Where is the evidence that animal research benefits humans?
BMJ 2004; 328 doi: https://doi.org/10.1136/bmj.328.7438.514 (Published 26 February 2004) Cite this as: BMJ 2004;328:514
All rapid responses
Rapid responses are electronic comments to the editor. They enable our users to debate issues raised in articles published on bmj.com. A rapid response is first posted online. If you need the URL (web address) of an individual response, simply click on the response headline and copy the URL from the browser window. A proportion of responses will, after editing, be published online and in the print journal as letters, which are indexed in PubMed. Rapid responses are not indexed in PubMed and they are not journal articles. The BMJ reserves the right to remove responses which are being wilfully misrepresented as published articles or when it is brought to our attention that a response spreads misinformation.
From March 2022, the word limit for rapid responses will be 600 words not including references and author details. We will no longer post responses that exceed this limit.
The word limit for letters selected from posted responses remains 300 words.
Dear Editor,
We read with some concern the paper by Pound et al on the benefits of
animal research to humans1. The authors’ analysis was entirely descriptive
and their arguments were focussed on only six papers, all reviews, that
covered areas of research whose findings could be directly translated to
clinical trials. They found that clinical trials either went ahead despite
the contrary results of animal studies, or ran concurrently with them. The
authors therefore stood by their statement that there is little evidence
to support the view that animal research has contributed to the treatment
of human disease. Clearly this is a reasonable conclusion bearing in mind
the question being asked.
However, most animal research is directed at understanding the basic
mechanisms of disease from which future therapies may arise. The Medical
Research Council has summarised many areas where such research is leading
to treatments or prevention, including cystic fibrosis, deafness,
depression, tuberculosis, malaria and cancer2. In our area of interest,
much of the understanding of the pathogenesis of emphysema has come from
studying rodent models3,4, and this is leading to the development of new
therapies such as the retinoic acid receptor agonists5 that are now in
clinical trials. Only a minority of animal studies are a direct precursor
to clinical trials, although obviously such studies are also important.
The fact that sometimes the results are either ignored or not awaited by
clinical researchers is surely a criticism of the speed with which such
studies are undertaken, often ahead of true understanding, rather than of
the importance of the animal research itself. That is not to say that
coordination of research efforts cannot be improved, or that quality of
study design does not need addressing.
However, the fact is that this paper in a high profile journal will
be interpreted by the media and the general public as evidence that animal
research is not benefiting medical science. This is unfortunate since if,
instead of this quite narrow meta-analysis, the BMJ had published a review
of the high quality translational animal research that is produced in the
UK and internationally, they probably would draw a quite different
conclusion.
1. Pound P, Ebrahim S, Sandercock P, Bracken M.B, Roberts I. Where is
the evidence that animal research benefits humans? BMJ 2004;328:514-517.
2. Mice and Medicine: animal experiments, medical advances and the
MRC. Medical Research Council 1993.
3. Shapiro S.D, Goldstein N.M, Houghton A.M, Kobayashi D.K, Kelley D,
Belaaouaj A. Neutrophil elastase contributes to cigarette-induced
emphysema in mice. Am. J. Path. 2003;163(6):2329-35.
4. Hautamaki R.D, Kobayashi D.K, Senior R.M, Shapiro S.D. Requirement
for macrophage elastase for cigarette-induced emphysema in mice. Science
1997;277(5334):2002-4.
5. Massaro G.D, Massaro D. Retinoic acid treatment abrogates elastase
-induced pulmonary emphysema in rats. Nat. Med 1997;3(6):675-7.
Yours sincerely,
Paul A. Dawkins (Clinical Lecturer, Respiratory Medicine)
Robert A. Stockley (Professor of Medicine)
Competing interests:
None declared
Competing interests: No competing interests
Pound et al. (1) take an extremely narrow approach to the question:
'Where is the evidence that animal research benefits humans?', and they
misinterpret their own data.
They open their paper with the extraordinary statement: 'Clinicians
and the public often consider it axiomatic that animal research has
contributed to the treatment of human disease, yet little evidence is
available to support this view'. This is simply untrue and is seriously
misleading. There is a huge amount of evidence for the contribution made
by animal research. Summaries are included, for example, in the MRC
booklet 'Mice and Medicine' (2), in evidence provided to the House of
Lords Select Committee on Animals in Scientific Procedures (3), on the
website of the Research Defence Society (4), and in a very recent Royal
Society publication (5), which starts with the statement:
'Humans have benefited immensely from scientific research involving
animals, with virtually every medical achievement in the past century
reliant on the use of animals in some way.'
