Biosimilars of Monoclonal Antibodies: A Practical Guide to Manufacturing, Preclinical, and Clinical Development

Preface

With the expiration of patent protection for a range of antibody therapeutics, there is a rising expectation within the global biopharmaceutical community concerning the construction of knock‐offs, so‐called biosimilar antibodies, that has huge ramifications to the multibillion‐dollar industry and the health of millions of patients.

Biologic agents are increasing economic demand on healthcare systems worldwide, and it is widely recognized that they represent an opportunity to increase access and reduce costs for patients and healthcare systems. A systematic approach to the characterization of the structural and functional similarities of a biosimilar molecule to the originator molecule provides specifications to guide the development of biosimilars. Identical sequence and similar host cell lines are key to development of biosimilars. A well‐designed development program establishes biosimilarity based on structural, functional, preclinical, and clinical evaluation. Biosimilar endpoints are defined for each tier of characterization, and evaluation of successive tiers addresses additional relevance of biosimilarity.

Biosimilars are considered to be one of the fastest‐growing sectors in the pharmaceutical industry, accelerated by the expiry of patents on multiple brands of biologics. Biosimilars in development include a large number of candidates from established and many new entrants to the biopharmaceutical industry. Nearly 20 biosimilars are already marketed in the EU and some other major market countries having implemented biosimilar approval pathways. A database tracking the biosimilars/biobetters pipeline reports nearly 700 biosimilars and 500 biobetters in development worldwide, with the great majority targeted to enter the US and other major markets where patents and other granted exclusivities are expiring. Many companies view biosimilars (and to a lesser extent, biobetters) as a more affordable, lower‐risk way to enter the lucrative US and EU markets and gain credibility as a biopharmaceutical developer. With only comparative testing and trials required for biosimilar approval, this mechanism enables market entry at a cheaper and faster pace, when compared with the pursuit of innovative products receiving traditional full approvals. Risks are also much lower, generally involving much the same active agent that has been on the market for more than two decades.

Monoclonal antibody biosimilars are expected to dominate the biosimilar field, even though it is a newcomer into the market. The value of the 10 top selling products, the majority of them antibodies and antibody‐like molecules, had reached more than $50 billion in 2009, and the value of new antibodies and biosimilars is estimated to rise 7–15% per year. In the last three decades, protein yields from recombinant cell lines in bioreactors have increased 10‐ to 100‐fold, as the result of improvements in media, bioprocess design, and cell culture process control. At the same time, the largest vessels in use for CHO cells have a working volume of about 20,000 l. This improvement of yield efficiency and production scale has lowered the production cost of making biosimilar antibodies. The strong disposable trend has emerged, which uses single‐use bioreactors (SUB) and presterilized plastic bags, and dramatically reduces the capital investment in manufacturing biosimilar products. The huge market potential, reduced production cost, and lowering of capital investment in manufacturing facility have made the field of antibody biosimilars an attractive business.

However, there are multiple challenges facing the burgeoning monoclonal antibody biosimilars industry. As detailed by the authors in this book, the complexities and methods of manufacture create an important difference between biosimilars and conventional generic small‐molecule drugs: while chemical generics can be fully characterized as identical to the originator product, biosimilars cannot. Full guaranty of similarity is only granted after equivalence parallel trials assessing PK, efficacy, safety, and immunogenicity comparing the biosimilar candidate and the originator. As the first European Medicines Agency (EMA) approval monoclonal biosimilar (infliximab) shows, biosimilars are licensed through a comparability exercise with the reference product and clinical studies to ensure equivalence of efficacy and safety profiles. Guidelines produced by the EMA detail manufacturing process requirements and the range of protein structure, isoform, aggregate, receptor binding, and biological activity assays necessary to demonstrate biological equivalence. It also outlines the required clinical and nonclinical pharmacokinetic (PK), pharmacodynamic (PD), and pharmacotoxicological evaluations necessary to assess safety and efficacy before approval. In the United States, the FDA released draft guidance for the regulatory review of biosimilars in early 2012, which is a significant step toward the first approved monoclonal antibody biosimilar in the US market.

This volume is comprised of a series of chapters dealing with this dynamic and diverse topic. It summarizes work in a number of aspects of the field to bring together the present‐day state of the art. Here we review the major points laid out by the authors of this work.

Dr. Regis Sodoyer provides us with an introductory chapter describing the history of therapeutic monoclonal antibody technology. As he explains, although the power of the immune system to protect the body from disease has long been recognized, murine hybridoma fusion technology, developed in the early 1970s of the last century, has dominated drug development since its inception. Easy access to the production and engineering of murine monoclonal antibodies drove a revolution of drug discovery. The first two decades between 1975 and 1995 were marked by many challenges and the development of the field of molecular engineering. At this critical time, antibodies, still omnipresent in both diagnostic and research domains, have come to dominate the field of immunotherapy.

New technologies, such as phage display, humanized transgenic mice, and repertoire mining, have been perfected and standardized, allowing for the isolation of fully human antibodies. The natural complexity of the antibody molecules and the rapid implementation of engineering methodologies helped make them preferred candidates for the solution of complex immunotherapeutic challenges. Sodoyer updates the different antibody‐derived molecules as well as a survey of the latest antibody engineering technologies. In addition, the chapter reviews the critical issue of the development of expression systems suitable for large‐scale and cost‐effective production of recombinant antibodies.

In Chapter 2, Dr. Zhinan Xia describes the structure, classification, and naming of therapeutic antibodies. He explains the evolution of antibody discovery, moving after the first wave of innovation based on mouse mAbs to the recombinant DNA technology providing the tools for chimeric, humanized, and fully human mAbs as well as recombinant antibody fragments and bispecific therapeutic mAbs. Over the decades, therapeutic mAbs have evolved as innovative pharmaceutical compounds, not only for the treatment of cancer but also of autoimmune and infectious diseases. A new generation of antibodies including anti‐PCSK9 mAbs, now in clinical development by Amgen, Pfizer, and Regeneron (in partnership with Sanofi), respectively, may represent a class of blockbuster drugs aimed at cardiovascular disease. We are now entering a third wave of scientific advancement based on antibody architecture modification and engineering, taking advantage of new insights into Fc effector function, glycoengineering, and antibody‐drug conjugation.

Among the new wave of antibody technology platforms, antibody‐drug conjugates are the most prominent, with their main focus on oncology indications. Because they can offer targeted delivery of chemotherapeutic or radioactive agents, they will increase the performance of mAbs and offer a toolkit for product differentiation and life cycle management.

In Chapter 3, Dr. Yu Zhou and Dr. James Marks cover the mechanisms of action of therapeutic antibodies. Elaborating on the development of multiple antibody‐based drugs directed at a single target, in the case of the autoimmune disease targeting TNFa, the differences in Ab–Ag binding properties, the ability to induce effector functions, and other factors resulted in three full‐length IgG mAbs, one TNFR–Fc fusion protein, and one pegylated Fab. Meanwhile, a better understanding of antibody mechanism of action has led to the expansion of antibody drugs with increased potency and greater precision.

