Up to Speed: Cell Culture Development. European Biopharmaceutical Review July 2016, pages 36-40. © Samedan Ltd

Up to Speed: Cell Culture Development. European Biopharmaceutical Review July 2016, pages 36-40. © Samedan Ltd

• 1.
  Cell Culture Development Upto Speed Despite the strong impetus towards accelerating novel biotherapeutics in clinical trials, the development of cell culture processes remains a labour-intensive and time-consuming procedure. High-throughput methods are showing huge potential in the mission to reduce both time and costs 36 July 2016 IND Preclinical Process development Manufacturing for tox Clinical manufacturing Commercial process development Process characterisation Process validation BLA and PAI Process monitoring and improvement FIH process ambrTM Commercial process Tox Phase 1 Phase 2 Phase 3 Submission Lifecycle managementand approval POBA Discovery support Cell line development Design space and optimisation Manufacturing First in human (FIH) process • Deliver clinical process quickly • Platform process • Clinical supply Commercial process • Deliver manufacturing process for registrational trials and market • Design keeping large-scale manufacturing • Improve productivity, efficiency, robustness, manufacturability, cost of goods (CoGs) Process characterisation and validation •Developin-processcontrol(IPC)strategythrough understandingofprocessinputsandoutputs(designspace) •Scale-downcharacterisationandvalidationstudies •Large-scaleprocessvalidationtodemonstrate processconsistency •Biologicslicenseapplication(BLA)preparation •Supportingdocumentsforlicensureinspections •Post-commercialprocessimprovement(CI) •Post-commercialprocessmonitoring With the growing prominence of biopharmaceuticals in the clinic (which currently amounts to more than 900), and a steady increase in approvals totalling more than $100 billion annually, there is a strong impetus within the industry to implement strategies that will accelerate clinical entry (1). In the current regulatory landscape, it often takes 10 years and billions of dollars to bring a drug candidate from development to the shelves (2). While it is typically desired to keep chemistry, manufacturing and controls (CMC) activities off the critical path for drug development, this situation cannot be avoided prior to clinical entry. Hence, the industry has shown increased interest in pursuing methodologies that can shorten the window for both process development and manufacturing. Some of these have arisen in the form of platform processes, high-throughput methods and single-use manufacturing technologies (3-5). In this article, we will focus on increasing experimental throughput in process development utilising high-throughput methodologies. Understanding the Processes Platform approaches have been successfully adapted for the rapid development of certain classes of therapeutics, such as monoclonal antibodies (mAbs). However, even for this well- established product class, what is gained in terms of speed is often lacking in terms of process knowledge, and the influence Shahid Rameez and Abhinav A Shukla at KBI Biopharma Figure 1: Clinical and process development/manufacturing activities during biopharmaceutical development and role of ambr™ in accelerating product development during the FIH phase of the biopharmaceutical development lifecycle
• 2.
  www.samedanltd.com38 of various parameterson procedure and product quality outcomes. Biosimilar developments present an even greater challenge. In this situation, a comparable bioanalytical profile is critical to achieve, and is significantly, influenced by cell culture process parameters.Thus, the obstacle in development is finding the right conditions to produce a molecule which matches the product quality attributes of the innovator. With conventional laboratory-scale bioreactors and shake flasks being the dominant forms of experimentation, the cell culture development stage becomes a resource- and time-intensive step. Mammalian methods typically have the longest experimental duration, with inoculum seed train and production culture stretching between four and six weeks. In order to test critical process parameters such as pH, dissolved oxygen and agitation, bioreactors must be used as shake flasks lack the necessary control capabilities. During the optimisation of a typical cell culture procedure, at least three to four rounds of 10-12 bioreactor runs need to be performed. This combination of experimental duration, and the extensive resources required to run multiple reactors in parallel, makes the cell culture process development phase a key bottleneck stage during process development. More importantly, to develop a robust cell culture process that ensures batch to batch product quality consistency, design of experiment based studies must be implemented in order to reveal the effect of cell culture changes on homogeneity, purity and post-translational modifications. These studies provide for a comprehensive process understanding, which subsequently aids the production of more consistent batches. However, employing this approach produces a large number of bioreactor runs and a vast amount of samples. This, in turn, can exceed the resources and capacity of cell culture and analytical labs, which primarily depend on conventional small-scale glass bioreactors (1-15L in size) and high-performance liquid chromatography (HPLC) and capillary electrophoresis-based separations to monitor protein quantification