Pound et al. used a Medline search to identify 277 reviews of animal
experiments but they chose to describe just six systematic reviews
conducted to discover whether animal research had informed particular
clinical studies. One pointed out that there is no simple animal analogue
of the established relationship between social status and coronary heart
disease in humans. This is hardly surprising in view of the complexities
of human society, which have no clear parallel in animal hierarchies. The
other five papers all described clinical trials that had apparently been
started without full analysis of prior animal studies or even in parallel
with animal work. In each case the putative therapy turned out to have no
benefit and subsequent systematic review showed that animal research
revealed exactly the same problems. Far from providing evidence that
animal research doesn't work, these studies revealed excellent agreement
between animal results and clinical experience.
It is, of course, imperative that animal research is properly
evaluated before the results are transferred to medical practice. The
authors have identified some clinical trials of ineffective treatments
that were based on inadequate analysis of results from animal research.
One would have expected the relevant Ethics Committees and the regulatory
authorities to have identified this problem. Results on animals can show
when it is not appropriate to move to human trials, as much as when it is.
If these animal studies had been evaluated properly, it is likely that
they would not have proceeded to clinical trials. It is not the animal
studies that were at fault but the interpretation of them.
In the UK, animals can be used in research only where it is strictly
necessary, and the minimum number of animals must be used to answer the
question. This requires good experimental design; an area in which the MRC
is leading through high-quality peer review and our Centre for Best
Practice for Animals in Research. The vast majority of the animal research
funded by the MRC is not trials of new drugs but is aimed at studying
normal bodily function and disease processes. By their own admission,
Pound et al. did not even consider the important research involving
animals that makes a scientific contribution, as opposed to animal-based
trials of putative treatments. This is a curious distinction to make: it
is not meaningful and it ignores all the research on normal life processes
and the natural history of disease, not to mention safety-testing. All
these make essential contributions to the development of new therapies for
humans (and animals), and much of this work is required by law.
Some of the authors of this paper and their supporting organization
(Scientists for Accountability in Biomedical Research and Education) have
called publicly for a 'moratorium' on animal research (6). This is
irresponsible and is totally unjustified by the results that they
describe.
Colin Blakemore (Chief Executive) and Tony Peatfield (Head of
Policy), Medical Research Council
1. Pound P, Ebrahim S, Sandercock P, Bracken MB, Roberts I. Where is
the evidence that animal research benefits humans? BMJ 2004;328:514-7.(28
February.)
2. Medical Research Council 'Mice and Medicine' (2000):
www.mrc.ac.uk/pdf-mice_and_medicine.pdf
3. House of Lords Select Committee on Animals in Scientific
Procedures (2002). www.publications.parliament.uk/pa/ld/ldanimal.htm
4. Research Defence Society: www.rds-online.org.uk
5. Royal Society (2004): The use of non-human animals in research: a
guide for scientists.
www.royalsoc.ac.uk/framer.asp?page=/policy/AnimalsResearch.htm
6. Scientists for Accountability in Biomedical Research and
Education. Press Release, 27 February. www.S-A-B-R-E.org
Competing interests:
None declared
Competing interests: No competing interests
I have many reservations about the recent article by Pound et al.
Let me start with the title. Not ALL animal research can be summarised or
represented, adequately, by the six studies this article has considered.
However, the title certainly appears to make that claim. The subtitle
“Much animal research into potential treatments for humans is wasted
because it is poorly conducted and not evaluated through systematic
reviews” is similarly unjustifiable. Do they consider the total of animal
research summarised in their examples to be such a high proportion of all
animal research, that it could be correctly labelled as “Much...”? I
would have thought “Some..” was a more appropriate quantifier in this
context.
Without question, there is poorly conducted animal research (and many
of us as editors of Pharmacology journals do our best to keep it from
appearing in print) but I am not convinced that there is any more of this
quality of research with animals than, say, with cell cultures or with
patients. I would imagine the BMJ and most other clinical Journals have
seen and rejected many poorly conducted clinical studies which have indeed
been a waste of money, time, effort, and opportunity.
