Following an alarming episode in London last month, in which six drug-trial participants needed emergency treatment, the UK Medicines and Healthcare Products Regulatory Agency (MHRA) says it will change the way it regulates clinical trials, at least temporarily. But this may prove more easily said than done.

In the trial on 13 March, six healthy subjects suffered violent reactions within minutes of ingesting an antibody drug candidate, TGN1412, which was being developed to treat autoimmune diseases such as rheumatoid arthritis. Initial investigations suggest that the antibody itself was responsible for the side effects (see Nature 440, 855–856; 2006). On 5 April, the MHRA said it will seek advice from outside experts in determining whether drug candidates with novel modes of action should be allowed to enter clinical trials.

The incident has drawn attention to the limitations of preclinical animal trials in determining the safety of drugs in humans.

The incident at London's Northwick Park Hospital has drawn attention to the limitations of preclinical animal trials in determining the safety of drugs in humans, especially for ‘humanized’ antibody drugs that are targeted at mimicking human biological processes. It has also sparked some debate about whether the participants were sufficiently aware of the dangers they faced.

For the regulator, the immediate question is whether the existing rules strike the right balance between safeguarding trial participants and promoting the study of potentially valuable cures. Previously, the MHRA allowed initial, small-scale human safety trials to go ahead on the basis of successful animal trials and a description of how the compound works.

Now the agency says it will allow such trials to proceed only after review by a panel of outside experts. However, companies that have drug candidates up their sleeves don't want information on them to be shared, and any outside panel worth its salt is bound to contain people who work with rival companies. So such a provision could lead drug developers to turn their backs on Britain as a location for early-stage clinical trials.

The best approach is probably that practised by the US Food and Drug Administration (FDA), the only drug regulator in the world with the in-house expertise to conduct such reviews by itself in strict confidence. The FDA, which is partly supported by fees levied on drug-makers eager to enter the lucrative US market, has 9,000 staff compared with the MHRA's 800 (although the FDA does handle food as well as drug safety).

One theoretical option would be a Europe-wide body set up to regulate and approve clinical trials, but the practical problems of constructing and operating such an agency would be daunting. In the interim, the MHRA may struggle to perform additional screening while satisfying confidentiality requirements.