The Animals (Scientific Procedures) Act 1986¹ makes provision for the protection of animals used for experimental or other scientific purposes.

This Guidance, produced jointly by the Home Office, the Department of the Environment, Food and Rural Affairs (DEFRA), the Department of Health (DH), the Health and Safety Executive (HSE), the Medicines and Healthcare Products Regulatory Agency (MHRA), the Pesticides Safety Directorate (PSD), the Veterinary Medicines Directorate (VMD), the British Toxicology Society (BTS) and the British Society of Toxicological Pathologists (BSTP) is aimed at those with responsibility for the design and conduct of animal studies carried out under the authority of the 1986 Act for regulatory toxicology and safety testing.

The Home Office and other UK regulatory authorities have endorsed a statement of principles (Annex I) concerning welfare and are committed to the highest scientific and welfare standards and to the effective implementation of the 3Rs of Russell and Burch² That is, to, where possible -

• refine the experimental procedures to minimise suffering;

• reduce the number of animals used to the minimum required to support sound risk assessments to protect humans, animals and the environment;

• replace the use of animals, for example, by in vitro methods.

This Guidance sets out advice, principles and recommendations aimed at minimising the pain, suffering, distress and lasting harm caused to animals, without compromising the validity of the test data generated in the specific area of regulatory toxicology and safety evaluations.

It offers advice primarily to those seeking Project Licence authority for the conduct of safety evaluation studies regulated by the Animals (Scientific Procedures) Act 1986. Project Licence applications submitted to the Home Office must contain a plan of work and procedures that demonstrably implement the 3Rs. The general principles of humane experimental technique hold true regardless of the class of test material investigated (e.g. chemicals, pharmaceuticals, biocides, materials used in medical devices etc) or the regulatory requirement to be fulfilled.

The Guidance does not list, interpret, compare or justify regulatory requirements for regulatory toxicology and safety evaluations. It supplements, rather than replaces, the formal Home Office Guidance on the Operation of the Animals (Scientific Procedures) Act 1986³.

1.1 The Home Office is responsible for the implementation of the Animals (Scientific Procedures) Act 1986, as amended, regulating the use of protected animals for experimental or other scientific purposes when the procedures performed may have the effect of causing pain, suffering, distress or lasting harm.

1.2 The 1986 Act requires that regulated procedures only be authorised and performed when the scientific objectives, in whole or in part, cannot be achieved by means not requiring the performance of regulated procedures; and that those regulated procedures which can be justified entail the least animal suffering required to produce scientifically satisfactory results. There is a requirement for all new and existing project licences protocols to be subject to a local ethical review process⁴.

1.3 For the purposes of the Animals (Scientific Procedures) Act 1986, the justification for animal-based regulatory toxicology and safety testing is the need for regulatory authorities to have sufficient information to assess the risks to which humans, animals, plants or the environment are exposed when the test substances are produced, transported or used. Thus, in the case of regulatory toxicology and safety testing, the cost/benefit assessment performed under the 1986 Act generally assumes that the principal benefit is the facilitation of sound regulatory decisions, rather than the utility or profitability of the end-product.

1.4 Individual Project Licence applications are drafted to cover specific, coherent categories of test materials (e.g. industrial chemicals, pharmaceuticals, biocides etc.) and to identify the relevant range of specific animal test requirements, the test methods and strategies, and the species involved. Test methods to be used should normally be internationally recognised. These are principally the Organisation for Economic Co-operation and Development (OECD) test guidelines or in the case of medicines the guidance provided by the International Conference on Harmonisation of Technical Requirements for the registration of Pharmaceuticals for Human Use (ICH). Occasionally the regulatory agencies may require specific information that is not covered by such guidelines, in which case the preferred method should be identified in dialogue with the relevant regulatory department or agency and the scientific justification given in the project licence.