One place to look for systematic evaluations of animal research is in
the processes of testing and validating a new drug. In order to permit
clinical trials of a new compound, licensing authorities require evidence
of a considerable amount of animal research in more than one model and
species to show both pharmacological activity and lack of toxicity. ALL
this evidence IS (one hopes) critically evaluated by the licensing body
before clinical studies can begin. As Andre MacLean has already pointed
out, this evidence and its assessment is not published but that is, I
believe, a matter of “commercial confidentiality” and not evidence for
absence of evaluation.
The first line of Pound et al’s Conclusion asserts - “The
contribution of animal studies to clinical medicine requires urgent formal
evaluation”. This assertion misses the point. The problem highlighted by
their examples is not that the animal research was “valueless” but that it
was either not done at the right time or it was disregarded or selectively
presented. In four of the six examples quoted, the animal studies agreed
with the clinical results. So there would have been a good correlation
between animal and clinical research and the clinical trials should not
have been started, if the results from animal studies had been known.
Surely the criticism should be aimed more at the reasons for the clinical
trial, not at what the animal results showed. The problem of selective
referencing (example 5) is not an example of animal research giving the
wrong signals, it is an example of the Nelson Syndrome –choosing which
signals to see.
I should add here that some of the faults ascribed to the animal
models are inherent in what can or should be done with animals. Animal
models cannot reproduce all the features of human disease; some models are
closer than others but all are approximations. For instance, the long-
term consequences of stroke (case 1) are difficult if not impossible to
model in animals and the ischaemia, infection and necrotic debris in
wounds (case 3) appear to me to be unethical conditions for animals.
Finally, it seems to me that the real message of the article is not
about the value of animal research but about the design of, and
justification for, clinical trials. Surely, before a clinical trial is
started, all relevant existing research IS critically examined, regardless
of its origins, animal or human. If the trial is supported by non-
industry funds (MRC / Wellcome Trust / charities, etc) that critical
examination is what the grant reviewers perform. One possible reason for
the relative lack of funding of clinical research might be that few such
proposals are of the high quality the reviewers demand.
Nowhere in this article is there adequate justification of the
wholesale rejection of animal research in providing benefit for humans, as
implied in this article. The evidence that animal research IS a benefit
to humans is like Christopher Wren’s monument –it is all around us. Many
of us are benefiting now from, say, ACE inhibitors for hypertension,
selective beta-agonists for asthma, neuromuscular blockers and
anaesthetics for surgery or statins for lowering cholesterol. All will
have been introduced as a result of animal research, leading to assessment
in humans.
I would support the authors in asking for better designed clinical
research but this is not a reflection on the value of the preceding animal
research and its results but of what those who undertake clinical research
do with the results from animal research.
Competing interests:
I held a Home Office Licence for more than 30 years.
Competing interests: No competing interests
Here are two related and recetn publications. I post them in hopes that they might help advance the ethical and scientific debates here:
Review of Why Animal Experimentation Matters: The Use of Animals in Medical Research, American Journal of Bioethics, Vol. 3, No. 1, online Bioethics Education Network.
"Carl Cohen's 'Kind' Argument FOR Animal Rights and AGAINST Human Rights," Journal of Applied Philosophy, March 2004
Competing interests:
None declared
Competing interests: No competing interests
The article by Pound and colleagues raises many important issues that
need to be discussed. Of course animal experiments should only be
undertaken if they address an important question, if there is no other
reasonable way to answer the question, if the experimental design is
correct, if the number of animals to be used is the minimum necessary to
provide adequate statistical power, if suffering is kept to an absolute
minimum and if the study isn't an unecessary repeat of earlier work. I
suspect they are right that there is room for improvement in many of these
areas. However where I disagree strongly with the article is in the
inference drawn from their brief survey.
In the introduction they attempt to distinguish between "clinical"
and "scientific" contribtuions. Indeed the article is largely based on
observations on the potential of effects in an animal model to predict
therapeutic effect in patients. This is an interesting area that merits
discussion. However the title is "Where is the evidence that animal
research benefits humans?". This is a much broader question and not
tackled at all in the article. Jim Black explored the effects of
adrenaline and potential antagonists on the beating rates of animal
hearts. Thus emerged beta blockers. John Vane explored the release of
inflammatory mediators from tissues and identified the mechanism of action
of aspirin and a whole class of anti-inflammatory agents, as well as
identifying potential cardioprotective effects of aspirin. Presumably the
authors would accept that these drugs have benefitted humans? In such a
sensitive area as animal research in which articles such as this are bound
to be used in support of political and campaigning aims, it is essential
that scientists and journals report their findings accurately. A more
appropriate title would have been "A preliminary survey of the ability of
animal models to predict clinical outcomes" and the discussion should have
reflected this.
