Home page
Site map
about RDS animal research facts medical benefits animal welfare hot topics & policy news & resources
the 3Rs animal welfare prize alternatives 1986 animals act
in vitro research

non-animal replacement methods in safety testing

Despite much effort, the development of methods that directly replace animals is an extremely slow and difficult process. The most progress has probably been made in safety testing. Some of the techniques developed are in vitro (ie they do not use living animals at all, although they often use animal serum or blood), some use lower animals or bacteria, some use animal embryos.

Current spurs to the development of replacements for safety testing of chemicals are the new European legislation on chemicals (REACH), which foresees the testing of some 30,000 substances to minimise risk to our health, and from 2009, the European Cosmetics Directive which will mean a complete ban of cosmetics if they are tested on animals. Cosmetics testing on animals is already outlawed in the UK.

If a non-animal replacement test is developed, validated and accepted by all the regulatory authorities, it will be used instead of the animal method. In the UK and many other countries this is a legal requirement. Even in countries where it is not, there are still ethical, economic and scientific obligations not to use animals unnecessarily.

In our outline of in vitro alternative methods, we concentrate on the development of non-animal replacements in safety testing.

Replacement and reduction in safety testing

This is probably the area where most progress has been made in the development of direct replacements for animals, although it is a slow and difficult process. This is partly because a battery of 20 different types of test is required to assess the risk of any new substance. In addition, replacement involves not only development of the methods themselves, but also checking them against the animal methods to make sure they work - a process known as validation.

To be widely used, national and international regulatory authorities must accept a non- animal replacement technique. These authorities tend to be conservative and it can take many years to get a new technique written into their guidelines. Even then, some countries may still insist that the animal tests are carried out if they haven't been written out of the guidelines.

All this may affect the testing carried out by UK companies regardless of acceptance of methods by UK regulators. All companies must comply with regulations in the countries where their product will be manufactured, transported or sold.

Some of the non-animal tests have limitations, so they result in reductions in animal use rather than being complete replacements. Some tests still use animals, but fewer of them. Some in vitro methods can act as pre-screens so that fewer compounds go forward to the animal testing stage. Use of cell cultures for example has considerably reduced use of rodents in initial screening of potential new medicines while speeding up the process so that 10 or even 20 times the number of compounds can be screened in the same period.

Non-animal methods progress:

Batch testing of insulin
Mutagenicity & carcinogenicity
Skin absorption
Skin sensitising potential
Skin corrosion & irritation
Acute oral toxicity
Eye irritancy
Phototoxicity (UVA)
Acute ecotoxicity

Batch testing of insulin

Insulin is one example of a biological substance for which the quality control tests can now be done using high tech equipment rather than animals. For over 60 years following its discovery in the 1920s, insulin for the treatment of diabetics was produced from pig and cow pancreases. Because of the possibility of impurities, it had to be tested for purity and activity in mice and rabbits. Each batch had to go through three tests using 160 animals.

Today, a fermentation process using genetically engineered bacteria produces human insulin. As a result it is much purer. For this reason it can be tested by a technique called high performance liquid chromatography (HPLC). The peak on the resulting chromatograph shows how much pure insulin is present. Any impurities would show as separate peaks.This is a much faster technique than the previously used animal method, as well as avoiding the use of animals.

Status: accepted by USA Pharmacopeia in 1991

Mutagenicity and carcinogenicity

There are several in vitro tests to determine whether chemicals cause mutations in the genetic material (DNA) of cells. Chemicals that cause mutations (mutagens) can often also cause cancer (carcinogens). A test using bacteria was developed and validated by Bruce Ames and colleagues in the mid 70s. The Ames test uses strains of the bacterium Salmonella typhimurium to test for mutagens. This test and others are now widely used as pre-screens to partially replace rodent testing for cancer-causing compounds. The in vitro tests can produce false results, and tend to be used as way of understanding aspects of mutagenicity and carcinogenicity rather than as replacements for the animal assays.

A micronucleus test that improves assessments for mutagenicity and genotoxicity has been validated. There are moves to replace the standard mouse carcinogenicity assay with other animal assays which cause less suffering because they use fewer animals and do not take so long.

