Ethical review of regulatory toxicology guidelines involving experiments on animals: the example of endocrine disrupters.

The safety assessment of new chemicals (including medicines, pesticides, food additives, and industrial chemicals) relies on the results of animal experiments. Because the safety of those exposed to these products and the welfare of the experimental animals used are considered critically important, both testing requirements and the welfare of experimental animals are controlled by law. In the U.K., projects that propose to use animals for experimental purposes, including for the testing of chemicals, have been controlled by law for over a century, with the most recent legislation (Animals [Scientific Procedures] Act of 1986) requiring a cost/benefit assessment before it may proceed. New regulations introduced in 1998 will require an ethical review process for all projects from April 1999. Such ethical review will have to take account of the toxicity testing methods and schemes that are required by the legislation aimed at protecting human health. Neither national nor international proposals for toxicity testing methods and schemes are generally subjected to ethical review from the point of protecting animal welfare. The international nature of the chemical and pharmaceutical industry means that testing requirements from one of the major national regulatory agencies (USA, EU, or Japan) or the international organizations (Organization for Economic Co-operation and Development [OECD]or the International Conference on Harmonization [ICH]) have an impact on the testing carried out by industrial organizations in all countries. The recent proposals for screening and testing chemicals to identify endocrine disrupters (ED) from the Endocrine Disrupter Screening and Testing Advisory Committee (EDSTAC) of the U.S. Environmental Protection Agency (EPA) are used as an example of the interaction between regulatory proposals and animal welfare issues. The current proposals are the most extravagant in the use of animals. Between 0.6 and 1.2 million animals would be required for each 1000 chemicals tested. The EPA, before incorporating them into regulation, is subjecting the recommendations to further review. This will undoubtedly moderate the number of animals actually used from the worst-case calculation. The variables that have the greatest impact on the number of animals required for testing are the prevalence of ED chemicals in the chemicals to be tested, and the sensitivity and specificity of the testing methods. The modeling demonstrates, for example, that increasing the prevalence from 10 to 50% reduces the number of animals used to detect one ED from 10,000 to 2700. Knowledge of the prevalence of EDs in the chemicals to be tested would allow rational selection of tier one screening based on the sensitivity and specificity of the screening tests. The EDSTAC proposals are difficult to justify from an ethical perspective, as equally effective detection rates may be achieved with fewer animals. National and international regulatory testing proposals should be subjected to formal independent ethical review before they are finalized, with a view to improving animal welfare.