Philosophy of AR > Animal Testing Index > Anti-vivisection Index
101 Misleading Vivisection Results

101 Misleading Vivisection Results - printer friendly


101 documented examples of failures, fatalities, misleading results, mistakes and missed opportunities which have occurred through our acceptance of vivisection. By Dr Robert Sharpe. Scientific Director of International Association Against Painful Experiments on Animals

Dr Robert Sharpe was a Senior Research Chemist at the prestigious Royal Postgraduate Medical School in London when he found himself at odds with colleagues who tested his chemicals on animals. He resigned his position and began to investigate, what until then he had taken for granted, that animal experiments were vital for medical progress.

The facts proved disturbing and posed fundamental questions about the use of animals. For how could a method that produced such conflicting results be so vital to our health?

Dr Robert Sharpe presents a powerful body of evidence documenting the failures, misleading results and missed opportunities of animal research. Is our acceptance of vivisection the biggest medical error in human history?

The suffering is real, and so it seems are the mistakes.


The life-saving qualities of a new anti-rejection drug, FK506, could have been missed when animal experiments suggested it was too toxic for human use.1 The tests were carried out at Cambridge University in England and showed that "...animal toxicity was too severe to proceed to clinical trial".2 US researchers, however, decided it was worthy of further investigation but nevertheless did not feel justified in first giving the drug to healthy volunteers, the usual practice in drug development, since this could be "potentially dangerous."3 Instead, FK506 was administered as a last chance option to liver transplant patients in "desperate plight". So far clinical experience with FK506 has been very promising.4

Animal tests also proved misleading in suggesting that FK506 would give better results if combined with another antirejection drug, cyclosporin. However, clinical trials revealed the opposite, with FK506 actually increasing the kidney damage caused by cyclosporin.3


I ) R Allison, Journal of the American Medical Association, 1990. April 4, 1766.

2) R.Y.Calne et al, Lancet. 1989, July 22, 227.

3) T.E.Starzl et al, Lancet, 1989, October 28, 1000-1004.

4) J Neuberger, Hepatology, 1991, vol.13, 1259-1260.


In 1983 a television documentary programme drew attention to an increased number of childhood leukemia cases in the vicinity of the nuclear reprocessing plant at Sellafield in Britain. Although the incidence of leukemia was 10 times the national average, the official  Committee of Inquiry decided the nuclear facility was not the cause.

Their conclusions were based on calculations from animal experiments. By preferring animal data to direct human observations, the effect was to minimise the risks of radiation.1 Subsequently, a major investigation concluded that radiation was indeed to blame, for those at highest risk of leukemia were born to fathers who worked at the nuclear plant.2 Not all studies supported these findings and clarification must await further epidemiological research. Nevertheless, the observations linking leukemia clusters to nuclear plants did persuade the Ministry of Defence and the government's Health and Safety Executive to recommend major cuts in the maximum radiation doses to which workers are legally exposed.3


1) E.Millstone in Animal Experimentation: The Consensus Changes, Ed. G.Langley (MacMillan, 1989)

2) M.J.Gardner et al, British Medical Journal, 1990, February 17, 423-429.

3) The Guardian, 1991, March 22 and April 30.


The British National Formulary (1993) warns that methysergide, a treatment for migraine, should only be administered under hospital supervision because of dangerous side-effects resulting from abnormal formation of fibrous tissue. This condition, known medically  as retroperitoneal fibrosis, can lead to obstruction of abdominal blood vessels and blockage of the tube carrying urine from the kidneys to the bladder. Fibrotic damage to the heart valves has also been reported and can result in heart failure.

Methysergide's life-threatening side-effects were not predicted by animal tests,1 nor could they be induced during subsequent experimentation, and a report in the British Medical Journal notes that "Attempts to reproduce these fibrotic lesions in animals have been unsuccessful."2


I ) R.Heywood in Animal Toxicity Studies: Their Relevance for Man, Eds. C.E. Lumley & S.R.Walker (Quay Publications, 1990).

2) K.A.Misch, British Medical Journal, 1974, May 18, 365-366.


The arthritis drug suprofen (Suprol) was withdrawn worldwide in May 1987 following reports of kidney problems and pain in the side of the body.1 Patients experiencing these side-effects had to have their kidney function monitored for 2 years after they stopped taking the drug.2 The dangers were unexpected because "In animal studies suprofen has been shown to have an excellent safety profile. No significant effects were observed on cardiac, renal [kidney] or central nervous system parameters in several species."3


1) Drug Withdrawal from Sale, C.Spriet-Pourra & M.Auriche (PJB Publications, 1988)

2) FDA Drug Review: Postapproval Risks 1976-1985 (US General Accounting Office, April 1990).

3) A.Yeadon et al, Pharmacology, 1983, vol.27, Suppl.l, 87-94.



For centuries, alcohol has been regarded as poisonous for the liver.1

That is, until the first half of the 20th century when it was cleared of

liver toxicity following experiments on animals.1,2 In 1934, a summary

of animal tests concluded that "experimental evidence has not

substantiated the belief that alcohol is a direct cause of cirrhosis."3

Based largely on experiments with rats, researchers later argued that "there is no more evidence of a specific toxic effect of pure ethyl alcohol upon liver cells than there is for one due to sugar."4 Today, alcohol is once again considered a liver toxin but since it has proved so difficult to induce cirrhosis in laboratory animals, there are still some who doubt the evidence.5

Animal experiments have proved misleading in other areas of alcohol research. Although it has been known for decades that too much alcohol can cause cancer, this well established clinical fact has been questioned because it proved impossible to induce the disease in animals. Indeed, some insist that alcohol should not be classified as a human carcinogen since there is no evidence from animal experiments!6

Alcohol seems more toxic to the circulatory system of humans than animals, and whereas prolonged consumption raises the blood pressure in alcoholics, this is not usually the case in rats.7 And whilst alcohol can damage the human heart, "Studies on a variety of animals being given large amounts of ethanol (alcohol) over long periods of time did not lead to heart failure. Also, until recently when the heart of the Nicholas turkey was shown to be susceptible to alcoholic damage, there has been no animal model of alcoholic cardiomyopathy (heart muscle damaged) as it is seen in man."7

During the early 1970s researchers described how alcohol could induce physical dependence in mice. The experiments showed that the tranquillizing drug Librium could reduce the severity of withdrawal convulsions, but also suggested that the treatment had a lethal sideeffect with some of the animals dying.8 Fortunately, clinical studies carried out 6 years earlier had already shown that Librium was effective 9 and the drug remains an important treatment for alcohol withdrawal symptoms.

Despite the known effects of alcohol and the availability of human tissues to supplement clinical observations, there seems no shortage of funds for animal experiments. A report by the National Research Information Centre, compiled by Murry Cohen MD and Constance Young, revealed that the US Government funded 284 alcohol research projects involving animals during 1986, costing nearly $24 million. 10 "Animal research", they concluded, "has had no significant effect on our knowledge of alcohol-use disorders."


1) H.J.Zimmerman, Alcoholism: Clinical & Experimental Research, 1986, vol.l0, 3-15.

2) C.S Lieber & L.M.DeCarli, Journal of Hepatology 1991, vol 12, 394-401.

3) V.H.Moon, Archives of Pathology, 1934, vol.l8, 381-424.

4) Reported in Ref 2.

5) R.F.Derr et al, Journal of Hepatology, 1990, vol.10, 381-386.

6) L.Tomatis et al, Japanese Journal of Cancer Research, 1989, vol.80, 795-807.

7) J.V.Jones et al, Journal of Hypertension, 1988, vol.6, 419-422.

8) D.B.Goldstein, Journal of Pharmacology & Experimental Therapeutics, 1972, vol 183. 14-22.

9) G.Sereny & H.Kalant, British Medical.Journal, 1965, January 9, 92-97.

10) M.Cohen & C.Young, Alcoholic Rats, The National Research Information Centre, 1989.


In February 1986 the British Parliament's Agriculture Committee

began an enquiry into pesticides and human health. The Committee

learnt that great reliance is placed on animal experiments but that

"...similar tests in different animal species often yield quite different

results".1 An example is the organophosphate pesticide dipterex

which produces nerve damage in people but not in the animal tests

specially designed to detect such injuries.2 In fact, Dr Murray of the

National Poisons Unit informed the Committee that one well

documented case of human poisoning is equivalent to 20,000 animal


The Committee concluded that "It cannot be satisfactory to rely on

animals so much as a means of testing and, as other forms of testing

become available, we recommend that they be adopted...we are

satisflied from the evidence that we have received that animal testing

can produce misleading results."1


1) Special Report of the House of Commons Agriculture Committee, reproduced in FRAME

News, 1987, No.16,p.2.

2) A.N.Worden in Animals and Alternatives in Toxicity Testing, Eds. M.Balls et al (Academic

Press, 1983).


It was fortunate that so much human evidence linked arsenic to

cancer because for over 70 years, researchers were unable to

"confirm" the dangers in laboratory animals. Suspicions that arsenic

might cause cancer date back to 1809 when its harmful effects in

drinking water were first noted.1 In 1887/88, Sir Jonathon Hutchinson

described the earliest cases of cancer resulting from medicinal use

of arsenic1 and subsequently, others have reported cancers in

chemical, agricultural and metallurgical workers exposed to arsenic.2

Animal tests began in 1911 and an historical analysis of the subject,

published during 1947, described how dozens of experiments had

been performed.1 However, these had given "only doubtful results."

The tests continued but still proved negative, and in 1969

researchers at America's National Cancer Institute stated that

"arsenic has been suspected by many investigators as a carcinogen

in man, though there is no supporting evidence from animal

experiments."3 And in 1977 a further summary of the data concluded

that "there is little evidence that arsenic compounds are carcinogenic

in experimental animals."2

Finally, in the late 1980s, scientists managed to produce cancer in

animals. This was 180 years after arsenic was first suggested as a

human carcinogen. Despite decades of failure, animal researchers

had at least been correct about one thing: in 1962 Heuper and Payne

wrote that "With perseverance and some luck arsenicals one day

may be shown to cause cancer in animals."4


I ) O.Neubauer, British Journal of Cancer, 1947, vol 1. 192-251.

2) F.W.Sunderman Jr. in Advances in Modern Toxicology, vol.2, Eds. R.A.Goyer &

M.A.Mehlman (Wiley, 1977).

3) A.M.Lee & J.F.Fraumeni Jr. Journal of the National Cancer Institute, 1969, vol.42, 1045-1052.

4) W.C.Heuper & W.W.Payne, Archives of Environmental Health, 1962, vol.5, 459


Benzene is used as a starting point for the production of industrial

chemicals and for the manufacture of detergents, explosives and

pharmaceuticals. It is also present in gasoline and was once

commonly employed as a chemical solvent. Because benzene is so

widely used, there has been considerable debate over the safety of

exposed workers, especially since experience has shown it to be a

cancer hazard.

Tragically, human evidence was once again undermined by the

animal laboratory. According to Lester Lave of the Brookings Institute

in Washington DC, "although there are reliable human data linking

benzene to leukemia, scientists have been reluctant to categorise

benzene as a carcinogen because there are no published reports

that it induces leukemia in rodents."1

In fact, 14 separate animal trials, starting in 1932, failed to show that

benzene caused cancer.2 Only during the late 1980s were

researchers finally able to induce cancer in laboratory animals by

dosing them with benzene.


1) L.B.Lave, The American Statistician, 1982, vol 36, 260-261.

2) D.M.De Marini et al, in Benchmarks: Alternative Methods in Toxicology, Ed.

M.A.Mehlman (Princeton Scientific Publishing Co.lnc.,1989).



Chymotrypsin is widely used in ophthalmic surgery for the treatment

of cataract. Although recommended for human use,1 chymotrypsin is

harmful to the rabbit eye. In his book Toxicology of the Eye (1974),

Morton Grant states that "the rabbit cornea appears to differ

significantly from the human cornea in its reaction to � -chymotrypsin.

It has been noted repeatedly that introduction of � -chymotrypsin into

the (rabbit's) corneal stroma...leads to severe swelling reaction of

cornea, much more than is seen in human beings, and in some

instances leading to perforation of the cornea."


1) British National Formulary, No.26(BMA and The Royal Pharmaceutical Society of G.B., 1993).


Lindane is probably best known as an agricultural insecticide but

very dilute lotions, creams and shampoos are used therapeutically

for treating lice and scabies. Nevertheless such preparations can

cause "excessive eye irritation" and conjunctivitis,1 and the British

National Formulary (1993) warns users to "avoid contact with eyes".

In rabbits, however, application of a far more concentrated solution

produced only minimal effects.1 Furthermore, exposure to lindane in

the form of a dust proved non-irritating to the eyes and nasal mucosa

of rabbits but caused irritation to the eyes and respiratory passages

of sensitive people.1


1) W.M.Grant, Toxicology of the Eye, 2nd edition (Charles Thomas, 1974).


In 1933, following "thorough" experiments on animals, researchers

described the use of dinitrophenol as a treatment for obesity.

However, doctors soon noticed that the drug unexpectedly caused

cataracts in some of their patients, and nearly 200 cases were

reported before the drug was prohibited for internal use. Attempts

were made to replicate the clinical findings in rats, rabbits, guinea

pigs and dogs but none of the experiments produced any change in

the lens of the eye.1 In 1942, a summary of the tests stated that "All

attempts to produce experimental cataracts in laboratory animals by

various and repeated doses of dinitrophenol have been


Although birds are rarely used for the safety testing of drugs, later

experiments accidentally discovered that cataracts could be induced

in chicks dosed with dinitrophenol in their food.1

Similar problems were encountered with triparanol (Mer-29), a drug

used to lower cholesterol levels. The cataracts observed in human

patients could be induced in rats and dogs after very high doses but

not in rabbits and monkeys.3 Triparanol was withdrawn in 1962.


1) B.H.Robbins, Journal of Pharmacology, 1944, vol.80, 264-269.

2) Reproduced in ref. 1.

3) W.M.Grant, Toxicology of the Eye,2nd edition (Charles Thomas,1974).


For years, high doses of corticosteroids have been recommended for

the treatment of septic shock, a condition which leads to heart,

kidney and respiratory failure in a high proportion of patients. The

idea was based on animal experiments where corticosteroids

improve survival when given before shock 1 or shortly afterwards.2

It has been pointed out, however, that "...extrapolation of data from

experimental models of shock to the clinical setting may be

dangerous and misleading."3 So it is not surprising that an analysis

of clinical trials by the Drug & Therapeutics Bulletin found that "highdose

corticosteroids are ineffective for the prevention or treatment of

shock associated with sepsis. They do not improve outcome, and

make secondary infection worse. They may harm patients with

impaired renal (kidney) function."1 For instance, one trial found that

corticosteroids not only failed to prevent or reverse shock but actually

seemed to increase deaths amongst patients, even though treatment

was initiated within 2 hours.4


I ) Drug & Therapeutics Bulletin, 1990, vol.28, 74-75.

2) S.G.Hershey in Anaesthesiology: Proceedings of the Vl World Congress of

Anaesthesiology, Mexico City, April 1976, Eds. E.Hulsz et al (Excerpta Medica, 1977).

3) A.S.Nies in Clinical Pharrnacology: Basic Principles in Therapeutics, Eds. K.L.Melmon &

H.F.Morrelli (MacMillan,1978)

4) R.C.Bone et al, New England Journal of Medicine, 1987, September 10, 653-658.