Therapeutic monoclonal antibodies and their targets are the topic of the fourth chapter by Dr. Jose Figueroa and his colleagues. These investigators state that neoplastic diseases are currently treated by a variety of mAbs that use diversified mechanisms against cancer‐specific targets, including natural cytotoxic mechanisms as well as target‐neutralizing functions, in order to produce clinical activity against various cancers. Targets presently used in mAb treatments include CD20 for patients with chronic lymphocytic leukemia and non‐Hodgkin’s lymphoma; CD30 for HL and ALCL; CTLA‐4 and PD‐1 for melanoma; HER‐2 for HER‐expressing breast cancer and gastroesophageal cancer; EGFR‐1 for CC and squamous cell carcinoma; VEGF for CC, glioblastoma, non‐small cell lung cancer, ovarian cancer, and metastatic renal cell carcinoma; and VEGFR‐2 for GE cancer, gastric cancer, and NSCLC. While no specific cancer‐related antigens have yet been identified, novel tumor targets, for instance, cancer/testis antigens (CTAs) such as SP‐17, are highly restricted to cancer cells with normal expression limited to testis and placenta. The authors predict a bright future for antibody development and application. They assert that the use of new protein expression platforms and glycoengineering techniques, the design of multispecific binding domains, heterologous fusion constructs, molecular safety switches, and effective linkers will result in an explosion of more effective products with fewer and less serious side effects.

Dr. Roy Jefferis authored the fifth chapter on the topic of antibody posttranslational modifications. He explains in detail that the clinical efficacy of recombinant antibody therapeutics has resulted in their worldwide adoption as drugs of choice, for example, in oncology and inflammation. Antibodies are often referred to as adapter molecules as they form a bridge between a specific target and the activation of downstream molecular and cellular effector functions. Immune complexes of each of the four human IgG subclasses may bind and activate multiple host ligands with varying downstream outcomes. Recognition and activation of host ligands are dependent on interactions with the IgG‐Fc region and can vary depending on the precise structure of the naturally occurring isoforms, for example, subclasses and glycoforms. Antibody therapeutics must therefore be fully characterized, structurally and functionally, as must a candidate biosimilar. Jefferis enumerates and discusses the multiple posttranslational and chemical heterogeneities that may arise within recombinant antibodies and possible consequences for mechanisms of action in vivo.

In Chapter 6, Drs. Ningning Xu, Meimei Liu, and Margaret Liu present the pharmacology, pharmacokinetic, and pharmacodynamic aspects of recombinant antibodies. The authors point out that pharmacological studies of anticancer mAbs focus on the interaction between a patient population and the biopharmaceutical agent. Specifically, these investigations describe the interaction between specific mAbs and deranged cancer cells as opposed to those dynamics involving normal cells. In sum, the branch of pharmacological antibody studies deals with the uses, effects, and modes of action of mAbs on the body.

Pharmacokinetics (PK) on the other hand is concerned with the biological processes that a drug undergoes in the intact organism, that is, what the body does to a drug.

The third branch of pharmacological studies is referred to as pharmacodynamics. It focuses on the therapeutic action of a drug on an organism, that is, pharmacologic response, as well as the duration and the magnitude of the response associated with the concentration of drug at an effectual site of the organism. The authors lay down a set of general principles describing in vivo antibody behavior to serve as a guide for initial drug development leading to future clinical applications.

In Chapter 7, Jeanne Robson outlines the clinical performance of currently approved anticancer antibodies, considering their developmental history, target biology, mechanisms of action, clinical efficacy, and FDA‐approved indications.

Dr. Reena Nair discusses the development of biosimilar rituximab and her institution’s clinical experience with their biosimilar in Chapter 8. Rituximab was the first monoclonal antibody approved for cancer treatment. Since its launch in 1997, it has been at the forefront of treatment for B‐cell non‐Hodgkin’s lymphoma. Reditux™ (Dr. Reddy’s Laboratories Ltd.) is the first in a series of biosimilar products that dramatically expand the affordability of antibody therapy for patients in India and beyond. Regulatory requirements for biosimilars are evolving. Education of physicians and healthcare providers, patients, and payers about biosimilars will assist in informed decision making and promote acceptance of biosimilars into clinical practice. Scientific bodies should formulate practice guidelines and position statements to establish biosimilarity in efficacy, safety, comparability, and interchangeability with the reference originator biologic molecule.

In Chapter 9, Dr. S.J. Projans discusses monoclonal antibodies for infectious and cardiovascular diseases. The advent of the age of antibiotics in the 1940s had a major impact on extending lives, but it also brought with it a false sense of security. The overenthusiastic use of antibiotics, sometimes as a substitute for good sanitation and hygiene, has resulted in unintended consequences. We now know that drug resistance developed by these life‐threatening bacteria has become a big problem for hospitals treating infectious disease. Finding new ways to prevent and treat infections caused by resistant organisms will be a major challenge in the twenty‐first century. Identifying at‐risk patients (e.g., those colonized with Staphylococcus aureus) and taking preemptive measures (e.g., decolonization) or perhaps developing a protective vaccine may well obviate the use of the antibiotics that have made the practice of medicine of the late twentieth and early twenty‐first centuries possible. A new term of art has been coined precision medicine, and we now have the ability to identify an infecting pathogen within 2 h. This will allow for the use of narrow‐spectrum agents that, in and of themselves, cannot obviate either toxicity or the development of resistance but may well prevent cross‐resistance to current antibiotics as well as ameliorate, if not prevent, horizontal gene transfer. In addition to the continued hunt for effective vaccines, currently there are numerous monoclonal antibodies in clinical development to prevent and/or treat infections caused by Pseudomonas aeruginosa, Staphylococcus aureus, Clostridium difficile, and others. To date, these monoclonal antibodies, mainly of human origin or design, have had an enviable safety record, and these may represent the future of antibacterial treatments.

In Chapter 10, Drs. Yie and Wu discuss monoclonal antibodies directed against markers for musculoskeletal, central nervous system, and other diseases. There are seven approved monoclonal antibodies available for the treatment of various nonmalignant conditions. They are natalizumab for autoimmune disease such as multiple sclerosis and Crohn’s disease, eculizumab for rare diseases associated with the dysfunction of the complement system, ranibizumab for macular diseases, denosumab for bone diseases, and daclizumab, basiliximab, and muromonab‐CD3 as immunosuppressive agents for solid organ transplantations. The approval of these seven monoclonal antibodies spanned more than two decades, reflecting the evolution of the investigative process, from murine to human/murine chimeric antibody to humanized antibody and finally a fully human antibody. In this chapter, drug development history, target biology, mechanism of action, clinical efficacy, and approved indications of each monoclonal antibody are covered thoroughly. Appropriate tables, factsheets, and sequences of the relevant biologicals round out the discussion.

