Preparing for first-in-man studies: the challenges for translational immunology post-TGN1412 (original) (raw)

Abstract

Current immunology research is generating many new approaches to immunotherapy. However, the recent disaster surrounding the testing of TGN1412, has unsettled regulators and the pharmaceutical industry regarding new immunotherapies and highlighted the complexities of conducting clinical trials with agents that target the immune system. Here we discuss the critical role for immunologists in ensuring that the development of new immunotherapies continues.

Keywords: clinical trials, first-in-man, immunotherapy, phase 1, TGN1412

We are fortunate to live in an era of unprecedented success in medical research. Within the last 35 years – less than half a lifetime − smallpox has been eradicated, diphtheria, tetanus, whooping cough, measles, peptic ulceration, malignant hypertension and rheumatic heart disease dramatically reduced, rising cure rates for several cancers achieved, unprecedented low levels of cholesterol have become achievable with very low-risk medication, and solid organ transplantation, operations on the beating heart and laparoscopic surgery have become routine. But this is just the beginning. In the last 10 years very sophisticated science and industrial pharmacy has led to new generations of treatments that, although not inexpensive, are dramatically effective for common diseases, including anti-tumour necrosis factor (TNF) for rheumatoid arthritis and Crohn's disease, trastuzumab (Herceptin) for breast cancer, anti-vascular endothelial growth factor (VEGF) for age-related macular degeneration (the most common cause of previously untreatable blindness in older people) and highly active retroviral therapy for human immunodeficiency virus (HIV).

Even this recent wave has only just begun to scratch the surface of the knowledge that has become, and is becoming, available through the cloning of the human genome, recent advances in proteomics, cell cycle and cell signalling research, and most recently the genetics of common polygenic disorders. These knowledge houses are generating new therapeutic targets that are being coupled with high-throughput screening of new therapeutic compounds. Already, not only are many young adults alive because of the eradication of traditional killer diseases of childhood, but an increasing number of 80-year-olds are alive only because of life-saving cardiac surgery or chemotherapy in their 70s. Average life expectancy in many parts of the developed world has exceeded 80 years (meaning that many people live substantially beyond this in good health) and this statistic cannot yet fully take into account the benefits of better nutrition and preventive medicine that have occurred in the last 20 years. The concept of ‘care of the elderly beginning at 65’ prevalent in the 1980s has become wholly inappropriate, and increased life expectancy is causing a major crisis for pension schemes. We cannot predict the future, but even a conservative estimate would accept that the growth in medical research is exponential rather than linear, indicating that progress in the next 10 years will be, at the very least, greater than in the last 10 (which was already impressive). Put simply, ‘we ain’t seen nothin' yet'.

Such progress, of course, comes at a price. Public expectation of medicine has increased in parallel with, if not in advance of, the pace of progress so that people's perception is more often disappointment with the ‘slow’ pace of progress rather than admiration at what has been achieved. Secondly, there is a very real increase in cost for treatments that simply did not exist before and may have a price tag of £10 000 per year or more. To deal with this, economic growth must continue its rapid pace and health-care systems remain under great pressure to be as efficient as possible. Thirdly, there is an increasing need for clinical studies of new treatments. In particular, there is an increasing need for first-in-man studies of new compounds and therapeutics, agents the like of which, in many cases, have not been seen before.

This last is a particular problem for immunologists. Until recently, immunology has been viewed as promising a great deal but delivering very little other than vaccines. The advent of the recent wave of monoclonal antibodies targeting immunologically active molecules has begun to change this perception. However, as researchers become bolder we see that these benefits do not come without real risks at times − namely progressive multi-focal leucoencephalopathy observed with natalizumab [1], the reactivation of Epstein–Barr virus (EBV) with anti-CD3 [2] and the unexpected near-fatal cytokine storm with an anti-CD28 superagonist (TGN1412 [3]). No healthy volunteers have died so far in Phase 1 trials of immunological agents, but TGN1412 came very close [3].

Quite appropriately, there has been much official and unofficial analysis of ‘what went wrong’ in the case of TGN1412 [4–6]. In fact, there was nothing wrong with the drug manufacture and the study itself, begun at 0800 h on 13 March 2006, was conducted according to the approved protocol. ‘What went wrong’ appears to relate to five aspects, summarized in (Table 1), relating to prediction of risk based on preclinical studies and the breadth of the margin of safety required (in dose and trial design) when using a treatment with a novel mode of action.

