Characterization by Andre Grebe, Christel Fenge, Jean-Francois Chaubard

uring the past decade, single- Photo 1: UniVessel SU bioreactor with optical of conventional glass bench-top use bioreactors have become holder and connection box bioreactors to ensure comparability widely accepted as an with previous data generated using alternative to conventional such systems. Each unit is delivered stainlessD steel or glass bioreactors for irradiated and ready to use right out of clinical manufacturing and process the box. It is equipped with development. In the biopharmaceutical noninvasive, single-use pH and industry, glass bioreactors are used dissolved oxygen (DO) sensors to mainly for process development and eliminate the need for manipulation of optimization, but also scale-down the vessel in a laminar-flow bench to models for process characterization. So introduce sensors (Photo 1) before it is of significant importance that such initiating a cell culture experiment. vessels replicate the design of Initially, reliability and robustness production-scale bioreactors for both were sometimes challenging with such reusable and single-use applications. systems — especially for pH Stirred-tank bioreactors with 2-L, 5-L, optochemical patches. But significant and 10-L working volumes have proven Table 1 compares glass and single- improvements of the chemistry and to be particularly well adopted across use options, estimating a 25% increase control of single-use patches now the industry. The 2-L version is the of bioreactor controller run time for ensure reliability and robustness (1). work horse of process development, with the latter. Otherwise, additional glass a volume sufficiently large to serve as a vessels would be necessary to reach EASILY UPGRADING representative small-scale model that equally high use rates; however, they EXISTING CONTROLLERS allows sampling, yet is easy to handle. would increase efforts related to UniVessel SU vessels can be easily change-over activities as well as integrated into both new and existing INCREASING EFFECTIVENESS investment cost. Furthermore, when bioreactor control units, reducing OF DEVELOPMENT STAFF working with microcarriers, single-use investment costs of moving to single- Project timelines and workloads can bioreactors eliminate cumbersome and use bench-top systems. Also, the change dramatically, especially during hazardous siliconization. In essence, integrated single-use sensor signal can process development. It can be a single-use bench top bioreactors be applied to control pH with a challenge to keep enough bench-top simplify the normal daily life of UniVessel SU connection box. It reactors on hand at all times. laboratory staff much like the makes cumbersome and risky Cleaning, setting up, and autoclaving introduction of single-use shaker integration of reusable probes glass vessels requires extra time and flasks did in mammalian cell culture unnecessary. Moreover, the same pH effort on the part of laboratory staff, about 20 years ago. and DO measurement principle can be relegating them to regular used as in larger-scale BIOSTAT maintenance rather than performing EMULATING THE PROVEN STR single-use bioreactors that other, more beneficial tasks. Adding GLASS VESSEL DESIGN contain the same optochemical probes. single-use culture vessels that can be Glass bioreactors have been used for Figures 1 and 2 compare classical and used interchangeably with glass vessels decades and are proven as reliable single–use sensors with simulated pH to a development laboratory provides scale-up and scale-down models of and DO step changes — significant flexibility, especially during stainless steel and state-of-the-art, demonstrating good correlation capacity peaks or maintenance periods. larger-scale, single-use, stirred-tank between the probes. Conventional That reduces downtime of bioreactor bioreactors. The UniVessel SU single- probes can be used in both systems. controllers to an absolute minimum. use culture vessel emulates the design

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 7 EXCELLENT COMPARABILITY Chinese hamster ovary (CHO) cell lines. But when dealing Suspension culture of recombinant CHO cells (typically with complex, posttranslationally modified recombinant under serum-free or protein-free conditions) is widely used proteins and vaccines, special focus on scale-up and for modern large-scale production of monoclonal antibodies comparability is typically necessary. Especially when cells (MAbs) and other therapeutic proteins. A significant body are grown on microcarriers to produce vaccines, careful of knowledge has been produced using conventional bench- consideration must be given to scale-up and comparability scale glass culture vessels. So any new product-development of cell culture systems used (4). or troubleshooting exercise of commercial processes should build on historical data and knowledge, limiting the need to CASE STUDY: A MICROCARRIER-BASED VACCINE PROCESS provide supporting data that demonstrate comparability of A significant number of vaccine processes still rely on cell culture systems in use. adherent cell lines, which require microcarriers as a growth Figure 3 shows excellent comparability of the most support in stirred-tank bioreactors. This adds complexity to important process engineering parameter — kla and scale-up and scale-down of vaccine processes from the mixing time — of classical glass and UniVessel SU increased shear sensitivity of cells grown on carriers, the bioreactors at typical impeller tip speeds used in larger size of the carriers compared with single suspension mammalian cell culture (3). In modern MAb production, cells, and the higher density of carriers. Very often that most companies use platform technologies and robust requires special low-shear impeller designs and careful consideration of the arrangement of probes, dip tubes, and Table 1: Comparing single-use and glass bench-top bioreactors other inserts to prevent formation of dead zones (3). Glass UniVessel SU Therefore, bioreactor designs and volumes typically used Process Step Bioreactor Bioreactor for suspension culture might be unsuitable as scale-up or Pre- and postrun preparation time 2 days ~1 hour -down model systems for microcarrier-based processes. (vessel assembly, sensor calibration, autoclaving, medium fill, harvest, Within the cell and viral technologies department of cleaning) GlaxoSmithKline Vaccines, special attention is paid to Sterility test 1 day None such challenges. Culture time 12 days 12 days Maintaining comparable process conditions to Possible runs per year 24 runs 30 (+ 25%) production scale with regard to shear, homogeneity, and per bioreactor controller microenvironment for microcarrier-based processes will require particular working volumes and appropriate Figure 1: Comparing classical and single-use pH measurement in Britton-Robinson buffer solution at 37 °C in the UniVessel SU bioreactor Figure 3: Comparing process-engineering parameters of conventional

glass and UniVessel SU bioreactors — mixing time (dashed lines) and kLa 7.5 (solid lines) 7.4 Optical pH Sensor 60 60 7.3 Classical pH Sensor Glass 2-L 7.2 50 Glass 5-L 50 Glass 10-L 7.1 UniVessel SU 40 40 k L

7.0 a pH 6.9 30 30 (1/h) 6.8 Reference pH Meter (i) 20 20 6.7 6.6 10 10 6.5 0 20 40 60 80 100 120 140 160 (seconds) Time Mixing 0 0 Time (hours) 0.0 0.6 1.2 1.8 2.4 Impeller Tip Speed (m/s)

Figure 2: Comparing classical and single-use dissolved oxygen (DO) Figure 4: Growth characteristics of an adherent cell line grown on measurement in reverse-osmosis (RO) water at 37 °C in the UniVessel SU microcarriers in different Sartorius single-use stirred-tank bioreactors bioreactor 120 (UniVessel SU 2L, BIOSTAT STR 50L, BIOSTAT STR 200L) and the 10-L stirred- Optical DO Sensor tank glass bioreactor typically used as a scale-up/scale-down model Classical DO Sensor 100 3.5 3.0 80 UniVessel 2L 2.5 10-L Glass Vessel

cells/mL) Cultibag STR 50L 60 6 2.0 Cultibag STR 200L 40 1.5 1.0 Air Saturation (%) 20 0.5

0 DensityCell (10 0 5 10 15 20 25 0.0 J1 J2 J3 J4 J5 J6 Time (hours) Duration (days)

8 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Photo 2: UniVessel SU vessel connected to KEY BENEFITS Figure 5: Contour plot of a multivariate existing bioreactor controllers at GSK Vaccines experimental design to optimize the viral (Rixensart, Begium) Immediately available, helps to manage production phase; the response is measured peak manufacturing demands using an ELISA test (ELISA) and shown as a function of infection time (A: growth time) and Completely single-use, increases multiplicity of infection (B: MOI). effectiveness of laboratory Same design as existing glass vessels, full comparability and scalability Compatible with existing bioreactor controllers, limited additional

investment ELISA cells on microcarriers in different (B) MOI single-use bioreactor systems — 2-L (A) GrowthTime designs. In addition, it is often UniVessel SU, 50-L and 200-L necessary to purify a product so its BIOSTAT STR — with the 10-L quality and activity can be determined glass vessel. Comparable cell growth adherent-cell culture process for viral and process performance can be was obtained in all evaluated stirred- vaccine production. We could assessed. Because purification steps tank bioreactor vessels. Comparable demonstrate excellent comparability of are primarily based on product metabolite profiles also were obtained cell growth and metabolite profiles for amount, a certain volume is required (data not shown), proving the the single-use 2-L vessel, a 10-L to achieve required amounts for suitability of the UniVessel SU 2-L stirred-tank glass vessel historically further purification and product system for scale-up and -down studies. used as a scale-up/-down model, and characterization. In this case study, Based on comparable process larger-scale single-use (BIOSTAT the purification team required a performance and the UniVessel SU STR 50L and 200L) and stainless- volume of at least 1 L to perform system’s ease of use, GSK installed a steel stirred-tank bioreactors. Based on purification experiments. new process-development platform these data, GSK established a new Taking into account the above- consisting of 12 parallel 2-L single- process development platform mentioned considerations, the team use, stirred-tank bioreactors (Photo 2). consisting of 12 multiparallel 2-L decided to use bench-top bioreactors of This set-up is primarily used for single-use vessels to support 2–L scale to optimize process multivariate design of experiment multivariate experimental designs for parameters. We expected this scale to (DoE) studies, for example that efficient and fast optimization of emulate larger-scale, stirred-tank illustrated in Figure 5. It was possible vaccine production processes. bioreactor conditions. It would supply to optimize viral production by the required volume for purification evaluating five different process REFERENCES activities, while still allowing for easy parameters performing only three runs 1 Weichert et al. Integrated Optical and straightforward parallel processing. of 12× 2-L single-use vessels Single-Use Sensors. BioProcess Int. 12(8) 2014: S20–S25. To simplify the experimental set-up demonstrating the effectiveness of such 2 De Wilde D, et al. Superior Scalability and increase the throughput of runs a parallel set-up of single-use vessels. of Single-Use Bioreactors. BioProcess Int. 12(8) per bioreactor controller, the GSK Thanks to this new single-use 2014: S14–S19. team decided to look for a single-use process development platform, the team 3 Fenge C, Lüllau E. Cell Culture solution. The team took into significantly increased the number of Bioreactors. Cell Culture Technology for consideration the single-use UniVessel experiments that could be run without Pharmaceutical and Cell Based Therapies. Ozturk SU from Sartorius Stedim Biotech increasing laboratory personnel. The S, Hu WS, Eds. Taylor and Francis Group: Oxford, UK, 2006; 155–224. because of its design flexibility and approach especially helped to further 4 Chaubard J-F, et al. Disposable geometrical similarity to classical accelerate development timelines and Bioreactors for Viral Vaccine Production: stirred-tank bioreactors (2). will therefore be extended to other Challenges and Opportunities. BioPharm Int. To demonstrate the suitability of development projects. 2 November 2010 (supplement).  the 2-L scale UniVessel SU bioreactor for microcarrier-based processes, the CONCLUSIONS Andre Grebe is head of product team compared its process We have shown excellent management for multi-use bioreactors, and performance with that of 10-L stirred- comparability of the single-use 2-L Christel Fenge is vice president of tank glass bioreactors as a benchmark. stirred-tank vessel (UniVessel SU 2L) marketing fermentation technologies at Those 10-L glass bioreactors have with proven 2-L, 5-L, and 10-L Sartorius Stedim Biotech GmbH, August- proven to be a representative scale- stirred-tank glass vessels regarding kla Spindler-Straβe 11, 37079 Göttingen, down model of the GSK larger-scale and mixing time. The team also Germany. Jean-Francois Chaubard is stainless-steel or single-use evaluated biological comparability director of cell and viral technologies at (BIOSTAT STR) bioreactors. Figure under the most demanding process GlaxoSmithKline Vaccines, Rue de I’Institute 4 compares the growth of adherent conditions using a microcarrier-based 89, 1330 Rixensart, Belgium.

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 9 B IOP ROCESS PROCESS DEVELOPMENT

Design of Experiments with Small-Scale Bioreactor Systems Efficient Bioprocess Development and Optimization by Andree Ellert and Conny Vikström

esign of experiments (DoE) is one of the most valuable techniques for organized and efficient planning, execution, andD statistical evaluation of experiments. Although a DoE investigation can be completed using several runs in one bioreactor, small- scale bioreactor systems designed for parallel operation (such as the ambr15 or ambr250 systems) provide the optimal basis to economically realize a series of experiments. Because of the multitude of interdependent parameters involved in applications such as cell line development, culture media screening, and the optimization of bioreactor operating parameters, Photo 1: The ambr250 scale-down bioreactor system controls 12 or 24 disposable small-scale DoE evolved as an essential and bioreactors (100–250 mL working volume) and offers parallel processing and evaluation of multiple indispensable method to create process experiments while maintaining the characteristics of a larger scale bioreactor. knowledge. It has improved speed of development and has helped define Spearheaded by the FDA initiative to understanding of processes. By contrast manufacturing processes for high- use PAT for greater control and to changing one factor at a time, the quality products. process understanding, statistical DoE DoE procedure delivers optimized methods are extensively applied to information content by using the least DESIGN OF EXPERIMENTS: look for the best process conditions number of experiments, thereby THE EFFICIENT STRATEGY while reducing expensive and time- reducing development time and labour. In pharmaceutical bioprocessing, consuming experiments to a minimum The improvement of production decreased development time and (1). and feed media components is crucial production costs are key objectives. The concept of DoE is to vary to providing an environment optimal Within this context, the combined process parameters simultaneously over for growth of a used recombinant cell application of process analytical a set of planned experiments and then line and formation of a functionally technology (PAT) tools and reliable interpret the results by means of a active therapeutic protein (2). Another flexible bioprocess equipment is the proven mathematical model. This key target is the optimization of basic basis for an efficient optimization of model can be used subsequently for process parameters such as existing production processes and the interpretation, prediction, and temperature, pH, and dissolved development of new ones. optimization, which allows for greater oxygen (3). Often, optimization of

10 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Photo 2: The BIOSTAT B-DCU II system is designed for advanced process optimization and characterization and is available with working volumes of 0.5–10 L; it features independent process control for up to six culture vessels culture media and state variables are medium development is to use high- The automated workstation combined, whereas DoE studies can throughput cell culture scale-down features an integrated liquid handler, significantly facilitate such tasks (1). systems in combination with statistical which can remove the cap of a single- DoE approaches. use vessel to enable initial media load, TRANSITION FROM MEDIUM In addition to improving the seeding, sampling, and bolus AND FEED DEVELOPMENT medium composition, optimization of additions. Culture or feed medium can TO PROCESS OPTIMIZATION the feed-control strategy can be be transferred preformulated by the Apart from the productivity of a beneficial for fed-batch processes. The pipetting module, or the system can specific cell line, the culture medium practical approach depends on the be used to automatically make up has an important impact on both yield infrastructure of the bioreactor system media from different components and quantity of therapeutic protein itself and the cultivated cell line. using imported experimental designs. and monoclonal antibody produced. Whereas a classical fed-batch with Besides single bolus additions (10 μL The cells require a combination of continuous feeding strategies has been to 10 mL), integrated pumps allow macro- and micronutrients as sugars, proven to be successful for Escherichia continuous feeds of linear or trace salts, vitamins, amino acids, and coli and Pichia pastoris cultivations, the exponential profile ranging from other components to support cell method of adding a stepwise bolus of 20 nL/h to 20 mL/h. growth and protein production. feed solution to a production Once the most prominent medium Because of the diverse nutritional bioreactor is most widely used in components and their optimal ranges requirements that are unique to every industry for mammalian cell culture are identified, the process can be cell and the large number of medium because of its simplicity and scalability transferred to a larger laboratory-scale factors, small-scale systems — with (5, 6). system for experimental verification of their ability to perform a large number The new ambr250 workstation the identified optimum to test of experiments in parallel — are with 12 or 24 single-use stirred tank scalability and to continue with needed to unlock development reactors (TAP Biosystems, a Sartorius optimizing the bioreactor’s operating bottlenecks. company) represents an efficient, parameters as well as process Historically, media and feed high-throughput, scale-down model characterization studies. formulations are derived through for initial process development of The BIOSTAT B-DCU II numerous empirical tests by changing microbial fermentation or cell culture. benchtop bioreactor (Sartorius Stedim one factor at a time (4). This The system provides the capability to Biotech) is specifically designed for methodology is simple and convenient, continuously monitor and control laboratory-scale parallel operations of but it ignores interactions between critical parameters in real-time as well up to six multiuse or single-use culture components, can miss the optimum as extended feed and sampling options vessels with one control tower. It is completely, and is fairly time- to enable a more effective scale-up and used widely in the industry for process consuming. A superior approach to to demonstrate equivalent process optimization and characterization (10). tackle the problem of media design performance in comparison to Each vessel has independent process and shorten the time needed for laboratory and pilot scale (7–9). controls with a wide range of

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 11 Figure 1: Comparison of two protein production phases with different factor settings 1 and 2; Sk = fluorescence signal of internal soluble (k = sol) and insoluble (k = IB) protein fraction; θ = cultivation For advanced process monitoring temperature; μ = online observed cell specific growth rate and control, the BIOSTAT Q plus

θ μ system was equipped with O2 and Sk Induction (103 RFU) (˚C) (h−1) CO off-gas analysers. Values from 2 400 40 0.60 μ = 0.18 h−1 θ = 30 ˚C w_1 w_1 those sensors and with the BioPAT μ = 0.21 h−1 θ = 28 ˚C w_2 w_2 SIB_2 MFCS/win bioprocess software 320 35 048 allowed for online calculation of cell-specific growth rates as well as data storage and supervisory control. 240 30 0.36 θ 1 The DoE software BioPAT MODDE θ was used for experiment definition, 2 μ 2 S 160 25 0.24 IB_1 statistical evaluation of raw data, and μ construction of an easy to interpret 1 Ssol_2 model. 80 20 0.12

Ssol_1 THE WORKFLOW TO SUCCESS 0 15 0.00 An experimental design was created to 0 1 2 3 4 5 6 screen for variables that would have t (h) ind the highest effect on the space–time yield. Given the known dependence of LEFT RIGHT Figure 2: Response surface plots for soluble ( ) and insoluble ( ) space-time yield; STYk = growth rate on temperature, the DoE soluble (k = sol) and insoluble (k = IB) space-time yield; θ = cultivation temperature; μ = cell-specific software was used to design a set of growth rate experiments with varying levels of each factor. The final experimental 17,500 60,000 plan included four runs on a centred 14,000 ) 48,000 level to determine the variability

−1 ) within the overall system. −1 10,500 36,000 Using the BIOSTAT Q plus

7,000 (RFU h

IB 24,000 system, it was possible to screen each

(RFU h (RFU sol STY E. coli strain in as few as 12 runs to

3,500 12,000 STY 31 30 29 28 27 26 31 30 29 28 27 26 identify the most promising strain for further studies. In addition, the θ (˚C)μ (h−1) θ (˚C) μ (h−1) inducer concentration was shown to 0.16 0.20 0.24 0.16 0.20 0.24 have no significant effect on product yield, so that the lowest investigated concentration could be used for measurement and automation features. ingredients (APIs) or parts thereof further studies. Once the initial Systems are preconfigured for (e.g., antibody-drug conjugates, screening procedure was completed, immediate use, and they can be used ADCs). Testing novel APIs requires higher levels for growth rate were for microbial fermentation or cell the physiologically active form of the tested in combination with lower culture packages. protein for preclinical toxicology temperature set-points. studies. Although protein solubility Figure 1 shows the optimization OPTIMIZATION OF RECOMBINANT does not necessarily correspond to potential of the DoE approach while PROTEIN EXPRESSION: A CASE STUDY active protein, fast product availability simultaneously varying factor Within process development divisions, often is limited by low soluble protein conditions. Using the lowest inducer the goal is to identify key parameters yields. concentration, different set-points for that maximize target yield and product Growth rate, cultivation growth rate (μw) and cultivation quality within process parameter ranges temperature, and IPTG inducer temperature (ew) resulted in a higher that can be reliably attained at pilot and concentration were investigated for soluble and insoluble protein yield for production scale. In the following case each E. coli strain. The effect of each the high/low factor level combination study, a BIOSTAT Q plus six-fold factor and strain on the space–time subscripted with 2. system was used in conjunction with yield of soluble protein and inclusion Through this set of optimization DoE to optimize recombinant protein bodies was used for process evaluation. experiments and continued expression in two E. coli BL21 (DE3) Expressed protein was tagged with a interpretation within the DoE strains. green fluorescent protein, allowing for software, the process could be further E. coli is one of the most used simple and rapid quantification of understood finally leading to a prokaryotic organisms for the protein expression with a common substantial increase of soluble protein production of active pharmaceutical fluorescence reader. concentration. Highly predictive

12 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT models gave a reliable direction to REFERENCES obtain high space–time yields at low 1 Mandenius CF, Brundin A. Bioprocess temperature in combination with high Optimization Using Design-of-Experiments Methodology. Biotechnol. Prog. 24, 2008: 1191– growth rate. Figure 2 displays the 1203. predicted space–time yields as a 2 Dong J, et al. Evaluation and response surface spanned by the two Optimization of Hepatocyte Culture Media factors. Factors By Design of Experiments (DoE) The final test in this experiment Methodology. Cytotechnol. 57, 2008: 251–261. was a robustness trial, testing 3 Fricke J, et al. Designing a Fully parameters for the maximum Automated Multibioreactor Plant for Fast DoE Optimization of Pharmaceutical Protein allowable process parameter ranges Production. Biotechnol. J. 8, 2013: 738–747. without influencing product quality. 4 Weuster-Botz D. Experimental Design Using only six experimental for Fermentation Media Development: conditions, a safe operating range Statistical Design or Global Random Search? J. could be confirmed in which the Biosci. Bioeng. 90(5) 2000: 473–483. desired space–time yields were 5 Shiloach J, Fass R. Growing E. coli to achieved. The design space tools High Cell Density: A Historical Perspective on Method Development. Biotechnol Adv. 23(5) provided by the BioPAT MODDE 2005: 345–357. software can be used to visualize an 6 Li F, et al. Cell Culture Processes for operating range for the investigated Monoclonal Antibody Production. mAbs 2(5) process parameters in a unique 2010: 466–479. probability contour plot considering 7 Hsu WT, et al. Advanced Microscale risk analysis specifications. That Bioreactor System: A Representative Scale- guides engineers in determining how Down Model for Benchtop Bioreactors. Cytotechnol. 64, 2012: 667–678. likely it is that their experiments will 8 Bareither R, et al. Automated truly identify the most reliable Disposable Small-Scale Reactor for High operating parameter ranges. Throughput Bioprocess Development: A Proof of Concept Study. Biotechnol Bioeng. 110(12) CONCLUSIONS 2013: 3126–3138. High-throughput cell-culture 9 Ngibuini M. Automated Mini scale-down bioreactor systems such as Bioreactor Technology for Optimizing Early Process Development in Fermentation and the ambr 15 and the ambr 250 systems Bioprocessing. BioProcess Int. 12(8) 2014: S3– can increase efficiency and S6. effectiveness of screening, process 10 Tsang VL, et al. Development of a Scale development and process Down Cell Culture Model Using Multivariate Analysis as Qualification Tool. Biotechnol. Prog. characterization exercises. They offer  the advantage that a significant 30(1) 2013: 152–160. number of experiments can be performed simultaneously at small Corresponding author Andree Ellert is culture volumes. At the same time, product manager, bioprocess software at the culture environment and Sartorius Stedim Biotech GmbH (andree. bioreactor characteristics of small-scale [email protected]), and Conny single-use stirred tank vessels provide Vikström is senior application specialist, excellent comparability to larger scale product manager MODDE at Umetrics AB, bioreactors (7–9). Leaving the 46 90-184849; conny.vikstrom@umetrics. traditional way of trial-and-error com. optimization behind, advanced DoE approaches enable fast and effective identification of critical process parameters and provide significant cost savings and reduction of development timelines.

ACKNOWLEDGMENTS Experimental work has been performed at the Research Center of Bioprocess Engineering and Analytical Techniques from the Hamburg University of Applied Sciences, headed by Prof. Dr-Ing. R. Luttmann.

