Antibody Purification: Drivers of Change

ONLINE EXCLUSIVE |

By Narahari Pujar, PhD; Duncan Low, PhD; and Rhona O’Leary Shortening development times while better understanding process
Editor’s Note: This article is excerpted from “Antibody Purification: Drivers of Change,” a chapter in Process Scale Purification of Antibodies. The book, which was edited by Uwe Gottschalk, PhD, was published last year John Wiley & Sons, which also publishes Pharmaceutical Formulation & Quality.

Antibody purification will have a robust future based on the strong pipeline of products in development and the wealth of diverse strategies for purification, as discussed in the preceding chapters. These strategies will continue to be refined, and in some cases replaced, as technologies are developed offering superior performance, or the same performance with lower costs. That being said, what factors are likely to drive continued change?
At the highest level, the eternal business challenge is to continue to do more with less. For antibody purification, this means continuously shortening development times while simultaneously gathering more and more data to support process understanding and to improve process economics, all with increasing resource constraints and zero tolerance toward compromises on safety. Economic pressures are increasing at a time when public perception of the biopharmaceutical industry is lower in confidence with regard to safety, and cynical about costs. The largest biotechnology companies increasingly resemble large pharmaceutical companies as they face the inevitable problems, associated with scale, of slowing growth rates and increased competition. In this environment, many companies are forced to make process development decisions earlier in the product life cycle, and as such are making more risk-adverse decisions with respect to technology implementation.
Upstream development has responded heroically, and titers as high as 10 g/L have been reported, although lower titers may be preferred if they can be achieved in shorter time frames.¹ Increasing titer reduces material costs and increases equipment and facility utilization upstream, but increases the relative costs downstream. From a purification perspective, costs are driven by the mass of material to be purified, and as titers increase, so must the scale of columns (and to a lesser extent filters) to keep pace.
The current strategy used to commercialize monoclonal antibodies (mAbs) is to develop a robust platform process that can be applied as broadly as possible to different candidate molecules in order to avoid spending time refining details when the most important priority is to test the candidate in the clinic.^2,3 Ideally, once the efficacy of the molecule has been established, the commercial process can be developed and characterized.

The largest biotechnology companies increasingly resemble large pharmaceutical companies as they face the inevitable problems, associated with scale, of slowing growth rates and increased competition. In this environment, many companies are forced to make process development decisions earlier in the product life cycle, and as such are making more risk-adverse decisions with respect to technology implementation.

However, since the major drivers are speed and efficient use of resources, in many cases the approach has been to continue to scale up the platform approach (on which the clinical data is based) rather than to develop a truly cost-effective commercial process. Many companies are also choosing to mitigate risks by postponing optimization until the potential for commercialization has been assured, and then submitting post-approval modifications. The area of post-approval process changes will be discussed later in this chapter.
Most mAb processes today are based on the large-scale use of Protein A.
Scaling up these processes to cope with larger quantities has clearly been shown to be technically feasible, and indeed sensitivity analyses have shown that there are opportunities for significant savings in material costs.⁴ However, purification remains the most significant cost and is dominated by the cost of Protein A resins, which over the lifetime of a product may amount to hundreds of millions of dollars.⁵
Non- affinity-based antibody processes are discussed earlier in the book. From a technical and regulatory perspective, eliminating Protein A is perfectly feasible. However, most companies choose to continue with the expensive Protein A resin, since it is a very efficient and reliable step in the process, achieving high product quality and yield, and removing host cell protein (HCP) and viruses. If Protein A is not used, additional steps (resins or filters) are often needed in the process to achieve the same product quality. At the end of the day, many manufacturers will say that additional steps (equating to cleaning of equipment, capital, labor, etc.) are more costly than the cost of the Protein A resin. In that situation, many manufacturers will argue that the merits of Protein A outweigh the costs.
Innovations in process technology will also have a significant impact on current and future facilities in terms of the fit compared to current processes, and the relative allocation of space required for purification and associated buffers and cleaning solutions. Facility costs remain the most significant production costs, and any improvements in productivity and facility utilization have a significant benefit, assuming they can be captured without excessive disruption.
Technical alternatives to the standard platform approach to antibody purification are certainly emerging, but the reluctance of manufacturers to adopt them is a reflection of the conservatism in our industry. This conservatism is driven by two factors—the regulatory burden of making changes and whether a new technology/approach fits within existing facility designs. Few manufacturers will embrace a technology if it makes their current facilities obsolete, or if the technology is for a single process and is not applicable across multiple products. The process economics have to be very compelling to embrace change.
This conservatism has led to an incremental approach to process improvements and technology adoption. Along with continued improvements in the current technology, this conservatism will remain a barrier to innovation for some years to come. Some of the technical alternatives have been reviewed recently, so rather than reiterate and reassess these alternatives, we shall instead focus on some of the topics that influence change and that may facilitate the stepwise incorporation of improvements.^6,7 These include, but are not limited to, the changing regulatory environment, coupled with the ever-increasing ability (and requirement) to generate data, the emergence of innovative analytical and control technology, process economics, the globalization of our industry, and finally, follow-on biologics (FOBs).

The Changing Regulatory Environment

During the first part of this decade, the Food and Drug Administration (FDA) set out to reshape the way the industry approached manufacturing science and technology. In particular, the industry was seen as wasteful and inefficient, and quality was perceived as being achieved through test and rejection rather than from well-designed and well- controlled processes. A recent study supports this view.⁸ This has led to Pharmaceutical cGMPs for the 21st Century, an initiative for product quality regulation that utilizes a science- and risk-based approach, thus facilitating innovation.
The elements of this approach are based on improvements in process design, where quality is designed into the process from the outset [quality by design (QbD)]. This is based on a thorough scientific understanding of the relationship between the product and the process by which it is made. Critical product quality attributes (CQAs) are monitored during the process, either directly or through surrogate markers, and critical process parameters (CPPs) are controlled through a responsive process to reduce variability and to provide greater quality assurance. The aim is to encourage manufacturers to innovate and implement the most recent scientific advances into manufacturing processes by reducing the regulatory burden, based on increased understanding that the manufacturer has gathered throughout the life cycle of the product from development through to commercial manufacture. Shorter cycle times, less waste, improved automation and reduced human error, and real-time product release are all benefits that could be gained.