An important area of medical research is pharmacology where

scientists study exactly how drugs and natural body substanccs exert

their effects on the tissues. An understanding of the chemical

processes involved can be valuable in providing a more rational

basis for the design of new treatments. Unfortunately, many

pharmacologists rely on animals despite numerous contradictory

results. As a result of experiments with dogs, acetylcholine, a

chemical produced by nerve endings, was widely believed to dilate

the coronary arteries. But in human coronary tissue it causes a

narrowing of the vessels which is thought to lead to heart spasm in a

living person.1 Another body chemical, bradykinin, relaxes blood

vessels in human brain tissue but contracts them in dogs.2

Further species differences have been found with the leukotrienes

(LT), natural substances involved in inflammation. Leukotrienes

known as LTC4 and LTD4 constrict blood vessels in the guinea pig's

skin but dilate corresponding tissues from people and pigs.3 Yet

another case is the prostaglandins (PG), a family of substances

discovered over 50 years ago in human seminal fluid: in heart tissue

from cats and rabbits, PGE1 has no effect on contractile force or

heart rate but increases them in rats, guinea pigs and chickens.4

Some pharmacologists have recognised that "direct extrapolation

from animals to humans is frequently invalid," so that "recently much

interest has focussed on use of human autopsy or biopsy tissue as

a means of overcoming these limitations."5


1) S.Kalsner, Journal of Physiology, 1985, vol.358, 509-526.

2) K.Schror & R.Verheggen, Trends in Pharmacological Sciences, 1988, vol 9, 71-74.

3) P.J.Piper et al, Annals of the New York Academy of Sciences, 1988, vol.524, 133-141.

4) S.Bergstrom et al, Pharmacological Reviews, 1968, vol.20, 1-48.

5) Trends in Pharmacological Sciences, 1987, vol.8, 289-290.



During the l960s animal experiments suggested that clonidine might

be a useful drug for preventing migraine.1 Using cats, experimenters

found that clonidine interfered with physiological processes thought

to cause headaches. The drug was introduced in 1969 but clinical

experience now suggests that clonidine is largely ineffective and little

better than a dummy pill.1

Clonidine proved more successful in the treatment of high blood

pressure. Its ability to lower blood pressure was discovered

accidentally when it was given to people as a nasal decongestant.2

Although effective, there were serious unexpected side-effects when

patients stopped taking the drug: the "clonidine withdrawal

syndrome" is characterised in extreme cases by sweating, trembling,

rapid heart beat and a dangerous rise in blood pressure.The

symptoms may occur after only one or two missed doses or even

after gradual withdrawal over 3 days.

Attempts to replicate the condition in dogs and cats produced

inconsistent results 3 whilst in the rat "..attempts to reproduce the

clonidine discontinuation syndrome...have met with even more

difficulties and controversy than those encountered in dogs and

cats."4 'Success' was only achieved when researchers implanted a

special pump into the rat's body to maintain adequate levels of

clonidine in the bloodstream prior to withdrawal.4

In view of its serious side-effects, the Drug and Therapeutics Bulletin

considers clonidine obsolete for the treatment of high blood



1) Drug & Therapeutics Bulletin, 1990, vol.28, 79-80.

2) A.S.Nies in Clinical Pharmacology: Basic Principles in Therapeutics, 2nd edition,

Eds. K.L.Melmon & H.F.Morrelli (MacMillan, 1978).

3) L.Hansson et al, American Heart Journal, 1973, vol.85, 605-610.

4) M.J.M.C. Thoolen et al, General Pharmacology, 1981, vol.l2, 303-308.

5) Drug & Therapeutics Bulletin, 1984, vol.22, 42-43.


A major disaster occurred in the UK during the 1960s when at least

3,500 young asthma sufferers died following use of isoprenaline

aerosol inhalers.1 Fatalities were reported in countries using a

particularly concentrated form of aerosol that delivered 0.4mg of

isoprenaline per spray.2,3 Fortunately, the death rate declined rapidly

when the drug was made "prescription only" and warnings were

issued to doctors.

Attempts to replicate the effects in laboratory animals proved difficult.

In 1971 researchers at New York's Food and Drug Research

Laboratory reported that "Intensive toxicologic studies with rats,

guinea pigs, dogs and monkeys at dosage levels far in excess of

current commercial metered dose vials... have not elicited similar

adverse effects."4

Experimenters persisted in their attempts however, and eventually

found that by artificially reducing the amount of oxygen in the

animal's tissues, they could increase the toxic effects of isoprenaline

on the heart.5


1) W.H.lnman in Monitoring for Drug Safety, Ed. W.H.lnman (MTP Press, 1980).

2) P.D.Stolley, American Review of Respiratory Diseases, 1972, vol. 105, 883-890.

3) P.D. Stolley & R.Schinnar, Lancet, 1979, October 27, 896.

4) S.Carson et al, Pharmacologist, 1971, vol.l8, 272.

5) British Medical Journal, 1972, November 25, 443-444.


Since 1953 when doctors first drew attention to the kidney damage

associated with prolonged use of combination painkillers, there have

been many animal experiments to try and clarify the effects seen in

people.1 In fact, these have only obscured the issue. For example,

interest centered on which ingredient was responsible, and although

suspicion naturally fell on phenacetin since this was present in most

analgesic mixtures, the characteristic kidney damage seen in

patients could not be reproduced in animals.1

The experiments also suggested that aspirin rather than phenacetin

was to blame in painkillers containing the two drugs.2 This is

because, unlike phenacetin, aspirin readily induces kidney damage

in laboratory animals. Eventually, human studies showed that

phenacetin was indeed a major culprit.3

So contradictory were the experiments that a major analysis of the

subject concluded that if doctors had not first observed the effects in

patients, they would never have been suspected, foreseen or

predicted by animal tests.1 Phenacetin was finally withdrawn in 1980

when there were also suspicions that it caused cancer.


I) I.Rosner, CRC Critical Reviews in Toxicology, 1976, vol.4, 331-352.

2) British Medical Journal, 1970, October 17, 125-126.

3) K.G.Koutsaimanis & H.E. de Wardener, British Medical Journal, 1970, October 17, 131-134.


During the 1960s, Japan suffered a devastating epidemic of druginduced

disease associated with clioquinol, the main ingredient of

Ciba-Geigy's antidiarrhoea medicines Enterovioform and Mexaform.

At least 10,000 people, and perhaps as many as 30,000, were

victims of SMON (subacute myelo-optic neuropathy), a new disease

whose symptoms include numbness, weakness in the legs, paralysis

and eye problems, including blindness.1 In 1970 Japan's Ministry of

Health and Welfare banned the drug and 15 years later clioquinol

was withdrawn worldwide.

Clioquinol's harmful effects result from nerve damage yet animal

experiments performed by the company revealed "no evidence that

clioquinol is neurotoxic", tests being carried out on rats, cats, beagles

and rabbits.2

Although some argue that "Animal tests have consistently failed to

reproduce the effects seen in humans,"3 researchers at the Okayama

University Medical School say they have induced clioquinol toxicity in

mongrels.4 Nevertheless, they note that different species respond

differently, with monkeys, hens, cocks, and mice only mildly affected

even after higher doses. They also found that beagle dogs were 3-4

times less sensitive to clioquinol than mongrels, and concluded that

"These facts suggest strongly differences in strains as well as

species of animals for the neurotoxicity of clioquinol."


1) Lancet, 1977, March 5, 534.

2) R.Hess et al, Lancet, 1972, August 26, 424-425.

3) W.Sneader, Drug Development: From Laboratory to Clinic (Wiley)

4) J.Tateishi et al, Lancet, 1972, June 10, 1289-1290.


Careful observation of women taking the pill has shown that the most

serious side-effects are on the circulatory system: there is an

increased risk of blood clots leading to heart attacks, strokes and

lung diseases. By 1980, Britain's Committee on Safety of Medicines

(CSM) had received reports of 404 deaths.1 Further studies found

that 1-5% of women taking the pill have raised blood pressure.

None of these problems had been identified by animal experiments.2

Furthermore, in some species oral contraceptives produced the

opposite effect, making it more difficult for the blood to clot!3 As

Professor Briggs of Deakin University in Australia points out, "Many

experimental toxicity studies have been conducted on contraceptive

oestrogens, alone or in combination with progesterones. At multiples

of the human dose no adverse effect on blood clotting was found in

mice, rats, dogs or non-human primates. Indeed, far from

accelerating blood coagulation, high doses of oestrogens in rats and

dogs prolonged clotting times. There is therefore no appropriate

animal model for the coagulation changes occurring in women using

oral contraceptives." 4

In 1972, the CSM described tests on over 13,000 animals which

showed that very high doses of oral contraceptives cause cancer. 5

But the rats and mice used in these experiments were so susceptible

to cancer that even those not dosed with the pill (the "control"

animals) suffered high levels of disease: for instance, lung and liver

tumours were found in 25% and 23% of control mice, and adrenal,

pituitary and breast tumours were found in 26%, 30% and 99% of

control rats. Under these circumstances, the British Medical Journal

noted, "It is difficult to see how experiments on strains of animals so

exceedingly liable to develop tumours of these various kinds can

throw any useful light on the carcinogenicity of any compound for

man."5 The Journal believed the tests neither incriminated nor

exonerated the pill and concluded that we would have to wait for the

results of human studies.

The uncertainty of animal experiments has meant that, effectively,

oral contraceptives have been tested by women themselves during

long term use.


1 ) G.R.Venning, British Medical Journal, 1983, January 22, 289-292.

2) R.Heywood in Animal Toxicity Studies: Their Relevance for Man, Eds.

C.E.Lumley & S.R.Walker (Quay Publishing, 1990).

3) R.Heywood & P.F.Wadsworth in Pharmacology of Estrogens, Ed.

R.R.Chaudhury (Pergamon Press, 1981).

4) M.H.Briggs in Biomedical Research Involving Animals, Eds. Z.Bankowski &

N.Howard-Jones (CIOMS, 1984).

5) British Medical Journal, 1972, October 28, 190.


Animal experiments suggested that chloramphenicol was a very safe

drug but clinical experience soon revealed serious side-effects

making it no longer suitable for internal use, except for lifethreatening

infections such as typhoid fever. In France,

chloramphenicol has been completely withdrawn.1

In 1952, physicians in Baltimore drew attention to chloramphenicol's

effects on nerve cells in the body.2 They described a patient who

almost became blind and who suffered such severe pain in her feet

that she could only walk with the aid of pain-killing narcotic drugs.

She had been taking chloramphenicol for 5 months. This was the first

of many cases of optical and peripheral neuritis caused by

chloramphenicol yet animal experiments had shown the drug to be

practically free of side-effects, even after prolonged administration.2

Even more seriously, the drug caused aplastic anaemia, an often

fatal blood disease sometimes terminating in leukemia. Once again,

the effect had not been predicted by animal tests, and the British

Medical Journal records how chloramphenicol produced nothing

worse than transient anaemia in dogs when given the drug for long

periods by injection, and nothing at all when given orally.3

Today we know that chloramphenicol's deadly side-effect can be

identified by test-tube studies with human bone marrow cells.4


1) C.Spriet-Pourra & M.Auriche, Drug Withdrawal from Sale (PJB Publications,1988).

2)L.Wallenstein & J.Snyder, Annals of Internal Medicine, 1952,vol.36, 1526-1528.

3) British Medical Journal, 1952, July 19, 136-138.

4) G.M.L.Gyte & J.R.B.Williams, ATLA, 1985, vol.l3, 38-47.


Halothane was introduced into clinical practice in 1956 and

immediately hailed as a great advance in anaesthesia. Unfortunately,

the anaesthetic was soon found to harm the liver and within 5 years,

at least 350 cases of "halothane hepatitis" had been recorded. The

condition sometimes proves fatal and between 1964 and 1985, 180

British deaths were linked to the drug.1

The original animal tests had shown no evidence of liver damage,2

and "early attempts to produce an animal model of halothane

hepatitis proved disappointing," according to anaesthetists at

Edinburgh's Royal Infirmary.3 Nevertheless, there has been no

shortage of experiments: since 1976 five "animal models" have been

described though "their application to humans is of doubtful

significance." 3

By 1986, when Britain's Committee on Safety of Medicines

strengthened the warnings of liver toxicity in human patients, 4 it was

still not clear whether the same injuries could be induced in animals.5


1) British Medical Journal, 1986, April 5, 949.

2) Anaesthesiology, 1963,vol.24, 109-110.

3) D.C.Ray & G.B.Drummond, British Journal of Anaesthesia, 1991, vol.67, 84-99.

4) Scrip, 1987, October 2, 2.

5) C.E.Blogg, British Medical Journal, 1986, June 28, 1691-1692.


Butadiene is an important intermediate in the production of synthetic

rubber but causes cancer in the B6C3 F1 strain of laboratory mouse,

an animal widely used to assess the risk of chemicals. Tumours have

also been found in rats although the dose was very high.

Based on the experiments with B6C3F1 mice, America's National

Institute of Occupational Safety and Health (NIOSH) has classified

butadiene as a carcinogen, estimating that exposure to 2 parts per

million for 45 years would result in 597 cancers per 10,000 people.

However, careful observation of butadiene plant workers employed

since 1945, and exposed to much higher levels of the chemical,

revealed no extra cancers. On the contrary, overall cancer deaths

were considerably less than among the ordinary public.1

The NIOSH findings have been criticised since there are many

differences between people and the cancer prone B6C3F1 mouse.

According to an editorial in the journal Science ,1 "with trillions of

dollars, loss of competitiveness, and jobs at stake, a searching

review of the risk assessment methodology of the regulatory

agencies is overdue."


1) P.H.Abelson, Science, 1992, June 19, 1609.


During the 20th century, there has been much debate over the actual

cause of pneumoconiosis, a lung disease suffered by coal miners

because of their occupation. For many years, scientists believed that

inhalation of coal dust was "completely innocuous" and that any

respiratory disease arose from the silica that sometimes

contaminated the coal.1 In bituminous coal pits, where there is little

exposure to silica, mining was not considered dangerous and

consequently few observational studies were carried out in the US

between 1900 and 1960. As a result, there was almost no

information on the amount of coal workers' pneumoconiosis until the

Public Health Service conducted studies in 1962/63.1

The idea that coal dust was harmless originated primarily from the

vivisection laboratory. According to an editorial in the British Medical

Journal,2 scientists who believed silica to be the responsible

contaminant, "take their strongest stand on the fact that animal

experiments... have with few exceptions shown that pure coal dust

produces no fibrogenic reaction." Fibrosis is the formation of scar

tissue and a clear sign of damage to the lung. In fact, the

experimental evidence exonerated pure coal dust and pointed to

silica as the cause of respiratory disease.3

However, the animal data were contradicted by the discovery that

men who worked with pure coal dust or carbon alone, also

developed pneumoconiosis.1,2 Such evidence shows that coal dust

can cause lung disease even in the absence of silica. The

experimental results were further undermined when coal dust,

collected at a coal face where pneumoconiosis among miners was

high, proved innocuous to laboratory rats!2


1 ) W.K.C.Morgan in Occupational Lung Diseases, Eds. W.K.C.Morgan & A.Seaton

(Saunders, 1982).

2) British Medical Journal, 1953, January 17,144-146.

3) L.U.Gardner, Journal of the American Medical Association, 1938, November 19,

1925-1936; Chronic Pulmonary Disease in South Wales III Experimental Studies,

Medical Research Council Special Report Series No.250 (HMSO, 1945).


Many people suffer dermatitis when they come into contact with

nickel compounds as they are considered potent skin sensitizers.1

Nickel is recognised as the single most common cause of contact

dermatitis in women and many of those who suffer prolonged

eczema receive disability pensions.2 In people exposed

occupationally, the condition is known as "Nickel itch".

In contrast, nickel is not a potent skin sensitizer in most of the animal

tests used to predict allergic responses.3 The Draize guinea pig test,

for instance, suggests that nickel does not cause allergic reactions.

Even in the two most widely used animal procedures, nickel

produces either no response (the Buehler Test) or only a moderate

response (the Maximization Test). Both methods also use guinea



1) Medical Toxicology, Eds. M.J.Ellenhorn & D.G.Barceloux (Elsevier, 1988).

2) Textbook of Dermatology VoL 1, 5th edition, Eds. R.H.Champion et al (Blackwell

Scientiflc Publications, 1992).

3) P.A.Botham et al, Food & Chemical Toxicology, 1991, vol.29, 275-286.