1.5 Regulatory toxicology and safety testing is a complex and constantly changing field, where flexibility, specialist requirements and technical progress must be accommodated. It is hoped that demonstrable adherence to the key considerations, principles and recommendations contained in this document, will provide consistent guidance for those seeking Project Licence authority for this area of animal use, ensure the principles of humane experimental technique are properly applied and simplify the process of preparing Project Licence applications for consideration by the Home Office.

2.1 Project Licence applicants must satisfy the Home Office on four essential requirements: that

• there is no appropriate validated alternative replacement to animal tests;
• animal testing will be performed only when there is a reasoned, sustainable justification for the generation of new test data;
• the protocols proposed cannot be further refined based on current information and knowledge;
• the protocols will be likely to produce data which will meet the specified objective.

2.2 Applicants must not propose 'over-testing' or 'check-list approaches', (where a comprehensive battery of test methods is indiscriminately applied for all test materials).

2.3 Project Licence applicants must possess up-to-date knowledge in the field of regulatory toxicology, particularly where there is discretion in the application or selection of test requirements. They are expected to show how professional judgement and flexibility are to be applied with respect to designing and selecting test strategies and protocols to best suit the nature of the test materials and the purpose of the tests.

2.4 Project Licence applicants must demonstrate an awareness of the scope and limitations of the available animal-based tests and knowledge of possible alternative methods. They must review and revise their practices as appropriate alternative or refined methods are developed, validated and gain regulatory acceptance.

2.5 The majority of testing programmes and regulatory assessments are intended to make provision for product development world-wide. Testing strategies and methods should always be scientifically justified and, whenever possible, designed to satisfy the relevant regulatory authorities to which data is likely to be submitted to minimise the need for subsequent duplication or supplementary testing.

2.6 Timely consultations with the relevant regulatory bodies should help prevent unnecessary or inappropriate animal tests and help refine and modify tests already in progress.

2.7 Consideration should be given to the need for pilot studies prior to finalising the design of definitive studies. Pilot studies should be performed when they can provide important information to refine, optimise the design of definitive studies, generate relevant preliminary kinetic data, determine the need for specific mechanistic studies, identify likely biological activity and appropriate endpoints for definitive studies and possibly help in choosing the appropriate species for the definitive study.

2.8 All potential notifiers are encouraged to share information in order to avoid duplicate testing on animals. Project Licence applicants are expected to confirm that all reasonable efforts will be made to take advantage of existing data sharing requirements and opportunities.

2.9 It is strongly recommended that all animal studies where test results are intended for regulatory submission be conducted in compliance with the Principles of Good Laboratory Practice⁵ or other recognised quality assurance systems.

2.10 Opportunities should be sought to provide systems of care and accommodation for the animals on test that enhance their welfare without invalidating the data generated. For example, provision for environmental enrichment and interactions between the animals and humans are expected to be in place. Unless precluded on justifiable welfare, husbandry or scientific grounds, pair- or group-housing of social animals, including dogs and non-human primates, should be the norm

2.11 The following guidance for the design and conduct of studies should ensure the 3Rs are applied optimally to animal tests for regulatory toxicity and safety evaluation.

3. Review of Existing Data

3.1 All relevant data, propriety or in the published literature, must be identified and evaluated to avoid unnecessary duplication of existing data and to reduce and refine the tests required for safety evaluation purposes. In addition regulatory authorities should be consulted as to the availability of existing data. An inter-departmental concordat on data sharing has been agreed by UK regulatory authorities (Annex 2).

3.2 It is essential to determine at the outset whether or not an adequate regulatory assessment can be performed using existing data before additional animal testing is considered.

3.3 In order to comply with the requirements of the 1986 Act, laboratories which conduct studies on behalf of clients are obliged to take all reasonable steps to obtain and evaluate all relevant data before planning and undertaking animal studies.

3.4 Project Licence applications should describe the measures to be taken to obtain and evaluate existing data, and to determine the specific gaps in available knowledge, before requesting additional animal tests. Applicants must demonstrate an awareness of, and commitment to, the opportunities that exist for data sharing.

