Proceedings of the Production of Monoclonal Antibodies Workshop

August 29, 1999 | Bologna, Italy

This document has been adapted with permission from a publication created by the Alternatives Research & Development Foundation (ARDF)

Commerical Production of Monoclonal Antibodies

Simon J. Y. Saxby, BSc
Director of Contract Services, Unisyn Technologies
25 South Street, Hopkinton MA 01748

Introduction

Commercial manufacturing of monoclonal antibodies has experienced a renaissance in the last few years due to the success of a number of antibody-based therapeutics, including Cytogen's ProstaScint MAb, Centocor's Reopro and Remicade antibodies and IDEC's Rituxin. There are an estimated 750 protein/antibody based biotechnology products in development, with approximately 200 of these in late-stage clinical trials.

The most extensively used technique to date for the production of MAbs is in vivo production in ascitic fluid. The ascites production of hybridoma-derived monoclonal antibodies has been in routine use since the late 1970's following the Nobel Price winning Köhler and Milstein1 publication in 1975. Some twenty years later, there is still debate over the practical, financial and moral issues of manufacturing monoclonal antibodies in vivo compared with in vitro methods. The extensive use of the ascites method is due to the fact that the technique is relatively simple and was initially inexpensive when compared to the in vitro alternatives available at the time. The technique does however require specialist animal handling and aseptic technique skills, and contrary to popular perception, also has a number of significant disadvantages when compared to many modern in vitro technologies. These issues will be addressed, but only with regard to the manufacture of MAbs for commercial purposes.

The overall efficiencies and practicalities of the ascites and in vitro methods depend not only on the upstream production phase of the product, but also on the subsequent commercial use of the antibodies and whether any downstream processing of the antibodies is required or desirable. This discussion assumes that the hybridomas are "average" secretors and secrete antibodies at a rate of 20 mgs per Liter, (20 (g/ml), in static culture. It is also assumed that the antibody yields post-purification are 80% for diagnostics and 50% for therapeutics.

There has recently been a number of extremely informative articles published describing the numerous in vitro products and systems available for the production of monoclonal antibodies. A number of these papers provide comprehensive data comparing the ascites production of monoclonal antibodies with the in vitro alternatives available2,3,4,5. For the purposes of this discussion only those technologies and methods that are most appropriate for commercial production of MAbs will be highlighted.

Monoclonal Antibody Production for Diagnostic and Human Use

Monoclonal antibodies are manufactured commercially in considerably variable quantities for use as research tools, diagnostics, (in vitro), and in vivo and ex vivo purposes. It is difficult to categorize the quantities produced as small, medium and large, or to assign specific quantity ranges to research, diagnostic and therapeutic needs. There are always exceptions to these generalizations that are important in the context of the discussion about the validity of the various production methods available. Nevertheless, based on personal industrial experience, I will attempt to define some of these categories, quoting examples where there may be doubt surrounding the inclusion of a particular category or definition.

With the possible exception of many of the antibodies produced for the research reagent market, the commercial production of MAbs for the diagnostic and therapeutic markets requires that the manufacturers follow a number of stringent regulatory guidelines and controls. Many MAbs aimed at the commercial research reagent market are produced by companies and research organizations that are not in compliance with regulatory guidelines. It is doubtful whether the quantities and applications warrant such compliance. The implementation of these types of controls for many of these reagents would price them out of the range of many researchers' budgets and severely limit their availability to the research community.

The controls for therapeutic and diagnostic antibodies for human use are published in the FDA "Points to Consider in the Manufacture and Testing of Monoclonal Antibody Products for Human Use", (FDA, 1997). These "Points to Consider" (PTC) guidelines are designed to ensure the safety of the products for human use by guaranteeing lot-to-lot consistency and supply for the commercial lifetime of the product. The FDA, and many of the World regulatory authorities will generally not approve a therapeutic or clinical diagnostic product for commercial use if there is any doubt about the long-term availability of the product.

The guidelines governing the manufacture of MAbs for in vitro diagnostics are less stringent, requiring fewer safety tests. However, MAbs produced for use in in vitro clinical assay applications are nevertheless required to be produced in compliance with Good Manufacturing Practices (GMPs) to ensure traceability of the product and reliability of the results that are frequently indications of life-threatening diseases, such as cancer.