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 13 B IOP ROCESS SCALE-UP

Superior Scalability of Single-Use Bioreactors by Davy De Wilde, Thomas Dreher, Christian Zahnow, Ute Husemann, Gerhard Greller, Thorsten Adams, and Christel Fenge

uring the past several years, Scale-up is an important and production processes (6). Typical single-use bioreactors have potentially time-consuming step in the parameters of concern are oxygen been gradually established in development of industrial processes. It transfer, mixing, and heat-transfer modern biopharmaceutical involves much more than just doing characteristics as well as the generated processesD (1, 2). This adoption is the same at a larger volume. It requires shear forces. directly linked to their unique ability the generation of solid process During the past 30 years, to enhance flexibility and reduce understanding at different scales to stainless-steel stirred-tank bioreactors investment and operational costs. ensure consistent quality and titer have evolved as the gold standard, Furthermore, production output can throughout scale-up from early clinical especially as a result of their be increased, and time to market is trials to final production scale (6). straightforward scale-up. Multiple shortened (3). Today a wide variety of Today, many companies use times, their well-understood design single-use bioreactors exists for the chemometric tools such as design of principles have proven successful in cultivation of mammalian and insect experiments and multivariate data development and scale-up to safe and cells (4), whereas only limited solutions analysis to establish critical process robust commercial processes. are available for microbial cultures (5). parameter ranges that define the Furthermore, they enable users to Typically, processes are established design space of a robust production implement their existing knowledge and optimized in stirred-tank process. Especially during late-phase — especially with platform processes benchtop bioreactor systems. One development of a commercial process, — into production processes of new challenge during the development of a the availability of a properly drugs and to set-up experiments in a robust cell culture process is the representative scale-down model of way that can shorten development straightforward scale-up to final full production scale is essential to timelines. production scale. This is especially allow efficient process development (7). However, many commercially critical when using less-characterized Detailed understanding of available single-use bioreactors differ bioreactor designs that deviate from bioreactor characteristics at different from this gold standard. Vessel design, the well-known and understood scales significantly facilitates the stirrer design, and gassing strategy classical stirred-tank principle. development and scale-up of robust especially may differ from the classical

Figure 1: ambr250, UniVessel SU, and BIOSTAT STR family; working volume ranges from 250 mL to 2000 L

14 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Table 1: Summary of geometrical dimensions of the ambr250, UniVessel SU, and BIOSTAT STR family AMBR UniVessel SU BIOSTAT STR 0.25 2 50 200 500 1000 2000 Total volume (L) 0.36 3 68 280 700 1,300 2,800 Maximum working volume (L) 0.25 2 50 200 500 1,000 2,000 Minimum Working volume (L) 0.06 1 12.5 50 125 250 500 Vessel diameter D (min) 62.5 130 370 585 815 997 1,295 Vessel height H (mm) 126 240 666 1,055 1,467 1,800 2,330 Ratio H/D 2 1.8 1.8 1.8 1.8 1.8 1.8 Liquid height (mm) 90 177 480 783 1,005 1,360 1,670 h1 Ratio / 1.44 1.36 1.3 1.34 1.23 1.36 1.29 h1 D Impeller diameter (mm) 26 54 143 225 310 379 492 d2 Ratio / 0.42 0.42 0.38 0.38 0.38 0.38 0.38 d2 D Distance between impellers Δz (mm) 30 70 186 300 403 493 640 Size holes ring sparger part (mm) 2 0.5 0.8 0.8 0.8 0.8 0.8 Number holes ring sparger part 1 14 5 25 100 100 200 Size holes micro sparger part (μm) NA NA 150 150 150 150 150 Number holes micro sparger part NA NA 25 100 500 500 1,000

Table 2: Comparison of process engineering parameters suitable for scale-up from the BIOSTAT simple and straightforward scale-up of STR 50 to the BIOSTAT STR 2000; for scale-up, a CHO process performed at 50 L scale was assumed, a process (10). This is especially which was performed at 150 rpm equivalent to a tip speed of 1.1 m/s, a commonly used tip speed for cell culture applications critical as design changes across scales might influence mixing behavior, 3 Process Engineering Parameter N (rpm) Tip Speed (m/s) Re P/VL (W/m ) BIOSTAT STR 50 150 1.1 49,420 22.3 oxygen transfer, bubble dispersion, and various other key parameters. On Equal N for BIOSTAT STR 2000 150 3.9 605,406 293.1 Equal tip speed for BIOSTAT STR 2000 42 1.1 171,936 6.7 the other hand, a homogenous culture Equal Re for BIOSTAT STR 2000 12 0.31 49,420 0.15 environment across scales — where important cultivation parameters such Equal P/V for BIOSTAT STR 2000 63 1.6 254,270 22.3 L as pH, oxygen partial pressure, temperature, and nutrient supply are Figure 2: Combisparger for the BIOSTAT STR scale and beyond to facilitate process well controlled — is a key prerequisite 2000L system transfers to existing legacy production to establish a robust and safe facilities. production process. To characterize Ring Sparger Here we detail the different design bioreactor performance across scales aspects of those bioreactors and their and govern scale-up, an appropriate Micro sparger impact on critical process engineering criterion should be defined and kept parameters such as power input per constant during scale-up. In general, volume, mixing time, and volumetric the power input per volume is used as mass transfer (9). We compare the scale-up criterion (10). Also, the tip characteristics of the large scale speed or other shear-related single-use bioreactor family parameters are often used, especially BIOSTAT STR to small-scale single- when using shear-sensitive cells (11) or stirred-tank design principles and do use bioreactor vessels such as the when growing cells on microcarriers. not necessarily offer consistency and ambr250 (250 mL) and the single-use Based on the very well- geometrical similarity across scales (8). UniVessel SU 2L, which are typically characterized, reusable, stirred-tank So the scale-up exercise might be used during process development, bioreactors, it is possible to assign complicated, and additional risk might optimization, and characterization. relevant design criteria to single-use be added to the process transfer. bioreactors for animal and microbial To offer a solution for that, DESIGN PRINCIPLES AND cells. One such criterion is the height- Sartorius Stedim Biotech has SUITABLE SCALE-UP CRITERIA to-vessel-diameter ratio (H/D or developed a range of single-use OF STIRRED-TANK BIOREACTORS aspect ratio), which should be kept bioreactors from 250 mL to 2,000 L Geometrical similarity of vessel design within a range of 1:1 to 3:1 for working volume. Its designs are (amongst others defined by the height- stirred-tank reactors (10). A low value entirely based on proven stirred-tank to-diameter ratio and the for an H/D ratio results in an bioreactor principles. This ensures impeller-to-vessel-diameter ratio) is increased ratio of headspace surface to straightforward scale-up to 2,000 L commonly considered important for filling volume, which enables an

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 15 Figure 3: Stirrer speed as a function of the tip speed for the UniVessel SU and BIOSTAT STR reactors. type disk impellers to achieve effective homogenization. Also, because of the lower dissipative energy transferred, this impeller type 600 is more adequate for shear-sensitive UniVessel SU cell lines. In comparison, disk 500 impellers generate a radial flow, BIOSTAT STR 50 leading to higher power input per 400 BIOSTAT STR 200 volume at a given stirrer speed and enhanced gas-bubble dispersion (11). BIOSTAT STR 500 300 It is common to install multiple impellers in bioreactors with an H/D BIOSTAT STR 1000 ratio above 1:1.4 to ensure efficient Stirrer Speed (rpm) Stirrer 200 BIOSTAT STR 2000 mixing throughout the entire cultivation chamber. 100 To achieve suitable power input, the distance between the impellers is 0 important. It is recommended to use 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 a distance that is 1.2–1.5times Tip Speed (m/s) greater than the impeller diameter to guarantee that the impellers act independently of each other. Historically cell culture processes Figure 4: Power input per volume (P/VL) for the UniVessel® SU and BIOSTAT® STR family often used two three-blade segment 400 impellers, but today many companies are using a combination of a 350 Rushton-type impeller and a three-blade segment impeller. The UniVessel SU 300 disk impeller (installed right above BIOSTAT STR 50 the sparger) ensures good dispersion 250 of the gas bubbles. A three-blade )

3 BIOSTAT STR 200 segment impeller serves as a superior 200 axial mixer and ensures homogeneity BIOSTAT STR 500 (W/m

L in the entire vessel, thus supporting 150 high–cell-density processes (17, 18). P/V BIOSTAT STR 1000 That approach has been facilitated by 100 BIOSTAT STR 2000 modern, more robust recombinant 50 cell lines to grow at higher shear rates, which have been selected for 0 commercial manufacturing at large 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 scale (19). Tip Speed (m/s) Another determining factor for a successful cell culture process is the amount of gas transfer. In conventional bioreactor designs, a improved gas exchange at the gas– typically be between 0.33 and 0.5 for sparger is installed directly below the liquid interface. On the other hand, a animal cells (14) and influences mixing lower impeller, which ensures proper larger aspect ratio offers advantages in efficiency and the generated shear gas-bubble dispersion (14). Spargers case of direct sparging due to the forces. Three-blade-segment impellers with small holes generate small longer residence time of gas bubbles in or marine-type impellers are bubbles and improve oxygen transfer the liquid and hence a higher oxygen- commonly used for animal cell because of their high gas–liquid transfer rate (12). For animal cell cultures (15). They efficiently interface areas. Reusable bioreactors cultivations, often a ratio of 2:1 is transform the transferred energy into use sintered stainless steel recommended (13). hydrodynamic power and generate microspargers. But this design has Another parameter to consider is large circulation loops because of their the disadvantage of nondefined pores, the ratio of impeller diameter to vessel axial flow patterns (16). Therefore, leading to coalescence of bubbles. diameter. This parameter should they are often preferred over Rushton- Therefore, Sartorius Stedim Biotech

16 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 5: Mixing times for the different scales of the UniVessel SU and BIOSTAT STR family as a function of the tip speed for 2 x 3-blade-segment impeller configuration fits perfectly to the gold standard 70 derived from reusable bioreactors for mammalian cell culture. Table 1 provides a detailed description of the 60 geometrical dimensions of those closely linked single-use bioreactor families. 50 The impellers are installed on a UniVessel SU rigid, central shaft. For agitation, 40 BIOSTAT STR 50 2 =3-blade-segment impellers are

BIOSTAT STR 200 available. For the BIOSTAT STR, a 30 combination of a six-blade-disk BIOSTAT STR 500 (bottom) and three-blade-segment 20 BIOSTAT STR 1000 (top) impeller can be installed as an Mixing Time Mixing Time (s) BIOSTAT STR 2000 alternative. 10 The study presented here focuses on process engineering 0 characterization of a configuration 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 based on 2 = 3-blade-segment Tip Speed (m/s) impellers for different single-use bioreactor volumes. Gas transfer has been characterized for a microsparger has developed a special microsparger For efficient, cost-conscious process (hole diameter = 150 μm) or a design with 150 μm holes that development, highly automated, single- ringsparger (hole diameter = 0.5 mm provides a uniform bubble swarm of use multiparallel bioreactors are for the Univessel SU or 0.8 mm for the small bubbles for effective gas available at 250 mL scale (ambr250) BIOSTAT STR) positioned below the transfer. Spargers with large holes (20). This high-throughput process lower impeller. The BIOSTAT STR is have a relatively low oxygen transfer development bioreactor system allows available with a combisparger (Figure but offer improved performance for fast and effective establishment of 2) — consisting of both a microsparger

CO2 stripping because bigger bubbles optimal process conditions early in (0.15 mm holes) and a ringsparger (0.8 typically rise to the gas–liquid process development. With the mm holes). The microsparger supports interface and carry excessive CO2 classical stirred-tank design, high oxygen transfer, and the from the cell suspension to the straightforward scale-up is possible ringsparger enhances CO2 stripping. headspace. either directly for production of All single-use bioreactors from 250

At small-scale, CO2 stripping is material for toxicological studies or mL scale to 2,000 L are equipped less of a challenge. It is more critical through step-wise scale-up through with precalibrated, single-use at larger volumes because of the 2 L scale using the UniVessel SU optochemical probes for pH and pO2 higher hydrostatic pressure and thus technology and 50 L scale using the measurement. Alternatively, improved solubility of CO2. Together BIOSTAT STR. With the UniVessel conventional probes can be introduced with excessive foaming, that limits SU 2 L model, conventional glass if desired for scales ≥2 L. All single- the efficiency of conventional 10 to vessels can be replaced easily even with use bioreactors are available with 20 μm microspargers for large already existing bioreactor controllers standard digital control units. bioreactor volumes. in development laboratories. The UniVessel SU design is similar to its PROCESS ENGINEERING SARTORIUS STEDIM BIOTECH glass counterpart and the large-scale CHARACTERIZATION SINGLE-USE BIOREACTORS EMULATE single-use BIOSTAT STR design, We performed in-house process CLASSICAL STIRRED-TANK DESIGN thereby enabling straightforward scale- engineering characterization of the Most large-scale, single-use up all the way from development to UniVessel SU and BIOSTAT STR bioreactors do not rely on established commercial production and offering a bioreactors. For the ambr250 system, design criteria of reusable bioreactors, single-use scale-down model for we used data published by Bareither which can add risk to scaling-up process characterization. et al. (20). The process engineering processes. To overcome this, Sartorius The BIOSTAT STR has a characterization of the UniVessel SU Stedim Biotech offers a range of cylindrical shape with an H/D ratio of and BIOSTAT STR family was stirred single-use bioreactors (Figure 2:1 and a semi-torispherical bottom performed at parameters typical for 1) based on classical, well-proven and top. The impeller-to-bag- mammalian cell culture (tip speeds design principles. Different scales diameter-ratio is 0.38 with a distance between 0.6–1.8 m/s). For the exist, allowing to work from 250 mL between both impellers of 1.3= the characterization of the ambr250 to 2,000 L culture volumes. impeller diameter. So the vessel design Bareither et al. used tip speeds

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 17 Figure 6: Characteristics of the volumetric mass transfer for the different scales of the UniVessel SU A B and BIOSTAT STR family using 2 = 3-blade-segment impellers; ( ) results of the ringsparger and ( ) speed. To maintain a constant P/VL results of the microsparger value during scale-up, the tip speed will need to increase with increasing 30 A scale. That clearly demonstrates that users have to choose their scale-up criterion because it is not possible to keep both the tip speed and the power 20 input per volume constant when increasing the bioreactor scale. We determined mixing times for UniVessel SU ring (0.5 mm) (1/h) the BIOSTAT STR and UniVessel SU a L BIOSTAT STR 50 ring (0.8 mm) k bioreactors using the decolorization 10 BIOSTAT STR 200 ring (0.8 mm) method (14) to characterize the mixing BIOSTAT STR 500 ring (0.8 mm) capabilities of the single-use bioreactor BIOSTAT STR 1000 ring (0.8 mm) systems. It is obvious that the mixing BIOSTAT STR 2000 ring (0.8 mm) time decreases with increasing tip speed because of the higher power 0 input per volume. In addition, we 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 obtained higher mixing times with Tip Speed (m/s) increasing scales. Figure 5 shows the mixing B efficiency for the different single-use 50 bioreactor scales. At 2 L scale, a mixing time of 7 s at a tip speed of 40 0.6 m/s and 2 s at a tip speed of 1.8 m/s can be achieved. At 50 L scale, mixing times increase to 19 s at 30 BIOSTAT STR 50 micro (0.15 mm) tip speed of 0.6 m/s and 8 s at a tip BIOSTAT STR 200 micro (0.15 mm) speed of 1.8 m/s. For 2,000 L scale, a (1/h) L BIOSTAT STR 500 micro (0.15 mm) k mixing times of 20 s can be obtained. 20 BIOSTAT STR 1000 micro (0.15 mm) Overall, it is possible to ensure mixing BIOSTAT STR 2000 micro (0.15 mm) times <30 s for all scales, from 2 L to 10 2,000 L, thus demonstrating the superior performance of the UniVessel SU and BIOSTAT STR family for 0 mammalian cell culture (9). 0.5 0.7 0.9 1.1 1.3 1.5 1.7 1.9 For the quantification of the oxygen-transfer rate, we determined Tip Speed (m/s) the volumetric mass transfer coefficient 21 (kLa) using the gassing out method ( ). 8 ranging from 0.27 m/s to 1.02 m/s, measurements ( ). For 2 = 3-blade For most common cell cultures, kLa which resulted in corresponding stirrer segment impellers, we calculated a Ne values of 5–10 h–1 are required. Figure 20 speeds from 200 to 800 rpm ( ). of ~1.3. The fact that the Reynolds 6 shows the kLa characteristics for Figure 3 graphs stirrer speed as a number (Re) is above 10,000 at the different tip speeds and scales of the function of tip speed for the UniVessel chosen tip-speed range implies that BIOSTAT STR and UniVessel SU SU and BIOSTAT STR systems. The turbulent flow conditions are present, bioreactors. Both ring and microsparger increasing impeller diameters requires thus the Ne value is constant. elements were used for the BIOSTAT lower stirrer speeds at increasing scale The power input per volume (P/VL) STR system and a ringsparger for the to maintain the same tip speed. is an important process engineering UniVessel SU and ambr studies.

During in-house trials for the parameter and can be calculated based The kLa-values increase with BIOSTAT STR and UniVessel SU on the experimentally determined increasing tip speed at any given scale bioreactors, the Newton number (Ne) Newton number. Figure 4 shows the and configuration. As expected, the was determined to characterize the power input per volume for the kLa values increase also with the scale. different impeller types and BIOSTAT STR and the UniVessel That can be explained by the increased configurations and to quantify the SU bioreactors. For the ambr system, liquid height and the longer residence power input per volume. The Newton Bareither et al. (20) reports a power time of the gas bubbles in the liquid — number was determined by torque input of 10–445 W/m³ for the tip resulting in a more efficient oxygen

18 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT transfer (22). The ambr system had k a REFERENCES Challenges in Extremely High Cell Densities. L Single-Use Forum, Istanbul, Turkey, 5 July –1 1 Eibl R, Werner S, Eibl D. Bag values of 8.5 h at a tip speed of 2013. 20 Bioreactor based on Wave-Induced Motion: 1.02 m/s ( ). For the other bioreactor 18 Zijlstra G, et al. High Cell Density XD –1 Characteristics and Applications. Adv. Biochem families kLa values >10 h can be easily Engin. Biotechnol. 115, 2009: 55–87. Cultivation of CHO Cells in the BIOSTAT CultiBag STR 50 L Single-Use Bioreactor with achieved at all scales for both 2 Brecht R. Disposable Bioreactors: Novel Microsparger and Single-Use Exhaust microsparger and ringsparger. Hence, Maturation into Pharmaceutical Glycoprotein Cooler; http://microsite.sartorius.com/ the ambr, UniVessel SU, and Manufacturing, Applications. Adv. Biochem. fileadmin/Image_Archive/microsite/ Engin. Biotechnol. 115, 2009: 1-31. BIOSTAT STR bioreactors meet the BIOSTAT®_cultibag_str/pdf/11-06-21/ 3 oxygen-transfer requirements of Eibl D, Peuker T, Eibl R. Single-Use DSM_ESACT_Sartorius.pdf. Equipment in Biopharmaceutical Manufacture: mammalian cell cultures. With a 19 Ozturk S. Cell Culture Technology: An A Brief Introduction. Single-Use Technology in -1 Overview, Cell Culture Technology for microsparger, kLa values up to 40 h Biopharmaceutical Manufacture. Eibl R, Eibl D, Pharmaceutical and Cell-Based Therapies, 2006. can be reached at 2,000 L scale, Eds. Wiley: Hoboken, NJ, 2010. 20 Bareither R, et al. Automated thereby demonstrating the superior 4 Eibl R, et al. Disposable Bioreactors: Disposable Small-Scale Reactor for High performance of this bioreactor type and The Current State-of-the-Art and Throughput Bioprocess Development: A Proof Recommended Applications in Biotechnology. making this bioreactor type the ideal of Concept Study. Biotechnol. Bioeng. 110(12) Appl. Microbiol. Biotechnol. 86 2010: 41–49. choice for high cell-density processes. 2013: 3126–3138. 5 Dreher T, et al. Microbial High Cell 21 Wise WS. The Measurement of the Density Fermentations in a Stirred Single-Use SUPERIOR SINGLE-USE BIOREACTOR Aeration of Culture Media. J. Gen. Microbio. Bioreactor. Adv. Biochem. Eng. Biotechnol. 138, 167–177, 1951, DOI: 10.1099/00221287 SCALABILITY DUE TO CLASSICAL 2013: 127–147. 22 Ochoa FG, Gomez E. Bioreactor Scale STIRRED TANK PRINCIPLES 6 Junker BH. Scale-Up Methodologies for Up and Oxygen Transfer Rate in Microbial Their excellent performance Escherichia coli and Yeast Fermentation Processes: An Overview, Biotech. Adv. 27, 153– Processes. J. Bio. Bioeng. 97(6) 2004: 347–364. characteristics make the single-use 176, 2009, DOI: 10.1016/j.biotechadv.2008. 7 ambr250, UniVessel SU, and Mire-Sluis A, et al. New Paradigms for 10.006.  BIOSTAT STR bioreactors ideal for Process Validations. BioProcess Int. 11(9) 2013: 28–39. mammalian cell culture — even for 8 Löffelholz C, et al. Bioengineering Corresponding author Davy De Wilde is very demanding, high cell-density or Parameters for Single-Use Bioreactors: microcarrier-based processes. director of marketing, fermentation Overview and Evaluation of Suitable Methods. technologies; Thomas Dreher is scientist, Low-shear agitation with three-blade Chemie Ingenieur Technik 85(1–2) 2013: 40–56. R&D upstream technology; Christian 9 segment impellers provides Lara A R, et al. Living with Zahnow is scientist, R&D upstream Heterogenities in Bioreactors. Mol. Biotech. homogeneous mixing. With a technology; Ute Husemann is R&D 34(3) 2006: 355–381. nonparticle-shedding microsparger, manager, upstream technology; Gerhard 10 Platas Barradas O, et al. Evaluation of operators can reach oxygen transfer Greller is R&D director, upstream –1 Criteria for Bioreactor Comparison and rates of up to 40 h at 2,000 L scale. Operation Standardization for Mammalian technology; Thorsten Adams is product Like other single-use bioreactors, such Cell Culture. Eng. Life Sci. 12, 2012: 518–528. manager, fermentation technologies; and systems make cell culture operations 11 Storhas W. Bioreaktoren und Periphere Christel Fenge is vice president of more flexible, cost-effective, and less Einrichtungen. Vieweg & Sohn marketing, fermentation technologies at time-consuming. Their classical Verlagsgesellschaft: Braunschweig/Wiesbaden, Sartorius Stedim Biotech. stirred-tank design allows relying on Germany, 1994. 12 well-known and established scale-up Catapano G, et al. Bioreactor Design and Scale-Up. Cell Culture and Tissue Reaktion criteria such as power input per Engineering: Principles and Practice. Eibl R, et volume or tip speed. In addition, their al., Eds. Springer-Verlag Berlin Heidelberg: geometrical similarities across all Berlin, Germany, 2009. scales facilitate successful scale up and 13 Stanbury PF, Whitaker A, Hall J. scale down as well as process transfer, Principles of Fermentation Technology. Pergamon thus derisking scale-up and process Press Oxford: New York, NY 1995. 14 transfers significantly. These single- Marks D. Equipment Design Considerations for Large-Scale Cell Culture. use stirred tank bioreactor solutions Cytotechnol. 42, 2003: 21–33. enable seamless scale-up from 250 mL 15 Fenge C, et al. Agitation, Aeration, and to 2,000 L. This makes them ideal Perfusion Modules for Cell Culture solutions for any modern Bioreactors. Cytotechnol. 11, 1993: 233–244. biopharmaceutical development and 16 Zlokarnik M. Rührtechnik Theorie und production facility for monoclonal Praxis. Springer-Verlag Berlin Heidelberg, antibodies, recombinant proteins, and Berlin, Germany, 1999. 17 vaccines. Mol O. Meeting Increased Media Throughput and Enhanced Process-Control with Single-Use Solution for Meeting

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 19 B IOP ROCESS UPSTREAM/PRODUCTION

Integrated Optical Single-Use Sensors Moving Toward a True Single-Use Factory for Biologics and Vaccine Production by Henry Weichert, Julia Lueders, Joerg Weyand, Thorsten Adams, and Mario Becker

hrough the past decade, single- Photo 1: Different scales of single-use bioreactors with integrated, single-use pH and DO sensors; use bioreactors for culturing (LEFT TO RIGHT) UniVessel SU, BIOSTAT RM 20, and BIOSTAT STR 2000 systems from Sartorius Stedim mammalian and insect cells Biotech have been widely adopted in Tpreclinical, clinical, and production- scale biopharmaceutical facilities (1, 2). With such bioreactors in operation, monitoring and control of process parameters is vital for ensuring critical quality attributes (CQAs) of biologicals or vaccines are met for production of a safe product. Traditionally, bag-based and bench-top vessels have been fitted with conventional pH and dissolved oxygen (DO) probes similar to those used in stainless steel or bench-top bioreactors. DO and pH are the most commonly monitored physicochemical Single-use sensors offer a number and be comparable to conventional parameters measured in real time. For of advantages to scientists using probes. Case studies show that in real many processes, pH measurement and disposable bioreactors: lessening large-scale cell culture runs up to control is critical because small contamination risks and eliminating 1,000-L scale. Advances in single-use deviations can influence culture cleaning, associated validation, sensor design and performance could growth and metabolism, particularly sterilization, and probe-calibration enable the biopharmaceutical industry glucose consumption and lactate steps. Despite the benefits of these to move closer to completely single- production (3). DO must be monitored sensors, the biopharmaceutical use production of biologicals of and controlled to prevent changes in industry has been slower to adopt consistent quality. oxygen concentration, which can cause them than single-use bioreactors. One problems such as excessive lactate reason for that was stability issues INTEGRATED DESIGN production, reduced antibody related to irradiation, drift, and sensor For integration into single-use glycosylation, or cytotoxicity (4, 5). lifetime of optical pH sensors bioreactors, optical single-use sensors Reusable sensors usually are compared with traditional probes. must be designed so that they can be calibrated separately, mounted in Also, few single-use bioreactors come fitted to single-use bags or containers probe assemblies, autoclaved, and then ready equipped with these sensor and irradiated. Then they have to be fitted by means of an aseptic types at different scales. ready to calibrate easily and function connector to a single-use bag. Such Here, we describe Sartorius Stedim accurately after irradiation in real-time time-consuming manual tasks can Biotech’s integrated single-use sensor cell culture conditions. Optical sensors reduce many benefits of single-use assemblies for different scales of designed for integration into single- bioreactors and could even introduce single-use bioreactors. They are use bioreactors described herein contamination. designed to ensure robust performance determine pH and DO concentration

20 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Table 1: Technical features of optical single-use pH and DO sensors (from PreSens Precision Sensing GmbH) used in Sartorius single-use bioreactors. Single-Use pH Sensor Single-Use DO Sensor Patch composition Polymer matrix with two linked dyes Silicon matrix with integrated dye Dyes pH-sensitive fluorescein derivate with a nanoparticle Platinum derivate encapsulated reference dye Wavelengths Excitation at 480 nm, emission at 570 nm Excitation at 505 nm, emission at 630 nm Measurement method Dual lifetime referencing method (DLR): This method Fluorescence quenching: Light from a blue-green LED excites enables internally referenced measurements. A an oxygen sensor to emit fluorescence. If the sensor spot combination of different fluorescent dyes detects encounters an oxygen molecule, excess energy is transferred intensity changes in the time domain. to that molecule in a nonradiative transfer, decreasing or quenching the fluorescence signal.

Calculation of pH or DO Based on existing pH, a pH-sensitive dye undergoes a Dye emission depends on O2 concentration. O2 quenches a values phase shift, which is compared with the reference dye. fluorophore and reduces light emission. Final values are Final pH is calculated according to the Boltzmann calculated using the Stern–Volmer equation, applying equation and a calibration curve. calibration parameters as constants. Measurement ranges pH 6–8 0–110% at 37 °C based on fluorescence and Photo 2: Sensor fitted to stirred, single-use bioreactor with sensor port in “closed” (LEFT) air- luminescence principles, which are protected position and “open” position (RIGHT) for cell culture use well-established methods of determining those parameters for cell culture (6, 7). When fitted to a single- use bioreactor, these sensors are connected to a measurement amplifier that transmits light of a specific wavelength to a sensor patch. Fluorescent dyes in that patch are stimulated by sinusoidally modulated light, allowing pH or O2-dependent emission intensity to be captured by a phase shift from an excitation to a single-use sensors need to be flexible optical fiber cable. On a fluorescent signal (8). integrated into at laboratory-, pilot-, BIOSTAT STR bioreactor, sensors Table 1 summarizes the main and manufacturing-scale bioreactors. are integrated into sensor ports that features of optical single-use DO and That reduces the risk of differences in protect them during gamma pH probes (from PreSens Precision cell culture parameters that can come irradiation. All BIOSTAT STR bags Sensing GmbH) that are used in with different measurement principles. can be equipped with a second pair of Sartorius Stedim single-use bioreactors. It enables process development and sensors to provide back-up probes DO and pH patches are fixed on the biologics production with consistent because such bioreactors are designed inner surface of a single-use bag or CQAs at each scale throughout a for good manufacturing practice container, allowing both parameters to product’s life-cycle, so processes can (GMP) production. be measured noninvasively from be scaled and transferred without outside through the polymer wall. needing time-consuming additional PREPARATION OF INTEGRATED That eliminates contamination risk investigations. SINGLE-USE SENSORS because the sensors are introduced to Sensor patches often have to be Sterilization: To produce fully the bioreactor assembly before located in different places and integrated, single-use bioreactors and sterilization. The single-use DO sensor connected up differently for optimal sensors that are ready-to-use and offers additional advantages over measurement of pH and DO, consistent in quality, the assembly traditional, reusable, glass Clark depending on the bioreactor design must be provided as one sterile unit. electrodes because it requires no and materials and/or the type of cell Bioprocess scientists then can begin electrolytes. That significantly reduces culture mixing used (Photo 1). For using such integrated bioreactors equilibration times, which saves time example, on a rigid, stirred, 2-L without having to fit sensors, which in use by making the integrated single- single-use UniVessel SU bioreactor, can introduce the possibility of use bioreactor and sensor ready to single-use sensor patches are contamination. calibrate with cell culture media at the integrated into the vessel bottom, and Gamma and beta irradiation are point of installation. read-out comes directly through free- established methods of sterilizing beam optoelectronics. In the rocking single-use bioreactors. However, SCALABILITY motion BIOSTAT RM system, single-use pH sensors can be affected For seamless scalability of biologics sensors are installed by means of tubes during irradiation by acidic gases and and vaccine production processes, and their read-out comes through a ozone generated by gamma or beta

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 21 Figure 1: Calibration curves of single-use pH sensors protected from Figure 3: Calibration function of optical DO sensors residual air by the sensor port and subjected to different irradiation doses

60 1.0 6 (A) I W (B) 55 25 kGy 0 = 0 = 1 + K ([O ]) 30 kGy I([O ]) ([O ]) SV 2 2 W 2 5 50 35 kGy 0.8

40 kGy 0

45 I 0

45 kGy /

4 I 40 0.6 τ or or τ/τ 0 I

35 0 / /τ I 3

Phase (°) 30 0.4 25 2 20 0.2 1 15 0 20 40 60 80 100 10 Oxygen (%) 5 6 7 8 9 10 pH beams in the surrounding air. That Figure 2: Calibration function of optical pH proportional, only two points may cause a loss of active dyes in sensors measured at 100% and 0% air 56 r sensor patches and impair min saturation are necessary for calibration. 52 measurement performance, especially 48 To calibrate a single-use pH sensor for pH dyes. When producing 44 pH0 for specific cell culture process integrated bioreactor and pH-sensor 40 conditions, a one-point calibration is 36 assemblies, methods are needed to φ (°) necessary after the medium reaches a 32 d protect sensor patches from residual pH temperature and CO equilibrium. 28 r 2 air in those assemblies. For example, 24 max This must be performed because of BIOSTAT STR stirred systems 4 5 6 7 8 9 differences between cell culture media contain internal stirrers and spargers, pH and reference systems, the latter which prevent bags from folding flat having been used to determine and thus create air pockets. calibration parameters. We A highly specialized design of together and provide a fully sterile and recommend daily off-line sampling for sensor port protects sensors during functional assembly, which is ready to pH measurement to determine irradiation of such containers (Photo use from installation. whether deviations go beyond defined 2). It has a closed (irradiation, storage) Calibration: From each batch of criteria, typically 0.1 pH units. If it position that isolates sensors from the ~1,000 patches, a defined number of has done so, then a single-use pH air and a working position in which sensors are used to generate calibration sensor should be recalibrated. they come into contact with the bag values after irradiation. This is part of A recalibration is also necessary lumen. By contrast, rocking-motion routine quality control (QC) and when the ionic strength of a medium single-use bioreactors can be flattened ensures that sensors in each batch is altered (e.g., through addition of to contain marginal residual air meet stringent quality acceptance base to control pH). Figure 4 volumes, making no additional criteria. An integrated quality demonstrates a calibration-curve shift measures necessary to protect them management system ensures full caused by changes in ionic strength during irradiation. Similar to such traceability of all production steps, over a range of 50 mM to 200 mM, bioreactors, the residual air volume in including raw-material lots, which represents the most common UniVessel SU vessels is small, so intermediates, and sensor components. range for mammalian cell culture optical sensors remain unaffected. Calibration of single-use PreSens processes and related feeds. Also, UniVessel SU bioreactors are pH sensors involves four calibration Ionic strength represents electric beta irradiated, which is a gentler type parameters approximated from the field strength based on the total r of sterilization. sigmoidal calibration functions: min, amount of ions in solution. They r Gamma irradiation does not affect max, dpH, and pH0 extrapolated by a influence conductivity of electrolytes, single-use pH probe performance curve fit according to the Boltzmann including the electrolyte gel in optical alone. Figure 1 demonstrates equation (Figure 2). For calibration of patches. By comparison, osmolarity reproducible and consistent pH a single-use PreSens DO sensor, the measures the total amount of osmotic measurements of optical pH sensors correlation between oxygen active substances in solution, that have been subjected to a broad concentration and luminescence representing not only ions, but also range of irradiation doses. The data lifetime is expressed by the Stern– carbohydrates and proteins, which do show that it is possible to sterilize a Volmer equation (Figure 3). Because not influence ionic strength of a single-use sensor and bioreactor oxygen and luminescence are directly medium.

