In January 2016, six healthy men were rushed to hospital with severe brain injuries, and one subsequently died, while they were testing a new drug for treating pain and mood disorders in a research laboratory in Rennes, France.
Experts convened by the French National Agency for Medicines and Health Products concluded that the compound had caused an “astonishing and unprecedented” reaction in the brain. Yet none of these risks had been picked up in early trials of animals, something the experts found “inexplicable”.1 The drug had been tested on mice, rats, dogs and monkeys with no ill effects, even though doses 400 times stronger than those given to the volunteers had been administered.2
This is just the latest example of how animal testing tells us little—if anything—about the safety of new drugs before they’re tested on people. Instead, more relevant tests based on human samples are already available.
The inadequacies of animal testing were highlighted in 2006 when a drug trial in London was stopped after six young men ended up in intensive care with heart, liver and kidney failure, with the worst affected losing all his toes and several fingertips.
At one stage, their heads swelled up in what became known as the ‘elephant man’ trial, after Joseph Merrick, whose deformities made him a curiosity in Victorian England. Yet the drug, to treat rheumatoid arthritis and leukaemia, had looked to be safe in monkeys at doses 500 times higher than those that nearly proved fatal to the volunteers.3
Not many dead
Officials are quick to point out that such catastrophes are very rare, and assert that animal tests are essential to protect the volunteers and patients who participate in clinical trials, without which there would be no new medicines.
But there’s no way of knowing for sure how rare these life-threatening reactions are because drug companies are not required to release the results of early-stage (phase-I) trials, in which drugs are tested for safety.
In an attempt to quantify the risks to healthy volunteers, a study published in the British Medical Journal in 2015 found 11 serious adverse events happened in 394 phase-I trials.4 This suggests that many notable problems do occur in phase-I trials, but are not reported.
And these were compounds that had been given the green light on safety following animal tests.
Data obtained by Freedom of Information legislation shows that, from 2010 to 2014 in the UK, 7,187 people suffered serious and unexpected adverse drug reactions (ADRs) during clinical trials and 761 died, although none of the deaths could be proven to have been “directly caused” by the test drug.5
In 1993, a combined phase I and II clinical trial (to test both safety and effectiveness) of a potential hepatitis B treatment, conducted by the National Institutes of Health (NIH) in the US, caused unexpected and devastating reactions, including jaundice, liver failure and multiple organ failure.
Five of the 15 participants died, while emergency liver transplants saved the lives of two others. Previous toxicity tests in animals, including a six-month trial in dogs, had given the drug the okay for testing in humans.6
But it’s not just the early phases of clinical trials that are potentially risky. In February 2016, the US Food and Drug Administration (FDA) shut down an international phase-III clinical trial of pacritinib, a blood-cancer drug, after several patients died of cardiac arrest, cardiac failure or brain haemorrhage.7 The actual number of deaths has not been revealed. The drug had demonstrated a good safety profile in mice and dogs.8
Ironically, the ‘elephant man’ trial of 2006 may not have gone ahead at all had a previous trial, in which a patient became ill after being given a very similar drug, been published.9
India’s Supreme Court in 2013 banned clinical trials in India until provisions were adopted to record and report all adverse events and deaths, and participants’ informed consent. This followed several scandals and public-interest litigation petitions filed by non-government organizations in the wake of government data showing that more than 2,600 patients participating in clinical trials had died between 2005 and 2012, and nearly 12,000 had suffered serious adverse effects. Of these, 80 deaths and more than 500 serious side-effects were directly attributed to the drug being trialled.10
Supreme Court Justices R.M. Lodha and A.R. Dave said, “It really pains us that illiterate people and children of India are being used as guinea pigs by the multinational drug companies.”11 Indeed, India had attracted thousands of clinical trials, owing to its large, unmedicated patient population and relatively low costs, but since the introduction of stricter regulations, many drug companies are now conducting their trials in China, Turkey, Indonesia, Malaysia and Eastern Europe.
Although there are not many publicly available examples of deaths or severe injuries in clinical trials in Europe or America, India’s experience suggests they are not isolated events; in fact, 95 per cent of new drugs fail during clinical trials.12 This is usually because they turn out not to work in humans and/or they cause unpleasant or dangerous reactions, even though the drugs have previously been tested in at least two species of animals and appeared to be both effective and safe.