In Chapter 11, the manufacture of recombinant therapeutic proteins using Chinese hamster ovary (CHO) cells in large‐scale bioreactors is presented in detail by Drs. Wűrm and De Jesus. Immortalized mammalian cells in suspension culture have been used in large‐scale bioreactors for the production of recombinant protein therapeutics since the mid‐1980s. All recombinant antibody products with greater than $1 billion US sales/year on the market today are produced in CHO cells. Clearly they have achieved superstar status since more than 50% of all protein pharmaceuticals on the market are produced now in these cells. Eight of the ten top selling biologicals in the market today are derived from CHO cells. When considering products made in animal cells only, CHO cells are probably used in more than 90% of cases. Other immortalized cell lines, such as NS0, BHK, or HEK‐293 cells, are used for a few individual products, but they play an insignificant role in the overall sales picture. All processes here considered involve stable cell lines, that is, cells that were engineered by inserting the desired gene(s) of interest into the genome (chromosomes) of the host cells.

The contribution of expression vector constructs and host cell engineering is, contrary to what numerous publications and even recent reviews are claiming, relatively minor. The reason for this surprising and counterintuitive statement is clear. While the academic world has been very active in solving problems in CHO technology by addressing them through cell engineering, the industry was bound by the conservative approaches necessary in pharmaceutical manufacturing and by the benefit of sticking to approaches that worked. Significant investments (cell banking, testing, established operational handlings) into a relatively solid approach with a proven track record from a nonmodified cell host could not easily be abandoned just because an improving cell cycle or antiapoptotic gene might eventually provide some (minor) benefit. In fact, the backlog of knowledge concerning a given cell host provided ample opportunity to reap the benefits from modifications of media and process, with resultant dramatic improvements.

In general, the same vector components are used today as 20 years ago. Codon optimization of gene sequences coding for the desired protein of interest have improved overall expression of some proteins, but did not result in breakthrough yield increases. The identification and selection of the suitable clonal populations from transfected cells are facilitated by the widely accessible and cost‐efficient equipment that allows high‐throughput screening, including the use of flow cytometry and sorting. Some DNA elements in plasmids and novel vector/gene‐transfer approaches have pushed primary expression higher. Currently marketed protein therapeutics are exclusively produced in large‐volume steam‐sterilizable, stainless steel bioreactors, and the majority of production processes therein can be characterized as fed‐batch cultures, with processing times of 10–20 days. However, over the last 10 years, a strong disposable trend has emerged, particularly not only with innovative products that are entering the clinic but also with biosimilar products under development. Single‐use bioreactors (SUB) and presterilized plastic bags, mounted into or onto containers or platforms, serve as a physical barrier between the nonsterile environment and cell culture process liquids. While stirring with marine and pitched blade impellers is still the main method for mixing of both sterilizable stainless steel and single‐use bioreactors, other impeller‐free approaches have emerged as well that are now applied in single‐use bioreactors.

In Chapter 12, Stimple and Wood consider process development at the downstream end. They summarize several advances in antibody purification at industrial scale, with an emphasis on alternatives to conventional protein A affinity batch processing. At the molecular level, these alternatives include the use of newly developed affinity ligands, as well as reemergence of conventional chromatographic processes. The chapter also covers several novel process configurations, including the use of disposable technologies and the rapidly increasing focus on continuous processing. The general trend of these advances is toward highly flexible manufacturing environment, which will provide rapid process development as well as production on demand of antibody therapeutics in multiproduct facilities. The authors cover each of these developments, with the relative strengths, weaknesses, and potential impacts evaluated.

The impact of biosimilars and biobetters on biopharmaceutical manufacturing and contract manufacturing organizations is covered in Chapter 13 by Ronald A. Rader. He discusses the fact that biosimilars are subjected to an abbreviated but rigorous approval process involving direct comparisons of the biosimilar to its reference product, including analytical profiles and head‐to‐head comparative pharmacokinetics and clinical trials. Biosimilars will raise the bar for product quality expectations for all biopharmaceuticals, with regulators, healthcare professionals, patients, and the public focusing on the details of manufacturing processes and the analytical, bioprocessing, and related quality differences between products.

Dr. Rader believes that by the end of the decade, we can anticipate 10–12 entries for every successful reference product. While this may seem a chaotically competitive market, there already are many or more sources for some major selling generic drugs. The United States is projected to be the dominant biosimilars market. Many new and established companies will be moving aggressively into this market, even if this results in overcrowding and a market so fractured that it becomes impossible to realize desired profit margins. The biosimilars, biobetters, and biogenerics market will be much like the generic drug market. This includes active pharmaceutical ingredient sources and competing products originating from a range of international companies. As international competition and the available number of biosimilars and other follow‐ons increase, buyers will favor the US and EU manufactured products. As a result, unlike the generic sector, development and manufacturing will be located in the industrialized countries.

Dr. Jianguo Yang’s Chapter 14 covers cell line production and cell culture development for biosimilar antibody‐drug manufacturing. Almost all major pharma companies are heavily committed to biosimilar product development, resulting in many active players within the industry. Likely, pharma companies with well‐integrated drug R&D divisions and robust clinical trial and sales networks will come to dominate the field. Due to a wealth of current biosimilar antibody‐drug candidates in various development stages, biopharmaceutical companies are increasingly interested in a platform process for mammalian cell line and cell culture development to meet productivity and quality attributes with maximum efficiency.

Cell line and cell culture processing are the most critical steps for biosimilar antibody development since both can determine productivity, quality, and cost/benefits, which are of vital importance to the bottom line. For the last several years, average productivity of mammalian cell lines has reached 3–6 g/l. These gains were achieved using fed‐batch process, high‐productivity cell lines, and optimized media and bioreactor platforms. These innovations are dramatically cutting costs and speeding drug development. Meanwhile, product quality is the key factor for success as many companies compete against one another in a cutthroat landscape.

Drs. Weidong Jiang, Scott Liu, and Ziyang Zhong review in Chapter 15 the analytical and bioassay methods used to characterize the finished antibody products based on predetermined specifications for biosimilar antibodies. For their identification, it is typical to have the primary sequence confirmed to the reference standard using the LC/M method. There are many options available on the market for this purpose, including quadrupole time of flight (QTOF) and Orbitrap. Both Waters and Agilent make excellent TOF instruments, while Thermo/Finnigan is the undisrupted leader in the Orbitrap technology sphere. All of these instruments offer techniques to provide both primary sequences, as well as information concerning disulfite bonds. Their purity and integrity, on the other hand, are demonstrated by conventional chromatography, such as SDS‐PAGE and IEF. Similarly, HILIC is the standard method to characterize the glycan profile, which plays a key role in ADCC.

In Chapter 16, Rafiq Islam reviews the development of bioanalytical tests of biosimilar antibodies, including glycosylation pattern, ADCC/CDC activity, and FcγR binding profile. Critical features of such characterizations are the test results required to demonstrate similarity to the original product, as stated in the regulatory guidelines issued by the FDA and EMEA.

Drs. João Fonseca and João Gonçalves cover preclinical and clinical development of biosimilar antibodies in Chapter 17. They touch on a recurrent theme in this book: that biosimilars offer a highly attractive strategy for reducing medical costs and increasing accessibility to targeted biologic therapies. However, unlike small molecules, the development and production of candidate original biologics and biosimilars can be hampered by unpredictable variability that should be tackled as far as possible during the preclinical development process. Biological therapies are inherently variable, creating unavoidable differences between even subsequent batches of the same product.