Competing interests:
I undertake experiments using animals. The drugs I use in patients have all been tested on animals.
Competing interests: No competing interests
Pound and colleagues(Feb 28th p514) ask the question whether there
is evidence that animal research benefits humans. They try to answer their
question by looking for published systematic reviews of animal and
clinical research on the same topic, to see if these two approaches give
the same answer. They find plenty of poorly designed and executed
experiments, and conclude that animal experiments are useless.That is not
surprising,they were looking in the wrong place.
First, most clinical research cannot even be thought of without the
previous accumulated understanding of how the body works and how pathology
develops, and how much of biochemistry is similar between species. All
that knowledge is as basic to research as learning to read or understand
numbers, but there are no systematic reviews of literacy as an essential
for research.
Secondly much animal work is concerned with deciding whether it is
reasonably safe for a study to go ahead in humans , after an idea has
developed from basic science. Much of the animal safety data is in
pharmaceutical industry files, and never published,and that is a separate
scandal. I doubt that even the authors would be prepared to try, say , a
new anti-malarial drug in patients , without some animal data to give a
reasonable expectation that the patients would not drop dead, or the new
drug stay in the body for ever. Unfortunately there are no reliable in
vitro methods to predict overall toxicity, and even animals tell us little
about chronic toxic effects, so we are left with imperfect safety
assessments based on animal studies, which are the best we have to tell us
if it is reasonable to proceed to human studies.
Let me cite two examples from my own field of interest . There are
observations that rats fed a diet deficient in vitamin E and Selenium
develop fatal liver necrosis . No such disease exists in humans, but Kesan
heart disease found in selenium deficient areas of China killed thousands
of children,over many years. The idea that selenium was an essential
element , and deficiency could cause cell damage led to the large scale
trials of selenium supplementation , and successful prevention of the
disease. It shows that the link is more complex than Pound et al
imagine.It is a vital link but not so direct as to be measurable by the
simple methods that they use.
In the treatment of paracetamol overdose the effective methods are
giving SH compounds that are glutathione precursors. No one could have
guessed this from epidemiology, or bedside observation. It needed a long
line of animal experiments on essential amino acids, and the glutathione
depleting effects of a number of poisons.Later on we needed epidemiology
to show that treatment with methionine or n-acetyl cysteine worked in
the first 8 – 12 hours. Now, to develop treatment for the later stages ,
we need to go back to experiments using cells , receptors, molecular
biology, and back to mice and then people again.
The conclusions that I draw from their paper is that poorly designed
experiments give no useful information, bad science does not produce good
medicine , and that in the long run we benefit from the ideas about how
molecules may be organised to form people, and illnesses, or poetry .
Competing interests:
None declared
Competing interests: No competing interests
Penman is not mistaken about the aetiology of stomach ulceration, but
did he read (or understand) Fiddian-Green's response?
It seems to me that F-G used his understanding of the aetiology of
peptic ulceration to develop (with the help of animals) a test for 2
seperate conditions. They were: bleeding from an ulcer (distinct from just
ulceration); or bowel damage folowing vascular surgery.
He then alludes to the needless repetition of animal experiments. I
think many doctors would disapprove of unnecessary animal suffering,
accept the limitations of animal studies, and hope to work towards other
experimental methods. It isn't black or white however. Animal experiments
have led to the alleviation of much human suffering.
Perhaps the gulf of mutual incomprehension between the medical
establishment and the journalistic movement is as wide as ever.
Competing interests:
None declared
Competing interests: No competing interests
Pound et al ask, "Where is the evidence that animal research benefits
humans?" The authors need look no further than the fields of perinatology
and neonatology for an answer to this provocative question. Two
therapies, prenatal treatment with glucocorticoids to improve neonatal
pulmonary function and the use of surfactant to treat respiratory distress
syndrome (RDS), both solidly based on animal research, have markedly
improved survival and decreased morbidity of premature babies.
