Photo: GSK Marburg
Past, Present and Future of Monoclonal Antibodies
(last update: June 2013)
Antibodies from Natural Sources
Antibodies were used for decades as antisera or as passive vaccines. Prominent examples are tetanus immunoglobulin (from human), snake venom antiserum (from horse), Digitalis antitoxin (from sheep). In contrast to the complex mixtures of antisera, Monoclonal Antibodies (mAbs) are obtained as pure protein molecules with help of genetic engineering and animal cell culture techniques.
Discovery of Monoclonal Antibodies
Since 1975, when George Köhler and César Milstein discovered the new cell type, named hybridoma, resulting from the fusion of B-lymphocytes with a myeloma cell, mAbs became the most important biopharmaceuticals produced by cell culture techniques. In 1986, the CD3-specific monoclonal antibody muromonab (Orthoclone OKT3) was approved by the US Food and Drug Administration for use in the treatment of acute transplant rejection, making it the first therapeutic antibody to be used as a drug in humans. Since then, therapeutic antibody development provided new therapies for the treatment of cancer, autoimmunity and inflammatory diseases.
Research Tools and High Efficient Biopharmaceuticals
The monoclonal antibody technology was quickly adopted by scientists in both industry and academia due to the unlimited usefulness of mAbs for research and therapy. In 2008, the therapeutic monoclonal antibody generated revenues of over US$15 billion, the highest earning category of all biological drugs. Cancer and autoimmune diseases are the main indications for mAbs, collectively constituting ~78 % of the total antibody sales in the US. It led to a hype in industry and academia about the almost unlimited usefulness of monoclonal antibodies.
Their usefulness in research and diagnostics has by now been proven: it is difficult to find a research laboratory where monoclonal antibodies are not generated for research purposes, in particular for selection of compounds. Industrially, mAbs are used to recognize proteins from a culture broth and thereby assist in purification processes. In diagnostics, numerous test kits are on the market featuring monoclonal antibodies that specifically recognize certain molecules. Among the over the counter products featuring monoclonal antibodies, easy-to-use pregnancy or fertility tests are best known among the public. This widespread use of monoclonal antibodies was based on such inherent characteristics as
- a broad range of potential targets,
- high selectivity,
- high affinity and,
- ease of generation.
Therapeutic Disappointments in the 1980s
Monoclonal antibodies extended to the field of biotherapeutics in the early 1980s. The concept of the ‘magic bullet’ was born: a combination of a specific drug or toxin, bound to a monoclonal antibody that specifically zeroes in on its target and delivers the drug or toxin there where it is needed.
Beside the approval of OKT3 (as the first mAb for therapeutic use) the development of therapeutic monoclonal antibodies suffered a number of serious disappointments, which reduced faith in the therapeutic applicability of monoclonal antibodies considerably.
These disappointments were caused by:
- the need for high therapeutic doses,
- poor penetration in solid tumors,
- potential cross reactivity with other tumors,
- high production costs,
- potential viral safety problems,
- technical difficulties in large scale production and,
- relatively poor patent protection.
In addition, all mAbs were derived from rodent cells, and these rodent-derived monoclonal antibodies suffered from:
- a very short half life,
- poor recognition of the rodent IgG constant region by human effector function,
- HAMA response (generation of human anti-mouse antibodies).
Continuous Improvements in Safety and Efficiency
All these drawbacks led to sophisticated technological developments and improvements, primary consisting of strategies to generate human(ised) monoclonal antibodies. Indeed, to reduce the immune response of mAbs from mouse and to improve the mAb half life and binding properties, chimeric, humanized and human mAbs were developed and introduced as therapeutic proteins. The goal was to exchange the mouse effector domains of the mAb by human ones. There are 5 possible strategies for that purpose:
- generation of true human mAbs (but these are instable and have only a limited number of targets),
- chimerisation (the resulting monoclonal antibodies are 60-70 % human),
- humanization (the resulting monoclonal antibodies are 90-95 % human),
- use of phage display (resulting in fully human mAbs),
- use of transgenics (resulting in fully human mAbs).
The extent of human parts of the mAb is reflected in the nomenclature: mAb names with ending „-ximab“ are chimeric, „-zumab“ means humanized, „-umab“ is for human mAbs. All therapeutic mAbs have the ending „-mab“. Pure mouse mAbs end with the suffix „-omab“. Such technological progress led to the development and approval of an increasing number of initially chimeric, then humanized therapeutic monoclonal antibodies.
Chimeric antibodies have a constant part from human peptide sequences, but the variable part of the antigen binding fragments (Fab) contains mouse protein. Reopro® (Abciximab) was approved as the first chimeric mAb in year 1995. It binds to the GPIIb/IIIa receptors of thrombocytes and it is given as an adjunct to percutaneous coronary intervention for the prevention of cardiac ischemia.
