Assessing the toxicity of chemicals
We have an innate understanding that despite many differences living things are fundamentally alike. We do not allow our children to play with rat poison because we reason that a substance which is harmful to animals may also be toxic to humans. While animals and humans do not respond to all substances in the same way, the data gathered from animal tests does allow potentially harmful chemicals to be identified and people protected from the risks they present.
Animal studies are a vital part of risk assessment for toxicology, but are only valuable within the context of the studies’ limitations. The protection of humans, animals and the environment relies on these limitations being taken into account, and studies being constantly improved to provide the best information using the fewest number of animals.
When toxicity testing is needed
Toxicology is the study of poisonous and harmful substances. Toxicological testing is used to determine the toxicity of those chemicals we use or are exposed to, and to give information about the potency of their toxic effects. These tests give information about industrial chemicals, pharmaceuticals and natural products, such as those formed by plants, bacteria and fungi. Knowing whether a chemical can cause cancer, allergic reactions or abnormalities in unborn children is important to human health, and the process of discovering this information is known as assessment of hazard.
The vast majority of toxicity testing in animals is carried out for pharmaceutical products, but other chemicals of concern to human health, animal health or the environment are also tested. Household products are rarely tested on animals unless there is specific reason: a weedkiller which is labelled ‘safe for pets’ will have been tested to determine its toxicity to animals. However, these products often contain chemical ingredients which do undergo toxicity tests, including tests on animals.
A single chemical substance may have a variety of uses, and legislation requires that new chemicals must be tested to assess their potential to harm people, animals or the environment before they can be used in manufacture.
REACH legislation now means that toxicity data is required for chemicals which are already in use but have never been tested to current requirements. Importers and manufacturers of these chemicals must now supply data on their potential hazards, and carry out safety tests as needed.
As well as showing whether a chemical is potentially harmful, toxicity tests show how potent it is in producing toxic effects. A substance which illustrates the importance of potency is Vitamin A; a low dose is essential, but at doses only five times greater it causes abnormalities in the formation of a child’s eyes.
Assessment of potency is an important aspect of toxicology, as the dose of the most potent carcinogen that causes cancer is 100,000,000 times less than that of the least potent. Chemicals such as dioxins may cause cancer at concentrations 10 or 100 million times lower than some synthetic carcinogens. Without knowledge of the potency, the degree of exposure which would be harmful cannot be determined.
Assessment of risk is the likelihood of harm occurring on exposure to the chemical, and is another important aspect of toxicology. Assessing risk requires knowledge of both the type of toxicity (the toxic hazard) and the dose at which the toxic effect occurs (potency). This is compared to the dose of a chemical that people are actually exposed to. In summary: a chemical can be hazardous, but if people use small amounts of they may be at little or no risk of toxicity.
Protecting the health of workers
Workers who are exposed to chemicals are protected by keeping their exposure below a particular limit which is set for each toxic chemical. The limit is known as the Occupational Exposure Standard (OES), and is the concentration in the air which is low enough to protect people from that chemical’s toxic effects. In most cases the OES is established by the results of animal toxicity studies.
Damage to health caused by chemical exposure in the workplace has decreased substantially over the years, and this has undoubtedly been due to the appropriate OES being set and the enforcement of these limits. Improved health in the workplace is often viewed as “employers being generally more careful about exposing employees,” but to provide a standard of care for employees the OES must be known so that exposure can be limited.
For example, phosgene is a toxic gas which is fatal to humans. There is no systematic Health or exposure data from humans but a safe exposure level is needed to protect people. The OES is based on the results of animal toxicity tests, and is set at 0.08 mg/m3 in the air. This level is recognised through its use to be an effective control level.
Similarly, there is insufficient human data to set a safe exposure level for chlorodifluromethane, but animal studies have allowed the Health and Safety Executive to set the OES level of 2590 mg/m³. This has proved to be a safe level of exposure although it is 4,500 times higher than that of phosgene. These examples demonstrate that specific knowledge is required. It is very unlikely that human health would be adequately protected from the effects of these two different chemicals simply by general improvements to care in the workplace.
Similar arguments hold in the development of medicines and in basic research and in other areas where animal studies are of value.
All methods available for assessing hazard have problems of interpretation, precision and reproducibility. They also apply to particular circumstances. An understanding of the limitations associated with a given method for hazard assessment increases its value as it can be used more effectively.
The limitations of any animal tests used need to be understood if they are to make a valid assessment of hazard; often the first stage of examining the potential harm that a particular chemical could cause. The validity of the animal tests can frequently be judged by whether a particular chemical produces a particular type of toxicity in the animal models as well as in humans. Each animal test will only be valid for assessing particular types of hazard, and part of the toxicologist’s job is to take this into account when assessing chemicals.
The following example shows how animal tests can provide information on hazard.
Some chemicals can cause allergy, either causing a skin reaction or asthma. This is classed as a severe effect because once a person is sensitised to a chemical they are likely to remain sensitised all their life and their reaction to it can be life threatening. The dose of a substance needed to produce an allergic reaction can be very low, so there are real difficulties in protecting the health of an allergic person (for example, in the case of peanut allergy).