22 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 4: Calibration-curve effect of changing in ionic strength Figure 5: Calibration curves after accelerated aging of single-use pH sensors up to three years after gamma irradiation 55 IS = 50 mM 60 50 IS = 100 mM 55 Reference sensor IS = 200 mM Three years after irradiation 45 Two years after irradiation 50 One year after irradiation 40 45 35 40 30 Phase (°) 35 25

Phase (°) 30 20 25 15 20 4 5 6 7 8 9 pH 15 5 6 7 8 9 10 pH Figure 6: Continuous pH measurement by a single-use optical sensor and classical pH probes in buffer at two different pH values Figure 7: Calibration curves of used pH sensors compared to unused 7.5 sensors; good comparability is achieved between a sensor exposed to 7.4 20.000 measuring cycles and an unused sensor. 7.3 Hamilton 55 Unused sensor 7.2 pH electrode PreSense Sensor after 20,000 cycles optical pH Sensor after 40,000 cycles 7.1 Lot 2 50 7.0 45 pH 6.9 Reference 6.8 pH meter 40 (Sartorius)

6.7 PreSense 35 optical pH 6.6 Lot 1 Phase (°) 30 6.5 0 50 100 150 200 250 Time (hours) 25 20

Operational Considerations: Fluorescence dyes in optical 15 sensors are sensitive to light. Exposure to intense light can 5.0 5.5 6.0 6.5 7.0 7.5 8.0 8.5 9.0 cause irreversible damage to those sensors, making them pH inoperable. So bioreactors with integrated optical sensors must be stored in light-proof packaging until bioprocess using a Sartorius bench-top pH meter. That pH probe was runs begin to keep the effect of light to a minimum kept in potassium chloride (3M KCl) and calibrated before established for all Sartorius Stedim Biotech single-use every single measurement. containments. And during operation, users should After 140 hours, we shifted the pH from 6.7 to 7.3 using minimize direct exposure of optical sensors to light 1 molar sodium hydroxide (1M NaOH) and recalibrated the sources. optical single-use sensors. Deviation between the optical Stability: The shelf life of single-use sensors has been single-use pH sensor and the reference pH probe was <0.03 established at two years before and up to three years after pH units for about 200 hours. That equates to ~12.000 gamma irradiation. So they can be stored for up to two single measurement points and <0.1 pH units for about 270 years before being assembled into an integrated single-use hours, ~16.000 single measurement points. This test bioreactor, which then has a shelf life of two years after demonstrated long-term stability and accuracy of the single- irradiation. The single-use sensors will maintain their use pH sensors compared with conventional bench-top and calibration parameters within strict acceptance criteria even inline pH probes with a buffer system. for three years (Figure 5), so they are no longer the limiting factor for shelf life of single-use bioreactors. CASE STUDY: CELL CULTURE In cell culture, pH control is important to maintaining CASE STUDY: LONG-TERM STABILITY IN BUFFER SYSTEM optimum growth and metabolism. Frequent pH As part of implementing optical sensors into different measurement is not required with Chinese hamster ovary single-use bioreactors, we executed detailed performance (CHO) and other mammalian cell lines because their tests. The first test measured pH every minute for >11 days growth and metabolism are slower than those of microbial at 37 °C with a buffer system at an ionic strength 150 mM cultures, with typically simple step changes in pH involved. (Figure 6). The pH value was controlled using classical pH So a measuring frequency of once per minute is again probes (Hamilton easyferm), with reference pH measured effective to provide sufficient data for robust control loops. SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 23 Figure 8: Measurement of pH by a single-use, optical pH sensor and a Figure 9: Measurement of pH by a single-use, optical sensor and classical online and offline pH probes of a 50-L fed-batch CHO cell culture classical on-line and off-line pH probes of a 1,000-L fed-batch CHO cell grown in a BIOSTAT STR 50 bioreactor; the single-use pH sensor was culture; the single-use sensor was recalibrated on days 1, 3, and 6. recalibrated daily. Fed-batch 8.0 5 8.0 pH on-line 7.5 Fed-batch 7.8 optical pH on-lineclassical 7.4 4 7.6 Daily feed additions pH off-line Daily feed additions (ActiCHO Feed-B: pH 11) 7.3 (ActiCHO Feed-B: pH 11) 7.4 7.2 3 7.2 7.1 7.0 pH pH 7.0 2 6.8

Flow (L/m) Flow Recalibrations 2 6.9 6.6 6.8 CO pH on-lineoptical 1 6.4 6.7 pH on-lineclassical 6.2 6.6 pH off-line 0 6.0 6.5 0 2 4 6 8 10 12 14 16 18 6.0 Time (days) 0 5 10 15 20 Time (days) REFERENCES Optical pH sensors have a life time of ≥20,000 read- 1 Eibl R, et al. Disposable Bioreactors: The Current State-of-the- outs. At that measuring frequency, continuous measurement Art and Recommended Applications in Biotechnology. Appl. Microbiol. Biotechnol. 86 (1) 2010: 41–49. over 14 days is possible without any significant loss of 2 Scott C. Single-Use Bioreactors: A Brief Review of Current performance during cultivation (Figure 7). Technolog y. BioProcess Int. 5(5) 2007: S44–S51. Single-use pH sensors integrated into BIOSTAT STR 3 Wayte J, et al. pH: Effects on Growth Productivity of Cell Lines 50 and 1000 bags provide comparable results to on-line and Producing Monoclonal Antibodies: Control in Large Scale Fermenters. off-line pH probes, as shown with 50-L and 1,000-L fed- Cytotechnol. 22, 1997: 87–94. batch CHO cell cultures over 17 days (Figures 8 and 9). 4 Fleischaker RJ, Sinskey AJ. Oxygen Demand and Supply in Cell Our results demonstrate that integrated single-use pH Culture. Appl. Microbiol. Biotechnol. 12, 1981: 193–197. sensors are reliable and robust and provide a scalable 5 Kunkel JP, et al. Dissolved Oxygen Concentration in Serum-Free method for monitoring pH during a full cell culture run. Continuous Culture Affects N-Linked Glycosylation of a Monoclonal Antibody. J. Biotechnol. 62, 1998: 55–71. 6 CONCLUSION Shabbir B, et al. Phase Fluorometric Sterilizable Optical Oxygen Sensor. Biotechnol. Bioeng. 43(11) 1994: 1139–1145. In bioprocess applications, single-use sensors provide 7 Agayn VI, Walt DR. Fiber-Opic Sensor for Continuous consistent and precise measurement that compare well to Monitoring of Fermentation pH. Bio/Technol. 11, 1993: 726–729. traditional online and offline probes. Single-use bioreactors 8 Naciri M, et al. Monitoring pH and Dissolved Oxygen in with integrated optical single-use sensors are available at Mammalian Cell Culture Using Optical Sensors. Cytotechnol. 57(3) 2008: different scales (from 15 mL to 2,000 L). The sterilization 245–250.  and design of optical probes have been optimized for the type of single-use bioreactor to which they will be fitted. Henry Weichert is head of product management in process That allows for robust measurement performance analytics; Julia Lüders is a PAT scientist in sensors and throughout each cell culture run. The sensors provide chemometrics; Mario Becker is director of marketing and product process development scientists with good data management for PAT and automation; Thorsten Adams is comparability across scales: from small-scale rigid vessels fermentation technologies product manager; and Joerg Weyand is used in process development and characterization all the an application specialist in fermentation technology at Sartorius way to 2,000-L bag-based bioreactors for clinical and Stedim Biotech GmbH, Goettingen, Germany. commercial production (including seed-train bioreactors such as rocking-motion systems). Easy and straightforward single-use sensor operation and calibration provides accurate data as long as optical sensors are protected from direct light. Such sensors are integrated into single-use bioreactors to create a single, sterile, sealed unit. That allows for rapid set-up with easy alignment with automated control systems and prevents bioreactor contamination by requiring no fitting of reusable probes. Development of robust, integrated, single-use sensors and bioreactors thus enables the biopharmaceutical industry to move closer to implementing automated, fully single-use factories to produce safe and effective biologicals and vaccines.

24 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT B IOP ROCESS DOWNSTREAM PROCESSING Diatomaceous Earth Filtration Innovative Single-Use Concepts for Clarification of High-Density Mammalian Cell Cultures

by Tjebbe van der Meer, Benjamin Minow, Bertille Lagrange, Franziska Krumbein, and Francois Rolin

n the past decade, Figure 1: Filtration principle of dynamic body-feed filtration (DBF) with diatomaceous earth (DE) biopharmaceutical manufacturers (LEFT) and conventional filtration (RIGHT) have demonstrated major Conventional DBF Filtration improvements in monoclonal or Filtration

antibodyI (MAb) production, Flow Flow exhibiting product titers frequently in the range of 5–10 g/L using standard fed-batch mammalian cell cultures (1, 2). Increased product yields allow for smaller-scale production vessels. With

2,000-L single-use bioreactors already low solids capacity commercially available, single-use Low volume and manufacturing of biomolecules becomes more and more an option. Other recent developments in the

biopharmaceutical industry — e.g., High and volume Flow higher solids capacity drugs for smaller indications and more Flow potent drugs allowing for lower dosages — will further stimulate the demand for smaller and more flexible High specific flow Low specific flow rates, rate, no clogging rapid filter clogging single-use manufacturing facilities. Although single-use technology in MAb Cells and debris Diatomite general has matured considerably over the past few years, some unit operations (e.g., cell removal) still compensate for fluctuating filtration clarification of solutions with high need more attention to become more capacities. That drives up costs and solids content. The first use of economical and robust. High product increases waste. Other new diatomaceous earth (DE) as a filter titers often result from increased cell commercially available single-use cell- aid in fractionation of human plasma densities rather than increased specific removal technologies such as was reported over five decades ago (3). productivities per cell, and the centrifuges still lack capacity. For Since then, a number of resulting solids content poses harvesting higher–cell-density manufacturing processes for considerable challenges on commonly cultures, a major technical production of intravenous applied harvesting technologies. breakthrough would be welcomed. immunoglobulin (IVIG), albumin, Currently the most prevalent single- and clotting factors have been use harvesting technology, depth BODY FEED FILTRATION SUCCESSFUL developed based on that technology filters block at lower loading capacities FOR PLASMA FRACTIONATION (4–6). Fractionation uses the principles with higher biomass concentrations. When we looked at other industries of selective precipitation by pH Higher contaminant concentrations that have similar process needs — adjustment, ionic strength, addition of also make depth filters more sensitive such as the plasma fractionation alcohol, and temperature shifts. to batch variation, which can lead to industry — we found that they often Precipitates are removed by depth 50% oversizing of filter area to use body-feed filtration for filtration often in combination with

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 25 Figure 2: Mean particle-size distribution of cell-free Chinese hamster ovary (CHO) culture Photo 1: Scanning electron microscopy (SEM) supernatants from three different harvest days determined at different pH values; green curve = pH shows the porous structure of Celpure 300 7.0, red curve = pH 5.0; measurement performed using a Mastersizer 2000 system (Malvern diatomaceous earth (magnitude 1,000 times). Instruments)

15

10

5

Number of Particles (%) 0 0.1 1.0 10.0 30.0 Particle Size (μm)

Figure 3: Turbidity measurement of the DBF Figure 4: Mean residual DNA amount after DE filtrates at pH 7.0 and pH 5.0 for 10 different filtration at pH 5.0 for 10 different cell culture host-cell proteins (6) decrease along cell cultivations with an initial turbidity of supernatants, expressed in percentage of initial with pH. Figure 2 shows the mean 2,396–3,235 NTU DNA concentration at pH 7.0, measured in the particle-size distribution for three 120 cell free supernatant; initial concentrations were 475–730 ppm (DNA quantitation with different cell-free cell culture 100 PicoGreen assay from Life Technologies). supernatants at neutral pH (green line) 80 and after the pH of those cultures was 100 60 lowered to pH 5 (red line). Lowering 80 the pH leads to formation of bigger 40 60 particles and makes the typical 20 submicron particles (<1 μm) present at 40 0 pH 7.0 completely disappear. Average Turbidity (NTU) Turbidity Average HCCF, pH 7 HCCF, pH 5 20 Therefore, body-feed filtration would clearly benefit from lowered pH. 0 diatomaceous earth, which is added as (NTU) Turbidity Average HCCF, pH 7 HCCF, pH 5 In addition to the significantly an aid to increase filter throughputs. reduced filter-aid concentration, all Recently, the principles of body-feed tested acidified cell culture filtration have been tested for The minimum amount of DE to supernatants showed much clearer harvesting cell cultures, with promising guarantee smooth filtration depends on filtrates than their neutral results (7). Our objective was to the particle concentration. Laboratory- counterparts (Figure 3). When we evaluate the technology as a potential scale experiments with many different analyzed the particle-size distribution single-use alternative to centrifuges and mammalian cell lines, culture media, of neutral body-feed filtrates, we depth filters. Here we describe the and viabilities revealed a correlation detected only small particles most interesting findings we obtained between the wet cell weight (WCW) (<0.4 μm). Filtration with a 0.2-μm using a number of cell lines and culture and the required amount of DE at membrane barely reduced those media, which were kindly provided by constant pH (data not shown). For all turbidity values, indicating that very different biotech companies. Together tested cultures, the optimum DE small particles are mainly responsible with Rentschler Biotechnologie concentration was in the range of for the higher turbidity values of the GmbH, Sartorius Stedim Biotech 40–50% of WCW. In all cases, the neutral body-feed filtrations. tested the optimized conditions in a filter-aid ratio could be reduced The absence of smaller particles in 600-L cell culture production process significantly to a range of 20–30% acidified culture supernatants probably to evaluate the scalability of body-feed when pH was lowered to pH 5. is also the reason for higher filtration filtration technology. capacities at constant filter-aid LOW PH PRECIPITATION IMPROVES concentrations. We assume that small HOW BODY FEED FILTRATION WORKS CLARIFICATION RESULTS particles deposit in the pores of DE When the concentration of solids or Performance differences between particles, lowering the permeability colloids is too high in turbid biological acidified (pH 4.3–5.5) and neutral cell and filtration capacity of the DE process fluids needing clarification, the culture fluids for other cell-removal overall. By increasing the filter-aid filter cake forming on the surface of a technologies such as microfiltration (4) concentration, we could increase its filter becomes impermeable and blocks and depth filtration (5) can be capacity for retaining small particles the filter (Figure 1, right). Adding explained by precipitation of smaller without losing the required cake highly porous DE creates a more particles at lower pH levels. The permeability. permeable filter cake, which prevents solubility of cell debris and negatively Significant reduction of filter-aid blockage (Figure 1, left). charged impurities such as DNA and consumption at low pH is promising

26 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 5: Schematic representation of pilot-scale body-feed filtration Figure 6: (TOP) 3-D illustration of commercial DBF process-scale module holder; (CENTER) a Celpure C300 drawing of one single-use DBF module; diatomaceous BOTTOM earth ( ) flow path inside one module indicates in red the feed flow and in green the filtrate Dust-free flow path; dotted lines indicate the filter plates. transfer

Clarified Mixing 1,000-L SUB Body-feed 0.2-μm filter sterile system filter filtrate

Controlled pH adjustment

Acetic acid (2M) stock solution from both economical and handling purposes (12). In total, 1,000 L of a points of view. Another important high–cell-density (17.6 × 106 cells/mL, economical factor to consider is the 95% viability) Chinese hamster ovary recovery rate at lower pH levels. (CHO) cell fed-batch culture were Although most antibodies are stable at available for the depth-filtration and acidic conditions (9), in some cases body-feed filtration runs. In the low-pH precipitation of cell culture largest single run, we subjected 600 L fluid will lower recovery rates, most to body-feed filtration at pH 5.0. likely due to coprecipitation of the WCW on that day was 8%. Feed antibody (6). However, some authors For the scale-up experiment, we have described that lowering the pH installed seven process-scale modules value influenced overall antibody with a total filtration area of 1.61 m2 recovery positively by preventing in a universal stainless steel holder. enzymatic reduction of the product The filter cassettes consist of two (10, 11). In all our laboratory-scale polyethylene filter plates, which retain Cake Space Cake Space Cake body-feed tests at low pH, we the DE and biomass (Figure 6). achieved a recovery rate >85% (data In total, we added 12 kg of not shown). Cellpure C300 DE to the 600-L bulk Filtrate harvest. Prefilled DE bags were POSITIVE IMPACT ON RESIDUAL DNA connected with a dust-free adapter to a considerably higher than turbidity had AND HOST-CELL PROTEIN LOAD mixing bag for fast and safe DE been during filtration. Inadequate With the low-pH precipitation and transfer. Just five minutes of gentle dosing of the neutralization buffer subsequent retention of contaminants mixing was sufficient to dissolve that probably had resulted in local pH by body-feed filtration, we DE powder in the cell suspension. excursions and caused the turbidity accomplished a significant reduction Before filtration, we adjusted the pH increase in that final pool. In a small- of DNA in the primary recovery step of the resulting mixture to a final level scale parallel test, the turbidity from our cell culture runs. Figure 4 of 5.0 and gently mixed it for two increase was prevented through more shows the mean residual DNA content hours at 140 rpm. gentle neutralization of the filtrate of 10 different body-feed filtrates after Pressure increased steadily during pool. To improve that neutralization low-pH precipitation at pH 5.0. We filtration (Figure 7, left). We step, an integrated ready-to-use first measured the initial DNA terminated filtration when the pressure process skid is under development that content in the neutral, cell-free reached 1.3 bar and the crude harvest will enable controlled inline pH supernatant of each cell culture at pH had been filtered. During the entire adjustment and prevent overshooting. 7.0. After a low-pH precipitation step filtration process, the system We found IgG1 recovery to be followed by DE filtration, DNA maintained a high and stable flux acceptably high at 85% (Figure 7, content in the cell culture supernatants slightly above 300 L/m2/h. Overall, a right). In the future, an optimized was reduced by 65%. capacity of 311 L/m2 was achieved. We neutralization procedure and larger monitored a very low turbidity of 5–8 postfiltration flush should further PILOT-SCALE TEST RESULTS NTU (Figure 7, left) in the clarified improve MAb recovery. We monitored At Rentschler Biotechnologie, we harvest stream during filtration. contaminants such as a host-cell performed tests at pilot production After neutralization, pool 3 (the protein and DNA throughout the scale to evaluate the applicability of final pool) exhibited a turbidity of process. In the final pool (after buffer DBF technology for manufacturing 41 NTU (Figure 7, right), which was flush and neutralization), those levels

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 27 Figure 7: Results of DE body-feed scale-up experiment at reduced pH (5.0) with seven filter modules; filtration performance (LEFT), pressure (bar), and flux (L/m2) as well as the course of turbidity during filtration; recovery of IgG1, blue columns (RIGHT) measured from crude harvest and harvest pool without buffer flush (pool 1), with buffer flush (pool 2), and after neutralization of harvest fluid (pool 3)

2 600 240 10,000 Pressure pH 5.0 DBF × 7 Flux Turbidity during filtration 1,128 NTU 1,000 Turbidity (NTU) Flux (L/m 160

1 300 100 (bar) 2 p /h)

Δ 41 NTU 41 NTU

IgG1 (g) × 10 80 16 NTU 10

7 NTU 8 NTU 6 NTU 5 NTU 5 NTU 5 NTU IgG1 IgG1 IgG1 IgG1 0 0 0 1 0 120 240 360 Crude Pool 1 Pool 2 Pool 3 Capacity (L/m2) Harvest were reduced from 841 to 629 mg/mL currently the method of choice for 9 Singh N, et al. Clarification of and 13.8 to 5.0 μg/mL, respectively. large-scale cell removal. Even very Recombinant Proteins from High Cell Density Mammalian Cell Culture Systems Using New dense crude cell harvests could be Improved Depth Filters. Biotechnol. Bioeng. CONCLUDING REMARKS clarified quickly at high flow rates. 110(7) 2013: 1964. Our aim was to demonstrate the Moreover, this method has the 10 Brodsky Y, et al. Caprylic Acid universal applicability of a new single- additional benefit of efficiently reducing Precipitation Method for Impurity Reduction: use harvest method for mammalian contaminants in a single step. An Alternative to Conventional Chromatography cell culture, suitable even for high– Scalability — one of the most for Monoclonal Antibody Purification. Biotechnol. Bioeng. 109(10) 2012: 2589. cell-density cultures. Tests using crude important requirements in bioprocessing 11 Kao Y, et al. Mechanism of Antibody harvests from different cell lines and — is easily attainable following a Reduction in Cell Culture Production culture conditions allowed us to generally linear approach with very Processes. Biotechnol. Bioeng. 107(4) 2010: 622. determine the optimal concentration consistent process performance. 12 Trexler-Schmidt M, et al. Identification of DE as a filter aid in relation to and Prevention of Antibody Disulfide Bond WCW, which is an easily accessible REFERENCES Reduction During Cell Culture Manufacturing. process specific for cell removal and 1 Kelley B. Industrialization of MAb Biotechnol. Bioeng. 106(3) 2010: 452. 13 harvest processing. With respect to Production Technology: The Bioprocessing Minow B, et al. High-Cell-Density Industry at a Crossroads. MAbs 1(5) 2009: Clarification By Single-Use Diatomaceous process economics, the 50% reduction 443–452. Earth Filtration. BioProcess Int. 12(4) 2014: of filter aid required with low-pH 2 Buttler M. Animal Cell Cultures: S36 –S47.  filtrations is promising. Recent Achievements and Perspectives in the Generally, the pilot-scale results Production of Biopharmaceuticals. Appl. MSc, is a product confirmed our findings from DBF Microbiol. Biotechnol. 2005, 283–291. Tjebbe van der Meer, manager at Sartorius Stedim Biotech GmbH, 3 Cohn EJ, et al. Preparation and filtration trials at laboratory scale: August-Spindler-Straße 11, 37079, Göttingen, Reducing pH to 5.0 after addition of Properties of Serum and Plasmaproteins, IV: A System for Separation into Fractions of the Germany, tjebbe.vandermeer@sartorius- DE to crude cell-culture supernatant Protein and Lipoprotein Components of stedim.com. Benjamin Minow, PhD, is gives the best performance in terms of Biological Tissue and Fluids. J. Am. Chem. Soc. director of cell culture disposable filtration capacity, flux, and 68, 1946: 459–475. manufacturing at Rentschler Biotechnologie contaminant removal. 4 Curling J, et al. Production of Plasma GmbH, Erwin Rentschler Straße 21, 88471 The applicable flux rate of DBF Proteins for Therapeutic Use. Wiley-Blackwell: Laupheim, Germany, benjamin.minow@ technology is very advantageous. A Hoboken, NJ, 2012. rentschler.de. Bertille Lagrange, MSc, is a 5 600-L harvest was processed within More J, et al. Purification of Albumin scientist at Sartorius Stedim Biotech GmbH, from Plasma. Blood Separation and Plasma only an hour of filtration using just August-Spindler-Straße 11, 37079, Göttingen, Fractionation. Harris J, Ed. Wiley-Liss: Germany, bertille.lagrange@sartorius- seven modules. A module holder Hoboken, NJ, 1991. stedim.com. Franziska Krumbein, Dipl. Ing allows arrangement of a maximum 33 6 Stucki M, and al. Investigations of (FH), is director of process technologies at modules, so we estimate that a harvest Prion and Virus Safety of a New Liquid IVIG Sartorius Stedim Biotech GmbH, August- volume of ~3,000 L could be filtered Product. Biologicals 36, 2008: 239–247. Spindler-Straße 11, 37079, Göttingen, 7 in the same time. Westoby M, et al. Effects of Solution Germany, franziska.krumbein@sartorius- In conclusion, this method allows Environment on Mammalian Cell stedim.com. And Dipl. Ing, Fermentation Broth Properties. Biotechnol. Francois Rolin, is with ChangeXplorer Production SAS, Z.I. effective clarification of high–cell- Bioeng. 108(1) 2011: 50. des Waillons 02220, Braine, France density, crude cell culture harvests in a 8 Delavaille D. How to Apply Old [email protected]. single-use set up at large scale. Techniques to New Processes. European Economically and competitively, it can Downstream Technology Forum, 7–8 September replace centrifugation, which is 2010, Göttingen Germany. Sartorius Stedim Biotech AG: Goettingen, Germany.

28 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT B IOP ROCESS INTERVIEW

Designing Single-Use Solutions for the Future A Conversation with Christel Fenge by Brian Caine and S. Anne Montgomery

ecently, BPI publisher Brian Caine and editor in chief Anne Montgomery had the opportunity to talk with RChristel Fenge, Sartorius Stedim’s VP of marketing for fermentation technologies, in the Göttingen, Germany, Sartorius facility. They began by talking about fermentation technology, a topic that led them to touch upon a number of key issues in the biopharmaceutical industry.