Anti-antibodies can be developed against chimeric mAbs. This was the main driver to reduce the mouse part further. In 1997, Zenapax® (Daclizumab) was approved as first humanized mAb. All parts of this mAb were human, except for the hypervariable loops (CDR), i.e. the small portions of the mAb which are essential for the antigen recognition. Zenapax® was approved to diminish kidney transplant rejection. The humanized mAb binds to the IL-2 receptor complex of activated T-cells and suppresses the immune response.
A milestone for the generation of mAbs from human origin was the phage display technology. This novel method used bacteriophages to connect proteins with the genetic information that encodes them. Large libraries of phages were generated where a protein (phenotype) and its encoding DNA (genotype) occurred jointly in the same phage. This allowed a specific screening of phages with good binding properties to a desired binding partner (the immobilized antigen), out of a large variety of different transgenic phages. Due to the genotype-phenotype complex, the DNA of the binding antibody epitope is also available. The DNA sequence can be extracted, multiplied by PCR and sequenced. Antibody libraries displaying millions of different antibodies on phage can be applied to isolate specific therapeutic antibody lead candidates.
Besides the large technological improvements, some new challenges were faced with highly effective mAbs which showed adverse effects in patients. Prominent examples were Tysabri® (withdrawn after FDA approval temporarily due to suspect of side effects and re-launched later), TGN1412 (discontinued after severe inflammatory reaction occurred in six healthy volunteers at start of clinical evaluation in year 2006), and Raptiva® (withdrawn after FDA and EU approval due to suspect of side-effects).
Regulatory authorities such as the United States Food and Drug Administration (FDA) and the European Medicines Agency (EMA) reacted and strictly regulated the prescription and administration of monoclonal antibody drugs as soon as particular risks were suspected. For some drugs, use was restricted to patients that meet well defined demands. But in many cases the medical benefit for the individual is to rate much higher than the related risk. A very large number of patients suffering from a wide variety of diseases already benefit from the possibility of the treatment with mAbs, where formerly no medication was available.
Monoclonal Antibodies in Development
Regardless recurring throwbacks, technological developments and better understanding of physiological pathways and valid targets have led to a whole series of new projects in industry, resulting in an increasing number of mAbs in clinical trials. Currently about 200 different antibody-based candidate therapeutics are in clinical trials targeting about 70 different molecules. Presumably one or several antibodies against each relevant target are developed concurrently at different companies.
Approx. 30 mAbs are approved by the FDA for therapeutic use. Most of them are also approved in the EU. For a listing, see the table of approved antibodies on the ACTIP website.
Monoclonal Antibodies Approved for Diagnostic Application
Monoclonal antibodies are also used as highly specific diagnostic tools. The first approved in vivo radio-labelled mAb diagnostic product was OncoScint® in 1992 (a mAb targeting for the tumour associated glycoprotein 72 antigen, a cell surface protein generally over-expressed in colorectal cancers, labelled with the radioisotope Indium-111) for detection and evaluation of cancers.
Other examples were MyoScint® (a mAb targeting for myosin, labelled with the radioisotope Indium-111) for cardiac imaging, OncoTrac® (a mAb for the detection of lung cancer, labelled with the radioisotope Technetium-99), and CEA-Scan® (a murine monoclonal Fab element, joint to a radioactive Technetinum-99 label approved as imaging agent in metastatic colorectal cancer).
Polyclonal Antibody Preparations for Therapeutic Application
Also polyclonal antibody preparations were approved for therapeutic use. However, they are mainly derived from blood of animals or humans. Examples are CroFab® for the treatment of rattlesnake evenomation, Gammagard® for passive immunisation, and Respigam® for the treatment of respiratory syncytial virus (RSV). Respigam® was later replaced by Synagis®, a mAb targeting the RSV.
Engineering of Monoclonal Antibodies
Besides the development of humanised and fully human mAbs, various optimisation strategies were pursued to improve their effector functions. Direct effector functions, mediated by the Fab part, can be binding, blockage or linking on surface molecules, stimulating or inhibiting signalling effects for cells, or induction of apoptosis.
Indirect effector functions, mediated by the fragment crystallisable region (Fc) domain, are antibody-dependent cellular cytotoxicity (ADCC), complement-dependent cytotoxicity (CDC), and attraction of killer cells. ADCC activates immune cells like macrophages, granulocytes and monocytes. CDC acts by the complement cascade, which is mediated by aprox. 20 serum proteins. In addition, pharmacokinetics can be largely influenced by the Fc part.
At first, the choice of an appropriate mAb isotype (IgG1, IgG2, IgG3, IgG4, IgGA1, IgGA2, IgGD, IgGE, IgGM) gave different options to modulate binding, interactions and pharmacokinetics. IgG1 is the most abundantly used antibody isotype. It activates complement and natural killer (NK) cells effectively, it is the most effective subclass at high antigen concentrations, and it has a long serum half-life (approx. 21 days). If interactions with the human immune system are not desired, IgG4 or IgG2 isotypes are preferred. For therapeutic use, both subclasses were engineered to reduce instability of the mAb and to shut down the binding properties to cellular Fc receptors.