Animal trials unjustifiable
Scientists have been questioning the value of animal tests for years. In 1984, Professors Laurence, McLean and Weatherall observed: “The methods of assessing toxicity in animals are largely empirical and unvalidated . . . It is urgently necessary to know whether the tests as in fact conducted have sufficient predictive value to be justifiable, or whether they are a colossal waste of resources to no good purpose.”13
More than 30 years later, it seems their worst fears have been confirmed. Dr Jarrod Bailey and colleagues have published a review of studies involving more than 3,000 drugs, in what is considered the most comprehensive analysis ever compiled, which shows that even if a medicine appears to be safe in tests using mice, rats, rabbits, dogs and monkeys, none of the results provides any guarantee that the medicine is also safe for humans.14
One study found that animal tests missed 81 per cent of the serious side-effects of 43 drugs that went on to harm patients.15 Another review of studies “using tens of millions of animals” led scientists from the US Agricultural Research Service in Beltsville, Maryland, and Israel’s Agricultural Research Organization to conclude: “The extraordinary high failure rate of animal models to predict human responses in the context of clinical trials argues for an alternative approach to assessing safety and efficacy”.16
It’s not only participants in clinical trials who are at risk from medicines with an incomplete safety profile. Once a new medicine has been granted a licence, it can be prescribed to millions of people within months of coming on the market. Yet clinical trials aren’t able to detect side-effects that only become apparent when much larger numbers of people are taking the drug. Some patient groups suggest that a good rule of thumb is to avoid taking any new drug until at least seven years after its release.17
As clinical trials are not designed to detect rare side-effects, the only safety net to protect the public from new medicines is preclinical testing—which mainly relies on the use of laboratory animals. Post-marketing surveillance, or pharmacovigilance, is the next important gatekeeper to identify dangerous drugs, but it’s undermined because only around 5 per cent of adverse reactions (ADRs) are ever reported.18
ADRs kill 197,000 people in the EU each year, costing €79 billion euros.19 In the US, around 2.3 million serious ADRs put people in hospital and cause 106,000 deaths each year—and these are only the ones that get reported.20
There’s a better way
A new generation of more relevant and predictive toxicological tools is now being developed by scientists and researchers in academia and industry. One crucial feature is that they use human cells or tissues, or computer models of human organs or systems, or non-invasive imaging of human volunteers, or combinations of these. They include ‘a living system on a chip’ devices and low-risk microdose studies, and they provide invaluable insights
into the functioning of the integrated human system.21
Such human-based tests are frequently able to detect side-effects that were missed by the animal tests used to support the human testing and marketing of medicines that later proved to be deadly.
A micro-liver (called HepatoPac) comprising human liver cells can predict liver damage from fialuridine,22 the potential hepatitis B treatment that killed five patients in the devastating 1993 clinical trial (see page 29). Liver damage had not been predicted by preclinical animal tests.
A laboratory method using human cells was rapidly developed following the ‘elephant man’ trial to model the “cytokine storm” experienced by the volunteers, which did not show up
The US government’s Tox21 initiative tested around 10,000 chemicals using a panel of human-cell-based assays. These assays were able to identify important safety aspects of drugs and chemicals “markedly better” than toxicity tests
Assays based on human genetic markers are able to identify serious problems with the drugs Vioxx and Avandia, which both caused many tens of thousands of deaths due to toxicities that were not predicted by animal tests.25
Computer models are now available that can predict the bioavailability of medicines far more accurately and reliably than animal tests. This will increase the prospects of successful clinical trials in humans.26
Non-animal tests are often faster and cheaper, as well as more accurate and reliable, so enabling accelerated access to safer medicines for patients. Human tissue company Biopta estimates an average saving of £7 for every £1 invested in predictive human assays.27 The director of the US NIH Dr Francis Collins recently told the US Senate Appropriations Committee that, within 10 years, human biochips “will mostly replace animal testing for drug toxicity and environmental sensing, giving results that are more accurate, at lower cost and with higher throughput.”28
The human factor
Animal tests may also suggest that a compound might be a successful treatment for people, only to reveal that the benefits don’t translate across species.
Over 100 treatments for acute stroke have proved ineffective or harmful in patients, despite success in animal models of stroke.29
Around 150 drugs have failed in patients with sepsis, the leading cause of death in intensive care. Researchers warn that false leads from misleading animal models cost many years and billions
For 30 years, serious head-injury victims were given steroids—to reduce the risk of death as predicted by animal studies. But, in 2004, it was found that they actually increase that risk in humans,
and had probably killed more than 10,000 people.31
Not only are animals poor models of safety for humans, but they are also unreliable for demonstrating the effectiveness of treatments too. Just as many drugs fail in clinical trials because they turn out to cause side-effects in humans, many others turn out to be ineffective in humans, despite performing well in animals. This makes drug development extraordinarily expensive because companies need to recoup the costs of clinical trials not only for successful drugs, but also for the nine others that fail for each one that succeeds.