4.1 Results of in vitro screening can help determine the in vivo testing strategy.

4.2 Computer modelling and structure-activity relationships can serve as a guide to potential effects of close chemical analogues.

4.3 Physico-chemical properties of the test substance should be taken into account when designing animal studies. For example, inert insoluble substances (if shown to be non-bioavailable) do not usually require further toxicity testing. For example, pH values of less than 2.0 or higher than 11.5 may predict severe irritation and corrosion properties and additional tests may not be required.

4.4 Known toxicological profiles of related compounds may be used to predict the toxic properties of the substance and hence help in the study design.

4.5 Tiered or hierarchical approaches to testing allow the satisfactory prediction of likely effects whilst minimising and refining the animal testing required to generate results

4.6 For example, in the OECD test guideline 404 and 405 on skin and eye irritancy/corrosivity there is now a detailed testing strategy which includes in vitro methods. The OECD has also now adopted in-vitro test guidelines for investigating skin corrosivity (TG 430 and 431). Positive results in these assays, or the demonstration of corneal damage in isolated eye models (not yet included in regulatory guidelines) may preclude the need for investigating eye irritancy in animals.

4.7 Project Licence holders must demonstrate that they employ a structured, staged approach in their design of safety evaluation or toxicity studies in animal.

4.8 Test strategies must reflect and incorporate up-to-date scientific knowledge and ethical standards in animal experimentation. Input in the study design from experts (regulatory specialists, information technologists, statisticians, named veterinary surgeons, named animal care and welfare officers etc.) must be available to Project Licence applicants and holders.

4.9 The need for long-term rodent carcinogenicity studies is described in published guidelines but it should be emphasised that, in the presence of clearly positive in vivo mutagenicity, animal carcinogenicity studies are unlikely to be justified.

4.10 Project Licence applicants should supply an algorithm summarising in-house tiered and hierarchical approach and practices to gather and consider any existing data, and determine the most appropriate and refined animal test requirements. This should be incorporated within the plan of work section of the application and may be in narrative or diagrammatic format. One example is shown below:

5. Choice of Method

5.1 Where more than one scientifically satisfactory method exists, acceptable to the regulatory authorities, for a regulatory toxicology or safety evaluation then the method of choice should be the one that overall uses animals of the lowest degree of neurophysiological sensitivity, causes the least amount of pain, suffering, distress or lasting harm and requires the least number of animals. The use of more severe protocols must be justified.

5.2 The OECD, for example, now recognises two methods for investigating skin sensitisation namely TG 406, based on use of the guinea pig, and the new TG 429, the Local Lymph Node Assay (LLNA) in mice. The LLNA has advantages with regard to animal welfare considerations and therefore specific scientific justification should be given if TG 406 is to be used.

5.3 Similarly, the justification for tests that exceed the minimum numbers specified by the appropriate regulatory test protocols should be clearly stated.

6.1 The 1986 Act requires that the animals subjected to regulated procedures be of the lowest neurophysiological sensitivity required to meet the stated objectives.

6.2 Scientific considerations should dictate the choice of species for toxicity and safety evaluation studies; in particular, the relevance to humans, the species sensitivity and metabolism, and the availability of background data. Practical considerations such as the size of the animal, availability and length of gestation may also have a bearing.

6.3 In some instances, specific test requirements or the need to use 'target species', allow little discretion in the species selected.

6.4 In other circumstances, regulatory systems allow judgement to be exercised when selecting appropriate non-rodent species. It is expected that the most appropriate species for the regulatory assessment will be chosen on scientific grounds rather than by custom and practice.

6.5 The applicant should summarise and justify how the second species will be selected.

6.6 Specific justification is required for the use of cats, dogs, equidae and non-human primates. Additional restrictions apply to the acquisition and use of endangered species as listed in Appendix 1 of the Convention on International Trade in Endangered Species of Fauna and Flora. These are set out in Annex A to Council Regulation (EEC) No.3626/82, as amended by Commission Regulation (EEC) No. 969/88 and Commission Regulation (EEC) No. 1970/92 and Annex 1 to the Council Regulation.

