In March, phase I clinical trials were subject to unprecedented media scrutiny when six healthy volunteers taking part in a first-in-man administration of a monoclonal antibody (mAB) were admitted to an intensive care unit, all critically ill with multi-organ failure. They had been part of a group of eight study subjects taking part in a study testing TeGenero’s TGN1412, a fully humanised monoclonal superagonist of the CD28 T cell surface receptor, at Parexel’s clinical pharmacology research unit at Northwick Park Hospital in North London. 

TGN1412 was being developed for the treatment of serious immune-mediated disease. Two subjects were given placebo and remained well. pA2 has asked some of those involved in conducting healthy volunteer clinical pharmacology studies in the UK to reflect upon events.


The TGN1412 disaster was a tragedy for the volunteers and their families, and we all extend our sympathy to them. We should also be concerned about the potentially devastating impact that the disaster might have on the staff of the research unit – a consideration that has received little attention thus far.

The effects on our Phase I work (HMR is a contract clinical pharmacology unit) have been interesting and, perhaps, unexpected. For example, more volunteers have come forward: some with altruistic motives, and some who had learned of the opportunities to make money as a human guinea pig (to borrow a phrase from the tabloid press). Our volunteers now ask more questions about safety, and more of them withdraw between screening visit and first dosing day. But nervousness at the time of dosing has not increased, nor has our drop-out rate changed from its usual very low level.

Some of our ward staff expressed concern about the safety of phase I studies, but were comforted by the fact that our current studies all involve “small molecules”. Both of the ethics committees that review our projects seem (understandably) more nervous than before, and have queried the arrangements for compensation of injured subjects.

Remarkably, neither our sponsors nor we (at HMR, at least) have changed our routine procedures for first-in-human studies of conventional pharmaceuticals. In particular, we continue to dose subjects in cohorts of 8–10, and at intervals of 5–10 minutes. That’s surely because those practices are supported by an excellent safety record: in first-time-in-human studies of small molecules, no serious injury or death has ever occurred that would have been prevented by using a longer inter-subject interval or smaller dosing groups.

The media coverage of the events was, in my view, of a high standard. We received many requests for interviews, and we granted as many as we could do without compromising our work. It is disappointing that no physician from Parexel made a statement of any kind; that presumably reflects company policy but I feel that this was a serious mistake.

Should we now change the way that we develop mAB as medicines? The TGN1412 disaster has shown, in the most harrowing manner, what we feared – that animal toxicology predicts poorly what mAB will do in humans. Moreover, the occurrence of six life-threatening adverse reactions means that the safety record of mAB in healthy volunteers is now poor – perhaps even unacceptable. It could be argued that we should not be unduly influenced by this single episode, because the toxicity of the molecule – a superagonist at CD28 receptors – could have been predicted, given that TeGenero were aware that anti-CD3 stimulation caused a “cytokine storm” in animals. But the truth is that we cannot be certain that we can predict similarly serious acute toxicity of another mAB in future. Moreover, there is justified concern over potential long-term adverse effects of mAB: for example, infliximab is associated with an increased risk of malignancy and infection, including reactivation of latent TB, and perhaps also hepatitis and granulomatous dermatitis.

Of course, it’s easy to say, with the benefit of hindsight, that TGN1412 was a high-risk mAB, and that the disaster might have been predicted and thus averted. But the immunologists in Wurzburg University, the scientists at TeGenero, the MHRA assessors, the Parexel physicians, and the ethics committee all failed to predict it. So, on whose advice can we rely when we decide whether to give an mAB to healthy volunteers or not?

The idea that mAB study proposals could be referred to a panel of acknowledged experts appears attractive at first sight, but seems to me fundamentally unsatisfactory because the ultimate responsibility for giving the product remains with the physician in charge. Moreover, whose opinion can I rely on? As the physician with the final responsibility for my experimental subjects, I’m in an uncomfortable position.