Clarinda Islam covers regulatory issues in Chapter 18 providing an overview of the regulatory process with a focus on biosimilar antibody‐based therapeutics. Of major concern are the challenges faced by regulatory authorities globally and at the national level as they seek to evaluate biosimilar products for safety and potency. Some of the issues include the question of how similar is similar enough and what are the implications of the observed differences for antibody performance and patient safety. This chapter will also address limitations of the regulatory process and the patients’ responsibility to report safety concerns and symptoms.

Legal considerations concerning biosimilar antibodies are subjected to detailed analysis in Chapter 19 by K. Lance Anderson, Jennifer R. Moore, and Jonathan Ball. Follow‐on biologics, also known as biosimilars, face unique statutory and regulatory frameworks, the understanding of which is essential for successful commercialization. Modeled after existing schemes for generic drugs, such as the Hatch–Waxman Act, biosimilars are afforded the opportunity for special treatment due to their similarity to an existing innovator product. However, due to the variable nature of a biosimilar product, whether from its origin, impurities, or method of manufacture, this process has become codified in its own legislation designed to address the additional inherent product variation. The Biologics Price Competition and Innovation Act of 2009 (BPCIA) was signed into law by President Barack Obama on March 23, 2010, as a subchapter of the Patient Protection and Affordable Care Act. The objective of the legislation was to amend the Public Health Service Act (PHS Act) to provide the FDA a framework for approving follow‐on biologics. Other regulatory authorities such as the European Medicines Agency (EMA) have developed similar guidelines, although such statutory frameworks do vary.

In essence, the BPCIA provides an abbreviated approval, or licensure, pathway for biological products shown to be biosimilar to, or interchangeable with, an FDA‐approved regulatory product. The FDA‐licensed biological product, or reference product, may be relied upon so long as similarities can be shown between the biosimilar product and the reference product. In doing so, an abbreviated regulatory pathway is available for the biosimilar product.

In focusing on the BPCIA as it amends the PHS Act, and related FDA guidances issued periodically with regard to the applicable legislation, the basic framework of the licensure process will be appreciated in light of the reference products to which they pertain. Additionally, certain product exclusivity periods may be obtained, or must be permitted, within the biologics licensure process. As discussed further, these aspects will continue to remain of importance in coming years due to known expirations of biological products approved under Section 351(a) of the PHS Act. The exclusivities afforded may vary due to the nature of the biologic products and the sponsors involved. Further, the consideration of patent analysis and procedures critical to the BPCIA process will be discussed, along with recent guidance documents issued by the FDA, submissions of biosimilar applications, and early court cases regarding such submissions.

In Chapter 20, Dr. Cheng Liu, editor of this volume, considers antibody‐dependent cell cytotoxicity enhancement technologies for next‐generation therapeutic antibodies. ADCC enhancement is a key strategy for improving therapeutic antibody‐drug efficacy. Recent clinical studies provided further evidence in support of the technology. It has the potential of lowering effective drug dosage, hereby benefiting patients through lower drug cost. Antibodies with ADCC enhancement are expected to eliminate variations in patient response to antibody treatments caused by genetic polymorphism and improve survival of cancer patients. The commercial value of the technology has been demonstrated by the licensing agreement between Amgen and Kyowa Hakko Kirin on an anti‐CCR4 antibody, which includes $100 million up‐front fee plus $420 million milestone payment and double‐digit royalties. The FDA approval of obinutuzumab and its superior efficacy compared with the first‐generation anti‐CD20 antibody, rituximab, in leukemia was a milestone success of ADCC enhancement technology in new mAb drug discovery and development. Meanwhile, ADCC enhancement is also becoming a core technology for developing next‐generation therapeutic antibody drugs with favorable clinical outcomes.

Dr. K. John Morrow, Jr., coeditor of this book, is the author of Chapter 21 which considers technologies for engineering improved antibody half‐life. Biosimilar antibodies include a subcategory of biobetters, that is, antibodies that actually represent an improvement over the original product. The loss of patent protection and technological improvements are predicted to generate a wave of biosimilar products in the coming years that will force deep inroads into the market share of the original products. The competition brought about through the introduction of biosimilars may cut costs of antibody‐based therapeutics by 50% or more.

Because the IgG framework is able to sustain great variability, it lends itself to a range of engineering possibilities. Advances in cloning and engineering technologies accomplished in recent years have opened the door to improving antibody performance, in particular increasing the half‐life of therapeutic mAbs. Fundamental investigations of the mechanism of half‐life extension have shown that the complexing of the Fc region of the antibody with the FcRn protein is critical to the prolongation of serum half‐life. This receptor protects the IgG molecule from cellular catabolism by way of a pH‐dependent recycling and transcytosis pathway.

Antibody half‐life can be extended by increasing the affinity of the antibody for the FcRn molecule. Extensive molecular mapping studies have revealed the mechanism of binding of the two molecules and the best mutational substitutions to the Fc region to optimize affinity. Alternatively, in some cases it may be desirable to interfere with the binding in order to shorten the half‐life of the antibody. This is the strategy that is being applied in therapeutic application to various autoimmune diseases.

The FcRn molecule complexes with immunoglobulin and also with a totally unrelated molecule, albumin. This property allows for the development of stable albumin fusion molecules with notable therapeutic applications. As new biosimilars are developed, some preclinical and clinical evaluation will be required in order to establish their safety and efficacy. Since preclinical studies involve the use of mice, the recent controversy concerning the accuracy of the mouse as a model for human disease conditions should be mentioned. While there is reason for concern, the weight of the evidence indicates that the mouse is an accurate model of the human immune system, with certain caveats.

One of the most promising approaches to protecting the half‐life of engineered antibodies is the use of nanotechnological devices. In its most fundamental state, the design concept is a drug‐containing core in a soft sheath covered with polyethylene glycol, which attracts water molecules surrounding the nanoparticle with a liquid halo. Currently there are a number of nanotech applications that have received FDA approval, although none involving an antibody therapeutic.

Despite the problems associated with the development of biosimilars and biobetters, the future of the industry appears bright. There is an extremely robust market for biologics, now 27% of pharmaceutical sales in Europe. In the United States alone, there are 907 biologics in development, targeting more than 100 diseases. Extension of the half‐life of these molecules could constitute a direct and rapid approach to improving their performance, so it is certain that this will be a major part of an R&D biobetters program.

In the final Chapter 22, Patrick G. Holder and David Rabuka discuss technologies for antibody‐drug conjugation. As described in the preceding chapters, therapeutic mAbs leverage their specificity to induce cellular cytotoxicity through a variety of mechanisms. Small‐molecule chemical pharmacophores can also induce cytotoxicity. An antibody‐drug conjugate (ADC) is prepared by chemically joining therapeutic mAbs with these potent chemical cytotoxins. In vivo, a circulating ADC binds its target antigen on the cell surface and is internalized, carrying with it the chemical pharmacaphore. This payload can then be delivered through a variety of release mechanisms and potentiate cell death. The combination of a mAb that targets a carefully chosen antigen (Chapter 4) with a highly potent toxin can produce ADCs that significantly increase the therapeutic index of targeted treatments.