Table 1.

Summary of learning points from the TGN1412 phase 1 study.

TGN1412 study problem Detail Learning point
Interpretation of preclinical (primate) studies Low-level cytokine release in primate studies should have promoted more caution Minor but potentially important effects in preclinical studies should raise caution in crossing the species barrier
Use of human in vitro studies Insufficient in-vitro human studies on PBL were performed In vitro studies on human material as close as possible to the target tissue can be important
Choice of starting dose Subtle difference between primate and human target ligand may explain marked difference in potency – the calculation of an initial dose based on a fraction of predicted ‘no adverse effect level’ (NOAEL) proved dangerously wrong Prediction of risk and dose range from animal studies may prove unreliable: extra caution with wider margins of safety are required with ‘potentially risky modes of action’
Dosing interval between subjects No ‘proper interval’ allowing for the observation of possible side effects was left between the dosing of one subject and the next In first-in-man studies, investigators should expect the unexpected
Preparation for adverse events Preparation for possible adverse events (cytokine storm) was inadequate − investigators did not expect it, recognize it or treat it early Where there is a known theoretical risk, investigators should plan for its potential occurrence

These issues are especially relevant in immunology. The immune system is designed to operate on ‘trigger pharmacology’ − i.e. responses, if they occur, self-amplify rapidly to mount a response to infection rather than representing a predictable dose–response relationship. This feature of the immune system is best exemplified by type 1 hypersensitivity, in which minute doses can have enormous systemic effects that vary between individuals. These problems are less apparent in cardiovascular or cancer pharmacology, on which most trial design and drug development is based. Essentially this signifies the need for an appropriate degree of respect for the power of the immune system.

Almost certainly the most important issue in the design of Phase 1 clinical trials with drugs targeting the immune system is the question of starting dose. Currently the estimation of first-dose-in-man is based on the ‘no observed adverse effect level’ (NOAEL), as determined in toxicity studies in relevant animal species. The starting dose for human intervention is then reduced by a substantial safety margin. It is suggested that an additional approach be used for high-risk drugs, including those targeting the immune system. Every effort should be taken to calculate the ‘minimal anticipated biological effect level’ (MABEL). This would take into account receptor binding and occupancy data arising from in vitro human and in vivo animal studies. In addition, the calculation should incorporate dose–response data from in vitro studies with human cells and in vivo studies in suitable animal species, where appropriate. A starting dose for first-in-man trials would be set at a level substantially below the NOAEL and also below the MABEL [7].

Calculating NOAEL and MABEL levels will be particularly complicated for drugs targeting the immune system. First, if there is a relevant animal species in which to test them, and this is certainly not always the case, then it is to be expected that such drugs will have an effect on the immune system. A major issue arises because regulatory pharmacologists consider any immune response, following introduction of a potential medicinal product, to be an adverse affect. If this and the NOAEL guidelines are applied to drugs targeting the immune system, then the first-dose-in-man would have to be set at a level far below the level at which any biological or beneficial effect might be observed. As such, many potentially life-saving treatments may die in development, as it would appear that they have no biological activity in man. Secondly, while every effort should be taken to calculate the MABEL, this may prove difficult with drugs targeting the immune system. Quite clearly, the immune system responds more efficiently in vivo than in vitro. This is not surprising, given the elaborate architecture upon which the immune system depends. This cannot be reproduced in vitro following the removal of immune cells from the body. As such, in vitro studies may grossly overestimate the MABEL in vivo. Thirdly, we cannot be confident that even the use of MABEL will guarantee safety. As pointed out above, the immune system is different to other drug targets in that it depends on amplification mechanisms. Once triggered, the immune system is designed to produce a strong and rapid response. For certain immune targets, any response could trigger a potentially dangerous cascade. One could therefore argue that in the case of TGN1412, even the minimal biological effect would have been damaging to health. Non-human primate preclinical studies were conducted with TGN1412 to determine the safe starting dose and proved misleadingly reassuring. However, minor toxicity (cytokine release) in primates was overlooked, and with hindsight the MABEL for humans should have been calculated differently (Table 1 in Hehrishi et al. [8]). This emphasizes that animal studies (including primate studies) can be valuable to warn of toxicity, but that lack of toxicity can give false reassurance, especially with regard to immunomodulatory agents developed for the human immune system [7].