Dr. Avery has elegantly summarized the history of antenatal therapy
with glucocorticoids (1). This story began with the seminal observation
by Liggins that after prenatal treatment with glucocorticoids preterm
lambs exhibited precocious aeration of the lungs. It evolved through
further animal experiments that delineated the effects of steroids on
synthesis of pulmonary surfactant. Then large controlled human studies
documented similar effects in preterm babies. The risk of RDS decreased
(Odds Ratio [OR] <0.4) and mortality decreased (OR <0.7). These
studies further demonstrated that prematal steroids substantially reduced
the risk of severe complications of prematurity: intraventricular
intracranial hemorrhage, a cause of death or severe brain injury) (OR
0.4), and necrotizing enterocolitis., a cause of death and long-term
morbidity) (0R <0.4) (2). An NIH concensus conference later that year
endorsed prenatal treatment with steroids. No adverse long-term effects
are known when steroids are employed as a single course of treatment in
women with threatened preterm delivery. This use of steroids has become
"standard of care" worldwide.
The prophylactic and therapeutic use of surfactant for RDS has a
similar history and denouement. The seminal discovery that lungs of
infants dying with RDS are deficient in surfactant (3)led to animal
research that described the ontogeny of surfactant synthesis in the fetus,
defined the role of pulmonary surfactant in treating experimental RDS, and
led to development of exogenous surfactant preparations that could be used
to treat RDS. Multiple prospective controlled trials of surfactant
therapy in premature babies ensued. They demonstrated that treatment is
effective in reducing mortality from RDS and is devoid of long-term
complications (4). Exogenous surfactant preparations are in use worldwide
to prevent or treat RDS in preterm infants.
Advancing "from the bench to the bedside to the pharmacy" (5), these
two therapies provide a paradigm of the applicability of animal research
to the improvement of human health. They provide an unequivocal answer to
the question posed by Pound et al.
Many other problems of the newborn infant remain to be addressed.
These questions demand investigation in appropriate animal models to
provide a better understanding of mechanisms of disease, studies which are
underway in centers around the world. These investigations in turn should
generate therapeutic concepts that require animal testing before advancing
to human trials. For example, ongoing animal studies in this institution
are addressing the pathophysiology of bronchopulmonary dysplasia and the
mechanisms of respirator-induced lung injury in premature animals. The
goal of these studies is to identify potential therapies to prevent
chronic lung disease in premature babies and to improve respirator
management of these fragile patients.
In no way do I intend to suggest that inadequately designed and
poorly conducted research in animals is justifiable or useful. Pound et
al raise important issues about the quality and utility of some animal
studies. However, it is obvious that all research in animals cannot be
tarred with the same brush. Important, significant improvements in human
health and well being have resulted and continue to evolve from
appropriately conducted research in animals.
1. Avery ME. Historical overview of antenatal steroid use.
Pediatrics 1995;95:133-135.
2. Crowley PA. Antenatal corticosteroid therapy: A meta-analysis of
the randomized trials, 1972-1994. Am J Obstet Gynecol 1995;173:322.
3. Avery ME, Mead J. Surface properties in relation to atelectasis
and hyaline membrane disease. Am J Dis Child 1959;97:517.
4. Jobe A. Surfactant treatment in Fetal and Neonatal Physiology,
3rd ed, (Polin RA, Fox WW, Abman SH, eds), Saunders, Philadelphia, 2004,
p1074.
5. Avery ME. Surfactant: Historical perspective in Fetal and
Neonatal Physiology, 3rd ed, (Polin RA, Fox WW, Abman SH, eds), Saunders,
Philadelphia, 2004, p1003.
Competing interests:
None declared
Competing interests: No competing interests
Rather than "[shedding] the most light on the contribution that
animal research makes to clinical medicine" most of the systematic reviews
of animal models in this article were prompted by unsuccesful clinical
trials in humans. The remainder appear to have been prompted directly by
inconsistency between animal models and clinical data. This suggests a
publication bias in systematic reviews of animal models (under this search
stategy) which renders the conclusions reached in the paper invalid. It
seems intuitively plausible that animal models in those fields which lead
to unsuccesful clinical trials or inconsistency are of poor quality but
the findings cannot generalise to all animal models.
Reviews following unsuccesful clinical trials:
Horn et al 2001; Lucas et al 2002; Roberts et al 2002, Mapstone et al
2003; Lee et al (2003).
The unpublished Ciccone & Candelise review followed "the finding
that clinical trials of thrombolysis for acute stroke had found a
substantial excess risk of intracranial haemorrhage that had not been
predicted by individual animal studies".
The Petticrew & Davel Smith paper is also a review highlighting
the inconsistency of primate models of stress and coronary heart disease
with human epidemiology.