The current strategy used to commercialize monoclonal antibodies (mAbs) is to develop a robust platform process that can be applied as broadly as possible to different candidate molecules in order to avoid spending time refining details when the most important priority is to test the candidate in the clinic.2,3 Ideally, once the efficacy of the molecule has been established, the commercial process can be developed and characterized.

Implementing change in pharmaceutical processes, even as an improvement, has been seen as difficult, and often requires clinical data to permit implementation. QbD introduces the concept of design space, which describes the relationship between process inputs in terms of variables and process parameters, and product CQAs. A systematic approach to process characterization is required, which begins with risk assessment and prioritization, process modeling (at the laboratory scale) and characterization, and the setting of acceptance criteria.^9,10 Theoretical approaches are useful for modeling chromatographic steps.^11,12 Also, experience from earlier, similar processes should be leveraged. Development of process design space takes place after the decision to commercialize a molecule during process characterization since there is little point in performing extensive experimentation before efficacy is established.
Once an acceptable design space (within which satisfactory product quality is achieved) has been established, a manufacturer should then have flexibility to modify the process and to move within that design space with less regulatory burden. This provides significant benefits to the manufacturer, permitting greater freedom to make improvements to the process and facilitating the transfer of processes from one site to another and even from one manufacturer to another, provided the appropriate analytical tools and assays are available.
Exactly how these initiatives will be implemented is still the topic of debate.
In addition to scientific literature, a framework of practices, standards, and guidelines is being worked on by professional societies and regulators. The International Conference of Harmonization (ICH) develops guidelines for quality, safety, efficacy, and multidisciplinary topics. Section Q8, Pharmaceutical Development, describes the concept of design space, whereas Section Q9 provides guidelines for risk management. International standards organizations such as the American Society of Mechanical Engineers (ASME) and American Society for Testing and Materials (ASTM) International are developing consensus standards for implementation approaches, and professional societies such as the Parenteral Drug Association (PDA) and the International Society for Pharmaceutical Engineering (ISPE) are developing technical reports, guides, and white papers for specific operations. These documents provide an important framework for the more detailed interpretation of ICH Guidelines, and consensus standards in particular give users a level of confidence in how practices will be accepted by regulatory authorities.¹³
Combining QbD with process analytical technology (PAT) allows for the development of a dynamic process that can respond to input variability. It follows that such a process should be able to respond to changes better and to facilitate incorporation of improvements in process technology, be they changes in equipment or even in critical processing aids such as resins and filters.⁶
Dr. Pujar, is in bioprocess R&D at Merck & Co.; Dr. Low is in process development at Amgen; and Dr. O’Leary is in early stage purification at Genentech. For more information, e-mail Dr. Pujar at hari_pujar@merck.com.

REFERENCES

1. Luan YT, Wang W, Nolan R, Drapeau D. (2006). Defined medium development for high yielding mammalian cell culture processes. Presentation at the IBC and Bioprocess International Conference, San Francisco, CA, 2006.
2. Fahrner RL, Knudsen HL, Basey CD, Galan W, Feuerhelm D, Vanderlaan M, Blank G. (2001). Industrial purification of pharmaceutical antibodies: development, operation, and validation of chromatography processes. Biotechnology and Genetic Engineering Reviews 18, 301-327.
3. Shukla A, Hubbard B, Tressel T, Guhan S, Low D. (2007). Downstream processing of monoclonal antibodies: Application of platform approaches. Journal of Chromatography B 848, 28-39.
4. Kelley B. (2007). Very large scale monoclonal antibody purification: the case for conventional unit operations. Biotechnology Progress 23, 995-1008.
5. Farid S. (2007). Process economics of industrial monoclonal antibody manufacture. Journal of Chromatography B 848, 8-18.
6. Low D, O’Leary R, Pujar H. (2007). The future of antibody purification. Journal of Chromatography B 848, 48-63.
7. Thömmes J, Etzel M. (2007). Alternatives to chromatographic separations. Biotechnology Progress 23, 42-45.
8. Macher J, Nickerson J. (2006). Pharmaceutical Industry Wastes $50 Billion a Year due to Inefficient Manufacturing. McDonough School of Business, Georgetown University, Washington, DC, and Olin School of Business, Washington University, St. Louis, MO.
9. Seely R, Seely JE. (2003). A rational, step-wise approach to process characterization. BioProcess International 1, 24-34.
10. Van Hoek P, Harms J, Wang X, Rathore A. (2008). Case study on definition of design space for a microbial fermentation step . In Rathore AS, Mhatre R. (eds.) Quality by Design For Biopharmaceuticals: Perspectives and Case Studies, Wiley Interscience (in press).
11. Kaltenbrunner O, Giaverini O, Woehle D, Asenjo J. (2008). Application of chromatographic theory for process characterization towards validation for an ion-exchange operation. Biotechnology and Bioengineering 98, 201-210.
12. Mollerup J, Hansen T, Kidal S, Staby A. (2007). Quality by design — Thermodynamic modeling of chromatographic separation of proteins. Journal of Chromatography A 1177, 200-206.
13. Watts C, Clark J. (2006). PAT: Driving the future of pharmaceutical quality. Journal of Process Analytical Technology 3, 6-9 .