Chapter 5
Beyond animal testing – developing and using non-animal methods
print

It is widely accepted throughout Europe that the testing of cosmetics products and ingredients on animals must end, but there are differences of opinion as to how and when this should happen.
COLIPA, the European Cosmetic, Toiletry and Perfumery Association, states that “animal testing on finished products has come to an end” . However, OneVoice discovered disturbing evidence to the contrary in an undercover investigation in a French laboratory, which took place during 1999-2000. The contract testing laboratory appeared to be testing finished cosmetic products, such as face creams and moisturisers, on animals (see Chapter 2).

An ethical stance

The European Parliament and the European public take the view that animal testing for cosmetics is ethically unacceptable and should stop – whether or not there are currently enough non-animal methods to replace animal tests. This view is also supported by OneVoice and by the European Coalition to End Animal Experiments.
This approach is pragmatic because the European inventory of cosmetics ingredients already contains 8 400 substances with a long history of use in cosmetics and which are generally accepted as safe. Novel combinations of these existing ingredients would enable many thousands of new products to be developed, without any risk to consumer safety – a key concern of the Commission’s DG Health and Consumer Protection.
In fact, Britain and the Netherlands have already ended the testing of cosmetics products and ingredients on animals, and Austria and Germany have partial bans.
COLIPA argues against this ethical approach, and instead supports an end to tests on animals only when these can be replaced by validated non-animal methods. Of course, Directive 86/609/EEC which protects laboratory animals, does not permit animal tests to be conducted when non-animal methods are available. So COLIPA’s position is nothing more advanced than legislation requires.

The legislative situation

The seventh amendment to the EU Cosmetics Directive 76/768/EEC , agreed in 2003, has introduced deadlines for implementing marketing bans for cosmetics containing animal-tested ingredients, as well as a deadline for an animal testing ban within EU member states.
The seventh amendment states that, by 2009, no further animal testing of cosmetics ingredients shall be conducted in the EU for the purposes of the Cosmetics Directive. This is regardless of the status of alternative, non-animal testing methods.
As of September 2004, there will be an EU marketing ban on cosmetics containing animal-tested ingredients, if EU-validated non-animal testing methods exist. This partial ban would apply to products from anywhere in the world.
In 2009, a further partial marketing ban will prohibit the sales of cosmetics in the EU containing any ingredients tested on animals, whether or not non-animal methods are available. The ban applies to tests conducted anywhere in the world as long as they are done for the purposes of the Cosmetics Directive. Testing under the Dangerous Substances Directive or because of the safety testing requirements of non-EU countries are unaffected by the seventh amendment. Exceptions are for three types of toxicity only: reproductive toxicity, toxicokinetics and repeat-dose toxicity. Cosmetics containing ingredients tested on animals for these three toxic endpoints can be marketed until 2013. If non-animal methods have not been finalised by that time, the deadline can be postponed yet again, by means of the EU co-decision process.
The EU cosmetics testing and marketing bans should encourage the cosmetics industry and its chemical suppliers to increase their effort to find non-animal testing methods. However, the bans will not affect the testing on animals of chemicals produced mainly for other product categories, such as for industrial or household purposes, which will continue to be regulated under the dangerous substances legislation.
Thus it is likely that companies doing animal tests for chemicals will simply re-classify their tests from the cosmetics category into the dangerous substances category. This would provide a smokescreen to hide a continuing use of animal tests for cosmetics ingredients which would not appear as such in the EU statistics of animal experiments. If companies perceive this to be in their economic interests, they may not hesitate to evade EU restrictions.

OneVoice urges the French government, and other national authorities throughout Europe, to advise laboratories that the re-classification of animal tests to avoid the cosmetics testing ban would be unacceptable and a betrayal of the will of the public.

In 2002 a leaked internal memo from the Procter and Gamble company revealed the industry’s cynical view of animal suffering and public opinion. The company, which markets brands such as Max Factor, Vidal Sassoon, Camay soap, Hugo Boss fragrance and Nice n’ Easy hair colours, has been lobbying behind the scenes to postpone testing and marketing bans. The memo stated: “It would be damaging to be seen as the company lobbying to test on animals, against public opinion.” It revealed that the company conducts most of its safety tests outside Europe and thus expects to avoid the impact of the EU animal testing ban, while still marketing its products in Europe.