In order to comply with the FDA "Points to Consider" requirements dealing with guarantee of supply and consistency, manufacturers must perform the basic cell culture and clonal selection work that will also invariably guarantee good productivity in in vitro culture. When this work is performed correctly at all stages of the cell line and product development, manufacturers invariably overcome one of the major justifications for the continued use of ascites, namely that the cell lines do not "perform" well or cannot be cultured in vitro.

In Vitro Technologies

The production requirements and the appropriate production technologies (based on quantities required) can be broadly divided into the following groups:

The list of in vitro technologies that are available today for the production of MAbs is extensive, but can be broadly categorized according to the descriptions below. These are by no means the only in vitro technologies available that would provide the capacity needed for commercial production of many diagnostic and research reagent antibodies. Some of these additional technologies are as simple as sealed sterile bags, the Wave Bioreactor, and the miniPerm bioreactor. The miniPerm is a modified roller bottle with a semi-permeable membrane that separates the cells and secreted antibodies from the bulk feed medium -- thereby raising the concentration of the antibodies in the harvested medium. These and many other systems are excellently described in publications by Grant Meisenholder6 and Jackson et al2,3,4.

T-Flasks

These are the conventional static culture flasks with surface areas of up to 225 cm2 per flask and a potential culture volume of approximately 400 mls for most hybridoma cell lines. This would provide an approximate yield of 8 - 10mgs / flask. Flasks are easily scaleable in standard CO2 and "dry" incubators. The skills required are basic sterile culture techniques, no more advanced than those required to initially generate the hybridoma cell lines, and it is relatively simple for most tissue culture laboratory staff to handle up to 50 or more flasks simultaneously.

Roller bottles

These are the standard roller bottles with a standard surface area of 850 cm2 per roller bottle. Roller bottles are easily scaleable with a roller "machine" with an approximate yield of 20 - 30 mgs per roller bottle. Again, the skills required are similar to those required to culture cells in flasks, and it is relatively simple to routinely produce several hundred milligrams of antibodies in roller bottles. One must acknowledge the fact that cold room storage capacity is required to store the bulk supernatants before the antibodies are purified or incorporated into kits, but these facilities are standard for most commercial manufacturers of antibodies. Robotic systems are available and in commercial use that enable the economic production of multi-gram quantities of antibodies. A number of therapeutic proteins are manufactured on a very large scale in roller bottles using robotic systems, and therefore roller bottles have to be considered as a technology capable for the production of hundreds of grams of protein per year. One such example is Johnson and Johnson's multi-kilogram and multi-billion dollar drug EPO.

Spinner cultures

These are small-scale stirred culture vessels that work well for suspension cell cultures, including most hybridomas. These systems can again be easily scaled up to a 2.0-liter vessel size with stirrer bases that can maintain up to four vessels per base. With minimal additional training it is easy to generate several hundred milligrams of antibodies using stirrer systems, with the possible additional requirement for a warm (37°C) room for further scale-up.

Perfusion and hollow fiber bioreactor systems

These are also known as high cell density bioreactor systems. Specialized training and skilled operators are required to operate these systems efficiently, and systems are available in scales ranging from hundreds of milligrams to hundreds of grams per year. State-of-the-art hollow fiber bioreactor systems are arguably the most efficient systems available for the production of up to 1-2 kilograms of antibody.

Stirred tank and Airlift bioreactor systems

These are also referred to as "fermentors" even when used for mammalian cell culture. Specialized training and skills are required to operate these systems efficiently, together with a significant capital investment corresponding to the size of the facility. These systems represent the highest scale of in vitro antibody production routinely available today, and stirred tanks are regularly used to produce multi-kilogram amounts of single products every year.

Current Trends

It is generally acknowledged that in vitro production of MAbs is preferable to producing antibodies in ascites. One must distinguish between the practical reasons for producing antibodies in ascites and the desire to continue with ascites production simply because it is an "established" technique. Many of the arguments in favor of continuing with routine ascites are based on historical data that do not take into account recent technology developments and an abundance of new data that demonstrate the superiority of in vitro systems when compared with ascites production.

