Surprisingly, most scientists do not regularly review the predictability of animal tests as part of their work- there is an assumption that animal tests work based on certain similarities with humans and long history of use. But it is not enough to come up with one or two examples of where animal tests appear to have been helpful. Just because animals were used as part of the drug discovery process also does not mean that they had to be used, nor that they were a key part. To be an effective scientific approach, animal testing needs to be consistently predictive – but this is far from the case.
“The claim that animal experimentation is essential to medical development is not supported by proper, scientific evidence but by opinion and anecdote. Systematic reviews of its effectiveness don’t support the claims made on its behalf.” (Pound, P. et al. 2004. British Medical Journal 328, 514-7.)
Where reviews of the predictability and utility of animal tests have been done, the results have been damning:
A review of highly cited animal studies of 76 potential new drugs found that, despite the studies indicating that the drug would work, only just over half were followed up in humans and of these, one third were found to produce conflicting results. Only 11% of the drugs were subsequently placed on the market. The authors concluded that “patients and physicians should remain cautious about extrapolating the findings of prominent animal research to the care of human disease”. (Translation of research evidence from animals to humans. Journal of the American Medical Association 2006; 296, 1731-2.)
A review of six new drugs found that for two of the treatments human trials were conducted at the same time as the animal studies, while for another three human trials went ahead despite evidence of harm from the animal studies. The authors concluded; “This suggests that the animal data were regarded as irrelevant, calling into question why the studies were done in the first place and seriously undermining the principle that animal experiments are necessary to inform clinical medicine”. (Where is the evidence that animal research benefits humans? British Medical Journal 2004; 328:514-7.)
A follow up review of a further six interventions for various human diseases found that the ‘animal models’ failed to accurately predict the human outcome in four cases. For two of these the animal tests actually suggested the drug would be helpful when it was in fact harmful. (Comparison of treatment effects between animal experiments and clinical trials: systematic review. British Medical Journal 2007; 334; 197-200)
A review of 101 “high impact” basic science discoveries based on pre-clinical (animal) research claiming human relevance found that only 5% successfully translated into approved treatments within 20 years. The authors concluded that, “Even the most promising findings of basic research take a long time to translate into clinical experimentation and adoption into clinical practice is rare”. (Translation of highly promising basic research into clinical applications. American Journal of Medicine 2003, 114: 477-484).
In its key report on how to improve the development of drugs for people, the Food and Drug Administration (FDA) noted that the chances of a drug being suitable for human patients even after it had passed all the animal tests and other laboratory studies, was only 8%. (Innovation or stagnation: Challenge and opportunity on the critical path to new medicinal products. US Department of Health and Human Services, Food and Drug Administration; March 2004).
The US drug industry invests $50 billion per year in research and yet the number of new drugs that are approved annually is no greater than it was 50 years ago. Only 6% of the 4,300 companies that are engaged in drug innovation have registered a new drug since 1950. (Lessons from 60 years of pharmaceutical innovation. Drug Discovery 2009, 8: 959-968).
A review of the ability of animal tests to predict acute toxic effects in humans conducted by the drug industry found that out of 150 drugs, tests on rats and mice only predicted 43% of human effects. (Concordance of the toxicity of pharmaceuticals in humans and in animals. Regulatory Toxicology and Pharmacology 2000; 32: 56-67).
Researchers from Utrecht University found that out of 93 serious adverse reactions prompting action by the regulatory authorities only 19% could have been predicted by the animal tests done before the drugs were released onto the market. (The ability of animal studies to detect serious post marketing adverse effects is limited. Regulatory Toxicology and Pharmacology 2012, 64: 345-349.
A review of tests to find out whether chemicals affect the development of unborn animals found that the animal tests only predicted damage to human foetuses just over 50% of the time – in other words, virtually no better than tossing a coin. (The future of teratology research is in vitro. Biogenic Amines 2005; 19:97–145.)
Only 5% of the hundreds of compounds that are tested in human trials for use against cancer actually make it to the pharmacist’s shelf, a success rate 50% lower than most other types of drug. (Can the pharmaceutical industry reduce attrition rates? Nature Reviews: Drug Discovery 2004, 3: 711-725).
Dr. Homer Pearce, a cancer researcher at drug company Eli Lilly, said “If you look at the millions and millions and millions of mice that have been cured, and you compare that to the relative success, or lack thereof, that we’ve achieved in the treatment of metastatic disease clinically, you realise that there just has to be something wrong with those models”. (Why we are losing the war on cancer. Fortune Magazine, 22 March 2004).