A technology for antibody‐drug conjugation describes any of the modular components that are chemically assembled in order to produce a functional ADC. These include the location on the antibody to which drug is attached, the chemical reaction used for conjugation, the components of the linker joining mAb and cytotoxin, and the chemical functional groups that can trigger release of the cytotoxin. The exact choice of cytotoxin will depend upon the target cell and its sensitivity to a specific mode of action.

Researchers have been designing and building ADCs for over 50 years. As the construction of ADCs has changed, so have the underlying technologies. Pertinent reviews of ADC technologies have been prepared along the way; this chapter assembles current knowledge and critiques best practices.

Cheng Liu, Ph.D., and K. John Morrow, Jr. Ph.D.

1

The History of Therapeutic Monoclonal Antibodies

Regis Sodoyer

Expression Systems and Protein Engineering, Technology Research Institute Bioaster, Lyon, France

1.1 Summary

Antibodies, a main component of the immune response, have been recognized, more than a century ago, for their proven therapeutic value. The hybridoma fusion technology, proposed in the early 1970s, for the first time gave easy access to the production and engineering of murine monoclonal antibodies. The potential of these new molecules, as laboratory tools, was largely exploited during the two following decades. At present antibodies, still omnipresent in both diagnostic and research domains, have progressively come to dominate the field of immunotherapy. New technologies, such as phage display, humanized transgenic mice, and repertoire mining, have been proposed, allowing for the isolation of fully human antibodies. The natural complexity of the antibody molecules and the rapid implementation of engineering methodologies helped in making them ideal candidates for new applications and for the solution of complex immunotherapeutic challenges. The first chapter is a current update on the different antibody‐derived molecules as well as a survey of the latest antibody engineering technologies. In addition the chapter reviews the critical issue of the development of expression systems suitable for large‐scale and cost‐effective production of recombinant antibodies.

1.2 Introduction

The historical roots of immunotherapy trace their origins to the end of the nineteenth century. In collaboration with Shibasaburō Kitasato, the first bacteriologist to succeed in cultivation of Clostridium tetani, Emil Adolf von Behring demonstrated the efficacy of so‐called antitoxins to protect animals against tetanus and diphtheria [1, 2]. Although not immediately recognized, the discovery was adapted to the treatment of diphtheria‐stricken children and was found to significantly reduce mortality [3]. For some years serum therapy was widely adopted before being gradually replaced by active immunization. Several decades passed before the composition, physiological behavior, and chemical nature of the antitoxin components were thoroughly characterized.

These components, referred to as gamma globulins according to their electrophoretic mobility and as immunoglobulins and antibodies according to their immunological function, were at the center of multiple investigations involving a steadily increasing number of renowned scientific teams. From the early 1950s until now, the definition of antibody (Ab) became more precise, as well as the genetic aspects behind the creation of their molecular diversity. In parallel, advances in cellular biology were under way around the world, allowing a better understanding of the precise role of B cells in the immune response. Paving the way toward the advent of modern immunotherapy, hybridoma technology described by Köhler and Milstein in 1975 gave, for the first time, easy access to murine monoclonal antibodies (mAbs) [4]. mAbs were recognized as revolutionary laboratory tools from their inception, although the in vivo applications and therapeutic potential of these molecules were still controversial, raising some skepticism in the early 1980s. Nevertheless, significant progress in molecular biology techniques allowed mAbs to move from research to diagnostics and applications in therapy.

More recently, monoclonal and recombinant antibodies have become the focus of new technologies, such as bacteriophage or other in vitro display techniques, mice or larger animal transgenesis, and other technologies permitting direct access to fully human antibodies. Importantly, the pressing need for large quantities of mAbs was a major driver for the development and optimization of recombinant protein production systems. The contribution and complementarity of these different approaches will be considered in the context of large‐scale industrial production of therapeutic mAbs.

1.3 New Markets for Old Antibodies, Old Markets for New Antibodies

To date, more than 40 mAbs have been approved by the US Food and Drug Administration (USFDA) or European Medicines Agency (EMA) for therapeutic applications: 5 are of murine origin, including 2 bispecific constructs, 8 chimeric, 18 humanized, and 13 fully human. Thanks to the novel technologies available [5], fully human antibodies are rapidly taking over the market, and it seems probable that even humanized molecules will be marginalized in the coming years. The composition of the early‐stage antibody pipeline is indicative of this trend. In terms of revenue, substantial returns have been realized both on existing and new markets. A good illustration is provided by recent data from Lawrence and Lahteenmaki: [6] among the top 10 selling biological drugs of 2014, 5 are mAbs. Humira® (adalimumab, AbbVie Inc., North Chicago, IL, the United States), with applications to several conditions (such as rheumatoid arthritis, juvenile rheumatoid arthritis, Crohn’s disease, psoriatic arthritis, psoriasis, ankylosing spondylitis, and ulcerative colitis), is number one, with a total revenue exceeding $10.5 billion. Remicade® (infliximab, Janssen Biotech, Inc., Titusville, NJ, the United States) for multiple indications (rheumatoid arthritis, psoriatic arthritis, ulcerative colitis, Crohn’s disease, ankylosing spondylitis, and severe or disabling plaque psoriasis) is in second place with a total revenue over $9.2 billion. Rituxan® (rituximab, Biogen Idec Inc., Cambridge, MA, the United States, and Genentech USA, Inc., San Francisco, CA, the United States) for multiple indications (rheumatoid arthritis, chronic lymphocytic leukemia/small‐cell lymphocytic lymphoma, non‐Hodgkin’s lymphoma, antineutrophil cytoplasmic antibody‐associated vasculitis, indolent non‐Hodgkin’s lymphoma, and diffuse large B‐cell lymphoma) is found in the sixth position with a total revenue over $7.5 billion. Avastin® (bevacizumab, Genentech USA, Inc., San Francisco, CA, the United States) and Herceptin® (trastuzumab, Genentech USA, Inc., San Francisco, CA, the United States) are, respectively, ranking seventh and ninth positions with individual revenues fluctuating between $6.8 and $7.0 billion.

Beside the success stories of some blockbusters, mAbs are set to play a role in the rapid control of emergent diseases. A striking example is the 2014 Ebola outbreak, for which more than 20 laboratories and research groups around the world, including those from Canada, Japan, Israel, Uganda, and the United States, are working simultaneously to develop therapeutic mAbs against the virus. The dire state of emergency will no doubt facilitate and shorten the approval process. A mixture of three mAbs known as ZMapp (LeafBio, Inc., San Diego, CA, the United States), never tested in humans, was exceptionally accepted despite the fact that very little is known regarding the safety and effectiveness of this treatment.

1.3.1 Intellectual Property

One possible explanation for the great commercial success of therapeutic antibodies is the proper management of intellectual property and accurate designation of strategies for protecting antibodies or antibody‐derived products [7]. This is particularly meaningful if we consider the emblematic and well‐known Cabilly patent filed by Genentech, one of the most ubiquitous patents in biotechnology, which covers a fundamental method for the production of therapeutic recombinant mAbs that cannot be ignored by anyone planning to commercialize an antibody [8]. Besides the uniqueness of this example, one should not underestimate the inextricably complex patent situation governing the recombinant antibody world. As a matter of illustration of this complexity, according to Sandercock and Storz [9], it might be possible to patent the precise definition of an epitope with the possibility to claim later‐generation antibodies targeting the same epitope.