and product quality. As a result, there is a compelling demand for a high-throughput process development (HTPD) platform which facilitates key decisions during the early process development phase. Case in Point The ambr™ system was employed to make crucial process decisions during the development of a biopharmaceutical manufacturing process (6).The capability to fine-tune process controls with 24-48 single-use miniature bioreactor vessels provides a platform to use fractional factorial and minimum-run designs, enabling identification of key process parameters and interactions of those parameters. Moreover, the reproducibility and scalability of the system allows for high-throughput experiments in cell culture process development during the FIH phase of biopharma drug development, offering a significant possibility of decreasing the development timeframe prior to clinical entry (see Figure 1, page 36). In addition to this, two further high-throughput analytical technologies were integrated to enhance performance, workflow and ease of use. Table 1: Cell culture performance comparison between bioreactor systems (ambrTM , 3 and/or 15L glass bioreactors and 200L single-use bioreactor) for viability at harvest (%), titer (normalised), cell-maximum growth rate (1/d) and cell-specific productivity (pg/cell/d) for a mAb and a non-mAb. a: n = 3, b: n = 1, c: n = 4, d: n = 1, e: n = 2, f: n = 1, g: n = 1, h: measured from days 0-8 for mAb and from days 0-7 for non-mAb Figure 2: Comparison of time courses for viable cell growth for recombinant CHO cell lines in ambr™ vessel and other scales classical bioreactors; 3L and/or 15L glass bioreactors and 200L single-use bioreactor for (A) mAb and (B) non-mAb. The experimental data for ambr™ shows an average of 3 and 2 vessels in parts A and B respectively Viablecellcount(106 cells/mL) Viablecellcount(106 cells/mL) 0 02 24 46 68 Days mAbA B ambr ambr3L 15L 15L200L 200L non-mAb Days 810 1012 1214 1416 16 Bioreactor system Viability at harvest (%) Titer (normalised to 200L titer values) Cell maximum growth rateh (1/d) mAb ambra 90.27 ± 0.14 0.96 0.37 3-Lb 98.70 1.06 0.37 15-Lc 91.38 ± 2.19 0.88 0.34 200-Ld 90.20 1.00 0.34 non-mAb ambre 81.20 0.99 0.46 15-Lf 61.40 0.94 0.51 200-Lg 84.20 1.00 0.47
• 3.
  www.samedanltd.com 39 The HTPDapproach utilises high-throughput microbioreactors and analytics to accelerate product development. The HTPD method can be deployed throughout the entire process, starting with the selection of a clone during the cell line development. However, due to limitations in time and resources, relatively few top clones (the top 1-4) are evaluated in conventional bioreactors; this decreases the chance of identifying a high-producing clone with desired quality attributes. HTPD overcomes this restriction, while offering the capability to evaluate a larger number of clones (the top 24-48) in parallel under representative stirred tank bioreactor conditions. In particular, this broader screening benefits biosimilar programmes in which the objective is to identify a clone that is capable of producing specific product quality features. Furthermore, during the cell culture process development phase, HTPD allows the investigation of factors including pH, temperature, dissolved oxygen, nutrients in media and feeds, glucose, ammonia, salt and other metabolites that have shown to affect the productivity and product quality of proteins. In Figure 4 (page 40), data is shown from two case studies demonstrating the HTPD approach during cell culture process development for a biosimilar.The aim of case study 1 (part A) was to evaluate eight different feeds for Chinese hamster ovary (CHO) cell line producing a biosimilar.This was followed by case study 2 (part B), which was a DOE. As an example, Figure 4 shows one specific glycan structure (G0F) from these case studies – a critical quality attribute in this biosimilar – and the change it undergoes during various tested process conditions. Based on these results, the conditions that do not allow the G0F to remain within the value ± (more than, or equal to the) variability of the originator molecule were not carried forward. Therefore, feeds 3, 7 and 8 (see Figure 4, part A) were not evaluated further. Moreover, the selected feed showed strong interaction with respect to process pH Figure 3: Comparison of (A and B): mAb X and mAb Y concentrations in eight 3L glass bioreactors using Octet™ and Protein A HPLC methods; (C and D): mAb C charge variants using LabChip™ and conventional cation exchange chromatography (CEX) method; (E and F): mAb C N-glycan species using LabChip™ and conventional fluorescence-based hydrophilic interaction liquid chromatography (HILIC) method. The variability between HTPD and conventional methods was less than 5% mAb X ProA HPLC 3L bioreactor number 3.50 A B 4.00 3.50 3.00 2.50 2.00 1.50 1.00 0.50 0.00 3.00 2.50 2.00 1.50 1.00 0.50 0.00 B01 B01B02 B02B03 B03B04 B04B05 B05B06 B06B07 B07B08 B08 3L bioreactor number ProA HPLCOctet Octet Antibodyconcentration(g/L) Antibodyconcentration(g/L) mAb Y Fluorescence 600 500 400 300 200 100 0 30 35 Time (sec) Free Dye Basic 50 10 0 20 40 30 60 70 80 Microchip CZE (mAb C) Percentage CEX (mAb C) Main Acidic Main AcidicBasic 40 45 Fluorescence 4000 3000 2000 1000 0 8 9 10 11 12 Multiplate overlay electropherogram (mAb C triplicate measurements) Man5 50 10 0 20 40 30 60 70 MicroChip N-glycan Percentage HILIC N-glycan Man5G2G0G2FG1FG0F G0 G0F G1F G1F G2F Size (CGU) C E D F
• 4.