Status: Ames test formally adopted into international (OECD) guidelines in 1983, as part of mutagenicity testing programmes. Micronucleus test currently under consideration by OECD.

Skin absorption

As many cosmetics and toiletries are applied to the skin, it is obviously important to find out to what extent they are absorbed. Some drugs are designed to be absorbed through the skin, and other chemicals, for instance those used in agriculture, may be accidentally absorbed. There has been an international effort by cosmetics scientists to replace the use of live rats in such tests. Isolated human, pig and rat skin preparations can be used, although such methods need to provide reliable information on the effect of carrier compound, the duration of application and dose levels as well as skin characteristics.

Status: Adopted in international guidelines. Research and development continues on tests for the effect of repeat exposure.

Skin sensitising potential

The murine local lymph node assay (LLNA) provides an alternative to previous guinea pig tests - the guinea pig maximisation test and the Beuhler assay. For decades guinea pig tests were the standard methods accepted by regulators worldwide for assessing skin sensitisation potential or skin allergy. These tests required the shaving and bandaging of animals and usually lasted several weeks. By comparison, the mouse LLNA causes less stress; there is no need for shaving or bandaging and animals receive the test chemical on the surface of the skin. In addition the LLNA requires half the number of animals.

The LLNA provides robust objective and quantitative data which is more readily understood by other laboratories. The data generated provides a much sounder basis for the ultimate development of non-animal models.

Status: validated and accepted by OECD.

Skin corrosion and irritation

There has been much scientific research into production of artificial skin, both for transplants (for burns victims particularly) and for skin irritancy testing. In 1996 and 1997 four in vitro tests using isolated animal or reconstituted human skin were subject to international validation studies (3 laboratories, 60 chemicals) to test for chemicals showing unexpected skin corrosivity. Such tests are designed to replace the use of live rabbits. The rat skin transcutaneous electrical resistance (TER) assay and EPISKIN tests were scientifically validated. However, EPISKIN was the only test able to distinguish between two particular types of chemical on an acceptable number of occasions. As these tests are likely to be used in conjunction with other tests, strictly speaking they represent reduction rather than replacement. Similar in vitro tests are being developed to test for skin irritation.

Status: TER and EPISKIN scientifically validated, adopted by the EU and OECD in international guidelines for testing skin corrosion. The guidelines stipulate a stepwise process in which only those chemicals shown to be non-corrosive pass to the next stage. The steps start with assessment of likely corrosivity based on structure (QSAR, or quantitative structure-activity relationship) followed by chemical testing, followed by the in vitro tests, and only then by in vivo tests on rabbits.

Acute oral toxicity

Scientists of the British Toxicological Society developed the Fixed Dose Procedure in the 1980s. It is designed to replace the LD50 test, a somewhat crude test devised by statisticians in 1927 to assess the acute oral toxicity of substances. The LD50 test involves feeding single doses of the test substance to animals to find out what dose kills 50% of them within a given time. It can take 60-80 animals to find the LD50 of one substance.

The Fixed Dose Procedure uses a dose selected to produce signs of toxicity but not death. It is sometimes necessary to re-test using a slightly higher or lower dose. The aim is to set a maximum dose level beyond which no further testing is needed because the lethal dose can be predicted. It takes between 10 and 20 animals to test a single substance. It is an example of reduction and refinement rather than replacement, as animals are still used but in smaller numbers and with less suffering.

Status: validated and internationally accepted (by OECD, EPA, FDA) in the early 1990s. The LD50 guideline still exists, although the OECD announced in June 1999 that the test should be discouraged and it is now working towards deletion by November 2001. Use of the LD50 (OECD guideline 401) was banned in the UK in October 1999.

The Up and Down Procedure (UDP) and the Acute Toxic Class (ATC) methods are further refinements of acute oral toxicity testing, following on from the Fixed Dose Procedure. They further limit the maximum dose and require the use of fewer (6-10) animals.

Status: validated and adopted by OECD in 1996 (ATC) and 1998 (UDP). The LD50 guideline still exists, but should be phased out by November 2001.