It is also generally accepted that the following criteria are nevertheless valid for the continued use of the ascites method:

The main reason for the continued use of routine in vivo produced antibodies is the perceived economic advantage of ascites production when compared with the production costs using an in vitro method. The National Research Council13 report states that "when all fully loaded production and pre-production and post-production costs are considered for a commercially viable line, economics usually favor in vivo production." However, the report also reveals that there is considerable doubt about whether the fully loaded costs for in vivo production are included in the comparative calculations. The report also does not define the scale at which the in vivo method is more cost effective, and states that as the quantity required increases, in vitro technologies become more economical.

Technology Summary

PRODUCTION SCALE
  1.0 gram
85% of MAbs produced		 1.0-10.0 grams
10% of MAbs produced		 10.0-100 gram
3% of MAbs produced		 100 gram-2.0 Kg
1% of MAbs produced		 >2.0 Kg
<1% of MAbs produced
Applicable In vitro Production Technologies T- Flasks, Roller bottles, Spinner cultures, Perfusion and Hollow fiber bioreactors. Roller bottles, Spinner cultures, Perfusion and Hollow fiber bioreactors. Roller bottles, Hollow fiber bioreactors, Stirred tank bioreactor systems. Roller bottles, Hollow fiber bioreactors and Stirred tank and Airlift bioreactors. Roller bottles, Stirred tank and Airlift bioreactors
Number of Mice required -- assuming 20 mgs/mouse <50 mice 50 - 500 mice 500 - 5,000 mice 5,000 - 100,000 mice >100,000 mice

The report acknowledges two highly significant factors with regard to ascites production costs. Firstly, it states that one of the reasons why in vitro technologies "can become" more economical is because the costs associated with selecting the best clone in terms of growth and secretion can be amortized over a larger production amount. This is certainly reasonable. However, any commercial organization that initiates a product development program without assuming that the product is going to be a commercial success is likely to fail. This being the case, the selection of the best clone should be one of the priorities at the initiation of the product development cycle. When this selection process is performed using all of the appropriate selection procedures, the resulting cell line will not only be stable for commercial production, but it will be stable and adapted to commercial production in an in vitro system. In over sixteen years of producing and cloning hundreds of hybridomas and manufacturing similar numbers of MAbs for commercial use, I have never been unsuccessful in selecting a clone that will grow and secrete antibodies in in vitro systems by following this principle. A lack of effort or good science are not justifiable reasons for allowing the continued routine use of ascites production, and scientists working in research organizations and academia must also be encouraged to follow these guidelines when embarking on hybridoma/monoclonal antibody programs.

Allowing routine ascites production because "...it is a much more forgiving procedure..." -- as stated in the NRC13 report, is equally unacceptable for a commercial product. Similarly, the belief that even if a cell line has not been stabilized antibody production in ascites is guaranteed is unsound. An unstable cell line is equally susceptible to decreasing antibody yields due to loss of antibody-secreting clones and overgrowth by non-secreting clones (negative variants) whether the cells are routinely grown in ascites or in vitro culture. Safety, long-term supply, lot-to-lot consistency and product quality cannot be sacrificed in favor of incomplete product development.

Secondly, when discussing the relative costs associated with ascites and in vitro production, the NRC13 report acknowledges that the costs for ascites "...might not include all factors, such as animal housing costs and technician time." This is undoubtedly true in many cases, and raises the concern of the validity of much of the in vivo cost data that are frequently used to justify the continued routine use of the ascites method. These costs are substantial when MAbs are produced in compliance with the "Points to Consider" guidelines. The rigorous animal stock controls, health monitoring and animal and testing programs are expensive and labor intensive. The products are then subject to at least the same, if not more stringent, safety testing requirements.

The cost data available for the production of MAbs in ascites largely assume the use of a common strain of mice, namely Balb/c mice. This is because the myeloma fusion partners in general use around the world for the generation of hybridoma cell lines are derived from Balb/c mice. The production of MAbs from non-Balb/c derived donor lymphocytes can be considerably more expensive. In order to optimize yields, and in some cases achieve any yields at all with "cross-strain" hybridomas and heterohybridomas, cross-strain F1 generation mice or SCID mice are required. Not only are these animals more expensive to buy, but SCID mice are also more expensive to maintain and are highly susceptible to infections by adventitious agents such as viruses and bacteria. This increases the concerns of product safety and suitability for use in humans.