“That said, personally I don’t think most mouse model studies translate immediately to human disease. Ninety-nine per cent of them don’t” said Lynda Chin, a cancer researcher at the University of Texas. (Recapitulating human cancer in a mouse. Nature Biotechnology 2013, 31: 392-395).
See Cancer briefing for more information and criticisms of animal experiments.
Derek Hill, professor of medical imaging science at UCL, said, “The drug development world is littered with [Alzheimer’s] drugs that seemed to work on transgenic mice but don’t work in people”. (Cancer drug reverses Alzheimer’s in mice. New Scientist 2012, 18:17).
A review of 160 potential treatments for brain diseases including Parkinson’s disease, Alzheimer’s disease, multiple sclerosis, stroke and spinal cord injury concluded that, “there is a great discrepancy between the intervention effects found in preclinical animal studies and those found in clinical trials of humans with most of these interventions rarely achieving successful translation”. (Evaluation of excessive significance bias in animal studies of neurological diseases. PLoS Biology 2013, 11;7).
Dr. Phillipe Menasche, a professor of cardiovascular surgery at the University of Paris, speaking about the hype of stem cell therapy for heart disease, said “Once again in medicine, clinical outcomes have not matched the hopes raised by animal data”. (Skeletal myoblasts and cardiac repair. Molecular Cell Cardiology 2007, 45: 545-553).
British Heart Foundation researcher Dr. Paul Williams, said “Despite a huge amount of basis science research, promising animal studies, and numerous clinical trials, to date no gene therapy [for heart disease] has demonstrated unequivocal benefit in the clinical setting”. (Gene therapy for cardiovascular disease – white knight of white elephant? British Cardiovascular Society Editorial, 8 Feb 2010).
See Heart disease briefing for more information and criticisms of animal experiments.
Of over 1,000 potential neuroprotective stroke treatments that had been successful in animal models, only approximately 10% progressed to human trials but none were successful. Of these approximately 50% were tested in humans before the animal tests were published. (1,026 Experimental Treatments in Acute Stroke. Annals of Neurology 2006; 59, 467-77.)
See Stroke briefing for more information and criticisms of animal experiments.
An analysis of 174 trials of neuropathic pain treatments found that the majority of patients did not report significant pain relief, and, in some trials, a high number of people dropped out early because of the adverse drug reactions. (The evidence for pharmacological treatment of neuropathic pain. Pain 2010, 150: 573-581). “The animal models of neuropathic pain just don’t seem to be that efficient,” said Dr. Andrew Rice, pain researcher at Imperial College. (Animalgesic effects. Nature Medicine 2010, 16: 1237-1240).
A review of the animal and human trials of HIV vaccines found that: “To date, 85 candidate AIDS vaccines have been tested in 197 clinical trials. Just 12% of these trials have reached Phase II, only seven (3.5%) have reached Phase III, and altogether, 18 trials were prematurely terminated. None has been successful”. (An Assessment of the Role of Chimpanzees in AIDS Vaccine Research. Alternatives to Laboratory Animals 2008; 36, 1–48).
Out of 150 clinical trials testing potential drugs to treat sepsis, not one has been successful. When a team of researchers compared gene responses in mouse models with the patterns found in human sepsis patients, they were unable to find a single similarity; in fact the responses in mice were “close to random” in matching the equivalent human gene response. (Genomic responses in mouse models poorly mimic human inflammatory diseases. PNAS 2013, 110:3507-12). “Clearly, current animal models seem to be incapable of predicting human trials of new agents,” says Dr. Mitchell Fink, a surgeon at the University of California. (Rodent models of sepsis found shockingly lacking. Nature Medicine 2012, 18:998).
Dr. Mark Davis, an immunologist at Stanford University in California, said that mice make a “lousy model” for the human immune system. Davis says: “Hundreds of clinical trials have been based on curing mice, but almost none led to clinical treatments”. “Studies on mice are very elegant and beautiful, but they aren’t reflecting the needs of the [human] population,” says Dr. Jacques Banchereau, head of the Baylor Institute for Immunological Research in Texas. (Immunology uncaged. Science 2010, 327:1573).
Daniel Sarewitz, co-director of the Consortium for Science, Policy and Outcomes at Arizona State University, said “The technology to produce unlimited numbers of identical transgenic mice attracts legions of researchers and abundant funding because it allows for controlled, replicable experiments and rigorous hypothesis-testing – the canonical tenents of ‘scientific excellence’. But the findings of such research often turn out to be invalid when applied to humans”. (Beware of the creeping cracks of bias. Nature 2012, 485: 149).