Competitiveness and trade disputes

COLIPA opposes a testing ban because it fears this would affect the competitiveness of the industry in the global market, by weakening the ability of European companies to ‘innovate’ and to market their products overseas. This assumes that product innovation based on new combinations of the 8 400 existing ingredients will not satisfy consumers, at least in the short term. COLIPA says that the deadlines in the seventh amendment for developing non-animal methods and banning animal tests are too close.
COLIPA also opposes an EU ban on the marketing of cosmetics containing animal-tested ingredients, as it believes this would risk a trade dispute with other sectors, such as Japan, the USA or Canada. Some member states have agreed with this interpretation. However, Member of the European Parliament Caroline Lucas and others believe that a marketing ban could be defended under the rules of the World Trade Organisation (WTO) and should be pursued vigorously by the EU. Public concern about animal welfare is one of the permitted general exceptions to free trade under Article XX of the GATT 1994 . Together with an expression of US public abhorrence for the trade in dog and cat fur, animal health and welfare was a key argument used by the USA when it banned the import and sale of these products in December 2000.

Other precedents for exceptions based on non-trade objectives – such as the protection of people, animals and the environment – have been set. They include the unofficial moratorium since 1998 on approving genetically modified foods in the EU. The USA has only recently asked the WTO to rule on whether that was a breach of free trade rules.

In another relevant instance, at the WTO meeting in November 2001, trade ministers declared that intellectual property rights (i.e. patents held by drug companies) should not stand in the way of poor countries obtaining cheaper medicines. Finally, in 2003, the WTO has agreed that poor nations may continue to import cheap, generic versions of patented drugs to help fight diseases such as AIDS and tuberculosis. This was described by the director-general of the WTO as proof “that the organisation can handle humanitarian as well as trade concerns” .

The UK-based BUAV has received high-level legal advice indicating that WTO members may impose trade restrictions if the issue is one of public morality (European polls consistently show public opposition to animal testing for cosmetics) and it is necessary to protect animal health (which is obviously damaged by testing).
Despite the need for ethical issues, including animal protection, to be dealt with justly under free trade agreements, the French government and the European Federation for Cosmetics Ingredients, representing the industry, have attempted to block the agreed EU prohibitions on animal testing (see Chapter 6).
Of course, public opposition to animal testing is growing worldwide, cosmetics industries in third countries (outside the EU) are aware of the issues involved. Laboratories in these countries are participating in the development and validation of non-animal tests. Testing and marketing bans in Europe will send very strong signals to those industries that they will need to change their testing methods too if they want to market their products in the EU.

The role of non-animal methods

One solution to the testing of cosmetics ingredients and to potential trade disputes, is the rapid development of non-animal methods to replace animal testing. The cosmetics industry – under pressure from consumers and under the threat of a ban – has probably done more than any other, except the pharmaceutical industry, to develop non-animal methods. However, there is much more that can and must be done. The chemical industry collectively has made almost no contribution to funding or researching non-animal tests. However, proposed changes to the EU chemicals strategy (see Chapter 2) mean that there is now a two-fold pressure on chemical companies to take positive action in this regard.
The development and validation of new, humane testing methods will certainly take time and investment in the short term. But the pay-off will be a full range of non-animal techniques, enabling new and existing ingredients to be tested rapidly, reliably and more cheaply for potential health effects. This will maintain or enhance consumer and environmental protection, and will also meet consumer demand that animals should not suffer for the sake of cosmetics.
An added spin-off will be that other chemicals – such as industrial chemicals, biocides, pesticides and pharmaceuticals – can also be tested more quickly and cheaply using the new, non-animal methods. This will save many millions of animals from distressing and sometimes lethal tests. Finally, by investing funds and expertise in developing computational methods and in vitro techniques (cell culture, genomics, proteomics, instrumentation etc.), the EU will benefit from new scientific knowledge and skills in these technologies. The technologies have wide applicability to important areas including medicine; thus the worldwide standing and competitiveness of the EU and its member states in these areas will be enhanced.