Scientists have argued that humans, not animals, should be the subjects of biomedical research. Cliff Barry, chief of tuberculosis (TB) research at the US National Institute of Allergy and Infectious Diseases, has observed that testing everything first in the mouse “has cost us a new generation of medicines . . . We keep getting led down the garden path . . . This isn’t just true for TB; it’s true for virtually every disease”.32
Dr Azra Raza, professor of medicine at Columbia University, argues that, at least as far as cancer research is concerned, “we have to stop studying mice because it’s essentially pointless, and we have to start studying freshly obtained human cells”.33
Turkish Journal of Gastroenterology Editor-in-Chief Professor Hakan Sentürk challenges other scientific journals to follow his lead and avoid publishing animal research, saying: “Given the limitations of animal models, publishing animal studies would mislead the scientific community into futile research and give the general public false hope. This is unethical . . . Human-relevant approaches should be more aggressively developed and utilized instead. Fortunately, non-animal research methods like established clinical, computational and in-vitro models abound, and new technologies like guts and other organs-on-chips are constantly being developed and validated.”34
Dr Fiona Godlee, Editor-in-Chief of the British Medical Journal, asks: “Where would you place the balance of effort: investment in better animal research or a shift in funding to more clinical research?”35
Dr Elias Zerhouni, former Director of the US NIH, echoes: “Researchers have over-relied on animal data. The problem is that it hasn’t worked, and it’s time we stopped dancing around the problem . . . We need to refocus and adapt new methodologies for use in humans to understand disease biology in humans.”36
We are on the cusp of a new era of biomedical research, where we will all reap the benefits of more effective and safer medicines, designed and tested specifically for humans. The evidence is compelling that animal models can impede medical research and put patients at risk.
Far from jeopardizing progress, a shift to advanced techniques based on human biology would accelerate biomedical research, and deliver safer and better medicines at lower costs: a win–win situation that should be supported by everyone. Investment in human-relevant methods must be prioritized. International cooperation, bringing together the best scientists in industry and academia, is vital. Some excellent programmes, such as the Innovative Medicines Initiative in Europe, are already underway, and need to be supported and expanded.
Crucially, the regulatory guidelines that govern how drugs are developed must be updated and improved to encourage adoption of the best new approaches. The current regulations are stifling innovation by failing to keep pace with scientific progress.
For the past 50 years and in the wake of the thalidomide tragedy, governments worldwide have insisted on animal testing, despite the irony that more animal tests would not have prevented the release of that drug because very few species are harmed by it.37
Despite the availability of many new tests that show a greater ability to predict drug safety in humans, governments still fail to demand that pharmaceutical companies use them.
The Safer Medicines Campaign is currently running an online petition—for UK residents only—that calls on the UK government to mandate more reliable safety tests for new medicines. The goal is to reach 10,000 supporters. In a survey, 75 per cent of healthcare professionals agree that pharmaceutical companies should be legally obliged to test new medicines using methods demonstrated to be the most predictive of safety for humans.
To sign up, visit: www.safermedicines.org
How new drugs get tested
Drugs pass through various testing stages before they become available to the public. It’s reckoned that just one in 5,000 compounds discovered in a lab ever makes it to market. If it does, the whole process will have cost the drug company around £150 million, and will have taken 12 to 15 years.
Preclinical phase. Before a compound is tested on humans, its effectiveness must first be established in the lab. The first tests are usually on cell lines in test-tubes in what are called ‘in-vitro’ (literally, ‘in a glass’) tests. The safety of the new drug is then usually tested on laboratory mice, rats or monkeys (in-vivo tests, ‘in the living’). Only if the compound passes these tests will it then be considered safe and effective for tests in people. These tests are almost never made public, usually because of commercial reasons.
Phase-I trials. The purpose of these trials is solely to test the safety of the new compound. A very small group of young healthy men and women (often medical students) are recruited, and dosages may be increased during the course of the trial. Again, the results of these tests are rarely made public unless they go catastrophically wrong.
Phase-II trials. These usually involve a larger number of people, including those who have the condition the drug is meant to treat. This is the first test of whether the drug works in people.
Phase-III trials. If the new drug has passed the first two phases, it can then be tested in much larger groups of patients, and its effectiveness and safety will often be compared against an existing drug or a placebo (sugar pill). Ideally, these trials should last for at least a year and include several thousands of patients, although much shorter trials involving far fewer people are common. If this phase is passed successfully, the new drug is given a licence and becomes available on prescription.
Phase-IV trials. Monitoring the safety and effectiveness of the new drug continues even after it’s become available on prescription and is being taken by many thousands of people. But it’s far from being a fail-safe system, usually because the vast majority of side-effects and adverse reactions go unreported. Sometimes, worrying data are also hidden, as happened with the painkiller Vioxx. Early signals of cardiovascular toxicity in patients were brushed aside after the animal tests had produced evidence that the drug was safe, and even had heart-protective qualities.