  Abhinav A Shuklais Senior Vice President for Development and Manufacturing at KBI Biopharma, a rapidly growing CDMO organisation. Abhinav has over 18 years of experience in developing biopharmaceuticals from early development through commercialisation launch. Most recently, he was Director of Manufacturing Sciences andTechnology at Bristol-Myers Squibb, where he led the successful commercialisation of Yervoy® , Nulojix® and Orencia® , in addition to other late-stage programmes. He also led the creation of a platform approach for mAbs at Amgen. Abhinav is a known industry expert in biopharmaceutical development and has over 40 publications in this area. He serves on the editorial boards of several journals, including Biotechnology and Applied Biochemistry and Bioprocess International. He has a PhD in Chemical and Biochemical Engineering from Rensselaer Polytechnic Institute. Email: ashukla@kbibiopharma.com Shahid Rameez, PhD, is a Principal Scientist for Process Development at KBI Biopharma Inc. Shahid has experience in CMC activities responsible for design, development and implementation of efficient and robust cell culture processes for small- and large-scale manufacturing using various expression systems. Prior to joining KBI in 2012, Shahid completed his doctoral and postdoctoral studies at Ohio State University on designing cellular protein therapeutics. He has over 10 publications in the area of process development and protein therapeutics. Email: srameez@kbibiopharma.com About the authors 40 to control the critical quality attribute in this biosimilar (see Figure 4, part B). Both these studies helped to assess product quality metrics from cell culture process development, while identifying the right conditions to produce the molecule with matching product quality attributes to the innovator. Method Analysis The multi-stage nature of process development and the long duration of mammalian cell culture experiments make it a time- and resource-intensive procedure. Of the approaches tested, the HTPD method offers a realistic possibility of decreasing the timeline for process development experimentation, which, in turn, decreases the timeframe for manufacturing clinical material prior to clinical entry. In addition, material needs and other resources are minimised, and thus a larger number of drug candidates can be advanced into the clinic faster, in order to address unmet clinical needs. Acknowledgements The authors would like to thank Joe McMahon, Chief Executive Officer of KBI Biopharma Inc, for his support for this work. Members of the process development, analytical development and formulation sciences teams at KBI Biopharma Inc are thanked for providing support. We would also like to thank Srivatsan Gopalakrishnan, Carl Zhang, Jaspreet Notey, Christopher Miller, Derek Ryan, Nathan Oien, James Smedley and Sigma Mostafa for their help with this article. References 1. Walsh G, Biopharmaceutical benchmarks, Nature Biotechnology 28(9): pp917-924, 2010 2. Gottschalk U, Brorson K and Shukla AA, The need for innovation in biomanufacturing, Nature Biotechnology 30(6): pp489-492, 2012 3. Shukla AA and Thömmes J, Recent advances in large-scale production of monoclonal antibodies and related proteins, Trends in Biotechnology 28(5): pp253-261 4. Shukla AA and Gottschalk U, Single-use disposable technologies for biopharmaceutical manufacturing, Trends in Biotechnology 31(3): pp147-154, 2013 5. Rege K, Pepsin M, Falcon B, Steele L and Heng M, High-throughput process development for recombinant protein purification, Biotechnology and Bioengineering 93(4): pp618-630, 2006 6. Rameez S, Mostafa SS, Miller C and Shukla AA, High-throughput miniaturized bioreactors for cell culture process development: Reproducibility, scalability, and control, Biotechnology Progress (30): pp718-727, 2014 Figure 4: Percentage (normalised to innovator value) of specific glycan structure (GoF) in case studies 1 (A) and 2 (B), a critical quality attribute in the biosimilar, and change it undergoes under various tested process conditions. Based on these results, the conditions which do not allow the GoF to remain within the value ± vavariability (blue region) of the originator molecule were not carried forward 120 120 A B 110 110 100 100 90 90 80 80 70 70 60 60 1 Feeds %GoFofinnovatorproduct %GoFofinnovatorproduct 2 3 4 5 6 7 8 7.00 ± 0.2 FDSA FDSB FDSC FDSD 7.00 ± 0.1 7.00 ± 0.05 www.samedanltd.com pH range