The use of monkeys to investigate malaria led to the suggestion that

coma in human patients is due to an increased amount of protein in

the cerebro-spinal fluid, and that this leakage could be corrected with

steroids.1 But in people, steroids do not help with coma. On the

contrary, they actually prove harmful.2 Among survivors, for instance.

coma is prolonged by 16 hours, while complications such as

pneumonia, urinary tract infections, convulsions and gastrointestinal

tract bleeding, developed more frequently in patients receiving

steroids. Subsequent clinical observations of malaria victims have

shown that "the monkey model may simply not be relevant."1


1) Lancet, 1987, May 2, 1016.

2) D.A.Warrell et al, New England Journal of Medicine, 1982, February 11, 313-319


The antidepressant drug mianserin can cause potentially fatal blood

disorders and the British National Formulary recommends that

patients should have full blood counts every 4 weeks during the first

3 months of treatment.1 By early 1988 the World Health Organisation

Collaborative Centre for International Drug Monitoring had collected

321 reports referring to white blood cell disorders. The effects had

not been predicted by animal tests,2 but subsequent studies showed

that they could be observed in test tube experiments with human



I) British National Formulary, No.26 (BMA & the Royal Pharmaceutical Society of

GB, 1993).

2) H.M.Clink, British Journal of Clinical Pharmacology, 1983, vol.15, 291S-293S.

3) P.Roberts et al, Drug Metabolism & Disposition, 1991, vol.l9, 841-843.



Dogs are favourite animals in cardiac research and many

experiments were carried out to develop an artificial mitral valve.

However, the artificial valves almost always produced fatal blood

clots in these animals,1 with the result that many surgeons were

deterred from carrying out human trials.2

Like other experimental surgeons, Starr and Edwards encountered

the familiar problem of blood clots but eventually decided on a

"caged-ball" device.3 Other designs were uniformly fatal to the

animals and whilst 6 of the 7 dogs receiving the caged ball valve died

within 17 days, one did survive for some months. Fortunately, the

new valve proved far more successful in clinical trials where blood

clotting was not a problem.4 The surgeons concluded that "the

marked propensity of the dog to thrombotic occlusion (blood clotting)

or massive embolization from a mitral prosthesis is not shared by the

human being."5

Starr and Edwards wanted to carry out further animal testing of their

new caged ball device but could not use the valve that proved so

successful in patients because it nearly always killed the dogs.

Instead, they designed a different valve specifically for use in these

animals! The modified valve did not kill the animals so frequently:

even so, 78% still died within 46 days. It was noted that "species

differences have therefore led to the use in this clinic of an

unshielded ball valve for human mitral replacement and a shielded

ball valve as the prosthesis of choice for further testing in the dog."5

The successful clinical application of another early design of mitral

valve replacement cast further doubt on the value of animal

research, since none of the dogs used in preclinical tests survived

beyond 40 hours!6


1) A.V.Doumanian & F.H.Ellis, Journal of Thoracic & Cardiovascular Surgery, 1961,

vol.42, 683-695.

2) G.H.A.Clowes Jr, Annals of Surgery, 1961, vol. l54, 740.

3) A.Starr, American College of Surgeons, Surgical Forum, 1960, vol. 11, 258-260.

4) A.Starr & M.L.Edwards, Annals of Surgery, 1961, vol.l54, 726-740.

5) A.Starr & M.L.Edwards, Journal of Thoracic & Cardiovascular Surgery, 1961,

vol.42, 673-682.

6) N.S.Braunwald et al, Journal of Thoracic & Cardiovascular Surgery, 1960,

vol.40, 1-11.



A major hazard of the anti-inflammatory drugs used to treat arthritis

is that they damage the stomach.1 So serious is the problem that any

drug free of this side-effect would have an enormous advantage over

its competitors.

The anti-inflammatory drug Surgam appeared to have these

advantages because it was promoted by the company, Roussel

Laboratories, as giving "gastric protection". However, the claims

were made on the basis of animal tests and could not be confirmed

in clinical trials. As a result of their promotional claims, Roussel were

found guilty of misleading advertising and fined �20,000. A report of

the case in the Lancet described how expert witnesses for both

sides, "...agreed that animal data could not safely be extrapolated to



1) R.Cockel, Gut, 1987, vol.28, 515-518.

2) J.Collier & A.Herxheimer, Lancet, 1987, January 10, 113-114.


The diuretic drug Selacryn was introduced in 1979 but withdrawn

from the US market only a year later after 363 reports of liver

damage including 24 fatalities.1 In many other countries, including

the UK, development of the drug was cancelled.2 Selacryn's harmful

effects were unexpected since they had not been detected in animal



1) S.Takagi et al, Toxicology Letters, 1991, vol.55, 287-293.

2) C.Spriet-Pourra & M.Auriche, Drug Withdrawal from Sale (PJB Publications, 1988).


Perhexiline was first marketed in France during the 1970s as a

treatment for angina. But concern over its side-effects, especially

fatal cases of liver damage, led to withdrawal in the UK, while in

some countries it was never licensed at all. Indeed, some argue that

"its use should be completely avoided."1

The dangers were not predicted by animal tests 2 and administration

of high doses to several species for up to 2 years produced no effect

on the liver.3 According to Richardson Merrell, the company

marketing perhexiline, "...there has been an inordinate amount of

animal work done..At this point we simply have been unable to

induce hepatic (liver) disease in any species." 4

Perhexiline's harmful effects arise in individuals whose body

chemistry has been altered by genetic factors, making them more

sensitive to the drug. Reliance on animal tests can therefore be

seriously misleading since they provide no basis for such subtle



1) D.G.McDevitt & A.M.MacConnachie in Meyler's Side Effects of Drugs, IIth edition, Ed.

M.N.G.Dukes (Elsevier,1988).

2) C.T.Eason et al, Regulatory Toxicology & Pharmacology, 1990, vol.11, 288-307.

3) J.W.Newberne, Postgraduate Medical Journal, 1973, vol.49, April Suppl., 125-129.

4) ibid, p.l30.


Menthol is an ingredient of many cough and cold remedies and is

used as an inhalent to relieve symptoms of bronchitis, sinusitis and

similar conditions. It can also be used as an ointment for application

to the chest or nostrils. If menthol accidentally comes into contact

with the eye, it produces a temporary burning sensation lasting 15-

30 minutes, but there are no after effects. In contrast, menthol

causes "severe damage" to the rabbit's eye.1


1) W M Grant, Toxicology of the Eye, 2nd edition (Charles Thomas,1974).


The success of selenium disulphide (Selsun) as an antidandruff

shampoo, led to the suggestion that it might also be useful for the

treatment of blepharitis, a similar but painful condition involving the

eyelids. Trials were carried out in which an ointment containing 0.5%

selenium disulphide was applied to the lid margins. However, the

ointment proved irritating if it accidentally came into contact with the

conjunctiva and one patient developed "moderately severe

conjunctivitis."1 In contrast, animal experiments have shown that

"Selenium disulphide 0.5% ophthalmic ointment is nontoxic to rabbit

corneas or conjunctivas" (emphasis added).2


1) G.C.Bahn, Southern Medical Journal, 1954, vol.47, 749-752.

2) J.W.Rosenthal & H.Adler, Southern Medical Journal, 1962, March, 318.


Researchers have discovered that coconut soap causes skin

irritation in rabbits. During a comparison of human and animal test

data for a selection of household and industrial products, Proctor and

Gamble scientists found that while coconut soap had a "negligible"

effect on the skin of volunteers, it produced "moderate" irritation in

rabbits.1 Pine oil cleaner also produced "moderate" irritation in

rabbits (and guinea pigs) but only a slight effect on human skin.

Other substances which produced insignificant effects on human

skin but irritation in animals included high and low carbonate

detergents, phosphate detergents, enzyme detergent, sodium

carbonate and even lemon juice! Overall, only 6 of 24 products

tested had the same effects in people, rabbits and guinea pigs. The

report concluded that "Neither the rabbit nor the guinea pig provides

an accurate model for human skin. The skin responses of these

animals differ in both degree and in kind from those found in human


Similar conclusions have been reached for cosmetic ingredients.

Scientists at the Warner Lambert Research Institute in New Jersey

note that "...animal skin is entirely different from human skin and that

there may be no correlation between the mildness of a raw material

on a rabbit's back and its safety during use on a human face." They

describe how the cosmetic ingredient isopropyl myristate is

considered safe for use on the human body but causes irritation to

rabbits. 2


I) G.A.Nixon et al, Toxicology & Applied Pharmacology, 1975, vol.31, 481-490.

2) M.M. Rieger & G.W.Battista, Journal of the Society of Cosmetic Chemists, 1964,

vol.15, 161-172.


Morphine remains the most valuable analgesic for severe pain 1 yet

has such a peculiar effect in some species that had it been tested on,

say cats, prior to human studies, it could have been rejected. In

these animals the drug produces a condition known as "morphine

mania" which leaves them highly excitable and apprehensive. Their

movements are irregular and jerky, and their pupils are abnormally

dilated.2 While morphine produces hyperexcitement in cats, it has the

opposite, calming effect in people.3 Furthermore, their pupils may be

contracted.1 Fortunately the drug was discovered through human

studies and only later tested on animals. 4


1) British National Formulary, no.26 (BMA & the Royal Pharmaceutical Society of


2) F.M.Sturtevant & V.A.Drill, Nature, 1957, June 15, 1253.

3) B.Brodie, Clinical Pharmacology & Therapeutics, 1962, vol.3, 374-380.

4) J.T.Litchfield in Drugs in our Society, Ed. P.Talalay (Johns Hopkins, 1964).


Domperidone is used for the treatment of nausea and vomitting,

particularly that caused by anticancer drug therapy. In 1986 the

injectable form of the drug was withdrawn worldwide1 because of

potentially hazardous heart rhythm disturbances. The dangers had

not been predicted by animal experiments 2 and in dogs, the animal

traditionally used to assess effects on the heart, more than 70 times

the recommended human dose produced no changes in the



1) C.Spriet-Pourra & M.Auriche, Drug Withdrawal from Sale,( PJB Publications,1988).

2) R.Heywood in Animal Toxicity Studies: Their Relevance for Man, Eds. C.E.Lumley &

S.R.Walker (Quay Publications, 1990).

3) R.N.Brogden et al, Drugs, 1982, vol.24, 360-400.


Squalene is a natural constituent of human sebum, the substance

formed by sebaceous glands around the roots of hairs to keep the

skin lubricated and supple. Although a natural human product,

squalene has still been applied to the skin of rabbits and guinea pigs,

where it actually produced hair loss. This is obviously not the case in

people,1 and it has been extensively and safely employed in



1) B.Boughton et al, Journal of Investigative Dermatology, 1955, vol.24 179-189.

2) M M Rieger & G W Battista, Journal of the Society of Cosmetic Chemists, 1964,

vol 15, 161-172


Prenylamine, a treatment for angina, was withdrawn from the UK

market in 1988,1 the main problem being that the drug caused

ventricular tachycardia, a condition in which the heart beats

abnormally fast. The side-effect caused patients to faint. In contrast,

animal experiments carried out at the University of Eoteborg in

Sweden revealed that in cats, rabbits and guinea pigs, prenylamine

reduced the heart rate by up to 25%.2 In cats, for instance, a dose of

prenylamine reduced heart rate from 225 beats per minute to 171.


1) C.Spriet-Pourra & M.Auriche, Drug Withdrawal from Sale (PJB Publications, 1988).

2) H.Obianwu, Acta Pharmacology et Toxicology, 1967, vol.25, 127-140.


In 1951, physicians at the University of California Medical School in

San Francisco, warned ophthalmologists against the prolonged use

of furmethide in the treatment of glaucoma.1 They noted that

permanent obstruction of the tear passages occured in over 70% of

patients where the drug was used for more than 3 months.

Eleven years earlier, researchers had reported experiments on

animals' eyes, pronouncing the drug "entirely safe" and worthy of

clinical trial.2 The tests were performed on rats, guinea pigs and

rabbits and continued for several months.


1) R.N.Shaffer & W.L.Ridgway, American Journal of Ophthalmology, 1951, vol.34, 718-720.

2) A.Myerson & W.Thau, Archives of Ophthalmology, 1940, vol.24, 758-760.


Phenylbutazone (Butazolidine) was once widely employed for the

treatment of arthritis but reports of aplastic anaemia, an often fatal

blood disease caused by damage to the bone marrow, led to the

drug's withdrawal in some countries and to its restriction in others,

notably America, France and the UK.1

On the basis of animal tests, phenylbutazone had seemed a safe

drug with no toxic effects observed in rats even after administration

of 5-10 times the dose used for people.2 In particular,

phenylbutazone's harmful effect on the bone marrow had not been

predicted,3 and one year after marketing, researchers noted that

"there have been no published reports of serious effects...on the

hemopoietic (blood forming) the experimental animal."4

Later research showed that the dangers could be identified by testtube

experiments with human bone marrow cells.5

It has been estimated that phenylbutazone and oxyphenbutazone, a

closely related drug that also causes aplastic anaemia, have been

reponsible for 10,000 deaths worldwide.6 Oxyphenbutazone

(Tanderil) was withdrawn altogether in 1985.


1) C.Spriet-Pourra & M.Auriche, Drug Withdrawal from Sale (PJB Publications, 1988).

2) C.Hinz & L.M.Gaines, Journal of the American Medical Association, 1953, vol.l51, 38-39.

3) R.Heywood in Animal Toxicity Studies: Their Relevance for Man, Eds. C.E.Lumley &

S.R.Walker (Quay Publishing, 1990).

4) O.Steinbrocker et al, Journal of the American Medical Association, 1952, November 15,


5) C.S.Smith et al, Biochemical Pharmacology, 1977, vol.26, 847-852.

6) Estimate by Dr Sidney Wolfe in Lancet, 1984, February 11, 353.


The anaesthetics ether, nitrous oxide and chloroform originated from

experiments carried out by physicians and scientists on themselves,

and, together with the introduction of hygienic conditions, enabled

surgery to emerge from the dark ages.1 Because of their high safety

profile, nitrous oxide and ether have stood the test of time. In the

case of chloroform, entrenched attitudes and contradictory animal

experiments allowed a toxic drug to outlive its value and remain in

use for over 100 years. 2

Deaths from chloroform were reported almost weekly during the

second half of the l9th century and between 1887 and 1896 there

were 376 fatalities in England and Wales. Many believed the deaths

resulted from respiratory failure but that risks could be minimised by

appropriate administration of the drug and by devoting attention to

the patient's breathing in order to detect early warning signs. The

alternative (correct) explanation, that chloroform has a direct effect

on the heart, was discounted.

Unfortunately, animal experiments carried out by the Hyderabad

Commissions of 1888 and 1889 supported the view that chloroform

affects the respiration rather than the heart.2 In a famous telegram to

the Lancet,3 Lauder Brunton summarised results from the Second

Commission: "Four hundred and ninety dogs, horses, goats, cats

and rabbits used...Results most instructive. Danger from chloroform

is asphyxia or overdose: none whatever heart direct." Anaesthetists

must have been reassured to hear Brunton's conclusion that

chloroform "never causes sudden death from stoppage of the heart."

In 1893, clinical observations completely contradicted the

conclusions from Hyderabad and showed that heart failure is the

commonest cause of death from chloroform.2 Nevertheless, use of

the drug continued until the 1950s and the Hyderabad Commissions

were later blamed for failing to recognise species differences.2


I ) R.Sharpe, The Cruel Deception: the use of animals in medical research (Thorsons, 1988).

2) K.B.Thomas, Proceedings of the Royal Society of Medicine, 1974, vol.67, 723-730.

3) Lancet, 1889, December 7, 1183.


When treating iron-deficiency anaemia, doctors prefer their patients

to take iron by mouth, but should oral therapy fail, the iron is

administered by injection.1 Injectable iron remedies were introduced

during the 1930s but could easily have been discarded. At that time,

experiments in which anaemias were artificially induced in animals

by iron deficiency or by repeated haemorrhage, led to the conclusion

that injecting iron had no therapeutic value.2 Fortunately, clinical

studies proved that anaemic patients could be cured in this way.