A test known as the Local Lymph Node Assay (LLNA) is now used to detect whether chemicals are contact sensitisers. This test was adopted in the EU in 2004 as a refinement over previous skin sensitisation tests. It is a less aggressive procedure which uses fewer animals. This in vivo test uses mice, and has been validated with known human allergens. A review of the method states:
‘Investigations show that the LLNA can be used to provide quantitative estimates of relative skin sensitising potency EC3 values that correlate closely with No Observed Effect Levels established from human repeat patch testing and from our clinical experience’
This means that the measure of potency in the LLNA test (the EC3) correlates very closely with the measure of potency in humans. Clear experimental evidence that the LLNA method in mice is valid for testing chemicals for human allergenicity. However, even this test fails to identify some human allergens such as nickel. There are reasons why this is so, and no other method, whether in vitro or in vivo, provides better prediction of human effects.
The really valuable animal models of human conditions are those where the limitations are known, acknowledged and taken into account. These limitations need to be understood when using the results to predict likely effects in humans. In many cases, although not in all, the results of toxicity testing provide valuable information for assessing human health effects which is not available from any other source.
The process of risk assessment needs to be understood to determine the effectiveness of data obtained from animal toxicity tests. Risk assessment is used to decide whether it is necessary to control a risk. When a risk is small it may not be necessary to control it, but if a risk is great considerable lengths may be considered reasonable.
The methods available to scientists who assess risks are far from perfect, and ultimately well-conducted studies in humans provide the best information for assessing risks to human populations. However, there are problems associated with assessing risks solely through epidemiological studies. The effects being studied might occur so rarely that they cannot be observed in the study-population, or there may not be adequate information available on exposure.
Studies in humans are difficult to carry out and, for diseases that are serious or take many years to develop, cannot be used as the sole method of assessing risk to humans. There are ethical problems of exposing human populations to chemicals which are not properly tested, and then waiting to see if toxic effects occur. This is the rationale for the REACH regulations to cover existing chemicals in the environment. When new chemicals are introduced or when the toxic side effect is serious, an assessment of risk using animal tests is essential to avoid chemical exposure causing human disease.
Accounting for variation in susceptibility
There are differences in susceptibility to risks, both between individual humans and between humans and animals, so the methods used to assess risks for general toxic effects take account of these differences.
Acceptable or ‘safe’ doses of chemicals are usually based on an assessment of the dose that produces no toxic effect in animal tests (known as the No Adverse Effect Level). This dose is divided by a ‘safety factor’, usually 100: a factor of 10 to take into account the differences between the animals and humans, and another factor of 10 to take differences in individual susceptibility into account. The safety factor gives a dose with little likelihood of causing toxicity in humans, and this system has been used to derive safety standards and limits for many years. It is an arbitrary mechanism, but has withstood the test of time. Such a calculation is necessary irrespective of whether safety tests are carried out on animals or with non-animal methods.
The use of these two factors has been researched and the safety factor that takes individual differences into account has been demonstrated adequate to deal with differences in susceptibility for the majority of the population.
Work on the differences between animals and humans is less well developed. Studies suggest that one component of this factor may be too small, and a larger value may be necessary to take differences into account., Additional factors may be also be needed to improve the assessment of risk to infants and children.
Animal studies are a vital part of risk assessment for toxicology, but are only valuable within the context of the studies’ limitations. To protect the welfare of humans and the environment, it is important that these limitations are taken into account and that studies are constantly improved to ensure that toxicity tests provide the best information using the fewest animals.
- Gerbick GF, Robinson MK, Ryan CA, Dearman RJ. Kimber I, Basketter DA, Wright Z and Marks JG. (2001) Contact Allergenic Potency: Correlation of Human and Local Lymph Node Assay Data. American journal of contact Dermatitis, 12, 156-161.
- ECETOC (2000) Skin Sensitisation Testing for the Purpose of Hazard Identification and Risk Assessment. European Centre for Ecotoxicology and Toxicology of Chemicals, Brussels. Monograph No 29.
- HSE (2000). Guidance note EH40/00 "Occupational Exposure Limits" ISBN 0 7176 1730 0, Health and Safety Executive, Rose Court, London.
- Dorne, J. L., Walton, K., and Renwick, A. G. (2001). Human variability in glucuronidation in relation to uncertainty factors for risk assessment. Food Chem. Toxicol. 39, 1153–1173.
- Walton, K., Dorne, J. L., and Renwick, A. G. (2001). Uncertainty factors for chemical risk assessment: Interspecies differences in glucuronidation. Food Chem. Toxicol. 39, 1175–1190.[Medline]
- Walton, K., Dorne, J. L., and Renwick, A. G. (2001). Categorical default factors for interspecies differences in the major routes of xenobiotic elimination. Hum. Ecol. Risk Assess. 7, 181–201.
- Renwick, A. G., Dorne, J. L., and Walton, K. (2000). An analysis of the need for an additional uncertainty factor for infants and children. Regul. Toxicol. Pharmacol. 31, 286–296.