Expert View: Drugs trial hospital could have started by saying sorry

Most of us are not interested in how and why lab mice reacted to a test

Eight volunteers participated in a drug trial at Northwick Park hospital in north-west London on 13 March. Six suffered an adverse reaction to the bio-engineered protein drug TGN-1412, produced by TeGenero, a tiny German biotech firm. Thankfully, all are now making progress in their recovery.

A month on, the Medicines and Healthcare products Regulatory Agency (MHRA) has concluded that the drug trial company, Parexel, was not at fault, yet it appears baffled as to the cause of the catastrophe.

Neither in-vitro nor animal tests can replace human tests. In this case, the volunteers were administered 0.2 per cent of the dose that was seen as toxic to laboratory animals - a very high safety margin - yet this still left two of the men critically ill. One of the possible explanations for these reactions is that the drug was designed to tackle a human protein - nothing else. Hence, in cases such as this, where animal testing continues to provide valuable toxicity data, and indeed is required by law, the pace of innovation may actually demand new types of testing procedure.

At the same time, regulators, drug companies and hospitals have to become more proficient in communicating with the outside world. From a non-medical perspective, this case could have been handled much better. Early on, no one took charge. Rather, those involved quickly tried to distance themselves by stating either that they were simply following orders, or that the dose was correct, or that the drug had not produced any adverse reactions in animals. The impression was that they all wanted to plead not guilty.

Even if there was no attributable human error, everyone responsible needed to be clearer, more honest and humbler in stating immediately that they were most sorry for what had happened. The communication process was not helped by the contradictory messages put forward by the hospital early on in the crisis, informing distraught relatives that their sons would be better in a couple of days, and then, the next moment, putting out a press statement warning that the patients could die within hours.

The TGN1412 trial has led to calls for more openness on the part of the drug companies, including invoking the Freedom of Information Act. From a communications perspective, this makes sense. Transparency increases trust while secrecy erodes it. Since the drugs industry as a whole is now experiencing a crisis in public confidence, changes plainly need to be made.

The issue, however, is not clear- cut. Yes, there should be transparency, and of course the drug companies in question should co-operate fully with the regulators when information is demanded of them. Does this mean putting all the data on the internet and having all of us look at it? This may not be necessary if the companies in question are willing to open their books to a "blue ribbon panel" of neutral third-party experts, who can then interpret the results on behalf of the public - something that the MHRA is now proposing.

If the drugs companies agree to this, and if the public take the view that these experts can be trusted, then such a panel will suffice. Most of us do not have an interest in how exactly and why laboratory mice reacted to a certain test. And too much information can increase public anxiety with unintended consequences, such as patients giving up taking a drug.

What the TGN1412 case has shown is that the medical profession has to become better at communicating both the risks and the benefits of its work to the rest of us - and that there is a need for more involvement by independent experts in the regulatory process. If this does not happen, regulatory bodies will continue to lose public trust. And if this occurs, these same regulators, as we have seen in the past, will become more risk averse in order to regain credibility. Thereby they may deny patients life-saving treatments.

Professor Ragnar Lofstedt is director of the King's Centre for Risk Management, King's College London