The improved efficiencies of many of the in vitro technologies now available enable them to compete very effectively on an efficiency and cost basis with ascites for commercial production of MAbs. At a production scale of less than two grams, technologies like hollow fiber bioreactors can significantly lower the upstream production costs of MAbs. This is particularly true when the overheaded costs of the facilities and animal husbandry are included for ascites production as described above.

The in vitro alternatives available for MAb production are cited in the NRC13 report as being complex and requiring skilled technicians to operate them. Contrary to being a disadvantage, it is in the best interests of the industry and the public that products for human use are manufactured by skilled staff. The report states that "ascites production is a simple procedure, once proper technique is learned". This statement applies equally well to in vitro systems, and in cases of flasks and roller bottles, the staff that are capable of culturing cells for injection into mice would require no additional training to produce reasonable quantities of antibodies.

Regulatory Changes

Recent changes in the FDA regulations now make it much easier for healthcare companies to have their products produced by one or more contract manufacturers. This has led to a rapid growth in the contract manufacturing industry, which offers significant advantages to companies trying to commercialize new antibody-based products. The contract manufacturers have the skills, experience and technology to develop the manufacturing process and produce the MAbs using a variety of in vitro technologies. In some cases, the contract manufacturers have the flexibility to use technologies desired or selected by the client. The client companies avoid the risks associated with the capital expenditure, the expansion and training of staff to develop an in-house manufacturing process for a product that may not reach commercialization. Because the contract manufacturer is likely to have extensive experience in manufacturing process development for MAbs, it is also likely that they will be able to help accelerate the product through the clinical trial process and early commercialization using a more regulatory and consumer "friendly" technology.

Production Yields and Functionality

Many MAbs can now be manufactured in serum or protein-free media in vitro more cost effectively than they can be produced in vivo. This has been demonstrated by a number of companies that have produced antibodies in ascites for early screening and then developed and scaled up production in vitro7. An optimized purification process, that may require fewer purification steps, will frequently yield higher quantities of antibodies when produced in such an in vitro system. This has a significant effect on reducing the manufacturing costs of the antibodies.

There are concerns over the possible loss of functional activity in antibodies produced in vitro. While some changes may occur when the method of production is altered or scaled-up, glycosylation being one example, these are generally not associated with a loss of functional activity. As the antibodies that are initially selected on the basis of functional activity and affinity to the target antigen, (amongst a number of selection criteria), are selected from in vitro cultures, it is highly unlikely that a suitable in vitro production method cannot be developed from the range of technologies available. Any loss of functional activity is more likely to be associated with cell line instability due to inadequate clonal selection during development. In contrast to in vitro production adversely affecting the quality and immunoreactivity of antibodies by altering glycosylation patterns, glycosylation can be deliberately influenced by the in vitro culture conditions. This can result in the production of antibodies with a specific glycosylation structure and enhanced functionality8.

Product Quality and Safety

There are additional disadvantages to the ascites method of production that are also related to functional and safety issues. It has been reported by a number of groups that MAbs produced in ascites can be less immunoreactive when compared with the same antibodies produced in vitro8.

In vitro systems are capable of producing antibodies with enhanced immunoreactivity when compared to batches produced in ascites. It has been reported in a number of instances that it has been possible to double the number of diagnostic kits produced from a given quantity of antibody due to the enhanced activity (unpublished data). This enhanced immunoreactivity is almost certainly due to the absence of contaminating mouse immunoglobulins and cytokines present in the ascites produced material.

In vitro processes make it easier to control lot-to-lot consistency as required under GMPs. A number of GMP approved products are still produced in ascites, but the manufacturers of these products will acknowledge lot-to-lot yield inconsistencies. These yield inconsistencies can lead to product delivery delays and increased manufacturing costs.

Product safety is a major issue for manufacturers of biological drugs for therapeutic use. The recent advances in in vitro culture technologies, as well as the development of a number of very productive serum-free and protein-free media have made it possible to manufacture antibodies in the absence of any animal-derived proteins. While the FDA and other regulatory authorities do not expect manufacturers to convert previously licensed products to a serum-free or protein-free manufacturing process, they will expect new drug development to incorporate these technologies into the processes because of the enhanced safety of the products. These technologies offer two main advantages when considering product safety.