FERMENTATION TECHNOLOGY Left to right: Brian Caine (publisher of BioProcess International), Christel Fenge Caine: Because this special issue (vice president of marketing for fermentation technologies at Sartorius Stedim Biotech), focuses on fermentation technology, and S. Anne Montgomery (editor in chief of BioProcess International). let’s begin by talking about some recent technological improvements Also, process analytical automated bioreactor systems that can and what impact these developments technologies (PAT) have been significantly increase the speed and will have. increasingly adopted by the efficiency of process development. Fenge: The major change or biopharmaceutical industry, improving An Acquisitions Strategy: Fenge “breakthrough” of the past decade is development times and providing a describes the acquisition of TAP and the increasing use of single-use structured approach to fully its multi-parallel bioreactor systems as technologies, which started in the understanding the production process. being “very deliberate.” She comments: early 2000s with the implementation Other less regulated sectors such as “Our customers have a large number of of the Wave bioreactor (now known as “white biotechnology” have adopted development projects in their pipelines, the BIOSTAT RM [rocking motion] PAT to a much wider extent to save and we wanted to be able to help them bioreactor). Given the increased costs, and I believe this will happen to with process development. Clearly, attrition rates and development costs, our industry as well. For Sartorius highly automated, multi-parallel better tools were needed to make Stedim Biotech (SSB), this means bioreactors fill a last remaining gap clinical trial materials more efficiently. offering design of experiments (DoE) that hadn’t been addressed in the past. Subsequently, single-use stirred tank and multivariate data analysis software We want to make process development bioreactors were introduced, which led tools from Umetrics, enhanced with faster, more effective and more cost- to a huge improvement in the way we single-use sensors for pH, oxygen and efficient. For me, this acquisition fits develop biopharmaceuticals, online glucose and lactate perfectly. It provides full and seamless particularly with the large monoclonal measurement. Therefore, we also scalability from 15 mL all the way up antibody (MAb) portfolios that many decided to acquire TAP Biosystems, to 2,000 L in single-use stirred tank big pharmas have today. and we can now supply ambr multi- bioreactors: from cell line development parallel mini-bioreactors, highly and media screening at 15 mL scale, to

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 29 process development at 250-mL scale, for this large-scale single-use process characterization and scale-up at bioreactor. 2 L, preclinical and clinical production Fenge: During the past two years, at 50–1000 L, and finally, full we have been working on our “film production scale at 2,000 L. The ambr strategy.” This is a materials science bioreactor has been very quickly project for future single-use adopted across a number of different manufacturing. As the use of single- product segments, which demonstrates use bioreactors in commercial a very real industry need; if they didn’t production increases, our customers need it, they wouldn’t have adopted it.” the 2,000 L version), fix it at the are looking for new levels of assurance Caine: Tell me about the bottom of the bag holder and at the of supply and business continuity development and maturation of your motor drive, and off you go,” she adds. when it comes to components and raw single-use bioreactors. The 2,000-L STR bioreactor is materials. They are concerned about Fenge: We started with the Wave- based on the same design principles as change control; namely, how we as a style bioreactor — a simple pillow- the smaller 50-L to 1,000-L versions vendor control and manage changes in shaped bag. Other companies were that have been on the market since our production and supply chain. exploring the use of modified single- late 2008. With the development of Companies are now “outsourcing” use mixers with noncentric mixing the smaller BIOSTAT STR family, activities that, in the past, they used to devices as stirred-tank bioreactors, and members started the interaction with control themselves. And given the they launched the first single-use customers and what functionality is single-use nature of today’s bioreactors (SUBs). We went one step required for their various processes. bioreactors, manufacturing them is a further with the BIOSTAT STR, “Collaboration partners in the vaccine recurring process and not a one-off offering a true stirred-tank bioreactor industry and the MAb space discussed event covered by installation and based on the established design aspects such as impeller and sparger operational qualification (IQ/OQ ) as principles of glass and stainless steel design, port types and sizes, it was in the past when commissioning vessels. For us, it was very important connectivity, exhaust lines and stainless steel production capacity. to create a bioreactor that provides the biosafety concepts with us, and what As such, we are committed to same scale-up characteristics and capabilities and technical solutions understanding and controlling the process performance that they are looking for. Our process whole production process and the biotechnologists have been using for scientists frequently liaise with quality requirements that go with it. decades to make vaccines and drugs. customers to gain a deep insight into One aspect is raw material traceability, Seamless Scalablity: The major their processes and needs, which is right down to the polymer resins, to benefit for the customer, Fenge extremely useful,” said Fenge. where the polymers are manufactured, comments, is that it provides seamless Montgomery: What are your establishing traceability and control scalability from process development thoughts about the scalability of over the entire process from the resin (2–10 L) to commercial production single-use bags: in practical terms, is and the additives to the final bag. We and straightforward process transfer to 2,000 L the uppermost limit? want to ensure, for the long-term, that existing large-scale production Fenge: Yes, we believe so for the film material we use in our bags is capacities. She adds that the various reasons. High titers (5–10 g/L) of consistent quality. We achieve this BIOSTAT STR compares extremely and the trend toward smaller by establishing specifications and well with classic stainless steel stirred- indications, personalized medicine, controls for the polymer resins and tank bioreactors, even in demanding and regional production make 2,000 L additives that are used to manufacture processes such as cell culture on a very attractive option. Another trend the new film and by establishing microcarriers or high cell density, is the move toward concentrated fed- process understanding and control of concentrated fed-batch processes. The batch, a type of continuous culture in the film extrusion, bag assembly, and company is now in the process of which the product remains in the gamma irradiation processes. launching the 2,000 L BIOSTAT bioreactor, and that enables very high STR for commercial production. product titers. Typically, companies Ease of Use: “A key aspect for us want to benefit from single-use and during development was ease of use. smaller bioreactors to make their Installing a 1,000-L or 2,000-L processes more manageable, make single-use bioreactor bag needs to be capacity adaptation more flexible, and carefully considered; users are looking derisk process transfers. for easy, fast, and straightforward procedures that avoid complicated and FILM STRATEGY risky bag manipulations. With the Caine: In connection with the 2,000-L STR, you open the door, insert the BIOSTAT STR version, I understand bag (with a special lifting device for you developed a new film, especially

30 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT aspect is robustness; bags need to still rather naïve regarding how the reliably work in day-to-day films are made, what the sourcing operations. issues are, and how variations in Our new PE film is the thickest in composition might affect their the industry and the most robust. We products. Fenge comments that try to balance its stiffness or rigidity “manufacturing your own film helps with flexibility. A bag has to be to provide greater customer assurance. flexible, but also rigid enough to work If a company is going commercial, it in a three-dimensional large-scale will need to supply patients for years. bioreactor, mixer, or shipping bag. It We’re not talking about a clinical trial; must be resistant to both hydrostatic we’re talking about guaranteeing the pressure and liquid movement, which is supply for full-scale, commercial why we spent almost two years production, for the long-term. developing the new PE film. We Companies should talk to us about the wanted to put the key elements of materials science involved, about the traceability, control, assurance of film manufacturing process, and about supply, and robustness firmly in place. the controls we have in place. As And right from the beginning, we single-use technologies mature and wanted to have a PE film that works move into commercial production, the for a broad range of different industry is demanding assurance of From a few milliliters up to 2,000 L, Sartorius Stedim Biotech covers the bioprocessing applications, including supply. They want to ensure that their complete product range of bioreactors stirred bioreactors, Wave-type bags, vendors can provide them with single- for upstream processing. large mixing and shipping bags, storage use bags that show consistent quality and freeze-and-thaw bags, to reduce and reliable performance over and over Montgomery: Can you give us the validation effort of our customers. again,” says Fenge, “Putting raw some specific characteristics about Caine: How do you address material specifications in place and your Flexsafe bags and the new customer concerns about extractables polyethylene film? and leachables? Fenge: The new Flexsafe STR bags Fenge: We typically provide our were an important component in the customers with an E&L guide. Under development of the 2,000-L bioreactor a confidentiality agreement, they get and its application in commercial access to model liquid (water and production. It was important for SSB ethanol) data. An E&L guide is to ensure the consistent supply of a available for each bioprocessing bag dependable polymer film that would type and covers the E&L analysis for provide the robustness and assurance the complete bag, including the film, of supply that people look for when tubing, connectors, and/or, in the case using the product in full-scale of the Flexsafe STR, the agitation and manufacturing. sparger system components. One key driver was the sheer size Customers can then decide what data of the 2,000-L bag, which forced the to submit to the regulatory agencies, company to look into a new film and using either our model data or data for  our desire to stick to the same their specific process fluids. SSB can geometrical design principles of the also perform process-specific E&L It was important to existing STR product range. We also studies with customer liquids. With ensure consistent knew that our customers had the new film, we’ve taken an even supply of a experienced poor cell growth with more rigorous approach; as a supplier, certain (other) bags, so we — and our you cannot provide reliable E&L data DEPENDABLE industry partners — focused on that if you don’t have full control of your film that would when we developed the new polymer resin, additive package and provide the polyethylene (PE) film. It turned out film and bag manufacturing process. robustness and that certain antioxidants degrade to Only by doing so can you assess substances that are toxic to cells, even whether the E&L profile for one assurance of supply at very low levels, so we optimized batch is representative of future that people look for the antioxidant package to ensure batches. With full control, we can take when in full-scale that we can consistently supply the E&L discussion to a new level. manufacturing. single-use bioreactor bags that deliver Materials Composition and excellent cell growth. Another key Sourcing Issues: Many companies are

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 31 establishing operating ranges for the and the other is the business extrusion process reflects the cGMPs continuity element, which is based on that our customers use in their own long-term agreements with our manufacturing. extrusion and resin partners, safety “We feel it is extremely important stocks, and back-up capabilities.” that we, as suppliers of single-use components and solutions, fully MEDIA ADVANCES understand what they will be used for, Caine: You acquired Lonza’s cell as well as the significant value of the culture media business at the end of pharmaceutical intermediates and helps to achieve price security, and 2012. Tell us about the role that’s products being used in them. most of all, customers can rely on tested going to play and how that helps you Understanding the product profile is process solutions whereas bespoke to become a total solution provider. important: What is the intended use, solutions might include unidentified Fenge: One aspect that we couldn’t what are people doing with it, what do process risks, especially for more supply in the past was cell culture they expect? As such, we communicate complex single-use applications or media. But given our desire to provide with our end users. Some companies whole unit operations such as a “total solution,” customers can now even come to our production site to bioreactors. buy liquid media from us in our bags, talk about their drug manufacturing Assessing Risks in Single-Sourcing: made with our filters. Or alternatively, processes and products; we try to For an end user, from a validation they can use powder media, making establish a connection with the end standpoint, it does mean releasing their media at the point of use using users,” adds Fenge. control, particularly if they’re using our Flexact MP unit, filters, and bags. Sartorius as a single-source solution We are also developing further STANDARDIZATION provider. “It might require a leap of applications in this area based on the Montgomery: Do you get many faith,” admits Fenge, “and this is acquisition of the ambr multi-parallel questions about the compatibility of something that definitely needs bioreactors. Our goal is to bring added your materials with those of other ongoing work. But single sourcing is value to our customers, and so far the single-use manufacturers? not new to the industry; it’s common introduction of a media offering has Fenge: Yes, absolutely. We currently practice for critical components such gone extremely well, especially with incorporate components from various as microcarriers, chromatography gels, our key accounts. Very often, they will vendors. Our customers want and virus filters. It’s really about risk use their own cell culture medium but interchangeable systems that can use assessment, balancing risks and might be looking for a second supplier, both our bags and our competitors’ benefits, and ultimately, risk which provides an opportunity to bags, so it’s all about standardization. management through (for example) discuss buying it from us. Of course, And as much as we try to accommodate the use of safety stocks, close we are not well known as a cell culture the needs of our customers, it makes collaborations, and quality agreements media company. But Lonza has years the supply chain very challenging and between suppliers and end users. As a of media (powder and liquid) involves a lot of manual work and risk supplier, we are committed to building development and manufacturing assessments, which adds to the cost. In trust in our materials strategy and experience. Working together, our the past, this was not a major concern quality assurance. We’ve not yet combined market reach, expertise, and because most single-use bioreactors and reached the end of the innovation capabilities enable us to support our bags were used in clinical stage in single-use and/or the level of customers even better. manufacturing. And because it is a maturity that is necessary to Long-Term Growth Strategy: The relatively young technology, there is a standardize, so there’s still quite a lot media offering is part of the lot of innovation, and people want to of work to be done.” company’s 2020 strategy, a key test different solutions or simply come Long-term film supply and disaster component for future growth. And up with their own designs. But preparedness is another critical Fenge acknowledges that having the achieving the level of assurance and consideration. But Fenge feels long-term strategy to be a total robustness of supply that’s needed for confident about addressing business solution provider means that Sartorius commercial production when faced continuity through a mix of has been quite aggressive in the short- with all these different components and establishing back-up raw materials for term to put all the pieces of that modifications — that’s a different the polymer resins, back-up film strategic plan — the vision — matter. Customers should consider this extrusion and bag manufacturing together. aspect very seriously as their use of capabilities, and safety stocks of resins single-use solutions increases, which is and film rolls. “There are two EXPANDING CAPABILITIES why we talk to our customers about elements to the new film strategy,” she Looking ahead, Fenge expects the standardization. It makes the supply says. “One element is controlling how company’s single-use capabilities to chain easier and more robust, not just the film is manufactured, which grow and expand into other areas, for us, but also for our customers. It ensures batch-to-batch consistency; such as the production of ADCs, cell

32 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT technologies are also improving, and so I think we will see more industrialized processing, comparable with the routine production of chemical APIs. We will see a push toward the so-called ballroom concept, which will help to address cost pressures. With robust data on the safety of closed single-use systems, we will also be able to help our customers to present the ballroom facility concept to the regulatory agencies.

ENABLING GLOBAL EXPANSION Finally, Fenge describes the use of single-use technologies as an enabler that will provide the opportunity to SSB’s ambr15 system enables expand drug development and rapid evaluation of 24 or 48 multiple bioreactor cultures at microscale. manufacturing applications around the world for many types of disease therapies, biosimilars, and so on. about how else Sartorius Stedim treatments. “As emerging economies “We’re keeping a close eye on those Biotech could help customers to derisk grow stronger, their population wants applications, and in fact, we’re already their use of single-use systems. All access to the kinds of medications that very active in biosimilars and the things considered, it’s a bag. But we are already enjoying today in our ADC field. We’ve recently engaged a special considerations are necessary for developed world. So absolutely, single- Regenerative Medicine Taskforce to the transport, storage, unpacking, and use is definitely here to stay,” she work specifically with that market installation of that bag, so we have concludes.  segment and position our existing developed a device that allows products there more strategically than operators to pressure test their we’ve done in the past. We also bioreactor bag assembly, envision selected developments or postinstallation but preuse; it’s very adaptations. Regarding cell therapy, similar to the approach that many for example, I don’t think you can users take for their steel tanks. So far, really do cell therapy without using we have qualified the bag tester for closed, single-use systems. We are also our CultiBag STR bags, from 50 to working with Lonza and other 1,000 L, and qualification for the new collaboration partners to demonstrate Flexsafe STR bag range (up to good stem cell propagation. We 2,000 L) is in progress. In addition, strongly believe in the suspension we are shipping our STR bags route for scale-up, which doesn’t overseas to destinations such as North necessarily mean that you have to and Latin America, Australia, China, grow the cells in single cell Korea, and India, to places where the suspension. But ultimately, we believe roads are not always in the best shape. very strongly in stirred tanks. You can Our new packaging concept helps to control them. It will be interesting to ensure safe transport and mitigates see what develops and becomes any risks of damage during Christel Fenge is Sartorius Stedim standard procedure in the years transportation. It has passed the very Biotech’s VP of marketing for fermentation ahead.” stringent ASTM shipping validation technologies, based in Göttingen, Germany. is cofounder and conditions. Brian Caine publisher, and S. Anne Montgomery is BAG TESTING AND PACKAGING Caine: What you think the future is cofounder and editor in chief of BioProcess Montgomery: How does your new bag going to hold for single-use International, amontgomery@ tester ensure process safety, and how technologies? bioprocessintl.com. does your new packaging concept fit Fenge: I believe that we will see an into that picture? increased level of automation and a Fenge: Beyond talking about move toward automated single-use materials science, consistency, control, facilities. This will reduce manual and business continuity, we thought handling. Single-use sensor

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 33 B IOP ROCESS SINGLE-USE/POLYMERS

Consistently Superior Cell Growth Achieved with New Polyethylene Film Formulation by Christel Fenge, Elke Jurkiewicz, Ute Husemann, Thorsten Adams, Lucy Delaunay, Gerhard Greller, and Magali Barbaroux

uring the past decade, single- Figure 1: Polyethylene film development and manufacturing strategy use bioprocessing bags and bioreactors have gained a Resins significant foothold in the Assembly Film and bag biopharmaceuticalD industry because Extrusion Welding specifications they offer a number of advantages over Additives Irradiation traditional stainless steel equipment, Formulation Qualification especially for clinical production, development Process development multiproduct facilities, and emerging economies. At the same time, some companies are concerned that plastic Specifications Operating ranges and controls and controls materials might release potentially toxic substances that could affect cell growth and product titers (1). In a erucamide, and stearamide. Such TARGET PRODUCT PROFILE OF worst-case scenario, they could even agents act as lubricants and prevent NEW S80 MULTILAYER compromise drug safety when a stickiness of polymer film layers, which POLYETHYLENE FILM company uses disposable bags for in turn facilitates bag fabrication and Reproducible cell growth (comparable intermediate or drug substance storage. in-process use. Ideally, slip agents to that in borosilicate glass) Despite the broad acceptance of should be omitted to further reduce the Traceable and controlled polymer and single-use bags in clinical and risk of leaching substances that could additive formulation commercial biomanufacturing, some potentially affect cell growth. Robust and suitable for all single-use researchers have reported recently on bioprocessing bag applications in INNOVATIVE STRATEGY inconsistent and poor cell growth of upstream, downstream, and final filling: some production cell lines in various When Sartorius Stedim Biotech began rocking-motion bioreactor bags single-use bags (2–4). Hammond et al. development of its innovative S80 ∙ ∙ stirred-tank bioreactor bags identified a degradation product polyethylene (PE) film for the new ∙ two- and three-dimensional (2-D derived from a commonly used Flexsafe range of bioprocess bags, the and 3-D) storage bags antioxidant that can be present in some company’s vision was to establish full ∙ mixing bags commercially available single-use bags control and traceability of resins and ∙ shipping bags at levels that negatively affect cell additives, to fully control the ∙ freeze and thaw bags growth (5, 6). However, such manufacturing process (including film antioxidants are necessary to keep extrusion), and to guarantee industry- single-use bags stable: They protect the leading security of supply to ensure upstream and downstream. So it polymer film during gamma irradiation business continuity for the should meet the various physical, and protect it from oxidative biopharmaceutical industry (Figure 1). chemical, and biological requirements degradation during extrusion and The aim was to move away from the of culture media and chromatography storage. Hence, it is extremely current industry practice of using resins buffer mixing and storage, stirred-tank important to optimize and closely or films based on trade names, which and rocking-motion bioreactors, control the concentration of such in essence provides no true control of intermediate and bulk drug product antioxidants to prevent batch variability. polymer and additive formulations. storage, shipping, and freeze–thaw Other common additives are so This new film ideally should work for applications (as described in the called “slip agents,” typically oleamide, all single-use bag applications, both in “Target Product Profile” box).

34 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT STANDARDIZED Figure 2: Cytotoxicity of bDtBPP; CHO-DG44 viability (dotted line) and cell growth (solid line) as a CELL-GROWTH ASSAY function of bDtBPP concentration (7) 125 Protein-free, chemically defined cell culture medium incubated for three days at 37 °C in freshly γ-irradiated 100 sample bags at a volume/surface ratio of 3 cm2/mL Control: medium incubated three days 75 at 37 °C in Duran glass bottle

Method: rCHO DG44 cells grown in six- 50 well plates using the above treated medium

Cell Growth (% control) 25 To achieve those goals, Sartorius Stedim entered into a strategic 0 partnership with Südpack, Europe’s 0.01 0.10 1.00 leading film-extrusion company. bDtBPP Concentration (mg/L) Furthermore, we established a strong collaboration with leading polymer and additive manufacturers to access and Most important, we established the Figure 3: Cell-growth behavior in some selected film formulations; a number of film lots control the formulation of our polymer dose dependency of cell growth and and sample bags were evaluated. The “error and additive package. That is crucial to viability for a model recombinant bar” shows minimum and maximum cell a strong assurance-of-supply concept. Chinese hamster ovary (CHO) cell growth observed. line as a function of bDtBPP 120 DEVELOPMENT OF NEW FILM concentration (Figure 2) to ensure 100 Standardized Cell-Growth Assay: that our cell growth assay would During development, 25 different identify nonsuitable formulations, 80 polyethylene multilayer film which were evaluated during formulations (based on widely used development of the polyethylene film. 60 polymers and European Figure 3 illustrates results from some 40 Pharmacopoeia listed antioxidants) selected film formulations. We found were created and evaluated against the the “S80” formulation to be the most 20 defined target product profile (TPP) suitable, showing no impact on cell Cell Growth control) (% 0 of the new film. To identify growth in our standardized assay or Control NC Q S80 4, n = 12 n = 30 n = 18 potentially harmful leachables, we with other production cell lines ( 1 lot 5 lots 4 lots established a standardized cell culture film 6). assay to examine gamma-irradiated sample bags, as described in the EXTRUSION PROCESS OPERATING Figure 4: Graphical representation of full RANGES ESTABLISHED factorial experimental design to establish “Standardized Cell Culture Assay” operating ranges of critical extrusion process 3, 4 box ( ). We understood early on that Next, we assessed the impact of parameters (extruder temperature, chill roll the typically conducted USP <87> extrusion process parameters on cell temperature, and output); L, M, and H represent cytotoxicity testing would not be growth behavior within a defined low, middle, and high settings. Center-point 7 predictive of cell growth behavior. window. We also performed an expert (CP) settings were performed in triplicate. ( ) No Slip Agents and Optimized risk assessment of those parameters 4 7 Antioxidant Package: Instead of slip and their likely effects on film quality agents, we applied nontoxic mechanical attributes. 58 antiblocking based on silicon dioxide, The main influencing factor during H which creates a slightly rough surface film extrusion is the amount of heat CP to counteract stickiness. We optimized transferred to process polymer M 9 6 H the antioxidant package and minimized materials. This heat transfer is directly Output the concentration of a pharmacopeia- proportional to the oxidation of L M 23 L listed, trisarylphosphite-based additives and the quantity of L M H processing stabilizer (sold under a degradation products that are Chill roll Extruder temperature temperature number of trade names). That generated to protect the polymer component is known to degrade into a chains. This heat transfer is linked to reported growth inhibitory substance three critical extrusion process cooling temperature of the film (chill (5): bis(2,4-di-tert-butylphenyl) parameters: the melting temperature roll temperature), and the quantity of phosphate (bDtBPP). of resins (extrusion temperature), the material extruded per hour (output).

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 35 We used the best film formulation NO AGING EFFECTS DETECTED months (Figure 6) — whereas the determined in our initial studies and Because degradation of antioxidants competitor’s film results showed an conducted a full multifactorial continues during storage of film rolls, initially poor growth behavior that did experimental design (Figure 4). We after a-irradiation and dry storage of recover after six months of dry examined cell growth using our bags, we investigated the cell-growth storage. That suggests further established cell-growth testing performance of our selected film degradation of bDtBPP into nontoxic approach, demonstrating that the formulation in an accelerated aging compounds. So our data support bag evaluated critical process parameter study and compared the results with reliability from day one until the end (CPP) ranges provide consistent cell those of a competitor’s bag. Most of shelf life — batch-to-batch growth comparable to the control notably, our S80 film formulation consistency, in other words — whereas sample results (Figure 5). ensures excellent growth behavior other vendor bags may show right after a-irradiation until the end deviations depending on storage time. of the aging study — representing 36 NO INTERACTION WITH CULTURE MEDIA Figure 5: Influence of critical film extrusion process parameters on cell growth examined using the established cell-growth testing approach; error bars represent ±SEM (standard error of mean). The This new polyethylene S80 film should numbers correlate to different experimental conditions of the factorial design; 1a, 1b, and 1c work across the full range of represent the center-point setting. (7) bioprocessing applications. Therefore, 125 we were especially interested in understanding cell-growth performance in cold-storage applications with 100 protein-free, chemically defined cell culture media. We considered that to be a worst-case scenario given that 75 serum, hydrolysates, and protein additives can mask detrimental effects of toxic leachates. Gamma-irradiated 50 test bags and borosilicate bottles (control) were filled with chemically

Cell Growth Control) (% defined medium and stored at 2–8 °C 25 for up to six months. No negative effects on cell growth were detected using our cell growth assay (Figure 7) 0 during the storage period. Reference 1a 1b 1c 2 3 4 5 6 7 8 9 Center Point QUALITY BY DESIGN APPROACH TO ESTABLISH FILM SPECIFICATION Figure 6: Cell-growth behavior of γ-irradiated bags stored under accelerated aging conditions (36- Using a quality-by-design (QbD) month representing 337 days at 40 °C); optimized film formulation (S80) showed excellent growth approach, we established a film right after γ irradiation until the end of the study (representing 36 months). The negative control (NC) and a bag from another vendor showed initial poor cell growth that recovered after about six specification that ensures excellent months. (7) and reproducible cell growth, proven 120 by a broad panel of different production cell lines and cell culture media. Together with our strategic 100 collaboration partners in the polymer and film industry, Sartorius Stedim 80 Biotech established full control of the new, state-of-the-art, polyethylene Control 60 multilayer film formulation as well as traceability and control of polymers NC and additives used in it. We applied a 40 multifactorial experimental-design

Cell Growth control) (% S80 approach to determine operating 20 ranges of the film extrusion. And we Vendor X established operating ranges for film- 0 welding parameters and a-irradiation 0 3 6 9 12 15 18 21 24 27 30 33 36 of Flexsafe bags (data not shown). Time (months) Based on this unprecedented material

36 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 7: Cold storage (2–8 °C) of chemically defined cell culture medium in S80 polyethylene film and biological science and process bags; excellent cell growth (comparable to control) was detected over the entire storage period (7). control approach, we can ensure 120 excellent and reproducible growth behavior for the most sensitive cell 100 lines with Sartorius Stedim Biotech’s Flexsafe bags. This has been confirmed by a recently published 80 interlaboratory test (4), in which a number of bioprocessing bags from 60 different vendors were evaluated using a panel of different production cell lines and media. Other authors have 40

demonstrated the suitability of Cell Growth control) (% Flexsafe bags for mesenchymal stem 20 cells (8). Controlling the complete 0 manufacturing process — from resin Control 0.0 0.5 1.0 2.0 3.0 4.5 5.0 6.0 to final bag — ensures consistent and Time (months) reproducible quality for the new generation of Flexsafe bioprocessing bags made with the innovative S80 This unprecendented traceability and 6 Hammond M, et al. A Cytotoxic film. This has been proven by lot-to- control aproach answers key needs of Leachable Compound from Single-Use Bioprocess Equipment That Causes Poor Cell lot consistency in cell-growth the biopharmaceutical industry for Growth Performance. Biotechnol. Progr. 22 performance of this new multilayer single-use factories of the future. January 2014 (epub ahead of print). polyethylene film. Understanding and Sartorius Stedim Biotech’s new S80 7 Jurkiewicz E, et al. Verification of a controlling the film formulation and polyethylene film sets the standard New Biocompatible Single-Use Film manufacturing process are key with respect to cell growth, Formulation with Optimized Additive Content prerequisites to establish meaningful robustness, assurance of supply, and for Multiple Bioprocess Applications. Biotechnol. Progr. 30 May 2014; http://onlinelibrary.wiley. and reliable extractable and leachable batch consistency that are com/doi/10.1002/btpr.1934/full. data as needed for toxicological unprecedented in the industry. 8 Imseng N, et al. Towards a assessment of intermediate and bulk Standardized Biological Test for the drug substance storage and stability ACKNOWLEDGMENTS Identification of Cell Growth Influencing studies. The authors thank Amgen Inc. (USA) for Leachables (poster). ESACT 2013: Lille, All that combined with a robust providing bDtBPP. We also acknowledge Hanni France, 23–26 June 2013. Sun, Oliver Scheibe, and Tanja Frick for 9 assurance of supply concept Cappia JM, et al. Enhanced Assurance of excellent technical assistance. Supply for Single-Use Bags: Based on Material guaranteeing uninterrupted supply of Science, Quality By Design, and Partnership bioprocessing bags (9) is absolutely with Suppliers. BioProcess Int. 12(8) 2014: S43– crucial to pave the way toward fully REFERENCES S46.  single-use bioprocessing facilities of 1 Wood J, Mahajan E, Shiratori M. the future. It ensures that a Strategy for Selecting Disposable Bags for Cell Culture Media Applications Based on a Root- is vice president of reproducible quality of bioprocessing Christel Fenge Cause Investigation. Biotechnol. Progr. 29(6) marketing fermentation technologies, Elke bags can be supplied throughout the 2013: 1535–1549. Jurkiewicz is a senior scientist in upstream entire life cycle of a biologic drug — 2 Gammell P, et al. The Impact of Lot- technology R&D, Ute Husemann is from clinical development, when to-Lot Variability of Disposable Cell Culture manager of upstream technology R&D, single-use bioreactors and bags might Bags on Cell Growth During the Scale-Up of a Thorsten Adams is product manager of Mammalian Production Cell Line. Cell Culture be used for the first time to make or fermentation technologies, Lucy Delaunay Engineering XIII, 22–27 April 2012. store investigational medicinal is an R&D film and material scientist, 3 Steiger N, Eibl R. Interlaboratory Test Gerhard Greller is director of upstream products, to many years after market for Detection of Cytotoxic Leachables Arising launch of the final drug product. technology R&D, and Magali Barbaroux is from Single-Use Bags. Chem. Ing. Tech. 85(1–2) director of film and material R&D at 2013b: 26–28. Sartorius Stedim Biotech, August-Spindler- CONCLUSION 4 Eibl R, et al. Recommendation for Straße 11, DE-37079 Goettingen, Germany. The material science and process Leachable Studies: Standardized Cell Culture understanding applied during Test for the Early Identification of Critical Films for CHO Cell Lines in Chemically development of the new S80 film — Defined Culture Media. Dechema January 2014. enabled by strategic partnerships in 5 Hammond M, et al. Identification of a the polymer industry — were Leachable Compound Detrimental to Cell cornerstones to delivering the new Growth in Single-Use Bioprocess Containers. Flexsafe range of bioprocessing bags. PDA J. Pharm. Sci. Tech. 67(2) 2013: 123–134.