1.3.2 Biosimilars

A consequence of the astronomical commercial value of mAbs associated with the patent expiry date (or estimated expiry date) is their attractiveness as candidates for biosimilar drugs [10]. The European Union (EU) has been the first to establish a regulatory framework for marketing authorization application (MAA) and has named copycat biotherapeutic products with the term biosimilars, a term also recently adopted by the USFDA [11].

Biosimilars offer a highly attractive mechanism for reducing the cost of medical care and thus should be considered positively [12]. Nevertheless, the abbreviated approval pathway associated with their adoption requires reinforced pharmacovigilance after a biosimilar mAb is approved, in order to ensure its long‐term safety and efficacy. Unfortunately, it has been observed that the studies carried out to obtain approval of the reference product are not always adequate to ensure comparability of biosimilar mAbs. A similar efficacy does not necessarily imply a similar safety profile between the innovator and biosimilar products [13]. A major challenge to be addressed is the prediction and assessment of immunogenicity of subsequent entry biologicals (SEBs) and, in particular, biosimilar products [14]. It is nevertheless important to keep in mind that the global cost of manufacturing and licensing a biosimilar product remains high, and the reduction in cost may be more limited than for a nonbiological small‐molecule drug and its generic version. Vital et al. provide a good illustration for biosimilar versions of the antibody rituximab [15]. In addition to the industrial challenges of development, testing, and marketing, biosimilars and biobetters [16] are raising new analytic challenges concerning product characterization and immunogenicity profiling [17a, 17b].

1.3.3 Modified Antibodies, Nontherapeutic Applications

Recent technological innovations will facilitate access to alternative antibody formats including bispecific antibodies (BsAb), conjugated antibodies, and antibody fragments. The demand for these molecules is expected to rise in significance [18]. mAbs are set to play a significant role in the treatment of a wide number of indications in various therapeutic domains, although oncological therapeutics will continue to dominate the majority of applications [19, 20].

The goal of cancer therapy is to cause the direct or indirect destruction of cancer cells, by specifically targeting the tumor (Rituximab®) or the vasculature that nourishes the tumor (Avastatin®). Numerous anti‐inflammatory antibodies already on the market (Remicade, Humira) have also shown the great potential of mAbs. Asthma, transplant rejection, inflammatory conditions, autoimmune disease, cardiovascular conditions [21], atherosclerosis [22], and infectious disease [23] are new domains of treatment. While treatment of infectious disease will increase in the future, full‐blown success will require the development of original screening strategies to obtain broadly neutralizing antibodies [24] and/or the combination of several antibodies targeting different epitopes or stages of the pathogen life cycle [25].

1.4 Antibody Engineering: A New Approach to the Treatment of Disease

In 1975, Köhler and Milstein developed and described but did not patent a revolutionary technique for the generation of mAbs [4], based on the immortalization of mouse B cells through fusion with myeloma tumor cells. Mouse hybridomas, generated upon fusion of the two parental cell partners, became the first reliable source of mAbs, facilitating the laboratory‐scale production of murine mAb (Mumab) specific for a given antigen. The first Mumab, as therapeutic agent, was unsatisfactory due to their very short half‐life in serum, poor or absent activation of human effector functions, and undesirable stimulation of human anti‐mouse antibody (HAMA) responses in patients when repeated administration protocols were applied [26].

These issues were addressed using genetic engineering or chemical coupling techniques but have not been entirely resolved. The different strategies chronologically developed to avoid, mask, or redirect this human immune response are chimerization by fusion of mouse variable regions to human constant regions [27], humanization, [28] and deimmunization by removal of in silico predicted T‐cell epitopes [29] or introducing regulatory epitopes to induce tolerance [30].

1.4.1 Chimerization: The Stone Age of Antibody Engineering

In chimeric antibodies, the murine constant regions are replaced with human equivalent regions, since the constant region significantly contributes to the molecule’s immunogenicity. In addition, the presence of human constant domains guarantees a more robust interaction with human effector cells and the complement system. This strategy led to therapeutic successes such as basiliximab (Simulect®: IgG1 anti‐CD25, developed by Novartis Pharmaceuticals, East Hanover, NJ, the United States) or cetuximab (Erbitux®: IgG1 anti‐EGFR, developed by ImClone, LLC, New York, NY, the United States). Nevertheless, chimeric antibodies, even if perceived as less foreign and therefore less immunogenic than mouse mAbs, have been shown to induce human anti‐chimeric antibody (HACA) responses. This has been seen with infliximab/Remicade [31]. The advent and increasing sophistication of humanization technologies and the direct access to fully human antibodies will likely represent the death knell for chimeric antibodies.

1.4.2 Humanization: Improvement Never Ends

In order to move forward to reduce both HAMA and HACA, antibody humanization technologies were developed and made possible by, at the least, transferring all xenogeneic complementary determining regions (CDRs) onto the framework of a human antibody. Different approaches to CDR grafting have been tested according to the human template used as a matrix: sequences with known crystalline structure [32], rearranged somatic sequences, unmodified germ line sequences [33], or consensus sequences [34].

The first humanized antibodies were constructed based on human sequences with known crystalline structures, which permits the identification of residues contributing to antigen binding. In the best fit strategy, the closest human sequence, usually rearranged, is used as a framework to receive the murine CDRs. Another approach for humanizing an antibody is to choose the closest human germ line sequence [33]. Indeed, human antibody genes are formed in vivo by rearrangements of germ line gene segments. Later in B‐cell ontogeny, the hypermutation process takes place, tailoring the initial sequences to improve recognition of the specific target antigen (Ag). The body thus may be theoretically more tolerant to germ line‐encoded Abs. It is likely that there will be a reduction in clinical issues if germ line sequences are used for constructing humanized Abs.

The consensus method utilizes variable light (VL) and variable heavy (VH) domain frameworks derived from the most common amino acid found at each position within a given human subgroup. Whatever the method, CDR grafting might not result in the complete retention of antigen‐binding properties since some framework residues can directly or indirectly interact with the antigen [35] or may affect the conformation of CDR loops [32c]. In this case, the antibody must be re‐engineered or back‐engineered to fine‐tune the structure of the antigen‐binding loops and restore its original high affinity. On the other hand, humanization of a xenogeneic Ab does not necessarily abolish the immunogenicity of the molecule, since the humanized Ab can still induce response against its xenogeneic CDRs. Not all residues within the CDRs of an Ab are essential for binding to its Ag. In fact, the Ag‐binding site of an Ab usually involves only 20–33% of the residues [36]. These residues have been designated as specificity‐determining residues (SDRs) [37]. Therefore, a murine Ab can be humanized through restrictively grafting only its SDRs onto the human template, minimizing the immunogenicity [38].