The European Initiative

The Council Directive 86/609/EU9 and the European Convention for the Protection of Vertebrate Animals Used for Experimental and other Scientific Purposes10 require that in vitro alternatives be used whenever reasonably and practically available within EU countries. The Animals (Scientific Procedures) Act 198611 effectively implements the Directive 86/609/EU in the United Kingdom where a Home Office project license is required by all individuals wishing to carry out animal experiments, including the production of antibodies in ascites. In addition to the 1986 Act, the UK Home Office issued further advice on protocols requiring the use of animals12, stating that ascites production may be allowed on a one-off basis requiring less than 20 mice. Despite this "allowance", the UK Home Office still requires that in vitro alternatives for MAb production be used whenever they are available and practical.

In addition to these regulations, there exist many other laws and recommendations within individual European countries that govern animal experimentation with the specific purpose of restricting the use of animals in experiments to an absolute minimum. In most instances the production of ascites is not regarded as "experimentation" as there is a known or pre-defined outcome, and permission to use animals for ascites production requires extensive justification on the part of the petitioning scientists.

The overall effect of these regulations and the banning of the "routine production" of MAbs in ascites in many European countries have led to some interesting, and in some cases disappointing developments. The use of the word "production" in the legislation has led to continued debate and has provided a loophole for many organizations to continue using ascites produced antibodies. Organizations and individuals in countries that are subject to the ban on the routine production of ascites are able to export cells to contract ascites producers outside their own borders, and simply import the antibody-containing ascites for further in-house processing and/or incorporation into diagnostic kits etc. As a result, many cell lines have been shipped to countries like the U.S. and Belgium for ascites production, and a number of contract ascites facilities have appeared in Eastern European countries still free from a legislative ascites ban.

It is expected that the EU will introduce legislation requiring that products be clearly labeled to show whether they contain in vivo or in vitro produced components. The likely outcome will be a loss of market share for companies that continue to have their antibodies produced in ascites whenever an in vitro alternative is available. This possibility has already provided a number of US companies with the incentive to begin converting their MAb-based diagnostic products to in vitro production methods. Their goal is to avoid loss of market share in Europe and to ensure that they have the time to develop the in vitro manufacturing processes well in advance of the legislation.

Emerging Technologies

There are a number of emerging technologies that are in various stages of development that will undoubtedly have an impact on the options available for the commercial production of antibodies and recombinant proteins in the future. Many of these emerging technologies are the subject of patent protection, but will become more widely available and used in the future through collaborations, licensing and royalty partnerships within the healthcare industry.

Phage Display

Techniques have been developed that enable the expression of large numbers of antibody molecules or "libraries" on the surfaces of filamentous bacteriophage particles. Using affinity selection with a variety of target antigens such as recombinant proteins and/or intact cells, antibodies with pre-defined specificities and high affinities can be selected. The technique enables the construction of libraries from immunoglobulin genes of any species, including human, and in many cases antibody selection can take place without the need for prior immunization of a lymphocyte donor. The selected antibody fragments can be recloned into a variety of vectors that allow the production of custom-designed molecules, including whole immunoglobulins of pre-defined isotypes.

Transgenic Animals

Transgenic animals are currently used for the production of recombinant proteins, including monoclonal antibodies. The FDA has yet to approve a therapeutic antibody that is produced in transgenic animals, but there are a number of transgenic products currently in late-stage clinical trials, including Genzyme Transgenic Corporation's (GTC) Anti-thrombin III. The technique has the potential to produce multi-kilogram quantities of proteins at low cost. It is also the only recombinant technique capable of producing naturally amidated peptides, eliminating the need for in vitro enzymatic amidation. ("Transgenic Animals for Production of proteins" GEN, 19 Number 9, May 1, 1999).

Transgenic Plants

Transgenic plant technology is also being developed for the production of recombinant proteins, but the technology may also lend itself to the economical production of whole antibodies or antibody fragments for diagnostics or therapeutics. Indeed, the first plant-derived therapeutic proteins to be injected into humans for clinical trials have already been produced.

