Animal Testing: Science or Fiction?
24th March, 2009
MPs, medical professionals and scientists unite in demanding a thorough evaluation of the utility of vivisection. By Kathy Archibald, Science Director of Europeans for Medical Progress
Most of us know that cancer, heart disease and stroke are the leading causes of death in the West. But many people would be surprised by the next biggest killer: side effects of prescription medicines. Adverse drug reactions kill more than 10,000 people a year in the UK (and more than 100,000 in the US), costing the NHS alone £466m per year.
The pharmaceutical establishment constantly reassures us that all drugs are tested for safety and efficacy on animals before they can be administered to humans. When challenged about the ethics of vivisection, their defence typically goes like this: ‘Which do you think is more important: your child’s life or a rat’s?’ Given this choice most people would thankfully sacriﬁce the rat.
But what if you were told that the current animal testing procedures are seriously ﬂawed? Consider the following evidence:
■ Arthritis drug Vioxx, withdrawn from the global market in September 2004, appeared to be safe and even beneﬁcial to the heart in animals, but caused as many as 140,000 heart attacks and strokes in the US alone. The associate safety director of the US Food and Drug Administration (FDA) described it as the ‘single greatest drug-safety catastrophe in the history of the world’.
■ Many studies published in the scientiﬁc literature comparing drug side effects in humans and animals have found animal tests to be less predictive than tossing a coin. One review of human-animal correlation in drugs that had been withdrawn because of adverse reactions found that animal tests predicted the human side effects only six out of 114 times.
■ Hundreds of drugs to treat strokes (eg, Cerestat, MaxiPost, Zendra, Lotraﬁban, gavestinel, nimodipine, clomethiazole) have been found safe and effective in animal studies and then injured or killed patients in clinical trials.
■ Hormone replacement therapy (HRT), prescribed to many millions of women because it lowered monkeys’ risk of heart disease and stroke, increases women’s risks of these conditions signiﬁcantly. The chairman of the German Commission on the Safety of Medicines described HRT as ‘the new thalidomide’. In August 2003 The Lancet estimated that HRT had caused 20,000 cases of breast cancer over the past decade in Britain, in addition to many thousands of heart attacks and strokes.
■ Dr Richard Klausner, former director of the US National Cancer Institute (NCI), lamented: ‘The history of cancer research has been a history of curing cancer in the mouse. We have cured mice of cancer for decades, and it simply didn’t work in humans.’ The NCI also believes we have lost cures for cancer because they were ineffective in mice.
■ Cigarette smoke, asbestos, arsenic, benzene, alcohol and glass ﬁbres are all safe to ingest, according to animal studies.
■ Of 22 drugs shown to have been therapeutic in spinal cord injury in animals, not one is effective in humans.
■ Of 20 compounds known not to cause cancer in humans, 19 do cause cancer in rodents.
■ Dr Albert Sabin, the inventor of the polio vaccine, swore under oath that the vaccine ‘was long delayed by the erroneous conception of the nature of the human disease based on misleading experimental models of [it] in monkeys’.
■ Penicillin, the world’s ﬁrst antibiotic, was delayed for more than 10 years by misleading results from experiments in rabbits, and would have been shelved forever had it been tested on guinea pigs, which it kills. Sir Alexander Fleming himself said: ‘How fortunate we didn’t have these animal tests in the 1940s, for penicillin would probably never have been granted a licence, and possibly the whole ﬁeld of antibiotics might never have been realised.’
■ Thalidomide, the infamous cause of birth defects in more than 10,000 children in the early 1960s, induces birth defects in very few species. Dr James Schardein, the doyen of birth defect studies, says: ‘In approximately 10 strains of rats, 15 strains of mice, 11 breeds of rabbits, two breeds of dogs, three strains of hamsters, eight species of primates, and in other such varied species as cats, armadillos, guinea pigs, swine and ferrets in which thalidomide has been tested, teratogenic effects have been induced only occasionally.’ Ironically, if thalidomide, the drug whose side effects made animal testing obligatory, were assessed exclusively on its results in such tests it would still be passed today. Even the Handbook of Laboratory Animal Science admits that ‘uncritical reliance on the results of animal tests can be dangerously misleading and has cost the health and lives of tens of thousands of humans’.
So why use animals to test new drugs?
Animal testing became legally enshrined in response to the thalidomide tragedy. The UK Medicines Act 1968 followed the US Kefauver-Harris Act, which was implemented in 1961 in the midst of the thalidomide furore to ensure that the FDA received proof of safety and efﬁcacy for all new drugs. The intention was good but the reliance placed on animal tests to ensure safety was tragically ill-informed.