Iron sorbitol is one form of injectable iron that might have been

rejected for a different reason. Administration to rats and rabbits

caused cancer at the injection site and the implications for human

therapeutics appeared serious. However, clinical experience has

revealed no real hazard to patients.3


1) British National Formulary, No.26 (BMA & Royal Pharmaceutical Society of G.B., 1993).

2) G.N.Burger & L.J.Witts, Proceedings of the Royal Society of Medicine, 1934, vol.27, 447-455.

3) M.Weatherall, Nature, 1982, April 1, 387-390.


During the twentieth century, extensive research has been carried

out to develop an animal model that mimics spinal cord injuries (SCI)

in people.1 A common procedure is to drop weights onto the spinal

cord of cats.2 By using animals, researchers hoped to devise

promising therapies and discover new insights into the

condition.However, virtually no treatments have been developed that

work in human patients.1 In 1988 for instance, Dennis Maiman of the

Department of Neurosurgery at the Medical College of Wisconsin,

Milwaukee, noted that "In the last two decades at least 22 agents

have been found to be therapeutic in experimental SCI...

Unfortunately, to date none of these has been proven effective in

clinical SCI."1 The failure to accurately predict human responses is

attributed to the artifical nature of the animal model.

In 1990, however, clinical trials did show that high doses of steroids

could be beneficial. Some have credited animal tests with the

discovery but the claim has been challenged. It is argued that the

animal experiments were not only unnecessary but they gave

inconsistent results, with some tests suggesting the therapy would

actually fail!2


I) D.Maiman, Journal of the American Paraplegia Society, 1988, vol. 11, 23-25.

2) S.R.Kaufman, Perspectives on Medical Research, 1990, vol.2, 1-12.


When animal researchers tested a newly discovered substance,

psicofuranine, for anti-cancer activity, they found contradictory

evidence in rats and mice.1 The drug proved active against several

tumours in laboratory rats but had no effect on 3 different cancers in

mice. Unfortunately doctors could not properly assess the drug

against human cancer since psicofuranine produced severe and

unexpected side-effects in early human trials, thus terminating any

further investigation in people. The drug damaged the heart yet no

cardiac toxicity had been found in mice, rats, dogs or monkeys. 1

Although clinical study of psicofuranine was abandoned, further

animal experiments were carried out in an attempt to reproduce the

heart problems seen in people. Once again, no cardiac toxicity could

be observed even when dogs and monkeys were given 5-10 times

the harmful human dose.


1) C.G.Smith et al, Journal of International Medical Research, 1973, vol.1, 489-503.


During clinical trials, the anticancer drug sparsomycin produced eye

damage, resulting in serious blind spots in 3 of the 5 patients.

Although sparsomycin was highly toxic to several animal species, as

would be expected for an anticancer drug, no specific effect on the

eye had been found.1 After the eye problems had been reported,

further attempts were made to induce the condition in rats and

monkeys but these also failed even though rats were dosed every

day for 2 weeks with 30-300 times the amount found to harm

people.1 No retinal toxicity was observed in additional animal tests

and further experimentation was adandoned, as was the drug.


1) C.G.Smith et al, Journal of International Medical Research, 1973, vol.l, 489-503.


Contrary to human experience, experiments on pregnant mice and

rabbits would suggest that corticosteroids are very dangerous to the

unborn child. In some strains of mice cortisone produces cleft palate

in up to 100% of the offspring.1 With rabbits, corticosteroids mainly

affect the heart but can also cause severe growth retardation in the

uterus and death of the foetus. However, scientists have found "very

wide species variation"2 and cortisone is not considered harmful to

human babies.1 Rats and monkeys are also "very tolerant of

corticosteroids in pregnancy, abnormalities or growth retardation only

occurring uncommonly, with high doses of the most potent



I) R.M. Ward & T.P.Green, Pharmacology & Therapeutics, 1988, vol.36, 326.

2) R.K. Sidhu in Drugs & Pregnancy: Human Teratorgenesis & Related Problems,

Ed. D.F.Hawkins (Churchill Livingstone, 1983).


Iproniazid was originally developed as a treatment for tuberculosis

but found use as an antidepressant. Although considered "harmless"

on the basis of animal tests,1 iproniazid produced fatal cases of liver

damage in human patients and the drug was eventually abandoned.2


I ) J.Boyer in Clinical Pharmacology, Basic Principles in Therapeutics, 2nd edition,

Eds. K.L.Melmon & H.F.Morrelli (MacMillan, 1978)

2) B.Blackwell & J.S.Simon in Side Effects of Drugs Annual 13, Eds. M.N.G.Dukes

& L.Beeley (Elsevier, 1989).


Thalidomide was first introduced as a sedative by the German drug

company Chemie Grunenthal in 1957, and by the Distillers company

in Britain a year later. Although animals could tolerate massive doses

without ill-effect,1 thalidomide was soon found to cause peripheral

neuritis in human patients: feelings of numbness were followed by

severe muscular cramps, weakness of the limbs and a lack of


The Australian obstetrician William McBride was first alerted to

thalidomide's most notorious side-effect after seeing 3 babies born

with very unusual birth defects. Unfortunately, his warnings to the

medical profession were delayed because he tried to "confirm" his

observations by testing the drug in mice and guinea pigs, both of

whom proved resistant to the drug.2 Only after seeing further human

cases did McBride publish his findings.

Although not specifically tested for birth defects prior to marketing,

subsequent experiments revealed "extreme variability in species

susceptibility to thalidomide."3 For instance, mice could safely

tolerate 8000 times the dose found harmful to human babies.4 In his

book Drugs as Teratogens, Schardein writes, "in approximately 10

strains of rats, 15 strains of mice, eleven breeds of rabbit, 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 (birth

defects) have been induced only occasionally." Scientists eventually

discovered that birth defects similar to those found in people could

be induced in certain types of rabbit and primate. Nevertheless, New

Zealand white rabbits had to be dosed with 300 times the amount

dangerous to humans.5

The thalidomide disaster prompted additional,extensive testing of

drugs and chemicals in pregnant animals,but some scientists believe

that "animal malformations seldom correlate with those of humans."6

Furthermore, " animal model has been found which responds

satisfactorily to all known teratologic agents in humans to permit

reliable screening of substances for their teratologic potential.

Careful surveillance, reporting and prospective study ...remain the

mainstays for detection of adverse effects following foetal drug



I) R.D.Mann, Modern Drug Use, an Enquiry on Historical Principles (MTP Press, 1984).

2) The Sunday Times "Insight" Team, Suffer the Children - The Story of Thalidomide

(Andre Deutsche, 1979).

3) T.H.Shepard, Catalogue of Teratogenic Agents (Johns Hopkins Press, 1976).

4) S.K.Keller & M.K.Smith, Teratogenesis, Carcinogenesis & Mutagenesis, 1982, vol.2,


5) New Zealand White rabbits were sensitive to doses of 150mg/Kg of thalidomide

(ref.6) whilst the dangerous human dose was O.5mg/Kg (ref.4).

6) R.M.Ward & T.P.Green, Pharmacology & Therapeutics, 1988, vol.36, 326.


Mitoxantrone was developed in the hope of providing effective

cancer treatment without side-effects on the heart. Animal

researchers were presumably reassured when tests on beagle dogs

"failed to demonstrate cardiac failure."1 But in clinical trials several

patients suffered side-effects including heart failure, and more

widespread use of the drug confirmed that cardiac toxicity is a major

problem. For instance, data from 3,360 patients receiving

mitoxantrone included 88 reports of cardiac side effects with 29

cases of heart failure.2 And a recent Chinese study suggested that

20% of patients developed cardiotoxicity following treatment with



1) R.Stuart Harris et al, Lancet, 1984, July 28, 219-220.

2) Martindale: The Extra Pharmacopoeia, 29th edition, Ed. J.E.F.Reynolds

(Pharmceutical Press,1989).

3) A.Stanley & G.Blackledge in Side Effects of Drugs, Annual 15, Eds.

M.N.G.Dukes & J.K.Aronson (Eisevier,1991).



Carbenoxalone was introduced during the 1960s for the treatment of

peptic ulcers but caused salt and water retention in some patients

leading to high blood pressure, swelling, weight gain, muscle

weakness and heart failure. Other drugs are now preferred, says the

British National Formulary, but if carbenoxalone is to be used,

patients should be monitored carefully during treatment.1

Prior to marketing, animal tests had given the impression that

carbenoxalone was safe, having revealed no harmful effects.2 These

tests were carried out on rodents but scientists then realised that

people metabolised carbenoxalone quite differently to rats, mice and

rabbits. Further experiments were therefore undertaken with dogs

and monkeys but again, there was no evidence of toxicity. 2


1) British National Formulary, no.26 (BMA & the Royal Pharmaceutical Society of G.B., 1993).

2) C.T.Eason et al, Regulatory Toxicology & Pharmacology, 1990, vol. 11, 288-307.


Britain's Committee on Safety of Medicines has alerted doctors to the

dangers of clindamycin, an antibiotic whose most serious side-effect

is an intestinal disease called pseudomembraneous colitis. The

condition leads to diarrhoea and sometimes proves fatal. By 1980,

12 years after the drug was marketed in the UK, 36 deaths had been

reported.1 Although the problem can occur with other antibiotics, it is

most frequently seen with clindamycin, and the British National

Formulary warns that patients should stop taking the drug

immediately if diarrhoea develops.

In contrast, rats and dogs given clindamycin every day for a year,

could tolerate 12 times the maximum recommended human dose.2


1) G.R.Venning, British Medical Journal, 1983, January 15,199-202.

2) The British National Formulary (No.26,1993) lists the maximum oral dose for

severe infections as 450mg every 6 hours i.e. 25mg/kg for a person weighing 70

kg taking 4 doses in 24 hours. Rats and dogs could tolerate more than 300mg/kg

(J.E.Gray et al, Toxicology & Applied Pharmacology, 1972, vol.21,516-531)


For decades, America's National Cancer Institute (NCI) has used

animals in the search for new drugs. Tens of thousands of chemicals

have been assessed in mice given leukemia but the method has

proved highly inefficient. One scientist estimates that for every 30-40

drugs effective in treating mice with cancer, only one will work in

people,1 which suggests that during clinical trials many cancer

patients will be exposed to the severe toxicity of anticancer drugs

without any corresponding benefit. During the 1980s, researchers

acknowledged that the NCl�s traditional approach was failing to

identify promising new treatments against any of the main cancers.2,3

In the new strategy, mice have been replaced by test-tube studies

with human cancer cells, at least for preliminary experiments. Drugs

showing promising activity are then subject to further animal tests so

there is still the risk of misleading predictions.4 As an alternative,

drugs could be further assessed using fresh human tumour tissue

from biopsies or therapeutic operations.5 Results would then be

directly relevant to people.4


I) D.D.Von Hoff, Journal of the American Medical Association, 1979, August I0, 503.

2) R.Kolberg, Journal of NIH Research, 1990, vol.2, 82-84.

3) A.Pihl, lnternational Journal of Cancer, 1986, vol.37, 1-5.

4) S.E.Salmon, Cloning of Human Tumor Stem Cells (Alan Liss, 1980).

5) C.W.Taylor et al, Journal of the National Cancer Institute, 1992, vol.84, 489-494.


Beta-blockers were developed for the treatment of heart conditions

and the first agents to be administered to human patients were

pronethalol and propranolol. Ironically, pronethalol proved generally

safe and effective in laboratory animals but failed the clinical test,

while propranolol appeared toxic in many animal experiments yet is

widely used in clinical practice.

Pronethalol was "well tolerated" by rats and dogs in prolonged

toxicity tests at high doses, except for occasional effects on the

central nervous system.1 However, clinical trials revealed an

unacceptable number of side-effects2 including heart failure, a

hazard not predicted by animal experiments.1 Shortly after, long term

tests in a certain (Alderley Park) strain of laboratory mouse produced

cancer of the thymous gland but no carcinogenic effects were ever

found in rats, guinea pigs, dogs, monkeys or other types of mouse.1

Pronethalol was quickly replaced by propranolol but tests in rats,

dogs and mice put further development in jeopardy.3 Moderate to

high doses caused rats to collapse and dogs to vomit severely.1

Deaths were also seen in mice shortly after dosing. When the

amount of drug was reduced to that used clinically, propranolol was

said to be "well tolerated". Even so, some of the rats still had heart


Later clinical observations showed that propranolol could also lower

the blood pressure,4 and today beta-blockers are widely used for the

treatment of high blood pressure.


1) J.M.Cruickshank et al in Safety Testing of New Drugs, Eds. D.R.Laurence et al

(Academic Press,1984)

2) W.Sneader, Drug Discovery: the evolution of modern medicine (Wiley, 1985)

3) D.R.Laurence et al (Eds.), Safety Testing of New Drugs (Academic Press,1984)

4) E.S.Snell, Pharmacy International, 1986, February, 33-37.


Librium and Valium were the first of a new type of tranquillizing drug

to be introduced during the early 1960s. They were called "minor

tranquillizers" (benzodiazepines) and many similar drugs quickly

followed. They soon became the most widely used of all prescribed

drugs. Almost immediately after the introduction of Librium and

Valium, doctors reported cases of dependence but it was generally

assumed that high doses were necessary.1 At the usual therapeutic

amounts, dependence was thought to be uncommon and not a

serious problem. The idea prevailed for 20 years and received

support from laboratory research since "animal not

indicate the potential for the development in the human of

dependence at therapeutic dosage levels."2

It is known, however, that "animal not predict clinical

dependence potential reliably,"3 and more careful human

observations revealed that tranquillizers could induce dependence at

ordinary doses. By the mid-1980s, an estimated 500,000 people in

Britain alone may have been addicted to their treatment.4


1) H.Petursson & M.Lader, Dependence on Tranquillizers (Oxford University Press, 1984).

2) J.Marks, The Benzodiazepines (MTP Press, 1978).

3) Drug & Therapeutics Bulletin, 1989, vol.27, 28.

4) The Benzodiazepines in Current Clinical Practice, Eds. H.Freeman & Y.Rue (Royal Society

of Medicine Services, 1987).


In 1971 doctors reported unexpected pregnancies among women

taking the "pill".1 Of 88 women taking oral contraceptives in addition

to the antituberculous drug rifampicin, 75% suffered disturbances to

their menstrual cycle, and 5 became pregnant. The rifampicin had

stimulated the patient's liver to metabolise, or breakdown, the pill

more rapidly. Consequently, far less contraceptive remained to

protect the women from pregnancy. The British National Formulary

(1993) now tells doctors prescribing rifampicin to "advise patients on

oral contraceptives to use additional means (of contraception)."

Further reports showed that rifampicin accelerates the breakdown of

many other medicines.2 An example is methadone where rifampicin

led to withdrawal symptoms by reducing the amount of drug. Another

patient rejected their kidney graft because rifampicin had diminished

the dose of immunosuppressive drug cyclosporin.

Rifampicin's peculiar effect had not been predicted by animal

experiments.3 Following discovery of the effects in people, further

animal tests were carried out but these proved contradictory. For

instance, the drug's action could not be reproduced in rats.4 In mice,

however, prolonged treatment with rifampicin did stimulate the liver's

metabolic processes but a single dose had the opposite effect,

slowing down metabolism.4 Nevertheless, the problems with

rifampicin might have been predicted had scientists used human liver

tissue for their tests.5


1) Reported in J.P.Mumford, British Medical Journal, 1974, May 11, 333-334.

2) H.Meyer et al in Meyler's Side Effects of Drugs, 11th edition, Ed. M.N.G.Dukes

(Elsevier, 1988).

3) E.Nieschlag, Pharmacology & Therapeutics, 1979, vol.5, 407-409.

4) D.Pessayre & P.Mazel, Biochemical Pharmacology, 1976, vol.25, 943-949.

5) A.M.Jezequel et al, Gut, 1971, vol.l2, 984-987.