Recombinant Protein Production

In addition to the advantages of producing monoclonal antibodies using the in vitro technologies described, many if not all of these technologies can also be used for the production of recombinant antibodies and proteins. The cost of the technologies is therefore spread over a number of different products.

Summary

Much debate has ensued over the banning of ascites production in the U.S., leading to the NIH sponsored Committee on Methods of Producing Monoclonal Antibodies and the subsequent publication of the National Research Council (NRC) report in April 199913. The report largely overlooks the word "routine" when considering whether there should be a ban on the production of MAbs in ascites. The legislation in Europe is based on an assumption of "reasonableness" and reality. It is understood by the lobbying parties that there will probably be cell lines and antibodies that may not adapt well to in vitro production methods, and assuming that sufficient data are available to demonstrate this fact, exemption status will be granted. Nobody involved in this debate has any desire to unnecessarily prevent a highly beneficial biologic drug or diagnostic from reaching the market when it is proven that the in vivo route is the only realistic option available. This will equate to the situation in Europe where it is similarly acknowledged that in vivo production will sometimes be necessary. In fact, a number of monoclonal antibodies have been granted exemption status by the European authorities and are currently produced in ascites. It is worth pointing out that "cost" is not considered as a criterion for granting exemption status in Europe. There is great emphasis on the requirement of the organization seeking permission to use ascites to provide comprehensive data demonstrating that multiple in vitro methods have proven to be technically unsuccessful.

Considering the product safety issues associated with ascites production, the available alternative technologies and the preference of the world-wide regulatory authorities for in vitro manufacturing processes, the ascites production method should only be available as an option in extreme circumstances.

This legislation will endorse the FDA recommendation that investigators and manufacturers of MAbs should adapt the production process to in vitro conditions as early as possible in the product life cycle. This will not avoid the requirement to demonstrate product equivalence during scale-up, but it will avoid having to demonstrate equivalence late in the product development if a change from ascites production to an in vitro method becomes necessary due to regulatory or market pressure.

References

  1. Köhler, G and Milstein, C. 1975. Continuous Cultures of Fused Cells Secreting Antibody of Predefined Specificity. Nature 256: 495-497.
  2. Jackson, LR, et al.1999. Small-Scale Monoclonal Antibody Production In vitro: Methods and Resources. Lab Animal 28(3): 38-50.
  3. Jackson, LR, et al. 1999. Monoclonal Antibody Production in Murine Ascites. I. Clinical and Pathologic Features. Laboratory Animal Science 49(1): 70-80.
  4. Jackson, LR, et al. 1999. Monoclonal Antibody Production in Murine Ascites. II. Production Characteristics. Laboratory Animal Science 49(1): 81-86.
  5. Campbell, AM. 1984. Monoclonal Antibody Technology. Laboratory Techniques in Biochemistry and Molecular Biology 13: Elsevier.
  6. Grant Meisenholder. 1999. "Postmodern Culture, Maximizing cell culture output at every level". The Scientist, July 5: 21-23.
  7. Hoppe H. 1996. Monoclonal Antibody Development From Ascites to Serum Free Production. Waterside Monoclonal Antibody Production Meeting.
  8. Marx U. et al. 1997. Monoclonal Antibody Production. The Report and Recommendations of ECVAM Workshop 23. ATLA 25: 121-137.
  9. Anon. 1986. Council Directive 86/609/EEC, on the Approximation of Laws, Regulations and Administrative Provisions of the Member States regarding the Protection of Animals used for Experimental and other Scientific Purposes. Official Journal of the European Communities L358, 1-29, 24 November.
  10. Anon. 1986. European Convention for the Protection of Vertebrate Animals used for Experimental and other Scientific Purposes, 51 pp. Strasbourg: Council for Europe.
  11. Anon. 1986. Animals (Scientific Procedures) Act 1986, 24 pp. London: HMSO.
  12. Anon. 1992. Report of the Animal Procedures Committee for 1991, Appendix II, Cmnd 2048, 37 pp. London: HMSO.
  13. National Research Council.1999. Monoclonal Antibody Production. National Academic Press. Washington, DC.