The team studied only one area, the use of animal models for disease treatments. They did not look at more basic research such as physiology, nor did they cover safety testing, which tends to involve standardised tests. Additionally, they concede that the sample size “was far too small to give precise statistical estimates of the extent of concordance”.
No-one claims that animal experiments are perfect, and indeed it is well-known that at every stage of drug development – computer modelling, modern robotics, test tube studies, animal research, clinical trials – a large proportion of candidate drugs are discarded. That’s why developing new treatments is such a long and difficult process and only two or three out of 10,000 potential treatments make it as far as clinical trials (see the ABPI’s web pages for more explanation of the process).
So it’s remarkable that in this study of six experimental treatments, chosen at random, as many as half of them worked well in both animals and humans. Two other treatments gave negative results in humans but positive results in animals. In the sixth example, the animal experiments gave insufficient data to draw conclusions.
In fact, the authors were critical of the design of the experiments they looked at, rather than the validity of animal research in general. They stressed that their aim was to find ways of improving experimental methodology, such as
• Prospective registration of animal experiments that might prevent publication bias.
• Research to identify the design features of animal experiments that introduce bias and that could improve validity.
• Reporting standards for animal experiments that might improve the quality of published reports.
• Promoting reviews of both animal and human studies that would encourage researchers to develop a shared language and better communication.
It was perhaps inevitable that antivivisectionists would draw their own, invalid conclusions. Rather than taking the trouble to analyse the results of the review, Animal Aid used it as a pretext to rehearse its usual unsubstantiated claims :
“animal experiments aren’t just cruel, they don’t appear to work”
and
“across a range of important human ailments, animal research provides misleading and conflicting information and is therefore dangerously unreliable”.
The observation by Animal Aid that “clinical trials with human patients get underway even before the animal research is completed” is plain silly as a critique of animal testing.
One wonders, though, whether the study authors were trying compare apples with oranges. Is it reasonable to expect that many scientists at many centres studying similar treatments in animals would adopt closely matched methodologies and protocols? Applied research can be used to answer many different questions, not just “does this drug work?” As a result, animal studies are not always conducted using the same standardised procedures as clinical trials. Should they be? Can they be? This is an area for more deliberation, and relates to work currently being carried out by the National Centre for the 3Rs.
*Research Monograph from the NHS National Coordinating Centre for Research Methodology, RM04/JH18/IR: Testing treatments on animals: Relevance to humans, Pablo Perel, Ian Roberts, Emily Sena, Philipa Wheble, Peter Sandercock, Malcolm Macleod, Luciano E Mignini, Pradeep Jayaram and Khalid S Khan (2006).
<Wrap up...>