It is a minor miracle that tamoxifen overcame a succession of

conflicting animal data to find a place in clinical practice. The drug

was developed by ICI during the 1960s as an oral contraceptive and

in rats tamoxifen can prevent ovulation or terminate pregnancy. 1 In

women however, tamoxifen stimulates ovulation and is listed as a

treatment for infertility!2

Tamoxifen is also used in breast cancer therapy where it works by

blocking the action of oestrogen in breast tissue. The drug is

therefore called an "anti-oestrogen". In monkeys, and rats at low

doses, tamoxifen also acts as an anti-oestrogen but in mice, dogs,

and rats at high doses, the drug has the opposite effect, behaving

like an oestrogen.1 The use of animals to investigate these effects is

bedevilled with problems since "significant species variation has

been observed in target tissue response to oestrogens and

antioestrogens making it hazardous to predict therapeutic activity in

the human by extrapolation of effects in experimental animals..."3

Animal tests have also given conflicting results in assessing the

drug's harmful effects, with tamoxifen producing liver tumours in rats

but not mice.4 Liver cancer does not seem to be a problem for human

patients and only two cases have been reported in around 3 million

women who have received the drug. Furthermore, tamoxifen is

known to be processed differently by rats and people. John

Patterson, medical director at ICI, explains that "If this was a new

chemical entity those findings in rats would have caused us to stop

its development, but the human experience gave us confidence."5

Nevertheless the rat data led to a major row between Britain's two

biggest cancer charities and the Medical Research Council over

trials of tamoxifen for the prevention of breast cancer, in which the

drug is given to healthy women. Most proponents of the trial

considered the rat experiments nothing to worry about 6,7 but the

MRC actually withdrew its support and initiated new animal tests.

Embarrassed by the split, one of the cancer charities, the Imperial

Cancer Research Fund, stated that "We are going to be in a position

where the animal rights people are going to be saying to us �You

ignore animal data when you choose to'."1

Ironically, fresh doubt was cast on tamoxifen�s preventive role by a

subsequent clinical study suggesting an increased risk of womb

cancer amongst breast cancer patients being treated with the drug.

Whilst questioning whether it should be given to healthy women, the

report stressed that for the treatment of breast cancer, the benefits

far outweighed the risks.8

Overall, tamoxifen has comparatively few serious side effects and

according to ICI, the main reason patients stop taking the drug is

nausea and vomitting.1 This must have surprised the company

because "None of the toxicological studies produced any evidence of

vomitting even though high doses were used in dogs which we

consider to be a predictive species for vomitting in man."1


1) M.J.Tucker et al in Safety Testing of New Drugs, Eds. D.R.Laurence et al (Academic Press,


2) British National Formulary, No.26 (BMA & the Royal Pharmaceutical Society of GB, 1993)

3) P.K. Devi in Pharmacology of Estrogens, Ed. R.R.Chaudhury (Pergamon Press, 1981)

4) I.N.H. White et al, Biochemical Pharmacology, 1993, vol.45, 21-30

5) Reported in P.Brown, New Scientist, 1992, February 29, 11.

6) Editorial, New Scientist, 1992, March 21, 9

7) P. Brown, New Scientist, 1992, March 21, 12.

8) F. E.van Leeuwen et al, Lancet, 1994, February 19, 448-452.


Because of their potent effects on the immune system, corticosteroid

drugs are widely used in medicine. They also have many side-effects

which limit their usefulness, and much research has been carried out

to discover exactly how the drugs work. However, there are said to

be "remarkable differences in susceptibility to glucocorticosteroids

between various species," with animals being classiffed as steroid

resistant or steroid-sensitive.1 In mice, a steroid-sensitive species, a

single dose of cortisone produces a 90% decrease in the thymus, an

organ that plays a crucial role in immunity. By contrast, the same

dose of cortisone given every day for a week, produced only a 37%

decrease in the steroid-resistant guinea pig's thymus. And while

steroids inhibit the production of circulating antibodies in sensitive

animals, the same effect is difficult to achieve in resistant species.1

Most of the research on corticosteroids has been carried out on

steroid sensitive species such as rats, mice, rabbits and hamsters

whereas human beings are steroid resistant.1 As researchers at the

University of Dundee point out "The mode of action of these drugs is

very complicated, so it is regrettable that most of the extensive

literature on animal experimental work is irrelevant to human

therapeutics since many species respond in a very different manner

from man."2 Consequently they concentrated on human clinical

studies and test-tube experiments.


1) H.N.Claman, New England Journal of Medicine, 1972, August 24, 388-397.

2) J.S.Beck & M.C.K.Browning, Journal of the Royal Society of Medicine, 1983, vol. 76, 473-



In 1956 British doctors drew attention to a link between X-rays during

pregnancy and subsequent childhood cancers.1 Within a few years

similar findings were reported in American children. But for a quarter

of a century, scientists questioned whether X-rays were actually the

cause and cited animal experiments to show that the foetus is not

especially sensitive to radiation.2 However, it seems that compared

with other species, the human foetus is more susceptible to the

carcinogenic effects of X-rays,2 and during the 1980s further

observational studies confirmed the hazards, particularly in early



I) A.M. Stewart et al, Lancet, 1956, September I,447; British Medical Journal, 1958,

June 28,1495-1508.

2) E.B.Harvey et al, New England Journal of Medicine, 1985, February 28, 541-545.

3) E.G.Knox et al, Journal of the Society of Radiological Protection, 1987, vol.7,3-15;

E.A.Gilman et al, Journal of Radiological Protection, 1988, vol.8, 3-8.


Methanol is employed in a wide variety of consumer products

including solid fuels, antifreeze, windshield wiper fluid, paint remover,

varnishes and as a solvent in photocopying machines. It is also

imbibed as a cheap alternative to alcohol.

Although methanol is a highly poisonous, potentially lethal

substance, this was not realised for many years.1 Common

laboratory species such as rats and mice are resistant to its effects,2

and experiments during the early years of the 20th century gave the

impression that methanol was only slightly toxic, and far less

poisonous than alcohol.3 In fact, methanol is ten times more toxic

and a single bout of drinking methanol can lead to temporary or

permanent blindness in people.4 This does not happen in rats, mice,

dogs, cats, rabbits or chickens.3 Eventually, in the 1950s, and again

during the 1970s, scientists found that the horrifying symptoms of

methanol poisoning could be induced in monkeys.2

Animal experiments also proved misleading in devising treatment.

During the 1920s, good results were achieved using bicarbonate in

cases of human poisoning, but tragically the results were

undermined by animal experiments. In 1955 an analysis of the

subject stated that "it is indeed deplorable that about 30 years

elapsed before the good effects of this treatment became commonly

known, and unfortunately some still doubt its value. It seems that the

authors of medical textbooks have paid more attention to the results

of animal experiments than to clinical observations."3 The treatment

not only failed in animals but generally proved fatal, prompting some

researchers to advise against it.

Another approach is to administer alcohol in order to reduce the

toxicity of methanol. While this is effective in people. animal tests

suggested that it would actually increase the danger of methanol

As a result, some discouraged its use in cases of human poisoning.3

However, both bicarbonate and alcohol have withstood the clinical

test and are still recommended for the treatment of methanol



1) M.J.Ellenhorn & D.G.Barceloux, Medical Toxicology: Diagnosis & Treatment of

Human Poisoning (Elsevier, 1988)

2) T.R.Tephly, Life Sciences, 1991, vol. 48, 1031-1041

3) 0.Roe, Pharmacological Reviews, 1955,vol.7, 399-412.

4) P.Wingate, Medical Encyclopedia (Penguin, 1983).


During the 1960s Swiss doctors noticed a sudden and unexpected

rise in a dangerous lung disease called obstructive pulmonary

hypertension. The cause was traced to aminorex which had been

used since 1965 for the treatment of obesity.1 The drug produces an

increase in lung pressure leading to chest pains, difficulty breathing,

fainting spells, heart problems and, in some cases, death.2

Aminorex�s deadly side effect had not been predicted by animal

experiments 3 and in 1968 the drug was withdrawn from sale.

Animal experiments continued even after withdrawal but long term

administration to rats still failed to induce the disease.2 In dogs,

aminorex did increase lung pressure 1 but its relevance to the human

condition is unclear since a later analysis concluded that "pulmonary

hypertension cannot be induced in experimental animals even with



1) F.Follath et al, British Medical Journal, 1971, January 30, 265-266.

2) E.H.Ellinwood & W.J.K.Rockwell in Meyler�s Side Effects of Drugs, 11th edition, Ed.

M.N.G.Dukes (Elsevier, 1988)

3) A.D.Dayan in Risk-Beneft Analysis in Drug Research, Ed. J.F.Cavalla (MTP Press, 1981).

4) P.H.Connell in Side Effects of Drugs Annual - 3, Ed. M.N.G.Dukes (Excerpta Medica,



On the basis of animal experiments, the synthetic oestrogen

diethylstilbestrol (DES) was suggested as a means of preventing

miscarriage.1 Although no proper human (clinical) trials were carried

out,2 the procedure nevertheless became widely accepted, and

between 1948 and 1971, DES was given to some 2-3 million

pregnant women in the US alone.

However, DES was ineffective. In 1953, properly controlled clinical

trials showed that DES did not work.3 Tragically, the study failed to

report that DES increased abortions, neonatal deaths and premature

births, a conclusion that could have been made from the data

available in the trial.4 DES was not only ineffective, it was also

unsafe. Just how unsafe was only revealed in 1971 when

researchers traced a link between exposure to DES and a previously

rare form of vaginal and cervical cancer in daughters of women who

had taken the drug during pregnancy.5 Almost 600 cases have been

reported 6 but DES has proved a biological timebomb as side-effects

continue to surface in sons and daughters of women who took the


It has been suggested that animal tests provided an early warning of

the problems. It is true that in 1938 DES was found to cause breast

cancer in male mice, but since the cancer-causing potential of other

oestrogens varied according to the strain of mouse used,7 the results

could hardly be a serious basis for action. Furthermore, the

consensus among animal researchers at the time was that

oestrogens did not produce cancer,7 rather they gave male mice

mammary glands and thus made them susceptible to the same

cancer-causing factors that operated within female animals. In fact,

a summary of the animal data in 1941 found "only meagre evidence"

that oestrogens cause cancer of the cervix.7 Not until the 1970s did

it become clear that in contrast to the majority of animal experiments,

DES was a potent cause of cervical cancer in women.


1) Health Action International, "Problem Drugs" pack, 1986, May 13

2) D.Brahams, Lancet, 1988, October 15, 916.

3) W.J.Dieckmann et al, American Journal of Obstetrics & Gynaecology, 1953,

vol.66, 1062-1081.

4) Y.Brackbill & H.W.Berendes, Lancet, 1978, September 2, 520.

5) A.L.Herbst et al, New England Journal of Medicine, 1971, April 22, 878-881.

6) C.Vanchieri, Journal of the National Cancer Institute, 1992, vol.84, 565-566.

7) S.Peller, Cancer in Man (McMillan, 1952).


The vital heart drugs digoxin and digitoxin are the pure substances

extracted from digitalis whose value in treating heart failure and

cardiac arrhythmias originated from studies of human patients.1,2

However, doctors must be careful not to give too high a dose as they

can then be toxic. Fortunately the drugs did not derive from animal

experiments since doses considered safe for rats, guinea pigs, dogs,

and cats can actually kill human patients.3 Today we know that

digoxin's lethal dose is more accurately predicted by test-tube

studies with human cells.4

Animal tests also suggested that digitalis raised the blood pressure,

and as a result, it was once widely taught that the drug would be

dangerous for certain patients and should not therefore be given.

Thankfully, clinical observations eventually showed this to be

incorrect and digitalis can be used with great benefit.2


I) W.Sneader, Drug Discovery: The Evolution of Modern Medicine (Wiley, 1985)

2) T.Lewis, Clinical Science (Shaw & Sons Ltd, 1934)

3) G.T.Okita, Federation Proceedings, 1967, vol.26, 1125-1130.

4) R.Jover et al, Toxicology In Vitro, 1992, vol.6, 47-52.


On the basis of experiments with dogs, the narcotic analgesic

pethidine was once thought to be non-addictive in people.1 The side

effect was not anticipated because pethidine is metabolised, or

broken down, much more quickly in dogs resulting in less exposure

to the drug. In fact, dogs metabolise pethidine more than 6 times

faster than people.2

Such differences in metabolism are the rule rather than the

exception2,3 and according to Miles Weatherall, former Director of the

Wellcome Research Laboratories, "every species has its own

metabolic pattern, and no two species are likely to metabolise a drug

identically." 4


1) B.Brodie, Pharmacologist, 1964, vol.6, 12-26.

2) R.Levine, Pharmacology: Drug Actions & Reactions (Little, Brown & Co.,1978)

3) G.Zbinden, Advances in Pharmacology, 1963, vol.2, 1-112.

4) M.Weatherall, Nature, 1982, April 1, 387-390.


Detergents are not only used for domestic and industrial cleaning. In

research aimed at increasing penetration of therapeutic drugs across

the cornea, a number of dilute detergents were assessed in the eyes

of volunteers. Although considered "generally harmless to rabbit

eyes", some caused pain and irritation in people. For instance, a

detergent called Brij 58 produced "alarming" changes to the surface

of the human eye, together with discomfort and blurred vision.1 In

rabbits Brij 58 is classified as a "non-irritant".2

A 3% solution of a similar product, Brij 35, caused delayed irritation

in volunteers but was also non-irritating to the rabbit eye, even when

undiluted.1 And although another detergent, dupanol, caused

immediate severe pain in human subjects,1 it was considered to have

only moderate effects in the eyes of rabbits.3


1) R.J.Marsh & D.M.Maurice, Experimental Eye Research, 1971, vol. 11, 43-48.

2) M.Cornelis et al, ATLA, 1991, vol.l9, 324-336.

3) L.W.Hazleton, Proceedings of the Scientifc Section of the Toilet Goods

Association, 1952, vol.17, 5-9.


Many cancer patients have suffered unnecessarily because

researchers believed large doses of anticancer drugs were

necessary for efficient treatment. The widely held view was that to be

effective in reducing tumour size, cancer chemotherapy must also be

toxic:l only then did doctors think they had given sufficient drug. The

idea was based on animal experiments l,2 yet there were early

warning signs that patients survived longer when given

comparatively nontoxic doses, even though the drugs had a smaller

effect on tumour size.3

The high dose concept has been challenged by clinical researchers.

During the 1960s, a series of statistical studies by the Rosewell Park

Memorial Institute for Cancer Research in New York, concluded that

toxicity is not necessary and can be counterproductive.2 In 1976,

London cancer specialists found that the animal data on which the

high dose concept is based, are not always valid for human patients.1

They argued that "Since patients given large doses of antineoplastic

(anticancer) agents are often at greater risk of toxicity, alternative

methods of improving the selectivity of cancer chemotherapy must

be explored."


1) M.H.N.Tattersall & J.S.Tobias, Lancet, 1976, November 13, 1073-1074.

2) I.D.Bross, Perspectives On Animal Research, 1989, vol. 1, 83-108.

3) M.A.Schneiderman & M.J.Krant, Cancer Chemotherapy Reports, 1966, vol.50, 107-112.


The fact that even chimpanzees do not develop AIDS when infected

with HIV, casts serious doubt on the validity of animal experiments.1

Some AIDS researchers seem to recognise this since vaccines

which failed to protect chimpanzees from infection with HIV, were

nevertheless tested in human trials!2 Certainly, faith in animal tests

could have serious repercussions. For instance, failure to induce

AIDS in laboratory animals has been used to support arguments

against HIV as the cause.3

Attempts to produce "animal models" of AIDS could be dangerous in

other ways. By inserting parts of the human immune system into

mice, scientists believed they had developed an animal model of

AIDS. But fears have been expressed that interaction of HIV with

viruses commonly found in mice may not only make the "model"

irrelevant to people but promote hazardous changes in the AIDS

virus. The new HIV variants could then spread in different ways,

possibly even through the air. 4


1) P.Newmark, Nature, 1989, October 19, 566-567.