7.1 To minimise experimental variables within and between studies the animals should be standardised as far as possible with regard to source, age, weight and health status.

7.2 The number of animals used in regulatory toxicology and safety evaluation studies must be sufficient to allow meaningful interpretation of the data generated.

7.3 Although test requirements are often couched in terms of 'minimum numbers' to be used, the actual numbers of animals required should depend on the expected nature, frequency, intensities and variability of the effect. Expert advice may be required on a study-by-study basis to determine the appropriate numbers. Justification for the proposed numbers (that is, how they will be determined) should be made in section 18 of the Project Licence application form.

7.4 The judicious use of satellite groups (e.g. to obtain toxico-kinetic data), that require small increases in the numbers of animals used, is justified when the additional information generated may refine the subsequent testing programme.

7.5 Many studies require the use of contemporary negative control groups whenever large numbers of candidate compounds are being screened. Efforts should be made to use a single control group as the comparator for a number of test materials. The use of positive controls is occasionally required but it is commonly possible to use them intermittently to check the responsiveness of a test system.

7.6 The responsible reuse of animals may be considered when scientifically justified. This will require the approved by the Secretary of State.

8. Dose Selection and Administration

8.1 Single dose rather than multiple dosing over a 24 hour period should be the norm in acute toxicity tests. Special justification will be required for multiple dose acute toxicity tests.

8.2 Steps should be taken to avoid the use of death as an endpoint (see section 10).

8.3 In repeat dose toxicity and safety evaluation studies, the highest dose should be selected with the intention of allowing the scientific objective to be realised without seriously affecting the wellbeing of the experimental animals. When available, it is essential that dose selection and the dosing schedule for repeat dose studies are based on toxico-kinetic data. ‘Principles for the selection of doses in chronic rodent bioassays’ by Foran J A is a useful guide⁶.

8c. Routes of Administration

8.4 The justification for the choice of the route of administration should be provided. In principle this should be the same as the intended or expected exposure to the test material. However the route, formulation and dosing regime selected should be the one consistent with meeting the scientific objectives and causing the least pain, suffering or distress to the test animals, taking into account the kinetics to be experienced in humans. In general, however, the oral route is likely to be the method of choice.

8.5 A number of expert groups have published tables of maximum dosing volumes and rates of administration designed to minimise discomfort and optimise controlled uptake in the experimental animal⁷. Project Licence applicants should provide tables for typical and maximum dose volumes, frequency, and administration rates by routes of administration to be used. Applicants may seek authority to exceed the stated levels when a justifiable case can be made.

9.1 Regulatory guidelines may require that haematological, biochemical and pharmacokinetic parameters are monitored.

9.2 The sampling times for collecting body fluids should be based on relevant data obtained from earlier preliminary investigations; should be kept to a minimum; and should not cause undue stress to the experimental animal.

9.3 When multiple sampling of body fluids is required, consideration should be given as to whether samples need to be taken from all the dosed animals or whether sampling subgroups would suffice. If satellite groups are included in the study design, they could provide the required samples. When sampling involves the restraint of animals, its duration must be kept to the shortest time possible.

9.4 There are well-characterised differences between venous, arterial and cardiac samples. The method chosen should be the least stressful available that does not compromise the scientific aims of the study.

9.5 A number of publications⁸ recommend sampling regimens designed to minimise the likelihood of compromising the welfare of the animals or the quality of the sample.

9.6 Project Licence applicants should specify limit sample volumes, routes and frequencies, periods of food and water withdrawal, or supply in-house tables (and their derivation) as a single technical appendix relevant to all licensed protocols and studies. They may seek authority to go beyond the stated levels when a justifiable case can be made.

10.1 In every case animal test protocols must be designed and conducted in a manner that causes the minimum suffering to the animals consistent with the scientific objectives. The identification and implementation of early, endpoints is one of the means of achieving this objective. In this regard the OECD has published a Guidance Document on the Recognition, Assessment and Use of Clinical Signs as Humane Endpoints for Experimental Animals Used in Safety Evaluation⁹.