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 37 B IOP ROCESS ROBUSTNESS Robust and Convenient Single-Use Processing The Superior Strength and Flexibility of Flexsafe Bags by Elisabeth Vachette, Christel Fenge, Jean-Marc Cappia, Lucie Delaunay, Gerhard Greller, and Magali Barbaroux

ith the increased use of Figure 1: Film and bag development based on a quality by design (QbD) approach, with expertise in material science, application know-how, film, weld, and bag making disposable bioprocessing bags in all critical process steps of the biopharmaceuticalW drug production, there is a growing requirement for Material Science high-quality, robust, and easy-to- handle bioprocessing bags. The new generation of films and bags must combine multiple mechanical, physical, Application Robustness Film and chemical properties to make these Knowledge Expertise products suitable and scalable for all processing steps in upstream, downstream, and final filling Bag- operations, including cell culture in Making rocking motion and/or stirred-tank, Expertise single-use bioreactors as well as storage, mixing, shipping, and freezing. In partnership with our resin and film suppliers, Sartorius Stedim using model bags and worst-case INDUSTRY DEMANDS Biotech’s polymer and film experts application testing of two-dimensional The need for product quality, robustness, have selected the best raw materials, (2-D) rocking-motion cell culture bags and integrity has been ranked in a optimized the design space of the and three-dimensional (3-D) stirred number of surveys as the primary extrusion process, and defined the bioreactor bags to confirm the bioindustry requirement (1). Because critical welding parameters following robustness of new Flexsafe bags in single-use bags offer multiple technical the principles of quality by design different applications. Notably, our and economic benefits, they are broadly (QbD). With this approach, we have new Flexsafe 3-D shipping bags passed adopted throughout the developed a new polyethylene film the most stringent ASTM D4169 test biopharmaceutical drug industry to help and new bags that combine strength for air and truck shipment. companies achieve rapid and flexible and flexibility, the two prerequisites The thickness, strength, and development and commercial production that will provide for robust and flexibility of S80 film improve its of monoclonal antibodies (MAbs), other convenient single-use bioprocessing. mechanical robustness and make it recombinant proteins, and vaccines. The The robustness of our new S80 suitable for all bioprocessing biotechnology industry is now polyethylene film structure (used in applications. Its strength significantly expanding its implementation of single- manufacturing new Flexsafe bags) has reduces the risk of accidental damage use bags into all bioprocess steps for been demonstrated by means of to bags due to inappropriate handling. applications such as cell culture, storage, standard flex-durability, tensile And the flexibility enables convenient shipping, mixing, freeze and thaw, and strength, elongation, and energy at installation and self-deployment of final filling (2). break testing. Furthermore, we applied each bag in its container. With that growing implementation a water-burst test developed in house of single-use bags, requirements are

38 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 2: QbD approach to ensuring consistent robustness of Flexsafe bioprocessing bags increasing for outstanding product quality, assurance of supply, change control, robustness, ease of use, and 1) Define application targets for the bag: cell culture, storage, shipping, mixing, freezing, and filling scalability. These are important throughout all applications of single- 2) Establish critical quality attributes (CQAs) for the film and bags: use bioprocessing, from cell culture to strength, flexibility, weldability, biocompatibility, purity, stability, and gas barrier shipping of bulk drug substances and final filling of drug products. Film and Resins and Film Film Film and bag bag CQAs additives extrusion welding specifications PARTNERSHIPS WITH RESIN AND FILM SUPPLIERS Formulation Design Space Process Developing innovative film and Development Development Development bioprocessing bags that meet all those Specifications Operating Operating emerging demands requires different and controls range, controls range, controls disciplines and skills and expertise. So the new S80 multilayer polyethylene specifications and controls, definition Further to bag robustness, routine (PE) film from Sartorius Stedim of the film-extrusion process design use in production facilities calls for Biotech has been specified and space, and validation and control of easy, convenient, and straightforward designed in partnership with Südpack, the welding and overall bag-making installation and robust handling a leading film manufacturer, and in process. tolerance. Here, film flexibility and close collaboration with polymer and easy bag deployment during additive suppliers. This collaboration QUALITY BY DESIGN APPROACH installation and filling are the main combines material science, application As described in ICH guidelines Q8, targets. know-how, film-extrusion knowledge, Q9, and Q10 and applied to new The critical quality attributes and welding and bag-making product development, a QbD (CQAs) we identified for providing expertise. approach ensures that developed robustness and ease of use in all The resins in each layer of the S80 products consistently meet intended bioprocessing applications were film and their associated extrusion performance criteria (3–6). Well- strength, flexibility, and weldability of parameters were selected after established and open supplier polymer film. Strength significantly extensive prototyping in which various relationships and knowledge sharing reduces the risk of accidental bag polymers and film compositions were are critical for selecting appropriate damage from inappropriate handling. evaluated to achieve the best raw materials, identifying relevant Flexibility enables easy handling and mechanical properties. Our main specifications, and establishing installation and bag self-deployment objective was to obtain the highest necessary controls to achieve in containers. film robustness: an optimal consistent robustness of the final combination of strength and flexibility single-use bag. MATERIAL SCIENCE with the most reliable and robust Our QbD approach to achieving AND FILM EXPERTISE welds. That in turn enabled us to consistent robustness for Flexsafe bags The superior robustness of single-use design completely new bioprocessing started with an in-depth Flexsafe bags resulted from combining Flexsafe bags suitable and scalable for understanding of the intended different competencies and capabilities all bioprocessing steps in upstream, application and definition of the URS to develop a completely new film that downstream, and final filling. Flexsafe (Figure 2). Today, nearly all unit covers all aspects of their creation: bags meet user-requirement operations in biopharmaceutical drug from the choice of raw materials to specifications (URS) of both rocking development and production can be final manufacturing of bioprocessing motion and stirred-tank single-use performed using single-use solutions. bags. Development of S80 PE film bioreactor applications as well as However, the required mechanical and started with translating target medium, buffer, and intermediate physical properties can vary greatly by application needs into URS, then into storage; shipping, mixing, and application. For example, a bag used functional specifications for the bag, freezing; and aseptic filling of bulk in rocking-motion cell culture or in the film, and the resin and additive drug products. liquid shipping applications requires a package. Development of this innovative very flexible film that resists material Deep understanding of the Flexsafe bioprocessing bag range fatigue. A bag used in large-scale underlying material science and followed a QbD approach to achieve mixing applications or in a stirred- expertise in polymers is required to reliable and robust bioprocessing bags tank bioreactor design must absorb the select the right plastic resin materials suitable for all applications. This significant hydrostatic pressure of and additives that will provide the approach included selection of 2,000–3,000 L of fluid. In such an desired quality attributes. As a first appropriate resins and additives and application, film and weld strengths step, based on theoretical subsequent establishment of are more critical. considerations grounded in material

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 39 Figure 3: Tensile strength data expressed in Figure 4: Elongation at break data of S80 Figure 5: Energy-at-break data for S80 Newtons (N) for S80 and other PE films compared with other PE films; elongation at compared with other PE films break is shown in % of initial length of a film Film D sample at the start of the experiment. Film D

Film C Film D Film C

Film B Film C Film B

Film A Film B Film A

S80 Film A S80 Tensile Strength S80 Energy at Break Elongation at Break Experiment: Once the best film expertise, robustness is designed into structure has been identified, film- the polymer by those who know the break, and so on compared with films extrusion expertise is also critical to plastic materials and how they behave used for other single-use bioprocessing ensuring lot-to-lot consistency of when brought together in a multilayer bags on the market. CQAs and reproducible robustness of a film. Our polymer experts tested and Expressed in Newtons (N), tensile film. This is achieved by identifying selected the plastic materials that strength describes the force required and controlling the critical process confer the best film properties: to break a film. Tensile-strength data parameters (CPPs) of the extrusion strength, flexibility, weldability, are expressed in machine direction process. During development of the biocompatibility, purity, gas-barrier and transversal direction to take into extrusion process, the team explored properties, and relationship with drug consideration the manufacturing process variability in a multivariate stability. This material selection — process influence. This specifically design of experiment (DoE). As a the number and arrangement of plastic characterizes the film strength criteria for robustness, the ultimate layers and their thickness together required for large-volume bags to tensile strength of prototype films was with the extrusion approach of the resist hydrostatic pressure. When measured and shown to be consistent multilayer film — ultimately compared with other PE films, S80 throughout an investigated range of determined the properties and overall film structure shows the best strength parameters evaluated in the DoE performance of the final multilayer (Figure 3). (Figure 6). As for cell growth, we were film structure. Expressed in percentage form, able to define a design space for the In-depth knowledge of film elongation at break is the maximum extrusion process that provides behavior helped us choose the right elongation that a film can withstand assurance of achieving reproducibly the plastic material mix, considering that before breaking. These data represent desired robustness for our S80 film. some of those necessary properties the behavior of a film with regard to The DoE allows for creation of a could be antagonistic. For example, a deformation and resistance to model to predict results depending on strong and stiff film is highly resistant breakage. A high value for elongation the CPP settings. The contour plot to impact and puncture but shows less at break characterizes a highly flexible shows the ultimate tensile strength in flexibility and resistance to fatigue. A film that resists to material fatigue. machine direction (MD), when the too-flexible film will not withstand High resistance to material fatigue is chill-roll parameter is kept constant hydrostatic liquid pressure in large- required, for example, by bags used in but output and extrusion temperatures scale bags. Our film experts designed shipping applications or in rocking- vary within a given range. The several film formulations and motion cell culture applications contour plot illustrates the design combinations and thoroughly tested involving movement of liquid. space in which the predefined them against predefined CQAs. For Compared with other PE films, S80 specification is achieved. the S80 film, the main CQAs for film structure shows the highest value As a result of our QbD approach, characterizing our robustness target of elongation at break (Figure 4). S80 film used in Flexsafe bags is are tensile strength, elongation at Expressed in Joules (J), the energy composed of a PE contact layer and a break, and energy at break (5). at break is the total energy required to coextruded PE-EVOH-PE backbone Film expertise is critical for break a film. This combines the structure (Figure 7). The combined development of robust single-use maximum strength and maximum strength and flexibility of each layer bioprocessing bag solutions. flexibility of the film in one direct with a total thickness of 400 μm Robustness comes from multiple measurement. It can be modeled by provides extraordinary overall properties such as film strength, film integrating the area under a strength/ robustness. The resin-additive flexibility, seal-strength resistance, elongation curve to characterize the formulations for both the contact layer film weldability, and material puncture general robustness of a film. and backbone are completely known or tear resistance. So we characterized Compared with other PE films, S80 and controlled according to S80 film with regard to its tensile film structure shows the best energy specifications. Furthermore, the strength, elongation at break, energy at at break (Figure 5). extrusion process parameters are

40 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 6: Example of a DoE experiment (using MODDE software) on film-extrusion design space with the ultimate tensile strength as a criterion

20

15 H

10 M H Output 5 L M L Chill Roll Temperature

L M H Ultimate Strenth Tensil (MPA) 0 Chil Roll 1 2 3 4 5 6 7 8 9 10 11 Extruder Temperature Temperature Experimental Runs controlled within established ranges. steps in biopharmaceutical Figure 7: PE S80 film structure of the new This ensures long-term lot-to-lot manufacturing has been confirmed Flexsafe bag family Internal layer: LLDPE consistency for all targeted critical against stringent applications such as Core layer: mechanical quality attributes of cell culture, liquid shipping, long-term EVOH Flexsafe bags — specifically, strength, storage, and freezing applications. flexibility, and (hence) robustness. Robustness trials were conducted to test the performance of the product at WELDING AND its limit. As part of our film BAG-MAKING EXPERTISE development, extensive robustness External layer: LLDPE 400-μm total The third critical expertise required trials were performed under worst-case thickness for achieving robustness of Flexsafe application conditions with rocking- single-use bioprocessing bags is in motion bioreactor bags (100-L welding, sealing, and bag-making. The working volume) and stirred bioreactor Figure 8: In-house water-burst test used to challenge is to define appropriate bags (2,000-L working volume). qualify the strength of bags and welds CPPs for consistently achieving the Typically, these bags were filled with defined CQAs with a robust process. water to maximum their working Our bag-making experts have volume. The bags were either rocked optimized the welding, sealing, and at maximum rocking rate or stirred at bag manufacturing process and maximum stirrer rate for 21 days at qualified these parameters based on a 40 °C and maximum operating process risk analysis. Bringing together pressure. Weber et al. further explain the knowledge of welding experts, this approach elsewhere in this special application scientists, and production issue (7). engineers ensures an adequate level of In addition, Flexsafe bags were process understanding of the bag tested for their robustness in air and CONCLUSION assembly and welding process. truck shipping applications by With the increased implementation of CPPs for bag welding and sealing applying the very stringent ASTM single-use bags in biopharmaceutical such as temperature and time have D4169-09 guideline for shipping drug production, there is a growing been considered in our QbD approach. applications (6). The testing sequence requirement for robust and easy-to- Operating ranges of the welding is designed to reflect worst-case handle bioprocessing bags that will be parameters have been tested and are conditions for air or truck shipment suitable and scalable for all processing routinely controlled to ensure process and performed at several temperatures. steps in upstream, downstream, and consistency. To confirm robustness, In addition to the handling test, the final-filling operations. especially of critical welds, we have bags’ truck-transport behavior is tested In partnership with our resin and established a water-burst test in which through investigating horizontal film suppliers, we combined material model bags are filled with water in a impact (Figure 9a) or rotational science and film and bag-making controlled manner. Thus, when shocks. Air-shipping behavior is tested expertise, applying the principles of evaluating sample bags, we found that by exposing a shipper to vibrations QbD to select the best raw materials, typically our bags break at a certain (Figure 9b), compression, or a low- optimize the design space of the overfilling level at the film rather than pressurized environment. extrusion process, and define the the welds (Figure 8). Our new Flexsafe bags successfully critical welding and overall bag passed the most stringent ASTM making parameters. With this PROVEN ROBUSTNESS D4169 test for air and truck shipment approach, our film experts designed The performance robustness of (Figure 9). several film formulations and Flexsafe bags in different bioprocess combinations and were able to achieve

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 41 2 Figure 9A: Horizontal impact ASTM D880-92, Figure 9B: Vehicle vibration ASTM D4728-06, Large P. MCC Fujifilm Billingham. method B method A (air spectrum) Single-Use Workshop, 19 September 2013. Parenteral Drug Association UK Chapter: Uxbridge, UK, www.pda.org/Chapters/ Europe/United-Kingdom.aspx. 3 ICH Q8(R2). Harmonized Tripartite Guideline: Pharmaceutical Development. US Fed. Reg. 71(98) 2009; www.ich.org/fileadmin/ Public_Web_Site/ICH_Products/Guidelines/ Quality/Q8_R1/Step4/Q8_R2_Guideline.pdf. 4 ICH Q9. Harmonized Tripartite Guideline: Quality Risk Management. US Fed. Reg. 71(106) 2006: 32105–32106; www.ich.org/ fileadmin/Public_Web_Site/ICH_Products/ Table 1: Strength and flexibility of film material and welds qualified using multiple methods Guidelines/Quality/Q9/Step4/Q9_Guideline.pdf. Qualification Test Proven Performance 5 ICH Q10: Pharmaceutical Quality Standard flex durability of film High resistance to fatigue and pinhole formation during System. US Fed. Reg. 74(66)2009: 15990– handling, shipping, rocking, and mixing 15991; www.ich.org/fileadmin/Public_Web_ Tensile strength of film and welds High strength of film and welds; high resistance of bags Site/ICH_Products/Guidelines/Quality/Q10/ to breakage under hydrostatic and air pressure Step4/Q10_Guideline.pdf. 6 Elongation and energy at break High flexibility and resistance to fatigue; ease of Gerighausen S, Barbaroux M, Hackel H. installation and use An Approach to Quality and Security of Supply ASTM D-4169-09 Film, weld, and bag robustness; safe and robust air and for Single-Use Bioreactors. Advances in Biochemical truck liquid shipping Engineering/Biotechnology, Volume 138. Springer- Verlag: Berlin, Germany, January 2014. In-house water-burst test Strength of film, welds, and overall bag 7 ASTM D882/ISO 527-3. Standard Test Extensive worst-case testing Robustness for stirred-tank and rocking-motion bioreactor bags and shipping applications Method for Tensile Properties of Thin Plastic Sheeting. ASTM International: West Conshohocken, PA, 2012. a new 400-μm thick film structure motion in cell culture or liquid- 8 ASTM D4169. Standard Practice for composed of a PE contact layer and a shipping applications requires a very Performance Testing of Shipping Containers and coextruded PE-EVOH-PE backbone flexible film that will resist material Systems. ASTM International: West that combine strength and flexibility. fatigue, whereas large-scale mixing Conshohocken, PA, 2009. 9 This optimal combination makes our applications with significant Weber A, et al. Development and Qualification of a Scalable, Disposable new S80 film the most robust and hydrostatic pressure call for strength Bioreactor for GMP-Compliant Cell Culture. suitable for all bioprocessing in both film and welds. BioProcess Int. 12(8) 2014: S47–S52.  applications. The strength That is why we have also applied a significantly reduces the risk of bag water-burst test developed in house is Flexel and Flexsafe damage due to inappropriate handling, and worst-case application testing of Elisabeth Vachette product manager at Sartorius Stedim Biotech and the flexibility provides ease of rocking-motion cell culture and France (elisabeth.vachette@sartorius-stedim. installation and self-deployment of stirred bioreactor bags. We confirmed com). Christel Fenge is vice president of bags in their containers. the extraordinary robustness of marketing for fermentation technologies at The formulation of resins and Flexsafe bags in various applications Sartorius Steidm Biotech Germany. Jean- additives are completely known and and at maximum operating pressure, Marc Cappia is vice president of marketing controlled by specifications, and the working volume, and rocking or stirrer for fluid management technologies at extrusion process and welding rate over 21 days at 40 °C. Sartorius Stedim Biotech France (jean-marc. parameters (e.g., temperature and Finally, Flexsafe bags have been [email protected]). Lucie time) are controlled within established tested for robustness in shipping by Delaunay is R&D project leader at Sartorius process parameter ranges. This both air and truck, applying the very Stedim Biotech France. Gerhard Greller is R&D director of upstream technology at ensures long-term lot-to-lot stringent ASTM D4169-09 guideline Sartorius Stedim Biotech Germany. consistency of strength, flexibility, and (designed to reflect worst-case Magali Barbaroux is R&D director at Sartorius (hence) robustness of Flexsafe bags. conditions for air or truck shipment) at Stedim Biotech France. The superior robustness of Flexsafe several temperatures. bags was first demonstrated using standard flex durability, tensile ACKNOWLEDGMENTS strength, elongation, and energy-at- Thanks to Sartorius Stedim R&D teams in break testing. The S80 film structure Aubagne, France and Göttingen, Germany. shows the best strength, the highest value of elongation at break, and the REFERENCES best energy at break when compared 1 Langer E. Ninth Annual Report and Survey with other PE films. of Biopharmaceutical Capacity and Production. BioPlan Associates, Inc.: Rockville, MD, 2012; Mechanical property requirements www.bioplanassociates.com/images/9th_ largely vary by application. Rocking Annual_2012_ReportTable_of_Contents.pdf.

42 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT B IOP ROCESS SUPPLY CHAIN

Enhanced Assurance of Supply for Single-Use Bags Based on Material Science, Quality By Design, and Partnership with Suppliers by Jean Marc Cappia, Elisabeth Vachette, Carole Langlois, Magali Barbaroux, and Heiko Hackel

rowing adoption of single-use Figure 1: Material science and quality by design (QbD) approach for development of Flexsafe bags bags in commercial production 1) Establish critical quality attributes (CQAs). of biopharmaceutical drugs 2) Validate resin specifications and process parameters raises new challenges for bag that meet CQAs. suppliersG and drives the need for 3) Control resin specifications, process parameters, and product CQAs. consistent product quality, improved 4) Develop a life cycle for continuous process improvement and change control. assurance of supply, robust change management, and business continuity planning. In close collaboration with Film and Resins and Film Film Film and bag resin and film suppliers, polymer bag CQAs additives extrusion welding specification scientists and biologists at Sartorius Formulation Design space Process Stedim Biotech have followed a development development development stringent material science and quality Specifications Operating range Operating range by design (QbD) program to develop a and controls and controls and controls completely new polyethylene film and Continuous process improvement and change control to achieve consistent performance of new Flexsafe bags for all bioprocess steps and applications. rather than simply using resins based development and commercial Reliability of the supply chain for on their trade names provides robust production of biological drug products Flexsafe bags is based on partnerships change control and facilitates change and vaccines. Single-use technologies and agreements with suppliers and our management. Our company can rapidly offer safer, cheaper, faster, smaller, complete understanding and control of qualify equivalent resins sourced from and “greener” biomanufacturing the bag manufacturing process — preferred resin suppliers should one options and meet the main challenges from resin and film extrusion to final supplier discontinue or introduce of the biopharmaceutical industry for sterile bioprocessing bags. Consistent changes to necessary raw materials. capacity adaptation, cost savings, risk product performance is ensured by Business-continuity planning is finally mitigation, and process transfer. For establishing specifications, traceability, addressed through a mix of already smaller biotechnology companies, and control of resins and additives as available redundant equipment and single-use technology also offers a fast well as definition of a design space for manufacturing locations, qualification and economic approach to rapidly film extrusion. of back-up equipment, and implement flexible manufacturing Assurance of supply relies on a long- maintenance of safety stocks for resins capacity (1). term contract and partnership with our and film rolls. The technical and economic film supplier and its large extrusion benefits of using single-use capacity, which allows us to absorb the FACING NEW CHALLENGES technologies are well known, and the projected double-digit growth for Single-use bioprocessing bags are biotechnology industry is broadening single-use solutions. In addition, widely adopted today by established their adoption into all bioprocess steps establishment of resin specifications biopharmaceutical companies for such as cell culture, storage, shipping,

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 43 Figure 2: PE S80 multilayer film structure of mixing, freezing and thawing, final initial raw materials to the final the new Flexsafe bag family (patent pending) 2 filling, and sampling ( ). In the past, bioprocessing bags. Internal layer: LLDPE Core layer: most of the single-use bags were used We selected our resin–additive EVOH for noncritical media and buffer package and processing parameters for storage applications. Those bags did the S80 film after testing several raw not necessarily address fully the more materials and film compositions to critical quality demands of identify one with the best physical and biopharmaceutical process mechanical properties and the most 5, External layer: intermediates, bulk drug substances, consistent cell growth performance ( LLDPE 400-μm total and final drug products. 6). The film specification was thickness The growing use of disposable bags established to enable reproducible in all critical process steps — both for manufacturing of Flexsafe bags that process development and commercial are suitable and scalable for all by combining thickness, strength, and production of high-quality biological upstream, downstream, and aseptic flexibility for each layer of S80 film. drugs — raises new challenges for bag bioprocessing steps, including cell Together with application knowledge, suppliers. Modern single-use solutions culture, storage, shipping, mixing, our resin, film, and bag expertise need improved quality, robustness and freezing, thawing, and filling allowed us to develop a 400-μm thick assurance of supply; more robust applications. Key user requirements structure that offers outstanding change control; and well-established cover, for example, consistent cell robustness of film, welds, and bags for business continuity plans (3). growth, biocompatibility, purity, all bioprocessing applications in robustness, gas barrier properties, upstream production, downstream A PARTNERSHIP, MATERIAL SCIENCE, cleanliness, and sterility of final processing, and final filling. AND QBD APPROACH single-use Flexsafe bags. The S80 film used in Flexsafe bags To meet those new challenges for a The biocompatibility of films is comprises a polyethylene (PE) contact more robust quality of single-use bags, normally assessed using conventional layer and a coextruded backbone Sartorius Stedim Biotech developed a USP chapter <87> and chapter <88> structure of PE, ethylene vinyl alcohol new, proprietary, PE S80 film in in vitro and in vivo biological (EVOH), and PE. The formulation of partnership with Südpack, our film reactivity tests. For the new Flexsafe resins and additives for both the manufacturer. That new film is the bags, a standardized cell-growth assay contact and backbone layers are core component of our new Flexsafe was used at the design stage to completely known, and the extrusion family of single-use bags for all optimize the resin, the additive process parameters are controlled bioprocess applications. Through our package, and the film formulation and within established process-parameter collaboration with Südpack, we have to ensure their compatibility with a ranges. That ensures lot-to-lot established direct contacts, trust, and broad range of cell lines (5). Cell- consistency of all critical quality long-term relationships with selected growth testing also allowed us to attributes (CQAs) of Flexsafe bags: industry-leading suppliers of plastic determine the operating ranges for cell growth, robustness, barrier resins. Direct access to (and the extrusion, welding, and a-irradiation properties, and extractable profiles. openness on the part of) our raw- processes and establish specifications material suppliers provides an and critical process controls (4). ENHANCED QUALITY ASSURANCE unprecedented level of information on Purity of films and bags is AND ASSURANCE OF SUPPLY the initial resin and additive package characterized by their extractable and The entire supply chain for single-use formulation. That is the central part leachable profiles. It is determined by bags is complex and requires material of our new strategy to achieve the nature of the polymer, the amount science and film expertise to achieve complete control over our entire and nature of additives used to enable consistent quality attributes of film; supply chain, from raw materials to processing, and the processing bags; and final, sterile, single-use finished products. Our resin suppliers parameters themselves. With a well- assemblies. support our assurance-of-supply known and well-controlled resin, our First, the key aspect of quality strategy through their open mind-set polymer experts can thoroughly assurance consists in collaboration with and strong understanding about (and characterize the extractable profiles of manufacturers of resins and additive contribution to) risk-analysis and Flexsafe bags and ensure their packages used for extrusion of polymer change-control concepts. reproducibility. End users can be films. Working directly with suppliers The partnership with Südpack, assured that their own initial of raw materials and plastic resins is of open attitude, and long-term extractable and leachable qualification paramount importance in establishing relationship with our resin suppliers work and data will remain valid every specifications, operating ranges, and have allowed us to combine material time they operate single-use process controls and obtaining full science and QbD expertise to achieve bioprocesses using Flexsafe bags. understanding and traceability of the a complete understanding and control The robustness and barrier initial resin and additive package used of our manufacturing process from the properties of Flexsafe bags are obtained for making a film.

44 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 3: Overall supply chain for single-use bioprocessing bags required to sustain an expected yearly Single-Use Supply Chain double-digit growth of the demand for Resin Film Bag Component Assembly Sterilization single-use bioprocessing bags. Here again, complete control of our supply chain and manufacturing processes — from raw materials to final assemblies — is critical to offering both quality Material control Process control Quality control and assurance of supply.