2) A.S.Fauci & P.J.Fischinger, Public Health Reports, 1988, vo1.103, 230-236.

3) New Scientist, 1988, March 3, 34.

4) J.Marx, Science, 1990, February 16, 809; P.Lusso et al, Science, 1990,

February 16, 848-852.


The diuretic drug furosemide is well established in clinical practice as

a treatment for cardiovascular and kidney disease. In mice, however,

the drug produces massive liver damage. Similar effects have also

been found in rats and hamsters.1 Yet liver toxicity is not a major

problem for human patients,2 and the harmful effects in mice have

been traced to a breakdown product of furosemide which is not found

to any serious extent in the human body.3 Fortunately, the effects in

mice were reported after furosemide's safety in people had been

determined.3 Otherwise the drug may never have been introduced.

A comparison of human and animal test data shows that furosemide

is not an isolated example. At most, only one out of every four sideeffects

predicted by animal tests actually occurs in people.4 These

findings suggest that reliance on animal experiments must lead to

the rejection of potentially valuable medicines.


1) R.M.Walker & T.F.McElligott, Journal of Pathology, 1981, vol. l35, 301-314.

2) M.N.G.Dukes in Meyler�s Side Effects of Drugs, 11th edition, Ed. M.N.G.Dukes

(Elsevier, 1988).

3) M.Weatherall, Nature, 1982, April 1, 387-390.

4) A.P.Fletcher, Journal of the Royal Society of Medicine, 1978, vol.71, 693-698.



Cyclosporin is used to prevent rejection of transplanted organs and

although hailed as a major advance over existing drugs, it is not a

panacea: side-effects are common and sometimes dangerous. The

most serious hazard is kidney damage,1 an effect not predicted by

the initial animal tests.2 Ironically, kidney toxicity has been reported

in almost 80% of kidney transplant patients receiving the drug.2

Some heart transplant patients treated with cyclosporin required

dialysis because their kidneys had failed.3

Subsequent animal experiments showed that only extremely high

doses of cyclosporin could induce kidney toxicity in rats 1 although

dogs and rhesus monkeys were still unaffected.2 Researchers

believe that "...failure to produce renal dysfunction (kidney damage)

experimentally that is similar to that seen clinically may result from

species differences in metabolism."2

Although cyclosporin can prevent rejection of transplanted organs in

both animals and people, an early review of the drug found sufficient

variation in experimental results to suggest that "The

immunosuppressive effects of cyclosporin have ...differed

considerably between species, limiting any direct inference that may

be made regarding use in human organ transplantation..."1


1 ) D.J.Cohen et al, Annals of Internal Medicine, 1984, vol. 101, 667-682.

2) W.M.Bennett & J.P.Pulliam, Annals of Internal Medicine, 1983, vol.99, 851-854.

3) Lancet, 1986, February 22, 419-420.


In September 1983, the antidepressant zimelidine (Zelmid) was

withdrawn worldwide following potentially serious side-effects

including nerve damage, leading to loss of sensation or paralysis.1

Some patients also suffered hypersensitivity reactions such as fever,

headache, muscle or joint pains, and liver problems. The drug had

been introduced only a year earlier but Britain's Committee on Safety

of Medicines had received over 300 reports of adverse reactions, 60

of which were serious: there were 7 deaths.2 Prolonged tests in rats

and dogs had shown no evidence of toxicity at 5 times the human



1) B.Blackwell in Side Effects of Drugs Annual, vol.8, Eds. M.N.G.Dukes & J.Elis

(Elsevier, 1984).

2) R.D.Mann, Modern Drug Use, an Inquiry on Historical Principles (MTP Press, 1984).

3) R C Heel et al, Drugs, 1982, vol 24, 169-206


In 1984 a Milan Poison Control Centre reported 32 patients with

severe neurological side-effects following an overdose of zipeprol,

the cough suppressant.1 Symptoms included seizures and coma,

and the Centre stated that "Zipeprol should be much more strictly

controlled..." Animal tests had given no warning of severe

neurological problems despite the use of higher doses.2


1) C.Moroni et al, Lancet, 1984, January 7, 45.

2) D.Cosnier et al, Drug Research, 1976,vol.26, 848-854; G.Rispat et al, Drug

Research, 1976, vol.26, 523-530.


Evicromil (code name FPL 52757) was submitted for clinical trial as

an antiasthmatic drug following safety evaluation in mice, rats,

hamsters, rabbits, ferrets, squirrel monkeys, cynomolgus monkeys,

stump-tail monkeys and baboons. Despite using doses many times

greater than the amount intended for human use, no harmful effects

were seen, especially with respect to the liver.1 Yet 20% of patients

participating in the trial had symptoms of liver damage, precluding

any further development of the drug.2 Subsequent tests showed that

liver toxicity could only be induced in dogs.1,2


1) D.V.Parke in Animals & Alternatives in Toxicity Testing, Eds. M.Balls et al

(Academic Press, 1983).

2) C.T.Eason et al, Regulatory Toxicology & Pharmacology, 1990, vol.11, 288-307.


During clinical trials, ICI's arthritis drug fenclozic acid unexpectedly

produced jaundice in some of the patients. Researchers were

surprised since tests with rats, mice, dogs and monkeys had given

no hint of liver problems.1 Not content with these results, further

experiments with rabbits, guinea pigs, ferrets, cats, pigs, horses,

neonatal rats and mice, together with a different strain of rat, were

carried out but still no evidence of liver damage could be found.1 The

ICI researcher commented that "The quite unexpected onset of

jaundice in a few patients caused withdrawal of the drug from

humans and initiated a vast programme of experimental work. This

search for hepatotoxicity (liver damage) in different species or any

indication of its likelihood has so far been unrewarding."1


1) S.J.Alcock, Proceedings of the European Society for the Study of Drug Toxicity,

1971, vol.12, 184-190.


In 1985 Britain's Committee on Safety of Medicines issued a special

warning of serious liver damage associated with antifungal drug

ketoconazole (Nizeral)1. The Committee cited 82 cases with 5

deaths. The warnings followed similar action by the US Food and

Drug Administration in 1982.2 Doctors are advised to monitor their

patients carefully and perform regular liver function tests throughout

treatment with ketoconazole. No evidence of liver toxicity had been

found in the original animal tests.3


1) Lancet, 1985, January 12, 121.

2) C.B.M.Tester-Dalderup in Meyler's Side-Effects of Drugs, 11th edition, Ed.

M.N.G.Dukes (Elsevier, 1988).

3) J.K.Heiberg & E.Svejgaard, British Medical Journal, 1981, September 26, 825.


Reliance on animal experiments rather than human observations

delayed a full realisation that lack of food early in life can harm the

brain. During the first quarter of the 20th century, there was

considerable interest in the possibility that lack of food during

childhood might interfere with the proper development of the brain

and therefore affect later achievement of the individual.

Unfortunately, almost all the research was carried out on animals and

showed that starving baby or adult rats had no effect on the brain.

Not surprisingly, the topic was abandoned and only resumed in the

late 1950s when children with histories of undernutrition were

persistently found to underachieve, both in school and in formal


Researchers then realised that the early animal tests had failed since

no account had been taken of the "brain growth spurt". This is the

period of fastest growth when the brain is at its most vulnerable.

Furthermore, the exact timing varies between the species: in human

babies the brain growth spurt begins during the final stage of

pregnancy and proceeds through to at least a year; in guinea pigs, it

occurs almost entirely during the foetal period; and in rats it happens

during the first 3 weeks after birth.2

Despite millions of underfed and malnourished people,"early life

undernutrition" remains a popular subject among animal

researchers. Unlike current aid levels to developing nations, there

seems no shortage of funds for such research: indeed, one

justification is that, someday, it might better enable us to give relief to

the starving!3


1) J.Dobbing in Early Nutrition & Later Behaviour, Ed. J.Dobbing (Academic Press, 1987).

2) J.Dobbing & J.L.Smart, British Medical Bulletin, 1974, vol.30, 164-168.

3) J.L.Smart in ref.1.


Practolol (Eraldin), marketed by ICI during the early 1970s for the

treatment of heart conditions, was "particularly notable for the

thoroughness with which its toxicity was studied in animals, to the

satisfaction of the regulatory authorities."1 Nevertheless, unforeseen

side-effects began to emerge including serious skin, eye and

abdominal problems. Some patients suffered dry eyes, conjunctivitis

and corneal damage leading to blindness. There were also cases of

stomach damage with obstruction of the intestine, a condition known

as sclerosing peritonitis which led to 23 reported deaths.2 Overall, ICI

compensated more than 1000 victims.3

The "practolol syndrome" had not been predicted by animal

experiments 4 and even after the drug was withdrawn in 1976, no one

could replicate the harmful effects in laboratory animals.1


1) M.Weatherall, Nature, 1982, April 1, 387-390

2) G R.Venning, British Medical Journal, 1983, January 15, 199-202; January 22, 289-292.

3) A Question of Balance, Office of Health Economics, 1980.

4) F.H Gross & W.H.lnman (Eds.), Drug Monitoring (Academic Press, 1977).


One of the most serious side-effects of steroid eye therapy is

glaucoma. An abnormally high pressure builds up within the eye and

can lead to permanent loss of vision if the effects are prolonged.

During the early 1950s, when corticosteroids were first employed in

ophthalmology, animal tests suggested that cortisone had no effect

on pressure within the eye.1 Subsequent attempts to induce

glaucoma in rabbits and monkeys proved difficult or impossible,2 and

researchers at Britain's Porton Down laboratories refer to "the

differing response of the eye of man and animals to repeated topical

(surface) application of corticosteroids. Such a procedure is without

effect on tension of the eye of many experimental mammals, but

increases tension in the human eye."3

Another side-effect of steroid therapy that is difficult to replicate in

laboratory animals is cataract. Although scientists have produced

slight changes in the lens of the rabbit's eye after repeated

application of high doses, they did not mimic the more serious

condition found in human patients.2


1) L H.Leopold et al, American Journal of Ophthalmology, 1951, vol.34, 361-371.

2) W.M.Grant,Toxicology of the Eye,2nd edition,(Charles Thomas, 1974).

3) B.Ballantyne & D.W.Swanston in Current Approaches in Toxicology, Ed.

B.Ballantyne (Wright Sons, 1977).


In 1991, doctors at Southampton General Hospital warned that

inhaling babies' talcum powder could be fatal,1 representing "an

unappreciated hazard." They state that "talcum powder can cause

severe respiratory symptoms in infants: its use should be

discouraged and containers should carry a warning and have child

proof caps." Eight deaths have been attributed to inhalation of talc.

Concerns over the safety of talc have been raised before and studies

of talc miners and millers have shown that it can damage the lungs.2

But experiments in which huge amounts of the commercial product

were administered to animals, seemed to suggest no hazard to

consumers. For instance, in 1977 experimenters exposed hamsters

to high grade cosmetic talc at doses nearly 2000 times higher than

that experienced by babies during toilet care. There was no effect on

survival or damage to the lungs.3 In the same year, other scientists

forced rats to breathe talc at doses approaching 6000 times those

used in baby care. Despite the massive amounts, there was only a

slight effect on the lungs.3


1) P.W.Pairaudean et al, British Medical Journal, 1991, May 18, 1200-1201.

2) A.Seaton in Occupational Lung Diseases, Eds. W.K.Morgan & A.Seaton (Saunders, 1982).

3) Lancet, 1977, June 25, 1348-1349.


Studies with human volunteers have shown that animal experiments

can seriously underestimate the likely effect of riot control gases on

the eye. The tests found that people are 18 times more sensitive to

CS than rabbits, and 90 times more sensitive to another sensory

irritant, CR.1

When applied to the rabbit's eye, a solution of CR produced only

"minor transient changes" in pressure within the eye. But instillation

of a smaller amount into the human eye produced a 40% rise in

pressure within 5 minutes compared with only a 3% rise after 10

minutes in rabbits.2

Species differences have also been found when CS and CR are

applied to the skin. A method known as the human blister-base

technique allows volunteers to classify irritants according to the level

of discomfort they produce. The procedure showed that CR is a more

potent irritant than CS which is confirmed by other human test

systems, yet is the reverse of that found from experiments on

rodents.3 The study also found that a further sensory irritant,VAN, is

less potent than CR which is again the opposite of that found from

animal tests. In a masterpiece of understatement, the researchers

conclude that "data derived from humans thus appears to be of

importance when assessing irritant potency."3


1) D.W.Swanston in Animals & Alternatives in Toxicity Testing, Eds. M.Balls et al

(Academic Press 1983).

2) B.Ballantyne et al in Current Approaches in Toxicology, Ed. B.Ballantyne

(Wright & Sons, 1977).

3) R.W.Foster et al, Pain, 1986, vol.25, 269-278.



Widespread fluoridation is thought to be a key factor in the decline of

dental caries, and originated from a dentist's observation that

children with mottled teeth, caused by a naturally high concentration

of fluoride in the water supply, seemed to have less tooth decay than

usual.1 Although fluoride has been added to public water supplies for

over 3 decades with apparently no ill effects,2 experiments with

laboratory rats raised concerns that it may cause cancer.3 The

findings prompted an in-depth analysis of over 50 human health

studies conducted over the previous 40 years. The resulting report

by America's Department of Health and Human Services (DHHS)

found no evidence of a link with cancer and gave fluoride the "allclear."

Animal tests have indicated other harmful effects of fluoride but the

DHHS report notes that sensitivity to fluoride varies widely between

species, making results difficult to apply to humans.4


1 ) A.M.Lilienfield & D.A.Lilienfield, Foundations of Epidemiology (Oxford University

Press, 1980).

2) Eg. see R.Peto & R.Doll, The Causes of Cancer (Oxford University Press, 1981).

3) Journal of NIH Research, 1991, vol.3, 46.

4) C.Anderson, Nature, 1991, February 28, 732.



Coumarin is a naturally occurring product derived from the Tonka

bean and has been employed for over 100 years. It is used in

consumer products and as a therapeutic agent, for instance in the

treatment of cancer and infectious disease. During the 1950s, doubt

was cast on coumarin's safety when experiments produced liver

damage in laboratory rats. As a result, coumarin was banned as a

food flavouring agent.1

Subsequent research, however, showed wide species variation in

response to coumarin. While dogs also suffer liver toxicity, there are

only minimal effects in baboons.2 And doses which damage the rat's

liver are harmless to gerbils.1 Even different strains of the same

species react differently and coumarin is less harmful to the DBA/2J

strain of laboratory mouse than the CH3/HeJ strain.3

Among patients receiving relatively high doses of coumarin for

therapeutic purposes, liver toxicity is said to be "very rare,"1 and rats

and dogs are now considered poor "models" for assessing the drug

because they metabolise coumarin in a completely different way. 1,2,3


1) J.H.Fentem et al, Toxicology, 1992, vol.71, 129-136.

2) J.G.Evans et al, Food & Cosmetic Toxicology, 1979, vol. 17, 187-193.

3) W.Endell & G.Seidel, Agents & Actions, 1978, vol.8, 299-302.



Fears for the safety of formaldehyde workers followed reports that

the chemical causes cancer in rats.1 Formaldehyde is widely used as

a laboratory fixative and as an embalming fluid but human

epidemiological studies had revealed no evidence of cancer. The

animal tests led to further observations of exposed workers but these

were also negative.1

The rats had been forced to breathe such high doses (7-15 times that

inhaled by workers) that the formaldehyde caused tissue damage

which led to the cancers. Nevertheless, "...there are still some who

believe that the positive results in the rat are the dominant factor to

be taken into account and overrides the epidemiology but there is

always some hope that common sense may prevail."1


1) P.Grasso, Journal of the Royal Society of Medicine, 1989, vol. 82, 470-473.


Scientists have devised more than 50 ways of inducing fits in

laboratory animals. One reason for the large number is that "none of

the models is fully trustworthy as an imitation of clinical epilepsy,"1

and indeed results can vary depending on the "model" chosen.