10.2 Endpoints must make provision for action to be taken to end animal suffering in three situations:

• when the degree of suffering exceeds that required by the protocol and cannot be justified by the objective; or
• when the objective has been or cannot be realised; or
• when the quality of the results that will be produced has been compromised.

10.3 Endpoints may indicate when the animal should be removed from the study; when specific, symptomatic or supportive therapy should be given; or when the animal should be promptly and humanely killed. Every effort should be made to institute appropriate supportive and symptomatic treatment without invalidating the scientific rationale for the study.

10.4 Nothing must over-ride the basic principle that animals, which are suffering severe pain, or distress that cannot be alleviated must be promptly and humanely killed.

10.5 In regulatory toxicology and safety testing, the consequences of the physical interventions (e.g. dosing and sampling methods) on experimental animals are generally well characterised. However, licensees should be alert to the unpredictable or unexpected side effects of test materials and take timely action to prevent undue or unnecessary suffering of the experimental animals.

10.6 Behaviour, appearance, food consumption, body weight changes, body temperature changes, physical examination, clinical observation and laboratory investigations (including haematology and serum biochemistry) should be used as required to assess animal suffering.

10.7 Although unrelated disease is rare in toxicological studies, pain or distress thought to be of that origin should be treated in the same way as those arising from regulated procedures. Serious consideration should be given to the scientific value of the experimental results obtained from such animals.

10.8 A number of species-specific 'severity scoring systems' and 'endpoint guidelines’ have been devised and published.^{10, 11} These can be adapted to match the severity limits of Project Licence study protocols and as a practical guide to licensees and animal care staff for limiting the suffering experienced by animals on experiment and implementing agreed endpoints.

10.9 If ‘severity scoring systems’ are supplied as appendices to Project Licence applications (where they set out common endpoints to be applied across all of the licensed studies), these can simplify Licence amendments and can be integrated directly into in-house staff training, Good Laboratory Practice (GLP) systems and Standard Operating Procedures (SOPs).

10.10 Endpoints must be described in meaningful terms, reflecting findings that can be recognised by those entrusted with the welfare of the animals on experiment. Observation schedules should be tailored to allow the early detection of endpoints, and there should be no delay between the detection and implementation of the endpoints. Decisions should not be based solely on the use of any checklist, but must also involve expert judgement of those responsible for the management of the study.

10.11 Studies likely to produce significant acute welfare problems should commence early in the working day in order to maximise the time experienced staff are available

to monitor progress and provide care for the experimental animals. Observation schedules should be determined by the nature and the likely time of onset of welfare problems. Timely supportive, symptomatic or specific treatment for adverse effects must be considered in consultation with the Named Veterinary Surgeon. Animals must be checked at least once a day by a competent person. Studies must be scheduled so that animal welfare and scientific objectives are not compromised by inadequate resources of facilities and staff.

10.12 All reasonable steps should be taken to avoid using death as an endpoint. As death may often occur from the secondary or tertiary effects of a test material, the scientific justification for using death as a meaningful endpoint will be difficult to justify.

10.13 The incidence of animals found dead should be reduced to the minimum. When deaths are encountered, or animals are required to be killed on welfare grounds, records should clearly distinguish between these two events. Means of further reducing the numbers of animals found dead should be determined and implemented. All such animals should be autopsied and the results reviewed to determine whether their death added to scientific information, or identify further scope for refinement.

Home Office

2003

References

2. ‘The Principles of Humane Experimental Technique’: Russell WMS, & Burch RJ (1959) London: Methuen.

3. Guidance on the Operation for the Animals (Scientific Procedures) Act 1986: The Stationery Office (March 2000)

4.The Ethical Review Process: http://www.homeoffice.gov.uk/comrace/animals/reference.html#ethical

5.Report of the Animal Procedures Committee for 1997: Chapter 2, Annex I: The Stationery Office (November 1998)

6. Principles for the selection of doses in chronic rodent bioassays’: Foran JA and The International Life Sciences Institute (1997) Env. Hlth. Persp. 105:18-205.