ROBUST CHANGE CONTROL Assurance Partnership, Quality of Change Business Vertical PROCEDURES contracts assurance control continuity integration supply Our long-term contracts with resin and film suppliers involve six months’ notification for changes in raw materials and/or manufacturing Figure 4: Global Sartorius Stedim Biotech supply chain and manufacturing infrastructure for processes of the resin and two years’ single-use bags notification for changes to the film. In ~7,000 m2 of cleanroom capacity with four plants for single-use bioprocessing bags worldwide addition, we have a last-buy time Back-up France, Tunisia, cracker in option of unchanged material for development Back-up France, Puerto Rico, China extruder in Tunisia meeting two years of resin demand. development France, Multiple Tunisia, sources Even more important is Germany, in the Puerto Rico, United establishment of resin specifications and States and instead of using resins by their trade strategic partners European and suppliers Union names. This provides robust resin Resin Film Bag Component Assembly Sterilization change control and facilitates change management. With our established specifications and understanding of the resin’s CQAs, we can rapidly qualify an equivalent resin with our preferred resin suppliers. Releasing Second, robust product quality and relevant extractable profiling, data and controlling resins against necessitates a full understanding of the reporting, and leachable validation specifications — not trade names — complete process, definition of support. Knowing the compounds in enhances the reliability of film quality, operating ranges, and implementation the films helps us interpret our facilitates change management, and of multiple controls to guarantee lot- extractable data very precisely and improves long-term assurance of to-lot performance of films, bags, and accurately. This approach has been supply with fast implementation of single-use bag assemblies. Such an implemented to consolidate the supply alternative resins in case of change or overall knowledge and understanding of our existing S71 ethylene vinyl discontinuation of some raw materials. of the entire process — from resin to acetate (EVA) film used for making And the design space we have final product — can be achieved only Flexboy and Celsius bags and in established for the film-extrusion through strong relationships and development of our new S80 PE film process provides a fast and easily partnerships among single-use bag used for the new Flexsafe bag family. qualified implementation of new manufacturers, raw-material suppliers, Assurance of Supply Through extrusion equipment should a change and film manufacturers. Partnership and Long-Term Supply or a capacity ramp-up be required. Our knowledge about the Agreements with Resin and Film formulation of the resins, additives, and Suppliers: Supply assurance for BUSINESS CONTINUITY PLANNING catalysts as well as the critical process Flexsafe bags is guaranteed first by The terms business continuity planning parameters (CPPs) for film extrusion our long-term supply contracts and assurance of supply often, makes it possible for us to understand established with different resin incorrectly, are used interchangeably. which CPPs influence the CQAs of suppliers and with Südpack, our film Assurance of supply reflects our our final product. This QbD approach manufacturer. We have a 10-year capability as a leading single-use helped us develop more detailed resin supply agreement in place for the S80 supplier to consistently provide specifications for both the contact layer PE film and a last-time buy option for products according to defined and film backbone. a minimum of two years of resin specifications and quality Also, our detailed knowledge demand in case of changes or requirements under normal operating regarding the formulation of discontinuation. We also benefit from conditions. This is the result of our compounds used in manufacturing our Südpack’s large, state-of-the-art, film- long-term contracts, our large films enables us to establish consistent extrusion capacity, which will be manufacturing capacity for all process

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 45 Table 1: Global approach to quality, assurance of supply, change control, and business continuity steps, and the control of our entire Challenges Solutions Results process from raw materials and films Consistent Relationships with resin suppliers Specifications, traceability, and control of to final sterile products. quality Partnership with film manufacturer resin and additive package ensure resin Business continuity planning is lot consistency. required to cover unexpected and Design space and control of extrusion unlikely events such as raw-material parameters ensure film lot consistency. discontinuation or manufacturing Assurance Contracts with resin suppliers 10-year contract with film manufacturer of supply Supply contract with film Last-time buy clause for minimum of two shut-down due to earthquake, fire, manufacturer years of resin demand flood, storms, or other potential Large extrusion capacity aligned with disasters. This is best achieved by growing demand Control of complete resin-to-bag process ensuring that every critical process Change Partnership and contracts with Six-month notification for resin change step can be performed with at least control suppliers Two-year notification for film change two different pieces of equipment and/ Specifications and design space for Resin specification and extrusion design or at two different manufacturing resin space provide for fast implementation of alternative resin/film. locations. When neither redundant Business Back-up resin crackers and film Future redudant resin and film processes process equipment nor an alternative continuity extruder in development Two years of resin and film safety stock production site is available, another Safety stocks of resins and films available alternative is to maintain safety stocks Multiple bag manufacturing Capability to balance capacity between locations and γ-sterilization multiple sites in Europe, Tunisia, Puerto that cover demand for the time that contractors Rico, and China would be necessary to build and qualify alternative equipment or a new production line. contingency planning. This strategy Chapter; www.pda.org/Chapters/Europe/ Sartorius Stedim Biotech is currently has been applied both in development United-Kingdom.aspx. 3 qualifying a backup resin cracker and a of our new S80 polyethylene film and Langer E. Ninth Annual Report and Survey of Biopharmaceutical Capacity and second film extruder at our suppliers Flexsafe bag family and in Production. BioPlan Associates, Inc.: Rockville, and partners to further enhance our strengthening the supply chain of our MD, 2011. business continuity planning. In the existing S71 film made of an EVA 4 Imseng N, et al. Towards a meantime, we maintain at least two contact layer and used for Standardized Biological Test for the years of safety stock for resins and manufacturing Flexboy and Celsius Identification of Cell Growth Influencing films. Multiple bag-making equipment bags. Leachables (poster). ESACT 2013, 23–26 June 2013, Lille, France. and assembly lines are installed in Establishing resin specifications 5 Fenge C, et al. Consistently Superior multiple manufacturing locations in and film-extrusion design space for Cell Growth Achieved with New Polyethylene France, Tunisia, and Puerto Rico to controlling the entire manufacturing Film Formulation. BioProcess Int. 12(8) 2014: ensure business continuity planning for process provide long-term supply S34–S37. our final bag assemblies. We also have assurance, consistent product 6 Vachette E, et al. Robust and contracted multiple gamma-irradiation performance, and robust change Convenient Single-Use Processing with sterilization sources and suppliers control. Flexible business continuity Superior Strength and Flexibility of Flexsafe Bags. BioProcess Int. 12(8) 2014: S38–S42.  worldwide to ensure final sterility planning is guaranteed by establishing assurance. back-up equipment for resin cracking and film extrusion, multiple bag Corresponding author Jean Marc Cappia CONCLUSION manufacturing locations, and safety is vice president of marketing for fluid- Sartorius Stedim Biotech has been stocks of resins and extruded film management technologies (jean-marc. working over the past few years on a rolls. In addition to the current EVA [email protected]); Elisabeth is global product manager for strategy to enhance quality, supply and new PE bags, we also control the Vachette Flexel and Flexsafe (elisabeth.vachette@ assurance, and change control for its manufacture of other critical sartorius-stedim.com); Carole Langlois is single-use bioprocessing bags and components such as filters, connectors, senior global product manager for Flexboy, intelligent single-use solutions tubing, and sensors and assemble ATS, and DDS at; and Magali Barbaroux is consisting of bags, filters, tubes, them into our single-use solutions at R&D director at Sartorius Stedim Biotech connectors, sensors, automation, and multiple manufacturing locations France. Heiko Hackel is vice president of hardware. This strategy is based first worldwide. global sourcing for Sartorius Group at on partnerships and long-term Sartorius Stedim Biotech, August-Spindler- contracts with suppliers, which ensure REFERENCES Straße 11, DE-37079 Goettingen, Germany. control of raw materials and films as 1 Duffy K. Introduction of Technical well as extension of resin and film- Innovations to Improve Sterility Assurance in Vaccines Production. BioProcess International extrusion capacity. Second, it is based European Conference and Exhibition, 18 April on the continuous expansion of our 2012. own manufacturing capacity in 2 Large P. MCC Fujifilm Billingham. multiple sites to allow flexible business Single-Use Workshop, 19 Sept 2013. PDA UK

46 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT B IOP ROCESS UPSTREAM/DOWNSTREAM

Development and Qualification of a Scalable, Disposable Bioreactor for GMP-Compliant Cell Culture by Anne Weber, Davy De Wilde, Sébastien Chaussin, Thorsten Adams, Susanne Gerighausen, Gerhard Greller, and Christel Fenge

uring the development of Figure 1: Phases during product single-use, stirred-tank development of a single-use bioreactor bioreactors (e.g., BIOSTAT STR bioreactors), different User requirement specification phasesD can be distinguished (Figure 1). First, a clear definition of the Single-use bioreactor design intended application and all related requirements should be captured in a Material selection user requirement specification (URS). Based on that, the single-use bioreactor design phase and the Proof-of-concept studies material selection phase are initiated, both closely linked to each other. Qualification of the manufacturing process During the proof-of-concept phase, relevant component- and product- Final product qualification based tests are established and realized Photo 1: BIOSTAT STR 2000 bioreactor to ensure URS compliance. Finally, the qualification can be divided into product and process qualification, in USER REQUIREMENT SPECIFICATION sterilization. Single-use bioreactor set- which the latter includes qualification At the start of BIOSTAT STR up and change-over require a higher of the production equipment bioreactor development, key user level of manual handling. Therefore, procedures and parameters. This requirements are defined covering bag handling must be considered process, together with stringent basic requirements of cell growth and during the development process to change-control procedures, will productivity, robustness, reliability, make it easy-to-use and limit risks for positively influence batch-to-batch and security of supply. Leachables misuse leading to loss of bag integrity. consistency and security of supply. from plastic materials should be Final product qualification should considered to ensure a reproducible SINGLE-USE BIOREACTOR DESIGN include qualifying the product’s biological performance because such Because the characteristics of stirred individual components and the final substances can affect cell growth (1). stainless steel bioreactors are well product using mechanical and Other key requirements are proper understood, this culture vessel design biological tests emulating the intended mixing and oxygen transfer (2). The has been the gold standard for application or a combination of both. stirrer design should allow for superior bioprocessing for decades. Based on Those tests should be performed on mixing while minimizing shear forces. that, the BIOSTAT STR single-use bioreactor bags derived from regular Stainless steel processes typically bioreactor family (Photo 1) was manufacturing. include highly automated cleaning and developed and designed from

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 47 12.5 L to 2,000 L (2). A detailed selection for the individual bag manufacturing, transport, description of the single-use bioreactor components is guided by regulatory installation, and use. Therefore design is given by De Wilde et al. (2). considerations, mechanical properties mechanical properties of the raw The BIOSTAT STR bioreactor is needed for the bioreactor’s intended material (e.g., hardness and tensile equipped with fluorescence-based use, and the need for a biological modulus) must be carefully considered, single-use sensors for pO2 and pH performance comparable with glass not only with regard to component measurement and control. Sensor and stainless steel bioreactors (5). We performance (e.g., tensile strength), patches are installed and sterilized followed guidelines that are applicable but also for component manufacturing together with the cultivation bag. to polymeric materials that come into (e.g., melt flow index). Once the Modern single-use sensor technology contact with final pharmaceutical components are fabricated out of a ensures the same level of control as products (e.g., requirements selected raw material, they have to be traditional sensors but avoids risky established for pharmaceutical evaluated, typically in the final insertions of classical probes (3). The packaging). Guidelines related to the assembly during the proof-of-concept functionality of the probes is tested food and medical-devices industries phase. successfully by Reglin et al. (4) in a also can facilitate assessment of the high–cell-density, 1,000 L STR fed- suitability of a material for bioprocess PROOF-OF-CONCEPT STUDIES AND batch CHO cultivation using applications (5). Table 1 presents PRODUCTION PROCESS QUALIFICATION chemically defined media. Reusable examples of relevant regulations that A key element in biopharmaceutical probes can be inserted as an should be taken into account when development is transfer of a cultivation alternative if desired. qualifying raw materials of polymeric process from laboratory to production nature for single-use bioreactors. scale while ensuring comparable MATERIAL SELECTION The Flexsafe STR bag contains process characteristics (6). Although a A single-use bioreactor bag mainly internal moving components (e.g., the design-space approach can support consists of various components made agitation system) that are subject to successful process transfer (7, 8), of polymer resins. Raw material substantial mechanical forces during scale-up remains a challenge and is “as much an art as a science” (9). Table 1: Examples of relevant guidelines Therefore, extensive process Reference Title understanding and knowledge of the Biocompatibility ISO 10993-5 Biological Evaluation of Medical Devices: bioreactors and equipment in use is a In Vitro Cytotoxicity prerequisite (10). US Pharmacopeia <87> Biological Reactivity Test, In Vitro During the proof-of-concept phase, US Pharmacopeia <88> Biological Reactivity Test, In Vivo the design space with regard to TSE/BSE European Medicine Agency Minimizing the Risk of Transmitting Animal agitation and aeration rate was A410_01_re2 or European Spongiform Encephalopathy Agents via Human determined, and a system Pharmacopoeia 5.2.8 and Veterinary Medicinal Products characterization was conducted, in

Table 2: Example approach to qualify and characterize a film for bioreactor applications

Film Properties Objective Tests Performed Biocompatibility To guarantee a biocompatible film In accordance with the criteria of USP<87> Biological Reactivity Test, In Vitro; USP<88> (including overall materials and Biological Reactivity Test, In Vivo; ISO10993-5 Biological Evaluation of Medical Devices, In manufacturing process) Vitro Cytotoxicity; and in-house cell culture trials

Robustness Evaluation and qualification of During a single-use bioreactor application the film is exposed to puncture, traction, mechanical properties required flexion, and fatigue (Photo 2). In accordance with ISO527-3 or ASTM D 882, tensile test during the applications allows characterizing the tensile properties of the film and the ability to resist to the elongation. In accordance with ASTM D 3787 (an adaption), ball bursting allows characterizing the ability to resist puncturing of nonsharp items through the film. And in accordance with ASTM F 392 condition C, flex durability test allows characterizing the ability of the film to resist to flexion and fatigue in flexion.

Interaction Evaluation of physical–chemical In accordance with the criteria of USP<661> for aqueous extraction for nonvolatile properties residue, residue of ignition, heavy metals, and buffering capacity

To guarantee compliance In accordance with the criteria of European Pharmacopoeia 3.1.5: for polyethylene with of the contact layer composition additives for containers for parenteral preparations and for ophthalmic preparation with the European Pharmacopoeia and to evaluate physical–chemical properties of the contact layer Characterization of the In accordance with a Sartorius model: Identification and semiquantification of volatile extractables and leachables profile compounds, semivolatiles compounds, nonvolatiles compounds and metal elements of film including an evaluation of extracted/leached from the film; representative solvents are used to create the interaction with the properties of extractable guide; characterization of the pH shift, conductivity, total organic carbon water for injection (TOC) and nonvolatile residue (NVR) of water for injection stored in bags made with the film

48 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Photo 2: Tensile and flex durability tests (photos left and middle from Reference 19)

Figure 2: Water burst test with a filled bag chamber

Digital Manometer

Water Source Semiautomated Bench

which typical scale-up parameters components in direct contact with materials and assemblies during the were evaluated (11). De Wilde et al. process fluids should apply a quality performance qualification at the describe a detailed characterization system meeting international application level. Therefore, a approach (2). And Reglin et al. (4) standards such as ISO 9001, ISO combined approach of qualifying describe a successful transfer of a 13485, or CFR 21 Part 820 (5). individual components as well as high–cell-density CHO fed-batch Finally, specific design parameters qualification at application level of the process to a Flexsafe STR 1000 (e.g., critical dimensions and final product is considered the most bioreactor finalizing the proof- component delivery times) are meaningful and effective way to of-concept phase. established for all critical parts. qualify such a complex system (2). After a successful proof of concept, the final design was transferred into FINAL QUALIFICATION APPROACH SELECTING AND QUALIFYING THE production, including operator To perform a proper qualification, it is PROPER FILM PROPERTIES training and qualification of the not sufficient to focus only on the Films used in bioprocess applications production equipment and procedures. qualification of the components and commonly possess a multilayer Production parameters were the manufacturing process. structure. Currently marketed films monitored, and requalification of Qualification of the final product include a backbone and a contact layer. them was performed in regular under relevant conditions for the The backbone provides the barrier intervals. A stringent change-control intended use is absolutely critical. structure and consists of one or more process ensures consistent quality, By contrast with pharmaceutical layers, which determine the overall performance, and reliability. plastic packaging, no regulatory mechanical behavior and barrier As part of ensuring a high level of guidelines have been specifically properties. The contact layer is the supply assurance, current regulatory developed for the qualification of surface that comes into contact with guidelines and standards for quality single-use bioreactors. Therefore, the culture broth. It must be inert and systems such as ISO 9001 (12) were qualification approach of STR bags is have good sealing characteristics (5). followed. These guidelines instruct based on general good manufacturing Examples for properties typically manufactures of single-use bioreactors practice (GMP) principles and tested in accordance with regulations to control components and services relevant parts of existing guidelines applicable to biopharmaceutical obtained from subsuppliers. During associated with other products. For packaging are shown in Table 2. Some supplier qualification, their quality the component-based qualification of those tests are defined to qualify system, their experience in offering approach, test procedures for the film with pass–fail criteria. Others materials for pharmaceutical or individual components are are performed to characterize the film, medical applications, and the suppliers implemented. The qualification without specifications. reliability are critically reviewed. approach for the entire final product is The influence of substances leached Suppliers of raw materials for based on the verification of the from the film on the growth and

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 49 Figure 3: Pressure plot obtained during a robustness trial of a Flexsafe STR 1000, including final aeration and pressure stress test bursting — caused by a breakage of 120 the film rather than at the welds — of the model S80 bag (14). Comparing 100 these results with those obtained from 80 other commercially available films 60 (not shown) confirms superior robustness of the new S80 40 polyethylene film used for the Pressure (mbar) Pressure 20 manufacturing of Flexsafe bags. For the conversion of the STR bag family 0 510 520 530 to the new S80 film, each weld of the Time (h) single-use bioreactor chamber was subject to tensile strengths, color invasion, and water-burst testing. The Table 3: Assessment of the robustness trial color invasion test allows identification Step General Acceptance Criteria of the presence of microchannels in a Before Unpacking No damages to the packaging weld. Different film lots were used to installation Visual inspection of No film or component defect the bag No damages to welds, inner components, film, ports conduct those tests. Installation of the bag in No deviations to the steps in the BIOSTAT STR manual the bag holder SUCCESSFUL REALIZATION OF THE Final visual Agitation trial No bag damage FINAL BIOREACTOR PRODUCT inspection Aeration trial No internal component/stirrer defect as per visual Ensuring State-of-the-Art Bioreactor inspection from outside the bag Bag Robustness: Qualification of the Functionality of the agitation system Functionality of the aeration system STR bags with S80 film was based on No external component defect a worst-case rationale relevant to the final application. Because bigger bags are exposed to larger mechanical parameters settings of the design forces, they can be considered as a space worst case comparison for smaller  • guarantee film quality and sizes. Three film lots were used for reproducibility the final qualification approach, and a Because bigger bags • guarantee that film properties representative bag quantity was tested. are exposed to larger reach the acceptance criteria when Additional robustness trials will be films are produced within the design performed routinely on bags from MECHANICAL space. regular production to confirm the forces, they can be The properties of the film were results of the initial qualification. considered as a also verified after aging it three years The parameters of the robustness worst-case to establish the film’s shelf life. Both trial were chosen based on the typical accelerated aging regulated by ASTM characteristics of a mammalian comparison for F 1980 (13) and natural aging cell-culture process. In the 21-day smaller sizes. conditions were considered to qualify trial, a maximum filling volume, the three-year shelf life of the film. maximum stirrer speed, 40 °C, and 30 mbar overpressure were applied. QUALIFICATION OF ASSEMBLIES Additionally, an aeration test and a The various connections between pressure stress test were conducted productivity of mammalian cells has individual components of the single- following the 21-day period. been identified as one of the major use bioreactor were qualified to During robustness qualification concerns related to the use of single- demonstrate physical and mechanical and routine testing, a pressure of up to use bioreactors. Fenge et al. (1) integrity of the final bag assembly. To 50 mbar was evaluated. Additionally, a summarize the approach to investigate demonstrate reliability, reproducibility, worst-case pressure stress test on a biocompatibility of the new S80 and robustness, seal strength and selected number of single-use polyethylene film used in the Flexsafe integrity were qualified. Robustness bioreactor bags was performed. That product range. was qualified both on final-product consisted of increasing the pressure A design space (process window) level and on component level. from initially 30 mbar to 110 mbar in for critical process parameters of film A water-burst test (Figure 2) was 10-mbar steps (Figure 3). Every extrusion was defined to developed to evaluate the robustness pressure-step increase was maintained • guarantee extrusion process of critical welds. A certain nominal constant for one hour. Before and after reproducibility within established volume could be reached before the robustness trial, a visual inspection

50 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Table 4: Stability of water for injection in accordance with the European and US pharmacopoeias robustness trial, comparable to a Reference Title Total organic carbon (TOC) US Pharmacopeia <643>, European/US Total Organic Carbon worst-case cell-culture application. In Pharmacopoeia Monograph for “Sterile accordance with US Pharmacopeia Water for Injections” <85>, endotoxin values were Conductivity US Pharmacopeia <645>; European/US Conductivity determined. Pharmacopoeia Monograph for “Sterile Water for Injections” Extraction with Water and Ethanol: Extractables can directly influence the pH US Pharmacopeia <791>; European/US pH Determination Pharmacopoea Monograph for “Sterile quality of a final product. Therefore, Water for Injections” in each pharmaceutical production Ions and oxidizables European/US Pharmacopoeia Monograph Stability of Sterile step, potential contamination with for “Sterile Water for Injections” Water for Injection extractables has to be assessed. Nonvolatile residue (NVR) US Pharmacopeia <661>; European/US Polyethylene Detailed knowledge of the chemical Pharmacopoeia Monograph for “Sterile Container Water for Injections” nature of extractables is essential even for trace amounts. Particulate matter US Pharmacopeia <788> and European Nonvisible Particles Pharmacopoeia 2.9.19 The analytical scheme was developed in line with the concept that a bag is made up of a large of the whole bag, seams, and the protection against mechanical number of different polymeric agitation system was performed. Table handling (impacts and drops), vehicle materials. Potential extractables could 3 lists the defined pass criteria for the vibration (truck and aircraft cargo), consist of respective breakdown robustness trials. and stacking. The Flexsafe STR products, residual monomers, and Sterilization Validation to Ensure 1,000-L unit was tested and was taken oligomers from manufacturing Risk-Free Cell Cultivation: The as the representative for the STR residues and additives of the different Flexsafe STR system is sterilized by 500-L unit because of their nearly polymers. Such a broad spectrum of gamma irradiation. The first step in identical packaging dimensions. The substances represents a complex the validation of a gamma sterilization Flexsafe STR 50-L and 200-L models analytical challenge that can be cycle with a security assurance level were tested because they differ in managed only by combining different (SAL) of 10–6 described by the ISO terms of unit per pallet and height. analytical methods. Typically, various 11137 (15) is assessment of the final Furthermore, the Flexsafe STR mass spectrometry methods are used bag product’s bioburden load. 2000-L unit was successfully tested as and extractables are investigated, with Determination of the bioburden of a worst-case configuration in terms of the amount and nature determined. the Flexsafe STR family is based on height and weight. For developing the analytical an in-house method in accordance to Stability of Water for Injection: scheme, the two most commonly used ISO 11737 (Sterilization of Medical Apart from the mentioned solvents (reverse osmosis water and Devices: Microbiological Methods, Part 1 qualification approaches for robustness absolute ethanol) were chosen to — Determination of a Population of and sterility, leachable data are create a database covering all Microorganisms on Products) (16). provided. The stability for water for extractables. For this study, a rational A dose-mapping study of the injection is specified by the approach was used that considered the Flexsafe STR product family was monographs of the US and European different components of the Flexsafe performed according to ISO pharmacopoeias. Tests were applied to STR family and a worst-case ratio of regulations (15) and based on a gamma-irradiated Flexsafe STR bags film surface to volume. The extraction worst-case rational. Because it is (Table 4) and documented in a conditions (40 °C for 21 days) are directly linked to product density, validation guide. Results from the considered to be comparable to a including its packaging, we defined stability of water for injection are not typical mammalian cell-culture the packaging and performed a considered as release criteria but serve process. transportation test (described below) as comparable data for the Long-Term Flexsafe Bag Stability: before dose mapping. The verification pharmaceutical industry because Shelf-life qualification of the Flexsafe dose is calculated to achieve SAL 10–1 mammalian cell-culture media are STR bags is comparable to the according to ISO 11137 (15). used during the application. verification of STR bags made of S40 Individual sterility testing was For the stability of water- film (18). Based on a risk assessment, performed on a specific number of for-injection trials, three subfamilies the STR 2,000-L, 1,000-L, and STR bags. are defined, considering the 200-L models were chosen to Packaging Validation to Ensure Full construction of the agitation system. determine robustness, stability of Robustness in a World without Three bags per subfamily from three water for injection, and functionality Boundaries: Packaging was qualified different film lots were tested. The of the optical sensors after accelerated by a transportation test simulation in conditions were 40 °C for 21 days ageing in accordance to ASTM F accordance to ASTM D 4169-DC 2 with identical working volume, 1980 (13) and natural ageing after two (17) to assure robustness and pressure, and agitation speed as for the years. Storage conditions for

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 51 accelerated aging are 225 days at bioreactor. That bioreactor is not only 13 ASTM F1980:2007 Standard Guide for 40 °C and 75% humidity. This highly comparable with existing Accelerated Aging of Sterile Barrier Systems for Medical Devices. American Society for Testing correlates to two years natural aging. stainless-steel bioreactors, but also and Materials: West Conshohocken, PA, 2007. meets industry requirements with 14 Vachette E, et al. Robust and CONCLUSIONS regard to quality, robustness, and Convenient Single-Use Processing with Unlike other single-use bioreactor assurance of supply. Superior Strength and Flexibility of Flexsave designs available on the market, the Bags. BioProcess Int. 12(8) 2014: S38–S42. BIOSTAT STR bioreactor family is ACKNOWLEDGMENT 15 ISO 11137:2006 Sterilization of Health unique in its design and performance Thanks to the R&D team in Aubagne and Care Products. International Organization for Standardization: Geneva, Switzerland, 2006. because its design is similar to that of Göttingen. 16 ISO 11737-1:2006 Sterilization of conventional stainless steel stirred Medical Devices: Microbial Methods, Part 1 — tank bioreactors. For development and REFERENCES Determination of a Population of Microorganism qualification, specific tests were 1 Christel F, et al. Consistently Superior on Products. International Organization for selected or developed to ensure full Cell Growth Achieved with New Polyethylene Standardization: Geneva, Switzerland, 2006. Film Formulation. BioProcess Int. 12(8) 2014: 17 ASTM D4169:2009 Standard Practice robustness, reliability, and lot-to-lot S34–S37. consistency. Robustness, cell culture for Performance Testing of Shipping 2 De Wilde D, et al. Enhanced Scalability Containers and Systems. American Society for performance, and quality of this in Single-Use Bioreactors. BioProcess Int. 12(8) Testing and Materials: West Conshohocken, disposable bioreactor family are based 2014: S14–S19. PA, 2009. primarily on the selection of suitable 3 Noack U, et al. Single-Use Stirred Tank 18 Weber A, et al. Development and polymer materials and a well-defined Reactor BIOSTAT® CultiBag STR: Qualification of a Scalable, Disposable and controlled production process to Characterization and Applications. Single-Use Bioreactor for GMP-Compliant Cell Culture Technology in Biopharmaceutical Manufacture. ensure that critical quality attributes BioProcess Int. 11(4) 2013: S6–S17. Eibl R, Eibl D, Eds. John Wiley & Sons, Inc: 19 are consistently met. Furthermore, an Teniers C. How Laminates with New York, NY, 2011; 226–2404. EVAL™ EVOH Film Improve the intelligent, application-driven 4 Reglin R, et al. Validation New Flexsafe Performance of VIPS; http://oisd.brookes.ac. qualification approach both on Film in a CHO Fed Batch MAb Process in the uk/ivisnet/resources/papers/3A_Cynthia component level as well as on final BIOSTAT STR 1000L. BioProcess Int. 12(8) Teniers - revised copy.pdf.  product level was established that 2014: S53–S57. 5 emulates the intended application of Barbaroux M, Gerighausen S, Hackel H. An Approach to Quality and Security of Anne Weber is a scientist, upstream the advanced single-use bioreactor Supply for Single-Use Bioreactors. Adv. technology, R&D; corresponding author platform. This ensures a holistic Biochem. Eng. Biotechnol. 138, 2014: 239–272. Davy De Wilde is director of marketing, qualification approach starting from 6 Vorlop J, Lehmann J. Scale-Up of fermentation technologies (Davy.deWilde@ the raw materials to the final Bioreactors for Fermentation of Mammalian Sartorius-Stedim.com); Sébastien single-use bioprocessing bag product Cell Cultures, with Special Reference to Chaussin is R&D program leader, fluid in its intended application. Oxygen Supply and Microcarrier Mixing. management technology; Thorsten Chem. Eng. Technol. 11(1) 1988: 171–178. Adams is product manager, fermentation Full functionality and consistent 7 Xing Z, et al. Scale-Up Analysis for a technologies; Susanne Gerighausen is and reproducible performance could CHO Cell Culture Process in Large-Scale director of quality; Gerhard Greller is R&D be confirmed, and biological Bioreactors. Biotechnol Bioeng. 103(4) 2009: compatibility was proven by 733–746. director, upstream technology; and Christel Fenge is vice president of marketing, mammalian cell-culture trials (1, 4). 8 Varley J, Birch J. Reactor Design for fermentation technologies at Sartorius Superior and consistent robustness was Large-Scale Suspension Animal Cell Culture. Stedim Biotech. confirmed in worst-case application Cytotechnol. 29(3) 1999: 177–205. 9 trials and pressure tests. All relevant Ma N, Chalmers JJ, Mollet M. Aeration, Mixing, and Hydrodynamics in and recommended tests according to Bioreactors. Ozturk SS, Hu WS, Ed. Cell US and European pharmacopoeial Culture Technology for Pharmaceutical and Cell- monographs and an extractable study Based Therapies. CRC Press, New York, NY, were conducted and passed 2006; 225–248. successfully. Finally, qualification of 10 Catapano G, et al. Bioreactor Design and the BIOSTAT STR single-use Scale-Up. Cell and Tissue Reaction Engineering. Eibl R, et al., Eds. Springer, Berlin-Heidelberg: bioreactor family confirmed that the Berlin, Germany, 2009; 173–262. user-requirement specification (which 11 Löffelholz C, et al. Dynamic Single- was defined at the starting point of Use Bioreactors Used in Modern Liter and m³ the product development) was Scale Biotechnological Processes: Engineering consistently met. Characteristics and Scaling-Up. Adv. Biochem. The risk-based approach of testing Eng. Biotechnol. 138, 2014: 1–44. 12 at component level and on final- ISO 9001:2008 Quality Management Systems — Requirements. International product level, in the intended Organization for Standardization: Geneva, application, ensured the successful Switzerland, 2008. development of a robust single-use