An example is the artificial sweetener aspartame. In research

sponsored by the NutraSweet Company and the Wellcome Trust,

researchers at London's Institute of Psychiatry carried out

experiments with photosensitive baboons in which fits are induced by

flashing lights. The tests followed suggestions that high doses of

aspartame may produce seizures in sensitive people. Aspartame

had no effect in the baboons but conflicting data has been found in

other animal models: aspartame enhances chemically-induced

convulsions in mice, for instance, but has no effect on electric shockinduced

or sound-induced seizures in these animals.2

Similar species differences are found in drug development. Although

reducing convulsions in mice and baboons, the drug THIP proved

ineffective when tried in patients with epilepsy.3


1) R.S.Fisher, Brain Research Reviews, 1989, vol.l4, 245-278.

2) B.S.Meldrum et al, Epilepsy Research, 1989, vol.4, 1-7.

3) Lancet, 1985, January 26, 198-200.


In 1991 the US Occupational Safety and Health Administration

decided that glass fibre products should be labelled as a potential

cancer hazard.1 The decision followed studies of glass fibre workers

that showed an increased risk of lung cancer.

Glass wool products have been manufactured for about 60 years

during which animal experiments seemed reassuring. In the 1950s,

experiments with rats, guinea pigs, rabbits and monkeys produced

no lung damage when the animals were forced to breathe the fibres.2

And an analysis of further tests conducted during the 1980s noted

that "An increase in lung tumours or mesothelioma has not been

observed following long-term inhalation studies in several animal

species including rats, hamsters, guinea pigs, mice, monkeys, and

baboons exposed to glass fibres, glass wool or mineral wool."3

Ironically, experiments in which rats did develop cancer have been

dismissed as unlikely to have any relevance to the human condition.

This is because the glass fibres were artificially implanted into the

tissue membrane lining the animal's lung, whereas in people the

usual means of exposure is through breathing. Furthermore, it is well

known that rats are especially prone to cancer when solid

substances are surgically implanted into their bodies.2 In his book

Occupational Lung Disorders, Raymond Parkes concludes that "the

production of malignant tumours in animals by direct implantation

experiments is unlikely to have any relevance to human exposure."


I ) Letter from G.F.Scannell, Assistant Secretary for Occupational Safety and Health,

Washington DC, to Richard Munson, Chairman of Victims of Fibreglass (May

6,1991); The Guardian, July 20, 1991.

2) Reported in R.Parkes, Occupational Lung Disorders (Butterworths, 1982).

3) C.S.Wheeler, Toxicology & Industrial Health, 1990, vol.6, 293-307.


During the 1960s, doctors noticed that women receiving the steroid

drug Depo-Provera as a treatment for premature labour, experienced

a delay in the return of fertility after the birth of their babies. The

observation led to clinical trials of the drug as a possible long-acting

contraceptive.1 Injectable preparations of Depo-Provera are now

known to be as effective as oral contraceptives and are available in

Europe, Asia, Africa and the Far East. In America however, approval

was delayed for many years.2

Much of the controversy surrounding Depo-Provera relates to

experiments with beagle dogs that indicated a host of disturbing sideeffects.

2 There were abnormal growth problems, cases of breast

cancer, and many animals died of pyometra, a condition in which pus

accumulates in the uterus. None of these effects have been

observed in women taking Depo-Provera1,2 and scientists point to

physiological differences between human beings and dogs which

make beagles especially sensitive to certain kinds of steroids.1

High doses of Depo-Provera can also cause cancer in monkeys but

again their relevance has been questioned since the tumours arise

from a type of cell not found in women. Furthermore, the kind of

cancer produced in monkeys is successfully treated by Depo-

Provera in women!1

In l991 an editorial in the Lancet entitled "DMPA (Depo-Provera) and

breast cancer: the dog has had its day", argued that "Countries such

as the USA, Australia and Japan would do well to reassess their

existing policies on injectable preparations, otherwise they may

deprive their female citizens of a reliable, effective and safe method

of contraception."2 One year later, America's Food & Drug

Administration finally decided to approve Depo-Provera as a long

acting contraceptive.


1) Bulletin of the World Health Organisation, 1982, vol.60, 199-210.

2) Lancet, 1991, October 5, 856-857.


In 1939, animal researchers devised an astonishing treatment for

silicosis, the debilitating lung disease caused by exposure to silica

dust. They found that inhalation of metallic aluminium could prevent

silicosis in laboratory rabbits,1 and from the early 1940s to the mid-

1950s, the technique was widely employed by industry in an attempt

to treat or prevent the condition amongst workers.2

Before beginning work, men whose occupations exposed them to

silica, passed through an aluminium dusting chamber where they

breathed a daily dose of the powder. But in 1956, studies of pottery

workers showed that the method did not work and the Industrial

Pulmonary Disease Committee of Britain's Medical Research

Council recommended that the technique should not be used.3

Today we know that the treatment itself carried risks. Although large

doses of aluminium proved harmless to animals,4 cases of lung

damage and cancer have been reported amongst aluminium

workers.5 Furthermore, studies of Canadian miners who breathed

aluminium powder to prevent silicosis, have revealed symptoms

consistent with the current theory that aluminium may cause

Alzheimer's Disease!6


I) J.J.Denny et al, Canadian Medical Association Journal, 1939, vol.40, 213: reported in ref 3.

2) W.R.Parkes, Occupational Lung Disorders (Butterworths,1982).

3) M.C.S.Kennedy, British Journal of lndustrial Medicine, 1956, vol.13, 85-101.

4) L.U.Gardner et al, Journal of Industrial Hygiene & Toxicology, 1944, vol.26, 211-223.

5) M.J.Ellenhorn & D.G.Barceloux, Medical Toxicology (Elsevier, 1988).

6) Lord Walton of Detchant, Journal of the Royal Society of Medicine, 1992, vol.85, 69-70.



For decades animals have been deliberately poisoned to death in

lethal dose (LD50) toxicity tests, yet the results are of little value in

the prevention and treatment of overdose patients and can be

misleading.1 According to their lethal doses in rats, aspirin would

seem safer than another common pain killing drug, ibuprofen. In fact,

human overdose experience reveals that ibuprofen is the safer drug.1

As physicians at London's National Poisons Centre point out, "The

�natural experiment' of cases of self poisoning has to be taken as the

starting point as the results of experiments on animals cannot reliably

be extrapolated to man..." 2


1) G.N.Volans in The Contribution of Acute Toxicity Tests to the Evaluation of

Pharmaceuticals, Eds.D.Schuppan et al (Springer-Verlag, Berlin, 1986).

2) S.Cassidy & J.Henry, British Medical Journal, 1987, October 24, 1021-1024.



Comparisons of people living in different countries, together with

other human studies, have shown that too much fat in the diet can

lead to cancer of the colon, with saturated fat the chief culprit. Animal

tests agree that too much fat can be dangerous but suggest it is the

polyunsaturated fats that are mostly to blame.1

Clinical studies have also suggested that a high fibre diet is

beneficial and the idea has been tested by animal researchers. Again

the results are conflicting, some experiments showing a reduced risk

of cancer and others an increased risk.2 And although population

studies have identified diets high in animal protein as most risky,3

much laboratory research suggests that the type of protein is


Human studies have consistently shown that diets rich in fruit and

vegetables can protect against colon cancer. In contrast, many of the

natural substances evolved by fruit and vegetables to protect

themselves from predators and parasites, actually cause cancer

when tested in rats and mice!4


1) J.L.Freudenheim & S.Graham, Epidemiologic Reviews, 1989,vol.11,229-235.

2) D.Galloway, Cancer Surveys, 1989, vol.8, 169-188.

3) B.Armstrong & R.Doll, International Journal of Cancer, 1975, vol.l5, 617-631.

4) P.H.Abelson, Science, 1990, September 21, 1357.


Although prednisone is a valuable drug for the treatment of leukemia

and other human cancers, it fails to work in a range of animal

tumours including two different types of leukemia in mice.1 Ironically,

these experimental cancers were once commonly used by America's

National Cancer Institute in attempts to identify promising new drugs!

Prednisone can be even more effective when used in conjunction

with certain other anticancer drugs but once again animal tests have

proved misleading: of 6 drug combinations showing an improved

clinical effect, only one was correctly predicted by animal


The incentive to develop prednisone stemmed from encouraging

results with the closely related steroid, cortisone, a hormone derived

from the adrenal gland. In 1930 Californian physicians claimed they

had cured human cancers with extracts of adrenal gland.

Unfortunately, these findings led to animal rather than human trials

and when the former proved negative, the treatment was

abandoned.2 Only when the tests were repeated, a decade later, did

researchers confirm that adrenal extracts could be beneficial against

some forms of cancer. Promising human trials led to the

development of analogues such as prednisone.


1) R.J.Johnson & A.Goldin, Cancer Treatment Reviews, 1975, vol.2, 1-31.

2) B.Reines, Cancer Research on Animals: Impact and Alternatives (NAVS,

Chicago, 1986).


Pro-vivisection organisations often cite the conquest of polio as a

triumph of animal experiments. In fact, emphasis on animal research

rather than human studies delayed a proper understanding of the

disease for over 25 years.1

In 1908 Landsteiner and Popper announced they had discovered the

polio virus: tissue from an infected patient who had died produced

spinal cord disease when injected into two monkeys. The animals

died with one developing paralysis in both legs. Negative results had

been obtained with rabbits, guinea pigs and mice and indeed the

researchers had been "fortunate" in choosing Old World monkeys

such as the rhesus who are highly susceptible to the disease. New

World monkeys are relatively resistant.

Although obviously important to discover the virus, the means by

which it was achieved had a devastating impact on polio research.

Believing they now had an exact replica of the human infection,

scientists focussed their main attention on the artificially induced

disease in monkeys. Based on these experiments it was generally

believed that poliovirus enters the body through the nose and that it

only attacks the central nervous system (CNS) producing spinal cord


Yet by 1907 careful epidemiological analysis of actual human cases

had shown that poliomyelitis was not entirely or even chiefly a

disease of the CNS. The studies, based on over 1000 Swedish

cases, were carried out by Dr. Ivar Wickman, who also correctly

concluded that the gastrointestinal tract was the probable route of

infection.1 By 1912 other clinical studies also established the

intestinal tract as the means of infection.

Tragically, animal experiments so dominated research that prior to

1937 most scientists rejected the notion that polio is an intestinal

disease. As Dr Paul explains in A History of Poliomyelitis, "...with the

discovery of the virus and the rush of enthusiasm for experimental

work, the mainstream had soon been diverted away from Wickman's

correct concepts of the human disease gained from clinical

epidemiological work carried out so painstakingly in the field."

Whether the virus entered the body by the mouth or nose was of

great practical importance, for it determined strategies for preventing

the spread of disease. By 1937 researchers had produced a nasal

spray that prevented infection in monkeys. It was widely promoted for

human use but inevitably failed.1 The only result was to abolish the

children's sense of smell, in some cases permanently.2

Support for the nasal route of infection gradually waned after further

clinical observations while animal researchers were no doubt

reassured with the finding that chimpanzees, unlike the rhesus

monkeys used earlier, could be infected via the intestinal tract. It was

only when scientists understood that poliovirus enters the mouth and

first resides in the intestines that it was possible to develop an orally

administered vaccine, and this formed the basis of Sabin's approach.

For years, monkeys were also used for diagnostic purposes, to test

for the presence of virus. Tissue samples from patients (or other

monkeys) were innoculated into the animals who were then

assessed for damage to the spinal cord. The procedure was

laborious, time consuming and expensive but during the early years

of the 20th century, scientists had only learnt to grow viruses in living

animals. However, in 1949 Enders, Weller and Robbins showed that

polio virus could be grown in human tissue culture. Most significantly,

the virus produced a specific change in the infected cells which could

be recognised under the microscope. It was therefore easy to detect

the presence of polio virus in tissue samples. Had such a quick and

simple alternative been introduced at an earlier stage, progress

would surely have been more rapid. Indeed, by suggesting that virus

would only grow in the CNS, the misleading monkey model of polio

delayed the development of tissue culture techniques which were

ultimately critical to the discovery of a vaccine.1,3


1) J.R.Paul, A History of Poliomyelitis (Yale University Press,1971).

2) H.F.Dowling, Fighting lnfection (Harvard University Press, 1977).

3) A Critical Look at Animal Research (Medical Research Modemization

Committee, New York, 1990).


Years of experimentation have taught scientists that guinea pigs and

hamsters are especially sensitive to the harmful effects of antibiotics.

For instance, widely prescribed human antibiotics such as ampicillin,

amoxycillin and oxytetracycline are considered "toxic" and therefore

inappropriate for use in these species.1 Another example is

erythromycin where the usually recommended human dose is

enough to kill a hamster!2

Today, "it is generally recognised that the guinea pig is peculiarly

sensitive to the lethal effects of antibiotics,"3 but this was not always

realised. In his book Drug Development: From Laboratory to Clinic,

Dr Walter Sneader describes how "it was fortunate that Florey and

Chain did not decide to use guinea pigs when first testing penicillin,

for they may then have abandoned the project as these animals are

hypersensitive to penicillin." Florey and Chain were the Oxford

scientists who carried out animal tests following Fleming's discovery

of penicillin. Florey later commented "...mice were tried in the initial

toxicity tests because of their small size, but what a lucky chance it

was, for in this respect man is like the mouse and not the guinea pig.

If we had used guinea pigs exclusively we should have said that

penicillin was toxic, and we probably should not have proceeded to

try to overcome the difficulties of producing the substance for trial in



1) A.A.Tuffery (ed.), Laboratory Animals - An lntroduction for New Experimenters (Wiley,


2) A single, minimum recommended dose of erythromycin is 250-5OOmg every 6 hours

i.e. 3.5-7.0 mg/kg for a 70kg person. The lethal dose for hamsters is 3.5mg/kg (ref.3)

3) S.J.Desalva et al, Toxicology & Applied Pharmacology, 1969, vol.14, 510-514.

4) H.Florey, Conquest, January, 1953.



Prevention is always better than cure, particularly for diseases like

cancer where treatment can be both difficult and unpleasant. But

first, doctors must discover the causes so people know how to avoid

ill-health. This is the primary role of epidemiology - the study of

disease in human populations. Tragically, a preference for laboratory

research and animal experiments diverted attention from

epidemiology, and for decades little was known about the main

causes of human cancer.

Before World War I, epidemiology had identified several causes of

the disease.1 For instance, pipe smokers were more likely to develop

cancer of the lip; workers in the aniline dye industry often contracted

bladder cancer; and skin cancer was an occupational hazard of

radiologists. It was also known that combustion products of coal

(soot and tar) could cause the disease, an observation dating back

to 1775 when the English surgeon Potts identified soot as a

carcinogen in chimney sweeps.

Attempts to replicate Potts' findings in laboratory animals repeatedly

failed2 but finally, in 1918, Japanese researchers reported that

cancer could be produced on a rabbit's ear by continually painting it

with tar, a discovery that changed the course of cancer research.

According to the renowned British epidemiologist Sir Richard Doll,

human observational data were now commonly dismissed because

it was confidently assumed that laboratory experiments held the key

to success.1 Crucial epidemiological studies like those of Percy

Stocks at London University, who reported in 1933 that people

consuming larger amounts of fruit and vegetables were less likely to

develop cancer,3 received little attention,1 yet today we know that

Stocks was right.4

The absence of human epidemiological data allowed mistaken ideas

based on animal research to flourish. Although we now know that

only about 5% of Western cancers are linked to viral infection,5 some

scientists believed that most, if not all cases vvere caused by viruses,

a view derived from experiments on animals where it is easy to

transmit the disease in this way.6 One animal researcher even

argued that women should not breast feed their babies: he believed

that in humans, as in mice, a virus is the prime cause of breast

cancer, and that the virus is acquired in the mother's milk! 7

Following World War II, interest in epidemiology was reawakened

with the striking discovery that smoking causes lung cancer. This

breakthrough led to further population studies which identified the

causes of many other types of'cancer. The result is that 80-90% of

cases are now considered potentially preventable. And it is revealing

that the 1980 US Congress Office of Technology Assessment Report

on the causes of cancer, relied far more on epidemiology than

laboratory tests because these "cannot provide reliable risk

assessments." 5


1) R.Doll, Cancer 1980. vol.45 2475-2485

2) W.H.Woglom, Archives of Pathology, 1926 vol 2. 533-576

3) P.Stocks & M.N.Karn, Annals of Eugenics, 1933, vol 5, 237-280.

