Kirk Leech extols the alleged benefits of animal research. It is to be hoped that scientists bring greater intellectual rigour to their research than he does to his arguments.

Mr Leech, who used to work for Understanding Animal Research (the staunchly pro-animal research pressure group), relies on the fact that novel treatments for muscular dystrophy, cystic fibrosis and heart disease (stem cells) have reached clinical trials, having been shown to be effective in animals. But this proves nothing. As the Food and Drug Administration in the US has acknowledged, greater than 90% of clinical trials fail - in other words, the vast majority of drugs which were found to be sufficiently efficacious and safe in animals to justify the huge expense of clinical trials are found to be insufficiently efficacious or safe for use in people.

And the reason is clear. Different species are biologically different and so react differently to disease and treatment. Indeed, there are huge differences even within species. Anyone who knows anything about cancer knows that different people with the same cancer, even the same subtype, react very differently to the same drug. Little surprise that a former director of the US National Cancer Institute (by no means opposed to animal experiments) felt compelled to observe wryly some years ago 'The history of cancer research has been the history of curing cancer in the mouse. We have cured mice of cancer for decades and it simply didn't work in human beings'.

Mr Leech falls into another polemical trap. He claims that monoclonal antibody therapies would not have happened without animal research. But he cites no evidence in support. This is the classic logical fallacy post hoc ergo propter hoc ('after this, therefore because of this') which all scientists are taught to avoid. A variant is 'correlation does not equal causation'. Animal researchers, when on the defensive, routinely point out that use of animals is but a small part of a research project. The mere fact that animals were used in particular research does not prove that their use was necessary or useful. Researchers are usually creatures of habit and it is easier to stick to familiar paradigms, which are also more comforting for avoiding product liability claims.

Of course, some effects in some animal species will coincide with effects in some people for some drugs. The law of averages dictates that that should be so. But the point is that no one can have any real idea when they will. And, in the process, it is inevitable that potentially valuable treatments will be missed - for example, had penicillin been tested in guinea-pigs or hamsters, in whom it precipitates death, or rats, in whom it is teratogenic, it might never have seen the light of day, as Alexander Fleming himself is said to have acknowledged.

Mr Leech claims that research on primates led to the development of deep brain simulation (DBS) for Parkinson's disease sufferers. In fact, DBS has been used in humans since the 1940s, with early brain-implanted stimulators employed in human patients with Parkinson's and other movement disorders years before the first ever description of the nonhuman primate model of Parkinson's. DBS was discovered and developed in human patients, and we believe owes nothing to monkeys for its advancement.

There are any number of examples of the absence of correlation between animal tests and experience with people. It is not generally known that the monoclonal antibody which caused such terrible and unexpected side-effects in human volunteers at Northwick Park a few years ago had been tested on primates - the species closest to humans - at 500 times the dose and considered to be safe. Subsequent laboratory tests showed that the effects in people could have been predicted using in vitro methods.

Similarly, not one of the 85 or more candidate AIDS vaccines tested successfully on primates has worked in patients. Over 1,000 potential neuroprotective stroke treatments have been tested in animal "models" but none of the 100 which have progressed to human trials has proved successful. So many more examples could be given.

None of this is a great advert for animal research. Where systematic retrospective analyses of the ability of animal "models" to predict effects in humans have been carried out, they are predictive less than half the time for reproductive toxicity, acute toxicity and drug efficacy. Animal research has undoubtedly generated an enormous amount of data. The real question, however, is how useful that data is and whether other methods, whether existing or with potential for being developed, could do a lot better (as well as be humane).

What is really needed is an intelligent, reasoned debate. Such a debate can, axiomatically, only take place if it is informed. Sadly, many animal researchers and the organisations that support them such as Understanding Animal Research talk the talk about transparency but always find a reason why detailed information about particular research cannot be disclosed. Mr Leech seeks to downplay the significance of the BUAV's recent Information Tribunal win against Newcastle University. What is extraordinary is the lengths to which the university and its researchers went to block the BUAV's FOIA request - despite the fact that the researchers had voluntarily published some information about the research (brain experiments on primates) in journals. To date the university has spent a staggering £230,000 on its legal defence, with more to come - and this at a time when universities and their researchers are facing a tight squeeze on their finances.

Will Mr Leech join us in urging far greater transparency in animal research (anonymised and with genuine confidential information withheld)? After all, what is there to hide? Quite apart from the terrible suffering of animals in laboratories, we all have a massive stake in ensuring that medical research is scientifically sound and that scarce research resources are wisely targeted.

BUAV

Follow Michelle Thew on Twitter: www.twitter.com/@thebuav