52 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT B IOP ROCESSUPSTREAM/PRODUCTION

Verification of New Flexsafe STR Single-Use Bioreactor Bags Using a CHO Fed-Batch Monoclonal Antibody Production Process at 1,000-L Scale by Regina Reglin, Sebastian Ruhl, Jörg Weyand, Davy De Wilde, Ute Husemann, Gerhard Greller, and Christel Fenge

n the past decade, single-use Figure 1: Seed train of the 1,000-L fed-batch cell culture run starts with one cryogenic vial before bioreactors have gained wide six consecutive cell-expansion steps using single-use shaker flasks and bioreactors. A 17-day fed- acceptance for biomanufacturing. batch production process followed in a BIOSTAT STR 1000 system. The entire duration from The biopharmaceutical industry is cryogenic vial to 1,000-L harvest on day 17 took 35 days. Iincreasingly interested in performing Main Step 6 Culture modern production processes in (n – 1) single-use facilities. That trend is Seed Culture driven by the time and cost benefits of Steps 1–3 Steps 4 and 5 (n – 6, n – 5, n – 4) (n – 3, n – 2) single-use technologies, as well as the Cryoculture enhanced manufacturing flexibility they offer (1). With single-use bioreactors increasingly used in late-phase clinical 1-mL 0.5-L and 1-L 10-L and 50-L 200-L 1,000-L cryogenic shake flasks BIOSTAT RM 20/50 BIOSTAT STR BIOSTAT STR trials and commercial production, vials optical system system system their quality, reliability, and assurance of supply becomes more critical. Many industry experts consider process film (“S80”) specifically optimized for established process parameter ranges. control of film and bag manufacturing such applications (2, 3). This This approach based on specifications and traceability of raw materials to be collaborative concept enabled us to and controls prevents detrimental key to ensuring batch-to-batch develop a completely new range of effects on cell growth that can be consistency with single-use bioprocessing bags that meet industry caused by extractables released by film bioprocessing bags. Especially crucial needs for future biomanufacturing: materials during bag manufacture and for such bags is excellent and consistent cell-growth and extractable storage (3). Measured by a Chinese consistent cell-growth performance for profiles, superior robustness and ease of hamster ovary (CHO) cell–based cell- a broad range of different use, unprecedented assurance of supply, growth assay developed by Sartorius commercially relevant cell lines. and suitability for all applications from Stedim Biotech, the new film has Most commercially available film upstream production to downstream demonstrated superior biological materials used for single-use bioreactors processing and final filling (3–5). performance in accelerated aging today were not specifically developed The S80 polyethylene film used in studies and media-storage trials (3). for cell culture applications. Therefore, Flexsafe bags has an optimized Similarly excellent results were Sartorius Stedim Biotech partnered additive package controlled by recently published by the “Single-Use with Südpack (Europe’s leading film- specifications rather than trade names Technology in Biopharmaceutical extrusion company) and closely of resins and additives. Film Manufacturing” temporary working collaborated with resin and additive extrusion, bag assembly, and group (4). In their study, nine different suppliers to develop a new polyethylene a-irradiation are all controlled within cell lines and related media were used

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 53 Table 1: Cultivation conditions of the 1,000-L production process comprising set points for GmbH to produce an IgG1-type MAb

pH, pO2, and temperature and including in a high–cell-density fed-batch process. ranges of gas flow rates and stirrer speeds as  Final MAb titers >7 g/L are typically part of the pO control cascade 2 This fed-batch reached with this cell line. We fed Parameter Set Point/ Range process is considered cultures using the chemically defined Temperature 36.8 °C ActiCHO cell culture media system to be a suitable pH 7.15 developed by Cellca GmbH: ActiCHO system for SM for cell expansion and ActiCHO P pO 60% MODEL 2 process verification for production, with ActiCHO Feed-A, Stirrer speed 65–85 rpm ActiCHO Feed-B, and a 400-g/L N 0–20 Lpm of Flexsafe STR bags 2 d-glucose solution from Sigma Aldrich. Air flow 0–15 Lpm regarding cell The seed train for the 1,000-L fed-

O2 0–60 Lpm growth, productivity, batch production process consists of six and robustness consecutive cell-expansion steps in typical (Figure 1) starting with cryogenic to evaluate different single-use vials. The inoculum cell density of all bioprocessing bags from different biopharmaceutical seed train steps was 0.2 × 106 cells/mL. vendors with regard to their ability to applications. For the first three expansion steps, we support cell growth. used disposable shaker flasks (n – 6, In addition to biocompatibility, the n – 5, and n – 4) incubated in a robustness of Flexsafe bags has been CERTOMAT CTplus CO2 incubation confirmed through a number of shaker. We further expanded the cells different test methods, including an bioreactor. This fed-batch process is in 5-L (n – 3) and 25-L (n – 2) in-house water-burst test and worst-case considered to be a suitable model working volumes using a BIOSTAT application trials (2, 5). Finally, system for process verification of RM rocking-motion bioreactor. The functionality and consistent, Flexsafe STR bags regarding cell last seed expansion step — at 200 L in reproducible performance of these growth, productivity, and robustness a BIOSTAT STR 200 (n – 1) system Flexsafe bags has been qualified in typical customer cell culture — generated the required inoculum following industry accepted methods (2). applications because such processes volume of ~100 L for the large-scale In the study reported below, we using CHO cells are widely used in run. All single-use bioreactors used in discuss the performance of Flexsafe the biopharmaceutical industry (7). our study were equipped with Flexsafe STR bags in a high–cell-density fed- The recombinant CHO cell line we bags. batch culture of recombinant CHO used is the same one involved in the The 1,000-L production process cells producing a monoclonal antibody cell-growth assay our company started with a three-day batch phase (MAb). Fed-batch cultures ran in established to support development of followed by a 14-day fed-batch phase. 50-L and 1,000-L BIOSTAT STR the new polyethylene film (8). We inoculated a BIOSTAT STR 1000 single-use, stirred-tank bioreactors for production bioreactor at an initial cell 17 days, with the results compared PROCESS VERIFICATION density of 0.3 × 106 cells/mL. Table 1 with those obtained using a 5-L We applied a recombinant CHO lists the cultivation conditions set for BIOSTAT B stirred-tank glass DG44 DHFR– cell line from Cellca the production bioreactor. We used

Figure 2: Viable cell density (VCD) and viability of high–cell-density fed- Figure 3: Monoclonal antibody (MAb) titer from high–cell-density fed- batch processes in a BIOSTAT B 5-L glass vessel and BIOSTAT STR 50 and batch processes in a BIOSTAT B 5-L glass vessel and BIOSTAT STR 50 and BIOSTAT STR 1000 Flexsafe bags; 5-L data include a standard deviation of BIOSTAT STR 1000 Flexsafe bags; 5-L data include a standard deviation of cell density and viability for a total of four runs. product titer for a total of four runs.

100 35 10

30 BIOSTAT B 5 (n = 4) 80 8 BIOSTAT STR 50 25 BIOSTAT STR 1000 cells/mL) 6 60 6 20

40 15 4 Viability BIOSTAT B 5 (n = 4) Viability (%) IgG Titer (g/L) Titer IgG 10 VCD BIOSTAT B 5 (n = 4) Viability BIOSTAT STR 50 2 20 VCD BIOSTAT STR 50 5 Viability BIOSTAT STR 1000 VCD BIOSTAT STR 1000 0 0 Viable Cell Density(×10 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Time (days) Time (days)

54 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Figure 4: Quality of IgG1 produced during 17 days of high–cell-density Figure 5: Glucose, lactate, and osmolality profiles of high–cell-density fed-batch culture at 1,000-L scale assessd using sodium-dodecyl sulfate fed-batch processes in BIOSTAT STR 50 and BIOSTAT STR 1000 systems polyacrylamide gel electrophoresis (SDS PAGE) under reduced conditions and Coomassie blue staining; in addition to samples from the process Glucose BIOSTAT STR 50 Glucose BIOSTAT STR 1000 start and days 5, 7, 9, 11, 13, 15, and 17 post-inoculum, the graph shows a Lactate BIOSTAT STR 50 Lactate BIOSTAT STR 1000 Osmolality BIOSTAT STR 50 Osmolality BIOSTAT STR 1000 marker (T851 from Roth) and control (IgG from Bayer). 500 4.0 10

Sample (d) M C 0 5 7 9 11 13 15 17 3.5 Concentration (g/L) 10.0 0 0.4 1.1 2.5 4.1 6.2 7.0 7.8 450 8 3.0 Feed start

2.5 400 6 212 2.0 66 43 Heavy 350 1.5 4 chain 1.0 300 2 29 0.5 Light Osmolality (mOsmol/kg) Lactate Concentration (g/L) Glucose (g/L) Concentration 20 chain 250 0.0 0 Protein Size (kb) 14 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 Time (days) M = protein marker (T851, Roth); C = control (IgG, Bayer) two three-blade segment impellers and mL at >97% viability (measured by a a ring sparger based on a previously Cedex HiRes cell-counting system performed characterization study on  from Roche Diagnostics) on day 9 the bioreactor system, which suggests post inoculum. Growth and viability sufficient oxygen transfer using that Our focus was to profiles follow each other very closely set-up (9). As Table 1 shows, the demonstrate not at all scales. Notably, cell growth data system controlled pO in a cascade- from the BIOSTAT STR 50 and 2 only that the film control approach involving stirrer 1000 were within the standard delivers excellent speed and sparging with nitrogen (N2), deviation of the data from 5-L air, and oxygen (O2). Carbon dioxide growth results, but reference runs using conventional glass (CO2) sparging controlled pH levels. also that the stirred-tank vessels. Hence, we bag observed no cell-growth–inhibiting SUPERIOR BIOLOGICAL ENTIRE effects with Flexsafe bags in the large- PERFORMANCE AT ALL SCALES assembly performs scale, stirred-tank format, which The main aim of our study was to well in a true and demonstrated excellent comparability confirm the suitability and representative and scalability of the bioreactor performance of Flexsafe bags for their process application. systems. final process application in both Accelerated aging studies of sample rocking-motion and stirred-tank bags showed no impact on biological formats up to the 1,000-L scale. performance of the S80 film Previous studies had already formulation during an investigated confirmed that the S80 polyethylene 36-month shelf-life period (3). Our film material ensures superior and densities (VCDs) of ~2 × 106 cells/mL study involved Flexsafe RM and STR reproducible cell growth based on with viabilities >99% were reached bags that had been stored between one medium extraction trials using test (data not shown). and 12 months. The consistent and bags (3). So our focus was to Figure 2 compares cell growth and superior cell growth obtained with the demonstrate not only that the film viability data of the high–cell-density different scales and bioreactors further delivers excellent growth results, but fed-batch run at 1,000-L scale with confirms those accelerated aging data also that the entire bag assembly data from 50-L scale using the single- using sample bags. performs well in a true and use BIOSTAT STR bioreactor set-up. In addition to cell growth and representative process application. Our large-scale, single-use, bag-based viability, product formation (in this During seed expansion (n – 3 to n – 1) bioreactors achieved excellent case an IgG1 MAb) is especially using the BIOSTAT RM 10 (n – 3) comparability in both runs. Those critical to establishing comparability of and 50 (n – 2) as well as in the n – 1 results are further supported by different cell culture systems. Figure 3 BIOSTAT STR 200 seed bioreactor, excellent comparability seen with shows that, at all investigated scales, we observed typical growth behavior small-scale reference data generated IgG1 titers were highly comparable that was comparable to what we have during four 5-L cultures in BIOSTAT and within the standard deviation of seen in cell expansions using shaker B glass vessels. At 5-L, 50-L, and the 5-L glass vessel results. A final flasks. At the end of the three-day 1,000-L scales, all cultures achieved MAb titer of 7.8 g/L was achieved batch growth phase, viable cell peak cell densities of ~28 × 106 cells/ with the high–cell-density fed-batch

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 55 Figure 6: Figure 7: pO2 profile obtained in a high–cell-density fed-batch process Comparing pH profiles of single-use optochemical probe (pH in a BIOSTAT STR 1000 system; in addition to pO2 control performance, on-line optical) and conventional glass pH probe (pH on-line classical) the graph shows stirrer rate, air flow, and nitrogen (N2) and pure oxygen with off-line pH measurement (pH off-line) for a 1,000-L high–cell- (O2) gas flow rates (cascade control strategy). density fed-batch culture in the BIOSTAT STR 1000 system; further, CO2 gas flow rates and off-line values of pCO 100 100 80 16 20 2 90 70 14 150 5.0 8.0 pH on-lineoptical pH on-lineclassical pH off-line 80 80 16 4.5 7.8 125 60 12 4.0 7.6 70 3.5 7.4 10 100 60 50 60 12 3.0 7.2

(%) 50 8 2 40 75 2.5 7.0 2.0 6.8 pO 40 40 8 Flow (Lpm) pH Value pH Flow (Lpm) 2 30 Flow (Lpm) 6 (mmHg) Value

2 50 2 2 1.5 6.6 O N Air Flow (Lpm)

30 CO 4 1.0 6.4 Stirrer Speed (rpm) Speed Stirrer 20

pCO 25 20 20 4 0.5 6.2 2 10 10 0 0.0 6.0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 0 0 0 0 0 0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 Time (days) Time (days) run at 1,000 L, which is comparable to Adding antifoam and changing from 8.2 g/L achieved at the 50-L scale. air to oxygen flow control caused

Specific production rates ranged from minor variations in the pO2 value, 11 ng/cell on day 1 to 45 ng/cell on day  especially at very low oxygen flow 13. We used reduced sodium-dodecyl Both our batch data rates (Figure 6). We successfully sulfate polyacrylamide gel and our fed-batch controlled the pH value to 7.15 during electrophoresis (SDS-PAGE) to the entire run (Figure 7) using signals analyze product purity and potential data are highly from a single-use optochemical pH degradation at different culture days of comparable across probe. During the fed-batch phase, the 1,000-L high–cell-density fed- ALL investigated daily supply of the highly alkaline batch run and compared the results scales and in line Feed-B (pH 11) caused minor with an IgG1 control (Figure 4). variations. For increased accuracy in Product-specific bands found at with the results of pH control, we twice recalibrated the ~25 kDa and 50 kDa showed our 5-L glass-vessel single-use probe signals (t = 3d, consistency with the control. reference data. t = 6d ) based on off-line Moreover, we discovered no significant measurements (ΔpH > 0.05 units). impurities. Based on those data, we The single-use and conventional glass excluded the possibility of detrimental pH probes demonstrated excellent effects on productivity with Flexsafe comparability during the 1,000-L run, bags. Negative impacts on product SUCCESSFUL CONTROL OF CRITICAL which we confirmed with off-line pH purity also seem unlikely. PROCESS PARAMETERS measurements. This demonstrated the Furthermore, we monitored glucose Minow et al. showed that detailed capability of single-use sensors for and lactate concentrations as well as understanding of bioreactor stable control of pH and pO2 during a osmolality during the fed-batch runs performance enabled fast scale-up of a 17-day fed-batch production process. 11 (Figure 5). Lactate concentration similar CHO fed-batch process ( ). We monitored pCO2 values off-line remained <2 g/L, below previously For our study, the extensive using samples that were analyzed by reported inhibitory levels (10). Glucose characterization of the BIOSTAT an ABL800 Basic blood gas analyzer concentration was maintained at >3 g/L STR family with regard to oxygen from Radiometer. Maximum values of during the entire process. Osmolality transfer, mixing time, and power input 107 mmHg were reached on day 12 ranged 300–470 mOsmol/kg and was summarized by de Wilde et al. (9) and — well below a critical value of kept below a previously established the classical stirred-tank bioreactor 150 mmHg reported by Zhu et al. 12 critical level of 550 mOsmol/kg (data design facilitated and simplified ( ). The pCO2 values obtained during not shown). scale-up from 5 L to 1,000 L, as our 1,000-L high–cell-density fed- In essence, both our batch data and demonstrated by our highly batch run were comparable to data our fed-batch data are highly comparable results. obtained in the BIOSTAT STR 50 comparable across all investigated We ensured successful pO2 control fed-batch run (data not shown). Thus, scales and in line with the results of at 60% ± 10% during the entire high– successful CO2 stripping was achieved our 5-L glass-vessel reference data. cell-density fed-batch process (Figure using the ring-sparger gassing strategy This demonstrates the superior 6) using a multistage cascade-control at the applied gas flow rates. biological performance of Flexsafe approach based on pO2 signals from a STR bags. single-use optochemical probe.

56 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT CONCLUSION Partnership with Suppliers. BioProcess Int. Our study successfully demonstrates 12(8) 2014: S43–S46. 6 the new Flexsafe STR and RM bag  Eibl R, et al. Recommendation for Leachable Studies: Standardized Cell Culture family’s suitability using a state of the We hope to pave the Test for the Early Identification of Critical art, high–cell-density, fed-batch MAb Films for CHO Cell Lines in Chemically production process using CHO cells. way for full adoption Defined Culture Media. Dechema January We have confirmed data obtained of large-scale, single- 2014. from a-irradiated sample bags during use bioreactors in 7 Walsh G. Biopharmaceutical development of S80 polyethylene film Benchmarks 2010. Nat. Biotechnol. 28(9) 2010. late-phase and 8 (3) in the final bag assembly with a Jurkiewicz E, et al. Verification of a New Biocompatible Single-Use Film typical cell-culture application COMMERCIAL Formulation with Optimized Additive Content growing recombinant CHO cells in manufacturing, in for Multiple Bioprocess Applications. chemically defined, protein-free cell which quality, Biotechnol. Progr. in press. 9 culture media. In light of recently reproducibility, and De Wilde D, et al. Superior Scalability published Dechema results (6), it of Single-Use Bioreactors. BioProcess Int. 12(8)s becomes obvious how important assurance of material 2014. 10 optimizing polyethylene films for cell- supply are critical. Zhang Y. Approaches to Optimizing Animal Cell Culture Process: Substrate culture applications is to ensuring Metabolism Regulation and Protein Expression reproducible growth performance. We Improvement. Adv. Biochem. Eng./Biotechnol. hope to pave the way for full adoption 113, 2009: 177–215. of large-scale, single-use bioreactors in 11 Minow B, de Witt H, Knabben I. Fast late-phase and commercial scale-up. Finally, the performance of Track API Manufacturing from Shake Flask to manufacturing, in which quality, the optochemical pH and pO sensors Production Scale Using a 1000-L Single-Use 2 Facility. Chem. Ing. Technik 85(1–2) 2013: reproducibility, and assurance of demonstrates that a single-use 87–94. material supplies are critical for safe bioreactor concept does not need to be 12 Zhu MM, et al. Effects of Elevated and uninterrupted drug availability. compromised by insertion of classical pCO2 and Osmolality on Growth of CHO Direct scale-up from a 5-L probes that increase the risk of Cells and Production of Antibody-Fusion conventional glass stirred-tank vessel contamination. Our results Protein B1: A Case Study. Biotechnol. Progr. (BIOSTAT B 5), through a 50-L convincingly show that a fully single- 21(1) 2005: 70–77. 13 stirred-tank single-use bioreactor use cell culture process can be Dreher T, et al. Design Space Definition for a Stirred Single-Use Bioreactor (BIOSTAT STR 50) to the final implemented easily and scaled reliably Family from 50 to 2000 L Scale. Eng. Life Sci. 1,000-L single-use stirred-tank system to production scale confirmed by special issue (Single-Use Technologies in (BIOSTAT STR 1000) verified again excellent comparability of the 5-L, Biopharmaceutical Manufacturing). Eibl R, the unique scalability of our entire 50-L, and 1,000-L data. Eibl D, Eds. In press.  BIOSTAT family, which is based on ACKNOWLEDGMENTS classical stirred-tank principle and Corresponding author Regina Reglin and Our thanks are due to the entire R&D upstream single-use materials with superior Sebastian Ruhl are scientists in upstream growth-promoting properties. These technology team for their technical contributions. technology R&D; Jörg Weyand is a results are also in line with previous fermentation technologies application studies that showed good agreement specialist in Central Europe; Davy De Wilde of the engineering design space and REFERENCES is director of marketing fermentation key process parameters of BIOSTAT 1 Eibl D, Peuker T, Eibl R. Single-Use technologies; Ute Husemann is manager stirred-tank bioreactors for Equipment in Biopharmaceutical Manufacture: of upstream technology R&D; Gerhard A Brief Introduction. Single-Use Technology in 13 Greller is R&D director of upstream mammalian cell culture ( ). In Biopharmaceutical Manufacture. Eibl R, Eibl D., addition to achieving the same cell technology; and Christel Fenge is vice Eds. John Wiley and Sons: Hoboken, NJ, 2010. president of marketing fermentation culture performance across all 2 Weber A, et al. Development and technologies at Sartorius Stedim Biotech, investigated scales, we also Qualification of a Scalable, Disposable August-Spindler-Straβe 11, 37079 Bioreactor for GMP-Compliant Cell Culture. demonstrated efficient control of key Goettingen, Germany; regina.reglin@ process parameters such as pH and BioProcess Int. 12(8) 2014: S47–S52. 3 sartorius-stedim.com. pO in our very first high–cell-density Fenge C, et al. Consistently Superior 2 Cell Growth Achieved with New Polyethylene fed-batch cultivation at 1,000-L scale. Film Formulation. BioProcess Int. 12(8) 2014: In addition to similar design S34–S37. principles within the BIOSTAT 4 Vachette E, et al. Robust and stirred-tank bioreactor family, the Convenient Single-Use Processing with Superior Strength and Flexibility of Flexsafe combination of very effective CO2 stripping with high oxygen mass Bags. BioProcess Int. 12(8) 2014: S38–S42. 5 transfer provided by a ring sparger Vachette E, et al. Enhanced Assurance of Supply for Single-Use Bags Based on enabled our direct and successful Material Science, Quality By Design, and

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 57 B IOP ROCESS PRODUCT INTEGRITY

Pressure Decay Method for Postinstallation Single-Use Bioreactor Bag Testing by Magnus Stering, Martin Dahlberg, Thorsten Adams, Davy De Wilde, and Christel Fenge

ingle-use technology is well Figure 1: Porous fleece prevents masking effect accepted today, and Porous Direct Contact manufacturers’ quality Fleece assurance programs ensure Sleak-free single-use bags upon Bag Holder Bag delivery. But what about risks involved Holder with installation and other handling errors? Operator training and implementation of suitable standard Hole Hole operating procedures (SOPs) are mandatory, but should they be the only ways to mitigate the risk of failures? In addition, more companies are advocating the use of ballroom concepts (1) for the manufacture of biopharmaceutical drug substances and drug products. However, how do Air leaking out without restriction Air leaking out with restriction you prove that applied unit operations are closed and intermediates are not subjected to risks of cross- POSTINSTALLATION PREUSE detecting typical leaks that might have contamination? TESTING HELPS MITIGATE RISKS been introduced by operator-handling The high value of growth media In terms of regulatory framework and errors — would greatly improve risk- and the duration of typical cell culture feasibility, there is a clear difference mitigation capabilities in single-use processes call for the highest degree of between integrity testing sterilizing- production facilities. That could scrutiny when setting up such runs. A grade filters and testing single-use improve operator safety and financial leaking bioreactor would generate a bags. According to current European security associated with maintained significant financial loss and GMP regulations (2) — which are the production capacity and on-time jeopardize a carefully timed most stringent among international delivery. A typical example of a production schedule in a good guidelines — sterilizing-grade filters potential risk is the improper manufacturing practice (GMP) should be tested after sterilization both connection of different tubing production facility. In vaccine before use and immediately after use. elements during bioreactor setup, processes, it might also pose a risk to Single-use bags have no such media preparation, and inoculum operator safety as well as the general regulatory requirement. Nonetheless, a transfer. An operator error during such environment and therefore must be postinstallation preuse test of the critical operations might result in a mitigated under all circumstances. entire single-use bioreactor system leak, potentially causing contamination (including tubing) — capable of at a later time and loss of a run.

58 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT Table 1: Minimum detectable leak sizes for different Biostat STR bag volumes SCANNING DIFFERENT TEST METHODOLOGIES 50 L 200 L 500 L 1,000 L 2,000 L Detectable minimum leak 50 μm 100 μm 200 μm 400 μm TBD More than 40 different test methods size at 20 min test time have been proposed for leak detection (e.g., according to EN1779 “Leak testing: Criteria for Method and Table 2: Test program parameters for different Biostat STR bag volumes Technique Selection”) based on a 50 L 200 L 500 L 1,000 L 2,000 L number of technologies such as gas Test pressure 50 mbar 50 mbar 50 mbar 50 mbar 50 mbar leak detection (integral or by probe), Filling time 4 min 20 min 35 min 65 min TBD thermal imaging, flow measurement, Stabilization time 20 min 20 min 20 min 20 min TBD and pressure decay and pressure Test time 20 min 20 min 20 min 20 min TBD increase. Each technology has its own strengths and weaknesses in terms of technology, but it requires a very • accuracy and reproducibility precise pressure control throughout • rest time the test. Slightly overshooting or  • handling aspects undershooting the initial test pressure • condition of the bags after testing would directly affect the measured A specifically • possibility of testing connections flow and result in false estimation of designed, patent- • investment costs for equipment the test value. The bigger the volume pending • cost per test the more difficult it is to accurately polyethylene fleece • footprint of the test equipment adjust the pressure. This methodology • contamination risks therefore would not be applicable to acting as a porous • IP situation larger volumes. SPACER between • feasibility of in-place testing Pressure-decay technology, on the the plastic film of the (point of use). other hand, is a robust, easy to bioreactor and the Those factors have to be carefully implement, cost-effective, and considered and balanced in developing recognized technology that can be holder prevents a suitable bag-testing device for single- used to test a single-use bioreactor bag masking effect. use bioreactors for GMP production after installation in its final of vaccines, monoclonal antibodies, configuration directly in its holder. and recombinant proteins. We have therefore used this approach to develop a reliable and predictive the bag from the flat surface, would FEW METHODS QUALIFY FOR risk-mitigation tool. not represent a reliable test set-up POSTINSTALLATION PREUSE TESTING because of the masking effect. There is a risk of potential handling DEVELOPING A RELIABLE AND Using a specifically designed, errors when installing single-use REPRODUCIBLE METHOD patent-pending polyethylene fleece bioreactors into their holders. So a The objective of any test method must acting as a porous spacer between the reliable leak test for such bags must be be to reliably identify potential plastic film of the bioreactor and the performed at the point of use (preuse damage of installed bioreactor bags holder prevents this masking effect but postinstallation). Testing a single- covering bag seals, port welds, (Figure 1). It delivers reliable and use bioreactor bag in a separate device connections, and all bag surfaces. Any reproducible test values comparable to (as is usually done when applying gas damage could result in a loss of results obtained with the same test leak detection) and then installing it bioreactor content or pose a biosafety method on bag areas that are not into its bag holder would not permit risk. covered by any hard surface. detection of operator-handling errors. We identified early on during the It would merely compensate for development of a method based on PRESSURE-DECAY potential shipping incidences. Also, pressure decay that a leak in a plastic TESTING METHODOLOGY the complexity and costs associated bag might be totally masked when The developed pressure-decay test with gas leak detection technology pressed against a smooth, hard method is based on the protocol from would be unreasonable to make it a surface. Very little of the test gas can the American Society for Testing and routine testing method. However, escape through the leak, and the Materials ASTM F2095-01 (3).The proper packaging (by the reliability of the pressure-decay Sartocheck 4 plus Bag tester from manufacturer) and inspection of the detection is nil. Therefore, performing Sartorius Stedim Biotech is used cardboard box (by the customer) pressure decay or flow measurement together with the above-described bag would detect shipping incidences. on a bag in a holder or even on a bag tester fleece that prevents masking of Mass or volumetric flow placed freely between two plates, leaks that might have been potentially measurement is a well-known without any porous spacer separating introduced during installation. It

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 59 Figure 2: Point-of-use testing of singe-use bags ensures reliable point-of-use leak testing of single-use bioreactor bags Top Surface postinstallation and preuse in its final bag holder. Pressure-decay measurement is very sensitive to temperature changes. Therefore, the Sartocheck4 plus bag tester provides a temperature measurement of the environment and Connections generates a warning message in case of Bag Seals excessive temperature drift during the test phase. This pressure-decay leak test has been developed to reliably detect defects in BIOSTAT STR single-use bioreactor bags over the complete range from 50 L to 1,000 L. Qualification of the method at the Tank Surface 2,000 L scale is ongoing. The bag is tested after installation, so potential leaks are detected in the bag walls, seals, or connections over the entire flexible bag system, including tubing. The test method is nondestructive and enables the implementation of a reliable and reproducible point-of-use test in bioproduction facilities (Figure 2).