Contamination of intravenous medicines by bacteria, even bits of dead bacteria, can cause a toxic reaction and fever. These fever-causing contaminants are called pyrogens. Testing for pyrogens used to be carried out using a very mild test in live rabbits. However, a pyrogenicity test - the Limulus Amoebocyte Lysate (LAL) test - has now been developed using blood from the horseshoe crab, Limulus. Blood is taken from the crabs and they can be released unharmed back into wild - the blood from one crab can substitute for several live rabbits. The test uses cells called amoebocytes, extracted from the blood, which are incubated in a test tube with small amounts of the test substance. The blood clots when bacterial toxins are present. However, the LAL test is not perfect and may produce false results. A suite of six methods using whole diluted human blood has been developed which may overcome this problem.

Status: The LAL test has been scientifically validated and is accepted by many regulatory authorities. The rabbit pyrogenicity guideline still exists for use in circumstances where the LAL test is not appropriate. The tests using human blood have yet to be validated or accepted by the regulatory authorities.

Eye irritancy

The Draize eye test, in which rabbits are used to test the irritancy of substances that could come into contact with the eye, has gained unwarranted notoriety amongst animal rights activists. In its current form it is a very mild test, in which small amounts substances believed to be non-irritants are used, and are washed out of the eye at the first sign of any irritation. Nonetheless, due to high levels of public concern, there has been a major international effort to develop non-animal replacement methods.

A battery of 10 tests are being evaluated which may replace the Draize eye test. One of these is the chorio-allantoic membrane (CAM) test that uses 10-day old hen's eggs. One of the membranes that forms round the chick embryo is exposed via a window in the shell and used to model what happens when a substance comes into contact with the eye. It is treated for 20 seconds with 0.2 ml of the test substance if liquid or 0.1g if solid. It is then rinsed off with 5ml of warm water and the damage, if any, done by the test substance is assessed under the microscope, 30 seconds, 2 minutes and 5 minutes after treatment. Another test is EYTEX, which uses a clear gel from jack bean protein to mimic the cornea of the eye. Eye irritancy is indicated by coagulation of this protein.

Efforts are continuing to understand better the mechanisms involved in human eye irritation so that better non-animal test systems might be developed.

Status: two tests have been validated for the detection of severe eye irritants. Eight other assays for mild irritants are currently under evaluation.


Since the thalidomide tragedy, all medicines and many other substances must be tested for their ability to cause damage to the developing foetus (teratogenicity). While pregnant mammals are still required in the later testing stages, it is possible to use simpler organisms or embryos as pre-screens, so that substances identified as potential teratogens are abandoned at an early stage rather than going forward to animal testing. A common teratogenicity pre-screen uses Hydra, a simple animal related to sea anemones and jellyfish. Embryonic cells, chick embryos, frog tadpoles and rat foetuses are also used. Such pre-screens are a good example of the use of techniques to reduce, rather than replace, tests using adult or higher animals.

Status: Pre-screen for reduction of animal use only.

Phototoxicity (UVA)

The European Commission and the Cosmetic, Perfumery and Toiletry Association agreed in 1991 to develop and validate non-animal replacement methods for phototoxic substances that may cause skin damage in the presence of light. They developed and validated a test (3T3 NRU PI test) which examines the ability of certain mouse cells (3T3) to take up a substance called neutral red (NR), which indicates whether they are damaged. This involved strict standardisation of cells and UV light sources, as previous attempts to develop phototoxicity tests had been marred by inter-laboratory variation. The validation involved testing 30 carefully selected test chemicals in 11 laboratories in a blind trial.

Status: scientifically validated and accepted by OECD for international regulatory guidelines.

Acute ecotoxicity

A test strategy which saves 60% of fish has been validated in Europe. The strategy has been submitted to OECD for acceptance into international regulatory guidelines.

April 2007

top of page
The alternatives:
Potential of in vitro replacements in research
Non-animal replacement methods in safety testing
Microdosing: the facts
Related links
LD50 test ban welcomed
Safety testing of non-medical products
Safety testing of medicines
email page to a friend