Research is essential for progress, and research on animals has contributed to almost every medical advance of the last century. The NHS would be unable to function effectively were it not for the availability of medicines and treatments that have been developed or validated through research using animals. Thus the public health - in its widest sense - is the ultimate beneficiary of medical research using animals.

The thalidomide tragedy of 1960 prompted governments around the world to play an increasing role in the development of new drugs. This occurred principally by the introduction of legislation forcing companies to test new drugs for safety and efficacy using animal models and randomised controlled trials. A product licence (the alternative term for which is marketing authorisation) could be granted only following a successful review of data by government-appointed experts.

In the UK the Medicines Act of 1968 was the primary legislation; its provisions have now been largely superseded by EU legislation, and internationally agreed guidelines, developed under the aegis of the International Committee on Harmonisation (ICH), play an important quasi-legal role in drug development.

Requirements on animal studies are set out in EU legislation. Before a new medicinal product can be marketed, its developer must obtain a Marketing Authorisation (sometimes called product licence) by submitting a Marketing Authorisation Application (MAA) to the appropriate competent authority (either the national regulatory agency or the European Agency for the Evaluation of Medicinal Products - EMEA). The MAA must include, among other things, results of pharmacological and toxicological tests, unless:

• data are already available on an identical or essentially similar product, or
• an acceptable justification for the omission of particular studies can be provided in the expert report on the toxico-pharmacological documentation.

A medicinal product can be placed on the market in the EU only when the developer has been granted either a national or European marketing authorisation. Council Directive 65/65/EEC details the particulars and documents that must be provided when making a marketing authorisation application. A mandatory part of the documentation for a pharmaceutical product containing a new active substance is the results of pharmacological and toxicological tests (Article 4.8). These tests involve a variety of biological studies employing both non-animal and animal models.

The utility of animal models to assess the potential effectiveness (i.e. pharmacological properties) of new drugs has long been recognised. For example, the curative power of prontasil (the prototype sulphonamide antibiotic) was discovered using groups of mice infected with streptococcus. Animal testing of drug safety began in the 1950s, particularly in the USA. An added impetus was provided by the thalidomide disaster resulting eventually in changes in the UK law, as above, and the evaluation of reproductive toxicity testing (especially teratogenicity) as part of the standard non-clinical test battery.

The last 20 years have seen increasing international rationalisation and harmonisation of drug toxicological testing requirements. In EU Member States legislation on pharmaceutical products is of European origin, and there are no longer any purely national guidelines on drug safety testing. In the 1990s the International Committee on Harmonisation (ICH), with representatives from the US, EU and Japan, achieved its goal of harmonising critical guidelines on quality, safety and efficacy worldwide. Thus there is a global consensus of approach to the appropriate methodologies for assessing drug safety, which parallels the increasing globalisation of the pharmaceutical industry.

The current pharmacotoxicological testing guidelines, developed over the last 40-50 years, represent a distillation of the scientific knowledge acquired over this period applied to ensuring patient safety both during and following drug development. New guidelines are developed, as appropriate, in response to therapeutic and other advances such as in biotechnology, gene therapy and transgenics. The current pre-clinical safety assessment of new chemical entities (NCEs) is largely dependent on in vivo mammalian toxicity data. Reasons for the continued use of animal models include:

• Reliability: Whole-body toxicity responses rather than responses of individual cell and / or organ types are obtained
• Comprehensiveness: Numerous toxicity markers can be incorporated, e.g. macro / micro-pathology, haematology, clinical chemistry. Responses to metabolites are in-built and separate experiments on metabolites are not normally necessary.
• Flexibility: Different routes and durations of administration, and of dosing schedule are possible to match the anticipated drug-dosing regimen. Different laboratory animal species can be used as appropriate.
• Dose and Exposure: Using a spread of dose levels, that normally range from the anticipated clinical dose to a considerable multiple of it, enables an assessment of the "no-adverse-effect dose" and the "maximum tolerated dose", both of which are useful in data interpretation. In vitro toxicity data, which are normally concentration-based, are much more difficult to interpret unless detailed information on in vivo cell / organ drug concentrations is available.
• Comparability: There is a high degree of physiological and biological commonality across mammalian species. Biochemical and toxicological mechanisms are similar.
• Predictivity: Conventional repeated-dose animal studies in a rodent and non-rodent species can predict about 70% of human drug toxicities, human skin reactions being the worst predicted.
• Prior Art: Use of animal models and interpretation of results are underpinned by the vast repository of information in the published literature. For example, it is often possible to assess with reasonable certainty whether animal toxicity findings are or are not likely to be relevant to humans.