Non-animal testing methods

The European Commission has established a series of meetings between stakeholders to discuss the timetable for developing non-animal methods for the range of toxicity testing endpoints. The European Centre for the Validation of Alternative Methods (ECVAM) is playing a key role.
As discussed in Chapter 4, the data required by the Scientific Committee on Cosmetics and Non-Food Products (SCCNFP) for a cosmetic ingredient, under the Cosmetics Directive, can include:

1. Acute toxicity (if available) 5. Skin sensitisation & photo-
2. Skin absorption sensitisation
3. Skin corrosion/irritation 6. Sub-chronic toxicity
4. Eye irritation 7. Mutagenicity/genotoxicity

If the chemical absorbs ultraviolet light:
8. Phototoxicity
9. Photomutagenicity/photogenotoxicity
10. Human data (if available)

And, if the ingredient is likely to enter the bloodstream in significant amounts, through the skin or by mouth:
11. Toxicokinetics
12. Metabolism studies
13. Long-term toxicity studies (e.g. teratogenicity, reproductive toxicity and carcinogenicity)

However, the SCCNFP claims that it takes a flexible approach to testing and indeed it accepted in vitro results from skin penetration studiers in advance of this method becoming an EU guideline.

OneVoice recommends that the SCCNFP does not limit its acceptance of test data only to non-animal tests which have received Europe-wide regulatory approval, especially if this is delayed for bureaucratic reasons. If the SCCNFP is convinced that a non-animal test method is scientifically valid, it should not wait for EU agreement (although it should facilitate such agreement).

There are already valid non-animal tests accepted by the SCCNFP and listed in either the OECD or EU test guidelines, for providing some of these data (see Table 1). Within the EU, if a non-animal method is listed in the guidelines, it is mandatory to use it in place of an equivalent animal test.

Table 1
Non-animal testing methods which are validated and approved in the OECD and/or the European Union

Toxic endpoint	Non-animal method	OECD and/or EU test guideline number*
Skin absorption	Assay using isolated skin fragments	TG428
Skin corrosion	Human reconstructed skin	TG431, B40
Skin corrosion	TER assay	TG430, B40
Mutagenicity/genotoxicity	Ames test	TG471, B13/14
Mutagenicity/genotoxicity	Gene mutation in yeast	TG480, B15
Mutagenicity/genotoxicity	Mitotic recombination in yeast	TG481, B16
Mutagenicity/genotoxicity	Mammalian chromosome aberration test	TG473, B10
Mutagenicity/genotoxicity	Mammalian gene mutation test	TG476, B17
Mutagenicity/genotoxicity	Unscheduled DNA synthesis	TG482, B18
Mutagenicity/genotoxicity	Sister chromatid exchange	TG479, B19
Phototoxicity	Neutral red uptake assay	TG432, B41
Carcinogenicity	Mammalian cell transformation test	B21

* A ‘TG’ number indicates an OECD test guideline. A ‘B’ number indicates an EU test guideline in Annex V of the Dangerous Substances Directive 67/548/EEC.

The status of non-animal methods and priorities for research

Here, we review available non-animal testing methods for different toxicities, and identify research which needs to be done as a priority in order to replace animal experiments as soon as possible.
The SCCNFP does not demand that animal tests are conducted specifically to provide [1] acute toxicity data for cosmetics ingredients . The data need only be provided where already available. For many ingredients, the data will already exist especially if the chemical has been approved under the Dangerous Substances Directives. To determine acute toxicity without using animals, a combination of tests is required:
– Results from in vitro skin absorption studies (see above) and from in vitro gut absorption tests (widely used in the drug industry), show whether there would be skin or oral absorption of a chemical.

– If this is likely, studies of in vitro cell toxicity (cytotoxicity) can be conducted. Two methods are undergoing a joint European and US validation study during 2003, and result should be available in 2004. These tests would identify chemicals which are directly very toxic, or else non-toxic.

– Studying chemical metabolism in the test-tube, using metabolic activity from human liver cells or genetically modified cell cultures, can identify toxic metabolites. These tests are widely used in the drug industry, and prevalidation studies started in 2003.