7. Administration of Substances: LASA Good Practice Guidelines, series 1/ issue 1 - October 1998.

8. ‘A good practice guide to the administration of substances and removal of blood including routes and volumes’: Journal of Applied Toxicology (2001) 21, 15-23.

9.‘The assessment and control of the severity of scientific procedures on laboratory animals’: Report of the Laboratory Animal Science Association Working Party, Laboratory Animals (1990) 24: 97-103

10.‘Guidelines on the recognition of pain, distress and discomfort in experimental animals and on hypotheses for assessment’: Morton D, and Griffiths P, Veterinary Record (1985) 116: 431-436.

11.‘Guidance Document on the Recognition, Assessment, and use of Clinical Signs as Humane Endpoints for Experimental Animals Used in Safety Evaluation’: OECD Paris, November 2000.

This Agency has as its primary duty the protection of [humans]/[animals] and their environment. The safety assessment of human or veterinary pharmaceutical products and of other substances is an essential element in providing this protection.

Some animal use will remain necessary for these safety evaluation studies for the foreseeable future. To ensure that animal welfare is protected during such use, this Agency is committed to the following principles in line with the Animals (Scientific Procedures) Act 1986:-

Any suffering caused to animals and the number of animals used in a safety evaluation protocol should be the lowest necessary to achieve its identified objectives.

To this end, we shall:

1. encourage the development of alternative methods through co-operation with or participation in national and international initiatives aimed at refining, reducing or replacing the use of animals;

2. seek to influence international harmonisation initiatives to ensure that policies and practices take full account of the ethical duty to protect the welfare and minimise the numbers of animals used in safety assessments;

3. alert applicants/clients to instances where data submitted in support of registration or safety evaluation are considered excessive in terms of animal suffering or numbers or where internationally accepted alternatives replacing, reducing or refining animal use have not been employed;

4. accept for assessment internationally recognised, non-animal alternative tests conducted to appropriate standards.

5. not require repetition of data produced by the applicant for other safety evaluation agencies where such data accord with internationally accepted guidelines and have been generated in compliance with appropriate standards.

---

UK regulatory authorities will:-

(a) minimise data requirements for animal tests as far as possible¹;

(b) review their relevant powers and procedures and identify any procedural and legal barriers to data sharing², looking for opportunities to extend data sharing;

(c) encourage data sharing between clients, wherever appropriate, but particularly at the intent to notify (or similar) stage and earlier³;

(d) accept adequately conducted animal test data⁴ generated under different

assessment requirements or regimes (ie. support mutual acceptance in UK and

international settings);

(e) as lead agencies, press for agreement on behalf of the UK Government for fullest provisions and procedures which enable data sharing when negotiating, updating and transposing relevant European Directives and when taking part in other international harmonisation processes⁵;

(f) discuss data sharing with relevant trade associations and other parties, in order to help resolve practical and legal obstacles and to set a climate of expectation that data sharing should take place, whilst recognising the need not to place unreasonable additional burdens on business;

(g) actively encourage companies to search for relevant test data (directing potential notifiers to sources⁶ which reduce the chances of substances and products being tested under more than one national or international requirement), to collaborate over animal testing, and to place data in the public domain.

1 consistent with the principle, agreed across Government in 1996, that any suffering caused to animals and the number of animals used in a safety evaluation protocol should be the lowest necessary to achieve its identified objectives

2 for example, reviewing discretionary measures and exclusivity periods

3 especially in the notification of new substances

4 within statistical and legal limitations

5 this may mean developing procedures and systems which ensure another EU regulator has not been presented with relevant animal test data under notification for a separate requirement or regime; this would be in the interests of "joined-up Government" as well as data sharing

6 for example, the UK Chemicals Regulatory Atlas and equivalent summaries which set out requirements for testing and notification under different regimes