The development of a new pharmaceutical product is a lengthy stepwise process with the purpose of establishing quality, safety and efficacy for the product in its intended indication. Possibly the most critical aspect of the development programme is a sequence of clinical trials starting with Phase I trials (exploratory investigations in human volunteers and patients), continuing through Phase II trials (initial investigations of efficacy in patients) and culminating in Phase III trials (extensive investigations of safety and efficacy in patients).

Various non-clinical investigations, both preceding and running concurrently with the clinical trials programme, are also necessary for a number of reasons. The key functions of non-clinical studies in drug development relate to:

• Drug Discovery: Although in vitro screening tests are often able to establish presumptive evidence of pharmacodynamic activity, data from animal models of the particular disease or condition are required to generate sufficient in vivo evidence to provide a sound pharmacological rationale for further development.
• Protection of Clinical Trial Patients: Before an investigational drug is administered to patients or volunteers, sufficient information must be available to characterise the potential toxic effects in the intended clinical trials. In almost all cases, the only reliable means of obtaining such data is through the use of animal studies of appropriate type and duration.
• Safety Endpoints not Amenable to Clinical Evaluation: As a clinical development programme progresses, sufficient human safety data are normally obtained from clinical trials to supersede animal toxicity data. However, several endpoints such as genotoxicity, carcinogenicity and reproductive toxicity, for both practical and ethical reasons, are investigated in non-clinical studies. For example, to avoid the administration of a drug under development to a pregnant woman without a prior assessment of fetotoxicity the only reliable techniques for carrying out such an assessment involve the use of animal models. Periods of 20 to 40 years from initiation to diagnosis are the norm for human tumours. Thus, even if it were ethical and practicable to evaluate carcinogenic potential in humans, the process would take several decades and provide results of questionable validity (due to the confounding effects of smoking and other lifestyle factors in human populations).

A number of in vitro safety tests have been proposed but only a few (e.g. those used in the assessment of genotoxicity) have proved sufficiently robust and reliable for regulatory purposes. A lower standard of validation is acceptable when screening candidate drugs during development.

The key goal of non-clinical safety evaluation prior to commencing clinical trials is the assessment of potential toxicity with respect to target organs, dose dependence, relationship to systemic exposure and reversibility. Use of in vitro systems such as different types of cultured cells cannot replicate the complex dynamic, interactive and multi-organ events that occur in vivo. Moreover, in vivo systems, by their nature, test the safety of parent drug and metabolites (which are often responsible for toxic effects). Assessment of metabolite toxicity in a cell-culture system is highly problematic: since metabolic capacity is often diminished or absent in cultured cell lines, an approach involving the use of metabolic activating systems (e.g. liver microsomes) or of individual metabolites, would need to be adopted. The former approach would lead to exposure of all cell types to the same range of (possibly unrepresentative) metabolites, and the latter approach would require the appropriate animal / human studies to be undertaken in order to achieve metabolite identification.

In the UK there are three schemes under which clinical trials in patients can be undertaken. These are:

• Clinical Trial Certificate (CTC): This scheme is used only in those exceptional cases where a thorough evaluation of the data is necessary in order to ensure the safety of the clinical trial patients. Data requirements are similar to those for a Marketing Authorisation Application (MAA).
• Doctors' and Dentists' Exemption (DDX): Under this scheme doctors and dentists notify the Medicines Control Agency of their intention to carry out a clinical trial. The trials are normally of the "proof-of-concept" type in which a well-established drug undergoes a preliminary assessment of efficacy (and safety) in a new indication.
• Clinical Trial Exemption (CTX): The Medicines (Exemption from Licences and Certificates)(Clinical Trials) Order 1995 (SI 1995/2809) is made under the Medicines Act 1968 and provides the statutory basis for the CTX Scheme. Any company or other organisation wishing to undertake a clinical trial under this scheme must apply to the Medicines Control Agency and provide details of the intended trial and summaries of pharmaceutical data and of reports and evaluations of any experimental and biological studies and of the preclinical or laboratory studies carried out with each product or its constituents which, in the view of the supplier, are relevant to the assessment of safety, quality and efficacy of the product, together with references to relevant publications or other clinical trials. Monitoring for safety should be designed bearing in mind the pattern of toxicity in animal studies, i.e. more extensive monitoring should be conducted with relation to the system showing target organ toxicity.

Animal studies perform several vital functions in drug development. Although a number of in vitro techniques are now employed, overwhelming technical difficulties remain to be solved before suitable and validated alternatives can be established to replace current practice, particularly for studies on target organ toxicity, reproductive toxicity and carcinogenicity.