oa Editorial [Immunogenicity of Biotherapeutics: A Need for Consensus on Flexibility of Approach (Guest Editor: Stewart Barker)]

Biotherapeutics are a burgeoning class of drugs that encompass three main categories of biopharmaceuticals. The first group includes early biotherapeutics such as insulin as replacement therapy, and more recently protein molecules such as the interferons. These represent equivalents to endogenous proteins and may therefore to a large extent be regarded, and recognised, as normal protein and not as foreign protein. A second group which is becoming more prevalent in the biopharmaceutical sphere is that of biosimilars. These are generally recombinantly produced proteins that are made as generic products once patent lifetime for the original biotherapeutic has expired. Similarly, these do not in theory present as foreign protein, though in practice, issues do materialise in manufacture that must be carefully addressed. The third and perhaps the largest, most diverse and most rapidly expanding group of biotherapeutics comprises therapeutic antibodies. Therapeutic antibodies are generally seen by the patient's immune system as foreign protein, and thus present the most challenging issues in their development, manufacture and eventual clinical use. It is important to distinguish between drug effects that are related to mechanism of action and those that derive from certain physicochemical aspects of the drug protein structure. The former include drug-related adverse effects that result, for example, from non-exclusivity of target (lack of complete specificity). The latter include immunotoxicity and immunogenicity - immunogenicity being largely the focus of this collection of reviews. The following questions are proposed. 1) What is the magnitude of the problem clinically with respect to the immunogenicity of biotherapeutics and in particular therapeutic antibodies? 2) Can clinically significant immunogenicity be predicted pre-clinically? 3) How can the potential of a protein biological to give rise to immunogenicity be minimised? 4) Is there justification for a pharmaceutical industry/regulatory authority co-ordinated program to identify best practise for pre-clinical prediction and monitoring of actual immunogenicity for the development of biologicals such as monoclonal antibodies? The first of the articles in this hot topic issue provides a comprehensive overview of the most predominant class of biotherapeutic in development - monoclonal antibodies. These were seen as the proverbial “magic bullet” when first concepts of their clinical use emerged. In more recent times we have seen the reality that, just as with any other drug, they have multiple side effects. These may be related to their lack of absolute specificity and also their propensity to elicit secondary immune responses involving antibody-dependent cellular cytotoxicity (ADCC) and the recruitment of the complement system (complement-dependent cytotoxicity; CDC). Although, this sometimes has beneficial effects - see trastuzumab (Herceptin) [1, 2]. Niebecker and Kloft have managed thoroughly to describe these issues for the main therapeutic areas, currently the targets of monoclonal antibodies; chronic inflammatory disease and oncology/haematology. Immunogenicity of monoclonal antibodies features in a dedicated section and clearly sets out the problems that have arisen in monoclonal therapeutic antibody development from the original murine full-length antibodies to wholly human antibodies generated in specially engineered “humanised” (relative to antibody production) mice [3]. The questions I would put to the field are the following. a) How much of the drive towards fully human antibodies has been evidence based? b) To what extent is this a pharmaceutical industry trend that has placed an undeservedly high value on identifying and eliminating potential immunogenicity, in the face of the much more substantial and lifethreatening drug mechanism-of-action related adverse effects? There is a case for minimising the risk of immunogenicity. However, since the first chimeric human/murine antibodies became clinically validated (e.g. cetuximab (Erbitux), infliximab (Remicade), rituximab (Rituxan) and abciximab (ReoPro) [4-11]) it has surely been proven that the impact of immunogenicity, where murine variable regions may constitute 30% of a whole antibody, is of a much lesser significance than their wholly murine predecessors. The success of these chimeric molecules should indicate that development of entirely human antibodies, though perhaps theoretically ideal, ought not to be an absolute requirement in therapeutic antibody development. Indeed, it is clear from evidence, that even fully human antibodies suffer the risk of generating anti-antibody antibodies with the consequent potential loss of efficacy [12]. Other approaches for antibody humanization have also led to promising drug candidates [13, 14]. These have included changing murine variable regions (retaining murine complementarity-determining-regions (CDRs)) for human variable domains. However, this kind of humanization approach has potentially detrimental effects on function that sometimes can only be restored by re-murinization of certain amino acid sequences [15, 16]. From Niebecker and Kloft we can also conclude that prediction and effective monitoring of immunogenicity are reasonable aims and that management of immunogenicity of therapeutic antibodies once in humans (e.g. by altering dosing schedules and routes of administration) is of equal importance. The article by Büttel et al., though not expressing the views of the regulatory authorities, gives a flavour of the kind of approach being taking in relation to potential immunogenicity of drugs in development. This outlines very much a flexible approach, with clear emphasis on treatment of drug candidates on a case-by-case basis. This is helpful, as the risk of a more prescriptive system would be that many more promising molecules would be excluded from further development, based on no firm evidence that clinically significant immunogenicity might ever materialise. Although no-one would dispute the high prevalence of immunogenic responses to fully murine antibodies, a more prescriptive approach would most likely lead to only fully human antibodies being authorised for further development. This may in fact be the current industry trend, but once again Büttel et al. indicate that immunogenicity of fully human antibodies is not an unusual occurrence. The questions this raises in my view are: a) what pre-clinical assessment criteria could be used to predict immunogenicity once in humans? b) what methods can be used to distinguish between neutralising (effecting drug efficacy) and non-neutralising (perhaps clinically irrelevant, or at least clinically manageable) forms of immunogenicity?..................