One of the most common statistics used by activists to justify why animal research should be phased-out, or banned, is the frequently-quoted claim that ‘90% of animal tests are ineffective’. Here Dr Becky Jones discusses the truths and mistruths behind this statement and its variations.
The claim of activists that only around 10% of research leads to drugs and treatments that are effective in humans (a 90% failure rate) is actually not disputed by the biomedical community.
While that might sound surprising, the fact is that science involves trial and error, observation and understanding of those two things. The 90% figure has been used to imply that researchers have been wasting their time with pointless experiments on animals, but taken globally, a 10% success rate in biomedical research is still producing a remarkable number of long-term benefits and life-saving treatments for human patients.
These breakthroughs only happen by using the best models available to study disease – in most cases that requires the use of animals. The fact remains that our understanding of the human body, how disease progresses, and the wide availability of so many medicines and treatments, relies heavily on the use of animals in research. And while pressure to phase out animal research grows, it is more important than ever that we recognise the vital contribution of animal research to modern medicine.
Take the Covid-19 pandemic as a recent example. The incredible speed with which three vaccines were approved within less than a year of the emergence of the virus would have been impossible without the use of animals at every stage.
From the decades of research into mRNA and vaccine technology in mice, to the development of specific rodent and ferret models to study the disease, to preclinical vaccine and antibody tests involving mice and monkeys – the role of animals in biomedical research has never been more clearly illustrated.
Below we analyse the various ways that animal rights groups have used this 90% figure and show that it is both inaccurate and deliberately misleading when used to describe animal research.
92% of drugs fail in human trials even though they passed preclinical tests – Cruelty Free International
ANSWER: Drugs fail for many different reasons at every stage of the drug development process, but often thanks to pre-clinical testing with animals, potentially harmful or ineffective drug candidates are rejected. What is true is that almost all approved medicines today have required animal research to be successful.
If you hear that 90% of drugs tested in animals fail once they reach clinical trials, you might wonder how it is that so many are approved every year. The first indication that this statistic might be misleading, is that we don’t hear of hundreds of tragic stories about human volunteers dying every year from fatal side effects in drug trials. In fact, adverse events are incredibly rare, thanks to this initial safety testing in rodents and then either dogs, minipigs or monkeys.
In fact, the majority of drugs that make it to human clinical trials, fail not on the grounds of safety concerns, but because the effect of the new drug is not as good as was hoped – or are only marginally better than existing drugs. It is true that more drugs fail after the first round of human testing than at any other point in the drug discovery process. Yet we don’t suggest that humans are not a good model for testing human drugs– instead we recognise that a diverse population of humans might mean that results seen in a small, controlled study might vary from what we see in a much larger trial.
The majority of drugs that make it to human clinical trials, fail not on the grounds of safety concerns, but because the effect of the new drug is not as good as was hoped – or are only marginally better than existing drugs.
Some activist groups use a quote from Mike Leavitt, former Health and Human Services Secretary of the US Government, who in 2006 said: “Currently, nine out of ten experimental drugs fail in clinical studies because we cannot accurately predict how they will behave in people based on laboratory and animal studies.” However, this quote is taken out of context, and is often incorrectly attributed to the FDA. In the statement, Leavitt was reacting to publishing of new guidelines from the FDA which aimed to improve the drug discovery process as a whole, and does not single out animal research as the sole reason for failure of new drugs.
89% of preclinical studies, most of which involve animals, could not be reproduced - PETA
ANSWER: A failure to reproduce studies accurately is a big problem for biomedical research, but picking out the use of animal studies as the main problem is inaccurate.
In biomedical research, animal studies may be difficult to recreate, but the problem is more to do with laboratory practices than an innate problem with using animals. Even looking at the paper that is often referenced by activists on this claim, it considers preclinical research as a whole – including problems with reproducing cell-based or computer research. It would be wrong therefore to suggest that non-animal methods are the only solution to this – as they too can be difficult to replicate.
It is true that there is what scientists call a ‘reproducibility crisis’, and it affects many different areas of science, not just animal studies. A Nature study found that 77% of chemists, and more than 60% of physicists struggled to reproduce results obtained in another lab. Another study of some of the most influential cancer sciences papers showed that more than half of them could not be reproduced, blaming “underlying data, protocols, statistical code, and reagent sources…and tumour cells not behaving as expected.”
It is true that there is what scientists call a ‘reproducibility crisis’, and it affects many different areas of science, not just animal studies.
Often, the 89%/90% figure is used alongside claims that animal research is unreliable and cannot be replicated, and therefore it has no value. EARA has attended several conferences where activist groups claim that as much as 50% of all animal research is not reproducible. However, it’s clear when looking more closely at all the studies above we have referenced, that the problem is wider than just animal research.
90% of animal tests are not predictive of human disease - Humane Research Australia
ANSWER: We have medicines, vaccines and medical devices to improve and extend the quality of life for countless patients, and have eradicated some diseases completely, with research using animals an important part of that process.
In an article written by Associate Professor Brett Lidbury, of Humane Research Australia, he referenced a study questioning whether animals have been relevant in any medical discovery throughout history. If this were true, then treatments for diabetes, heart rhythm issues, cancers and even bacterial infections would not exist today. One of the most used cancer treatments, Herceptin, was developed thanks to identification of a cell receptor involved in breast cancer development in mice.