4) J.Robbins, Diet for a New America (Stillpoint. 1987).

5) R.Peto & R.Doll, The Causes of Cancer (Oxford University Press, 1981).

6) E Northrup, Science Looks at Smoking (Conard-McCann, 1957).

7) J.Furth, Bulletin of the New York Academy of Medicine, 1964, vol.40,421-431.


In a meticulous study at the Vienna General Hospital, Ignaz Phillipe

Semmelweiss discovered that expectant mothers were more likely to

die of childbed (puerperal) fever if their attendants had previously

been working in the dissecting and post-mortem rooms. The disease,

he reasoned, must be caused by an infection carried from the

dissecting room on the hands of doctors and students. When

Semmelweiss insisted on strict hygiene, the death rate promptly

dropped from 1 in 8 confinements to 1 in 100.1

Tragically, the Hospital professors responded with such hostility that

Semmelweiss was forced to leave. Only 4 years earlier, in 1843, the

American researcher and humanitarian Oliver Wendell Holmes had

reached the same conclusion by careful observation, but had been

similarly villified. According to medical statistician Dr Sigmund Peller,

"In a world that had not been stultified by the idea that only animal

experimentation and only the laboratory can provide proof in matters

of human pathology, the battle against puerperal fever would not

have needed to wait for the discovery of cocci (the responsible

bacterium,discovered during the 1860s). The experts who, during the

1840s, opposed and prevented the initiation of a rational programme

for combatting the disease should have been charged with a

negligence that resulted in mass killings. But they were not."2

Proper recognition of Semmelweiss and Holmes, and the central role

of cleanliness, would surely have hastened the introduction of

lifesaving, hygenic measures in surgery. But these had to wait at

least another 20 years until Lister developed his antiseptic



I) R Sand, The Advance to Social Medicine (Staple Press, 1952)

2) S.Peller, Quantitative Research in Human Biology (J.Wright & Sons, 1967).


The arthritis drug Opren (Oraflex in the US) was withdrawn from the

world market in August 1982 following British reports of deaths and

serious liver damage in people taking the drug.1 Since 1980, when

Opren was first introduced in the UK, there had been 3,500 reports

of harmful effects with 61 deaths, mainly in elderly patients.2

Scientists list Opren as a drug whose injuries were not predictable

from animal tests,3 and note that "despite searching preclinical

animal toxicity studies...administration to rheumatoid patients

resulted in adverse reactions including onycholysis (nail damage)

and skin phototoxicity (light sensitivity) and finally in fatal

hepatotoxicity (liver damage) whereupon the drug was withdrawn."

And Dista, the subsiduary of Eli Lilly who marketed the drug in

Britain, stated in their literature that "the effects of benoxaprofen

(Opren) in the rhesus monkey were studied for one year...There were

no apparent adverse effects on survival."

Researchers believe that the fatal cases of liver damage might have

been averted by more extensive clinical trials,3 especially in the

elderly who take much longer to eliminate Opren from the body than

either young people or laboratory animals.


1) E.M.B.Sorensen, Toxicology Letters, 1986, vol.34, 277-286.

2) British Medical Journal, 1982. August 14, 459-460.

3) C.T.Eason et al, Regulatory Toxicology & Pharmacology, 1990, vol.11, 288-307.


Following experiments on rabbits, dogs, gerbils and monkeys,

animal researchers suggested that barbiturates could provide

protection against the effects of a stroke.1 In human stroke victims,

however, barbiturates had little or no protective effect.2 This failure of

animal tests is not an isolated example: between 1978 and 1988, 25

drugs were found useful in treating animals with artificially-induced

stroke yet none has come into general clinical use.2

Stroke researchers are divided over the relevance of animal

experiments3 and some argue that "over-reliance upon such (animal)

models may impede rather than advance scientific progress in the

treatment of this disease... Each time one of these potential

treatments is observed to be effective based upon animal research,

it propagates numerous further animal and human studies

consuming enormous amounts of time and effort to prove that the

observation has little or no relevance to human disease or that it may

have been an artifact of the animal model itself."2

Although defending the role of animal experiments, researchers at

the Mayo Clinic conclude that "Ultimately... the answers to many of

our questions regarding the underlying pathophysiology and

treatment of stroke do not lie with continued attempts to model the

human situation perfectly in animals but rather with the development

of techniques to enable the study humans."2


1) Stroke, 1975, vol.6, 28-33; Stroke, 1974, vol.5, 1-7; Neurology, 1975, vol 25,

870-874; Stroke, 1972, vol.3, 726-732; Annals of Neurology, 1979, vol 5, 59-64.

2) D.O.Wiebers et al, Stroke, 1990, vol.21, 1-3.

3) C.Millikan, Stroke, 1992, vol.23, 795-797.


A new animal test raised fears that Astra's ulcer treatment,

omeprazole, may cause stomach cancer. In the test, developed by

Glaxo pharmaceutical company, rats are dosed with the suspect

drug or chemical, after which tissue samples are removed from the

animal's stomach and analysed for effects on DNA, the substance

which controls proper development of the cells. Interference with

DNA is regarded as a possible first step to cancer.

The experiments showed that omeprazole damaged the DNA but

that ranitidine, Glaxo's own antiulcer drug, did not.1 On the basis of

these results, Glaxo halted comparative clinical trials of ranitidine

(Zantac) and omeprazole, an action, according to the Lancet, that

seemed certain to influence prescribing habits.2

In response, Astra, the makers of omeprazole, argued that "the

method used by Glaxo is scientifically unsound and the results

therefore have no clinical consequences."3 They noted that "long

term studies in which omeprazole was administered for up to 2 years

in rats, 18 months in mice, and 1 year in dogs yielded no evidence

for a direct carcinogenic potential, in the stomach or elsewhere."


1) B.Burlinson et al, Lancet, 1990, February 17, 419.

2) Lancet, 1990, February 17, 386

3) L.Ekman et al, Lancet, 1990, February 17, 419-420.


A US study has found that two drugs designed to prevent irregular

heart beats can actually cause heart attacks in certain types of

patient. The cardiac arrhythmia suppression trial (CAST) began in

June 1987 but was halted in April 1989 when doctors found more

deaths among patients treated with encainide and flecainide than in

those receiving a placebo (dummy pill).1 Based on the findings it has

been estimated that, nationwide, 3000 people may have died

prematurely after taking the drugs.2

In contrast, the animal research had indicated that encainide and

flecainide were both safe and effective.3


1) CAST Investigators, New England Journal of Medicine, 1989, August 10, 406-412.

2) Dr. J.Morganroth reported in Washington Times, 1989, July 26.

3) Flecainide: B.Holmes & R.C.Heel, Drugs, 1985, vol.29, 1-33; encainide: D.C.Harrison

et al, American Heart Journal, 1980, Vol.100, 1046-1054, and J.E.Byrne et al, Journal of

Pharmacology & Experimental Therapeutics, 1977, vol.200, 147-154.


Although olive oil has been used to anoint the human body for

thousands of years without any apparent ill effects,1 tests carried out

at New York University showed that olive oil actually had a harmful

effect when applied to the skin of rats, causing swelling, proliferation

of cells and a great shedding of large, loose flakes of skin! 2


1) M.M.Rieger & G.W.Battista, Journal of the Society of Cosmetic Chemists, 1964,

vol.15, 161-172.

2) E.O.Butcher, Journal of Investigative Dermatology, 1951, vol.16, 85-90.


Rabbits and guinea pigs are commonly used to assess irritancy but

neither provides an accurate model for human skin.1 For instance,by

the criterion of animal experiments, hypochlorite bleach should be

considered comparatively safe for human use, since it only produces

"slight visible irritation" in rabbits and guinea pigs.1 However, in

human volunteers bleach causes severe skin reactions.


I) G.A.Nixon et al, Toxicology & Applied Pharmacology, 1975, vol.31, 481-490.


In 1954, Richard Doll and Bradford Hill published their famous

investigation into the smoking habits of British doctors which clearly

revealed that the chances of developing lung cancer increased with

the number of cigarettes smoked.1 More than a dozen similar

(human) studies had already been published but some scientists still

argued that the link between smoking and lung cancer was

unwarranted since no-one had produced the disease in laboratory


Two years after publication of Doll and Hill�s findings, the British

Empire Cancer Campaign - the forerunner of today's Cancer

Research Campaign - reported nearly two years of experiments

during which mice, rabbits and other animals were exposed to

tobacco derivatives by direct inhalation, feeding, injection into the

lungs, and skin painting. None developed cancer.3 And in 1957,

American pathologist Eric Northrup concluded in his book Science

Looks at Smoking that the "....inability to induce experimental

cancers, except in a handful of cases, during 50 years of trying, casts

serious doubt on the validity of the cigarette-lung cancer theory."

Health warnings were delayed for years and Northrup describes how

"it is reassuring...that public health agencies have rejected the

demand for a mass lay educational programme against the alleged

dangers of smoking. Not one of the leading insurance companies,

who consider health hazards in terms of monetary risk, has raised

the life insurance rates for heavy smokers."

Despite years of further experimentation, it has proved "difficult or

impossible" to induce lung cancer in animals using the method

(inhalation) by which people are exposed to the smoke.4


1) R.Doll and A.B.Hill, British Medical Journal, 1954, June 26, 1451-1455.

2) Reported in S.Peller, Quantitative Research in Human Biology (J.Wright &

Sons, 1967).

3) Reported in E.Northrup, Science Looks at Smoking (Conard-McCann, 1957).

4) Lancet, 1977, June 25, 1348-1349. See also F.T.Gross et al, Health Physics,

1989, vol.56, 256.


Asbestosis, the lung disease caused by inhaling asbestos, was first

recognised in 1907. The reports were so disturbing that 11 years

later, the Prudential Insurance Company in New York refused to

issue life policies on asbestos workers. Animal research began in

1925 but much of the early experimentation proved contradictory. For

instance, during the 1930s, one group of scientists wrongly classified

the chrysotile, amosite and crocidolite forms of asbestos as harmless

on the basis of animal tests.1 Others found that chrysotile caused

lung damage in guinea pigs but not rabbits.2

In 1931 and again in 1951, experimenters reported that the injuries

caused by asbestos start to heal when the animals are removed from

the dusty atmosphere.2 This is contrary to human experience where

asbestosis progresses even when workers are no longer exposed.

Only later were researchers able to mimic this aspect of the disease

in animals.3

The fact that asbestos could harm the lungs was serious enough but

doctors soon discovered a more alarming threat - cancer. The first

reports of an association between asbestos and lung cancer came

from America, England and Germany during the 1930s following

examination of people who had died with asbestosis. But attempts to

induce cancer in animals repeatedly failed and despite further

evidence from exposed workers, the carcinogenic action of asbestos

was doubted until the 1960s.4,5 Only then were researchers able to

mimic the disease in animals.

Prior to this "...a large literature on experimental studies has failed to

furnish any definite evidence for induction of malignant tumours in

animals exposed to various varieties and preparations of asbestos

by inhalation or intratracheal injection."6


1) Reported in L.U.Gardner, Journal of the American Medical Association, 1938,

November 19, 1925-1936.

2) J.C.Wagner, British Journal of Industrial Medicine, 1963, vol.20, 1-12

3) J.C.Wagner et al, British Journal of Cancer, 1974, vol.29, 252-269.

4) P.E.Enterline in Epidemiology & Health Risk Assessment, Ed. L.Gordis (Oxford

University Press, 1988)

5) P.E.Enterline, American Review of Respiratory Diseases, 1978, vol.118, 975-978.

6) W.E.Smith et al, Annals of the New York Academy of Sciences, 1965, vol.l32, 456-



The key problem for transplant scientists has always been to

overcome the body's natural defence mechanism whereby a

transplanted organ is rejected. Most of the animal research directed

is towards this end has relied on rodents, with rats by far the most

commonly used species.1 Yet scientists have discovered important

tissue differences which mean their results are of questionable

relevance to people, and could be misleading.

For instance, if experiments with rats were used as a guide, patients

receiving heart or kidney grafts would only need a very brief period

of immunosuppression with drugs like cyclosporin, after which they

would never reject their new organ.2,3 In fact, such a course would be

disastrous, for unlike rats, human patients need lifelong

immunosuppression to prevent organ rejection.

The reason, scientists suspect, is that within a few days of

transplantation, the rat's kidney has no cells to stimulate the immune

system, so the animal does not reject a transplant when

immunosuppressive drug treatment is stopped. In contrast, the

human kidney does have these cells as an integral part of its

structure and transplant patients must therefore have lifelong drug

treatment to suppress the cells' immune-stimulating effects.4

During the 1960s and 1970s much research focussed on rat

"models" of kidney and heart transplants, and according to John

Fabre of Oxford University's Nuffield Department of Surgery, "The

many encouraging results raised hopes that a major advance in

clinical immunosuppression for transplantation was in the offing, but

these hopes have now faded and nothing of the great mass of work

has been translated into clinical practice." Fabre suggests that the

tissue differences between people and rats may be responsible.2


1) According to British figures for 1986, 66% of experiments performed in transplant

research used rats, 26% used mice, 7% used rabbits, dogs, primates or other species.

Source: Statistics of Experiments on Living Animals, Great Britain, 1986 (HMSO,1987)

2) J.W.Fabre, Transplantation, 1982, vol.34, 223-224.

3) D.J.Cohen et al, Annals of Internal Medicine, 1984, vol.101, 667-682.

4) P.J.Morris (Ed.), Tissue Transplantation (Churchill Livingstone, 1982). See also ref.2.


The anti-inflammatory drug Ibufenac was marketed in Britain during

1966 but withdrawn two years later following 12 deaths, mainly

through liver damage. Although submitted to "extensive" tests in

mice, rats and dogs, no evidence of liver damage was detected

except for a slight effect in rats exposed to lethal doses of the drug.1

Dr Cuthbert of the Medicine's Division at Britain's Department of

Health and Social Security, explained that "Evidence of liver damage

is sometimes detected in animal studies of non-steroidal

antiinflammatory drugs but usually no such evidence is forthcoming

even in circumstances where a drug is eventually shown to be

hepatotoxic (damaging to the liver) in man."1


1) M.F.Cuthbert in Current Approaches in Toxicology, Ed. B.Ballantyne (Wright & Sons,



In June 1993, researchers at America's National Institutes of Health

abruptly halted trials of a new drug to combat hepatitis B virus,

following deaths and serious complications among participants.

Although the drug, fialuridine (FIAU), was intended to improve liver

disease, many of the patients undergoing prolonged treatment were

getting worse with several dying from liver failure.1

The liver toxicity surprised researchers, for the drug seemed safe

and effective in animal experiments.1 It reduced the amount of

hepatitis virus in infected woodchucks, the "preferred" animal model,

and was also tested for toxicity in mice, rats and rhesus monkeys.

However, one of the trial's chief investigators later asked "...why

didn't the animal toxicity studies show any abnormality at all due to

the drug?"2

The metabolism of anti-viral drugs of this type is said to be very

different in animals and people,3 and the tragedy has prompted a

closer look at related drugs to see if other patients are experiencing

similar harmful effects.


I) N.Touchette, The Journal of NIH Research, 1993, 5, 33-35.

2) J.Hoofnagle, reported in ref. 1.

3) C.Macilwain, Nature, 1993, July 22, 275.



Fair Use Notice and Disclaimer
Send questions or comments about this web site to Ann Berlin,