Port Welds The pressure decay during the test is measured and compared with acceptance criteria established during method qualification.

Figure 3: Engineering study results (BIOSTAT STR 200L bag) QUALIFICATION LIMITS 10,000 ESTABLISHED DURING INITIAL STUDY 9,000 0 μm The validation approach consisted of 50 μm 8,000 100 μm two parts: 200 μm • an engineering study to establish 7,000 300 μm 6,000 the detection limit for different bag volumes 5,000 • formal qualification to verify the 4,000 minimum detectable leak size and 3,000 establish acceptance criteria of the 2,000 test. 1,000 To determine the minimum Pressure Drop (mbar) Drop Pressure 0 detectable leak size, a simplified bag 0 600 1,200 1,800 2,400 3,000 3,600 4,200 4,800 5,200 −1,000 with only one bottom port connection for test gas application was prepared −2,000 Time (s) with multiple “defect patches” representing different defined leak sizes. A defect patch is a circular sheet Table 3: Summary of maximum pressure decay for different bag volumes established during qualification study of film with a laser-drilled and flow- calibrated leak diameter that was 50 L 200 L 500 L 1,000 L welded to the bag surface. Test pressure 50 mbar 50 mbar 50 mbar 50 mbar For reliable detection of a given leak Maximum pressure decay during test time 1.2 mbar 1.5 mbar 2.5 mbar 3.2 mbar size there must be a significant difference in pressure decay between a bag with the given leak size and a bag

60 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT without defects. By repeated testing, equipment is both robust and the standard deviation of the test value cost-effective. was calculated for the various leak sizes  One saved batch can provide an and for bags without defects. The The test method instant return on investment. The leak minimum detectable leak size was test helps to prevent derailing of defined to be the leak size of which the using pressure decay production schedule and project lowest pressure decay value was at least combined with the delays. Especially in processes where six times higher than the standard fleece to prevent biosafety is critical (e.g., vaccine deviation of the highest pressure decay masking of leaks was production), it is an indispensable tool value of a bag without leaks. for protecting operator safety. Figure 3 shows results for different successfully qualified defect patch sizes for a 200 L bag and PROVEN to REFERENCES volume. The pressure decay measured be a robust and 1 Chalk S, et al. Challenging the for a 50 μm defect patch was Cleanroom Paradigm for Biopharmaceutical predictive method Manufacturing of Bulk Drug Substance. overlapping with a bag without leaks. for reliable detection BioPharm Int. Aug. 2011: 1–13. Therefore, a 50 μm leak cannot 2 EudraLex: The Rules Governing reliably be differentiated to a bag of leaks. Medicinal Products in the European Union. without leaks. On the contrary, a Volume 4: EU Guidelines to Good Manufacturing 100 μm leak size shows a clear Practice Medicinal Products for Human and differentiation to a bag without leaks. Veterinary Use; http://ec.europa.eu/health/ documents/eudralex/vol-4/index_en.htm. In a second step, one standard bag All tested bags showed expected 3 ASTM F2095-07: Standard Test without a defect patch and one results: The nondefect standard bags Methods for Pressure Decay Leak Test for standard bag with a defect patch of passed the test, and the standard bags Flexible Packages with and without the previously determined minimum prepared with a single defect failed. Restraining Plates; www.astm.org/Standards/ detectable leak size were used to Hence, the test method using pressure F2095.htm.  establish the final test parameters. decay combined with the fleece to Table 1 shows the minimum prevent masking of leaks was Corresponding author Magnus Stering is detectable leak sizes for the different successfully qualified and proven to be product manager, integrity testing BIOSTAT STR bag volumes. Based a robust and predictive method for solutions; Martin Dahlberg is manager, on the results of the engineering reliable detection of leaks during R&D instrumentation and control; Thorsten study, the final test parameters for the routine operation. Adams is product manager, fermentation qualification of the method were technologies, Davy De Wilde is director of established (Table 2). Those FAST AND ROBUST BAG TESTING marketing, fermentation technologies; parameters were used to generate a METHOD FOR ROUTINE USE Christel Fenge is vice president of test program for each BIOSTAT STR This new, qualified single-use marketing, fermentation technologies. bag volume using the Sartocheck 4 bioreactor leak test based on pressure All authors are from Sartorius Stedim plus Bag tester. decay using a specifically designed Biotech. porous fleece enables reliable and METHOD VALIDATION CONFIRMS robust routine postinstallation and RELIABLE DETECTION OF LEAKS preuse testing. With this novel test During the final qualification study, method, the same level of risk the minimum detectable leak size for mitigation and assurance of operating the different bag volumes was procedures can be achieved for single- confirmed on a statistically significant use bioreactors as previously achievable number of standard bags from only with conventional stainless steel different routine production lots. bioreactors that could easily be Standard, nonmodified bags of pressure-tested before use. different bag volumes and standard The pressure-decay approach is bags modified with a single defect easy to implement on a routine basis. patch representing the minimum The automated testing sequence and detectable leak size were used. All user guidance ensure a low risk of bags were gamma irradiated. A operator-handling errors. Monitoring minimum of 10 test repeats were the environmental temperature during performed per bag. Table 3 the full test sequence further enhances summarizes the maximum pressure reliability of the test results and decay measured for nondefect bags prevents false conforming and false during the qualification study. nonconforming test results. The

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 61 B IOP ROCESS INTERVIEW

Total Solutions Support the Growth of a Dynamic Industry A Conversation with Reinhard Vogt and Stefan Schlack by Brian Caine and S. Anne Montgomery

hile attending a conference at Sartorius Stedim Biotech in Göttingen, Germany, BPI publisher Brian Caine Wand editor in chief Anne Montgomery spoke with Reinhard Vogt (executive vice president of marketing sales and services, and member of the administrative board) and Stefan Schlack (senior vice president, marketing and product management). They discussed Sartorius’s forward- thinking business strategies, its position as a total solution provider, Left to right: Stefan Schlack (senior vice president of marketing and product management, and how the company’s strategic goals Sartorius Stedim Biotech), Brian Caine (publisher of BioProcess International), mesh with its assessment of current S. Anne Montgomery (editor in chief of BioProcess International), and Reinhardt Vogt (executive vice president of marketing, sales, and services, industry directions. administrative board member of Sartorius Stedim Biotech).

SINGLE-USE AS AN ASSURANCE OF SUPPLY and other indications? Do you see ENABLING TECHNOLOGY Caine: Single-use technologies have applications for your technologies in Caine: Your company positions itself as reopened once-stalled markets, such as cell therapy, for example? a total solution provider for the vaccines. As a total solution provider, Vogt: We’re watching the market biopharmaceutical industry, with a how do you work with clients to assure carefully and have made some recent special focus on single-use them that Sartorius Stedim Biotech is investments. We believe that we applications. Can you tell us what the company that can take them from have an excellent portfolio already impact single-use has had on the start to finish in their process? supporting this market. For example, market and what the future holds for Vogt: Single-use technologies have our wave-style bioreactors are the technology? reached a quality level similar to already used in this area, and very Vogt: For filter cartridges and stainless steel. Today, we offer a promising data have been published capsules, single-use is quite an old comprehensive technology and product recently for our single-use, stirred- concept. Today, though, single-use portfolio. We source and monitor our tank bioreactors. We’re also processes are much more complex and raw materials, and we control the actual cooperating with Lonza, which has a comprise part or indeed the whole production of the single-use solutions as lot of experience in this area with its production concept; they are replacing well. We bring our customers to our contract manufacturing services. stainless steel containers and piping. suppliers to ensure them that our Montgomery: Where are we However, it has taken a while for suppliers’ quality-management systems regarding the reality of completely single-use to develop and for the match our quality system and that they single-use facilities? solutions to become robust and scalable can rely on it. Schlack: Actually, this is already enough to operate at production level. Caine: What about other reality for quite a few of our applications for single-use technologies customers. In this respect, one of our

62 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT  To be a true ALTERNATIVE to stainless steel, we have to move away from focusing on components and individual products and offer complete single-use process steps. customers just received the ISPE We have developed very useful “Facility of the Year” award for a enabling technologies for single-use completely single-use MAb process.We drug manufacturing facilities. For actually delivered the unit operations example, our new Virosart HF system for upstream and downstream is gamma sterilizable. You avoid processing. sanitization, and at the same time you And we expect to see further get an excellent virus filtration significant activity especially driven by performance comparable with other the so-called emerging economies. We market leading technologies. Another believe that our industry faces four key example is our FlexAct UD single-use challenges: capacity adaptation, smart solution for ultra- and diafiltration. It process transfers, risk mitigation, and saves you up to six or more hours of cost savings. For example, the Chinese preparation. This is especially critical biopharmaceutical market grows on when you think about product average by 30% per year; some products change-overs. Our customers typically grow by 70% or more. Fast capacity look for solutions that reduce change- and bag, as well as tubing. And all adaption will be a key success factor to over times. Our membrane adsorbers that has to be combined with meet the drug demand of the growing are yet another example. They are easy application and process know-how to market. On one hand, it is important to operate and fully scalable, so you make it easy for our customers to use. for our customers to avoid sunk costs don’t need to invest in installing large Furthermore, our customers require due to underused capacity. On the chromatography columns for your assurance of supply and transparent other hand, they need sufficient polishing steps. change-control procedures for the capacity to secure market share. single-use components of the process Especially for biosimilars, I believe that PROVIDING A TOTAL SOLUTION steps. Sartorius Stedim Biotech is flexibility will be crucial and will be Caine: In terms of single-use, how do capable of doing this. However, provided by single-use technologies you define being a total solution managing this complexity is certainly Another example is process provider, and what are the main a challenge, and we are prepared to transfer. When you put the right and drivers and hurdles in that market? develop our company to adapt to it. scalable single-use technologies in Schlack: We just talked about the Vogt: A total solution means your process development, you will be challenges of the pharmaceutical exactly that — a total solution — not able to accomplish process transfers in industry. Single-use is a solution to just the supply of components and 12 months or less. By contrast, process address them. However, to offer a the availability of scalable products transfers based on stainless steel can true alternative to stainless steel, we for each process step, but risk easily take twice as long or even have to move away from focusing on mitigation and logistics as well. longer, depending on whether you components and individual products Some customers are looking for have to build a new facility or adapt and offer complete single-use process upstream solutions, some are looking an existing one. No biopharmaceutical steps. Using our FlexAct UD as an for downstream solutions, and others company can afford that amount of example, we provide a mixing bag, a may want a total production time: you have to be first to market to mixer, vent filters, a controller, solution, wherever they are in the take full advantage of your patent or, pressure transmitters and pumps, world. Some want an entire single- in the case of biosimilars, take your crossflow cassettes, a holder and a use factory; others want a hybrid share of the market. recirculation bag, a storage container solution. Some want as much as

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 63 supply chain even down to the resin manufacturer, with long-term contracts and true partnerships. Business continuity is achieved with a robust contingency plan that includes backup resin crackers and film extruders, multiple bag manufacturing and sterilization sites and safety stocks of resins and film. Caine: How does the BIOSTAT STR 2000 affect your portfolio, and how does it benefit your clients? Will bioreactors need to be developed above a 2,000-L capacity? possible to be single-use, but they Vogt: With increasing titers, I don’t might already have a 5,000-L stainless think that single-use bioreactors will steel bioreactor. Sometimes, no single- need to go beyond 2,000 L. In some use technology is available for a  cases, that volume may not go above particular application. Our “total You have to offer a 1,000 L. Vaccine producers, for solution provider” strategy means that example, don’t need a 2,000-L we understand what our customers COMPLETE range fermentor. expect from us: single-use products (2–2,000 L) so that Schlack: I can agree to that; and integrated single-use process you can ensure 2,000 L will be the manufacturing solutions, and perhaps most straightforward and scale of the future, especially importantly, a partnership to jointly considering recent developments achieve the goal in the most effective fast scale-up and regarding intensified fed-batch way. scale-down, process processes. But it’s equally important Montgomery: Talking about some transfer, and that you have a complete range. No of your single-use products, can you troubleshooting. customer buys a 2,000-L bioreactor tell us a little more about your new from Sartorius Stedim Biotech and Flexsafe bag concept and the new S80 then another from a different supplier. polyethylene film? You have to offer a complete range Schlack: Sartorius Stedim Biotech (2–2,000 L) so you can ensure has developed a new polyethylene (PE) This is a major issue for today’s straightforward and fast scale-up and film and family of bioprocessing bags biopharmaceutical companies. What’s scale-down, process transfers, and to meet the single-use manufacturing the cause? It is related to a cytotoxic troubleshooting. needs of the future. Unlike any other degradation product derived from a We also incorporate the same film on the market used to make bags commonly used antioxidant. To single-use sensors in each single-use for biomanufacturing, our new S80 PE overcome this challenge, you need a bioreactor for complete comparability. film, which is used in Flexsafe bags, is deep understanding of polymer and We’ve worked hard to provide a the result of close collaborations material sciences in combination with complete upstream solution, which is between Sartorius Stedim Biotech and cell biology. We have exactly why we acquired TAP Biosystems. The our polymer and film suppliers. With accomplished this deep understanding this unprecedented partnership, we in Flexsafe development. have developed a completely new Flexsafe ensures excellent and polyethylene film structure and reproducible growth, even with the achieved excellent and consistent cell most sensitive production cell lines. growth, robustness and unprecedented Cell-growth reproducibility is assurance of supply. With these unique guaranteed by controlling the resins benefits, Flexsafe enables the and the additives specifications and by implementation of single-use setting a design space for the film- bioprocessing throughout the entire extrusion parameters. Independent drug manufacturing process: from laboratories have demonstrated that development to production, from Flexsafe bags are free of cytotoxic upstream to downstream, including cell leachables. No known toxic culture, storage, shipping, mixing, and degradation product of common freezing and filling applications. additives (bDtBPP) is detectable in For example, take the issue of cell- water-for-injection (WFI) extracts. growth inhibition in single-use bags. Furthermore we’ve established a secure 64 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT ambr15 multiparallel minibioreactor has already become the industry choice for cell line and medium development. The ambr250 system that we recently launched is based on the same design principles as conventional glass stirred- tank systems and our BIOSTAT STR single-use bioreactor. At the same time, it is a multiparallel process- development bioreactor system enabling effective and fast quality by design (QbD) studies to determine the operating ranges of critical process parameters and the design space of the production process. Vogt: Imagine that you have a single-use process and a single-use fermentor, and you source tubing from one supplier, the filter from someone else, the controller from yet another use technologies to improve quality, made no sense for us to buy the company, and so on. And then, the reduce operator errors, and decrease remaining 20%. Instead, we created a components don’t fit together. Small, costs. The bag is comparatively cheap, virtual joint venture without capital or start-up companies aren’t in a position but putting all the components legal components. Lonza does the to manage multiple vendors and together is expensive. production and oversees logistics; support FDA inspections based on this together we handle sales, marketing level of complexity. And large TO PARTNER OR ACQUIRE? and R&D; and at the end, we share biopharmaceutical companies are Caine: Having mentioned your the profits. looking for sourcing efficiencies as well. association with SüdPack and the Montgomery: Acquisition is clearly Montgomery: Not to mention the importance of collaborations, how do an important part of your business challenges of monitoring the supply you decide whether to acquire or strategy. Can you offer more chain for each of those components. partner with a company? And how do background on the purchase of TAP Vogt: Yes, and we’re also trying to you assess the value of one kind of Biosystems? educate our customers about that. relationship versus another? Schlack: We believe that TAP Schlack: In process development, Vogt: First, we determine what we Biosystems is the final piece in our aspects such as assurance of supply, consider to be a component rather upstream portfolio. When you do change control, and pricing security than a key technology. Our key QbD studies, you have to do multiple are not really critical. But when our technologies are cell culture, small-scale trials. There’s no better customers move toward commercial membrane and single-use system available for that than the production based on single-use technologies. Once we decide on a key ambr250, which is truly scalable. We processes, they become critical. Your technology, we need to develop or now can offer fully automated, manufacturing costs could acquire it so that we have R&D and multiparallel bioreactors, sensors and significantly increase if those elements the whole manufacturing under controllers, offering full scalability up aren’t in place when you go into control and can ensure a sustainable to 2,000 L. So it’s a really complete current good manufacturing practice supply chain. Sometimes, however, we package. I’m very glad that it’s now (CGMP) production. identify a key technology but cannot part of the Sartorius family. As Reinhard said, to reach this level buy the company for whatever reason, Vogt: The acquisition became of assurance of supply, standardization or we might be interested in only 10% interesting when TAP Biosystems is needed. But standardization will be of a company’s portfolio. That’s when combined its automation expertise possible only when the biotech we think about cooperation models. A with single-use technology. TAP’s industry understands that is has to good example is our agreement with know-how regarding stirred reconsider the R&D mindset of Lonza. We wanted a media offering, microreactors enables automated selecting bits and pieces from multiple and we also wanted to work with a process optimization and the different sources — such as tubings company that had many years of unparalleled development of cell from vendor A, filters from vendor B, experience in that area, with products lines for biopharmaceutical sensors from C and so on. on the market and a good brand production by reproducing classical Vogt: If you don’t standardize a name. Lonza uses 80% of its media for bioreactor conditions. The process, you can’t automate it. We its contract manufacturing combination of automation and need more standardization in single- organization (CMO) activities, so it single-use is unique, and

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 65 phone, you might have said no. Today want to be best in class. We want to be everybody uses them. That’s a good the best in cell culture, in downstream  example of need versus want. processing, in filtration, and so on. The TAP Biosystems Schlack: Timing is also important. Schlack: We want to be a partner It’s often not about reinventing the for the biopharmaceutical industry. acquisition, together wheel, but rather offering a complete We offer solutions that address key with our chemometrics solution. In our industry, you have industry needs and challenges. And offering, lays the technology adoption cycles of six to we work hard to close the final  foundation for QbD eight years or more, so you have to remaining gaps. believe in your analysis of future approaches and trends — and keep on investing — effective process even if you don’t immediately get a optimization and good return on that investment. characterization in a COMPANY GROWTH, SUSTAINABILITY VERY SHORT Montgomery: Can you speak about time frame. planning for the future, how you recruit and train new staff? Vogt: That’s very important for us. We have a very good staff retention multiparallel process development rate and good training programs. But translates into enormous time and I think it’s also about company culture cost savings for customers. and giving young people a chance. You have to create a culture of THE FUTURE OF PROCESS appreciation and encouragement that ANALYTICAL TECHNOLOGY (PAT) goes beyond the framework of typical Caine: Let’s switch to another talent management. important topic. During the Schlack: We need openness and conference, we’ve heard a lot about room for people to gain experience. process analytical technology (PAT), We have excellent people. As QbD, and design of experiments managers, we have to ensure that they (DoE). What do you think is the key can share their knowledge and to their successful adoption and exchange ideas. Of course, it helps to commercialization? have a dynamic company in a vibrant Schlack: Our vision is to offer market. We send people all over the seamless transfer from process world, which encourages them to development to commercial production develop their skills and expertise and in upstream processing. We are very keeps them loyal to the company. To close to reaching this point. The TAP lose people is to lose potential. Biosystems acquisition, together with our chemometrics offering, was crucial THE TAKE-HOME MESSAGE in this respect because it lays the Caine: Given that your market-facing Reinhard Vogt is Sartorius Stedim Biotech’s foundation for QbD approaches and statement for Sartorius Stedim Biotech executive vice president of marketing, sales, effective process optimization and is to be a total solution provider, what and services and a member of the characterization in a very short time message are you trying to send to the company’s administrative board. Stefan frame. market? Schlack is Sartorius Stedim Biotech’s senior Caine: What’s more important: Vogt: First, customers should know vice president of marketing and product technological advancement or that if they want to incorporate single- management. Both are based in Göttingen, additional education on existing use processes, they should come to us! Germany. Brian Caine is cofounder and technology? We have invented it, we are the market publisher, and S. Anne Montgomery is Vogt: Both. Unlike the mature leader. And, of course, we go beyond cofounder and editor in chief of BioProcess chemical industry, the bioprocess components, technologies and International, amontgomery@ bioprocessintl.com. community is still on a learning curve. applications into supply chain Our strategy has always been driven management. So customers can review by finding out what our customers our assurance of supply and business need, not what they want. That’s a big continuity plans. The other key difference. If you were asked 20 years message is that we have added ago whether you needed a smart products to our portfolio because we

66 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT B IOP ROCESS THE FINAL WORD

Disposables for Biomanufacturing A User’s Perspective by Dr. Berthold Boedeker

he supply scenario for many disposables, standardization and biopharmaceutical drugs such as qualification of bags and connections, monoclonal antibodies (MAbs) and validation of leachables and is changing. With the extractables, as well as dependency on Timplementation of personalized individual solutions from different medicine resulting in drugs for vendors. specific, high-responder subsets of patients, market volume per drug will ADVANTAGES OF DISPOSABLES decrease. In addition, increasing What are the major advantages of fermentation titers of up to 10 g/L for using disposables-based processing MAbs are leading to smaller compared with standard production in fermentation volumes necessary to a hard-piped, steel-tank–based accommodate individual setting? One major advantage is that biopharmaceutical market demands. presterilized single-use systems can be That results in approaches such as used in a laboratory-like environment. flexible production in campaigns or This is well suited for small-scale decentralization of manufacturing research and development activities, using similar facilities with low risk of because no supporting engineering tech-transfer issues for regional infrastructure regarding utilities, markets. hard-piping, or automation (for In this regard, single-use example) is needed to set up and run technologies (disposables) play an such operations. This enables important role in how bioprocessing to be performed at a biopharmaceutical development and reasonable scale even in university production, particularly from laboratories. Another advantage is the mammalian cell culture, can nowadays SARTORIUS AG WWW.SARTORIUS.COM time and cost savings in plant be performed. Except for some large- construction and operation. The main scale unit operations such as contributors here are lower capital centrifugation, chromatography skids, part of biomanufacturing for a long costs; reduced consumption of utilities and UF/DF operations, all process time now through integration into such as gas, electricity, and water steps can be performed in disposables hard-piped set-ups (filters, and so on), (purified, WFI); no or limited hard- up to a bag volume of around 3,000 L. the real progress toward completely piping; and less-complex automation. Such steps include mixing, holding, disposable processing came with Time savings vary depending on and distribution of culture media and development of single-use bioreactors the extent of disposables use. Most buffers; cell seed expansion; (SUBs). Several systems are now facilities still contain nondisposable production fermentation; and cell available up to a fermentation scale of unit operations. In hybrid designs, removal by depth filtration, disposable 2,000 L. However, there are still time savings in the early engineering chromatography columns, and UF/ limitations with single-use project up to mechanical completion DF/virus filtration. Although many technologies, particularly in the areas are sometimes marginal. But during single-use units have been an integral of pretesting and the quality of start-up, including during

SUPPLEMENT SEPTEMBER 2014 12(8)S BioProcess International 67  New processing approaches include use of disposables in DOE studies, feeding strategies, concentrated cell banking, and applications for CONTINUOUS processing.

SARTORIUS AG WWW.SARTORIUS.COM qualification and validation, time unit. To comply with the desired are the use of new plastic surface films savings can be very pronounced (up to second-supplier concept in with fewer side effects on cell growth, 70%) because equipment qualification biomanufacturing for SUBs, a improvements in film robustness, and using disposables is very limited, and company has to show biochemical supply-assurance strategies. Other no steam-in-place (SIP) and clean-in- comparability and consistent product contributions address quality-control place (CIP) processes are needed. Also quality in two bioreactor types before issues such as point-of-use bag testing the sometimes very lengthy cleaning they can be used for commercial and SUB qualification strategies. validation of vessels and pipes that production. This is a substantial Finally, new processing approaches are contact a product is not necessary for additional development and validation shown such as use of disposables in single-use because the bags are effort until two adequate systems are design-of-experiments (DOE) studies, discarded after each run. Another licensed. In addition, it is desired to feeding strategies, concentrated cell advantage is the possibility to get improved quality control by the banking to avoid open cell handling, efficiently perform short product suppliers. For example, bags should be and applications of single-use systems campaigns in multipurpose facilities, pressure tested before delivery to for continuous processing. including fast product turnover by reduce failure rates. It also would be All together, these new simply using new bags. advantageous to get full supporting developments show that single-use validation packages, including technologies are maturing so that the ADDRESSING LIMITATIONS extractables and leachables data as desired situation may become reality: a OF SINGLE-USE well as regulatory support files from fully disposable production facility On the other hand, a number of the suppliers to make regulatory filing with closed systems in a GMP-lab– disadvantages, risks, and limitations of simpler. like environment as an alternative or using disposables have to be addressed supplement to standard hard-piped, to make single-use-systems–based RAPID EVOLUTION OF IMPROVED steel-tank based production. This production a reliable alternative to SINGLE-USE TECHNOLOGIES would fulfill the desire for easy and standard production. First, there will Many new developments for fast plant construction, simple and always be a volume limit for handling continuously improving technical reliable operation, high flexibility, fast and operating disposables. For support, materials, and quality of product turnover, low cost of goods, fermentors and larger hold bags, that disposables are shown in the articles and easy technology transfer to is expected to approach 3,000 L. For of this supplement, illustrating that different regions.  portable systems, the limit is currently single-use technology is still a rapidly in the range of 1,000 L. Another issue evolving area. Promising contributions Dr. Berthold is standardization of single-use units are being made by vendors to assure Boedeker is chief and connections among vendors. end users that manufacturing can be scientist at Bayer Pharma AG, Several integrated systems are being performed reliably and with high GDD-GB-Biotech developed by individual suppliers, but quality as needed for pharmaceutical Development, those are not always compatible; that applications. Wuppertal, Germany; is, it is not possible to interconnect There are several contributions in berthold.boedeker@ systems from different suppliers to a this issue to addressing quality and bayer.com. large, functionally closed processing reliability of bag supply. Among these

68 BioProcess International 12(8)s SEPTEMBER 2014 SUPPLEMENT New Flexsafe Bag Family. New PE Film. New Benchmark.

Passed ASTM Shipping Test

Flexsafe meets your requirements for outstanding robustness and ease of use at ONE FILM FOR ALL all steps of your manufacturing process. The film's strength and flexibility provide consistent performance and easy handling – even in the most stringent applications like cell culture, long-term storage or shipping of drug products. www.sartorius-stedim.com/flexsafe USP – DSP – F+F