The next step focused on the design and engineering of the glycosylation profile, leading to glyco-engineered antibodies. Carbohydrates are placed at conserved positions in the Fc region of the mAb and their structure often vary significantly. High-mannose, multiply branched or biantennary complex oligosaccharides can be formed during co-translational and posttranslational processing. A basic option to obtain a favourable glycosylation profile is the selection of an appropriate expression cell line, e.g. from hamster or mouse or human origin. Subsequently, the glycosylation profile can be engineered to enhance the ADCC activity, for example by low-fucosylated or non-fucosylated glycovariants.
Another way to enhance or to reduce binding to Fc receptors is based on the modification of the protein backbone of mAbs. If Fc receptor binding is not favoured, Fab2 fragments could be used also. Due to their short plasma half life, entire mAbs with reduced binding properties to Fc receptors were engineered. Such antibodies can be useful as immunosuppressive agents, for example to prevent transplant rejections or to treat T-cell mediated autoimmune diseases. Protein-engineered mAbs with a low Fc receptor binding capability are also attractive for immunotoxins. In contrast, a high Fc receptor binding capability is advantageous to increase the potency up to 10 or even 100 fold, e.g. by enhancing the ADCC with effector cells. The engineered Fc part should have a high affinity to the Fc receptor in order to displace inhibitory isoforms and to boost the activation. For clinical use, a good balance of activation and inhibition of receptors is important.
Antibody Drug Conjugates
Antibody Drug Conjugates (ADC) can bring highly potent drugs closely and precisely to the desired target. Conventional cancer therapeutics like cytotoxic chemicals or high energetic radiation are effective inside and outside of tumour tissue and can therefore show very adverse side effects for patients. One goal of modern tumour therapy is to bring cell-destructing agents more exactly to the sites of action. For this purpose, a toxin or a chemotherapeutic chemical or a radioactive isotope is linked to a mAb. This should allow a more specific tumour targeting and the administration of much higher dosing of cytotoxic agents. Furthermore, precursors of tumour-destructing substances can be designed which are converted to the active conformation only at the binding site of the mAb. New therapies based on ADCs bear tremendous opportunities for a better treatment of many diseases.
A prominent ADC was Mylotarg® for treatment of acute myeloid leukaemia (AML). It was approved by the FDA in the year 2000, but withdrawn from the US market in 2010 based on the negative results of a post-approval trial. Other approved ADC include, Adcetris® (Brentuximab) for the treatment of systemic anaplastic large cell lymphoma (ALCL) or Hodgkin lymphoma (HL) approved by FDA in 2011, and Kadcyla® (Trastuzumab emtansine) for treatment of HER2-positive metastatic breast cancer approved by the FDA in 2013.
Bispecific antibodies can bind two different antigens simultaneously. Hereby, they can be applied for tumour targeting (similar to ADCs) or for the direction of immune effector molecules or effector cells (“retargeting”) to a mAb binding cell, e.g. a tumour cell. Other promising indications for bispecific mAbs are infectious or allergic diseases, inflammatory and autoimmune diseases. The first bispecific antibodies were generated by cross-linking or cell fusion. Subsequently, various formats of bispecific mAb fragments or whole mAb molecules were generated by means of genetic engineering techniques.
The different types of bispecific mABs developed included:
- trifunctional antibodies (two different Fab binding sites and one Fc docking site),
- chemically linked Fabs (two different Fab binding sites are linked, no Fc part),
- bivalent or trivalent single-chain variable fragments (scFvs),
- fusion proteins mimicking the variable domains of two different antibodies.
Bispecific mAbs can connect target cells to cells of the immune system, resulting in the destruction of the target cells. The effective dose is several orders of magnitude lower compared to conventional mAbs, because bispecific mAbs can also bind to antigens that are expressed in very low amounts. Bispecific mAbs can also be combined with additional components such as toxins, enzymes or cytokines. This allows dual targeting plus delivery of the fusion partner. A fusion with plasma proteins (e.g. serum albumin) may be used to extend the plasma half-life of bispecific mAbs.
More than 40 different formats of bispecific antibodies were designed and generated. Actually, the Dual-Affinity Re-Targeting (DART) and the Bispecific T cell Engager (BiTE) are two promising formats for the therapeutic application of bispecific mAbs. The first bispecific antibody approved was Removab® (Catumaxomab), by the EMA in 2009, for the treatment of malignant ascites.
Reasons for Optimism
Currently, the context for developing monoclonal antibodies is much better than it was foreseen 10 years ago. Knowledge increase, technological improvements and growing experience with marketed mAbs as well as promising sales volumes have accelerated the development of therapeutic mAbs. Not only small Biotech companies, but also the global players in “Big Pharma” push the development of these promising therapeutic tools.
Important groundbreaking elements for further successful developments are:
- the continued development of human(ised) mAbs,
- proven financial success,
- more experience in the selection of good targets,
- more experience in the selection of good antibodies,
- techniques to create new antibody formats,
- improvements in manufacturing (technologies, cost of goods),
- well established registration procedures.
Updated by Prof. Dr. Roland Wagner and Dr. Ralph Kempken (June 2013).