It has been known among scientists and the pharmaceutical industry for decades that animal testing is scientiﬁcally unreliable. As long ago as September 1962 The Lancet commented: ‘We must face the fact that the most careful tests of a new drug’s effects on animals may tell us little of its effect in humans.’ In 1964 Dr J Gallagher, the medical director of Lederle Laboratories, admitted: ‘Animal studies are done for legal reasons and not for scientiﬁc reasons.’
So, pharmaceutical companies conduct animal tests simply to satisfy government regulators. Crucially, animal data also provide liability protection when drugs kill or injure people. Industry can point to the rigorous animal tests they have performed and claim that they have done their best to ensure against tragedies occurring, thus minimising any damages awarded against them.
From the perspective of satisfying the regulators, pragmatic selection of species will demonstrate whatever is required of a drug, whether it is favourable safety or efﬁcacy. And companies are not required to submit all their animal data, but only that from any two species (one rodent and one higher mammal). Dr Irwin Bross, former director of the world’s largest cancer research institute, the Sloan-Kettering, observed: ‘Whenever government agencies or polluting corporations want to cover up an environmental hazard, they can always ﬁnd an animal study to “prove” their claim. They can even do a new animal study which will come out the way they want by choosing the “right” animal model system.’
Placing massive emphasis on animal-safety data has also allowed pharmaceutical companies to avoid the expense of conducting clinical trials as extensively as they should. Since the 1950s doctors have been saying that clinical trials should involve more people, last for a longer period of time and use representatives of a broader swathe of society than the young, white males of standard practice. Women are generally not utilised in case they might be pregnant: the manufacturer would be held liable for any unanticipated birth defects. Very often trials do not even include representatives of the patient population the drug is designed to treat. This absurd situation clearly needs to be addressed.
There is no getting away from the fact that people have to be the ultimate guinea pigs for testing new treatments. Clearly, the health and safety of research volunteers and patients should be paramount and the best safeguards should be in place to protect them.
Testing drugs safely on people
New drugs go through three basic testing phases: in vitro (test-tube) and in silico (computer) modelling; animal testing; and, ﬁnally, human trials.
Before a drug is tested in humans, there should be persuasive evidence that it is safe and effective. No method, neither animal, human nor test-tube, can predict the reactions of every patient with 100 per cent accuracy. Reactions differ between sexes, ages, ethnic groups, even between family members. We are all different, but not as different from each other as we are from animals, with which the differences are so great that they render extrapolation hazardous. Non-animal methods are not completely fail-safe, but do offer more security.
There are excellent in silico and in vitro testing methods available today. Many companies specialise in virtual screening of drugs for potentially toxic effects. A wide range of predictive software is available, including complete clinical trial simulations. Other companies focus on safety and efﬁcacy assessments in human tissues. A 10-year international study proved that human cell culture tests are more accurate and yield more useful information about toxic mechanisms than traditional animal tests.
In place of animal-based pre-clinical studies, subsequent clinical trial patients and volunteers would be better protected by the adoption of preliminary microdosing studies (or ‘phase 0’ clinical trials). Microdose studies involve the administration of ultra-small (and safe) doses of the test drug to volunteers monitored by scanners. Human microdosing, based on the concept that the best model for man is man, helps in selecting the best drug candidates before advancing into full development, thereby reducing the chances of failure in later, more risky and more expensive phases.
During clinical trials, relevant pharmacological measurements should be made, which would give early warning of potential problems. It is true that some rare side effects will only be detected when drugs are prescribed to large numbers of people. This is why post-marketing drug surveillance is so important and should be strengthened, in order to pick up these effects as quickly as possible. Reports of adverse reactions to drugs are currently soaring in the US, where a record 422,500 adverse events were reported to the FDA in 2004. The FDA cautions that the actual number is likely to be between 10 and 100 times greater because of under-reporting.
Many drugs that have been pronounced safe on the basis of animal tests have gone on to injure or kill people.
■ Vioxx, for arthritis, caused up to 60,000 deaths between its launch in 1999 and its withdrawal in 2004;
■ Baycol (Lipobay), used to treat cholesterol, caused more than 10,000 cases of serious muscle-wasting or death between 1997 and its withdrawal in 2001;
■ Rezulin (troglitazone), for diabetes, killed more than 400 people between 1997 and its withdrawal in 2000;
■ Propulsid (cisapride), for heartburn, killed more than 300 adults and children before being withdrawn in 2000;
■ Opren, for arthritis, caused 61 deaths and 3,500 serious injuries and was withdrawn in the 1980s;
■ Eraldin, a heart treatment, killed 23 people and blinded many more, and was withdrawn in the 1970s. Its devastating side effects were not reproducible in any species except man;
■ Isoprenaline, for asthma, killed 3,500 young people in Britain alone and was withdrawn in the 1960s. Intensive studies with rats, guinea pigs, dogs and monkeys, at huge dosages, failed to elicit similar results.
This article first appeared in the Ecologist May 2005