– If necessary, more organ-specific tests can be conducted in vitro to see if a chemical might damage the liver, brain or kidneys. Methods for doing this are under development and need resources for further development.
As Table 1 shows, a valid in vitro method exists for [2] skin absorption. The SCCNFP has been accepting data from this method for some years; it also accepts human data for skin absorption which it considers as “ideal” . There are animal tests for skin absorption but they have never been properly validated.
No animal tests should now be conducted for [3] skin corrosion, as in vitro methods have been validated and accepted. Two methods for assessing [3] skin irritation in vitro will start a validation study in 2003 and are expected to complete this successfully. However, an international group of expert toxicologists has already recommended that risk assessments for skin irritation can be completed safely and effectively without any animal testing. In addition to in vitro methods, in certain cases they propose validated ethically-conducted human volunteer studies (patch tests), and the SCCNFP agrees with this approach . Therefore there is no justification for animal tests for skin irritation.
OECD guidelines already recommend a stepwise testing strategy for [4] eye irritation, requiring an animal test only when a chemical has tested negative through the non-animal stages, such as physico-chemical properties, acid/alkaline properties, etc. The EU test guideline also says that such data are useful.
Additionally, there are four well-established methods using isolated animal tissues which can distinguish mild from severe eye irritants without using living animals . Some of these methods have regulatory approval at the national level, for example in Germany, Belgium, Britain and the Netherlands.
In France, all four of these tests are accepted for positive classification of severe eye irritants, and the neutral red release assay and the agarose-diffusion assay are also accepted for evaluating cosmetic products .
French representatives on the SCCNFP and in other EU fora should urge the acceptance of data from these methods throughout the EU. Irritant chemicals should not be tested on rabbits in France, where alternatives have been accepted as valid for distinguishing mild from severe irritants.
Finally, there are a number of candidate in vitro tests which could classify non-irritants. These tests should be prioritised for rapid development and validation, and then speedy regulatory acceptance by the European Commission.
There are as yet no fully validated non-animal methods to assess chemicals for [5] skin sensitisation. However, skin sensitisation cannot occur unless a chemical can both penetrate the skin and bind to proteins. There is an OECD-accepted method for in vitro skin absorption (Table 1, above), and protein binding is reliably measurable in the test tube. COLIPA considers protein-binding to be a useful in-house method for hazard identification . There are also computer systems which can predict skin sensitisation on the basis of chemical structure, and these will be undergoing validation within the next year or two. Therefore, many chemical ingredients could be classified as non-sensitisers or likely sensitisers, on the basis of these three approaches.
Additionally, cell-based tests are being developed to predict sensitisation, and L’Oréal has done considerable research in this area. ECVAM envisages these tests as being technically ready to complete validation studies by about 2006. The SCCNFP accepts results from skin sensitisation studies in volunteers. While not actually recommending that human studies be undertaken, the SCCNFP points out that human testing is advantageous because species differences in reactions are thus avoided . For all these reasons, testing for sensitisation could be conducted for many cosmetics ingredients or products without animal experiments even today; and certainly by 2006.
The SCCNFP states that certain new cosmetics ingredients (such as preservatives, colourings or ultraviolet filters) should be tested for [6] sub-chronic toxicity (repeat-dose toxicity) in animals, even if they are exempted under the Dangerous Substances Directives (for example because the chemical is manufactured only in small quantities). Sub-chronic toxicity tests involve daily dosing of animals over a period of 28 or 90 days (see Chapter 4). The SCCNFP does not explain why such tests should be necessary if the ingredients do not penetrate the skin and are not likely to be swallowed or absorbed into the bloodstream. Both of these properties can be assessed without using animals (see under Acute toxicity, above).
The seventh amendment to the Cosmetics Directive has postponed a marketing ban on cosmetics containing ingredients tested on animals for repeat-dose toxicity. The envisaged date for such a ban is 2013.
In vitro methods for testing repeat-dose (e.g. sub-chronic) toxicity are not yet fully developed or validated, but research shows that they can be achieved. The slow progress reflects a lack of research effort and investment rather than insurmountable technical difficulties. Organ-specific tests can also be conducted in vitro to see if a chemical might damage the liver, brain or kidneys. Again, methods for doing this are under development and need resources for more rapid progress.
This work should be prioritised so that companies will be able to introduce safe, novel ingredients in due course. But there is no reason to permit animal testing to continue in the meantime, because new products can be created safely using the inventory of
8 400 existing ingredients.