There are countless examples of how animal research has played a part in ensuring that once life-threatening diseases can be managed or eradicated altogether. Discovery of insulin was only possible thanks to essential early studies in dogs, and larger animals such as pigs and sheep have helped our understanding of how to treat circulatory diseases such as heart attacks and strokes.
Activists also argue that as we are not 70kg mice why would we use a mouse to understand how something works in a human? This is another common argument, and it can seem pretty convincing. After all, surely you wouldn’t test a new drug for a horse on a hamster? Except, maybe you would.
Mammals, in evolutionary terms, resemble each other very closely. Between species, many of their bodily systems and tissues are very similar, if not the same. Smaller animals such as mice and rats, share as much of 85% of our DNA, and can help explain how individual genes function in a healthy body, or what happens when they go wrong.
There are countless examples of how animal research has played a part in ensuring that once life-threatening diseases can be managed or eradicated altogether.
Mice and zebrafish are especially useful for studying genetics – and as they breed and age quickly, we are able to see the changes that an individual gene fault might have on the body, and use this to help diagnose or develop treatments for rare conditions that otherwise would be difficult to understand. Larger animals can help our understanding of how whole systems work, thanks to their similar ratios of size and organ size.
90% of basic scientific discoveries, most of which involves animals, fails to lead to human treatments - PETA
ANSWER: There are many examples of studies decades old that have been dusted off to prove invaluable to modern day research. The very nature of basic research means that it is not always obvious how the research is useful immediately.
Basic research helps researchers understand more about how a healthy bodily system works, which is essential when deciding what to do if that system is not functioning properly. For that reason, not all basic animal research leads directly to a cure or a treatment for a human disease and individual pieces of this research may not directly have an immediate impact on biomedical progress.
Science doesn’t get it right first time, every time – some of the greatest achievements and Nobel Prize winning work has come through many failures before producing the celebrated results we know today. Many examples of this are celebrated in the yearly Golden Goose Awards, dedicated to fundamental research which has later has a large impact on society.
PETA’s claims, in their statement, misunderstand, the scientific process. Using animals in basic research is usually just one part of a project where a research group will look through past literature to develop a hypothesis, and then often go through lengthy in vitro studies using cells, tissues or computer simulations, before even considering looking at an animal model.
Science doesn’t get it right first time, every time – some of the greatest achievements and Nobel Prize winning work has come through many failures before producing the celebrated results we know today.
Animal studies fail to predict real human outcomes in 50 to 99.7 percent of cases …non-animal replacement methods provide faster and more human-relevant answers – Humane Society International
ANSWER: The non-animal alternatives offered certainly play a role in reducing the numbers of animals needed – but there is a very long way to go before they can reliably replace animal studies in many areas of research.
Organoids, 3D cultures of cells and tissues that can replicate organ structures are improving and the technologies are certainly providing more opportunities to reduce the numbers of animals needed in research. The 2021 World Congress on Alternatives, in Maastricht, Netherlands, featured many presentations showing how they can be applied to different areas of biomedical research with great success.
However, not even experts on organoid culture think that this technique can fully replace animals. Writing to The Scientist magazine, prominent organoid researcher Professor Hans Clevers has said that while organoids may replace animals in some areas. “There will always be the need for confirmation of any finding . . . in vivo (in animals),” he said.
In future, just as now, non-animal methods might still need to be validated in animal models, and there are areas of basic or behavioural research which might never be completely replaced. While conducting research for this article, EARA found no data to show how predictive non-animal experiments are, after an extensive search. Perhaps with greater use of alternatives it might be possible to see how non-animal models compare with the animal data in terms of predictability.
In future, just as now, non-animal methods might still need to be validated in animal models, and there are areas of basic or behavioural research which might never be completely replaced.
What would be the effect if researchers were banned from using animals for research?
It is clear then that where animals are used now, they are used because there is simply no other safe and effective way to produce the same results. While the biomedical research community actively works to reduce the numbers of animals needed in research, we are still a long way from achieving this aim. If we were to ban animal research tomorrow – the sector would simply not be able to develop, produce or safely test any new medicines or vaccines.
Directive 2010/63/EU, which governs the use of animals in for scientific purposes, states that animals cannot be used if a suitable alternative is available. The ultimate goal enshrined in this Directive is to ensure that the scientific community works towards a time when animals are no longer needed in research.
Using animals in research is often time consuming and expensive, and so even without regulations such as EU Directive 2010/63 and other ethical considerations, there is no incentive to use an animal where a non-animal alternative exists. And although alternatives are slowly becoming more widely available, the unfortunate fact is that animals are still necessary for biomedical research that benefits both humans and animals. The EU scientific organisation that validates alternative methods to animal testing, EURL ECVAM, has approved just 50 alternative methods in the last 25 years.
Rising pressure on governments from vocal animal activists makes it seem like the alternatives are all there, that animal research is useless, and that the only thing preventing science making the switch to non-animal methods are stubborn researchers who don’t want to change their ways. But the more complicated truth is that we will still rely, for the foreseeable future, on a choice between animal and non-animal models, and will use whichever is the most appropriate, to advance our knowledge of human and animal health.
End
See also: What does Will Young have to do with animal testing? Understanding Animal Research 26/10/2023