Another humanization strategy, termed resurfacing, was proposed by Padlan and involves the replacement of solvent exposed murine framework residues in the variable regions with human residues [39]. More recently, Hanf et al. [40] have proposed humanization by redesign of CDR residues in close proximity to the acceptor framework, yielding an antibody with better binding capacity than simple CDR graft with reduced immunogenicity compared to framework redesign.

Almagro and Fransson [41] have divided the field of humanization into two main trends: rational methods based on design cycle and precise knowledge of the antibody structure and sequence information and empirical methods using large combinatorial libraries and selecting the desired variants, a so‐called rational strategy.

In contrast, a prime example of an empirical approach is guided selection, a process that transfers the specificity of a Mumab to novel human mAb (Humab) by creating a hybrid library of the murine heavy chain and random human light chains, subjecting them to a selection process of binding antibodies and repeating the protocol with the human light chains previously isolated and a library of human heavy chains. Adalimumab (Humira) is the first phage display‐derived human antibody and was generated by guided selection starting from a Mumab [42].

Humaneering™ is a KaloBios Pharmaceuticals’ proprietary method for converting nonhuman antibodies into engineered human antibodies. The resulting antibodies are as close to the human germ line sequences as the products of fully human antibody generation techniques. The authors assert this platform maintains epitope specificity and increases affinity.

Finally, humanization technology has taken advantage of the recent advances in bioinformatics and in silico modeling. Zhang et al. [43] proposed a novel antibody humanization method based on this strategy, including an epitope‐scanning algorithm designed to identify antigenic residues in the framework regions (FRs) that are mutated to their human counterpart in the humanization process.

The process of humanization is not restricted to murine antibodies and can similarly be applied to antibodies from other species. Nishibori et al. [44] have described a simple method for humanizing chicken mAb by CDR grafting followed by framework fine‐tuning using a chicken phage‐displayed mAb, phAb4‐31, as a model antibody.

Despite the reduction in the proportion of exogenous sequences, the level of immunogenicity of humanized Abs ranges from negligible to highly detrimental. The humanized anti‐HER2/neu Ab Herceptin gave as low as 0.1% of human anti‐human antibody (HAHA) in breast cancer patients [45], while the humanized A33 Ab elicited 49% of HAHA among colon cancer patients treated with this Ab [46]. In this case, immunogenicity might result from HAHA responses, in particular to the paratope of the antibody (anti‐idiotype antibody response) [47].

To date, it is difficult to determine whether fully human mAbs are less immunogenic than humanized mAbs as full immunogenicity data are available only for a limited number of mAbs in each category. The answer is complex and most probably will be resolved through a case‐by‐case consideration and detailed analysis.

Finally, anti‐allotype reactions are predicted to occur during therapy of a genetically diverse population with a single antibody reagent. Moreover, allotypic variation is not equally distributed to all antibody isotypes and might be taken into consideration before designing a therapeutic strategy and effector antibody [48].

1.4.3 Deimmunization

Biovation (the United Kingdom, www.biovation.co.uk) has developed the DeImmunization™ technology consisting of identification and removal of T helper (Th) cell epitopes from antibodies [29]. An example of a product of DeImmunization actually under clinical trials, J591 is a modified antibody binding to prostate‐specific membrane antigen (PSMA) [49].

Antitope, a subsidiary of Abzema, is conducting the same kind of approach to limit the development of antidrug antibodies (ADAs), which can reduce efficacy through rapid clearance. The associated EpiScreen™ immunogenicity assessment technology is used to confirm that T‐cell epitopes have been removed [50].

The company Epivax is sharing the same field and proposes in silico epitope discovery, immunogenicity assessment, and protein deimmunization. They include Tregitope Analysis based on the identification within each submitted sequence of putative regulatory T‐cell epitopes. These are subregions contained within the submitted sequences, which may relate to natural regulatory T cells and which may help to dampen the immune potential of the submitted antibody sequence [51, 52].

1.5 Fully Human Antibodies, What Else?

mAbs of human origin may have greater therapeutic value; thus several methods have been developed to generate Humab: selection from human hybridomas, selection from humanized transgenic mice, construction of in vitro combinatorial libraries, or direct cloning from immunized individuals.

1.5.1 Human B‐Cell Hybridoma and Immortalized B‐Cell Lines

The recovery of stable human B‐cell hybridoma producing high‐affinity IgG mAbs has rarely been achieved, due to the lack of a suitable human myeloma cell line. The most satisfactory results were obtained using heteromyelomas (mouse–human hybrid myelomas) as fusion partners. However, the mouse–human heteromyelomas that have been used for fusion with human lymphocytes are often unstable. The situation was ameliorated by the recent development of the novel SPYMEG cell line, a human lymphocyte fusion partner, codeveloped by Naomasa Yamamoto, of Ohu University and MBL. SPYMEG was established by the cell fusion of MEG‐01 with a murine myeloma cell line. The SPYMEG cell line overcomes the problem usually encountered by hybridoma cells of human origin that are prone to chromosome deletions.

As an alternative, human antibody‐secreting cells (ASCs) can be immortalized by Epstein–Barr virus (EBV) infection. However, EBV‐infected cells are somewhat difficult to clone and usually produce only very low yields of immunoglobulin [53]. An improved method of B‐cell immortalization by EBV involving the addition of a polyclonal B‐cell activator (CpG/TLR‐9 agonists) has been described [54, 55]. In contrast to plasma cells, which are terminally differentiated, memory B cells retain substantial growth potential and can be immortalized by EBV [56].

The final disadvantage restricting the human B‐cell hybridoma approach (and, more generally, all the technologies giving access to human antibodies) is the fact that the human circulating antibody repertoire does not generally retain specificity to self proteins, which represent a significant number of targets for human antibody therapeutics.

1.5.2 Ex Vivo Stimulation

A recent technology using a process of ex vivo stimulation has been proposed by Duvall et al. [57]. Naive human B cells, isolated from tonsil tissue, are immortalized. The ex vivo stimulation of these cells induces class switching and somatic hypermutation. The resultant human library of different IgG antibodies can then be screened against any antigen using a standard limiting dilution protocol. By eliminating immunization and humanization steps, this platform should reduce both cost and time in producing a therapeutic mAb of interest.

1.5.3 Mice and Other Animal Species Producing Human Antibodies

1.5.3.1 SCID and Other Immune‐Deficient Mice

A potential source of Humab can be obtained upon transplanting a functional human immune system into immunocompromised mouse strains, such as severe combined immunodeficient (SCID), SCID‐bg, Trimera, or γc−/−/RAG2−/− mice.

SCID mice, lacking mature T and B cells and virtually devoid of endogenous serum immunoglobulins, can be successfully reconstituted with human peripheral blood lymphocytes (PBLs). Such mice reconstituted with a competent human immune system would represent an invaluable tool for producing human immunoglobulin, after immunization with antigen. However, the use of SCID mouse can be limited by shortened life spans, spontaneous production of functional lymphocytes with aging, and residual innate immunity leading to variable levels of engraftment. Mouse natural killer (NK) cells in particular would be detrimental to engraftment of human lymphoid cells [58]. For example, the natural depletion of NK activity in SCID‐bg mice facilitated engraftment of human PBL from anthrax‐vaccinated (AVA) donors. Stable recombinant cell lines producing Humab were generated by hybridoma formation, and human anti‐protective antigen (PA) neutralizing mAbs of high affinity were obtained [59]. Other immunodeficient strains of mice were developed, for instance, the γc−/−/RAG2−/− mice. In addition to T‐ and B‐lymphocyte deficiency, γc−/−/RAG2−/− mice are completely deficient in NK activity [58]. In recent years, the passing fad for such technologies has been completely masked by the gain in maturity of transgenic mice expressing human antibodies.