Competing interests:
None declared
Competing interests: No competing interests
Human holocaust - Animal holocaust
The paper by Pound, Ebrahim, Sandercock, Bracken & Roberts is a
welcome and timely development apropos the controversial issue of animal
experimentation and its contribution to human health.
Defenders of animal experiments perpetuate the view that there has
hardly ever been a cure for any disease which was not due to animal
research and that medical progress would be threatened by a
moratorium.Since it has been a tradition of western medical science to
use animals widely for over a century and a half, it is easy for
proponents to claim the use of animals has been vital to progress. But
this by no means proves that the animal experiments themselves were the
real key to the most important discoveries.Animals have been killed in
their billions in medical research, but this neither automatically proves
that the experiments were vital or irreplaceable,nor that medical progress
will be severely hampered by their abandonment.
Not so well publicised are the historical and long-term consequences
of this legislated protocol that secures funding, tenure or promotion and
protects drug companies rather than human beings. Thalidomide is but one
high profile disaster.Medical history is strewn with hazardous medications
and human fatalities traceable to drug development's dependency on the
animal model.
Diethylstilbestrol[1][2][3]and Zimeldine [4][5][6]were
predicted safe by animal tests but resulted in severe adverse effects in
humans. Benzene was not withdrawn from use as an industrial chemical
despite clinical and epidemiological evidence that exposure caused
leukemia in
humans because manufacturer-supported tests failed to reproduce leukemia
in mice.[7]
Animal testing can also delay progress and lead to the
rejection of medicines which are potentially valuable for treating human
illnesses.The plant digitalis has been used for centuries to treat heart
disorders but clinical trials of the digitalis-derived drug were delayed
because
it caused high blood pressure in animals. Fortunately, human evidence
overrode and as a result,digoxin, an analogue of digitalis, has saved
countless lives. Many more people could have survived had the animal
testing
been ignored and digitalis been released earlier.[8][9][10][11]More
recently,Multiple Sclerosis sufferers must have been pleased to learn that
scientists studying human brain tissue may have advanced MS research along
a different path from misleading earlier animal studies on which all
subsequent research has been based.[12]
What explanation and words of comfort would proponents of animal
experiments offer victims of the disastrous effects, delays in progress
and misleading results of animal based medicine?
Put enough money in the slot machine and you will occasionally
triumph but this does not make it either a reliable or logical method of
pursuing your goal.Who can be certain whether or not as much or more
useful knowledge
could have been obtained from other sources if medical science had been
force to take an alternative direction? It is impossible to unravel
every medical discovery in order to measure precisely the part paid by
animal experiments but the doom predicted by research scientists if a ban
were to be implemented is irrational.
Animal experiments have exhausted resources that could have been
dedicated to educating the public about health hazards and health
maintenance, thereby diminishing the incidence of diseases that require
treatment.
Proponents of vivisection claim the welfare of lab animals in the UK
is strictly regulated, but documentary evidence and the recent legal
victory over Imutran/Novartis by Uncaged campaigns proves otherwise.The
lack of effective
regulatory oversight provided by the Animal Procedures Committee was a
factor in Uncaged's successful legal argument for the public interest in
disclosure of the documentation.In addition to the barbaric cruelty, the
documents reveal a shameful record
of broken regulations, exaggerated claims and haphazard science. The whole
research project was eventually
abandoned in 2000 as a total failure.[13]
[1]NEJM, 333;1099-1105, 1995.
[2]J. NIH Res., 1993, 5, pp33-35.
[3]Nature, 1993, July 22, p275.
[4]B. Blackwell in Side Effects of Drugs Annual, vol. 8, eds M. N. G.
Dukesand J. Elis (Elsevier, 1984).
[5]R. D. Mann, Modern Drug Use; An Inquiry on Historical Principles (MTP
Press, 1984).
[6]R. C. Heel, et al, Drugs, 1982, vol. 24, pp.169- 206.
[7]Lancet, June 25 1977, pp1348-9.
[8]W. Sneader, Drug Discovery: The Evolution of Modern Medicine,
Wiley,1985.
[9]T. Lewis, Clinical Science, Shaw and Sons Ltd., 1934.
[10]Federation Proceedings, 1967, vol.26, pp1125-30.
[11]Toxicology In Vitro, 1992, vol.6, pp.47-52.
[12]New Scientist (26 February 2004)
[13]Uncaged campaigns:www.xenodiaries.org
Competing interests:
None declared
Competing interests: No competing interests