One thing is certain – development of mAB as medicines will now be more difficult, slower, and more expensive. That’s a pity, because there’s no question that mAB as medicines can confer tremendous benefits, and perhaps offer our best hope of treating some of the most serious and distressing of illnesses that afflict human beings.


Steve Warrington
Medical Director, Hammersmith Medicines Research


In the recent past, all of the major headline stories involving drug safety have involved licensed products – the rofecoxib (Vioxx®) saga was, for example, front-page news. It was therefore unparalleled when the first-ever trial of a new drug was the cause of such dramatic toxicity, evoking fevered interest in the media.

Whilst some of the coverage has been quite sensational and some more guarded, the public reaction has generally been considered and rational, recognising the absolute need for studies of new drugs in healthy volunteers to continue, whether as “human guinea pigs” or not.

On 1 May 2004 the legislative loophole allowing clinical trials in healthy volunteers to take place without regulatory approval was finally closed, and thus the TGN1412 study was con-ducted in an ever-tighter research governance climate, undergoing both ethical and MHRA review. It is hard to imagine any regulatory developments that would further improve volunteer welfare and safety. It appears quite likely that, albeit with the much relied-upon benefits of hindsight, the clinical events following TGN1412 were not idiosyncratic after all, but based on predictable pharmacology.

There was also a system reliant on all the steps in a process, from review of the Investigator Brochure to formal approval, being sufficiently sensitive to flag up a potential problem – to stop a study with a new drug becoming just another “routine” first-in-man administration.

The nature of the informed consent process has rightly been questioned – how much information about risk is conveyed to volunteers? No drug should be viewed as 100% safe – even those available over-the-counter – and nowhere is there more uncertainty than when the drug is a human mAB that has never been given to humans before.

Informed consent should theoretically present a balanced view – the potential risks and the unknown, without scaring the volunteers so much that no-one who has capacity to consent would take part. Consider how difficult it would be to consent a volunteer for a study involving the administration of a mAB at the current time. Furthermore, it would be naïve to assume that money is not a prime motivator for volunteers, even if some are also more altruistic than others. No ethics committee would allow payment to incentivise and make up for risk – phase I studies are by no means as lucrative as some of the media speculation would suggest, particularly given the significant time and inconvenience involved.

The presence of the Parexel unit on the Northwick Park campus was quite probably life-saving. Much has also been made of the decision to dose all eight study subjects on the same day and at relatively short dosing intervals. Such procedures are more frequently based on practical considerations such as clinical unit space, dispensing capabilities and sample handling, rather than through any safety concerns. In many cases these need not necessarily be specified in the study protocol anyway, unless it is considered to be a potential unique safety factor.

In this study, dosing of four subjects on one day at a time would have of course immediately halved the size of the catastrophe. Arguments for only dosing one subject at a time should be met with even more caution – if you are so concerned, should you be doing it at all, let alone in healthy volunteers, when the risk-benefit ratio is completely skewed? Bearing this in mind, the relevance of healthy volunteer studies in the subsequent use of a drug in disease is often contested. However, with TGN1412 the outcome in patients with serious illness would surely have been even worse than that observed in the healthy young men taking part here, with all their robust physiology enabling them to put up a fight in the face of a terrific insult.

The events at Northwick Park undoubtedly feature as one of the worst nightmares for all of those involved in conducting healthy volunteer studies. No investigator does so lightly and without the welfare and safety of the subjects under their care being of paramount importance. Our first thoughts are for those men who did receive TGN1412 but surprisingly little has been heard from or about the staff at Parexel. Surely, it is possible to express regret without admitting liability: this is undoubtedly the right thing to do.

Many commentators lament the decline of clinical pharmacology. However, in the current governance framework and in light of the TGN1412 catastrophe, the pivotal role of clinical pharmacologists in the discovery and development of new medicines has never been more relevant.


Simon Constable
Lecturer in Clinical Pharmacology & Therapeutics, The University of Liverpool

David Webb
Professor of Clinical Pharmacology & Therapeutics, The University of Edinburgh

If you would like to comment on this article, or anything else in pA2, then please write to the editorial team