1.5.3.2 Humanized Transgenic Mice

The availability of transgenic strains of mice expressing human Ig genes (XenoMouse®, HuMab™ Mouse, TransChromo Mouse) provides a breakthrough for isolating Humab [60]. The occurrence of rearrangements and hypermutations confirmed that the endogenous cell signaling machinery of the mouse is compatible with human immunoglobulin sequence elements [61] and would constitute a marked advantage.

XenoMouse strains from Abgenix (Fremont, CA), were engineered by functionally inactivating the murine heavy‐chain and k light‐chain Ig loci and incorporating megabase‐sized inserts of human DNA yeast artificial chromosome (YAC) carrying Ig heavy‐chain and k light‐chain loci that express the vast majority of the human Ab repertoire [62]. Three different strains of XenoMouse mice have been produced, constrained to class switch from IgM to IgG1, IgG2, or IgG4. The immune repertoire of XenoMouse strains was further increased by the introduction of the entire human Igλ locus [61]. To date, results obtained in preclinical and early clinical trials with human antibody from XenoMouse mice confirm their relative lack of immunogenicity [63]. However, human antibodies from mice can be distinguished from human antibodies produced in human cells by their state of glycosylation, particularly with respect to their Galα1‐3Gal residue [64]. This carbohydrate residue is widely distributed among nonprimate mammals. Anti‐Gal antibodies are produced in humans throughout life, as approximately 1% of circulating Ig [65]. Thus, antibodies bearing that residue would probably be subject to an accelerated immune clearance [66].

In a parallel strategy, human minichromosomes (derived from human chromosome 2 and 14) containing the complete germ line clusters for heavy and k light chains were introduced into TransChromo mice [67]. These mice, developed by the Kirin Brewery Company (Japan), are capable of producing every subtype of fully human Ig, including IgA and IgM. However, the instability of the transchromosome carrying the Igk locus was particularly detrimental, as hybridoma production was less than 1% of that seen in wild‐type mice. An instant solution to this problem was to crossbreed the Kirin TC mouse carrying the human chromosome fragment 14 (locus IgH), with the Medarex YAC transgenic mouse carrying about 50% of the Igk locus. The resulting mouse (KM) performed as well as normal mice with regard to immune responsiveness [67b].

In 2006, panitumumab, the first fully human antibody generated from transgenic mice, was approved for clinical use by the USFDA. In 2012, seven of such antibodies were approved, which clearly indicates the important contribution of humanized transgenic mice in the pipeline of novel therapeutic mAbs [68].

Over the years, models of transgenic mice have risen in sophistication. According to Murphy et al. [69] from Regeneron Pharmaceuticals, a common weakness of humanized transgenic mice is a lack of fully functional immune systems intrinsically due to the strategy used for their genetic humanization. The VelocImmune mice model efficiently produces human–mouse hybrid antibodies in which the mouse constant regions (which are rapidly convertible to fully human antibodies) are kept intact and have fully functional humoral immune systems. The rationale was that the introduced human variable region gene segments would function indistinguishably in their new genetic location, whereas the retained mouse constant regions would allow for optimal interactions and selection of the resulting antibodies within the mouse environment.

The strategy adopted by Kymab is quite similar; Kymouse™ transgenic mice have been generated upon insertion of the genetic regions corresponding to the variable genes from all three human immunoglobulin loci, including λ genes, into precise locations in the corresponding loci of mouse embryonic stem (ES) cells using a proprietary technology called sequential recombinase‐mediated cassette exchange (S‐RMCE) [70].

Besides the aforementioned transgenic mice, MeMo® mouse developed by Merus produces, upon immunization, human antibodies composed of a single common light chain and diversified immunoglobulin heavy chains. These antibodies serve as building blocks for the easy and direct generation of BsAb for therapy (Biclonics™). Simultaneous transfection of two cLC VH genes gives greater than 99% pure BsAb from a single cell and might additionally provide an innovative solution to the complex problem of coexpressing a mixture of several antibodies.

1.5.3.3 Humanized Transgenic Chicken

Transgenic chickens expressing human antibodies are designed to access human targets and epitopes that have been intractable in mammalian hosts because of tolerance to conserved proteins. The major difficulty lays in the mechanism for the generation of diversity, which in the chicken is accomplished through gene conversion, a very different system from that occurring in the human and murine immune response. To resolve this issue it was necessary to confirm that a knockout chicken B‐cell line lacking VH and VL loci was able to undergo gene conversion with inserted human VH and VK regions. Gene targeting in avian primordial germ cells may provide a unique approach to antibody engineering, particularly in the field of antibody discovery [71, 72].

1.5.3.4 Humanized Transgenic Bovine: Polyclonal Revival?

Antigen‐specific human polyclonal antibodies (hpAbs), produced by hyperimmunization, have proven to be useful for treating many human diseases or infections. However, yields from available transgenic mice carrying human immunoglobulin loci are too low for direct therapeutic applications. Kuroiwa et al. [73] have produced transgenic bovine antibodies after transferring a human artificial chromosome bearing the entire unrearranged human immunoglobulin heavy (hIGH) and kappa light (hIGK) chain loci to bovine fibroblasts in which two endogenous bovine IgH chain loci were inactivated. Up to 2 g/l of hIgG, either associated with human kappa light chain or with bovine kappa or lambda light chain (chimeric), can be obtained from these animals. Polyclonal sera from cattle immunized with anthrax PA proved to be highly active in an in vitro toxin neutralization assay and protective in an in vivo mouse challenge assay. The transgenic bovine platform was recently fine‐tuned to further increase the proportion of fully hpAbs produced after triple genetic knockout eliminating the bovine lambda cluster [74]. A first clinical trial for pathogen‐specific fully human antibodies derived from Tc bovine (currently in preparation) will assess the value of the approach. The possibility of a yearly production, of approximately 450 l of plasma per transgenic animal, makes this strategy especially appealing.

1.5.4 Antibody Display

A major step forward in the production of fully human antibodies after selection from antibody gene repertoires expressed either in vivo on the surface of cells or filamentous bacteriophages [75, 76] or in vitro took place during the early nineties. Display technologies made possible the selection of antibodies from large repertoires, and the physical linkage between genotype and phenotype enabled the recovery of the DNA encoding the selected antibody fragment.

1.5.4.1 Phage Display

The bacteriophage platform is still the most widely used and well‐established technique for antibody display and library screening [77]. Large repertoires of single‐chain Fv fragment (scFv) or antigen‐binding fragment (Fab) antibody genes are cloned into phage or phagemid vectors as a fusion product to one of the phage coat protein genes. Expression of the fusion product and its subsequent incorporation into