Regarding [7] mutagenicity/genotoxicity, despite the good range of OECD- and EU-accepted in vitro methods it has been customary under the Dangerous Substances Directives to further test chemicals with positive results in animal studies. The SCCNFP says in its Notes of Guidance that two in vitro assays generally provide “…sufficient evidence of mutagenic and/or genotoxic potential”, but then adds that animal tests can be conducted to confirm “…a mutagenic activity already observed in vitro”.
This practice of re-testing a positive chemical in animals is not ethically supportable. If an ingredient has two positive results from the standard in vitro tests (usually the Ames reverse mutation test and the mammalian chromosome aberration test), it should be classified as a mutagen or genotoxic carcinogen without making animals suffer in further tests. Two negatives indicate that the chemical is not a mutagen or genotoxic carcinogen. If necessary, results can be clarified by varying the conditions of the assays (e.g. further studies of metabolism) or by conducting additional in vitro tests.
For chemicals which absorb ultraviolet light, a cell-based technique for measuring [8] phototoxicity (photo-irritancy) is accepted in the OECD and the EU guidelines, and therefore under Directive 86/609/EEC, animal tests should not be performed for this endpoint. Anyway, animal tests for this purpose have not been validated.
Also needed only for ultraviolet-light-absorbing chemicals, [9] photo-mutagenicity/photo-genotoxicity tests in vitro are accepted by the SCCNFP. Recommended are the Ames test plus the cell test for chromosome aberrations, conducted with and without ultraviolet light.
In their Notes of Guidance, the SCCNFP provide details of situations in which [10] human studies can be conducted. These include patch tests for skin irritancy (Annex 11) and compatibility testing for cosmetics products (Annex 12). The SCCNFP also accepts data from skin sensitisation studies in volunteers (although it stops short of recommending such tests), pointing out that human testing avoids the problem of species differences in reactions (Annex 13).
If a cosmetics ingredient is likely to be ingested or absorbed through the skin in significant amounts, the SCCNFP expects [11] toxicokinetic studies and [12] metabolism studies to be conducted. These are aimed at understanding the absorption, distribution, metabolism and excretion of a chemical. The in vitro skin absorption test is in the OECD test guidelines, as discussed above. Gut absorption assessed by means of human cell cultures has been very widely used by the pharmaceutical industry and by academic researchers.
Computer models called physiologically-based biokinetic (PBBK) models have been developed to predict the distribution of chemicals within the body, and how they are excreted . These are based on knowledge of the chemical itself, on test-tube data and on the known physiological features of the human body. These computational systems could be validated by 2007.
Regarding metabolism, according to ECVAM the best in vitro methods use either enzyme activity in human liver cells, or genetically modified cell lines. A prevalidation study on these techniques started in 2003. If necessary funds and expertise are provided, validation should be completed by 2007.
If a cosmetics ingredient might be absorbed into the body in substantial amounts, the SCCNFP may want to see information from [13] long-term toxicity tests. These could include carcinogenicity, fertility and reproductive studies.
The embryonic stem cell test in vitro has been validated by ECVAM as suitable for identifying chemicals which cause embryo damage (teratogenicity). This test should be incorporated into regulatory guidelines without delay. Effects of chemicals on fertility can be screened by studying sperm and other germ cells in the test tube. These methods need final development and validation.
Meanwhile, the complex area of fertility and reproductive toxicity needs more research and development, which should aim to produce results within five or six years at the maximum. This work should be prioritised so that companies will be able to introduce safe, novel ingredients in due course. But there is no reason to permit animal testing to continue in the meantime, because new products can be created safely using existing ingredients.
Genotoxic carcinogens have already been identified by tests described above (see item [7]). Non-genotoxic carcinogens act by a range of processes not directly affecting the genes or chromosomes. Mammalian cell transformation assays are already in the EU guidelines (B21) but, although widely used to study carcinogenesis, they have not yet received international regulatory support. The cell tests are very much quicker and cheaper to perform than equivalent animal tests, and do not cause animal suffering.
A guideline for these assays is under development at the OECD and finalisation of this should be rapidly progressed, with support from French representatives at the OECD.
As novel products can be created from the existing inventory of ingredients accepted as safe, the suffering involved in animal testing could be ended today without jeopardising public safety. However, the cosmetics industry is dissatisfied with this ethical approach. The seventh amendment to the Cosmetics Directive has made the proposed deadlines for marketing bans dependent on the availability of non-animal testing methods.
Since the industry and the EU have insisted on allowing animals to continue suffering in the name of vanity, they bear the responsibility for putting maximum financial and human resources into developing and validating non-animal testing methods. Moreover, at the OECD, the European Commission and EU member states must actively lobby other member countries of the OECD to accept, without delay, the results of validated new methods so that these replace animal tests around the world.