So how do we proceed? Agonist drugs targeting the immune systems will inevitably elicit an immune response; this is their intended function. It is clearly nonsensical to consider this response as an adverse effect unless, of course, the response is so violent as to be dangerous. The case for any drug targeting the immune system must be considered by experts with knowledge of the probable consequence, i.e. is this likely to trigger a cascade of events and could this be harmful to health? There seems little point in trialling drugs at levels way below a biologically effective dose. It is, after all, only when the biologically effective dose is reached that complications are likely to arise. We simply have to be more selective and knowledge-based in the choice of drugs that are taken to the clinic and these decisions must be made by experts in the field.

The regulatory authorities have responded to the TeGenero experience bringing these issues to light by increasing the degree of regulation. For first-time-in-man trials of ‘high-risk’ investigational products in the United Kingdom, for example, the Medicines and Healthcare Regulatory Agency (MHRA) will seek advice from an Expert Advisory Group (EAG)/Commission on Human Medicines (CHM) before approval for the trial can be given. Sponsors will be requested to make contact with the Agency before making the clinical trial authorization (CTA) application for such trials and to make available a data package which will allow that advice to be obtained. The normal CTA application process will then follow. New therapies are defined as potential high-risk medicinal products when there are concerns that serious adverse reactions in first-time-in-man trials may occur. These changes, viewed as insufficient by some commentators, appear to have already led to companies switching their Phase 1 clinical trials to countries other than the United Kingdom [9]. If true, this is not only a loss to UK research but also, rather than reducing the risk to human subjects, will simply transfer it to subjects in other countries.

How should immunologists respond more generally to these challenges? It is very important that we do not lose the ability to conduct translational immunology due to fear of the regulatory process, of companies that shy away from drug development or even of patients themselves. Immunologists need to regain the initiative in four ways. First, we should acknowledge the public anxiety about new treatments [10], and be as honest as we can about potential risks − not dismissing a possible but unlikely event too readily. We should be more aware and have better understanding of these risks than anyone else, and even though addressing them slows the drug development process, hiding or understating them may carry a much higher price. All available preclinical and clinical data that might impact on risk should be made available to regulators. Research effort should be directed at developing better preclinical assessment techniques, such as artificial immune systems and in silico assessment tools. Secondly, at the same time as accepting and openly addressing what could theoretically happen, we should also be clear on what, in view of the current knowledge of immunology, is not likely. For example, a cytokine neutralizing monoclonal is not likely to cause a cytokine storm, whereas an antibody to a co-stimulatory molecule, albeit designed to be inhibitory, might. Regulators and the general public are not (usually) immunologists and may consider two treatments to be related when to an immunologist they are clearly not. It seems likely that greater independent immunological input into the development and approval of TGN1412 for human testing would have added an important extra element of caution over and above routine testing of blocking monoclonal antibodies. Thirdly, where risk exists, we should show our respect for the immune system's ‘trigger pharmacology’ by exploring it at the preclinical and ex-vivo experimental level as far as we can, making an estimate of the risks that still remain (e.g. if it was not possible to study in a human isolated cell system) and designing our first-in-man study with a wide margin of safety in terms of initial dose chosen, delay between subjects tested and preparations for treating an adverse reaction should it occur (Table 1, Learning points). Finally, we must involve ourselves in the regulatory review process. It is vitally important that the most knowledgeable experts in the field contribute to this process. The new process will undoubtedly help to ensure safety (at least in the United Kingdom), but the review process must be driven by knowledge and not hampered by ignorance. Immunologists have an enormous contribution to make towards the development of new drugs and ensuring that the safety of first-time-in-man clinical trials is a vitally important part of that process.

What we should not do is to shy away from these tasks, remain in the laboratory and avoid translation, as this will lead to a missed decade of opportunities and large numbers of patients with debilitating and potentially lethal immune disorders missing out on the rapid progress in medical research occurring in other fields. Only immunologists have the understanding to provide both the direction and confidence that the public and the commercial world need for future progress.

References