Ian Le Guillou recently posted an essay at Understanding Animal Research titled: “We need to stick with research through the ups and the downs.” The essay begins by discussing the failure of animal models to find a cure or treatment for Alzheimer’s disease (AD). Le Guillou claims that all of the failures were not a waste of time and money as: “Targeting the plaques was the obvious thing to do, the approach most likely to yield a cure and now we know that it is not going to work.”
In some circumstances Le Guillou would be correct but this is not one of them. I wrote the following in Is the use of sentient animals in basic research justifiable?
When Thomas Edison was asked about all his failures in trying to invent a light bulb, he supposedly said that he had not failed 100 times (or 1000 times, sources vary on the exact number) but that he had succeeded in finding 100 ways that would not work and that when he had eliminated the ways that would not work, he will have found the one way that will work. That is what basic research is.
This is a cute little story and might even be true (again sources vary), but it has essentially nothing to do with our discussion. First, Edison was not spending resources that society valued beyond their monetary face value. Society values children more than orange juice and endangered plants more than those not endangered. Society does have a hierarchy of value. Scientists can spend resources that are largely of no or very little value to society, like common chemicals in a beaker, on the off chance something will result from it. But society mandates that researchers cannot spend, with impunity, resources it does value. Society values sentient animals more than inorganic materials.
Second, Edison was spending his own time and his own funds, and using resources society did not find to have inherent value. Therefore, society had essentially no legitimate grounds for telling Edison what to do with any of the above. Third, society did not fund Edison over other options. Fourth, Edison made no promises to society in exchange for its resources. His failures were largely irrelevant in that regard.
In addition, Le Guillou makes the assumption that because animal models targeting the plaques failed to translate to humans, this means that targeting the plaques in general is not the answer to AD in humans. Maybe targeting the plaques is the answer, maybe it is not, but the results of targeting the plaques in animal models have no predictive value for humans with AD.
Le Guillou argues that pessimism regarding animal models because of failures, like those from AD research using animal, is dangerous. Again, he has a point but the point is not applicable to this case. One cannot judge the value of practice or test or model based solely on a few failures. Doing so would be an example of the fallacy of insufficient statistics. But Le Guillou refutes his own claim by quoting the British Union for the Abolition of Vivisection (BUAV) as stating:
Due to the lack of predictability of animal tests and sheer volume of such tests that are conducted, the entire process is extremely ineffective, wasteful and expensive. Reliance on such a system would not be tolerated in other sectors of professional business.
Yep, that pretty much sums it up. A more scientific summary would be:
While trans-species extrapolation is possible when perturbations concern lower levels of organization or when studying morphology and function on the gross level, one evolved, complex system will not be of predictive value for another when the perturbation affects higher levels of organization.
The paradigm of animal modeling has been extensively explored and the conclusion, both on empirical grounds and from theory, is that animal models currently have no predictive value for human response to drugs and disease, nor will they in the future. Pointing out that animal models have no predictive value is not an example the fallacy of insufficient statistics as it is based on science, not anecdotes, and supported by scientific theory. The track record from animal-based basic research is horrible (Crowley 2003) and, in an attempt to offset that, most of animal-based basic research is currently pretending to be applied research: research that has a high predictive value. (See Is the use of sentient animals in basic research justifiable? for more on this.) Animal modeling thus fails regardless of how it is categorized. (Note that this does not mean basic research per se fails. Basic research in chemistry and physics has produced results, as has much of the non-animal basic research in the biomedical sciences.)
Lastly, Le Guillou attempts to use the intact systems argument to justify animal models. (I have addressed the intact systems argument in More Misrepresentations, Fallacies, and Other Lies. Part II and Vivisection Or Death: Part III, No Other Options.) Briefly, while animals and humans are both intact systems, they are differently complex and thus differently intact. This invalidates the use of animals as models for predicting human response to drugs and disease. To suggest that one evolved complex system has predictive value, at higher levels of organization, for a second is to deny evolution and what we have learned from complexity science. The animal modelers, who pose as the stalwart defenders of science, are the science-deniers on this one. (For more depth on why the intact systems argument fails, see our articles Questions regarding the predictive value of one evolved complex adaptive system for a second: Exemplified by the SOD1 mouse and Animal models and conserved processes.)
Le Guillou’s last sentence states: “we can’t afford to be complacent about research, because who knows where the next potential cure will come from?” If this were true, we should be funding psychics to tell us which drugs to develop. In fact, we have a very good idea of where the next medical advances will come from: human-based studies. THAT is what society should be funding!
Crowley, W. F., Jr. (2003). "Translation of basic research into useful treatments: how often does it occur?" Am J Med 114(6): 503-505. http://www.ncbi.nlm.nih.gov/entrez/query.fcgi?cmd=Retrieve&db=PubMed&dopt=Citation&list_uids=12727585